The Oxford Handbook of the South African Economy 9780192894199, 0192894196

Citation preview

T h e Ox f o r d H a n d b o o k o f

THE SOUTH A F R IC A N E C ON OM Y

The Oxford Handbook of

THE SOUTH AFRICAN ECONOMY Edited by

ARKEBE OQUBAY, FIONA TREGENNA, and

IMRAAN VALODIA

1

3 Great Clarendon Street, Oxford, OX2 6DP, United Kingdom Oxford University Press is a department of the University of Oxford. It furthers the University’s objective of excellence in research, scholarship, and education by publishing worldwide. Oxford is a registered trade mark of Oxford University Press in the UK and in certain other countries © Oxford University Press 2021 The moral rights of the author‌s have been asserted First Edition published in 2021 Impression: 1 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form or by any means, without the prior permission in writing of Oxford University Press, or as expressly permitted by law, by licence or under terms agreed with the appropriate reprographics rights organization. Enquiries concerning reproduction outside the scope of the above should be sent to the Rights Department, Oxford University Press, at the address above You must not circulate this work in any other form and you must impose this same condition on any acquirer Published in the United States of America by Oxford University Press 198 Madison Avenue, New York, NY 10016, United States of America British Library Cataloguing in Publication Data Data available Library of Congress Control Number: 2021936256 ISBN 978–​0–​19–​289419–​9 DOI: 10.1093/oxfordhb/9780192894199.001.0001 Printed and bound by CPI Group (UK) Ltd, Croydon, CR0 4YY Links to third party websites are provided by Oxford in good faith and for information only. Oxford disclaims any responsibility for the materials contained in any third party website referenced in this work.

Preface

The South African economy has evolved from a pre-​colonialist subsistence economy into being an economic powerhouse in Africa in the twentieth century—​by official statistics, the second largest economy in Africa. The discovery of diamonds in 1870 and the discovery of the largest gold deposits in the world in the Witwatersrand in 1886 shaped much of the growth of the economy, and its associated process of capitalist development and conquest of the indigenous populations of the country. Colonial conquest and later its apartheid history shaped South Africa’s racially determined economic growth trajectory which generated some of the highest rates of economic growth and the most severe forms of economic and social inequality. The economy today, while having some significant capabilities in agriculture, mining and minerals, manufacturing, finance, and services, faces serious economic challenges. These include: low levels of economic growth; a dependence on exports of commodities; extremely high levels of unemployment; the highest levels of inequality in the world; large backlogs in infrastructure and social services; high levels of crime and gender-​ based violence; and a highly unequal education system with poor outcomes. The skewed economic structure and dynamics of the apartheid period continue to pose a key structural constraint on growth and economic development in the post-​apartheid period. Notwithstanding these economic and social challenges, since coming to power in 1994, the democratically elected government has followed economic policies that have attempted, with limited success and arguably within a narrow economic paradigm, to develop economic opportunities for the previously disadvantaged segments of the population and extended social services to large segments of the society. South Africa has by far the most extensive social services sector on the continent, with significant transfers through the pension system. By most developing country standards, its tax to GDP ratio at 29 per cent is fairly high. Yet, growth rates have been very weak with manufacturing contracting with the challenge of premature deindustrialization and a lack of structural transformation. Furthermore, the economy remains dependent on exports of mineral products, unemployment remains intractable, and ownership remains concentrated. The economic challenges have been accentuated by corruption and political uncertainty.

Overview of the Volume While South Africa shares some characteristics with other middle-​income countries, it has a unique economic history and the economy has distinctive characteristics. Perhaps

vi   Preface surprisingly in light of the extensive research on various aspects of its economy, there is no comprehensive handbook providing in-​depth analysis of a wide range of aspects of the South African economy. This volume is intended to fill this important gap, and provides detailed and wide-​ranging coverage of the key economic questions in South Africa. Though covering the important historical antecedents, the intention of this volume is to concentrate on the more recent economic challenges facing South Africa. The preparation of the chapters for this volume followed an extensive process of consultation, collaboration, and peer review. The project involved key workshops where initial ideas were developed and then subjected to peer input. The editors approached leading scholars on various economic issues in South Africa to contribute chapters. Authors were asked to prepare detailed abstracts on their proposed chapters. Thereafter, an inception workshop was held to review chapter abstracts and discuss common themes, gaps, overlaps, and data sources. Authors then prepared drafts of the chapters, all of which were subject to at least two peer reviews. With these reviews, draft chapters were discussed at a chapter review workshop. Authors then revised the chapters. The handbook provides a comprehensive treatment of all key aspects of the South African economy. Part I begins with a historical background, segmented into the periods before and after the formal introduction of apartheid in 1948. The next two chapters discuss issues of political economy and economic policymaking, cross-​cutting issues that are an essential backdrop to all the issues discussed in the remainder of the volume. The remaining four chapters in Part I focus on what are arguably the central challenges facing the South African economy: growth and the triple challenges of unemployment, poverty, and inequality. Part II begins with three chapters on agriculture, agro-​processing, and the land question, followed by chapters on mining and minerals and energy, thus providing a treatment of the primary sectors and associated issues. The remaining two chapters focus on the key issues of climate change. Part III of the volume deals with a set of topics linked broadly to structural transformation, industrial development, trade and regulation. These chapters include focuses on value chains; South Africa’s place in the region and on the continent; the fourth industrial revolution; innovation and technological change; competition policy and regulation; state-​owned enterprises; black economic empowerment; entrepreneurship and small, medium, and micro-​enterprises (SMMEs); and urbanization and cities. In Part IV, attention turns to the labour market, various aspects of distribution, and the economics of social policy. The multiple and far-​reaching dimensions and effects of the triple challenges discussed in Part I are laid bare here, including in chapters on different aspects of the labour market, gender, and the economics of education, health, food security and hunger, and social development and social security. The macroeconomy is the focus of Part V. While constraints on economic growth were discussed in Part I as a central challenge facing the economy, here we turn to a macroeconomic analysis of growth. The next chapters discuss the key macroeconomic issues of investment, public finance and fiscal policy, debt, monetary policy, banking and finance and financialization.

preface   vii

In Memorium Two esteemed contributors to this Handbook, professors Bill Freund and Vishnu Padayachee, sadly passed away during the production of the volume. Freund was emeritus professor at the University of KwaZulu-​ Natal, while Padayachee was distinguished professor and Derek Schrier and Cecily Cameron Chair in development economics at the University of the Witwatersrand. Both Freund and Padayachee were highly original thinkers, unafraid to challenge dominant views. Each of them made profound contributions to economic thought in South Africa, particularly in economic history and political economy. In the case of Padayachee, this contribution extended to a prominent role in economic policymaking over a long period of time. Freund supervised Padayachee’s doctoral thesis, and they collaborated in various publications. When Freund passed away in 2020, the chapter that he had begun writing for this volume was completed by Padayachee, who passed away in 2021 while the volume was in press. Their co-​authored chapter provides an authoritative account of the economic history of South Africa over the period 1948–​94.

Acknowledgements

Editing the Oxford Handbook of the South African Economy has been an exciting and challenging venture. It has been a complex project to manage, with eighty authors contributing to the volume. Additional challenges were brought by the fact that the project coincided with the COVID-​19 pandemic, which meant that our inception and review workshops had to be conducted virtually rather than in person as originally planned, and that authors had to manage new ways of working and many competing demands on their time. Nonetheless, we are very pleased with the speed with which this volume has come together, and with the quality of the product. This would not have been possible without the contributions of a number of people. We would like to thank our commissioning editor at Oxford University Press (OUP), Adam Swallow, for his support and guidance of the project, as well as to the entire OUP production team. We also wish to thank the anonymous reviewers of our book proposal, and the delegates of OUP who approved it. Several people gave helpful inputs and suggestions in the course of the project, but we especially appreciate the invaluable contributions of Christopher Cramer (University of London, SOAS) throughout the project. Arabo Ewinyu, Meron Tilahun, and Deborah Kefale provided dedicated and efficient support, without which a project of this magnitude could not have succeeded. The volume benefited tremendously from a number of national and international peer reviewers, whose detailed and insightful comments on individual chapters contributed significantly to their ultimate quality. Finally but most importantly, we acknowledge the chapter authors, who participated in the inception and review workshops and delivered such high-​quality chapters. The editors are very grateful to all the contributors of this Handbook and thank them sincerely for their exceptional efforts and their personal commitment to ensuring the success of this volume. 16 February 2021

Contents

List of Figures List of Tables List of Abbreviations List of Contributors

xv xxi xxv xxxv

PA RT I   H I STORY, P OL I T IC A L E C ON OM Y, A N D K E Y C HA L L E N G E S 1 Challenges and Complexities of the South African Economy Fiona Tregenna, Arabo K. Ewinyu, Arkebe Oqubay, and Imraan Valodia

3

2 The Economic History of South Africa before 1948 Stefan Schirmer

26

3 The Economic History of South Africa 1948–​94 Bill Freund† and Vishnu Padayachee†

47

4 Politics and Economic Policymaking in South Africa since 1994 Alan Hirsch, Brian Levy, and Musa Nxele

66

5 South Africa’s Post-​apartheid Economic Development Trajectory David Francis, Adam Habib, and Imraan Valodia

91

6 Constraints to Economic Growth in South Africa Kenneth Creamer

111

7 Unemployment in South Africa James Heintz and Karmen Naidoo

135

8 Poverty in South Africa Vusi Gumede

157

9 Inequality in South Africa Murray Leibbrandt and Fabio Andrés Díaz Pabón

175

xii   Contents

PA RT I I   T H E P R I M A RY SE C TOR S , E N E RG Y, A N D T H E E N V I RON M E N T 10 Agriculture in South Africa Wandile Sihlobo and Johann Kirsten

195

11 Agro-​processing Industries in the South African Economy Horman Chitonge

217

12 Land and Agrarian Development in South Africa Ruth Hall and Farai Mtero

240

13 Mining and Minerals in South Africa Neva Seidman Makgetla

261

14 Energy in South Africa Rod Crompton and Ruwadzano Matsika

283

15 Socio-​economic Aspects of Energy and Climate Change in South Africa Roula Inglesi-​Lotz 16 Climate Change and the Green Transition in South Africa Channing Arndt, Sherwin Gabriel, Faaiqa Hartley, Kenneth Marc Strzepek and Timothy S. Thomas

305 323

PA RT I I I   T R A DE , I N D U ST RY, A N D R E G U L AT ION 17 Corporate Structure, Industrial Development, and Structural Change in South Africa Pamela Mondliwa and Simon Roberts

349

18 Value Chains and Industrial Development in South Africa Mike Morris, Justin Barnes, and David Kaplan

375

19 Southern African Regional Value Chains and Integration Reena das Nair

396

20 South Africa’s Economic Role in Africa Mills Soko and Mzukisi Qobo

419

21 South Africa’s International Trade Lawrence Edwards

441

Contents   xiii

22 Innovation and Technological Change in South Africa Erika Kraemer-​Mbula and Rasigan Maharajh

467

23 South Africa and the Fourth Industrial Revolution Bhaso Ndzendze and Tshilidzi Marwala

489

24 Industrial Policy in South Africa Anthony Black

510

25 Competition Policy in South Africa Liberty Mncube and Nicola Theron

531

26 Regulation of Network Industries in South Africa James Hodge and Tamara Paremoer

553

27 State-​Owned Enterprises in South Africa Mark Swilling and Nina Callaghan

576

28 Black Economic Empowerment in South Africa Thando Vilakazi

599

29 Entrepreneurship and SMMEs in South Africa Boris Urban

622

30 Urbanization, Agglomeration, and Economic Development in South Africa Ivan Turok

646

PA RT I V   T H E L A B O U R M A R K E T, DI S T R I BU T ION , A N D S O C IA L P OL IC Y 31 Changing Dynamics in the South African Labour Market Haroon Bhorat, Ben Stanwix, and Amy Thornton

673

32 The Youth Labour Market in South Africa Cecil Mlatsheni

690

33 The Economics of Education in South Africa Nicola Branson and David Lam

707

34 Gender and Work in South Africa Daniela Casale, Dorrit Posel, and Jacqueline Mosomi

735

xiv   Contents

35 The South African Informal Economy Michael Rogan and Caroline Skinner

757

36 Migration and Remittances in South Africa Mark A. Collinson and Mduduzi Biyase

777

37 The Economics of Households in South Africa Dorrit Posel and Katharine Hall

800

38 Food Security, Hunger, and Stunting in South Africa Julian May

823

39 The Economics of Health in South Africa Ronelle Burger and Mosima Ngwenya

844

40 Social Security and Social Development in South Africa Leila Patel

869

PA RT V   T H E M AC ROE C ON OM Y 41 The Macroeconomics of South African Economic Growth Philippe Burger

891

42 Investment in South Africa Ciaran Driver and Laurence Harris

913

43 Public Finance and Fiscal Policy in South Africa Tania Ajam

935

44 Public Debt in South Africa Manoel Bittencourt

956

45 Monetary Policy in South Africa Nicola Viegi

974

46 Banking and Finance in South Africa Penelope Hawkins

992

47 Financialization in South Africa Ewa Karwowski

1013

Index

1031

Figures

1.1 GDP per capita in South Africa and comparator economies, 1994–​2019

5

1.2 Gross fixed capital formation in South Africa and comparator economies, % of GDP, 1994–​2018

7

1.3 Carbon dioxide emissions (metric tons per capita) in South Africa and comparator economies, 1994–​2016

10

4.1 The market matrix: alignment of elites into rentiers, powerbrokers, magicians, and workhorses

76

4.2 South Africa’s GDP growth relative to upper-​middle-​and middle-​income countries, 2008–​19

83

5.1 GDP growth and GDP per capita in South Africa

92

6.1 Real GDP growth rate 1946–​2019, annual percentage change

113

6.2 Rising employment and unemployment, 1994–​2019

117

6.3 Commodity cycle and the real GDP growth rate, 1994–​2015

120

6.4 Investment level and GDP growth, 1994–​2019

122

6.5 Composition of investment (as % of GDP), 1994–​2019

123

7.1 Labour force participation and unemployment rates, 1993–​2019

137

7.2 Growth rate, fixed capital stock, South Africa 1950–​2019

140

9.1 Average monthly wages between different sets of categories (in South African Rands-​ZAR)

181

9.2 Socio-​economic class sizes, 2008–​17

184

9.3 The intergenerational transmission of earnings advantages or disadvantages 186 10.1 Agricultural land capability map of South Africa

196

10.2 Long-​term trend of South Africa’s agricultural labour

208

10.3 Total agricultural exports by selected major products, South Africa, real 2010 values

212

10.4 Annual food inflation rate, January 2009 to July 2020

213

11.1 Manufacturing employment by subsector 2019

221

11.2 Agro-​processing output and employment by subsector, average for 2011–​19

222

11.3 Agro-​processing in manufacturing employment, 2000–​19

232

xvi   Figures 12.1 South African Index of Multiple Deprivation 2011 at ward level showing former homeland boundaries

242

12.2 Land redistribution and restitution, by province, in hectares, 2009–​18

249

12.3 Land redistribution under five presidents, 1994–​2018

250

12.4 Land reform budgets, 1996–​2019, inflation adjusted

251

13.1 Indicators of the significance of mining for the economy, 1994 and 2019

262

13.2 Annual percentage change in GDP compared to annual percentage change in international price of South Africa’s main mining exports

263

13.3 Contribution of mining to GDP in constant (2020) prices, by commodity

265

13.4 Employment by commodity, 1980 to 2019, in thousands

266

13.5 Mining exports excluding gold by destination, 1990 to 2019

268

13.6 General government and public corporation (SOC) investment as percentage of GDP and average annual metals price index (2011 = 100), 2000 to 2018

276

14.1 GDP, energy, and emissions intensity, certain countries

285

15.1 Energy consumption per capita and GDP per capita 2010, South Africa, 1971–​2019

307

15.2 Primary energy shares panel A (top): averages 2007–​17; panel B (bottom): percentage change between 1997 and 2017

308

15.3 Electricity access—​new connections and transmissions lines, South Africa

311

15.4 Access to electricity (total, urban, and rural), 1996–​2018, South Africa

313

15.5 CO2 emissions per capita and fossil-​fuel share of total energy consumption, 1971–​2014, South Africa

315

16.1 Annual rainfall distribution for Republic of South Africa.

325

16.2 The six hydro-​climatic zones which are the aggregation of the 9 Water Management Areas with percentage of total national annual.

326

16.3 SACReD –​Systematic Analysis of Climate Resilient Development Framework for LTAS.

328

16.4 Cultivated land use by agricultural system in southern Africa.

333

16.5 Estimated density functions of the total runoff in South Africa for the periods Base, 2025, 2045, and 2065 under a) Paris Forever (PF) scenario and b) 2C scenario.

337

17.1 Sectoral composition of top 100 JSE-​listed firms

356

18.1 South African light vehicle domestic market performance versus production, 1995 to 2018

381

19.1 South Africa’s exports of diversified manufactured goods to SADC

402

20.1 Composition of South Africa’s exports in Africa, 2019

433

figures   xvii 20.2 South Africa’s imports composition in Africa, 2019

434

20.3 Intra-​regional economic community trade in Africa, 2010–​12 and 2014–​16 (billions of dollars and percentage of total African trade)

434

21.1 Trade volumes (goods and services) and real GDP, 2010 = 100

443

22.1 South African publications and world ratio (count)

474

22.2 South African business expenditure on R&D (BERD) and the BERD/​GERD ratio, 2006–​18

476

22.3 South African patent applications and patents granted, by country of origin, 1994–​2019

477

22.4 R&D personnel, 2001–​18

479

23.1 4IR-​related and non-​related industrial wages in South Africa (annual, in ZAR) 498 26.1 Average electricity prices (c/​kWH)

564

27.1 Assets by SOC and DFI (excluding PIC), in billions of Rand, 2019

583

27.2 A tumultuous decade leading up to the second coal-​power-​system emergency in 2018 and the liquidity insolvency in 2019

588

30.1 South Africa’s population distribution, 1975–​2030

650

31.1 Value-​added and employment growth by sector, 2000–​19

676

31.2 Annual average growth rate of employee earnings, 2000–​17

680

31.3 Wage distributions by union status and public/​private sector, 2017

683

31.4 Proportion of workers earning below the national minimum wage by sector, 2017

684

31.5 Cost per beneficiary and jobs created by ALMPs in South Africa

686

33.1 Mean years of education by year of birth and population group

711

33.2 Proportion completing secondary schooling by year of birth and population group

712

33.3 Proportion completing post-​secondary education (of those with twelve+ years of education) by year of birth and population group

713

33.4 Proportion completing post-​secondary qualifications by level and year of birth 717 33.5 Education distribution of working-​age population (aged 25–​59) by year

725

33.6 Relationship between schooling and employment, 1994/​95 and 2017/​18, ages 25–​59

726

33.7 Relationship between schooling and earnings, 1994/​95 and 2017/​18, ages 25–​59

727

34.1 Labour-force participation rates by gender, 1994–​2019

738

34.2 Employment and unemployment rates by gender, 1994–​2019

740

34.3 Real monthly earnings by gender for the employed (mean and median), 1994–​2017

745

xviii   Figures 35.1 Total non-​agricultural employment in the South African informal economy, by gender (2008–​19)

765

35.2 Percentage of non-​agricultural employment in informal employment, by gender (2008–​19)

766

35.3 Percentage of non-​agricultural informal employment in the informal sector, by gender (2008–​19)

766

35.4 Percentage of total non-​agricultural employment in the informal sector, by gender (2008–​19)

767

35.5 Status in employment within the informal economy, by gender (2008–​19)

768

35.6 Sectoral distribution of informal employment, by gender (2019)

769

35.7 Sectoral distribution of informal employment within the formal sector, by gender (2019)

770

36.1 Migration prevalence, by age, sex, and period, NIDS, 2008–​14

784

36.2 Age-​sex profiles of temporary migrants of the combined SAPRIN populations and each of the three nodes: AHRI, Agincourt, and DIMAMO

786

36.3 Age-​sex profiles of the four migration types in the combined SAPRIN population 787 36.4 Trend in migration incidence rate for the four migration types in the combined SAPRIN population

788

37.1a Percentage of households with at least one co r​ esident heterosexual couple, 1995–2018

807

37.1b Percentage of households that are female dominated, 1995–2018

807

37.1c Percentage of households that are male dominated, 1995–2018

807

37.1d Percentage of children living with both biological parents, 1995–2018

807

38.1 Dietary energy supply, 2000–​17

828

38.2 Self-​assessed hunger, 2002–​17

830

38.3 Prevalence of undernourishment, 2002–​19

832

39.1 Shifts in mortality trends in South Africa, 1997–​2017

847

39.2 Public-and private-sector hospital accreditation scores, COHSASA

852

39.3 Real per capita health expenditure

854

39.4 Health expenditure as a percentage of GDP

854

39.5 Out-​of-​pocket health expenditure as percentage of health expenditure in purchasing power parity

856

41.1 Contributing factors to economic growth

894

41.2 Manufacturing relative to the aggregate economy

898

figures   xix 42.1 Gross fixed capital formation as a percentage of GDP (nominal and real ratios) and growth of the real capital stock

914

42.2 Composition of real gross capital formation (R million at constant 2010 prices), private and public sectors

917

42.3 Asset composition of private-​sector and public-​sector gross fixed capital formation (R million at constant 2010 prices)

918

42.4 Gross and net nominal capital formation as a ratio of nominal GDP, private and public sectors

919

42.5 Ratio of nominal gross capital formation to gross operating surplus

924

42.6 Broad sectoral shares of the real capital stock

928

43.1 Debt outlook scenarios—​trajectories of gross debt to GDP in South Africa, 2016/​17–​2028/​29

942

44.1 Government debt, the macroeconomy, and political regime characteristics, South Africa, 1970–​2016

961

44.2 Income inequality, and government expenditure on education and health, South Africa

962

44.3 OLS regression lines, government debt, growth, inflation, and polity, South Africa, 1970–​2016

965

44.4 Government debt, inequality, and redistribution, South Africa

967

44.5 Government debt and government wastefulness, South Africa

970

45.1 South African government debt and EMBI+SA risk premium

985

46.1 Distribution of financial assets between financial intermediaries

997

46.2 Bank assets and loans and advances

998

46.3 Assets of the pension fund industry

1001

47.1 GDP shares by sector, 1960–​2017

1018

47.2 Real house price inflation in the United States, United Kingdom, and South Africa, 2000–​16

1021

Tables

1.1 GDP per capita in South Africa and comparator economies, 1994–​2019

6

1.2 Sectoral composition of GDP (%) in South Africa and comparator economies, 1994–​2019

8

1.3 Selected measures of R&D, innovation, and technology intensity in South Africa and comparator economies, 1994–​2018

9

1.4 Headline measures of unemployment, poverty, and inequality in South Africa and comparator economies, 1994–​2019

12

3.1 Conventional and recalculated economic growth rates before and after 1948

48

4.1 South African real annual economic growth, 1990–​2019

75

4.2 South Africa’s 2014 population distribution, by ethnicity and class

83

5.1 Labour force overview

103

7.1 Broad unemployment rates by demographic characteristics and educational attainment

138

7.2 Output-​employment elasticities

142

7.3 Sectoral shares of employment

143

8.1 Poverty trends, 2006–​15

161

8.2 Poverty by gender

162

8.3 Estimated poverty rates

165

8.4 Estimated poverty rates by province

166

8.5 Estimated poverty rates by gender

167

8.6 Multidimensional headcount

167

8.7 Contribution of each indicator, percentage

168

8.8 Estimated poverty rates by geotype

170

10.1 The landscape of organized agriculture and the various national farmer organizations

199

10.2 Structural changes in the South African agricultural economy, 1990–​2019

201

10.3 The output mix for South African agriculture, top ten commodities in selected years according to production value

202

xxii   tables 10.4 Agriculture’s contribution to GDP according to the agricultural census and national accounts

202

10.5 Statistics on farm numbers and farm employment from the agricultural censuses and selected agricultural surveys, 1990–​2017

204

10.6 Farm structure in South Africa, 2007

204

10.7 Farm structure in South Africa, 2017

205

10.8 Categories of agricultural households according to population groups, 2016 205 10.9 Number of farms/​farming units in the commercial agriculture industry by activity and size group, 2017

206

10.10 Share of Black farmers in commercial agricultural output, average 2015–​19

207

10.11 Annual sales of tractors and combine harvesters

210

10.12 South Africa’s agricultural exports and imports, 2010 prices

211

11.1 Manufacturing output by subsector (%), 1970–​2019

220

11.2 Food and beverage output composition, 2011–​17

224

11.3 Clothing and textile output and formal employment trends, 1995–​2019

227

11.4 Leather, leather products, and footwear employment

229

11.5 Employment and output composition by subsector, 2000–​19

231

11.6 Agro-​processing formal employment by subsector (%)

233

12.1 Areas of land by land use, ownership, and tenure category

246

12.2 Land redistribution and restitution 1994–2020 (in hectares)

247

15.1 Economic output and employment of selected energy-​related economic sectors 310 16.1 Mean daily maximum temperature for the warmest month in the wettest three consecutive months

331

16.2 Total rainfall in the wettest three consecutive months

332

17.1 Manufacturing and services performance: selected sectors

352

17.2 Summary of control of JSE market capitalization (% of total)

354

18.1 SSA apparel exports to South Africa (in US$000)

388

18.2 Clothing and textiles key performance indicator data

390

19.1 Country share of intra-​SADC and total trade

409

21.1 South African export composition by technology classification (values and share structure)

445

21.2 Indicators of export concentration and complexity

453

21.3 Contribution to output and employment growth in manufacturing as share initial total output or employment, 1992–​2019

459

24.1 Economic performance of selected middle-​income countries, 1994–​2019

512

tables   xxiii 24.2 A chronology of industrial policy

514

24.3 Import penetration ratios in key sectors

516

24.4 Manufacturing employment by sub s​ ector 2000–​19, categorized according to factor intensity

518

24.5 Share of manufacturing value added in textiles and clothing, 1996–​2018

523

27.1 Government guarantee exposure

585

27.2 Oversight of priority SOEs and DFIs

594

29.1 Types of entrepreneurial activity, by region and select country as percentage of adult population

631

31.1 Sectoral employment distribution and change, 2000–​19

677

31.2 Occupational employment distribution and change, 2000–​19

678

33.1 Enrolment, educational attainment, and grade repetition by age

715

34.1 Key labour market statistics by gender, selected years from 1994–​2019

739

34.2 Occupational distributions by gender, 1994 and 2019

743

34.3 Mean total time and activity participation rates by gender, 2010

749

36.1 Households with labour migrants and non-​resident members

785

36.2 Population, migration, and remittance characteristics, 2002, 2007, 2012, 2017, in Agincourt HDSS data

789

36.3 Quantile estimates of the effect of remittances on expenditure patterns

791

37.1 Household types and union formation in South Africa, 1995–​2018

804

37.2 Economic well-​being across household types

816

40.1 Social assistance: Type, target group, level of grant, and number of beneficiaries as at end of July 2020

876

41.1 Contributing factors to sectoral output growth, average per annum growth

895

41.2 Sectoral characteristics (average 2010–​19 or values in either 2010 or 2019)

903

42.1 Gross fixed capital formation as a percentage of GDP

916

43.1 Actual and projected main budget revenue, expenditure, and budget balances in South Africa 2008/​09–​2020/​21

941

43.2 Main budget revenue by source 2008/​09 to 2020/21

945

44.1 Descriptive statistics and the correlation matrix

964

44.2 Government debt and the macroeconomy

965

44.3 Government debt and political regime characteristics

966

44.4 Descriptive statistics and the correlation matrix: Public goods

967

44.5 Government debt, inequality, and redistribution

968

46.1 Overview of the South African financial sector

995

46.2 South African banks: Two decades at a glance

1000

Abbreviations

4IR

Fourth Industrial Revolution

AAC

Anglo American Corporation

AEC

African Economic Community

AfCFTA

African Continental Free Trade Area

AGOA

Africa Growth and Opportunity Act

AMCU

Association of Mineworkers and Construction Union

ANC

African National Congress

APDP

Automotive Production Development Programme

APRM

African Peer Review Mechanism

ARF

African Renaissance Fund

ARM

African Rainbow Minerals

ARV antiretroviral ASGISA

Accelerated and Shared Growth Initiative for South Africa

ASUF

Agri Sector Unity Forum

ATC

average total cost

ATNS

Air Transport Navigation Services

AU

African Union

AV

autonomous vehicle

BBC

Black Business Council

BEE

Black Economic Empowerment

BERD

business enterprise R&D

BEV

battery electric vehicle

BIS

Black Industrialists Scheme

BPS

business process services

BRT

bus rapid transport

CAC

Competition Appeal Court

CAPS

Cape Area Panel Study

CATI

Computer Assisted Telephone Interviewing

xxvi   Abbreviations CBM

Consultative Business Movement

CeSTII

Centre for Science, Technology and Innovation Indicators

CGE

computable general equilibrium

CHW

community health worker

CI

contractual intermediaries

CID

Center for International Development

CIP

Competitiveness Improvement Programmes

CIS

collective investment scheme

CIT

corporate income tax

CLP

corporate leniency policy

CMS

Council for Medical Schemes

CoCA

Census of Commercial Agriculture

COI

Complexity Outlook Index

COMESA

Common Market for Eastern and Southern Africa

COS

cost of sales/cost of services

COSATU

Congress of South African Trade Unions

CPI

consumer price inflation

CRAM

Coronavirus Rapid Mobile Survey

CREFSA

Centre for Research into Economics and Finance in Southern Africa

CS

Community Survey

CSG

Child Support Grant

CSIR

Council for Scientific and Industrial Research

CSP

Cities Support Programme

CSR

corporate social responsibility

CTCP

Clothing and Textiles Competitiveness Programme

CWP

Community Work Programme

DAC

Department of Arts and Culture

DACST

Department of Arts, Culture, Science, and Technology

DAFF

Department of Agriculture Forestry and Fisheries

DALRRD

Department of Agriculture, Land Reform and Rural Development

DBSA

Development Bank of South Africa

DCDT

Department of Communications and Digital Technology

DDS

dietary diversity score

DEA

Department of Environmental Affairs

Abbreviations   xxvii DFA

Dark Fibre Africa

DFI

development finance institution

DG

Disability Grant

DHET

Department of Higher Education and Training

DMRE

Department of Mineral Resources and Energy

DoT

Department of Transport

DP

Democratic Party

DPE

Department of Public Enterprises

DRDLR

Department of Rural Development and Land Reform

DSI

Department of Science and Innovation

DSMI

Data Services Market Inquiry

DST

Department of Science and Technology

EAC

East African Community

EAF

energy availability factor

ECA

Electronic Communications Act No.36 of 2005

ECD

early childhood development

ECNS

electronic communications network services

EDB

Ease of Doing Business index

EDI

Electricity Distribution Industry Holdings

EEA

Employment Equity Act

EFF

Economic Freedom Fighters

EFTA

European Free Trade Association

EPA

Economic Partnership Agreement

EPWP

Expanded Public Works Programme

EROSA

Economic Research on South Africa

ESA

East and Southern Africa

ESI

electricity supply industry

ETI

Employment Tax Incentive

EV

electric vehicle

EWC

expropriation without compensation

FABCOS

Foundation for African Business and Consumer Services

FAO

Food and Agriculture Organisation

FBE

Free Basic Electricity

FCI

Federated Chamber of Industries

xxviii   Abbreviations FDI

foreign direct investment

FEDUSA

Federation of Unions of South Africa

FMI

financial market infrastructure

FTE

full-​time equivalent

FTTH/​FTTB fibre to the home/​business GCI

Global Competitiveness Index

GEAR

Growth, Employment, and Redistribution

GEDI

Global Entrepreneurship and Development Institute

GEIS

General Export Incentive Scheme

GEM

Global Entrepreneurship Monitor

GEPF

Government Employees Pension Find

GETS

general-​to-​specific

GFB

General Freight Business

GFCF

gross fixed capital formation

GHG

Greenhouse gas

GHI

Global Hunger Index

GM

General Motors

GNI

gross national income

GPI

global performance indicator

GPN

Global Production Network

GRMI

Grocery Retail Market Inquiry

GUM

general unrestricted model

GVA

gross value ​added

GVC

global value chain

HDS

high demand spectrum

HDSS

Health and Demographic Surveillance Systems

HFIAS

Household Food Insecurity Access Scale

HPDD

Historical Public Debt Database

HS

Harmonized System

HSRC

Human Sciences Research Council

IBTS

Inclining Block Tariff System

ICASA

Independent Communications Authority of South Africa

ICE

internal combustion engine

ICLS

International Conference of Labour Statisticians

Abbreviations   xxix ICOR

incremental capital output ratio

IDC

Industrial Development Corporation

IDZ

industrial development zone.

IEA

International Energy Agency

IES

Income and Expenditure Survey

IFP

Inkatha Freedom Party

IGDP

Integrated Growth and Development Plan

IID

innovation for inclusive development

IIS

Impulse Indicator Saturation

IMVP

Institute for Motor Vehicle Productivity

INEP

Integrated National Electrification Programme

IP

intellectual property

IP

Internet Protocol

IPAP

Industrial Policy Action Plans

IPP

independent power producer

IPPO

Independent Power Producers Office

IRP

Integrated Resource Plan

ISCOR

Iron and Steel Industrial Corporation

ISI

import substitution industrialization

ITA

invitation to apply

ITU

International Telecommunications Union

JSE

Johannesburg Stock Exchange

JV

joint venture

KIDS

KwaZulu-​Natal Income Dynamics Study

LAMOSA

Landless Movement of South Africa

LCOE

levelized cost of energy

LCP

local content programme

LCS

Living Conditions Survey

LEDS

Low Emission Development Strategy

LFPR

labour force participation rate

LFS

Labour Force Survey

LMD

Labour Market Dynamics

LRA

Labour Relations Act

LTAS

Long-​Term Adaptation Scenarios

xxx   Abbreviations LTE

Long Term Evolution

LVPS

large value payment system

M&A

mergers and acquisitions

MALR

market-​assisted land reform

MARS

Migration and Remittance Survey

MDS

Market Demand Strategy

MEC

minerals-​energy ​complex

MERCOSUR Southern Common Market MERG

Macroeconomic Research Group

MIDP

Motor Industry Development Programme

MNO

mobile network operator

MPI

multidimensional poverty index

MPRDA

Mining and Petroleum Resources Development Act

MTBPS

Medium Term Budget Policy Statement

MW megawatt MYPD

multi-​year price determination

NACI

National Advisory Council on Innovation

NACTU

National Council of Trade Unions

NAFCOC

National African Federated Chamber of Commerce and Industry

NCR

National Credit Regulator

NEDLAC

National Economic Development and Labour Council

NEG

New Economic Geography

NELM

New Economics of Labour Migration

NER

National Energy Regulator

NERSA

National Energy Regulator of South Africa

NFC

non-​financial company

NSFAS

National Student Financial Aid Scheme

NGO

non-​governmental organization

NGP

New Growth Plan

NHN

National Hospital Network

NIBUS

National Informal Business Upliftment Strategy

NIDS

National Income Dynamics Study

NIPF

National Industrial Policy Framework

NMS

National Migration Survey

Abbreviations   xxxi NMW

National Minimum Wage

NP

National Party

NPA

National Ports Authority

NPC

National Planning Commission

NPFNS

National Policy for Food and Nutrition Security

NPO

non-​profit organization

NQF

National Qualification Framework

NRDS

National Research and Development Strategy

NSC

National Senior Certificate

NSDS

National Skills Development Strategy

NSF

National Skills Fund

NSI

national system of innovation

NT

National Treasury

NTB

non-​tariff barrier

NUM

National Union of Mineworkers

OAP

Old Age Pension

OFI

other financial intermediary

OHS

October Household Survey

OSBP

One Stop Border Post

OSD

Occupation Specific Dispensation

PALMS

Post-​Apartheid Labour Market Series

PC4IR

Presidential Commission on the Fourth Industrial Revolution

PCT

Patent Cooperation Treaty

PEAC

Presidential Economic Advisory Council

PHC

primary health care

PIRLS

Progress in International Reading Literacy Study

PI

Production Incentive (programme)

PIC

Public Investment Corporation

PIT

personal income tax

PJ

a peta joule per annum

PPA

power purchase agreement

PSEC

Presidential State-​owned Enterprise Council

PSLSD

Project for Statistics on Living Standards and Development

PV photovoltaic

xxxii   Abbreviations QES

Quarterly Employment Statistics

QLFS

Quarterly Labour Force Survey

QR

Quick Response

RDP

Reconstruction and Development Programme

REC

Regional Economic Communities

RED

regional electricity distributor

REIPP

Renewable Energy Independent Power Producer Procurement Programme

RISDP

Regional Indicative Strategic Development Plan

RMB

Rand Merchant Bank

ROA

return on assets

ROE

return on equity

RR

required revenue

RVC

regional value chain

SAA

South African Airways

SAAM

South African Automotive Masterplan

SAB

South African Breweries

SACReD

Systematic Analysis of Climate Resilient Development

SACU

Southern African Customs Union

SADC

Southern African Development Community

SAM

social accounting matrix

SAMP

Southern African Migration Project

SAMRC

South African Medical Research Council

SANHANES South African National Health and Nutrition Examination Survey SANRAL

South African National Roads Agency

SAP

Het Volk/​South African Party

SAPRIN

South Africa Population Research Infrastructure Network

SAQA

South African Qualifications Authority

SARB

South African Reserve Bank

SARS

South African Revenue Service

SASOL

South African Coal, Oil and Gas Corporation

SASSA

South African Social Security Agency

SATRA

South African Telecommunications Regulatory Authority

SAVA

South African VANS Association

Abbreviations   xxxiii SBI

Small Business Institute

SciSTIP

DSI-NRF Centre of Excellence in Scientometrics and Science, Technology and Innovation Policy

SDG

Sustainable Development Goal

SDI Spatial Development Initiative SDP Supplier Development Programme SEA

social entrepreneurship activity

SEDA

Small Enterprise Development Agency

SEE

Survey of Employment and Earnings

SEP

strategic equity partner

SESE

Survey of Employers and the Self-​employed

SETA

Sector Education Training Authorities

SEZ

special economic zone

SHV

shareholder value

SIU

Special Investigating Unit

SLC

substantial lessening of competition

SLLDP

State Land Lease and Disposal Policy

SMP

significant market power

SNO

second national operator

SOC

state-​owned company

SPS

sanitary and phytosanitary standards

SRD

social relief of distress

StatsSA

Statistics South Africa

TDCA

Trade, Development and Cooperation Agreement

TEC

Transport Economic Council

TER

Transport Economic Regulator

TFP

total factor productivity

T-​FTA

Tripartite Free Trade Area

TIMSS

Trends in Mathematics and Science Study

TIPS

Trade and Industrial Policy Strategies

TNC

transnational corporation

TNPA

Transnet National Ports Authority

TPSF

Trade Policy and Strategy Framework

TPT

Transnet Port Terminal

xxxiv   Abbreviations TUS

time-​use survey

TVET

technical and vocational education and training (TVET) 33:2

TYIP

Ten-​Year Innovation Plan

UCT

United Conference Ticket

UIF

Unemployment Insurance Fund

UNFCCC

United Nations Framework Convention on Climate Change

USA

Universal Service Agency

USAASA

Universal Service and Access Authority of South Africa

USF

Universal Service Fund

VPN

virtual private networks

WBES

World Bank Enterprise Survey

WBS

Wireless Business Solutions

WDI

World Development Indicators

Contributors

Tania Ajam is an associate professor in public policy, finance, and economics at Stellenbosch University. An economist with broad experience in implementing fiscal policy and intergovernmental fiscal relations, she holds degrees in economics from the Universities of Cape Town and Cambridge, and a PhD in public management from the University of Pretoria. She served on the Financial and Fiscal Commission and as a non-​executive director of the South African Reserve Bank. She is a member of President Ramaphosa’s Economic Advisory Council. Fabio Andrés  Díaz Pabón is a researcher at the African Centre of Excellence for Inequalities Research (ACEIR) at the University of Cape Town and research associate at the Department of Political and International Studies at Rhodes University in South Africa. His most recent work looks at protests and armed conflict in Colombia and South Africa and the relation between conflict and development. His research is focused on issues of justice, development, inequality, and state-​building, aiming to connect theory with practice via different methods and methodologies. Channing Arndt currently heads the Environment and Production Technology Division at IFPRI. He began working in South Africa in 1998 and has worked closely with the Economic Policy Unit of the National Treasury since 2009. He has an established reputation for building institutional capacity in Mozambique, South Africa, Morocco, and Vietnam and within the African Economic Research Consortium. His programme of research has focused on agricultural development, poverty, growth, market integration, nutrition, gender, discrimination, HIV/​AIDS, technological change, trade policy, aid effectiveness, energy, bioenergy, climate variability, and the implications of climate change. Justin Barnes  is the chairperson of B&M Analysts, the executive director of the Toyota Wessels Institute for Manufacturing Studies, and an associate professor at the Gordon Institute of Business Science at the University of Pretoria. He is an international expert in global value chain dynamics in the automotive and apparel/​textile sectors. He has a long history of advising governments in sector industrial policy, and firms to build manufacturing competitiveness capabilities and skills. He has pioneered benchmarking and firm-​level upgrading competitiveness assessment methodologies in the clothing and autos value chains. He has also developed several industry–government partnerships through industry firm-​level clusters.

xxxvi   Contributors Haroon Bhorat  is professor of economics, and director of the Development Policy Research Unit (DPRU), at the University of Cape Town (UCT). His research interests cover labour economics, poverty inequality, and income distribution. He sits on the Presidential Economic Advisory Council, holds a national research chair, and is a non-​ resident senior fellow at the Brookings Institution, a research fellow at IZA, and an honorary research fellow at the Human Sciences Council. He has co-​authored and co-​edited a number of books on labour market and poverty issues in Africa, and has published more than two hundred academic journal articles and working papers. Manoel Bittencourt  is an extraordinary professor of economics at the University of Pretoria and a National Research Foundation-​rated researcher. He is also affiliated with the universities of Goettingen, Heidelberg, and Oxford, and with Economic Research Southern Africa. He holds a diploma in economics from the University of Warwick, and an MSc and PhD in economics from the University of Bristol. In addition, he is a member of the council of the Economic Society of South Africa and an associate editor of the South African Journal of Economics and Journal of Public Finance and Public Choice. Mduduzi Biyase  is director of the Economic Development and Well-​being Research Group (EDWRG) and is a senior lecturer at the School of Economics, University of Johannesburg. He is a researcher specializing in the field of development economics: exploring the remittances behaviour, poverty, unemployment, and inequality in South Africa. He has published widely on research relating to remittance behaviour, poverty, and economic growth. His research on poverty and economic growth has been published in a number of journals, including the Journal of Developing Areas and Frontiers in Finance and Economics. Anthony Black  is professor of economics at the University of Cape Town and director of Policy Research in International Services and Manufacturing (PRISM). His research is focused on industrial development, the automotive industry, regional integration, foreign direct investment, and employment. His latest book is a co-​edited volume entitled Value Chains in Sub-​Saharan Africa: Challenges of Integration into the Global Economy (2019). He was extensively involved in South Africa’s automotive policy development. He has also acted as an advisor or consultant to a number of other African governments as well as to international organizations including UNIDO and UNCTAD. Nicola Branson is a senior research officer in the Southern Africa Labour and Development Research Unit (SALDRU) in the School of Economics at the University of Cape Town (UCT). She holds a PhD in economics from UCT and has experience in quantitative research using household and longitudinal survey data and in managing longitudinal datasets such as that from the National Income Dynamics Study (NIDS). Currently, she heads up the Siyaphambili Post-​School Research group, where her research focuses on inequalities in access to post-​school education and subsequent transitions into the labour market.

Contributors   xxxvii Philippe Burger  is pro-​vice-​chancellor: poverty, inequality and economic development, professor of economics and vice dean of the Faculty of Economic and Management Sciences at the University of the Free State. In addition, he is a member of the Fiscal Policy and Financial Markets Task Force of the Lancet Commission on COVID-​19. He is a 2016/​17 Fulbright Exchange Scholar at the Center for Sustainable Development, Earth Institute, at Columbia University. From 2012 to 2014 he was president of the Economic Society of South Africa. He has also been a visiting scholar at the IMF and consultant to the OECD. Ronelle Burger is a professor at the Economics Department of Stellenbosch University. Her research considers strategies to protect and empower vulnerable and poor individuals, households, and communities. She is also a fellow at the Partnership on Economic Policy Laval University and also at Nottingham University’s Centre for Research in Economic Development and International Trade. She serves as an associate editor for Health Economics and Development Southern Africa. She has done consulting work for the World Bank, UNICEF, and the National Treasury. Nina Callaghan  is a researcher at the Centre for Complex Systems in Transition at Stellenbosch University. Her research focuses on governance, theory-​building of state capture, and geopolitical influence in Africa. She has authored chapters in and co-​edited Anatomy of State Capture (forthcoming 2021). She has coordinated a publication on the renewable energy transition in South Africa, provisional title, ‘South Africa’s Energy Transition: The REIPPPP’s Potential for Energy Democracy in South Africa’ (forthcoming). Her MPhil in sustainable development at Stellenbosch University focuses on governance dynamics in South Africa’s unfolding energy transition. Daniela Casale is a professor in the School of Economics and Finance at the University of the Witwatersrand. She is a development economist with twenty years of experience in research in the field of gender economics specifically. Other broad areas of study include labour, well-​being, health, and education. She has published widely across a variety of economics, development, and public health journals, among them, Feminist Economics, Economic Development and Cultural Change, Economics and Human Biology, Development Policy Review, Public Health Nutrition, and PLOS ONE. Horman Chitonge is professor of African Studies at the Centre for African Studies (UCT), a visiting research fellow at Yale University (Global Justice Programme), a research associate at PRISM, School of Economics (UCT), a visiting fellow at the African Studies Centre at Tokyo University of Foreign Studies. His most recent books include: Industrial Policy and the Transformation of the Colonial Economy in Africa (Routledge, 2021). Industrialising Africa (Peter Lang, 2019); Social Welfare Policy in South Africa: From the Poor White Problem to a Digitised Social Contract (Peter Lang, 2018); Economic Growth and Development in Africa: Understanding Trends and Prospects (Routledge, 2015).

xxxviii   Contributors Mark A. Collinson is co-​ director of the South African Population Research Infrastructure Network, a national research infrastructure aimed at strengthening population health, social and economic research, and links to public policy. He is a reader in population and public health at the SAMRC/​Wits Rural Public Health and Health Transitions Research Unit (Agincourt), University of the Witwatersrand. He has been PI of the INDEPTH Network Migration, Urbanisation and Health Working Group since 2003, and is co-​PI of a five-​year R01 award from the National Institutes of Health, called ‘Migration, Urbanisation and Health in Transition Settings’, which started in 2016. Kenneth Creamer  is a senior lecturer at the School of Economics and Finance at the University of the Witwatersrand (Wits) in Johannesburg, South Africa. He has master’s degrees in Law (Wits) and Financial Economics (SOAS, University of London) and a PhD in Economics (Wits). His research focuses on fiscal policy, monetary policy, open economy, and energy policy. He serves on the Management Committee of the South African Student Solidarity Foundation for Education, which raises funds for tertiary education students in need, and as a director of Creamer Media, publisher of Engineering News, Mining Weekly and Polity. He serves on South Africa’s Presidential Economic Advisory Council. Rod Crompton  is an adjunct professor at the Wits Business School, University of Witwatersrand, where he set up the African Energy Leadership Centre. Previously he ran Crompton Consulting, specializing in energy, regulation, and industrial policy. He has occupied the following positions: full-​time board member at the National Energy Regulator of South Africa, deputy director general at the Department of Minerals and Energy and has served on the boards of several companies. Currently he is a non-​ executive director at Eskom, Africa’s largest electricity utility. During the struggle years he was General Secretary of the Chemical Workers Industrial Union. Reena das Nair  is a senior researcher at the Centre for Competition, Regulation and Economic Development (CCRED); and programme coordinator and senior lecturer in the MCom in competition and economic regulation degree at the University of Johannesburg. She has worked for specialist economic consultancy Acacia Economics offering expertise in competition and regulatory economics. Prior to this, she was Programme Manager: Industrial Policy at Trade and Industrial Policy Strategies (TIPS), a not-​for-​profit research organization undertaking research for policymakers. Before TIPS, she worked as principal economist at the Competition Commission of South Africa. She holds a PhD (economics) from the University of Johannesburg. Ciaran Driver  is professor of economics at the School of Finance and Management, SOAS University of London. He works on capital investment, corporate governance, R&D, dividends, and economic policy. Publications include Driver and Temple, The Unbalanced Economy (Palgrave Macmillan, 2014) and Driver and Thompson (eds), Corporate Governance in Contention (Oxford University Press, 2018). He has taught eco­nomics at Imperial, Syracuse, Notre Dame, and NYU. Public policy

Contributors   xxxix work includes consultancy for the EC and the Government Office for Science (United Kingdom). Work on South Africa includes a study of the BER survey forecasts (Journal of Forecasting, 2019). Lawrence Edwards  is a professor in the School of Economics, University of Cape Town, and a research associate at the South African Labour and Development Research Unit (SALDRU), Policy Research on International Services and Manufacturing (PRISM), and the Institute of Labor Economics (IZA). His research interests focus on international trade, trade policy, firms, and labour markets. He has consulted widely on trade issues for institutions including the World Bank, the African Development Bank, the International Growth Centre, the Southern African Development Community Secretariat, and governments in Africa, including South Africa, Swaziland, Lesotho, Zimbabwe, and Zambia. Arabo K. Ewinyu  is a research manager at the Southern Centre for Inequality Studies, University of the Witwatersrand. She has an MCom in economics from the University of Witwatersrand. Her research interests include labour economics, public economics, and poverty and inequality studies. Prior to joining SCIS, she worked at the University of Cape Town and at Genesis Analytics. David Francis  is the deputy director at the Southern Centre for Inequality Studies, and a PhD candidate in economics at Wits University. His research interests focus on labour market economics, the informal economy, and inequality. He has previously worked as a researcher, and a policy and budget analyst at South Africa’s National Treasury. He has an MA in development studies from the University of KwaZulu-​Natal, and a bachelor of social sciences in economics and history from the University of Cape Town. Bill Freund  † was emeritus professor in economic history in the School of Built Environment and Development Studies at the University of KwaZulu-​Natal. He is the author of many books on African history and South African development, including Twentieth-​ century South Africa: A Developmental History (Cambridge University Press, 2019). Sherwin Gabriel is a scientist at the International Food Policy Research Institute, working in the Global Futures and Strategic Foresight team. He contributes to research using global models of agricultural commodities and trade, and on economy-​ wide models. He applies these models to questions of climate-​change risk, food policy interventions, energy planning, and inclusive economic development. Before joining IFPRI, he was a director at the National Treasury of South Africa, where he managed and conducted economic modelling and research for policy analysis, and macroeconomic forecasts to support government planning. Vusi Gumede  is Dean of the Faculty of Economics, Development and Business Sciences at the University of Mpumalanga. Prior to that he was a professor at the University of South Africa and director at Thabo Mbeki African Leadership Institute. Before that he was an associate professor at the University of Johannesburg. He occupied various

xl   Contributors positions in the South African government before joining academia. He is widely published, including fourteen books and over forty journal articles and book chapters. He holds a PhD in economics from the University of KwaZulu-​Natal. Before joining government he worked as a researcher for various institutions. Adam Habib is Director of SOAS, University of London. A professor of Political Science, he has over 30 years of academic, research and administration expertise, spanning five universities and multiple local and international institutions. Prior to his appointment as director of SOAS, he was, between 2013 and 2020, Vice-​Chancellor and Principal of the University of the Witwatersrand, Johannesburg, South Africa. He has published extensively over the last three decades in the thematic areas of democratization in South Africa, social movements, philanthropy, inequality, institutional reform, changing identities, and Higher Education. His latest book, Rebels and Rage was on the FeesMustFall student protests in South Africa. Katharine Hall is a senior researcher at the Children’s Institute, a policy research unit at the University of Cape Town. She has a PhD in development theory and policy from the University of the Witwatersrand and a master’s degree in sociology from the University of Cape Town. She conducts applied research in the areas of child poverty and inequality, household and family contexts, mobility, and care arrangements. She coordinates Children Count, an ongoing project that draws on large national datasets to monitor social indicators related to children’s rights, and provides a basis for shadow reporting to international rights bodies. Ruth Hall holds the South African research chair in poverty, land, and agrarian studies, supported by the Department of Science and Innovation and the National Research Foundation. She is a professor at the Institute for Poverty, Land and Agrarian Studies (PLAAS) at the University of the Western Cape. She holds a DPhil in Politics and International Relations from the University of Oxford and has over many years conducted research on land reform, land rights, and agrarian change. Laurence Harris  is an emeritus professor of economics at the School of Finance and Management, SOAS, University of London and founder of CeFiMS. He is academic leader of the Macroeconomic Modelling workstream of the UNU-​WIDER programme Southern Africa: Towards Inclusive Economic Development (SA-​TIED). He has taught economics at LSE, University of California, Berkeley; University of Zimbabwe; Harvard University; Birkbeck, University of London; The Open University, and elsewhere. Faaiqa Hartley  is a researcher at the University of Cape Town, whose research has centred on developmental issues in Africa, primarily South Africa. She has over ten years of experience in policy development and analysis. She is proficient in economy-​ wide modelling techniques and formed part of the team that developed and maintained the suite of computable general equilibrium (CGE) models (including the South African General Equilibrium, SAGE, model) at the National Treasury of South Africa. Her current research areas include the use of CGE models to understand and quantify

Contributors   xli the economic trade-​offs and implications of energy, waste, and climate-​change policies in South Africa and sub-​Saharan Africa. Penelope Hawkins is a senior economist at the Debt and Development Finance Branch of UNCTAD, within the Division on Globalization and Development Strategies. Previously she was the managing director and founder of Feasibility (Pty) Ltd, a policy research consultancy specializing in financial regulation, consumer credit, and consumer protection in sub-​Saharan Africa. She studied at the University of the Witwatersrand, University of South Africa, and the University of Stirling and has lectured at Wits, Rhodes University, and the University of the Free State. James Heintz is the Andrew Glyn Professor of Economics at the University of Massachusetts, Amherst. He has worked on collaborative projects with numerous national and international institutions, including the Office of the High Commissioner for Human Rights, the International Labour Organization, the United Nations Economic Commission for Africa, the United Nations Development Programme, the Human Development Report Office, the South African Human Rights Commission, the International Development Research Center (Canada), and UN-​Women. His policy work has included work in developing countries, primarily in sub-​Saharan Africa, including Ghana, Kenya, Liberia, the Gambia, Madagascar, and South Africa. Alan Hirsch  is a professor at The Nelson Mandela School of Public Governance at the University of Cape Town. He joined the Department of Trade and Industry in 1995 and later managed economic policy in the South African Presidency, represented the President in the G20, and was co-​chair of the G20 Development Working Group. He serves on the board of the European Centre for Development Policy Management, the International Advisory Board of the New Development Bank, and President Ramaphosa’s Economic Advisory Council. His books include Season of Hope—​ Economic Reform under Mandela and Mbeki and The Oxford Companion to South African Economics. James Hodge  is the chief economist of the Competition Commission. He was formerly a researcher at the Development Policy Research Unit (1995–​2000), a senior lecturer at the University of Cape Town (2000–​5) and managing partner of the Competition and Regulatory Economics practice at Genesis Analytics (2005–​19). Roula Inglesi-​Lotz is a professor in economics at the Department of Economics, University of Pretoria. She has authored and co-​authored more than seventy papers in peer-​reviewed academic journals of high quality. Her research interests focus on energy and environmental economics from a macroeconomic perspective. She is the founding president of the South African Association for Energy Economics (SAAEE), and served as a co-chair of the South African Young Academy of Sciences (SAYAS) in 2020 and a co-chair of the Global Young Academy (GYA) in 2021/22. She was named the Distinguished Young Woman Researcher in the Humanities and Social Sciences category of the 2017 DST Women in Science Awards.

xlii   Contributors David Kaplan  taught at the University of Cape Town for thirty-​eight years and before that at the University of Massachusetts. He was chief economist of the Department of Trade and Industry (2000–​3) and chief economist (part-​time), Department of Economic Development and Tourism, Provincial Government of the Western Cape (2004–​11). He served on the Presidential Commission to Investigate the Development of a Comprehensive Labour Market Policy and on the National Advisory Council on Innovation (NACI) and is a former board member of the Technology Innovation Agency (TIA). He is currently senior research specialist at the International Science Council in Paris. Ewa Karwowski  holds a PhD from SOAS, University of London. She is lecturer in Economics King’s College London, and senior research associate at the University of Johannesburg, South Africa. Her research focuses on finance and development, financialization, and the political economy of development. She has consulted for various international organizations, including the International Labour Organization and the OECD. During her ODI fellowship (2008–​10), she worked as policy adviser for the National Treasury, South Africa. Karwowski is on the board of the Post-​Keynesian Economics Society, member of FinGEO, IIPPE, and a founding member of Reteaching Economics. Johann Kirsten is professor and director of the Bureau for Economic Research (BER) at the University of Stellenbosch. He was previously a professor and head of the Department of Agricultural Economics at the University of Pretoria. He served as a council member of the National Agricultural Marketing Council in South Africa from 2001 to 2011 and was also appointed as chair of the Food Price Committee during 2003/​ 4. He also served as the vice-​president of the International Association of Agricultural Economists for the period 2006–​9. Erika Kraemer-​Mbula is professor of economics at the College of Business and Economics, University of Johannesburg, South Africa, where she heads the DST/​ NRF/​Newton Fund trilateral chair in transformative innovation, the Fourth Industrial Revolution and sustainable development. She holds a master’s degree in science and technology policy from the Science and Policy Research Unit (University of Sussex), and a doctorate in development studies from the University of Oxford. She specializes in science, technology, and innovation policy analysis and innovation systems in connection to equitable and sustainable development. She is visiting professor at the UCL Institute for Innovation and Public Purpose. David Lam  is professor of economics, and research professor in the Population Studies Center at the University of Michigan. He is honorary professor of economics at the University of Cape Town and a research associate of the National Bureau of Economic Research. He has worked extensively in Brazil and South Africa, where his research analyses link between education, labour markets, and income inequality. He has served as president of the Population Association of America and has served on the Council of the International Union for the Scientific Study of Population (IUSSP).

Contributors   xliii Murray Leibbrandt  is NRF research chair in poverty and inequality research, and director of the African Centre of Excellence for Inequality Research (ACEIR) and of the Southern Africa Labour and Development Research Unit (SALDRU) at the University of Cape Town. His research focuses on South African poverty, inequality, and labour market dynamics using survey data and, in particular, panel data. He is a UNU-​WIDER non-​resident senior research fellow and a member of several international and pan-​ African research initiatives. Brian Levy is a professor at the Mandela School at the University of Cape Town and at the School of Advanced International Studies, Johns Hopkins University in Washington, DC. He worked at the World Bank from 1989 to 2012. He has published widely on the interactions among institutions, political economy and development policy. His most recent books are Working with the Grain: Integrating Governance and Growth in Development Strategies (Oxford University Press, 2014; info at https://​www. workingwiththegrain.com) and The Politics and Governance of Basic Education: A Tale of Two South African Provinces, edited by Brian Levy, Robert Cameron, Ursula Hoadley, and Vinothan Naidoo. Rasigan Maharajh is concurrently Node Head of the Department of Science and Innovation and National Research Foundation’s Centre of Excellence in Scientometrics and Science, Technology and Innovation Policy; the founding chief director of the Institute for Economic Research on Innovation at the Tshwane University of Technology; and an associate research fellow of the Tellus Institute in Boston. Neva Seidman Makgetla  is senior economist: trade and industrial policy at Trade and Industrial Policy Strategies (TIPS). She was deputy director general for economic policy in the Economic Development Department from 2010 to 2015. Before joining EDD, she worked for the Presidency, the DBSA, and COSATU as well as other government departments. Prior to 1994 she worked in various universities in Africa and the United States. She is a member of the National Minimum Wage Commission. Her research and publications centre on industrial policy and value chain analysis, and on socio-​economic challenges facing South Africa, especially employment creation and inequality. Tshilidzi Marwala, PhD, is vice-​ chancellor and principal of the University of Johannesburg and a full professor in its Faculty of Engineering and the Built Environment. An author of over fifteen books and hundreds of articles on artificial intelligence, machine rationality, social science, and leadership among others. He also serves as deputy chairperson of the South African Presidential Commission on the Fourth Industrial Revolution and is a member of the Namibia Fourth Industrial Revolution Task Force. Ruwadzano Matsika  is a Carnegie research associate at the Global Change Institute at Wits University. Her research areas of interest and expertise are energy transitions, climate risk, and sustainable development. Her current research is investigating pathways

xliv   Contributors to greening industry in South Africa and managing emerging climate risks across different sectors in society. She holds a doctorate in ecology and environmental sciences (Wits), a postgraduate diploma in energy leadership (Wits Business School), and has completed a course in climate-​related financial risk from the University of Oxford. Julian May  is director of the Centre of Excellence in Food Security at the University of the Western Cape and holds the UNESCO chair in African food systems. He is an economist with a PhD in development studies. He has worked on options for poverty reduction including land reform, social grants, information technology, and agriculture throughout Africa. His current research focuses on food security, childhood deprivation and malnutrition. He has edited seven books and published over ninety papers in books and academic journals. He serves on the Academy of Science in South Africa (ASSAf) Council. Cecil Mlatsheni  is a senior lecturer at the School of Economics at the University of Cape Town. He is also a research associate at the Southern Africa Labour and Development Research Unit (SALDRU). His research interests include labour market analysis, with a particular focus on youth transitions from schooling to work. He has recently served as a principal investigator on the National Income Dynamics Study (NIDS) and as a commissioner on the Employment Conditions Commission (ECC). Liberty Mncube  is a managing director at FTI consulting and an associate professor of economics at the School of Economic and Business Sciences of the University of the Witwatersrand. He is a Member of the Presidential Economic Advisory Council (PEAC). He is a former chief economist of the Competition Commission of South Africa. He served as chief economist from January 2014 to February 2019. Prior to joining the Competition Commission in September 2007, he was a researcher at the Development Policy Research Unit (DPRU), University of Cape Town. Pamela Mondliwa  is a senior managing consultant at BRG and research associate at the Centre for Competition, Regulation and Economic Development (CCRED), University of Johannesburg. She is an experienced economist and has worked across consulting, academia, and policy. Prior to joining BRG she was a senior researcher at CCRED and led the research on industrial development. She has worked extensively on issues of industrial development, competition, and economic regulation in southern Africa. Mike Morris  is emeritus professor in the School of Economics, University of Cape Town and honorary research associate in the Institute of Development Studies at the University of Sussex. His major activities have centred on research and policy work, in academia and consulting to the private sector, government, and international agencies. He has published over seventy articles in international journals and over thirty-​five book chapters in the fields of development, global and regional value chains, political economy, industrial clusters, international competitiveness, industrial policy, commodity-​based industrialization paths for Africa, renewable energy, as well as on the apparel and automotive industries.

Contributors   xlv Jacqueline Mosomi  is a junior Research Fellow in the Southern Africa Labour and Development Research Unit (SALDRU) at the University of Cape Town. She holds a PhD in Economics, and her thesis focus was on the trends in female labour force participation, employment, and the gender wage gap in post-​apartheid South Africa. Her current research within SALDRU’s Siyaphambili Post-​School Research Group focuses on gender, education, and labour market outcomes. Farai Mtero  is a senior researcher at the Institute for Poverty, Land and Agrarian Studies (PLAAS) at the University of the Western Cape, South Africa, where he undertook research for his PhD. His research focus is on redistributive land reform, rural development, social differentiation, class formation, and agrarian change. His work uses political economy theory and analysis to understand processes of agrarian change in contemporary societies. He is currently leading research on land redistribution in South Africa and specifically focuses on the role of redistributive land reform in societal transformation. Karmen Naidoo is a doctoral candidate in economics at the University of Massachusetts, Amherst. She is also a senior research associate at the DST/​NRF South African research chair in industrial development (SARChI), University of Johannesburg. She holds master’s degrees in economics from the University of Cape Town and SOAS, University of London. Her research interests are in the areas of development economics, labour economics, and structural change. In particular, she is interested in the impacts of trade and technology on the labour market. Bhaso Ndzendze, PhD, is a senior lecturer and head of department in the Department of Politics and International Relations at the University of Johannesburg. His areas of research and teaching include technology dynamics in, international relations, digital policy, and the Fourth Industrial Revolution. He holds certificates in executive programs on artificial intelligence and blockchain from the Massachusetts Institute of Technology and a PhD in international relations from the University of the Witwatersrand. Mosima Ngwenya  is a lecturer and health economics PhD candidate at the Economics Department of Stellenbosch University. She is interested in health inequalities, mental health, and body image. She is also interested in the use of qualitative and mixed-​ method approaches in economic research. She obtained her economics honours and master’s degrees from North-​West University. Musa Nxele  is a lecturer at the Mandela School, where he convenes master’s courses and research related to the political economy of development. He is doing a joint PhD in economics and development policy and practice at Université de Paris 1 Panthéon-​ Sorbonne and UCT. He also holds master’s degrees from both universities, in economic development and globalization. Before moving to tertiary education teaching at Rhodes University in 2017, he worked in the private sector in the fields of investment banking and industrial development consulting.

xlvi   Contributors Arkebe Oqubay, PhD, is a senior minister and special advisor to the prime minister of Ethiopia and has been at the centre of policymaking for over twenty-​five years. He is the former mayor of Addis Ababa, winner of the Best African Mayor of 2005, and finalist in the World Mayor Award 2005. He is a recipient of the Order of the Rising Sun, Gold and Silver Star presented by the Emperor of Japan. His recent works include the path-​breaking Made in Africa: Industrial Policy in Ethiopia; The Oxford Handbook of the Ethiopian Economy; African Economic Development: Evidence, Theory, and Policy; The Oxford Handbook of Industrial Hubs and Economic Development; and The Oxford Handbook of Industrial Policy. Vishnu Padayachee  † was distinguished professor and Derek Schrier and Cecily Cameron Chair in development economics, in the School of Economic and Business Sciences at the University of the Witwatersrand. He was a lifetime fellow of the Society of Scholars at Johns Hopkins University through the Paul Nitze School of Advanced International Studies, Washington, DC. Tamara Paremoer  is an economist with more than fifteen years’ experience in the public and private sectors. She is the divisional manager of Mergers and Acquisitions at the Competition Commission of South Africa. Leila Patel  is professor of social development studies and DST/​NRF research chair in welfare and social development, University of Johannesburg. Her research interests include social welfare policy, social protection, poverty reduction, gender, care, social services, and children and youth. Her recent books are Development, Social Policy and Community Action: Lessons from Below, edited with Marianne Ulriksen (HSRC Press 2017); Social Welfare and Social Development (Oxford University Press 2005) and Social Protection in Southern Africa, edited with James Midgley and Marianne Ulriksen (Routledge 2014). The Handbook on Social Protection and Social Development in the Global South (ed) (Edward Elgar 2021). Dorrit Posel  holds the Helen Suzman chair and is a distinguished professor in the School of Economics and Finance at the University of the Witwatersrand. She specializes in applied microeconomic work, and much of her research over the past thirty years has explored the interface between households and labour markets in South Africa. She is an associate editor of Feminist Economics and has published in a wide range of journals including Economic Development and Cultural Change, Demographic Research Journal of Human Development and Capabilities, Nature, Journal of Economic Methodology, and African Studies Review. Mzukisi Qobo is Head: of the Wits School of Governance, University of the Witwatersrand. He serves on President Cyril Ramaphosa’s Economic Advisory Council and the ministerial task team on the Agriculture Master Plan. He was previously chief director for trade policy at the Department of Trade and Industry. He has extensive experience in the fields of governance, strategy, and political economy. He obtained his PhD in international political economy from the University of Warwick, United Kingdom; an MA from the University of Stellenbosch; and a BA from the University of Cape Town.

Contributors   xlvii Simon Roberts  is a professor of economics at the University of Johannesburg, where he founded the Centre for Competition, Regulation and Economic Development (CCRED). He has worked extensively on issues of industrial development, trade, regional and global value chains, competition and economic regulation in Southern and East Africa, advising governments, competition authorities, and regulators. He has been an economics director at the UK Competition and Markets Authority (2019 to 2020) and the chief economist and manager of the Policy and Research Division at the Competition Commission of South Africa from 2006 to 2012. Michael Rogan is an associate professor in economics and economic history, and an active member of the Neil Aggett Labour Studies Unit (NALSU)—​both at Rhodes University. Since 2011, he has been a research associate in the Urban Policies Programme of the global research-​ policy-​ action network Women in Informal Employment: Globalizing and Organizing (WIEGO). He holds a PhD and an MA in development studies (social policy) and a BA in international studies (development). His research and publications over the past five years have focused largely on informal employment, gender, poverty, food security, education and skills development, and survey design. Stefan Schirmer  has taught economics and economic history at Wits University since 1992. He has published widely and was a visiting scholar at Oxford University and the London School of Economics. He has served on the editorial board of African Studies and was the founding editor of the journal Economic History of Developing Regions. Wandile Sihlobo is the chief economist of the Agricultural Business Chamber of South Africa (Agbiz) and the author of Finding Common Ground: Land, Equity, and Agriculture (Pan Macmillan, 2020). He is a visiting research fellow at the Wits School of Governance, University of the Witwatersrand. He was appointed to Cyril Ramaphosa’s Presidential Economic Advisory Council in 2019. He served on the Presidential Expert Advisory Panel on Land Reform and Agriculture. He is a member of the Council of Statistics of South Africa and a commissioner at the International Trade Commission of South Africa. He is a columnist for Business Day and Farmers Weekly. Caroline Skinner  is a senior researcher at the African Centre for Cities at the University of Cape Town and urban research director for the global policy research network Women in Informal Employment: Globalizing and Organizing. For over two decades her work has interrogated the informal economy and processes of informality. She has published widely on the topic. She has a long track record of policy and advocacy work at a local, provincial, national, and international level and provides technical support for worker-​based movements. Mills Soko  is professor of international business and strategy at Wits Business School. He is a member of the editorial boards of Global Governance, Journal of Common Market Studies, and Africa Growth Agenda. He is a member of the advisory boards of Namibia Business School and TSIBA Business School. He is also a member of the board of the

xlviii   Contributors Field Band Foundation. He previously chaired the board of the South African Institute of Advancement. And he was a member of the Warwick Commission on the Future of the Multilateral Trading System. Ben Stanwix  is a senior researcher at the Development Policy Research Unit (DPRU) within the School of Economics at the University of Cape Town (UCT). His main research interests are in the area of labour economics, primarily focused on South Africa. He holds a master’s degree in economics from UCT and a master’s degree in social history from the University of Oxford. Prior to joining the DPRU in 2013 and, he spent time working in India with the Self-​Employed Women’s Association. Kenneth Marc Strzepek  is a research scientist at MIT’s Joint Program on the Science and Policy of Global Change, a faculty fellow at the MIT Office of Sustainability and professor emeritus of civil, environmental, and architectural engineering, University of Colorado and consultant with Industrial Economics. He has spent over forty years as a researcher and practitioner at the nexus of engineering, environmental, and economics systems. Currently he is focused on the role of civil infrastructure in sustainable economic development and the design of climate-​resilient infrastructure. Mark Swilling  is distinguished professor of sustainable development and co-​director of the Centre for Sustainability Transitions at University of Stellenbosch. His latest book is The Age of Sustainability: Just Transitions in a Complex World (Routledge, 2020). He is a member of UNEP’s International Resource Panel acting as coordinator of the Cities Working Group and is on the board of the Development Bank of Southern Africa. He has been a visiting professor at the Universities of Sheffield and Utrecht, and in 2018 was the Edward P. Bass visiting environmental scholar at Yale University. Nicola Theron  specializes in the application of competition and regulatory economics. She has provided expert witness and numerous economic reports for various large clients in proceedings before the Competition Commission and the Competition Tribunal. She is an extraordinary professor in economics at Stellenbosch University, as well as a senior managing director at FTI Consulting where she leads the Economic and Financial Consulting (EFC) team in South Africa. In addition to competition economics, Theron has specialized knowledge of healthcare, telecommunications, and energy markets. She often presents on health-care policy (such as the NHI) in various forums. Timothy S. Thomas  is a research fellow in the Environment and Production Technology Division at IFPRI. His latest work focuses on climate uncertainty and weather variability, and how that affects agricultural production and food security in low-frequency, high-impact events. This extends his earlier modeling work which examined the impact of climate change on agriculture and what adaptation options would be best suited in response. Prior to IFPRI, he worked at the World Bank, focusing on land use change and tropical deforestation issues. Amy Thornton  is a post-​doctoral research fellow at the Southern African Labour and Development Research Unit (SALDRU) in the School of Economics at the University

Contributors   xlix of Cape Town (UCT). Between 2016 and 2020, she worked as a researcher at the Development Policy Research Unit. Her doctoral work was in the field of economic demography, focusing on household formation and quality of household survey data in South Africa. Fiona Tregenna  holds the DSI/​NRF South African research chair in industrial development (SARChI) and is a professor of economics at the University of Johannesburg. She has a PhD in economics from the University of Cambridge. She is a part-​time member of the Competition Tribunal where she adjudicates competition (anti-​trust) cases, and serves on a number of boards, advisory panels, and councils, including the Presidential Economic Advisory Council. She consults with various research institutes and international organizations such as UNIDO, UNCTAD, and the International Labour Organization. Her primary research interest is in issues of structural change, deindustrialization, and industrial development. Ivan Turok  holds the DSI/​NRF South African research chair in city-​region economies at Free State University. He is also a distinguished research fellow at the HSRC and honorary professor at Glasgow University. He is former editor-​in-​chief of Regional Studies, and a current editor of Area Development and Policy and Development Southern Africa. He chaired Durban’s City Planning Commission, and has a PhD in economics from Reading University, United Kingdom. He is occasional advisor to the United Nations, OECD, African Development Bank, UNECA, and several governments. He contributed to South Africa’s Integrated Urban Development Framework and the National Development Plan. Boris Urban, PhD, is a professor at the Wits Business School, University of Witwatersrand and was the inaugural chair in entrepreneurship. He has more than thirty years of academic and professional experience where he has practised, taught, and researched organizational behaviour, strategy, and entrepreneurship. His principal research agenda is to develop understanding of entrepreneurial behaviour within a cohesive framework by connecting individual, organizational, and societal features. Based on more than 120 journal articles and books, his work is considerably cited. Imraan Valodia  is dean of the Faculty of Commerce, Law and Management and director of the Southern Centre for Inequality Studies (SCIS) at the University of the Witwatersrand. His research interests include inequality, competition policy, employment, the informal economy, gender, and industrial development. He is a part-​time member of the Competition Tribunal in South Africa, a commissioner of the National Minimum Wage Commission, and a member of the Academy of Science of South Africa. He was appointed by President Cyril Ramaphosa to chair the Advisory Panel on the National Minimum Wage. He is a member of Ramaphosa’s Presidential Economic Advisory Council. Nicola Viegi  is the South African Reserve Bank professor of monetary policy studies at the University of Pretoria. He has held positions at the University of Strathclyde,

l   Contributors the University of KwaZulu-​Natal, and the University of Cape Town. He has been visiting fellow at the Nederlandsche Bank, the University of Milan ‘la Bicocca’, Newcastle University, Fordham University, the University of Malta, and the Ecole Superieure de Commerce in Toulouse. His areas of research are economic policy theory, monetary policy, macroeconomic modelling, and international macroeconomics. Thando Vilakazi  is executive director of the Centre for Competition, Regulation and Economic Development (CCRED) at the University of Johannesburg (UJ), specializing in academic research, teaching, and advice on competition policy and industrial development issues. He is a senior lecturer at UJ and currently also serves as a part-​time member of the Competition Tribunal of South Africa. He holds a PhD (economics) from UJ, and recently co-​edited an HSRC Press book titled Opening the South African Economy: Barriers to Entry and Competition.

PA RT   I

H I S TORY, P OL I T IC A L E C ON OM Y, A N D K E Y C HA L L E N G E S

CHAPTER 1

Challeng e s a nd C om plexitie s of t h e Sou th African E c onomy Fiona Tregenna, Arabo K. Ewinyu, Arkebe Oqubay, and Imraan Valodia

1.1 Introduction The South African economy is unique, yet also shares some characteristics, complexities, and challenges with other economies on the continent, and with middle-​ income countries in other regions. The structure of the economy continues to be shaped by the country’s colonial and apartheid legacy, to which its basic structure and ongoing difficulties can be traced in part. The persistently low rate of economic growth, and in particular the ‘triple challenges’ of poverty, unemployment, and inequality, have their roots in the apartheid period, when the economy was deliberately structured as non-​inclusive. In addition to the central dimensions of race and class, inequality was also manifested along gender, spatial, and other dimensions, with patterns of unemployment and poverty also characterized along similar lines. At the time of democratization, in 1994, South Africa faced colossal challenges of addressing the high rates of poverty, inequality, and unemployment, as well as the broader challenges of raising the rate of economic growth and of transforming the economy. As discussed below and in various chapters of this volume, progress has been uneven. It seems indisputable that different policy choices could have yielded better outcomes in raising growth rates and in dealing decisively with the ‘triple challenges’. The level of inequality and the rate of unemployment are among the highest in the world, together probably the highest, and poverty is extremely high for the country’s level of income per capita. While these challenges are shared with many other countries

4    Fiona Tregenna, Arabo K. Ewinyu, Arkebe Oqubay, and Imraan Valodia on the continent, and with a number of other middle-​income countries, the specific configuration of the triple challenges is specific to South Africa. In other respects, challenges facing the South African economy are common to many middle-​income and African countries. The ongoing dependence on natural resources and the ways in which South Africa’s historical dependence on mining has shaped the current economic structure, is shared with some other countries on the continent, and to some extent with various other middle-​income economies. So, too, are the macroeconomic challenges of a balance-​of-​payments constraint, and the labour market challenges of generating sufficient levels of employment. More broadly, the low rates of economic growth and the failure to ‘catch up’ with advanced economies, are characteristic of countries stuck in a ‘middle-​income trap’ (Gill and Kharas 2015). South Africa has failed to attain the rates of productivity growth, technological progress, and, ultimately, of sustained high growth in income per capita that would be necessary to close the gap with advanced economies (Andreoni and Tregenna 2020, 2021). In this chapter, we frame key economic issues of the South African economy by reflecting on some of its central challenges and complexities. Each of these issues is explored in greater detail in individual chapters of this volume. We discuss South Africa with reference to six relevant comparator countries. These include four significant middle-​income countries from other regions of the world: Malaysia, Turkey, China, and Brazil. The latter two are also included in the Brazil, Russia, India, China, and South Africa (BRICS) grouping, with Brazil in particular sharing some pertinent common characteristics with South Africa. We include India, a major low-​income developing country that is also a member of BRICS. Our final comparator country is Botswana, a neighbouring country with both commonalities and differences with South Africa. Of course, any selection of comparator countries is necessarily limited, especially when the same set of comparators is used for a range of issues, as here. Each country has its own unique history, political economy, resource endowments, geopolitical position, level of economic development, and so forth; these and many more country-​specific factors are essential to understanding its own characteristics and indicators. In this chapter, these diverse countries are utilized as simple comparisons for some of South Africa’s economic indicators and trends. We begin, in section 1.2, with the key issue of growth. We first discuss the rate of economic growth and then some aspects of the ‘nature’ of this growth, in terms of investment, the sectoral composition of the economy and structural change, innovation and technology intensity, and environmental sustainability. In section 1.3, we focus on the inclusivity of growth and the ‘triple challenges’ of poverty, inequality, and unemployment. Next, section 1.4 briefly reflects on these challenges in the context of South Africa’s transition and political economy. Some key sources of economic data in South Africa, in particular microeconomic data, are reviewed in section 1.5. This section serves as a reference for most of the datasets that are used and not discussed in detail in individual chapters.

Challenges and Complexities of the South African Economy    5

1.2  Economic Growth in South Africa 1.2.1 The Rate of Economic Growth During the post-​apartheid period, the South African economy has grown at low rates. Figure 1.1 compares gross domestic product (GDP) per capita in South Africa with that in our comparator countries, from the time of democratization in 1994 until 2019. This shows that South Africa did experience some economic growth between 1994 and 2006, but at a low rate. The period of (weak) expansionary growth in the democratic era coincided with growth in sub-​Saharan Africa, driven largely by global demand for commodities and fuelled by strong growth in East Asia, primarily China (Fedderke 2014; ­Chapters 6 and 41 in this volume). Consequently, many African economies experienced positive growth rates over a similar period. Rising global demand for commodities in the 2000s improved South Africa’s terms of trade. Hence, it is apparent that growth in the early post-​apartheid years was driven by world demand for commodities rather than by structural shifts in the economy that would result in greater competitiveness and a transformation of productive resources to ensure sustained high growth rates. 16 000

GDP per capita (constant 2010 US$)

14 000 12 000 10 000 8 000 6 000 4 000 2 000

19

9 19 4 9 19 5 9 19 6 9 19 7 9 19 8 9 20 9 0 20 0 0 20 1 02 20 0 20 3 0 20 4 0 20 5 0 20 6 0 20 7 08 20 0 20 9 1 20 0 1 20 1 12 20 1 20 3 1 20 4 1 20 5 1 20 6 1 20 7 1 20 8 19

–

South Africa

Brazil

Botswana

India

China

Turkey

Malaysia

Figure 1.1  GDP per capita in South Africa and comparator economies, 1994–​2019 Source: World Bank World Development Indicators (WDI). Note: GDP per capita data are in constant 2010 US$.

6    Fiona Tregenna, Arabo K. Ewinyu, Arkebe Oqubay, and Imraan Valodia Table 1.1: GDP per capita in South Africa and comparator economies, 1994–​2019 GDP per capita

Compounded annual growth rate, GDP per capita

1994

2019

2019/​1994

1994–​2006

2007–​19

South Africa

5,563.50

7,345.96

1.32

1.95

0.05

Botswana

4,332.03

8,092.97

1.87

2.81

1.97

Brazil

8,311.56

11,121.74

1.34

1.39

0.65

China

1,116.03

8,254.30

7.40

8.78

7.46

India

639.27

2,151.73

3.37

4.68

5.18

Malaysia

5,861.75

12,486.68

2.13

2.89

3.15

Turkey

6,889.37

15,125.39

2.20

3.44

2.9

Source: World Bank WDI. Note: GDP per capita data are in constant 2010 US$.

The global financial crisis in 2007/​08 marked the beginning of a slowdown in the rate of growth and, thereafter, growth has either plateaued or declined. As shown in Figure 1.1, most of the comparator countries experienced a decline in their growth coinciding with the global financial crisis, yet recovered better than South Africa did. In contrast, the South African economy never really recovered from this crisis. This can be understood in terms of the underlying structural weaknesses of the economy and the failure to put in place a solid foundation for sustainable growth.1 Relating the growth performance of South Africa to the comparator countries over the full post-​apartheid period, Table 1.1 shows South Africa to have had the lowest growth in income per capita (see the ratio between GDP per capita in 2019 and 1994), followed closely by Brazil. Botswana and China initially had lower income per capita, but overtook South Africa.

1.2.2 The Nature of Economic Growth We now review the nature of economic growth in South Africa with a particular focus on the ‘quality’, composition, and characteristics of growth that have fundamental implications for future growth prospects as well as current and future developmental

1  Chapter 6 gives a fuller account of the constraints of growth faced by South Africa. Briefly, these include historic racial and gender exclusion, shifting global trends and the emergence of China’s industrial capacity, policy, and macroeconomic imbalances that resulted in lower investment levels, and weaknesses in state capacity. Chapter 41 discusses the determinants of growth and suggests pro-​growth policies.

Challenges and Complexities of the South African Economy    7 45 40 35

GFCF (%)

30 25 20 15 10 5 0

South Africa

Botswana

Brazil

China 1994

India

Malaysia

Turkey

2018

Figure 1.2  Gross fixed capital formation in South Africa and comparator economies, % of GDP, 1994–​2018 Source: World Bank WDI. Note: Data not shown for Malaysia in 1994 due to an apparent break in the series.

outcomes. Here, we focus on investment; the sectoral composition of the economy and structural change; innovation and technological upgrading; and the environmental sustainability of growth.

1.2.2.1 Investment Investment is necessary for economic growth, as it stimulates total demand and catalyses future productive capacity. Broadly, this has strong implications for the sustainability of economic growth and future growth prospects. High rates of productive investment have been one of the hallmarks of growth success stories internationally. Figure 1.2 compares gross fixed capital formation (GFCF) as a percentage of GDP in South Africa and the six comparator economies, in 1994 and 2018. At the time of democratization, in 1994, South Africa had the lowest investment rate in this group. At its peak, GFCF surpassed 20 per cent between 2007–​09 and again in 2013–​15, before declining. The drop in the investment rate was even more pronounced in Brazil. Chapter 42 in this volume discusses investment in more detail.

1.2.2.2 Sectoral Composition of the Economy and Structural Transformation The sectoral composition of the economy matters for growth and developmental outcomes. Especially from a structuralist perspective, industrialization—​a shift in composition towards the manufacturing sector—​is crucial for growth and for developing countries ‘catching up’ with the advanced economies of the world (Andreoni et al. 2021; Blankenburg et al. 2008; Oqubay 2015; ­Chapter 17 in this volume).

8    Fiona Tregenna, Arabo K. Ewinyu, Arkebe Oqubay, and Imraan Valodia Table 1.2: Sectoral composition of GDP (%) in South Africa and comparator economies, 1994–​2019 Agriculture (% GDP)

Manufacturing (% GDP)

Services (% GDP)

1994

2019

1994

2019

1994

2019

South Africa

4.2

1.9

19.3

11.8

55.3

61.2

Botswana

3.8

2.0

4.9

5.2

45.4

58.2

Brazil

8.5

4.4

23.2

9.4

43.5

63.3

China

19.5

7.1

N/​A

27.2

34.4

53.9

India

26.4

16.0

16.8

13.6

37.5

49.4

Malaysia

13.7

7.3

26.6

21.4

47.9

54.2

Turkey

15.5

6.4

22.1

18.3

48.9

56.5

Source: World Bank WDI. Notes: N/​A means that the data are unavailable for the indicator in the reference year.

Table 1.2 shows the sectoral composition of the South African GDP for the key sectors of agriculture, manufacturing, and services between 1994 and 2019. Relative to the comparator countries, South Africa in 2019 had the lowest share of agriculture, the median share of manufacturing, and the second highest share of services. This arises in part from growth in the financial services sector and increased internationalization of larger South African businesses (Andreoni et al. 2021; ­Chapter 17 in this volume). Furthermore, the commodity boom of the 2000s attracted short-​term capital investors to the Johannesburg Stock Exchange, which accelerated the expansion of the financial services sector. The expansion of the services sector may also be attributed partly to changes in the statistical treatment of workers in the temporary employment services sector, which was classified as belonging to the business services sub sector (Bhorat et al. 2014; Tregenna 2010). Over this period, other service sectors such as retail and wholesale, health-care, and telecommunications also witnessed major growth. The authors of ­Chapter 17 argue that this growth in the services sector was not accompanied by increased investment or by broader structural transformation shifts to high-​value activities. Between 1994 and 2019, Brazil and South Africa experienced the most deindustrialization when measured simply in terms of manufacturing share of GDP, as summarized in Table 1.2 (Andreoni and Tregenna 2021; ­Chapters 17 and 24 in this volume).

1.2.2.3 Innovation and Technological Upgrading Innovation and technological upgrading are crucial for productivity and competitiveness, for avoiding a ‘middle-​income trap’, and for long-​run economic dynamism and growth (Andreoni and Tregenna 2020; C ­ hapters 22 and 23 in this volume). The increasing

Challenges and Complexities of the South African Economy    9 uptake of technologies associated with the Fourth Industrial Revolution (4IR)—​such as digitalization and robotization—​present both opportunities and challenges for employment creation and for closing the digital and developmental gaps with advanced economies. Table 1.3 presents three indicators—​of R&D, innovation, and technology intensity, albeit imperfect measures and with incomplete data coverage. Gross domestic expenditure on R&D as a percentage of GDP is indicative of R&D investment and an indicator of intensity; only India ranks lower than South Africa in this. Patent applications are one indicator of innovation. Between 1994 and 2018, the absolute number of patent applications from South Africa declined. While differences in country population size and GDP make comparisons of absolute numbers difficult, we observe that South Africa had the lowest number of patent applications among the comparator countries in 2018, despite having significantly more applications in 1994 than those of Malaysia and Turkey. The third measure shown in Table 1.3, medium-​and high-​tech percentage of manufacturing value added, is an indicator of technology intensity in manufacturing. This reflects the sub ​sectoral composition of the manufacturing sector and is in part an outcome of prior investments in technological upgrading. By this measure, South Africa has the second lowest technology intensity in manufacturing, after Botswana. These indicators bode poorly for South Africa’s future competitiveness, economic dynamism, and growth prospects. This points to the need for greater investment in R&D and other forms of innovation, and in technological upgrading.

Table 1.3: Selected measures of R&D, innovation, and technology intensity in South Africa and comparator economies, 1994–​2018 Gross domestic Patent applications, expenditure on R&D as residents % of GDP

Medium-​ and high-​tech % of manufacturing value ​added

1994

2017

1994

2018

1994

2018

South Africa

N/​A

0.83

935

657

30.23

24.43

Botswana

N/​A

N/​A

N/​A

N/​A

7.30

7.76

Brazil

N/​A

1.26

2,269

4,980

49.79

35.02

China

N/​A

2.15

11,191

1,393,815

35.52

41.45

India

N/​A

0.67

1,588

16,289

41.80

41.47

Malaysia

N/​A

1.44

223

1,116

48.73

44.01

Turkey

N/​A

0.96

151

7,156

27.88

32.15

Source: World Bank WDI, UNESCO. Notes: N/​A means that the data are unavailable for the indicator in the reference year. In the case of gross domestic expenditure on R&D as a percentage of GDP, no data is available for these countries for 1994 or a similar period, so countries are compared for 2017.

10    Fiona Tregenna, Arabo K. Ewinyu, Arkebe Oqubay, and Imraan Valodia

CO2 emissions (metric tons per capita)

12

10

8

6

4

2

00 20 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 11 20 12 20 13 20 14 20 15 20 16

99

20

98

19

97

19

96

19

95

19

19

19

94

0

South Africa

Brazil

Botswana

India

China

Turkey

Malaysia

Figure 1.3  Carbon dioxide emissions (metric tons per capita) in South Africa and comparator economies, 1994–​2016 Source: from World Bank WDI.

1.2.2.4 The Environmental Sustainability of Economic Growth Over the post-​1994 period, we observe that not only has growth been low and not inclusive, but it has also been unsustainable. Sustainability could be considered in various dimensions; here we consider environmental sustainability. One key indication of the environmentally unsustainable growth path in South Africa is the economy’s heavy dependence on coal as a source of energy. Figure 1.3 shows that South Africa’s CO2 emissions per capita are the highest among the comparator countries. This is discussed in more detail in several chapters in this volume, notably ­Chapters 6, 14, 15, and 16.

1.3  The ‘Triple Challenges’ of Poverty, Inequality, and Unemployment 1.3.1 The Inclusivity of Economic Growth It seems clear that South Arica’s growth path has not been inclusive. To begin with, we compare the rates of growth in employment and GDP. In the 2000s, the period

Challenges and Complexities of the South African Economy    11 coinciding with South Africa’s highest rate of economic growth, employment growth lagged GDP growth. This worsened in the 2008/​09 recession following the global financial crisis, as employment contracted more than did GDP. Over the 2000s, growth in the labour force far exceeded the growth of employment, with an associated significant rise in unemployment. This is the case whether unemployment is defined broadly or narrowly,2 and implies that employment growth was insufficient to absorb the rapidly expanding supply of workers, thereby highlighting the inability of the economy to create jobs at the same pace at which the labour force has been growing (Oosthuizen and Bhorat 2005). Employment shifts at the occupational level indicate a bias towards high-​skilled workers compared to the demand for unskilled and semi-​skilled workers (Chapters 31 and 33 in this volume). This is further compounded by the lack of racial and gender transformation in the more skilled and managerial occupations, which remain male and white twenty-​five years into democracy. Furthermore, women and African workers in general remain over-​represented in the lower-​skilled and more precarious types of employment. The triple challenges are clearly structurally interconnected. For instance, C ­ hapter 9 in this volume, on inequality, shows that labour market income is by far the largest determinant of income inequality. This arises from the significant proportion of individuals in households that lack access to any labour market income (see also Cramer et al. 2020; Tregenna 2011; and Tregenna and Tsela 2012). Tregenna (2012) shows how fundamental distribution is to different poverty outcomes. Table 1.4 shows key statistics on unemployment, poverty, and inequality, discussed further below. Each of these have very clear racial and gender dimensions. These international comparisons also throw the severity of South Africa’s triple challenges into stark relief. South Africa has the highest unemployment and poverty rates in 2018 and is the most unequal of these countries. While comparison of poverty rates across countries is always fraught and should be interpreted with caution (even when using a common poverty line as here), it is striking that South Africa’s headcount poverty rate is higher than that of a country such as India with significantly lower income per capita. South Africa has not experienced the improvement in poverty observed in the other countries during this period.

1.3.2 Unemployment As shown in Table 1.4, South Africa’s labour force participation rate (LFPR) has remained stagnant over an extended period of time. This likely reflects the combination of new entrants into the labour force and the exit of the ‘discouraged’ unemployed workers (­Chapter 7 in this volume). 2 The

narrow definition of unemployment corresponds to the official unemployment rate that is calculated by expressing the share of unemployed individuals as a proportion of the total employed workers. The broad or expanded definition also includes ‘discouraged’ work seekers.

12    Fiona Tregenna, Arabo K. Ewinyu, Arkebe Oqubay, and Imraan Valodia Table 1.4: Headline measures of unemployment, poverty, and inequality in South Africa and comparator economies, 1994–​2019 Labour force participation rate (% of total population)

Unemployment (% of total labour force)

Headcount poverty (% of population)

Gini index (equivalized)

1994

2019

1994

2019

1994

2018

1994

South Africa

59.56

60.08

30.14

28.18

33.35

19.28

60.8

62.5

Botswana

61.99

73.32

21.20

18.19

33.48

13.43

57.6

N/​A

Brazil

66.96

70.39

6.22

12.08

16.52

4.42

54.1

47.9

China

84.06

75.61

2.90

4.32

52.60

0.27

35.2

41.2

India

60.25

52.15

5.63

5.36

47.36

8.67

41.7

N/​A

Malaysia

64.03

67.93

3.62

3.32

1.66

0.01

43.5

N/​A

Turkey

56.98

58.08

8.58

13.49

2.95

0.04

43.1

40.2

2017

Source: World Bank WDI; World Bank POVCAL; Standardized World Income Inequality Database (SWIID), Versions 8–​9. Notes: 1. Poverty line of PPPUS$1.9/​day. 2. The Gini index is equivalized, using a square root scale, on the household disposable (post-​tax, post-​ transfer) income. 3. N/​A means that the data is unavailable for the indicator in the reference year.

The female LFPR in South Africa increased from approximately 44 per cent in 1995 to 49 per cent in 2015 (Casale and Posel 2002; Mosomi 2019). This increase is attributed to rising education levels (Casale and Posel 2002), declining marriage rates (Posel and Casale 2013) and extensive amendments to labour and employment legislation that institutionalized changes that provided women with greater access to the labour market (Posel 1999; Mosomi 2019). Still, in the fourth quarter of 2019, the participation rate for female workers was 12 percentage points lower than the male participation rate (Statistics South Africa [StatsSA] 2019). Furthermore, at 31.3 per cent, female unemployment remains significantly higher than male unemployment (27.2 per cent) (final quarter of 2019; StatsSA 2019). Younger workers experience higher unemployment rates relative to older workers. Persistently high youth-​ unemployment rates are a feature of the South African economy, and these figures are unmatched amongst similar middle-​income countries. For example, in the first quarter of 2020, 53 per cent of workers aged between 15 and 34 were unemployed. Chapter 32 in this volume discusses the market dynamics and unemployment relating to youth labour.

Challenges and Complexities of the South African Economy    13 Unemployment trends in the post-​apartheid era mirror historic discrimination patterns, as Black individuals face higher unemployment rates compared to whites. In line with the structural change in the economy, individuals with a post-​secondary qualification have lower unemployment rates, as the current growth trajectory of the South African economy appears to favour a minority of high-​skilled workers relative to the masses of unskilled or semi-​skilled workers3.

1.3.3 Poverty Initial inequality affects the overall pace of poverty reduction, regardless of the rate of economic growth. Furthermore, countries experiencing average growth rates and rising income inequality will realize a decline in average poverty levels (Ravallion 2001). However, this decline will be lower than in similar countries experiencing inclusive growth. Initially high inequality rates that later remain stagnant will have the effect of stifling pro-​poor growth. In keeping with global trends, the reference countries, including South Africa, all reduced the share of their population classified as poor over the reference period. In 1994, almost half the population in China and India was classified as poor. In 2019, this proportion had reduced to 0.3 and 8.7 per cent respectively. While South Africa also reduced its poverty headcount by almost half, we note that, in 2019, it had the highest poverty rate of all the seven countries. High and persistent inequality rates dampen the likelihood of any pro-​poor growth. Poverty trends in post-​apartheid South Africa are both racialized and gendered. Racially, we observe that, although the poverty gap between white-​and Asian-​headed households and those headed by Africans has declined, the latter population group is still more susceptible to poverty and continues to face high poverty rates (Leibbrandt et al. 2009). By gender, we note equally that, while female-​headed households have realized a significant decline in their poverty gap ratio, these households continue to face poverty rates that are almost twice as high as those headed by males (Leibbrandt et al. 2009). Ravallion (2001) highlights the fact that ensuring pro-​poor growth is closely tied to reducing disparities in access to human and physical capital as the asset-​endowment structure of an economy. Failure to reduce these access rates and increase the rates of return from the relevant assets will perpetuate unequal growth. Chapter 33 in this volume discusses various education trends in post-apartheid South Africa with particular focus on the impact of education on employment and earnings for working-​age adults. Chapter 8 in this volume discusses poverty measures and trends in greater detail. Government transfers have contributed somewhat to lessening the effects of high 3 

Chapter 7 in this volume analyses the problem of unemployment in South Africa, demonstrating its structural nature. The authors focus on unemployment arising from structural change, as well as provide microeconomic reasons from both a supply and demand side. Chapter 31 discusses the labour market broadly.

14    Fiona Tregenna, Arabo K. Ewinyu, Arkebe Oqubay, and Imraan Valodia poverty and inequality. Chapter 40 discusses South Africa’s social security and social development programmes in greater detail.

1.3.4 Inequality Inequality in the post-​apartheid era has remained high and has even increased (Bhorat and van der Westhuizen 2012; Leibbrandt et al. 2009; Wittenberg 2017a; Francis et al. 2020). From Table 1.4, we observe that inequality increased in South Africa between 1994 and 2019. Additionally, South Africa had the highest level of inequality among the sample of countries across both time periods. Furthermore, we observe that growth in China was accompanied by rising inequality. Conversely, the data show that Brazil and Turkey experienced a period of growth that coincided with lowered inequality. The findings indicate rising income inequality in African and coloured households, with no significant change in the levels of inequality among white-​and ​Asian-​headed households in the post-​apartheid period (Leibbrandt et al. 2009). ‘Within-​race’ inequality is increasing in relevance over time, and the Gini coefficient among Africans has increased significantly (Wittenberg 2017b). The importance of inequality among whites is declining, as this racial group constitutes a smaller proportion of the overall workforce (Wittenberg 2017b). South Africa’s spatial segregation also has a bearing on the observed levels of inequality and poverty. The greatest levels of need and deprivation remain concentrated in the former Bantustans and townships. Spatial issues are discussed in detail in C ­ hapter 30 of this volume. Structural shifts within the economy and the relationship between the labour market and poverty and inequality necessitate the discussion of whether access to the labour market, and thereafter the distribution of labour income earned, are a source of rising inequality. Research estimates that labour market inequality accounts for approximately 90 per cent of total income inequality (Leibbrandt et al. 2009, 2012; ­Chapter 9 in this volume). The South African economy is characterized as being heavily biased towards highly skilled individuals. Hence it is important to consider whether growing differentials between individuals with different skills and education levels are a further source of rising inequality. Wittenberg (2017b) estimates that 25 per cent of overall inequality can be attributed to inequality in earnings among individuals with a post-​secondary school qualification. Differences in returns to skills highlight existing inequalities in the quality of education received between well-​resourced and under-​resourced schools. It is also a reflection of poor outcomes relative to the high budgetary allocation (see also C ­ hapter 33 in this volume, which discusses education in greater detail). Inequality is the focus of Chapter 9 in this volume, in which the authors provide an explanation for the processes that generate and reproduce such high inequality in South Africa. Inequality is discussed along categorical measures such as race, gender,

Challenges and Complexities of the South African Economy    15 and space. Panel data is also analysed to discuss intergenerational and intragenerational mobility.

1.4  The Challenges of Economic Development in South Africa In our view, the central challenges discussed above—​the low rate of economic growth and its non-​inclusive and unsustainable nature, and the triple challenges of poverty, inequality, and unemployment—​are integrally intertwined. South Africa has continued along a growth path that fails to utilize the capacity and capabilities of a large section of the adult population, and in which the benefits of economic growth, however inadequate, do not reach this section in any meaningful way. The extreme levels of poverty, inequality, and unemployment in South Africa are not just manifestations of a non-​inclusive growth path, but are also constraints to growth itself. The triple challenges bring wasted human resources, a lack of social cohesion, social instability, and poor developmental outcomes, all of which constrain South Africa’s economic growth. We do not see any viable path to sustained high rates of economic growth that does not include fundamentally addressing the triple challenges. This suggests that, for policy, addressing the triple challenges is important not just in its own right, but as central tenets of any shift towards higher economic growth. This conceptual approach also points to the inadequacy of simplistic binary trade-​offs between equity and efficiency or, for instance, between productivity and employment, in the South African context. South Africa’s growth path needs to be located within an understanding of the underlying political economy dynamics. Indeed, political economy is crucial to understanding the issues explored in the chapters in this volume. The nature of South Africa’s transition to democracy in 1994 was arguably one in which the majority of South Africans were carried along, yet were not at the heart of the transformation project. Certainly, the economic lives of the majority of South Africans have improved, including through the meeting of basic needs via the provision of infrastructure and services such as housing and sanitation, through the opening up of economic opportunities, and for some through the receipt of social grants. To characterize South Africa’s transition as purely elitist would be simplistic and inaccurate, yet elites have certainly been most able to protect and advance their interests. A series of different economic policies have been implemented in the post-​apartheid era to propel growth and other various outcomes. These have had various shortcomings, not least of which are the divergent incentives of key players such as the state and capital. One of the key challenges in the post-​apartheid period is the inability of the various interest groups to cohere around a set of policies which address the key trade-​ offs and consensus that would bring coherence and action on the policy front. Much

16    Fiona Tregenna, Arabo K. Ewinyu, Arkebe Oqubay, and Imraan Valodia of the debate about policy choices is re-​enacted in various policy documents, such as the National Development Plan, with very little implementation. At least a part of the implementation challenge is due to the lack of consensus and coherence in policy. These challenges and the policy options are explored further in various chapters of this volume, including in ­Chapters 4 and 5, which focus on various dimensions of political economy and policymaking. As at the time of writing, it is becoming increasingly clear that the COVID-​19 crisis has both exposed and deepened the existing fault lines of the South African economy. This poses new challenges for an economy that has been characterized as being stuck in a “middle-​income trap” and that has failed to catch up to advanced economies, instead falling further behind. As already discussed, the fallout from the 2007/​08 global financial crisis was far reaching and longer lasting in South Africa than was typical internationally. Similarly, the effects of this health and economic crisis are sure to be long lasting. This, and the likelihood of other crises of various sorts in future, bring to the fore the need for South Africa to move on to a different growth path, in which the economy is structurally transformed and is more inclusive and sustainable.

1.5  A Note on South African Economic Data The chapters in this volume have utilized a range of datasets at the household, intermediate, and macroeconomic levels. These are described briefly in this section, with a particular focus on microeconomic data, for which fuller details are provided in Table 1.5 in section 1.6, Appendix. South Africa has several household-​level datasets that provide detailed information on household demographics, living conditions, and access to services, as well as a selection of labour market outcomes. The first nationally representative household survey, administered in 1993, was the Project for Statistics on Living Standards and Development (PSLSD). This questionnaire was administered to approximately nine thousand households in the nine months leading up to the first democratic election in April 1994 (Project for Statistics on Living Standards, 1994). The key objective of the project was to collect data on the living conditions of South Africans in order to enable policymakers to develop relevant policies and strategies to meet the goals identified in the Reconstruction and Development Programme (RDP). In October 1993, Statistics South Africa (StatsSA) conducted the first of a series of annualized nationally representative household surveys known as the October Household Survey (OHS). The OHS was later replaced partly by the Labour Force Survey (LFS), which was administered from February 2000 to the first quarter of 2008, when it was replaced by the Quarterly

Challenges and Complexities of the South African Economy    17 Labour Force Survey (QLFS). These three surveys are each discussed in further detail below. Annual iterations of the initial OHS survey often changed their sampling frame in efforts to improve representativeness. The 1993 OHS excluded the former Bantustans of Transkei, Bophuthatswana, Venda, and Ciskei, which resulted in under-​sampling the total number of Black South Africans relative to others in that year (Yu 2007). Households in the former Bantustans were included in the 1994 OHS, but sampling of these households was unreliable, as only approximate estimates of population sizes were used and emerging informal settlements were not included in the sampling frame (Yu et al. 2017). Surveys undertaken before 1995 used sampling frames based on the 1991 census. This changed in the 1996 OHS (Yu et al. 2017). In 1998, the sampling frame was again adjusted to adequately cover individuals residing in mining hostels (Statistics South Africa 2000). The 1995 OHS coincided with the first Income and Expenditure Survey. In this, StatsSA utilized a more representative sample that included more of those households that had been omitted in previous surveys. This improved coverage in 1995 makes direct comparison with the previous two OHS surveys difficult, and researchers typically omit these two surveys and commence analysis from 1995. The first StatsSA master sample was developed in 1999 from the 1996 census. This master sample was relied upon to draw a sample for the 1999 OHS and the first LFS in 2000, until a new master sample based on the 2001 Census was introduced. Transitioning to the master sample in 1999 was significant, as it meant that enumerators would henceforth interview all households residing at the sampled dwelling unit, unlike the previous sampling procedure that mostly ignored any small households (Kerr and Wittenberg 2013, as cited in Yu et al. 2017). The LFS followed the OHS and was designed to capture all forms of work more rigorously than its predecessor. This was undertaken by emphasizing that all forms of small-​ scale activities, such as informal work and subsistence agriculture, could be classified as self-​employed work, provided that individuals had participated in the activity for even an hour in the previous week. This category of workers would otherwise have been classified as inactive or unemployed in previous surveys (Casale et al. 2005; Yu 2007). This shift is significant because, until 1996, the OHS did not provide a prompt for respondents explaining what was viewed as work. The majority of these additionally enumerated workers are characterized as working few hours and earning low wages (Wittenberg 2017b). Unlike the OHS, which had independent cross-​sections, the LFS was designed to include a rotating panel. This sampling methodology was maintained at the introduction of the QLFS, where selected dwelling units would remain in the survey for four consecutive periods and exit the survey thereafter. The QLFS replaced the LFS in 2008, partially in response to various criticisms related to the scope, coverage, timeliness, and frequency of the LFS survey. Beginning in 2005, StatsSA undertook significant revisions of the LFS. These resulted in changes

18    Fiona Tregenna, Arabo K. Ewinyu, Arkebe Oqubay, and Imraan Valodia to the survey questionnaire, methodology, frequency of data collection, and utilizing automated data capturing and processing systems. The QLFS is administered at the household level to individuals aged 15 and above to collect detailed data on the individual’s labour market status. Earnings data from the QLFS are released once a year in the form of labour market dynamics data (LMD). Sampling issues, as well as other limitations of the data, make comparability across different time periods and surveys difficult. This led to the creation of the Post-​ Apartheid Labour Market Series (PALMS). This valuable data compilation allows users to use a version of the data that is easily comparable and wherein definitions have been standardized. PALMS is a stacked cross-​sectional dataset consisting of sixty-​nine household surveys conducted by StatsSA between 1994 and 2019. It also includes the 1993 PSLSD. In addition to these household surveys, a growing body of research utilizes an administrative tax dataset provided by the South African Revenue Service (SARS), which is the country’s statutory tax authority, and the National Treasury (NT) in 2015 (the SARS-​ NT data). This dataset is available to researchers by application and under restricted conditions, and consolidates four sources of tax data (see Arndt et al. 2018; Pieterse et al. 2018), providing rich information on firms’ balance sheet variables in particular. Such administrative data enable researchers to study the reported mismatch between earnings reported in the QLFS and actual earnings (Wittenberg 2017b). This will have adverse effects on poverty and inequality estimates. Using administrative tax data also enables researchers to understand income dynamics at the upper end of the earnings distribution, as such individuals are either under-​sampled in the survey data or, where included, often refuse to provide their earnings. Non-​participation is also highest in the more affluent areas (Wittenberg 2017a). While South Africa enjoys relatively rich microeconomic data at the household level, this is sparse at the firm level. There are no publicly available, comprehensive, and recent national firm-​level datasets. StatsSA currently provides such data for selected sectors, for example for the manufacturing sector and for certain sub ​sectors, but this is not comprehensive. The Survey of Employment and Earnings (SEE) was a quarterly survey covering a sample of public and private enterprises. Participating firms were all registered for VAT with a minimum turnover of R300,000, indicating that they were within the formal non-​agricultural sector of the economy. Information so received is an input for the gross domestic product. This survey was discontinued and replaced by the Quarterly Employment Statistics (QES) in March 2006. The QES is administered to selected industries and provides information on the number of employees and gross salaries paid. The employment estimates received from this firm-​level survey will differ markedly from the QLFS data, as the latter includes employees working in the agricultural sector, the self-​employed, unpaid family workers and domestic workers. Unlike the

Challenges and Complexities of the South African Economy    19 QLFS, which sets a minimum age for inclusion of fifteen years, the firm survey does not. Further differences also arise in the definition of the formal sector, where the QES includes only those VAT-​registered firms with the stipulated minimum turnover (Statistics South Africa 2020). The Survey of Employers and the Self-​employed (SESE) is conducted by StatsSA. Thus far, SESE surveys have been undertaken in the following years: 2001, 2005, 2009, 2013, and 2017. The survey is undertaken to provide information on the size of the value-added within the informal sector, and the information so received is complementary to the QES. Changes in methodology over the years limit full comparability (Statistics South Africa 2017). A non-​statutory national firm-​level dataset is that of the World Bank Enterprise Survey (WBES). One advantage of this survey is its comparability with other countries; these surveys covered 164,000 firms in 144 countries at the time of writing (see https://​www. enterprisesurveys.org/​en/​enterprisesurveys). However, the last survey for South Africa was undertaken in 2007, further back than for most other countries, limiting its current usefulness. A number of other firm-​level surveys have been undertaken by universities, research institutions such as the Human Sciences Research Council (HSRC), and other bodies. These tend to be limited to a specific geographic area or sector, typically focus on a particular theme (e.g. innovation), and are generally not publicly available. The sparsity of representative, comprehensive, and current firm-​ level data has constrained research in this area, which is an important body of economics research in many other countries. It has also hampered the inclusion of South Africa in cross-​ country studies employing firm-​level data, such as those using the WBES across countries. This weakness has arguably also affected the development of evidence-​based policy on firm behaviour. The growing body of research utilizing the SARS-​NT data illustrates the rich possibilities of novel, policy-​relevant research with firm-​level data. Meso-​data at the sectoral and sub ​sectoral levels is also weak in South Africa. For instance, sectoral data are also provided through subscription to private data service providers, notably Quantec (see https://​www.quantec.co.za). Many researchers find the Quantec data useful, since they are standardized and balanced across industries and over time, and provide a wide range of valuable measures. However, while drawing on official sources, these data are not official and utilize imputation and other methods that are not always transparent to users. Macroeconomic data are relatively straightforward, with reliance on secondary data from official sources and international institutions. The standard national sources of macroeconomic data are the South African Reserve Bank (SARB) and StatsSA. Researchers also utilize data from the World Bank, notably the World Development Indicators (WDI); the International Monetary Fund (IMF); and sources such as Bloomberg.

20    Fiona Tregenna, Arabo K. Ewinyu, Arkebe Oqubay, and Imraan Valodia

1.6  Appendix 1 Table 1.5: An overview of micro-​datasets in South Africa (household and labour force) Periods available

Dataset

Details of the dataset

Project for Statistics on Living Standards and Development (PSLSD)

The PSLSD was a World Bank-​sponsored Living 1993/​94 Standards Measurement Survey covering approximately 9,000 households, drawn from a representative sample of South African households. The PSLSD covered demographic, economic, education, and health data of enumerated households.

Southern Africa Labour and Development Research Unit (SALDRU) at the University of Cape Town

KwaZulu-​ Natal Income Dynamics Study (KIDS)

Drawing on the nationally representative 1993 PSLSD, households in Kwazulu-​Natal province were re-​surveyed from March to June 1998 for the Kwazulu-​Natal Income Dynamics Study. Combining these two survey datasets yielded a panel dataset. The dataset continues information on household demographics, household environment, education, income, expenditures and remittances, employment and other labour characteristics, agricultural activities, health, and anthropometry.

School of Built Environment and Development Studies, University of KwaZulu-​ Natal (UKZN)

October Household Survey (OHS)

This was the first annual sample survey that collected 1994–​99 household and labour market information at the national level. The scope of the OHS includes: employment, unemployment, informal sector, internal migration, services available by type of dwelling, access to health and social services, safety and well-​being of household, households by average household size and type of dwelling, level of education, quality of life, health statistics, vital statistics.

1993, 1998, 2004

South Africa SAPRIN combines three longitudinal datasets various start Population from the following Health and Demographic dates until Research Surveillance Systems (HDSS): 2017 Infrastructure • MRC/​Wits University Agincourt HDSS in Network Bushbuckridge District, Mpumalanga, established (SAPRIN) in 1993; • the University of Limpopo DIMAMO HDSS in the Capricorn District of Limpopo, established in 1996; • the Africa Health Research Institute HDSS in uMkhanyakude District, KwaZulu-​Natal, established in 2000. A significant contribution of these data is to provide regular and updated longitudinal data on the status of South Africa’s poorer and rural communities. Individual-​level data on the following are collected: health care utilization, marital status, labour market and education status as well as a record of household assets.

Producing agency

Statistics South Africa

Department of Science and Innovation (DSI) and the South African Medical Research Council (SAMRC)

Challenges and Complexities of the South African Economy    21 Table 1.5: Continued Periods available

Dataset

Details of the dataset

Income and Expenditure Survey (IES)

The IES is a survey administered to a nationally representative sample of households in order to update the basket of goods and services required for the compilation of the Consumer Price Index. While the main variable is expenditure, the IES also provides additional insights on household income and other individual and household characteristics.

1995, 2000, 2005/​06, 2010/​11

Statistics South Africa

National Census

The first census in post-​apartheid South Africa was 1996, 2001, held in 1996. The data that existed prior to this were not 2011 nationally representative. The main objective of the census is to collect sufficient information on living conditions and access to basic services which then helps government and other departments to allocate resources.

Statistics South Africa

Time Use Data Time Use surveys seek to provide information on how 2000, 2010 different South Africans spend their time to provide nuanced information on paid and unpaid labour, a gendered breakdown of work, subsistence work, casual work, and work in the informal sector. The Survey collects household and demographic data on two people, ten years and older, selected as respondents within each household. The questionnaire also include a diary in which respondents record the different activities they perform in the day.

Statistics South Africa

Labour Force Survey (LFS)

The LFS was a biannual rotating panel household survey designed to measure the dynamics of employment and unemployment in South Africa. It measures a variety of issues related to the labour market, including unemployment rates.

2000–​07

Statistics South Africa

General Household Survey

This survey replaced the OHS. It is an annual 2002–​18 household survey which measures the living conditions of South African households to provide information on development trends. The GHS collects data on the following key service delivery related themes: education, health and social development, housing, access to services and facilities, food security, and agriculture

Statistics South Africa

Cape Area Panel Study (CAPS)

CAPS is a longitudinal study of the lives of youths in metropolitan Cape Town, South Africa. The first wave of the study collected interviews from 4,800 randomly selected young people aged between fourteen and twenty-​two in the period August–​ December 2002. The study collects data across the following outcomes: schooling, employment, health, family formation, and intergenerational support systems.

Population Studies Center in the Institute for Social Research at the University of Michigan, the Centre for Social Science Research, SALDRU, and the Research Program in Development Studies at Princeton University. (continued)

2002–​09

Producing agency

22    Fiona Tregenna, Arabo K. Ewinyu, Arkebe Oqubay, and Imraan Valodia Table 1.5: Continued Periods available

Producing agency

2007, 2016

Statistics South Africa

Dataset

Details of the dataset

Community Survey

The Community Survey is held between each Census to obtain data at the national, provincial, and municipal levels to measure progress and outcomes on the following indicators: education, health, sanitation, water supply, housing, and transport as well as other demographic indicators. Information collected from these surveys informs Integrated Development Plans and infrastructure investment budgeting.

Living Conditions Surveys (LCS)

The LCS is a national household survey of over 32,000 2008/​09, individuals that provides detailed information on 2014/​15 household’s living circumstances, as well as their income and expenditure patterns. Data collected are also used to update the consumer price index (CPI) basket of goods and services.

Statistics South Africa

Tax data

These are anonymized panel data at the firm or 2008 -​2016 individual level. These data are created by merging various administrative tax data, namely: company income tax, employee data from employee income tax certificates, value-​added tax, and customs records.

United Nations University World Institute for Development Economics Research (UNU-​WIDER), National Treasury, and the South Africa Revenue Services

Quarterly The QLFS is a household survey that collects data on 2008–​ Labour Force a quarterly basis on the labour market activities of present Survey (QLFS) individuals aged fifteen or older. Earnings data for the QLFS series are released once a year as the Labour Market Dynamics data.

Statistics South Africa

National Income Dynamics Study (NIDS)

Southern Africa Labour and Development Research Unit (SALDRU) based at the University of Cape Town’s School of Economics

NIDS data was initiated in 2008 to enable the South 2008–​17 African Presidency to intensively track dynamic changes in the well-​being of South Africans. The first wave of the data tracked approximately 28,000 individuals across 7,305 households. The movements of household members as they enter or exit the initial household or establish their own households will be captured in subsequent waves of the panel study. At the time, it was the first national panel study to document a sample of households in South Africa and report on changes in income, expenditure, assets, access to services, education, health and other measures of well-​being. Data are collected every two years and so far, five waves have been collected.

Challenges and Complexities of the South African Economy    23 Table 1.5: Continued Periods available

Dataset

Details of the dataset

Producing agency

Gauteng-​ City Region Observatory Quality of Life Survey

These biennial data measure the quality of life, socio-​ 2009–​17/​18 Gauteng-​City Region economic circumstances, attitudes to service delivery, Observatory psycho-​social attitudes, value-​base, and other characteristics

Post-​ Apartheid Labour Market Series (PALMS)

PALMS is a stacked cross-​sectional dataset created 1993–​2019 and updated by the DataFirst team that collates different labour market data from various sources and releases it in a comparable and reliable format. The data consist of microdata from sixty-​nine household surveys conducted by Statistics South Africa between 1994 and 2019, as well as the 1993 PSLSD. The Statistics South Africa surveys include the annual OHS, the bi-​annual LFS, including the smaller LFS pilot survey from February 2000, and the QLFS. While the data is at individual level, household-​level variables may be created using the unique household identity variable provided.

DataFirst, University of Cape Town

National Income Dynamics Study–​ Coronavirus Rapid Mobile Survey (NIDS-​CRAM)

NIDS-​CRAM is a rapid assessment survey that 2020–​ investigates the socio-​economic effects of the national lockdown instituted by the South African government in March 2020 in response to the Coronavirus pandemic. The sampling frame for NIDS-​CRAM is Wave 5 of NIDS which was collected in 2017. Continuing sample members and temporary sample members older than eighteen as at April 2020, when Wave 1 of the NIDS-​ CRAM fieldwork was undertaken, were re-​interviewed at the time of the NIDS-​CRAM Wave 1 fieldwork in April 2020. Respondents were interviewed using Computer Assisted Telephone Interviewing (CATI), with data collection repeated over several months. NIDS-​CRAM is a component of a broader study called the Coronavirus Rapid Mobile Survey (CRAM) which aims to inform policy using rapid reliable research on income, employment, and welfare in South Africa, in the context of the global Coronavirus pandemic.

University of Stellenbosch, University of Cape Town and University of the Witwatersrand

24    Fiona Tregenna, Arabo K. Ewinyu, Arkebe Oqubay, and Imraan Valodia

References Andreoni, Antonio and Fiona Tregenna. 2020. ‘Escaping the middle-​income technology trap: A comparative analysis of industrial policies in China, Brazil and South Africa’, Structural Change and Economic Dynamics, 54: 324–​40. Andreoni, Antonio and Fiona Tregenna. 2021. ‘The middle-​income trap and premature deindustrialization in South Africa’, in Antonio Andreoni, Pamela Mondliwa, Simon Roberts, and Fiona Tregenna (eds) Structural Transformation in South Africa: The Challenges of Inclusive Industrial Development in a Middle-​ income Country. Oxford: Oxford University Press. Andreoni, Antonio, Pamela Mondliwa, Simon Roberts, and Fiona Tregenna. 2021. ‘Framing structural transformation in South Africa and beyond’, in Antonio Andreoni, Pamela Mondliwa, Simon Roberts, and Fiona Tregenna (eds) Structural Transformation in South Africa: The Challenges of Inclusive Industrial Development in a Middle-​income Country. Oxford: Oxford University Press. Arndt, Channing, Rob Davies, Elizabeth Gavin, and Landon McMillan. 2018. ‘Introduction: Firm level analysis using administrative record data’, South African Journal of Economics, 86(S1):  3–​5. Bhorat, Haroon and Carlene van der Westhuizen. 2012. ‘Poverty, inequality and the nature of economic growth in South Africa’. DPRU Working Paper, Cape Town. Bhorat, Haroon, Aalia Cassim, and Derek Yu. 2014. ‘Temporary employment services in South Africa: Assessing the industry’s economic contribution’ , [online] http://​www.capes.org.za/​ Research/​TES-​Economic-​Impact-​on-​SA/​. Blankenburg, Stephanie, Gabriel Palma, and Fiona Tregenna. 2008. ‘Structuralism’, in Lawrence Blume and Steven Durlauf (eds) The New Palgrave: A Dictionary of Economics, 2nd ed. Basingstoke: Palgrave Macmillan. Casale, Daniela, and Dorrit Posel. 2002. ‘The continued feminisation of the labour force in South Africa: An analysis of recent data and trends’, South African Journal of Economics, 70(1): 156–​84. Casale, Daniela, Colette Muller, and Dorrit Posel. 2005. “‘Two million net new jobs”: A reconsideration of the rise in employment in South Africa, 1995–​2003’. University of Cape Town, Cape Town. Cramer, Christopher, John Sender, and Arkebe Oqubay. 2020. African Economic Development: Evidence, Theory, Policy. Oxford: Oxford University Press. Fedderke, Johannes. 2014. ‘South Africa’s growth performance’, in Haroon Bhorat, Alan Hirsch, Ravi Kanbur, and Mthuli Ncube (eds) The Oxford Companion to the Economics of South Africa. Oxford: Oxford University Press. Francis, David, Imraan Valodia, and Edward Webster. 2020. Inequality Studies in the Global South. London: Routledge. Gill, Indermit S. and Homi Kharas. 2015. The Middle-​income Trap Turns Ten. Washington, DC: World Bank. Leibbrandt, Murray, Arden Finn, and Ingrid Woolard. 2012. ‘Describing and decomposing post-​ apartheid income inequality in South Africa’, Development Southern Africa, 29(1):  19–​34. Leibbrandt, Murray, Ingrid Woolard, Arden Finn, and Jonathan Argent. 2009. ‘Trends in South African income distribution and poverty since the fall of apartheid’. OECD Social, Employment and Migration Working Papers no. 101, Paris.

Challenges and Complexities of the South African Economy    25 Mosomi, Jacqueline. 2019. ‘Distributional changes in the gender wage gap in the post-​apartheid South African labour market’. UNU-​WIDER Working Paper no. 2019/​17, Helsinki. Oosthuizen, Morné, and Haroon Bhorat. 2005. ‘The post-​apartheid South African labour market’. University of Cape Town. Oqubay, Arkebe. 2015. Made in Africa: Industrial Policy in Ethiopia. Oxford: Oxford University Press. Pieterse, Duncan, Elizabeth Gavin, and Carl Friedrich Kreuser. 2018. ‘Introduction to the South African Revenue Service and National Treasury firm-​level panel: Firm level analysis using administrative record data,’ South African Journal of Economics, 86(S1):  6–​39. Posel, Dorrit. 1999. ‘Intra-​family transfers and the household division of labor: A case study of migration and remittance behavior in South Africa.’ https://​scholarworks.umass.edu/​ dissertations/​AAI9950203/.​ Posel, Dorrit, and Daniela Casale. 2013. ‘The relationship between sex ratios and marriage rates in South Africa’, Applied Economics, 45(5): 663–​76. Project for Statistics on Living Standards. 1994. South Africans Rich and Poor: Baseline Household Statistics. SALDRU, School of Economics, University of Cape Town. Ravallion, Martin. 2001. ‘Growth, inequality and poverty: Looking beyond averages’, World Development, 29(11): 1803–​15. Statistics South Africa., 2000. Labour Force Survey 2000 September: Metadata. Pretoria: Statistics South Africa. Statistics South Africa. 2017. Survey of Employers and the Self-​employed. Pretoria: Statistics South Africa. Statistics South Africa. 2019. Guide to the Quarterly Labour Force Survey. Pretoria: Statistics South Africa. Statistics South Africa. 2020. Quarterly Employment Statistics. Pretoria: Statistics South Africa. Tregenna, Fiona. 2010. ‘How significant is the intersectoral outsourcing of employment in South Africa?, Industrial and Corporate Change, 19(5): 1427–​57. Tregenna, Fiona. 2011. ‘Earnings inequality and unemployment in South Africa’, International Review of Applied Economics, 25(5): 585–​98. Tregenna, Fiona. 2012. ‘What are the distributional implications of halving poverty in South Africa when growth alone is not enough?,’ Applied Economics, 44(20): 2577–​96. Tregenna, Fiona and Mfanafuthi Tsela. 2012. ‘Inequality in South Africa: A study of the distribution of income, expenditure, and earnings’, Development Southern Africa, 29(1):  35–​61. Wittenberg, Martin. 2017a. ‘Wages and wage inequality in South Africa 1994–​2011: Part 1—Wage measurement and trends’, South African Journal of Economics, 85(2): 279–​97. Wittenberg, Martin. 2017b. ‘Wages and wage inequality in South Africa 1994–​2011: Part 2–​Inequality measurement and trends’, South African Journal of Economics, 85(2): 298–​318. Yu, Derek. 2007. ‘The comparability of the statistics South Africa October household surveys and labour force surveys’. Stellenbosch University Economic Working Paper no. 17/​2007, Stellenbosch. Yu, Derek, Rulof Burger, and Neil Rankin. 2017. ‘South African employment trends: How reliable are the different data sources?’. REDI Working Paper no. 31, Pretoria.

Chapter 2

The Ec onomi c History of S ou t h Africa bef ore   194 8 Stefan Schirmer

2.1 Introduction This chapter outlines the evolution of the South African economy and analyses the institutions that shaped the economy’s structure and performance. It covers a vast, complex, and changing time span, from when settled agricultural communities first emerged in South Africa, around 400, until the gradual establishment of a significant manufacturing sector in the 1920s and 1930s. Rather than trying to provide a detailed account of all the economic events and changes that took place, the chapter focuses on the interactions that shaped crucial economic transformations. We start by examining sources of stability and dynamism during the pre-​colonial era. This section concludes with an examination of the rise and fall of Bokoni, a remarkable and innovative intensive farming system, the disappearance of which is associated with the rise of a number of predatory state systems in the northern and eastern parts of South Africa, at a time when colonialism was starting to encroach. The focus then shifts to the emergence of a new, racially defined agrarian order in the colonial Cape, followed by a description of the ways in which the mineral discoveries changed the South African economy and helped to usher in the emergences of a modern state, along with a highly oppressive and discriminatory migrant labour system. We look at how both Black and white farmers responded to these new circumstances and conclude with a description of policy shifts influencing the rise of manufacturing in the 1920s, 1930s, and 1940s.

The Economic History of South Africa before 1948    27

2.2  The Emergence of Pre-​colonial Farming Communities Settled agricultural communities emerged initially in the third and fourth centuries along the northern east coast of South Africa. They used iron tools to plant crops such as millet and sorghum. Such farming practices spread inland, during the fifth century, into valleys where soil was fertile, rainfall regular, and rivers plentiful. Iron working was extensive and crops were supplemented with the herding of sheep and goats. Cattle were only acquired rarely. During this period areas of economic specialization became established around rich salt and iron deposits. Salt manufacturing consisted of mixing naturally occurring salt crusts with sand, then filtering the mixture through grass to produce a clean brine, which was boiled down to form hardened cakes that could easily be transported. Salt producers exchanged their products for cattle and other foodstuffs difficult to grow in the relatively barren northern areas where the salt occurred. Similarly, in the area around modern Phalaborwa, communities came to specialize in mining, manufacturing, and trading iron and copper. From the eighth century onwards, deep shafts were cut into the volcanic hills, and copper ore was removed by cracking the rock with fire and working it with chisels and hammer stones. Iron ore came from the magnetic pebbles readily available on the slopes of the hills. Several hundred sites have been identified from which the ore was extracted before being forged into wire, beads, bracelets, arrowheads, spear points, woodworking tools, and agricultural hoes (Hall 1987).

2.3  The Settlement of the Interior and the Dynamics of Economic Change, 1000s–​1700s Pastoral communities, as well as groups who maintained a hunting and gathering lifestyle, were pervasive over most of the South African interior. Rather than forming distinctive, disconnected groups that can easily be labelled in terms of particular ethnic or genetic characteristics, these communities interacted with one another extensively. While there was probably some intermittent conflict between groups pursuing different livelihood strategies, the dominant modes of interaction were exchange, absorption, and movement from one mode to another. It was, in fact, by trading with pastoralists, that precarious farming communities originally confined to coastal regions and fertile

28   Stefan Schirmer valleys gradually acquired cattle herds. This then provided these communities with the additional security and nutrients that made settlement on poorer soils in areas with less predictable rainfall possible. Consequently, agricultural settlements across the large inland highveld plateau emerged in the early years of the second millennium. Over the course of the next two hundred years other communities moved into South Africa from the north and interacted with and absorbed, or were absorbed by, communities already living there. These processes then produced cultures that are broadly identifiable as Sotho and Nguni (Hall 2009). In the days when settled agriculturalists had few cattle, and they depended on hunting and gathering for additional nutrition, they developed cultural practices to ensure that when drought or disease struck in one area, the people in that area could count on the assistance and hospitality of people in other, unaffected areas. The adoption of cattle reduced the need for these practices as it increased population densities and made communities less vulnerable. Consequently, reciprocal interactions within geographically defined communities became more important than reciprocal cooperation between communities. However, strong traditions of cooperation across settlements remained in place into the nineteenth century (Keeton and Schirmer, forthcoming). Within these communities, control over cattle provided control over people. Economic superiority allowed people to become powerful as they drew large numbers of dependents towards them. In this way, political structures that revolved around the institution of chieftainship became established. However, ever-​present possibilities for people outside formal positions of power to accumulate cattle ensured that there were frequent changes in who ruled. Furthermore, movements in and out of established villages or chieftainships were relatively easy and a regular feature of pre-​colonial communities. Thus, while political centralization could emerge to a substantial extent, as it did in the case of Mapungubwe in the twelfth and thirteenth centuries, political structures were also highly unstable and frequently broke down (Antonites 2012). As Hall (1987) puts it, over time chiefdoms were constantly fragmenting and reforming as factions gained power, built up strength, and then lost control to other groups. Within these communities, individual families pursued wealth and power, but they were expected always to ensure that their own success benefited everyone in the society. Cattle loans and other forms of support were not just a way to build up political influence. They were also a moral duty and there were strong mechanisms, such as witchcraft accusations, that could be used to punish those who did not fulfil their duty. The farmers on the Highveld who were able to produce a surplus of food, but not very much iron and salt, were always prepared to trade cattle, which had the advantage of being easy to transport, for technologies like iron hoes for ploughing and salt for preservation. These technologies had the potential to improve the standard of living of households. By choosing to acquire them, households were placing the risk of economic gain over the certainty of social status. Thus, while the social role that cattle played within communities did create disincentives to trade cattle, such norms never acted to eliminate markets and extensive trading (Wilson 1969).

The Economic History of South Africa before 1948    29 African societies were never egalitarian. Witchcraft accusations may have acted as a disincentive to accumulate wealth by individual households, but wealthy households were in fact not, in themselves, the targets of witchcraft accusations. ‘It is not just wealth that caused resentment,’ Sansom (1974) maintains, ‘Accusations [were] incited by hoarding and the selfish use of wealth, by the failure to share one’s fortune with the less fortunate.’ The fact that early travellers reported witchcraft accusations as a common practice demonstrates that what constituted hoarding and selfish behaviour was the subject of regular debates. In the nineteenth century, African farmers who harvested a bumper crop were observed storing their surplus grain in secret hoards they would visit at night. In a hostile and unpredictable environment it was completely rational for everyone to commit to structures providing social security and risk minimization. However, some people—​mostly the poor—​would have had a greater commitment to security than others—​mostly the wealthy. Second, it was impossible to choose between accumulation and political standing within established social systems. Entrepreneurs depended on power-​enhancing structures. Thus, when power got in the way of profit, those that were negatively affected could only oppose the particular instances of power. They could not afford to oppose the broader social structure and accepted cultural practices. In these systems, forms of control within the household centring on the household head’s ability to control cattle created an important dynamic, but cattle accumulation also took place for other reasons (Guy 1987). Heads of households were concerned with the security that cattle provided, with the social status that cattle signified within the broader community and the access that cattle provided to the labour of dependants outside the household. In addition, while contradictions in the household probably created tensions, there is no evidence that these tensions brought about fundamental change during the pre-​colonial era. Conflicts between chiefs and commoners and between household heads and family workers definitely existed (Peires 1981). At the same time, cooperation between these groups as a way to ensure security was equally important, and became more so when survival became more precarious. Structures of cooperation, individual ambition, and the maintenance of political power were factors whose contradictory influences ensured that there were always many directions in which African societies could develop. In addition, shifts in the way these factors interacted linked to changes in nature, to exchanges between different social groups, and to broader economic transformations.

2.4  The Intensive Farming Terraces of Bokoni, 1500s–​1820 The significance of the area known as Bokoni has only relatively recently been raised to the level of importance it deserves. Knowledge about the terraced farming that took

30   Stefan Schirmer place on the escarpment in today’s Mpumalanga province has existed for many years, but, through collaborative work between historians and archaeologists, it is now evident that the area consisted of one of the largest intensive farming regions in sub-​Saharan Africa, and that extensive productive and institutional innovation made this form of agriculture possible. Agriculture in this region took place along stone-​built, hillside terraces spanning an area of 150 kilometres north to south, and 50 kilometres or more east to west between the modern towns of Ohrigstad in the north and Carolina in the south. Settlements consisted of stone-​built homesteads linked by walled cattle paths. These settlements emerged in the 1500s and prospered until they all but disappeared in the 1830s. Cattle manure, collected from central cattle enclosures, enhanced the fertility of the terraces and relatively high rainfall and proximity to seasonal streams permitted a kind of irrigated farming consisting of a controlled flooding of fields during periods of high rainfall. Extensive trading was a core feature of the area, as it produced and exported large surpluses of crops such as millet and sorghum and eventually maize, while importing in large quantities the iron implements needed to work the fields. By producing a tradable surplus, Bokoni was able to create an improved level of food security in areas to the north and east of it, which acted as a disincentive for neighbouring groups to raid it for crucial resources. Trading rather than raiding marked the way the area was integrated into the wider region (Maggs 2010). Bokoni had loose aggregations of homesteads, with little evidence of hierarchy or the centralization of political power. The long stretches of terracing in Bokoni were probably the work of young men—​perhaps mobilized through age-​set organization. Thus male, and perhaps also female, initiation groups emerged in this politically decentralized social system as an effective way to mobilize labour for major projects and to socialize youth into the long-​term commitment and range of skills required to sustain this system (Delius and Schirmer 2014). As the eighteenth century drew to a close, the Pedi Kingdom established a form of hegemony over Bokoni, but this had little impact on the production of food on the terraces. Dramatic changes that destroyed the Bokoni settlements were the result of predatory and expansionary states emerging with a propensity to employ violent means to obtain not just grain and cattle, but also women and children. These new states became known as the Ndwandwe, Swazi, Ndebele, and Zulu kingdoms. With very few defensive capabilities, the Bokoni settlements were not equipped to survive in this kind of violence-​dominated world. As a result the whole Bokoni system collapsed, and it, along with knowledge of its major economic achievements, all but disappeared for more than 150 years. Bokoni demonstrates beyond any doubt that African societies were extremely dynamic. True, the need for security was a powerful force. It ensured that redistributive social structures gave the vulnerable some control over the resources accumulated by individual initiatives. It also made political power an attractive option, both for the wealthy who turned resources into control over people and for the weak, who preferred to be dominated rather than vulnerable. This emphasis on power and control in

The Economic History of South Africa before 1948    31 pre-​colonial systems limited the levels and the impact of market-​responsive development, but aspects of that were always there and, under the right circumstances, commercialization would be unleashed amongst those who commanded sufficient resources, security, and confidence.

2.5  Colonialism and the Establishment of a Racial Agrarian Order in the Cape The white settlers who came to farm and trade in and around the Cape peninsula were initially just as vulnerable, if not more so, than many African communities, and they depended heavily on the Dutch East India Company. The Company fulfilled the dual role of commercial enterprise and government. In the 1650s, soon after the establishment of the station at the Cape, the directors decided to promote the settlement of ‘freemen’. However, despite their title, these settlers were subjected to tight control. The company regulated all trade in the Cape, keeping prices down and forcing settlers to sell only to the company. Control was never as complete as the company desired, but settlers generally relied on the company for support and protection in a hostile environment (Guelke 1989). The company provided cheap loans and imported slaves that farmers could obtain on credit. In districts further inland, Khoisan dispossessed of their land and livelihoods by colonial attacks and restrictions provided labour. Under these circumstances, settlers rarely complained about the company’s control. In the 1700s, as the economic capacity of some farmers expanded, however, growing discontent emerged. In 1738, resentment became evident in the ‘Barbier Rebellion’ and in 1778 the Cape Patriot Movement brought matters to a head. The members of this movement demanded to be freed from ‘economic subservience to company interests’ and to be allowed to trade freely with ships in the harbour (Schutte 1989). These settlers were attempting to create a more open society, where markets would become more dominant in allocating resources and incentives. In 1795, company rule ended and a wealthy farmer and merchant elite began to emerge, first in Cape Town, then in the wine-​growing regions of Stellenbosch and Drakenstein, and in the wheat-​ growing districts of Tijgerberg and the southern Swartland (Ross 1986). Subsequent colonial states forged alliances with these elites rather than seeking to control them directly. The Cape was a British colony in the nineteenth century and thus became an integral part of the largest, most economically powerful empire in the world. The dual effect was the establishment of many new economic opportunities for farmers, traders, and service providers, as well as a significant enhancement of the state’s capacity. British colonial policy was to keep down administrative costs as far as possible, but expanded financial and military resources allowed the new state to penetrate into areas, especially African occupied areas, in ways that had not been possible before. Initially, from the 1820s,

32   Stefan Schirmer established merchants and farmers, some of whom were engaged in both activities simultaneously, were determined to open up opportunities for capital accumulation and to expose as many producers to the market as possible. During this time, a ‘mercantile— humanitarian’ alliance emerged around the project of liberalizing the Cape economy and creating a greater equality of opportunity (Keegan 1996). The vision promoted by this alliance was one in which Black and white farmers would be drawn into the colonial economy—​thereby expanding trade and production opportunities for those with the capacity to take on the risks of market production without reservations. Members of the emerging Cape elite, with significant resources and diverse business interests, were not strongly dependent on the state and could only see benefits from an extension of the market. The abolition of slavery in 1834 alienated some elite wine and wheat producers, but also injected major supplies of investment capital, in the form of compensation payments, into the economy. This capital stimulated the development of urban commerce in Cape Town and investment in merino sheep farming. Sheep farming took off especially rapidly in the Eastern Cape, where recently arrived British settlers predominated. They constituted a new element, with generally little economic capacity and a stronger tendency to look towards the state to promote economic opportunities through military intervention. Subsequently, military violence was used to free up African land and labour for white exploitation. The new settlers benefited from the increased spending that accompanied any mobilization of the army and forged a closer alliance with British colonialism. Opposed to this project was a group within the more established settler elite who put long-​ term economic interests before the short-​ term strategies of the newly arrived British settlers. At the same time, merchants on the frontier were able, for a while, to promote a more accommodating state policy as they continued to maintain a ‘small tradition’ of liberalism in the pursuit of trade with Black entrepreneurs. These competing visions ensured frequent policy shifts occurred as the state attempted to play a mediating role between different interests. Nevertheless, the colonial economy was founded during this time on an alliance between white entrepreneurs, both English-​ and Afrikaans-​speaking, and a state whose interests were closely tied to the promotion of white accumulation. Despite the lack of industrialization and the avoidance of the market by many producers, there was steady and impressive growth in the output of the colonial economy’s main agricultural products during the period 1700–​1830. The growth in output was brought about primarily in response to growing demand within the colony, regional integration, and the emergence of Indonesian and Australian markets where South African producers enjoyed a cost advantage over European producers. In addition, the outbreak of war or harvest failures in Europe represented temporary sources of demand that could lead to high profits for South African producers but also made the market unstable and risky. The enterprises of the richest families, like the Van Reenens, became integrated and diversified, providing internal access to investment capital and facilitating risky

The Economic History of South Africa before 1948    33 ventures. Capital was increasingly made available to a broader band of entrepreneurs through the development of local credit markets. A state-​owned bank was established in 1790 and from the 1830s numerous regional, privately owned banks emerged. By 1860, twenty-​three Cape banks had a total circulating capital of £374,584 (Ross 1989). By the 1840s white commercial farmers were usually members of the settler elite based in productive wheat, wine, and merino districts, while defensive, risk-​avoiding farmers were to be found mostly amongst the frontier farmers known as ‘trekboers’. These travelling agriculturalists were mostly concerned to maintain a way of life that was regarded as superior to a landless existence. As one trekboer put it in 1834: ‘[on the frontier] every child is a farmer and gets his inheritance in stock, and in what country will people serve for hire if they can live as their own masters’ (Van der Merwe 1938). White settlers frequently used access to land and control over labour to restrict their dependence on market forces. Slavery and the less clearly defined relations with Khoisan operated in a ‘coercive framework’ within which ‘violence and the threat of violence were the primary means of control (Mason 1989). Employers did devise ways to co-​opt their workers by providing access to land and cattle and allowing them to maintain spheres of independent activity. There were ex-​slaves who accompanied their employers on the ‘great trek’ into the interior, accepting the risks and the hardships despite colonial officials who advised them to stay in the Cape as ‘freemen’. They made this choice because they wanted to maintain access to cattle and land and because their prospects in the Cape were extremely restricted. The labour system that was therefore sustained despite the abolition of slavery and the resistance of ex-​slaves to new forms of exploitation depended on coercion, some cooperation, and severe restrictions on alternatives available to workers. This system cut across economic differences between white farmers. It provided commercial farmers with cheap, controlled labour, thereby enhancing their profit margins, while defensive farmers were provided with dependent workers who helped to facilitate a particular kind of lifestyle in spite of market forces. Of course, there was a potential contradiction here as state regulation and defensive farmers tied up workers that could be more effectively employed by commercial entrepreneurs. However, as long as farmers believed that the state could provide them with a large and subservient workforce this contradiction remained mainly below the surface.

2.6  The Mineral Discoveries and the Making of a Modern, Racial State After a few false starts, the South African diamond rush began in earnest in 1871, quickly leading to the establishment of a new settlement of approximately fifty thousand men at Kimberley in West Griqualand. Once this pattern was repeated when gold-​bearing seams were uncovered in Johannesburg in the 1880s, a rapid transformation occurred.

34   Stefan Schirmer The city grew very quickly, new kinds of manufacturing, in particular explosives and large-​scale brewing, became established, an increasingly sophisticated banking sector emerged, and railways were built linking mineral centres to the coast. It is undeniable that the diamond and gold industries were new engines of export-​ led growth. By 1872, the diamond fields were already producing 1 million carats. The level of output had doubled by 1879, and reached 3.5 million carats by 1888. After that time, De Beers managed to gain control of most diamond production in South Africa, and restricted output so as to prevent a price collapse. Output remained at between 2.5 million and 3 million carats until the Great Depression when the market did finally collapse, albeit temporarily. By the 1890s, exports of gold stood at £4.5 million per year. In the period 1906–​10, the annual average soared to over £27 million and in the 1930s reached £80 million. At the end of the 1890s gold accounted for over 50 per cent of South Africa’s exports and by the end of the 1930s the percentage had risen to 70 (Feinstein 2005). This kind of revenue generation, unrestricted by the small local market, created massive new opportunities, around which urbanization, especially in Johannesburg, took off. As Van Onselen (2001: 2) has described it in his seminal study of everyday life in Johannesburg: By 1896 the 3,000 diggers of the original mining camp were lost in a town of 100,000 residents and, by 1914, the 100,000 were in turn becoming harder to find in a city of over a quarter of a million inhabitants. The inexorable pressure exerted by people, houses, shops, offices and factories pushed back the municipal boundaries from five square miles in 1898, to nine square miles in 1901, and then—more ambitiously still—to an enormous 82 square miles in 1903.

A crucial development was the establishment of large joint-​stock mining corporations on South African soil, along with a stock exchange that pulled in huge investments from locals and foreigners, even before the gold-​mining industry was fully established. During the period from 1886 to the end of 1895, when stock prices crashed as a result of over-​confident speculation, approximately £15 million was invested in the Rand. Many new institutional developments, especially the close relationship between gold mining and the financial sector, were driven by prominent business leaders who had, at least partially, put down roots in South Africa after making their fortunes in Kimberley. They included Cecil John Rhodes, Alfred Beit, Barney Barnato, the Ecksteins, George Farrar, and Julius Jeppe (Lukasiewicz 2017). Different mining companies had various levels of commitment to investing in, and becoming part of, the emerging, diversified economy taking shape around the gold diggings. Some concerned themselves, out of necessity, merely with survival, while others sought to make as much as they could in as brief a time as possible, and then got out (Mawby 1974). By contrast, the strongest, most established companies—​in particular Rand Mines—​had a long-​term vision. Thus, during a time of depression and widespread mining-​company failures in the late 1890s, when many smaller companies

The Economic History of South Africa before 1948    35 considered exit as the best strategy, Rand Mines’ Julius Wernher looked forward to a time when, in contrast to the boom and bust atmosphere then prevalent, a different kind of economy would emerge. This economy, he speculated, would be made up of strong, viable companies whose valuation on the stock exchange would reflect a firm’s real value, rather than artificial hype. He admitted that his company had been ‘not quite guiltless’ in pushing up values artificially, but, while he saw such behaviour as tempting, it was not, in his opinion, ‘worthy of a great firm’ (Kubicek 1972). It is also significant that attempts to extract rents by prominent ‘wheeler-​dealers’ like Barney Barnato, quickly met with organized resistance from those who demanded free and equitable access to the Johannesburg Stock Exchange. Such organized opposition to attempts by business elites to impose their narrow, short-​term interests, seems to have been a prominent and early feature of Johannesburg’s social make-​up. Mawby’s (2000) two-​volume study of the politics of Johannesburg, covering the period 1902–​07 provides a very detailed sense of both the complexity of politics and the levels of political engagement amongst the mostly white, mostly male residents of the emerging city. The broad conclusion to draw from a careful reading of this detailed study is that neither the state nor mine owners could easily impose their narrow interests on this segment of the population. Mine owners, in particular, needed to compromise, to become drawn into debates about the future of the society that was being constructed on the Highveld, to engage with the issues that concerned both the middle and the labouring classes. In this way, capitalists became, albeit gradually and very unevenly, engaged citizens rather than just pursuers of profit with narrow time horizons. Mine owners, often in pursuit of specific economic advantages, did play a role in political movements like the United Conference Ticket (UCT), founded to contest the elections of 1907, and the Progressive Association, formed to influence the direction of development in the Transvaal once self-​government was established. The UCT was the product of a number of meetings of delegates from a very wide range of more narrowly focused organizations. Out of an exhaustive and complex process of engagement, the body elected: ‘four members of the Chamber and three small independent mining men; the militant president and three other leading members of the Witwatersrand Trades and Labour Council, and at least one other working man; a considerable number of commercial retailers; two Afrikaners; a few professional men, and others’ (Mawby 2000). The politics of Johannesburg pulled the mine owners who saw the need to engage in it into a complex process of persuasion and compromise, which inevitably led them to think of long-​term commitments and benefits that mining could produce for the broader, albeit white, society. Interactions between mine owners and the state were often marked by similar characteristics. A close examination of these reveals a complex interaction between various owners and, at first, the Transvaal state under the leadership of Kruger. The Kruger state clearly recognized the important potential of this massive new industry 50 kilometres to the south of the capital, Pretoria. According to Van Onselen, Kruger’s republic had developed plans to promote industrialization before gold was discovered and exerted much effort to support the development of gold mining and a nascent

36   Stefan Schirmer manufacturing industry as well. Some mine owners, especially the Eastern C ​ ape-​born Joseph Robinson, established close relationships with the Kruger regime. In general though, the perspective of the state on how to promote development conflicted with the views of most mine owners, and acted as a significant break on general development. In Van Onselen’s terms, the Kruger state envisioned the emergence of an industry that grew out of and was dependent on the output of the agricultural sector. This conflicted with the realities of a new economy emerging around, and dominated by, gold mining. However, the main reason why the Kruger government mostly failed to promote development was because it was weak and ineffective (Marks and Trapido 1979). This prevented it from engaging with the new business elite on an equal and dynamic footing. As Jeeves (1975) points out, many interactions between mine owners and the state prior to the War of 1899–​1902 were dominated by the miners because government officials lacked the expertise to avoid having to rely upon Chamber of Mines data and Chamber expertise when any mine-​related investigations were undertaken. In 1902, after three years of war, the capacity of the Transvaal state was enhanced significantly, under the leadership of the British High Commissioner Lord Milner. The beneficiaries of this were equally the mine owners as they were the white farmers who had been the core constituency of the Kruger regime. The colonial regime was able to provide services, support, and finance far more effectively than its predecessor (Krikler 1993). And, while the Milner State has often been described as essentially a mine owners’ instrument, actual relationships were far more complex. The Milner and subsequent Selbourne regimes were clearly sympathetic to the needs of mining development in ways the Kruger state had not been, but they were equally determined to harness mining-​generated revenues and resources to promote a broader vision of development. This elicited numerous complaints from mine owners about what was seen as a heavy 10 per cent tax on all mining profits, made more irksome by the Milner State’s propensity to collect taxes efficiently. It also produced numerous demands from the state on mine owners to use their resources in ways that would promote a broader developmental process. In an interesting exchange suggestive of the to-​and-​froing of state–​ business relationships at this time, Lionel Phillips of Wehrner Beit and Co responded to Lord Selbourne’s request to invest in housing as a way to encourage British immigration: ‘while I am fully in sympathy with your views, the moment is not propitious to consider any large and avoidable capital outlay’ (Van Onselen 2001). When it was not British immigration, it was white poverty and unemployment that Milner and Selbourne worried about extensively. In this way, their concerns and the nature of their relationship with mining interests was very similar to the one that emerged after the Afrikaner-dominated Het Volk/​South African Party (SAP) took over first in the Transvaal in 1907 and then across the Union of South Africa in 1910. The leaders of this new government, Louis Botha and Jan Smuts, were in no way hostile to the existence of the gold-​mining industry as such, but feared its potential to defy the government as they sought its help in tackling their major development priorities. These were white poverty and, especially, Afrikaner unemployment. Consequently, the SAP used resources extracted from the mines to establish agricultural settlement schemes for poor whites

The Economic History of South Africa before 1948    37 and encouraged the mines to offer decent jobs for relatively unskilled Afrikaner work-​ seekers on the Rand. Yudelman (1983) characterized the outcome of these processes as a ‘symbiotic relationship’ between the state and mining capital. These complexities, which helped to create a mining industry and a state that contributed positively to the broader economic development of South Africa, fall away completely when it comes to the efforts of the mines to create what they saw as a dependable, affordable Black labour force. As early as 1890, the Chamber of Mines was making extensive and unyielding demands on the Kruger government to vigorously enforce pass laws and coercive ‘master and servants laws’, and to prevent any competition for labour between the mining houses. Many of the coercive laws and practices to control African workers had been pioneered in an agrarian context, but the mine owners saw the advantage of them clearly and demanded their strengthening. The Kruger state’s efforts to comply were deemed unsatisfactory by the mine owners. After the war, the Milner administration provided a much more elaborate system of control. The pass department was reorganized and expanded, inspectors were appointed, and a ‘finger impression’ branch was established to facilitate positive identification of deserters (Jeeves 1975). At the same time, the state acted on behalf of mine owners in securing a monopoly for labour recruiting in Mozambique. Apart from occasional compromises that took into account the labour demands of white farmers, the state continued to act as an instrument of the mines’ labour demands right until the 1940s, and beyond. Within the context of colonialism and racism, African workers creatively struggled to influence their working conditions, their pay, and their terms of employment, but they were never in a position to disrupt the narrow economic alliance between mine owners and the state around the creation of a coerced, cheap labour supply. As Feinstein has pointed out, mine owners did bear costs in terms of this arrangement, primarily as a result of the enormous labour turnover and constant re​introduction of the workforce required as migrants moved in and out of work. The owners also put themselves in a weak position with regard to the wage demands made by highly unionized white workers, who were often successful in mobilizing state support for their cause. However, in terms of immediate profits, the mines benefited greatly. They were able to pull in a large workforce at a very low cost. The total number of African workers on the mines increased from 54,000 in 1896–​98, to a pre-​apartheid peak of 318,000, in 1936. Of those workers, 52,000 came from within South Africa. The rest were recruited mainly from Lesotho, Swaziland, Bechuanaland, Mozambique, and Malawi. Throughout this time, this migrant labour system was frighteningly effective at holding down wages. Very few nominal wage increases occurred after 1911, and there were periods when real wages fell dramatically. Real wages in 1951, were in fact at the very same levels that they had been in 1916. All of this ensured that the modern South African economy emerged as an incredibly distorted and unequal system. These realities were then exacerbated further as a result of the way white and Black farmers responded, and were allowed to respond, to the expanding markets created by the mineral revolution.

38   Stefan Schirmer

2.7  Agricultural Responses to Expanding Internal Markets Initially, the strongest response to the new agricultural demands created by the mineral revolution came from African farmers. There were, to be sure, many Africans in the 1870s, probably the majority, whose main aim was to defend traditional structures and security. Defensive farmers mostly produced limited amounts for the market, primarily as a way to pay rents to white landlords or taxes to the state. The choice that an increasing number faced was which colonial market they would enter: the labour or the produce market. Most chose the produce market because it gave them more independence and allowed them to avoid the horrific conditions in the mines. In the Eastern Cape the wars of 1877–​78 and of 1881 were a last ditch effort to push back the intrusive colonial system. Instead, agricultural resources were decimated and a large number of Xhosas were forced into low-​paid employment as migrant workers. These kinds of responses were partially eroded after 1870, as a significant minority of Xhosas began to see impending wars as a threat to their property. They were more likely to appeal to the colonial authorities for assistance rather than fight. The colonial records increasingly mention ‘a spirit of emulation’ as Africans became increasingly willing to adopt commercial farming strategies previously confined to mission stations. The records talk of a growing willingness to experiment, to diversify in order to strengthen economic capacity, and to form associations that would protect the interests of, and allow the sharing of ideas with, fellow entrepreneurs. Agricultural fairs were held in which an astounding variety of crops were shown and discussed. This was accompanied by ongoing technological advances and an ever-​intensifying demand for private land ownership (Bundy 1988). A group originally organized around the idea that they were refugees from the North, but who probably came to include a variety of entrepreneurial individuals from surrounding societies, the Mfengu, were hiring machines known as strippers to harvest their corn as well as machinery to thrash and clean their wheat. By the late 1870s, African farmers were becoming increasingly active in land markets. In the Transkei, a group reportedly purchased 38 thousand acres in 1879. African farmers seeking increasingly deeper integration into the market economy and ready to cooperate with the colonial government where that was possible were growing in numbers in the 1870s and 1880s. By succeeding as market-​oriented farmers they developed the capacity to adopt risk-​taking attitudes and activities. They were always on the look-​out for new opportunities. In the North West, in the lands bordering the diamond fields, there were farmers seeking to take advantage of the economic opportunities emerging out of Kimberley (Shillington 1985). The poor soils and low rainfall of this region limited opportunities in agriculture and ensured that most entrepreneurs started outside of farming. Some earned their first profits by selling diamonds, while that was still permitted. They then

The Economic History of South Africa before 1948    39 invested in agriculture, and those who were able to take advantage of good rainfall years made further profits as cattle and maize farmers. Many then diversified into transport riding, while a few got access to irrigated land, which permitted further investment into agriculture. Very few of these committed accumulators displayed much sympathy for traditional leaders or for fighting wars with the colonists. In Natal, Kholwa farmers achieved great success between 1860 and 1880. With missionary support, which was particularly useful in dealing with colonial governments, and with savings derived from wage employment, the Kholwa were always on the look-​ out for economic opportunities. From the early 1860s, sugar production appeared to offer the best opportunity for profit and expansion. For twenty years, African sugar growers competed successfully against better-​ resourced white farmers, growing cane and manufacturing sugar in their own steam-​powered mills (Etherington 1985). However, circumstances turned against them in the 1880s and they found it increasingly difficult to compete with whites who had greater government support, access to larger amounts of credit, and better information networks. The Kholwa remained entrepreneurial, though, and switched into transport riding and other forms of agriculture. Around the turn of the century, in the wake of the war of 1899–​1902 and movements towards creating a unified South Africa, significant policy shifts took place within the various colonies. In the Cape, especially, the idea that African farmers who sought integration should be accepted and/​or encouraged was rapidly eroding. Increasingly the idea took root that Africans should ‘develop within their own societies’. The original idea, at least in the Cape, had been to view commercializing farmers as a way to build up local support for colonialism and as a way to expand the supply of labour. Officials believed for a while that entrepreneurship in the countryside would create stratification, which would then concentrate land ownership in the hands of the most dynamic farmers, leading subsequently to the creation of a landless working class. After all, that is the way it had worked in Britain during the course of the Industrial Revolution. This was the thinking behind the 1894 Glen Grey Act. However, the emerging shifts in official ideology were reflected in specific provisions of the Act. Whereas the original idea had been to promote individual forms of tenure for Africans, the Act ended up allowing only four morgen per allotment, which was hopelessly inadequate for profitable farming at any scale. The accumulation of additional land was expressly forbidden (Ally 1985). Thus, although the Act paid lip service to the previous policy of assimilating and developing Black entrepreneurs, it then ensured in its provisions that very little landlessness or development occurred. This policy direction initially threatened the large-​scale creation of an urban workforce because it gave Blacks the ability to remain on the land, away from wage employment. A crucial cause of this shift in direction was a realization in the bureaucracy that a cautious assimilationist policy could not deal adequately with the growing and potential strength of landless Black labourers. The emergence of industrialization and unprecedented levels of urbanization in the 1890s clearly intensified administrative concerns. This, in turn, combined with the defensive struggles of African chiefdoms located in the reserves, who often felt threatened by the movement towards commercialization

40   Stefan Schirmer amongst many of their followers. The demands from below for a protection of traditional structures allowed ‘the state to deflect conflict from the towns and entrench some of the forms of pre-​colonial society in the reserves’ (Beinart 1980). Thus defensive reactions to colonial markets presented the state with an opportunity to create, in their eyes, a less risky and probably cheaper form of administrative control. In this way ‘Segregationism’, which also became official ideology in Natal and had always appealed to the administrations operating in the Transvaal and Orange Free State, became the basis on which the Union of South Africa dealt with the African majority. The culmination of this was the 1913 Land Act, which designated a tiny percentage of the land as African areas and set the rest aside for white occupation and development. It also aspired to create a clearer, more rational ‘separation of the races’ but most of those clauses, or ‘chapters’ would only be properly implemented much later, in the 1950s, when a new, more determined state set out to take segregation to the next level. Nevertheless, the first four decades of the twentieth century no longer provided aspirant Black commercial farmers with sufficient encouragement. Apart from islands of agrarian entrepreneurialism scattered all over the country, the majority of Black farmers opted for defensive farming strategies, or moved to urban areas, or sought out the limited opportunities for educated Africans as teachers, nurses, or administrators in the growing bureaucracies of the African ‘reserves’. In the meantime, white farmers were finding it difficult to make headway. The South African state implemented numerous initiatives and provided massive support in an attempt to promote the economic development of the white agricultural sector, but constantly ran into difficulties when confronted with farmers who avoided development more than they embraced it. The size of this challenge was increased by the response of farmers to the very difficult conditions of the 1920s. By then it was clear that state aid had made it easier for farmers with land titles to stay on the land as the number of whiteowned farms rose by 23 per cent between 1918 and 1928. Simultaneously, and in contradiction to the aim of keeping whites on the land, the development of agriculture had made land a highly valuable commodity. This meant that owners frequently forced white tenants off land that could be sold or put into production. It also became increasingly expensive to get a start in agriculture, and land mortgages became a popular mechanism for raising the necessary funds, thus making farmers more vulnerable to foreclosure. By 1930, registered bonds on farms reached the £91 million mark, representing 36 per cent of total farm values and an average indebtedness of nearly £1,000 per holding (Bradford 1986). After three years of serious drought, the Great Depression hit and some agricultural products declined to 25 per cent of their 1928 level. A broad range of farmers faced the prospect of bankruptcy (Minnaar 1990). The proportion of whites defined as ‘poor’ rose by 150 per cent and constituted about one-​sixth of the entire white population. Of the farmers who managed to hold onto their land, only a small proportion were ready, under these circumstances, to reinvest their capital as part of an accumulation strategy. Many had increased their risks through overly enthusiastic lending, but the resources they acquired in this way were rarely invested into improving the productivity

The Economic History of South Africa before 1948    41 of their farms. Rather than enter into risky investments, the majority of farmers preferred to consume whatever income they got their hands on. The pressures on government to save poor rural whites became intense. In response, marketing interventions were launched along with various other drought and debt relief measures. Officials and policymakers saw these as emergency measures. Once the depression ended and they had time to consider the situation, a new Marketing Act was formulated and passed in 1937. The Act was based on the then widely accepted idea that single channel, state administered, and subsidized marketing boards were the best way to protect the industry from the chaos of the previous decade. Those who genuinely believed in this solution wanted officials to control the boards so as to transcend the narrow interests of agricultural producers (De Swardt 1983). Instead, the council, like the marketing boards established under it, became the tool of organized white farmers, who used their political clout to push the state into raising agricultural prices ‘well above the competitive level’ (Wilson 1971). By making the costs of production of the most inefficient producers the basis for price setting, and by introducing land prices into the calculations, the boards ensured that, rather than agricultural efficiency, the long-​term social aim of keeping farming incomes in line with those in town became the primary purpose of the new marketing controls (Van Waasdijk 1954). The state therefore massively supported white farmers in the 1930s and 1940s, which allowed them to survive, often in uneconomical ways, and provided them with huge advantages over any rural Africans in neighbouring reserves who were still trying to produce for markets.

2.8  The Rise of Manufacturing and the Evolution of Industrial Policy, 1900s–​1940s The picture usually presented of manufacturing in the early 1900s is a collection of small, sickly firms, subject entirely to the whims of the market and the state. There were impediments preventing the sector from growing and becoming internationally competitive, but there were also many solid, dynamic firms driven by a determination to expand (Schirmer 2008). Johannesburg and Cape Town contained the largest and most diverse manufacturing activities, but industry had also taken root in other parts of the country. In the Cape, small towns like Grahamstown and King Williams Town had long accommodated small manufacturing firms. In Natal, Pietermaritzburg emerged as an early manufacturing centre. In the Transvaal, thanks largely to the initiatives of Sammy Marks, Vereeniging was a small but dynamic industrial centre. Two major industrial companies, the Vereeniging Brick and Tile Company and the Union Steel Corporation, operated from the town during the 1920s. The most important industrial region, outside of Cape Town

42   Stefan Schirmer and Johannesburg, though, was Port Elizabeth, where the relatively large boot and shoe industry led experts to label the city ‘the Liverpool of South Africa’ in 1917. There was even a company in the town called British United Shoe Machinery, which supplied local firms with some of the equipment required for shoe-​making. In the 1920s, the city’s industrial development was further enhanced when Ford and General Motors established assembly plants. After the First World War, the state set out seriously to promote manufacturing. An industrial policy emerged largely under the leadership of Jan Smuts. The approach that he encouraged involved intense engagements between an Industrial Board housed in the Department of Commerce and Industry, and firms represented by various industrial associations, as well as their umbrella body, the Federated Chamber of Industries (FCI). The Industrial Board tackled calls for tariff protection by individual firms or industrial associations by considering the merits of each case. The principle on which they based their determination was that state intervention should provide as much support as possible to emerging industries while minimizing the harm done to the future health of the economy as a whole. The Board generally rejected requests that would benefit one sector or firm at the expense of another. Manufacturers, in turn, made their case for tariff adjustments on the basis of detailed arguments, focusing on both the broader benefits of tariff changes as well as how these would affect the competitiveness of the local industry. Some industries took the issue of competitiveness more seriously than others, but strong views were frequently expressed that government support should be elicited not for short-​term survival, but for long-​ term improvements. Demand for lower duties on raw and intermediate goods were made just as often as demands for more protection. There were also aspirations, both within government and amongst established industrialists, to enter into export markets. In 1924, the situation altered radically after the election of the Pact Government, which brought together the National and Labour Parties under the leadership of General Barry Hertzog. The new government’s policy was to promote industrial development through dramatically raised tariffs across the board. Some have described this change as ‘[manufacturing] capital taking over the state and imposing its interests directly on and through it’ (Yudelman 1983). This is highly implausible, and the facts suggest that 1924 in fact represented a change in the opposite direction. Given the strongly immigrant and, especially, British character of the manufacturing community, as well as the economic importance of the international links cultivated by many, most manufacturers were strongly opposed to a party that stood for South African self-​sufficiency and Afrikaner nationalism. Manufacturers, in fact, viewed with alarm the way the new Board of Trade and Industries was set up and undertook its deliberations. During a 1926 meeting, the changed way in which government and organized manufacturing interacted was readily apparent. The minister of labour addressed the members of the FCI, telling them they needed to accept that all future tariff protection would largely depend on their willingness to provide employment and livelihoods to the largest possible number of white workers at wages determined by the state. The manufacturers responded by objecting to ‘the increasing tendency of the Government to enact legislation of a drastically restrictive

The Economic History of South Africa before 1948    43 character in regard to the control and conduct of private business undertakings’. This was clearly not a scene consistent with a state captured by manufacturing capital. The stated intentions of governments very rarely translate neatly into on-​ the-​ ground realities. Nevertheless, the relationship between manufacturers and the state had changed fundamentally. By 1929, the articles and the opinion pieces in the manufacturer’s journals were, to a large extent, no longer about influencing and engaging with industrial policy. The approach adopted by Herzog’s government set a very negative precedent that would form the basis for industrial policy in decades to come. Manufacturers lost the voice that they once had, and they never regained it. For most of the subsequent decades before South Africa became a democracy, they resigned themselves to receiving more and more protection as compensation for living with governments that were essentially hostile to business and the principles it represented. The rate at which manufacturing expanded was not immediately affected by the shift in policy. In the eight years spanning 1916 to 1924, the gross value of manufacturing output in 1938 prices expanded by 40 per cent. In the subsequent eight years, that is, after the raising of tariffs, output expanded by only 27 per cent. It was in 1933, though, that manufacturing really took off, and by 1948 the value of output had increased by 340 per cent (Feinstein 2005). After 1933, manufacturing was ‘swept forward’ by the expansion of the gold-​ mining industry and the devaluation of the South African pound. This process was accompanied by rapid urbanization, which in turn stimulated further demand for manufactured goods. Then, in response to the trade disruptions caused by the outbreak of the Second World War, manufacturers responded vigorously to the opportunities created by shortages of imported products, and to the new demands of the war effort. From 1916 to 1948 a doubling in the average number of employees per establishment occurred, reflecting a significant movement to larger more capital-​intensive factories. Capital per worker was 15 per cent higher in 1948 than it had been in 1924. The most significant change, however, was the increasing employment of Black workers in semi-​skilled positions. During these boom periods, the employment of Black workers outstripped the employment of whites, and by 1948 Blacks made up two-thirds of the industrial workforce. Evidence from the Wage Board from 1937 to 1956 shows that 16 per cent of skilled and 72 per cent of semi-​skilled workers were Black (Feinstein 2005).

2.9 Conclusion Throughout South Africa’s history, institutions—​ consisting of rules and norms structuring people’s interactions—​ tended either to protect the position of the privileged or to provide the insecure with some form of protection. When accommodative spaces for socially engaged entrepreneurs expanded between these almost universal imperatives, so did the drivers of long-​term economic change. Innovative

44   Stefan Schirmer farmers seeking greater integration into market-​based systems emerged and were accommodated in both the pre-​colonial and colonial eras. Gold and diamond mining drove forward long-​term development in powerful ways, largely because mine owners sought an accommodation with the state, and were then integrated into negotiated visions of social development. Unfortunately, this accommodation in the context of colonialism and racial oppression quickly produced an economically and socially destructive labour policy, while racially defined alliances between white farmers and colonial states drastically undermined the prospects for Black commercial farmers and produced segregated, unjust land allocations. The state and manufacturers experienced a period of productive accommodation, but that eroded as early as 1925. Nevertheless, the rise of manufacturing represented a huge challenge to the viability of the segregated, cheap-​labour system that had been set up in the context of an economy dominated by mining and agriculture. In a world of rapidly changing living and employment circumstances, Africans organized and demanded better treatment and rights. Within white politics, there were tentative movements towards reform, but these were countered by proposals from right-​wing, nationalist factions who proposed to extend the power of the state to halt any development towards a more integrated society. These proposals won the day in the all-​white elections of 1948. That was the beginning of four decades of apartheid, which would prove to be even more oppressive and economically damaging than the previous Segregation era.

References Ally, Russell. 1985. ‘The Development of the System of Individual Tenure for Africans, With Special Reference to the Glen Grey Act, c.1894 to 1922’. Unpublished master’s thesis, Rhodes University. Antonites, Alexander. 2012. ‘Political and Economic Interactions in the Hinterland of the Mapungubwe Polity, c.AD 1200—​1300, South Africa’. PhD thesis, Yale University. Beinart, William. 1980. ‘Production and the material basis of chieftainship’, in Shula Marks and Antony Atmore (eds) Economy and Society in Pre-​industrial South Africa. London: Longman. Bradford, Helen. 1986. A Taste of Freedom. Johannesburg: Ravan Press. Bundy, Colin. 1988. The Rise and Fall of the South African Peasantry, 2nd ed. Oxford: James Currey. De Swardt, Stephanus Jansen Jacobus. 1983. ‘Agricultural marketing problems in the nineteen thirties’, South African Journal of Economics, 51(1–​2):  1–​18. Delius, Peter and Stefan Schirmer. 2014. ‘Order, openness and economic change in precolonial Southern Africa: A perspective from the Bokoni Terraces’, Journal of African History, 55(1):  37–​54. Etherington Norman. 1985. ‘African economic experiments in colonial Natal, 1845–​1880’, in Bill Guest and John Sellers (eds) Enterprise and Exploitation in a Victorian Colony. Pietermaritzburg: University of Natal Press. Feinstein, Charles. 2005. An Economic History of South Africa: Conquest, Discrimination, and Development. Cambridge: Cambridge University Press.

The Economic History of South Africa before 1948    45 Guelke, Leonard. 1989. ‘Freehold farmers and frontier settlers, 1657–​1780’, in Richard Elphick and Herman Giliomee (eds) The Shaping of South African Society, 1652–​1840. Cape Town: David Phillip. Guy, Jeff. 1987. ‘Analysing pre-​capitalist societies in southern Africa’, Journal of Southern African Studies, 14(1):  18–​37. Hall, Martin. 1987. The Changing Past: Farmers, Kings and Traders in Southern Africa, 200–​1860. London: James Currey. Hall, Simon. 2009. ‘Farming communities of the second millennium: Internal frontiers, identity, continuity and change’, in Carolyn Hamilton, Bernard Mbenga, and Robert Ross (eds) The Cambridge History of South Africa, Volume 1: From Early Times to 1885. Cambridge: Cambridge University Press. Jeeves, Alan. 1975. ‘The control of migratory labour on the South African gold mines in the era of Kruger and Milner’, Journal of Southern African Studies, 2(1):  3–​29. Keegan, Tim. 1996. Colonial South Africa and the Origins of the Racial Order. Cape Town: David Phillip. Keeton, Lyndal and Stefan Schirmer. forthcoming. ‘Migration and state formation in pre-​ colonial South Africa: Why the nineteenth century was different’, in Ewout Frankema and Michiel De Haas (eds) A History of African Migration. Cambridge: Cambridge University Press. Krikler, Jeremy. 1993. Revolution from Above, Rebellion from Below: The Agrarian Transvaal at the Turn of the Century. London: Clarendon Press. Kubicek, Robert. 1972. ‘The Randlords in 1895: A reassessment’, Journal of British Studies, 11(2): 84–​103. Lukasiewicz, Mariusz. 2017. ‘From diamonds to gold: The making of the Johannesburg Stock Exchange, 1880–​1890’, Journal of Southern African Studies, 43(4): 715–​32. Maggs, Tim. 2010. ‘The 2009 FYI workshop and excursion: Valuable lessons from Eastern Africa’, African Studies, 69(2) 213–​18. Marks, Shula, and Stanley Trapido. 1979. ‘Lord Milner and the South African state’, History Workshop, 8:  50–​80. Mason, John Edward. 1989. ‘Fortunate slaves and artful masters: Labour relations in the rural Cape Colony during the era of emancipation, ca.1825–​1838’, in Wilmot G. James and Mary Simons (eds) The Angry Divide: Social and Economic History of the Western Cape. Cape Town: David Phillip. Mawby, Arthur Andrew. 1974. ‘Capital, government and politics in the Transvaal, 1900–​1907: A revision and a reversion’, The Historical Journal, 17(2): 387–​415. Mawby, Arthur Andrew. 2000. Gold Mining and Politics: Johannesburg, 1900–​1907. London: Edwin Mellen Press. Minnaar, Anthony. 1990. ‘The effects of the Great Depression (1929–​1934) on South African white agriculture’, South African Journal of Economic History, 5(2):  80–​94. Peires, Jeff. 1981. ‘Chiefs and commoners in pre-​colonial Xhosa society’, in Jeff Peires (ed.) Before and after Shaka: Papers in Nguni History. Grahamstown: Rhodes University Press. Ross, Robert. 1986. ‘The origins of capitalist agriculture in the Cape Colony: A survey’, in William Beinart, Peter Delius, and Stanley Trapido (eds) Putting a Plough to the Ground: Accumulation and Dispossession in Rural South Africa, 1850–​1930. Johannesburg: Ravan Press. Ross, Robert. 1989. ‘The Cape of Good Hope and the world economy’, in Richard Elphick and Herman Giliomee (eds) The Shaping of South African Society, 1652–​1840. Cape Town: David Phillip.

46   Stefan Schirmer Sansom, Basil. 1974. ‘Traditional economic systems’, in David Hammond Tooke (ed.) The Bantu Speaking Peoples of Southern Africa, 2nd ed. London: Routledge. Schirmer, Stefan. 2008. ‘The contribution of entrepreneurs to the emergence of manufacturing in South Africa before 1948’, South African Journal of Economic History, 23(1–​2), 184–​215. Schutte, Gerrit. 1989. ‘Company and colonists at the Cape, 1652–​1795’, in Richard Elphick and Herman Giliomee (eds) The Shaping of South African Society, 1652–​1840. Cape Town: David Phillip. Shillington, Kevin. 1985. The Colonisation of the Southern Tswana, 1870–​1900. Johannesburg: Ravan Press. Van der Merwe, Petrus. 1938. Die Trekboer in die Geskiedenis van die Kaapkolonie. Cape Town: African Sun Press. Van Onselen, Charles. 2001. New Babylon, New Nineveh: Everyday Life on the Witwatersrand, 1886–​1914. Johannesburg & Cape Town: Jonathan Ball Publishers. Van Waasdijk, Tom. 1954. ‘Agricultural prices and price policy’, South African Journal of Economics, 22(1–​2) 160–​173. Wilson, Francis. 1971. ‘Farming, 1866–​1966’, in Monica Wilson and Leonard Thompson (eds) The Oxford History of South Africa, Volume II. Oxford: Oxford University Press. Wilson, Monica. 1969. ‘ “The Nguni People”: “The Sotho, Venda, and Tsonga” ’, in Monica Wilson and Leonard Thompson (eds) The Oxford History of South Africa, Volume 1. Oxford: Oxford University Press. Yudelman, David. 1983. The Emergence of Modern South Africa: State, Capital, and the Incorporation of Organised Labour on the South African Goldfields, 1902–​1939. Westport, CT: ABC-​CLIO.

Chapter 3

The Ec onom i c H i story of Sou th Afri c a 194 8 –​9 4 Bill Freund † and Vishnu Padayachee †

3.1 Introduction The post-​war period under review here for South Africa can be split into two episodes: the first, roughly twenty-​five years until c.1970, was one of almost uninterrupted economic growth, and rising standards of living, especially for whites. Foreign reserves rose dramatically, on the back of growth in trade with the United States, the United Kingdom, and Japan. The country was the recipient of major loans from the World Bank for infrastructure development in the 1950s. The Rhodes University economist Hobart Houghton observed that few would have predicted that 1961 would be the ‘prelude to one of the greatest waves of economic expansion that this country has ever experienced’ (Dubow 2014: 100). Mining, manufacturing, and commercial agriculture all benefited. But there remained some key differences between social groups within the country. Blacks in general did not benefit to the same extent as whites. By the hundredth anniversary of the Great Trek in 1938, Afrikaners were rapidly urbanizing but they still formed a small proportion of the professional classes and were under-represented in the business and financial sectors. By the 1960s, Afrikaner entrepreneurs emerged in numbers. Jan Marias formed Trust Bank; Anton Rupert’s business empire (Rembrandt) grew strongly and diversified; Albert Wessels started in the clothing industry but made his name and fortune when he acquired the Toyota franchise in 1961. In 1961, the then prime minister established an Economic Advisory Council made up of state bureaucrats, academics, white trade unions, and leading businessmen (both Afrikaans and English) to advise the government on matters of economic policy. A widely cited but unproven statistic is that put forward by Hobart Houghton, who boldly claims that the South African economy grew as fast as that of any other country in the world, alongside Japan (Moll 1991: 271–​2). This narrative of high growth rates following Hobart Houghton has not gone unchallenged. Terence Moll, for one, argues persuasively

48    Bill Freund and Vishnu Padayachee Table 3.1: Conventional and recalculated economic growth rates before and after 1948 Real GDP growth rates Conventional

Recalculated

1920–​29

4.3

4.2

1929–​36

3.7

4.6

1936–​48

4.0

3.8

1948–​54

4.7

4.4

1954–​63

4.4

3.9

1963–​74

5.1

4.6

Source: Moll (1991: 275).

that a recalibrated calculation of post-​war economic growth shows instead a steady rather than a spectacular trend in global terms (Moll 1991). See Table 3.1. But the end of the Bretton Woods international regulatory system in 1971, the rise in the international price of oil after 1973, and rising Black wages following the strikes and Black worker mobilization that began in Durban in 1973, changed all that. The balance of payments, once described as the Achilles heel of the South African economy, became unstable once more. From the mid-​1980s and following the passing of the US Comprehensive Anti-​apartheid sanctions bill and the ending of all new IMF and World Bank loans, the international economic and financial pressure on apartheid South Africa was truly intensified, and forced the National Party tentatively to engage with the banned ANC, so setting the stage for the negotiations that were to lead to democracy in 1994.

3.2  Perspectives on the Academic Study of the South African Economy and the Economic History of South Africa We frame this chapter in terms of an analytical fusion between a ‘developmental state’ paradigm and a more traditional ‘economic growth’ narrative. Varying interpretations of South African economic history in this period, such as that based on the notion of the minerals-​energy complex (MEC) are also critically reviewed. And the role of the giant Anglo American Corporation (AAC) in shaping the economic history and development of the South African economy in this period is emphasized.

The Economic History of South Africa 1948–94    49 On balance we would describe our approach as non-​reductionist, non-​Marxist but materialist, in which issues of class and power are given some, but not a determining, role. The economic history of South Africa, we argue, going against the grain of the literature that downplays its place in academia, ‘has been [such] a significant domain of study for a relatively long time within South African history writing . . . that it has been enriched by a considerable amount of newer literature and debates, including those that straddle the social/​economic divide’ (Freund n.d.: 1) Studies of South Africa’s pre-​conquest African society ignored the economy except as a ‘factor’ among many, and focused on a static anthropological account based on still photographs. In the first half of the twentieth century, ‘liberal’ scholars, including D. Hobart Houghton, W. H. MacMillan, and C. W. de Kiewiet (1941), dominated the discourse, an analysis in which apartheid was seen as being irrational and dysfunctional to economic success and likely to fade in the course of economic development. In this tradition AAC executive Michael O’Dowd, following W. W. Rostow, argued that South Africa’s pattern of economic development was normal, and if that pattern continued radical constitutional reform would lead to the end of apartheid, around the year 1980 (1977: 35). A more muscular liberalism associated with Jan Lombard of the Reserve Bank1 and with academics like W. H. Hutt from the UCT Economics Department, the author of The Economics of the Colour Bar prevailed in some circles from the late 1930s to the 1980s. Hutt’s view has also been defined as neo-​liberal constitutionalism, where market order rather than democracy was seen as key to social and legal change. In that approach, inspired by Milton Friedman and by F. A. Hayek, order was prioritized above democracy, an idea that appealed to (white) South Africans who were grappling with models for constitutional change. From the 1960s, this market-​privileging, liberal approach was driven by people like Michael O’Dowd, Leon Louw, and Jan Lombard, who began to experiment with a range of ideas in part inspired by the arrival in the country of F. A. Hayek in 1963. Lombard was a leading member of the government and influential in the government’s economic thinking. But that focus of thinking changed with the arrival of the revisionist scholars including Johnstone, Legassick, and Wolpe in the early 1970s. These scholars tried to historicize African history in South Africa, by introducing a mode-​of-​production approach. In his seminal 1972 paper ‘Capitalism and Cheap Labour Power in South Africa’, Harold Wolpe argued that the articulation between capitalist and pre-​capitalist modes of production, manifest most clearly in the racially based migrant labour system, was critical to South Africa’s capitalist accumulation project. Far from being an irrational bunion on the face of an otherwise rational economic system, as liberals had argued or implied, race for Wolpe was key to South Africa’s capitalist development. Wolpe argued that capitalism found in the reserves of the cheap labour power that underwrote or subsidized accumulation (see also Hart and Padayachee 2013). A few years into that decade the young

1 

Author of Freedom, Order and Welfare, 1978.

50    Bill Freund and Vishnu Padayachee Poulantzians-​influenced South African scholars, including David Kaplan, Rob Davies, Dan O’Meara, and Mike Morris, then at Sussex University in the United Kingdom, introduced the ‘state’ into this revisionist literature in novel ways and for a while the study of ‘fractions of capital’ and their relationship to the state became a major feature of this new literature. Part of this new wave of economic history studies of South Africa focused on the history of labour as a leading force in the resistance to capital and the state. Merle Lipton (1986) is arguably the most sophisticated proponent of the liberal tradition. Her emphasis is on sectoralism in South African business and she argues that even if the gold-​mining sector needed cheap labour, given narrow margins, this was not the case for a growing secondary industry. She appeared alive to the conflicts and ructions amongst capitalists and with the state in the late apartheid era. Stephen Gelb in his seminal edited volume (which was commissioned by the labour federation COSATU and driven largely by its metal affiliate NUMSA) used a variant of French regulation theory to explain the origins of the economic crisis in South African capitalism. Racial Fordism was Gelb’s description of South Africa’s industrial economy, in which apartheid acted as the social and political form of regulating its capitalist system. In 2005, Charles Feinstein published his An Economic History of South Africa: Conquest, Discrimination and Development, the first economic history of South Africa in sixty years. Many chapters powerfully cover the story of the economy in the apartheid era. A key finding of the book was that both the apartheid government and radical, revisionist historians such as Wolpe failed to understand that modern economic growth does not ‘depend on crude exploitation of wage labour . . . it is achieved by means of technical progress, better human capital and advances in technology’ (2005: 248). For these reasons, he argues, economic growth in South Africa in the 1950s and 1960s, while impressive, was not as high as may have been possible with a more skilled and productive Black labour force. These are just a few of the ideas and trends which have influenced the study into the economic history of South Africa. In both direct and indirect ways they have shaped our interpretation of the narrative of the economic history of the period 1948–​94 that now follows.

3.3  The Afrikaner Revolution of 1948 and the Construction of the South African Developmental State In 1948 South Africa was primed in the direction of a state-​led development for whites through the earlier policy actions of the South African Party coalition government of General Smuts, which was defeated in the election that year. This included a series of

The Economic History of South Africa 1948–94    51 major planning initiatives through state commissions, the creation of the Council for Scientific and Industrial Research (CSIR) and the Industrial Development Corporation (IDC); the substantial expansion of ISCOR which gave South Africa largely self-​ sufficiency in iron and steel; and the establishment of a national grid as Eskom achieved virtual monopoly status. In the private sector, the dominant gold-​mining industry was in the process of developing huge new finds in the northern Orange Free State and the far west Rand. The Bretton Woods agreement assured South African gold mining of steady and predictable prices that allowed for major investments and ensured profits were substantial in the long term. The National Party in power generally carried this approach further albeit without the brain trust gathered around Smuts. The most significant achievement was the creation of new state-​owned enterprises (SOEs)—​Foskor, Safmarine, SAICCOR, and others, and especially SASOL, which led to a very substantial expansion of South Africa’s metal and chemical industries. SAICCOR was interesting because of its focus on exporting its product (rayon pulp), unlike most other state-​led investments which were aimed at the domestic market. The creation of Gencor as a mining company dominated by Afrikaners from the SANLAM stable represented a major concession from AAC that set the stage for cooperation between the Afrikaner-​run state and the largely Anglo-​ dominated private sector despite tensions sometimes emerging between them. In return, for instance, AAC was able to get involved in steel-​making through highveld on a relatively level playing field, and indeed created companies in sectors such as chemicals and paper pulp that competed with the state. Yet the National Party remained throughout this period suspicious of English capital as represented by companies like AAC (a commission of enquiry was even set up at a very high level in the 1960s to investigate the threat that AAC posed to the future of apartheid) and the government displayed a strong degree of anti-​Semitism and an ongoing concern about ‘foreign’ interference in its domestic policies. The term ‘Hoggenheimer’ was widely used within Afrikaner nationalist circles in this period to caricature local ‘Jewish’ capitalists as ‘rapacious and manipulative’ (Rajab 2017: 14). Links with the state was a major reason for the development of national oligopolies dominant in so many lines of business. Manufacturing for the home market attracted notable foreign investment. This included the substantial automobile industry, hallmark of consumerism. German and Japanese marques tended to replace their US predecessors. The dependence on foreign design and initial investment was coupled with state regulations on local content to promote manufacturing activity in South Africa. The net outcome was that the South African economy at the end of the Second World War was in a far stronger position than it had been at Union in 1910, largely because of the PACT government’s (the alliance between the National Party and the Labour party) industrial policy and its support for SOEs. But the economy remained weak and skewed in many important areas. These included the constraints arising from a low-​ wage, low-​productivity economy; the failure to extend training and skills development to the majority population; and the limited size of the home market and its heavy

52    Bill Freund and Vishnu Padayachee dependence on gold to pay for imported materials and capital goods and so balance its external accounts. From the early 1970s there was a sharp deterioration in the performance of the economy including in its balance of payments, as the country’s appeal to foreign investors diminished, political unrest grew, and financial sanctions began to bite. The government’s response included steps to tighten its apartheid measures. Evictions intensified (3,500,000 between the 1960s and 1983), homelands were consolidated, and so-​called ‘black spots’ were ruthlessly cleared. In short, African urbanization was strictly controlled; the government, under the influence of H. F. Verwoerd, rejected many of the more ‘liberal’ recommendations of the 1956 Tomlinson Commission. The Commission had recommended a large-​scale, ten-​year programme of land reclamation and development, expenditure of R200 million on rural development, major reforms of agriculture, and the establishment of white-​owned industries both inside and on the borders of the reserves (Feinstein 2005: 155). In general the government instead adopted a series of programmes to strengthen its apartheid programme; these included legislation that excluded Africans from the definition of ‘employee’, and Black trade unions, while not prohibited, were not granted official recognition. Racially mixed union membership was prevented. Job reservation was introduced and the whole approach to African education was transformed, through which church and Mission schools, previously responsible for African education, were stripped of state financial support and control was moved to the Department of Native Affairs (Feinstein 2005: 154–​8). Contrary to widespread pre-​war expectations, gold mining flourished after the war, with the discovery of new reefs in the Transvaal and Orange Free State (Feinstein 2005: 165). The simultaneous devaluation of the UK and South African pound raised the sterling price of gold by over 40 per cent, a level maintained until the early 1970s. The discoveries were made on the basis of scientific exploration and new towns such as Welkom sprung up in and around the new mines. The new mines operated with advanced mechanization. By the end of the 1950s these new gold mines produced the majority of the country’s gold production. A major beneficiary of these developments was AAC under the leadership of Ernest and Harry Oppenheimer who were able to raise vast sums of overseas capital, especially from the United States. With its profits on the rise AAC even briefly contemplated improvements in the standard of housing it provided for its African workers until this was strongly opposed by Verwoerd (Feinstein 2005: 167). Gold mining was not the only mining that boomed. Uranium production for nuclear weapons in the environment of the Cold War became a significant export; the diamond industry, for both gemstones and industrial stones, also grew strongly. Copper, manganese, vanadium, antimony, and chrome production also took off. Alongside these mining developments, manufacturing based on an Import Substitution Strategy also experienced an ‘unprecedented boom’ (Feinstein 2005: 172), its contribution to total output rising from 23 per cent in 1948 to almost 31 per cent in 1970. Industry also diversified: AECI, an explosives manufacturing company, was established; other manufacturing companies set up in the post-​war era included Board and Hard Metals (industrial diamonds) and Scaw Metals (cast-​steel grinding balls). The National Finance Corporation (NFC) was set up, with AAC playing a big role along with

The Economic History of South Africa 1948–94    53 City of London firms and the support of the Reserve Bank in developing a more comprehensive money market, something that the Reserve Bank had failed to support earlier. It is worth noting the new impetus from the state to promote scientific and industrial research linked to economic performance. Here the Smuts appointee Basil Schonland played a vital role in setting up the Council for Scientific and Industrial Research (CSIR), insisting on its ‘independent’ status and making it imperative that it work collaboratively with other research organizations including those within the major universities. The CSIR was based on existing British, Canadian, and especially Australian models, which gave it autonomy and authority (Dubow 2006: 243–​4). An important development after 1948 was the growth of Afrikaner capital. Here the Rembrandt company group (led by Anton Rupert) was a noteworthy development. Its initial business was very locally based and built around tobacco and cigarette manufacturing, but it grew to take control of the international cigarette company Rothmans, while also branching out to other more diverse and luxury activities and growing its share of the market capitalization on the local stock exchange. By 1985 the company controlled 3.8 per cent of the share capital on the Johannesburg Stock Exchange, doubling this figure to 7.6 per cent within just three years (Fine and Rustomjee 1996: 103). In the early 1980s a major feud broke out between the former friendly Afrikaans heavyweights, Rembrandt and the insurance company Sanlam, over attempts by Sanlam’s executive to strengthen its control of the mining house General Mining (see Dommisse 2005: 254ff.). Another important development after 1960 was the government’s decentralization policy, a version of regional industrial policy that (according to Trevor Bell, the foremost scholar in this research area) was initially aimed less at economic growth and efficiency goals and more on the roll-​out of the state’s grand apartheid strategy (1997: 1). The aim was to stem the flow of Black job-​seekers to urban areas. Later, the policy adopted a more industrial-​policy character and was supported by local business as evidence of the government’s commitment to free enterprise. Large incentives were in fact offered to firms to operate in these industrial zones in the so-​called Border Areas. The creation of the Development Bank of South Africa in 1979 to promote regional investment was an important supporting development (Freund 2019: 194).

3.4  The MEC and Some of its Critics Fine and Rustomjee (1996) have characterized the post-​war South African economy as being dominated by an MEC, that is, incorporating a core of industries associated with large-​scale minerals extraction backed by cheap energy, a point first made by Renfrew Christie (1984). But linkages between the state and private corporations also played a vital role in the evolution and consolidation of the MEC, as Hein Marais has argued, pointing to the development of large-​scale electricity capacity and an indigenous fuel-​chemical industry (2011: 19). It was not just the weight of this subset of interrelated economic

54    Bill Freund and Vishnu Padayachee activity but also its determining role throughout the rest of the economy that mattered (Fine and Rustomjee 1996). The MEC was not without its critics. Arguably the most powerful criticism came from Trevor Bell and Greg Farrell (1997). While praising the notion of the MEC as employed by Fine and Rustomjee for highlighting the minerals-​rich nature of the South African economy, Bell and Farrell argue that there is no historical evidence to support the contention that the MEC as a system of accumulation prevented diversification of manufacturing. They show that manufacturing did in fact diversify between the wars and in the post-​war period. Here is their central point of difference with Fine and Rustomjee: Neither in the inter-​war period, nor through to 1972 in the post-​war period, has South African industrialisation been characterised in any relevant sense by a failure to diversify out of MEC manufacturing, and in so far as performance deteriorated after 1972, this has not been due to the MEC as a system of accumulation. (1997: 610)

Bell and Farrell, it can be argued, are representative of a tradition of writing about the evolution of South African economic history that could be characterized as ‘liberal’. Theirs is a kind of (benign) South African nationalism which emphasized the positive aspects of manufacturing and commerce on South African economic development. An (early) orthodox Marxist view associated with Jack and Ray Simons shared some of these views. The South African narrative on the MEC and on Import Substitution Industrialization (ISI) echoes the policy debate over ISI in Latin America, in which towering figures such as Albert Hirschman engaged scholars such as Rosenstein-​Rodan (see Hirschman 1968). A body of literature linking Latin American policy debates to South Africa may have helped here but has been imperfectly developed in the literature. We are not convinced that this dispute between Fine and Rustomjee and their critics could or can ever have been resolved by recourse to statistics and data. The use of the MEC as employed by Fine and Rustomjee (1996) to describe the system of South African accumulation, initially dismissed by its neo-​liberal critics, is now widely accepted and sometimes used (uncritically) in the economic history literature. Sampie Terreblanche (2012) used it extensively though also somewhat naively in his later work as did many other more sophisticated analysts such as Bill Freund (2009), who follows Fine and Rustomjee in his chapter on energy (2019: 171ff.). David MacDonald also uses this notion in his book Electric Capitalism (2008). We are ultimately sympathetic to the Fine and Rustomjee understanding of MEC as a system of South African accumulation, while accepting that its value in the later part of our period as the economy globalized, ‘financialized’, and gradually moved away from minerals and mining, has waned. In important ways, however, this post-​war development strategy and policy (argued from within or outside an MEC thrust) lacked some of what could prove essential to economic growth. The state did not create a very significant financial sector aimed at development. The IDC was not permitted more capital than its initial establishment covered.

The Economic History of South Africa 1948–94    55 The Reserve Bank was a powerful force for traditional banking, with the de Kocks, father and son, promoting an approach aimed largely at interest rates and regulating foreign exchange in the shadow of Klasie Havenga. In 1932, and again in 1949, the South African government faced a tough choice about whether or not to follow the United Kingdom in devaluing the local (South African) pound in line with the United Kingdom. Klassie Havenga was the minister of finance on both occasions, getting it wrong (by most accounts) in 1932 at great cost to local farmers, and getting it right in 1949 by devaluing at the same time as the United Kingdom and to the same degree (de Kock 1952: 312). AAC and Goldfields dominated financial markets and banking otherwise. Related to this was the absence of an orientation towards competitive international manufacturing, often stated bluntly initially. It was believed that the gold-​mining revenues were enough to promote a good balance-​of-​payments picture. This would be critiqued by later government commissions but never led to successful action. There was little consideration about sources of power (electricity). It was assumed Eskom, and the coal mines that serviced Eskom, could be counted on for the inexpensive electricity that led South Africa to becoming a very heavy user and a massive polluter. It was assumed with little questioning that manufacturing, mining, and agriculture should depend on a coterie of highly rewarded executives and a body of skilled men who, with their families, were increasingly able to participate in lucrative consumer society. Beyond this was a mass of dehumanized and cheap Black labour. It is conventionally assumed that cheap labour provided the foundation for South African economic success. This is partly true, but it is important to note that as National Party rule carried on, the wages of whites, often in labour unions, formed a significant part of the wage bill. The new SOEs employed many whites—​until 1990—​and white unemployment was largely eliminated. Indeed, for whites, something of a welfare state was created with inexpensive medical care for all, national pensions, considerable public housing and rent controls. Maybe most important was the expenditure on technical and higher education in the Afrikaans language, which led to the wide diffusion of industrial skills. In some respects, the Black population benefited from being understood as a factor in the process of industrialization. While under Smuts there was almost a total failure in the provision of urban housing, the National Party moved to find the means to create the big townships, notably Soweto. Primary education was substantially expanded and health care for active workers improved, after Smuts’ rejection of a national health programme, through the facility of some big urban health centres such as Baragwanath Hospital. However, it should be noted that Blacks, an increasingly large majority of the population despite white immigration being encouraged, had only an extremely limited purchase on the expanding consumer society. They formed a very underdeveloped market. Moreover, they were more or less excluded from acquiring the skills or business opportunities (or the right to form and join licit trade unions) critical to making a better life in these circumstances. It is true that after 1960 the numerous Indian and coloured South African minorities started to benefit from enhanced opportunities in significant areas, but only at the very end of the apartheid era did things begin to change (marginally) for Black Africans.

56    Bill Freund and Vishnu Padayachee Major housing projects were funded and completed for Indians and coloureds. Schools were built for them and two universities, one for coloureds in the Western Cape and one for Indians in Durban. Businessmen such as Vivian Reddy (who owns the electricity and construction company Edison Power), who had close relationships with the Indian Affairs Department and the Zulu king and traditional prime minister (Buthelezi), were major beneficiaries of such infrastructure projects. The economic growth path went hand in hand with extremely high levels of inequality correlating with the racial categorization used by the state. As argued by Wolpe (1972) Black male workers coming from so-​called locations as migrants in low-​skilled jobs in profusion and returning to those locations where little monetized economic activity existed, was ceasing to work effectively. Black rural life was breaking down sociologically and economically. Moreover, the state was concerned with fighting the beswarting van die platteland (the ‘blackening’ of the rural areas). It promoted the development of large well-​organized and lucrative capitalist agriculture, with small-​scale white farmers turned into urban skilled workers and the mass of Black workers, often enjoying non-​cash paternalist relations with the landowners, being forced off the land into urban areas. Agriculture was heavily supported by the state but did not produce significant export income.

3.5  AAC and the Mining Industry It is important to say a little more here about the mining industry and the role of AAC and its relationship with the state in this post-​war period. AAC’s relationship with the apartheid state is frequently portrayed as one of tension and conflict but this was not the case, as the relationship between them was far more nuanced and complex. As argued by Hart and Padayachee (2013) Afrikaner nationalist concerns about AAC and its disproportionate influence over South Africa’s economy (and politics) did not diminish over the following decades. By the late 1930s, Oppenheimer realized his dream of taking complete control of de Beers and the London-​based diamond syndicate; and AAC significantly extended its operations in mining, finance, and industry across the whole of southern Africa. Oppenheimer skilfully managed his company through many political minefields, even through the depression years when threatened production cut-​backs in diamonds and gold were fiercely resisted by the Nationalist Government of Hertzog because of the unemployment that would result among its Afrikaner mine-​ working supporters. Oppenheimer was able to ignore or circumvent these tensions, relying on the strength of English capital, especially gold-​mining-based capital, which was the single-​most important source of government revenues, and his direct and ‘considerable influence at the political centre’ (Pallister et al. 1987: 60). The period 1930–​60 was a remarkable period in the history of the mining industry, which in turn had linkages into both manufacturing and finance. The most significant event was the discovery of new gold fields in the Orange Free State province in the 1940s.

The Economic History of South Africa 1948–94    57 AAC was at the forefront of these developments. As Innes remarks, ‘Anglo’s takeover of groups like SA Townships and Lewis and Marks was an important manifestation of the growing tendency towards increasing centralization of capital and control in the industry (Innes 1984: 137). By the 1950s AAC had become active in developing the local money market in South Africa. Ernest Oppenheimer helped to set up the National Finance Corporation in 1949, and in 1955 AAC set up its own private merchant bank Union Acceptances Ltd (UAL)—​and further developed and diversified its industrial interests. Despite some crude characterizations of the relationship between the new National Government and the AAC as poor or even antagonistic (a view rejected by Terreblanche 2012: 305), there is evidence that each side recognized the importance of the other to future success. Thus, apart from the role that government and AAC played in the establishment of a local money market in the 1950s, there are other such examples of their mutually supportive relationship. Both the leading state corporations, Eskom and ISCOR, maintained close working relations with AAC, and ‘the NP maintained these partnerships despite its suspicion of mining capitalists’ (Terreblanche 2012: 344). In a footnote Terreblanche (2012: 369) points out that one result of the ‘close co-​operation between HJ van der Bijl of Eskom and Ernest Oppenheimer of the AAC’ was the expropriation early in 1948 of the privately owned Victoria Falls Power Company, ‘with huge financial support from the AAC’. As the major mining house and consumer of electricity, the AAC benefited enormously from the cheaper electricity supplied to it by Eskom. With the support of Oppenheimer, ISCOR was also making headway towards becoming a steel monopoly before 1948 (Terreblanche 2002: 369). Nancy Clarke points out that ISCOR had strong partnerships with AAC from just after the Second World War (Clarke 1994). Recent efforts have been made to compare post-​1948 Afrikaner and post-​1994 Black economic empowerment. One specific issue is around AAC’s sale of its (approximately) 23 per cent share in General Mining and Finance Corporation (Gencor) to Federale Mynbou (Fedmyn). Was this an exercise in Afrikaner empowerment? Hermann Gilliomee (2008) has correctly argued that this had little to do with AAC’s concern with growing Afrikaner participation in gold mining, or because of the poor management at Gencor. Rather the decision to sell was made following an undertaking sought by Harry Oppenheimer from the Sanlam Group that they would limit their existing diamond business and allow de Beers to have the controlling interest in any new diamond business that Sanlam established. So it was de Beer’s monopoly in diamonds that AAC was aiming to secure and protect through these deals. And neither were the shares ‘given’ over to Fedmyn (Gilliomee 2008). This was no handout, shares were bought at market price (Gilliomee 2008: 777), a sharp contrast with the current post-​apartheid mantra, as Durban businessman Neville Kerdachi notes, that ‘Black people do not pay for shares’ (Neville Kerdachi, Interview 10 October 2017). While the massive state interventions and controls of the immediate post-​war era had ended and a shift to market-​oriented economic reforms was being rolled out, an equally powerful network of insidious state security interventions and organizations emerged

58    Bill Freund and Vishnu Padayachee aimed at countering what the Prime Minister P. W. Botha referred to as a ‘total onslaught on the state’ (see Morris and Padayachee 1988).

3.6  The Economic Crisis of the Early 1970s The 1970s represented a turning point when the developmental-​ state thrust lost traction as economic growth noticeably slowed. Forced removals in the interests of making different racial trajectories practicable, was costly and engendered bitterness. Dependence on steady gold-​mining revenues ceased to be possible once Bretton Woods was abandoned by the US President Nixon in 1971. Thereafter profits generated by gold-​ mining, the expansion of which involved ever-​increasing expenditure on technology, became unpredictable. The mining sector pivoted to base metals, notably to coal and to iron ore. However, transportation to the new ports of Saldanha and Richards Bay was a major additional expense and financing these ventures involved important debts to mainly European banks. The state became very anxious about dependence on foreign loans especially from Switzerland, and Germany and Britain as the United States reduced its exposure to South Africa (see Padayachee 1988). Dresdner Bank, Union Bank of Switzerland, and Swiss Banking Corp were the main supporters of apartheid South Africa (Padayachee 1988: 371). The state managed this through financial juggling but initially without propounding a real new economic strategy beyond trying to fall in with the growing neo-​liberal international consensus as best it could. The impulse to look beyond mining weakened substantially. Eventually by the mid-​1980s the government could no longer avoid a unilateral debt standstill which was followed by a debt rescheduling agreement brokered by Fritz Leutwiler, the former head of the Swiss National Bank, and supported technically by Price-​Waterhouse and partners, a London-​based firm of accountants. Leutwiler was a long-standing supporter of South Africa, personal friend of former finance minister Owen Horwood, and had brokered South African loan applications before in troubled times. The Swiss banks were deeply implicated in supporting apartheid South Africa through the Zurich gold pool, in marketing South African gold, and in melting down and restamping South African gold as Swiss. Leutwiler was in an ideal position to ease tensions between the Swiss and German banks on the one hand and on the other, the American banks, such as Chase Manhattan, accused by the Europeans of taking overhasty action against South Africa (Padayachee 1989: 264). Various commissions of inquiry were set up to advise on the direction and content of economic policy. These included commissions on monetary policy (de Kock), trade and industrial policy (Kleu, du Plessis), and labour market policy (Wiehahn) among others. The broad thrust of all these commissions was to shift the thrust of economic strategy from the state-​led, developmental strategy characteristic of the immediate post-​war era

The Economic History of South Africa 1948–94    59 to a market-​oriented one, and one recommending the fostering of a stable, permanent Black, semi-​skilled workforce (see Morris and Padayachee 1988). The mass strikes in and around Durban in 1972/​73 were among the most unsettling for the apartheid regime. Jay Naidoo, later general secretary of COSATU and a minister in Mandela’s first cabinet, maintains that the 1973 strikes were in fact more important than Soweto, ‘because it wasn’t in the townships, it was in the cities, it was in the factories, it was hitting at the core of white power, which was economic power’ (Waldmeir 1997: 27). These strikes inspired worker mobilization and spawned a number of emergent unions across the country, mostly independently of the ANC’s reach. Some of these unions were industry based, some craft based, others were generalist in character. A regrouping followed along more political lines: some developed along Black consciousness lines and foregrounded Black leadership. The Council of Unions of South Africa (CUSA), which included the powerful National Union of Mineworkers, was one. Another grouping organized was within the formation that became the Federation of South African Trade Unions (FOSATU). FOSATU championed worker control and democratic accountability. It eschewed close political relations with more ‘populist’ political formations such as the United Democratic Front and gained as a result the pejorative label of being ‘workerist’. These and other unions later merged into the Congress of South African Trade Unions (1985), which after a fierce debate became more aligned to the ANC (Dubow 2014: 226–​30). The state responded to these strikes and mobilization by setting up the Wiehahn Commission into labour relations, and Nic Wiehahn’s report was published in 1979. It recommended the abolishment of statutory job reservation along racial lines; proposed that Africans be included in the definition of ‘employee’; and it recommended the official recognition of Black trade unions, as well as the legalization of collective bargaining in the hope that this step would reduce industrial conflict (Dubow 2014; Baskin 1996, especially pp. 21–​40; Feinstein 2005: 241). The state, moreover, was taken up with new political challenges that impacted on the economy. First was the exterior border challenge. The collapse of Portuguese colonialism led to the dramatic expansion of expenditure on the military and defence in general. South Africa even acquired the nuclear bomb. The Border Wars were considered of the greatest importance especially given the involvement of the Soviet Union, a substantial global military power. A large military-​industrial complex was created that took in much of the private sector as well as the SOEs and new state initiatives. South Africa was fairly effective at fighting on the borders and defeating Umkhonto we Sizwe, the ANC army, but at some cost. Some South African military equipment was very competitive on quality and price in global markets but these markets involve heavy state involvement and are hard to break into so this proved largely to be economically a dead-​ end orientation. The state created initiatives that were unprofitable and only useful for assuaging its obsession with security. Second and related, the state equally feared the impact of sanctions as apartheid became less and less acceptable in the West. The emphasis on protectionism and the home market intensified constantly and finally in the late 1980s foreign exchange controls became paramount. Third, the state was unable to

60    Bill Freund and Vishnu Padayachee dam up and suppress growing violent unrest in the townships, the emergence of large squatter settlements, and the failure of state-​created administrative bodies to function. In the early 1960s, and in response to the earlier crisis associated with Sharpeville, the state had virtually stopped construction apart from hostels in urban townships, as they were called, and intensified its commitment to developing the ‘Bantustans’. The homelands system made it possible to deflect social and economic crises, including rising unemployment, into the backwater of the reserves, well hidden from ‘white’ South Africa. This move in particular was responsible for introducing unprecedented corruption into the system and the state half paid the wages of firms prepared to—​and partially forced to—​move, first to the edge of these homelands, and then to industrial parks set up within them. Trevor Bell was the notable economist who imagined that this could be a shift to the use of cheap rural labour that would lead to a form of industrial growth. Frame Textile Group, a huge textile empire established by Philip Frame in 1928 in the province of Natal, is an example of such a company but it did this without an export component of importance. With hindsight, this was not a formula that was going to work; the industrial parks collapsed and virtually died after the end of the old regime. Finally, it is worth noting that relations between the private sector and the apartheid state in our period were both complex and changing. There were growing tensions but also co o ​ peration with the private sector including Afrikaner businessmen. Rembrandt’s Anton Rupert largely restricted himself to propagating the economic development of the homelands while Andries Wassenaar attacked the financial profligacy of the state. Wassenaar, a top SANLAM executive, denounced the emphasis in the system on state control and guidance in a notorious book in 1977 entitled Assault on Private Enterprise: The Freeway to Communism. Both Rupert and Wassenaar enjoyed a good relationship with Verwoerd’s successor, B. J. Vorster.

3.7  Growing International Financial Pressure in the 1980s By the time we get to the 1980s South Africa’s economy was in a vastly weakened state. Growth had slowed since the mid 1970s, political tensions had escalated, and international financial pressure was mounting. Chief and arguably the most fatal of these pressures was that exerted by the international banks and by the IMF and World Bank. South African borrowing from the World Bank for infrastructure development had been critical to the country’s high rate of economic growth in the 1950s and 1960s. But the last loan from the World Bank had been repaid in 1967 and no new loans were taken out after that. South Africa borrowed heavily from the IMF in the mid 1970s when it experienced severe economic and political difficulties, and again in 1982. IMF assistance to South Africa in the mid-​1970s was greater than the combined assistance to all other African countries during that time. So the curtailment of IMF loans to South

The Economic History of South Africa 1948–94    61 Africa in 1983, was to prove decisive, and forced the apartheid regime to rethink its strategy. The flow of international credit (loans and bonds) to the apartheid state, especially in the period 1972–​76 and again in 1980–​82, in support of the state’s strategic and infrastructural development had been crucial to keeping the economy and the political system on track. Citibank international bank (CIBL) had raised massive loans for both the private and public sector in South Africa since its establishment in London in 1972. US banks were particularly active in South Africa in the decade of the 1970s and early 1980s but it was not the only region so involved. Thirty-​six banking groups participated in mobilizing eurocurrency credits for South Africa between 1972 and 1976. Non-​US banks were mainly from Britain, West Germany, and Switzerland and the banks involved included such banking heavyweights as Citibank, Chase Manhattan, Manufacturers Hanover Trust, and Barclays Bank International. Borrowing from private bank international sources during this time could best be described as an ‘orgy’ (Padayachee 1989: 227). But technology transfers were as important as loans. John Suckling has estimated that 60 per cent of South Africa’s GDP growth between 1957 and 1972 originated from exogenous technical change and technology (Suckling 1977). As we have noted, these loans and credits and technology transfers began to dry up by the mid 1980s, together with foreign direct investment after 1976 (Padayachee 1989: 223). By 1982 the gold-​led boom was over and the South African economy fell into a recessionary phase almost continuously from 1981, not helped by the international crisis brought about by the Latin American debt crisis, among other factors. South Africa abolished exchange controls and received plaudits from the IMF among others. But the upheaval in South African townships and factories from September 1984 forced international bankers to rethink their lending to South Africa. On 1 August the New York Times reported that Chase Manhattan Bank, the second-​largest lender to South Africa, had ceased extending credit to South Africa. Other banks followed suit during August and the Rand dropped to a record low on 27 August 1985. ‘South Africa’s long and mostly favoured relationship and status with private international banks and its international financial market in general, had reached its nadir’ (Padayachee 1989: 242). Gelb and Innes (1985), in an important intervention about the post-​1980 economic crisis confronting South Africa, argued that the state’s efforts at adopting a monetarist approach to resolve the crisis would fail, as it depended on high economic growth which looked impossible as long as the political–​economic crisis persisted—​a double bind (1985: 39). For the reasons already referred to, including tightening international sanctions and domestic resistance to apartheid, South Africa in the final five years of National Party rule, the government was forced to consider opening negotiations with the liberation movements, including the African National Congress. On balance, we would argue that the initial conditions of the South African economy in 1994, viewed from a macroeconomic-​balance perspective, were not those of extreme instability of the kind which had characterized some developing countries at the point of their economic and political transition (Padayachee and Van Niekerk, 2019).

62    Bill Freund and Vishnu Padayachee

3.8 Conclusion We would argue that growth rates in the 1960s and early 1970s, the two decades from 1973 were for South Africa ones of declining economic growth, falling net investment, rising unemployment, falling average real wage rates, and unacceptably high levels of poverty and income inequality. The levels of provision of physical and social infrastructure for the majority population were totally inadequate. However, a tight monetary regime ensured that there was relative price stability and that the inflation rate was relatively low; and foreign debt, due to special circumstances, had been reduced to very low levels compared with other indebted middle-​income developing countries. The real macroeconomic problem lay in the fiscal side and in the external account, where, since 1985, the apartheid regime had been forced to run trade account surpluses to meet debt repayments and to compensate for large net capital outflows. The years of National Party rule (1948–​94) brought about an affluent consumer society with sophisticated institutions and oligopoly c​ontrolled businesses, largely orientated to and enjoyed by a white minority, a proportion of the population which was steadily shrinking. This minority on the whole had succeeded fairly well in surviving its lost state power thanks to the advantages accrued before 1990. For the majority, the verkrampte social vision meant that Blacks were dramatically deficient in capital and skills, and their social needs in an industrializing and urbanizing society were poorly catered for. This was the inheritance with which the post-​1990 negotiations had to cope. It is harder to make the case quantitatively but the continued reliance on natural resources and heavy industry—​the minerals-​energy complex as it was dubbed by Fine and Rustomjee tellingly—​took a large environmental toll as well on South Africa’s previous natural state; costs here are only incrementally being put on the table as critical environmental analysis of energy use and the state of water supplies takes shape. Greenpeace has estimated that ‘the economic impacts of air pollution from fossil fuels is important to consider in the context of Eskom’s massive debt crisis, which amounts to over R400 billion’. According to this report, the estimated total cost of air pollution from fossil fuels in South Africa is a staggering Rand 94.7 billion every year.2 South African capital was hungry to be mobile and free from state constraints, to join in the new wave of ‘globalization’ sweeping the globe but which the opprobrium generated by South Africa’s version of capitalism made difficult or impossible. It was ironically the fruit of the very process that had empowered South African private capital. This led business leaders to defy the government and begin discussions with the ANC. Even Afrikaner business leaders such as Anton Rupert, for example, as well as aligned Afrikaner academics, did the same. Some of the classic developmental initiatives of the mid-​century were failing. ISCOR, for example, had become very unprofitable and in 2 

https://​www.greenpeace.org/​africa/​en/​press/​8939/​air-​pollution-​from-​fossil-​fuels-​costs-​the-​world-​ r120-​billion-​every-​day-​greenpeace/​.

The Economic History of South Africa 1948–94    63 1989 the state was thrilled to be able to sell it to an international buyer. The huge 1987 gold mine strike was successful in leading to union recognition for Black miners but it also led to many dismissals. The gold-​mining industry, increasingly mechanized, has declined in gold production and number of employees ever since. Eskom, for long the poster boy of National Party development, gets pushed aside in the late 1980s by P. W. Botha, a prime minister eager to favour nuclear power in Koeberg, and a growing military industrial complex represented by Armscor as icons of South Africa’s technological sophistication (see Freund 2019: 174). Here the need to show off the country’s national capabilities may have run counter to the liberalizing trends being pushed onto the policy agenda by some key figures in the state. To some extent the election of F. W. de Klerk in 1989 resolved this tension in favour of the market-​oriented reformers. We conclude by looking at the interpretations of this period by two well-​known historians, each with a different take on the overall performance of the South African economy. The Yale historian Jacob Dlamini reminds us that some view the second half of the twentieth century in South Africa with a certain nostalgia, as an era of fairly robust (even speculative) economic growth, efficiency, a capable state bureaucracy, and order. In this vein he quotes the Stellenbosch historian Herman Giliomee who argues that apartheid delivered dynamic economic growth between 1948 and 1994 ‘accompanied by the development of a sophisticated infrastructure and a steady increase in the life expectancy of all population groups’ (2008: 666). Dlamini challenges this view, arguing that the system was ‘messy, inefficient and violent. Apartheid did not work and, as a result, eventually collapsed under its own weight’ (2020: 14).

Acknowledgements Bill Freund sadly passed away on 13 August 2020 and was not able to complete this chapter. The editors of the volume asked me to do so. We found a two-​thousand-​word draft of text and notes, which I used as a framework for the chapter and tried to write it up in the style of Bill Freund as best as I could (Padayachee 1996).

References Baskin, Jeremy. 1996. Against the Current: Labour and Economic Policy in South Africa. Johannesburg: Ravan Press. Bell, Trevor. 1997. ‘South African regional industrial development policy: Critical issues’, Transformation, Critical Perspectives on Southern Africa, 32:  1–​30. Bell, Trevor, and Greg Farrell. 1997. ‘The minerals-​energy complex and South African industrialisation’, Development South Africa, 14(4): 591–​613. Christie, Renfrew. 1984. Electricity, Industry and Class in South Africa. London: Macmillan. Clark, Nancy. 1994. Manufacturing Capital: State Corporations in South Africa. New Haven, CT: Yale University Press.

64    Bill Freund and Vishnu Padayachee De Kiewiet, C.W. 1941. A History of South Africa, Social and Economic. London: Oxford University Press. De Kock, Gerhard. 1952. A History of the South African Reserve Bank. Pretoria: JL van Schaik. Dlamini, Jacob. 2020. Askari, a Story of Collaboration and Betrayal in the Anti-​apartheid Struggle= Auckland Park: Jacana. Dommisse, Ebbe. 2005. Anton Rupert, a Biography. Cape Town: Tafelberg. Dubow, Saul. 2006. A Commonwealth of Knowledge, Science, Sensibility and White South Africa 1920–​2000. Oxford: Double Storey. Dubow, Saul. 2014. Apartheid 1948–​1994. London: Oxford University Press. Feinstein, Charles H. 2005. An Economic History of South Africa, Conquest, Discrimination and Development. Cambridge: Cambridge University Press. Fine, Ben, and Zavareh Rustomjee. 1996. The Political Economy of South Africa, from Minerals-​ Energy Complex to Industrialisation. Johannesburg: Wits University Press. Freund, Bill. (n.d). ‘Economic history in South Africa: An introductory overview’. Unpublished, Durban. Freund, Bill. 2009. ‘The significance of the minerals-​energy complex in the light of South African economic historiography’, Transformation: Critical Perspectives on Southern Africa, 71. Freund, Bill. 2019. Twentieth-​century South Africa, a Developmental History. Cambridge: Cambridge University Press. Gelb, Stephen, and Duncan Innes. 1985. ‘Monetarism’s double bind’, Work in Progress, 36: 31-39. Gilliomee, Hermann. 2008. The Afrikaners, Biography of a People. Cape Town: Tafelberg. Hart, Keith, and Vishnu Padayachee. 2013. ‘A history of South African capitalism in national and global perspective’, Transformation, Critical Perspectives on Southern Africa, 81(1):  55–​85. Hirschman, Albert. 1968. ‘The political economy of import-​substituting industrialization in Latin America’, Quarterly Journal of Economics, 82(1):  1–​32. Innes, Duncan. 1984. Anglo American and the Rise of Modern South Africa. London: Heinemann Educational Books. Legassick, Martin. 1974. ‘South Africa: Capital accumulation and violence’, Economy and Society, 3: 253–​291. Lipton, Merle. 1986. Capitalism and Apartheid, South Africa, 1910–​ 1986. Aldershot: Wildwood House. Lombard, Jan. 1978. ‘Freedom, welfare and order: Thoughts on the principles of political co-​operation in the economy of southern Africa’. Bureau for Economic Research, Bantu Development, Pretoria. MacDonald, David. 2008. Electric Capitalism, Re-​colonizing Africa on the Power Grid. Cape Town: HSRC Press. Marais, Hein. 2011. South Africa, Pushed to the Limit: The Political Economy of Change. Cape Town: UCT Press. Moll, Terence. 1991. ‘Did the apartheid economy fail?’, Journal of Southern African Studies, 17(2): 271–​291. Morris, Mike and Vishnu Padayachee. 1988. ‘Hegemonic projects, accumulation strategies and state reform policy in South Africa’, Labour, Capital and Society, 22(1): 65–​109. O’Dowd, Michael. 1977. ‘The stages of economic growth and the future of South Africa,’ ‘in Lawrence Schlemmer and Eddie Webster (eds) Change, Reform and Economic Growth in South Africa. Johannesburg: Ravan Press.

The Economic History of South Africa 1948–94    65 Padayachee, Vishnu. 1988. ‘Private international banks, the debt crisis and the apartheid state’, African Affairs, the Journal of the Royal African Society, 87(348): 361–​76. Padayachee, Vishnu. 1989. ‘South Africa’s International Financial Relations: History, crisis and transformation’. Unpublished PhD thesis, University of Natal, Durban. Padayachee, Vishnu, and Robbie van Niekerk. 2019. Shadow of Liberation, Contestation and Compromise in the Economic and Social Policy of the African National Congress. Johannesburg: Wits University Press. Pallister, David, Sarah Steward, and Ian Lepper. 1987. South African Inc: The Oppenheimer Empire. Johannesburg: Lowry House. Rajab, Kalim. 2017. A Man of Africa: The Political Thought of Harry Oppenheimer. Cape Town: Zebra Press. Simons, Jack, and Ray Simons. 1969. Class and Colour in South Africa. Harmondsworth: Penguin. Suckling, John. 1977. ‘Factors affecting the promotion and payment of black labour in South Africa’. Seminar paper, October, Centre for Southern African Studies, York, England. Terreblanche, Sampie. 2012. A History of Inequality in South Africa, 1652–​ 2002. Pietermaritzburg: UKZN Press. Wassenaar, Andries. 1977. The Assault on Private Enterprise. Cape Town: Tafelberg. Wolpe, Harold. 1972. ‘Capitalism and cheap labour power’, Economy and Society, 1(4):425–​456  .

Chapter 4

P olitics and E c onomi c P olicymaking i n S ou t h Africa sinc e   19 94 Alan Hirsch, Brian Levy, and Musa Nxele

4.1  Historical Introduction For a little over two-​thirds of the twentieth century, South Africa was a profitable location for investment in repression. Government-​enforced exploitation of Black labour kept wages low in the resource sectors, centrally in the extraction of gold. From the late 1960s, as the gold fields reached and passed peak production, the integrated system of racial repression and labour exploitation began to break down. But the rigidities of the political system inhibited the kinds of reforms that could continue to produce large surpluses (Gelb 1991). The growing incompatibility of the economic and political systems and the inevitable, resilient struggle for freedom of Black South Africans led to a series of crises that culminated in a political transition. This transition, coinciding not by chance with the end of the Cold War, culminated in the political settlement enshrined in South Africa’s 1994 elections and its democratic constitution agreed in 1996.1 Struggles for power have continued through the short history of the young democracy. In this chapter we consider the formulation, adoption, and implementation of economic policies of the new Republic within the evolving political economy of post-​ 1994 South Africa and observe how the allocation of political and economic power has influenced economic policy.

1  The end of the Cold War, through weakening the remaining ‘Western’ support for South Africa as a ‘bastion against communism’, put pressure on F. W. de Klerk and the National Party to engage with the African National Congress (ANC), as well as putting pressure on the ANC to negotiate (Guelke 1996) .

Politics and Economic Policymaking in South Africa since 1994    67

4.2  Key Players in South African Politics South Africa is a complex society. The schema presented here is a simplification of the social actors around 1994.

4.2.1 Big Business/​Business The economy was dominated by a few very large firms owned mainly by white South Africans. ‘The corporations that grew to dominate the South African economy were formed during colonialism and apartheid. They grew around a core of finance, mining and minerals-related activities. By the 1980s, four diversified conglomerates and two financial companies dominated ownership and control over most of the economy’ (Seeraj Mohammed 2019). During the 1980s, conglomerates absorbed subsidiaries of multinationals which were divesting under political pressure to observe trade and investment sanctions. The conglomerates had few investment outlets for their profits which were penned into South Africa due to capital controls, and their investment targets were fire-​sale bargains. By 1992, the five largest firms owned 85.7 per cent of the market capitalization of the Johannesburg Stock Exchange (JSE). These firms, and similar ones, were notorious for their predatory and collusive behaviour and form part of the reason for the weak mittelstand (medium-​size industrial companies) and the relative absence of dynamic SMEs in South Africa (Lewis 1995; Kaplinsky and Manning 1998). Formal business organization has tended to be fragmented, but relatively briefly during the mid-​2000s business united in the peak organization Business Unity South Africa.

4.2.2 Black Business Almost until 1994, Black South Africans were hugely restricted by law and were blocked from owning large businesses or accumulating wealth. Though most severe for African South Africans, Black South Africans, those classified coloured and Indian were also affected. All Black South Africans were restricted to owning business and property in their ‘own group areas’, which excluded white-​owned central business districts, suburbs, and shopping zones. Nor were African South Africans allowed to own shares in public companies. Even in their ‘own’ areas African South Africans were severely restricted regarding the number and size of establishments. Few Black South Africans could become prominent businesspeople, and when they did, it was through front companies with white ‘paper owners’. In 1994, African Black-​owned businesses were organized into

68    Alan Hirsch, Brian Levy, and Musa Nxele two main associations, National African Federated Chamber of Commerce and Industry (NAFCOC) and Foundation for African Business and Consumer Services (FABCOS), and among other pressure groups they lobbied for access to economic power.

4.2.3 Organized  Labour The largest formation in the organized working class was the mostly blue-​collar Congress of South African Trade Unions (COSATU), the largest union federation with a membership of about 1.3 million in 1993. The majority of COSATU members were African. COSATU, forged during two decades of struggle for the freedom to associate and negotiate, was ANC-​aligned, with powerful constituent industrial unions. A much smaller blue-​collar union federation was the National Council of Trade Unions (NACTU), which had been associated with the Black consciousness movement. Many white-​collar workers belonged to centrist groupings. In 1997 they formed the unaffiliated Federation of Unions of South Africa (FEDUSA) with 515,000 members including many white workers.

4.2.4 Black Middle Class The Black middle class consisted of teachers, nurses, priests, police, shopkeepers, ‘taxi’ owners, and a few professionals. Only 12.2 per cent of Black Africans in 1995 had incomes above the mean national income, of whom less than a quarter were in the top income decile. Twenty-one per cent of those in the highest decile were African and 71 per cent were white (Levy, Hirsch, and Woolard 2015). In some definitions the middle class consists of white-​collar workers, professionals, and managers, in others it includes greater parts of the organized working class. What is most striking about class formation in South Africa is the income cliff within the third decile (Levy, Hirsch, and Woolard 2015). One way of describing the cliff is that this is the division between ‘insiders’ and ‘outsiders’. Elsewhere we have defined the Black middle class as being people with more than the mean income. If we exclude the ‘rich’, the top decile of income, there were three million Black African people in the middle class in 1995 out of 31 million Africans. In the top decile there were nearly 900,000 Africans—​2.7 per cent of the total Black African population (Levy, Hirsch, Naidoo, and Nxele 2020).

4.2.5 The Precariat A very large segment of the South African population can be described as the precariat—​those in precarious forms of employment without long-​term contracts or benefits. They include construction labourers, domestic workers, some service-​sector

Politics and Economic Policymaking in South Africa since 1994    69 workers, low-​level private security workers, some farmworkers, and much of the informal sector.

4.2.6 The  Poor The remainder of the population were simply poor. These are households with no one in regular employment while some received remittances and/​or social grants. One analysis put this group as 50 per cent of the population in 1997, almost all of whom are Africans, and an unusual proportion of whom (by global benchmarks) were located in rural areas (Leibbrandt, Woolard, and Woolard 2007).

4.2.7   Whites In the 1996 population Census, 4,434,697 whites made up 10.9 per cent of the total population of 40.6 million. Average per capita income for whites was double that of Indians, 5 times that of coloureds and 7.5 times that of African South Africans. The national Gini coefficient was estimated at 0.69. While 62 per cent of Africans fell below a poverty line set at R250 in 1996 Rands, only 3 per cent of whites did. Not all whites were rich, but almost none were poor (Leibbrandt, Woolard, and Woolard 2007).

4.2.8 The  State In 1994, the ANC came into power in a Government of National Unity—​this was the negotiated form of government until the final constitution was adopted two years later. The ANC won the national vote and control of seven of the nine newly demarcated provinces which replaced four provinces and nine Bantustans in the old regime. Yet, the central state and the four legacy provinces still reflected the old regime in the composition of employment (managers and professionals were largely white) and in structure. Around 1.27 million people were employed in national and provincial government in 1995. Nine puppet Bantustan regimes, designed in part to build a pliant Black bureaucratic middle class, still had to be dismantled and absorbed as were many racially defined parallel structures such as in security services, health, and education. Employment in puppet authorities trebled between 1980 and 1994 to nearly 250,000 people (Botha 1995). There were several hundred thousand employed in the municipalities (Hassen and Altman 2010).

4.2.9 The Political Parties The ruling party in white-​controlled South Africa from 1948 to 1994 was the National Party (NP)—​the party that formalized segregation into the rigidly oppressive apartheid

70    Alan Hirsch, Brian Levy, and Musa Nxele system. The NP won 20.4 per cent of the vote in 1994. The main white opposition parties were the Freedom Front to the right of the National Party and the Democratic Party (DP) on its centre-​left—​together they won 4 per cent of the vote in 1994. More successful was the Inkatha Freedom Party which was based in the largest single ethnic group—​Zulus. The IFP won 10.5 per cent of the vote, also attracting conservative non-​ Zulus in the KwaZulu-​Natal province. The ANC, which won 62.6 per cent of the vote, had a clear majority in the national assembly and most of the provinces, but in terms of the interim constitution a government of national unity was to rule until 1999. The ANC included cabinet members from other parties even beyond 1999. The centrist DP never joined the cabinet, and the NP withdrew soon after the finalization of the constitution in 1996.

4.3  Mapping Out Strategies Pre-​1990 By the 1980s, apartheid was under attack by organized Black opposition groupings in the workplace and in communities, countered with violent repression and states of emergency. The apartheid state closed ranks in a ‘total strategy’ in response to a ‘total onslaught’ (Morris and Padyachee 1988). Economic policy shifted selectively towards neo-​ liberal economic policies pioneered in the United Kingdom and the United States in the 1980s and included some liberalizing labour market reforms. International economic relations remained protectionist and state controlled through tariffs, capital controls, and subsidies for large companies (Hirsch 2020). State-​owned enterprises (SOEs) and agricultural cooperatives were corporatized and directed towards privatization (van Rooyen et al. 1996). As the apartheid state weakened in the face of uprisings, dissent within the white elite, boycotts, and economic sanctions, the politics of negotiation began. In 1985, some business leaders met with the ANC and there were secret meetings between minister of justice Kobie Coetsee and Nelson Mandela (Evans 2020). Manoeuvring for position, big business built bridges to the ANC and began to advocate neo-​liberal lassez-​faire policies. The ANC issued two documents, ‘Constitutional Guidelines’ published in 1988 and the 1989 ‘Harare Declaration’, which began to position the ANC closer to the centre compared with its more radical economic stance during much of the period of exile (Hirsch 2005: ­chapter 3). In the late 1980s and early 1990s, several research groupings supporting the development of economic policy for the ANC emerged, including the COSATU-​linked Economic Trends and Industrial Strategy research groups, the London-​based Economic Research on South Africa (EROSA) and Centre for Research into Economics and Finance in Southern Africa (CREFSA), and the Macroeconomic Research Group or MERG (see Hirsch 2005 and Padayachee and van Niekerk 2019).

Politics and Economic Policymaking in South Africa since 1994    71

4.4  Economic Policy and the Political Settlement, 1990–​96 On 2 February 1990, the National Party government announced that the ANC and other illegal political organizations were unbanned, along with their leaders, some of whom were released from prison. To the consternation of big business, Mandela strongly advocated the nationalization of large companies, partly in response to NP moves towards the privatization of SOEs. Mandela successfully halted the privatization process by threatening to renationalize those privatized during the transition period. From 1990, the ANC Department of Economic Planning published several documents culminating in the 1992 policy document ‘Ready to Govern’ and the 1994 election manifesto the ‘Reconstruction and Development Programme’ (RDP). Combining growth with redistribution was the challenge, and various formulations were put forward within ANC policy circles including Asian-​style development funding in the MERG report (MERG 1993), Asian-​style industrial policy in the ‘Industrial Strategy Project Report’ (Joffe et al. 1994), Keynesian-​style demand-​led policies in the 1990 ‘Discussion Document on Economic Policy’, and conventional positions on trade and investment in the ‘Ready to Govern’ document. Nationalization receded from the agenda. Though the RDP election manifesto seemed to be a coherent document balancing macroeconomic caution with major reforms, the reality was that the ANC and its allies, the South African Communist Party and COSATU, did not agree on many aspects of economic policy and there was no agreed vision of what economic development would look like in South Africa (Natrass 1994; Hirsch 2005; and Padayachee and van Niekerk 2019). Big business mobilized to influence the direction of economic policy. ‘In late 1988, following a high-​level meeting between business leaders and representatives of the Mass Democratic Movement in Broederstroom, the Consultative Business Movement (CBM) was formed. The initial focus was on consultation and relationship building with key political players, such as the African National Congress (ANC) and Inkatha Freedom Party (IFP)’ (Eloff and Fourie 2006). Corporations undertook scenario exercises drawing in leadership from the ANC and other influential constituencies. Ten years of declining per capita income in South Africa between 1984 and 1993, the conservative messages coming from business, and the influence of the international financial community contributed to the ANC’s aversion to ambitious reforms. The ANC repeatedly voiced a commitment to a conservative macroeconomic policy, not wanting to repeat the errors of the Latin American debt crises of the 1980s and 1990s which it saw as risking national sovereignty (ANC 1990, 1992, 1994; Kahn, Walton, and Senhadji 1992). Economic policy during the 1990–​94 transition was hampered by government’s weak legitimacy and by the continuation of an economic slump. The need for transitional decision-​making gave rise to the Transitional Executive Council (comprising government and ANC) and the National Economic Forum (government, business, labour, and

72    Alan Hirsch, Brian Levy, and Musa Nxele informally ANC). Some issues addressed were participation in the Uruguay Round of the GATT, corruption in the customs services, and obtaining short-​term finance from the International Monetary Fund. Signalling that the new government would be cautious in its economic policies, Derek Keys, the finance minister in the outgoing NP government was reappointed by new President Nelson Mandela. Months later Keys left for a top position in London and was replaced by a white banker, Chris Liebenberg. Mandela made this decision because ‘the boys from the stock exchange and elsewhere seem to be very jittery’. He assured journalists there had been ‘no change whatsoever in government policy’ (Hirsch 2005: 67). When Liebenberg was replaced with an ANC minister of finance, Trevor Manuel, the Reconstruction and Development Ministry and Office were closed, the rationale being that now that the ANC directed the budget, separate offices were not required. Some saw this as a conservative step, though in practice the RDP Office achieved very little (Marais 1998: 191). While the ANC and its allies researched economic policy during the transition, the new government did not have a clear economic development policy. President Mandela was preoccupied with peace, stability, and reconciliation in the face of a range of real and present dangers. The national army and police were as yet untransformed, and there had been violent dissent from conservative whites and the IFP around the elections. In an era of violent conflict in Rwanda, Northern Ireland, and former Yugoslavia, the key task was to preserve peace (Mandela and Langa 2017).

4.5  From GEAR to Polokwane, 1996–​2007 The apartheid system was founded on conflict, constructed as a zero-​sum game. Until 1972, the wages of Black African mineworkers, the core labour force, failed to rise in real terms—​they generally remained below 1911 levels. The ratio of white mineworkers’ wages to Black rose from 11.7 in 1911 to 20.9 in 1971 (Wilson 1972; Lipton 1986). In 1995, 71 per cent of the richest 10 per cent of South Africans were white. The average per capita income of Black South Africans was R5,144.68 while whites had an average per capita income of R 35,907.41. The bulk of the ANC’s constituency were the poor, the dispossessed, and the oppressed. The remainder were organized workers and much of the relatively small Black middle class. A successful economic programme needed to address all their needs. This meant jobs or relief for the poor. Unemployment had risen to 20 per cent by 1994 and was rising rapidly as employment in gold-​mining and other primary sectors declined in the absence of growth in other sectors. The unemployment rate was 29 per cent for Africans and 4 per cent for whites (Statistics South Africa 1998: 3–​4). COSATU unions were a powerful force in the ANC alliance and expected significant improvements in wages and workplace conditions. Economic success for the ANC also meant a considerable expansion and improvement for the Black middle class.

Politics and Economic Policymaking in South Africa since 1994    73 The ANC feared antagonizing white South Africans because of political risks, but also because white South Africans possessed much of South Africa’s wealth and skills, which were critical for economic recovery and growth. When Mandela visited Washington, DC, in 1994, accompanied by a group of South African business leaders who had been prominent in the apartheid era, the UK ambassador to Washington Robin Renwick congratulated him on his magnanimity. ‘Mandela replied, with understandable bitterness, that he forgot nothing, nor did he forgive, but that he needed them now’ (Renwick 2015).

4.5.1 Building Structures of Cooperation One of the first steps of the new government was the establishment of the National Economic Development and Labour Council (NEDLAC). This resembled the transitional National Economic Forum, but it also represented the idea that, unlike the apartheid system, benefiting from the economy was not a zero-​sum game—​rents would not be monopolized by one interest group. NEDLAC embodied the idea that a successful economy could be built through cooperation rather than conflict. The main purpose of NEDLAC was to act as a forum at which economic policies were agreed to by the major social partners—​government, organized labour, and business, mainly big business. A community constituency was also present but was the least representative of the delegations. NEDLAC is required by law to act as the forum at which labour law reform is negotiated, but other social and economic policies and laws were also to be discussed at NEDLAC. The NEDLAC agreements on labour law (and other laws which are voluntarily negotiated through NEDLAC) must go to Parliament for further debate but need to return to NEDLAC if any of the tripartite agreements are challenged in Parliament. NEDLAC has four chambers: the labour chamber, the trade and industry chamber, the macroeconomic chamber, and the development chamber. The labour chamber negotiated four reforming labour laws during Mandela’s presidency. The composition of the bargaining parties was critical to the outcome of the negotiations. The strongest elements in the union delegation were the major industrial unions of COSATU, the dominant federation, closely aligned to the ANC. The leading negotiators on the side of business were drawn from business federations and large corporations which had the resources to hire industrial relations managers. The government delegation to the labour chamber was composed largely of former union officials. The result of NEDLAC negotiations over labour regulation reflected the point of agreement between big business and big labour. Small firms and competitive industries were thinly represented, and their voices faint. The outcome was progressive legislation for industrial relations and employment conditions which large profitable firms could accommodate. Exemptions for small firms were marginal, and only for very small firms. No formal allowance was made for labour-​intensive firms in regions of high unemployment; instead it became accepted practice that sectoral agreements

74    Alan Hirsch, Brian Levy, and Musa Nxele between unions and employers were extended by the minister of labour to ‘non-​ parties’. Relatively high wages and high standards of employment conditions were also required of less competitive firms with lower unionization, often outside of the major urban centres. This put poorly skilled and poorly educated workers in rural towns at a disadvantage and probably discouraged labour-​intensive investments (Nattrass and Seekings 2014). While NEDLAC was implementing labour laws, reflecting an orientation towards what COSATU called a ‘high-​wage, low-​cost economy’ (Masiya 2014), trade and macroeconomic policy headed in another direction. Early salary negotiations by government had led to expensive commitments in a still mediocre economic environment. Attempts at restructuring government finances in the context of false rumours about the health of President Mandela, and an adverse reaction in financial circles to the appointment of ANC leader Trevor Manuel as finance minister, led in mid-​1996 to the introduction of the Growth, Employment and Redistribution strategy (GEAR). It was a home-​grown structural adjustment programme—​most notably an ambitious fiscal deficit-​reduction target, trade tariff reduction, privatization, relaxation of foreign exchange controls, moderation of public-sector wage growth, and an anti-​inflationary monetary policy. GEAR also had stimulatory elements such as a commitment to a competitive exchange rate, tax holidays for investment, an expansion of infrastructure investment, and a levy system to fund employee training, but the overwhelming impact came from the structural adjustment elements. Tension between a competitive exchange rate policy and an inflation-​wary stance was resolved, de facto, in favour of the latter. Trade liberalization went beyond the 1994 Marrakesh Agreement of the GATT/​WTO in the belief that getting the prices right would encourage investment in export-​oriented labour-​ intensive businesses. While some analytical work found positive dynamic outcomes from the trade reform (Jonsson and Subramanian 2001), labour-​intensive industries such as garments, footwear, and leather were decimated by the medium-​term effects of the reforms despite subsidies and incentives designed to cushion the blow (Erten, Leight, and Tregenna 2019). The Treasury had little appetite to support active measures to build global competitiveness. A further contributing factor to industrial decline was that the South African Reserve Bank, pre ​occupied with the fast-​growing money supply, hiked rates sharply to defend the currency against depreciation to protect itself from exposure in forward currency markets after ill-​advised currency swaps intended to reduce the rate of depreciation (Hirsch 2005: 103–​5). Reserve Bank reaction to a currency collapse at the end of 2001 again led to excessively high interest rates and an overvalued currency which, unfortunately, coincided with China’s accession to the WTO to the severe detriment of the non-​ traditional tradable sector of the economy. Competition policy reform was embodied in the Competition Act of 1998 which established a Competition Commission and Tribunal with sharper teeth than the previous regime. Still, it was limited by the inhibition of the competition authorities to act against prohibited practices outside of a proposed merger. In mergers, workers’ rights were protected. Negotiated in NEDLAC, it may be reductionist, but outcomes defending

Politics and Economic Policymaking in South Africa since 1994    75 the interests of big labour and big business could be interpreted as a deal at the expense of smaller, competitive companies. Tight macroeconomic policies provided a greater share of government revenue to fund significant social improvements in the late 1990s and the 2000s. Mass housing programmes were an initial target for government. Investments led to much improved social infrastructure services, notably water supply and electricity access, while education and health services broadened rapidly. Pension payments (which during apartheid years had been generous for whites) were universalized and equalized. A Child Support Grant—​the main form of support for indigent families—​was introduced in 1998 and already reached 3 million children by 2003. By 2010, after the maximum eligibility age was raised several times, more than 10 million of the poorest children in a country with a total population of less than 50 million at that time were receiving child support grants. Social security transfers rose to 3.5 per cent of GDP by 2009, representing a concerted attempt to ameliorate severe poverty in the absence of comprehensive unemployment insurance (Levy, Hirsch, Nxele, and Naidoo 2020: 14–​15). A further development after the GEAR macroeconomic targets were met was a considerable improvement in growth performance, depicted in Table 4.1. Government and private-sector investment rates picked up with consequent improvements in growth, per capita income, and employment. Unemployment fell from over 32 per cent in 2003 to under 22 per cent only five years later, in 2008.

4.5.2 The Politics of Growth Pritchett and Werker (2012) introduced a conceptual framework, structured in part around ‘the rents space’, dividing firms into four different types according to a two-​ dimensional quadrant. This is depicted in Figure 4.1. The two dimensions are exporters

Table 4.1: South African real annual economic growth, 1990–​2019 Year

Growth rate1

Growth rate per capita income2

Private-sector fixed investment3

1990–​93

–​0.56%

–​2.97%

13.81%

1994–​2000

2.90%

1.06%

13.89%

2001–​04

3.48%

2.18%

13.20%

2005–​08

4.86%

3.52%

17.00%

2009–​12

1.75%

0.25%

16.86%

2013–​19

1.18%

–​0.29%

16.21%

Source: World Bank national accounts data and OECD National Accounts data files (2020). Notes: 1 Annual percentage growth rate of GDP at market prices based on constant local currency. 2 GDP per capita growth (annual %). 3 Gross fixed capital formation, private sector (% of GDP).

76    Alan Hirsch, Brian Levy, and Musa Nxele High-rent RENTIERS Natural resource exporters, Exportagricultural concession oriented exporters POWERBROKERS Legislative monopolies or Domestic oligopolies, natural monopolies market or oligopolies, government services

Competitive MAGICIANS Manufacturing and service exporters, other agricultural exporters WORKHORSES Importers, traders, retailers, subsistence farmers, local manufacturers, producers of nontradeables

Figure 4.1  The market matrix: alignment of elites into rentiers, powerbrokers, magicians, and workhorses Source: Pritchard and Werker (2012 53).

versus domestic market-oriented, and firms dependent on regulatory rents versus those relying on success in market competition. ‘Rentiers’ are export oriented but reliant on regulatory rents—​typically they are natural resource extractors, though they could export other products into world commodity markets. They are reliant on governments to give them the right to sell resources belonging to the state and the people (Pritchett, Sen, and Werker 2018: 22). ‘Magicians’ are exporters in competitive markets—​ they rely on innovation, branding, or productivity and can be in manufactured goods or services markets. ‘Powerbrokers’ are domestic-​market suppliers relying on favourable regulation and can include sectors like telecommunications, power, or petroleum-​refining. Finally, ‘workhorses’ are oriented towards the domestic market and operate in a competitive environment. In the hothouse environment of sanctions and economic isolation, the late apartheid state had strongly favoured the rentiers and the powerbrokers, while the workhorses operated behind protective barriers and magicians were few and far between. This placed firms reliant on regulatory favour in a powerful position during and after the transition to a more open economy. Powerbrokers, which in South Africa include the SOEs and the financial sector (see Chapter 47 on financialization by Karwowski), tend to favour red tape ‘where the impetus may come from strong politicians or bureaucrats creating rents for themselves or their cronies, or from strong businesspeople buying off politicians and bureaucrats to entrench their market position. The problem is that they advocate for policies that are detrimental to firms in the other quadrants’ (Pritchett and Werker 2012: 56). Competition policy, innovation policy, and investment incentives were founded on the desire for more competitive workhorses and magicians. But they were not sufficiently well designed or given adequate resources to tilt the balance away from the rentiers and powerbrokers. Moreover, the commitment of key actors to these reforms had shallow roots (Hirsch and Levy 2018: 3–​4).

Politics and Economic Policymaking in South Africa since 1994    77 The relative strength of the rentiers and powerbrokers relative to the magicians and workhorses (reflected in the business delegation to NEDLAC) provides some of the explanation for the peculiar combination of policies in the new democratic state.2 The lack of an alternative economic development strategy was exemplified in unresolved competition and contestation within government and within the ANC Alliance. During 1996, three separate, overlapping, and conflicting economic policies were put forward by different parts of government. In addition to GEAR which was driven by the Treasury, the Reconstruction and Development Ministry produced a draft National Growth and Development Strategy, and the Ministry of Labour published the report of the Presidential Labour Market Commission. The RDP office’s draft never saw the light of day, the Labour Market Commission saw the less ambitious aspects of its proposals in the formation of NEDLAC, while only the structural adjustment elements of the GEAR strategy were implemented with vigour. As NEDLAC narrowed into a forum mainly to negotiate labour law, President Mbeki sought to engage major interest groups in another structure—​what the Presidency called ‘working groups’. President Mbeki and his immediate successor Motlanthe met regularly with different working groups representing big business, Black business, labour, religious leaders, and university leaders, and sometimes he met with combinations of these groups. Mbeki also established an International Investment Advisory Council and an International Information and Communications Technology Council made up of weighty global players in their fields and a local presidential economic advisory panel to provide him and his ministers with advice and feedback. But this helped relatively little in achieving an economic development programme that committed the whole of cabinet to a coherent set of goals and actions. The decline of the manufacturing sector’s contribution to employment and output is symptomatic of the weakness of firms which rely on market competition (Roberts 2014). Within export-​oriented manufacturing only rentier companies thrived—​good examples are the aluminium smelters that relied on a cheap electricity deal with the state-​owned electrical power supplier Eskom, and the automobile sector, which relied on a hugely subsidized Motor Industry Development Program. The high margins earned by oligopolistic upstream suppliers and manufacturers in South Africa are widely documented. The historical pattern has been for rent-​distributing industries to share their high margins with unionized workers (Worgetter 2014; Mahajan 2014; Fedderke 2014). For significant periods, the rate of investment in R&D was relatively high in South Africa, an indicator normally associated with a dynamic, innovating economy. However, research was not strongly steered towards industrial innovation (Kaplan 2014). During the economic boom of 2004–​08 when the growth rate was around 5 per cent per annum and about two million jobs were created, few of them were created in the

2 

The NEDLAC Executive Council is dominated by big labour and big business—​mining and finance in particular. See e.g. NEDLAC (2007: 19).

78    Alan Hirsch, Brian Levy, and Musa Nxele competitive export-​oriented sector or in the workhorse sector. Though capacity utilization remained at record levels for an extended period of time, investment in new production was disappointingly low. Foreign investment into South Africa has tended to flow as portfolio investment into the high-​yielding rentiers and powerbrokers or as the purchase of cash cows in these sectors. Funds have also flowed to finance government debt, but foreign investment in workhorses and magicians is rare in South Africa (Black 2014).

4.5.3 BEE—​A Strategy for Elite Transformation The effort to shift power within the elite and to support growth in competitive sectors resulted in the early shape of the Black Economic Empowerment programme. The government had to contend with an incumbent (white) business class which recognized the necessity of altering the colour of ownership and control but sought to do so on its own terms. At least as challenging was the task of navigating between disparate ideological factions within the ANC. Nor were the political pressures only ideological. There were also new waves of political leaders, with strong incentives to use their new ​found political power over state resources to create business opportunities for their allies. The initial response of policymakers and political leaders was to try and contain the struggle over ownership within a broader discourse around economic transformation. They sought ways of addressing the seemingly contradictory interests of the various protagonists without undercutting efforts to reincorporate South Africa into the global economy and to foster economic growth. Efforts focused on new firm development, affirmative action in the workplace, and measures to support the enhanced education of Black people who had the potential to be professionals, managers, and business leaders. Indeed, the term ‘empowerment’ was not included in ANC economic policy documents before the 50th National Conference in Mafikeng in December 1997 where the ANC called for a National Empowerment Policy that will focus on those who have been historically disadvantaged and particularly black people, women, youth and the disabled and rural communities. The empowerment process must constitute part of a more radical and profound change in social relations. Changing ownership and workplace relations are part of this wider process of empowerment. [. . .] The ANC government should ensure the implementation of a vigorous affirmative procurement policy which will ensure that government and parastatals facilitate awarding of tenders to our people through approved mechanisms.

This was still largely a growth-​oriented approach, focusing on new firm development, affirmative action, and the skilling of Black South Africans, though voluntary empowerment deals in the 1990s foreshadowed patterns accentuated under Mbeki.

Politics and Economic Policymaking in South Africa since 1994    79 The BEE process accelerated, spurred in part by a 1998 economic contraction and stock market decline crisis which left many Black beneficiaries of initial rounds of deals without the cash flow needed to meet their BEE-​related financial obligations and led to a decline of Black ownership on the stock exchange from 8 per cent to less than 4 per cent. BEE became increasingly central to the objectives of the ANC, and its protagonists and beneficiaries became increasingly influential within the party. A central pillar of this accelerated response was an effort to establish a system of rules within which to manage BEE expectations and pressures. Milestones in this effort included:3 • The establishment in 1998, at the urging of the Black Management Forum, of a BEE Commission, with the task (as President Mbeki put it) ‘to answer the question—​how do we promote the formation of a Black bourgeoisie which will itself be committed and contribute to Black economic empowerment?’ Chair of the commission was Cyril Ramaphosa, former unionist, later to be president. • The passage in 2002 of a Mineral and Petroleum Resources Development Act (MPRDA) which required owners of ‘old-​order’ mineral rights to convert to ‘new-​ order’ rights subject to conditions of meeting BEE imperatives. • Finalization of the BEE Commission’s report and release of an official strategy document in 2003, followed by promulgation in 2004 of the Broad-​Based Black Economic Empowerment (BBBEE) Act. The 2004 BBBEE Act went beyond a narrow pre o ​ ccupation with ownership. Its implementing framework laid out a balanced scorecard which set targets for ownership and management control, for skills development, and support for subcontractors and other SME(small and medium enterprises), as well as socio-​economic development more broadly. The expectation was that each economic sector would detail a quasi-​voluntary sectoral ‘charter’ aligned with the framework laid out in the legislation. The only enforcement mechanism was that to have access to public procurement opportunities, companies needed to be BEE compliant. The combination of voluntarism and rule-​focused BEE remained largely consistent with the continuation of strong formal institutions. But there was also a shadowy side to the new government’s use of economic power. There was the desire of top political leadership to ensure that trusted allies would be the beneficiaries of ongoing efforts to change the colour of business. There was a need on the part of the ANC to finance its activities. Inevitably, there also was an opportunistic desire on the part of some influential insiders to use their political influence as a platform for private economic enrichment and empire building. Examples of how this played out in practice include:

3 

For details, see Hirsch (2005), Tangri and Southall (2008), Mpanza (2016), Plaut and Holden (2012), and Harvey (2015).

80    Alan Hirsch, Brian Levy, and Musa Nxele • Arms deals in the 1990s: after a defence review was completed, arms RFPs were issued for a range of conventional naval and air force equipment. The exercise was muddied by several kickbacks and side-​deals under the cover of helping the ANC and supporting empowerment, including one which implicated the then deputy-​ president, Jacob Zuma. • A series of murky deals in the early 2000s involving the newly formed SOE PetroSA, the United Nations ‘oil for food’ programme in Iraq, and a little-​known BEE consultancy company—​all seemingly linked to the ANC’s efforts to raise finance for the 2004 election campaign. • A multi-​billion-​Rand procurement contract, awarded in 2004 by the state-​owned electricity company, Eskom, through an obscure process, to a joint venture between Hitachi Power Systems and Chancellor House (the investment arm of the ANC) for the boilers for the massive coal-​fired Medupi electricity-​generating plant. • Also in 2004, the purchase by the BEE ‘elephant consortium’ (headed by the former director general of the Ministry of Telecommunications, and dominated by close allies of then-​president Thabo Mbeki) of a controlling shareholding of South Africa’s leading fixed-​line telecommunications operator, Telkom, with subsidized support from the state-​owned Public Investment Corporation. • A series of 2007–​09 deals by Vodacom (one of South Africa’s two largest mobile telecommunications operators, partly owned by Telkom, but with control in the hands of UK-​based Vodaphone ever since its establishment in 1993) which resulted in: i. A massive windfall to the ‘elephant consortium’ (as part owner of Telkom); ii. Establishment of a new BEE consortium to partner with Vodacom, to be headed by Bulelani Ngcuka, a close ally of President Mbeki and, earlier, director of public prosecutions, a lead actor in efforts to prosecute Jacob Zuma for corruption; iii. After Thabo Mbeki lost an ANC leadership struggle to Jacob Zuma, the setting aside of the Ngcuka-​headed consortium in favour of a more politically neutral BEE arrangement with Royal Bafokeng Holdings. Notwithstanding the multifaceted character of BEE and some signs of its corrosive effect on formal institutions, the balance between rules and deals was still good enough to provide an increasingly robust platform for the economy. Up to 2008 or so, growth accelerated and extreme poverty declined. The inclusion of Black South Africans into the capitalist class began through an ad hoc set of private initiatives in the early 1990s and transformed into a rule-​based set of programmes during the Mbeki presidency, in the early 2000s. Many of the transactions of this era in practice straddled the boundary between rules-​ based and more personalized deal-​making. While it might be argued that the seeds for a more expansive series of deals and manoeuvres to enrich Black South Africans (and others) were planted during the Mandela and Mbeki presidencies, prior to the assumption of power by Jacob

Politics and Economic Policymaking in South Africa since 1994    81 Zuma, legitimate order remained dominant in a system where rules outweighed deals. Government effectiveness, regulatory quality, and control of corruption all declined considerably from 2009 onwards (World Bank 2020). The period 1996 to 2008 was as volatile as South Africa’s open economy tends to be, but there was more growth and, in the latter part of the period, faster growth than South Africa had experienced for nearly fifty years. However, the ruling elite had failed to shift the economy from one mainly dominated by large and powerful corporations which were able to capture regulatory rents. Ownership had altered little (except for greater foreign ownership, mainly through the stock exchange) and, if anything, the gap between the white business elite and the Black governing elite had grown. Corporations and investors sought to immunize themselves from the vagaries of policies rather than joining hands with what the ANC called the ‘developmental state’. Inequality remained extremely high, by some measures even higher than before. Social transfers, social services, and social infrastructure investment had done relatively little to alter the fundamental circumstances of the poor. Despite higher growth, higher employment, and higher wages, the populist surge in the ANC, which led to the ousting of the centrist Mbeki-​led grouping, is understandable in part as a failure of inclusion. The ANC (2004) noted that ‘many, many South Africans do not have jobs or decent self-​employment; poverty is a reality for millions, while many cannot get credit to start or improve their own businesses’. Pressure also arose for the inclusion of a broader segment of emerging Black elite into the corridors of economic power.

4.6  Crisis and Decline, 2007–​17 In 2007, well into the first real economic growth spurt in nearly three decades, the ANC and its allies turned away from Thabo Mbeki and his leadership team. Mbeki had alienated too many elements of the ANC alliance. Organized labour and the Communist Party blamed Mbeki for pushing through the neo-​liberal GEAR strategy in 1996 and having remained unsympathetic to their causes. Some in the ANC saw Mbeki as an arrogant leader. Former deputy-​president Jacob Zuma, who Mbeki had ousted after he was implicated in a corruption trial where he was not among the accused, mobilized Zulu nationalism and rural populism. The discontented unified behind Zuma (then recently acquitted of rape in a marginal decision) and defeated the Mbeki faction at the elective conference in Polokwane in December 2007. After a placeholder president, Kgalema Motlanthe, was inserted to complete Mbeki’s term, Zuma took office following the general elections in April 2009. Zuma won nearly 66 per cent of the vote, which represented a drop from Mbeki’s nearly 70 per cent majority in 2004. At first Zuma seemed to be carrying on the practices of the ANC government that preceded him; in some respects, he improved them. Zuma’s policy to combat HIV/​AIDS was far superior to his predecessor’s and represented a great success. Zuma also took a

82    Alan Hirsch, Brian Levy, and Musa Nxele tough stance against Moammar Ghaddafi’s imperial ambitions in the African Union, and later succeeded in getting a South African leader elected chair of the African Union. A series of economic programmes were announced in the latter part of the first decade of the twenty-​first century. First, the Accelerated and Shared Growth Initiative (ASGISA) was announced by the Mbeki presidency in 2006. Its main objective was to expand the growth capacity of South Africa through investment in human capital (especially artisans and engineers) and to improve the performance of the economic infrastructure. After Zuma came to power in 2009 ASGISA was shelved and replaced with two parallel initiatives—​the New Growth Path led by minister of economic development Ebrahim Patel, and the National Development Plan led by planning minister Trevor Manuel who had been finance minister under the previous three presidents. In addition, the minister of trade and industry had a national industrial strategy framework and the finance minister had priorities of his own. As in previous administrations, the president failed to knit these initiatives together or to choose among them and each initiative continued to represent its own constituencies in a cacophony of policies.

4.6.1 The Impact of the Global Financial Crisis on South Africa Shortly before Zuma became president, the financial crisis in the United States morphed into a global finance crisis. While 2008 was one of moderate GDP growth at 3.1 per cent, the following year saw South Africa crash into recession with a 2 per cent decline in GDP. In the course of 2009 nearly half of the two million jobs created between 2003 and 2008 were lost (Ravinder and Msomi 2017). South Africa experienced a combination of the First World credit crunch (as its financial markets are deep as developing countries go) and a Third World trade crunch as exports and trade finance evaporated. The credit crunch followed an excess of lending that preceded the coming-​into-​force of tighter lending laws in 2008. While few financial institutions were dangerously stressed, the tightening up of credit markets had widespread effects. Added to this was an electricity-​generation bottleneck which began early in 2008. All this does not sufficiently account for the downward divergence of South African economic growth after 2014 when compared with earlier trends (see Figure 4.2). The country remained deeply unequal along racial and economic lines. Table 4.2 shows that twenty years of ANC government had not yet addressed the challenge of inclusion of the poor and, in relative terms, inclusion into the middle class and the elite. One partial measure of the incorporation of Black South Africans into the elite is market capitalization in the Johannesburg stock exchange (JSE) commanded by ‘Black-​ controlled’ companies.4 Black-​controlled companies declined from 9.6 per cent of JSE 4 

Black-​controlled companies are defined by an empowerment holding exceeding 26 per cent with no other dominant shareholder.

Politics and Economic Policymaking in South Africa since 1994    83 8.00 6.00 4.00 2.00 0.00

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

2018

2019

–2.00 –4.00 Upper middle income

South Africa

Middle income

Figure 4.2  South Africa’s GDP growth relative to upper-​middle-​and middle-​income countries 2008–​19 Source: World Bank national accounts data and OECD National Accounts data files (https://​data.worldbank.org/​ indicator/​NY.GDP.MKTP.KD.ZG?locations=XT).

capitalization in 1998, at the peak of the first wave of empowerment, to 5.5 per cent in 2007 and to 1.2 per cent in 2019, representing a poor rate of inclusion. Foreign ownership through the JSE rose from 4 per cent in 1998 to a peak of 42 per cent in 2016, declining to 22 per cent in 2019 (Who Owns Whom 2020: 46). The Zuma challenge to Mbeki exploited the failure of inclusion. Following Zuma’s ascension to leadership, another populist element emerged in the ANC which challenged Zuma on his left. After expulsion from the ANC in 2012, Julius Malema formed the Economic Freedom Fighters political party. Left-​wing populism rooted in the realities of inequality acted as a cover for corruption.

Table 4.2: South Africa’s 2014 population distribution, by ethnicity and class Total %

African %

Other Black %

White %

Chronic poor

49.5

46.9

2.5

0

Transient poor

12

10

2

0.1

Vulnerable

15

13

2

0

Middle class

20

9.5

4

6.5

Elite

3.5

0.6

0.5

2.4

% population

100

80

11

9

Source: Schotte, Zizzamia, and Leibbrandt (2017).

84    Alan Hirsch, Brian Levy, and Musa Nxele At the 2007 conference in Polokwane where Zuma was elected president of the ANC, a resolution on economic transformation that emphasized the role of SOEs also contained the following paragraph, which gave a sense of the orientation of the new ANC leadership: The developmental state should maintain its strategic role in shaping the key sectors of the economy, including the mineral and energy complex and the national transport and logistics system. Whilst the forms of state interventions would differ, the over-​riding objective would be to intervene strategically in these sectors to drive the growth, development and transformation of the structure of our economy.

An era had begun where interventions in the name of economic transformation became the staple of patronage within the political party machine called the ANC. Corruption and abuse of state resources did not begin in the Zuma era, as earlier noted. Initially such deals were rationalized as means of raising money for the ANC, but individuals came to benefit greatly too. SOEs became the locus for many such deals after 2010 when Malusi Gigaba was appointed minister for public enterprises by Jacob Zuma. Tracking the governance and operation of South Africa’s SOEs from 2010 to 2017 reveals the ‘formula’ used to try and repurpose them to serve the overlapping agendas of ‘radical economic transformation’ and private enrichment. This formula comprised the following steps:

• • • •

Put in place a supportive minister of public enterprises Reshuffle the SOEs’ boards of directors Appoint a compliant/​supportive chief executive Override established rules governing decision-​making, especially those concerning procurement—​including via direct intervention by the board in operations, as necessary for the over ​rides.

SOEs are responsible for a large part of government procurement; in 2010/​2011 when state capture took off, SOEs were responsible for nearly 25 per cent of government procurement—​the two largest SOEs, Eskom and Transnet, were each responsible for more than 8 per cent of government procurement (Chipkin, Swilling et al. 2018: 113). Overriding the rules requires not only capture of the SOEs but also compliant institutions beyond the SOEs themselves. This helps to explain why efforts to deploy loyal functionaries in both the justice system and the National Treasury were part of the ‘state capture’ agenda. Many of the actions supporting state capture were justified in terms of economic transformation in response to ‘white monopoly capitalism’. This rationale was promoted by a secret propaganda campaign funded by the Gupta family, facilitators and beneficiaries of corruption (Levy, Hirsch, Naidoo, and Nxele 2020: 35–​6). Nine years of rule by the Zuma regime did not completely undermine state institutions, but many, from small municipalities to important policy departments in government, were critically damaged through the appointment of compliant/​complicit

Politics and Economic Policymaking in South Africa since 1994    85 servants of the patronage machine. Urgent decisions such as digital migration, electromagnetic spectrum allocation, and energy policy were held hostage to the needs of the patronage machine. Money was wasted on overpriced and misdirected contracts, and policy processes stalled. Government finances were damaged and global credit ratings began to decline. Business confidence seeped away, and the period after 2014 saw six successive years of negative per capita growth (Bureau for Economic Research 2020). Figure 4.2 shows the exceptional deterioration of South Africa’s economic performance since 2014. South Africa’s growth path diverged negatively from comparator countries in this period. Relative and absolute deterioration continued post-​Zuma.

4.7  Can South Africa Start Over? By the time of its December 2017 elective conference, enough delegates believed that Zuma’s tarnished leadership threatened the electability of the ANC. The municipal elections of 2016 showed an 8 per cent decline in the popular vote for the ANC, to 53 per cent, and many ANC public representatives feared losing power and access to patronage machines. Zuma’s candidate for leadership lost by a small margin, handing the presidency of the ANC to Cyril Ramaphosa’s reform ticket. Nevertheless, three of the ‘top six’ elected were Zuma supporters and Ramaphosa’s margin of leadership was very small—​this was also reflected in the divided composition of his cabinet, even after an initial reshuffle. Ramaphosa was able to oust Zuma from the presidency of the country without a pause and he announced his programme at the opening of Parliament in February 2018, invoking the values of Nelson Mandela: ‘We should put behind us the era of diminishing trust in public institutions and weakened confidence in our country’s leaders’ (Ramaphosa 2018). Ramaphosa promised to tackle growing corruption in the state sector and in the private sector. He sought a return to a model of cooperation based on trust and mutual benefit which Mandela and Mbeki, in different ways, had pursued. This was reflected in the many consultative processes he set up. Ramaphosa retained some of the flagship policies of Jacob Zuma, such as the development of a National Health Insurance system, the national minimum wage (which Ramaphosa had steered), and free higher education for students from poor backgrounds, but his approach most resembles the status quo ante. Some reforms have begun to take root. The justice and crime prevention sector which, save for the judiciary itself, had been gravely damaged in the Zuma era, began to be reformed and judicial enquiries and prosecutions of the corrupt have since made significant progress. Yet, movement on many of the simple economic reform goals Ramaphosa committed to has been painfully slow. Delayed policy execution is exemplified in the information and communications technology sector where the digital migration of broadcasting

86    Alan Hirsch, Brian Levy, and Musa Nxele scheduled to be completed by 2015 has still not started, the issuing of new spectrum is delayed repeatedly, and the appointment of members of the communications regulator by Parliament was repeatedly delayed and subject to excessive interference from the minister of communications. The charitable explanation is incompetence, but the underlying issue in this sector, in mining, in transport, in energy, and in other regulated economic sectors is the struggle over the distribution of rents. Such rents had become critical to the welfare of corrupt individuals and to different parts of the ANC patronage machine. Ramaphosa chose to address the corruption of the ANC through the slowly grinding legal machinery of the state, which had been undermined in the Zuma era but was slowly being rebuilt. He might instead have thrown his fate to the wind and drawn on the support of a sympathetic media and public in a head-​to-​head confrontation. But Ramaphosa, once a firebrand trade unionist, chose a more cautious route of avoiding confrontation with the party machine unless it was unavoidable. More than a year after he became president, at the 2019 elections the major social partners, the major trade unions, and much of the business community supported Ramaphosa’s ANC. Overall, support seeped from the centrist ANC and Democratic Alliance to the political fringes.

4.8 Conclusion South Africa’s mid-​1990s ‘political settlement’ seemingly transformed pre-​existing patterns of inequality and power both economically and politically, with profound consequences for inter-​elite relationships and the modes of incorporation of non-​elites. The settlement included implicit promises to: • rebalance political power and economic benefits between established and emerging elites • navigate potential rivalries among emerging elites (especially those within the ANC’s broad tent) in a way that sustains a platform of economic and political stability • create opportunities for upward mobility for non-​elites, beyond a favoured few • and address the country’s massive legacy of extreme poverty with the promise of ‘a better life for all’. The political settlement struggled to live up to its promise of sustaining hope, stability, and a virtuous spiral in the face of the daunting challenge of transforming and redressing the country’s difficult political and economic legacy. Failure to deliver enough on the promise of shared gains set in motion a cascading downward spiral of a decline in co-​ operation, growing pressure on institutions, and worsening economic performance.

Politics and Economic Policymaking in South Africa since 1994    87 The failure of the ANC to address the challenge of racial and economic inequality, and the persistence of poverty in the quarter of a century since the democratic transition, gave cover to economic looting of the state in the name of economic transformation. The inability of the reformers to have an impact on the underlying cause of instability threatens the survival of any kind of genuine reform programme. The likelihood of South Africa moving to a sustainable developmental path appears to rest on the ability of reformers to dislodge enough of the obstacles to reform put in place by the ANC’s patronage machine and at the same time to obtain broad support for a truly inclusive growth programme.

References ANC. 1990. ‘Discussion document on economic policy’, African National Congress, September 1990. ANC. 1992. ‘Ready to govern: ANC policy guidelines for a democratic South Africa’, Policy Unit of the African National Congress. ANC. 1994. The Reconstruction and Development Programme’, African National Congress. ANC. 2004. ‘A people’s contract to create jobs and fight poverty’, http://​www.anc.org.za/​ elections/​2004/​manifesto/​manifesto/​html. Black, Anthony. 2014. ‘The evolution and impact of foreign direct investment into South Africa since 1994’, in Haroon Bhorat, Alan Hirsch, Ravi Kanbur, and Mthuli Ncube (eds) The Oxford Companion to the Economics of South Africa. Oxford: Oxford University Press. Botha, R.F. 1995. ‘Affirmative action in South Africa’, PhD thesis, University of Johannesburg. Bureau for Economic Research. 2020. ‘Business Confidence Index’, https://​www.ber.ac.za/​ BER%20Documents/​ R MB/​ B ER-​ B usiness-​ C onfidence-​ I ndex/​ ? doctypeid=1095&y ear=2020#11590. Bhorat, Haroon, Carlene van der Westhuizen, and Toughedah Jacobs. 2009/​2012. ‘Income and non-​income inequality in post-​apartheid South Africa: What are the drivers and possible policy interventions?’. TIPS. http://​www.tips.org.za/​files/​u65/​income_​and_​non-​income_​ inequality_​in_​post-​partheid_​south_​africa_​_​bhorat_​van_​der_​westhuizen_​jacobs.pdf. Bhorat, Haroon, Aalia Cassim, and Alan Hirsch. 2017. ‘Policy coordination and growth traps in a middle-​income country setting: The case of South Africa’, in John Page and Finn Tarp (eds) The Practice of Industrial Policy: Government–​Business Coordination in Africa and East Asia. Oxford: Oxford University Press. BusinessTech. 2019. ‘South Africa unemployment: 1994 vs 2019’, https://​businesstech.co.za/​ news/​business/​353051/​south-​africa-​unemployment-​1994-​vs-​2019/​. Chipkin, Ivor, Mark Swilling, Haroon Bhorat, Mbongiseni Buthelezi, Sikhulekile Duma, Hannah Friedenstein, Lumkile Mondi, Camaren Peter, Nicky Prins, and Mzukisi Qobo. 2018. Shadow State: The Politics of State Capture. Johannesburg: Wits University Press. Eloff, Theuns, and Andre Fourie. 2006. local business, local peace: The peacebuilding potential of the domestic private sector case study South Africa’, International Alert, http://​www. international-​alert.org/​sites/​default/​files/​publications/​27_​section_​2_​South_​Africa.pdf. Erten, Bilge, Jessica Leight, and Fiona Tregenna. 2019. ‘Trade liberalization and local labor market adjustment in South Africa’, Journal of International Economics, Repository: Datafirst https://​www.datafirst.uct.ac.za/​dataportal/​index.php/​catalog/​434.

88    Alan Hirsch, Brian Levy, and Musa Nxele Evans, Gavin. 2020. ‘South Africa: Book reveals new, surprising nuggets about Nelson Mandela’s last years in jail’, The Conversation. https://​allafrica.com/​stories/​202009120386. html. Fedderke, Johannes. 2014. ‘South Africa’s growth performance’, in Haroon Bhorat, Alan Hirsch, Ravi Kanbur, and Mthuli Ncube (eds) The Oxford Companion to the Economics of South Africa. Oxford: Oxford University Press. Gelb, Stephen (ed.). 1991. South Africa’s Economic Crisis. Cape Town: David Phillip; London: Zed Books. Guelke, Adrian. 1996. ‘The impact of the end of the Cold War on the South African transition’, Journal of Contemporary African Studies, 14(1): 87–​100. Hassen, Ebrahim-​Khalil, and Miriam Altman. 2010. ‘Public service employment and job creation in South Africa’, Human Sciences Research Council. Hirsch, Alan. 2005. Season of Hope: Economic Reform under Mandela and Mbeki. Ottawa, Canada and South Africa: IDRC and University of Kwa-​Zulu Natal Press. Hirsch, Alan, 2020. ‘Fatal embrace: How relations between business and government help to explain South Africa’s low-​growth equilibrium’, South African Journal of International Affairs, 27(4): 473–​92. Hirsch, Alan, and Brian Levy. 2018. ‘Elaborate scaffolding, weak foundations: Business–​ government relations and economic reform in democratic South Africa’. ESID Working Paper no. 105, University of Manchester. Joffe, Avril, David Kaplan, David Lewis, and Raphael Kaplinsky. 1994. Improving Manufacturing Performance in South Africa: Report of the Industrial Strategy Project. UCT. Jonsson, Gunnar, and Arvind Subramanian. 2001. ‘Dynamic gains from trade: Evidence from South Africa’, IMF Staff Papers 48: 197–​224. Kahn, Brian, Abdel Senhadji, and Michael Walton. 1992. ‘South Africa: Macroeconomic issues for the transition’. Southern Africa Department, World Bank. https://​documents. worldbank.org/​en/​publication/​documents-​reports/​documentdetail/​459221468759566031/​ south-​africa-​macroeconomic-​issues-​for-​the-​transition. Kaplan, David. 2014. ‘Technology and innovation: Performance, policy, and prospects’, in Haroon Bhorat, Alan Hirsch, Ravi Kanbur, and Mthuli Ncube (eds) The Oxford Companion to the Economics of South Africa. Oxford: Oxford University Press. Kaplinsky, Raphael, and Claudia Manning. 1998. ‘Concentration, competition policy and the role of small and medium‐sized enterprises in South Africa’s industrial development’, Journal of Development Studies, 35(1)139–​161. Leibbrandt, Murray, Ingrid Woolard, and Christopher Woolard. 2007. ‘Poverty and inequality dynamics in South Africa: Post-​apartheid developments in the light of the long-​run legacy’. https://​ipcig.org/​conference/​ems/​papers/​ENG/​Leibbrandt_​Woolard_​Woolard_​ENG.pdf Levy, Brian, Alan Hirsch, Vinothan Naidoo, and Nxele Musa. 2021. ‘South Africa—​ when strong institutions and massive inequalities collide’. Carnegie Endowment for International Peace. https://​carnegieendowment.org/​2021/​03/​18/​ south-​africa-​when-​strong-​institutions-​and-​massive-​inequalities-​collide-​pub-​84063. Levy, Brian, Alan Hirsch, and Ingrid Woolard. 2015. ‘Governance and inequality: Benchmarking and interpreting South Africa’s evolving political settlement’. Effective States and Inclusive Development Working Paper no. 51, Manchester University. Lewis, David. 1995. ‘Markets, ownership and manufacturing performance’, in Avril Joffe et al. (eds) Improving Manufacturing Performance in South Africa: The Report of the Industrial Strategy Project. Cape Town: UCT Press/​IDRC.

Politics and Economic Policymaking in South Africa since 1994    89 Lipton, Merle. 1986. Capitalism and Apartheid: South Africa, 1910–​1986. Aldershot: Wildwood House. Mahajan, Sandeep. 2014. ‘South Africa’s suboptimal political economy equilibrium’, in Haroon Bhorat, Alan Hirsch, Ravi Kanbur, and Mthuli Ncube (eds) The Oxford Companion to the Economics of South Africa. Oxford: Oxford University Press. Mandela, Nelson, and Mandla Langa. 2017. Dare Not Linger: The Presidential Years. London: Macmillan. Marais, Hein. 1998. South Africa Limits to Change: The Political Economy of Transformation. London: Zed Press. Masiya, Tyanai. 2014. ‘Social movement trade unionism: Case of the Congress of South African Trade Unions’, South African Journal of Political Studies, 41: 443–​60. MERG. 1993. ‘Making democracy work: A framework for macroeconomic policy in South Africa’. Macroeconomic Research Group, Centre for Development Studies. Mohamed, Seeraj. 2019. ‘The political economy of accumulation in South Africa: Resource extraction, financialization, and capital flight as barriers to investment and employment growth’’. Doctoral dissertation, University of Massachusetts. https://​scholarworks.umass. edu/​dissertations_​2/​1533. Morris, Mike, and Vishnu Padayachee. 1988. ‘State reform policy in South Africa’, Transformation, 7:  1–​26. Nattrass, Nicoli. 1994. ‘Politics and economics in ANC policy’, African Affairs, 93(372): 343–​59. Nattrass, Nicoli, and Jeremy Seekings. 2014. ‘Job destruction in Newcastle: Minimum wage-​ setting and low-​wage employment in the South African clothing industry’, Transformation Critical Perspectives on Southern Africa, 84(1):  1–​30. NEDLAC. 2007. Annual report. http://​nedlac.org.za/​wp-​content/​uploads/​2019/​05/​2007-​part1. pdf. Padayachee, Vishnu, and Robert van Niekerk. 2019. Shadow of Liberation: Contestation and Compromise in the Economic and Social Policy of the African National Congress, 1943–​1996. Johannesburg: Wits University Press. Pritchett, Lance, Kunal Sen, and Eric Werker. 2018. ‘Deals and development: An introduction to the conceptual framework’, in Lance Pritchett, Kunal Sen, and Eric Werker (eds) Deals and Development: The Political Dynamics of Growth Episodes. Oxford: Oxford University Press. Pritchett, Lance, and Eric Werker. 2012. ‘Developing the guts of a GUT (Grand Unified Theory): Elite commitment and inclusive growth’. ESID Working Paper no. 16, University of Manchester. Ramaphosa, Cyril. 2018. ‘2018 State of the Nation Address’. https://​www.gov.za/​speeches/​ president-​cyril-​ramaphosa-​2018-​state-​nation-​address-​16-​feb-​2018-​0000. Ravinder, Rena, and Msoni Malindi. 2017. ‘Global financial crisis and its impact on the South African economy: A further update’, Journal of Economics, 5(1):  17–​25. Renwick, Robin. 2015. Mission to South Africa: Diary of a Revolution. Johannesburg: Jonathan Ball. Roberts, Simon. 2014. ‘Industrialisation strategy’, in Haroon Bhorat, Alan Hirsch, Ravi Kanbur, and Mthuli Ncube (eds) The Oxford Companion to the Economics of South Africa. Oxford: Oxford University Press. Simkins, Charles. 2019. ‘Left and right in South African politics today: I—​The left’. Helen Suzman Foundation HSF Briefs. https://​hsf.org.za/​publications/​hsf-​briefs/​left-​and-​right-​ in-​south-​african-​politics-​to-​day-​i-​the-​left#_​ftn7.

90    Alan Hirsch, Brian Levy, and Musa Nxele South African History Online. 2017. Federation of Unions of South Africa (FEDUSA). https://​ www.sahistory.org.za/​article/​federation-​unions-​south-​africa-​fedusa. Southall, Roger. 2012. ‘South Africa’s fractured power elite’. https://​wiser.wits.ac.za/​system/​ files/​seminar/​Southall2012_​0.pdf. Statistics South Africa. 1998. ‘Unemployment and employment in South Africa’. Van Rooyen, Johan, Johan F. Kirsten, Johan Van Zyl, Nick Vink, and Tracy Simba. 1996. ‘Structural adjustment and agricultural policy reform in South Africa’. USAID Regional Trade Agenda Series. Wilson, Francis. 1972. Labour in the South African Gold Mines, 1911–​1969. Cambridge: The University Press. Worgetter, Andreas. 2014. ‘Industrial structure and competition policy’, in Haroon Bhorat, Alan Hirsch, Ravi Kanbur, and Mthuli Ncube (eds) The Oxford Companion to the Economics of South Africa. Oxford: Oxford University Press. World Bank. 2020. Worldwide Governance Indicators. https://​info.worldbank.org/​governance/​ wgi/​Home/​Reports.

Chapter 5

Sou th Af ri c a ’ s P ost-​a parthei d E C ONOMI C Deve l opment T raj e c tory David Francis, Adam Habib, and Imraan Valodia

5.1 Introduction South Africa’s economic and social development trajectory in the post-​apartheid period remains a controversial subject. Notwithstanding an extensive literature, many of the key issues of the growth and development strategy, impact, and evidence remain contested. Our intention in this chapter is briefly to cover some of this debate, and to offer new insights and analysis of the issues. The literature is broad, with, at the one extreme, views from the far left, arguing that the African National Congress (ANC) has essentially ‘sold out’, and uncritically followed a neo-​liberal path (Bond 2000), to defenders of the ANC’s policies from active participants in the process, at the other (Hirsch 2005). Between these two views, there are a number of significant contributions (Seekings and Nattrass 2005; Bhorat, Kanbur, and Human Sciences Research Council 2006; Habib 2013; Padayachee and van Niekerk 2019, among others). Reviewing South Africa’s post-​apartheid economic trajectory and economic policy causes a pervasive sense of déjà vu. Writing about South Africa in the early 1990s, Alan Hirsch commented: To any economic observer in South Africa in the early 1990s, it was clear that the country had entered an economic cul-​de-​sac. The economy was shrinking. Its assets were being run down—​gross fixed investment was negative for four consecutive years to 1994, and capital was in full flight. National income was stagnating, and per capita income had declined every year since 1982, except 1988. Government debt was

92    David Francis, Adam Habib, and Imraan Valodia rising to dangerous levels, with the general fiscal deficit over 9% of gross domestic product in 1993. (Hirsch 2005: 112)

Despite describing a period almost thirty years ago, this accurately sketches the current macroeconomic situation in South Africa in 2020. And so it goes with post-​apartheid economic policy. We find a cycle which proceeds in a fairly consistent manner through the post-​apartheid period: an identification of the economic problems (which remain largely the same—​high unemployment, poverty, and inequality, and chronically low investment), the suggestion of policy solutions which cover a wide spectrum, the inability of the key actors to reach consensus, the failure of the state to effectively implement these policies, and so begins the cycle once again (Kantor and Rees 1982; Gelb 1991; Abedian and Standish 1992; Freund 2010; Habib 2013). The result of this policy malaise is, among other things, growing unemployment, rising poverty, and persistently high inequality of wealth and income. There is increasing evidence, even in the absence of any counterfactual, that South Africa’s developmental project, if that is the appropriate term, has been a failure by many important metrics. At the highest level, the country’s recent macroeconomic failure is evident in Figure 5.1, which shows how growth has not recovered from the 2008 financial crisis, and that there has been a corresponding decline in per capita income in recent years. It would, however, be a mistake to assume that because many of the economic challenges, like unemployment, inequality, and poverty, remain unresolved, there has been no economic progress. As Figure 5.1 above shows, there have been periods of economic growth and, for much of the 1990s, growth in GDP per capita. Despite little progress achieved in terms of solving the key structural problems, there have been other areas of success, too. These include the widespread roll-​out and provision of free basic services across the country, and near-​universal electricity access 60,000

6.0 5.0 4.0 3.0

50,000

2.0

45,000

1.0 –

40,000

GDP Growth %

Per Capita GDP (real LCU)

55,000

(1.0)

35,000

(2.0) 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019

30,000

Per Capita GDP (Constant LCU)

(3.0)

GDP Growth

Figure 5.1  GDP growth and GDP per capita in South Africa Source: World Bank Data Portal.

South Africa’s Post-apartheid ECONOMIC Development Trajectory    93 (Dinkleman 2011). There has also been a significant increase in the social wage, with the provision of free basic services, such as water, to poor households, and near-​ universal schooling attendance (Sulla and Zikhali 2018). In addition, one of the most impressive successes of the post-​apartheid state is the provision of social grants to almost 17 million people. But this further complicates the economic and development picture and raises the issue of the South African inequality paradox (Francis and Webster 2019). South Africa has a progressive constitution which enshrines socio-​economic rights, and this has been the basis for the development and successful implementation of progressive socio-​ economic policy, such as free schooling, social grants, and free health care for all. But this has not translated into any meaningful reduction in the levels of inequality of wealth and income in the post-​apartheid period. In this chapter we review the history and contestation of economic policy in South Africa and offer some explanations for why the country’s economic progress has been so uneven. We argue that this cycle of doom is the result of a number of persistent problems in the political–​economic structure of South Africa. The first is the failure of politicians and policymakers to account for the limits of South African state capacity to implement even simple economic reforms. Post-​apartheid economic policymaking is characterized by an assumption that the South African state is able to carry out complex economic coordination—​by 2020 this assumption looks increasingly unrealistic. Second, the impasse is really a political one caused by ideological contestation within the ANC, which has no mechanism to resolve the impasse.

5.2  Review of Key Economic Policies In this section we review the historical evolution of the contestation of ideas about post-​apartheid South Africa’s possible development trajectories, beginning before the transition but certainly including the positions of key actors during the political transition and post ​the transition into the Mandela and Mbeki presidencies. We see this contestation in key policy documents that were generated during the period 1985–​2018, including ‘Ready to Govern’, the Macroeconomic Research Group (MERG) report, the Reconstruction and Development Programme (RDP), the New Growth Path (NGP), and the National Development Plan (NDP), among others. The evolution of economic policy in post-​apartheid South Africa had four distinct phases: (i) the implementation of the Growth, Employment and Redistribution Strategy (GEAR) from 1996 to 2001, pioneered by Mbeki and Trevor Manuel, and implemented mainly during Mandela’s presidency; (ii) the massive expansion of the social ​support grant system, and the building of the Black middle and upper classes during Mbeki’s years as president (1999–​2007); (iii) the post-​Polokwane period under Zuma (from 2008); and (iv) the Ramaphosa period which began in 2017/​18.

94    David Francis, Adam Habib, and Imraan Valodia

5.3  The Shift to GEAR and the Mbeki Project Apartheid bequeathed a mixed economic legacy to South Africa. On the one hand it created appalling levels of marginalization and exclusion structured along racial lines. Not only was this morally reprehensible, but it also placed limits on the sustainability of the country’s economic growth. Bantu education—​one of apartheid’s most notorious policies—​ compromised the development of South Africa’s people, creating skills shortages that continue to constrain the economy. In addition, the oppression and economic exploitation of the Black majority limited growth in the manufacturing sector. As importantly, the industrialization and economic development that was spurred on by the mining booms of the late nineteenth century, and the peculiarly racialized forms of social and political rule that were developed to support these mines, meant the country’s modernization was driven by the mineral and energy sectors (Fine and Rustomjee 1996; Freund 2019). As the end of apartheid approached, two economic visions vied for supremacy in the battle to transform South Africa’s racially skewed economic inheritance in the post-​apartheid era. The first, advanced by the former ruling alliance, but most articulately argued by the corporate sector, was a traditional neo-​liberal economic programme. This emphasized economic growth, market solutions to social problems, and curtailing of the role of the state. The programme stressed the deregulation of financial and labour markets, the privatization of state-​owned enterprises (SOEs), and the integration of South Africa into the global economy. Aiming for a GDP growth target of 4.5 per cent and a 3 per cent increase in employment, the corporate sector proposed a set of supply-​side reforms, and the removal of political and social barriers that hindered the market (Hirsch 2005). Proponents argued that this would enable the economy to realize its maximum growth potential. The need for redistribution was acknowledged, but it was expected to occur through the trickle-​down effect as economic growth accelerated. In the language of the ANC in the 1990s, this was labelled ‘growth with redistribution’. The second economic vision was advanced by the ANC and its allies. Rooted in a ­social democratic vision of the new society, and understood as offering ‘growth through redistribution’, the aim was essentially a social-​democratic one that involved a mix between market and state planning. The vision was by no means a coherent one and evolved as the ANC came closer to assuming power. Indeed, ANC leaders issued an abundance of contradictory messages. Moreover, even the processes and stance of key policy documents was hugely contested. For example, the Macro Economic Research Group (MERG), established by the ANC, was quite marginal to the final policy position of the ANC as it came to power (see Padayachee and van Niekerk 2019). Other initiatives, for example, the Industrial Strategy Project, established by COSATU, were quite influential in the initial years of the ANC government.

South Africa’s Post-apartheid ECONOMIC Development Trajectory    95 As a result, on the eve of South Africa’s first democratic election in April 1994, two distinct economic visions were being advocated. The pre-​election Reconstruction and Development Programme (RDP), which reflected the vision and many of the policies and targets put forward by MERG, served as the ANC’s electoral manifesto. Many South Africans and external observers expected that the ANC would implement this developmental economic agenda after winning the election, and for a short while it seemed as if this was the plan. But soon after the April elections ushered in the government of national unity, evidence began to emerge (for example, through the White Paper) that the ANC in government was pursuing an economic agenda quite distinct from the pre-​ election RDP (Padayachee and van Niekerk 2019). Defenders of this shift argued that the critics had not understood the difference between the policies of the ANC as captured in its preparatory and election documents, and that of a party that had to confront the realities of being in government (see, for example, Hirsch 2005). This shift, many have argued (Habib 2013; Padayachee and van Niekerk 2019) was formalized in June 1996, when the Cabinet adopted the GEAR policy following significant pressure on the Rand, and a concern, as has historically been the case for South Africa, about the vulnerability of the economy to a balance-​of-​payments crisis. GEAR proposed cuts in state expenditure and the rationalization of the public sector so as to reduce South Africa’s budget deficit to 3 per cent by 1999. Deregulation measures were introduced, including the further liberalization of foreign exchange and new allowances intended to attract foreign investment. Support for the privatization of state assets was also reiterated. All of this took place under the guise of supporting the RDP but, in its essential thrust, GEAR represented a departure from the ANC’s 1994 election manifesto. GEAR’s economic impact was mixed. The programme did enable South Africa to strengthen its finances, lower its interest rates, and bring inflation under control. Inflation came down from an average of 15 per cent in the late 1980s and early 1990s, to an average annual rate of 5.2 per cent by 1999. Government debt as a percentage of GDP, which stood at 49.5 per cent in 1996, hovered in this range until about 2000, and then dropped significantly to 33 per cent by 2004, and 22.3 per cent by 2007. Debt-​ service expenditure decreased accordingly from a high of 5.7 per cent in 1999 to 3.9 per cent in 2004. These fiscal gains came at a social cost, however. Levels of inequality and poverty increased dramatically in the years immediately after the adoption of GEAR (Leibbrandt, Woolard, and Woolard 2007). Meanwhile, inequality, as measured by the Gini coefficient, actually increased from 0.672 in 1993 to 0.685 in 1999 (Leibbrandt, Woolard, and Woolard 2007). The realization of GEAR’s consequences facilitated a gradual but significant shift in economic policy during Mbeki’s presidency. First, social expenditure expanded massively, particularly in the form of social grants. The number of recipients of such grants increased from 2,687,169 in 1999 to 6,476,587 in early 2004, and then to 12,386,396 by late 2007. Second, state intervention came back into vogue in relation to Black economic empowerment (BEE), and a major infrastructure development programme was initiated. With regard to the former, the government became increasingly sceptical of the corporate sector’s commitment to deracialization. It therefore introduced new legislation

96    David Francis, Adam Habib, and Imraan Valodia that included the establishment of equity and transformation targets in industry charters. Importantly, GEAR did allow for the management of South Africa’s debt burden and created the fiscal space for a large infrastructure investment programme in the latter years of the Mbeki administration. For example, a Rand 780 billion programme was partly directed towards preparations for hosting the 2010 soccer world cup. Infrastructure investment increased rapidly. Much of this was planned through SOEs, and the privatization rhetoric of the late 1990s receded. On the production side of the economy, many of South Africa’s large corporations shifted their focus to the global economy, with the largest, such as the Anglo American Corporation, opting to shift their primary listing to London. Unlike some other developing countries, while government policy promised large benefits from a shift towards export o ​ rientation, the payoff from globalization and export ​orientation was limited (see Edwards, ­Chapter 21 in this volume). As the government’s rhetoric changed, the need to correct market failures became a priority, with poverty alleviation and even inequality reduction being given as much emphasis as growth. By 2004, the left within the ANC, including in COSATU and the SACP, were convinced that a significant reappraisal was underway within government. There was indeed a shift in favour of investment in infrastructure and increased social expenditure during Mbeki’s presidency. Notwithstanding these gains, the essence of the structural problems in the economy continued to challenge policymakers. There is some irony in the Mbeki government declaring budget surpluses of 0.3 per cent and 0.8 per cent of GDP in 2006 and 2007, while unemployment hovered at 36.4 per cent. Under Mbeki, the idea that South Africa was constituted of two economies—one modern, efficient, and internationally competitive; and the other informal, marginalized, poor, and overwhelmingly located in Black communities (Hirsch 2005; Devey, Skinner, and Valodia 2006)—gained prominence and allowed Mbeki to explain the impasse in solving the key structural challenges. While this analysis was appealing at face value, it enabled government to assume that the first economy required no intervention, and that only the second was in need of policy reform and social assistance. Yet, Mbeki’s dual-​economy analysis did not enable the presidency to understand that the policies and functioning of the first economy were precisely what was creating poverty and discontent in the second (Devey, Skinner, and Valodia 2006). Apartheid had established two sectors in South African society: one white and privileged, the other Black and disadvantaged. The ANC’s explicit mandate was to facilitate the transcendence of this racial divide. Instead, while the economic and social policies it pursued in the first decade of its rule began a process of deracializing the first sector, these same policies simultaneously increased the size and aggravated the problems of the second. Increasing the state’s social expenditure did not solve the problem. In fact, as long as South Africa’s industrial, trade, monetary, and fiscal policies stayed as they were, they worked against the social expenditure components of the national budget. This contradiction has lain at the heart of the ANC government’s policy ensemble since 1996, and it continued throughout both of Mbeki’s terms in office.

South Africa’s Post-apartheid ECONOMIC Development Trajectory    97 The contradiction manifested itself very clearly in the Mbeki administration’s performance. Annual GDP grew progressively from 2000 and reached 5.1 per cent in 2007, driven in part by the booming global economy. Formal employment began to take off from 2003, and broad unemployment decreased from 42.5 per cent in early 2003 to 34.3 per cent in late 2007. Poverty levels dropped, partly because of the new jobs, but mainly as a result of the massive increases in the number of people receiving social grants. Between 2003 and 2007, the poorest 10 per cent of the population increased their per capita real income from R921 to R1,032, but their share of total income remained at 0.6 per cent. Thus, the excessive levels of inequality, which increased in the second half of the 1990s because of GEAR, have since remained constant.

5.4  Groping towards Social Democracy—​the Zuma Years There is a widespread assumption that President Zuma introduced no radical changes in economic and social policy during his time in office. Yet there is evidence to suggest that South Africa’s economic and social policies shifted in important ways. The question is whether this shift was effective in addressing the challenges of inclusive development and reducing inequality, poverty, and unemployment. This question must also be considered in the context of how the configurations of power have evolved in South Africa since 1994. The ejection of Mbeki from the presidency inevitably empowered the grassroots within the ANC, who had realized that political leaders should be accountable to them, and could be replaced if they were not responsive to their interests. More importantly, COSATU and the SACP’s leverage increased because they had provided a significant share of the institutional platform required for Zuma’s ascendance to power. This is not to imply that COSATU and the SACP have ruled the roost in the past decade; the corporate sector retained substantial power and leverage, too. In effect, the power differential in favour of business that prevailed under Mbeki’s reign eroded slightly in favour of a more equitable balance of power—​both within the ruling party and in the country as a whole. This, in turn, constrained the behaviour and choices of political elites, and began to condition the evolution of political and economic life in South Africa. This new, more equitable balance of power was immediately evident in the Zuma administration’s responsiveness to public and stakeholder opinion, especially in his appointments to Cabinet and other important economic portfolios. Business was concerned that Trevor Manuel would be ejected, and that there would be too fundamental a change in economic policy. The left, mainly represented by COSATU and the SACP, were concerned that too little change would take place—​that Manuel and company would retain control of economic policy and that alternative voices would not be heard.

98    David Francis, Adam Habib, and Imraan Valodia Zuma’s early appointments appeased both sides. Manuel was retained in a new portfolio within the presidency (national planning), and Pravin Gordhan, who had completed a sterling performance as commissioner of the South African Revenue Service, was brought in as finance minister in an effort to appease the markets. These appointments were coupled with those of Rob Davies as trade and industry minister, and Ebrahim Patel as minister of economic development. Davies and Patel had strong connections with the SACP and COSATU respectively. A similar appeal to multiple stakeholders informed Gill Marcus’s appointment as governor of the Reserve Bank in 2009. A stalwart of the ANC, Marcus had previously served in the Treasury, and her sojourn in the private sector involved not only chairing the board of a mining company, but also that of South Africa’s largest bank (ABSA). Marcus’s appeal transcended the boundaries of the corporate sector, however, and both COSATU and the SACP welcomed her appointment. Her strength lay in the fact that she had never been fundamentalist in her economic or ideological perspectives. She was perceived as pragmatic, consultative, and less pompous than her predecessor, and appealed to multiple stakeholders. The result of this mix of appointments was to introduce a plurality of thought in the corridors of economic power that gave several stakeholders the prospect of influencing policy. This more equitable balance of power was also reflected in a deepening of the shift to the left in economic and social policy that had begun with Mbeki. Social-​ support grants continued to be expanded, the higher education and training sector saw a renewed focus on making tertiary education more affordable and on rebuilding the training college sector, and the health ministry became much more responsive, especially in relation to the HIV and AIDS pandemics, and in its commitment to developing a national health insurance scheme. But the most dramatic evidence of the shift to the left emerged on the economic front and became apparent at multiple levels, including both fiscal and monetary policy. In the midst of the global economic crisis, the last national budget drawn up by Manuel as finance minister and the first two budgets drawn up by his successor, Pravin Gordhan, expanded the fiscal deficit to 7.3 per cent, 6.2 per cent, and 5.3 per cent of GDP in 2009, 2010, and 2011 respectively. Spending on social and economic development increased in areas such as the maintenance of the infrastructure programme, social ​support grants, rural development, education, and health care. Similarly, interest rates maintained a downward trajectory under the Zuma administration. The management of South Africa’s monetary policy had subtly changed under Marcus to include variables such as employment and growth in order to prioritize price stability. Perhaps the strongest indication of the government’s shift to the left was the adoption of the Industrial Policy Action Plan 2011/​12–​2013/​14 and the New Growth Path, pioneered by Rob Davies and Ebrahim Patel respectively. The Industrial Policy Action Plan (informally known as IPAP2) had, as its goal the creation of 2.4 million jobs by 2020. The New Growth Path, under Patel’s leadership, squarely set as its target the reduction of South Africa’s economic inequalities. To achieve this, a set of policies with two

South Africa’s Post-apartheid ECONOMIC Development Trajectory    99 distinct goals were recommended: enhancing the livelihoods of those at the bottom of the economic ladder; and temporarily and voluntarily constraining the incomes of the upper-​middle classes and the elite. To enhance livelihoods at the bottom, policies that facilitated increased employment were proposed. These ranged from infrastructure and skills-​development programmes, through to a competitions policy, a looser monetary agenda, as well as small-​business promotion and financing. At the same time, under Manuel’s leadership, the National Development Plan (the NDP) was adopted (National Planning Commission 2011). The NDP, strong on the diagnosis of the structural problems in the economy, aimed to eliminate poverty in South Africa by 2030, by doubling employment, raising incomes, addressing key impediments to growth in the energy sector, and developing a capable state. All noble intentions, but the NDP was always viewed by COSATU, the SACP, and other left forces, as linked with Trevor Manuel. While there was a lot of fanfare about the NDP, it never really truly became government policy, and its implementation has been lacklustre to say the least. Nevertheless, there can be no doubt that the collective mix of economic and social policies pursued by the Zuma administration had a social-democratic flavour, as distinct from the economic agenda pursued since 1994. Its aim was job creation and inclusive development. Clearly policy rupture did not occur between the Mbeki and Zuma eras, but this does not mean that significant change did not take place. We need to understand the policy developments for what they are: a continuation of the leftward shift in economic and social approaches that began under Mbeki, and which was deepened, strengthened, and acquired an increasingly social-democratic flavour under the Zuma administration. These nuances are significant if only because the new policies began to explicitly and directly target inequality, poverty, and unemployment. In this sense, they proved significantly different to the mix of policies pursued since 1994. This drift towards a neo-​Keynesian economic agenda was also being enabled by a similar shift globally. The global economic crisis weakened the global corporate sector and this enhanced the leverage of states and national political elites. This new configuration of power enabled a substantive change in the macroeconomic policy environment. US banks were virtually nationalized via the largest bailout in history. Similar bailouts were facilitated for the motor industry and other conglomerates. Effectively, the American state addressed the financial crisis in exactly the way it, and the International Monetary Fund (IMF), has prevented developing nations from managing theirs. And the Americans were not alone. The governments of almost every other major economy in the world undertook similar interventions. The net effect has been to open up possibilities for the emergence of development agendas across the world. This does not mean that the neo-​ Keynesian orientation of economic and social policies in South Africa was or is uniformly supported. There are a number of countervailing trends. The first, and most worrying, is corruption. As the National Development Plan recognizes, this is reaching a scale that threatens the state’s entire developmental programme. The second is that many technocrats within the state continue to support important aspects of GEAR. There are many in the Treasury, the Reserve

100    David Francis, Adam Habib, and Imraan Valodia Bank, and the National Planning Commission who remain tied to rigid deficit and inflation targets that have the potential to compromise the neo-​Keynesian agenda.

5.5  Contradictory Approaches to Inequality In spite of the overlap in focus between the New Growth Path and the National Development Plan, especially on improving the livelihoods of the poor, a fundamental philosophical difference was evident in the two plans on the issue of inequality. The New Growth Path hoped to address inequality by expanding livelihood opportunities at the lower end of society and containing enrichment at the upper end. The National Development Plan, by contrast, supported the expansion of livelihood opportunities at the lower end, but was silent on containing inequality at the top end. Those who compiled the National Development Plan seemed to imagine that it is possible to reduce economic inequalities simply by growing the economy, expanding employment opportunities, and addressing poverty—​a typical trickle-​down strategy. Mainstream business, its economists, and aligned research agencies shared this view. The resolution of this dilemma, we argue later, will be determined by the cracks that have emerged in the ruling party and within the Tripartite Alliance. At the heart of these conflicts, which are ongoing, lie serious differences between nationalists and the left about where South Africa is with regard to its development, the goals that should be pursued, and the policies that should be advanced in order to achieve them. Over the last decade, the debate has revolved around the meaning of national democracy. Nationalists within the ANC maintained that COSATU and the SACP ignore the fact that South Africa is in the national-​democratic rather than the socialist phase of its transition. Members of the left wing held that the national democratic phase should not focus on developing conservative Black capitalists, but should instead have moved towards creating a social democracy and a mixed economy, in which inclusive development was at the core of the state’s agenda. For the drift to social democracy to be maintained, it was essential that COSATU, the SACP, and social-democratic elements in the ANC retained their political momentum. The moulding of an economic consensus within the ruling party required a bridging of the nationalist–​left divide, and an engagement with intra-​party stakeholders who were advancing different economic agendas. This required the establishment of parameters for acceptable trade-​offs, and the moulding of an economic consensus within the ANC. Despite initial optimism about the ability of President Ramaphosa, who became president of the ANC in 2017 and of South Africa the following year, recent years have seen the economic situation worsen, and any economic consensus within the ruling party collapse. There are three distinct but interrelated reasons why these opportunities were squandered. The first is that the country has seen a significant decline in the

South Africa’s Post-apartheid ECONOMIC Development Trajectory    101 quality of appointments to senior positions in the public service and parastatals, and this has dramatically eroded the capabilities of the state. While this was most starkly illustrated by Zuma’s firing of Nhlanhla Nene and later Pravin Gordhan, and their replacement by unknown Des van Rooyen and the compromised Malusi Gigaba respectively, the pattern of such appointments was far reaching and included key government institutions, including the revenue authority, SARS. This occurred concurrently with a consolidation of power by the Zuma administration, leading to a more powerful president but a less capable state. Second, while corruption has been a feature of the South African state for decades, there was a startling escalation in corruption under the Zuma administration, which was most visibly manifest in the capture of the state by the Gupta family but extended through most of the state (Swilling et al. 2018). Third, these two developments drove a wedge between the ANC and the state, and provoked widespread divisions as contestation emerged over the beneficiaries of corruption. These three developments paralysed the implementation of economic policy, escalated the economic and social crises, and dramatically reduced livelihoods, resulting in a fall in economic growth, significant increases in unemployment, and increases in levels of poverty. Moreover, South Africa’s fiscal position worsened significantly as Zuma made far-​reaching decisions, for example to extend free higher education, with no consideration for the fiscal implications of his actions.

5.6  Evidence for Developmental Outcomes A key issue in the debate about economic policy is what evidence of development outcomes is available, and the interpretation thereof. While everyone agrees that because of the social grants system development outcomes have been better than they would otherwise have been, and that unemployment remains an intractable issue, there is a lot of disagreement about the evidence itself; the conclusions that may be reached, and the causal mechanisms (Seekings and Nattrass 2005, 2015; Leibbrandt, Woolard, and Woolard 2007; Hundenborn, Leibbrandt, and Woolard 2016; Sulla and Zikhali 2018). Here we review these data and findings and offer our interpretation of the issues. At the macroeconomic level, South Africa has not performed well. Unemployment in 2020 is largely at the same level as it was in 2000. Income inequality remains largely unchanged and ranks as the highest in the world among countries for which there are good data (Francis and Webster 2019). Similarly, the macro-​fiscal situation remains precarious, and is deteriorating (Sachs 2021). But this is not to suggest that the country has made no developmental progress at all. Indeed, there have been notable successes in improving access to basic services and utilities, the expansion of social assistance in the form of social grants, and the provision of housing and infrastructure. This has resulted in a decline in multidimensional

102    David Francis, Adam Habib, and Imraan Valodia poverty in South Africa from 17.9 per cent in 2001 to 7.0 per cent in 2016 (Sulla and Zikhali 2018). Social infrastructure has also improved steadily. For example, the roll-out of electricity has been widespread in post-​apartheid South Africa. In 1994, only 62 per cent of the population had access to electricity, and by 2014 this had risen to 87 per cent (Sulla and Zikhali 2018). Whereas in 2002, 4.5 million households had access to piped water in their dwelling, by 2018, this had increased to 7.7 million (Statistic South Africa General Household Survey 2018). The provision of housing has been equally successful, with the state building 3.2 million homes between 1994 and 2018. While poverty rates improved throughout the Mbeki years, unfortunately, mainly due to HIV/​Aids, South Africa’s general health indicators worsened. The infant mortality rate worsened from 46.9 per cent in 1990 to 52.1 per cent in 2000, before improving to 38.4 per cent in 2010. Similarly, life expectancy fell from 61.6 per cent in 1995 to 57.7 in 2010, before improving to 62.6 per cent in 2015, as policy changed and a highly successful anti-​retroviral programme was rolled out (UNU HDR Database 2020). But as Sulla and Zikhali note, while there was rapid improvement in access to services in the 2000s, economic progress slowed after 2010 (Sulla and Zikhali 2018). This is largely due to the inability of the South African economy to generate jobs at a magnitude which would alleviate some of the socio-​economic problems the country faces. This problem has a long historical arc. As argued by Hirsch in 2005, Two of the most important lessons gradually learned by the government during the first decade of freedom were that investment dynamics often were not national— rather they were frequently regional and local—​and that the manufacturing sector was not ever likely to be a major supplier of jobs in South Africa, though it remained an important dynamo for growth. (Hirsch 2005: 148)

And this continues to be the case. Between March 2008 and March 2020, the manufacturing industry shed 405,000 jobs (Statistics South Africa 2020). The only sectors to add significantly to employment were finance (737,000 new jobs) and community and social services (1.05 million new jobs). While there is consensus that some aspects of social policy have been effective, there is also consensus that, because of the pattern of economic growth, inequality has increased, and this in a society with already extremely high levels of inequality. As Stephen Gelb argued in 2010, and remains true ten years later: Not only has recent growth been low, but its pattern has been ‘unequalising’. Significant shifts in the sectoral composition of output and of trade between 1990 and 2003 have led to a ‘skills twist’ in the labour force, where jobs have been created for high-​skilled workers at a relatively rapid rate while unemployment amongst low-​ skilled workers has grown. (Gelb 2010: 51)

It is uncontentious that the triple challenge of poverty, inequality, and unemployment has, at its heart, the South African labour market. The extent of the problem is evident

South Africa’s Post-apartheid ECONOMIC Development Trajectory    103 Table 5.1: Labour force overview Q1 2008

Q1 2020

Q2 2020

Thousand

Thousand

Thousand

Population aged 15–​64

31,544

38,874

39,021

Labour Force

18,808

23,452

18,443

Employed

14,438

16,383

14,148

Formal sector (non-​agricultural)

9,934

11,282

10,064

Informal sector (non-​agricultural)

2,433

2,921

2,280

Agriculture

838

865

799

Private households

1,233

1,316

1,005

Unemployed

4,371

7,070

4,295

Not economically active

12,736

15,422

20,578

Discouraged work-​seekers

1,202

2,918

2,471

Other (not economically active)

11,534

12,504

18,107

Unemployment rate

23.2

30.1

23.3

Employed /​population ratio (absorption)

45.8

42.1

36.3

Labour force participation rate

59.6

60.3

47.3

Rates (%)

Source: Statistics South Africa (2020).

if we examine some key labour market statistics over the past decade.1 We use, as reference, the period before the global financial crisis in 2008 (Q1 2008). We compare this to the latest QLFS data.2 Between Q1 2008 and Q1 2020, the labour force grew by 4.6 million people, while employment grew by 1.95 million. Most striking is that, over this period, the unemployed grew by 2.7 million, and there were an additional 1.7 million people who became discouraged work-​seekers. This means that by March 2020 there were 9.99 million people who were either unemployed or had given up looking for work altogether (Statistics South Africa 2020). Table 5.1 also includes the latest results from the quarterly labour forces survey (QLFS), which gives an indication of the significance of the COVID-​19 pandemic and the lockdown on the labour market. These data have prompted furious debate. This is largely because the lockdown prevented people from looking for work, which resulted in the unemployment rate falling significantly in the second quarter of 2020. Much like the GDP data, they reflect a significant economic shock, but the longer-​term effects will only be visible in subsequent quarters. 1 

Unemployment in South Africa is discussed in Chapters under Part IV in this volume. The quarter 2 QLFS generated furious debate about the definitions and reporting of the narrow unemployment rate—​for our purposes it is more useful to examine the actual extent of job losses. 2 

104    David Francis, Adam Habib, and Imraan Valodia There is, of course, no definitive single answer to the puzzle of South Africa’s persistently high unemployment. However, it is uncontentious that one of the leading causes has been the country’s poor growth performance which has characterized the post-​ apartheid period, with the exception of the commodity boom years of the early 2000s. Given the labour market picture, it is unsurprising that inequality has not decreased. But what is surprising is that when we did grow, inequality increased (Sulla and Zikhali 2018). This is largely due to the fact that the growth period was fuelled by a resource boom, and the gains of such a boom are distributed highly unequally (Sachs 2021). Why, then, has South Africa’s growth performance been so poor? Hirsch highlighted a number of factors including a lack of investor confidence, skills shortages at the higher end of the labour market, and what he rather obliquely terms ‘a range of related institutional factors’ (Hirsch 2005: 160). These issues recur again and again in South Africa’s developmental trajectory. This leads us to restate the questions Hirsch posed in 2005: ‘How can we sustain a higher level of investment by the private sector? How can we afford not to focus resources on high potential growth and/​or employment sectors?’ (Hirsch 2005: 160).

5.7  South Africa as a Developmental State? One of the weaknesses in many of the existing reviews about post-​apartheid economic policy is a focus on the content of these policies, with less attention paid to how successfully (if at all), they have been implemented. One of the most striking observations which emerges when reviewing South Africa’s post-​apartheid economic policies is the persistent optimism about the capacity of the South African state to play a positive developmental role (Hirsch 2005). There has been mounting evidence which should disabuse policymakers and academics about the possibility of the South African state to effect meaningful change (see, for example, Freund 2010). This is not a generalized critique about the role states can play in effecting economic change, but about the South African state in this current moment. There are a number of reasons why the South African state has been unsuccessful in many aspects of its developmental agenda. The first is that much economic policy does not account for the lack of skills and experience within the civil service—​unlike the Asian Tigers, South Africa does not possess the machinery within the civil service to implement the complex policies designed to support investment and growth. Second, there are a number of important cases where pragmatic interventions have been stalled because they have become proxies for ideological battles (Eskom and SAA being the two most important examples). Third, rampant and widespread corruption has diverted resources and political and public attention away from policy implementation (Swilling et al. 2018).

South Africa’s Post-apartheid ECONOMIC Development Trajectory    105

5.8  Breaking the Impasse? Economic Growth and Inclusive Development In recent years, the unseating of Jacob Zuma, the apparent undermining of the state capture project, and the rise of Cyril Ramaphosa, have created a renewed focus on the development project in South Africa. High levels of debt—​a result of the large increases in expenditure under Zuma without any significant impact on the growth rate—​has generated a more complex set of macroeconomic challenges. Moreover, state capacity has been severely undermined. Added to this, in 2020, COVID-​19 has had a devastating impact on economic output, public health, and livelihoods. How might it be possible for South Africa to address its development challenges? A key problem to address is the continuing and increasing levels of inequality in society. Notwithstanding the gains in addressing some of the infrastructure backlogs that characterized apartheid, and the massive growth in social grants, South Africa continues to be one of the most unequal societies in the world by a number of metrics. This is highlighted by research on income and wealth (Hundenborn, Leibbrandt, and Woolard 2016; Espi, Francis, and Valodia 2019; Webster and Francis 2019; Chatterjee, Czajka, and Gethin 2020). On income measures, South African households in the poorest income quintile have a total monthly income of just R2,600, which has to support five members of the household (Francis and Valodia 2020). About 18 million South Africans live in these households. In contrast, on average, the highest income quintile has an income of R38,000, which has to support two household members. Some 7 million South Africans live in these households. Data on the distribution of wealth are even worse. The poorest 50 per cent of South Africans have an average net wealth of negative R16,000. That means their assets are less than their liabilities; they are deeply in debt. By contrast, the richest 10 per cent of South Africans have an average net wealth of R2.8 million per person. The top 1 per cent of holders of wealth in South Africa have an average net wealth of R17.8 million per person (Chatterjee, Czajka, and Gethin 2020). Addressing these social and economic challenges will, we argue, require a socialdemocratic political economy that is able comprehensively to deal with the challenges of inequality, poverty, and unemployment. Although this was acknowledged at many ANC national conferences and some elements of a social-democratic platform have gradually been adopted, consensus on this economic perspective remains elusive within the ruling party with elements in the ANC continuing to advocate for fiscal conservatism in the guise of financial prudence. The battle for the economic soul of the ANC has to be resolved, the economic divides bridged, and a consensus built around a socialdemocratic political economy. And consensus has to be built within the wider society. Related to this is the need for a new social pact to be cemented between business, labour, and the state (Francis, Valodia, and Webster 2020). However, various ANC administrations have failed to manage the expectations of both the economic elite and

106    David Francis, Adam Habib, and Imraan Valodia the general populace, with the result that the other essential foundation for a successful pact—​a willingness by all to defer the immediate realization of their desires—​has not been achieved. It is cause for concern that the country’s diverse set of leaders, with their impressive track records and experience, do not yet seem to understand the basis of successful social pacts. Pacts require compromise. They are established to manage the dilemmas presented by competing interests. A central dilemma that a South African social pact would have to address is how to enable economic competitiveness and the accumulation of work experience by new entrants to the labour market, without losing the hard-​won gains of the labour movement as expressed in the Labour Relations Act of 1995. The unions fear that compromises might weaken the 1995 Labour Relations Act, and that this would allow employers to roll back the gains won by workers in the formal sector. There is precedent for this in the post-​apartheid era, and to argue that this is impossible and unrealistic is to be seriously out of touch with the economic dynamics of the last twenty years. For example, in the mid-​1990s, internal employees carried out cleaning services in most companies and public institutions. Since then, most private companies and public institutions have subcontracted cleaning services to external companies that employ workers at much lower wages and provide far fewer benefits. Similarly, a study on transformation in South African mines demonstrated that as the racial ownership of mines has changed, so working conditions have worsened (Capps and Mnwana 2015). Essentially, South Africa’s more marginal mines have been sold to Black entrepreneurs who have derived profits from squeezing wages and reducing benefits. This has been achieved mainly through the use of labour brokers, which have served as a ‘safety valve’ for business owners, enabling them to avoid the obligations made mandatory by the Labour Relations Act. Consensus will also require the business elite to shift its focus towards a social pact that addresses the high levels of inequality (Francis, Valodia, and Webster 2020). This will require, inter alia, commitments on investment, addressing the high levels of earnings premia paid in the upper echelons of the corporate sector, and addressing the issue of democracy in the workplace through representation of workers on boards, share ownership schemes, and the like. A critical factor on the part of government is addressing the vexing problem of state capacity. Two issues are especially important here. First, realigning the state towards effective regulation and service delivery must involve building a meritocratic state. This necessarily involves addressing the political question of how the state operates. The ANC policy of deploying its cadres into the upper echelons of the state has to make way for a more skills-​based approach to management in government. Second, many of South Africa’s SOEs remain locked in their histories of being key elements in the apartheid state. A major overhaul of these institutions is needed to meet the exigencies of a state steering an economy in the digital age (for a broader discussion of this see Swilling and Callaghan, ­Chapter 27 in this volume). More systemically, however, any social pact to resolve the impasse in development strategy and create a more equitable and sustainable society will need to address four challenges that characterize South Africa’s development problems in

South Africa’s Post-apartheid ECONOMIC Development Trajectory    107 the post-​apartheid period. First, while political forces have converged to improve livelihoods and incomes at the bottom end of the distribution, all previous attempts at a social pact have not been able to address the growth in incomes and wealth at the top end of the distribution. Second, addressing poverty has been based almost entirely on social grants and other forms of transfers, rather than by creating sustainable economic opportunities and employment. In comparable contexts, for example Brazil under Lula, state transfers were combined with bringing the poor in the informal economy into the productive and mainstream economy. In South Africa there is little appetite, especially among the left, to grow the informal economy. Third, the imperative to deracialize society creates a tension between opening up opportunities for the political elite and addressing inequality. Black economic empowerment is one of the key drivers of growing inequality. Finally, there is the issue of which social forces the political and economic elite represent. For all of the fanfare associated with the public articulation of social pacts, the reality is that in an economy with such high levels of unemployment and low levels of investment, the trade unions and business organizations represent only a small element of the key actors needed to bring consensus for a more inclusive development path, and the implementation of such a consensus.

5.9 Conclusion The literature on South Africa’s post-​apartheid development trajectory tends to be focused in two binaries—​either that the ANC government has ‘sold o ​ ut’ to a neo-​liberal project, or that the inability of the ANC to bed down a conservative economic stance characterized by GEAR is the Achilles heel of the post-​apartheid period. We argue that neither of these does justice to the nuance of the actual development trajectory post-​1994. Our argument is that significant advances have been made, particularly in areas of social infrastructure and in the extension of social grants to large numbers who have not been incorporated into an economic growth trajectory. However, the key structural problems which characterized the economy at the end of apartheid—​high levels of poverty, high levels of unemployment, and growing levels of inequality—​have not been solved. There is no shortage of policy documents—​from the RDP to the NDP to Ramaphosa’s post-​COVID-19 economic recovery plan—​that analyse these issues and propose solutions. The problem, we argue, is not only an economic one. Rather, the political processes that have shaped post-​apartheid South Africa, especially within the ANC and its alliance partners, are key to understanding why South Africa’s development trajectory continues to be so uneven. While a broad social pact in the society has been elusive and the different factions of the ANC continue to contest the direction of development policy, consensus on some issues has emerged during the final years of the Zuma administration. For example, in 2019, South Africa introduced a national minimum wage after agreement was reached

108    David Francis, Adam Habib, and Imraan Valodia by all the major social partners in NEDLAC. The fall ​out from COVID-​19 has severely set back the development trajectory of South Africa and exacerbated an already difficult fiscal challenge. The debate on the post-​COVID-​19 economic and social policy trajectory has generated some level of consensus on the need to focus on addressing the lacklustre growth record and on the need for fiscal consolidation. However, as has been the case for much of the post-​apartheid period, two trends appear to characterize much of the narrative. First, how growth will be generated, the pace of the fiscal consolidation, and who will bear the cost of this adjustment continue to divide both the ANC and its alliance partners. Second, the lack of capacity in the state to deliver on the sometime-​ambitious programme now appears more limited than ever. The Perennial development quagmire remains: how to unlock a politics that would enable inclusive development?

References Abedian, Iraj, and Barry Standish (eds). 1992. Economic Growth in South Africa: Selected Policy Issues. Oxford: Oxford University Press. Bhorat, Haroon, Ravi Kanbur, and Human Sciences Research Council (eds). 2006. Poverty and Policy in Post-​apartheid South Africa. Cape Town: HSRC Press. Bond, Patrick. 2000. Elite Transition: From Apartheid to Neoliberalism in South Africa. London: Pluto Press; Sterling, VA: Pietermaritzburg, South Africa: University of Natal Press. Capps, Gavin, and Sonwabile Mnwana. 2015. ‘Claims from below: Platinum and the politics of land in the Bakgatla-​ba-​Kgafela traditional authority area’, Review of African Political Economy, 42(146): 606–​24. Chatterjee, Aroop, Leo Czajka, and Amory Gethin. 2020. ‘Estimating the distribution of household wealth in South Africa’. SCIS Working Paper no. 3. Southern Centre for Inequality Studies, Wits University [online] https://​www.wits.ac.za/​media/​wits-​university/​ faculties-​and-​schools/​commerce-​law-​and-​management/​research-​entities/​scis/​documents/​ Estimating%20the%20Distribution%20of%20Household%20Wealth%20in%20South%20 Africa.pdf. Devey, Richard. Caroline Skinner, and Imraan Valodia. 2006. ‘Definitions, data and the informal economy in South Africa: A critical analysis’, in Vishnu Padayachee (ed.) The Development Decade? Economic and Social Change in South Africa, 1994–​2004. Cape Town: HSRC Press. Dinkleman, Taryn. 2011. ‘The effects of rural electrification on employment: New evidence from South Africa’, The American Economic Review, 101(7): 3078–​108. Espi, Gabriel, David Francis, and Imraan Valodia. 2019. ‘Gender inequality in the South African labour market: Insights from the Employment Equity Act data’, Agenda, 33(4):  44–​61. Fine, Ben, and Rustomjee, Zaareh. 1996. The Political Economy of South Africa: From Minerals Energy Complex to Industrialisation. London: Hurst. Francis, David, and Imraan Valodia. 2020. ‘South Africa needs to mitigate the worst of its inequalities in tackling coronavirus’, The Conversation, 5 April [online] https://​ theconversation.com/​south-​africas-​budget-​to-​deal-​with-​covid-​19-​fails-​to-​pave-​way-​for-​ more-​equal-​society-​141458.

South Africa’s Post-apartheid ECONOMIC Development Trajectory    109 Francis, David, and Webster, Edward. 2019. ‘Poverty and inequality in South Africa: Critical reflections’, Development Southern Africa, 36(6): 788–​802. Francis, David, Imraan Valodia, and Webster, Edward. 2020. ‘Politics, policy and inequality in South Africa under Covid-​19’, Agrarian South: Journal of Political Economy, 9(3). Freund, Bill. 2010. ‘Development dilemmas in post-​apartheid South Africa: An introduction’, in Bill Freund and Harold Witt, (eds) Development Dilemmas in Post-​apartheid South Africa. Scottsville: University of Kwazulu-​Natal Press. Freund, Bill. 2019. Twentieth-​century South Africa: A Developmental History. Cambridge: Cambridge University Press. Gelb, Stephen (ed.) 1991. South Africaʾs Economic Crisis. Cape Town: D. Philip; New Jersey: Zed Books. Gelb, Stephen. 2010. ‘Macroeconomic policy and development’, in Bill Freund and Harold Witt, (eds) Development Dilemmas in Post-​apartheid South Africa. Scottsville: University of Kwazulu-​Natal  Press. Habib, Adam. 2013. South Africa’s Suspended Revolution: Hopes and Prospects. Athens, OH: Ohio University Press. Hirsch, Alan. 2005. Season of Hope: Economic Reform under Mandela and Mbeki. Scottsville, South Africa; Ottowa: University of KwaZulu-​Natal Press; International Development Research Centre. Hundenborn, Janina., Murray Leibbrandt, and Ingrid Woolard. 2016. ‘Drivers of inequality in South Africa’. SALDRU Working Paper no. 194, Cape Town. Kantor, Brian, and David Rees. 1982. South African Economic Issues. Cape Town: Juta. Leibbrandt, Murray, Ingrid Woolard, and Christopher Woolard. 2007. ‘Poverty and inequality dynamics in South Africa: Post-​apartheid developments in the light of the long-​run legacy’, in IPC-​DRCLAS Workshop, Brasilia, pp. 1–​51. https://​ipcig.org/​conference/​ems/​papers/​ ENG/​Leibbrandt_​Woolard_​Woolard_​ENG.pdf. National Planning Commission. 2011. National Development Plan: Vision for 2030. Pretoria, South Africa: The National Planning Commission. Padayachee, Vishnu, and Robert van Niekerk. 2019. Shadow of Liberation: Contestation and Compromise in the Economic and Social Policy of the African National Congress, 1943–​1996. Johannesburg: Wits University Press. Sachs, Michael. 2021. ‘Fiscal dimensions of South Africa’s Crisis’. Public Economic Project SCIS Working Paper Number 15. Online https://​www.wits.ac.za/​media/​wits-​university/​faculties-​ and-​schools/​commerce-​law-​and-​management/​research-​entities/​scis/​documents/​Sachs-​ 2021-​Fiscal%20dimensions%20Working%20Paper%2015.pdf. Seekings, Jeremy, and Nicoli Nattrass. 2005. Class, Race, and Inequality in South Africa. New Haven, CT: Yale University Press. Seekings, Jeremy, and Nicoli Nattrass. 2015. Policy, Politics and Poverty in South Africa. London: Palgrave Macmillan. Statistics South Africa. 2020. Quarterly Labour Force Survey Quarter 2: 2020. Statistical Release P0211, Statistics South Africa, Pretoria [online] http://​www.statssa.gov.za/​publications/​ P0211/​P02112ndQuarter2020.pdf. Sulla, Victor, and Precious Zikhali, 2018. ‘Overcoming poverty and inequality in South Africa: An assessment of drivers, constraints and opportunities’. International Bank for Reconstruction and Development/​ The World Bank. http://​documents.worldbank.org/​

110    David Francis, Adam Habib, and Imraan Valodia curated/ ​ e n/ ​ 5 30481521735906534/ ​ p df/ ​ 1 24521- ​ R EV- ​ OUO- ​ S outh- ​ A frica- ​ Poverty- ​ and-​ Inequality-​Assessment-​Report-​2018-​FINAL-​WEB.pdf. Swilling, Mark et al. 2018. Shadow State: The Politics of State Capture. Johannesburg: Witwatersrand University Press. Webster, Edward, and David Francis, 2019. ‘The paradox of inequality in South Africa: A challenge from the workplace’, Transformation: Critical Perspectives on Southern Africa, 101(1):  11–​35.

Chapter 6

C on straints to E c onomi c Grow th in Sou t h   A fri c a Kenneth Creamer

6.1 Introduction This chapter will analyse the drivers and constraints on the rate of economic growth in South Africa from the 1950s’ apartheid-​era through to the democratic period post-​ 1994. The analysis will include discussion of the impact on economic growth of the country’s history of racial injustice and exclusion, of South Africa’s industrial structure including linkages to the global commodity price cycle, of South Africa’s evolving macroeconomic imbalances and infrastructure investment failures, and of the impact of poor state capacity weakened by corruption. The chapter will conclude with a reflection on policy perspectives for overcoming South Africa’s economic growth constraints.

6.2  The Apartheid Growth Model and its Crisis After centuries of colonial conquest and land dispossession, the implementation of the policy of apartheid from 1948 sought, ultimately without success, to tie Black South Africans to tribally based Bantustans. Conditions in the Bantustans did not permit viable economic activity and migrant labourers were forced to seek employment in mining, agriculture, the metropoles, and other regions of ‘white’ South Africa. Typically, migrant labourers were paid very low wages based on the false presumption that their families back in the Bantustans—​and equivalently in nearby

112   Kenneth Creamer states such as Lesotho, Mozambique, and Malawi—​were to some degree economically self-​sustaining.1 Long before apartheid was launched in 1948, the South African state developed an extensive apparatus of repression based primarily on control over movement. Passes and permits ‘allowed the ruling classes to keep African works in a permanent state of subordination’.2 This was the social basis which underpinned South Africa’s apartheid economic growth model. Super-​exploitation of Black workers, facilitated by ‘grand’ apartheid’s regulation of the flow of Black workers into the formal economy, reinforced ‘petty’ apartheid’s laws of racial segregation aimed at keeping apart Africans, coloureds, Indians, and whites—​allowing differential access to social services and protections, and regulating, among many other things, who could own a business, who could attend university, who could live where, and who one could marry. As can be seen from Figure 6.1—​which periodizes post-​Second World War-South African economic growth into four distinct episodes, namely, ‘apartheid growth’ (1946–​70), ‘crisis of apartheid’ (1971–​93), ‘transition to democracy’ (1994–​2010), and ‘persistent low growth’ (2011–​19). Through the period of the 1950s and 1960s the apartheid economy grew rapidly with GDP growth rates comparable to other fast-​growing economies worldwide, although apartheid policies ensured that the lion’s share of that growth flowed to the white population. The real GDP growth rate of 4.9 per cent from 1946 to 1971, fell in the 1970s and 1980s to an average annual growth rate of 2.0 per cent as certain inherent contradictions and limitations in the apartheid growth model began to come to the fore. This decline in the rate of economic growth in the 1970s fed into theoretical debates between liberal and Marxist writers in South Africa about how best to understand the relationship between capitalism and apartheid. Liberal theorists had argued that apartheid’s racially discriminatory laws served to limit the development of capitalism, in that apartheid disallowed the emergence of an urbanized Black middle class and related consumption and wealth-​generating potential (Marquard 1952). Apartheid’s racist limitation of Black access to education, and strict job reservation of certain skilled employment categories for whites only, limited the productivity-​enhancing skills and know ​how transfers required for the South African economy to compete along the skills-​intensive

1 

‘The logic of domination dictated that the majority of the Black population should remain isolated and dispersed in rural areas and mine compounds; the logic of industrialization dictated that they should concentrate in urban areas to provide the labour necessary for economic expansion . . . In 1922, the Stallard Commission attempted to defy the logic of industrialization by laying down the principle that: “The Native should only be allowed to enter urban areas, which are essentially the white man’s creation, when he is willing to enter and to minister the needs of the white man, and should depart therefrom when he cease so to minister”. Migrant labour would thus have to be the means by which the white economy obtained its workforce while white society maintained its dominance of the urban areas’ (Feinstein 2005). 2  Bundy (2016: 56).

Constraints to Economic Growth in South Africa    113 GDP Real Growth Rate - annual % change 10 Apartheid Growth (1946-71) Ave Growth 4,9%

8

Crisis of Apartheid (1972-93) Ave Growth 2,0%%

6

Transition to Democracy (1994-2010) Ave Growth 3,2% Persistent Low Growth (2011-19) Ave Growth 1,5%

4

2016 2018

2014

2012

2006 2008 2010

2002

2004

1998 2000

1992 1994 1996

1990

1988

1986

1984

1982

1980

1978

1976

1972 1974

1970

1968

1964 1966

1960

1962

1956 1958

1952 1954

1948

1950

0

1946

2

-2

-4

Figure 6.1  Real GDP growth rate 1946–​2019, annual percentage change Source: SARB GDP growth series, KPB6006Z, and author’s own calculations.

growth path being followed by many other countries in the global economy (Lipton 1985).3 Along a different vein, Marxist theorists contended that the apartheid system in fact facilitated capitalist development through providing for the super-​exploitation of Black labour drawn from the country’s Bantustans (Wolpe 1972). This racialized form of capitalism allowed for the significant extraction of surplus value from Black labour, flowing into profits which could be reinvested in profitable activity, for the enrichment of a rent-​seeking, white capitalist class and for tax-​and-​spend transfers to service the needs mainly of the privileged white minority. The arrival of the crisis of the apartheid growth model in the 1970s and 1980s informed Wolpe’s Marxist-​oriented characterization of the relationship between apartheid and capitalism as being historically and materially contingent, rather than always being either a priori functional or dysfunctional (Wolpe 1988). In this formulation, loosely analogous to a reading of history where social forms of earlier epochs become fetters due to the evolution of productive forces—​say as feudalism gives way to capitalism, and thereafter capitalism gives way to socialism—​apartheid could be described as 3  Acemoglu and Robinson (2012) offer the following description, ‘It should be no surprise that the type of economic development that white South Africa was achieving was ultimately limited, being based on extractive institutions the whites had built to exploit the Blacks . . . Economic institutions were extractive . . . This economic growth without creative destruction from which only the whites benefited, continued as long as revenues from gold and diamonds increased. By the 1970s, however, the economy stopped growing’ (p. 270).

114   Kenneth Creamer initially having been functional to capitalist development in South Africa and thereafter as dysfunctional. From the mid 1970s—​commencing with the 1972–​73 wave of ‘wild-​cat’ strikes by Black workers, who at the time, unlike white workers, had no legal right to form trade unions—​it became increasingly apparent that apartheid’s unjust policies were beginning to constrain economic growth in South Africa. This inflection point in the mid-​ 1970s occurred against, and was impacted by, the backdrop of momentous changes taking place in the global economy at the time, such as the collapse of the Bretton Woods system of fixed and regulated exchange rates, the oil price shock that accompanied the formation of the Organization of the Petroleum Exporting countries (OPEC) cartel, and the stagflation that preceded the global ideological economic policy shift from a position where Keynesian demand-​side management was predominant to the dominance of Friedman-​inspired, neo-​liberal, supply-​side policies, which aimed at smaller government, deregulation, the reduction of trade union power, and policies to reduce and control inflation. The international constraint on apartheid South Africa’s economic growth potential was elucidated by Kahn in his explanation of the ‘balance-​of-​payments constraint’ faced by the country (Kahn 1991). If the economy grew, then South Africa would increase its imports more quickly and it could generate hard currency through exports or through capital inflows. This constraint was deeply structural in nature, given South Africa’s reliance on the import of certain intermediate and consumer goods and it being a ‘price-​taker’ in volatile non-​oil, global commodity export markets. This international constraint was intensified in the 1980s by rising economic sanctions on exports from, and finance to, apartheid South Africa, which was becoming increasingly politically and economically isolated in the international community. Unable to resolve its economic crisis, and after ultimately unsuccessful attempts during the 1980s at superficial political reforms—​such as establishing a tricameral parliament for whites, coloureds, and Indians and Black local authorities for limited African political representation at local government level—​combined with intense internal and external repression of anti-​apartheid activity, the apartheid regime through a negotiated settlement capitulated to one-​person-​one-​vote rule in 1994. This democratic breakthrough eased a number of the more immediate ‘political’ constraints on economic growth—​opening up the South African economy to increased international trade and investment, creating an enabling environment for economic reconstruction and development, and promising a democratic dividend as relative peace and stability came to replace the cycle of violence, instability, repression, and resistance which had intensified in the mid-​to-​late 1980s and early 1990s. From Figure 6.1, it can be seen that the average annual real GDP growth rate from 1994 to 2010 increased to 3.2 per cent, from the 2.0 per cent average growth rate of the apartheid economy in crisis, but from 2011 onwards GDP growth fell sharply to an average annual growth rate of only 1.5 per cent.

Constraints to Economic Growth in South Africa    115

6.3  Post-​apartheid Constraints and Performance The constraints on the post-​apartheid economy’s growth can be understood through the analysis of a number of inter ​related domestic and global, structural and political-​ economy factors. First, the unaddressed legacy of racial and gender exclusion resulted in sustained under-​development and lack of opportunity for significant sections of the population. Second, the post-​1994 economic growth rate was influenced and constrained by developments in the global economy, such as, global financial and health crises, the rise of China’s industrial capacity, and the related movements in the global commodity price cycle. Third, policy choices and macro ​economic imbalances resulted in declining investment levels, particularly investment in growth-​enhancing social and economic infrastructure. Fourth, the vision of a process of state-​led development which would be capable of guiding transformative, inclusive growth was undermined by state-​ capture, corruption, and related weaknesses in state capacity.

6.4  Legacy of Racial and Gender Exclusion The post-​apartheid government faced political constraints different from those of the apartheid regime, such as managing the legitimate expectations and hopes of the Black majority for a better life post liberation, managing white fears of a deteriorating standard of living in order to avoid the outflow of skills and capital, as well as taking actions which would establish the new government’s credentials and ability to manage effectively a modern democratic state and a large, relatively sophisticated, mixed, open economy. During the negotiations to end apartheid in the early 1990s, many of these issues played out in the terrain of economic policymaking. Within the anti-​apartheid forces, there was intense contestation about economic policy. The Macroeconomic Research Group published detailed policies on how to advance economic transformation objectives (MERG 1993), and the policy clashes between business and labour (Adelzadeh 1996) and political economy contestations (Marias 1998) have been well documented. Moderating forces were at work within South Africa’s liberation forces, typified, for example, by Mandela’s re-​written speech for the 1991 World Economic Forum in order to better mollify, and ultimately attract, foreign investors (Hirsch 2005). For a fuller discussion of the politics of economic policymaking, see C ­ hapter 4 in this volume and for a discussion on the impact of such contestation on South Africa’s post-​apartheid development trajectory, see ­Chapter 5 in this volume.

116   Kenneth Creamer Despite these areas of intense contestation, the broad strategic thrust towards a democratic dispensation and economic transformation was achieved through negotiated settlement and the country’s post-​1994 constitutional dispensation explicitly mandated government to take corrective action in order to overcome racial, gender, and other forms of inequality. The constitution through its entrenchment of social and economic rights placed an obligation on the democratic state progressively to expand access to education, health, and housing.4 Furthermore, the constitution also sought to correct the injustice of historical land dispossession by mandating that, in the public interest, land reform should be undertaken by the state. In a careful balancing of interests in favour of accelerated land reform and so as to avoid the process being caught up in a market-​determined dynamic of willing buyer and willing seller, the Constitution mandated that compensation for land expropriation should be ‘just and equitable’, rather than simply being based on current ‘market value’.5 While there have been important advances in securing certain social and economic rights for South Africans—​access to education, housing, water, electricity, and land—have improved, the quality of such services has been questionable and the scale of these changes has not been sufficient to bring about a definitive, qualitative, and transformational shift in order to overcome decisively South Africa’s inherited racialized patterns of privilege and under p ​ rivilege. Similarly, while programmes of land restitution and redistribution have increased Black land ownership, Black ownership of farmland and well-​situated and serviced urban land remained inadequate. Although informal settlements, particularly those around the big urban centres, have grown, housing and urban development has failed to keep up with the rate of internal urbanization and inward migration from other countries. South Africa’s 1994 negotiated settlement could not immediately resolve the many deep-​rooted structural impediments to economic growth resulting from the country’s history of colonialism and apartheid. Racialized inequalities of wealth, land, income, and opportunity were entrenched, just as access to education, skills, and health services continued to be racially skewed. ‘Dispossession has in fact become the source of major developmental handicaps for at least some and possibly many countries’,6 including South Africa (Arrighi et al. 2010). Hirsch (2005) takes this argument further, suggesting that a significant growth constraint, and a fundamental cause of the lack of competitiveness of the South African economy, has been due to the fact that ‘apartheid raised living costs for all South Africans, especially the poor . . . The working poor were located miles from their potential places of work, and often equally far from commercial and public services . . . The deterioration of access to public services such as education, health, and social security for Africans under apartheid meant the diminution of the social wage.’7 4  Chapter

2 of South Africa’s Constitution outlines a Bill of Rights, including the rights to housing (S.26), to health care, food, water, and social security (S.27), and to education (S.29). 5  S.25 of South Africa’s Constitution. 6  Arrighi (2011: 411). 7  Hirsch (2005: 182–​3).

Constraints to Economic Growth in South Africa    117 Employment and unemployment 1994 to 2019 10500000

31

10000000

29 27

9500000

25 23

9000000

21

8500000

19 8000000

Unemployment (% annual) (LHS)

2017 2018 2019

2016

2013 2014 2015

2005 2006 2007 2008 2009 2010 2011 2012

2003 2004

2000 2001 2002

15

1994 1995 1996 1997 1998 1999

17

7500000

Employment-formal non-agricultural (RHS)

Figure 6.2  Rising employment and unemployment, 1994–​2019 Source: SARB data series KPB7009J and KPB7019J.

The ongoing reality of racial and gender exclusion severely constrains growth in South Africa. Poor rural villages and urban informal settlements with high rates of unemployment and poverty are not conducive locales for new business development or even for investment in municipal infrastructure reliant on the collection of municipal rates and user ​fees for water, electricity, and other services. During the first two and a half decades of South Africa’s democratic dispensation, the absolute size of the Black middle class increased, and the absolute number of people in employment has grown, but neither at sufficient pace to keep up with wider demographic changes. In the labour market, the increase in the numbers of new entrants into the labour market each year resulted in a situation where both the number of people employed and the number of people who are unemployed have risen at the same time, as outlined in Figure 6.2 It can be seen that the official rate of unemployment—​not taking into account those who had become discouraged and had given up looking for employment—​rose from about 20 per cent in 1994 to about 30 per cent in 2019 as the number of formal non-​agricultural jobs rose, too slowly, from about 8.6 million to over 10 million during this period.8 In general, the labour force participation rate rose in the post-​apartheid era but remained low compared with other countries (OECD 2008). The number of women

8  As a reference, full employment in South Africa, including the sizeable agricultural sector and informal employment, increased from 14.4 million in March 2008 to 16.4 million in March 2020. The impact of the COVID-​19 pandemic resulted in a 2.2 million decrease in total employment in the second quarter of 2020 to a total employment level of 14.1 million (Statistics South Africa 2020).

118   Kenneth Creamer who had historically been excluded, but who began entering the labour market was a key driver of the growing size of the labour force in the post-​apartheid era. Typically, Black women continue to be the lowest paid, experience the highest level of unemployment, and endure the highest level of poverty in South Africa. While training and skills development for the middle and upper classes is of a high quality, children of the poor and unemployed are not given access to the skills that enable effective economic participation. For example, even though South Africa spends a comparable amount of its GDP on education, the country’s education outcomes have continued to be disappointing. For example, despite the fact that reading outcomes among South African schoolchildren improved between 2006 and 2011, the transfer of reading skills stagnated between 2011 and 2016 with Spaull and Pretorius reporting that ‘78 per cent of Grade 4 children cannot read for meaning in any language’ based on data from the Progress in International Reading Literacy Study of 2016 (Spaull and Pretorius 2019).

6.5  Impact of Global Factors Including the Global Commodity Cycle As a relatively open economy, a number of important channels exist between South Africa and the global economy. Most significantly, despite having a comparatively diversified production base including sizeable secondary and tertiary sectors, economic growth in South Africa has been historically linked to the global commodity price cycle.9 This is due to the fact that South Africa’s mining sector is comparatively large, producing a significant range of globally traded minerals. Furthermore, the mining and energy sectors—​through what has been characterized as the minerals–​energy complex (MEC)—​is significantly integrated into secondary and tertiary activity in the country. Fine and Rustomjee’s (1996) seminal work identified the importance of South Africa’s MEC as both a driver of economic growth and over time as a fetter to further industrialization.10

9 

Based on the combined value of imports and exports measured at around 50 per cent of GDP, South Africa has had a comparatively open economy over the full post-​Second World ​War period. This has been driven by South Africa’s extractive, colonial history, which has had the effect of linking economic performance to the global minerals cycle, but at the same time South Africa has experienced a high degree of industrialization—​and related international trade patterns—​compared with other African and colonized countries. 10  ‘Although the MEC’s impact has differed over time and across sectors, it has effectively led to policies which both supported its core sectors and precluded the adoption of other industrial policies of diversification away from economic dependence on South Africa’s resource base’ (Fine and Rustomjee 1996: 14).

Constraints to Economic Growth in South Africa    119 The combination of the economy’s linkage to the global commodity price cycle and the amplification of the effects of this cycle on the wider economy through the economy’s MEC-​linkages, result in a situation where downturns in the global commodity price cycle pose constraints on economic performance. When the terms of trade are in South Africa’s favour, this helps to lift the economic growth rate, but when the terms of trade turn against South Africa—​for example when gold, platinum, coal, iron o ​ re, and manganese prices decline and oil prices rise—​this negatively impacts on the country’s economic growth. For a fuller analysis of the intersection between the mining and energy sectors and economic growth, see ­Chapters 13 and 14 in this volume, respectively. The MEC was described as being based on close linkages and multiple inter-​ penetrations between mineral extraction, energy provision, and industrialization. For example, coal mined in South Africa has been used to generate the electricity required for the mining of other minerals and other economic activity and for the production of liquid fuels by Sasol. This mining and electrification activity would facilitate various upstream and downstream industrial activities, would provide the basis for financial activity, and would provide fiscal linkages in the form of state revenue. The MEC facilitated the ‘growth of manufacturing around primary production’11 and also allowed for the evolution of a large and sophisticated financial sector. ‘Despite its origins and productive basis in the mining and energy sectors, the MEC revolves around a separate but intimately related epicentre—​that of finance’.’12 As can be seen from Figure 6.3, during South Africa’s early democratic period, broadly speaking from after the 1997 Asian crisis to around 2010, the global commodity cycle rose significantly and continuously, so the period was described as the commencement of a seemingly ineluctable commodities super-​cycle.13 Due to these favourable terms-​ of-​trade effects, South Africa’s mineral exporting economy grew each year, without any intervening periods of recession, from around 1997 until the 2007–​09 financial crisis, dubbed the Great Recession. During the upswing in the commodity cycle, South Africa’s annual real GDP growth rate rose above 5 per cent per annum, employment levels increased, and the rate of unemployment began to fall (see Figure 6.2). Despite the early 2000s being a period of relatively high economic growth, certain longer-​run structural forces continued to impact on the economy’s performance and on the quality, composition, and distributional effects of the country’s economic growth.14 11

  Fine and Rustomjee (1996: 5). Fine and Rustomjee (1996: 10). 13  In what turned out to be a case of remarkable bad timing an IMF Staff Paper published in 2008 argued that ‘the evidence is consistent with the hypothesis that there have been three super cycles in the past 150 years or so, and that we are currently in the early phase of a fourth super cycle . . . [due] primarily to Chinese urbanisation and industrialisation’ (Cuddington and Jerrett 2008). 14  The quality and composition of growth meant that in South Africa, as in a wide range of economies around the globe, there were structural forces at play which had the effect of increasing inequality both of income and of the ownership of assets. Some, such as Piketty (2014), have ascribed rising inequality to the fact of weakening public fiscal institutions and long-​term dynamics of the capitalist system whereby the return on assets held by the wealthy (r) is greater than the rate of economic growth (g). Others, such as Brynjolfsson and McAfee (2011), have argued that the ongoing wave of technological change is 12 

120   Kenneth Creamer Commodity cycle and the Real GDP growth rate 1994 to 2015 115

6

110

5

105

4

100

3

95

2

90

1

85

0

80

–1

75 70

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 Commodity Cycle (2015 = 100) (LHS)

–2

GDP Growth Rate (Annual % Change) (RHS)

Figure 6.3  Commodity cycle and the real GDP growth rate, 1994–​2015 Source: Author’s own calculations based on Sachs 2020 and Jacks 2019; the commodity cycle is based on a weighted commodity price index (where 2015 = 100) including the following commodities gold, platinum, coal, iron ore, chromium, copper, aluminium, and steel.

Minerals beneficiation and industrialization continued to face global headwinds, mainly as a result of rising Chinese manufacturing power. This resulted in increased competition from global trade and the inward flow of imported manufactured goods into the South African market—​in turn resulting in de-​industrialization pressures, with sustained decline in both the share of manufacturing in total employment and the share of manufacturing in GDP (Tregenna 2008). It was also during the early 2000s that a major policy error was made with regard to the investment in the country’s electricity-​generation capacity. Complacency about high rates of economic growth, together with uncertainty over whether the government’s policy framework should favour public or private investment in future electricity power generation plus a mis-​estimation of the long lead-​times involved in building new electricity-​generation capacity, meant that by 2007 South Africa was experiencing electricity shortages. The 2007–​09 financial crisis impacted negatively on the rate of economic growth in South Africa mainly as a result of falling demand for South African exports as the country’s trading partners all went into recession. After the 2007–​09 financial crisis, the global commodity cycle began a downward trend from 2011, placing a significant tantamount to a Great Restructuring in which the acceleration of technology has negative consequences for wages and jobs and that, while digital progress grows the overall size of the economy, it does this while leaving the majority of people in a poorer position. In the South African context specifically, Burger (2015) has suggested that labour’s falling share of income is due to financialization and aggressive returns-​oriented investment strategies that have resulted in greater investment in capital-​augmenting labour-​saving technologies.

Constraints to Economic Growth in South Africa    121 constraint on South Africa’s rate of economic growth, which continued to fall precipitously from that time until the country fell into recession prior to the impact of the COVID-​19 pandemic that further accelerated the economic contraction. The downturn in the commodity price cycle was due in no small part to China losing faith in the ongoing reliability of the Western world’s consumption of its exports in the wake of the 2007–​09 financial crisis and deciding to rebalance its growth model away from prioritizing investment to increasing domestic consumption. This rebalancing of the Chinese growth model, which resulted in reduced demand for mineral inputs, was a key factor in bringing down global commodity prices,15 with significant negative consequences for mining revenues in South Africa. This, through the numerous MEC-​ linkages, began to constrain the overall economic growth performance of the South African economy, as well as the country’s fiscal potential. From 2011 onwards South Africa’s economic growth rate fell. Even when there was a degree of recovery of global commodity prices in 2016, South Africa’s economic growth rate did not respond as expected. By that time the linkage to the global commodity cycle had been weakened by a number of internal factors in the South African economy which meant that mining output was on the decline and was not able to respond to rising international prices. These factors included the shortage of electricity which limited the expansion of mining output and meant that new mining investment could not take place. Political and policy uncertainty about mining rights and the recognition of earlier Black empowerment deals in the mining sector also were not conducive to investment in the sector. In particular, electricity constraints and policy uncertainty meant that mining investment was curtailed and economic infrastructure—​associated with rail, port, and logistical networks—​became costly and inefficient. Likewise, the potential for downstream mineral beneficiation was severely curtailed by the country’s electricity shortage. The slowdown in the mining sector and the shortage of electricity transmitted through the various MEC linkages accelerated the process of de-​industrialization that had been a characteristic of the post-​apartheid economy. South Africa, like many other economies, found it very difficult to compete with industrial and other imports from China and other Asian countries.16 Industrial capacity shifted to those countries who pursued aggressive export-​oriented industrialization policies that were 15 Other

factors included demand-​subduing crises in a number of European countries, centred around Greece, as well idiosyncratic developments in the world platinum market leading to a falling price of platinum. This resulted in reduced sales of diesel-​powered vehicles due to the exposure of fraudulent claims around the environmental benefits of diesel engines, which used platinum in their catalytic converters. 16  ‘The reason behind these trends has to do both with technology and trade. Rapid global technological progress in manufacturing has reduced the prices of manufactured goods relative to services, discouraging newcomers in developing countries from entry. At the same time, manufacturing has become much more capital and skill intensive, substantially reducing the labour-​absorbing potential of the sector for workers from agriculture of informality. On the trade front, competition from China and other successful exporters combined with the reduction in protection levels means that few poor countries now have the opportunity to develop simple manufactures for home consumption. The room for import-​ substitution was squeezed out’ (Rodrik 2018: 90–​1).

122   Kenneth Creamer fundamentally more competitive as they began to include millions of new, skilled, lower-​cost workers into global trade. This fundamental challenge from new global industrial competitors, together with a ‘perfect-​storm’ of factors, including the downturn in the global commodity cycle, significant policy uncertainty, state corruption and incapacity, and an electricity shortage, meant that from 2011 onwards South Africa’s economic growth rate began to fall sharply with no clear signs of recovery on the horizon.

6.6  Macroeconomic Imbalances and Policy Uncertainty Constraining Growth and Investment Causality runs both ways between investment and growth—​economic growth creates an environment conducive to increased levels of investment and rising investment drives growth. From Figure 6.4 it can be seen that South Africa has experienced a downward spiral in gross fixed investment since the 2007–​09 financial crisis, with investment falling from a level of over 23 per cent of GDP in 2008, to below 18 per cent of GDP in 2019. Persistent macroeconomic imbalances and the dependence of the South African economy on foreign savings increases the fragility of the economy, making it vulnerable to capital flow reversals, which would lead to Rand currency depreciation and to rising Investment as % of GDP 1994 to 2019 24,0 23,0 22,0 21,0 20,0 19,0 18,0 17,0 16,0

19 9 19 4 95 19 9 19 6 97 19 98 19 9 20 9 00 20 01 20 02 20 0 20 3 04 20 0 20 5 06 20 07 20 0 20 8 09 20 10 20 11 20 12 20 13 20 1 20 4 15 20 16 20 1 20 7 18 20 19

15,0

GDFI (as % of GDP) (LHS)

Figure 6.4  Investment level and GDP growth, 1994–​2019 Source: SARB data series KPB6009J and author’s own calculations.

Constraints to Economic Growth in South Africa    123 interest rates. Foreign savings also allow for the persistence of both the country’s current account deficit and its budget deficit. Unlike export-​oriented growth models, persistent current account deficits, whereby the value of imports consistently exceeds the value of exports, tend to constrain economic growth rates and lead to increased foreign ownership of domestic economic assets leading, in turn, to the ongoing outflow of dividends to foreign-​based owners. In response to the 2007–​09 financial crisis, counter-​cyclical fiscal policy meant increased fiscal deficits in an effort to boost aggregate demand and growth. The sustained period of low growth in the post-​2011 period meant that fiscal deficits, which were intended to be cyclical, in fact morphed into structural deficits in the sense that due to the unexpectedly reduced level of economic growth, tax revenues were not sufficient to cover government’s expanded expenditure obligations. Unlike cyclical deficits which are reduced in the medium run as growth recovers—​leading to increased tax revenues and a degree of expenditure containment—​structural deficits became entrenched as the lower-​than-​expected rate of economic growth resulted in an entrenched gap between tax revenue and expenditure, fuelling rapidly rising national debt, from a level of about 50 per cent of GDP in 2016 to around 80 per cent of GDP in 2020 (South African National Treasury 2020b). Rising public-​sector wages and social security payments effectively ‘crowded out’ general government infrastructure spending to the extent that increased budget deficits and national debt were used to pay increased wages and debt-​repayments rather than drive an increase in government-​funded investment. Following the 2007–​09 global economic crisis, South Africa significantly increased state-​guaranteed investment by state-​owned companies just as private-​sector investment and general government investment levels fell, as indicated in Figure 6.5. Investment by state-​owned companies

Composition of investment (as % of GDP) 1994 to 2019 15,0 13,0 11,0 9,0 7,0 5,0

State Owned Companies

2019

2018

2017

2016

2015

2014

2013

2011

2012

2010

2009

2007

2008

2006

2005

2004

2003

2001

Government

2002

2000

1999

1998

1997

1996

1995

1,0

1994

3,0

Private Sector

Figure 6.5  Composition of investment (as % of GDP), 1994–​2019 Source: SARB data series KPB6100J, KPB6106J, and KPB6109J and author’s own calculations.

124   Kenneth Creamer in this period—​such as the building of two large coal-​fired power stations Medupi and Kusile—​was poorly managed and in many cases these investments were tainted by inefficiency and corruption. Poorly conducted investment by state-​owned companies saw state-​owned-​company investment levels decline from 2013 onwards; these failings also weighed heavily on business confidence and private-​sector investment has remained at subdued levels over the same period. For a fuller analysis of the intersection between macroeconomic policy and growth see ­Chapter 41 in this volume, as the current chapter will focus on the impact of falling investment levels in specific key economic sectors, such as electricity, roads, and telecommunications infrastructure.

6.7  Electricity Infrastructure With regards to electricity supply, government signalling from the late 1990s that power utility Eskom should not invest in new generation capacity as private-​sector investment would be facilitated, and then not opening the policy space for such investment, has been the fundamental reason for post-​apartheid South Africa’s electricity shortage. Thereafter, ongoing contestation and policy ambivalence has continued to cloud attempts to overcome the country’s electricity shortages since the first occurrences of load-​shedding in 2007. In response to the electricity shortage, Eskom built two large coal-​fired power stations—​Medupi and Kusile—​unfortunately at the very time when the global energy sector was transitioning away from large-​scale vertically integrated, coal-​based electricity-​ generation models towards a low-​ carbon, relatively de-​ centralized and lower-​cost renewable energy model. With serious delays and cost over-​runs at Medupi and Kusile, South Africa did embark on a programme of wind and solar renewable-​ energy investments based on its Integrated Resource Plan 2010. Future renewable-​ energy investments were thereafter disrupted for a number of years from 2016 to 2019, due to local and global vested interests pushing for coal and nuclear power options, despite the fact that renewable-​energy technologies, supplemented by gas, had become the lowest-​cost and most rapidly deployed option to plug South Africa’s electricity gap (Bischof-​Niemz and Creamer 2019). The immediate impact of the contestation and lack of policy certainty around South Africa’s electricity investments was to stall pent-​up investment in electricity infrastructure. The direct costs in the form of lost economic output were estimated at between R60 billion and R120 billion a year in 2019 and 2020 (CSIR 2020). The economic costs of the prolonged electricity shortage in the form of reduced investment, reduced exports revenues, and lost employment opportunities, were far greater as they placed a significant constraint on the country’s economic growth and development and left a permanent scar on post-​apartheid South Africa.

Constraints to Economic Growth in South Africa    125

6.8  Road Infrastructure Road infrastructure has been another sector in which policy uncertainty and contestation has resulted in declining investment. The decision to fund the upgrading of those parts of the national road network located in Gauteng triggered a de facto tax revolt as most Gauteng road users refused to pay. Even though the policy was designed so that business vehicles and private vehicle owners would contribute to the costs of the road network—​and that those using public transport such as mini-​bus taxis would not have to pay—​there was widespread rejection of the toll payments swept up in the anti-​ corruption sentiment of the period. Government commitment to the policy framework has wavered, with the Gauteng provincial government campaigning against the toll road policy. The consequence of such lingering policy uncertainty had devastating implications for the maintenance and development of South Africa’s wider road network—​and the important economic efficiency and trade and logistics that such a network provides. The situation has pushed up the indebtedness of the South African National Roads Agency Limited (Sanral) raising fears of a debt default which would have wider fiscal implications for the country. The failure of the Gauteng user-​pays model has meant that planned road upgrades in other provinces have not been undertaken and resources that would have been freed up for rural roads are not available. As of 2020, Sanral estimates a road maintenance backlog valued at approximately R150 billion (Barron 2020).

6.9 Telecommunications Infrastructure Despite the fact that worldwide the latest phase of industrialization is being driven by a revolution in digital technology which continues to shape economic activity and determine competitiveness, South Africa’s key enabling policy on the utilization of the electromagnetic telecommunications and broadcasting spectrum has been at a stalemate for over a decade. With regard to the release of the electromagnetic spectrum for the users of telecoms and data services, the ongoing delay in the auction of this spectrum has resulted in delays in the rolling-​out of more efficient and quicker 5G technology (McLeod 2019). Furthermore, the lack of supply of such a broadband spectrum has meant relatively high data prices for South African consumers and business, placing another constraint on economic growth. The competition authorities have sought to penalize the telecommunications companies operating in South Africa for the country’s relatively high costs of service, particularly for poorer consumers who purchase smaller packets of data, but at a more

126   Kenneth Creamer fundamental level it has been the limitation of supply of the electromagnetic spectrum that has pushed up prices. If markets operate competitively, effectively limiting pricing power, then the increased supply of the electromagnetic spectrum would have had the effect of reducing data prices, thereby releasing cost constraints on consumers and on business development, particularly small business development. The interests of an entrenched near monopoly in the broadcasting sector have effectively stymied the digitization of the country’s television signals, as the incumbent has sought to avoid the entry of any potential competitors into its market by campaigning against rivals being allowed to have the type of encryption technology that it uses itself to ensure that customers pay to access its services. Ironically, the speed of technological change is now beginning to surpass this incumbent, as subscriptions become web-​based, rather than based on satellite ​encryption-​based services, but nonetheless through the manipulation of government policy this incumbent was able to secure near-​monopoly profits for over a decade as it blocked the entry of others into its market. Policy uncertainty, the power of vested ​interests, and the prevalence of contradictory and unsustainable development models extended beyond the electricity, road, and telecommunications sectors, constraining growth and investment in a range of key infrastructure sectors, including, housing, water and sanitation, commuter rail, and rural development. The democratic state, upon which were political expectations and constitutional obligations to drive economic growth and transformation, faltered as it was not able to put in place sustainable strategies to expand infrastructure and foster long-​ term growth through a combination of fiscal commitments, user-​pays frameworks and guiding and disciplining local and foreign investors. Given the country’s history and the persistence of inequality and economic exclusion, the democratic state’s failure effectively to lead a process of inclusive development has in itself represented a formidable constraint to economic growth.

6.10  State Capacity and the Developmental State Through numerous policy frameworks, such as the country’s National Development Plan (2012), the post-​apartheid state has asserted a theoretical committment to state-​led development and to the building of the kind of capabilities and institutions reminiscent of East ​Asian ​miracle-​style developmental states (Johnson 1982 and Wade 1990). Despite this theoretical, or perhaps at times merely rhetorical, commitment to a decisive state role in guiding post-​apartheid economic growth and development, in practice there have been many failings, even to the extent that in the post-​apartheid South African context poor state capacity has emerged as a constraint to, rather than the driver of, economic growth.

Constraints to Economic Growth in South Africa    127 Notwithstanding this reality, there seems to be no alternative pathway, or historical precedent, that would indicate that South Africa can possibly develop into a more equal and prosperous society without a machinery of state which is capable of playing a guiding and developmental role. For there to be a sustained period of economic development will require a state capable of transforming the economy’s historically embedded structure of opportunity and removing entrenched structural constraints, in order to put the economy onto a new, more inclusive, growth path. Wade (2018) summarizes the key elements of the ‘developmental mindset’ as follows: • High and sustained economic growth rates, so as to catch up with developed countries ‘quickly’ (within a few decades) • High rates of investment to gross domestic product (GDP) so as to achieve rapid movement of the production structure into more productive sectors • The state to coordinate the catch-​up strategy and promote some sectors and functions ahead of others, whether through public enterprises or through steering private actors into sectors they would otherwise not enter • The state to curb the growth of consumption by the urban labour force and farmers, so as to free up more resources for investment • The state to promote exports intensively, so that the high investment could be profitable despite restrained growth of consumption at home; but at the same time, state industrial policy must target feasible replacement of imports and concentrate foreign exchange on imports of capital goods, intermediate goods, and raw materials (not consumer goods) by means of a managed trade policy, not free trade and not ‘neutral incentives’ between export promotion and import substitution • The state to invest heavily in education (people being the primary endowment, not natural resources), especially engineering, including sending students to the West for university education and putting them under obligation to return • The state to boost the take-​up of Western technology, including by establishing public R&D centres able to ‘domesticate’ technologies purchased or imitated from abroad or brought in by branches of Japanese or Western companies. Developmental state theory sees the state empowered to play a role in changing inherited economic structures. State ownership is combined with state action to shape and regulate market activity. This was regarded as providing superior outcomes to that of neo-​liberal policy which would most likely entrench current structures and racial inequality hence rendering the framework incapable of addressing the historical and structural realities of colonialism and apartheid. On the other hand, statist theory which seeks to maximize state ownership in the economy would lack the capacity for growth and dynamism and would tend towards centralization of power and corruption. The creation of a developmental state and the commitment to ‘inclusive growth’ seemed to resolve some of the policy tension on the balance to be struck between growth and redistribution. On the one hand, there was an argument that what was needed was ‘growth through redistribution’, which emphasized the need for policy interventions

128   Kenneth Creamer to change racialized patterns pre-​distribution and widen access to land, education, and business entry in order to generate a new growth dynamic in the South African economy. On the other hand, there was an argument in favour of ‘redistribution through growth’, which emphasized the need for policies aimed at promoting new investment and job creation so that the benefits of growth would become available to the wider, hitherto excluded population (Moll et al. 1991). The theory of the developmental state and the commitment to ‘inclusive growth’ recognized that both growth and redistribution need to be taken forward simultaneously and that policy has a role in shaping economic performance to achieve the reinforcing objectives of growth and redistribution. In practice, the role of a capable developmental state, inclusive growth, and a balance between growth and redistribution has not been easy to achieve, and at times has resulted in contradictory and constraining outcomes. Insights into the process of corruption and state capture have been offered by Von Holdt (2019) who recast the kind of corruption that saw state officials replace public development purposes with their own financial interests as a form of ‘primitive accumulation’ on the part of a nascent elite.17 The vision of a developmental state has been severely undermined by state capture and the corrupt practices of a new political class of insiders. Members of this elite may not have had opportunities in white-​dominated business and hence the corruption has the potential for the formation of a new elite class. However, this argument fails to fully foreground the constraint posed by such state capture and corruption as it repurposes the state’s role towards private accumulation and away from inclusive growth and development, a negative cycle which is accelerated if it is accompanied by falling levels of growth and investment, as has been the case in South Africa in recent years. Given apartheid’s strict limitations on Black business participation, post-​apartheid South Africa necessarily embarked on a process of de-​racializing ownership of the commanding heights of the economy. Through a process called Black Economic Empowerment (BEE) regulatory frameworks were put into place which required Black business ownership and other forms of Black participation in the formal economy. The issuing of new mining licences required a percentage of Black ownership of the firms concerned, preference was given to Black ownership in government and private procurement, and employment equity and training programmes favoured the advancement of Black people as managers and executives in the business sector. A critique of this process has been that it has been too narrowly focused, with only a small number of Black business leaders benefiting from the BEE process. Also, over time, as Black shareholders sought to sell their shares in particular companies, these companies came to be in danger of losing their BEE credentials. The BEE process, while politically and historically a necessary form of corrective intervention, served to entrench the insider–​outsider character of much of the South African economy. Cronyism and political connectedness played into the dynamic whereby a small number of individuals were given privileged access to business profits and frequently

17 

Von Holdt (2019: 17).

Constraints to Economic Growth in South Africa    129 these were in the same concentrated, uncompetitive business structures in which firms had pricing power—​entrenching overall inequality, although not necessarily racial inequality, in the wider society. The state has not been able to develop the kind of ‘embedded autonomy’ (Evans 1995) required from the new BEE class, historical white capital, or even to an extent the trade unions necessary to be a technocratically effective developmental state capable of planning in the long-​term national interest and free of short-​term insider, rent-​seeking, and vested interests. The state has not been capable of rising above distributional conflicts and mapping out relatively ‘autonomous’ growth strategies. This has weakened the planning and policymaking capacity of the post-​apartheid state to a significant degree. Furthermore, the state will not be in a position to lead a process of social compacting until it is able to achieve a sufficient degree of ‘embedded autonomy’ and has the capability of developing and implementing plans in the national interest to simulate inclusive growth and national development.

6.11  Policy Perspectives—​O vercoming South Africa’s Growth Constraints Moving beyond diagnosis, the chapter concludes with a focus on policy perspectives which seek to address and overcome South Africa’s growth constraints. First, the fundamental competitiveness of the South African economy needs to be re-​established. The MEC-​based growth structure, which has linked the economy’s fortunes with the global commodity prices cycle, should over time be replaced with new structures of competitiveness and innovation. At its techno-​economic heart, South Africa’s historical MEC must be replaced by new drivers of economic growth. Fundamental to this vision would be the establishment of a new green energy complex, which will seek to restore the country’s competitive advantage through lower-​cost, lower-​carbon, and reliable electricity supply. For a fuller analysis of the potential for a green transition in South Africa, see ­Chapter 16 in this volume. To be successful, the business model for energy investment requires the payment of cost-​reflective tariffs by industrial users. As South Africa has world-​class wind and solar assets (Bischof-​Niemz and Creamer 2019), this will put downward pressure on South Africa’s future electricity prices and assist in promoting export competitiveness. This process would require the restructuring of Eskom into three distinct entities—​each respectively responsible for electricity generation, transmission, and distribution. Following such a restructuring, increased competition will bring down generating costs and a national transmission company will play a strategic role in facilitating the energy transition and introducing new levels of efficiency and competitiveness. The alignment of South Africa’s industrial policy with the programme of electricity infrastructure investment would allow integrated exploitation of upstream and

130   Kenneth Creamer downstream industrial linkages to the green energy complex, including through localization of renewable-​energy technology and through the promotion of new export-​ oriented green industries. This should include the accelerated implementation of Renewable Energy Development Zones—​ including in coal-​ producing areas—​ to streamline environmental authorization processes and assist in the establishment of new upstream and downstream renewable-​energy industries in such zones. The correct alignment of South Africa’s industrial and energy policy in such a manner would unleash significant investment and employment potential and would be a key driver to overcoming the structural constraints currently weighing down the country’s rate of economic growth. Second, South Africa will need to restore the country’s external macroeconomic balance through export promotion and its fiscal balance by bringing national debt under control through growth-​enhancing infrastructure investment. Broadly speaking, public spending can be broken into four main categories: wages of public-​sector employees, welfare transfers, debt service costs, and infrastructure spending. In a context of rising public-​sector wages, welfare transfers, and debt service costs, at the same time as poorly performing government revenue over a number of years, spending on infrastructure has tended to be ‘crowded-​out’.18 This is most concerning as the prioritization of public spending on infrastructure and facilities is key to achieving inclusive growth. Spending on assets, such as school and health facilities, textbooks, sports equipment, libraries, toilets, clinics, scanners and x-​ray machines, as well as on municipal and provincial services, roads, and transport infrastructure serves to increase the social wage, particularly for those in poorer communities. Employment programmes linked to infrastructure investment are preferable to the expansion of social grants, as such programmes offer work experience, transfer skills, and capabilities, and result in the expansion of community services and assets. As such, South Africa’s limited public finances should be prioritized and be focused on the building of social and economic infrastructure that will increase the social wage for poor communities and facilitate inclusive growth and job creation. Three mutually reinforcing interventions are required for fiscal stabilization—​the control of the budget deficit, the management of borrowing costs, and the acceleration of economic growth. Although politically the most challenging, control of the budget deficit is the policy lever most directly controlled by government as it can be achieved through a combination of expenditure control, reprioritization, and tax policies. The management of borrowing costs, which are ultimately determined not directly by government, but in local and global capital markets, requires the implementation of an

18  As per South Africa’s Budget Review 2020 (South African National Treasury 2020a), ‘Government recognises that public-​service employees should be fairly remunerated, but is obligated to balance compensation demands with the broader needs of society as reflected in the budget. Civil servants’ salaries have grown by about 40 per cent in real terms over the past 12 years, without equivalent increases in productivity. Growth in the wage bill has begun crowding out spending on capital projects for future growth and items that are critical for service delivery.’

Constraints to Economic Growth in South Africa    131 effective deficit reduction strategy, impactful pro-​growth reforms, and credible monetary policy. Accelerating the rate of economic growth will require that government implement a focused economic recovery strategy based on a specific set of pro-​growth reforms capable of stimulating significantly increased levels of private-​sector investment and job creation. Decisive interventions are urgently needed to rekindle and accelerate growth. Such interventions must be designed to achieve an ‘inclusive’ form of growth, capable of overcoming apartheid’s persistent, racialized, structural exclusions. The chief instruments for driving ‘inclusive growth’ are, first, sectoral reforms aimed at unlocking investment in infrastructure and business activity across a wide range of sectors, and second, budget reprioritization aimed at improving access to public infrastructure and services, effectively increasing the social wage. Accelerated implementation of policies to unlock pent-​up demand for expanded investment in electricity-​generation capacity—​particularly lower-​cost and more ​rapidly deployed wind, solar, and gas generation technologies—​would serve to stimulate significant direct and indirect employment and BEE opportunities, as well as create the energy security which is a necessary precondition for investment and job-​creating activities across a wide range of economic sectors. Similarly, a correctly designed and executed release of the telecommunications spectrum would set in motion a process of lowering data costs and increasing access to data services—​putting money into consumers’ pockets and allowing for increased employment and business activity including township, village, and small business activity. Comparable reforms aimed at stimulating inclusive growth should be undertaken across a range of sectors, including mining and mining exploration, manufacturing, and agriculture and agro-​processing. Third, the capability of South Africa’s state and public service must be improved so that it can begin to play a role in guiding and leading inclusive economic growth. Skilled and ethical public servants are needed for this task. As stated in the National Development Plan (2012), in order to build a capable and developmental state, policy must inter alia: stabilize the political–​administrative interface through establishing a professional public service that is ‘sufficiently autonomous to be insulated from political patronage’; and make the public service a career of choice where ‘recruitment and management should be based on experience and expertise’, and where the state is able to develop and reproduce ‘the specialist and technical skills to fulfil its core functions’.19 Such policies would enable the state to be ‘developmental’ and to guide service delivery and economic growth and development with the required degree of ‘embedded autonomy’ (Evans 1995). A developmental state must be ‘embedded’ in the sense that the state is knowledgeable about the workings, strengths, and weaknesses of the economy and has the technical capability to articulate a transformative vision and carry out such plans. It should have ‘autonomy’ in the sense that the state is not captured by any particular foreign or domestic vested interests but is able to articulate and drive a

19 

NDP (2012: 410).

132   Kenneth Creamer programme of national economic development in the national interest. Any vision for social compacting by government, business, labour, and the wider community to advance such a development programme will require, as a precondition, that government is capable of developing and implementing such a programme of national economic development, as this is not a role that can be played by any of the other social partners. As has been the case in other countries that have pulled themselves out of poverty through sustained periods of economic growth, South Africa will require the apparatus of an effective developmental state capable of guiding the growth and development of key sectors and network industries if it is to overcome the country’s structural growth constraints. Furthermore, as a consequence of the country’s particular colonial and apartheid history, an effective developmental state in the South African context will also be required to re-​shape service delivery—​and undertake redistributive programmes—​ in order to lead the kind of ‘inclusive growth’ process that will be needed to overcome the country’s enduring racialized and gendered patterns of inequality of income, wealth, and opportunity.

References Acemoglu, Daron, and James Robinson. 2012. Why Nations Fail—​The Origins of Power, Prosperity, and Poverty. New York: Crown Publishing Group. Adelzadeh, Asghar. 1996. ‘Growth and development: Labour and business perspectives on economic development’, in Jeremy Baskin (ed.) Against the Current: Labour and Economic Policy in South Africa. Randburg: Ravan Press. Arrighi, Giovanni, Nicole Aschoff, and Ben Scully. 2010. ‘Accumulation by dispossession and its limits: The Southern African paradigm revisited’, Studies in Comparative International Development, 45: 410–​38. Barron, Chris. 2020. ‘Sanral runs out of road on e-​toll indecision’, Sunday Times, 20 September [online] https://​www.timeslive.co.za/​sunday-​times/​business/​2020-​09-​20-​sanral-​runs-​out-​ of-​road-​on-​-​-​-​-​e-​toll-​indecision/​. Bischof-​Niemz, Tobias, and Terence Creamer. 2019. South Africa’s Energy Transition—​A Roadmap to a Decarbonised, Low-​cost and Job-​rich Future. Oxford: Routledge. Brynjolfsson, Erik, and Andrew McAfee. 2011. The Race against the Machine: How the Digital Revolution Is Accelerating Innovation, Driving Productivity, and Irreversibly Transforming Employment and the Economy. Lexington, MA: Digital Frontier Press. Bundy, Colin. 2016. Poverty in South Africa Past and Present. Johannesburg: Jacana Media. Burger, Philippe. 2015. ‘Wages, productivity and labour’s declining income share in post-​ apartheid South Africa’, South African Journal of Economics, 83(2): 159–​73. CSIR. 2020. ‘Setting up for the 2020’s: Addressing South Africa’s electricity crisis and getting ready for the next decade . . . and now Covid-​19’. Centre for Scientific and Industrial Research Report (August). Cuddington, John, and Daniel Jerrett. 2008. ‘Super cycles in real metals prices?’, IMF Staff Papers, 55(4): 541–​65. Evans, Peter. 1995. Embedded Autonomy—​States and Industrial Transformation. Princeton, NJ: Princeton University Press.

Constraints to Economic Growth in South Africa    133 Feinstein, Charles. 2005. An Economic History of South Africa—​Conquest, Discrimination and Development. Cambridge: Cambridge University Press. Fine, Ben, and Zavareh Rustomjee. 1996. The Political Economy of South Africa—​From Minerals-​Energy Complex to Industrialisation. Johannesburg: Wits University Press. Hirsch, Alan. 2005. Season of Hope—​ Economic Reform under Mandela and Mbeki. Pietermaritzburg: University of KwaZulu-​Natal Press. Jacks, David. 2019. ‘From boom to bust: A typology of real commodity prices in the long run’, Cliometrica, 13: 201–​20. Johnson, Chalmers. 1982. MITI and the Japanese Miracle: The Growth of Industrial Policy, 1925–​ 1975. Stanford, CA: Stanford University Press. Kahn, Brian. 1991. ‘The crisis and South Africa’s balance of payments’, in Stephen Gelb (ed.) South Africa’s Economic Crisis. Claremont: David Philip. Lipton, Merle. 1985. Capitalism and Apartheid. Aldershot: Wildwood House. Macroeconomic Research Group (MERG). 1993. Making Democracy Work: A Framework for Macroeconomic Policy in South Africa. University of the Western Cape, Bellville: Centre for Development Studies (CDS). Marais, Hein. 1998. South Africa: Limits to Change: The Political Economy of Transformation. Cape Town: University of Cape Town Press. Marquard, Leo. 1952. Peoples and Policies of South Africa. London: Oxford University Press. Mcleod, Duncan. 2019. ‘No reason 5G spectrum can’t be allocated now: Vodacom’, Tech Central, 15 August [online] https://​techcentral.co.za/​no-​reason-​5g-​spectrum-​cant-​be-​allocated-​ now-​vodacom/​91825/​. Moll, Peter, Nicoli Nattrass, and Lieb Loots. 1991. Redistribution—​How Can It Work in South Africa. Cape Town: David Philip. National Development Plan. 2012. National Development Plan 2030—​Our Future, Make It Work. https://​www.gov.za/​sites/​default/​files/​gcis_​document/​201409/​ndp-​2030-​our-​future-​ make-​it-​workr.pdf. OECD. 2008. ‘South African economic assessment’, OECD Economic Surveys 2018/​15, July. Piketty, Thomas. 2014. Capital in the Twenty-​first Century. Cambridge, MA: Harvard University Press. Rodrik, Dani. 2018. Straight Talk on Trade—​Ideas for a Sane World Economy. Princeton, NJ: Princeton University Press. Sachs, Michael. 2020. Fiscal dimensions of South Africa’s crisis [Public Economic Project SCIS Working Paper, Number 15, Southern Centre for Inequality Studies University of the Witwatersrand. www.wits.ac.za/scis (March 2021)]. South African National Treasury. 2020a. ‘Budget review 2020’. 26 February. South African National Treasury. 2020b. ‘Supplementary budget review 2020’. 24 June  Spaull, Nic, and Elizabeth Pretorius. 2019. ‘Still falling at the first hurdle: Examining early grade reading in South Africa’, in Nic Spaull and Jonathan Jansen (eds) South African Schooling: The Enigma of Inequality. Dordrecht: Springer. Statistics South Africa. 2020. Quarterly Labour Force Survey (QLFS). Second Quarter 2020, StatsSA: P0211, Pretoria. The Constitution of the Republic of South Africa. 1996. Act 108 of 1996. Tregenna, Fiona. 2008. ‘Characterising deindustrialisation: An analysis of changes in manufacturing employment and output internationally’, Cambridge Journal of Economics, 33: 433–​66. Von Holdt, Karl. 2019. ‘The political economy of corruption: Elite formation, factions and violence’. Working Paper 10 February, Society, Work and Politics Institute, Johannesburg.

134   Kenneth Creamer Wade, Robert. 1990. Governing the Market: Economic Theory and the Role of Government in East Asian Industrialisation. Princeton, NJ: Princeton University Press. Wade, Robert. 2018. ‘Escaping the periphery—​The East Asian “mystery” solved’. WIDER Working Paper no. 2018/​101, Helsinki. Wolpe, Harold. 1972. ‘Capitalism and cheap labour power in South Africa: From segregation to apartheid’, Economy and Society 1(4): 425–​56. Wolpe, Harold. 1988. Race, Class and the Apartheid State. DTrenton, NJ: Africa World Press.

Chapter 7

U nem pl oyme nt i n Sou th A fri c a James Heintz and Karmen Naidoo

7.1 Introduction A remarkable feature of South Africa’s labour market is the persistently high level of unemployment. Since the end of apartheid, when reliable statistical measurements became available, the country has exhibited rates of open unemployment that are extreme by any comparative standard. Wage employment is the single most important source of income for most South Africans. Therefore, the lack of access to decent employment opportunities contributes to ongoing and deep-​seated economic inequalities (Leibbrandt et al. 2010). These disparities in employment outcomes reflect broader vulnerabilities across multiple dimensions: race, gender, age, educational attainment, and location. The extreme problem of joblessness also has shaped debates over the desirability of post-​apartheid labour market and social protection policies, such as the national minimum wage, which came into effect on 1 January 2019 and which aims to increase the wage floor for workers not covered by other wage regulations and for those under sectoral wage determinations deemed too low (Elsley 2019).1 Understanding the nature and causes of unemployment in South Africa is therefore essential when considering the country’s future trajectory. This chapter examines the reasons behind South Africa’s high unemployment rates. After an analysis of unemployment trends and patterns, it discusses alternative explanations of the country’s employment problems, with a focus on structural causes arising from historical and institutional factors. It also looks at microeconomic explanations of open unemployment based on the functioning of the labour market.

1 

The national minimum wage was initially set in 2019 at R20 per hour for general workers, R18 per hour for farmworkers, and R15 per hour for domestic workers.

136    James Heintz and Karmen Naidoo The chapter further examines how policy choices post-​apartheid have affected employment outcomes, including macroeconomic policies, trade policies, and labour market policies and concludes with a summary of the most important take-​away messages from this review.

7.2  Unemployment Trends and Patterns Unemployment rates are calculated as the number of unemployed persons expressed as a percentage of the total labour force. In South Africa, two unemployment rates are commonly used: a narrow, or strict, definition of unemployment and a broad, or expanded, definition. Under the strict definition of unemployment, an individual is only included in the ranks of the unemployed and as part of the labour force if they have actively been seeking employment over the seven days prior to being surveyed. Under the expanded definition, unemployed individuals who are no longer actively seeking employment, but who are willing and able to work if a paid job became available, are included in unemployment and labour force numbers. The inclusion of discouraged work-​seekers and other non-​searching unemployed individuals means that the broad unemployment rate is significantly higher than the narrow unemployment rate. Reliable labour market statistics do not exist for the apartheid era in South Africa. Therefore, we only examine trends for the post-​apartheid period (from 1994 onward). At the end of apartheid, unemployment rates were high—​around 30 per cent of the labour force using the expanded definition.2 Since then, the number of unemployed has grown significantly, even by the narrow definition. For instance, the number of unemployed individuals, using the narrow definition, has increased from about 2.4 million in 1994 to 6.7 million in 2019—​an almost three f​ old increase. When discouraged work-​ seekers are included, the number of unemployed doubles from 4.6 million in 1994 to 9.5 million in 2019. Depending on the definition of unemployment used, aggregate unemployment rates ranged between 25 and 40 per cent of the labour force for most of the post-​apartheid period (Figure 7.1). A portion of the initial increase in unemployment in the late 1990s has been attributed to a rapid rise in labour force participation, particularly among African women (Casale and Posel 2002). The removal of influx controls that prohibited the permanent urban settlement of Africans during apartheid was associated with increased mobility in the immediate post-​apartheid period, particularly for African women (Posel and Casale 2003). In this period, employment creation did not keep up with the growth in the size of the labour force. Since 2000, labour force participation rates have remained fairly stable and aggregate unemployment showed signs of improvement over the 2000s.

2  The

broad labour force is defined as individuals aged 15–​64 who are either employed, strictly unemployed or non-​searching unemployed.

Unemployment in South Africa    137 80 70 60 50 40 30 20 10 0 1993

1997

2001

2005

2009

2013

2017

Narrow unemployment rate (% of LF) Broad unemployment rate (% of LF) Labour force participation rate (broad, % of WAP) Labour force participation rate (narrow, % of WAP)

Figure 7.1  Labour force participation and unemployment rates, 1993–​2019 Source: Authors’ calculations using PALMS (Kerr, Lam, and Wittenberg 2019). Notes: The narrow unemployment rate is calculated as the number of unemployed divided by the size of the active labour force (LF), excluding non-​searching unemployed. The broad unemployment rate includes non-​searching unemployed. WAP stands for working age population, defined as individuals aged 15–​64. The narrow labour force participation rate excludes non-​searching unemployed from the size of the labour force.

However, the 2008/​09 global financial crisis and subsequent recession had a significant impact on the South African labour market. As a result, some 800,000 net jobs were lost over the 2009–​10 period and since then, unemployment has increased. An important effect of the recession was an increase in the number of discouraged workers (Verick 2012). Searching and non-​searching unemployment increased more during the recession for African men with low levels of education relative to women with similar characteristics (Verick 2012). Patterns of unemployment in South Africa reflect broader economic inequalities, including along the lines of gender, race, age, and location. Table 7.1 presents unemployment rates, using the expanded definition, by social and demographic characteristics. Unemployment rates among women have remained substantially higher than for men over the twenty-​five years since the end of apartheid. There has been a narrowing of the gender unemployment gap during the post-​apartheid period, which can be attributed to the relatively more rapid rise in unemployment rates among men. As a legacy of apartheid, unemployment rates continue to be structured along racial lines: non-​white groups continue to face systematically higher rates of unemployment in South Africa. In particular, unemployment among Africans has remained the highest at 36 per cent in 1995, rising to 41 per cent in 2019. Coloureds experience the second highest rates of unemployment, at 22 per cent in 1995 rising to 27 per cent by 2019.

138    James Heintz and Karmen Naidoo Table 7.1: Broad unemployment rates by demographic characteristics and educational attainment 1995

2000

2005

2010

2015

2019

Women

38.16

38.65

44.75

35.69

36.74

40.28

Men

21.74

28.17

30.90

29.50

29.60

33.19

African

36.18

39.22

43.51

37.44

36.98

41.10

Coloured

21.73

25.01

30.69

26.48

28.04

26.68

Indian

13.68

19.73

22.31

12.18

17.84

13.36

White

5.27

7.52

7.96

8.14

7.60

8.95

15–​34

38.07

45.49

50.31

44.32

43.77

48.75

35–​49

21.08

22.95

26.09

21.78

24.07

27.82

50–​64

15.48

14.29

16.71

14.36

15.60

18.07

High school or less

34.89

36.66

41.48

36.45

36.80

40.71

Post-​schooling diploma or degree

22.97

12.56

12.13

12.43

15.45

18.60

Gender:

Race:

Age groups:

Education:

Source: Authors’ calculations using PALMS (Kerr et al. 2019). Notes: Broad unemployment rates include discouraged work-​seekers.

Excessively high rates of youth unemployment have persisted over time—​in 2019 almost half of those aged 15–​34 who are willing and able to work could not find paid employment. A closer examination of youth unemployment is provided in ­Chapter 32 of this volume. Educational attainment also influences the risk of joblessness. Among individuals with high school completion or less, the unemployment rate has increased from 37 per cent in 2000, to 41 per cent in 2019. Comparatively, among individuals with a post-​schooling diploma or degree, unemployment has increased from 13 per cent to 19 per cent over the same period.

7.3  Structural Causes of Unemployment What might explain these extremely high and persistent rates of unemployment in South Africa? The economist Edmond Malinvaud produced a framework for thinking

Unemployment in South Africa    139 about the various factors contributing to unemployment that can be usefully applied to South Africa (Malinvaud 1978). In his typology, unemployment that can be attributed to a lack of aggregate demand for goods and services is referred to as Keynesian unemployment. Under Keynesian unemployment, the supply of goods and services is rationed due to inadequate levels of demand and this leads to a shortfall in employment opportunities. In contrast, unemployment that can be attributed to insufficient accumulation of fixed capital is labelled classical unemployment (or alternatively Marxian unemployment). In this scenario, a lack of investment, or overly capital-​intensive investment, limits employment opportunities. Even when the productive capacity of the economy is fully utilized, there is excess supply of labour, since the economy fails to generate an adequate number of paid jobs. In Malinvaud’s framework, involuntary unemployment, either ‘Keynesian’ or ‘classical’, has multiple causes and is structural in the sense that market prices do not seamlessly adjust to eliminate joblessness (Malinvaud 1977, 1978). We expand on this foundational approach to disaggregate the factors that determine the total number of paid employment opportunities. When these jobs are not sufficient to absorb the entire labour supply, unemployment results. Within this framework, we can disaggregate the determinants of total employment into four factors (see 7.6 Appendix for a more formal presentation): • The productive fixed capital stock • The output–​capital ratio (an indicator of capital intensity) • Labour productivity (the amount of output produced with a given amount of labour) • Capacity utilization (a measure of aggregate demand—​actual output expressed as a fraction of potential, or full-​employment, output). Structural unemployment can be explained by a combination of these factors: insufficient accumulation of fixed capital, adoption of capital-​ intensive production technologies (i.e. a falling output–​capital ratio), growth in labour productivity, or insufficient aggregate demand. We look at each of these potential sources of unemployment in the context of the South Africa economy. The idea that insufficient capital accumulation can yield an excess supply of labour and involuntary unemployment is not new. It is central to Lewis’s classic theory of labour surplus economies (Lewis 1954). Low rates of investment mean that, even if productive capacity is fully utilized, there will not be enough paid jobs to absorb the labour supply. Researchers have argued that low rates of capital accumulation during the final two decades of apartheid contributed to high rates of involuntary unemployment in the post-​apartheid era (Heintz 2009). Moreover, others have noted that the systematic underdevelopment of the apartheid-​era homelands and restrictions that prevented investments in informal businesses also limited opportunities for paid work (Kingdon and Knight 2004). The relatively small size of the informal sector in the post-​apartheid

140    James Heintz and Karmen Naidoo 7

Annual percentage change (%)

6 5 4 3 2 1

68 19 71 19 74 19 77 19 80 19 83 19 86 19 89 19 92 19 95 19 98 20 01 20 04 20 07 20 10 20 13 20 16 20 19

65

19

62

19

59

19

56

19

53

19

19

19

50

0

Figure 7.2  Growth rate, fixed capital stock, South Africa 1950–​2019 Source: South African Reserve Bank.

period is further seen to contribute to the sustained high levels of unemployment over time (Heintz and Posel 2008; Kingdon and Knight 2004).3 Figure 7.2 shows the growth rate of the aggregate fixed capital stock in South Africa from 1950 to 2019. The growth rate of fixed capital reflects net investment or capital accumulation in the country. Rates of capital accumulation were high during the 1950s, 1960s, and part of the 1970s. However, beginning in the mid-​1970s, the growth rate of fixed capital began to decline sharply. For approximately twenty years, from the mid-​ 1980s to the mid-​2000s, rates of fixed capital accumulation have been very low by historical standards. Since 2006, there has been a recovery in the growth rate of the capital stock. Nevertheless, the long period of low rates of investment would have impacted the economy’s ability to generate paid jobs. Others have stressed the type of investments made, not simply the overall rate of accumulation. Specifically, overly capital-​ intensive production technologies limit the number of jobs created by an additional unit of capital investment (Nattrass and Seekings 2012). Despite high rates of unemployment, production in South Africa remains highly capital-​intensive. Counter-​intuitively, market production economizes on the country’s abundant factor of production, that is, its labour force. There are numerous reasons given for this paradox. During the apartheid years, priority was given to the goal of generating sufficient decent jobs for the minority white population while maintaining a surplus of labour elsewhere in the economy. This outcome could be achieved by supporting capital-​intensive investments through policies that effectively 3 

Further discussion of the informal sector in South Africa is provided in C ­ hapter 35 of this volume.

Unemployment in South Africa    141 lowered the cost of capital. With the growing strength of organized labour, beginning in the 1970s, firms had an incentive to adopt capital-​intensive production techniques to avoid the ‘hassle’ of employing a labour force whose bargaining power was growing. Others have stressed the role of labour market institutions, such as bargaining councils, in creating a capital-​intensive bias (Nattrass and Seekings 2012). A critical form of investment, omitted from the Malinvaud-​inspired framework laid out here, is the development of the productive capacities of human beings. Not only was there underinvestment in physical capital during the later years of apartheid, South Africa also underinvested in the country’s human resources for decades. Vast historical inequalities with regard to access to education and training continue to persist within the labour market. The unequal investments in human capital represent another structural explanation of unemployment. South Africa exhibits both high rates of open unemployment and a skills shortage (Davies and van Seventer 2020). As shown previously, unemployment rates are inversely related to educational attainment and joblessness is significantly more prevalent among the less educated. Research has shown that young South Africans who complete secondary education are much more likely to be able to transition from school into paid employment (Lam, Leibbrandt, and Mlatsheni 2010). Along with physical and human investments, changes in labour productivity affect employment levels. Higher labour productivity means that less labour is required to produce a given amount of output. If output is fixed and labour productivity is growing, employment will necessarily fall. More generally, if output grows at a slower rate than labour productivity, jobs will be lost. The impact of growing labour productivity on paid jobs in South Africa has been documented with respect to patterns of deindustrialization (Tregenna 2009, 2011; Rankin 2012). Within South Africa, manufacturing employment as a share of total employment has fallen much more rapidly when compared to manufacturing output as a share of total output. This suggests that efforts to increase sectoral output may not result in an equivalent increase in paid jobs. The explanation for these patterns lies in the dynamics of labour productivity—​with productivity growth outstripping the growth in output. Increases in labour productivity are reflected in the elasticity of employment with respect to output—​that is, the percentage increase in employment associated with a given percentage change in output. If labour productivity does not change, a doubling of output should yield a doubling of employment (the elasticity would be equal to 1, all things equal). With increases in labour productivity, less labour is needed to produce a given amount of output and a doubling of output would be associated with less than a doubling of employment (the elasticity would fall below 1). Table 7.2 presents estimates of the elasticity of employment with respect to output during the post-​apartheid period.4 The elasticities are particularly low in mining and manufacturing, suggesting

4 

We calculate output-​employment elasticities for the aggregate economy and by sector for the 1994–​ 2019 period, according to the following equation:

lnLt = β0 + β1lnYt + ε t

142    James Heintz and Karmen Naidoo Table 7.2: Output-​employment elasticities 1994–​2019 Mining and quarrying

0.568

Manufacturing

0.440

Utilities

0.816

Construction

0.954

Wholesale and retail trade

0.952

Transport, storage, and communications

0.659

Finance, insurance, and real estate

1.288

Community, social, and personal services

1.182

Total GDP

0.852

Source: Authors’ estimations using PALMS and South African Reserve Bank online statistics. Notes: These estimates represent regression-​based elasticities. Constant and standard errors are not shown. Sector fixed effects are used for the total GDP-​employment elasticity.

that increases in productivity have limited the numbers of jobs created for a particular increase in output. Finally, inadequate demand contributes to open unemployment in the standard Keynesian fashion. Specifically, actual output will fall below potential output when producers are limited with regard to how much they can sell. In this case, rationing of sales in the final markets for goods and services leads to a rationing of employment opportunities. Policy changes, such as trade liberalization, that increase competitive pressures and displace domestic production, can also contribute to a decline in aggregate demand. The relationship between such policies and unemployment in South Africa is discussed later in this chapter. In addition, vast income inequalities, such as those evident in South Africa, limit market development and domestic demand (Stiglitz 2012). These aggregate changes are only part of the story of a shifting structure of employment. Changes in the composition of employment are also apparent at the sectoral level (Table 7.3). During the post-​apartheid period, the share of employment in primary and secondary industries, including agriculture, mining, and manufacturing, declined. Employment shifted into construction and certain service sectors, including retail and

Where lnLt is log employment, lnYt is log real GDP at 2010 constant prices, and ε t is the error term. We summarize the results in Table 7.3. For the analysis in this chapter, sectoral data on real output and the capital stock are obtained from the South African Reserve Bank. Sectoral employment and wages are obtained from PALMS, which is the primary source of labour market data and since these data are not available from the South African Reserve Bank at a sufficiently disaggregated level (i.e. for the nine main sectors).

Unemployment in South Africa    143 Table 7.3: Sectoral shares of employment ∆(1995–​2019) % pts.

Industry

1995

2005

2019

Agriculture, forestry, and fisheries

13.21

7.21

5.06

–​8.15

4.67

3.19

2.31

–​2.36

14.93

13.77

10.96

–​3.97

0.9

0.8

0.93

0.03

Mining and quarrying Manufacturing Utilities Construction Wholesale and retail trade

4.7

7.61

8.33

3.63

17.6

24.22

21.14

3.54

Transport, storage, and communications

4.89

4.93

6.05

1.16

Finance, insurance, and real estate

6.18

10.65

15.62

9.44

22.17

17.91

22.01

–​0.16

8.66

9.48

7.55

–​1.11

Community, social, and personal services Domestic services Source: Authors’ calculations using PALMS.

Notes: Community, social, and personal services relate primarily to public-sector employment.

wholesale trade and the financial sector. These changes in the structure of employment represent a shift away from what was once the traditional base of the South African economy into new types of employment. While the framework presented here helps clarify the structural sources of unemployment in South Africa, it is important to recognize that these factors interact with one another. For instance, increases in labour productivity are employment-​displacing, but only if they do not affect investment or aggregate demand. Productivity increases have the potential to raise profitability and returns to investments in ways that could support the overall rate of capital accumulation. Higher productivity can also lower unit labour costs, enhancing trade competitiveness and the demand for domestically produced output. Sustainable long-​run wage increases are also associated with improvements in productivity over time. As labour income grows, so does the potential aggregate demand. These interconnections need to be considered when analysing South Africa’s unemployment challenge. This overview provides a number of insights into the structural sources of unemployment in South Africa. To sum up: uneven and slow rates of capital accumulation limited the development of productive capacity in the country. Many of the productive investments made had a capital-​intensive bias to provide a relatively small number of quality jobs to a privileged minority. Underinvestment in the country’s human potential further compounded the effect on paid job opportunities. At the same time, labour productivity has been increasing faster than the demand for output, particularly in industrial sectors, leading to labour displacement. Such increases in labour productivity could be driven, in part, by growing competitive pressures due to import penetration

144    James Heintz and Karmen Naidoo as South Africa reintegrated into the global economy (Rankin 2012). Export competitiveness, import penetration, and constraints on the development of domestic markets also exacerbate unemployment by curtailing aggregate demand. Taken together, these structural factors contribute to South Africa’s high and persistent levels of unemployment. A historic perspective further clarifies the origins of these structural sources of unemployment. The apartheid labour market system created a large, inexpensive, and relatively well-​disciplined low-​skilled labour force. Outside of the white population, job opportunities were limited and rationed through various mechanisms, such as the ‘colour bars’, that reserved the best jobs for the white labour force, the pass laws (i.e. influx controls), that regulated residency rights in urban areas where workers could find industrial wage employment, and unequal access to education and training. Many regions in the country, such as the homelands, were systematically excluded from the apartheid development model, leading to gross under ​investment. The capital-​intensive bias, partly due to an artificially low cost of capital, further limited opportunities for paid work. Large labour surpluses and repression of organized labour, particularly in the first two and a half decades of apartheid, kept wages low. The laws limiting access to job opportunities, the underdevelopment of areas not dominated by whites, and unequal investments in the population’s productive potential raised the cost of job loss to Black workers. If workers lost their jobs, the apartheid labour market system made it difficult to find equivalent work elsewhere. The threat of job loss thereby served as a disciplinary device, helping to secure the productivity of this segment of the labour force and the profitability of the apartheid economy, particularly in the 1950s and 1960s (Heintz 2009; Wintrobe 1998). However, the system was predicated on maintaining insufficient numbers of decent jobs for the majority of the population. While it seems obvious that high rates of unemployment would emerge from the apartheid economic system, it is less clear why unemployment continues to persist at such elevated levels in the post-​apartheid period. One explanation is that the apartheid system, through a range of interventionist policies, created an economic structure that is slow to change and is not self-​correcting. Simply removing the apartheid-​era laws and distortions is not sufficient to transform these economic structures. The discussion of structural causes suggests that the key to South Africa’s unemployment challenge lays beyond the labour market and involves policies to support investments, appropriate technologies, skills gaps, and market development. This is not to say that labour market policies are irrelevant. Clearly, they affect employment levels and trends. But solving South Africa’s unemployment problem would involve far more than relying on wage and price adjustments. Nevertheless, the dynamics of the South African labour market cannot be dismissed when examining the country’s serious unemployment problem. For this reason, we look at these microeconomic dynamics, focusing on labour market rigidities, next.

Unemployment in South Africa    145

7.4  Microeconomic Perspectives The structural causes of unemployment in South Africa represent one set of factors contributing to the persistence of high rates of joblessness. However, alternative explanations have been advanced that are strongly rooted in neoclassical economic theory. These alternatives focus on the possibility that microeconomic dynamics are behind South African unemployment. On the demand side, these approaches argue that labour market rigidities keep labour costs artificially high, lower labour demand, and hinder job creation. There are also supply-​side microeconomic theories based on individual preferences and characteristics—​for example, high reservation wages due to unrealistic earnings expectations and a system of social grants lead to a situation in which individuals remain unemployed rather than accepting a job that falls short of these expectations. We review these demand-​side and supply-​side microeconomic theories in turn. The labour market rigidities explanation of unemployment in South Africa points to various factors that prevent labour costs in general, and wages specifically, from adjusting in the face of high rates of unemployment: collective bargaining agreements, bargaining council agreements, regulation of working conditions (such as those contained in the Basic Conditions of Employment Act of 1997), and sectoral minimum wages. There is a fairly sizeable body of research that suggests that these sources of labour market rigidity reduce the overall level of employment by dampening labour demand (e.g. Flowerday, Rankin, and Schöer 2017; Bhorat, Kanbur, and Stanwix 2014; Magruder 2012; Nattrass and Seekings 2012; Fedderke 2012; Moll 1996). Labour demand is reduced when market interventions and collective agreements raise the cost of employing labour. Labour market regulations and bargaining agreements are not the only source of rigidities in labour markets. Asymmetric information and imperfect markets can also generate wage premiums (Shapiro and Stiglitz 1984; Bowles 1985). This occurs when wages influence productivity, often referred to as efficiency wages (Solow 1979). Workers are likely to put in more effort when they fear losing a good job. Employers typically contract for a given volume of work (e.g. hours of work), but, in many cases, may not be able to contract for a particular level of effort. By paying a wage premium, employers can elicit more effort from their workforce, since workers face a higher cost of losing a well-​paying job if the employer discovers that they are not putting in sufficient effort and dismisses them. These dynamics produce wage rigidities even in the absence of collective bargaining or legislation. If labour market rigidities stifle labour demand, then lowering wages should increase employment and lower the measured rate of unemployment. This raises the question: how responsive is labour demand to changes in wages? Estimates of wage elasticities provide a partial answer to this question. Wage elasticities measure the responsiveness of employment to wages, when all other factors are held constant. The wage elasticity

146    James Heintz and Karmen Naidoo shows the expected percentage change in employment of a 1 per cent change in real (i.e. inflation-​adjusted)  wages. Estimates of wage elasticities for South Africa suggest that labour demand is relatively inelastic—​changes in wages, even sizeable ones, would not be expected to result in large changes in employment. For instance, von Fintel (2017) produces elasticity estimates of around –​0.1 to –​0.2, suggesting that a 10 per cent reduction in wages may only increase employment by 2 per cent. Other estimates for South Africa suggest that employment may be more responsive, for example, an average elasticity of –​0.7, but there appears to be a wide variation by sector (Fedderke 2012). Given this evidence on the relationship between wages and employment, can high wages fully explain the high rates of involuntary unemployment in South Africa? The elasticity estimates suggest that paid employment is relatively unresponsive to wages. Because of this, wages would have to fall drastically in order for labour demand to increase to a point where it absorbs the unemployed labour force.5 In other words, wage premiums would have to be extremely large in order to fully explain the persistently high rates of unemployment. Furthermore, wages are not the relevant variables—​labour costs are. Rising wages may not translate into increasing labour costs if productivity rises sufficiently with wages. Along similar lines, slashing wages may not lower labour costs if productivity suffers, as efficiency wage arguments would lead us to believe. A sizeable cut to wages, when employment is inelastic, would also reduce incomes with negative consequences for aggregate demand. Elasticity estimates assume nothing other than wages and employment change, but if productivity and domestic demand are linked to wages, the impact on employment becomes more complex. The institutions that push up wages, such as bargaining councils, minimum wages, and labour laws, have been shown to have uneven coverage and enforcement (Flowerday, Rankin, and Schöer 2017; Bhorat, Kanbur, and Mayet 2012). If these restrictions are not binding in all cases, their impact on employment would be muted. Despite these caveats, it is important to recognize that elasticity studies for South Africa have consistently shown a negative relationship between wages and employment. Therefore, wage premiums and upward pressure on wages, in themselves, will likely do little to mitigate the current high rates of unemployment. Studies of the impact of bargaining council agreements and minimum wages have suggested that these institutions have had some negative impact on employment, although the estimated size of this effect is fairly modest (e.g. Bhorat, Naidoo, and Yu 2014; Dinkelman and Ranchhod 2012). Research has revealed evidence of significant dis-​employment effects of the introduction of

5 

For example, over the period October–​December 2019, the Quarterly Labour Force Survey, using the narrow definition, estimated the labour force to be 23.15 million and unemployment to be 6.73 million, for an unemployment rate of 29.1 per cent. Reducing unemployment to 5 per cent (in terms of the narrow definition) would require generating an additional 5.57 million paid jobs, or a 24.1 per cent increase in employment. If we assume a wage elasticity of –​0.5, for the sake of argument, well within the range of –​0.1 to –​0.7 discussed in the text, this suggests that wages would have to fall by 48.2 per cent. In other words, average wage incomes would need to be cut in half.

Unemployment in South Africa    147 sectoral minimum wage laws in certain contexts, for example, agriculture (Bhorat, Kanbur, and Stanwix 2014), but such legislation cannot explain the persistence of high rates of unemployment prior to the introduction of these policies. Given this evidence, wage rigidities, by themselves, cannot be seen as the primary cause of high rates of unemployment in South Africa. However, they could contribute to slower rates of job creation if labour market interventions are not coordinated with efforts to address the other structural causes of unemployment. Labour market regulations typically are assumed to affect labour demand by raising the cost of labour per unit of time (i.e. hourly wages) or in terms of unit labour costs (i.e. the cost of labour in one unit of output). However, other restrictions impact on hiring decisions by raising the overall non-​wage costs of employing workers. For instance, employment protection legislation, such as restrictions on terminating employment, can raise the overall expected cost of employing workers or limit the employer’s ability to adjust to changing economic conditions (Bhorat, Naidoo, and Yu 2014). In addition, policies that restrict the ability of employers to fire or lay off employees effectively reduce the cost of job loss to paid workers. According to efficiency wage theories, this may result in employers having to pay more to elicit a certain amount of effort or reduce on-​ the-​job productivity, as workers are less concerned about being fired. The higher costs of employing workers combined with these efficiency wage effects would dampen labour demand. In South Africa, institutions that extend employment protections may raise hiring and firing costs in ways that go beyond standard measurements, such as severance pay and prior notification (Benjamin, Bhorat, and Cheadle 2010). Drawn-​out dispute resolution processes, the treatment of probationary periods, and cumbersome procedures around dismissals can also raise the non-​wage costs of employing workers with potentially detrimental effects on employment (Bhorat, Jacobs, and van der Westhuizen 2013). Studies of temporary employment services in South Africa suggest that extensions of employment protection legislation displace workers in temporary employment (Cassim 2020; Bhorat, Magadla, and Steenkamp 2015). The 2014 Labour Relations Act (LRA) Amendment extended certain employment protections to temporary workers whose earnings fell below a certain threshold. Analysis of data before and after the LRA Amendment was adopted showed that more temporary workers moved out of paid employment or into informal forms of employment than moved into permanent jobs (Cassim 2020). In addition, there appears to have been a positive wage effect for temporary workers receiving these new protections. These outcomes are consistent with expectations of the impact of employment protections that lower the expected cost of losing a job. Like labour market rigidities that raise labour costs, employment protection laws and processes may have a negative impact on job creation. However, the estimated effects are not large enough to explain the high and persistent rates of unemployment in South Africa. Moreover, given the apartheid labour market system, with the pass laws, insecure urban residency rights, the ‘colour bar’, and limited or no protections from unfair dismissal, the introduction of employment protection measures in the post-​apartheid

148    James Heintz and Karmen Naidoo period is a natural response to the country’s economic history. The apartheid labour market, with no real protections for the majority of the labour force, did not generate adequate numbers of good job opportunities. It is unlikely that removing such protections in the post-​apartheid period would, in itself, solve the challenge of joblessness. Some micro-​level explanations of unemployment in South Africa focus on the supply-​ side of the labour market. One critical supply-​side factor has already been discussed—​ the underinvestment in education and skills development. However, other supply-​side explanations focus on the ability and willingness of unemployed workers to accept employment opportunities that become available. One such argument centres around the reservation wages of South Africa workers. The reservation wage refers to the minimum acceptable wage for an individual participating in the labour force. If an available job offers a wage or salary that falls below the reservation wage, an unemployed worker will refuse to accept that job. If reservation wages in South Africa tend to be higher than the prevailing wages, this could contribute to the high rates of measured unemployment. There are a number of reasons to be sceptical of reservation wage explanations of high rates of unemployment. At a basic level, if workers refuse jobs that pay the prevailing wage and are not subject to substandard working conditions, it raises questions of whether such unemployment should be considered involuntary. Voluntary unemployment, when workers choose to remain unemployed, is deemed less serious of an economic failing than involuntary employment, that is, the absolute lack of job opportunities. Research into reservation wages in South Africa has shown reported reservation wages to be unreliable (i.e. they may be aspirational wages rather than true reservation wages) and that reservation wages are often not higher than predicted prevailing wages (Kingdon and Knight 2004). Finally, it is unclear why workers would hold onto unrealistic wage expectations for an extended period of time. Expectations should adjust over time to reflect actual labour market conditions, raising questions about how reservation wages could explain high rates of sustained unemployment over decades. Another, related, factor that could affect the supply side of the labour market is the system of transfer payments that exists in South Africa, including the Older Persons Grant and the Child Support Grant. Private transfers are also important, such as remittances from migrant workers. When these transfer payments raise household incomes, they may reduce the imperative to seek out paid employment to meet the household’s needs. In effect, they have the potential to raise reservation wages, with possible impacts on measured unemployment as described above. Some researchers have found a negative correlation between the presence of pension recipients (Older Persons Grants) in the household and participation in paid employment (Abel 2019). But other studies have not found the same negative correlation between labour supply and receipt of a transfer payment. Indeed, since labour force participation is not costless and often requires real resources, cash transfers can actually have a positive impact on participation in paid work. This is particularly relevant for South Africa, with its history of spatially fragmented labour markets. For instance, research has shown that the receipt of a pension facilitates the ability of household members, specifically young men, to migrate for paid work (Ardington et al. 2016)

Unemployment in South Africa    149 although this effect depends on the educational attainment of the migrant worker. Other studies have shown similar positive impacts of cash transfers on migration for paid work among women (Posel et al. 2006). Moreover, even if the argument that cash transfer payments cause individuals to willingly leave the labour force were accepted, this would be considered a voluntary withdrawal from paid employment. It should not be considered as an explanation for involuntary unemployment. There may be other reasons to be concerned about a decline in labour force participation. However, as Figure 7.1 shows, South Africa has experienced a modest growth in the labour force participation rate, not a decline in participation. There does not appear to be a meaningful connection between South Africa’s system of cash transfers and the involuntary unemployment rate. Indeed, South Africa’s system of cash transfers have been shown to have a significant impact on reducing economic inequality (Leibbrandt et al. 2010). In this respect, the transfer programmes help address one of the more serious socio-​economic consequences of high rates of unemployment—​the country’s dramatic economic inequalities.

7.5  Post-​apartheid Economic Policy and Unemployment South Africa’s historical development trajectory, structural factors, and labour market institutions all contribute, to varying degrees, to the persistence of high rates of open unemployment. But the story of South Africa’s unemployment challenge does not end there. Other policy choices affect aggregate demand and investment patterns and, through these channels, influence levels of unemployment. South Africa’s economic policy in the post-​apartheid era was centred on liberalization and re-​integration into the global economy. Along with liberalizing financial flows and loosening exchange controls, the policy regime reduced trade protections and removed export incentives. In addition, from 2000 onward, the central bank’s primary mandate of price stability was implemented through an inflation-targeting framework, despite the fact that inflation rates had been moderate throughout the country’s history. These new policy directions could be seen as attempts to address some of the structural sources of unemployment discussed earlier. A stable, low-​inflation environment might encourage greater long-​term investment, faster capital accumulation, and more consumer spending if other factors (e.g. real interest rates) were not affected. Trade liberalization could increase aggregate demand through exports, if South African products were sufficiently competitive. Integration into global capital markets could attract foreign productive investment that has the potential to generate new job opportunities. However, in contrast to expectations, after adopting these policy measures, South Africa did not attract substantial levels of investment and, in most cases, trade liberalization proved to have net negative impacts on certain sectors, particularly manufacturing.

150    James Heintz and Karmen Naidoo South African monetary policy has been based on an inflation-​targeting framework since 2000, when the consumer price inflation target was set as a band of 3 per cent to 6 per cent. Through adjustments in policy interest rates—​using interventions into the market for government bond repurchase agreements—​the South African Reserve Bank has largely managed to keep inflation within the target band. This also means that the approach has sustained a high interest rate environment, with potentially negative consequences for aggregate demand and fixed capital investment. Between 2000 and the first quarter of 2010, the prime lending rate exceeded 10 per cent and between then and the end of 2019, has remained between 8 and 10.5 per cent (SARB 2020). Researchers have questioned the relevance of a strict inflation-targeting regime for developing countries due to variations in the sources of inflation, the role of exchange rate pass-​through, and the negative impact of higher interest rates and over-​valued real exchange rates on economic activity and employment (Brito and Bystedt 2010; Epstein and Yeldan 2008; Heintz and Ndikumana 2011). For South Africa, Epstein (2008) estimates that over the 2001–​04 period, a 4 p ​ ercentage point reduction in the prime lending rate is associated with a 0.6 per cent increase in GDP growth and only a 1 p ​ ercentage point increase in inflation. In a cross-​country panel data analysis that includes South Africa, Brito and Bystedt (2010) find that the adoption of inflation targeting had a negative impact on output growth, with consequences for paid employment. Regarding the employment effects in developing countries, Epstein and Yeldan (2008) show that many countries have experienced higher levels of unemployment after implementation of inflation targeting. Braunstein and Heintz (2008) find that countries that raise real interest rates above the long-​run trend to curb inflation experience a slow-​ down in employment growth, with a greater negative effect on the employment rates of women relative to men. Therefore, the high interest rate environment required to maintain the inflation target and to manage capital flows may be a contributing factor to lower rates of output growth and higher levels of unemployment in South Africa. As South Africa transitioned to democracy in 1994, economic policy focused on opening up the economy to foreign capital flows and trade. After a period of tight exchange controls in the 1960s to the 1970s to manage capital outflows during the period of political unrest, the 1980s ushered in a period of gradual exchange control relaxation. Liberalization continued in the post-​apartheid period, which included relaxing exchange control regulations and a removal of the two-​tier exchange rate system in favour of a floating exchange rate. These policies, along with tax amnesties and voluntary tax disclosure programmes, were implemented with a view to attracting foreign investment, encouraging residents to repatriate capital held abroad without excessive penalties, and reducing capital flight. Despite the positive expectations in the post-​apartheid period, net FDI to South Africa has remained low and capital flight has accelerated. The 1990s saw a net FDI outflow of $7 billion (in constant 2018 US dollars), which improved to a net inflow of US$43 billion in the 2000s, followed by almost equal FDI inflows and outflows in the 2000–​ 18 period (Ndikumana, Naidoo, and Aboobaker 2020). In addition, capital flight has

Unemployment in South Africa    151 accelerated in the post-​apartheid period. In the 1990s, it is estimated that South Africa lost a combined $49.4 billion in capital flight and trade mis-​invoicing, which increased to $130 billion over 2000–​09, and further to $158 billion over 2010–​17 (Ndikumana et al. 2020). Trade policy also affects employment opportunities. The process of trade liberalization began when the South African government agreed to the comprehensive tariff reductions under the General Agreement on Tariffs and Trade (GATT), which was then rapidly implemented from 1995 onwards through the World Trade Organization (WTO), of which South Africa was a founding member. The average tariff rate in the manufacturing sector declined from 22 per cent in 1994 to 7 per cent in 2004 (Erten, Leight, and Tregenna 2019). In addition, the 1990 General Export Incentive Scheme was phased out between 1995 and 1997 (Cassim, Onyango, and Seventer 2004). The weak performance of South Africa’s manufacturing sector has been linked with the rapid liberalization of trade and the accompanying increased import competition. Early research analysing the impact of trade liberalization over the 1993–​97 period finds that manufacturing employment lost through import competition was matched by employment generated through exports (Edwards 2001). Over this period, there was also a significant shift away from low-​skilled elementary employment, potentially driven by the nature of export competitiveness (Edwards 2001). However, later studies uncover a much stronger link between trade liberalization and reductions in manufacturing employment. Jenkins and Sen (2006) find that import competition explained about 10 per cent of the decline in manufacturing employment. While the authors find that productivity growth explained considerably more of the decline than import competition, the approach does not preclude the idea that productivity upgrades are also a response to increased imports. Furthermore, the rapid rise in imports from China in particular is also linked to substantial job losses in manufacturing in the 2001–​10 period (Edwards and Jenkins 2015). Erten et al. (2019) study the impact of tariff reductions on local labour market outcomes in South Africa in the first decade of the post-​apartheid period and find that districts that are more exposed to tariff declines experience greater declines in manufacturing employment relative to less affected districts. In addition, more affected districts experience an increase in the proportion of the working-​age population that are discouraged work-​seekers. Other research has investigated the poor performance of export-​ oriented manufacturing. South African manufacturing sectors have been found to have higher unit labour costs—​ indicating lower levels of productivity and competitiveness—​ compared to other developing countries over the 1970s to the 1990s, which negatively impacted export performance (Edwards and Golub 2004). Rodrik (2008) points to the declining relative prices, a proxy for profitability, of South African manufacturing in the 1994–​2004 period as an important determinant of the weak manufacturing performance and declining manufacturing employment. In further analysis, Rodrik (2008) finds that higher levels of import competition and the real exchange rate exert a negative impact on manufacturing profitability.

152    James Heintz and Karmen Naidoo South Africa’s post-​apartheid policy regime, with respect to monetary, financial, and trade policies, would have affected demand for domestic production, investment, and employment. While these policy changes cannot fully explain the country’s extreme unemployment problem, to the extent that they have dampened labour demand and job creation, they would have contributed to the persistence of unemployment in the post-​ apartheid period.

7.6 Conclusion Numerous factors contribute to South Africa’s high and sustained rates of unemployment: the structure of the economy, the country’s apartheid history, labour market institutions, and the policies implemented post-​apartheid. One striking observation that can be made based on the review presented in this chapter is that the determinants of unemployment extend well beyond the labour market itself. Therefore, an exclusive focus on labour market reforms—​such as deregulation, rolling back social protections tied to employment, and enhanced flexibility—​is not sufficient to address the country’s problem of joblessness. Labour market dynamics are important, but they must be considered along with other policy interventions, institutions, and structural features of the economy. This overview points to factors beyond the labour market that have impacted on and continue to affect South Africa’s persistent unemployment challenge. These include the rate of productive investment, the changing sectoral composition of economic activity, factor intensity, educational attainment, skills development, capital flows, macroeconomic management, and trade policy. The economic structures that impede employment creation were built up during more than four decades of interventionist policies under the apartheid regime. These structures are slow to change and, if left alone, are unlikely to automatically adjust to lower the country’s elevated unemployment rates. Instead, deliberate directive policies are needed to transform these structural impediments to job creation in order to tackle South Africa’s unemployment challenge.

7.6 Appendix Modification of Malinvaud’s (1978) Framework:  Let N represent the total level of paid employment and L the total labour force. The rate of unemployment, U, can therefore be given by:

U=

L−N L

Unemployment in South Africa    153 For a given level of L, the unemployment rate varies inversely with the aggregate level of employment. We can define the level of employment as follows:

N=

1 Y Y K λ KY

with

λ = labour productivity (output per unit labour) K = fixed capital stock Y = output/​capital ratio (indicator of capital intensity) K

Y = rate of capacity utilization (actual output as a fraction of potential output). Y From these expressions, the impact of different factors on the unemployment rate can be determined (all other factors remaining constant):



• • • •

Higher labour productivity will raise the unemployment rate Growth in the capital stock will lower the unemployment rate Less capital intensity (a higher output/​capital ratio) will lower unemployment Greater capacity utilization (aggregate demand) will lower unemployment.

References Abel, Martin. 2019. ‘Unintended labor supply effects of cash transfer programs: New evidence from South Africa’s pension’, Journal of African Economies, 28(5): 558–​81. Ardington, Cally, Till Bärnighausen, Anne Case, and Alicia Menendez. 2016. ‘Social protection and labor market outcomes of youth in South Africa’, ILR Review, 69(2): 455–​70. Benjamin, Paul, Haroon Bhorat, and Halton Cheadle. 2010. ‘The cost of “doing business” and labour regulation: The case of South Africa’, International Labour Review, 149(1):  73–​91. Bhorat, Haroon, Elne Jacobs, and Carlene van der Westhuizen. 2013. ‘Do industrial disputes reduce employment? Evidence from South Africa’. Africa Growth Initiative Working Paper no. 6. Washington D .: Brookings Institution. Bhorat, Haroon, Ravi Kanbur, and Natasha Mayet. 2012. ‘Minimum wage violation in South Africa’, International Labour Review, 151(3): 277–​87. Bhorat, Haroon, Ravi Kanbur, and Benjamin Stanwix. 2014. ‘Estimating the impact of minimum wages on employment, wages, and non-​wage benefits: The case of agriculture in South Africa’, American Journal of Agricultural Economics, 96(5): 1402–​19. Bhorat, Haroon, Sibahle Magadla, and Francois Steenkamp. 2015. Employment Effects in the Temporary Employment Services (TES) Sector: Post Labour Relation Amendment Act s198. Published mimeograph, Development Policy Research Unit: University of Cape Town Bhorat, Haroon, Karmen Naidoo, and Derek Yu. 2014. ‘Trade unions in an emerging economy: The case of South Africa’, in Celestin Monga and Justin Yifu Lin (eds) The Oxford Handbook of Africa and Economics, Volume 2: Policies and Practices. Oxford: Oxford University Press. Bowles, Samuel. 1985. ‘The production process in a competitive economy: Walrasian, neo-​ Hobbesian, and Marxian models’, America Economic Review, 75(1):  16–​36.

154    James Heintz and Karmen Naidoo Braunstein, Elissa, and James Heintz. 2008. ‘Gender bias and central bank policy: Employment and inflation reduction’, International Review of Applied Economics, 22(2): 173–​86. Brito, Ricardo D., and Brianne Bystedt. 2010. ‘Inflation targeting in emerging economies: Panel evidence’, Journal of Development Economics, 91(2): 198–​210. Casale, Daniela, and Dorrit Posel. (2002). ‘The continued feminization of the labour force in South Africa: An analysis of recent data and trends’, South African Journal of Economics, 70(1): 156–​84. Cassim, Aalia. 2020. ‘The impact of employment protection on the temporary employment services sector’. WIDER Working Paper no. 2020/​79, Helsinki. Cassim, Rashad, Donald Onyango, and Dirk Ernst van Seventer. 2004. The State of Trade Policy in South Africa (TIPS Monograph). Johannesburg: Trade and Industrial Policy Strategies (TIPS). Davies, Robert, and Dirk Ernst van Seventer. 2020. ‘Polarization in the South African labour market: Economy-​wide scenarios’. SA-​TIED Working Paper no. 138, UNU-​WIDER, Helsinki. Dinkelman, Taryn, and Vimal Ranchhod. 2012. ‘Evidence on the impact of minimum wage laws in an informal sector: Domestic workers in South Africa’, Journal of Development Economics, 99:  27–​45. Edwards, Lawrence. 2001. ‘Globalisation and the skills bias of occupational employment in South Africa’, The South African Journal of Economics, 69(1): 40–​7 1. Edwards, Lawrence, and Stephen S. Golub. 2004. ‘South Africa’s international cost competitiveness and exports in manufacturing’, World Development, 32(8): 1323–​39. Edwards, Lawrence, and Rhys Jenkins. 2015. ‘The impact of Chinese import penetration on the South African manufacturing sector’, Journal of Development Studies, 51(4): 447–​63. Elsley, Trenton. 2019. ‘National Minimum Wage Commission: A report on the proceedings of the Commission’. Labour Research Service, Cape Town, South Africa. Epstein, Gerald, and Erinc Yeldan. 2008. ‘Inflation targeting, employment creation and economic development: Assessing the impacts and policy alternatives’, International Review of Applied Economics, 22(2): 131–​44. Erten, Bilge, Jessica Leight, and Fiona Tregenna. 2019. ‘Trade liberalization and local labour market adjustment in South Africa’, Journal of International Economics, 118: 448–​67. Fedderke, Johannes. 2012. ‘The cost of rigidity: The South African labour market’, Comparative Economic Studies, 54: 809–​42. Flowerday, Wayde Thomas, Neil Rankin, and Volker Shöer. 2017. ‘Employment effects of bargaining council decisions’. Swiss Programme for Global Issues for Development, Working Paper no. 2017/​04,World Trade Institute, University of Bern, Switzerland. Heintz, James, 2009. ‘Social structures of accumulation in South Africa’, in David Kotz, Terrence Macdonough, and Michael Reich (eds) Social Structures of Accumulation. Cambridge: Cambridge University Press. Heintz, James and Leonce Ndikumana. 2011. ‘Is there a case for formal inflation targeting in sub-​Saharan Africa?’, Journal of African Economies, 20(Supplement 2): ii67–​ii103. Heintz, James, and Dorit Posel. 2008. ‘Revisiting informal employment and segmentation in the South African labor market’, South African Journal of Economics, 76(1):  26–​44. Jenkins, Rhys, and Kunal Sen. 2006. ‘International trade and manufacturing employment in the South: Four country case studies’, Oxford Development Studies, 34(3): 299–​322.

Unemployment in South Africa    155 Kerr, Andrew, David Lam, and Martin Wittenberg. 2019. Post-​Apartheid Labour Market Series: 1993–​2019 (Version No. 3.2). University of Cape Town: DataFirst https://www. datafirst.uct.ac.za/dataportal/index.php/catalog/434. Kingdon, Geeta G, and John Knight. 2004. ‘Unemployment in South Africa: The nature of the beast’, World Development, 32(3): 391–​408. Lam, David, Murray Leibbrandt, and Cecil Mlatsheni. 2010. ‘Human capital, job search, and unemployment among young people in South Africa’. IZA/​World Bank Conference on Employment and Development, Cape Town. Leibbrandt, Murray, Ingrid Woolard, Hayley McEwen, and Charlotte Koep. 2010. ‘Employment and inequality outcomes in South Africa’. Southern Africa Labour and Development Research Unit, University of Cape Town. Lewis, Arthur. 1954. ‘Economic development with unlimited supplies of labour’, The Manchester School, 22(2): 139–​91. Magruder, Jeremy. 2012. ‘High unemployment yet few small firms: The role of centralized bargaining in South Africa’, American Economic Journal: Applied Economics, 4(3): 138–​66. Malinvaud, Edmond. 1977. The Theory of Unemployment Reconsidered. Oxford: Basil Blackwell. Malinvaud, Edmond. 1978. Profitability and Unemployment. Cambridge: Cambridge University Press. Moll, Peter G. 1996. ‘Compulsory centralization of collective bargaining in South Africa’, American Economic Review, 86(3):  326–​9. Nattrass, Nicoli, and Jeremy Seekings. 2012. ‘Institutions, wage differentiation and the structure of employment in South Africa’. Centre for Social Science Research, Working Paper no. 307, University of Cape Town. Ndikumana, L., Karmen Naidoo, and Adam Aboobaker. 2020. ‘Capital flight from South Africa: A case study’. PERI Working Paper Series no. WP 516, University of Massachusetts, Amherst. Posel, Dorrit, and Daniela Casale. 2003. ‘What has been happening to internal labour migration in South Africa, 1993–​1999?’, South African Journal of Economics, 71(3): 455–​79. Posel, Dorrit, James Fairburn, and Frances Lund. 2006. ‘Labour migration and households: A reconsideration of the effects of the social pension on labour supply in South Africa’, Economic Modelling, 23: 836–​53. Rankin, Neil. 2012. ‘Labour productivity, factor intensity and labour costs in South African manufacturing’. REDI3x3 Working Paper no. 21, Pretoria. Rodrik, Dani. 2008. ‘Understanding South Africa’s economic puzzles’, Economics of Transition, 16(4): 769–​97. SARB. 2020. Economic Statistics Online Database. Pretoria [online] https://​www.resbank. co.za/​Research/​Statistics/​Pages/​MonthlyReleaseOfSelectedData.aspx. Shapiro, Carl, and Joseph Stiglitz. 1984. ‘Equilibrium unemployment as a worker discipline device’, American Economic Review, 74(3): 433–​44. Solow, Robert. 1979. ‘Another possible source of wage stickiness’, Journal of Macroeconomics, 1(1):  79–​82. Stiglitz, Joseph. 2012. The Price of Inequality. New York: W. W. Norton. Tregenna, Fiona. 2009. ‘Characterising deindustrialisation: An analysis of changes in manufacturing employment and output internationally’, Cambridge Journal of Economics, 33(3): 433–​66.

156    James Heintz and Karmen Naidoo Tregenna, Fiona. 2011. ‘Manufacturing productivity, deindustrialization, and reindustrialization’. WIDER Working Paper no. 2011/​57, Helsinki. Verick, Sher. 2012. ‘Giving up job search during a recession: The impact of the global financial crisis on the South African labour market’, Journal of African Economies, 2(3): 373–​408. Von Fintel, Dieter. 2017. ‘Institutional wage setting, labour demand, and labour supply: Causal estimates from a South African pseudo panel’, Development Southern Africa, 34(1):  1–​16. Wintrobe, Ronald. 1998. The Political Economy of Dictatorship. Cambridge: Cambridge University Press.

Chapter 8

P overt y in Sou t h   A fri c a Vusi Gumede

8.1 Introduction Prior to democracy, poverty was extremely high in South Africa and mainly Africans or the Black population group experienced poverty. It is a given that poverty (and inequality) worsened in South Africa as a result of the coronavirus (COVID-​19) pandemic. Various estimates that are analysed in this chapter indicate that prior to COVID-​19 many people were already living below the poverty line and that the share of those who were experiencing different dimensions of poverty was already high. Estimates based on the lockdown that was introduced as a result of COVID-​19 confirm that many people lost jobs and extreme poverty increased. It is critical to examine poverty in more detail so as to present a better picture of the evolution of poverty in the post-​apartheid period, and to consider possible interventions that can better tackle poverty, especially because most poverty estimates end in 2015. The chapter provides updated estimates regarding poverty within the limits of the available data. It uses the NIDS data (2008–​17). Among other issues, results confirm that the feminization of poverty and hardships experienced by youth have to be prioritized when initiatives to reduce poverty are considered. Although the chapter focuses on poverty, from a policy perspective, it is important to examine both poverty and inequality jointly instead of focusing on only one. They are both as a result of the mediocre performance of the South African economy and the very character, structurally, of the economy and unemployment as a consequence of South Africa’s political history. This chapter, however, mainly deals with the extent of poverty and its character. Chapter 9 deals with inequality. It confirms that South Africa is one of the most unequal societies in the world as a result of colonialism and apartheid. Chapter 9 also confirms that it is mainly Black South Africans that experience extreme poverty and that they have lower educational attainment, live in larger households, come from female-​headed households, and are predominantly based in rural areas.

158   Vusi Gumede

8.2  Conceptual and Theoretical Issues Definitions, measures, and the conceptualization of poverty (and of inequality) elicit many debates. The debates are about accuracy of the measures and assumptions that inform conceptualizations and definitions. Money-​metric measures are used to measure income poverty, to assess the level of income of a person against a poverty line. The multidimensional index assesses the levels of deprivation, taking into account the various dimensions. The various conceptualizations of poverty are indeed useful, although it is not always easy to measure poverty the way it has been conceptualized. For instance, fully capturing the voices of the poor in measuring poverty has not been satisfactorily done. Similarly, to some extent, so-​called service poverty and asset poverty remain controversial in terms of how these are measured. As indicated above, money-​metric measures (i.e. measures of poverty based on income) are the ones that are widely used and debates are more about the poverty line (e.g. whether it is set appropriately) for determining the headcount (P0), poverty gap (P1), and squared poverty gap (P2) measures of poverty. In the main, the measures of poverty indicate the minimum living level for an individual or household. There has to be sufficient income to meet essential items for survival, including food, housing, water, clothing, and employment (Jansen et al. 2015). It is important to highlight that poverty can also be viewed in different terms. Poverty can also be thought of as being either relative or subjective. The relative poverty measure is one such that even if a certain individual or household meets the necessary income for basic needs, an individual or household can still be considered poor compared to others within the society they live in. Subjective poverty, on the other hand, is when an individual or household determines whether their own living standard is better or worse. To measure poverty in money-​metric terms, as indicated above, a poverty line is needed or should be determined. The most-​used poverty line worldwide is the $1.25 per individual per day, largely because it allows comparisons between countries or regions. In the context of national poverty lines, these are based on the actual consumption and spending patterns of a particular country. Arguably, it is important that each country has an official poverty line because it is better that the discussions about those who are said to be poor are based on an agreed national poverty line (Gumede 2014). The measures of income poverty were first introduced by James Foster, Joel Greer, and Erik Thorbecke in the 1980s, hence literature often makes reference to FGT indices as a family of poverty metrics. Foster et al. (1984) suggested the FGT measure that uses the deprivation gap of each individual as her or his shortfall weight as shown below:

Pa ( y;f ) =

1 N

q

 f − yi  a f 

∑  i =1



Poverty in South Africa    159 where Pa denotes the generalized FGT measure, y is a vector of achievement q x 1, where yi is the expenditure of household i(i = 1, . . . ,q) and f is the predetermined poverty line or simply the cut-​off level, q = q(y;f) is the overall number of households that are deprived or simply poor. Moreover, the headcount ratio (H) that shows the share of poor individuals in the total population can be obtained if α = 0. There have been other measures, largely aimed at quantifying the various dimensions of poverty. The multidimensional poverty index (MPI) takes into account people living under conditions which do not meet the minimum nationally agreed standards in indicators of basic functioning (e.g. being properly fed, being educated, or drinking clean water) and people living under situations where they do not meet the minimum required standards in many aspects at the same time. In other words, the MPI quantifies multiple deprivations that individuals are experiencing; people who, for example, are both undernourished and do not have clean drinking water nor adequate sanitation. Alkire et al. (2011) explain that the MPI approach uses a dual cut-​off in order to consider both the prevalence as well as the intensity of multidimensional poverty. The prevalence of poverty is the proportion of the population that is multidimensionally poor using a specific cut-​off. For an example, a certain individual is classified as poor if the weighted indicators in which that particular individual is deprived sum up to at least 40 per cent and the intensity reflects the average proportion of indicators in which poor people are deprived (Jansen et al. 2015). This chapter measures and analyses standard money-​metric measures of poverty (i.e. the income level of a person or the level of a multidimensional index relative to an income or multidimensional poverty line) in order to have better understanding of the extent and intensity of income poverty in South Africa during the post-​apartheid period (i.e. since 1994).

8.3  Selected Studies on Poverty in South Africa To start with, there have been numerous studies that examine income poverty in South Africa (see the reviews of some of those studies in Gumede (2014) and Gumede (2008) in addition to those discussed in this chapter). Statistics South Africa (StatsSA) (2019) has undertaken a review of non-​money-​metric dimensions of poverty as well as inequality. In all the studies, essentially, it is confirmed that poverty has remained very high during the post-​apartheid period. Access to assets and services as well as social transfers in the form of grants and pensions have helped in reducing poverty and mitigating against inequalities. As discussed in C ­ hapters 7 and 29 of this volume, the labour market has not absorbed enough people to dent unemployment and thereby reduce poverty and inequality. The inability of many South Africans to access quality education largely as a result of the country’s political

160   Vusi Gumede history has also contributed to the slow progress that has been made in reducing poverty. Access to quality education is critical for interrupting intergenerational transmission of poverty (Gumede 2008). The MPI calculated in this chapter however does not show the critical importance of education because the contribution of access to education to the MPI appears low. It could be because access to education is quantified through child enrolment and years of schooling in the data used. That said, there are various factors that have conspired in making it difficult for the country to reduce poverty significantly. One of the key issues is the structure of the South African economy. Chapter 6 in this volume discusses relevant issues. The political history of the country characterized by the oppression of Africans or the Black population group and the inability to address (or slow progress in redressing) the past injustices have arguably contributed in poor traction as far as reducing poverty is concerned. As indicated earlier, there are many studies that have evaluated poverty in South Africa and they all confirm that extreme poverty has somewhat declined in post-​ apartheid South Africa but poverty remains very high. Maisonnave, Mabugu, and Chitiga (2019) used a dynamic computable general equilibrium (CGE) model and data based on the social accounting matrix (SAM) to estimate people living under $1 a day. Maisonnave et al. (2019) found that those living under $1 a day had fallen by over half, from 11.3 per cent to 4.0 per cent in the period 1994–​2011. Although the decline can be considered significant, it is arguably not good enough given that poverty has remained very high. The context that it is mainly Africans or the Black population group that experience poverty also requires that the declines in poverty are treated circumspectly. The history of the country should dictate that poverty afflicting Africans who are the majority and who are original inhabitants in South Africa should have been tackled with vigour and drastically reduced during the democratic dispensation. Schotte et al. (2018) fitted a multivariate regression to account for initial conditions in four waves of the NIDS to investigate poverty dynamics to social stratification and the study finds that although poverty reduced, for those remaining deprived there are key indicators that are more likely to lead people into poverty (e.g. households headed by young people, households headed by females, and households located in rural areas). There are other studies that support this. For instance, David et al. (2018) found that poverty is severe in the former Bantustans and in townships. Schotte et al. (2018) also highlights that the probability for those initially poor to fall into or remain in poverty was 25.9 per cent. Zizzamia et al. (2016) also estimated the risk of falling into poverty over time using the NIDS data and found that while 3.7 million people had become poor in 2014/​15, about 6.7 million of those who were poor in 2008 had moved above the poverty line in 2014/​15. This confirms the view that poverty has been decreasing in democratic South Africa. The issue is whether the estimated decline in poverty is significant enough given the history of the country. In addition, how severe poverty is for those who are below the poverty line. It is instructive that Burger et al. (2004) found that provinces that were initially poor remained the poorest. This finding is consistent with

Poverty in South Africa    161 earlier studies such as the one by Noble et al. (2013). Schotte et al. (2018), using the NIDS data, found that across provinces poverty was lower in the Western Cape and Gauteng compared to other provinces. This result is consistent with what Basarir (2011) found using the General Household Survey (GHS). StatsSA (2017) shows that poverty appears to have declined between 2010 and 2015. Table 8.1 below summarizes the poverty trends since 1996. Having increased during the period 2006–​09, the headcount ratio declined during the period 2009–​15. However, for the poverty gap and the squared poverty gap measures, there have been increases during the period 2011–​15, implying that the severity and intensity of poverty remain significant. To be sure, the poverty gap measures the minimum level of income needed for basic survival (i.e. intensity of poverty) while the squared poverty gap shows the severity of poverty. As presented in Table 8.1, StatsSA (2017) estimated poverty using the food poverty line (FPL), the lower-​bound poverty line (LBPL), and the upper-​bound poverty line (UBPL). For the FPL, poverty declined during the period 2006–​15 (from 28.4 per cent to 25.2 per cent) while poverty declined then increased for the LBPL and UBPL. The increase in poverty seems linked to the decline in economic growth and the increase in unemployment, particularly from 2015. Unemployment continued to increase and economic growth continued to decline, and, indeed, worsened due to the COVID-​19 pandemic. It will take a long time for the economy to return to the levels of economic performance of the late 2000s. Similarly, unemployment will remain very high for a while until the effects of the pandemic subside. Therefore, the gains associated with any decline in poverty have been reversed and poverty remains very high. As indicated earlier, it is important to also examine poverty by gender instead of broad figures, in order to deal with the notion of feminization of poverty. Many studies on poverty confirm that women are more prone to poverty than men (Nwosu and Ndinda 2018) and that there are more female-​headed households in poverty (Posel et al. 2016). This phenomenon has existed for a long time, similar to the reality that larger households are more likely to be in poverty.

Table 8.1: Poverty trends, 2006–​15 Poverty Headcount ratio line 2006

2009

Poverty gap 2011

FPL

28.4% 33.5% 21.4%

LBPL

51.0%

UBPL

2015

2006

2009

Square poverty gap 2011

2015

2006

2009

2011

2015

9.3% 12.3%

6.8%

9.0%

4.2%

6.0%

3.0

4.5

47.6% 36.4% 40.0% 22.2% 21.0%

14.3%

16.6%

12.2%

11.7%

7.3%

9.1%

66.6% 62.1% 53.2% 55.5% 35.6% 33.5% 25.5%

27.7%

22.5%

21.3%

15.0%

17.0%

Source: StatsSA (2017).

25.2%

162   Vusi Gumede Table 8.2 confirms that, although poverty is declining for both males and females, there are always more women than men below the poverty line. Chant (2006) argues that women are not only prone to suffer more from poverty, this standard of living tends also to persist over a longer period and can thus become intergenerational. Nwosu and Ndinda (2018) analysed the effect of transitioning from a male-​headed to a female-​headed household on changes in the probability of falling into poverty and their results suggest that the transitioning from a male-​headed to female-​ headed household is highly correlated with the probability of that household falling into poverty. This confirms the hypothesis that poverty is feminized. Rogan (2012) used StatsSA surveys to investigate the poverty status of female-​headed and male-​headed households in South Africa and concluded that female-​headed households had a far higher risk of falling into poverty. In terms of multiple dimensions of poverty, there are not many studies that have adopted the MPI approach in South Africa. StatsSA (2014) calculated the MPI based on 2001 and 2011 Census data. The finding was that MPI decreased during the period 2001–​11. Mushongera et al. (2015) used the Gauteng City-​Region Observatory survey data for 2011 and 2013 to investigate the multidimensional poverty index for Gauteng province in South Africa and their results point to a low MPI in Gauteng—​many studies find that Gauteng, similar to the Western Cape, performs better than other provinces with regards to poverty. Ntsalaze and Ikhide (2016) applied the multiple correspondence analysis using the NIDS and found that the lack of employment as well as financial commitment (over-​indebtedness) are the core dimensions of poverty as they constrain households from participating in essential activities. Rogan (2016), on the other hand, assessed gender and multidimensional poverty in South Africa also using NIDS and found that the poverty gap is largely similar to the poverty gap measured by the traditional money-​metric measures. Omotoso, Adesina, and Gbadegesin (2019) investigated children’s vulnerability to multidimensional poverty in post-​apartheid South Africa using the South African representative surveys following the Alkire-​Foster methodology and found that overall children’s vulnerability to poverty has reduced. Lack of employment opportunities and poor access to basic schooling are found to be the main factors that contribute to children’s probability of falling into poverty. There is also an issue of the extent to which social grants alleviate poverty. There are different views. For instance, Kelly (2013) and Webster et al. (2014) conclude that

Table 8.2: Poverty by gender Measure

Headcount ratio

Year Male Female

2006 26.5% 30.1%

Source: StatsSA (2017).

2009 32.0% 35.0%

2011 20.2% 22.6%

2015 23.7% 26.5%

Poverty in South Africa    163 social grants do not yield the expected results. In other words, social grants do not necessarily get people out of poverty. They argue that social grants are said to be trapping people by constraining them from improving their circumstances themselves. Bahre (2011) also makes the same argument. Essentially, social grants are said to increase government spending while on the other hand perpetuating dependency on government cash transfers. Other scholars conclude the opposite. Woolard and Klasen (2010), Patel (2013), Patel and Plagerson (2016), Delany et al. (2016), and Xaba (2016) argue that social grants have a positive impact on poverty reduction. For example, Patel and Plagerson (2016) investigated the trajectories of social protection in South Africa covering the period from 1994 to 2017 and found that child support grants reduced hunger, increased food security, led to better school performance, reduced income inequality, and improved growth monitoring as well as health outcomes. Furthermore, Woolard and Klasen (2010) analysed the evolution and the impact of social security in South Africa and concluded that the South African social assistance programme is highly associated with poverty reduction, improvement in education as well as health with particular attention to children living under extreme poverty. Along the same lines, Blake (2018) analysed the social cost of child support grants for female caregivers and their extended networks and found that the grant system has a positive impact on poverty reduction. There are some chapters in this volume that deal with social grants. The overall view is that social grants have helped in denting poverty and inequality. In other words, without social grants poverty and inequality would have been worse. The fundamental issue, arguably, is that there should be better interventions to decrease poverty and inequality in South Africa. Better, in a sense that such interventions are comprehensive and sustainable. So, even though social grants help in mitigating against poverty there should be, say, job opportunities as one example for how the country might comprehensively deal with poverty. The policy implications section goes into detail about possible policy options for reducing poverty in the country. In any case, people should be in a position to do what they value instead of waiting for a social grant at the end of each month. To briefly reflect on inequality, for a chapter on poverty cannot be silent on inequality, there is consensus that South Africa is a highly unequal society and ­Chapter 9 goes into details about this. In terms of provinces and municipalities, Schotte et al. (2018) found that Gauteng, the Western Cape, and Mpumalanga exhibited high levels of local social inequality. This is consistent with findings in Leibbrandt, Finn, and Woolard (2012), while Ozler (2007) and David et al. (2018) examined the spatial distribution of poverty and inequality in South Africa using the 2011 South African Census and found that overall inequality measured by the Gini coefficient was about 0.57 in all municipalities. However, this was differently distributed across municipalities. For example, the large municipalities are those characterized by high levels of inequality, ranging from 0.73 to 0.76. To be specific, the findings suggest that income inequality in the ten poorest municipalities varies from 0.65 in Nqutu, 0.68 in Maphumulo, to 0.77 in Mbizana and Ngquza Hill. Along the same lines, to illustrate gender dimensions as far as poverty and inequality are concerned, Bhorat et al. (2016) analysed the share of women in total employment

164   Vusi Gumede in several sectors and the results point to a high degree of variation. For instance, some sectors employ far more men than women (e.g. the taxi industry and the private security sector). Sectors such as domestic work, hospitality, and contract cleaning employ more women than men. Other scholars, for instance Fredericks and Yu (2017), investigated differences in employment attainment and findings suggest that between 1997 and 2015 men and white people were favoured relative to their female counterparts. Espi et al. (2019) used the Employment Equity Act (EEA) data to investigate gender and pay gaps in 2015 and 2016: results confirm what the earlier studies found, that women are under-​represented at the high-​skilled levels. Espi et al. (2019) also found that the gender pay gap has widened. Bhorat and Goga (2013), Rogan and Alfers (2019), and Mosomi (2019) also analysed the gender wage gap and the results are consistent in that the gender wage gap is high. For example, Mosomi (2019) finds that the gender wage gap remained at an average of 23 per cent and 25 per cent. According to Bezuidenhout et al. (2019) the gender wage gap also depends on the status of a firm. For example, the authors used a unique employee–​employer matched data panel from 2011 to 2016 for South Africa and employing fixed effects regressions and found that the gender wage gap of trading firms was higher than that of domestic firms. Women also suffer various forms of discrimination in the economy. Kaggwa (2020) used a survey approach to analyse gender equality in the mining sector and found that women suffered more discrimination compared to men. In addition, Kaggwa (2020) also found that the main challenge faced by women was the lack of participation in decision-​making processes and lack of career progress. This confirms gender inequality as well as the inequality of opportunities between men and women. See also Chapter 34, which reviews gender dynamics in depth.

8.4  Examining Poverty Dynamics As indicated earlier, the chapter uses the NIDS to estimate poverty in the post-​apartheid period. The NIDS is useful in many ways, including that it was specifically designed for the purpose of understanding poverty dynamics in democratic South Africa and that it is a longitudinal study tracing the same households and individuals over time. In brief, the NIDS is a household panel study first conducted in 2008. There are five waves of the NIDS: first wave (2008), second wave (2010), third wave (2012), fourth wave (2014), and last/​fifth wave (2017). For this chapter, income is used as a measure of economic welfare. NIDS adopted a recall method to derive household income, where first households were asked to declare total household income received over the last month. Aggregate household income is used by compiling the respondents’ answers on each income component based on the adult questionnaire. Regarding missing data due to non-​responses, the imputation method was adopted and the imputed income from the six sources was then added to the implied rental income, guided by the approach that Derek Yu (2013) followed.

Poverty in South Africa    165 To calculate the per capita income measure, household income is divided by household size (see Budlender et al. 2015 and StatsSA, 2017). StatsSA’s food poverty line of R515 is used. Considering that effectively a cross-​sectional approach (wave-​to-​wave) is used, the waves are weighted with their correspondent post-​stratified weights (see Zizzamia et al. 2018). Post-​stratified weights are also used to ascertain that the weighted distribution of household by geographical location (i.e. provinces), race, and gender are the same as the distribution baseline survey. In other words, to control for changes in price levels over time the food poverty line is deflated by multiplying it with the consumer price index (CPI) of the corresponding wave. The NIDS first wave had 26,776 respondents and by the last survey conducted there were 24,758 respondents. As explained in Gumede (2008), P1 can be interpreted as a measure of how much (income) would have to be transferred to the poor to bring their expenditure up to the poverty line. Unlike P0 (which captures the share of those who fall below the poverty line), P1 does not imply that there is a discontinuity at the poverty line. However, both P0 and P1 cannot capture differences in the severity of poverty amongst the poor. P0 and P1 do not consider possible inequalities among the poor, unlike P2 (the weighted sum of poverty gaps as a proportion of the poverty line where the weights are the proportionate poverty gaps themselves). It is recommended that all the three money-​metric measures are estimated because they help better understand poverty from an income or expenditure point of view. P2, for instance, takes inequalities among the poor into account. These measures also highlight the importance of having an official poverty line so that the assessment of whether poverty is changing or not can be based on a nationally agreed poverty line. Gumede (2014) explains the importance of an official poverty line and discusses efforts made in government to come up with a nationally agreed poverty line—​but South Africa still does not have an official poverty line. An official poverty line means a poverty line that is discussed across society and consensus is reached on what are the minimum living levels for a country such as South Africa. The estimates (as shown in Table 8.3) imply that poverty has been declining, at least prior to the economic downturn that worsened in 2019 and before the COVID-​19 pandemic which obviously increased poverty.

Table 8.3: Estimated poverty rates Poverty measure

2008

2010

2012

2014

2017

Headcount ratio (P0)

48%

45%

39%

32%

29%

Poverty gap ratio (P1)

22%

22%

17%

12%

11%

Squared poverty gap ratio (P2)

13%

13%

9%

6%

5%

Source: Author’s own calculations based on National Income Dynamics Study (NIDS). Here is the link for the NIDS data: http://​www.nids.uct.ac.za/​

166   Vusi Gumede The headcount ratio implies that poverty has been declining relatively significantly while the poverty gap ratio shows small improvements, and even smaller improvements in the squared poverty gap ratio, implying that the severity and intensity of poverty has not changed much in the democratic period of the country. For both the poverty gap ratio and the squared poverty gap ratio, poverty rates did not change during the first two waves and changes in the last two waves are negligible. Table 8.4 shows estimates by province. Except for Northern Cape and North West where poverty rates increased between wave 4 and wave 5, for all other provinces poverty rates declined. Looking at poverty rates by population group (i.e. race), there was a slight increase in poverty rates of the coloured population group during the last two waves. There are also negligible increases for the Asian/​Indian and white population groups. This is in line with other studies and it is intriguing if not counter-​intuitive. Whites have had over 350 years of privilege thereby making it hard to imagine that they can experience poverty. In the context of the notion of the feminization of poverty, Table 8.5 shows poverty rates by gender. Although poverty rates for both men and women have declined during the successive waves of the NIDS since 2008, the poverty rates of women remain higher than those of men. It is not enough to only examine the standard money-​metric measures of poverty, hence the chapter also examines the MPI. The MPI is useful as it captures the proportion of people who experience multiple deprivations and the intensity of such deprivations. In short, the MPI is made up of three different dimensions: education, health, and living standards. Education includes years of schooling and school attendance. Health takes into account nutrition and disability and the living standard dimension includes cooking fuel, sanitation, water, electricity, refuse, and dwelling type. For the estimation, all the indicators are converted into dummy variables coded as 1 if a certain individual is below the cut-​off point associated to that indicator and 0 otherwise.

Table 8.4: Estimated poverty rates by province Province

2008

2010

2012

2014

2017

Western Cape

25%

23%

20%

17%

18%

Eastern Cape

68%

61%

55%

48%

42%

Northern Cape

43%

43%

31%

23%

27%

Free State

46%

45%

37%

23%

22%

KwaZulu-​Natal

60%

58%

51%

43%

38%

North West

42%

44%

38%

30%

39%

Gauteng

31%

29%

23%

18%

17%

Mpumalanga

49%

42%

41%

33%

29%

Limpopo

63%

62%

54%

43%

35%

Source: Author’s own calculations based on National Income Dynamics Study (NIDS).

Poverty in South Africa    167 Table 8.5: Estimated poverty rates by gender Gender

2008

2010

2012

2014

2017

Male Female

44%

41%

35%

28%

26%

52%

49%

42%

35%

32%

Source: Author’s own calculations based on National Income Dynamics Study (NIDS).

Applying the Alkire-​Foster (AF) method, the MPI is derived through a two-​step cut-​off approach to identify those who are multidimensionally poor. Before deciding on cut-​offs, main indicators were identified and then classified into domains. Using the weighting scheme suggested by the early multidimensional poverty measurement studies, the indicators within each domain are equally weighted:

wdj =

1 1 χ T d



Each domain carries an equal weight of one-​third and the total weights for all domains equal 1. Each indicator j, with j = (1,2, . . . d) is linked to a minimum level of satisfaction which is denoted as the deprivation cut-​off point, represented as z = (z1z2 . . . zd). An individual or household is deprived if his/​her or its total satisfaction is below that cut-​off point. The overall MPI reflects the product of both the percentage of population that is multidimensionally poor (H, the poverty headcount ratio) and the mean proportion of weighted deprivation an individual or household experiences (A, the severity of poverty). Table 8.6 shows the total MPI, the contribution of each indicator and the contribution of each domain. Different from the headcount ratio, the figures suggest that multidimensional poverty exhibits different trends in the covered period. For instance, as can be noted in Table 8.6, the ratio of individuals that were below the cut-​off of one-third increased from wave 1 to wave 2, from 0.116 to 0.217 respectively. The trend changes from one wave to another in the subsequent waves. For example, there was a decrease in the ratio of people who do not have enough income to fulfil the minimum food requirement and have at least three deprivations from wave 2 to wave 3, followed by a slight increase from wave 3 to wave 4 and finally a decrease from wave 4 to wave 5. Essentially, multiple

Table 8.6: Multidimensional headcount MPI

2008

2010

2012

2014

2017

0.116

0.217

0.114

0.135

0.103

Source: Author’s own calculations based on National Income Dynamics Study (NIDS).

168   Vusi Gumede deprivations remain prominent in South Africa although there were improvements for some years, as far as the period of analysis for this chapter is concerned. From the first wave to the last wave of the NIDS, as estimations indicate, there have not been major improvements in multiple deprivations in South Africa. The MPI was 0.116 in the first wave and it was 0.103 in the last wave. Arguably, as unemployment has increased and the economy’s performance has worsened, the MPI would deteriorate. In other words, multiple deprivations would worsen in South Africa, also because of COVID-​19. A useful feature of the MPI is that once the total MPI is computed, we can still assess the contribution of each indicator to the total MPI. This approach was followed and the results, as shown in Table 8.7, suggest that the share of years of schooling as well as child enrolment is almost zero across the waves, except in waves 2 and 5. Moreover, the portion of people deprived in the dimensions, such as being disabled, not having access to electricity, poor dwelling type, not having fuel for cooking, as well as not living in a place with piped water, appears to have an impact on overall multidimensional deprivation. Although these factors appear to contribute to the total MPI, their impact is not significant compared to factors such as malnutrition, sanitation, and refuse removal. The aforementioned factors are found to play a significant role in overall multidimensional poverty. Although fuel for cooking does not have a significant contribution to the total MPI, something that is worth highlighting is that in wave 4 it played a significant role. The results imply that more attention should be given to malnutrition, sanitation, and refuse removal as their share in the total MPI is very significant compared to other factors in all the domains. Chapter 38 in this volume deals with the issue of malnutrition. Although there have been improvements in terms of hunger and food insecurity in the post-​apartheid dispensation, malnutrition is still a big challenge, hence the severity and intensity of poverty

Table 8.7: Contribution of each indicator, percentage Domain

Indicators

2008

2010

2012

2014

2017

Education

Child enrolment

0.000

0.175

0.000

0.000

0.016

Years of schooling

0.000

0.095

0.000

0.000

0.000

Disability

0.047

0.008

0.000

0.034

0.094

Malnutrition

0.450

0.365

0.469

0.446

0.439

Electricity

0.063

0.058

0.059

0.064

0.031

Sanitation

0.141

0.089

0.156

0.140

0.130

Refuse

0.141

0.103

0.156

0.136

0.125

Dwelling type

0.016

0.013

0.014

0.019

0.016

Fuel for cooking

0.086

0.055

0.073

0.102

0.093

Water

0.055

0.039

0.073

0.057

0.057

Health Living standard

Source: Author’s own calculations based on National Income Dynamics Study (NIDS).

Poverty in South Africa    169 remaining significant even in the democratic period. From each indicator’s contribution to the total MPI, the contribution for each domain was also computed. It is instructive that the health and the living standard domains appear to contribute more to the total MPI compared with education, which confirms the point made earlier. Estimates indicate that health was significant in wave 2, and wave 5, while the living standard domain had significant contributions in waves 1, 3, and 4. This means that access to health and the standard of living are the core domains in which people do not meet the necessary/​minimum living level and thus they are considered as multidimensionally poor. This finding suggests that there is probably sufficient access to education in the country. The issue is more about whether children are accessing good quality education. Last, another critical issue when examining poverty, especially in the context of African countries, relates to the question of whether poverty is predominantly a rural or an urban phenomenon. To investigate the determinants of poverty in rural and urban areas of South Africa the chapter uses a mixed effect probit model specified in the following manner:

(

)

Pr yij = 1| xit , ui = H ( xit β + zit ui )

where the subscripts i and t show a certain household at time t, with i = 1, . . . ,M clusters, with cluster t consisting of i = 1, . . .,nt observations. As the equation shows, yit = 1 if poverty ≠ 0 and yit = 0 otherwise. In other words, this is a binary variable taking a value of 1 if a certain household or individual is below the poverty line and 0 otherwise. β is a vector of elasticities given a vector of explanatory variables X (denoting income in the context of this chapter). In addition, 1xp row vector xit and 1xq vector zit, are the covariates for the fixed effect and random effect respectively. While in the first vector, the covariates are those that can be found in a normal probit model with regression coefficients (i.e. fixed effects), in the second vector the covariates correspond to the random effect that can be either random coefficients or random intercepts. In other words, it can be treated as the scalar 1 in the random intercept model. Furthermore, the random effects ui denotes M realizations from a multivariate normal distribution with mean 0 and qxq variance matrix ∑, which can also be called variance components. H represents the standard normal cumulative distribution function. As Table 8.8 shows, poverty in South Africa has remained predominantly a rural phenomenon. Zimbalist (2017), among others, demonstrated that poverty was much higher in rural areas than in urban areas for the period 1997–​2012. The picture has not changed much from what Zimbalist (2017) found. Arguably, poverty in rural areas might have intensified as a result of a weakening economy and it would most likely have worsened due to the COVID-​19 pandemic. Although poverty appears to have been declining, as this chapter shows, rural poverty has not declined much. In the last NIDS wave, rural poverty was 45 per cent while urban poverty was 19 per cent. So, over and above that female-​headed households are predominantly in poverty, those based in rural areas are mostly in poverty. This supports those who have argued that policy interventions must not forget about the countryside.

170   Vusi Gumede Table 8.8: Estimated poverty rates by geotype Settlement Type

2008

2010

2012

2014

2017

Rural Urban

69% 32%

65% 31%

57% 26%

48% 21%

45% 19%

Source: Author’s own calculations based on National Income Dynamics Study (NIDS).

8.5  Policy Implications It is clear from the ensuing analysis that, as expected, poverty affects the Black/​ African population group more. Within that race group, women are the most affected. Intuitively, we know also that the youth is the most affected because of high levels of youth unemployment and because there are no specific government programmes that target youth. Put differently, there is no evidence of the dent on youth unemployment from existing government programmes as far as the youth is concerned. Therefore, more should be done and better efforts made to deal with the poverty that youth is experiencing. Similarly, more must be done to deal with the notion of the feminization of poverty. It is not surprising that South Africa has not sufficiently addressed the feminization of poverty because the interventions have been very broad and did not directly take into account the reality that there are far more women than men experiencing poverty. For Goldberg and Kremen (1990) the feminization of poverty would occur when there are not enough initiatives aimed at equalizing wages and salaries for men and women. In addition, there must be deliberate programmes for redress given the dual role for many women (i.e. work and family), as well as support to single mothers. There are other critical policy interventions that should assist in mitigating poverty (and reducing inequality): labour market reforms and ensuring that social policy is comprehensive. It is important to indicate that social security and/​or social protection are aspects of social policy but are not social policy (in and of themselves) (see Gumede 2019). To be sure, social protection is viewed as a broader set of interventions aimed at preventing and mitigating social harm while social security has to do with income protection schemes. Similarly, there must be more effort to facilitate rural development.

8.6 Conclusion Poverty remains very high in South Africa. There have been different interventions aimed at reducing poverty (and inequality) but interventions pursued so far have

Poverty in South Africa    171 been timid. Social protection mechanisms have helped make extreme poverty manageable, but they have not been effective in sustainably reducing poverty. Arguably, government hoped that by growing the economy and creating jobs poverty would decline. Overall, the required fundamental intervention as far as reducing poverty (and inequality) in South Africa is concerned has to do with the restructuring of the South African economy. The character and structure of the economy makes it difficult to sustainably reduce poverty (and inequality). So, over and above other interventions such as ensuring that there is a comprehensive social policy and that relevant laws and other policies are properly pursued, it is critical that the economy is not only made to perform better but that it is also restructured in favour of the majority of South Africans. Various interventions, such as those focusing on land and agrarian reforms, should be vigorously, but carefully, pursued.

References Alkire, Sabina, and James Foster. 2011. ‘Understandings and misunderstandings of multidimensional poverty measurement’. OPIH Working Paper no. 43, Oxford University. Bahre, Erik. 2011. ‘Liberation and redistribution: Social grants, commercial insurance, and religious riches in South Africa’, Comparative Studies in Society and History, 53(2): 371–​92. Basarir, Hasan. 2011. ‘Poor, multidimensionally speaking: Evidence from South Africa’, Journal of African Economies, 20(3): 463–​504. Bezuidenhout, Carli, Caro Janse Van Rensburg, Marianne Matthee, and Victor Stolzenburg. 2019. ‘Trading firms and the gender wage gap: Evidence from South Africa’, Agenda: Empowering Women for Gender Equity, 33(4): 79–​90. https://​www.tandfonline.com/​doi/​ abs/​10.1080/​10130950.2019.1677162. Bhorat, Haroon, and Sumayya Goga. 2013. ‘The gender wage gap in post-​apartheid South Africa: A re-​examination’, Journal of African Economies, 22: 827–​48. Bhorat, Haroon, Tara Caetano, Benjamin Jourdan, Ravi Kanbur, Christopher Rooney, Benjamin Stanwix, and Ingrid Woolard. 2016. ‘Investigating the feasibility of a national minimum wage for South Africa’. DPRU Working Paper no. 201601, University of Cape Town. Blake, Rosemary. 2018. ‘The price of the grant: The social cost of child support grants for female caregivers and their extended networks’. Working Paper no. 412, Cape Town: Center for Social Science Research. Budlender, Joshua, Murray Leibbrandt, and Ingrid Woolard. 2015. ‘South African poverty lines: A review and two new money-​metric thresholds’. SALDRU Working Paper no. 151, University of Cape Town. Burger, Ronelle, Servaas Van der Berg, Sarel Van der Walt, and Derek Yu. 2004. ‘Geography as destiny: Considering the spatial dimensions of poverty and deprivation in South Africa’, Proceedings of the DPRU/​TIPS (Development Policy Research Unit/​Trade and Industrial Policy Strategies) Conference, 13–​15 October, Somerset West. Chant, Sylvia. 2006. ‘Poverty begins at home? Questioning some (mis)conceptions about children, poverty and privation in female-​headed households’. UNICEF. Report on the State of the World’s Children. UNICEF, London, UK.

172   Vusi Gumede David, Anda, Nathalie Guilbert, Nobuaki Hamaguchi, Yudai Higashi, Hiroyuki Hino, Murray Leibbrandt, and Muna Shifa. 2018. ‘Spatial poverty and inequality in South Africa: A municipality level analysis’. Agence Francaise de Developpement (AFD), ISSN 2492-​2846. https:// ​ w ww.afd.fr/ ​ e n/ ​ r essources/ ​ s patial- ​ p overty- ​ a nd- ​ i nequality- ​ s outh- ​ a frica-​ municipality-​level-​analysis. Delany, Aislinn, Alejandro Grinspun, and Evelyne Nyokangi. 2016. ‘Children and social assistance: An introduction’, in Aislinn Delany, Selwyn Jehoma, and Lori Lake (eds) South African Child Gauge 2016. Cape Town: Children’s Institute, University of Cape Town. Espi, Gabriel, David Francis, and Imraan Valodia. 2019. ‘Gender inequality in South Africa labour market: Insights from the Employment Equity Act data’, Agenda: Empowering Women for Gender Equity [online journal]: 44–​61. Foster, James, Joel Greer, and Erik Thorbecke. 1984. ‘A class of decomposable poverty measures’, Econometrica, 52:  761–​6. Fredericks, Fadwah, and Derek Yu. 2017. ‘The effect of affirmative action on the reduction of employment’. ERSA Working Paper no. 681, Cape Town. Goldberg, Gertrude S., and Eleanor Kremen. 1990. The Feminization of Poverty: Only in America? Westport, CT: Greenwood Press. Gumede, Vusi. 2008. ‘Poverty and second economy dynamics in South Africa: An attempt to measure the extent of the problem and clarify concepts’. DPRU Working Paper no. 08/​133, University of Cape Town. Gumede, Vusi. 2014. ‘Poverty and poverty lines in South Africa’, in Haroon Bhorat, Alan Hirsch, Ravi Kanbur, and Mthuli Ncube (eds). Oxford Companion to the Economics of South Africa. Oxford: Oxford University Press. Gumede, Vusi. 2019. ‘Social policy in post-​ apartheid South Africa’, Journal of Public Administration, 54(5): 499–​511. Jansen, Ada, Mariana Moses, Stanford Mujuta, and Derek Yu. 2015. ‘Measurements and determinants of multifaceted poverty in South Africa’, Development Southern Africa, 32: 151–​ 69. http://​dx.doi.org/​10.1080/​0376835X.2014.984377. Kaggwa, Martin. 2020. ‘Interventions to promote gender equality in the mining sector of South Africa’, The Extractive Industries and Society, 7(2): 398–404. Kelly, Gabrielle. 2013. ‘We need to change how we think (and talk) about social grants’, Ground Up, 7 October. http://​groundup.org. za/​content/​we-​need-​change-​how-​we-​think-​ and-​talk-​about-​social-​grants. Leibbrandt, Murray, Arden Finn, and Ingrid Woolard. 2012. ‘Describing and decomposing post-​ apartheid income inequality in South Africa’, Development Southern Africa, 29(1):  19–​34. Maisonnave, Helene, Ramos Mabugu, and Margaret Chitiga. 2019. ‘Economywide consequences of attaining Millenium Development Goals in South Africa’, Economies Bulletin, HAL: 1118–​27. https://​hal-​normandie-​univ.archives-​ouvertes.fr/​hal-​02313227. Mbewe, Samson, Ingrid Woolard, and Dennis Davis. 2019. ‘Wealth taxation as an instrument to reduce wealth inequality in South Africa’, in Crain Soudien, Vasu Reddy, and Ingrid Woolard (eds) Poverty and Inequality: Diagnosis, Prognosis and Responses. Cape Town: HSRC Press. Mosomi, Jacqueline. 2019. ‘Distributional changes in the gender wage gap in the post-​apartheid South African labour market’. WIDER Working Paper no. 2019/​17, Cape Town. Mushongera, Darlington, Precious Zikhali, and Phindile Nguenya. 2015. ‘A multidimensional poverty index for Gauteng province, South Africa: Evidence from quality of life

Poverty in South Africa    173 survey data’, Social Indicators Research, 277–​303. https://​link.springer.com/​article/​10.1007/​ s11205-​015-​1176-​2. Noble, Michael, Wanga Zembe, Gemma Wright, and David Avenell. 2013. ‘Multiple deprivation and income poverty at small area level in South Africa in 2011’. SASPRI, Cape Town. Ntsalaze, Lungile, and Sylvanus Ikhide. 2016. ‘Rethinking dimensions: The South African multidimensional poverty index’, Social Indicators Research, 135(1): 195–​213. Nwosu, Chijioke O., and Catherine Ndinda. 2018. ‘Gender-​based household composition changes and implications for poverty in South Africa’, Journal of International Women’s Studies, 19(S): 82–​94. http://​vc.bridgew.edu/​jiws/​vol19/​iss5/​6. Omotoso, Kehinde O., Jimi O. Adesina, and Taiwo F. Gbadegesin. 2019. ‘Children on the edge: Estimating children’s vulnerability to multidimensional poverty in post-​apartheid South Africa’, Child Indicator Research, 13(4): 1155–​74. Ozler, Berk. 2007. ‘Not separate, not equal: Poverty and inequality in post-​apartheid South Africa’, Economic Development and Cultural Change, 3: 487–​529. Patel, Leila, 2013. ‘Do social grants create more problems than they solve?’. Hellen Joseph memorial lecture. Centre for Social Development in Africa, University of Johannesburg [online] http://​w ww.uj.ac.za/​EN/​Newsroom/​News/​Documents/​2013/​Helen%20Joseph%20 lecture%20Prof%20 Leila%20Patel%20speech.pdf. Patel, Leila, and Sophie Plagerson. 2016. ‘The evolution of the Child Support Grant’, Child Gauge 2017. Children’s Institute, UCT. Posel, Dorrit, Daniela Casale, and Erofili Grapsa. 2016. ‘Re-​estimating gender differences in income in South Africa: The implications of equivalence scales’, Development Southern Africa, 33(4): 425–​41. Rogan, Michael. 2012. ‘Poverty and headship in post-​apartheid South Africa, 1997–​2008’. ERSA Working Paper no. 288, Cape Town. Rogan, Michael. 2016. ‘Gender and multidimensional poverty in South Africa: Applying the global multidimensional poverty index (MPI)’, Social Indicators Research, 126(3): 987–​1006. Rogan, Michael, and Laura Alfers. 2019. ‘Gendered inequalities in the Southern African informal economy’, Agenda: Empowering Women for Gender Equity, 91–​102. https://​doi.org/​ 10.1080/​10130950.2019.1676163. Schotte, Simone, Rocco Zizzamia, and Murray Leibbrandt. 2018. ‘A poverty dynamics approach to social stratification: The South Africa case’, World Development, 88–​103. https://​ ideas.repec.org/​a/​eee/​wdevel/​v110y2018icp88-​103.html. Statistics South Africa. 2017. Poverty Trends in South Africa: An Examination of Absolute Poverty between 2006 and 2015. Pretoria: Statistics South Africa. Webster, Edward, M. Metcalfe, N. Van Niekerk, Michael Noble, John Reynolds, Ulandi Du Plessis, Xoliswa Dilata, Vijay Makanjee, Russel Grinker, and Jeff Peires. 2014. ‘Building a culture of service in the Chris Hani District Municipality: Household Survey report’. Chis Hani Institute project, paper on findings presented at Rhodes University. Woolard, Ingrid, and Stephan Klasen. 2010. ‘The evolution and impact of social security in South Africa’. Cape Town, European University Institute (EUI) https://socialprotection.org/ discover/publications/evolution-and-impact-social-security-south-africa. Xaba, Mzingaye Brilliant. 2016. ‘Bandaging a broken arm: The effectiveness of the child support grant in alleviating poverty in poor Grahamstown communities’, Africa Insight, 46(3): 153–​7 1.

174   Vusi Gumede Yu, Derek. 2013. ‘Poverty and inequality estimates of National Income Dynamic Study revisited’. Stellenbosch Economic Working Papers 05/​13, Bureau for Economic Research, University of Stellenbosch, Stellenbosch. Zimbalist, Zack. 2017. ‘Analysing post-​apartheid poverty trends by geo-​type, 1997–​2012: The understated role of urbanisation and social grants’, Development Southern Africa, 34(2): 151–​67. Zizzamia, Rocco, Simone Schotte, and Murray Leibbrandt. 2016. ‘Social class, life chances and vulnerability to poverty in South Africa’. Paper presented at the Development Studies Association Conference, Oxford, 12–​14 September, Oxford Poverty and Human Development Initiative (OPHI) Working Paper no. 43 ISSN–​2040-​8188. Zizzamia, Rocco, Simone Schotte, and Murray Leibbrandt. 2018. ‘Snakes and ladders and loaded dice: Poverty dynamics and inequality in South Africa between 2008–​2017’. NIDS, SALDRU, University of Cape Town.

Chapter 9

Inequali t y i n Sou th Af ri c a Murray Leibbrandt and Fabio Andrés Díaz Pabón

9.1 Introduction South Africa remains one of the most unequal countries in the world. The current levels of inequality are the legacy of the segregation and marginalization of the vast majority of its citizenry in every aspect of South Africa’s socio-​economic development before and during apartheid. Contemporary inequality literature is unambiguous about the fact that such extremely high levels of inequality are detrimental to a country’s development path and stifle a country’s potential in multiple dimensions (Wilkinson and Pickett 2010). In South Africa, poverty and inequality are often mentioned in the same breath. This elides over important conceptual differences between poverty (measured by incomes at a particular threshold) and inequality (measured by gaps between groups or individuals). Poverty is analysed as an individualized metric, the income level of a person, or the level of a multi ​dimensional index relative to an income or multidimensional poverty line (for a fuller discussion on poverty in South Africa, see ­Chapter 8 in this volume). On the other hand, inequality is a relational concept; one is more or less equal in relation to another (Tilly 1998). As such the analysis of poverty cannot necessarily explain the forces reproducing intergenerational inequality or poverty. Analysing inequality is better suited to unveiling the societal processes driving the provision and allocation of assets and services, and the determinants of well-​being, and can better describe the forces reproducing intra-​generational and intergenerational inequality. This is the task of this chapter.

176    Murray Leibbrandt and Fabio Andrés Díaz Pabón Although much of the literature that foregrounds the centrality of inequality is more recent, even in 1994 there was an intuitive sense in South Africa of the daunting conditions that its social and economic inequalities bequeathed South Africa’s new democracy. It is hardly surprising then that, from the outset, substantial progress in transitioning to a more equal society was flagged as a key metric of progress in South Africa. There has been much work in understanding the levels and patterns of income inequality and how these have changed over time in South Africa. Most of this work has been informed by the use of survey data to measure levels and changes in income and earnings inequities. Such measurement has been important in setting the context for analysis and, more recently, has been enriched by looking also at the role of wealth and asset inequalities as illustrated by the work of Piketty (2014). This work is reviewed in section 9.2. While the contribution from understanding levels and patterns of inequality remains important, for understanding the social structures behind the creation and reproduction of inequality we need also to account for the mechanisms creating inequality. Incorporating the insights from other fields in the social sciences improves our understanding of the drivers of inequality. Indeed, we go on to argue that, for understanding inequality in South Africa better, it is imperative that progress is made in mapping the mechanisms that reproduce and generate inequality along categorical distinctions such as gender, race, and class (Tilly 1998). Looking at the interaction between inequalities (e.g. income inequalities) and different sets of categories (e.g. gender, race, and class) helps in better describing the specific mechanisms that create and reproduce inequities for different groups. We then go on to review a new literature that analyses these interactions using South Africa’s panel data. The focus of this promising work is the study of intra-​generational and intergenerational social mobility. The concluding section pulls together this understanding of the connection between wealth and incomes and their interactions with categorical inequalities. It argues that this approach has led to a better understanding of the forces that drive inequality in South Africa and a sounder foundation for assessing policies to overcome inequality.

9.2  Income, Wealth, and Assets The analysis of household income and wealth inequality over time in South Africa has provided the flagship indicators of the levels of South Africa’s inequalities. The changes in these variables over the post-​apartheid period are used to describe overall progression of income inequality in the country (Leibbrandt et al. 2012; Hundenborn et al. 2018a). The Gini coefficient of incomes was 0.67 in 2006, 0.65 in 2009, and 0.65 in 2015, using data from Statistics South Africa’s Income and Expenditure, and the Living Conditions Surveys (Statistics South Africa 2019). These levels and trends are robust to the choice

Inequality in South Africa    177 of different inequality indexes used (Atkinson, Theil, and Palma indexes) and data sources.1 The consistency of these findings along different indicators and data sources has prevented any inequality denialism in South Africa and this picture has been mainstreamed in different policy documents. For example, the South African government National Development Plan sets a target of reducing the income Gini coefficient from these levels to 0.60 by 2030 (National Planning Commission 2012). Having been tabled as part of national plans and addressed in different policy proposals, the fact that the levels of inequality remain so high is a sign of the failures of the South African government and society in addressing its inequalities. This persistently high level of income inequality, should not be interpreted as saying that the dynamics of South Africa’s income inequality are static. The rest of this chapter is focused on unpacking the important social dynamics that undergird these aggregate measures. In beginning to understand the drivers of inequality in South Africa, there is a tradition that focuses on the role of different sources of income (labour market, social grants, remittances, and capital income) that are coming into households for different groups along the income distribution.2 In a contemporary example, Hundenborn et al. (2018b) study 1993, 2008, and 2014, finding that labour market income is highly correlated with overall inequality and accounts for 84–​90 per cent of the overall Gini coefficient, being by far the most important determinant of household income inequality in South Africa. This high share is due almost to the high number of households with no access to labour market incomes as well as the inequality in labour incomes for those households with access to labour market incomes. Given the importance of labour income, employment and earnings dynamics remain central in driving South Africa’s inequality of income.3 Earnings inequality increased from a Gini of 0.552 in 2001 to 0.634 in 2014 (Finn and Leibbrandt 2018). Unemployment has risen dramatically due to changes in the supply of labour (Casale and Posel 2002; Mosomi 2019) as well as changes in the demand for labour (Bhorat et al. 2020; Tregenna 2012). Whereas average real gross earnings of employed workers rose, real wage increases went mostly to top earners, explaining the increase in earnings inequality (Wittenberg 2017a, 2017b, 2017c, 2018). These extreme earnings differentials are undergirded by a labour market that is segmented along racial and gender lines (Statistics South Africa 2019) coupled with large power and class asymmetries between employers and employees in a job-​scarce economy (Webster and Francis 2019). (For a fuller discussion on the changing dynamics of the South African Labour Market, see ­Chapter 31 in this volume.)

1  For example, Hundenborn et al. (2018a) use the 1993 PSLSD and 2008 and 2014 NIDS data to estimate inequality in 1993, 2008, and 2014; finding that there is a very small total change in the Gini, from 0.68 to 0.69, in the 1993–​2008 period and that there is a reduction in the Gini, from 0.69 to 0.655, in the 2008–​14 period. 2  See review in Leibbrandt et al. (2012). 3  See recent reviews by Finn and Leibbrandt (2018), Bhorat et al. (2020), Tregenna and Tsela (2012).

178    Murray Leibbrandt and Fabio Andrés Díaz Pabón Household income source decompositions show clearly that aggregate income inequality would have risen more than it did as a result of these labour market forces, had it not been for government grants and subsidies (cash transfers). Grants and subsidies have been crucial in dampening increasing inequality (Hunderborn et al. 2018a). However, the inequality dampening effect of such grants had plateaued before the economic crisis unveiled by the COVID-​19 pandemic (Mashekwa and Woolard 2018), raising questions about the long-​term impact of the pandemic on income inequalities. Income source decompositions also show that capital incomes are an important source of inequality. This makes it imperative to understand how top-​end earnings and income from the capital are creating and reinforcing income inequalities. This need to understand top-​end income dynamics relates to an international concern that household and labour market surveys, which have served as the basic data for the analysis of income inequality in South Africa, may underestimate top-​end incomes (Alvaredo et al. 2017). South African researchers have started to address this concern. Bassier and Woolard’s (2018) recent analysis of tax and household survey data provides a striking demonstration of the usefulness of this line of research in understanding the extent of top-​end inequality in South Africa. Whereas Credit Suisse in 2016 estimated that there were about 45,000 US dollar millionaires in the country, Bassier and Woolard (2018) put the number much higher, at about 182,000. The impact of top-​end incomes driving increases in inequality of incomes can be also illustrated by how, between 2003 and 2016, the real incomes of the top 5 per cent of income earners increased at a rate of 5.1 per cent annually, more than double the rate of growth of gross national income (GNI) after 2008. In contrast, the incomes of the other 95 per cent of the population either stagnated, or, in the case of the bottom of the distribution, showed only slight growth. Real incomes of the top 1 per cent of income earners almost doubled between 2003 and 2016 and, between 2010 and 2016, the income shares of the top 1 per cent increased from 10.5 per cent to 12.6 per cent of the country’s GNI (Bassier and Woolard 2018). The importance of understanding capital income and top-​end income dynamics meshes with a need to understand wealth inequality. In spite of its growth in importance as part of the GNI, South African wealth inequality remains under-​researched. This is mostly because of the lack of access to data on wealth, especially the tax data that researchers use in other contexts to research wealth, as well as the challenge of tracking capital flows across different countries. However, recent literature makes an important start. Orthofer (2016) shows that while the South African income Gini coefficient is around 0.67, for wealth, this index is at least 0.9–​0.95. Both of these values are higher than in any other major economy of the globe for which such data exist.4 Chatterjee et al. (2020) have triangulated survey data and tax data with national accounts data, to derive

4 

Using tax records and data from the National Income Dynamics Study (NIDS), Orthofer also finds that the wealthiest 10 per cent of the population own at least 90–​95 per cent of all wealth.

Inequality in South Africa    179 estimates of the distribution of personal wealth in South Africa. Even after recognizing that these data might still underestimate the value of capital, they find that South Africa has ‘unparalleled’ levels of wealth concentration: The top 10 per cent own 86 per cent of aggregate wealth, and the top 0.1 per cent close to one­third of aggregate wealth. The top 0.01 per cent of the distribution (3,500 individuals) own 15 per cent of household net worth, more than the bottom 90 per cent as a whole. Such high levels of inequality can be accounted for in all forms of assets, including housing, pension funds, and other financial assets. Our series show no sign of decreasing wealth inequality since apartheid; if anything, we find that inequality has remained broadly stable and has even slightly increased within top wealth groups. (Chatterjee et al. 2020: 2)

However, in the South African context, it is important to acknowledge that changes over time in well-​being, and in inequalities in well-​being, extend beyond a focus on the flow of income or expenditure into and out of a household and even the stock of money-​metric wealth. Such analysis must include, among other things, a broader set of assets accumulated by households over time. There is a literature measuring these other dimensions of inequality and their changes over time (Bhorat and Van der Westhuizen 2013; Wittenberg and Leibbrandt 2017). Extending access to these assets and basic services has been a major focus of government policy over the post-​ apartheid years and many government interventions have involved the provision of low-​cost housing, and have boosted access to water, electricity, and other public services. All of the studies measuring these asset inequalities see this large policy push reflected in declines in asset poverty and asset inequality irrespective of the period within the last twenty years since 1994 (Bhorat and Van der Westhuizen 2013; Wittenberg and Leibbrandt 2017). Statistics South Africa (2019) provides an exhaustive review of these non-​money-​ metric dimensions of poverty and inequality and makes the point that the major improvements in asset poverty and inequality took place in the immediate period after the transition from apartheid rather than since the 2000s (for a fuller discussion on poverty in South Africa, see ­Chapter 8 in this volume). Also, underneath the improvements, strong patterns of relative deprivation persist towards Black South Africans, women, and citizens in rural areas. In sum, this is a major policy achievement. Yet, the reality is that the resilience of earnings, income, and wealth inequities illustrates that, in spite of a greater provision of public services and improvements in the ownership of assets, their impact on the livelihoods across the distribution is not the one that was expected. The fact that the improvement in access to assets and public services has not changed livelihood opportunities for many South Africans not only relates to the challenge of such interventions to alter income distribution dynamics but is also connected to the important and continuing role of the categories imposed by apartheid in defining inequities in the country.

180    Murray Leibbrandt and Fabio Andrés Díaz Pabón

9.3  Categorical Inequalities and Household Composition The distribution of earnings, income, wealth, and assets remains very unequal in South Africa. Crucial in understanding the resilience of these inequalities is the role of social categories such as race, gender, and class and their intersection with spatial segregation. Understanding the intersection between social categories, inequalities, and location offers a key lens for understanding the generation and persistence of inequality in South Africa (Winker and Degele 2011). Such an analytical framework can allow us to understand the differences between different sets of groups, as opposed to general comparisons of the entirety of a population—​the latter is referred to in the literature as vertical inequality. The analysis of inequalities along sets of categories has been undertaken by different fields, such as sociology via the analysis of categorical inequalities (Tilly 1998), or in the development economics literature via the analysis of horizontal inequalities (Stewart 2016; Hirschman and Rothschild 1973). Such analysis complements our understanding of horizonal inequalities and is better suited for unveiling the mechanisms that limit the access and rewards to different groups along different sets of categories in different locations. It allows us to better describe how different-​generation mechanisms operate along different sets of categories and therefore can inform better policies to reduce inequities as opposed to general descriptions of the entirety of a diverse population. In this spirit, and understandably given South Africa’s history, a decomposition of income by the analysis of different racial categories has featured prominently in research on income inequality. For example, Leibbrandt et al. (2010), compared between and within racial group inequalities between 1993 and 2008, finding that racial categories in 1993, contributed between 42 per cent and 50 per cent to between-​group inequities.5 By 2008, the inequality between different groups (Black,6 coloured, Indian, and white South Africans) had declined to between 30 per cent and 38 per cent.7 Hino et al. (2019) suggest a further decline by 2015. This illustrates that the dynamics of inter-​group inequalities (categorical/​horizontal inequalities) have changed, while the levels of inequality between individuals (vertical inequalities) remain at their high aggregate levels. So, crucial context to this decomposition by race is that the between-​race share remains extremely high compared to international standards. Also, inequality within each of South Africa’s racial categories has risen sharply over time.

5 

Most importantly, this depends on whether racial population shares or income shares are used to weight inequality. 6  Referred to also in South Africa as ‘African’. 7  Generally, in other countries, between group inequalities ‘account for’ less than 10 per cent of the overall inequality of income, earnings, or any vertical measure.

Inequality in South Africa    181 Residents in urban areas Residents in rural areas

White South African Indian South African Coloured South African Black South African

Male Female 0

8000

16000 24000

Figure 9.1  Average monthly wages between different sets of categories (in South African Rands-​ZAR) Source: Own elaboration based on NIDS-​CRAM 2020 data.

These decompositions are useful in allowing us to understand the importance of categorical inequalities. However, they serve as calls for explanation rather than explanations themselves. The forces that reproduce categorical inequalities are dynamic forces that condition, create, and recreate dynamic traps (Tilly 1998; Stewart and Langer 2008); these forces are not necessarily visible as they operate via both institutional and non-​institutional mechanisms. As the transition towards democracy eliminated the overt mechanisms and tools for segregation implemented by apartheid, inequality is now produced and reproduced by subtler and less overt mechanisms that operate both as class markers (e.g. dress codes, educational attainment, accents, location of residence, names, and surnames), but also as access cards for networks, opportunities, and rewards. This change in the operation of mechanisms, from being enacted by the state to being enforced by markets and society, makes it harder for households or individuals to overcome inequality. For example, Pellicer and Ranchhod (2016) show how the relationship between education and the labour market has operated to reproduce an inequality trap. Also, as discussed before, the inequalities in earnings and assets by gender are compounded by complex dynamics by which gender inequities continue to affect the incomes of the majority of the population of the country (for a fuller discussion on gender and work in South Africa, see ­Chapter 34 in this volume). Figure 9.1 illustrates differences in inequalities in earnings in South Africa in 2017.8 Aside from illustrating the importance of each of the above categories, the stark 8 

We are grateful to Aidan Horn for his support in analysing the data used in this figure.

182    Murray Leibbrandt and Fabio Andrés Díaz Pabón rural and urban differences show that, in the case of South Africa, spatial segregation remains a central dimension that illustrates the reproduction of long-​term categorical inequalities. External and internal colonialism carved South Africa into spatially segregated spaces (Nelson Mandela Foundation 2018). In terms of measurement, this is shown most starkly by studies that have used census data to map inequalities in income and many other services across space. Noble et al. (2006) and then, a decade later, David et al. (2018) and Statistics South Africa (2019), show very clearly that the deepest poverty and deprivation in contemporary South Africa remains in the areas that were the ‘Bantustans’ and ‘townships’ under apartheid. While there are inequalities in all dimensions of life, the worst allocation of services and opportunities continues, to be observed in the areas that were created and used by apartheid for exploiting South Africans of colour. As understanding of the forces creating inequalities requires a recognition of other dimensions of inequality beyond income inequality, a start has been made by complementing this analysis through understanding inequality via the lens of categorical inequalities, including gender, racial, and spatial dynamics. By moving the studies of inequality beyond incomes to other markers of advantage and disadvantage we start to make visible in the South African context the connection between advantages and wages and income of which Tilly speaks (Tilly 1998). Yet, the diversity of a country coupled with the extent of the different inequality-creating mechanisms in which gender, race, class, and location intersect, makes it harder to come to simple descriptions that can account for the different forms in which access to rights, opportunities, and rewards are conditioned. Recent research has made clear the importance of understanding the dynamics of household size and composition as aspects of social mobility. In measuring income inequality, income per capita is generally used as the measure of well-​being by normalizing total household income by the number of members. Acknowledging that the same income flow coming into a large or a small household does not equal the same level of well-​being is essential. Yet the decision to normalize incomes in this way, while sensible, is not innocuous. Such an analysis departs from a very strong abstraction that assumes that each member of a household receives or needs the same income. Indeed, research shows that differences in the size and the composition of households across the income distribution bring about important inequalities that need to be accounted for. Posel et al. (2016) find that households at the bottom of the income distribution are larger, with more children and with higher percentages of households being female-​headed. Hundenborn et al. (2018a) undertake a series of dynamic income source decompositions in which some of these demographic changes are accounted for. These decompositions separate out the effects of demographic changes—​driving the denominator, from the income changes—​driving the numerator. One example of the usefulness of such partitioning is that it shows that declining household sizes since 2000 have led to a slightly more equal distribution of employed adults per household. Such household size adjustments lower per capita income inequality even when employment itself has not increased.

Inequality in South Africa    183 The research on equivalence scales by Posel et al. (2016 and 2020) provides another affirmation of the importance of giving detailed attention to household size and composition. An equivalence scale allows for transforming the income from a household with different numbers and ages of members into an equivalent household with one adult member. However, using per capita income assumes that there are no scale economies for larger households and that individuals of all ages require equivalent resources. These are strong assumptions which serve as an excellent example of the need to be alert to the value judgements implicit in the different indicators used to measure inequality. Yet, there is another crucial substantive point here too. The structure of households matters in imposing different constraints and barriers to social mobility for the members of each household. Given this, households adjust their compositions to respond to the needs of their members, in the light of the structural conditions of their contexts, and to government policies (Hamoudi and Thomas 2014). Recent analysis on the effective targeting of COVID-​19 relief measures has made it clear that understanding these connections is a matter of life or death (Bassier et al. 2020).

9.4  Social Mobility and Inequality Dynamics Recent work on social mobility in South Africa has added to our understanding of the conditions associated with the production and reproduction of South Africa’s inequality. There are two lenses. The analysis of intra-​generational mobility focuses on the social progress of individuals across time in contemporary South Africa. Intergenerational mobility focuses on the progress of children relative to that of their parents.9 Both lenses provide important insights into the texture of South Africa’s inequality.

9.4.1 Intra-​generational Inequality The NIDS data have been used to track social mobility across the income distribution between 2008 and 2017 in South Africa along five waves (Zizzamia et al. 2016, 2019; Finn and Leibbrandt 2016; Schotte et al. 2018). This research suggests a delineation of the population of the country into five social strata: the chronically poor, the transient poor, the vulnerable middle class, the stable middle class, and the elite. Figure 9.2 below reflects this picture. Statistics South Africa has found that 55 per cent of the population in 2015 lived in poverty (Statistics South Africa 2017). This is a high poverty rate. The analysis of social 9 

The measurement of these mobilities requires panel data, tracking individuals and households. This work has been greatly facilitated by South Africa’s NIDS panel study.

184    Murray Leibbrandt and Fabio Andrés Díaz Pabón 100

3.3

2.7

3.3

4.9

19.2

19.0

19.7

23.4

22.4

13.6

12.1

13.4

11.6

16.8

13.0

19.4

4.0

80 60 11.2

11.5

11.4

40 52.0

54.1

51.3

2010

2012

20 0 2008

Chronic

Transient

Vulnerable

45.0

42.0

2014 Middle Class

2017 Elite

Figure 9.2  Socio-​economic class sizes, 2008–​17 Source: Zizzamia et al. (2019).

mobility goes further to show that, for three-quarters of the population who are not the middle class or the elite, life is filled with the precariousness of a game of ‘snakes and ladders’ (Schotte et al. 2018). In the bottom half of the income distribution, poverty is much more pervasive across time than a cross-​sectional analysis tells us. Building off the analysis of poverty dynamics by Finn and Leibbrandt (2016), Schotte et al. (2018) differentiate between persistent poverty and transient poverty, finding that about 42 per cent of the population persistently live in chronic poverty, 11.4 per cent of the population can be classified as ‘transient poor’, and about 19 per cent of the population is part of what can be called a ‘vulnerable middle class’—​at risk of falling into poverty. Both of these groups (transient and vulnerable) are at real risk of falling back into poverty from one period to the next. This work on social mobility also adds an important dynamic perspective to a large body of literature, suggesting a growing (Black) middle class in South Africa (Visagie and Posel 2013; Burger et al. 2014). Most of this literature has focused on measuring the size of the Black middle class and its change over time. Burger et al. (2015) use Sen’s capabilities approach to usefully anchor the discussion of a middle class in the possibilities and capabilities of South Africans in their day-​to-​day lives. Within such a framing of lived realities, it is clear that some quantitative measures have overestimated the size of the middle class. Zizzamia et al. (2016) and Schotte et al. (2018) build on this perspective and confirm that the size of a stable middle class is relatively small. Figure 9.2 shows that only one in four people can be considered to be part of either the stable middle class or the elite. This is a stark contrast to the narrative of burgeoning black diamonds that has been portrayed in the media for much of the last twenty years. This social mobility analysis goes further to ascertain the forces that propel people from one socio-​economic class to another or keep them routed to their class. Employment changes (losing or gaining a job) and demographic changes (increases

Inequality in South Africa    185 or decreases in household size) emerge as crucial determinants. Job gains accounted for one-​third of all exits from poverty, while changes in household size accounted for half of all poverty entries or exits. Whereas employment is important in lifting people out of poverty, this is not sufficient. The quality of employment is important too. Those with unstable or precarious jobs remain more vulnerable to falling back into poverty as opposed to those with permanent and formal jobs. This work on social mobility illustrates clearly how categorical inequalities intersect in explaining persistent and recurring inequalities. Those trapped in chronic poverty are almost wholly Black South Africans, have lower educational attainment (below high school), live in larger households, come from female-​headed households,10 and are predominantly based in rural areas.11 In contraposition to this, the stable middle class and the elite are defined by a set of markers that set them apart from the vulnerable middle class and people living in chronic poverty. These markers continue to be defined by racial categories and class. For example, 93.6 per cent of white South Africans were observed to be consistently non-​poor. By contrast, about 63 per cent of Black South Africans were poor in four or five waves, with only about 9 per cent of Black South Africans remaining non-​poor in all five waves. Whereas there has been a decline in the inequities between different racial groups in South Africa, racial categories continue to condition and limit opportunities and remain markers that define access to housing, public services, transport, schools, employment, safety, and opportunities—​the access to the tools required to exercise agency and freedom. Limited intra-​generational mobility remains grounded in the legacy of apartheid.

9.4.2 Intergenerational Inequality The unequal distribution of assets, notably education, health, and housing, persists across time and across generations (Pellicer et al. 2011; Pellicer and Ranchhod 2016; Spaull and Jansen 2019; Lundberg and Startz 1998). These traps reproduce inequalities and also condition the lives of future generations in the country. The lack of intergenerational mobility and its precariousness and vulnerability provide a telling lens to describe the reproduction of inequality in South Africa. Using data from the NIDS, recent research reveals that there is not a level playing field in terms of equality of opportunity in South Africa as the opportunities available to most South Africans have much to do with the socio-​economic status of their families (Piraino 2015; Finn et al. 2016). Piraino (2015) reveals that in South Africa there is a strong correlation between a parent’s earnings and what will be their children’s expected income in the future, estimating an intergenerational earnings elasticity of between 10  Nearly three-​ quarters of all the female-​headed households remained in poverty in four or five waves, compared with only 29 per cent in the case of male-​headed households. 11  Only 2.5 per cent of rural households remained non-​poor through all five waves while nearly 83 per cent of all households were poor in four or five waves.

Strength of Relationship Between Parent’s Earnings and Child’s Earnings

186    Murray Leibbrandt and Fabio Andrés Díaz Pabón 1 .95 .9 .85 .8 .75 .7 .65 .6 .55 .5 .45 .4 .35 5

10

15

20

25 30 35 40 45 50 55 60 65 70 Position of Parents in the Earnings Distribution Fathers

75

80

85

90

95

Mothers

Figure 9.3  The intergenerational transmission of earnings advantages or disadvantages Source: Finn et al. (2016) relabelled by Leibbrandt et al. (2018).

0.62 and 0.67. Piraino finds that this persistence is driven by a limited set of inherited circumstances and especially due to the impact of the racial categories imposed by apartheid. Finn et al. (2016) show that what is particularly notable about the South African case is the exceptionally high absence of intergenerational mobility at both the bottom and the top of the earnings distribution. Figure 9.3 is taken from this analysis. It shows that nine out of ten children from the poorest families still occupy the same vulnerable place in the earnings distribution as their parents do (or did) once they themselves enter the labour market. At the top of the distribution, a privileged position is being passed on as children of top-​earning parents have a 70 per cent chance of being at the very top of the earnings distribution in the future. At the same time, there is slightly more mobility in the middle of the earnings distribution, reflecting slightly higher chances of both upward and downward mobility. This resonates with the findings from the intra-​ generational mobility analysis in indicating stability or persistence of two opposing poles—​wealth or extreme poverty with more volatility being experienced by those in the middle of the distribution.

9.5 Conclusion In its most simple understanding, inequality is a noun that speaks of the differences between individuals or groups in a society (Milanovic 2010). But inequality is a multidimensional concept and, as Sen noted insightfully thirty years ago, calling for equality

Inequality in South Africa    187 without specifying what we are referring to when we speak about equality and inequality leads to confusion (Sen 1992). When this lack of specificity takes place, interventions designed to reduce inequities lack clarity as to both their objectives and the means through which to achieve their objectives. South Africa has good cross-​sectional survey data and, more recently, panel data and tax data that has been used to understand the factors creating and reproducing some of South Africa’s inequalities. Research using these data has made it clear that, in terms of income, earnings, and wealth at least, South Africa continues to have extremely high levels of inequality. The evidence shows an engaged, active citizenry, responding to opportunities and trying to forge better livelihoods for themselves, that continues to be severely constrained in what is possible by the less visible forces that remain binding inequality traps. The exceptionally high correlations between racial categories, where people are situated in the income and wealth distributions, their gender, where they live, and where they are situated in the distributions of other dimensions of well-​being, give a sense of the compounding factors that explain the persistence of one of the highest inequalities in the world. But much more needs to be done to better understand the social processes and mechanisms through which social mobility is limited and inequities continue to be reproduced period after period. Some authors have framed South African inequality dynamics as being a move from race to class (or class to race). It is clear from the synthesis of existing research in this chapter that both are vital in understanding contemporary South African inequality. The key analytic challenge is to understand the evolving interactions between race and class and their role in conditioning access to networks, services, opportunities, and rewards. The chapter has highlighted the multidimensional nature of the inequalities that prevail across contemporary South Africa today. Every day, individuals carry these inequalities into the labour market, into the broader economy and into their daily lives as intersecting bundles that tightly constrain opportunities for most and that open up many opportunities for a few. Evidence from the psychological and health literature has revealed that this lived reality makes policies directed at reducing inequality even more daunting. The economic insecurity associated with being vulnerable and unemployed reduces people’s well-​being, and is compounded by the impact of belonging to specific categories even if a deterioration in material welfare does not materialize (e.g. Cafiero and Vakis 2006). In other words, and as highlighted by the impact of the COVID-​19 pandemic, it is not only current earnings or consumption that matter for actual welfare, ‘but also the risks a household faces, as well as its (in)ability to prevent, mitigate, and cope with these’ (Klasen and Povel 2013, 17). This picture is made more complex as vulnerability and precariousness have the potential to create further traps in which people opt for stable, low-​return sources of income, rather than to invest in activities with more lucrative but also more uncertain outcomes that could improve their livelihoods (Wilkinson and Pickett 2010). All policy documents have consistently spoken about the need to reduce inequality and have given the state a strong role in redressing the social, political, economic, and

188    Murray Leibbrandt and Fabio Andrés Díaz Pabón spatial disadvantages via different interventions since 1994 (Tregenna 2012). Yet, collectively, these policies have had a limited effect in reducing inequality. Some would argue that the actual attention and understanding of inequality in the policy sphere and the coherence of the strategies to address inequality remains at best debatable (Bond 2014; Hall and Kepe 2017). The review of policy within this chapter has been somewhat more generous than this. Nonetheless, it is clear that despite the general recognition of poverty, inequality, and unemployment as South Africa’s three key policy challenges, the articulation of inequality in this troika has been more rhetorical than substantive. Government has yet to truly confront inequality as a separate and perhaps overarching challenge. If South Africa aims to create a stable, progressive society, poverty traps must be broken, poverty escapes must be sustained over time and a growing, stable middle class must be supported. None of this is possible unless the country reduces inequality and the impact of the mechanisms that produce and reproduce inequality. We do not know the counterfactual of how much worse South Africa’s inequality would have been without these interventions. We know that the extensive system of social grants has helped in ameliorating the increase of income inequality and that there has been notable progress in lowering inequalities in access to housing and essential social services. Nevertheless, the combination of these policies and minimum wages in the labour market has not had the positive effects on earnings and income inequality that they had in other contexts such as in Latin America (Cornia 2014). Because of this, it is urgent to sharpen the understanding of the reasons why existing policies have failed in reducing inequality. Interdisciplinary research can provide new insights that enable us to better understand the forces behind existing inequality traps.

References Alvaredo, Facundo, Anthony Atkinson, Lucas Chancel, Thomas Piketty, Emmanuel Saez, and Gabriel Zucman. 2017. ‘Distributional National Accounts (DINA) guidelines: Concepts and methods used in WID world’. WID World Working Paper no. 6/​2017. World Inequality Lab. Bassier, Ihsaan, and Ingrid Woolard. 2018. ‘Exclusive growth: Rapidly increasing top incomes amidst low national growth in South Africa’. REDI3x3 Working Paper no. 47, SALDRU, University of Cape Town. Bassier, Ihsaan, Joshua Budlender, Rocco Zizzamia, Murray Leibbrandt, and Vimal Ranchhod. 2020. ‘Locked down and locked out: Repurposing social assistance as emergency relief to informal workers’. SALDRU Working Paper no. 261, Cape Town. Bhorat, Haroon, and Carlene Van der Westhuizen. 2013. ‘Non-​monetary dimensions of wellbeing in South Africa, 1993–​2004: A post-​apartheid dividend?’, Development Southern Africa, 30(3): 295–​314. Bhorat, Haroon, Kezia Lilenstein, Morné Oosthuizen, and Amy Thornton. 2020. ‘Structural transformation, inequality, and inclusive growth in South Africa’. WIDER Working Paper no. 50/​2020, Helsinki.

Inequality in South Africa    189 Bond, Patrick. 2014. Elite Transition: From Apartheid to Neoliberalism in South Africa. London: Pluto Press. Burger, Ronelle, Camren McAravey, and Servaas Van der Berg. 2015. ‘The capability threshold: Re‐examining the definition of the middle class in an unequal developing country’. Stellenbosch Economic Working Papers no. 16/​2015, Department of Economics, Stellenbosch University. Burger, Ronelle, Cindy Lee Steenekamp, Servaas Van der Berg, and Asmus Zoch. 2014. ‘The middle class in post‐apartheid South Africa: Examining and comparing rival approaches’. Stellenbosch Economic Working Paper no. 11/​2014, Department of Economics, Stellenbosch University. Cafiero, Carlo, and Renos Nicos Vakis. 2006. ‘Risk and vulnerability considerations in poverty analysis: Recent advances and future directions’. Social Protection Discussion Paper no. 0610, World Bank, Washington, DC. Casale, Daniela M. and Dorrit R. Posel. 2002. ‘The continued feminisation of the labour force in South Africa: An analysis of recent data and trends’, South African Journal of Economics, 70: 156–​84. Chatterjee, Aroop, Léo Czajka, and Amory Gethin. 2020. ‘Estimating the distribution of household wealth in South Africa’. WIDER Working Paper no. 45/​2019, Helsinki. Cornia, Giovanni Andrea. 2014. Income Inequality in Latin America: Recent Decline and Prospects for its Further Reduction. Oxford: Oxford University Press. David, Anda, Nathalie Guilbert, Nobuaki Hamaguchi, Yudai Higashi, Hiroyuki Hino, Murray Leibbrandt, and Muna Shifa. 2018. ‘Spatial poverty and inequality in South Africa: A municipality level analysis’. SALDRU Working Paper no. 221, Cape Town. Finn, Arden, and Murray Leibbrandt. 2016. ‘The dynamics of poverty in the first four waves of NIDS’. SALDRU Working Paper no. 174, Cape Town. Finn, Arden, and Murray Leibbrandt. 2018. ‘The evolution and determination of earnings inequality in post-​apartheid South Africa’. WIDER Working Paper no. 83, Helsinki. Finn, Arden, Murray Leibbrandt, and Vimal Ranchhod. 2016. ‘Patterns of persistence: Intergenerational mobility and education in South Africa’. SALDRU Working Paper no. 175, Cape Town. Hall, Ruth, and Thembela Kepe. 2017. ‘Elite capture and state neglect: New evidence on South Africa’s land reform’. Review of African Political Economy, 44(151): 122–​30. Hamoudi, Amar, and Duncan Thomas. 2014. ‘Endogenous coresidence and program incidence: South Africa’s old age pension’, Journal of Development Economics, 109(30): 30–​7. Hino, Hiroyuki, Arnim Langer, John Lonsdale, and Frances Stewart (eds), 2019. From Divided Pasts to Cohesive Futures: Reflections on Africa. Cambridge: Cambridge University Press. Hirschman, Albert O., and Michael Rothschild. 1973. ‘The changing tolerance for income inequality in the course of economic development: With a mathematical appendix’, The Quarterly Journal of Economics, 87(4): 544–​66. Hundenborn, Janina, Murray Leibbrandt, and Ingrid Woolard. 2018a. ‘Drivers of inequality in South Africa’. WIDER Working Paper no. 162/​2018, Helsinki. Hundenborn, Janina, Ingrid Woolard, and Jon Jellema. 2018b. ‘The effect of top incomes on inequality in South Africa’. WIDER Working Paper no. 90/​2018, Helsinki. Klasen, Stephan, and Felix Povel. 2013. ‘Defining and measuring vulnerability: State of the art and new proposals’, in Stephan Klasen and Hermann Povel (eds) Vulnerability to Poverty. London: Palgrave.

190    Murray Leibbrandt and Fabio Andrés Díaz Pabón Leibbrandt, Murray, Arden Finn, and Ingrid Woolard. 2012. ‘Describing and decomposing post-​ apartheid income inequality in South Africa’, Development Southern Africa, 29(1):  19–​34. Leibbrandt, Murray, Simone Schotte, and Rocco Zizzamia. 2018. ‘Tackling persistent poverty and inequality: A dynamic perspective’, in Michael Nassen Smith (ed.) Confronting Inequality: The South African Crisis. Johannesburg: Jacana. Leibbrandt, Murray, Ingrid Woolard, Arden Finn, and Jonathan Argen. 2010. ‘Trends in South African income distribution and poverty since the fall of apartheid’. OECD Social, Employment and Migration Working Paper no. 101, Paris. Lundberg, Shelly, and Richard Startz. 1998. ‘On the persistence of racial inequality’. Journal of Labor Economics, 16(2): 292–​323. Mashekwa, Maboshe, and Ingrid Woolard. 2018. ‘Revisiting the impact of direct taxes and transfers on poverty and inequality in South Africa’. WIDER Working Paper no. 79/​2018, Helsinki. Milanovic, Branko. 2010. The Haves and the Have-​nots: A Brief and Idiosyncratic History of Global Inequality. New York: Basic Books (AZ). Mosomi, Jacqueline Nyamokami. 2019. ‘Distributional changes in the gender wage gap in the post-​apartheid South African labour market.’ WIDER Working Paper no. 17/​2019, Helsinki. National Planning Commission. (2012). National Development Plan 2030: Our Future—​Make It Work. National Planning Commission, Pretoria. Nelson Mandela Foundation. 2018. ‘Mandela Initiative 2018. Grappling with poverty and inequality: A report on the processes and findings of the Mandela initiative’. Nelson Mandela Foundation, Johannesburg [online] https://​www.mandelainitiative.org.za/​images/​final-​ reports/​MI_​Report_​Final_​13SEP.pdf. Noble, Michael, Miriam Babita, Helen Barnes, Chris Dibben, Wiseman Magasela, Stefan Noble, Phakama Ntshongwana, Heston Phillips, Sharmla Rama, Benjamin Roberts, Gemma Wright, and Sibonguile Zungu. 2006. ‘The provincial indices of multiple deprivation for South Africa 2001’, Statistics South Africa [online] http://​www.statssa.gov.za/​ census/​census_​2001/​PMID/​PIMDReport2006.pdf. Orthofer, Anna. 2016. ‘Wealth inequality in South Africa: Evidence from survey and tax data’. REDI3x3 Working Paper no. 15, SALDRU, Cape Town. Pellicer, Miquel, and Vimal Ranchhod. 2016. ‘Inequality traps and human capital accumulation in South Africa’, in Anthony Black (ed.) Employment Intensive Growth in South Africa. Cape Town: UCT Press. Pellicer, Miquel, Vimal Ranchhod, Mare Sarr, and Eva Wegner. 2011. ‘Inequality traps in South Africa: An overview and research agenda’. SALDRU Working Paper no. 57, Cape Town. Piketty, Thomas. 2014. Capital in the Twenty-​first Century. Translated by Arthur Goldhammer. Cambridge, MA: The Belknap Press of Harvard University Press. Piraino, Patrizio. 2015. ‘Intergenerational earnings mobility and equality of opportunity in South Africa’, World Development, 67(March): 396–​405. Posel, Dorrit, Daniela Casale, and Erofili Grapsa. 2016. ‘Re-​estimating gender differences in income in South Africa: The implications of equivalence scales’, Development Southern Africa, 33(4): 425–​41. Posel, Dorrit, Daniela Casale, and Erofili Grapsa. 2020. ‘Household variation and inequality: The implications of equivalence scales in South Africa’, African Review of Economics and Finance, 12(1): 102–​22.

Inequality in South Africa    191 Schotte, Simone, Rocco Zizzamia, and Murray Leibbrandt. 2018. ‘A poverty dynamics approach to social stratification: The South African case’, World Development, 110: 88–​103. Sen, Amartya. 1992. Inequality Re-​examined. Cambridge, MA: Harvard University Press. Spaull, Nic, and Jonathan D. Jansen. 2019. South African Schooling: The Enigma of Inequality. Berlin: Springer. Statistics South Africa. 2017. Poverty Trends in South Africa: An Examination of Absolute Poverty between 2006 and 2015. Pretoria: Statistics South Africa. Statistics South Africa. 2019. Inequality Trends in South Africa: A Multidimensional Diagnostic of Inequality. Pretoria: Statistics South Africa. Stewart, Frances, 2016. ‘Changing perspectives on inequality and development’, Studies in Comparative International Development, 51(1):  60–​80. Stewart, Frances, and Arnim Langer. 2008. ‘Horizontal inequalities: Explaining persistence and change’, in Frances Stewart (ed.) Horizontal Inequalities and Conflict. London: Palgrave Macmillan. Tilly, Charles. 1998. Durable Inequality. Berkeley and Los Angeles, CA: University of California Press. Tregenna, Fiona. 2012. ‘What are the distributional implications of halving poverty in South Africa when growth alone is not enough?’, Applied Economics, 44(20): 2577–​96. Tregenna, Fiona, and Mfanafuthi Tsela. 2012. ‘Inequality in South Africa: The distribution of income, expenditure and earnings’, Development Southern Africa, 52(1):  35–​61. Visagie, Justin, and Dorrit Posel. 2013. ‘A reconsideration of what and who is middle class in South Africa’, Development Southern Africa, 30(2): 149–​67. Webster, Edward, and Francis, David. 2019. ‘The paradox of inequality in South Africa: A challenge from the workplace’, Transformation: Critical Perspectives on Southern Africa, 101:  11–​35. Wilkinson, Richard, and Kate Pickett, 2010. The Spirit Level: Why Equality Is Better for Everyone. London: Bloomsbury Press. Winker, Gabriele, and Nina Degele. 2011. ‘Intersectionality as multi-​level analysis: Dealing with social inequality’, European Journal of Women’s Studies, 18(1):  51–​66. Wittenberg, Martin. 2017a. ‘Measurement of earnings: Comparing South African tax and survey data’. REDI3x3 Working Paper no. 41, SALDRU, Cape Town. Wittenberg, Martin. 2017b. ‘Wages and wage inequality in South Africa 1994–​2011: Part 1—​ wage measurement and trends’, South African Journal of Economics, 85(2): 279–​97. Wittenberg, Martin. 2017c. ‘Wages and wage inequality in South Africa 1994–​2011: Part 2—​ Inequality measurement and trends’, South African Journal of Economics, 85(2): 298–​318. Wittenberg, Martin. 2018. ‘The top tail of South Africa’s earnings distribution 1993–​2014: Evidence from the Pareto distribution’. SALDRU Working Paper no. 224 Cape Town. Wittenberg, Martin, and Murray Leibbrandt. 2017. ‘Measuring inequality by asset indices: A general approach with application to South Africa. Review of Income and Wealth, 63(4): 706–​730. Zizzamia, Rocco, Simone Schotte, and Murray Leibbrandt. 2019. ‘Snakes and ladders and loaded dice: Poverty dynamics and inequality in South Africa, 2008–​2017’. UNU WIDER Working Paper, Helsinki. Zizzamia, Rocco, Simone Schotte, Murray Leibbrandt, and Vimal Ranchhod. 2016. ‘Vulnerability and the middle class in South Africa’. SALDRU Working Paper no. 188, Cape Town.

PA RT   I I

T H E P R I M A RY SE C TOR S , E N E RG Y, A N D T H E E N V I RON M E N T

Chapter 10

Agricu ltu re i n Sou th Af ri c a Wandile Sihlobo and Johann Kirsten

10.1 Introduction Black farmers produce between 5 and 10 per cent of total agricultural output in South Africa (NAMC 2019), a ratio that has not changed much for two generations despite evidence that it was higher in the early twentieth century (Bundy 1979; Simkins 1981; OECD 2006). Part I of this handbook shows how colonial rule, the segregationist era, and later apartheid provided the foundation for economic dualism in agriculture that excluded most Black South Africans from access to land ownership, to agricultural support services, and to economic opportunities in South Africa’s rural areas. This chapter provides a more recent review of the South African agricultural sector, from the transition years in the early 1990s to the current challenges of 2020. The discussion starts with an inventory of the natural resource base with which farmers have to work, then addresses the political economy of the sector, various aspects of the performance of the commercial agricultural sector, and finally the growth prospects for the sector against the targets set in the National Development Plan.

10.2  The Natural Resource Base South Africa is a semi-​arid country with a weak resource base for agriculture. This renders the sector inherently risky, and this is exacerbated by climate change. This shapes the structure of the sector, the composition of output and its growth prospects. Figure 10.1 highlights South Africa’s limited land use potential, with more than half of total agricultural land area classified as low-to-medium potential (Class V–​VIII), suitable only for extensive grazing systems.

196    Wandile Sihlobo and Johann Kirsten Agricultural Land Capability, Cash Crop Cultivated Land & Irrigated Land

Figure 10.1  Agricultural land capability map of South Africa Source: BFAP (2018).

Figure 10.1 shows land capabilities across the country. Ironically, some areas of high potential in the former homelands in KwaZulu-​Natal, Eastern Cape, Limpopo, North West, and Mpumalanga have high poverty, unemployment, and inequality levels (StatsSA 2014), another of the legacies of past policies which left these areas without the functioning physical, social, and institutional infrastructure required for successful farming. The land resource base also influences what is possible in terms of land reform discussed in ­Chapter 12. For this reason, it is useful to articulate the resource base in numbers and tenure status to determine the land potential and farming systems for sustainable land reform:

• • • •

Total surface area of South Africa: 122 million hectares (ha) Agricultural land: 93 million ha1 Areas under traditional tenure (former homelands): 15.5 million ha Farmland with private title deeds: 77.5 million ha.2

1  This is 4 million ha less than the estimates of the 1980s due to the loss of farmland to urban sprawl, towns, roads, and national parks. 2  Of this, 42 million ha is classified as extensive grazing land.

Agriculture in South Africa    197 In light of the important imperative of redistributive land reform within the agricultural sector, it is worth reporting on the progress with the redistribution of the 77,580  million ha farmland owned by white farmers. Our estimates, which include restitution, redistribution, private transactions, and State procurement, suggest that a total of 13.2 million ha (or 17 per cent) has already been transferred away from white landowners to the State (3.08 million ha) or Black owners (10.135 million ha) through private and State-​supported transactions, including land restitution (Vink and Kirsten 2019).3 If we add the hectares of land (2.339 million ha) that were successfully identified for restitution but for which communities elected to receive financial compensation as the means for restitution, then the total area of land rights that were restored since 1994 is 15.56 million ha. This is equivalent to 20 per cent of formerly white-​owned land, which is much closer to the 30 per cent target (of 23.25 million ha) than commonly believed.

10.3  The Political Economy of South African Agriculture The history of South African agriculture can be summarized in three words: segregation, suppression, and support. The first was the deliberate result of the Natives Land Act of 1913 and the second a process that took place throughout the twentieth century and resulted in the almost complete destruction of farming by Black people. The third came to an end in a process that started in the 1980s (when conservative white farmers stopped supporting the government of the day in favour of the right-​wing opposition), was given additional impetus by accession to the trade rules of the World Trade Organization in 1994, and was then cemented with deep liberalization and deregulation by the new government after April 1994. These changes left commercial farmers (large and small ​scale) without the considerable government support that they had enjoyed up to that time (OECD 2006). The withdrawal of this support to white farmers had two consequences. First, it allowed the growth of very large-​scale (‘mega’) farming operations (about 2,600 (or 6.5 per cent) of them), especially (but not exclusively) in intensive irrigated horticulture production (StatsSA 2020a). Second, it was accompanied by the abolition of support measures, from direct subsidies to indirect market interventions, from funding of research and extension to the withdrawal of subsidies on conservation works (Vink 2000). The result was that Black farmers were bereft of the support services that they had been denied under the previous regimes.

3 Updated

in November 2020 with latest data on land transfers recorded by the Deeds Office, obtained from Bornmann (2020).

198    Wandile Sihlobo and Johann Kirsten The many attempts to remedy this situation through a range of plans and programmes have been ex post, piecemeal, and unsuccessful. Disappointingly, implementation and execution were inadequate and/​or uncoordinated, resulting in frustration and failure. In response, new plans were developed without addressing the root causes of failure and frustration. As a result, the dualism that has divided the sector has not abated. These plans included, among others:

• • • • • • • • • • •



• • • • • • •

The White Paper on Agricultural Policy, 1995 The Strauss Commission Report into the Provision of Rural Financial Services, 1996 Broadening of Access to Agriculture Thrust, 1996 The Integrated Food Security and Nutrition Programme, 1996 Integrated Sustainable Rural Development Strategy, 2000 The Land Reform for Agricultural Development Programme, 2001 The Strategic Plan for South African Agriculture, 2001 Comprehensive Agricultural Support Programme, 2004 The Micro Agricultural Financial Institutions of South Africa initiative, 2004 The Proactive Land Acquisition Strategy, 2006 The agricultural and land chapters of the Accelerated and Shared Growth Initiative for South Africa, 2007 Recapitalisation and Development Programme, 2010 The agricultural and land chapters of the National Development Plan, 2011 The Zero Hunger programme, 2011 The Ilima Letsema project, 2011 The ‘one household one hectare’ programme, 2015 The Agri parks initiative, 2015 Operation Phakisa, 2017.

While these plans were being devised, drafted, and launched, but not implemented, the value of South African agricultural output more than doubled in real terms (DALRRD 2020). This growth has largely been driven by increased productivity, which has been underpinned by technological innovation (section 10.7), as well as growth in traditional export markets and access to new ones (see section 10.8), and has spanned all subsectors of agriculture (livestock, horticulture, and field crops). Black farmers were largely excluded from the benefits of this agricultural growth while the various programmes and plans were not sufficiently broad-​based to foster an inclusive and prosperous sector and, combined with the slow pace of land reform, contributed to frustrations amongst Black farmers. As a result, the organizations representing farmers in South Africa proliferated along racial lines (as shown in Table 10.1), all trying to extract rent from various government programmes without necessarily benefiting their (unaudited) memberships. As a result, debates and negotiations around government policies, programmes, and plans have become a contested space, with little chance of successful implementation as all programmes end up accommodating the opposing views of these vested interests.

Agriculture in South Africa    199 Table 10.1: The landscape of organized agriculture and the various national farmer organizations Name

Membership

Comments

Agri SA

25,000

Mainly white commercial farmers with some few Black members

National African Farmers’ Union (NAFU)

Not able to verify claims of membership

Used to be the main representative body for Black farmers. Infighting led to a split and the formation of AFASA.

African Farmers Association of South Africa (AFASA)

3,000

Focus on promoting the interests of African (Black) commercial farmers

TLU-​SA

5,500

Only white commercial farmers and not aligned to Agri SA and mostly on the right of the political spectrum

Southern Africa Agri Initiative (SAAI)

2,630

A network of family farms to protect the future and sustainability of family farms (160 Black farmers paid-​up). Most are also members of the other organizations listed here.

Source: Managers of the various organizations.

The fractious nature of organized agriculture is often repeated at the commodity level and stems not only from opportunism, but also from the fundamental mistrust between organizations, despite the attempt to form a broader organizational front through the establishment of the Agri Sector Unity Forum (ASUF) in November 2012. One can assume that this mistrust is partially responsible for the lack of unity in the sector.4 Moreover, there is a lack of collaboration between government and the private sector as manifested through commodity organizations, agribusinesses, commercial farmer organizations, etc. This results in slow implementation of the aforementioned farmer development plans. The farmer organizations are frustrated with the lack of delivery and extreme bureaucracy. When they do try and implement transformation initiatives, it is sometimes not supported by government officials, primarily at provincial level and in the municipalities, and as a result the farmers who need and deserve support the most are underserved. In all, the problems that confront South Africa’s agricultural sector will not be resolved without a unified vision and commitment. In its absence, the chances of creating a united, inclusive, and prosperous agricultural sector are getting slimmer by the day.

4  We base this view on an assessment of media reports of dissenting views of these organizations on policy matters such as land reform, farmer support programmes, and minimum wage policies, amongst others.

200    Wandile Sihlobo and Johann Kirsten

10.4  The Role and Contribution of Agriculture in  the South African Economy The fundamental role that agriculture plays in economic development and structural transformation has long been recognized. In the seminal work on the subject by Johnston and Mellor (1961) and Mellor (1989) and more recently Timmer (2002), agriculture is seen as a source of contributions that helped induce industrial growth and a structural transformation of the economy. In this process agriculture plays a number of key roles: increased food supply, release of surplus labour, a market for industrial output, supply of savings, and increased foreign exchange earnings. However, globalization, integrated value chains, rapid technological and institutional innovations, and environmental constraints have deeply changed the context for agriculture’s role. Byerlee et al. (2009) therefore argue that agriculture has in fact multiple functions for development:  triggering economic growth, reducing poverty, narrowing income disparities, providing food security, and delivering environmental services. Nevertheless, in this section we only discuss agriculture’s contribution to economic growth. The foreign exchange role is highlighted in the section on agricultural trade and the food security role is partially addressed in the section on food prices. We present a high-​level overview of the economic contribution of South African agriculture in Table 10.2. Despite a consistent real growth in the gross value ​added in the agricultural sector, agriculture’s relative share in the total economy continued to decline to around 2.1 per cent for most of the last decade. Annual growth of the sector is also extremely volatile, varying, for example, from a negative growth of –​12.1 per cent in 2016 to 22.7 per cent in 2017. Contraction of the agricultural sector happened again in 2018 and 2019 but there was a robust recovery of 13.1 per cent in agriculture gross value added in 2020 (StatsSA, 2021). It is helpful to further unpack the output mix of South African agriculture and establish the top ten commodities according to production value (Table 10.3). The list remains largely stable and the top ten commodities continue to represent 75 per cent of the value of agricultural output. In fact, the top three remained the same over the last three decades: maize, poultry, and beef production—​it is only their relative position that has changed over time. Well-​known industries such as wine grapes, sheep, potatoes, and wool do not feature in the top ten but at least appear in the top twenty. The rapid rise of citrus production

Agriculture in South Africa    201 Table 10.2: Structural changes in the South African agricultural economy, 1990–​2019 Decade averages Indicator

Unit

1990 –​99

2000–​09

2010–​19

Agricultural GDP contribution Rand (current)

million

19,398

37,989

69,071

Rand (2010 prices)

million

39,085

46,049

55,167

Agric share of GDP

%

3.4

2.6

2.1

Avg annual growth

% p.a.

2.7

2.6

0.9

Net farm income

R million

6,959

26,396

60,375

R million (2010 prices)

18,569

34,194

57,691

R million

33,933

96,220

215,662

R million (2010 prices)

83,352

134,043

185,127

Livestock

%

43

41

48

Crops

%

35

34

24

Horticulture

%

22

25

28

Value of production

Share of production

Source: Authors’ estimates using StatsSA National Accounts data as well as data from Abstract of Agricultural Statistics (DALRRD 2020).

since 1994 is remarkable—​from being ranked 13th in 1994, to increasing to fourth position in 2018. There is, however, reason to believe that there is a potential undercount of the economic contribution of the agricultural sector. This is partly due to the large subsistence and unrecorded commercial activity in the agricultural sector. In addition, the release of the 2017 Commercial Agricultural Census by StatsSA in March 2020 confirmed this view. Table 10.4 shows the official items included in GDP calculations compared to the new census numbers. Across the board, the numbers in the current national accounts for agriculture are underestimated by around 30 per cent. What makes this even more concerning is that the census excluded around 214,800 farmers with a potential gross farm income of up to Rand 80 billion (see section 10.5). This argument suggests that it is now an opportune time to readjust the national accounts for agriculture.

Table 10.3: The output mix for South African agriculture, top ten commodities in selected years according to production value Production years Rank

1994

2000

2010

2018

1

Maize

Maize

Fowls slaughtered

Fowls slaughtered

2

Fowls slaughtered

Fowls slaughtered

Maize

Cattle and calves slaughtered

3

Cattle and calves slaughtered

Cattle and calves slaughtered

Cattle and calves slaughtered

Maize

4

Milk

Milk

Deciduous fruit

Citrus

5

Deciduous fruit

Deciduous fruit

Milk

Deciduous fruit

6

Wheat

Wheat

Vegetables

Milk

7

Vegetables

Sugar cane

Eggs

Vegetables

8

Eggs

Vegetables

Citrus

Eggs

9

Sugar cane

Citrus

Sugar cane

Other livestock products

10

Hay

Eggs3

Other livestock products

Sugar cane

Share

74.85%

74.39%

74.36%

74.78%

Source: Authors’ deductions using StatsSA National Accounts data as well as data from Abstract of Agricultural Statistics (DALRRD 2020).

Table 10.4: Agriculture’s contribution to GDP according to the agricultural census and national accounts 2017 value in Rand (’000) Items

Census

Official StatsSA National Accounts

% deviation

Total income

332,756

231,251

–​30.5

Intermediate consumption

188,467

136,157

–​27.8

GDP

144,289

95,094

–​34.1

36,761

24,695

–​32.8

3.4

2.3

–​32.7

Compensation of employees % Contribution to GDP

Source: Own compilation from StatsSA (2020a).

Agriculture in South Africa    203

10.5  Size and Structure of Commercial Agriculture There is a long-standing debate about the real number of commercial farms in South Africa, largely caused by the way official statistics on farm numbers are gathered. The last comprehensive agricultural census was done in 1993, while the agricultural censuses in 2007 and 2017 used only the farm businesses on the StatsSA business register as the sample frame. The agricultural survey by StatsSA in the in-​between years also relied only on the business register as the basis for the survey (see Table 10.5). We now unpack the two latest agricultural censuses (2007 and 2017) to obtain an understanding of the structure of South African commercial agriculture (see Tables 10.6 and 10.7 below). In both cases the sample frame was restricted to farm businesses registered to pay VAT, despite pleas from academia for a comprehensive census, largely because of a lack of funding. We were nevertheless able to expand the findings from the censuses with numbers from the 2011 population census, and the 2016 Community Survey, in order to get to a better understanding of the total number of commercial farming units in South Africa (see Table 10.8). Data from the 2011 population census (extracted from three agricultural questions included in the census) show that 2,879,638 households out of South Africa’s total population, or 19.9 per cent of all households, are active in agriculture (StatsSA 2013). The census report on the agricultural households established that only 2 per cent of these households reported an annual income derived from agriculture above R307,000 (StatsSA 2013).5 This translates into 57,592 households that can be considered as commercial farmers, with agriculture as the main or only source of household income. This refutes the common belief that the number of commercial farms in South Africa is declining. There are also another 349,000 households with agricultural income between R38,000 and R307,000, probably part-​time farmers with farm income serving as an additional income source. The main point again is that the 2007 agricultural census undercounted the number of commercial farmers by at least 366,600 households, but certainly by at least 17,500 farming businesses that are not listed on the business register. The 2017 agricultural census report shows that 40,122 commercial farmers, almost exclusively small family-​based operations, produce more than 80 per cent of the total value of agricultural output (StatssA 2020a). However, as was the case with the 2007 census, the 2017 agricultural census excludes 92,634 households who practise commercial farming as their main source of income,

5  We need to point out that the income refers to income for the head of household which could be derived from other sources. Although this might be the case, the number of households in this category seems to correspond with reality.

Table 10.5: Statistics on farm numbers and farm employment from the agricultural censuses and selected agricultural surveys, 1990–​2017 No. of commercial farming units

Paid employees (permanent and seasonal)

Year

Source

Sample frame

1990

Agricultural Survey

All farms but excluded Black farmers in former homelands

62,084

1,184,676

1993

Agricultural Census

All farms but excluded Black farmers in former homelands

57,980

1,093,265

1994

Agricultural Survey

All farms but excluded Black farmers in former homelands

59,828

922,000

1995

Agricultural Survey

All farms but excluded Black farmers in former homelands

60,901

891,000

2007

Agricultural Census

VAT-​registered farms only

39,966

1,328,600

2008

Agricultural Survey

Business register: income tax and VAT-​registered

64,162

814,524

2009

Agricultural Survey

Business register: income tax and VAT-​registered

66,606

849,782

2016

Agricultural Survey

Active VAT-​registered farms only

Not available

723,767

2017

Agricultural Census

Active VAT-​registered farms only

40,122

757,628

Source: Relevant StatsSA reports. Note: The annual agricultural surveys used only 10 per cent of the sample frame.

Table 10.6: Farm structure in South Africa, 2007 Category (gross farm income)

Number of farming units

Share of commercial farming units (%)

R5 million and more (large)

2,927

7.3

R3 million to R5 million (medium)

2,172

5.4

R500,000 to R3 million (small)

12,290

30.7

Less than R500,000 (micro)

22,577

56.5

Total (VAT-​registered famers)

39,966

100

Source: StatsSA (2010).

Agriculture in South Africa    205 Table 10.7: Farm structure in South Africa, 2017 Category (gross farm income)

Number of farming units

Share of commercial farming units (%)

Share of total gross farm income (%)

R22.5 million and more (large)

2,607

6.5

67.0

R13.5 million to R22.5 million (medium)

1,847

4.6

9.7

R2.5 million to R13.5 million (small)

10,712

26.7

18.5

6,219

15.5

2.9

R1 million to R2.5 Million (micro) Less than R1,000,000 (micro)

18,737

46.7

1.9

Total (VAT-​registered famers)

40,122

100

100

Source: StatsSA (2020a).

and a further 122,200 households who practise commercial farming as a secondary source of income (estimates from the Community Survey, 2016 as shown in Table 10.8). One can therefore assume that around 214,800 farming households (Black and white), who practise some form of commercial farming, were excluded from the agricultural census. Most of these are micro-​enterprises with gross farm incomes below R500,000

Table 10.8: Categories of agricultural households according to population groups, 2016 Purpose of households’ agricultural activity (percentage of households)

Population group

Number of agricultural households

Main source of household income

Extra source of household income

Main source of household food

Extra source of For household leisure/​ food hobby Other

Total

Black African 2,116,281

4.4

4.4

46.1

38.6

5.1

1.4

100

Coloured

56,686

4.0

5.8

35.2

33.9

18.5

2.7

100

Indian/​Asian

12,716

4.5

2.8

28.5

36.5

25.7

2.1

100

143,362

22.2

8.4

19.4

25.1

21.7

3.1

100

2,329,045

5.7

4.7

43.7

37.5

6.8

1.6

100

White Total

Source: StatsSA (2020b), based on additional analysis of 2016 Community Survey. Note: The Community Survey of 2016 established that 83.8 per cent of the 2,329 million agricultural households practice agriculture in their backyards (i.e. for subsistence or for leisure purposes).

206    Wandile Sihlobo and Johann Kirsten per annum. The best guesstimate for the total gross farm income for these farms could be anything between Rand 23 billion and Rand 80 billion. Commercial farming in South Africa therefore consists largely of small-​scale family-​ based operations with almost 90 per cent of all VAT-​registered commercial farming businesses classified as micro-​or small-​scale enterprises, responsible for only 23 per cent of total farm income but 37 per cent of all farm employment. The 2,610 large farms (with a turnover on average above R22.5  million per annum) are responsible for 67 per cent of farm income and employed more than half the agricultural labour force of 757,000 farm workers in 2017. The distribution of the different farm sizes across the main farming activities is presented in Table 10.9. The facts presented here should allow a more nuanced interpretation of South Africa’s farm structure. Not all white commercial farm operations are ‘large ​scale’ and not all Black farmers are ‘small s​ cale’, ‘subsistence’, or ‘emerging’, and most farm operations can be classified as micro or small in scale. We now turn to the race dimension of the farm structure, given that transformation of the agricultural sector is also an important imperative for the future of South African agriculture. Based on the information provided in Table 10.8, there are 242,221 commercial farming households in South Africa of whom only 43,891 (18 per cent) are white commercial farmers. If we consider only the agricultural households with agriculture as their main source of income surveyed in the 2016 Community Survey (StatsSA, 2016), Table 10.9: Number of farms/​farming units in the commercial agriculture industry by activity and size group, 2017 Income size categories Activity

Large

Medium Small

Micro

Total

Growing of cereals and other crops

387 (15%)

351 (19%)

2,123 (20%)

5,698 (23%)

8,559 (21.3%)

Horticulture

649 (25%)

366 (20%)

1,600 (15%)

2,028 (8%)

4,643 (11.6%)

Farming of animals

703 (27%)

408 (22%)

3,023 (28%)

9,505 (38%)

13,639 (34%)

Mixed farming (growing of crops combined 812 with farming of animals) (31%)

689 (37%)

3,720 (35%)

7,237 (29%)

12,458 (31.1%)

Agricultural services and fertilizer production

59 (2%)

43 (2%)

247 (2%)

474 (2%)

823 (2.1%)

South Africa

2,610 1,857 10,713 (100%) (100%) (100%)

24,942 (100%)

40,122 (100%)

Source: StatsSA (2020a).

Agriculture in South Africa    207 we end up with a total of 132,700 households of whom 93,000 (70 per cent) are Black farmers.6 The 2017 agricultural census report did not present specific results on the population group of the farm owner/​operator, but additional analysis of the 2017 census shows that 2,826 (7 per cent) of the VAT-​registered farm businesses in 2017 are owned/​operated by Black people. They contribute: • 19 per cent of the total value of animal and animal products • 14 per cent of the total value of field crop production • 14 per cent of the total value of horticultural production. These estimates only refer to the VAT-​registered farm businesses, but somehow correspond to estimates by the National Agricultural Marketing Council obtained from reports of the various commodity groups as shown in Table 10.10 below.

Table 10.10: Share of Black farmers in commercial agricultural output, average 2015–​19 Commodity

Black farmer share in output

Maize

4.7%

Soybeans

3.1%

Wheat

1.3%

Cotton

2.4%

Citrus

12.8%

Deciduous fruit

10.2%

Viticulture

1.6%

Potato

1.0%

Tomato

8.6%

Wool

11.0%

Mohair

12.8%

Cattle

34.0%

Poultry

4.2%

Source: NAMC (2019).

6 

These numbers help to put the membership claims of the various farmers’ organizations (Table 10.1) into perspective.

208    Wandile Sihlobo and Johann Kirsten

10.6 Farm Employment Agriculture is the eighth largest employer in the South African economy and has been identified as a strategic area for employment creation. The National Development Plan (NDP), estimates that ‘agriculture has the potential to create close to one million new jobs by 2030, a significant contribution to the overall employment target’ (The Presidency 2012). Nevertheless, this target comes on the back of a declining trend in agricultural employment, especially if viewed over a long period, as illustrated in Figure 10.2. The number of people employed in the South African agricultural sector declined from an average of 1.04 million in the 1990s to around 775,000 recently (Liebenberg and Pardey 2012; StatsSA, 2020a). These numbers account for both seasonal, temporary, and permanent labour, with seasonal labour accounting for roughly half in 2020. Moreover, the share of agriculture in overall employment in South Africa has also declined from 10 per cent in the 1990s to 6 per cent in the 2010s. In addition to the actual decline in employment in the sector, an expansion in labour participation in other sectors of the economy has contributed to the declining share. This is partially the result of labour policy changes in the early days of democracy, as well as mechanization trends in the recent past (Liebenberg and Kirsten 2013). The NDP aims to change this declining job trajectory by implementing a three-​tier approach. This includes the development of underutilized land largely in the communal areas and among failed land reform farms (approximately 400,000 jobs); the expansion of export-​led high-​growth industries (approximately 250,000 jobs); and investment in

1,200 1,000

000 jobs

800 600 400 200 –

1990s

2000s

2010s

Figure 10.2  Long-​term trend of South Africa’s agricultural labour Source: Liebenberg and Pardey (2012) and StatsSA (2020b). Note: The entries in the chart represent decade averages and the ‘2010s’ data ends in 2019.

Agriculture in South Africa    209 agro-​processing with integrated up ​and ​downstream linkages (approximately 350,000 jobs) (The Presidency 2012). By 2020, a total of 250,000 new jobs had been created through an expansion of the export-​led and high-​value industries (BFAP 2020a), mainly in industries such as citrus, tree nuts, apples, table grapes, avocadoes, and blueberries (BFAP 2020a). These increases have offset the decline in jobs in field crops and livestock production. Unfortunately the number of jobs on underutilized crop land and in the agro-​processing sector declined over this period (BFAP 2020a).

10.7  Agricultural Input Industry South African farmers spend around Rand 154 billion annually on intermediate inputs, which include packing material, plant protection chemicals, fertilizer, seeds, veterinary medicine, and animal feed, and a further Rand 10 billion on machinery, tractors, and implements (DALRRD 2020). Most of these inputs are imported. For instance, it is estimated that more than 80 per cent of domestic fertilizer demand and approximately 98 per cent of agro-​chemicals and 80 per cent of machinery is imported (Sihlobo 2016). This implies that local prices are subject to the same supply-​and-​demand forces that drive international markets, currency exchange rates, and shipping and distribution costs. Products related to mechanization, such as tractors, implements, machinery, and parts account for the greatest share of imports (Rand 8.2 billion), followed by fertilizers (Rand 7.5 billion), animal feeds (Rand 6.1 billion), and plant protection chemicals (Rand 6 billion) (DALRRD 2020). At an aggregate level, the major sourcing regions are the European Union, the United States, China, and Argentina (BFAP 2020b). The animal–​maize production nexus remains a dominant part of South African agriculture. Maize makes up between 50 per cent and 60 per cent of the rations of poultry, pigs, dairy cattle, beef cattle, etc.). Another major ingredient is soybean oilcake which was historically imported. Due to major investments in crushing facilities and the expansion of soybean production since 2015, South Africa is now able to produce approximately 75 per cent of its local demand, with the shortfall imported from Argentina (BFAP 2020b). A large share of other inputs related to the animal feed industry, such as amino acids, vitamins, and minerals, are procured from China. In 1990, South Africa’s fertilizer consumption was 785,000 tonnes, but fell to 674,000 tonnes in 2010 before increasing to 705,000 tons in 2015, primarily supported by expansion in horticulture plantings and also soybeans. Major fertilizer plant closures in the late 1990s saw the share of imports in South Africa’s fertilizer consumption increase from 20 per cent in 1990 to 95 per cent in 2015. There have been considerable advancements in seed technology in South Africa, largely based on a combination of varietal innovations in the public agricultural research system and the adaptation of imported germ plasm (Nhemachena and Kirsten

210    Wandile Sihlobo and Johann Kirsten Table 10.11: Annual sales of tractors and combine harvesters

Tractors Combine Harvesters

2000

2005

2010

2015

2019

2,668

4,475

5,155

6,602

5,270

126

150

185

216

148

Source: SAAMA (2020).

2017). In addition, South Africa was one of the few countries in the world that quickly adopted genetically modified seeds and is now one of the world’s largest users as a percentage of total commercially planted crops (especially maize and soyabeans) (Gouse et al. 2016). Innovations in tractors and cultivation equipment have also contributed to productivity growth by helping farmers to increase yields (by managing soil moisture and soil quality better), save fuel and maintenance costs, and save labour costs. Gandidzanwa (2018) reflected on the evolution of the market for tractors in South Africa, noting that two-​wheel-​drive machines dominated the tractor market from 1987 to 1994, but from 1994 four-​wheel-​drive tractors became dominant. From 1987 to 1992, tractor sales declined at an annual average rate of –​11.97 per cent, then increased by 27.95 per cent between 1992 and 1996. As Table 10.11 illustrates, this increase continued after 2000.

10.8  Agricultural Trade Patterns and Trade Policies South Africa’s agricultural trade policy has gone through various changes since 1994. South Africa’s signing of the Marakech Agreement in 1994 on world trade provided the initial steps in agricultural trade liberalization (OECD 2006), supported by the process of agricultural market deregulation since the early 1990s and which eventually culminated in the Marketing of Agricultural Products Act (1996). In this process, South Africa undertook significant import tariff cuts from around 23 per cent in the early 1990s to as low as 8.2 per cent (DTIC 2010). Moreover, South Africa also simplified its tariff regime. According to DTIC (2010), in 1990, South Africa’s tariff schedule consisted of 13,609 tariff lines, with 28 per cent of these subject to import controls. By the early 2000s, the number of tariff lines under import controls had declined by more than half to 6,420. Meanwhile, export controls were eliminated. Also, worth noting is that the average tariff on agricultural products is around 9.4 per cent. In 2020, South Africa had about 1,043 agricultural tariff lines, but approximately half

Agriculture in South Africa    211 Table 10.12: South Africa’s agricultural exports and imports, 2010 prices Exports

Avg Annual Growth Real Avg value rate (R Billions)

Avg Annual Growth rate

37

5.3%

24

6.0%

2000–​09

61

3.5%

42

9.5%

2010–​19

124

5.3%

85

3.7%

Period 1990–​99

Real Avg value (R Billions)

Imports

Source: Quantec (2020) and authors’ calculations.

of these lines carry a zero per cent duty as South Africa continued to enter into trade agreements with various countries. While the reduction in import tariffs opened the doors for an increase in agricultural imports, South Africa also benefited from efficiency gains and access to global export markets. South Africa’s agricultural exports accounted for 30 per cent of the gross value of production in 1990, increasing to 54 per cent in 2018 (DALRRD 2020) (see Table 10.12). Given that over half of what South Africa produces in value terms goes to export markets, it can be argued that the country is de facto export-​orientated and the expansion in production after the deregulation of agricultural markets has, in part, been underpinned by the expansion in trade (along with advancement in technology which has enabled an improvement in yields of most commodities and crops). South Africa’s agricultural exports grew from Rand 6 billion in 1990 to Rand 150 billion in nominal terms in 2019. Imports have also increased over the same period, although at a considerably slower pace. The imports were mainly underpinned by products that South Africa doesn’t produce sufficient volumes of, including rice, coffee, tea, wheat, and palm oil. Most importantly, since South Africa deregulated its agricultural markets, one of the benefits that its trading partners have enjoyed is stability and certainty, which has not always been the case in neighbouring countries. Horticultural products, specifically citrus, wine, grapes, apples and pears, and macadamia nuts have been the key drivers of exports and grew notably from the 2000s as shown in Figure 10.3. Beverages and spirits have been the second leading subsector of agricultural exports and the key product is wine. Meanwhile, in field crops, maize remains the primary driver of agricultural exports. In terms of livestock, live sheep, wool, and cattle and increasingly beef in the 2000s, are the drivers underpinning growth in agricultural exports. South Africa’s agricultural export markets have increasingly become well diversified since the dawn of democracy, with the country deepening and broadening its cooperation through trade agreements. These include the Southern African Customs Union (SACU), the Southern African Development Community (SADC), the Tripartite Free Trade Area (T-​FTA), and the nascent African Continental Free Trade Area (AfCFTA)

212    Wandile Sihlobo and Johann Kirsten 50 45 40

Billion Rand

35 30 25 20 15 10 5 0

1990

1993

1996

1999

2002

2005

2008

2011

2014

2017

Meat and edible meat offal Dairy produce Edible vegetables Edible fruit and nuts Cereals Beverages and spirits

Figure  10.3 Total agricultural exports by selected major products, South Africa, real 2010 values Source: Quantec (2020) and authors’ calculations.

within the African continent. Within the European context, South Africa joined the Economic Partnership Agreement (EPA) between the European Union and the SADC EPA group (Botswana, Lesotho, Namibia, Mozambique, South Africa, and Eswatini) to establish a regional agreement with the European Union and to secure further market access especially in agriculture (Mlunmbi-​Peter 2019). In the post-​Brexit environment, South Africa joined the SACU-​Mozambique-​UK Economic Partnership Agreement, which is an economic partnership agreement between the United Kingdom, the Southern African Customs Union (SACU), and Mozambique which was signed in September 2019. Within the South America region, South Africa has access through the SACU–​MERCOSUR Preferential Trade Agreement (PTA). In 2019, the African continent, Europe, and Asia constituted the largest markets for South Africa’s agricultural exports, respectively absorbing 41 per cent, 26 per cent, and 24 per cent of the value of exports (Trade Map 2020). The leading products to these markets were beverages, fruit, vegetables, sugar, wool, and grains.

10.9 Food Prices Is the South African agricultural sector able to produce affordable and safe food for the largely urban population? Anecdotal evidence driven by an uninformed media, especially during periods of rapid food inflation, has often overshadowed public perceptions on this issue by arguing that the ‘concentration’ in agriculture and in the food processing and retail sectors has contributed to sharp increases in food prices.

Agriculture in South Africa    213 A case in point is when the sharply depreciating exchange rate at the end of 2001 and early 2002, combined with a shortage in staple foods in the SADC region caused the prices of basic commodities, such as maize, wheat, and sunflower seeds, to rise dramatically. This was subsequently followed by sharp increases in retail prices of basic foodstuffs. Sporadic drought conditions in later years (2011/​12 and 2015/​16) in the summer rainfall regions of South Africa, combined with extremely high international commodity prices, also pushed food price inflation higher (see Figure 10.4). The analysis of the consumer price index for food and non-​alcoholic beverages issued by StatsSA illustrates that South Africa continues to experience high volatility in food prices, characterized by periods of high food inflation and spells of low food inflation. Since January 2018, however, year-​on-​year food price inflation has stabilized on or below 4 per cent (Figure 10.4). The final report of the Food Pricing Monitoring Committee (FPMC 2003) shows that the long-​term trend in food prices has no abnormalities. Importantly, in real terms food prices remained almost the same, that is, the agricultural sector has to a large extent been able to supply food without any significant increases in real prices. Agriculture does however, experience noticeable volatility in prices due to many external factors, for example, international market shocks and the weather, and the effect on the year-​on-​ year increase in food prices is clearly shown in Figure 10.4. The general hype about food prices (granted that it is critical in the food security context) also shows that there is little appreciation for the fact that all food originates from the farming sector where farmers have to deal with rapidly rising input costs, extreme weather risks, and extreme variability of market prices caused by sharp fluctuations in production levels and world prices. Farmers produce the basic raw commodity to be processed for the final consumer product (i.e. wheat-​to-​bread; maize-​to-​maize

18.0% 16.0% 14.0% 12.0% 10.0% 8.0% 6.0% 4.0% 2.0% Apr-20

Nov-19

Jan-19

Jun-19

Aug-18

Oct-17

Mar-18

May-17

Jul-16

Dec-16

Feb-16

Apr-15

Sep-15

Nov-14

Jan-14

Jun-14

Aug-13

Oct-12

Mar-13

May-12

Jul-11

Dec-11

Feb-11

Apr-10

Sep-10

Nov-09

Jan-09

Jun-09

0.0%

Figure 10.4  Annual food inflation rate, January 2009 to July 2020 Source: Authors’ estimations based on StatsSA CPI for food and non-​alcoholic beverages.

214    Wandile Sihlobo and Johann Kirsten meal; sunflower seeds-​to-​cooking oil; milk-​to-​cheese) and when farmers sell these commodities they usually have no ability to influence market prices since so many produce the same product and sell to only a few big buyers. Because of the fully deregulated and liberalized domestic market for agricultural products, the prices farmers receive are either a function of the combination of world prices, exchange rates, and transport costs (in the case of maize, wheat, rice, sunflower, soybeans), or it is negotiated between farmers and agri-​food business (e.g. dairy, processed potatoes, sugar). In these circumstances farmers have little or no bargaining power. This, however, does not ignore the fact that higher up in the value chain—​at the processing and retail level, there are many cases of anti-​competitive behaviour such as the bread, flour, and maize meals cartels which were uncovered by the Competition Commission of South Africa in 2007 (Competition Commission 2015).

10.10  Summary and Conclusion The South African agriculture sector remains an important sector to drive growth and job creation in rural provinces. This is a vision that was highlighted in ­chapter 6 of the National Development Plan (NDP). But this broad vision will only be achieved if it is followed up with detailed operational plans to guide the officials and various stakeholders about implementation at the local level. This is a crucial part which has been lacking since the publication of the NDP in 2012. Despite remarkable growth during the last three decades, South Africa’s agricultural sector remained plagued by dualism, mistrust, and suboptimal performance. On the one hand, South Africa agriculture has surpassed the NDP targets in expanding a number of high-​value commodities (citrus, macadamias, apples, table grapes, avocados, dairy, and pork), but on the other hand, the country has not fully achieved the jobs target and expansion of agriculture in the former homelands. The dualistic nature of the sector remains therefore entrenched. The only way this can change is through a capable and effective State (provincial and national); a stable and conducive policy and investment environment; infrastructure development and services including electricity and water; and effective farmer support programmes, amongst other support measures. This chapter also provided a useful insight into the structural realities of South African agriculture. The expansion and growth of the industry over the last three decades was driven by new technology (irrigation, cultivation techniques, genetic material, etc.) and by the 2,607 large farm enterprises that have capital and systems in place to invest, to expand, and to export. Although these large enterprises are responsible for 67 per cent of total farm output, the lifeblood of the agricultural sector remains the many small family farms. More than 90 per cent of all commercial farming units in South Africa are small family-​based operations at different levels of commercial activity. This is an important point of this chapter and should guide the way government support systems and regulations are implemented to advance agricultural growth and employment.

Agriculture in South Africa    215 If this is understood and appreciated, the growth and employment contribution of the sector could be substantive and could be more than what is reported in this chapter. There remains a great potential for the sector to provide more full-​time livelihoods and more wage employment if the remaining constraints (political, infrastructure, regulatory, and service) are addressed.

References Bornmann, Johann. 2020. Personal communication based on analysis of farm land transactions recorded at the deeds office [interview], 15 November. Bundy, Colin. 1979. The Rise and Fall of the South African Peasantry. London: Heinemann. Bureau for Food and Agricultural Policy (BFAP). 2018. Baseline Agricultural Outlook 2018–​ 2027. Pretoria: Bureau for Food and Agricultural Policy (BFAP). Bureau for Food and Agricultural Policy (BFAP). 2020a. Baseline Agricultural Outlook 2020–​ 2029. Pretoria: Bureau for Food and Agricultural Policy. Bureau for Food and Agricultural Policy (BFAP). 2020b. The Use of Agricultural Inputs in South Africa. Pretoria: Bureau for Food and Agricultural Policy. Byerlee, Derek, Alain De Janvry, and Elisabeth Sadoulet. 2009. ‘Agriculture for development: Toward a new paradigm’, Annual Review of Resource Economics, 1: 15–​31. Competition Commission. 2015. Why Some Industries Seem Prone to Endemic Collusion. Paris: OECD Competition Division. Department of Agriculture, Land Reform and Rural Development (DALRRD). 2020. Abstract of Agricultural Statistics. Pretoria:  Department of Agriculture, Land Reform and Rural Development (DALRRD). Department of Trade, Industry and Competition (DTIC). 2010. Trade Policy and Strategy Framework. Pretoria: Department of Trade, Industry and Competition. Food Price Monitoring Committee (FPMC). 2003. Final Report of the Food Pricing Monitoring Committee. Pretoria: National Agricultural Marketing Council. Gandidzanwa, Colleta. 2018. ‘The changing nature and quality of machinery in South African agriculture’, unpublished PhD Thesis, University of Pretoria. Gouse, Marnus. Debdatta Semgupta, Zambrano Patricia, and José Falck Zepeda. 2016. ‘Genetically modified maize: Less drudgery for her, more maize for him? Evidence from smallholder maize farmers in South Africa’, World Development, 83: 27–​38. Johnston, Bruce and John Mellor. 1961. ‘The role of agriculture in economic development’, American Economic Review, 87(2): 566–​93. Liebenberg, Frikkie, and Johann Kirsten. 2013. ‘Statistics on farm labour in South Africa’. Unpublished Working Paper, University of Pretoria. Liebenberg, Frikkie, and Philip Pardey. 2012. ‘A long-​run view of South African agricultural production and productivity’, African Journal of Agricultural and Resource Economics, 7(1): 14–​38. Mellor, John. 1998. ‘Agriculture on the road to industrialization’, in Carl Eicher and John Staatz (eds) International Agricultural Development. Baltimore, MD:  Johns Hopkins University Press. Mlunmbi-​ Peter, Xolelwa. 2019. Update on South Africa’s Trade Agreements. Pretoria: Department of Trade, Industry and Competition.

216    Wandile Sihlobo and Johann Kirsten National Agricultural Marketing Council (NAMC). 2019. Status Report on Statutory Measures Implemented in terms of the Marketing of Agricultural Products Act (Act N0.47 of 1996). Pretoria: National Agricultural Marketing Council (NAMC). Nhemachena, Charity, and Johann Kirsten. 2017. ‘A historical assessment of sources and uses of wheat varietal innovations in South Africa’, South African Journal of Science, 113(3/​4), Art. #2016-​0008, 8 pages. http://​dx.doi.org/​10.17159/.​ Organisation for Economic Co-​ operation and Development (OECD). 2006. Review of Agricultural Policies—​South Africa. Paris:  Organisation for Economic Co-​operation and Development. Quantec. 2020. EasyData: International Trade Service Pretoria: Quantec. Sihlobo, Wandile, 2016. ‘An Evaluation of Competitiveness of South African Maize Exports’, unpublished MSc Thesis, Stellenboch University. Simkins, Charles, 1981. ‘Agricultural production in the African Reserves of South Africa, 1918–​ 1969’, Journal of Southern African Studies, 7(2): 256–​83. Southern African Asset Management Association (SAAMA). 2020. South Africa’s Tractors and Combine Harvesters Sales. Edenvale:  South African Agricultural Machinery Association (SAAMA). StatsSA. 2010. Census of Commercial Agriculture 2007. Pretoria:  Statistics South Africa (StatsSA). StatsSA. 2013. Census 2011:  Agricultural Households Report no.  03-​11-​01. Pretoria:  Statistics South Africa. StatsSA. 2014. Poverty Trends in South Africa. Pretoria: Statistics South Africa. StatsSA. 2016. Community Survey 2016:  Agricultural Households Report no.  03-​ 01-​ 05. Pretoria: Statistics South Africa. StatsSA. 2021. Gross Domestic Product, 4th Quarter 2020. Pretoria: Statistics South Africa StatsSA. 2020a. Census of Commercial Agriculture 2017. Pretoria: Statistics South Africa. StatsSA. 2020b. Personal communication and additional analysis on the Community Survey 2016 data [interview], 14 September. StatsSA. 2021. Gross Domestic Product, 4th Quarter 2020. Pretoria: Statistics South Africa. The Presidency. 2012. National Development Plan 2012. The Presidency in the Republic of South Africa. Timmer, Peter. 2002. ‘Agriculture and economic development’, in Bruce Gardner and Gordon Rausser (eds) Handbook of Agricultural Economics, Volume 2. Amsterdam: Elsevier Science. Trade Map. 2020. Agricultural Trade Data. Geneva: International Trade Centre. Vink, Nick, 2000. ‘Agricultural policy research in South Africa:  Challenges for the future’, Agrekon, 39: 432–​70. Vink, Nick, and Johann Kirsten. 2019. ‘Principles and practice for successful farmland redistribution in South Africa’. Working Paper no. 57, Institute for Poverty, Land and Agrarian Studies, University of the Western Cape.

Chapter 11

Agro-​p ro c e s si ng Indu stries i n t h e Sou th AfricaN E c onomy Horman Chitonge

11.1 Introduction This chapter provides an overview of the agro-​processing sector in the South African economy. The agro-​processing sector has been recognized by the South African government as a strategic set of industries in the economy, with great potential to contribute to promoting inclusive growth and job creation. For instance, the New Growth Path (2010), the National Development Plan (NDP 2011) and Industrial Policy Action Plan (IPAP 2016/​17–​2018/​19) have all recognized agro-​processing as an important cluster of industries with enormous potential to contribute to the structural transformation of the economy. One of the reasons why agro-​processing industries are believed to have the potential to make a significant contribution to economic and social transformation is that several agro-​processing industries are labour intensive, and this is critical for job creation. The other reason is that agro-​processing industries are seen to have the potential to promote the growth of small and medium enterprises, which can contribute to economic transformation by promoting broader participation in the economy. For instance, the Economic Strategy for South Africa document, launched by the National Treasury in 2019, argues that this cluster of industries has the potential to contribute to ‘the creation of opportunities for all South Africans to live productive, prosperous and dignified lives’ (National Treasury 2019: 11). This chapter highlights the potential for structural transformation in this sector; but also points out the enduring challenges which must be addressed to realize the sector’s full potential. The chapter provides an overview of the sector, including trends in employment, output, and international and regional trade. Various components of agro-​processing industries, including the size, structure, and some of the high value-​added activities

218   Horman Chitonge and products, are discussed in the chapter. Given the diverse nature of agro-​processing industries, it is beyond the scope of this chapter to discuss each of the subsectors in detail; the chapter focuses on the major agro-​processing clusters to highlight the potential to contribute to the national objective of creating an inclusive economy and society. An inclusive economy here refers to the ability to create opportunities for the majority of people to actively participate, contribute to, and share in economic growth through productive jobs or businesses. The current structure of the economy, though it reflects features of an advanced economy, is not inclusive because many people who would like to participate in and contribute to economic growth are without jobs and viable business opportunities.

11.1.1 The Agro-​processing Sector Agro-​processing industries in the South African economy constitute a diverse set of activities ranging from the simple processing of food and other agricultural products to more sophisticated activities such as processing of cotton, leather, timber, and rubber into finished and intermediate goods. According to the International Standard Industrial Classification (ISIC Rev. 4)  agro-​processing is a component of the manufacturing sector that transforms primary agricultural products into intermediate and final goods. Although ISIC provides an internationally accepted classification of agro-​processing activities, this set of industries is defined differently by different institutions. For example, IPAP 2016/​17–​2018/​19 (DTI 2016) restricts the term agro-​processing to the processing of food and beverages only (see also FoodBev-​SETA 2017: 3). The Department of Agriculture Forestry and Fisheries (DAFF) has, however, adopted a broader definition which includes the processing of all primary agricultural, forestry, and fishery products into intermediate and finished goods (DAFF 2012). DAFF has adopted the Food and Agriculture Organisation’s (FAO 1997)  definition, which defines agro-​processing as a subset of ‘manufacturing that processes raw materials and intermediate products originating from agriculture, forestry and fisheries’ (DAFF 2012: 1). It is important here to note that the definition of agro-​processing that one adopts can affect the scope, profile, and size of the sector. As illustrated below, if we restrict agro-​processing to food and beverages only, the sector will be smaller in size and significance in the economy (see Figure 11.2). The agro-​processing segment of the manufacturing sector is a complex cluster of manufacturing industries which consists of eleven out of the twenty-​four divisions of the manufacturing sector. Although it is the food and beverage industries which dominate agro-​processing industries in South Africa (and in many low-​and middle-​income countries), accounting for over 60 per cent of the agro-​processing output and about 50 per cent of employment on average since the 1970s, there are several other important industries which make up the sector (see Table 11.7 and Figure 11.2). While industrial classification seeks to clearly demarcate various economic activities and sectors, analysts have observed that the ‘boundaries’ between sectors,

Agro-processing Industries in the South AfricaN Economy    219 particularly agriculture and agro-​processing, have in recent times become fuzzy, making it hard to distinguish agro-​processing from agricultural activities (Cramer and Sender 2015). The blurring of boundaries has become even more pronounced due to the fact that there is a great deal of processing taking place on-​farm or within the farming enterprise, which often falls outside of conventional agricultural activities. Apple and orange farmers are now engaging in sophisticated processing activities within the fruit value chain than hitherto. The adoption of sophisticated technology and equipment in modern agriculture has led to the industrialization of agriculture, a phenomenon which is sometimes referred to as ‘industrializing freshness’, such that even simple products like fruits are increasingly becoming manufactured goods (Cramer and Sender 2015). However, even if the boundaries are blurred, it is still important to classify activities according to the broader economic category under which they fall in order to capture trends in value addition and changes occurring in different subsectors.

11.2  The Agro-​processing Sector in South Africa:  Profile and Trends If we look at the five-​year manufacturing output moving average for different subsectors between 1970 and 2019, it is clear that the agro-​processing sector (broadly defined) has consistently accounted for the largest share (Table 11.1). Although the share of agro-​processing industries in total manufacturing output declined slightly between 1990 and 2010, it has consistently been the largest subsector of the manufacturing sector, and its share has been growing since 2010. If we look at the 2017–​19 manufacturing sector output by subsector, agro-​processing industries’ share was more than the combined output of the metal and machinery and the transport equipment subsectors. The dominant contribution of agro-​processing industries to the manufacturing sector is also visible when we look at the industries’ share in manufacturing employment. The share of agro-​processing industries in manufacturing employment was even greater than its share of output (Figure 11.1). In terms of structural transformation and employment creation, the importance of agro-​processing industries is not just because of the strong backward linkages to the agriculture sector, it is also because of its size and diversity. In the South African context where employment creation has become a major challenge, agro-​processing industries are critical, given that most of these industries tend to be labour intensive relative to other manufacturing industrial clusters (Black et al. 2016). This makes agro-​processing industries strategically important when it comes to job creation, as highlighted in several policy documents mentioned above.

Table 11.1: Manufacturing output by subsector (%), 1970–​2019

Agro-​ processing

Metal machinery and equipment

Transport equipment

1970–​75

34

27.3

12.2

1976–​80

34.8

26.7

10.7

1981–​85

35

25.8

1986–​90

36.7

21.5

1991–​95

36.2

19.8

1996–​2001

32.7

20.4

2001–​05

29.6

19.7

2006–​10

29.3

20

2011–​16

32

2017–​19

33.7

Basic and other chemicals

Coke and refined petroleum

Other

7.2

2.5

16.8

8.7

3.3

15.7

10.4

8.9

3.9

16

9.3

10.1

4.6

17.8

9.6

10.5

5.1

18.9

10.9

12.2

6.7

17.1

14.2

13.1

7.7

15.7

13.8

13.2

8.1

15.7

19.5

12.3

13.4

8.9

13.9

17

13.2

13.8

9.9

12.4

Source: Author’s calculations based on data from DAFF (2012) and StatsSA (2019). Note: The reporting of sectoral classification of the manufacturing sector has not been consistent over the period. Figures for 2011–​ 15 averages are based on a different sectoral classification.

Agro-processing Industries in the South AfricaN Economy    221 Glass & non-metalic products Electric Equipment & 6% communication 6%

Coke, Refined Pretroleum & Chemical 15%

Other 3%

Agroprocessing 40%

Transport Equipment 9% Metal Machinery & Equipment 21%

Figure 11.1  Manufacturing employment by subsector 2019 Source: Author’s calculations based on data from StatsSA (P3041.2 ‘Manufacturing Production and Sales, 2019 Third Quarter’).

In terms of output, the sector grew from Rand 450 billion to Rand 762 billion (current prices) between 2010 and 2018 (StatsSA 2019). This is confirmed by the production volume index, which shows that the agro-​processing cluster grew steadily from 2009 up to 2015, but registered a slight decline in 2017 and 2018 (StatsSA 2019a). The decline in the volume of goods produced in the sector can be attributed to the persisting challenges in the textiles, clothing, and leather subsectors which declined at an average rate of about 2 per cent per year between 2009 and 2018. This is mainly due to the increasing pressure from imported textiles, clothing, and leather products from Asia, particularly China and Thailand (DAFF 2018). However, output volumes in the other subsectors, particularly the food, beverages, and tobacco cluster, grew steadily, increasing at an annual average rate of 2.7 per cent over the same period. In terms of trade, the sector remained stable although the trade deficit has been growing since 2015 (FoodBev-​SETA  2017).

11.3  Components of Agro-​processing Industries According to ISIC Rev 4 classification, there are eleven divisions which make up the agro-​processing sector as noted above. The average share in output between 2011 and 2019 for each of the eleven subsectors is shown in Figure 11.2 below. In terms of both output and employment, it is clear from Figure 11.2 that while the agro-​processing sector is made up of several industries, it is dominated by the processing of food and beverages. The food and beverage cluster contributed an average of

222   Horman Chitonge OUTPUT BY SECTOR Rubber Products 3% paper & paper Products 12% Textiles Wearing Apparel 4% 3%

Footware 0%

Leather & Leather Products 2%

Food 40%

Tobacco 2% Furniture 2% Wood & Wood Products 12%

Beverages 20%

Formal Employment by Sector paper & paper Products 6%

Rubber Products 2%

Textiles 9%

Footware 1%

Leather & Leather Products 0%

Food 38%

Wearing Apparel 3% Furniture 6% Tobacco 1%

Wood & Wood Products 27%

Beverages 9%

Figure 11.2  Agro-​processing output and employment by subsector, average for 2011–​19 Source: Author’s calculations based on data from Quantec (Excel spreadsheet).

two-​thirds of the sector’s total output between 2011 and 2019, and slightly less than half of employment over the same period. The dominance of the food and beverage cluster is evident in the fact that out of the eleven subsectors, the combined output of six clusters (footwear, leather and leather products, tobacco, furniture, rubber, and wearing apparel) is smaller than that of the beverage cluster alone. When we compare formal employment and output share by sector, employment share for wood and wood products, furniture, and textiles is double the share in the sector’s output, suggesting that these are using more labour per unit of output (bottom in the sector’s Figure 11.2). These are the most labour intensive subsectors of the agro-​ processing sector (Black et  al., 2016). On the other hand, the formal employment

Agro-processing Industries in the South AfricaN Economy    223 share for beverages and paper and paper products is half of the share in output. This is not surprising; paper and paper products and beverages tend to be capital-​intensive industries (Black et al. 2016).

11.3.1 The Food Processing Subsector The food processing industry is often divided into five clusters according to the type of activities carried out. The production, processing, and preservation of meat, meat products, fish, fruits, vegetable, oils, and fats form one cluster. Manufacturing of dairy products forms another cluster of the food processing sector. The third cluster is the manufacture of grain mill and related products, including starches and animal feed. The manufacture of confectionary and baking products, including sugar, chocolate, pastas, syrups, custards, and other related products, constitute another cluster. The manufacture of beverages, which is divided into alcoholic and non-​alcoholic beverages, is the fifth cluster of the food sector (FoodBev-​SETA 2017). Evidently, the food and beverage subsector is a huge industry in the South African economy. The food and beverage industries were in 2017 reported to have 11,011 registered companies (FoodBev-​SETA 2017:  11), although as we shall see later, the sector is dominated by a few large local and multinational companies that account for an overwhelming share of output, production capacity, investments, employment, and market share (Nhundu et al. 2017). It is estimated that out of the more than 11,000 registered companies in the food and beverage industry, 91 per cent are small-​scale enterprises, while 5 per cent are medium- and 4 per cent large scale (FoodBev-​SETA 2017). Out of the 4 per cent large food and beverage companies, production and market share are heavily concentrated in ten large firms including vertically and horizontally integrated companies such as Tiger Brands, Pioneers Foods, AVI Limited, Ocean Group Limited, Tongaat Hulett, Rhodes Food Group, and multinational giants such as Clover, Parmalat, Nestlé, and Cargill. Just about a third of registered companies are involved in the processing and preservation of meat, fish, fruits, and vegetables, followed by 18 per cent operating in the beverage cluster. Most of the large food and beverage companies are involved in many food and beverage value chains, pursuing a vertically integrated strategy from agriculture production to the manufacture of packaging materials and distribution of food and beverage products (Harcourt 2011). What is interesting is that small firms make up the bulk of the sector, and if, say, 20 per cent of the small firms can grow to become medium-​sized firms, the opportunity for promoting inclusive growth in the sector can be significant. This would also contribute to reducing the concentration of market power in a few large companies, as is currently the case. There is also a spatial concentration of food and beverage enterprises and employment in South Africa, with the Western Cape (mainly Cape Town and surrounding

224   Horman Chitonge districts) and Gauteng accounting for over 82 per cent of the total employment in this cluster, followed by KwaZulu-​Natal at 7 per cent.

11.3.2 Composition of the Processed Food and Beverage Subsector If we disaggregate the food and beverage cluster, it is also evident that dairy products, confectionaries, meat and meat products, constitute the largest component in terms of output. These three processed food clusters accounted for over half of the total sales value of the processed food sector between 2011 and 2017 (Table 11.2). In the beverage cluster, wines and mineral water take up a lion’s share, accounting for more than 70 per cent of the total sales value in 2017.

Table 11.2: Food and beverage output composition, 2011–​17 2011

2014

2017

2011

2014

2017

Food (current R million) % Meat and meat products

54,268

86,265

94,640

14.9

17.1

15.2

Fruits (fresh and dried)

38,738

46,947

61,626

10.6

9.3

9.9

Sugar (refined and raw)

9,262

12,374

12,667

2.5

2.4

2.0

Vegetable and animal oils

32,095

49,153

74,024

8.8

9.7

11.9

Grain and cereal

34,973

54,080

69,676

9.6

10.7

11.2

Dairy products

78,035

102,978

126,344

21.4

20.4

20.3

Confectionery

63,253

81,289

93,507

17.4

16.1

15.0

Other foods Total food

53,479

72,132

90,996

14.7

14.3

14.6

364,105

505,219

623,483

100

100

100

Beverages (current R million) % Wines

46,832

74,599

93,324

41.8

40.6

38.9

Beer

22,206

40,216

51,459

19.8

21.9

21.5

Spirits

4,656

14,019

15,465

4.2

7.6

6.4

Mineral water, soft drinks

38,348

55,126

79,554

34.2

30.0

33.2

Total Beverage

112,044

183,962

239,803

100

100

100

Source: StatsSA (Excel download).

Agro-processing Industries in the South AfricaN Economy    225

11.4  Processed Food and Beverage Export Trends At the top of the list of South Africa’s processed export products are wines, fruits and juices, soups, and food preparatory additives. The African market (mainly southern Africa) has become the largest destination of processed foods from South Africa, accounting for close to half of the total export of processed foods. Although the African market is currently small in terms of size compared to the traditional destination of South African exports (Cramer and Sender 2015; Black et al. 2020), the African demographic factor and income dynamics are slowly changing this picture, with some analysts predicting significant growth in the next decade or so (Visser and Ferrer 2015; ACET 2017). In the SADC region, Namibia, Botswana, Mozambique, and Zimbabwe accounted for over 35 per cent of all processed agricultural export products in 2017 (DAFF 2018). While the Netherlands, China, and the United Kingdom accounted for about 30 per cent of South Africa’s export of primary agricultural products (unprocessed), the SADC region is becoming the top export destination for processed agricultural products from South Africa. The advantage for South Africa in the regional trade dynamics is that the phytosanitary, health and safety requirements in SADC countries are less stringent than those of the European Union and other regions. The bulk of the processed food produced is consumed locally, though there is an increasing share of processed food in total agriculture export, particularly fresh fruit, which is one of the high value-​added products (ITAC, 2016; van Lin et al. 2018). In 2018, for instance, 72 per cent of the fruits produced were exported, leaving only slightly over a quarter for domestic consumption (van Lin et al. 2018). While the export of processed foods has been growing steadily, the trade balance in processed food has been negative, mainly as a result of the appreciation of the Rand (for the better part of the first decade of the 2000s) which encouraged imports of processed foods. The main factor behind the growing processed-​food trade deficit is the rise of imports, particularly of poultry meat from South America and the United States, and this has been attributed to the shift in diet as more people move to a protein-​based diet, targeting poultry meat which is cheaper than other meat products (BFAP 2019). The shift in diet is confirmed in terms of the consumption volume per capita, with poultry meat rising from an average of 9.5 kg in the 1970s to 36 kg per capita per year in 2014 (Greyling et al. 2015: 6). Export of fresh fruits has been recognized as a high value-​added segment, and has been one of the most profitable sectors of the agro-​industries (van Lin et  al. 2018). The fresh fruits and vegetable cluster has great potential to contribute to promoting inclusive growth and job creation not only because this is highly profitable but also because these activities are the most labour intensive. Fruits and vegetables such as strawberries, carrots, tomatoes, pawpaw, avocado, and cherries are multiple times more labour intensive than field crops such as maize, wheat, and barley (Zalk 2019).

226   Horman Chitonge There is potential for small and medium firms to enter these value chains, which can contribute to creating employment and inclusive growth, given that these activities are more labour-​intensive. The growing share of processed agriculture products in total agriculture export reflects the ability of this segment of agro-​processing to compete favourably in international markets. This is partly driven by the fact that the commercial agriculture and food processing sectors employ advanced technology which enhances the sector’s efficiency and competitiveness (see Chapter 10 on agriculture, in this volume). Visser and Ferrer (2015) have observed that South African commercial farmers have become more efficient, pushed by deregulation which removed state support, leaving the sector to learn to stand on its own feet. The perssiting challenge in this regard is that these value chains are dominated by well-​established and connected (domestically and internationally) commercial farmers, leaving little room for new entrants, especially the small and emerging farmers. If the sector has to realise its potential to contribute to structural transformation of the economy, these structural challenges have to be addressed through appropriate policies.

11.5  Textile and Clothing Industries The textile and clothing industries are another important cluster of agro-​processing industries. This is one of the old set of industries that have evolved into a now complex cluster. The sector has gone through three major phases, notably the phase prior to the deregulation of the economy (before 1989 when import substitution was the main policy strategy), the liberalization of the market resulting in massive import penetration (1994 to 2005) resulting in many firms struggling to become competitive (Vlok 2006), and from 2009 onwards, when interventions to save the sector from total collapse were initiated (Morris and Barnes 2014). Although there are signs of recovery from 2014 onwards, the textile and clothing industries in the country have been characterized by persistent decline in employment, in particular. For instance, between 1996 and 2005, total employment in the textile industries declined from 228,000 to 143,000 workers (Vlok 2006: 233). In the clothing sector, employment levels declined from 97,000 in 2003 to slightly over 52,000 in 2013 (Morris and Barnes 2014: 11–​12). Black et al. (2016: 12) estimate that the employment compound annual growth rate (CAGR) for apparel and textile industries were  –​3.6 and –​3.4 respectively between 1994 and 2011. As illustrated in Table 11.3 below, formal employment in this cluster has been decimated, falling to between half and a third of the 2000 levels by 2019 (Table 11.3). The sector is divided into two major components:  the textile and clothing industries, with the apparel industries constituting the secondary manufacturing tier, drawing from the textile products. In terms of output, the textile industry accounts for slightly less than three-​fifths, with the clothing industry taking up the remainder.

Agro-processing Industries in the South AfricaN Economy    227 Table 11.3: Clothing and textile output and formal employment trends, 1995–​2019 1995

2000

2005

2010

2015

2016

2017

2018

2019

Output by subsector (R million, current ) Clothing 12,627.6

13,875.4 17,252.2

16,020.7

19,693.2

20,916.6

19,608.0

18,742.1

18,698.2

Textiles

20,073.5

23,226.6 28,760.2

24,931.5

27,784.8

29,848.0

30,494.9

30,898.9

29,113.8

Total

32,701.1

37,102.0 46,012.4

40,952.2

47,478.0

50,764.6

50,102.9

49,641.0

47,812.0

Share in total output (%) Clothing 38.6

37.4

37.5

39.1

41.5

41.2

39.1

37.8

39.1

Textile s

62.6

62.5

60.9

58.5

58.8

60.9

62.2

60.9

61.4

Formal employment by subsector Clothing 101,355

88,754

76,183

50,643

41,840

41,226

37,583

37,936

38,205

Textiles

82,279

66,109

51,410

40,513

37,202

37,303

36,118

33,306

31,274

Total

183,634

154,863

127,593

91,156

79,042

78,529

73,701

71,242

69,479

Share in formal employment Clothing 55.2

57.3

59.7

55.6

52.9

52.5

51.0

53.2

55.0

Textiles

42.7

40.3

44.4

47.1

47.5

49.0

46.8

45.0

44.8

Source: Author’s calculations based on Quantec data.

However, in terms of employment, the clothing industries contribute an average of about 55 per cent (Table 11.3). It has been observed that there are a large number of informal ventures and workers in the clothing sector, most of them small businesses operating in sitting rooms, backyards, and small shops in city and township markets (Vlok 2006; CTFL-​SETA 2014). Even when informal employment is excluded, the clothing segment accounts for the largest share of employment, confirming the labour-​intensity of this sector. Employment and output trends between 1995 and 2019 confirm the popular sentiments that the textile and clothing industries in the country have been in crisis since the mid-​1990s. Part of this crisis is due to the increasing share of imported textile intermediates from other countries, mainly Lesotho, Mauritius, and Kenya. Similarly, output in the sector has been declining, particularly the export of apparel, which fell sharply from US$231 million in 2003 to US$6 million by 2012 (Morris and Barnes 2014: 10). However, it is formal employment in the sector which has sharply declined, with employment in 2019 dropping to just a third of the levels reported in 1995. This is worrying in terms of job creation given that the textile and clothing industries are some of the most labour-​intensive clusters of the manufacturing industry (Vlok 2006; CTFL-​SETA 2014). Starting from 2007, several policy measures have been introduced to revive the textile and clothing industries (Patel 2016). One of the early measures aimed at stimulating competitiveness and sustained growth in the sector is the Textile and Clothing

228   Horman Chitonge Industry Development Programme, which was launched in 2006 and administered by the Industrial Development Corporation (IDC). In 2010, the Clothing and Textile Competitiveness Improvement Programme was introduced to support established firms to upgrade by recapitalizing the sector (Morris and Barnes 2014). Although these policy interventions have helped to save the industry from total collapse (Patel 2016), there are still challenges which need to be addressed in order to create an environment where firms can respond effectively to the new challenges in the sector (Zalk 2014). The failure to generate more jobs in the clothing and textile sector is sometimes attributed to there actions of labour unions (Nattrass and Seekings 2013) but that is not the only factor. While there has been stability in the sector in terms of output from 2012 (SA Cotton 2019), the sector is still struggling to remain competitive, especially the export market. It has been reported that as a result of policy interventions introduced since 2009, 63,000 jobs have been saved and 8,000 new ones created (CTFL-​SETA 2014; Patel 2016). Although the sector has continued to face competitive pressure from outside, there is potential for high value-​added products in niche areas such as men’s formal shirts, suits, and other tailor-​made apparel. There is also an emerging market for African-​ inspired designs and fashions which has potential for capturing the growing African market. The major challenge in the sector remains around firms being unable to restructure to remain competitive. Here too, there is potential for small and medium firms to contribute towards transforming the structure of the economy and towards inclusive development.

11.6  Leather and Leather Products The other major cluster of agro-​processing industries is the leather and related products cluster. This segment of agro-​processing, although small, is an important sector which provides inputs to other manufacturing clusters, including the automotive sector and the manufacture of furniture, bags, belts, and shoe uppers. There are two major types of leather used in the industry. The biggest is bovine leather which is produced from cattle hides. In 2015, 2.5 million bovine hides were produced. Leather is also produced from sheep and goat skins (small skins), with 5 million goat and sheep skins produced in 2015 (DTI 2016). The second leather component is exotic leather which is produced from ostrich and crocodile skin. Exotic leather products feed into high value-​added chains, and demand for these products is reported to be growing fast (DTI 2016). The majority of companies in this cluster are in the footwear subsector, with 175 firms in 2011, followed by leather products with eighty-​four companies and thirty-​one in the production of leather. The sector has two major components: footwear and leather and leather products. In terms of output, the leather and leather products cluster dominates the sector, accounting for over 95 per cent of total output between 2011 and 2017.

Agro-processing Industries in the South AfricaN Economy    229 Although the footwear industries only account for less than 5 per cent of output in the sector, it is the largest employer in this cluster of industries, accounting for an average of over two-​thirds of jobs in the sector between 2000 and 2019 (Table 11.4). As Table 11.4 illustrates, the leather industry faces similar challenges to the textile and clothing industry. Formal employment in the sector in 2019 was only half the level reported in 2000. The footwear industry was the hardest hit, with employment declining from over 18,000 in 2000 to just 8,000 in 2019. The leather industry’s challenges include pressure from growing imports, the low quality and quantity of the leather, and irregular supply of inputs. Another major challenge in the sector is declining demand from the automotive seating and trim industry, which has increasingly resorted to using synthetic leather (DTI 2016). This has resulted in some of the tanneries closing, although there have been some signs of recovery since 2011, with twenty-​eight new factories opening between 2011 and 2014 (DTI 2016). The high value-​added produce in the sector includes exotic leather products which mainly supply the upper end of the domestic and export markets (CTFL-​SETA 2014). The Department of Trade and Industry announced in 2016 that Centres of Footwear and Leather Goods Entrepreneurship will be established to drive demand for bovine and exotic leather in the country by training small business entrepreneurs. The impact of policy measures implemented to improve productivity and growth in the sector is yet to be seen, though there are growing opportunities in the high value-​added segments.

Table 11.4: Leather, leather products, and footwear employment 2000

2005

2010

2015

2016

2017

2018

2019

Formal employment (number) Footwear

9,864

8,445

8,869

8,411

9,513

9,229

8,584

Leather 8,656 and leather products

18,254

5,229

5,615

4,725

5,204

5,207

5,039

4,807

Total

15,093

14,060

13,594

13,615

14,720

14,268

13,391

67.8

65.4

60.1

65.2

61.8

64.6

64.7

64.1

Leather 32.2 and leather products

34.6

39.9

34.8

38.2

35.4

35.3

35.9

26,910

Share in employment Footwear

Source: Author’s calculations based on Quantec data.

230   Horman Chitonge

11.7  The Wood, Wood Products, and Furniture Cluster The wood and wood products and furniture industries comprise another major cluster in the agro-​processing sector. This set of industries has been the third largest in the agro-​ processing sector, after food and beverages, accounting for an average of 12 per cent of the sector’s output since the 1990s (see Figure 11.2). Similarly, this is the second largest subsector after the food processing industries, in terms of employment, contributing an average of 25 per cent of jobs in agro-​processing since the 1990s. The wood and wood products and furniture cluster is one of the few agro-​processing clusters where employment has been growing steadily, from an average of 140,000 in the 1990s to over 170,000 in 2019 (see Table 11.5 below). The sector’s output has grown consistently at an annual average rate of 2.4 per cent between 2000 and 2017, except in 2009 when output declined by 4.5 per cent (TIPS 2017). The inputs utilized in the sector are harvested from woodlands, natural forests, and plantation forests (Pogue 2008). The three major forest products used in this sector are pine, accounting for 52 per cent, followed by eucalyptus at 39 per cent and wattle with 1 per cent, and a variety of other types. While there are many forest plantations owned by different companies and individuals, the industry is dominated by two big, vertically integrated companies: Mondi South Africa and Sappi Forests. The sector is divided into four major subdivisions, namely sawmilling and wood planing, wood products, paper and paper products, and furniture. In terms of output, the furniture industry has been the largest, contributing an average of 30 per cent of the sector’s output between 2000 and 2019 (Table 11.5). The wood products segment was the second largest between 2000 and 2005, accounting for an average of 27 per cent of total output in the cluster, but from 2010 to 2019, the paper and paper products cluster has been the second largest with an average of over 30 per cent. However, when it comes to jobs, it is the paper and paper products segment that has been the largest, with an average of over 60 per cent of total formal employment between 2000 and 2019. Employment share in the furniture industry has hovered around 15 per cent over this period while the share of wood products has varied from 17 per cent in 2000 to 22 per cent in 2005, and then declined steadily to 18 per cent in 2019. The major challenges in the sector are related to the stringent regulations imposed to ensure compliance with environmental standards and requirements. Extraction of forest products requires that consideration is given to environmental issues, including the effects of logging on biodiversity and climate change. Other challenges reported in the sector include the lack of research and development (R&D) funding to promote innovation and the growth of appropriate technology (FP&M-​SETA 2014). The opportunity in the sector lies in increasing value addition on export products, most of which have been exported with low value-​added (mainly in the form of wood pulp) and

Agro-processing Industries in the South AfricaN Economy    231 Table 11.5: Employment and output composition by subsector, 2000–​19 2000

2005

2010

2015

2016

2017

2018

2019

Share in employment (%) Sawmilling and wood planing

7.4

8.8

7.6

8.5

8.8

8.8

8.3

8.1

Wood products

16.9

22.4

20.1

18.2

18.3

17.6

17.5

17.4

Paper and paper products

60.2

52.2

57.9

58.3

58.8

59.1

61.0

61.9

Furniture

15.5

16.6

14.5

15.0

14.0

14.4

13.3

12.7

Output composition% Sawmilling and wood planing

16.9

14.5

16.9

14.7

15.8

16.0

15.4

14.9

Wood products

25.6

28.3

21.5

24.5

24.4

22.9

23.6

26.1

Paper and paper products

19.9

21.9

32.9

34.1

34.3

34.5

31.2

28.7

Furniture

37.6

35.3

28.6

26.6

25.5

26.6

29.7

30.4

Source: Author’s calculations based on data from Quantec (Excel download).

imported back after value is added from outside the country, especially in Asia. As a result of this, the sector’s trade deficit has been growing since 2011 (TIPS 2017). There are other segments of the agro-​processing sector which have not been discussed here, particularly rubber, and printing and publishing. However, the clusters discussed above are the major segments of the agro-​processing sector in the country.

11.8  Agro-​processing Employment Trends When we look at employment trends, as noted above, agro-​processing industries account for the largest share of employment in the manufacturing sector, which has been consistently above 40 per cent (Figure 11.1). Although the percentage share of agro-​ processing employment has remained relatively stable, the actual number of formal

232   Horman Chitonge

2,500,000

40.0 35.0 30.0 25.0 20.0 15.0 10.0 5.0 0.0

2,000,000 1,500,000 1,000,000 500,000 0

%

Employment (Number)

jobs declined from just above 611,000 in 2001 to slightly over 500,000 in 2018 and 2019, representing a decline of about 17 per cent over this period. This reflects the general decline of formal employment in the broader manufacturing sector, where formal employment declined from 1.3 million in 2000 to 1.2 million in 2019 (Figure 11.3; see also Black et al. 2016). The level of informal employment in the manufacturing sector as a whole has oscillated around the 30 per cent mark between 2007 and 2016, before declining from 2017 to 2019. The decline of informal employment follows the general pattern of employment in the larger manufacturing sector, which, between 1994 and 2011, shed jobs at an annual average rate of 1.3 per cent (Zalk 2014). In terms of labour intensity, the agro-​processing sector, as a whole, has been consistently the most labour-​intensive sector, although the labour intensity has varied between subsectors. Black et al. (2016) divided manufacturing subsectors into capital intensive, intermediate labour intensive, and ultra-​labour intensive. All the clusters of the agro-​ processing sector, except beverage and paper and paper products, fell either in the intermediate labour intensive or ultra-​labour intensive subsectors. Ultra-​labour-​intensive subsectors are largely agro-​processing industries such as footwear, wood and wood products, leather products, furniture, and apparel. It is in the labour-​intensity feature of these industries where the potential to contribute to creating employment and social transformation lies. But these industries, as shown above, have faced serious challenges in terms of remaining competitive, both domestically and internationally. If we compare the sectoral output to levels of employment, the agro-​processing sector’s ratio of output to employment was again the highest in the entire manufacturing sector, confirming the widely held view that the sector is the most labour intensive in the manufacturing sector. However, it is worrying that the most labour-​intensive clusters of agro-​processing industries (textiles and wood and wood products) are both recording declining output and employment. Between 2010 and 2018, wearing apparel lost close

00 01 02 03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20 20

Agroprocessing Employment (Fromal)

Manufacturing Employment (Formal)

Total Manufacturing Manufacturing Employment (Informal)%

Manufacturing Employment (Informal)

Figure 11.3  Agro-​processing in manufacturing employment, 2000–​19 Source: Author’s calculations based on data from Quantec (Excel spreadsheet).

Table 11.6: Agro-​processing formal employment by subsector (%) 2000

2005

2006

2007

2008

2009

2010

2011

2012

2013

2014

2015

2016

2017

2018

2019

2000-​2019 (Average)

Food Beverage

32.0

34.1

33.4

33.0

34.1

35.6

35.9

34.9

34.3

34.9

36.2

37.3

39.1

39.9

40.8

41.1

35.2

6.6

5.8

5.8

6.1

6.3

6.3

6.7

7.1

7.5

7.9

7.6

7.7

7.8

7.6

7.2

6.7

6.7

Tobacco

0.5

0.5

0.5

0.5

0.5

0.5

0.5

0.6

0.6

0.6

0.6

0.6

0.7

0.6

0.6

0.5

0.5

Textiles and wearing apparel

25.3

23.1

22.6

22.1

20.7

19.2

18.7

17.7

16.9

15.9

15.7

15.9

15.5

14.6

13.9

13.6

19.7

Leather and leather products

4.4

2.7

2.8

2.7

2.7

2.8

2.9

2.8

2.7

2.9

2.7

2.7

2.7

2.9

2.8

2.6

2.9

Wood and paper products

20.8

23.5

24.8

24.7

25.5

26.0

26.7

27.8

28.7

28.9

28.8

27.5

26.5

26.2

26.0

26.2

25.3

Rubber products

3.0

2.8

2.8

2.8

2.9

2.8

2.5

2.6

2.7

2.5

2.3

2.3

2.3

2.5

2.4

2.3

2.6

Furniture

7.5

7.5

7.4

8.1

7.2

6.7

6.2

6.4

6.7

6.3

6.0

6.1

5.5

5.7

6.5

6.9

6.9

Source: Author’s calculations based on data from Quantec data.

234   Horman Chitonge to 10,000 jobs, while the number of jobs in the textile industry dropped from 37,000 to 28,000 over the same period (DAFF 2018). The largest sector (food processing) has relatively lower labour intensity, with the lowest being in the beverage industry which, though being the second largest in terms of output, accounted for less than 6 per cent of agro-​processing employment in 2018 (Table 11.6). The employment ratio for the food cluster has stabilized while jobs in the beverage sector grew steadily after 2012. However, the pressure from international competition is pushing some manufacturers to adopt new technology and innovative business approaches, leading to some sectors becoming more capital intensive. The real potential for generating employment in the agro-​processing sector lies with the small and medium processors, which tend to be more labour intensive than the large firms which currently dominate the sector (Ncube et al. 2016). This is why there is a strong push for policy intervention to encourage the growth of small enterprises to enter the market. It is these small enterprises which can contribute to employment creation because they tend to be more labour intensive (IDC 2017). But, for small enterprises to fulfil this role, they have to be competitive and this is where strategic policy interventions are needed.

11.9  The Challenge of Creating an Inclusive Economy 11.9.1 The Structure of the Agro-​processing Sector As noted above, the distribution of production capacity and employment between firms in the agro-​processing sector reveals high levels of concentration. For instance, the four largest agro-​processing firms: Tongaat Hulett, Tiger Brands, Pioneer Foods, and RCL Foods accounted for 72 per cent of productive capacity in 2014 (Nhundu et al. 2017). In terms of employment, the five largest food processors account for almost 80 per cent of total employment in the food processing cluster (Dube et al. 2018). It has been suggested that the move towards concentration of production capacity and market share can be explained by several factors including the carry-​over of food cartels from the apartheid era, the global trend towards vertical integration of lead firms in the food value chains, the effect of liberalization and deregulation of the economy during the 1990s, and the tightening of social networks and relations between farmers, food processors, retailers, and suppliers in the sector (Ncube et al. 2016; Visser and Ferrer 2015; Nair and Landani 2020b). Though some analysts have suggested that, given the current global trade context, ‘big and not small is beautiful’ (Amsden 2012), concentration of market power leads to economic inefficiencies and makes it difficult to address the pressing challenges of promoting inclusiveness in the economy as Black and Roberts (2009) have highlighted. Concentration of market power and production makes it difficult to ‘make the circle

Agro-processing Industries in the South AfricaN Economy    235 bigger’ as large vertically integrated firms create high barriers to entry in order to maintain (perhaps even increase on) the existing market share and power (Nair and Landani 2020b). In general, ‘high concentration has reinforced the capital and resource-​ intensive industrial development path, while imperfect competition [has] distorted economic outcomes more widely’ (Black and Roberts 2009:  217). Concentration of market power and production capacity within a few large firms has created a number of challenges, such as rising costs of inputs, monopolistic tendencies, high barriers to entry for small and medium firms, and ring-​fenced markets controlled by a few large firms (IDC 2017). The high concentration of production and market share, which is a reflection of the structural challenges in the wider economy, needs to be addressed to release the potential of agro-​processing to contribute to building a resilient and inclusive economy. Therefore, the need to transform the agro-​processing sector is essential to creating conditions for inclusive and sustained growth in the economy. But transforming the sector is not going to happen through the introspection of those who are already in the ‘circle’; it will require deliberate and sustained policy commitment to implement strategies that can successfully stretch out the perimeter of the circle to make it more inclusive. Admittedly, ‘making the circle bigger’ has its own trade-​off (see National Treasury 2019), but on balance, the current situation does not serve the long-​term interest of anyone, including the owners of large firms, politicians, the ordinary people on the street. Empirical studies have now shown that it is difficult to sustain economic growth under conditions of extreme levels of inequality (UNRISD 2010), as is the case in South Africa (StastSA, 2019b).

11.9.2 Agro-​processing Sector Potential At this stage, the key question is:  how can agro-​processing industries contribute to addressing the pressing challenge of transforming the structure of the economy and the creation of jobs? There are many ways through which agro-​processing industries can contribute to these two critical objectives. One of them is that the labour-​intensive agro-​processing industries can contribute significantly to creating jobs, especially low-​ and semi-​skilled jobs. The other potential in the sector is due to the fact that stronger linkages between agro-​processing and agriculture can stimulate growth not just in the agriculture sector but also in other manufacturing industries through the supply of inputs to the agriculture sector. By adding value to raw materials from the agriculture sector, agro-​processing industries can contribute to industrialization and job creation, especially if growth is occurring in labour-​intensive industries. Further, by making agro-​ processed products readily available, agro-​ processing industries contribute to the growth of secondary food-​processing activities where small and medium enterprises are likely to participate, especially in value chains with a low capital outlay requirement such as wearing apparel, fashion design, preparation of food, and bakery. The growth of small businesses in agro-​related activities can enhance

236   Horman Chitonge inclusivity in the economy. Although the barriers to entry in some agro-​processing value chains are high due to the concentration of market power in a few Johannesburg Stock Exchange (JSE)-​listed companies (Nhundu et al. 2017), there are opportunities for small and medium ventures to be involved in activities such as the extraction of oils from seeds and fruits, confectionaries, grain-​milling, and the processing of animal feed (IDC 2017). In the agro-​processing sector, opportunities are growing as the demand for processed foods in the Southern Africa region grows, and South Africa can tap into these growing value chains and markets. Chingumira (2019), for example, has noted that due to the growing demand for food in the region as population, urbanization, and income rise, the market for agro-​processing equipment in the region is expanding rapidly, and South African equipment and parts manufacturers can exploit the proximity and free-​ trade opportunities to increase manufacturing production. Currently, most of the food manufacturers in the region are importing food-​processing equipment from various countries, including India, Italy, China, Germany, and Turkey. The strategy for economic structural reform announced by the National Treasury in 2019 has argued that South Africa has an advantage in the southern Africa region in several value chains, including agro-​processing, which should be exploited.1 However, in order to sustain the benefits of regional integration in Southern Africa, South Africa has to source an increasing share of inputs from other countries in the region, something that is not happening at the moment (Black et al. 2020).

11.10 Conclusion This chapter has provided an overview of the agro-​ processing industries in the South African economy, looking at trends in different clusters of the sector over time. The chapter has shown that the agro-​processing sector is made up of a diverse set of industries which, together, have consistently accounted for the largest share of manufacturing output, value ​added, and employment since the 1970s. Although levels of employment have varied over the years, agro-​processing industries as a whole have been the most labour intensive in the manufacturing sector relative to output and capital stock. This highlights the potential of the sector to contribute to job creation and to creating an inclusive economy. However, to realize the full potential of this sector there are several challenges that need to be addressed. One of the challenges is the concentration of production capacity and market power in a few large firms. This creates problems for achieving the objective of promoting inclusivity in the economy. Another challenge relates to the issue of

1 

Part of the growth of exports to SADC has been attributed to the change made in 2013 when South Africa started to report exports to SANCU members (see Dube et al. 2018).

Agro-processing Industries in the South AfricaN Economy    237 creating conditions where local firms, especially small and medium firms, can become competitive. Yet another challenge lies in lowering barriers to entry for new players in the market by addressing issues like access to finance and the uncompetitive tendencies manifested by the large agro-​processors. If some of the structural constraints mentioned are addressed, agro-​processing industries can play an important role in responding to the challenges of job creation and promoting inclusive growth.

References African Centre for Economic Transformation (ACET). 2017. ‘African Transformation Report 2017: Agriculture, Powering Africa’s Economic Transformation.’ Accra: ACET. Amsden, Alice. 2012. ‘Grassroots war on poverty’, World Economic Review, 1(3): 114–​31. Black, Anthony and Simon Roberts. 2009. ‘The evolution and impact of industrial and competition policies’, in J Aron, S.B. Kahn, and G. Kingdon (eds) South African Economic Policy since Democracy. Oxford: Oxford University Press. Black, Anthony, Simon Craig, and P. Dunne. 2016. ‘Capital intensity, industrial policy and employment in the South African manufacturing sector’. REDI 3x3 Working Paper no. 23. Black, Anthony, Lawrence Edwards, Ruth Gorven, and Willard Mapulanga. 2020. ‘Agro-​ processing, value chains, and regional integration in Southern Africa’. SA-​TIED Working Paper no. 103, Johannesburg. Bureau for Food and Agricultural Policy (BFAP), 2019. BFAP Baseline Outlook 2019–​2029. Stellenbosch: BFAP. Chingumira, Gillian. 2019. ‘Assessment of demand in agroprocessing machinery in the SADC region:  A case study of the maize-​milling machinery value chain in South Africa and Zambia’. WIDER Working Paper no. 2019/​70, Helsinki. Cramer, Christopher, and John Sender. 2015. ‘Agro-​processing, wage employment and export revenue: Opportunities for strategic intervention’. Trade and Industry Policy Strategies (TIPS) Working Paper, Department of Trade and Industry, Pretoria. DAFF. 2016. Briefing Note on Agropresssing Sector in South Africa. Cape Town: Parliament DAFF. 2018. South African Trade Profile through Revealed Comparative Advantage Index for Agriculture, Forestry and Fisheries. Pretoria: DAFF. Department of Agriculture, Forestry, and Fisheries (DAFF). 2012. Economic Profile of the Agro-​ processing Industry in South Afrcia: 1970–​2010. Pretoria: DAFF (Agro-​processing Support Directorate). Department of Agriculture, Forestry, and Fisheries (DAFF). 2019. Economic Review of the South African Agriculture 2019. Pretoria: DAFF. Department of Trade and Industry (DTI). 2016. A South African Leather Research Institute: A Technical Feasibility. Pretoria: DTI. Dube, Shingi Chisoro, Rena das Nair, Maria Nkhonjera, and Ndiadivha Tempia. 2018. ‘Structural transformation in agriculture and agro-​processing value chains’. Centre for Competition, Regulation and Economic Development, Working Paper no.  08/​ 2018, University of Johannesburg. Fibre Processing and Manufacturing Sector Education and Training Authority (CTFL-​SETA). 2014. A Profile of the Clothing, Textiles, Footwear and Leather Subsectors. Pretoria: SETA.

238   Horman Chitonge Food Agriculture Organisation (FAO). 1997. State of the Food and Agriculture. Rome: FAO. Food and Beverage (FoodBev-​SETA). 2017. Food and Beverage Manufacturing Sector Education and Training Authority Sector Skills Plan (2018–​2019): Annual Update. Pretoria: SETA. Greyling, Jan, Nick Vink, and Edward Mbaya 2015. ‘South Africa’s agricultural sector twenty years after democracy (1994–​2013)’, Professional Agriculture Workers Journal, 3(1): 1–​14. Harcourt, Dave 2011. ‘The South African food processing industry’. A  report of the study on agroprocessing commissioned by the Embassy of the Kingdom of the Netherlands. Pretoria: Dutch Embassy. Industrial Development Corporation (IDC). 2017. ‘Financing the agro-​processing and agriculture value chain’, [presentation]. International Trade Administration Commission of South Africa (ITAC). 2016. Agro-​ processing Markets and Related Trade Trends: Opportunities and Challenges for South Africa. Johannesburg: ITAC. Morris, Mike, and Justine Barnes. 2014. ‘The challenge to reversing the decline of the apparel sector in South Africa’. A paper presented at an international conference on Manufacturing-​ Led Growth for Employment and Equality in South Africa. Johannesburg 20–​21 May 2014. Nair, Reena, and Namhla, Landani. 2020a. ‘Technological developments in South Africa’s fruit industry and implications for market access and participation’. CCRED Working Paper no. 2019/​5, Johannesburg. Nair, Reena, and Namhla Landani. 2020b. ‘Making agriculture value chains more inclusive through technology and innovation’. SA-​TIED Working Paper no. 104, Johannesburg. National Planning Commission. 2011. National Development Plan:  Vision 2030. Pretoria: National Planning Commission. National Treasury. 2020. Economic Transformation, Inclusive Growth, and Competitiveness: Towards an Economic Strategy for South Africa.’ Pretoria: National Treasury. Nattrass, Nicolli, and Jeremy Seekings. 2013. ‘Job destruction in the South African clothing industry: How an alliance of organised labour, the state and some firms is undermining job intensity growth’. Centre for Development Enterprise (CDE Focus) Occasional Paper, Johannesburg. Ncube, Phumzile, Maria Nkhonjera, Tamara Paremoer, and Tatenda Zengeni, (2016). ‘Competition, barriers to entry and inclusive growth: Agro-​processing, CCRED Working Paper, no. 2016/​3. Nhundu, Nicolas, Anthea Paelo, Mamoletji Thosago, and Thando Vilakzi, 2017. ‘Growth strategies of large and leading firms: Food production and processing sector assessment’. CCRED Working Paper no. 10/​2017/​, University of Johannesburg. Patel, Ebrahim. 2016. ‘Unravelling the fabric of the industry: South Africa’s clothing and textile business’, The Journalist, 24 May. http://​www.thejournalist.org.za/​spotlight/​unravelling-​the-​ fabric-​of-​the-​industry-​south-​africas-​clothing-​and-​textile-​business accessed 20/​11/​2020. Pogue, Thomas. 2008. ‘A sectoral analysis of wood, paper and pulp industries in South Africa.’ Research report commissioned by the Department of Labour. Pretoria: IERI. SA Cotton, 2019. 2014–​2018 South African Cotton Textile Statistics [online] https://​cottonsa.org. za/​resources/​textile-​statistics/​. Statistics South Africa. 2019. [Excel download]. Statistics South Africa. 2019a. Statistics in Brief 2019. Pretoria: StatsSA. Statistics South Africa. 2019b. Inequality in South Africa: A Multi Dimensional Diagnostic of Inequality, Report No.03-​10-​19. Pretoria: StatsSA.

Agro-processing Industries in the South AfricaN Economy    239 Trade and Industry Policy Strategies (TIPS). 2017. ‘Manufacturing subsectors:  Wood and paper’. Pretoria: TIPS. United Nations Research Institute for Social Development (UNRISD). 2010. Combatting Poverty and Inequality: Structural Change, Social Policy and Politics. Geneva: UNRISD. van Lin, Micha, Aart van den Bos, and Nazeem Sterras. 2018. ‘The current state of fruit and vegetable agro-​processing in South Africa: Challenges and opportunities’. Report of a Study Commissioned by the Netherlands Enterprise Agency, Dutch Embassy, Pretoria. Visser, Margareet, and Stuart Ferrer. 2015. ‘Farm workers’ living and working conditions in South Africa: Key trends, emergent issues and underlying and structural problems’. Research Report commissioned by the International Labour Organisation, Pretoria. Vlok, Etienne. 2006. ‘The textile and clothing industry in South Africa’, in H. Jauch and R. Traub-​Merz (eds) The Future of the Textile and Clothing Industry in Sub-​Saharan Africa. Bonn: Friedrich-​Ebert-​Stiftung. Zalk, Nimrod. 2014. ‘Industrial policy in a harsh climate:  The case of South Africa’, in J. SalazarXirinachs, I. Nubler, and R. Kozul-​Wright (eds) Transforming Economies: Making Industrial Policy Work for Growth, Jobs and Development. Geneva:  International Labour Office. Zalk, Nimrod. 2019. ‘Hiding in plain sight: High value-​added agriculture’s large-​scale potential to grow jobs and exports’. ECON 3x3 [online] https://​www.econ3x3.org 21/​11. 2020.

Chapter 12

L and and Ag ra ria n Devel opme nt i n Sou th A fri c a Ruth Hall and Farai Mtero

12.1 Introduction South Africa’s high levels of economic inequality arise in part from, and are in turn reinforced by, persistent and racialized inequality in access to and control of land. Land reform—​ often equated with the redistribution of white-​owned commercial farms to Black smallholders—​has been a largely unfulfilled political promise during the first twenty-​five years of democratic rule in South Africa. Land ownership and control historically underpinned patterns of unequal development in South Africa. Forced land dispossession through colonial conquest and under apartheid rule formed the basis for ‘agrarian dualism’, by which we mean the coexistence of two distinct countrysides: a thriving white-​owned commercial farming sector across most of the territory and, amidst this, a landscape of impoverished and overcrowded Black ‘reserves’ or ‘homelands’. These two landscapes inherited and persistently display distinctive features, with salient differences in the types of land use, levels of capital investment and technology, tenure arrangements, land governance, and administration. Overcoming this stark divide and resolving ‘the land question’ has been a central political promise of successive democratic governments. South Africa’s negotiated transition produced a constitution that provides certain protections to property rights while simultaneously mandating land reforms through land redistribution, tenure reform, and restitution, including via expropriation. In contrast to this legal framework, the policy, institutional, and financial approach was shaped by the World Bank-​promoted model of market-​assisted land reforms (MALR). Unlike state-​driven land reforms, predominantly associated with peasant struggles under both capitalist and socialist rule in East Asia and Latin America, the MALR model has involved the policy choice to rely

Land and Agrarian Development in South Africa    241 on voluntary market transactions rather than compulsory acquisition or expropriation, and to limit state involvement in subsidies, regulation, or wider agrarian restructuring. Attempts to redress land inequality in South Africa have therefore centred on assistance via state and private-​sector financing for Black South Africans to acquire farmland. Alongside the redistribution of land to those the state chooses, and restitution of land to those who make historical claims, land reform promises ‘tenure reform’, meaning the transformation of regimes of property rights and property holding. The concept of tenure reform is to redress, in law and in practice, the inequality and hierarchical relationship between private title—​the tenure of the privileged white minority—​and forms of customary and informal tenure that most people hold to the land they occupy and use. An estimated 60 per cent of South Africans have no documented rights whatsoever to their property, which means that the entire edifice of the deeds registry, cadastral system, surveys, and mapping, fail to provide any core of land administration to underpin land uses for most (Hornby et al. 2018). Land dispossession of South Africa’s peasantry was central to the creation of cheap labour supply—​and in this sense can be considered to underpin capitalist development in mining, agriculture, and industry (Bundy 1998; Wolpe 1972). A defining feature of the twentieth century was the reliance of these sectors of the economy on the migrant labour system, with large numbers of especially young men being recruited from the rural areas for wage work on recurrent one-​year contracts. The Natives Land Act of 1913 marked a watershed moment in cementing racialized land-​ownership patterns, by setting aside 9 per cent of the land as native reserves (later formalized as ‘homelands’ and extended to 13 per cent of the land). One outcome of dispossession and forced resettlement was the large population of land-​poor Black rural producers farming on small parcels of land, coupled with pervasive landlessness and overcrowding in ‘rural slums’ (Murray 1987). These legacies remain imprinted on South Africa, with poverty and deprivation still most starkly evident in the communal areas previously designated as Black ‘homelands’ (see Figure 12.1) and formed the basis for widespread demands for land reform, to broaden Black access to land across the rest of the country, to redress inequality and injustice, and as a central element of liberation and democratization.

12.2  Political Economy of Transition Among the ‘compromises’ of South Africa’s political transition in the 1990s were the terms of the constitutional treatment of property and the framing of a market-​based land reform. Yet the presumption that the ANC ‘sold out’—​as widely alleged by opposition parties—​suggests that it had a coherent position to begin with, a view we contest. The policy direction was driven by geopolitical shifts, a state verging on bankruptcy, and rampant capital flight, combined with divisions within and between exile-​based liberation movements and the mass democratic movement at home on how to address land inequality (Drew 1996; Hall 2010). The result was a set of land reforms alongside

242    Ruth Hall and Farai Mtero South African Index of Multiple Deprivation 2011 at Ward Level showing former homeland boundaries

Figure  12.1  South African Index of Multiple Deprivation 2011 at ward level showing former homeland boundaries Source: SASPRI (2014).

economic policies that undermined their chances of success. A package of agricultural reforms centring on deregulation of internal markets and trade liberalization satisfied the need to cut costs, withdraw the subsidies to white commercial farmers, and respond to both domestic and international pressure to open up markets. The result by the mid-​1990s was a faster-​than-​required adherence to World Trade Organization targets for liberalization, including tariffication and the dismantling of the elaborate architecture of regulation that had underpinned and subsidized white capitalist agriculture: input subsidies, capital subsidies, transport subsidies, export subsidies, subsidized interest rates, and a single-​channel marketing system for all agricultural products (Vink and Kirsten 2003). In the space of just over a decade, from the mid-​ 1980s to the late-​1990s, this dramatically transformed one of the most state-​subsidized farming systems in the world (see also Chapter 10 in this volume). It also constrained the possibilities for resolving the ‘land question’ and prompted concentration in land ownership. Two parallel processes reshaped land and agricultural policy during the 1990s. The first is widely referenced but almost universally misunderstood, and this is the story of the constitutional protection of property rights. The ANC was not forced into accepting constitutional protection of property rights, nor did it fight for nationalization of land. Its long-standing ambivalence on ‘the land question’ is central to what transpired in

Land and Agrarian Development in South Africa    243 the crucial period of 1990 to 1994. Nationalization of the land, mines, and banks was adopted as a policy position only at the ANC’s Morogoro Conference in 1969, when its revolutionary programme declared that the land of ‘land barons, absentee landlords, big companies and State capitalist enterprises’ should be confiscated and redistributed to ‘small farmers, peasants and landless of all races who do not exploit the labour of others  . . .  to the landless and the land-​poor’, leaving no private ownership of land for commercial production that involved the use of hired labour (ANC 1969). This position—​always contested within the movement—​was abandoned by 1988 when, in its Constitutional Guidelines for a Democratic South Africa, the negotiating position for a new dispensation, the ANC proposed that only corporate and commercial property would be subject to state regulation and liable to confiscation, alongside retention of private property ownership ‘for personal use’ (ANC 1989). This mixed tenure system was abandoned by 1992 when the ANC’s Land Manifesto rejected any proposal for nationalization, instead advancing a two-​stage plan in line with a national democratic revolution, which sought first to consolidate democratic majority rule prior to transition to land nationalization and agricultural collectivization. Often inspired by memories and imagery of rural resistance, like the Bambatha rebellion and the Pondoland revolt (Mbeki 1964), the ANC’s thinking on ‘the land question’ was largely historical as its key thinkers had limited connection with rural resistance in recent times. Unlike Zimbabwe, South Africa’s liberation struggle was not fought primarily in the countryside, but rather in the urban townships and in the frontline states, and the ANC had largely dismissed the revolutionary potential of rural dwellers (Drew 1996). Urban and rural communities that resisted forced removals through the processes of homeland consolidation in the 1960s and 1970s, and black spot removals in the 1980s, were more strongly linked to non-​governmental organizations and academics within the country, with whom the ANC had relatively little contact. Instead, many of those coming into the ANC’s leadership were of the view that the ‘land question’ was bound up with labour control, and that its resolution required dismantling the Bantustans and the creation of a unitary farming system. But what this would look like was far from agreed. Returning exiles were key among those in the Land Commission who led the ANC’s thinking on the subject in the transition period. Among them were those from Eastern Europe and the Soviet Union many of whom, given their experiences, favoured collectivization and state farming, while others, particularly, from Mozambique and Zambia, were more optimistic about the potential of smallholder farming (Hall 2010). Having initially argued against constitutional provisions on property, the ANC proposed a property clause that would entrench a land reform programme, rather than the blanket protection of individual private property, which the National Party and others had sought. In this way, Section 25 of the Bill of Rights in the Constitution affirmed the transformation of property, including but not limited to land. It set out enforceable rights to restitution of land rights for the dispossessed, redistribution to ensure equitable access for the landless, and tenure reform for those with insecure rights, as justiciable constitutional rights (RSA 1996). The Constitution provides limited procedural

244    Ruth Hall and Farai Mtero protection to all property holders, including tenants and other non-​owners, while mandating land reform, empowering the state to expropriate for land reform purposes. The second process that unfolded simultaneously was a set of plans for agricultural policy, and the economy in general. The ANC opted, in the sphere of agriculture, for the World Bank’s Proposals for Rural Restructuring, which combined agricultural deregulation with market-​based land reform. After a short interregnum in the form of the Reconstruction and Development Programme (RDP) from 1994 to 1996, economic policy took a sharp turn to the right, with the ‘1996 class project’ and its expression in the Growth, Employment and Redistribution (GEAR) framework which confirmed market-​based policies including, crucially for land reform, confirmation of property rights and the avoidance of confiscation and expropriation. In this way, economic policy determined that the state’s constitutionally enshrined powers to expropriate property in the interests of redressing land injustice and land inequality would not be used. Shorn of any link to a wider plan for economic restructuring, the RDP from 1994 promised that a new land reform programme would be the ‘central and driving force of a programme of rural development . . . and through the provision of support services, the democratic government will build the economy by generating large-​scale employment, increasing rural incomes and eliminating overcrowding’ (ANC 1994). Such a programme never materialized, as ‘rural development’ was shuffled between various institutional homes—​as a provincial mandate, in the RDP office in the Presidency, then Land Affairs, then provincial and local government—​over the coming fifteen years, before settling in 2009 under the Zuma administration into the ministry responsible for land reform, at which point its arrival diverted funds away from land reform into flagship projects in areas unaffected by land reform. Finally, welfare in the poorest wards of the ex-​Bantustans would be ‘rural development’ while land reform would focus on transferring commercial farms from white to Black farmers, without any discernible decongestion of the communal areas, nor expansion of employment or livelihoods, nor poverty reduction. Against the backdrop of the end of the Cold War, a new model of ‘market-​based’ or, in more sanitized form, ‘market-​assisted’ land reform was conceived and energetically promoted to South Africa, from the transition years, and thereafter also to Colombia, Brazil, and numerous other developing countries (Deininger 1999). Coupled with some reinterpretation of experiences elsewhere, proponents of the market model advocated voluntary transactions in place of expropriation, and ‘demand-​led’ processes driven by beneficiaries rather than state planning. In these ways, the inefficiencies of state-​led reforms could be avoided. Yet such characterizations of landowners and the landless negotiating peaceful land transfers ignored power and politics and isolated technical solutions from political context. On the basis of experience in the Philippines, and also in Brazil and Colombia, Borras (2007) showed how this approach ignored the asymmetry in negotiations, pushback from landowners to get the state to carry more

Land and Agrarian Development in South Africa    245 of the cost, and wrongly assumed that decentralization would ensure accountability (Borras 2007). In the contested policy process, South African agricultural economists—​including some returning ANC exiles who later took up key positions in government—​worked with the World Bank to develop proposals aimed at reconciling agricultural and land reforms. While those who blame the Constitution for land reform failures focus on the ways in which the ANC’s hand was ‘forced’, we argue instead that policy alliances transcended party politics, as expertise from within South African universities, international advisors, non-​governmental organizations, and state and party structures coalesced, with people moving across these spaces and taking up new posts, working as consultants for one another, and shaping one another’s thinking. The combination of dramatic agricultural reforms alongside moderate land reforms was the result of convergence rather than compromise (Hall 2010). Equally, the social forces pushing for land reforms—​rural communities and their allies among the NGOs—​largely had no economic argument in favour of redistribution and had no space in the wider debates about the economy, while the ANC, at the centre of policymaking, had little on the table in terms of land reform beyond a need to limit costs, reduce subsidies, and produce a pro-​poor offering. In this context, it was the World Bank that was able to argue that the political necessity of redistributing land was also economically efficient as long as it prioritized market-​oriented smallholder farming, and used market mechanisms for acquisition, rather than resorting to coercive means (Hall 2010). Large farms, tightly integrated into the corporate food system, dominate the commercial farming sector (Hall and Cousins 2018). Successive agricultural censuses show evidence of increased consolidation. In 2017, 2,610 large farms (those with an annual income of more than R22.5 million) constituted 6.5 per cent of the total number of farms in the commercial agriculture industry, and accounted for 67 per cent of total income and 51 per cent of total employment (StatsSA 2017). As Bernstein (1996) predicted, and later confirmed (Bernstein 2013), deregulation led not to a level playing field, let alone advantages for small-​scale farmers and newcomers, but rather to re-​regulation in the form of market dominance by several concentrated agribusinesses. Among these are the privatized (former state) co-​operatives which have expanded into services and up-​ and ​downstream activities, beyond farming (Williams et al. 1998: 67). The removal of subsidies in the commercial farming sector not only spurred greater concentration but also raised barriers to entry for small-​scale farmers. This epitomizes the ‘policy disconnect’ in post-​apartheid land reform and agricultural policies (Cousins 2013: 57). This disjuncture between land reform and agriculture persists in post-​apartheid policy. Land reform proceeds in the absence of appropriate agricultural policies and meaningful support for smallholder producers. The absence of significant coordinated efforts towards rural development means that land reform has proceeded largely in isolation from other interventions in infrastructure and enterprise development.

246    Ruth Hall and Farai Mtero

12.3  Post-​apartheid Land Reform South Africa consists of 122 million hectares, of which about 14 million hectares (11 per cent) is arable. A significant proportion, about two-​thirds, is privately owned as commercial farms, while the state owns much of the remaining land, including the ‘communal’ areas held under customary land tenure and held in trust by the state (see Table 12.1). Land reform has proceeded slowly, making only modest inroads into the ‘dualist’ pattern of land use, ownership, and tenure. In the first twenty-​five years since democracy, about 10 per cent of formerly white-​owned commercial farmland was transferred to Black South Africans through various mechanisms. The quality of the outcomes has been mediocre, with some people benefiting, but substantial failures of land-​use planning and support, with the result that the projected benefits of land access, including poverty reduction, have largely not materialized. A 2009 Quality of Life Survey found that 52 per cent of the land reform beneficiaries are producing some income from the cultivation of their own land, and that the average income from agriculture production of the beneficiaries was extremely modest, at R19,270 per annum, with the median income only R4,234 due to the skewing effect of a small number of very high earners (QoL 2009: 79). Another national survey in the same year found that 28 per cent of the redistribution projects were stable, 21 per cent had improved in performance, 22 per cent showed marginal benefits, and 29 per cent had completely failed (Umhlaba Rural Services 2009). Amidst a discursive commitment to small-​scale farmers and poverty reduction, land reform has also undergone fundamental redesign over time, reflecting shifting politics.

Table 12.1: Areas of land by land use, ownership, and tenure category Land use, ownership and tenure category

%

Commercial farmland

67

Urban areas

8

Protected areas and other state land

10

‘Communal areas’

15

Ex ‘homelands’ (excluding former KwaZulu)

10

Ingonyama Trust (former KwaZulu)

2

Former ‘coloured’ reserves

1

Other customary lands held in trust by the state Total

2 100

Source: PLAAS (2013), citing Department of Land Affairs (DLA 2002).

Land and Agrarian Development in South Africa    247 The design and delivery of land redistribution has shifted in several ways since the 1990s. First, the ‘means test’ to target only poor households was removed and replaced with a race criterion. Second, the provision for both urban and rural redistribution was replaced with an exclusively rural focus. Third, instead of land being made available for multiple uses for diversified livelihoods, it was restricted to being provided for agricultural production alone. Fourth, joint ventures with commercial strategic partners were introduced as a primary means of securing capital for continued production post-​reform. Fifth, from 2011, the state abandoned its transfer-​of-​title model and instead opted to directly purchase land and lease it to potential beneficiaries through the State Land Lease and Disposal Policy (SLLDP). This means land reform beneficiaries neither own nor lease the land. In short, after a quarter of a century, land reform morphed from a programme for poor people to have access to and secure rights to land for whatever they needed it for, typically in groups and through democratic management of collectively but privately held property, into a programme for small groups of Black South Africans, mostly better-​off men, who could secure capital, to access state land on a long-​ term rental basis. Land redistribution came to be joined by a specific programme of land restitution, as people who had been unjustly and forcibly removed from their land demanded the right to return to their own land. Unlike the wider redistribution process, those claiming back their own land are required to show proof that they were unfairly dispossessed due to racially discriminatory laws or practices—​or that their forebears were, after the commencement of the Natives’ Land Act of 1913. Prior colonial conquest and subsequent dispossession are unaddressed through restitution. After nearly twenty-​ five years of land reform, 4.8  million hectares had been redistributed and 3.5 million hectares restored to the original owners and their heirs through restitution claims—​a total of just under 10 per cent of all commercial farmland (see Table 12.2 below). Restitution of land has proceeded unevenly, with a growing budget and large numbers of claimants, but making a modest contribution of about 5 per cent of white commercial

Table 12.2: Land redistribution and restitution 1994-​2020 (in hectares) Programme

1994–​2018

Restitution

4,119,782

Redistribution Total

% of commercial farmland: 86,186,026 ha

% of all land: 122,320,100 ha

4.8

3.4

5,167,818

6.0

4.2

9,287,600

10.8

7.6

Source: DRDLR (2016, 2018), DALRRD (2021): authors’ calculations. Note: These land reform delivery figures are from April 1994 to March 2020. Commercial farmland is land categorized as agricultural, outside the communal areas of the ex-​bantustans, regardless of whether it is publicly or privately owned.

248    Ruth Hall and Farai Mtero farmland, and a negligible inroad into urban spatial inequality. Unlike in other contexts where post-​conflict restitution has been more successful, the length of time, generational gaps, changes in demography, land use, and priorities between the time of dispossession and the time of restitution, have rendered South Africa’s restitution process intractably complex (Walker et al. 2010). Dispersed descendants of people dispossessed a century ago are now expected to agree on the terms of restitution—​whether to opt for land restoration or financial compensation and, where land is to be returned, how it will be held and managed, who will live on it, how it will be used, who will contribute what capital and labour, and how the proceeds are to be divided. Not surprisingly, conflicts emerge between better-​off claimants with alternative livelihoods, and those who aim to return to an ancestral home. Unlike redistribution, where any land can be acquired, depending on the willingness of the owner to sell, restitution is guided by the specific locations of historical claims. As a result, while much redistributed land has been marginal, much of it in the semi-​arid regions, or farms not under full production—​because the state finds these cheaper—​many highly capitalized farms have been the target of restitution. Within the first decade, it became apparent that transferring such properties to large groups of people without the capital with which to retain ongoing operations, and without adequate resources or support after transfer, saw devastating production failures, and the state responded by contracting private companies as ‘strategic partners’ to manage the farms. They were to keep production underway, with the promise of profit-​share for ‘beneficiaries’, who were usually prohibited from living on their own land under such arrangements, or making any use of it themselves (Lahiff et al. 2012). In some cases, the main benefit for those having their land returned was to get priority access to waged employment on the farms (Hall et al. 2013). Restitution, initially highly judicialized, was expedited when negotiated settlements between claimants and the state became the norm from the late 1990s, and only disputed claims were referred to the newly established Land Claims Court. Coupled with this, massive administrative settlement of claims with small cash payouts enabled the state to speed up resolution of claims, but without substantial contributions to land reform—​ claims were settled but the land was not changing hands (Hall 2010). The re-​opening of claims for a further five-​year-window period from 2014 to 2019 saw a massive leap in the lodged claims—​many orders of magnitude and more new claims were lodged than old—​and based on existing trends, it would seem that it will take decades to investigate and resolve these claims—​we estimated 144 years at current rates (Cousins et al. 2014). These bureaucratic challenges sit astride a toxic politics in which traditional authorities have been opposing claims, and launching counterclaims, on the basis that expansion of settlement onto restored land, outside of their authority, undermines their status—​ a concern upheld by President Zuma on numerous occasions. The central question is whether Black landholders must in some way be subject to chiefly authority (and so the geographic jurisdiction of chiefs extended onto formerly white-​owned farms) or whether people can choose whichever land governance structure they wish to administer their land. These disputes are infused with particular urgency when high-​value land, or land with the potential for mining or other commercial developments, is involved. The

Land and Agrarian Development in South Africa    249 Landless Movement of South Africa (LAMOSA) launched a successful Constitutional Court challenge on precisely this, and in 2016 obtained an order that defended the rights of claimants to choose their landholding institution (including a democratically elected one rather than a traditional institution) and won the rights of existing claimants to have their claims addressed prior to new counterclaims, including from chiefs aiming to extend their jurisdiction, being entertained. The geography of land control has changed only modestly, with much of the land acquired under redistribution being in the semi-​arid Northern Cape—​where it is cheaper and where extensive livestock production is the main land use—​while restitution claims are clustered in the north-​east of the country, where a larger territory remained under Black control at the time of the Natives’ Land Act, unlike the Cape where the Act had little impact as dispossession had preceded it by several centuries. Both redistribution and restitution, then, have had different spatial impacts on landholdings and land use (see Figure 12.2), but this is not indicative of the areas of the country where there is greatest need or demand for land, but rather due to supply-​side factors. Not only is the (re)distribution of land geographically uneven, but the pace of reform has varied greatly over time. Figure 12.3 depicts yearly delivery of land via redistribution, initially by calendar year and from 2000 onward by financial year. The vertical lines depict the start of the presidencies of, respectively, Mandela (1994), Mbeki (1999), Motlanthe (2008), Zuma (2009), and Ramaphosa (2018). The decline is generalized but also uneven. In the Eastern Cape, for instance, by 2018 the number of people getting land, and the number of projects, was around 5 per cent of 700000 600000 500000 400000 300000 200000 100000

Redistribution

pe

t

Ca rn W es te

rth No

rth

er n

Ca

W es

pe

a ng m pu M

No

ala

po Lim

Na aZ

ul u-

po

l ta

g te n Ga u

St at e Fr ee

Kw

Ea

ste

rn

Ca

pe

0

Restitution

Figure 12.2  Land redistribution and restitution, by province, in hectares, 2009–​18 Source: DRDLR (2018): author’s calculations. Note: This covers nine financial years from 2009/​10 to 2017/​18, coinciding precisely with the Zuma presidency.

250    Ruth Hall and Farai Mtero 600000 500000 400000 300000 200000

2016/17 2017/18 2008/19 2018/19

2015/16

2014/15

2013/14

2012/13

2010/11

2011/12

2009/10

2007/08

2008/09

2006/07

2005/06

2004/05

2003/04

2001/02

2002/03

2000/01

1999

Jan-Mar 2000

1998

1997

1996

1995

0

1994

100000

Figure 12.3  Land redistribution under five presidents, 1994–​2018 Source: DRDLR (2016: 4) and DRDLR (2018: 5): author’s calculations. Note: Figures are per calendar year until 1999 and, after the first-​quarter adjustment in 2000, are per financial year.

what it had been a decade earlier (DRDLR 2018). In contrast, a full 40 per cent of all those people who got land from the government during the Zuma decade were in KwaZulu-​ Natal. Political direction of budgets may underpin this; certainly, farmer support budgets, especially tractor and other equipment programmes, were shifted heavily to the province of KwaZulu-​Natal during this period. Women lost out throughout, and everywhere. The most promising data suggest that, by 2016, women constituted 23 per cent of all the beneficiaries in the first two decades of land redistribution (DRDLR 2016). Previously, women had accounted for about 16 per cent of beneficiaries—​but got far less of the land, and the budget, due to being clustered mostly in more modest group-​based projects compared to the bigger commercial enterprises (Kepe and Hall 2016: 37). Fluctuating budgets, rather than an absence of farms for sale, largely explain changes in the pace of land reform, and in particular the overarching downward trend from 2008/​09 onward, coinciding with the global financial crisis and the start of the Zuma presidency, as Figure 12.4 shows. While the state began to redistribute land from white to Black, private transactions saw growing numbers of Black people buying land—​a trend that some claim could be substantial, though the data are unclear (Lyne and Darroch 2003). Simultaneously, as the farming sector adapted to the new regulatory environment, and was exposed to the vagaries of market conditions and domestic and global pricing, a major shake-up saw vast numbers of heavily indebted white farmers leave the sector—​not, as had been thought, to be replaced by Black farmers, but rather bought out by more successful white farmers, including foreign buyers, and domestic and foreign companies. The immediate outcome was a steep decline in the number of commercial farming units from over 60,000 in 1996 to 45,000 just six years later in 2002, and then stabilizing around 40,000 by 2007 and remaining in that region for the decade (DALRRD 2020). This dynamic of land consolidation prompted changes in land use, production, inputs, technologies,

Land and Agrarian Development in South Africa    251 Land reform budgets 1996–2019 (inflation adjusted) 7,000.0 6,000.0 5,000.0 4,000.0 3,000.0 2,000.0 1,000.0

19

96 19 /97 97 / 19 98 98 /9 19 9 99 /0 20 0 00 20 /1 01 / 20 2 02 20 /3 03 20 /4 04 / 20 5 05 / 20 6 06 / 20 7 07 / 20 8 0 20 8/9 09 /1 20 0 10 20 /11 11 / 20 12 12 20 /13 13 / 20 14 14 / 20 15 15 20 /16 16 / 20 17 17 / 20 18 18 /1 9

–

Restitution

Land Reform

Figure 12.4  Land reform budgets, 1996–​2019, inflation adjusted Source: National Treasury (various).

and scale. Among the significant shifts were the reduction overall in the area of land cropped (and in the area planted to crops like wheat, oats, and maize); horticultural intensification (expanding land area under apples, grapes, oranges); and sharp declines in the sheep, goat, and cattle herds, and the conversion of livestock and mixed farming lands into game farms (DALRRD 2020). These shifts reflect both climatic pressures, specifically increasing water scarcity, alongside a changing policy and economic context, prompting calls for a ‘just transition’ towards a carbon-​neutral agricultural sector with climate-​resilient production systems, by reducing carbon emissions and moving towards low-​tillage cultivation (Presidential Advisory Panel 2019). Amidst these profound changes, a major flexible cost that farmers sought to cut was labour—​just as new labour and tenure laws to protect long-​term farm workers were introduced. Stripped of state support, old patterns of paternalism that embodied the double-​edged sword of protection and exploitation, fell away: as farmers and companies rationalized their labour forces, vast numbers of Black people lost their jobs, homes, and access to land, through forced evictions from commercial farms. The only national study found that over two million people were displaced from commercial farms in the first decade after democracy, of whom about 940,000 were forcibly evicted (Wegerif et  al. 2005). A  comprehensive failure for land reform meant that more Black South Africans lost access to land via farm evictions than the number of Black South Africans who acquired access through all forms of land reform combined. They were also, largely, not the same people, as the state did not prioritize rural workers, or farm workers, for access to land through redistribution. Alongside attempts to redistribute land have been rising land prices and debt—​ making market-​based land reform more costly and introducing more risk for those

252    Ruth Hall and Farai Mtero acquiring land who were often required to take out loans to contribute towards land purchase and development costs. The total value of agricultural land rose from Rand 37 billion in 1990 to Rand 285 billion in 2019—​while simultaneously total debt rose from Rand 16 billion in 1990 to Rand 288 billion (DALRRD 2020: 28–​9). About 60 per cent of the debt is with commercial banks, 29 per cent with the long-​embattled Land Bank, and the rest is spread between agricultural co-​operatives, private people and other institutions (Business Day 2020). The state has used its discretion to shift the target group of land redistribution in a manner it cannot replicate in restitution where it is obliged to respond to specific claimants. Indeed, despite being a constitutional right for all citizens, ‘access to land on an equitable basis’ has never been legally defined in any manner that directs the content of the right, the nature of demands people can make against the state or one another to realize this right, or the obligations of the state as duty-​bearer to respond (High-​Level Panel 2017). Among the ironies of this policy approach was that the ANC in government replicated many of the features of the Farmer Support Programmes that had been pursued by Bantustan governments, and supported by the Development Bank of Southern Africa—​precisely the approach which the ANC had criticized as attempting to build a Black bourgeoisie rather than to dismantle capitalist agriculture (Dolny 2001: 40). A key goal of land reform in the 1990s was to end the discrimination against forms of tenure other than private title, yet the state has vacillated on core questions, such as how to address the insecure rights of a third of the population who live in the ex-​ Bantustans and whose customary rights are protected only via procedural safeguards in a temporary holding law. Communal tenure reform is stalled entirely, stuck between a political push towards privatization through titling, and demands by residents for their community rights to be recognized. One attempt to pave the way to convert customary rights into private title deeds for ‘traditional communities’ collapsed when rural communities themselves challenged this approach, overturning the Communal Land Rights Act of 2004 in the Constitutional Court in 2010. No clear policy or legal framework exists either for customary and informal land tenure, or for redistribution—​ despite the pleas of two key political processes:  Parliament’s High-​ Level Panel chaired by former President Kgalema Motlanthe (High-​Level Panel 2017)  and the Presidential Advisory Panel on Land Reform and Agriculture (Presidential Advisory Panel 2019). In the absence of policy direction on land tenure, contradictions arise, such as the transfer of titles to restitution claimants while beneficiaries of redistribution are subject to thirty-​year leaseholds, renewable for a further twenty, meaning fifty years of tenancy on state land before being able to acquire ownership of it. Meanwhile, laws meant to secure the tenure of people living on farms, including workers, their families and dependants, and long-​term farm occupiers and labour tenants, continue to be largely unimplemented. More salient than the debates about ‘market’ or ‘state’-​led reform, which dominated the 1990s, and even the deflecting device of the EWC debate—​which all centre on how to acquire the land—​are the somewhat neglected questions of the role of land reform

Land and Agrarian Development in South Africa    253 in economic restructuring, what agrarian sector is to be created; its class character and labour absorptivity; and the ways in which the state and market can respond to and support citizens. On this, the state has consistently proclaimed that ‘small-​scale farmers’ are its primary commitment, and even, since 2007, has espoused ‘agrarian reform’, while simultaneously refusing to subdivide farms, redistribute water rights, or regulate markets. Broadly three perspectives are evident in the debates on the limitations of land reform over the past quarter of a century. First are the ‘market believers’, the liberals who prioritize the preservation of the commercial farming sector while acceding that it must be deracialized, and acknowledging a role for the state in enabling a market-​driven transformation process (Vink and Kirsten 2019). This perspective also privileges the option of privatizing the limited remaining customary lands occupied and held by Black communities—​what is left of the ‘native reserves’, later the Black Bantustans, which became the dumping grounds for people forcibly removed from elsewhere. Second are the ‘nationalist populists’ who call for ‘expropriation without compensation’ (EWC), a phrase that has come to symbolize a generalized rejection of private property, of colonial and apartheid legacies. EWC, as invoked by the Economic Freedom Fighters (EFF), the first political party in post-​apartheid South Africa to mount a significant political challenge from the left of the ruling party, refers to nationalization of all land and its vesting in the state. In the cases where this approach has been implemented to date, though—​in minerals (Minerals and Petroleum Resources Development Act of 2002)  and water (National Water Act of 1998)—​nationalization has not resolved inequalities in access. Rather, the outcomes were to perpetuate unjust appropriation of communal land for mining developments, and of water by a minority of landowners and industries (Capps 2012; Van Koppen et al. 2009). Third are the ‘agrarian radicals’ who object to market-​based land reform while also being deeply sceptical of the state and its pattern of abusing its power for patronage and narrow accumulation. Market processes reinforce land inequalities while the state could act to redress these but generally avoids doing so, for a range of intractable political reasons. Among these are many of the social movements and NGOs, which call not only for racial redress, but also an explicit pro-​poor class agenda, gender equity, and for changes in landholdings, land uses, and the wider agro-​food system (Jara 2019). For clarity, we count ourselves among the latter. This three fold categorization is shorthand to depict only the broad contours of a more complex set of debates, concerning what precisely land reform is for, why it is warranted, who should benefit, how land should be held and administered, what kinds of land uses should be promoted, and what its role should be in wider economic transformation. On each of these, there is no agreement. The only thing all can agree on—​the ‘market believers’, the ‘nationalist populists’, and the ‘agrarian radicals’—​is that land reform has failed thus far to bring about any profound changes in land inequality and in redressing the political and historical injustices that underpin broader economic and social inequality in the country.

254    Ruth Hall and Farai Mtero

12.4  Land Inequality, Financialization, and Elite Capture While the broad contours of the perspectives outlined above are well established, we point in addition to three significant ways in which land underpins not only continuities, but also new dynamics, in economic inequality. First is the rising evidence of elite capture of public resources and corruption in land reforms, with their distributive logic hijacked by narrow interests (Hall and Kepe 2017). Despite the ostensibly ‘market-​based’ approach, South Africa’s land reform has turned out to be bureaucratically constrained and mediated through rent-​seeking networks, with growing evidence of both systemic corruption and inherent policy bias towards elites. The state prosecuting authority’s Special Investigating Unit (SIU) investigated 148 land reform projects between 2011 and 2017 and found that one in four land reform projects were implicated in corruption. As part of these investigations, twenty-​four land reform farms valued at more than R382 million (US $28 million) were recovered (Business Day 2019). Partnerships brokered by the state between land-​ receiving communities and private companies have become a major conduit for corrupt financial flows, as ‘strategic partners’ derive dividends and management fees, and have been implicated in malpractices, such as transfer pricing, to capture value on land reform farms. Mtero et al.’s (2019) research confirms the prevalence of elite capture in land reform and details the different practices and strategies used, by agribusiness, mentors, and powerful interests in business and politics as well as state officials, to capture public resources in land reform. Of the sixty-​two land reform farms investigated, 44 per cent of the farms were allocated to wealthy, urban-​based business owners with business interests in other sectors of the economy. Second is a range of new pressures towards privatization of customary lands, with corporate, political, and traditional elites driving the expansion of large-​scale agricultural and mining investments in communal areas, a pattern driven by powerful corporate, political, and traditional elites into territories reserved for ‘communal’ Black occupation and use but now subject to new pressures of privatization: the powerful chiefly lobby and the pressures of mining companies and, to a lesser degree, agribusinesses seeking to expand in these territories. The democratic parliament has passed several laws that serve to reinforce the distinction between the former Bantustans, where traditional leaders exert control over land via traditional councils, and the rest of the country where land is privately titled. These include the Traditional Courts Bill and the Traditional Leadership Governance Framework Act, dubbed the ‘Bantustan Bills’ by critics including rural social movements, as they entrench different forms of governance, and more limited rights, for those citizens who are Black and live in the former ‘homeland’ areas. Third is the entry of new financial s​ ector actors into corporate landholding, speculation, and financialization of property portfolios spanning urban and rural, and residential, farming, and other land uses. Financialization of the South African economy

Land and Agrarian Development in South Africa    255 accelerated at the end of apartheid as large conglomerates sought reintegration into the global economy following decades of sanctions and international isolation. The establishment of growth funds that link the agricultural sector with financial markets is a growing trend not only within South Africa but also on the broader African continent. These farmland funds mobilize finance and invest in specific investment opportunities, progressively building investment portfolios (Anseeuw et al. 2017). The Africa Future Growth AgriFund, for instance, targets agricultural investments within South Africa and the wider continent. Land reform farms in South Africa have become a new frontier for financial institutions ‘drawn to agriculture by the promise of secure accumulation and wealth preservation amidst financial turmoil’ (Sommerville 2019: 306). Cheap land and significant production support by the state in land reform provides an attractive option for financial capital to invest in land reform farms under the auspices of ‘strategic partnerships’. Wider political economy factors, particularly the politics of transition and subsequent ‘elite pacting’ amongst powerful groups, have profoundly shaped the trajectory of post-​apartheid land reform (Hall 2004). Landed property, private agribusiness, and the nascent class of Black commercial farmers have coalesced around a common vision to deracialize the commercial farming sector through co-​option of Black commercial farmers (Hall 2004). This has served largely to entrench agrarian dualism as opposed to reconfiguring the agrarian structure through inclusion of Black smallholder producers (Cousins 2013). Taken together, elite capture in land reform, contestations over state, corporate, and traditional authority over land, and financialization have presented, and will likely continue to pose, formidable constraints on any attempts to broaden access to land as a basis for economic inclusion.

12.5 Conclusions Initially conceived as a pro-​poor programme to reconfigure the highly unequal and dualistic agrarian structure, South Africa’s land reform was reinvented over time, reflecting wider economic policy shifts towards the creation of a prosperous segment of Black commercial farmers. By 2000, its objective had shifted to deracializing the dominant commercial farming sector without restructuring landholdings and the agrarian economy. The disjuncture between land and agricultural policies was resolved through the promotion of the large-​scale commercial farming model in land reform, the selection of beneficiaries able to contribute their own capital and expertise, and a consistent refusal to subdivide large landholdings. Trade liberalization policies initiated by the apartheid government, and accelerated after democracy, exposed South Africa to global processes of agrofood restructuring while agricultural deregulation dismantled the comprehensive architecture of agricultural subsidies and protections, a regime that had for decades sustained white-​owned, large-​scale commercial farms. Simultaneously, efforts to support ‘emerging’ Black commercial farmers, faltered, largely due to hostile market conditions,

256    Ruth Hall and Farai Mtero contradictory policy directions, and uneven and inadequate resourcing. National data show the effects of the above on land consolidation and economic concentration, at odds with the policy direction; indeed, the broader political economy of agriculture in contemporary South Africa is largely not permissive to a smallholder path articulated in land reform policy. Three contradictions characterize the relationship between land and the economy over the past quarter century. First, the contribution of land to the economy has been driven more by concerns with its value for exchange rather than production. In both urban and rural areas, corporate ownership and land funds drove financialization of land markets, with the entry of new financial sector actors into landholding and land speculation and institutional actors like pension and hedge funds becoming more predominant. Farmland funds and property portfolios now span urban and rural land, and residential, farming, and other land uses. Also associated with this trend is the conversion of agricultural land into non-​agricultural uses, for purposes of leisure, tourism, and residential estates. Equated with ‘foreign owners’, such trends have prompted calls and even draft legislation to limit foreign land ownership, all of which have come to nought. In any case, the precise purpose of such restrictions has never been articulated beyond broad appeals to nationalist populism, and fallaciously equates such trends with international rather than South African-​based capital. Second, contrary to policy objectives, the overall structural shift has been towards land concentration. Elite capture of public resources and corruption in land reforms contribute but do not fully explain this. Rather, policy bias combined with market trends have deepened land inequality. Broadening access to land was to be the centrepiece of a land redistribution programme, which aimed not only to deracialize ownership of the economy but also to provide land ownership as a basis for broad-​based economic development. Yet, the modest steps that the government took towards redistribution did little to counteract the overarching trend towards land concentration. The consolidation and vertical integration of larger enterprises in agriculture, alongside land concentration, demonstrates the overall impact of land and agricultural reforms. Concentration and financialization have spurred rapidly rising land prices, posing a challenge not only to a market-​based land reform programme but more broadly to the vast majority of citizens who are excluded from access to land for residential, farming or other livelihood purposes. Third, contrary to the notion of tenure reform, the hierarchy between private ownership and customary and informal tenure was reinforced, with commodification and privatization of the latter rather than, as initially envisaged, the creation of durable and democratic systems to recognize, protect, and administer all land within a single legal and administrative framework. Rather, information about landholdings, their uses, and who has what rights to them, is either dispersed across different institutions or absent entirely. The asymmetry in land administration reflects continued dualism between the formal system of land title, deeds, and surveys, and customary tenure. Amidst this asymmetry in land rights have been resurgent large-​scale land deals on the small (and shrinking) communal areas, driven by powerful corporate elites in combination with

Land and Agrarian Development in South Africa    257 state and traditional authorities, usually for mining or agricultural purposes, and often vigorously contested by residents. These three contradictions suggest that the popular equation of land ownership with ‘white farmers’, as evident in recent debates over expropriation without compensation, while historically valid, is increasingly inaccurate and fails to capture the dramatic transformations in land ownership and its position in the economy. Rather than a failure to transform land relations, the past quarter century is better understood as one of massive transformation—​in a direction of concentrated ownership of land, control of value chains, and patterns of accumulation—​in short, financialization, concentration and commodification. This is the antithesis of land reform. The ‘land question’ in South Africa has clearly expanded in the past quarter century since the dawn of democracy. The first expansion is from the ‘land question’ to the ‘agrarian question’, with the now widespread acceptance that, far from attempting to preserve the agrarian structure, land reform should, by design, alter the size distribution of landholdings, promote greater opportunities for access to well-​located smallholdings for production both for household consumption and for sale, and should extend beyond ‘land reform’ to significant regulatory interventions to address the power of corporations in upstream activities, in input industries, in technologies like seed, in downstream processing, food manufacture, and retail. A second expansion of the ‘land question’ is to challenge its equation with farmland, to also insist on urban land reform, and opportunities for urban agriculture as well as for non-​farm uses of rural land. While some progress has been made in deracializing land access via redistribution, restitution, and private transactions, this has been slow. Where the land regime has changed, or where new owners and users have acquired land, the outcomes and economic benefits have been limited due to the absence of a coherent rural development strategy in which access to land forms the basis for wider economic opportunities. In this context, and with the commodification of customary land, and mass urbanization, the politics of land has ‘gone urban’. With well over 60 per cent of the population urbanized, in both large metropolitan areas and growing medium-​sized cities, the land question can no longer be equated with an agrarian question. Mounting demand for access to land for housing, but also for urban agriculture and other uses, has sparked numerous conflicts, particularly between landless urban dwellers and local authorities. These centre on access to land, spatial planning, and the provision of well-​located land, and demands for recognition of informal rights. For instance, even where people have obtained state-​subsidized housing under the RDP (Reconstruction and Development Programme), 50 per cent of them are without title or their titles are inaccurate or outdated (PAP 2019: 89). This means that, like the rural areas, urban land inequality consists not only of unequal access to land, but also a continued inequality in the kinds of tenure rights that people hold, with private title remaining out of reach for most urban residents, especially poor and Black city dwellers. In the face of lengthy delays in housing provision, with people waiting decades for a subsidized house, in recent years the government has responded to demands for land access by delinking housing from

258    Ruth Hall and Farai Mtero urban land reform and promising ‘rapid land release’ so that people can access land to build their own shelters. Redistributing land has not been, in South Africa, a mechanism used to address a crisis of surplus labour in a de-​agrarianized society. The path dependency of an economy founded on a mineral–​industrial complex, open for global trade, and with a shrinking farming sector characterized by the consolidation of capital and concentration of landholdings, is at odds with any plan for land reform, save for the deracialization of agrarian capital. A spectrum of efforts has been launched to achieve precisely this, through Black economic empowerment shareholding schemes or farm equity schemes for farm workers. Persistent land inequality continues to underpin broader asset and income inequality, and spatial and economic exclusion, both within the rural areas marked by agrarian dualism between commercial farms and communal areas, and in towns and cities bearing the scars of apartheid spatial planners. The fragile legitimacy of the post-​ apartheid order is challenged constantly, and the continued inability or unwillingness of either the state or capital to confront land inequality renders ‘the land question’ constantly at the ready for populist grandstanding, election-​time sloganeering, diplomatic troubleshooting, investor kneejerk jitters, tenderpreneur cashing-​in, and farmer defensiveness. In their haste to invoke it, or dispel it, both the ruling party and opposition parties do no service to the real issue: the ways in which imaginative, principled, and structural changes in the distribution, holding, administration, and use of land could expand horizons for people’s lives and livelihoods.

References ANC (African National Congress). 1969. ‘The Freedom Charter: Revolutionary programme of the African National Congress’, in ANC (ed.) Intensify the Revolution:  The ‘Morogoro Conference’, Morogoro, Tanzania, 26 April–​1 May. ANC (African National Congress). 1989. ‘Constitutional Guidelines for a Democratic South Africa’, South African Journal of Human Rights, 5(2): 129–​32. ANC (African National Congress). 1994. Reconstruction and Development Programme. A Policy Framework. Johannesburg: Umanyano Publishers. Anseeuw, Ward, Jean-​Marc Roda, Antoine Ducastel, and Norfaryanti Kamaruddin. 2017. ‘Global strategies of firms and the financialization of agriculture’, Sustainable Agriculture and Tropical Agri-​chains, 1: 321–​37. Bernstein, Henry. 1996. ‘South Africa’s agrarian question: Extreme and exceptional’, Journal of Peasant Studies, 23(2–​3): 1–​52. Bernstein, Henry. 2013. ‘Commercial agriculture in South Africa since 1994: Natural, simply capitalism’, Journal of Agrarian Change, 13(1): 23–​46. Borras, Jun. 2007. Pro-​poor Land Reform: A Critique. Ottawa: University of Ottawa Press. Bundy, Colin. 1998. The Rise and Fall of the South African Peasantry. London: James Currey Publishers. Business Day. 2019. ‘Special investigating unit (SIU): Millions looted in land reform scam’, 24 January. Business Day. 2020. ‘Downgrade could hit the agriculture sector hard, AgBiz says’, 30 March.

Land and Agrarian Development in South Africa    259 Capps, Gavin. 2012. ‘Victim of its own success? The platinum mining industry and the apartheid mineral property system in South Africa’s political transition’, Review of African Political Economy, 39(131): 63–​84. Cousins, Ben. 2013. ‘Land and agriculture uncoupled: The political economy of rural reform in post-​apartheid South Africa’, in Paul Hebinck and Ben Cousins (eds) In the Shadow of Policy: Everyday Practices in South African Land and Agrarian Reform. Johannesburg: Wits University Press. Cousins, Ben, Ruth Hall, and Alex Dubb. 2014. ‘The restitution of Land Rights Amendment Act of 2014: What are the real implications of reopening land claims?’, Policy Brief no. 34, PLAAS, University of the Western Cape. Murray, Colin. 1987. ‘Displaced urbanization: South Africa’s rural slums’, African Affairs, 86 (344): 311–​29. DALRRD (Department of Agriculture, Land Reform and Rural Development). 2020. Abstract of Agricultural Statistics. Pretoria: DALRRD. DALRRD (Department of Agriculture, Land Reform and Rural Development). 2021. Annual Report 2019/​2020. Pretoria: DALRRD. DRDLR (Department of Rural Development and Land Reform). 2018. ‘Annual statistical report for selected service delivery programmes’. DRDLR, Pretoria. Deininger, Klaus. 1999. ‘Making negotiated land reform work: Initial experience from Brazil, Colombia, and South Africa’. Policy Research Working Paper no.  2040, World Bank, Washington, DC. Dolny, Helena. 2001. Banking on Change. Johannesburg: Penguin Publishers. DRDLR (Department of Rural Development and Land Reform). 2016. ‘Memorandum: To provide a draft input on reporting on women’s empowerment and gender equality contribution to nation building and social cohesion’. Task no. T2016-​1015, 15 September 2015, DRDLR, Pretoria. Drew, Allison. 1996. ‘The theory and practice of the agrarian question in South African socialism’, Journal of Peasant Studies, 23(2/​3): 53–​92. Hall, Ruth, 2004. ‘A political economy of land reform in South Africa’, Review of African Political Economy, 31(100): 213–​27. Hall, Ruth. 2010. ‘The politics of land reform in post-​apartheid South Africa, 1990 to 2004: A shifting terrain of power, actors, and discourses’. PhD dissertation in politics and international relations, University of Oxford. Hall, Ruth, Paul Wisborg, Shirhami Shirinda, and Phillan Zamchiya. 2013. ‘Farm workers and farm dwellers in Limpopo Province, South Africa’, Journal of Agrarian Change, 13(1): 47–​70. Hall, Ruth and Ben Cousins, 2018. ‘Exporting contradictions: The expansion of South African agrarian capital within Africa’, Globalizations, 5(1): 12–​31. Hall, Ruth, and Thembela Kepe, 2017. ‘Elite capture and state neglect: New evidence on South Africa’s land reform’, Review of African Political Economy, 44(151): 122–​30. High Level Panel of Parliament Report. 2017. Assessment of Key Legislation and Acceleration of Fundamental Change. Parliament of the Republic of South Africa. Hornby, Donna., Rosalie Kingwill, Laurel Royston, and Ben Cousins 2018. Untitled: Securing Land Tenure in Urban and Rural South Africa. Johannesburg: Wits University Press. Jara, Mazibuko. 2019. ‘Land redistribution in South Africa: Pondering the solidarity economy alternative.’ Paper presented at:  Resolving the Land Question:  Land Redistribution for Equitable Access to Land in South Africa Conference. Institute for Poverty, Land and Agrarian Studies, Bellville.

260    Ruth Hall and Farai Mtero Kepe, Thembela, and, Ruth Hall 2016. ‘Land redistribution in South Africa.’ Commissioned report for High Level Panel on the assessment of key legislation and the acceleration of fundamental change, an initiative of the Parliament of South Africa, 28 September. Lahiff, Edward, Nerhene Davis, and Tshililo Manenzhe, 2012. Joint Ventures in Agriculture:  Lessons from Land Reform Projects in South Africa. London:  International Institute for Environment and Development. Lyne, Michael and Mike Darroch, 2003. ‘Land redistribution in South Africa:  Past performance and future policy’, in L. Nieuwoudt and J. Groenewald (eds) The Challenge of Change: Agriculture, Land and the South African Economy. Pietermaritzburg: University of Natal Press. Mbeki, Govan. 1964. South Africa: The Peasants’ Revolt. London: Penguin Books. Mtero, Farai, Nkanyiso Gumede, and Katlego Ramantsima, 2019. ‘Elite capture in land redistribution in South Africa.’ Institute for Poverty, Land and Agrarian Studies Research Report no. 55. Presidential Advisory Panel on Land Reform and Agriculture. 2019. Report of the Presidential Advisory Panel on Land Reform and Agriculture.’ Republic of South Africa, 28 July. Quality of Life Survey. 2009. ‘Monitoring and evaluating the quality of life of land reform beneficiaries and the impact of land transfers:  2005/​2006’. Department of Land Affairs, South Africa. RSA (Republic of South Africa). 1996. Constitution of the Republic of South Africa, No. 108 of 1996, Government Printers: Pretoria. Sommerville, Melanie. 2019. ‘Agrarian repair: Agriculture, race, and accumulation in contemporary Canada and South Africa’. PhD thesis, University of British Columbia, Vancouver. StatsSA. 2017. ‘Census of commercial agriculture: Financial and production statistics’. StatsSA Report no. 11-​02-​01, Pretoria. Umhlaba Rural Services. 2009. ‘National assessment of the comprehensive rural development pilots’. Unpublished report submitted to the Department of Rural Development and Land Reform. Barbara van Koppen, Hilmy Sally, Michael Aliber, Ben Cousins and Barbara Tapela. 2009. ‘Water resources management, rural redress and agrarian reform’. Development Planning Division. Working Paper Series no.7, DBSA, Midrand. Vink, Nick and Johann Kirsten. 2003. ‘Agriculture in the national economy’, in L. Niewoudt and J. Groenwald (eds) The Challenge of Change: Agriculture, Land and South African Economy. Scottsville: University of Natal Press. Vink, Nick, and Johann Kirsten 2019. ‘Principles and practice for successful farmland redistribution in South Africa.’ Working Paper no. 57, Institute for Poverty, Land and Agrarian Studies, University of the Western Cape, Bellville. Walker, Cherryl., Anna Bohlin, Ruth Hall, and Thembela Kepe (eds). 2010. Land, Memory, Reconstruction and Justice:  Perspectives on Land Restitution in South Africa. Athens, OH: Ohio University Press and Scottsville: University of KwaZulu–​Natal Press. Wegerif, Marc, Bev Russell, and Irma Grundling, Irma. 2005. ‘Still searching for security: The reality of farm dweller evictions in South Africa’. Nkunzi Development Association, South Africa. Williams, Gavin., Joachim Ewert, Johann Hamann, and Nick Vink. 1998. ‘Liberalizing markets and reforming land in South Africa’, Journal of Contemporary African Studies, 16(1): 65–​94. Wolpe, Harold. 1972. ‘Capitalism and cheap labour-​power in South Africa: From segregation to apartheid’, Economy and Society, 1(4): 425–​56.

Chapter 13

Mining and Mi ne ra l s i n Sou th A fri c a Neva Seidman Makgetla

For 150 years, the mining industry has shaped economic and social development in South Africa. As of 2020, it continued to disproportionately affect exports, the production structure, investment, climate change, economic power, and labour relations. Yet it accounted for only a limited share of GDP and employment. Its outsized impacts arose largely because it remained a central link between the South African economy and international markets. This article first reviews the role of mining in the South African economy. It then outlines the main changes in the sector since the transition to democracy as well as key elements in government regulation of the industry. A discussion of the socio-​economic impacts of these changes follows, illustrated by the experiences of the coal and steel value chains and the Marikana massacre. The impact of mining on the economy and society is best understood through analysis of the value chain, rather than through extraction alone. The mining value chain stretches from the production of inputs—​in South Africa, principally capital goods and construction services—​to mining itself, to the metals and coal refineries, to export markets and domestic manufacturing. It draws on a host of critical support services, notably infrastructure and finance.

13.1  Mining in the South African Economy The mining value chain (excluding construction and coal-​based chemicals) contributed 13 per cent of GDP in 2019. Of that, 8 per cent was in mining itself and 4 per cent in the metals refineries and electricity. In that year, the value chain absorbed 17 per cent of

262   Neva Seidman Makgetla investment. In 2018, it accounted for 6 per cent of employment, with only 4 per cent in mining and 1 per cent in the refineries. But it contributed over half of exports in 2019 and used 40 per cent of electricity and freight transport. The economic significance of the mining value chain remained largely intact through the democratic era, as Figure 13.1 shows. Until the 2000s, gold was not always included fully in the trade data, in part because it was considered part of the balance of payments and in part because of apartheid-​era sanctions. The importance of the mining value chain for South Africa emerged in the way economic growth tracked global metals prices. The economy saw an upswing during the unprecedented global commodity boom that lasted from around 2002 to 2011, then slowed sharply when it ended. In 2011, global mining prices reached a thirty-​year high in constant dollar terms, a level last seen in 1980 (see Jacks 2016). The impact of the upswing was increased by the growing synchronization of price cycles across commodities. After 2011, however, mineral prices plummeted and in 2019 they were 40 per cent below their peak. South Africa’s GDP grew at 3.5 per cent a year from 2002 to 2011 during the commodity boom, but slowed to an average of 1.9 per cent from 2011 to 2015, then dropped to 0.7 per cent annually from 2015 to 2019 (see Figure 13.2). Compared to other upper-​middle-​income economies, South Africa was unusually dependent on exports of mining-​based commodities, even if China is excluded from the comparison. The pattern emerges from figures for revealed comparative advantage. The revealed comparative advantage is the ratio of the share of an industry in South African exports to the share of the industry in exports by comparator economies, in this case upper-​middle-​income countries excluding China. From 2014 to 2016, the ratio was 1.1 for ores, fuels, and metals; between 1.25 and 1.45 for (largely coal-​based) chemicals and

Mining

Metals refineries

Metal fabrication

Machinery (a)

Electricity

Figure 13.1  Indicators of the significance of mining for the economy, 1994 and 2019 Source: Calculated from Quantec, EasyData, Standardised industrial series. Downloaded from https://​www.quantec.co.za in October 2020. Note: (a) Includes electrical equipment except cables; excludes appliances and transport equipment.

6%

25% 20% 15% 10% 5% 0% –5% –10% –15% –20% –25%

% Change in GDP

5% 4% 3% 2% 1% 0% –1% –2%

% Change in major mineral prices (a)

Mining and Minerals in South Africa    263

2019 2018 2017 2016 2015 2014 2013 2012 2011 2010 2009 2008 2007 2006 2005 2004 2003 2002 2001 2000 1999 1998 1997 1996 1995

GDP

Mineral prices in US dollars (right axis)

Figure  13.2  Annual percentage change in GDP compared to annual percentage change in international price of South Africa’s main mining exports Source: For GDP, Statistics South Africa, Quarterly and Regional Fourth Quarter 2019, Excel spreadsheet downloaded in March 2020; for metals price, Index Mundi and Kitco data, downloaded in January 2020; and for trade weights, Quantec EasyData, international trade data at 6-​digit HST level in current Rand, accessed at https://​www.quantec.co.za in March 2020. Note: (a) Index of US dollar prices for coal, iron ore, gold, and platinum, weighted by share in exports.

auto, which were South Africa’s only major export industries outside of the mining value chain; and 0.8 for capital goods and agriculture and food products. In contrast, for the labour-​intensive and design-​intensive products that kickstarted export-​oriented industrialization in Asia—​clothing and appliances—​it was only between 0.2 and 0.4. Internationally, the biggest increase in mineral sales and investment during the commodity boom occurred in copper and other industrial minerals, which did not rank amongst South Africa’s main mining products. This contributed to muted growth in investment and mineral rents in South Africa compared to other mining-​dependent countries. Still, the value of South African mining exports in constant Rand terms, deflated by CPI, rose by an average of 6 per cent a year from 2002 to 2011, compared to growth of 4 per cent in other goods exports. As metals prices plummeted from 2012, export revenues from mining fell back an average of 0.6 per cent annually from 2011 to 2019. Other exports, however, climbed 3 per cent a year (calculated from Quantec 2020b). The trade data do not adequately capture figures for exports of services, which means they overstate the importance of mining. With the opening of the economy from 1989, finance and, to a lesser extent, tourism grew disproportionately rapidly, although figures on trade in services are notoriously unreliable. That said, a significant share of the financial sector depended on mining, funding both investment and trade, and dealing in mining shares on the stock exchange. Mining was also critical for the economy as a result of South Africa’s unusually extensive reliance on coal to generate electricity. In the 2010s, over 90 per cent of South African electricity was fuelled by coal, compared to under 15 per cent in other upper-​ middle-​income economies excluding China, and 70 per cent in China. Before the

264   Neva Seidman Makgetla commodity boom, cheap coal fostered an unusually energy-​intensive economy, at almost nine megajoules per dollar of GDP in 2015 or twice as much as in other upper-​ middle-​income economies outside of China (and a third higher than China). In the 2010s, the increase in the price of coal contributed to soaring electricity prices becoming a significant drag on growth. Pressures to shift to greener energy sources offered opportunities for growth, but also created a series of conflicts and disruptions in the established coal value chain. Finally, mining played a central role in the political economy of the democratic era. Above all, the mining houses dominated big business. Companies in the mining value chain accounted for around a fifth of equity on the Johannesburg Stock Exchange (JSE) in 2020. The mining industry, and especially the labour movement, was also an important source of political leadership in the African National Congress (ANC). Two of South Africa’s five presidents in the democratic era (Kgalema Motlanthe and Cyril Ramaphosa) had been general secretaries of the National Union of Mineworkers (NUM). Still, as the labour movement was legalized, mining became less dominant in organized labour. At The Congress of South African Trade Unions (COSATUs) founding conference in 1985 NUM’s 100,000 members accounted for one-fifth of the total. In 2012, although it claimed over 300,000 members, its share in COSATU fell to one-seventh, mostly because of the rapid unionization of the public sector from 1994. In 2018, both the miners’ unions and COSATU itself had fragmented, but miners constituted 8 per cent of total union membership.

13.2  Evolving Production and Power Structures from 1994 While the mining value chain remained central to the South African economy and the political economy after the transition to democracy, its production and power structures evolved significantly. On the one hand, a range of new products emerged. On the other, ownership and worker organization underwent far-​reaching changes. These trends did not, however, lead to a more equitable industry, generate significantly more employment, or promote broader economic development.

13.2.1 The Production Structure In the 1980s, as South Africa’s historically dominant gold mines ran out of easily accessible ores, they began downsizing. As foreign trade and financing became more accessible from the early 1990s, mining companies consequently diversified into new minerals. Iron ore, platinum, and coal production outstripped gold for most of the democratic era, while ferroalloys and later chrome and manganese ore exports also grew rapidly. Gold output dropped to 105 tonnes in 2019 from 580 in 1994 (which was down from the

Mining and Minerals in South Africa    265 450 400

75

350

74

70

73

250 48

200 150 100 50

38 29 23 32 28 27

34

12

13

50 67

44

51

11 11 53 48

20 17 49

55 47

24 62

58 55 69 71

81

77

74 70 65

61

15 18 16 28 15 23 29 15 15 53 56 47 33 37 36 51 32

65

67 61

51

52 70 62 51 56 53 96 91 97 94 94 98 83 90 85 86 60

49

49 44 47 45 45 99 96 93 93 94 69 57 64 56 53 41 42 93 89 86 82 82 86 94 36 81 77 69 34 59 54 52 58 31 46 25 28 30 30 26 34

1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019

-

42

31 43

58

76 74

66

81

300

76

Other

Chrome

Manganese

Iron ore

Gold

Coal

Platinum

Figure 13.3  Contribution of mining to GDP in constant (2020) prices, by commodity Source: Calculated from Quantec, EasyData, Standardised industrial series. Downloaded from https://​www.quantec.co.za in October 2020. Note: (a) Reflated with CPI based to 2019. The CPI is used to indicate the value a​ dded from production of the commodity relative to the rest of the economy, given changes in relative international prices. In contrast, the GDP in constant Rand aims to measure volume, with no change in relative prices over time.

peak of 680 tonnes in 1984). In contrast, as Figure 13.3 shows, iron ore production more than doubled from 30 million tonnes in 1994 to 72 million in 2019; coal expanded by a third, from 195 million tonnes to 260 million; platinum, from 185 to 270 tonnes; chrome, from four million to 18 million tonnes; and manganese, from three million to 17 million tonnes. The changing commodity structure in mining was associated with shrinking employment. Except for platinum, the new mines were mostly open pit, and therefore less labour intensive than the underground gold mines. As Figure 13.4 indicates, the number of miners dropped from a high of almost 800,000 in 1980 to 415,000 in 2002. It recovered to 512,000 at the height of the commodity boom in 2011, but then fell to just over 450,000 in the first quarter of 2020. In gold, employment fell from half a million in 1984 to 400,000 in 1994 and 90,000 in 2019. In platinum, it climbed from 97,000 in 1994 to 195,000 in 2011 before falling to 170,000 in 2019. The platinum mines absorbed many of the skilled underground miners laid off from gold because they used similar technologies. Coal jobs fell from 130,000 in 1980 to 50,000 in 2002, but then recovered to around 95,000 in 2019. The iron ore mines accounted for 12 per cent of mining value-​ added, but employed fewer than 20,000 people in 2019. The new production structure saw a shift of mining away from the Free State and Gauteng to smaller towns in Mpumalanga (for coal), the North West and Limpopo

266   Neva Seidman Makgetla 900

Thousands

800 700

83 27

600

128

500

77

400 300

200

476

76

68

121

104

74

97

519

50 62

490

92 380

40

53

51

57

96

155

48 38 74

53 48 78

182

186

217 161

100

157

115

53 52 95 168

93

– 1980 1985 1990 1995 2000 2005 2010 2015 2019 Other

Iron, manganese and chrome ore

Coal

PMG

Gold

Figure 13.4  Employment by commodity, 1980 to 2019, in thousands Source: For 1980 to 2015, Department of Mineral Resources, Mineral Statistics Bulletin. Accessed at Quantec EasyData, Mineral Statistics: National Employment and Earnings, Annual. Interactive database accessed at https://​www.quantec.co.za. For 2019, average for year from Department of Mineral Resources, Mineral Statistics Bulletin. Accessed at Quantec EasyData, Mineral Statistics: National Employment and Earnings, Monthly. Interactive database accessed at https://​www.quantec.co.za.

(platinum), and the Northern Cape (iron ore and ferroalloys). The result was harshest on the mining towns of the Free State, while the North West platinum belt around Rustenburg saw a surge in in-​migration for which it was ill prepared. At provincial level, the decline in gold mining had far more severe impacts on the Free State than on Gauteng, which had a much more diversified economy. The Free State’s share in national mining sales fell from 15 per cent in 1994 to 5 per cent in 2017; for Gauteng, the drop was from 22 per cent to 10 per cent (calculated from DMR 2018). From 1994 to 2019, the Free State lost 95,000 mining jobs, or 15 per cent of its total formal employment. Gauteng lost some 120,000 mining jobs, but that was just 3 per cent of formal employment there. In large part because of the decline in gold, the Free State saw the slowest economic growth in the country, expanding just 1.1 per cent a year on average from 1994 to 2018, compared to 2.4 per cent for the rest of the country. In the same period, despite a similar decline in gold mining, Gauteng was the fastest growing province, with the provincial GDP climbing by 2.8 per cent annually (Quantec 2020c). People voted with their feet: the population of the Free State climbed just over 10 per cent from 1996 to 2019, while the national population increased by almost 50 per cent and Gauteng’s population doubled. Meanwhile, the historically rural platinum belt in the North West, and on a smaller scale the mining towns of Limpopo and the Northern Cape, saw extraordinary

Mining and Minerals in South Africa    267 in-​migration. The population of the North West’s platinum belt doubled from 1994 to 2019, although the province as a whole only grew 50 per cent. The overloading of local services emerged from figures on housing. Around a third of platinum-​belt households lived in informal housing in 2019. For comparison, the population of Johannesburg climbed 125 per cent in this period, but only a sixth of the population lived in informal housing; the population of Cape Town grew 85 per cent, with a fifth in informal housing (Quantec 2020c). As discussed below, the difficulty of accommodating the influx into the platinum belt contributed to a labour-​relations crisis that culminated in the Marikana massacre in 2012. The democratic era also saw sharp fluctuations in beneficiation, especially inputs to steel (iron ore, chrome, and manganese) and aluminium. The 1990s brought a step up in highly energy-​intensive ferroalloys and aluminium refining. The trend reversed from the early 2000s, however, thanks to soaring electricity prices combined with increased overseas demand for ores, mostly from China. Historically, South Africa beneficiated most of its mining output. For much of the past century, around half of coal, by weight, was used locally to produce electricity and liquid fuels, while gold and platinum were refined at the mines before export. In the 2010s, around a tenth of South African platinum was manufactured locally into catalytic converters, largely for export. A state-​owned company, ISCOR, began producing steel from local ore and coal in the 1920s; it was privatized in the early 1990s and in the early 2000s sold to Arcelor Mittal, but still dominated local production. From the early 1990s, Eskom promoted ferroalloys and aluminium plants to increase demand for electricity. These refineries are extremely energy intensive, and the plants were effectively designed to beneficiate coal as much as metal ores. Indeed, the aluminium plants came to South Africa solely to get low-​cost electricity, since the bauxite was imported from Australia. Three new aluminium plants established at Richard’s Bay and in Mozambique used almost a tenth of Eskom’s total electricity production in 2013. Eskom was responding to an electricity surplus that resulted from its overexpansion, despite a sharp economic slowdown from 1985 through 1994. The opening of the economy after 1989 enabled it to leverage foreign investment and export markets. It partnered with the state-​owned Industrial Development Corporation to partially finance the new plants, although all were owned by foreign mining companies. The trends towards energy-​intensive beneficiation reversed in part because the commodity boom increased the export price for ores and coal, and in part because of soaring electricity prices from 2008. As a group, the metals and coal refineries’ electricity use climbed roughly 80 per cent from 1990 to 2008, but then dropped by a tenth through 2019 as electricity prices more than doubled in real terms.1 Their share in Eskom sales rose from 20 per cent in 1990 to peak at 31 per cent in 2000 but fell back to 19 per cent in 2019 (calculated from Eskom Annual Reports for relevant years).

1  The

figures are for all sales to a group of large, energy-​intensive enterprises, around thirty-​five in total, but the bulk goes to metals refineries.

268   Neva Seidman Makgetla Stagnation in local refining meant increased exports of iron, chromium, and manganese ores. These trends meant that South Africa largely lost its dominant position in global ferroalloys markets to China, whose producers imported South African chrome and manganese ores. From 2002 to 2019, by weight, domestic sales of iron ore fell by a third while exports more than doubled. As a result, although domestic users (mostly AMSA) accounted for a third of total iron ore sales in value terms from 1994 to 2005, their share fell to a tenth from 2011 to 2019. In chrome, domestic sales (by weight) doubled from 2002 to 2019, but exports climbed seven ​fold. As a result, in value terms the share of chrome ore processed in domestic refineries dropped from two-​thirds in the decade from 1995 to one-half after 2011. The decline in manganese processing was even larger, with local sales by value falling from a third of the total in the first decade of democracy to less than a twentieth between 2015 and 2019 (calculated from DMR 2018). The changing production structure in the mining value chain was associated with shifts in the direction of trade. The most important development was the emergence of China as a key market for iron ore and later chrome and manganese, and India for coal. South Africa does not publish gold exports by destination, and the information is also not available for a significant share of platinum sales. Figure 13.5 shows the changing pattern of mining-​based exports excluding gold. The rest of Africa was a key market for South Africa’s exports of steel and mining capital goods, although it was insignificant as a destination for mining products. In the 2010s, it bought 8 per cent of South African mining exports, mostly coal and refined petroleum and diesel, up from 2 per cent from 1990 to 2009. But it accounted for a fifth of 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1990 to 1999 Other India

US EU

2000 to 2009

2010 to 2019

Japan, Korea and Taiwan Africa China and Hong Kong

Figure 13.5  Mining exports excluding gold by destination, 1990 to 2019 Source: Calculated from Quantec, RSA Trade HS 8-​digit. Interactive dataset. Accessed at https://​www.quantec.co.za in October 2020.

Mining and Minerals in South Africa    269 South African exports of iron and steel in the 2010s, compared to a tenth in the previous two decades, and four-​fifths of structural steel products, an increase from around half in the 1990s. Moreover, as of 2017, South African producers still supplied over a quarter of all imports by other SADC countries of capital equipment for mining, notably pumps and grinding, earthmoving and material-​moving machinery. They had, however, lost ground to Chinese companies (Levin et al. 2019: 24).

13.2.2 Corporate Structure After democracy, ownership of the South African mines remained highly concentrated, but ownership changed hands repeatedly. Before 1989, the industry was effectively shared amongst a handful of conglomerates, led by Anglo American and Gold Fields. Each of these mining houses held individual mines through subsidiaries that often specialized by commodity. Because of global opposition to apartheid combined with capital controls, from the 1970s the mining houses were increasingly unable to undertake investments outside of southern Africa. In response, they diversified into capital goods production and other manufacturing and financial services in the region. Estimates suggested that Anglo American effectively controlled over two-​fifths of the JSE in 1994,2 including major paper, auto, and foundry interests. Democracy freed the mining houses to move into overseas mining opportunities. In effect, the largest became global mining companies rather than South African conglomerates. They were listed on the London Stock Exchange, initiated mining ventures overseas—​largely in Australia, Latin America, Eastern Europe, and other African countries—​and shed most of their local manufacturing interests. By 2016, Anglo American only controlled around 2 per cent of the JSE.3 The mining conglomerates’ globalization strategy ultimately led to divestment from most of their South African gold mines and many of their other mining interests. This outcome had two roots. On the one hand, foreign shareholders pressured the companies to diversify their risk by investing outside of South Africa. On the other, as mines matured or global prices fell—​notably in the early 2000s and again after 2011—​the conglomerates turned to mining investments in other countries, rather than seeking non-​mining opportunities in South Africa. In 2020, Anglo American’s main holdings in South Africa had narrowed to De Beers, the diamond giant; Kumba; Anglo Platinum; and coal mining, although it only kept its export mines. It had disposed of all its gold and almost all its manufacturing interests. Two of its twelve board members were South Africans. Nonetheless, South Africa still accounted for two-​thirds of its direct employment and half of its tax payments.

2  3 

Information kindly provided by Who Owns Whom, Johannesburg, in 2016. Information kindly provided by Who Owns Whom, Johannesburg, in 2016.

270   Neva Seidman Makgetla Gold Fields divested three of its four remaining gold mines into a new company, Sibanye Gold, now owned by Sibanye Stillwater, in 2012. As of 2020, Harmony and Sibanye Stillwater had acquired most of South Africa’s gold mines; platinum was dominated by Anglo American and Sibanye Stillwater; and iron ore, chrome, and manganese were largely controlled by Anglo American (through Kumba) and Glencore Merafe (Who Owns Whom 2019: 7–​8 and 2020b: 8–​9). Except for Glencore, foreign companies had divested from South African coal almost entirely by mid-2021. Three locally owned and empowered companies - Exxaro, Seriti and Sasol (which operated coal plants primarily to feed its coal-to-liquid fuels refinery) dominated production. The mining houses were owned principally by institutional investors, including foreign investment companies like Black Rock; retirement funds, notably the public service pension funds managed by the state-​owned Public Investment Corporation; and employee ownership schemes. Except for the multinationals with their main listing outside of South Africa—​Anglo American, Glencore, and Gold Fields—​all except Harmony and Gold Fields reported some Black ownership, ranging between 5 per cent and 55 per cent in 2019 (Who Owns Whom 2020a). Black shareholdings by individuals and companies were typically heavily leveraged, however. As a result, they generally expanded when mining prices were high—​as in the commodity boom—​but declined when the mines became less profitable. Irrespective of the size of Black shareholding, the executives of the leading mining companies remained disproportionately white. A third of the seventy-​eight executive directors of the ten largest mining companies with a primary listing on the JSE were black. That compared to half of their ninety-​eight independent directors. A seventh of the executive directors were black women, and a tenth were white women. As the commodity boom ended and mining profits plummeted after 2011, the second-​ generation mining conglomerates also began to seek opportunities overseas. Sibanye’s first annual report identified it as ‘a producer of gold and a major holder of gold reserves in South Africa’, with holdings exclusively in South Africa (Sibanye 2012: 2). By 2019, it billed itself as ‘a leading international precious metals company’ based on its expansion into platinum operations in the United States, Argentina, and Canada, in addition to acquiring Lonmin’s South African mines (Sibanye-​Stillwater 2019: 1). The restructuring of mine ownership did not significantly reduce concentration in the value chain. In 2019, five mining companies with a primary listing on the JSE—​Sasol, South32, Anglo American Platinum, Sibanye Stillwater, and AngloGold Ashanti—​ accounted for two-​thirds of the reported assets of all listed mining companies. Anglo American and Glencore, which had their primary listing on the London Stock Exchange, alone had assets equal to 60 per cent of the mining assets on the JSE (calculated from Who Owns Whom 2020a). Similar data are not available for the big metals refineries because they are mostly subsidiaries of foreign companies, and except for AMSA are integrated with affiliated mines (in the case of aluminium, located in Australia). In 2017, however, Statistics South Africa found that in both steel and other metals, the top five refineries accounted for two-​thirds of sales (StatsSA 2019b: 33).

Mining and Minerals in South Africa    271 In contrast to the mines and refineries, downstream metals manufacturing was fairly competitive. Companies ranged from foundries building custom-made equipment for the mines and manufacturing, to aluminium windows manufacturers, to subsidiaries of foreign companies that imported most of their advanced components, although they often used local structural inputs such as bumpers and metal fittings (see TIPS 2017: 12ff.). In 2018, there were around 100,00 employers in the metals and machinery industries excluding the refineries, with an average of twenty employees (calculated from StatsSA 2018). In 2017, the top five companies only controlled a seventh of sales in structural metal products production, with the share ranging from 25 per cent to 50 per cent for more specialized products such as engines, pumps, and cutlery (StatsSA 2019b: 33).

13.2.3 Unions The miners’ labour unions also underwent substantial changes in the democratic era, although union density remained extraordinarily high. It was just under 80 per cent from 2002, the earliest reliable data, to 2019 (calculated from StatsSA 2002 and 2019). Union density ranged from 60 per cent in coal to 80 per cent in gold and 90 per cent in platinum (calculated from StatsSA 2018). The earliest reported strike by African miners occurred in 1913, with national strikes in 1920 and 1946. In 1982, black miners established a national union, the NUM, in the face of often vicious repression. Non-​African miners already had their own, much smaller, organizations, which largely fought to maintain their privileged position in the workplace. Unionization in the 1980s vastly improved miners’ conditions even before the transition to democracy. In the 1970s, miners, farmworkers, and domestic workers all ranked amongst the lowest paid, most insecure, and most abused formal employees. But their fortunes diverged sharply from the 1980s as miners formed unions, while farm and domestic workers remained mostly unorganized. By the 2000s, the median pay for miners was twice as high as for the rest of the formal sector, at R8,500 a month in 2018. In contrast, the median farmworker earned R2,900 a month, and the median domestic worker, R1,900. For other formal workers, the median wage came to R4,300 a month (calculated from StatsSA 2018). In the 2010s, the NUM’s near monopoly on miners’ representation came to an end with the emergence of the Association of Mineworkers and Construction Union (AMCU). AMCU was founded as a breakaway from NUM in 2001, and in the 2010s became the majority union on the platinum mines. In 2020, both AMCU and NUM claimed around 300,000 members, which, taken together, would mean that union density on the mines exceeded 100 per cent. In reality, the figure was likely closer to 200,000 apiece, reflecting NUM’s loss of around a third of its members during the 2010s. AMCU’s rise largely reflected the difficulties miners faced around work and living conditions in the platinum belt, combined with slower wage increases after the commodity boom ended, as discussed below in the context of the Marikana strikes.

272   Neva Seidman Makgetla

13.3  Industrial Policy and the Mining Value Chain Despite an official commitment to more advanced industrialization, in practice government decision-​making continued to favour the mines and refineries over downstream producers. The consequences could be seen in the coal and steel value chains, where during and after the commodity boom the mine owners increasingly raised prices to downstream industries, effectively capturing the rents from South Africa’s natural resources at the cost of competitiveness in manufacturing. Three key policies illustrate the ambiguity in government’s approach to the value chain: policies on mining rents and industrialization; the public investment programme that began in 2005 and petered out in the late 2010s; and efforts to promote black ownership in mining.

13.3.1 Domestic Prices of Mining Products South Africa’s rich mineral resources, combined with long-​standing infrastructure support and other government incentives, laid the basis for significant mining rents. That is, the South African mines could sell their output overseas at a price above the domestic competitive price, which, by definition, would equal the cost of production plus a normal rate of profit. In theory, if the mines sold their production domestically at the competitive cost-​plus price, it would increase the competitiveness of downstream manufacturing. That would, however, require that the mines sell at least a share of their output below the global price. The question of who should benefit from mineral rents—​the mine operators, the refineries, or downstream manufacturing industries—​became an area of contestation long before the transition to democracy. From at least the turn of the last century, with the formation of Eskom and later ISCOR, the South African state sought to ensure that local refineries obtained mining products at less than the global price. The democratic government did not, however, consistently pursue this goal. Indeed, over time it tolerated substantial increases in the domestic prices of coal and iron ore at the cost of local fabricators. The government did not develop any measures to moderate domestic prices for mining products until the 2010s. It then introduced amendments to the 2002 Mining and Petroleum Resources Development Act (MPRDA) that empowered the minister to compel mines (and arguably refineries) to sell a share of their production to downstream users at the mine-​gate price. The mine-​gate price was defined as the import-​parity price excluding freight costs, which in effect meant the mines would keep the core mineral rents but in the case of bulk minerals would reduce the price by up to a third. The minister could establish a cost-​plus price—​that is, the theoretical competitive price—​only

Mining and Minerals in South Africa    273 by agreement with the producers. In any case, as of 2020, there were no reports that the regulation was ever implemented. The implications of failing to manage the allocation of rents in the mining value chain can be seen in the crises that hit South Africa’s coal and steel value chains in the 2010s. Both were partially due to the ability of upstream mining companies and refineries to capture rents at the cost of downstream producers. Their power in turn was conditioned by government provision of low-​cost, efficient export infrastructure. In terms of the coal value chain, the crisis emerged around electricity, since Sasol owned its own mines and set prices internally. By 2020, soaring electricity prices and load-​shedding had become a critical constraint on the economy. From 2008 to 2019, the unit price for electricity climbed 128 per cent in real terms (deflated by CPI), and Eskom demanded increases at more than 10 per cent above inflation for 2020. Yet its electricity sales fell almost 7 per cent from 2008 to 2019 (calculated from Eskom annual reports). In effect, Eskom had entered a ‘utility death spiral’, where an inability to reduce its costs as demand shrank fed into higher tariffs, in turn further depressing demand (see Makgetla 2021a; Eskom 2019: 47). Five factors underpinned this outcome: • A sharp rise in domestic coal prices from 2002 to 2019 eliminated Eskom’s historic competitive advantage, which had disproportionately benefited the mines and metals refineries through cheap electricity tariffs. • Eskom’s two new huge coal-​based electricity plants were both delayed and faulty, increasing its financing costs while revenues fell short. • Eskom’s older plants began to break down as the new capacity was delayed, leading it to ration electricity repeatedly after 2016, further reducing its sales. • The economic slowdown after the commodity boom ended, and especially the downsizing at the metals refineries, meant demand in any case was unlikely to live up to Eskom’s projections. One of the aluminium refineries closed down in 2014, and AMSA reduced production primarily at its electric furnaces as electricity prices climbed. • Finally, from the early 2010s the government encouraged cheaper renewable energy, which effectively displaced around 5 per cent of Eskom’s output in 2019. The shift in rents after the commodity boom emerged from trends in coal prices. During the commodity boom, both domestic and export prices climbed around 70 per cent. Thereafter, the domestic coal price far outpaced exports. From 2011 to 2019, it climbed by 40 per cent in constant Rand terms while export prices dropped almost 30 per cent (calculated from DMR 2020; deflated with CPI). Various factors enabled the coal producers to capture a higher share of mining rents. In the 2000s, in an effort to bring in more black mine-​owners, Eskom moved towards shorter-​term contracts with smaller producers, displacing its earlier practice of long-​ term contracts with the major mining houses. That, in turn, opened the door to overpayment for politically connected suppliers. At the same time, Transnet continually

274   Neva Seidman Makgetla improved its facilities for export coal, effectively enabling producers to shift to overseas markets if Eskom could not meet their price. Fundamentally, the electricity crisis resulted from Eskom’s failure to adapt its business model as the economy evolved. That business model was effectively based on cheap coal; a continual increase in demand rooted largely in growth in refineries; and an ability to pass on higher costs to users. Through the twentieth century, its symbiotic relationship with the big mining companies meant coal rents lowered the cost of energy, underpinning the development of the mining value chain (see Fine and Rustomjee 1996). In the twenty-​first century, Eskom continued to rely on large coal plants, even as new technologies, including renewables, became both cheaper and more flexible. Meanwhile, it allowed the new investors in the coal mines to retain a higher share of the rents. As it became increasingly uncompetitive, it focused on maintaining its monopoly position and prices. Nonetheless, faced with a deepening crisis around electricity supply, in 2020 the government committed to enabling lower-​cost private generation on a small scale, although in practice it was slow-​walking the process. Contestation over mining rents also fuelled a crisis in steel production in the 2010s. Over the 2010s, the leading crude steel producer, AMSA, closed major plants and cut production. Its decline was due in large part to changing relations with its ore supplier. In the early 2000s, ISCOR sold its mines to an Anglo-American subsidiary, Kumba, and its refineries to the international steel conglomerate Arcelor Mittal. AMSA initially retained a share in the iron-​ore mines, which meant it continued to pay a cost-​plus price for ore rather than the import-​parity price. Downstream manufacturers and government officials argued that it nonetheless charged import-​parity prices for steel, which effectively reduced the competitiveness of both fabricators and the construction industry. In 2010, AMSA lost its share in the ore mines essentially due to management negligence. Kumba promptly started charging it international prices for iron ore. The result was a steep increase in AMSA’s costs. In constant Rand, the unit price for iron ore sold domestically doubled from 2002 to 2011, while the price for exports tripled. From 2011 to 2019, however, export prices dropped 37 per cent while domestic prices rose 15 per cent (calculated from DMR 2020; deflated with CPI). In effect, the rents from South Africa’s high-​quality iron ore shifted from the steel mills to the mines. AMSA’s woes were aggravated by a number of other factors, notably increasingly costly and unreliable electricity from Eskom; a glut on the global steel market that led to plummeting import prices; and its parent company’s apparent disinterest in maintaining production in South Africa. The government’s main response to the crisis at AMSA was to introduce import tariffs to protect it from foreign competition, starting in 2015. That effectively shifted the burden of higher domestic iron-​ore prices onto the far more labour-​intensive, competitive, and innovative downstream manufacturers as well as construction companies. The government did not compel Kumba to return to a cost-​plus price, which would ensure that the rents could benefit downstream producers. Yet domestic sales made up only a tenth of Kumba’s total output. Besides the MPRDA mandates, in theory the government

Mining and Minerals in South Africa    275 had considerable leverage since iron ore exports, like coal, depended on dedicated low-​ cost rail and port facilities provided by Transnet.

13.3.2 Public Investment and Mining South Africa kicked off a major public investment programme with the Accelerated and Shared Growth Initiative for South Africa in 2006, followed by the National Infrastructure Plan in 2012 (see AsgiSA 2008: 41ff; RSA 2020). Both programmes effectively relied on mining, flush from the global commodity boom, to pay for bulk investments. The argument was that sales to mining would fund gargantuan bulk infrastructure schemes while laying the basis for improved services across the economy. Critical projects included road and rail transport to the new mines in the Waterberg; improvements in bulk rail lines to the coast for iron and chrome from the Northern Cape and coal from Mpumalanga; new dams that relied on the platinum mines for offtake; and two of the biggest coal-​based power stations in the world, based largely on the expectation that the commodity boom would continue to boost demand from the rest of the mining value chain. The build programme did not fully neglect other kinds of infrastructure, with commitments, amongst others, to new schools and broadband, and to improving services in communities, in the case of the National Infrastructure Plan especially around the expanded transport corridors for the inland mines. In practice, however, the anticipated investments in communities were rarely implemented in full. This was particularly serious because, as municipalities struggled to improve services in historically underserved black communities after 1994, they had often allowed water, electricity, and roads for manufacturing companies to deteriorate. In contrast to the improvements provided for mining, however, the national build programme did not prioritize funds for other industrial or commercial sites. The bias towards mining in infrastructure plans was reinforced by the central role of state-​owned enterprises charged with improving freight transport and electricity—​ Transnet (for rail), Sanral (roads), and Eskom (electricity). All of these institutions were established before democracy, when promoting mining and later metals and coal refineries were systemic priorities. Both Eskom and Transnet had long-​standing dedicated bulk services for the mines, which generated a significant share of their revenues. As noted earlier, Eskom also supplied electricity, in some cases at favourable rates, for the electricity-​intensive refineries. The dependence of public investment on the mining value chain meant that its fortunes ultimately followed the global commodity cycle, as Figure 13.6 shows. As a share of the GDP, public investment peaked in 2009 at 7.9 per cent. It then plateaued through 2015, before falling to 5.4 per cent in 2019 (calculated from SARB 2020). The synergy with the mines was most pronounced for the state-​owned companies. Transnet cut back on its investments in ore lines as prices fell. In contrast, Sanral and Eskom failed

10.0%

100

9.0%

90

8.0%

80

7.0%

70

6.0%

60

5.0%

50

4.0%

40

3.0%

30

2.0%

20

1.0%

10

0.0%

–

2011 = 100

Public investment as % of GDP

276   Neva Seidman Makgetla

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

General government Public corporations Average annual metals price index (right axis) (a)

Figure 13.6  General government and public corporation (SOC) investment as percentage of GDP and average annual metals price index (2011 = 100), 2000 to 2018 Source: For investment, calculated from South African Reserve Bank. Online Statistical Query. Accessed at https://​www. resbank.co.za in December 2019. For metals price index, Index Mundi and Kitco data, downloaded in December 2019; and for trade weights, Quantec EasyData, international trade data at 6-​digit HST level in current Rand, accessed at https://​ www.quantec.co.za in December 2019. Note: (a) Trade-​weighted average of US dollar prices for platinum, gold, iron ore, and coal.

to adapt to the downturn and both ended up with large losses by the end of the 2010s as mining demand stagnated.

13.3.3 Ownership Like its positions on pricing and infrastructure, the government’s approach to mine ownership was conflicted. From the 1950s, the ANC argued that South Africa needed more equitable ownership of economic assets, rather than merely improving the representivity of shareholders in existing dominant companies (see ANC 1955 and 1969: 32ff). In practice, however, after 1994 it did little to challenge class relations in the mining value chain. Instead, most of its effort centred on leveraging increased interests for black investors, with comparatively limited benefits for miners and mining towns. Before 2002, South African mining law effectively vested ownership in the mine operator in perpetuity. Through the 2002 Minerals and Petroleum Resources Development Act (MPRDA), the government shifted ownership of the resources to the state. Mine owners were required to apply for thirty-​year production licences, which enabled the state to impose a variety of environmental, social, and economic requirements. These conditions were mostly defined through the sector-​wide Mining Charter. The Charter was consulted in detail with organized business and labour, but ultimately gazetted by the minister.

Mining and Minerals in South Africa    277 Through the MPRDA, the government made minority black ownership a precondition for mining licences. The initial target was set at 26 per cent by 2014 but the figure was increased in 2018 to 30 per cent. The 2018 ownership target required that 10 per cent be split between employee and community trusts, with the remaining 20 per cent held by individuals or black-​owned companies. In most cases, the historic owners had to finance the new black investors, who in turn repaid the loans out of their profits. Licencees also had to establish an environmental management plan and a social and labour plan. Under the Charter, they could also earn points towards an overall grade for black representation in management and for procurement from enterprises with at least 25 per cent black ownership; beneficiation (essentially defined as smelting and refining of ores, not metals fabrication); spending on skills development equal to 5 per cent of payroll; and improvements to workers’ housing. In contrast to the ownership targets, however, these requirements were mostly vaguely defined and rarely monitored. The value of BEE deals in mining declined from 2010, largely reflecting the end of the global commodity boom in 2011 as well as the success of many companies in achieving the initial ownership targets. The Minerals Council estimated the value of deals only in current Rand. It found that they dropped from Rand 113 billion across 168 transactions between 2002 and 2010 to Rand 30 billion for sixty-​seven transactions from 2010 to 2019 (Minerals Council 2019: 12). The mining companies’ pyramid structure complicated efforts to measure the government’s success in promoting black ownership. Most of the dominant firms brought black owners into subsidiaries that operated mines, rather than into the holding company. As a result, the listed mining houses had, on average, a lower share of black ownership than other sectors, although the percentage varied from over 50 per cent to under 5 per cent (Makgetla 2021b: 9). In contrast, according to a Minerals Council study that covered ninety-​three mining rights holders—​that is, the mine operating companies themselves—​black investors had, on average, a 40 per cent interest in the operating companies. Black individuals held 22 per cent; mine community trusts, 9 per cent; and employee schemes, 8 per cent. Most of the black investors were highly leveraged, however. As a result, two-​thirds of the companies did not achieve ‘meaningful economic participation’, defined as having clearly identifiable black investors that fully owned their shares, received dividends, and had voting rights (Minerals Council 2019: 11). As might be expected, the historic mining companies objected vigorously to the MPRDA’s requirements. In the media, they tended to blame slower investment and growth in mining in the 2010s largely on these regulations, although they also recognized the impact of the end of the commodity boom and the disruption of the electricity supply (see Baxter 2016). This contestation can be understood as a classic post-​colonial conflict between new and pre-​existing elites over resource ownership. In practice, the main benefits went to a relatively small group of private black investors rather than to communities, workers, or emerging new businesses. The approach to ownership in the Mining Charter has been termed ‘financialized’ (see Bowman 2019) because it centres on financial investments in existing producers

278   Neva Seidman Makgetla rather than more fundamental changes in power relations within the enterprise or the economy. It effectively prioritized raising the share in the mines’ financial returns that went to black investors, and to a lesser extent to employees and communities. Even where new owners enjoyed some voting rights, they typically had only limited impacts on day-​to-​day decision-​making. Work organization remained deeply hierarchical and inequitable, as discussed in the following section. Legally, the social and labour plans, which laid out non-​ownership obligations to workers and communities, had to be consulted, but not agreed, with stakeholders.

13.4  Inequality and the Mining Value Chain Inequality in the mining value chain can also be understood in terms of income distribution and power relations between workers and mine owners, as well as the impact on job creation. The functional distribution of income—​that is, the split of value a​ dded between remuneration and profits—​remained more unequal in mining itself than in the rest of the economy. In 2019, remuneration accounted for just 45 per cent of value a​ dded in the sector, compared to 55 per cent in other industries. The share of wages was lower only in electricity, agriculture, and transport, where it hovered around 30 per cent. Labour’s share in mining value ​added was even lower at the height of the commodity boom, when producers benefited from windfall profits. In 2011, workers got only 38 per cent of value added in mining (calculated from Quantec 2020a). The picture was more complicated in the rest of the value chain. In 2019, in both refineries and downstream production, workers got around 80 per cent of value added (calculated from Quantec 2020a). That was a very high share compared to the rest of manufacturing. In the refineries, remuneration equalled only a third of value-​added in the early 2000s; its share climbed because of declining profitability rather than rising wages. The median pay in metals refining in 2018 was just R5,900 a month (calculated from StatsSA 2019a). For downstream producers, the share of remuneration in value-​ added was stable over the previous twenty years. While miners’ pay improved, work organization did not change much after 1994. Historically, the sector had epitomized apartheid labour relations, which evolved to generate ‘European’ pay for qualified employees—​almost exclusively whites before 1994. To this end, the apartheid workplace effectively centralized skills in a few formally qualified positions, while de-​skilling work for the majority and leaving them with virtually no prospects of promotion. Typically, management relied on commands and even violence rather than consultation. This pattern of workplace relations remained entrenched in South Africa long after the transition to democracy brought legal rights for workers. One result was high levels of workplace conflict.

Mining and Minerals in South Africa    279 In 2018, mining employment remained both heavily de-​skilled and racialized, despite more representative lower and middle management. Only 10 per cent of miners counted as professionals, managers, or technicians. In contrast, in the rest of the economy high-​ skilled positions accounted for 30 per cent of employment. Africans made up around half of the high-​skilled group in mining, virtually the same as in the rest of the formal economy. But Africans accounted for 85 per cent of total employment in mining, compared to just 70 per cent in the rest of the formal economy. Although whites made up only 12 per cent of the mine labour force, they held 40 per cent of high-​skilled mining jobs (calculated from StatsSA 2018). The inequalities were aggravated by limited employment creation in the mining value chain, reflecting its capital intensity. As a result, the sector did little to remedy South Africa’s jobless crisis—​a central source of overall economic inequality. In the late 2010s, only around four out of ten working-​aged South Africans had income-​generating employment, compared to an international norm of six out of ten. Yet job creation in mining lagged behind the rest of the economy for most of the democratic era. The industry’s share in formal employment fell from 6.4 per cent in 1994 to 3.7 per cent in 2018 (calculated from Quantec 2020a). From 2012 to 2018, as the mines adjusted to lower global prices, they downsized employment by around 70,000 positions, or 13 per cent. The job losses had a disproportionate impact on both miners and their communities. For workers with matric or less, mining provided unusually high median pay, at R8,000 a month. The median monthly income for formal workers without matric in other industries was only R3,200 a month, and it was R4,300 for workers with matric. The relatively high pay was particularly striking since only 20 per cent of adults without matric were employed at all (calculated from StatsSA 2018). The loss of jobs was harder because mining was concentrated in relatively small rural towns, which offered few alternatives for retrenched miners. The months-​long strikes in the platinum belt in 2012 and 2014 underscored the difficulties caused by the persistence of apartheid relationships in much of mining despite the transition to democracy. Near Marikana, on 16 August 2012, police fired on strikers, killing thirty-​four miners in the worst massacre since the transition to democracy in 1994. In both platinum strikes, the miners and their families endured real hardship as they went months without pay.4 Given that platinum miners were well paid relative to most other workers, the question becomes why they experienced such bitter workplace conflict. Moreover, the gold and coal mines did not experience similarly prolonged and rancorous strikes. The main factors contributing to the strike included the following. First, despite relatively high wages, miners continued to experience arbitrary and oppressive workplace relations and very limited career mobility. Surveys and focus groups found that many miners still considered their supervisors and managers, whether white or black, to be racist.

4 

This section draws on Makgetla and Levin (2016).

280   Neva Seidman Makgetla Second, the platinum mines grew rapidly in historically remote rural areas, leading to an influx of workers and a housing shortage. Historically the South African mines had built substandard dormitory housing for the workers. After 1994, the mines stopped requiring workers to stay on site, and the government assumed responsibility for housing people who moved to town once they were no longer legally restricted to the historic labour-​sending regions. In the platinum belt, however, miners were still viewed as temporary migrants by both the elected local governments and the ‘traditional’ authorities that controlled much of the land. As a result, most ended up squeezed into informal houses with inadequate services and no land rights. In effect, the racial and ethnic residential restrictions imposed under apartheid persisted even though democracy had brought national citizenship. The fact that miners were relatively well paid only heightened their anger over substandard living conditions. Third, the end of the commodity boom brought a sudden decline in wage increases. From 2002 to 2011, platinum miners’ pay rose an average of 4.4 per cent a year above CPI. From 2011 to 2014, in contrast, their real increase averaged just 0.2 per cent a year (calculated from DMR 2018). Finally, the competition between AMCU and NUM significantly heightened tensions as AMCU sought recognition on the platinum belt. The stakes were high, since in addition to increasing membership fees recognition meant the mines paid for full-​time shopstewards.

13.5 Conclusion The evolution of mining in South Africa after 1994 underscored the challenges of path dependency. The value chain underwent significant changes in terms of its production structure and location as well as ownership and beneficiation. But deep-​seated developmental challenges persisted, ranging from the failure to develop mining-​based manufacturing on a large scale, to the exercise of monopoly power within value chains, to deeply inequitable and oppressive work organization. In this context the democratic state generally avoided the risky and disruptive interventions required to fundamentally re-​orient the value chain in line with its stated aim of more equitable, dynamic, and inclusive growth. As a result, the end of the commodity boom laid bare a range of economic, workplace, and policy conflicts and crises.

References ANC. 1955. The Freedom Charter. Kliptown. ANC. 1969. Strategy and Tactics. Morogoro. AsgiSA. 2008. Annual Report 2008. Pretoria: The Presidency.

Mining and Minerals in South Africa    281 Baxter, Roger. 2016. ‘Mining in South Africa: The challenges and opportunities’. [Powerpoint presentation] https://​www.mineralscouncil.co.za. Bowman, Andrew. 2019. ‘Black economic empowerment policy and state-​business relations in South Africa: The case of mining’, Review of African Political Economy, 46(160): 223–​45. DMR. 2018. Bulletin B1. Quantec EasyData. https://​www.quantec.co.za. DMR. 2020. Mineral Statistical Tables. Pretoria. Accessed through www.quantec.co.za in February 2021. Eskom. 2019. Annual Report 2019. Johannesburg. Fine, Ben, and Zav Rustomjee. 1996. The Political Economy of South Africa. London: Hurst and Company. Jacks, David S. 2016. Chart book for ‘From boom to bust’. February. Downloaded from https://​www.sfu.jacks.ca in June 2016. Update of David S.  Jacks. 2013. ‘From boom to bust: A typology of real commodity prices in the long run’. NBER Working Paper no. 18874, Cambridge, MA. Levin, Saul, Neva Makgetla, and Sthembiso Mtanga. 2019. ‘Moving up the copper value chain in Southern Africa’. UNU Wider Working Paper no. 2019/​52, Helsinki [online] https://​www. wider.unu.edu/​publication/​moving-​copper-​value-​chain-​southern-​africa. Makgetla, Neva, 2021a (forthcoming). ‘SOC mandates, outcomes and the COVID-​19 pandemic’. TIPS Working Paper, Pretoria. Makgetla, Neva. 2021b (forthcoming). ‘The impacts and outcomes of BBBEE’. TIPS Policy Brief, Pretoria. Makgetla, Neva and Saul Levin. 2016. ‘A perfect storm: Migrancy and mining in the North West Province.’ TIPS Working Paper, Pretoria [online] https://​www.tips.org.za. Minerals Council. 2019. ‘Mining industry transformation progress report for 2019’. November, Johannesburg [online] https://​www.mineralscouncil.org.za/​work/​transformation in July 2020. Minerals Council. 2020a. ‘Mining in SA’. https://​www.mineralscouncil.org.za. Minerals Council. 2020b. Facts and Figures  2019. October. Johannesburg. https://​www. mineralscouncil.org.za/​industry-​news/​publications/​facts-​and-​figures. Quantec. 2020a. ‘RSA standardized industry’. [Interactive dataset] Quantec EasyData. https://​ www.quantec.co.za. Quantec. 2020b. ‘International trade service’. [Interactive dataset] Quantec EasyData. https://​ www.quantec.co.za. Quantec. 2020c. ‘Regional service’. [Interactive dataset] Quantec EasyData. https://​www. quantec.co.za. RSA. 2020. ‘National Infrastructure Plan’. https://​www.gov.za/​issues/​ national-​infrastructure-​plan. SARB. 2020. On-​line Statistical Queries. [Interactive dataset] www.resbank.co.za. Sibanye. 2012. Annual Report 2012. Johannesburg. Sibanye Stillwater. 2019. Annual Report 2019. Johannesburg. StatsSA. 2002. Labour Force Survey. [Interactive dataset] https://​www.statssa.gov.za. StatsSA. 2017. ‘GDP P0441: Annual, quarterly and regional, fourth quarter 2017’. [Excel spreadsheet] https://​www.statssa.gov.za. StatsSA. 2019a. Labour Market Dynamics 2019. [Interactive dataset] https://​www.statssa.gov.za. StatsSA. 2019b. ‘Manufacturing industry: Financial, 2017’. Report no. 30-​02-​03 (2017). Pretoria. TIPS. 2017. ‘Manufacturing subsectors: Metals and metal products’. December [online] https://​ www.tips.org.za/​manufacturing-​data/​manufacturing-​sectors.

282   Neva Seidman Makgetla Who Owns Whom. 2019. Mining of Iron Ore and Chrome. Compiled by Alex Conradie. November. Johannesburg. Who Owns Whom. 2020a. Report Generator. [Interactive dataset] https://​reportgenerator. woweb.co.za/​Companies. Who Owns Whom. 2020b. Mining of Precious Metals and Minerals in South Africa: Gold and Uranium, Platinum and Diamonds. Compiled by Alex Conradie. November. Johannesburg.

Chapter 14

E nergy in Sou t h   A fri c a Rod Crompton and Ruwadzano Matsika

14.1 Introduction Global energy problems are ‘complicated, dynamic, and politically charged’ (Kuzemko et al. 2019: 3) in a world grappling with the energy trilemma of energy security of supply, energy equity, or increasing access for the poor and environmental sustainability or decarbonization (World Energy Council 2019). The same is true for South Africa. South African (SA) energy markets have been the target of many state interventions over a long period, which have shaped the nature of these markets. In the post-​ democratic era, government has grappled with the tensions between market-​orientated reforms and state interventions intended to achieve a variety of economic and social policy goals, the main ones being: import substitution industrialization (ISI), security of supply, Black economic empowerment and energy access for the poor. The pursuit of these objectives was expressed through continued extensive regulation of the energy sector under democratic rule, albeit more formalized and transparent, through seven regulatory Acts1 and five regulatory bodies,2 excluding the competition authorities. State ownership as an instrument of market intervention is most pronounced in the electricity supply industry (ESI), through Eskom and municipalities, but less so in the petroleum sector, notwithstanding the existence of several state-​owned companies (SOCs).3 This chapter explores these tensions between market-​oriented reform and state intervention, plotting the shifting paths in each of the major energy sub sectors as government support for market-​based reforms or intervention ebbed and flowed in the 1  Mineral

and Petroleum Resources Development Act, Petroleum Products Act, CEF Act, National Energy Regulator Act, Petroleum Pipelines Act, Gas Act and Electricity Regulation Act. 2  Petroleum Agency of South Africa, Minister of Energy, NERSA, Ports Regulator of South Africa, Transnet National Ports Authority. 3  CEF, iGas, SANEDI, PetroSA, Transnet Pipelines, and Strategic Fuel Fund.

284    Rod Crompton and Ruwadzano Matsika evolving political economy. The chief instruments of intervention have been state ownership and regulation, most inherited from apartheid. The democratic era was preceded by a long history of ISI, reinforced during the apartheid era by security of supply interventions on a massive scale and underpinned by an exploitative, racially based cheap-​labour system. These shaped the energy landscape inherited by the new democracy and left an imprint still evident today. This found expression in a long-​running dependence on cheap domestic coal and imported petroleum. The historical dominance of the minerals-energy ​complex4 (MEC) is still evident in 2020 in an unusually energy-​intensive economy that accounts for 15 per cent of GDP (Bekun et al. 2019). As the South African energy sector has evolved in the post-​democratic era it has not been immune to changes taking place in the global energy sector. Two of the powerful forces reshaping global energy markets are those associated with the Fourth Industrial Revolution and climate-​ change-​ driven moves to decarbonize energy. These twin transitions have impacted the SA energy sector, particularly the electricity sub ​sector and have begun to unravel the MEC. We deal with these impacts and pointers towards future trajectories at the end of this chapter where we pose the question: where will the next tranches of capacity come from? Before embarking upon a closer inspection of the SA energy sector we outline some of its unusual dimensions and endeavour to provide a sense of local and international proportion.

14.2  A Sense of Proportion Globally, 53 per cent of the world’s commercial energy is derived from oil and gas (IEA 2020a). In this regard, South Africa is something of an outlier. Historically, the SA energy sector supported large investments in mining and its associated processing and supply industries, which constituted parts of the MEC (Davidson and Winkler 2003). These energy-​intensive activities were supported by concurrent investments in the energy sector based on the abundant reserves of cheap coal used to generate cheap electricity and resulted in energy-​led economic growth (Bekun et  al. 2019). Energy supply evinces the historical predominance of the MEC with 90 per cent of power still generated from coal (Eskom 2019).5 The composition of the primary energy mix has not changed significantly since 1990. The impact of the renewable power generation programme only increased the renewable 4 The

minerals-​energy ​complex is an organizing concept for understanding the South African economy, developed by Fine and Rustomjee (1996). It posits a system of accumulation centred on powerful alliances between mining, energy, and finance interests underpinned by exploitative apartheid labour systems. For a further exposition of this concept, see ­Chapter 27 in this volume. 5  See ­Chapter 15 in this volume for a comparison of South Africa and global primary energy.

0.8 0.6 0.4 0.2 0.0

2019 GDP PPP billions constant 2017

Emissions intensity kgCO2/ GDP

1.0

25000

10.0

20000

8.0

15000

6.0

10000

4.0

5000

2.0

0

South Africa

Turkey

Brazil

United Germany Kingdom

India

China

Energy Intensity MJ/GDP

Energy in South Africa    285

0.0

Emissions intensity kg CO2 per GDP (2015) GDP, PPP billion constant 2017 $ (2019) Energy Intensity MJ/GDP (2014)

Figure 14.1  GDP, energy, and emissions intensity, certain countries Source: Authors’ calculations, data from World Bank (2020).

share from 5 per cent in 2012 to 6 per cent by 2017 (IEA 2020). This share is expected to increase further by 2030 in line with the government’s Integrated Resource Plan 2019 (IRP) as more renewable power from bid windows 3, 4, and 5 (more below) come on line and some of the coal-​fired power stations are decommissioned. The IRP forecasts that by 2030 renewables (excluding hydro) will account for 34 per cent of installed capacity and 25 per cent of electricity consumption. Energy demand is concentrated in the industry and transport sectors, accounting for 47 per cent and 27 per cent respectively of final consumption (Department of Energy 2020). The balance of energy demand is commerce and public services (8 per cent), residential (8 per cent), agriculture (6 per cent), and other non-​specified energy use (4 per cent) (Department of Energy 2020). Energy intensity is pronounced and carbon emissions intensive (Figure 14.1), attributable to a resource-​intensive economy, low energy efficiency, and coal dependence. China and South Africa have similar profiles that are significantly more intensive than other emerging economies such as Turkey, Brazil, and India. Caution is required in dealing with primary energy statistics as there are large differences between values reported by South Africa (DMRE 2020), the International Energy Agency (IEA 2020a), particularly in petroleum products and renewable energy,6 and BP’s data (BP 2020).

6  The

DMRE (2020) data reveal a marked increase in renewable energy in 2017 requiring further investigation.

286    Rod Crompton and Ruwadzano Matsika Although the SA electricity sector has dominated headlines over the last decade, it is much smaller (by the value of retail sales) than the petroleum products sector, which in 2018 was more than 60 per cent larger than electricity and accounted for 9.4 per cent of GDP.7

14.3 Policy Landscape State intervention in SA energy markets has a long history dating from the 1920s. The post-​democratic state inherited a ‘labyrinthine set of regulatory controls’ (RSA 1998: 5). It set out to bring order and transparency to this with the White Paper on Energy Policy (RSA 1998), which remains the most important policy document for the energy sector along with the National Development Plan 2030 (NDP; RSA 2012). The White Paper announced wide-​ sweeping market-​ orientated reforms. Unfortunately, those promising announcements were overshadowed by subsequent prevarication, implementation failures, impasses, malfeasance, and a return to supply-​ side solutions, leaving competition in energy markets virtually unchanged by 2020. The NDP gave more muted support to market-​oriented reform whilst advocating the efficacy of public–​private financing. Among a plethora of economic and social policy objectives, the White Paper targeted ‘Securing supply (of energy) through diversity’ (Department of Energy 1998: 9) and redress of apartheid ills. The pursuit of these policy goals has contributed to reshaping the political economy and had much to do with the lack of reforms in some markets, as we shall see below.

14.4  Downstream Petroleum Markets South Africa has never been a significant oil nor gas producer. Consequently, upstream oil and gas exploration and production has been treated as a poor relation in an economy where the extractive industries have been dominated by hard-​rock mining, except for a period during apartheid when oil sanctions prompted the government to establish and fund Soekor to search for oil and gas. It found sufficient natural gas off-​shore of Mossel Bay to warrant government building in 1992 the world’s then largest, gas-​to-​liquids refinery (now PetroSA). Three tiny oil fields produced light sweet crude from the 1990s

7  Authors’ calculations based on Eskom (2019) with value of sales to distributors escalated by 30 per cent. Petroleum products includes only petrol and diesel excluding LPG, paraffin, jet fuel, and bunker fuel. Sales based on Department of Energy volumes and prices. GDP calculation based on data from StatsSA Nominal GDP (Statistical Release P0441).

Energy in South Africa    287 until 2008.8 Despite these modest successes, South Africa remains a ‘frontier province’, something policymakers have been reluctant to recognize. The 1998 White Paper raised the possibility of separate legislation for oil and gas and announced the separation of the regulatory function out of Soekor into the Petroleum Agency of South Africa (PASA), a (state-​owned) Central Energy Fund (CEF) subsidiary. This partial reform was never carried through to the envisaged independent regulator due to contestation over the regulatory dispensation, principally the legislation. The White Paper contributed to the Mineral and Petroleum Resources Development Act (Act No. 28 of 2002)  (MPRDA) but the translation of policy into law deterred investors over concerns about the extent of the natural resource rent to be appropriated by the state (‘free carry’ or ‘farm-​in’ rights) and the Black economic empowerment (BEE) burden imposed. Contestation between government and principally the hard-​ rock mining firms but also oil companies, led to the publication and withdrawal of several drafts. Ultimately, in 2019 a new Upstream Petroleum Resources Development Bill was gazetted. Whilst an improvement on previous dispensations, there are still concerns about tax rates, security of tenure, ministerial discretion, ‘free-​carry’ and BEE that may stall yet another new dispensation. The twenty-​five-​year delay in finalizing an investor-​friendly dispensation has cost the country dearly in a lost opportunity. This unresolved regulatory dispensation may delay the development of Total’s 2019/​20 ‘world-​class’ gas condensate discoveries (more below) as it still has to ‘engage with South African authorities regarding the possible conditions of the gas commercialization’ (S&P Global 2020: 1), something that ought to have been clear to investors many years previously.

14.5  Downstream Petroleum Markets In the pre-​apartheid era, petroleum markets were dominated by international oil companies. State interventions were limited to ISI and a modicum of regulation by consensus with the oil companies.9 Market intervention increased under apartheid with the first state coal-​to-​liquids venture in 1954. A sudden and massive increase in state involvement in the sector occurred in the era of oil sanctions against apartheid. Investments were made in synthetic fuels, strategic stocks, and legislation proclaimed gave wide-​sweeping powers, including secrecy, to the ministry. At the start of democracy the IEA found: From exploration through to retailing, [petroleum] was enveloped in a complicated web of interdependent policies, informal arrangements, market sharing agreements, trade restrictions and pricing controls (only some of which were/​are subject to 8 

9 

The Oryx and Oribi fields had recoverable reserves of 25 million barrels (Marquard 2006: 249). See Marquard (2006).

288    Rod Crompton and Ruwadzano Matsika ‘regulation’ in the strict sense of the term). (IEA 1996:  171, quoted in Marquard 2006: 247)

The post-​democratic state set about regularizing and formalizing these arrangements and rationalizing the apartheid legacy assets in the context of the market-​orientated White Paper. The policy formulation arrived at in the White Paper was the product of a three-​way tussle between the international oil companies, organized labour, and the new democratic state. Oil companies fearing Freedom Charter-​based nationalization pushed hard for deregulation. Organized labour wanted job protection for forecourt attendants and the state wanted Black economic empowerment. The compromise struck gave each lobby what it wanted. Deregulation became a core policy but only to be implemented once jobs had been secured and Blacks owned, or there were plans in place for them to own, approximately a quarter of the industry. This three-​way contestation has continued through the democratic era with some notable shifts as the reallocation of regulatory rents has shaped the political economy. In the wake of their vociferous advocacy of deregulation, the oil industry voluntarily signed the first Black economic empowerment (BEE) Charter in 2000 which was intended to accelerate progress in that regard. Over time as the oil companies perceived the threat of nationalization to be receding and as they gained comfort from the manner in which the industry was regulated, including their share of the rents, the oil companies’ public profile has shrunk to almost nothing. Similarly, their advocacy for deregulation has disappeared entirely. Organized labour scored a significant victory in 2003 when an amendment to the Petroleum Products Act (No. 120 of 1977) made forecourt attendant jobs the only jobs in South Africa protected by legislation. Union interest in the sector then shifted to wages where they have managed to wangle a form of negotiation where ostensibly they negotiate with their employers while in fact they haggle with the minister of energy to regulate prices such that they achieve their desired increases. This shift in regulatory rents is discussed in more detail below. Labour has always opposed deregulation because from apartheid times regulation has shifted rents towards retailers and resulted in more service stations and jobs than market forces would permit. The shift in rents towards retailers gained new impetus from the White Paper which included social policy goals promoting small business and BEE. Small business also scored a significant victory when the 2003 amendment to the Petroleum Products Act prohibited vertical integration by oil companies into retailing. Retailing has the lowest barriers to entry in the petroleum sector and was the obvious target for BEE. As this has occurred, the BEE lobby for increased regulatory rents has strengthened. This lobby naturally opposes deregulation for the same reasons as organized labour. The BEE Charter has been a key plank in its argument (Slabbert 2018). Has the White Paper’s BEE milestone been achieved and is the industry able to proceed to deregulation? The Department of Mineral Resources and Energy (DMRE) was meant to monitor BEE progress in the industry annually but has only done so thrice in

Energy in South Africa    289 the twenty years to 2020. Although all of the oil majors have done 25 per cent BEE ownership deals, the sticking points are crude oil procurement and the retail sector. The BEE Charter was written in charter-​type language but when it suddenly became law by being appended to the Petroleum Products Amendment Act in 2003, the inadequacies of that non-​legal language became evident. Incomplete consultations to replace the Charter with provisions from the Broad-​Based Black Economic Empowerment Act (Act No. 53 of 2003) commenced in 2016. Effectively, the oil companies, retailers, and labour have now aligned themselves against deregulation leaving the state isolated in wishing, half-​heartedly, to pursue that objective in a confused and confusing fashion as it wobbles down the tightrope between deregulation and continuing regulation. The following examples demonstrate this tension. In 2018, when there was political concern about high petrol prices (the minister of energy regulates these) the departmental official who was heading regulation wrote to the major industry players cogently arguing for partial deregulation (i.e. price cap) of 93 octane petrol—​a very low volume fuel. In the face of firm opposition from the private sector, the initiative was abandoned. A year later the same official took the opposite view. In a legal matter, he swore that Government believes that the ‘prevention of competition’ in wholesale and retail markets remained a regulatory necessity (Maqubela 2019: 7.3). This contradictory behaviour is not confined to government, it is also evident in the private sector. Although oil companies and retailers support regulation they also seek to increase their market shares. Discounting the price of petrol at the point of sale is unlawful but that has not prevented retailers and oil companies from using a host of devices such as small gifts and discounts in their convenience stores to attract custom. Some say that loyalty-​card discounting is effecting back-​door price deregulation (Majola 2020). The lack of progress in advancing deregulation is a function of the wider political economy but is also partly attributable to the institutional design of the price regulator. Placing a politician and policymaker, the minister of energy, as both the implementer of policy and regulator stems from the Liquid Fuel and Oil Act No. 49 of 1947 and has continued through apartheid and democracy to 2020. It stands in sharp contrast to the National Energy Regulator of South Africa’s (NERSA) more modern design which uses an independent board consisting partially of full-​time experts to make regulatory decisions in a transparent fashion.10 The intent to transfer the minister’s regulatory functions to NERSA was never carried through, with long-​term consequences as we shall see below. Given the five-​year electoral term, the politician-​as-​regulator model is more susceptible to political influence and to short-​term decision-​making than a board. Energy infrastructure decisions often involve assets with long lifespans.11 This regulatory model also has undesirable conflicts in other functions. The minister is both the

10  11 

See National Energy Regulator Act No. 40 of 2004. See Jamison et al. (2004) and OECD (2014) for a fuller discussion of these points.

290    Rod Crompton and Ruwadzano Matsika shareholder of PetroSA (the state oil company) and its regulator. Ministers are responsible for industry development and its regulator. Will a conflicted policymaker, the minister of energy, be able to advance South Africa to deregulation even if the BEE target were to be obtained? The prospects appear slim given the rising power of the small business and BEE lobbies, the power of organized labour, and the general weakening of the democratic state during President Zuma’s administration. Examples of this are the minister’s endeavour in 2012 to compel the oil companies to produce and sell cleaner fuels. The oil companies just ignored the regulation and arrogantly demanded that government grant them the funds to upgrade their refineries. The ministry’s inability to withstand pressure from the regulated entities has resulted in regulated petrol margins increasing by a weighted average real compound annual growth rate of 2.9 per cent p.a. between 2010 and 2019.12 Crompton et al. (2020) found that in 2019 regulated petrol prices exceeded estimated market prices by as much as 90 cents per litre or 5.8 per cent of the average petrol price. The weakness of the minister-​as-​regulator is also evident in the inability to curb the proliferation of service stations which would be necessary to avoid a sudden reduction at the advent of deregulation. Again, the retail and BEE lobbies have prevailed. The petroleum intuitions inherited from the apartheid era were located within the CEF group of companies under the Central Energy Fund Act (No. 38 of 1977). Originally established to combat oil sanctions, the democratic state has tried to repurpose them as instruments of the developmental state, without much success. CEF and its main subsidiaries all made losses in 2019.13 PetroSA and the Strategic Fuel Fund responsible for strategic oil stocks are both associated with corruption scandals. Has the failure to advance to deregulation affected investment in the downstream petroleum sector? Answers to this question may be evident from an examination of imports and domestic refinery expansion and investment. Imports of refined products reached 20 per cent14 of demand by 2018. Although refineries are protected by IPP regulation, no new refining capacity has been added since 1992, apart from capacity creep in the ageing refinery fleet from 708,000 barrels per day (bpd) in 2007 to 718,000 bpd by 2018 (SAPIA 2019). Investment in import infrastructure has kept pace with demand and when Chevron disinvested in 2019, Glencore purchased 70 per cent of Chevron SA, including the Cape Town refinery in a USD1 billion deal. In 2016, Puma Energy15 began acquiring retail assets. All of these developments suggest that there is considerable international interest in the downstream liquid fuels market and that investors are optimistic about import

12 Author’s

calculations based on data from:  http://​www.energy.gov.za/​files/​esources/​petroleum/​ petroleum_​arch.html and https://​www.eskom.co.za/​CustomerCare/​TariffsAndCharges/​Pages/​Tariff_​ History.aspx. 13  In 2018/​19 CEF made a net loss of R471 million, PetroSA a net loss of Rand 2 billion and has an unfunded decommissioning liability of Rand 7.4 billion (CEF 2019). 14  Authors’ calculations (petrol, diesel, jet fuel) from data in Sapia (2019). 15  Puma’s main shareholder is a large international trading company, Trafigura.

Energy in South Africa    291 supplied market growth. The corollary is that they do not believe that investment in additional refinery capacity is likely. In summary, it appears that the White Paper’s decisive break with the past in advocating a more market-​orientated approach has been partly responsible for its own undoing. The prominence given to the simultaneous pursuit of social policy objectives: job preservation, small business, and BEE, has unleashed forces that have in a sense trapped the minister-​as-​regulator into a relationship that has made policy implementation difficult. The failure to carry through the reform from minister-​as-​ regulator to independent regulator has compounded the problem. Nevertheless, market developments in the forms of increasing imports (reducing the relative importance of protecting refinery capacity), Glencore’s market entry (a global trader perhaps willing to contest for market share), and the entry by independent retailer Puma Energy (also seeking market share) means that the best prospects for a competitive market since 1994 exist, in the event of deregulation.

14.6  Natural Gas Markets South Africa has never had a fully fledged natural gas industry. The limited gas resources found off-​shore of Mossel Bay were used to develop an apartheid-​era oil sanctions-​ busting gas-​to-​liquids facility, now operated by PetroSA. Its gas fields are expected to run dry in 2020. The then state-​owned Sasol provided limited coal-​gas (methane) from 1966 (Meintjes 1975: 92). In 2004, the gas industry experienced a step change when Sasol supplied natural gas from Mozambique to the industrial heartland for the first time. This came about as a product of a remarkable balancing act between market forces and state intervention achieved through a public–​private partnership, one of the few in the energy sector in the democratic era. The gas was transported from Mozambique by the 865km-​long Rompco pipeline. This project was a good example of both cross-​border infrastructure development and of a multi-​country public–​private partnership involving Sasol and SOCs from South Africa and Mozambique. In South Africa, a combination of commercial and regulatory agreements enabled the project, giving Sasol a sheltered monopoly for ten years. Legislation to regulate the new industry followed these agreements, one of which was appended to the Gas Act (No. 48 of 2001). The compromises made in these agreements reflected the more market-​friendly approach in the 1998 White Paper which was carried through to the light-​handed regulation envisaged by the subsequent Gas Act (No. 48 of 2001). Unfortunately, the impasses over the MPRDA (see above) have inhibited the search for new gas fields, leaving no ready SA replacement for the Mozambique gas fields which will begin depleting in 2023 (Engineering News 2020) ending the approximately 150PJ/​a supplied to Sasol and other industrial customers.

292    Rod Crompton and Ruwadzano Matsika

14.7 Electricity Markets State intervention in electricity markets also has a long history beginning with the establishment of the state-​owned Eskom in 1922. Eskom was corporatized in 2002 following the market-​orientated reforms envisaged in the White Paper (1998). These included customer choice of supplier, competition in the generation market, open access to the transmission network, the unbundling of Eskom, and encouraging private participation (RSA 1998). At the time, state ownership completely dominated the electricity supply industry (ESI) and has only marginally reduced since then. By 2020, Eskom still dominated the ESI, accounting for approximately 90 per cent of generation capacity, 100 per cent of transmission, and approximately half of distribution. The other half of distribution is owned by municipalities in accordance with schedule 4B of the Constitution.16 Despite its commercialization, Eskom has retained a dual commercial and developmental mandate which has bedevilled its governance and performance. The White Paper identified the distribution industry as a major problem and so government set about rationalizing it. It established Electricity Distribution Industry Holdings (EDI) in 2003 as a receptacle for the country’s distribution assets and began parcelling them out to six regional electricity distributors (REDs) as a precursor to competition in the distribution market. Despite good progress and expenditure of hundreds of millions of Rands, REDs were abandoned in 2010. They were opposed by some municipalities that feared losing their electricity surcharges, an important revenue stream. An attempt to amend the constitution to remove this obstacle fizzled out in 2010, effectively ending distribution market reform efforts. In 2002, government intended to privatize 30 per cent of Eskom but this did not happen as it ran into opposition from organized labour and the ANC. After its 2004 electoral victory, the ANC reversed its decision on partial privatization of Eskom and changed its focus from market reform to security of supply, limiting private ownership of generation capacity to 30 per cent (van der Heijden 2013). Nevertheless, some progress was made in other areas towards market-​orientated reform. An independent regulator, NERSA, was established in 2005 and in 2011 the Independent Power Producers Procurement Office (IPPPO) was established to conduct power generation capacity auctions. Its auction processes were well received by the market. In seven bidding rounds, it procured 6422MW of capacity and attracted private sector investment of R209.7 billion of which 20 per cent was foreign investment (Watermeyer and Phillips 2020). Eskom signed twenty-​year power purchase agreements (PPAs) with the winning independent power producers (IPPs) on behalf of government. However, even this limited movement towards a market has not escaped state interventions, some deliberate and others possibly driven by malfeasance. 16 Schedule

4B gives municipalities jurisdiction over the ‘reticulation’ of electricity and gas. The boundary between reticulation and distribution has not been officially established.

Energy in South Africa    293 Deliberate political intervention occurs through Integrated Resource Plans (IRPs) in which ministers of energy promulgate the country’s future capacity needs.17 This empowers the minister to pick higher-​cost technologies rather than abide by the lowest cost IRP methodology, evident in coal and nuclear selections. Unfortunately, Section six of the National Energy Act (No. 34 of 2008) requiring annual integrated energy plans has not been promulgated and IRPs have been irregular. The first IRP was published in 2011 and was not updated during President Zuma’s administration (2009–​18) for political reasons. It was finally updated in 2019 (RSA 2019b). Interventions possibly driven by malfeasance occurred in 2015 when the then CEO of Eskom refused to sign any more PPAs.18 This impasse prevailed until he left office and the minister of energy intervened in 2018. The delays in bringing new capacity online contributed to subsequent load-​shedding and illustrated Eskom’s vulnerability to corruption or ‘state capture’ and, consequently, South Africa’s vulnerability to Eskom’s fortunes. Despite these demonstrated risks government nevertheless clung to centralized state ownership, evident in its reluctance to unbundle Eskom into several independent entities in its Eskom Roadmap (Department of Public Enterprises 2019), an attenuated version of the reforms in the White Paper published twenty-​one years previously. The costs of reliance on centralized state ownership have been evident for a long time. Between 1981 and 1990, Eskom added a massive 17,300MW of new capacity. This turned out to have been poor planning resulting in excess capacity and over 10,000MW being mothballed over the following decade. Between 2001 and 2010 only 4,100MW of new capacity was added. This period of underinvestment and poor planning contributed to load-​shedding from 2008, in all a forty-​year period of poor supply/​demand alignment that has cost the country dearly. The cost to society of reliance on centralized state dominance of the electricity market is evident in the scale of load-​shedding. Between 2007 and January 2020 South Africa experienced the loss of 1,710 hours, comprising 3,867GWh with an estimated economic impact of between Rand 169 billion and Rand 338 billion (calculated from Wright and Calitz 2020). Load-​shedding and its economic consequences were partly attributable to non-​ implementation of the White Paper’s market reforms. Government prevaricated about the decision to build or procure new generation capacity until after the 2004 elections when it decided to abandon market options in favour of Eskom building additional capacity. By this time it was too late to avoid load-​shedding. With the benefit of hindsight, Eskom exacerbated matters by ambitiously opting to build two of the largest coal-​fired stations in the world, Medupi and Kusile (dubbed the ‘new build’). Eskom’s technology choice could hardly have been more ill timed, occurring just as the cost curves for renewable power generation began their steep descent. This ‘new build’, 9,600MW of 17  IRPs become ministerial directives issued in terms of Section 34 of the Electricity Regulation Act 2006 (Act No. 4 of 2006). 18  Brian Molefe and other ex-​Eskom officials were accused of being part of the state-​capture force and sued by Eskom in 2020.

294    Rod Crompton and Ruwadzano Matsika capacity, has a fifty-​year useful life that leaves South Africa with financial and environmental legacies that may be difficult to manage. Government’s decision to use a state instrument (Eskom) left it without the protection inherent in the IPP programme in which private investors carried the project risk. This proved to be a huge error. Eskom was ill prepared for such mega-​projects. The rush to start the projects led to poor planning, poor contracting strategy, poor execution and oversight, large cost and time overruns, skills shortages, poor delivery by contractors, and design faults. Medupi’s cost escalated seven f​ old between 2005 and 2020 and Kusile’s three fold between 2007 and 2020.19 Eskom attributes delays to eighteen months of strike action and eight months of welding problems (Eskom 2020:56). Political meddling and corruption exacerbated matters. Eskom has had eleven CEOs since construction commenced. Medupi and Kusile are ‘textbook studies on how not to execute large infrastructure projects’ (Watermeyer and Phillips 2020: 2) and caused ‘severe damage to the South African economy and to economic growth’ (Watermeyer and Phillips 2020: iv). Financial difficulties inevitably followed. In March 2014 Eskom’s external auditors referred to its Going Concern status for the first time. By 2017, Goldman Sachs Group Inc said: ‘Eskom is the biggest single risk to the South African economy’ (BusinessTech 2017). By 2020 Eskom’s debt reached an unsustainable Rand 488 billion (Eskom 2020). Government has been forced to bail out Eskom as it is ‘to (sic) vital to our economy to be allowed to fail’ (Ramaphosa 2019: 6). Eskom bail outs cost taxpayers Rand 188 billion over five years to 2020 with more scheduled. In addition, government has guaranteed Rand 350 billion of Eskom’s debt, negatively impacting sovereign credit ratings (Eskom 2019: 35) and simultaneously posing credit default risks, loan covenants and cross-​defaults risks that would cause government guarantees to be called up, something government is unlikely to be able to meet. Eskom is at risk of breaching its loan covenants on several fronts, the most critical of which is its Going Concern status which it is only able to meet with promises of government bail ​outs. Within these domino-​like financial arrangements lies South Africa’s key risk. The cumulative result of the ‘new build’ market intervention was that Eskom became an impediment to economic growth rather than an enabling instrument for the developmental state as intended. It also reduced a once-​efficient utility to a giant subsidy instrument, transferring taxpayers’ funds (or government debt) to electricity customers and increasing real costs in the economy by reducing allocative efficiency (Steyn 2003), the opposite of the White Paper’s envisaged self-​sustaining ESI, with security of supply. During this era of Eskom’s decline, the twin transitions have rendered its vertically integrated business model obsolete. It has entered a classic ‘utility death spiral’ of rising prices leading to falling volumes, leading to rising prices etc. Technological innovation in renewable power generation resulting in lower costs offers customers an alternative and, despite regulatory obstacles, a limited de facto market has emerged. Eskom’s

19 

Author’s calculations based on Eskom (2020: 66–​7).

Energy in South Africa    295 sales volumes declined by 4.7 per cent20 between 2009 and 2019. Government reluctantly recognized this development in its Eskom Roadmap but still refused to relinquish Eskom dominance of the market and importantly retained the transmission and market operator function within Eskom rather than establishing an independent one (more below). Eskom’s ‘death spiral’ difficulties have been exacerbated by corruption, poor management, and rising coal costs. Operating costs increased by 30 per cent over five years to 2020, the staff were ‘demoralized’ according to its CEO (de Ruyter 2020: 3), and its energy availability factor (EAF)21 declined from 94 per cent in 2000 to 67 per cent in 2019 (Wright and Calitz 2020). A low EAF is a key contributor to load-​shedding. Also, the president has prohibited Eskom from retrenching any of its well-​paid and bloated workforce (Phakathi 2019). To offset declining sales and performance as well as fund the ‘new build’, Eskom’s average standard tariffs increased by a real compound annual growth rate of 5.9 per cent between 2010 and 201922 resulting in restructuring for industries that had enjoyed low tariffs in the pre ‘new build’ era. Not all of the White Paper’s policies were market-​orientated, some were pro-​poor, in particular electrification. A large programme electrified over 7.4 million households between 1994 and March 2018,23 resulting in 91 per cent electrification, in contrast to sub-​Saharan Africa’s 48 per cent (World Bank 2020). Also, free basic electricity (50kWh/​ month) has been provided to about 900,000 households (RSA 2020).

14.8  Energy Transitions and Future Tranches of Capacity Globally, the energy sector is undergoing two interlinked and simultaneous transformations. The first concerns the rapid technological change associated with the Fourth Industrial Revolution bringing about the digitization of the energy sector, introducing modern technologies such as cloud computing, big data, Internet of Things, amongst others and enabling smaller-​scale, decentralized business models in energy extraction/​ generation/​ production, transmission, distribution, retail, and demand

20 

Author’s calculations from data in Eskom Annual Integrated Reports, various years. proportion of the generating fleet that is available for operation. It is partly a function of maintenance. 22 Author’s calculations based on data from:  http://​www.energy.gov.za/​files/​esources/​petroleum/​ petroleum_​arch.html and https://​www.eskom.co.za/​CustomerCare/​TariffsAndCharges/​Pages/​Tariff_​ History.aspx. 23   http://​ w ww.energy.gov.za/ ​ f iles/ ​ I NEP/ ​ i nep_ ​ o verview.html#:~:text=The%20National%20 Energy%20Regulator%20of,INEP)%20from%201999%20to%202001.&text=The%20main%20 funding%20of%20the,the%20National%20Budget%20(Fiscus). 21  The

296    Rod Crompton and Ruwadzano Matsika management (IEA 2018). The ‘Fourth Industrial Revolution technologies, and digitalization, in particular, allow for open, real-​time, automated communication and operation of a more efficient energy system’ (IEA 2018: 8). The second transformation involves the move to lower carbon intensity driven by international efforts, such as the Paris Agreement on Climate Change intended to limit global warming by reducing greenhouse gas emissions to achieve net-​zero emissions between 2050 and 2070. The decarbonization of the energy sector requires a move away from fossil fuels to cleaner and renewable energy sources, cleaner industrial production technologies, cleaner technologies for vehicle propulsion, and increasing energy system efficiency, and needs support by strong policy interventions (Rissman et al. 2020). The combined effect of these twin dynamics is a global shift towards electricity as a preferred energy carrier driven by low-​carbon and renewable energy power generation (IEA 2019). Global electricity demand is expected to increase from about 20 per cent of energy consumption in 2018 to just over half by 2050 (BP 2019), exceeding the consumption of crude oil by 2040 (IEA 2019). This will be partly driven by greater uptake of technologies for industrial production, domestic consumption, cooling, and electric vehicles (IEA 2020b). This shift is expected to encompass smart grids, demand-​side management, increased interlinkages and interdependence between energy systems resulting in more integrated energy systems. Digitalization is simultaneously enabling a third transition in the structure of energy markets to decentralized models, particularly in the electricity sector (IEA 2019). Energy consumers are increasingly becoming ‘prosumers’ with the ability to produce electricity and sell the surplus to other customers (IEA 2020b). These transitions are transforming the economy itself, especially the energy, transport, industry, and building sectors, and changing consumer behaviour (IEA 2020). South Africa’s energy-​intensive and coal-​reliant economy was poorly placed to address these transitions. The World Economic Forum has ranked South Africa 106 of 115 countries in its trilemma index which measures transition readiness based on assessments of energy security, energy equity, and environmental sustainability (World Economic Forum 2020). South Africa faces several obstacles. Path dependence on a coal-​based energy economy, technology lock-​in due to historical infrastructure investments, and vested interests in the MEC-​based political economy are retarding transition to sustainable energy policies (Morris and Martin 2015). Capital requirements may be large and if the state wishes to maintain its large presence, it is poorly placed to do so given its fiscal constraints. Alternatively, it must embrace a more market-​orientated approach which it has been reluctant to do. Government may receive some assistance from ‘Green Finance’ and has begun coordinating funding across government departments towards lower carbon intensity (National Treasury 2020a). SA financial markets have followed international trends of declining interest in new coal and hydrocarbon projects, which may delay stall South Africa’s planned coal-​fired generation capacity. South Africa’s ability to respond to international decarbonization pressures with emissions taxes is constrained by the preponderance of hydrocarbons in South Africa’s energy mix and the price-​raising impact of such interventions. The

Energy in South Africa    297 gradualist approach adopted had not materially changed the energy mix by 2020 partly because government has been preoccupied with security of supply since 2004 (van der Heijden 2013; Morris and Martin 2015). The scale of change required by these transitions may involve considerable social and economic dislocation. Trade unions have proffered the concept of a ‘Just Transition’ that may offer a fulcrum around which South Africa can ‘muddle through’ these transitions. Government has ‘recognised the need for a just transition’ (National Treasury 2020a:3) even if it is not yet clear what that means.

14.9  Future Tranches of Capacity The long-​running tensions between market-​oriented reform and state intervention will be made more complex by the transitions described above, partly because technological changes are bringing about more market-​orientated solutions whereas decarbonization requires market interventions. This complex interplay is beginning to shape answers to the question: ‘where will the next tranche of capacity come from?’ Technological developments are making it easier to progress towards the White Paper’s objective of ‘security through diversity of supply’ and may improve security of supply, as is argued below.

14.9.1 The Next Tranche of Liquid Fuels The White Paper set as a policy cornerstone ‘a coastal refining and petrochemicals hub for future investments’ (RSA 1998: 66). Subsequently, government’s National Planning Commission (NPC) has differed, preferring continued importation of liquid fuels as more cost-​effective (RSA 2019a:  xii). However, oil refinery investors will nervously watch the declining cost of batteries,24 electric vehicles (EVs), and other propulsion technologies, perhaps fearing stranded assets. Both battery and hydrogen propulsion offer South Africa a tantalizing alternative to continued reliance on petroleum. A switch to EVs at scale could reduce reliance on South Africa’s largest import (petroleum), reduce balance-​of-​payments pressures, improve security of supply, and allow South Africa’s comparative advantages (coal, wind, and solar-​powered electricity, battery component minerals) to work in tandem with ISI, unlike in the past. It would also de-​couple transport costs from oil prices. This switch could happen faster if incentives for the local manufacture of internal combustion engines (ICE) were extended to include EVs.

24 

Lithium ion battery costs for EVs fell 87 per cent between 2010 and 2019 and are expected to reach close to parity with internal combustion engines by 2023 (Henze 2020).

298    Rod Crompton and Ruwadzano Matsika These developments strengthen the NPC’s wait-​and-​see stance. Also as ICE vehicles lose market share to other propulsion technologies the rationale for continued regulation of petrol prices may erode even further.

14.9.2 The Next Tranche of Natural Gas Attention is focused on gas because it is a cleaner fuel than other hydrocarbons and often touted as the bridging fuel to more sustainable energy. It is also a necessary adjunct to intermittent renewable power. South Africa has several possibilities in the medium-​to-​long term including hydraulic fracturing of the Karoo shales, Total’s 2019/​20 gas and condensate finds south of Mossel Bay, and the gas fields in Northern Mozambique. There are considerable barriers to exploiting these resources including opposition from environmentalists and proving commercial viability for a gas pipeline from Northern Mozambique. Several years of further work are still required before any of these may become a commercial proposition. The IPP office has been considering the importation of LNG for several years at Saldanha Bay, Coega, or Richards Bay. Total and Gigajoule are developing an LNG terminal in Maputo designed to replace Sasol’s declining gas fields and scheduled for 2023 (Engineering News 2020). This will threaten Sasol’s SA monopoly and introduce competition in natural gas markets as will Sasol’s 2020 announcement of the sale of its share in that pipeline, heralding a new dynamic in the SA gas market. Small-​scale LNG imports licensed in 2020 may also increase competition. LPG markets are an important precursor to natural gas markets (WLPGA 2020). Recent significant investment in LPG import terminals in Saldanha Bay and Richards Bay, as well as a long overdue revision of the LPG price regulation methodology,25 augurs well for LPG and gas markets.

14.9.3 The Next Tranche of Power-​generating Capacity Between 2020 and 2023 South Africa will probably be short of power-​generating capacity. The IRP 2019 predicted a 3,000MW shortfall between 2019 and 2022 assuming Eskom’s EAF would be 71 per cent. Wright and Calitz (2020) predict 2,000GWh in 2020, 3,000GWh in 2021 and 4,600GWh in 2023. Eskom forecasts a 4,000MW shortfall in 2021 assuming EAF at 70 per cent (Paton 2020). The actual shortfall will depend upon Eskom’s EAF, how quickly government responds and customers switch suppliers. In the short term, government has four parallel processes underway to improve supply:

25  Media

June 2020.

Statement CEF Ltd. on Behalf of the Department of Mineral Resources and Energy, 26

Energy in South Africa    299 1. Renewable Energy Independent Power Producer Procurement Programme (REIPPP) bid window four includes twenty-​seven delayed projects with a capacity of 2200MW scheduled to commence operations in 2020. 2. DMRE’s Risk Mitigation Independent Power Producer Procurement (RMIPPP) emergency programme for 2000MW commenced in January 2020 followed by a Request for Proposals on 24 August 2020, indicative of the slow reaction time. Power must be dispatchable and delivered by June 2022 requiring renewables to have storage. A twenty-​year PPA is offered. One option is expensive floating power stations. 3. REIPPP bid window five of 11813MW for renewables, storage, gas, and coal is expected to open late in 2020. Procurement and construction will take eighteen to forty-​two months. 4. The quickest additions appear to be small units of embedded generation exploiting section 12B of the Income Tax Act allowing accelerated depreciation over one year for below 1MW generators. Spare capacity from IPPs and embedded generation could also be tapped but is constrained by the lack of a functioning market. South Africa’s wind and solar resource endowment place it in a favourable position to transition to a low-​carbon, renewables-​based electricity system at least cost (Wright et al. 2020) and one that can be deployed rapidly. Solar photovoltaic (PV) and wind-​ power costs have declined, making them cost-​competitive with coal-​power generation. Between 2010 and 2019 the global weighted-​average LCOE of onshore wind fell by 39 per cent, concentrated solar power by 47 per cent, and PV by an extraordinary 82 per cent (IRENA 2020). Quick ramp-​up gas and storage are needed to complement intermittent renewables. Looking beyond the immediate power capacity crisis, South Africa must consider the 3,670MW26 of Eskom’s capacity that will reach the end of its useful life by 2023 and thereafter other ageing capacity that requires replacement. Economic growth will require even more. Together the ESI faces a formidable investment hurdle that neither Eskom nor government are equipped to handle. South Africa’s ‘current debt trajectory is not sustainable’ (National Treasury 2020b: 86). In addition to Eskom’s financial troubles, the REIPPP programme has cost government over Rand 200 billion in twenty-​year PPA guarantees and contingent liabilities (National Treasury 2020b: 86). Using the fiscus to fund Eskom and de-​risk IPPs is not a desirable nor sustainable way in which to fund generation capacity. Given the dreadful performance by SOCs in energy sector mega-​projects over the past fifteen years (Watermeyer and Phillips 2020; Gregory 2020)  they would likely be poor implementation instruments. These considerations suggest that security of supply is going to be intimately tied to market reform and investor trust in the market structures developed.

26 

Grootvlei 1,180 + Komatie 990 + Kriel 1,500 (half) = 3,670MW.

300    Rod Crompton and Ruwadzano Matsika There are small signs that government is once again considering market-​orientated solutions. A  1MW limit on commercial and industrial generators for own use was removed and municipalities in good financial standing were allowed to procure their power from IPPs (Ramaphosa 2020). Nevertheless, security of supply may require going beyond those limited measures, beyond the Roadmap, and further towards those envisaged in the White Paper. As a minimal first step, investors are likely to require a standalone transmission and systems market operator rather than the Eskom-​owned proposition in the Roadmap. A market may require arrangements to be made for legacy generators (IPPs and Eskom stations) that will not survive market prices and become stranded assets for many years to come, as has been seen in other countries. Emerging technologies are dramatically lowering the economies of scale for energy conversion and completely undermining the old economies-​of-​scale justification for state ownership. They range from matchbox size to massive and enable market solutions. The advent of ‘prosumers’ can literally shift ‘power to the people’. Capacity additions in ‘small steps’ remove mega-​project risk and reduce the risk of large stranded assets in times of rapid technological change. This will allow the financing burden to gradually shift from the state to private investors and produce promising market-​related prices. Regardless of South Africa’s legacy problems, a declining generation-​cost curve also holds the promise of a way out of South Africa’s seemingly intractable electricity supply problems. Ironically, it exacerbates the distribution market crisis as customers leave the grid and municipalities lose revenue. Emerging generation technologies also offer local industrialization opportunities and continuous construction/​installation jobs for a longer time (Bishoff-​Niemz and Creamer 2019). Although government’s local content prescripts carry a price burden, the early rounds of the REIPPP did enable fledgling manufacturers to emerge. With time and scale, they may become competitive and the price premium reduced or eliminated, provided that no further Eskom-​type interventions result in further serious setbacks to local manufacturing. Furthermore, South Africa has minerals that feed into the battery and renewables value chains that offer promise (SANEA 2020). Emerging technologies also require investment in grid strengthening as the locus of generation shifts from coal in the east to renewables in the west, estimated at Rand 100 billion by 2030.

14.9.4 Additional Capacity from Energy-​efficiency Measures Demand-​side energy-​efficiency measures are ‘the most cost-​effective means of driving sustainable energy change’ (Kuzemko et  al. 2019:  6) and reduce the need for additional generation capacity. South Africa has implemented a suite of energy-​ efficiency programmes, including tax incentives (Section 12L) which, between 2013 and 2019, delivered 93TWh in energy savings, equivalent to 90Mt of CO2 (SANEDI 2020a). New technologies like digitalization can provide real-​time linkage between

Energy in South Africa    301 supply and demand and enable smart metering and smart grids, which tap into existing spare capacity that residential and industrial customers already have (SANEDI 2020b). Already smart technologies can be programmed to save energy, such as lights that only operate when movement is detected, the use of the Internet of Things, and smartphone apps to implement energy-​saving actions (SANEDI 2020b). Control systems can be used to shut down machinery, appliances, and plugs when not in use to save energy.

14.10 Conclusion In 2020 South Africa’s energy sector was not well positioned to transition from a heavy reliance on coal to lower energy intensity and decarbonization because of the economy’s resource-​based MEC nature and technological lock-​in. South Africa’s long history of state intervention in energy markets encountered a substantial shift in policy thinking in the 1998 White Paper’s market-​orientated goals. After a promising start, practice moved away from those goals. It was only in 2018 that there were indications of movement back towards those goals, albeit in attenuated form. These moves were driven by higher liquid fuel prices, electricity shortages, fiscal constraints, and the near collapse of Eskom and other energy SOCs. Those intended instruments of the developmental state have had the opposite result. Eskom dominated the electricity landscape and became the single largest risk to the economy. In contrast, the larger, mainly privately owned, petroleum sector performed better. Notable successes in policy implementation have occurred in electrification, electricity subsidies for the poor, gas infrastructure, and renewable power-​capacity auction processes. However, private investment in the state-​dominated electricity generation market has come at a price for government: twenty-​year government guarantees. Progress towards market-​orientated policy goals in petroleum markets reached an impasse, due to a combination of factors: the old-​style minister-​as-​regulator institutional model and social policies entangled in petrol price regulation resulting in above-​ market prices. Emerging technologies such as EVs and hydrogen propulsion may in future unlock this costly impasse. In upstream markets, recent discoveries of off-​shore gas also hold the prospect of more, if the long-​running impasse over the regulatory dispensation can be resolved in an investor-​friendly manner. Technological developments in renewable power generation hold out the promise to exploit natural comparative advantages, substitute imports, decouple transport propulsion costs from the oil price, mitigate the unfolding power crisis, allow development in lower risk ‘small steps’, shift the burden of capitalizing power generation from the state to the private sector, and reinvigorate domestic manufacturing. To realize such promise, the conspectus of evidence suggests that a more market-​oriented policy approach will be necessary, especially given the weakened state apparatus and its vulnerability to corruption. Harnessing wind and solar resources could boost the economy. However,

302    Rod Crompton and Ruwadzano Matsika such a transition will have to contend with path dependence and manage the socio-​ economic dislocations associated with it.

References Bekun, Victor, Emira Firat, and Samuel Sarkodie. 2019. ‘Another look at the relationship between energy consumption, carbon dioxide emissions, and economic growth in South Africa’, Science of The Total Environment, 655: 759–​65. Bishoff-​ Niemz, Tobias, and Terrence Creamer. 2019. South Africa’s Energy Transition. Oxford: Routledge. BP. 2019. ‘BP Energy Outlook’, 2019 edition. https://​www.bp.com/​content/​dam/​bp/​business-​ sites/​en/​global/​corporate/​pdfs/​energy-​economics/​energy-​outlook/​bp-​energy-​outlook-​ 2019.pdf. Accessed 19 September 2020. BP. 2020. ‘Statistical review of world energy’, 69th edition. https://​www.bp.com/​content/​dam/​ bp/​business-​sites/​en/​global/​corporate/​pdfs/​energy-​economics/​statistical-​review/​bp-​stats-​ review-​2020-​full-​report.pdf. Accessed 19 September 2020. BusinessTech. 2017. ‘Eskom is currently the single biggest risk to South Africa’s economy:  Goldman Sachs’, 22 September [online] https://​businesstech.co.za/​news/​ business/​200478/​eskom-​is-​currently-​the-​single-​biggest-​risk-​to-​south-​africas-​economy-​ goldman-​sachs/​. CEF. 2019. Integrated Annual Report 2018/​19. Johannesburg. Crompton, Rod, Midesh Sing, Vernon Filter, and Nonhlanhla Msimango. 2020. ‘Petrol price regulation in South Africa: Is it meeting its intended objectives?.’ WIDER Working Paper 2020/​140. Helsinki: UNU-​WIDER. [online] https://​doi.org/​10.35188/​UNU-​WIDER/​2020/​ 897-​9. Davidson, Ogunlade, and Harald Winkler. 2003. ‘South Africa’s energy future:  Visions, driving factors and sustainable development indicators’. Energy and Development Research Centre, UCT. Department of Public Enterprises. 2019. ‘Roadmap for Eskom in a reformed electricity supply industry’. Pretoria [online] https://​dpe.gov.za/​wp-​content/​uploads/​2019/​10/​ROADMAP-​ FOR-​ESKOM_​0015_​29102019_​FINAL1.pdf. de Ruyter, Andre. 2020. ‘ESKOM: Powering the country through extraordinary times’. Group Chief Executive. SAN SANEA Webcast, 25 June. DMRE. 2020. ‘Energy balances 2017’. DMRE, Pretoria [online] http://​www.energy.gov.za/​files/​ media/​Energy_​Balances.html. Engineering News. 2020. ‘Foot on the gas—​developers of Matola LNG terminal intensify talks with potential customers to reach 2023 supply date’. https://​www.engineeringnews.co.za/​ article/​foot-​on-​the-​gas-​developers-​of-​matola-​lng-​terminal-​intensify-​talks-​with-​potential-​ customers-​ to- ​ reach- ​ 2 023- ​ supply- ​ d ate- ​ 2 020-​ 0 8-​ 1 1#:~:text=The%20project%20is%20 being%20developed,to%20begin%20operating%20in%202023. Eskom. 2019. ‘Integrated report 2019’. Eskom Holdings SOC Ltd, Johannesburg. Eskom 2020. ‘Eskom submission to the Parliamentary Standing Committee on Appropriations’, 2 September, Eskom Holdings SOC Ltd, Johannesburg. Fine, Ben, and Zavareh Rustomjee. 1996. The Political Economy of South Africa, from Minerals-​ Energy Complex to Industrialisation. New  York:  Routledge. https://​doi.org/​10.4324/​ 9780429496004.

Energy in South Africa    303 Gregory, Julian. 2020. ‘Governance, scale, scope:  A review of six South African electricity generation infrastructure megaprojects’, Utilities Policy 66. https://​doi.org/​10.1016/​ j.jup.2020.101103. Henze, Veronika, 2020. ‘Battery pack prices fall as market ramps up with market average at $156/​kWh in 2019’, Bloomberg, 3 December [online] https://​about.bnef.com/​blog/​battery-​ pack-​prices-​fall-​as-​market-​ramps-​up-​with-​market-​average-​at-​156-​kwh-​in-​2019/​. International Energy Agency. 2018. ‘Fostering effective energy transition: A fact-​based framework to support decision-​making’. IEA, Geneva [online] https://​www.mckinsey.com/​~/​ media/ ​ mckinsey/​ i ndustries/​ oil%20and%20gas/​ our%20insights/​ a %20framework%20 for%20fostering%20effective%20energy%20transitions/ ​ f ostering- ​ e ffective- ​ e nergy-​ transition.pdf. International Energy Agency. 2019. ‘World energy outlook 2019 executive summary’. [online]https://​www.iea.org/​reports/​world-​energy-​outlook-​2019. International Energy Agency. 2020a. ‘World energy balances’. [online] https://​www.iea.org/​ subscribe-​to-​data-​services/​world-​energy-​balances-​and-​statistics. International Energy Agency. 2020b. ‘Energy technology perspectives 2020’. IEA, Geneva [online] https://​webstore.iea.org/​download/​direct/​4165. IRENA. 2020. ‘Renewable power generation costs in 2019’. International Renewable Energy Agency, Abu Dhabi. Jamison, Mark, Sanford Berg, Farid Gasmi, and Jose Távara. 2004. ‘Annotated reading list for a body of knowledge on the regulation of utility infrastructure and services’. World Bank Group [online] http://​documents1.worldbank.org/​curated/​pt/​397981468319747524/​text/​ 375570Infrastr1eading0list01PUBLIC1.txt. Kuzemko, Caroline, Andrew Lawrence, and Matthew Watson. 2019. ‘New directions in the international political economy of energy’, Review of International Political Economy, 26: 1–​24. Majola, Given. 2020. ‘Independent fuel retailers say loyalty cards are pushing them out of business’, Business Report, 26 January [online] https://​www.iol.co.za/​business-​report/​energy/​ independent-​fuel-​retailers-​say-​loyalty-​cards-​are-​pushing-​them-​out-​of-​business-​41265816. Maqubela, Tseliso. 2019. Opposing Affidavit in the matter between Auto Commodities (Pty) Ltd. (applicant) and The Controller of Petroleum Products (First Respondent), the Minister of Energy (Second Respondent) and the National Energy Regulator of SA (Third respondent), Case No. 32733/​2019, in the High Court of South Africa, Gauteng Division. Marquard, Andrew. 2006. ‘The origins and development of South African energy policy’, PhD thesis, University of Cape Town. Meintjes, Johannes. 1975. SASOL 1950–​1975. Cape Town: Tafelberg-​Uitgewers Beperk. Morris, Mike, and Lucy Martin. 2015. ‘Political economy of climate-​relevant policies:  The case of renewable energy in South Africa’, Cape Town, PRISM, Economics, University of Cape Town. National Treasury. 2020a. ‘Financing a sustainable economy’. Technical Paper, Pretoria [online] http://​www.treasury.gov.za/​publications/​other/​Sustainability%20technical%20paper%20 2020.pdf. National Treasury. 2020b. ‘Budget review 2020’. National Treasury, Pretoria, 26 February. OECD 2014. The Governance of Regulators: OECD Best Practice Principles for Regulatory Policy. Paris: OECD Publishing. Paton, Carol. 2020. ‘The electric light at the end of the tunnel is years away’, Business Day, 14 September.

304    Rod Crompton and Ruwadzano Matsika Phakathi, Bekezela. 2019. ‘Eskom jobs will not be cut, says Ramaphosa’, Business Day, 15 February. Ramaphosa, Cyril. 2019. ‘State of the nation address by President Cyril Ramaphosa’. Parliament, Cape Town, 20 June. Ramaphosa, Cyril. 2020. ‘State of the nation address by President Cyril Ramaphosa’. Parliament, Cape Town, 13 February. Republic of South Africa. 1998. White Paper on the Energy Policy of the Republic of South Africa. Department of Minerals and Energy, Pretoria. Republic of South Africa. 2012. National Development Plan 2030:  Our Future—​ Make It Work. National Planning Commission, Pretoria. https://​www.gov.za/​issues/​ national-​development-​plan-​2030. Republic of South Africa. 2019a. ‘NPC economy series:  Energy’. The Presidency, National Planning Commission Department, Pretoria. Republic of South Africa. 2019b. ‘Integrated Resource Plan (IRP 2019)’, Government Gazette, 18 October, No. 42778, Department of Energy. Republic of South Africa. 2020. ‘Eskom embarks on free basic electricity programme campaign’. South African Government News Agency, Pretoria. https://​www.sanews.gov.za/​ south-​africa/​eskom-​embarks-​free-​basic-​electricity-​programme-​campaign. Rissman, Jeffrey, Chris Bataille, Eric Masanet, Nate Aden, William Morrow III, Nan Zhou, and Helseth Jonas. 2020. ‘Technologies and policies to decarbonize global industry: Review and assessment of mitigation drivers through 2070’, Applied Energy, 266: 114848. S&P Global. 2020. ‘Total makes second “significant” gas condensate find off South Africa’. [online] https://​www.spglobal.com/​platts/​en/​market-​insights/​podcasts/​oil/​ 102220-​crude-​throughput-​hindered-​reduced-​refinery-​runs-​covid. SANEA. 2020. ‘South African energy risk report 2020’. SANEA, Johannesburg. SANEDI. 2020a. ‘Section 12l energy efficiency tax incentive status report 2013–​2019’. SANEDI, Johannesburg. SANEDI. 2020b. ‘Insights 2019/​2020’. SANEDI, Johannesburg [online] https://​www.sanedi. org.za/​img/​Events/​Sanedi%20Insights2020FINALNW4singles.pdf. Slabbert, Antoinette. 2018. ‘Capping fuel price could sink industry transformation’, The Citizen. 15 November [online] https://​citizen.co.za/​business/​2036641/​capping-​fuel-​price-​could-​ sink-​industry-​transformation/​. South African Petroleum Industry Association 2019. ‘Annual Report 2018’. SAPIA, Johannesburg. Steyn, Grové. 2003. ‘Administered prices: Electricity: A report for National Treasury’. [online] http://​www.treasury.gov.za/​publications/​other/​epir/​Electricity.pdf. van der Heijden, Tracy. 2013. ‘Why the lights went out: Reform in the South African energy sector’. UCT Graduate School of Development Policy and Practice, Cape Town. Watermeyer, Ron, and Sean Phillips. 2020. ‘Public infrastructure delivery and construction sector dynamism in the South African economy’. Background Paper, National Planning Commission, Pretoria. WLPGA, 2020. ‘Charter of Benefits’. https://​www.lpgas.co.za/​pdfs/​benefits-​of-​lpg/​. World Bank. 2020. ‘World Development Indicators’. https://​data.worldbank.org. World Economic Forum. 2020. Fostering Effective Energy Transition. Cologny/​Geneva: WEF. World Energy Council. 2019. World Energy Trilemma Index 2019. London:  World Energy Council. Wright, Jarrad, and Joanne Calitz. 2020. ‘Setting up for the 2020s: Addressing South Africa’s electricity crisis and getting ready for the next decade’. CSIR, Pretoria. https://​cisp.cachefly. net/​assets/​articles/​attachments/​81125_​rs_​setting_​up_​for_​2020.pdf.

Chapter 15

So ci o-​e c onomi c Aspe c ts of Energy an d C l i mat e Change in Sou t h   A fri c a Roula Inglesi-​L otz

15.1 Introduction In the literature, economists had focused only on patterns of economic growth and substitutability between factors of production, such as labour and natural capital, until recently when W. D. Nordhaus and P. Romer proposed the integration of climate change into long-​term macroeconomic analysis (Halsnæs and Garg 2011). The development of any economy and the improvement of human development has always been closely connected with the use of energy (Essex and de Groot 2019) as well as land use and has resulted in rising emission levels and subsequent changes in climate conditions that also have an impact on the socio-​economic activities of the world. Thus the relationship between socio-​economic development, energy use, and climate change cannot be treated as a linear causal link but as a cyclical, complex, system (Primer to Climate Scenarios 2018). Energy is a key determinant in consumption in aggregate national, industrial, and residential levels, depending on a country’s natural endowments. Developing countries exhibit institutional conditions that might make use of production factors, such as being energy efficient, implying high costs, supply constraints, and potentially high pollution intensities. Energy is crucial also for human living conditions and any policy framework needs to find ways that ensure energy access. In sub-​Saharan Africa and South Asia, more than one billion people do not have access to energy. ‘This represents a fundamental barrier to progress for a sizeable proportion of the world’s population, and has impacts on a wide range of development indicators, including health, education, food security, gender equality, livelihoods, and poverty reduction (World Bank 2018).

306   Roula Inglesi-Lotz Furthermore, the deterioration of climate conditions has a direct negative impact on physical infrastructure, human capital, natural resources, and ecosystem balance. The environmental reasoning for a clean energy transition has also opened the debate on whether its impact on economic growth and overall social development will be positive too. Studies such as the one by Dai et al. (2011) concluded that a clean energy transition has the potential to promote energy efficiency improvements, investment increases in the renewable sector, and all the indirect positive benefits on population well-​being from the reduction in emissions. A transition to cleaner, renewable forms of energy is related in the literature with socio-​economic and environmental benefits (Bohlmann et al. 2019; Consoli et al. 2016; Porter and van der Linde 1995). Some studies argue that such a substitution will create job opportunities (Cameron and van der Zwaan 2015; Lehr et al. 2008) while others argue that renewable-​energy generation is less labour intensive and, hence, jobs will be lost (Henriques et al. 2016; Sooriyaarachchi et al. 2015). In addition, World Bank (2018) makes a point that a transition to renewable-​based energies might boost more flexible technology solutions that have the potential to assist with the continent’s problem of access to energy. Globally, policymakers and governments have appreciated these impacts and have already started the transition towards cleaner forms of energy. South Africa is not an exception from these international trends—​being a strong representative of the developing nations at the Paris Climate Agreement as well as with the country’s Integrated Resource Planning (IRP). The role of access to reliable and affordable energy that does not harm the environment for all its citizens is at the heart of South Africa’s current and future path to economic prosperity. The central position of energy points to numerous interlinkages with factors in the economy, those which have an impact on energy and those on which energy has an impact. This chapter aims to systematically select and categorize important factors reflecting the economic, social, and environmental perspectives of the South African status quo in relation to the energy trends in the country. These factors are adopted by the framework suggested by Halsnaes and Garg (2011) (which also resembles closely the ‘triple bottom line’ approach used in the sustainability literature) and adopted for the purpose of the chapter: • Economic perspective (economic growth and development, employment, access to energy services) • Social perspective (poverty alleviation, education, health) • Environmental perspective (emissions, fossil-​fuel role in the supply mix). So, while all the relationship of energy with all the socio-​economic dimensions of South Africa are investigated, the chapter also appreciates that the energy supply mix is transitioning following international trends and due to a variety of reasons such as environmental concerns, ensuring security, and lower costs of alternative sources. Smil (2016) also discusses that the current energy transition is not about shifting between energy sources but towards achieving decarbonization of the energy supply internationally,

ENERGY AND CLIMATE CHANGE IN SOUTH AFRICA    307 via the most appropriate policies, technologies, and interventions and with minimal costs to society. Therefore, the chapter presents a historic description of South Africa’s energy transition that, at times, was market-​focused but since 2011 has become fuel-​ focused. The chapter argues that policies and interventions, if not designed properly and implemented consistently, will not only not be helpful for energy conditions but will intensify the already problematic socio-​economic conditions that exist.

15.2 Economic Perspective

3500

8000

3000

7000 6000

2500

5000

2000

4000

1500

3000

Energy per capita

2019

2016

2010

2013

2007

2004

2001

1998

1995

1992

1989

1986

1983

0

1980

1000 1977

500 1974

2000

1971

1000

GDP per capita (constant 2010 US$)

Energy use (kg of oil equivalent per capita)

Figure 15.1 shows that energy use per capita and GDP per capita in South Africa have similar patterns with a quite stable relationship over the period 1971–​2019. Assuming a production frontier approach, the relationship between energy and economic output of a country is influenced by the substitution between energy and inputs, the shifts in the composition of the energy input, the changes in the mix of the other inputs, changes in the composition of output, and technological change (Stern 2020). Analysing broadly the term ‘energy’ signifies its importance to the economy, but does not represent the differences between the energy quality of the various fuels used and the productivity of each of the fuels. The quality of a fuel is attributed to its physical characteristics (scarcity, cleanliness, density, storage ability) and its usage characteristics (cost of conversion if needed, flexibility of use, safety). Technology plays a role in the change of these characteristics over time. ‘It is generally believed that electricity is the highest quality energy vector, followed by natural gas and oil, and then coal, wood and other biomass in descending order of quality’ (Stern 2020). The evolution of the consumption mix follows similar trends historically from biomass to fossil fuels to cleaner

0

GDP per capita

Figure 15.1  Energy consumption per capita and GDP per capita 2010, South Africa, 1971–​2019 Source: World Bank (2020).

308   Roula Inglesi-Lotz

Percentage

fuels, highly linked with the developmental stages of the countries, as well as with differences in natural endowments (Csereklyei et al. 2016). South Africa is not an exception; as can be seen in Figure 15.2, the total primary energy supply is dominated by coal (75 per cent average in the decade between 2007 and 2017) while the world reported less than 30 per cent on average; shares for crude oil demonstrate the reverse: 14 per cent and 32 per cent for South Africa and the world respectively. The rest of the energy carriers also show South Africa utilizes them less than the rest of the world. Figure 15.2 Panel B gives a picture of the changes in the last decade: South Africa shows the same trends as the rest of the world in all energy carriers with the exception of nuclear. Ndlovu and Inglesi-​Lotz (2019) found that South Africa follows international trends of fossil-​fuel dependence both with its BRICS counterparts and with OECD countries; however, the more industrialized economies show a decreasing trend in fossil-​fuel share in their energy supply mixes compared to emerging and developing economies.

80% 70% 60% 50% 40% 30% 20% 10% 0%

75%

32%

28%

21% 14% 3%

3% Coal

Crude Oil

Natural gas

13%

5%

5%

Nuclear

Renewables and waste

Primary Energy Source South Africa

25% 20% 15% 10% 5% 0% –5% –10% –15% –20%

World 23%

21%

13% 15% 7% 0% –2%

–1% –7%

Coal

Crude Oil

Natural gas

South Africa, Change 1997–2017

–16% Nuclear

Renewables and waste

World, Change 1997–2017

Figure  15.2  Primary energy shares panel A (top): averages 2007–​17; panel B (bottom): percentage change between 1997 and 2017 Source: Data from IEA 2020 (percentages do not sum to 100 due to rounding). Note: For this table, the category for crude oil is a sum of the IEA categories crude, NGL, and feedstocks with oil products; the category renewables and waste is a sum of the IEA categories renewables and wastes and heat; and we excluded electricity as a primary energy source.

ENERGY AND CLIMATE CHANGE IN SOUTH AFRICA    309 This shift to higher-​quality energy fuels is considered essential for the efficient use of energy and, hence, the relationship between energy and economic output (Kaufmann 2004). More discussion on the role of petroleum in the supply and consumption mix can be found in ­Chapter 14 of this volume. Next, the structure of the economy and hence the economic output mix is an important factor in the interconnection between energy and economic growth. At early developmental stages, economies are primarily based on primary sector activities and especially agriculture. As economies progress, a natural path is to go through an industrialization period and hence a more energy-​intensive nature of the economy emerges. Then the next transition to the tertiary sector and services sees a decrease in energy usage as these sectors have lower energy requirements. Recently, it has been observed that many emerging countries have the potential to leapfrog from the first to the last stage, not going through the energy-​intensive industrial period. Some would translate this as an expectation for low-​energy usage due to the composition of the economic output in the country, but Stern (2020) warns that when the indirect energy use in manufactured goods and services is factored in, the services sector and households are more energy intensive than initially thought. Van Benthem (2015) examined a number of less-​developed economies and discovered that although these countries presented energy-​efficiency improvements, these savings in energy consumption have been offset by shifts to consumption of energy-​intensive products and outsourcing. The South African services sector’s value ​added as a percentage of GDP showed a steep increase from the 1980s without that being after a substantial rise in the industrial sector’s contribution (World Bank 2020). Although energy is mostly looked at as an input to the economic production of the country, the broader energy sector itself is a significant contributor to the growth and development of the economy. Table 15.1 presents the trends of total economic output and employment of selected energy-​related industries to demonstrate the importance of these economic sub ​sectors. The energy sector, broadly including the mining sector, employs a significant amount of the South African labour force. For the same selected industries, Table 15.1 shows the number of employed persons in the country. The overall contribution of specific industries might seem minimal but, in view of the 30 per cent plus unemployment rates of the country in 2020 and the impact that COVID-​19 will have in the long run, these hundreds of thousands of jobs are essential. The increase of workers in the coal mining industry by 55.5 per cent since the change in the political regime of 1993/​94 and the end of sanctions is also an indicator that the coal mining industry plays an important role in the economy. The energy transition from fossil fuels to a cleaner supply mix will have an impact on the employment of these sectors too (Bohlmann et al. 2019). Increases in employment are expected due to the construction and operation of renewable ​energy-​generation facilities, but the coal-​generation sector will experience severe cuts in employment that will cancel some (all?) of the positive effects:  ‘the minimization of the negative consequences of a transition to a less coal energy supply mix in the country does not

310   Roula Inglesi-Lotz Table 15.1: Economic output and employment of selected energy-​related economic sectors Final economic output 2019 (R millions, constant 2010)

Employment

Share to all 2019 Change Change industries, (number of 1993–​2019 1993–​2019 % 2019 (%) employees) %

Coal mining

40,142.13

39.8

1.1

Electricity and gas

52,483.73

97.4

1.5

53,591

35.4

0.3

Coke, petroleum products, nuclear fuel

35,531.9

19.5

1.0

27,886 46.8

0.2

3,781.99

360.4

0.1

8,259 74.8

0.1

3,589,119.23

19.5

100.0

16,126,900 44.3

100.0

Electricity distribution and control apparatus All industries

101,929 55.5

Share to all industries, 2019 (%)

0.6

Source: Statistics South Africa (StatsSA). Note: Coal mining (QSIC 21); electricity and gas (QSIC 41); coke, petroleum products, nuclear fuel (QSIC 331–​333); electricity distribution and control apparatus (QSIC 362).

only depend on the country’s energy demand and supply profile but also the reaction to global markets, particularly that of coal’ (Bohlmann et al. 2019). Access to goods and services and energy to meet basic needs is imperative for socio-​ economic development (Johansson and Goldemberg 2002). South Africa’s electricity network is well developed and energy access rates are among the highest on the continent (see Figure 15.3). In order to ensure universal access, particularly in rural areas, IEA (2019) suggests that the least-​cost way would be to connect 81 per cent of those in rural areas by extending the national grid while the remaining 12 per cent should access energy via mini-​grids and self-​generation technologies. To assess the full influence of electrification on income-​generation opportunities is more challenging than expected as the impact of expanding the grid and connecting new users in areas that were not connected before is different from ensuring all citizens are connected to the existing grid (Prasad and Dieden 2007). Prasad and Dieden (2007) in their study on South African small and medium enterprises (SMEs) found that approximately 40–​53 per cent of the rise in entrepreneurial activity was due to connectivity to the electricity grid after its expansion to their regional area. Their business growth was significantly higher than those SMEs already connected. Behind the positive picture of a well-​developed electricity network that ensures physical access, ability to pay and affordability, are factors which will influence the aftermath of the electrification programme. The group defined by Essex and de Groot (2019) as an

ENERGY AND CLIMATE CHANGE IN SOUTH AFRICA    311

New connections by 2030 On-grid Mini-grids Stand-alone systems Transmission lines (>69 kV) Existing Planned

Figure 15.3  Electricity access—​new connections and transmissions lines, South Africa Source: IEA (2019).

‘energy underclass’ is divided into two: those who do not have physical access to electricity (national electricity access: 92 per cent of the population in 2018 from 84 per cent in 1996 [World Bank 2020]), and those who are energy vulnerable1 (Bouzarovski and Petrova 2015). Policies that focus on electricity ignore the diversification of fuels that can be used to ensure the provision of access to energy. The goal as per Gonzalez-​Eguino (2015) is the provision of a variety of services appropriate for use for essential services (lighting, heating, and cooking). Essex and de Groot (2019) stress that the post-​apartheid regime implemented a variety of pro-​poor policies to assist with the previously disadvantaged part of the population. Those policies identified low-​income disadvantaged households and offered free access to a minimum amount of free electricity. In 2000, the Free Basic Electricity (FBE) programme was introduced which provided 50kWh per month for free to all households using less than 450kWh (Mohlakoana 2014). Later, in 2007, households not connected to the national grid could also benefit from FBE for using alternative energies. The amount of 50kWh is widely contested in policy circles and academic literature. Ruiters (2009) explains that first, although the scheme suggests the amount is enough to cover all basic needs (cooking, lighting, heating, and basic appliances), it actually covers only household lighting. In order to define a fair energy-​use level, Aydemir and Soytas (2020) stress that one needs to measure the current state of energy consumption and well-​being in conjunction with the availability and efficiency of appliances and their link to providing a certain standard of quality of life. The 50kWh free amount has remained unchanged since the 1990s while lifestyle, living standards, and dependency on energy have substantially changed since 1  Energy

vulnerable population in this context includes those that are unable to keep their homes warm or pay their bills or they don’t have access to energy (Energy poverty, n.d.).

312   Roula Inglesi-Lotz then. Ruiters (2009), second, suggests that FBE poses an extra burden on the low-​ income households which feel pressure to save electricity thus adding an extra burden to their challenging everyday lives. Thom (2000) also supports the notion that the ‘blanket’ programme implemented fails to demonstrate the system’s understanding of rural households’ needs. The Integrated National Electrification Programme (INEP) was initiated in 2002 to integrate efforts for electrification in the country and formalize the public sector’s involvement in the financing of those efforts (World Bank 2017). By promoting infrastructure development, INEP is considered by many studies to be the most impactful policy when it comes to electrification and access to energy. In 2010, the Inclining Block Tariff System (IBTS) was introduced to remedy the inequalities among electricity consumers. According to this subsidy model, high-​ consumption users get charged at higher rates to cross-​subsidize the low-​consumption users. Only 30 per cent of municipalities in the country implemented the IBTS fully. Although the tool looks effective and straightforward, it ignores the reality of many households using a single meter, particularly in townships, leading to an accumulated amount of consumption and to a higher-​tariff bracket. Eskom aimed to introduce pre-​paid meters (Essex and de Groot 2019) as a way towards an infrastructural model that provided electricity for all. Lack of ability to pay and the strain placed on households to put a ceiling on their consumption of electricity was linked to expanding inequalities further in rural areas (Makonese et al. 2012). However, by utilizing technology, Eskom and the municipalities can recover some of the costs from a system where households do not always have an address and therefore cannot be billed appropriately (McDonald 2012).

15.3 Social Perspective Following the discussion of the previous section, access to energy is but one of the main challenges faced by a poor population. Economic poverty is broadly considered to be the lack of sufficient income to purchase basic goods and services and the alleviation of economic poverty is a major objective of sub-​Saharan African policymakers and specifically of South African policymakers. The target is to develop an environment characterized by inclusive growth and efficient allocation of resources. Energy poverty within the same framework is the ‘exclusion of people from basic access to energy’ (Singh and Inglesi-​Lotz 2020). Reddy (2000) defines energy poverty as ‘the absence of sufficient choice in accessing adequate, affordable, reliable, high-​quality, safe, and environmentally benign energy services to support economic and human development’. Figure 15.4 shows the improvement in access to electricity in South Africa since democratization, however it also points to the existing inequalities between rural and urban population over decades.

ENERGY AND CLIMATE CHANGE IN SOUTH AFRICA    313 100 80 60 40 20

19 96 19 97 19 98 19 99 20 0 20 0 01 20 02 20 03 20 04 20 05 20 06 20 07 20 08 20 09 20 10 20 1 20 1 12 20 13 20 14 20 15 20 16 20 17 20 18

0

Access to electricity, urban (% of urban population) Access to electricity, (% of population) Access to electricity, rural (% of rural population)

Figure 15.4  Access to electricity (total, urban, and rural), 1996–​2018, South Africa Source: World Bank (2020).

These inequalities in electrification are one of, admittedly, many factors that have intensified the income inequalities in a country that has been consistently ranked amongst the highest in income inequality based on their Gini coefficient. South Africa’s inequalities intensified from 1993 until 2010 (data availability) contrary to expectations: the Gini coefficient was 59.3 per cent in 1993 and 63.4 per cent in 2010 (World Bank 2020). Huber (2015) boldly states that ‘electricity networks not only reflect the uneven geographies of cities but actively reproduce them’. The actual problem of the expansion of the national grid to urban areas is inherent to the system and has nothing to do with the availability of technologies and other resources. Planning guidelines stipulate that informal dwellings that are located on private, unstable land within zones prohibited for development cannot be electrified. South African municipal distributors have been grappling with the issue of how to provide electricity to informal households for more than fifteen years, and a significant body of experience has developed regarding how to undertake this successfully (Gaunt et al. 2012). Households in informal settlements are presented with two options: connect to the grid illegally or remain without access to electricity. Choosing the illegal connection option carries great risks, which the South African system has suffered since the end of the 1990s. The system can be overloaded, which leads to a lack of stable connections and trips—​negatively affecting not only those with the illegal connection but all users in the area. Illegal connections add to the burden of proper planning by Eskom and the municipal distributors. Finally, illegal connections are not safe—​the risk of electrocution is high due to a lack of electrical protection, maintenance not carried out by professionals, and illegally connected wires coming into contact with other household items and building parts.

314   Roula Inglesi-Lotz Such are the decisions low-​income households have to make. Even households who manage to have a connection have to deal with supply that is interrupted or for which they cannot afford to pay. Therefore, they all shift to other fuels for their basic needs for heating and cooking. That runs contrary to the patterns of the 1990s that followed the theory of an ‘energy ladder’ in households as their development phases progress from traditional fuels to electricity (Davis 1998). The study by Davis (1998) on South African households showed that in a period of relatively lower electricity tariffs, as incomes rose, dependency on electricity also increased for electrified households. In a 2012 study, Gaunt et al. concluded that universal ‘electrification requires a greater degree of flexibility, including greater on-​site planning rather than desk-​based planning, and making network layout and technology decisions in response to conditions encountered during implementation’. Access to electricity services has been found in the literature to be a contributory factor to fuel-​switching, particularly for cooking (Heltberg 2005). Barnes and Floor (1996) found that ‘(i) where electricity is available, fewer barriers constrain other modern fuels as well, and (ii) availability of lighting and other appliances spurs people to a greater acceptance of modernity and modern fuels.’ When electricity access is provided and at first used for lighting purposes, rural households put the emphasis on moving up the energy ladder more actively (Ouedraogo 2006). Particularly for South Africa, Bohlmann and Inglesi-​Lotz (2020) explain that ‘low-​income households (decile 1) consumed 0.9 per cent of the total consumption of housing, water, electricity, gas, and other fuels, compared to high-​income households (decile 10), which consumed 51.3 per cent of total consumption on housing, water, electricity, gas, and other fuels’. A factor that is underexamined in the academic debate on the energy–​economy nexus is the level of education of the country’s population. Efficiency in economic systems and processes can be improved by increasing education in quantity and quality. Consequently, education can also affect energy in a variety of ways. First, access to energy for the schooling institutions in rural areas and energy for transport purposes for the users to access education are crucial for developing countries such as South Africa with severely underdeveloped rural areas with low levels of educated population. Education infrastructure is essential for South Africa but unreliable electricity networks have intensified the problem. In poor nations that pursue convergence with the more developed world, increases in education lead to higher job opportunities and hence income generation, which subsequently leads to purchasing power for home appliances as well as for goods and services and hence, higher energy consumption. In contrast, in countries which exceed a certain developmental threshold, education leads to potential energy savings as users are educated adequately to seek environmentally friendly and more efficient appliances in order to reduce their footprint. Higher education levels enable consumers to process cognitively the benefits of their fuel choice, given that their income is theoretically and generally higher leading to access to funds for these choices. Rural populations traditionally use more accessible and traditional sources of energy such as wood, however as individuals get access to education,

ENERGY AND CLIMATE CHANGE IN SOUTH AFRICA    315 they take advantage of opportunities to migrate to urban areas where they can expect an uplift in job and further education opportunities, and hence their income-​generation capacity. All these have the potential to promote the substitution of traditional energy fuels for higher-​quality, safer ones. ‘Additionally, these improved “awareness” levels in society lead to more informed consumers and public planners who will make better energy purchasing, generation, usage, and distribution decisions, which may in turn reduce energy consumption levels’ (Inglesi-​Lotz and del Corral Morales 2017).

15.4 Environmental Perspective Academic and business literature mostly agrees that the climate of the planet changes rapidly vastly due to the human activity that leads to greenhouse gas (GHG) emissions. The emission concentration in the atmosphere has demonstrated an upward trend continuously since the first industrial revolution in the 1700s. The factors attributed to this increase are the burning of fossil fuels, industrialization of economies, and deforestation. These emissions have been proven scientifically to induce global warming, as demonstrated in the higher average temperatures experienced globally (Meyer et al. 2014). Figure 15.5, by presenting the CO2 emissions per capita and the share of fossil fuels to total energy consumption, demonstrates the correlation between the two: the emissions generated from burning fossil fuels is one of the most prominent factors in emissions’ increases. Note here that even with its decreasing pattern, since 1971, the share of fossil fuels to total energy consumption has never fallen below 84 per cent and that a decrease in share does not mean necessarily an increase in volume. The CO2 emissions per capita 91 90 89 88 87 86 85 84 83 82 81

10 8 6 4 2

79 19 81 19 83 19 85 19 87 19 89 19 91 19 93 19 95 19 97 19 99 20 01 20 03 20 05 20 07 20 09 20 11 20 13

77

19

75

19

19

19

19

71 73

0

% total energy consumption

Metric tones per capita

12

CO2 emissions per capita

Fossil fuel

Figure 15.5  CO2 emissions per capita and fossil-​fuel share of total energy consumption, 1971–​ 2014, South Africa Source: World Bank (2020). Note: It should be noted that fluctuations in the share of fossil fuels do not equal fluctuations in the volume of fossil fuels in the generation of electricity or their overall consumption in the economy.

316   Roula Inglesi-Lotz have remained between 8 and 10 metric tonnes per capita since 1983, even though the share of fossil fuels experienced a decrease in the 1990s and a subsequent increase from the end of the 1990s. South Africa along with the rest of the African continent are among the most vulnerable regions on the planet due to its geography and socio-​economic conditions. The Department of Environmental Affairs estimates that with no mitigation interventions, a warming of 5–​8 degrees should be expected in the country: drier weather in the west and more rainfalls in the east. The impact will be significant for water resources, food production, and consequently for the overall economy and society, particularly for poor communities (for more discussion see ­Chapter 16 in this volume). As a result, South African policymakers appreciate climate change as a risk to the sustainable future of the country and, hence, the general development of the economy and its citizens (Department of Environmental Affairs [DEA] 2018). South Africa ratified the United Nations Framework Convention on Climate Change (UNFCCC), thus demonstrating a commitment to monitor greenhouse-​gas emissions by preparing annual inventories and undertaking the right actions and programmes to reduce emissions. Mailula (2019) states that the reporting of emissions has been erratic and inconsistent over the years, and only recently has the Department of Minerals and Energy reported on annual sectoral emissions linked with the energy balances of the country. In 2018, South Africa’s Low Emission Development Strategy (LEDS) 2050 was published which aims to build on previous efforts such as the National Climate Change Response Strategy in 2004 in order to set and follow the most appropriate guidelines to keep the country on a low-​carbon trajectory, accounting for challenging economic conditions. The strategy proposes a combination of mitigation actions and adaptation measures. Thus, the South African LEDS recognizes this and outlines the research, technology development, and transfer needed for the country to transit to low ​carbon-​ emissions development. The strategy also stresses the need for ongoing reporting and evaluation in place of the one-​off and unlinked efforts of the past.

15.5  Energy Transition in South Africa Over the centuries the interconnection between humans and energy has transformed multiple times, typically from traditional energy fuels to modern ones (disregarding the environmental impacts and enhancing rapid industrialization) and subsequently, from exhaustible ‘dirty’ fuels towards cleaner ones (Grubb et al. 2008).2 One of the challenges in this transition will be to substitute fossil-​fuels usage in specific energy-​intensive economic sectors such as mining, iron and steel, and other metals. At the same time, these sectors are significant sources of job opportunities, investment, and growth. These

2 

Chapters 14 and 16 in this volume also present perspectives of the energy transition in South Africa.

ENERGY AND CLIMATE CHANGE IN SOUTH AFRICA    317 industries are ‘inflexible and slow to change. For example, once a major investment has been made in the construction of a smelter, opportunities to change to more energy-​ efficient technology or production process are limited’ (Inglesi-​Lotz 2015). Industrialized countries that boast a well-​established and universal energy/​electricity network have paved the way for systems that generate and distribute energy to all (the idea of ‘public good’). The emerging and developing countries of the Global South, including South Africa and India, make an effort to provide their continuously rising populations with access to affordable and reliable clean energy services too (as per Sustainable Development Goal 7) (Essex and de Groot 2019). In all these changes in global dynamics and the broader socio-​economic transitions of the population (rural to urban, low ​income to middle-​and high ​income, for example) and market structures towards liberalization, the world faces the challenge of climate change and urgency in making use of alternatives to fossil fuels and moving towards renewable energy (Grubler 2012). A triple challenge is faced by policymakers in the energy sector: (1) generate adequate energy for the rising population at competitive prices; (2) distribute to users considering its affordability; and (3) develop and make use of clean energies to mitigate the consequences of climate change (Cock 2004). To ensure all three in combination, the South African energy sector has slowly but surely undergone noticeable transformations. Essex and de Groot (2019) divided the evolution of the energy transitions from provision responsibility, market structure, and access in South Africa into four phases: 1. The country’s electricity network in colonial South Africa (1860–​1948): Policymakers founded the take-​off of the South African economy on cheap electricity generated by coal reserves that were in abundance in four of the nine provinces of the country (Limpopo, Mpumalanga, Kwazulu-​Natal, and the Free State). Diamond-​and gold-​mining companies acted as the first private electricity generating companies, focusing primarily around mining towns (Gentle 2009). Electricity generation also started at the urban municipal level as part of their mandate to provide electricity in public spaces (Essex and de Groot 2019). Towards the middle of the period, in 1910, the government intensified the effort to create an integrated system for generation and transmission, focused primarily on supplying low-​tariff electricity to the industrial sector of the country (leaving most of the South African population without access). The Electricity Act of 1922 formalized the nationalization of electricity generation and by 1923, the then Eskom was formed (Christie 1984) and the formation of the national grid became reality. 2. Electricity under apartheid (1948–​94): The National Party’s election in 1948 signified an era in South Africa characterized by ‘segregation, spatial inequality, exclusion, and marginalization’ (Beall et  al. 2000; Essex and de Groot 2019). Apartheid policies with their exclusionary nature resulted in less than a third of households in townships having electricity connections (Louw et  al. 2008). Electricity was provided only to white areas

318   Roula Inglesi-Lotz (the regime’s idea of universal access), thus creating a racially determined part of the population that had to rely on dirty and dangerous fuels to cover their everyday needs. 3. Electricity in the post-​apartheid era (1994–​2011): Soon after the 1994 democratic elections, the government concentrated efforts on rebuilding the economy with a focus on taking care of the inequalities created by the previous regime, putting into action the Reconstruction and Development Programme (Mohlakoana 2014). In the programme, electricity was identified as a key and convenient fuel to ensure universal access and minimize health incidents due to indoor air pollution, while at the same time creating employment and income-​generation opportunities and improving living standards. The municipalities acquired a more prominent role in the distribution of electricity to the biggest part of the South African population (Mosdell 2016). Since then, Eskom has been the main generator of electricity (more than 90 per cent of the total) and the main distributor for industrial and mining users and only some residential areas; the vast majority of households are serviced by municipalities which purchase electricity generated by Eskom and resell/​distribute at a profit. 4. Climate change and renewable energies (2011–​present). South Africa’s 14th position on the list of the largest emitters of greenhouse gases in the world (2018) has been mainly attributed to coal being the dominant fuel used for electricity generation. The country has now made commitments as a signatory of agreements such as the Kyoto Protocol in 2002 and the Copenhagen Climate Change Conference in 2009 and targets set for the reduction of emissions make the shift to supporting renewable energies more prominent. This shift has been promoted because energy policymakers see renewable energies as a solution to the load-​shedding experienced since 2008. In this phase, renewable-​energy generation in the country is classified into two broad categories:  (a) renewable-​energy generators participating in the Renewable Energy Independent Producers Procurement Programme (REIPPP) where producers can tender for licences to supply Eskom’s grid under a two-​decade agreement and (b) small-​ scale embedded generation programmes where ‘the electricity generated is used solely on site’ (Baker and Phillips 2019). In 2020, Eskom operated fifteen coal-​fired power plants accounting for 85 per cent of installed capacity (Bohlmann et al. 2019). Although, however, the role of renewable energies is expected to be greater in the future supply mix, new coal-​fired generation is also planned (Essex and de Groot 2019). Bohlmann et al. (2019) explain that these will come in to replace the ageing fleet of coal-​fired power plants due to be decommissioned by 2030. To add a further layer to the debate on energy transition, the impact of a transition to a low-​coal supply mix in South Africa will not be homogenous across all provinces as the bulk of coal generation is located in the Mpumalanga province while the expected

ENERGY AND CLIMATE CHANGE IN SOUTH AFRICA    319 renewable-​energy generation will take place mostly in the Western and Eastern Cape provinces. As coal-​fired power plants and coal mines are important employers for the population in the region, the socio-​economic impact on these provinces will be negative without proper planning. Although there will be an increase in job opportunities during the expansion of renewable energies in the short term, the net effect could potentially be negative in terms of the high number of unemployed and unskilled workers resulting from the closing of coal-​fired power plants. Therefore, careful regional planning is essential for a smooth and ‘just’ energy transition within the borders of South Africa (Bohlmann et al. 2019).

15.6 Conclusion This chapter examined the socio-​economic dimensions that interconnect with energy use and supply in South Africa, and placed the various types of transitions the South African energy sector has undergone into a historical perspective. The importance of the provision of energy services was a cornerstone in all the analysis, demonstrated particularly in the discussion of access to energy services and the broader achievement of universal access. The chapter shows that an expansion of the national electricity grid or provision of an amount of free electricity have been solutions under certain conditions but the current socio-​economic conditions and livelihood standards of South African households now require solutions that will not challenge their already difficult livelihoods and constrained budgets. ‘The cost of providing universal access to energy by 2030 would require an annual investment of $35 billion, i.e. much less than the amount provided annually in subsidies to fossil fuels (IEA in Gonzalez-​Eguino 2015). As with any investment, particularly by government in an emerging economy with severe macroeconomic challenges, the net benefit of any investment should be taken into account: in this case, the socio-​economic benefits from the improvement in households’ well-​being and any further economic development should be compared with the financial and opportunity cost of improving and upgrading the energy infrastructure. The future of South Africa’s energy sector is not predetermined—​more than one scenario can play out in the future, depending on the policy choices that will promote sustainable energy supply and use in a synergistic manner with economic growth and human development. These choices will have to reflect the resource endowments of the country and have a common goal with three interlinked objectives: (a) adequate cost-​ effective energy; (b) affordable energy; (c) energy generation that mitigates climate change (Cock 2004). Of course, this is easier said than done. There are opportunities for the energy sector and, with the right decisions and institutional foundations, energy can become the ‘cornerstone of economic development and make a huge difference in the lives (International Energy Agency [IEA] 2019) of the South African population.

320   Roula Inglesi-Lotz

References Aydemir, Mehmet Olcay, and Ugur Soytas. 2020. ‘Quality of life and energy use: Is there a fair energy use level?’, in Ugur Soytas and Ramazan Sari (eds) Routledge Handbook of Energy Economics. Abingdon: Routledge. Baker, Lucy, and Jon Phillips. 2019. ‘Tensions in the transition: The politics of electricity distribution in South Africa’, Environment and Planning C: Politics and Space, 37(1): 177–​96. Barnes, Douglas F., and Willlem M. Floor. 1996. ‘Rural energy in developing countries:  A challenge for economic development’, Annual Review of Energy and the Environment, 21(1): 497–​530. Beall, Jo, Owen Crankshaw, and Susan Parnell. 2000. ‘Local government, poverty reduction and inequality in Johannesburg’, Environment and Urbanization, 12(1): 107–​22. Bohlmann, Heinrich R., J. Mark Horridge, Roula Inglesi-​Lotz, Elizabeth L. Roos, and Lardo Stander. 2019. ‘Regional employment and economic growth effects of South Africa’s transition to low-​carbon energy supply mix’, Energy Policy, 128: 830–​7. Bohlmann, Jessika, and Roula Inglesi-​Lotz. 2020. ‘Examining the determinants of electricity demand by South African households per income level’, Energy Policy, 148: 111901. Bouzarovski, Stefan, and Saska Petrova. 2015. ‘A global perspective on domestic energy deprivation: Overcoming the energy poverty–​fuel poverty binary’, Energy Research and Social Science, 10: 31–​40. Cameron, Lachlan, and Bob van der Zwaan. 2015. ‘Employment factors for wind and solar energy technologies:  A literature review’, Renewable and Sustainable Energy Reviews, 45: 160–​72. Christie, Renfrew. 1984. Electricity, Industry and Class in South Africa. Basingstoke: Palgrave Macmillan. Cock, Jacklyn. 2004. ‘Connecting the red, brown and green:  The environmental justice movement in South Africa’. University of KwaZulu-​Natal, Durban [online] http://​ccs.ukzn. ac.za/​files/​Cock%20Connecting%20the%20red,%20brown%20and%20green%20The%20 environmental%20justice%20movement%20in%20South%20Africa.pdf. Consoli, Davide, Giovanni Marin, Alberto Marzucchi, and Francesco Vona. 2016. ‘Do green jobs differ from non-​green jobs in terms of skills and human capital?’, Research Policy, 45(5): 1046–​60. Csereklyei, Zsuzsanna, M. D. Mar Rubio-​Varas, and David I. Stern. 2016. ‘Energy and economic growth: The stylized facts’, The Energy Journal, 37(2): 223–​55  . Dai, Hancheng, Toshihiko Masui, Yuzuru Matsuoka, and Shinichiro Fujimori. 2011. ‘Assessment of China’s climate commitment and non-​fossil energy plan towards 2020 using hybrid AIM/​CGE model’, Energy Policy, 39(5): 2875–​87. Davis, Mark, 1998. ‘Rural household energy consumption: The effects of access to electricity—​ evidence from South Africa’, Energy Policy, 26(3): 207–​17. Department of Environmental Affairs (DEA). 2018. ‘South Africa’s low emission development strategy 2050’. Department of Environmental Affairs [online] https://​www.environment. gov.za/​sites/​default/​files/​strategic_​plans/​southafricas_​lowemission_​developmentstrategy_​ dec2018draft.pdf. Essex, Stephen, and Jiska de Groot. 2019. ‘Understanding energy transitions: The changing versions of the modern infrastructure ideal and the “energy underclass” in South Africa, 1860–​2019’, Energy Policy, 133: 110937.

ENERGY AND CLIMATE CHANGE IN SOUTH AFRICA    321 Gaunt, Trevor, Malcolm Salida, Roger Macfarlane, Sivuyile Maboda, Yachika Reddy, and Mark Borchers. 2012. ‘Informal electrification in South Africa:  Experience, opportunities and challenges’. Sustainable Energy Africa [online] http://​cityenergy.org.za/​uploads/​resource_​ 116.pdf. Gentle, L. 2009. ‘Escom to Eskom: From radical Keynesian capitalism to neo-​liberalism (1910–​ 1994)’, in David A. McDonald (ed.) Electric Capitalism: Recolonising Africa on the Power Grid. London: Routledge. Gonzalez-​Eguino, Mikel. 2015. ‘Energy poverty:  An overview’, Renewable and Sustainable Energy Reviews, 47(C): 377–​85. Grubb, Michael, Tooraj Jamasb, and Michael G. Pollitt. 2008. Delivering a Low Carbon Electricity System:  Technologies, Economics and Policy. Cambridge:  Cambridge University Press. Grubler, Arnulf. 2012. ‘Energy transitions research:  Insights and cautionary tales’, Energy Policy, 50: 8–​16. Halsnæs, Kirsten, and Amit Garg. 2011. ‘Assessing the role of energy in development and climate policies—​conceptual approach and key indicators’, World Development, 39(6): 987–​1001. Heltberg, Rasmus. 2005. ‘Factors determining household fuel choice in Guatemala’, Environment and Development Economics, 10(3): 337–​61. Henriques, Carla O., Dulce H. Coelho, and Natalie L. Cassidy. 2016. ‘Employment impact assessment of renewable energy targets for electricity generation by 2020—​An IO LCA approach’, Sustainable Cities and Society, 26: 519–​30. Huber, Matthew, 2015. ‘Theorizing energy geographies’, Geography Compass, 9(6): 327–​38. Inglesi-​ Lotz, Roula, 2015. ‘How South Africa can transition to a less energy-​ intensive economy’, The Conversation [online] http://​theconversation.com/​ how-​south-​africa-​can-​transition-​to-​a-​less-​energy-​intensive-​economy-​44240. Inglesi-​Lotz, Roula and Luis Diez del Corral Morales. 2017. ‘The effect of education on a country’s energy consumption:  Evidence from developed and developing countries’. Economic Research Southern Africa Working Paper no.  678, Economic Research South Africa (ERSA) [online] http://​www.econrsa.org/​publications/​working-​papers/​ effect-​education-​countrys-​energy-​consumption-​evidence-​developed-​and  International Energy Agency, (IEA). 2019. ‘Africa energy outlook 2019—​analysis’, [online] https://​www.iea.org/​reports/​africa-​energy-​outlook-​2019. Johansson, Thomas B., and Jose Goldemberg. 2002. Energy for Sustainable Development: A Policy Agenda. ETDEWEB. Kaufmann, Robert K. 2004. ‘The mechanisms for autonomous energy efficiency increases: A cointegration analysis of the US energy/​GDP ratio’, The Energy Journal, 25(1): 63–​86. Lehr, Ulrike, Joachim Nitsch, Marlene Kratzat, Christian Lutz, and Dietmar Edler. 2008. ‘Renewable energy and employment in Germany’, Energy Policy, 36(1): 108–​17. Louw, Kate, Beatrice Conradie, Mark Howells, and Marcus Dekenah. (2008). ‘Determinants of electricity demand for newly electrified low-​income African households’, Energy Policy, 36(8): 2812–​18. Mailula, Tabana. 2019. ‘South Africa’s GHG emissions data lacks consistency and timeliness’, Climate Scorecard, 6 June [online] https://​www.climatescorecard.org/​2019/​06/​south-​ africas-​ghg-​emissions-​data-​lacks-​consistency-​and-​timeliness/​ Pretoria, South Africa. Makonese, Tafadzwa, David Kimani Kimemia, and Harold John Annegarn. 2012. ‘Assessment of free basic electricity and use of pre-​paid meters in South Africa’, 2012 Proceedings of the 20th Domestic Use of Energy Conference, 165–​72.

322   Roula Inglesi-Lotz McDonald, David A. 2012. Electric Capitalism:  Recolonising Africa on the Power Grid. London: Routledge. Meyer, L., S. Brinkman, L. van Kesteren, N. Leprince-​Ringuet, and F. van Boxmeer. 2014. ‘Technical support unit for the synthesis report’. IPCC, Geneva, Switzerland . Mohlakoana, Nthabiseng, 2014. ‘Implementing the South African free basic alternative energy policy: A dynamic actor interaction, [online] https://​doi.org/​10.3990/​1.9789036537971. Mosdell, Susan C. 2016. ‘The role of municipalities in energy governance in South Africa’, [online] https://​open.uct.ac.za/​handle/​11427/​20812. Ndlovu, Vanessa, and Roula Inglesi-​Lotz. 2019. ‘Positioning South Africa’s energy supply mix internationally:  Comparative and policy review analysis’, Journal of Energy in Southern Africa, 30(2): 14–​27. Ouedraogo, Boukary, 2006. ‘Household energy preferences for cooking in urban Ouagadougou, Burkina Faso’, Energy Policy, 34(18): 3787–​95. Porter, Michael E., and Claas van der Linde. 1995. ‘Toward a new conception of the environment–​competitiveness relationship’, Journal of Economic Perspectives, 9(4): 97–​118. Prasad, G., and S. Dieden. 2007. ‘Does access to electricity enable the uptake of small and medium enterprises in South Africa?’, [online] https://​open.uct.ac.za/​handle/​11427/​16803. Primer to Climate Scenarios. 2018. ‘Primer to climate scenarios’, [online] https://​climatescenarios. org/​primer/​how-​are-​socioeconomic-​development-​and-​climate-​change-​connected. Reddy Amulya K. N., J. Ramakrishna, Q. Wodon, A. Kaniz, and M. Zaidi. 2000. ‘Energy and social issues’, in World Energy Assessment: Energy and the Challenges of Sustainability. New York: United Nations Development Programme Ruiters, Geavan. 2009. ‘Free basic electricity in South Africa:  A strategy for helping or containing the poor?’, in David A. McDonald (ed.) Electric Capitalism: Recolonising Africa on the Power Grid. London: Routledge. Singh, Kerschyl, and Roula Inglesi-​Lotz. 2020. ‘The role of energy poverty on economic growth in sub-​Saharan African countries’, Economics of Energy and Environmental Policy, 9(2): 1–​10. Smil, Vaclav, 2016. ‘Examining energy transitions: A dozen insights based on performance’, Energy Research and Social Science, 22: 194–​7. Sooriyaarachchi, Thilanka M., I-​Tsung Tsai, Sameh El Khatib, Amro M. Farid, and Toufic Mezher. 2015. ‘Job creation potentials and skill requirements in, PV, CSP, wind, water-​to-​ energy and energy efficiency value chains’, Renewable and Sustainable Energy Reviews, 52: 653–​68. Stern, David I. 2020. ‘Energy and economic growth’, in Ugur Soytas and Ramazan San (eds) Routledge Handbook of Energy Economics. London: Routledge. Thom, Cecile. 2000. ‘Use of grid electricity by rural households in South Africa’, Energy for Sustainable Development, 4(4): 36–​43. van Benthem, Arthur A. 2015. ‘Energy leapfrogging’, Journal of the Association of Environmental and Resource Economists, 2(1): 93–​132. World Bank. 2017. ‘State of electricity access report (SEAR) 2017’. World Bank, Washington, DC. World Bank. 2018. ‘Access to energy is at the heart of development’. World Bank, Washington, DC [online] https://​www.worldbank.org/​en/​news/​feature/​2018/​04/​18/​ access-​energy-​sustainable-​development-​goal-​7. World Bank. 2020. World Development Indicators (WDI). South African Dataset. World Bank, Washington, DC [online] https://​data.worldbank.org/​country/​south-​africa?view=chart.

Chapter 16

Climate Cha ng e a nd the Green Tra nsi t i on in Sou th A fri c a Channing Arndt, Sherwin Gabriel, Faaiqa Hartley, Kenneth Marc Strzepek, AND TIMOTHY S. THOMAS

16.1 Introduction This chapter explores one of the most complex challenges facing South Africa in creating plans and policies for future economic growth and development:  climate change. Higher temperatures and more variable rainfall have already affected the South African economy through (to give three examples) droughts and water shortages, which impact employment, poverty, and food security in both rural and urban areas; floods, which can destroy crops and infrastructure, disrupt power supplies, and lead to loss of life; and heatwaves, which affect human and livestock health and lower crop yields (Blignaut et al. 2009). Changes in the climate will continue for decades, probably at least through to the end of the century. The degree to which climate change affects different regions in South Africa is likely to vary with wide ranges in the direction and magnitude of change in key climate variables, especially precipitation. Regions with vulnerable households are often particularly sensitive to large climatic changes (Cullis et al. 2015; Earthlife Africa and Oxfam International 2009; Turpie et al. 2002). Because of these uncertainties and their interactions with South Africa’s growth and development challenges, formulation of plans and policies to support economic development is complicated. The South African government has acknowledged the threat climate change poses to economic development. The National Climate Change Response White Paper (RSA

324    ARNDT, GABRIEL, Hartley, STRZEPEK, AND THOMAS 2011) outlines government’s mitigation and adaptation interventions to respond to climate change. While some of the measures have been implemented, many priorities still need to be transformed into actionable policies that influence practices and planning (Ziervogel et al. 2014). A process to scope the long-​term impacts of climate change on South Africa has taken place through the 2013 South African Long-​Term Adaptation Scenarios Flagship Research Programme (DEA 2013). Local governments have also considered local responses to addressing climate vulnerabilities (DEDEA 2011; DEADP 2014; L-​DEDET 2016). Globally, responding to climate change will require, among other items, a transition to a low-​carbon economy. Global efforts to reduce emissions decrease the likely magnitude and ranges of uncertainty surrounding future climate change impacts (Cullis et al. 2015; Arndt et al. 2019a; Schlosser et al. 2020). In South Africa, the mitigation actions to enable such a transition will largely depend on transforming the energy sector. Due to its dependence on coal, energy supply is the largest source of greenhouse gas emissions for the country. Yet, large renewable resource endowments, falling costs of renewable energy, and improvements in systems integration have strengthened South Africa’s potential to reduce energy emissions and achieve a less carbon-​intensive economy (IRENA 2016; Hartley et al. 2019; Arndt et al. 2019a; McCall et al. 2019). However, the shift toward renewable energy and the proliferation of associated technologies, such as electric vehicles, do have negative implications for traditional activities in the economy tied to fossil fuels and the livelihoods that depend on those activities (Burton et al. 2018; Merven et al. 2019). Government intervention could ease the potential negative impacts of transitioning to a low-​carbon economy. This chapter presents and summarizes the key research on climate changes experienced in South Africa in recent years, along with projected changes in years to come. It illustrates the uncertainty related to climate change and the key channels through which climate change affects the economy. The economic and developmental impacts of such changes are presented along with the lessons for adaptation policy. As energy will be a primary focus of mitigation efforts in South Africa, this chapter also outlines the implications of such a transition and the factors that need to be accounted for in limiting the impacts on vulnerable populations.

16.2  The Hydroclimate of South Africa South Africa is a semi-​arid country with an average annual rainfall of 495mm, ranging from less than 100mm/​year in the western deserts to about 1200mm/​year in the eastern part of the country. Only 35 per cent of the country has a precipitation of 500mm or more, while 21 per cent has a precipitation of less than 200mm. Based on annual

Climate Change and the Green Transition in South Africa    325 rainfall, three climate zones can be distinguished as shown in Figure 16.1. The eastern parts of the country are summer rainfall areas with an annual precipitation of more than 500mm. The central and the western parts of the great plateau are semi-​arid to arid and are characterized by late summer rains, varying from less than 100mm/​year to approximately 500mm/​year. The Cape Fold mountains and the area between them and the sea have a winter rainfall season in the west and rainfall throughout the year in the more south-​easterly parts. Annual precipitation in this region varies from about 300mm to more than 900mm. Figure 16.1 also distinguishes nine water management areas (WMAs) (Cole et al. 2018). In sum, South Africa exhibits great spatial variability in climate and streamflow. There are also large variations in water use. It is thus misleading to report only national level hydro-​climatic data or analyses. An appropriate subnational spatial scale is needed. For this chapter we report results using six hydroclimatic zones as shown in Figure 16.2. Surface water resources are very unevenly distributed across the country with more than 60 per cent of its river flow arising from only 20 per cent of the land area. The distribution of total South Africa annual runoff among the six hydroclimatic zones is as follows: (1) Northeast: 17 per cent; (2) Pongola: 13 per cent; (3) Vaal: 4 per cent; (4) Orange 13 per cent; (5) Mnmvubu 28 per cent; (6)Western Cape: 24 per cent. The combination of low and highly variable rainfall (both inter-​and intra-​annually) with high evaporation and shallow dam basins leads to highly seasonal and varied runoff and water supply, even before considering climate change.

Figure 16.1  Annual rainfall distribution for Republic of South Africa Note: The spatial and seasonal variation leads to a wide-​spatial heterogeneity in agricultural production. Source: Cole, M.J., Bailey, R.M., Cullis, J.D. and New, M.G., 2018. Spatial inequality in water access and water use in South Africa. Water Policy, 20(1), pp.37–52.

326    ARNDT, GABRIEL, Hartley, STRZEPEK, AND THOMAS

Figure  16.2 The six hydro-​ climatic zones which are the aggregation of the 9 Water Management Areas with percentage of total national annual Source: DWA (2013)

16.3  Climate Change Impacts: Building on Past Analysis Southern Africa has been highlighted as being especially vulnerable to the impacts of climate change. Warmer temperatures, erratic rainfall, and shorter growing windows present substantial challenges for the agricultural sector and the households that depend on it. Insufficient water supply brings risks to agriculture, industries, households, and power generation. Flood and excessive heat can reduce the lifespan of infrastructure. Multiple climate stresses and a low ability to adapt have important implications for households and industries that depend largely on the environment. While direct implications of climate change focus on specific parts of economic activity, such as agricultural yield or runoff, these spill over to the rest of the economy. This has important implications for rural development, food security, risk management, energy planning, and infrastructure budgeting, to give a few examples. Climate effects are not uniform. A hotter and drier climate future for the Western Cape, for example, can coexist with a warmer, but wetter, KwaZulu-​Natal. Similarly, the economic effects differ because the profiles of industry, labour, and households are not the same across different regions in South Africa; and these differing profiles are affected by different climate risks. Climate change is already occurring and is expected to continue (details on climate change outcomes are presented in subsequent sections). Average global temperatures in

Climate Change and the Green Transition in South Africa    327 southern Africa have already increased by more than 1.5°C, compared with 1°C globally (Kusangaya et al. 2014). There is an urgency to develop and implement adaptation and mitigation measures now, as it may be more expensive, and potentially less effective, to respond later. To understand the potential risks of climate change to the economy, it is important to consider regional economic differences, and the channels by which climate change could affect them. Studies that focus on the economic impact of climate change in South Africa have mostly focused on the agriculture sector. Climate change negatively affects agricultural production mainly because of higher temperatures, with crops growing close to the upper limit of temperature tolerance levels (Gbetibouo, Ringler, and Hassan 2010; Deressa, Hassan, and Poonyth 2005). Dryland farming is generally found to be more vulnerable than irrigated agriculture, and the profitability of small-​scale farmers is affected more than that of large farmers (Turpie and Visser 2012; Blignaut et al. 2009). South Africa recognizes climate change among its many development challenges. Climate change is an important element of the National Development Plan, the country’s long-​term development agenda, as well as of a number of economic plans with shorter timelines, such as the Medium-​Term Strategic Framework. South Africa’s key policy document to respond to climate change is the South Africa National Climate Change Response White Paper (Department of Environmental Affairs 2012). It outlines the policy context and mandates for adaptation and mitigation and aims to ensure a just transition to a climate-​resilient and less carbon-​intensive economy.1 Building on the White Paper, the Long-​Term Adaptation Scenarios (LTAS) research programme was one of the first, large multidisciplinary efforts in South Africa to outline plausible development pathways under different climate futures. It focused on agriculture, water, health, and biodiversity, as well as the cross-​cutting and economic effects of these channels. Such interdisciplinary approaches are useful in integrating the insights and risks across different affected sectors. The translation of scientific and biophysical processes into coherent economic outcomes allows for identification and coordination of detailed policy responses grounded in evidence-​based scientific and economic analysis. To determine the economic effects of climate change, it is necessary to trace how climatic indicators influence economic value. Hence, under LTAS, a range of assessment tools were linked into an integrated framework, called SACReD—​Systematic Analysis of Climate Resilient Development—​making it possible to examine climate change impacts in the broader national economic context, and identify cross-​sectoral spillovers

1  Several

subsequent policies and legislation that consider climate change include the Air Quality Amendment Act (2014), National Energy Efficiency Strategy (2016), Integrated Resource Plan for Electricity (2019), and the Carbon Tax Act (2018). Other draft policies include the Integrated Energy Plan, National Adaptation Strategy, Climate Change Legal Framework, Green Transport Policy, and regulations for Greenhouse Gas Reporting and Pollution Prevention Plans.

328    ARNDT, GABRIEL, Hartley, STRZEPEK, AND THOMAS Climate Models Precipitation Temperature

Catchment Model Water Resources Model (WRYM)

Crop Models (IRRDEM)

Crop yields

Water Supply and Hydropower Models Water supply Energy supply

Economy-wide general equilibrium (GE) model

Urban and Industry Demands

Roads Impacts Model (IPSS) Rehabilitation costs and road lengths

Land inundation from sea-level rise

Figure  16.3  SACReD  –​Systematic Analysis of Climate Resilient Development Framework for LTAS Source: Cullis et al. (2015)

and vulnerabilities.2 Also, given the uncertainty in how exactly climate change will unfold, it is important to test a wide range of plausible outcomes. A schematic of the SACReD framework is shown in Figure 16.3. At the top are climate projections, mainly temperature and precipitation projections, for a given future climate. This climate is then passed to biophysical models—​crop, hydrological, and road damage models—​to determine the impact on crop yields, water and energy availability, and infrastructure costs for each climate scenario at fine levels of geographic and temporal detail to a given point in the future (LTAS went to 2050). These climate-​related effects on productivity, supply, and cost permit an economic evaluation of climate change. This broad array of impacts, at least some of which are likely to be positive, are then passed to an economy-​wide general equilibrium model, where market participants respond to changes in relative prices as a result of changes in supply due to climate effects. The economy-​wide model serves as a complex and coherent ‘adding machine’ that accounts for both multiple impacts and the reactions of agents. As will be discussed in more detail in the next section, climate futures are uncertain. To account for this uncertainty, multiple climates are passed through the SACReD framework, allowing for the generation of a range of potential outcomes. The results from the SACReD framework from the LTAS study suggest that the total impact of climate change on the level of real GDP by about 2050 is found to range between –​3.8 per cent and 0.3 per cent compared with a fictional ‘no climate change’ baseline. While positive GDP outcomes are possible, results indicate that, for the very large 2 The

SACReD framework has also been applied to climate change policy research in Ethiopia, Ghana, Malawi, Mozambique, Tanzania, Vietnam, and Zambia.

Climate Change and the Green Transition in South Africa    329 majority of climate futures, the impact on total GDP will be negative. The median result shows that by 2050, South Africa’s real GDP level will be about 1.5 per cent lower than in the baseline scenario. This translates into a 0.03 percentage point decline in average annual real GDP growth rate. The net present value of the potential impact on GDP out to 2050 is highly variable, ranging from losses of Rand 930 billion to gains of Rand 310 billion (real 2007 Rand). About 96 per cent of the climate scenarios show overall losses. The median loss in NPV is approximately Rand 259 billion which, at more than 10 per cent of 2007 GDP, is sizeable and should motivate action in terms of both mitigation and consideration and funding of potential adaptation scenarios (Cullis et al. 2015). The LTAS findings suggest that South Africa’s economy would very likely benefit from a low-​carbon energy future. While this requires a global effort to achieve, South Africa is Africa’s largest GHG emitter and, in 2018, was fourteenth largest in the world (Carbon Brief 2018). South Africa’s interests lie in reduced global GHG emissions. As will be discussed, South Africa has significant opportunities to achieve future economic growth and development in a sustainable and low-​GHG manner. The analyses under LTAS were conducted about seven years ago—​a long time, given the rapid development of capabilities for analysing climate change. Subsequent sections present updated analyses of elements of the SACReD schema, beginning with future climates.

16.4  The Uncertain Future Climate 16.4.1 Probabilistic Projections of Future Climates 16.4.1.1 Methodology Here we compare both future climate and weather based on two emissions scenarios, one with atmospheric GHG concentrations rising through the century and the other with emissions reductions in such a way as to limit global warming to no higher than a 2oC increase by 2100.3 For each of the emissions scenarios, we use 7,200 climates generated as described by Schlosser et al. (2020). Each of these climates show projected changes in monthly precipitation and near-​surface average temperature from 2020 to 2069. These climates vary for many reasons, including: • Imperfect human understanding of the global climate system • Uncertainties in emissions paths, especially for the higher emissions scenario • The inherent chaotic properties of the climate system (and climate models), which imply that small perturbations can lead to drastically different outcomes over time, especially at relatively fine spatial scales. 3 

Under LTAS, future climates were developed in a conceptually similar way. The treatment of weather was far less sophisticated and comprehensive than the approach presented here.

330    ARNDT, GABRIEL, Hartley, STRZEPEK, AND THOMAS For each of the climates, we consider 100 different weather realizations, each also spanning a fifty-​year period. The weather realizations are random draws of the de-​ trended monthly Princeton Global Forcing (PGF) database from 1948 to 2016 (Sheffield, Goteti, and Wood 2006).

16.4.1.2 Results In Table 16.1, which focuses on temperature, we see the median and the range from the 5th to 95th percentiles for the six hydroclimatic zones of South Africa, as well as for the nation for the warmest month in the wettest three months (computed at each pixel). The top part of the table shows results for the climate models by themselves while the bottom part of the table shows the results for climate and weather combined. We note the following: (1) Median temperatures rise over time, but there are key differences between the two emissions scenarios. In the ‘Paris Forever’ (PF) higher-emissions scenario, median temperature steadily rises over time, going from a base of 29.4oC for the nation in the 1981–​2000 period to 31.7oC by the 2060s, while the ‘2C’ lower-emissions scenario only reaches 30.7oC, which is less than one degree above the 30.0oC projection for the 2020s. This much of a temperature difference in the 2060s between the two emissions scenarios can be critical for crop yields and is indicative of the value of global GHG mitigation strategies. (2) There is considerable variation across WMAs in terms of median temperatures. This, in part, is due to much of the Western Cape sustaining peak rainfall in a different time of year than most of the rest of the country, but also reflects genuine differences in temperature across the nation. The coastal WMAs (Pongola, Mzimvubu, and Western Cape) have more moderate temperatures than the inland WMAs. (3) The uncertainty about future temperature is reflected in the range growing through time when looking at climate by itself. We also see that the WMAs on the coast have a lower range of uncertainty than the inland WMAs. (4) Taking weather into consideration, Table 16.1 shows that uncertainty and variability together are much larger in the inland WMAs than in the coastal WMAs. (5) For both climate only and climate and weather together, there is a higher range of uncertainty for the higher emission ‘PF’ scenario. Table 16.2 is similar to Table 16.1, except that it is focused on precipitation for the wettest three consecutive months in each pixel. This period is a reasonable proxy for the growing season for rainfed crops. The table shows regional differences in precipitation, with Pongola reflecting the highest rainfall, followed by Northeast, then Vaal and Mzimvubu close together, and finally Western Cape and Orange. As a nation, there is minimal projected change for median precipitation of all the climates. However, expansion of the range indicates rising uncertainty over time, with a larger increase associated with the higher emissions scenario. We also note some

Table 16.1: Mean daily maximum temperature for the warmest month in the wettest three consecutive months Year

Scen

Focus

Northeast range

Pongola (NE coast) range

Vaal range

Orange range

Mzimvubu (SE coast) range

Western Cape range

South Africa range

50

(5–​95)

50

(5–​95)

50

(5–​95)

50

(5–​95)

50

(5–​95)

50

(5–​95)

50

(5–​95)

1990

Base

C

30.8

0

27.9

0

31.6

0

32.1

0

27.5

0

24.1

0

29.4

0

2020s

2C

C

31.4

1.2

28.4

0.8

32.3

1

32.8

0.8

28

0.7

24.7

0.8

30

0.8

2020s

PF

C

31.6

1.2

28.6

0.9

32.6

1.1

33.1

0.9

28.3

0.8

24.9

0.8

30.3

0.9

2040s

2C

C

31.6

1.4

28.5

1

32.5

1.2

33

1

28.2

0.9

24.8

0.9

30.2

1

2040s

PF

C

32.2

1.8

29.1

1.3

33.2

1.6

33.6

1.3

28.7

1.2

25.3

1.2

30.8

1.2

2060s

2C

C

31.7

1.6

28.6

1.1

32.7

1.4

33.2

1.2

28.3

1

25

1

30.3

1.1

2060s

PF

C

32.9

2.4

29.6

1.7

33.8

2.2

34.3

1.7

29.3

1.5

25.9

1.6

31.4

1.6

2020s

2C

C&W

31.9

2.6

28.8

2.2

32.7

3.6

33.1

3.3

28.2

1.7

24.7

1.6

30.3

2.2

2020s

PF

C&W

32.2

2.6

29

2.2

33

3.7

33.3

3.3

28.5

1.8

24.9

1.7

30.6

2.2

2040s

2C

C&W

32.1

2.7

28.9

2.4

33

3.8

33.3

3.4

28.4

1.9

24.9

1.8

30.5

2.3

2040s

PF

C&W

32.7

2.8

29.5

2.5

33.6

3.9

33.9

3.5

28.9

2.1

25.4

2

31.1

2.5

2060s

2C

C&W

32.2

2.9

29.1

2.4

33.1

3.8

33.4

3.3

28.5

1.9

25

1.8

30.7

2.3

2060s

PF

C&W

33.4

3.3

30

2.7

34.3

4.1

34.6

3.5

29.5

2.3

26

2.2

31.7

2.6

Source: Authors’ calculations Notes: ‘C’ considers only the 7,200 climates per emissions scenario and reflects the uncertainty about the future climate. ‘C&W’ considers climate and weather together and draws on 720,000 combinations per emissions scenario and reflects both uncertainty about the future and annual variation in the weather. ‘Scen’ is used to mean emissions scenario. ‘1990’ represents the mean of the 1981–​2000 period. ‘2020s’ are values for 2021–​29; ‘2040s’ are for 2040–​48; ‘2060s’ are for 2060–​68.

Table 16.2: Total rainfall in the wettest three consecutive months Year

Scen

Focus

Northeast range

Pongola (NE coast) range

Vaal range

Orange range

Mzimvubu (SE coast) range

Western Cape range

South Africa range

50

(5–​95)

50

(5–​95)

50

(5–​95)

50

(5–​95)

50

(5–​95)

50

(5–​95)

50

(5–​95)

1990

Base

C

305

0

388

0

236

0

119

0

228

0

148

0

214

0

2020s

2C

C

304

89

384

97

233

39

112

43

228

41

147

24

213

26

2020s

PF

C

309

106

383

132

232

53

111

53

230

53

144

31

213

36

2040s

2C

C

306

109

385

118

233

48

111

52

229

49

147

30

213

32

2040s

PF

C

314

157

383

199

231

79

108

77

233

80

141

46

214

53

2060s

2C

C

308

122

386

132

233

53

111

58

230

55

146

34

214

36

2060s

PF

C

321

217

383

276

231

110

104

101

235

112

139

64

215

72

2020s

2C

C&W

302

197

384

282

225

184

104

156

224

141

144

84

208

121

2020s

PF

C&W

304

203

384

288

225

187

103

157

225

145

141

86

208

123

2040s

2C

C&W

304

201

385

273

230

193

106

163

229

150

145

92

211

123

2040s

PF

C&W

309

225

387

304

230

198

105

168

231

160

140

97

211

129

2060s

2C

C&W

307

214

382

297

226

177

103

150

228

152

145

90

210

118

2060s

PF

C&W

315

269

387

351

228

197

101

163

233

174

139

100

212

132

Source: Authors’ calculations Notes: ‘C’ considers only the 7,200 climates per emissions scenario and reflects the uncertainty about the future climate. ‘C&W’ considers climate and weather together and draws on 720,000 combinations per emissions scenario and reflects both uncertainty about the future and annual variation in the weather. ‘Scen’ is used to mean emissions scenario. ‘1990’ represents the mean of the 1981–​2000 period. ‘2020s’ are values for 2021–​29; ‘2040s’ are for 2040–​48; ‘2060s’ are for 2060–​68.

Climate Change and the Green Transition in South Africa    333 regional differences in median projected change with some WMAs getting slightly wetter and others getting slightly drier, but very little change in general, with the largest (but still small) changes appearing in the ‘PF’ scenario in the 2060s. It is important to highlight that many studies only focus on the uncertainty associated with climate but neglect to consider the variability of weather. When variability of weather is considered, it is possible to look at the occurrence of extreme events such as drought or floods. For example, for South Africa, what would be a one-​in-​a-​hundred-​ year flood event in the 2020s becomes a one-​in-​seventy-​three-​year flood event in the 2060s under the PF scenario. A one-​in-​a-​hundred-​year drought event in 2020s would become a one-​in-​forty-​two-​year drought event in the 2060s under PF. This is primarily due to the uncertainty in climate models, not because of increased variance in weather.

16.5  Impacts of Climate Change on Water and Food Systems 16.5.1 The Agricultural System Although only contributing about 3 per cent to the national GDP and 7 per cent of total employment, the agricultural sector has many knock-​on effects in the economy and is considered a critical sector in terms of future economic growth, job creation, and national food security. The spatial and seasonal variation in climate leads to a wide spatial heterogeneity in agricultural production. About one-​third of the country receives enough rainfall for successful rainfed crops. The remaining two-​thirds is used for grazing and some irrigated agriculture. In winter, frost is common on the great plateau, limiting the choice of crops and resulting in strong seasonal patterns for most crops grown (FAO 2016). The potential for future expansion in agriculture is limited as the

Legend Agricultural landuse Irrigated Area Rainfed Area Water Managed Area (non-irrigated) Country Boundary

0

380

760

1520 Km

Figure 16.4  Cultivated land use by agricultural system in southern Africa Source: International Water Management Institute’s (IWMI) Irrigated Area Map.

334    ARNDT, GABRIEL, Hartley, STRZEPEK, AND THOMAS majority of the arable land in the country is already cultivated either through large-​scale commercial farming or small-​scale subsistence farming. Agriculture currently utilizes about 63 per cent of available water resources. Irrigation is concentrated on approximately 19 per cent of all cultivated area and is distributed among the six hydroclimatic zones as follows: (1) Northeast: 22 per cent; (2) Pongola: 9 per cent; (3) Vaal: 14 per cent; (4) Orange: 11 per cent; (5) Mnmvubu: 13 per cent; (6) Western Cape:  31 per cent. This generally follows the distribution of surface water resources. There are only a few areas where there is surplus water to support increases in irrigation. In other areas of the country, increases in the demand for water from other sectors such as mining and urban consumption could drive a reduction in the allocation of water to agriculture. While only accounting for a small percent of harvested area, irrigated lands account for 48 per cent of crop revenues (Cullis et al. 2015). The agricultural sector is considered to be one of the most critical sectors in terms of potential impacts of climate change (DEA 2011; Schulze 2010). Agriculture is impacted directly by changes in precipitation, temperature, and evaporation. The impact of climate change, however, could be different for different crop types. This is due either to differences in the response of different crop types to changes in climatic variables as well as differences in the spatial impacts of climate change relative to where the different crops are currently grown or will be grown in the future. Rainfed agriculture is particularly sensitive to climate variability. For example, an analysis of the sensitivity of agriculture production to historical changes in climate in South Africa (Blignaut 2009) found that a 1 per cent decline in rainfall resulted in a decline in maize production of 1.16 per cent and a decline in wheat production of 0.5 per cent. While some crops may suffer, other crops such as sugar cane may benefit from both an increase in precipitation and an increase in temperature (Schulze 2010). In all cases, it is anticipated that future increases in temperature and evaporation will result in an increase in irrigation demands across the country. Climate change is thought to also have an indirect impact on crop production through impacts on pests and diseases (Schulze 2012).

16.5.2 Crops Yields 16.5.2.1 Methodology Crop models are often used to study the impact of climate change on crop yields (Rosenzweig et al. 2014). One of the most used models is the DSSAT model, which we use in this study. The DSSAT model considers daily weather and farming input use such as seed variety, planting date, and fertilizer use, and ‘grows’ the crop in daily time increments, keeping track of soil moisture and nutrients. While the software performs very well for many crops, the computer time is substantial, and would overwhelm the computation of yields based on 30km-​level pixels of daily weather data for southern Africa for 720,000 different weathers per emissions scenario.

Climate Change and the Green Transition in South Africa    335 To ease the computational demand, we do two things. First, we select a subsample from the 720,000 that exactly reproduces the first, second, and third moments of the distribution of key climate/​weather variables relevant to crop production (Arndt, Fant, Robinson, and Strzepek 2013). This reduces the number of climate–​weather pairs for each emission scenario from 720,000 to about 455. Each climate–​weather pair spans fifty years (from 2020 to 2069). Second, we build a crop yield emulator, which takes yields generated by DSSAT for a subsample of the data, and regress them on monthly weather values to generate parameters that can be used to quickly estimate yields for that sample.

16.5.2.2 Result There are no official agricultural statistics data on area harvested by hydrozone. So, we turn to a gridded dataset, MapSPAM, which uses 10km pixels and projects the amount of harvested land by crop type, based on a number of geographic features and national and subnational agricultural statistics (You, Wood, Wood-​Sichra, and Wu 2014). When we focus on rainfed maize in South Africa—​the most important crop by harvested area—​we find that of the roughly 2.3 million hectares cultivated circa 2010, 1.6 million of those are in the Vaal hydrozone, another 400,000 are in the Northeast, the Orange has around 140,000 hectares and Pongola has 120,000 hectares. MapSPAM shows very little rainfed maize being cultivated in either Western Cape or Mzimvubu. For the nation as a whole, climate change will result in a reduction in rainfed maize yields, but by only modest amounts at the median of the climate projections. By the 2060s, the median under the lower emissions scenario is reduced by only 3.5 per cent while under the higher emissions scenario the median declines by 7.4 per cent. Even in the 2020s, the emissions scenarios show a median reduction of 2.1 per cent for the lower emissions scenario and 1.2 per cent for the higher emissions scenario (the emissions scenario assumptions take different pathways, so the higher emissions scenario has a better outcome for the 2020s but then is worse by the 2060s). The range of outcomes for the middle 90 per cent widens over time, though it contracts between the 2040s and the 2060s for the lower emissions scenario. A declining median yield and a widening range indicate that the frequency of years with low production will increase in the later years, and this is especially so for the higher emissions scenario. This suggests that lowering GHG emissions globally (and domestically) would reduce the incidence of bad years and potentially reduce food insecurity or at least the need for social assistance for households relying on their own production for income and food. The Vaal hydrozone, the main area for rainfed maize production, will be especially hard hit. Even in the 2020s, the Vaal should see a loss of 6.0 and 4.8 per cent (depending on emissions scenario) relative to the climate of 1981 to 2000. By the 2040s, the losses will be 8.5 and 7.5 per cent, and by the 2060s, 6.8 and 11.2 per cent. The national average would be much lower, except that losses in Vaal will be partially offset by gains in production due to climate change in Pongola, and more modest losses in the Orange in particular but also in the Northeast.

336    ARNDT, GABRIEL, Hartley, STRZEPEK, AND THOMAS

16.5.3 Irrigation Demand 16.5.3.1 Methodology Irrigation demand estimates are comprised of three parts: (1) crop water requirements per hectare by location; (2) crop areas by location; and (3) irrigation system efficiency (field and conveyance). The crop water requirements were estimated using the IRRDEM model (Cullis et al. 2015) at the quaternary catchment level based on effective precipitation estimated using the USDA methodology (Hess 2010), potential evapotranspiration estimated using the Modified Hargreaves (Droogers and Allen 2002), and crop areas and irrigation efficiencies by crop from RSA data (Cullis et al. 2015). The IRRDEM uses monthly precipitation and temperature data times series of fifty years with 455 realization for each of the two emission scenarios described above.

16.5.3.2 Result The analysis provides three key findings for the nation as a whole: (1) there is a progressive increasing irrigation demand over time; (2) there is a consistent increase in the potential extreme amounts of irrigation demand; and (3) there is significant difference between the PF and 2C scenarios output with PF exhibiting increases in extreme events and a slight increase in mean irrigation demand compared with 2C.

16.5.4 Water Resources Streams and rivers are South Africa’s main water resource. Streamflow accounts for over 97 per cent of exploitable water resources. Streamflow is diverted directly or stored in reservoirs for use in economic activities or for ecosystem preservation. Streamflow is the collection of ‘runoff ’ in basin catchments. Runoff is the result of rainfall that is not consumed by vegetation or percolated to the groundwater. Runoff is a function not only of precipitation but also temperature and land surface slope, land cover, and land use. Since rainfall and temperature are highly variable over South Africa, runoff is as well and may not be available when/​where it is most in demand. Since climate change impacts are not uniform over South Africa, it is important that we model the catchment runoff process at an appropriately detailed spatial level.

16.5.4.1 Methodology The Pitman model (Pitman 1973)  was used to determine a time series of monthly catchment runoff for all quaternary catchments in South Africa for all climate futures considered. The Pitman model has become one of the most widely used monthly time step rainfall-​runoff models within southern Africa (Hughes et al. 2006) and is still the basis for current models used for all water resources planning studies in South Africa (Pitman et al. 2006).

16.5.4.2 Results Figure 16.5 provides probability density functions (pdf) of the total runoff summed over South Africa for four eras: (1) historic (1980–​2000); (2) 2025 (2020–​30); (3) 2045 (2041–​50);

Climate Change and the Green Transition in South Africa    337

Figure  16.5  Estimated density functions of the total runoff in South Africa for the periods Base, 2025, 2045, and 2065 under a) Paris Forever (PF) scenario and b) 2C scenario. Source: Authors’ calculations. Notes: The Base period refers to the historic period 1980–​2000; period 2025 refers to 2020–​30; 2045 to 2041–​50; and 2065 to 2061–​70.

338    ARNDT, GABRIEL, Hartley, STRZEPEK, AND THOMAS and (4) 2065 (2061–​70). Figures 16.5a and 16.5b display the results for the PF and 2C scenarios, respectively. Figure 16.5 reveals two key findings: (1) under the PF scenario there is a significant increase in mean and high extreme runoff over time, suggesting flooding is a potential risk of climate change and (2) there is significant difference between the PF and 2C scenarios output with PF exhibiting large increases in the tails of the distribution and a significant increase in the mean while 2C has a mild increase in tails and a slight increase in the mean. There is a very wide range of climate impacts to runoff across the six hydroclimatic zones. By the 2060s, the higher emissions scenario for the Western Cape is projected to a median increase of 54 per cent while the Vaal decreases by 9 per cent and the Orange increases by 1 per cent. The lower emissions scenario projection for the Western Cape is a median increase of 28 per cent with the Vaal decreasing by 7 per cent and the Orange increasing by 1 per cent.

16.5.5 Synthesis of Impacts of Climate Change on Water and Food Systems By 2060, decreases in rainfed food production are expected under both scenarios, which will increase the demand for irrigated crop production. Water demand per hectare is shown to increase under both scenarios; so, more water will be needed to produce the current irrigated production. With increased irrigated area, total demand for irrigation water will increase significantly. Agricultural water use accounts for approximately 63 per cent of South Africa’s water withdrawals, while municipal accounts for 27 per cent and industry 10 per cent. Municipal and industrial (M&I) water use is projected to increase with population and economic growth. There are currently local and regional water scarcity issues. With the economic value of water to M&I many times higher than almost all agricultural uses, reallocation of limited supplies has been putting pressure on agricultural water supplies. These pressures will be accentuated by climate change. With declining rainfed crop yields and reduced water supplies for irrigation, considerations for food security arise. Looking broadly across all of South Africa, increases in annual runoff appear to be more likely than decreases. However, the probability of decreases in runoff at the national level remains substantial, and it is highly likely that runoff will decrease in at least some specific geographies. In addition, the spatial and temporal distribution of projected median runoff increases do not directly match the locations of increased irrigation demands. Finally, the reservoir storage infrastructure does not exist to store these additional flows. Due to environmental and budget constraints, it will be difficult to increase reliable water supply via reservoirs or inter-​basin transfers even if streamflow increases.

Climate Change and the Green Transition in South Africa    339 In terms of water resources, the principal conclusions from the LTAS study appear to remain valid. Under LTAS, a national water resources model was developed to provide a modelling platform for assessing climate change impacts based on a selected combination of climate change scenarios, water requirement projections, water resources infrastructure developments, and system operating rules. This analysis suggested: • For South Africa as a whole the impact of climate change on urban and industrial water supply is fairly limited. This is specifically notable in the larger systems like the Upper Vaal, Crocodile (West), and Mooi-​Mgeni (in the Mvoti and Umzimkulu WMA) that, combined, supply water to more than 50 per cent of the country’s urban and industrial water-​users. • The agricultural sector may, however, be more vulnerable with a significant number of future climate sequences analysed, resulting in lower water supply characteristics compared to the baseline. Most notable in this regard is the Luvuvhu and Letaba, the Crocodile West and Marico, the Olifants, the Inkomati, and the Mzimvubu to Keiskamma WMAs. Looking forward, an enhanced national water model is under development at a detailed catchment level to understand the local impacts and potential need for additional storage and inter-​basin transfer infrastructure to utilize the additional runoff. The potential for substantial increases in runoff in particular zones brings with it increased flooding risk. This risk may pose a substantial climate risk to the South African economy. Recent studies in Mozambique (Arndt et al. 2019b) and Uganda (Strzepek et al. 2016) show that increases in flooding due to climate change result in losses of valuable capital and output disruptions. Flooding analysis requires modelling at a daily time step, which is part of an extension to current climate impact studies.

16.6  The Transition to a Low-​carbon Economy The South African government has adopted the United Nations Environmental Programme’s definition of a green economy which is defined as a ‘system of economic activities related to the production, distribution, and consumption of goods and services that result in improved human well-​being over the long term, while not exposing future generations to significant environmental risks or ecological scarcities’ (NDC 2012). As such, a key part of transitioning South Africa to a green economy involves decreasing the carbon intensity of the economy. In line with this goal, and driven by the need for climate response, the South African government has committed to reducing emissions to between 398 and 614 MtCO2e by 2030 and between 212 and 428 MtCO2e by 2050 relative to a business-​as-​usual reference case (DEA 2015). To align the

340    ARNDT, GABRIEL, Hartley, STRZEPEK, AND THOMAS country’s contribution to limit warming to 1.5°C, emissions need to fall to below 348 and 224 MtCO2e by 2030 and 2050 respectively. This would require more ambitious mitigation efforts (Climate Transparency 2020). The energy sector, which includes electricity and liquid fuels, accounts for nearly 80 per cent of total emissions and is key to reducing emissions in the country. The electricity sector on its own is responsible for half of South Africa’s emissions. Since 2007, the country has been plagued by load-​shedding due to lack of investment leading to capacity overuse, inadequate maintenance, and issues with coal supplies. Load-​shedding is estimated to have reduced GDP by Rand 338 billion over the past ten years (CSIR 2020). The construction of two mega power plants, Medupi and Kusile, has provided little relief to the constrained electricity sector due to significant delays and large cost overruns combined with design and technical problems (Bolhmann et al. 2016). The failures of Medupi and Kusile place increased pressure on ageing power capacity to meet demand. Much of South Africa’s coal power capacity will reach the end of its designed life within the next decade. The cost overruns at Medupi and Kusile, estimated at between Rand 300 billion and Rand 480 billion, combined with design problems which prevent them from running at planned capacity, contributed to rising electricity costs, which in real terms increased by 114 per cent between 2008 and 2017 (Deloitte 2017; Bolhmann et al. 2016). Given ageing infrastructure and the failure of existing capacity to meet demand, renewable energy technologies provide South Africa with an opportunity to significantly reduce its emissions and increase power supply in short time frames without hindering economic development (Arnd et al. 2017, 2019a; Merven et al. 2019). Studies have also highlighted that, because of favourable endowments of sun and wind and the drastically reduced cost of renewable power generation, such shifts in the energy system have a net positive impact on the economy, including employment (Inglesi-​Lotz 2016; CSIR 2018; Hartley et al. 2019; Merven et al. 2019, 2020; McCall et al. 2019). The combination of domestic and global shifts away from fossil fuels will have a negative impact on demand for South African fossil-​fuel production and exports, notably coal. Negative impacts on fossil-​dependent sectors underscores the need for a planned and coordinated transition to mitigate the impact on vulnerable workers and communities. One of the key lessons from past transitions is the early implementation of steps to prepare and mitigate the shock of the transition (Caldecott 2017) as disorderly transitions can have devastating impacts on workers and communities (see Marais 2013; Strambo et al. 2019). Given South Africa’s high level of unemployment, poverty, and inequality, there is an imperative that the low-​carbon emission transition be just, ensuring social interventions to protect livelihoods, especially for the most vulnerable (NPC 2019). Under the green transition this protection extends beyond the impacts of mitigation efforts and includes those affected by climate change. The coal transition is the first of a set of transitions that South Africa will have to make on the path to a low-​carbon and green economy and will thus set a precedent for the transition of other sectors. Transitions in transport and refinery production already appear to be nearing. Alternatives, like electrified transport, are both better for

Climate Change and the Green Transition in South Africa    341 the environment and are rapidly becoming a better deal for the consumer (Ahjum et al. 2020). Compared with coal, such shifts have larger implications for producers and consumers as the transport sector shares more linkages with the rest of the economy and affects more workers. Furthermore, relative to coal production and electricity generation, decision-​making in the transport sector broadly defined is far more diffuse, implying a need to manage by indirect manipulation of incentives (as compared with directly influencing the decisions of Eskom, the electricity generation and distribution parastatal). Numerous government plans have been developed to guide the green and low-​ carbon transition. These include the Green Accord, the National Development Plan, the National Strategy for Sustainable Development and Action Plan, and more recently the first Low Emission Development Strategy, which targets a net-​zero economy by 2050 (see Gupta and Laubscher 2018 for a review; DEFF 2020). The Green Economy Inventory for South Africa (PAGE  2017) identifies thirty-​two green-​related policies and over a thousand initiatives which have been concentrated in the energy, transport, and agriculture sectors and provinces of KwaZulu-​Natal, Western Cape, and Gauteng. Initiatives have been primarily funded by the public sector. The agriculture, food production, fisheries, and forestry sectors were found to be key for employment creation. While significant progress has been made over the past decade, a broadening and ramping-​up in activity is needed to reach committed targets, and more ambitious decarbonization levels, whilst promoting economic growth and achieving sustainable development goals. The 2018 Green Economy Barometer for South Africa (Amis et al. 2018) rates South Africa at 50 per cent in its green transition, highlighting that the country remains locked in ‘brown energy’ as government fails to fully commit to the transition away from coal. No detailed coal transition plan exists for South Africa (Climate Transparency 2020). Political will and leadership have often been cited as the key constraints to the shift away from coal (PAGE 2017; Amis et al. 2018; NPC 2019; Renaud et al. 2020). There are, however, segments within government that have progressed on initial steps to developing a just transition plan, for coal and the broader economy. This has included the two-​year Social Partner Dialogue for a Just Transition held by the National Planning Commission (Burton et al. 2018), the development of the Presidential Climate Change Coordinating Commission, and the development of the Eskom Just Energy Transition Office. Technology choices need to account for the impact of changing climate domestically and in the region. For example, SAPP (2017) estimates that from the mid to late 2020s South Africa will become a net importer of electricity from the Southern African Power Pool, which will source nearly a third of its electricity from hydropower by 2040. Climate change effects, particularly on runoff, can influence the reliability/​availability of hydropower generation. Thus, policies and plans necessarily need to consider the uncertainty and risks associated with climate change. The transition plan must account for the potential changing climate landscape and the uncertainties related to these in its solutions to ensuring continued economic development that is inclusive and green. As highlighted, the success of a just transition to a low-​carbon and green economy

342    ARNDT, GABRIEL, Hartley, STRZEPEK, AND THOMAS will require the efforts of all, including civil society, business, government, labour, and communities.

References Ahjum, F., Catrina Godinho, Jesse Burton, Bryce McCall, and Andrew Marquard. 2020. ‘A low-​ carbon transport future for South Africa: Technical, economic and policy considerations’. Policy Paper, March, Climate Transparency, Cape Town, South Africa. Amis, A. Mao, Gaylor Montmasson-​Clair, Sonwabile Lugogo, and Emily Benson. 2018. ‘The green economy barometer South Africa’. Green Economy Coalition, London. Arndt, Channing, Douglas Arent, Faaiqa Hartley, Bruno Merven, and Alam H. Mondal. 2019a. ‘Faster than you think:  Renewable energy and developing countries’, Annual Review of Resource Economics, 11: 149–​68. Arndt, Channing, Paul Chinowsky, Charles Fant, Sergey Paltsev, C. Adam Schlosser, Kenneth Strzepek, Finn Tarp, and James Thurlow. 2019b. ‘Climate change and developing country growth:  the cases of Malawi, Mozambique, and Zambia’, Climatic Change, 154(3):  335–​49. Arndt, Channing, Charles Fant, Sherman Robinson, and Kenneth Strzepek. 2013. ‘Informed selection of future climates’, Climatic Change, 130: 21–​33. Arndt, Channing, Mackay Miller, Finn Tarp, Owen Zinaman, and Douglas Arent (eds). 2017. The Political Economy of Clean Energy Transitions. Oxford: Oxford University Press. Blignaut, James, Liza Ueckermann, and James Aronson. 2009. ‘Agriculture production’s sensitivity to changes in climate in South Africa’, South African Journal of Science, 105: 61–​8. Bohlmann, Jessika, Heinrich Bohlmann, Roula Inglesi-​Lotz, and Jan van Heerden. 2016. ‘An economy-​wide evaluation of new power generation in South Africa: The case of Medupi and Kusile’, Energy Policy, 97: 450–​60. Burton, Jesse, Tara Caetano, and Bryce McCall. 2018. ‘Coal transition in South Africa:  Understanding the implications of a 2oC-​compatible phase-​out plan for South Africa’. IDDRI and Climate Strategies, Paris and London. Caldecott, Ben, Oliver Sartor, and Thomas Spencer. 2017. ‘Lessons from previous “coal transitions’’. Part of ‘Coal Transitions: Research and Dialogue on the Future of Coal’ Project. IDDRI and Climate Strategies, Paris and London. Carbon Brief. 2018. ‘The Carbon Brief profile: South Africa, country profile’. October [online] https://​www.carbonbrief.org/​the-​carbon-​brief-​profile-​south-​africa. Cole, Megan J., Richard M. Bailey, James D. S. Cullis, and Mark G. New. 2018. ‘Spatial inequality in water access and water use in South Africa’, Water Policy, 20(1): 37–​52. CSIR [Council for Scientific and Industrial Research]. 2018. ‘Economic impacts of South Africa’s energy mix’. Energy Centre, CSIR, Pretoria. CSIR [Council for Scientific and Industrial Research]. 2020. ‘Setting up for the 2020s—​ Addressing South Africa’s electricity crisis and getting ready for the next decade’. Energy Centre, CSIR, Pretoria. Cullis, James et al. 2015. ‘An uncertainty approach to modelling climate change risk in South Africa’. WIDER Working Paper no. 2015/​045, Helsinki. DEA (Department of Environmental Affairs). 2015. ‘South Africa’s intended determined contribution.’ DEA, South Africa.

Climate Change and the Green Transition in South Africa    343 DEA (Department of Environmental Affairs). 2011. South Africa’s Second National Communication under the United Nations Framework Convention on Climate Change. Department of Environmental Affairs, Republic of South Africa, Pretoria, RSA DEADP (Department of Environmental Affairs and Development Planning). 2014. Integrated Energy and Climate Change Strategy 2014 Province of the Western Cape Department of Environmental Affairs and Development Planning. http://​www.gardenroute.gov.za/​wp-​ content/​ uploads/​2018/​10/​GRDM-​Integrated-​Energy-​ and-​ C limate-​ C hange-​ Strategy-​ Document-​PDF.pdf. DEDEA (Department of Economic Development and Environmental Affairs). 2011. Eastern Cape Climate Change Response Strategy, Province of the Eastern Cape Department of Economic Development and Environmental Affairs. DEFF (Department of Environment, Forestry and Fisheries). 2020. ‘South Africa’s low emission development strategy 2050’. DEFF, Pretoria. Deloitte. 2017. ‘An overview of electricity consumption and pricing in South Africa: An analysis of the historical trends and policies, key issues and outlook in 2017’. South Africa: Deloitte. Deressa, Temesgen, R. Hassan, and Daneswar Poonyth. 2005. ‘Measuring the impact of climate change on South African agriculture: The case of sugarcane growing regions’. Agrekon, 44(4): 524–​42. Droogers, Peter, and Richard G. Allen. 2002. ‘Estimating reference evapotranspiration under inaccurate data conditions’, Irrigation and Drainage Systems, 16(1): 33–​45. DWA (Department of Water Affairs). 2013. ‘Climate Change Adaptation Strategy for Water’. DWA Report 000/​00/​18012/​5. Pretoria. Earthlife Africa and Oxfam International. 2009. ‘Climate change, development and energy problems in South Africa: Another world is possible’. https://​www.oxfam.org/​en/​research/​ climate-​change-​development-​and-​energy-​problems-​south-​africa-​another-​world-​possible Gbetibouo, Glwadys A., Claudia Ringler, and Rashid Hassan 2010. ‘Vulnerability of the South African farming sector to climate change and variability: An indicator approach’. Natural Resources Forum. Vol. 34. No. 3. Oxford: Blackwell Publishing Ltd. Gupta, Kapil, and R. F. Laubscher. 2018. ‘South African government initiatives toward a transition to green economy’, in Murat Â. Yülek (ed.) Industrial Policy and Sustainable Growth. Singapore: Springer. Hartley, Faaiqa, Bruno Merven, Channing Arndt, and Gregory Ireland. 2019. ‘Part 2: Quantifying the macro and socio-​economic benefits of a transition to renewable energy in South Africa’. Working Paper no. 26. Southern Africa—​Towards Inclusive Economic Development. [online] https://​sa-​tied.wider.unu.edu/​sites/​default/​files/​pdf/​SATIED_​ WP26_​January_​2019_​Hartley_​Merven_​Arndt_​Ireland.pdf. Hess, Tim, 2010. ‘Estimating green water footprints in a temperate environment’, Water. 2(3): 351–​62. Hughes, Dennis A., Lotta Andersson, Julie Wilk, and Hubert H. G. Savenije. 2006. ‘Regional calibration of the Pitman model for the Okavango River’, Journal of Hydrology, 333(1): 30–​42. Inglesi-​ Lotz, Roula. 2016. ‘The impact of renewable energy consumption to economic growth: A panel data application’, Energy Economics, 53: 58–​63. IRENA (International Renewable Energy Association). 2016. REmap:  Roadmap for a Renewable Energy Future, 2016 Edition. International Renewable Energy Agency (IRENA), Abu Dhabi, www.irena.org/​remap.

344    ARNDT, GABRIEL, Hartley, STRZEPEK, AND THOMAS Kusangaya, Samuel, Michael Lynn Warburton, Emma Archer Van Garderen, & Graham P.W. Jewitt. 2014. ‘Impacts of climate change on water resources in southern Africa: A review’, Physics and Chemistry of the Earth, Parts A/​B/​C, 67: 47–​54. L-​DEDET. 2016. Limpopo_​Climate_​Change-​Response_​Strategy_​-​2016_​2020, Department of Economic Development, Environment and Tourism:  Limpopo Province. http://​www. ledet.gov.za/​wp-​content/​uploads/​2016/​11/​Limpopo_​Climate_​Change-​Response_​Strategy_​ -​2016_​2020_​Final.pdf. Marais, Lochner. 2013. ‘The impact of mine downscaling on the Free State goldfields’, Urban Forum, 24: 503–​21. McCall, Bryce, Jesse Burton, Andrew Marquard, Faaiqa Hartley, Fadiel Ahjum, Gregory Ireland, and Bruno Merven. 2019. ‘Least cost integrated resource planning and cost-​optimal climate change mitigation policy—​Alternatives for the South African electricity system’. Working Paper no. 29. Southern Africa—​Towards Inclusive Economic Development. [online] https://​sa-​tied.wider.unu.edu/​sites/​default/​files/​pdf/​SATIED_​WP29_​February_​2019_​ McCall_​Burton_​Marquard_​Hartley_​Ireland_​Merven.pdf. Merven, Bruno, Faaiqa Hartley, Bryce McCall, Jesse Burton, and Jules Schers. 2019. ‘Improved representation of coal supply for the power sector for South Africa.’ Working Paper no. 24. Southern Africa—​ Towards Inclusive Economic Development. [online] https://​sa-​tied. wider.unu.edu/​sites/​default/​files/​pdf/​SATIED_​WP84_​Merven_​Hartley_​McCall_​Burton_​ Schers_​October_​2019.pdf. Merven, Bruno, Faaiqa Hartley, Sherman Robinson, and Channing Arndt. 2020. ‘Modelling the costs of constraining the transition to renewable energy in South Africa’. Working Paper no. 102. Southern Africa—​Towards Inclusive Economic Development. [online] https://​sa-​ tied.wider.unu.edu/​sites/​default/​files/​images/​SA-​TIED_​WP102.pdf. NPC (National Planning Commission). 2019. ‘2050 vision and pathways for a just transition to a low carbon, climate resilient economy and society’. Draft Proposal V2. Social Partner Dialogue for a Just Transition. [online] https://​www.nationalplanningcommission.org.za/​ assets/​Documents/​Vision%20and%20Pathways%20for%20a%20Just%20Transition%20 to%20a%20low%20carbon%20climate.pdf. PAGE. 2017. ‘Green economy inventory for South Africa: An overview’. PAGE, Pretoria. Pitman, W. V. 1973. ‘A mathematical model for generating monthly river flows from meteorological data in South Africa’. Report no. 2/​73, Hydrological Research Unit, University of the Witwatersrand. Pitman, W. V., A. K. Bailey, and J. P. Kakebeeke. 2006. ‘WRSM2000 water resources simulation model for Windows: Users guide’. Prepared for the Department of Water Affairs and the Water Research Commission. Renaud, Celeste, Emily Tyler, Adam Roff, and Grové Steyn. 2020. ‘Accelerating renewable energy industrialisation in South Africa:  What’s stopping us?’, Meridian Economics, Cape Town, South Africa. Rosenzweig, Cynthia, Joshua Elliott, Delphine Deryng, Alex C. Ruane, Christoph Müller, Almut Arneth, Kenneth J. Boote, et al. 2014. ‘Assessing agricultural risks of climate change in the 21st century in a global gridded crop model intercomparison’, Proceedings of the National Academy of Sciences of the United States of America, 111(9): 3268–​73. SAPP. 2017. ‘Southern Africa power pool plan 2017’. [online] http://​www.sapp.co.zw/​sites/​default/​files/​SAPP%20Pool%20Plan%202017%20-​%20Executive%20Summary_​0.pdf. Schlosser, C. Adam, Andrei Sokolov, Kenneth Strzepek, Timothy S. Thomas, Xiang Gao, and Channing Arndt. 2020. ‘The changing nature of hydroclimatic risks across Southern Africa’.

Climate Change and the Green Transition in South Africa    345 SA-​TIED Working Paper no.  101, March [online] https://​sa-​tied.wider.unu.edu/​sites/​default/​files/​images/​SA-​TIED_​WP101.pdf. Schulze, Roland E. 2010. Atlas of Climate Change and the South African Agricultural Sector: A 2010 Perspective. Department of Agriculture, Forestry and Fisheries, Pretoria, RSA, Chapter 7.4, 257–​66. Schulze, Roland E. 2012. ‘A 2011 perspective on climate change and the South African water sector’. WRC Report TT 518/​12. Pretoria: Water Research Commission. Sheffield, Justin, Gopi Goteti, and Eric F. Wood. 2006. ‘Development of a 50-​yr high-​resolution global dataset of meteorological forcings for land surface modeling’, Journal of Climate, 19(13): 3088–​111. Strambo, C., M. T. Aung, and A. Atteridge. 2019. ‘Navigating coal mining closure and societal change: Learning from past cases of mining decline’. SEI Working Paper, Stockholm Environment Institute, Stockholm. Strzepek, Kenneth, Brent Boehlert, Jacqueline Willwerth, and James Neumann. 2016. ‘The contribution of water resources development and environmental management to Uganda’s economy’. Prepared for the Uganda Ministry of Water and Environment. Industrial Economics Incorporated, Massachusetts, USA. [online] https://​www.mwe.go.ug/​sites/​ default/ ​ f iles/ ​ l ibrary/​ E conomic%20Study%202018-​ C ontribution%20of%20Water%20 Development%20and%20Environment%20Resources%20to%20Ugandas%20Economy_​ 0.pdf. Turpie, Jane, and Marine Visser. 2012. ‘The impact of climate change on South Africa’s rural areas.’ In Financial and Fiscal Commission. Submission for the 2013/​14 Division of Revenue. [online] http://​www.ffc.co.za/​index.php/​component/​docman/​doc_​download/​ 204-​submissionfor-​the-​2013-​14-​division-​of-​revenue-​3-​pdf-​reduced-​pdf?Itemid=120. Turpie, Jane, Harald Winkler, Randall Spalding-​Fecher, and Guy Midgley. 2002. ‘Economic impacts of climate change in South Africa: A preliminary analysis of unmitigated damage costs.’ Research Paper: Cape Town: University of Cape Town, Southern Waters Ecological Research & Consulting & Energy & Development Research Centre. You, Liangzhi, Stanley Wood, Ulrike Wood–​Sichra, and Wenbin Wu. 2014. ‘Generating global crop distribution maps: From census to grid’, Agricultural Systems, 127(May): 53–​60. Ziervogel, Gina, Mark New, Emma Archer van Garderen, Guy Midgley, Anna Taylor, Ralph Hamann, Sabine Stuart-​Hill, Jonny Myers, and Michele Warburton. 2014. ‘Climate change impacts and adaptation in South Africa’, Wiley Interdisciplinary Reviews: Climate Change, 5(5), 605–​20.

PA RT   I I I

T R A DE , I N DU ST RY, A N D R E G U L AT ION

Chapter 17

C orp orate Stru c t u re , I ndustrial Deve l opme nt, and Structura l C ha ng e i n Sou th A fri c a Pamela Mondliwa and Simon Roberts

17.1  The Importance of Structural Change and Corporate Structure for Industrial Development The South African economy remains highly concentrated in terms of the ownership and control of economic activity in many sectors being in the hands of a relatively small number of large corporations. At the same time, there have been important changes in the conglomerate nature and internationalization of the main businesses, alongside a growing importance of services relative to industrial activities. The orientation of the major corporations is critical to understanding the economy’s performance, including low levels of fixed investment and premature deindustrialization (Andreoni and Tregenna 2020, 2021). While financial services have grown strongly, they have neither created substantial employment nor supported higher levels of investment in the local economy. Within industrial activities, minerals-​related production has remained important in South Africa rather than the diversification into higher value-​added and more sophisticated areas of manufacturing which has been identified as critical for economic development (McMillan et al. 2014). Large and lead firms are clearly important in developing productive capabilities as they are able to make the investments necessary to realize economies of scale and scope, as well as make long-​term commitments in the learning and research necessary to build

350    Pamela Mondliwa and Simon Roberts capabilities required for industrial development (Lazonick 2010; Roberts 2020a). In many industries and sectors, they have key technologies, provide access to markets, and control material inputs, which shape the structure and performance of economies (Sutton 2012; Chandler 2001). The diversification strategies of large conglomerates have been linked to country success in East Asian industrializers in the development of technological capabilities and export competitiveness across industries (Amsden and Hikino 1994). Large and conglomerate firms were the locus for building capabilities to engage in technological upgrading within industries; entry into more complex activities; increasing local content involving local innovations and design; and mastering more complex technological tasks within industries (Lall 1993). By shifting resources into activities with superior scope for cumulative productivity increases, the firms led the structural changes which have been recognized as critical drivers of industrial development (Amsden 2001; Andreoni et al. 2021a). Large firms are also likely to have a degree of market power, the ability to govern value chains, and influence policy agendas and institutions (Dallas, Ponte, and Sturgeon 2019). Driving structural change therefore requires an understanding of how the configurations of economic power impact on the development of dynamic capabilities within and across firms (Pisano 2017; Mondliwa et al. 2021; Roberts 2020a). This includes the importance of competitive rivalry in the incentives to invest, innovate, and improve productivity as against the ability for entrenched dominant firms to protect their rents (Mathis and Sand-​Zantman 2014; Arrow 1962; Bloom, van Reenen, and Williams 2019; Buccirossi et al. 2013). Economic models which assume away scale economies and the embedded nature of technological change, and approach competition in terms of static allocative efficiency rather than dynamic capabilities, therefore do not engage with the core issues of the central role of major corporations in structural transformation. We thus need to squarely address what Lamoreaux (2019) terms the ‘problem of bigness’ and analyse the strategies of large firms in practice given their heterogeneity, and the relationship to the development of productive capabilities (Chandler, Amatori, and Hikino 1997; Chabane et al. 2006; Roberts 2020a; Bloom et al. 2019). These strategies are in turn influenced by the internal and external environment within which the firms operate, including the country’s policy environment, the global context, as well as the scope of the firm’s operations and shareholder interest. In this regard, vertical and conglomerate linkages are important factors influencing firm strategies and production is governed across different levels of supply and in related and unrelated activities. Value chain frameworks emphasize how the governance role played by lead firms shapes the distribution of value creation and capture as well as the capability-​upgrading prospects of firms (Gereffi and Lee 2016). These large and lead firms effectively shape markets, influencing what can be exchanged and on what terms, as well as the wider institutional and policy environment (Dallas et al. 2019). Changing models of shareholder ownership are also relevant as, with the rise of the ‘shareholder value movement’ and increased significance of institutional investors, short-​term financial performance can be privileged over longer-​term capabilities (Andreoni, Robb, and van Huellen 2021).

Corporate Structure, Industrial Development, and Structural    351 In section 17.2 we review the broad patterns in terms of industrial development and the significance of large companies in South Africa in the democratic period from 1994 to 2019. Section 17.3 assesses the restructuring of the large conglomerates in important sectors of the South African economy, the nature of internationalization of the large-​listed companies, and the implications for economic policies. Section 17.4 concludes.

17.2  Industrial Development and Structural Change in South Africa 17. 2.1 Overview of Patterns of Structural Change in Manufacturing and Services Successful industrial development, particularly in late industrializers, has been characterized by a high and sustained share of manufacturing in gross domestic product (GDP); a high rate of fixed investment in GDP, and rapid growth in the industry value-​ added and sophistication of manufactured exports. South Africa has performed poorly on all these benchmarks (see Chapter 24). Investment levels have been very poor, with total gross fixed capital formation being persistently lower than those of comparator countries (at similar levels of GDP per capita in the 1990s) such as Malaysia, Turkey, and Thailand (Andreoni et al. 2021a). In the period 1994 to 2019, the average capital formation as a percentage of GDP, was 18 per cent compared to the middle-​income-​country average of 28 per cent. The economy prematurely deindustrialized with the manufacturing contribution to GDP declining from 21 per cent in 1994 to 12 per cent in 2019 (Andreoni et al. 2021a; Andreoni and Tregenna 2020, 2021). Growth in manufacturing value a​dded has lagged GDP growth and has been substantially below most other upper-​middle-​income counterparts (Andreoni et  al. 2021a). Similarly, manufacturing exports have performed more poorly than most other middle-​income and upper-​middle-​income countries (Andreoni et  al. 2021a; ­Chapter 24 in this volume). Some of the weak manufacturing performance is due to the outsourcing of high value design, engineering, and marketing activities, meaning that these activities move to be captured in services even while reflecting advanced industrial capabilities. By examining trends in services and manufacturing, however, we observe that the strongest growth in value ​added has been in finance and insurance services, although with relatively little employment growth (Table 17.1). In manufacturing, there has been somewhat of a structural regression, as resource-​based sectors (also highly capital-​intensive, with very small employment shares) led by coke and petroleum refineries along with basic chemicals and basic iron and steel, have all performed better than the manufacturing

352    Pamela Mondliwa and Simon Roberts Table 17.1: Manufacturing and services performance: selected sectors Total employment

Value added

Growth

Growth

Share of Total, by broad sector

1994–​2019

1994

1994–​2019

Share of Total, by broad sector 1994

2019

2019

Coke and refined petroleum products

1.6%

1.1%

1.8%

5.0%

4.4%

9.3%

Basic chemicals

0.3%

1.5%

1.8%

3.0%

3.5%

4.6%

Plastics products

1.2%

2.6%

3.9%

1.5%

3.2%

3.0%

Basic iron and steel

-​3.3%

5.4%

2.6%

2.3%

3.8%

4.3%

Basic non-​ferrous metals

-​2.1%

1.8%

1.2%

1.7%

3.2%

3.1%

Metal products excluding machinery

-​0.2%

8.1%

8.6%

1.0%

6.9%

5.6%

Machinery and equipment

1.2%

6.3%

9.6%

2.3%

5.2%

5.7%

Motor vehicles, parts and accessories

-​0.4%

7.0%

7.1%

3.9%

4.4%

7.2%

0.1%

15.2%

17.5%

2.9%

11.4%

14.8%

Other diversified manufacturing

-​0.9%

51.1%

45.8%

0.9%

54.0%

42.5%

Total manufacturing

-​0.5%

100%

100% 1.9%

100% 100%

Wholesale and retail trade

3.0%

22.1%

26.6%

3.0%

19.7%

20.4%

Catering and accomm services

1.5%

5.6%

4.6%

3.2%

1.6%

1.1%

4.6%

3.0%

5.3%

1.6%

9.7%

9.3%

-​0.5%

2.0%

1.0%

2.9%

1.5%

4.3%

Food

Transport and storage Communication Finance and insurance

1.1%

4.6%

3.4%

7.6%

7.2%

10.1%

Business services

3.5%

15.1%

20.2%

4.5%

18.8%

21.6%

Government, community, and personal services

1.4%

47.7%

38.9%

3.6%

41.6%

33.1%

Total services

3.0%

100.0% 100.0% 3.0%

100% 100%

Source: Quantec, authors’ calculations. Note: Employment figures include formal and informal employment. Value-​added shares and growth rates are calculated from constant 2010 price series and growth rates as compound annual average growth rates.

average in terms of value added.1 Alongside this has been relatively strong growth in motor vehicles, reflecting the support provided through industrial policies, as well as food products. 1  This differs somewhat from the picture presented by Edwards (Chapter  21 in this volume). This is due to Edwards adopting a different classification for ‘non-​commodity’ manufactured exports which excludes what we term as diversified industries that have performed poorly, such as plastics; and different

Corporate Structure, Industrial Development, and Structural    353 There has also been outsourcing of low-​value security and cleaning services that are captured in the very heterogeneous business services category, which has recorded high rates of growth in value-​added and employment. Growth across most services has also seen relatively high rates of increased value-added in wholesale and retail trade, catering and accommodation, and communication services (Table 17.1). Given the high levels of market concentration in many manufacturing and services sectors (Buthelezi et al. 2019; Bell et al. 2018), the outcomes observed are largely the result of the decisions of relatively few firms, to which we turn now.2

17.2.2 Corporate Structure: Continuity and Change While South Africa has had sustained high levels of market concentration, the corporate structure has altered in fundamental ways with major implications for structural change and industrial development. In 1994, Anglo American was estimated to control businesses which accounted for 43 per cent of the capitalization of the Johannesburg Stock Exchange (JSE) and stretched across mining, banking, metals, chemicals, paper, sugar, beer, motor vehicles, and retail (Goldstein 2010). It undertook a series of restructuring and unbundling which appears to reflect a decline in the concentration of ownership and control of the economy as a whole, reflected in the dramatic decline in its share of JSE-​listed companies which it controlled (measured by capitalization) (Table 17.2). However, this did not mean concentration declined within specific markets and sectors (see Chabane et al. 2006). In some cases, the unbundling was accompanied by consolidation, rebundling, and increasing vertical integration within a sector consistent with persisting high levels of concentration at the level of different markets since 1994 (Mohamed 2020). For example, divesting of a company such as AECI in chemicals, Mondi in paper, or the Anglo American Bevcon stake in South African Breweries (SAB) did not make these companies any less significant in the explosives, paper, or beer markets respectively, or these markets less concentrated. As we consider in more detail in section 17.3, the second largest family-​owned conglomerate in 1994, Remgro, followed a quite different path and continued to invest in diverse South African businesses in food, beverages, technology, health-care, and logistics sectors (Goldstein 2010). While Anglo American itself became narrowly focused on mining in the post-​ apartheid period, its significance on the JSE increased from 2015 to 2019, reflecting the global operations across gold, coal, iron ore, platinum, and diamonds. There are also major southern African operations in each of these minerals except gold, consistent time periods where Edwards takes an earlier starting point in 1990 and therefore includes changes in the early 1990s where our focus is on the post-​apartheid period and on the continuity in the 2000s in particular. The growth in auto industry exports is a notable development in diversified manufacturing, which stems more from targeted and extensive industrial policy rather than simply trade liberalization. 2 

The high levels of concentration in South Africa have been widely observed, see also World Bank (2014, 2016), Thakoor (2020), and Dauda et al. (2019).

354    Pamela Mondliwa and Simon Roberts Table 17.2: Summary of control of JSE market capitalization (% of total) 1995–​ 2002 ANGLO AMERICAN CORP 24.8% BIDVEST GROUP BLACK GROUPS

2003–​07 2009–1​ 4 2015

2016

2017

2018

2019

20.0%

9.4%

1.6%

3.3%

3.6%

4.8%

6.8%

1.0%

1.0%

0.9%

1.0%

1.1%

1.2%

1.4%

1.3%

6.3%

5.5%

4.0%

0.6%

0.5%

0.7%

1.4%

1.2%

10.4%

7.2%

8.0%

11.3%

12.4%

11.2%

10.0%

7.6%

FOREIGN

4.7%

18.4%

30.8%

26.8%

42.0%

40.3%

25.1% 22.8%

INSTITUTIONS

4.9%

11.4%

19.0%

17.6%

18.1%

17.9%

30.8% 28.3%

INVESTEC

2.1%

0.9%

0.8%

1.0%

0.8%

0.7%

0.8%

0.7%

LIBERTY LIFE/​STD BANK

7.7%

4.0%

2.8%

2.1% 7.7%

6.9%

10.4%

9.9%

17.7%

10.8%

5.1%

2.9%

2.8%

2.1%

1.9%

1.4%

0.8%

0.6%

1.4%

1.5%

1.7%

2.0%

1.9%

DIRECTORS

NASPERS OLD MUTUAL (SA MUTUAL) PSG REMGRO (REMBRANDT)

9.8%

7.7%

6.6%

9.2%

7.2%

6.9%

7.1%

6.9%

RMB/​FIRSTRAND

3.0%

4.3%

2.9%

2.6%

2.4%

2.2%

3.3%

2.4%

SABMILLER

3.3%

4.9%

7.9%

12.5%

11.0%

2.3%

1.4%

1.3%

2.5%

4.4%

4.0%

SANLAM SASOL

Moved to Foreign, then to Institutions 1.2%

1.3%

1.7%

1.5%

Moved to Institutions

Source: McGregors Who Owns Whom Notes: Control is assessed by McGregor’s taking into account the various cross-​holdings of shares that exist and may be associated with a relatively small direct shareholding in any given company. The Black owned groups are identified as such by McGregor’s on the basis of all those companies which have significant black influence in their ownership. Naspers and PSG were too small to be separately identified by WOW in the earlier years.

with the continued importance of resource-​based products in South Africa’s merchandise exports. Minerals accounted for more than 35 per cent of merchandise exports in 2019, compared with just over 40 per cent in 1994 (Andreoni et al. 2021a). The relative unimportance of companies controlled by Black South Africans is an element of continuity (see ­Chapter 28 in this volume). After the initial Black Economic Empowerment (BEE) deals in the late 1990s, the share of these companies became negligible (Table 17.2; Chabane et al. 2006). Instead of opening up markets to challenge by independent Black-​controlled rivals, BEE became largely about sharing of rents, along with targets in skills development, management, and procurement (Vilakazi and Bosiu 2021; Mondliwa and Roberts 2020; Ponte et al. 2007). This has had the effect of reinforcing rather than changing the existing economic structure (Andreoni et  al. 2021b).

Corporate Structure, Industrial Development, and Structural    355 In terms of changes, there has been a substantial increase in firms where control is exerted by institutional shareholders, to a peak of 31 per cent in 2018. This mirrors the international trend of financialization (Andreoni, Robb, and van Huellen 2021). And, it was accompanied by growth in portfolio and FDI inflows, especially from 2005 (Bell et al. 2018). The capitalization of the JSE grew to the equivalent of almost three times South Africa’s GDP in 2007 and, in terms of this ratio, South Africa has been second only to Hong Kong from 2013 (Andreoni, Robb, and van Huellen 2021). The growth of foreign ownership reflected in Table 17.2 is, however, also due to deliberate outward internationalization strategies by the major South African companies (Goldstein 2010). Anglo American had been highly internationalized from an early age including through holdings of its Minorco business in Luxembourg (Mohamed 2020). In the post-​apartheid era, a number of companies were given permission to list overseas on the grounds that this would assist with raising capital for investment in South Africa. Instead, international expansions and mergers led to the companies becoming part of massive transnational corporations (TNCs) as in the case of South African Breweries (SAB) whose merger with ABInBev saw it become part of the world’s largest beer company. It was also the biggest JSE listed company in terms of its market capitalization in 2016, until it was eclipsed by the Prosus listing in 2019. We assess the implications of internationalization further in section 17.3. The growth in the value of listed firms with international operations needs to be born in mind when interpreting the percentage shares of other categories, whose shares decline simply because of the international operations reflected in, for example, the valuations of SABMiller and Prosus. This is why we move to consider the large companies by broad sector in section 17.3 and in terms of the top 100 companies by number, as follows.

17.2.3 Changes in the Top 100 Companies Looking at the top 100 listed companies, a major change had already taken place in the first decade of democracy (Chabane et al. 2006). In 1994, eighty-​three of the top 100 were owned or controlled by the top six conglomerates. In 2004, these conglomerates controlled forty-​seven. In 2020, it was less than twenty. This has coincided with major changes in the sectoral breakdown of the top 100 listed companies. In the first decade after 1994, the number of retail and other services companies in the top 100 increased as supermarkets, mobile phone and ICT companies, and health-care groups all grew on the back of consumer spending and middle-​class incomes, while the number of mining companies fell (Chabane et al. 2006). From 2004 to 2020, the big change has been in the growth in financial services companies (including property-​related funds) while companies classified as industrial activities declined substantially in number—​a complete reversal in importance of these two groupings (Figure 17.1). There were still seventeen mining companies in the top 100 in 2020 and the ongoing significance of these companies is reflected in the fact that ten of the seventeen mining

356    Pamela Mondliwa and Simon Roberts 1994 Other Retail, 7 services, 6

2020 (September) Financial services, 15

Retail, 16 Financial services, 38

Other services, 11 Mining, 31 Industrial, 41

Mining, 17 Industrial, 18

Figure 17.1  Sectoral composition of top 100 JSE-​listed firms Source: JSE. Note: Companies ranked by capitalization; share by sector is simple number of companies.

companies were in the top twenty (see Appendix Table).3 Four of these ten are Anglo American-​related—​Anglo American, AngloPlat, AngloGold Ashanti, Kumba.4 There are other strong elements of continuity in the other largest twenty listed businesses, with Mondi (unbundled from Anglo), two banks, and insurance companies (FirstRand, Standard, and Sanlam), SAB (now in the shape of ABInBev), and BATSA also figuring along with Richemont. Prosus and Naspers are also in the top five, while Vodacom is in the top twenty.

17.3  Large Firms and Structural Change In this section we consider in more detail how the orientation of the large businesses has impacted on the growth of major sectors, before analysing the different dimensions of internationalization of the businesses and their relationships with economic policy.

17.3.1 Key Sectors—​Industrial, Finance, Other Services 17.3.1.1  Industrial groups The continued importance of resource-​ based companies stands out, including mining, along with basic metals and basic chemicals companies. However, while these companies have accounted for substantial proportions of industry investments in South 3 

Counting Kumba as primarily a mining company. The other mining companies in the top twenty are BHP, Glencore, Gold Fields, Sibanye, Impala Plats, South 32 (spun off from BHP). 4 

Corporate Structure, Industrial Development, and Structural    357 Africa, they have not been the base for wider diversification. Their ongoing significance and growth internationally is therefore consistent with the low overall investment levels in the country and remains one of the core economic development challenges facing South Africa (Andreoni, Kaziboni, and Roberts 2021; Bell, Monaco, and Mondliwa 2021; Mondliwa, Roberts, and Ponte 2020; Zalk 2021). We distinguish three groups of large companies linked to South Africa’s resource-​ base which have followed different paths. The first group comprises the former state-​ owned firms in steel and petrochemicals (ISCOR and Sasol), which have followed different, firm-​specific paths. The second group comprises businesses which were spun off from the mining conglomerates and Anglo American, in particular, which failed to build on their capability base, including important businesses in the metals and engineering industries. The third group comprises those which have grown strongly since being divested. There is a fourth group of large industrial businesses not directly linked to resources, including large food businesses, and those in the Remgro stable. The basic metals and basic chemicals industries have developed around individual companies with strong links to mineral resources and coal-​based energy. These companies have changed substantially as a result of internationalization. In steel, formerly state-​owned ISCOR unbundled its mining operations (which evolved into the Anglo-​associated Kumba and Exxaro companies), consolidated the steel businesses through a series of mergers to substantially increase levels of industry concentration, and became a subsidiary of the world’s largest global steel producer ArcelorMittal (Andreoni, Kaziboni, and Roberts 2021). Investment and production decisions in ArcelorMittal South Africa (AMSA) are thus the result of the global strategy of the parent company, for which South Africa is a relatively insignificant market, with profits being extracted in good years and lobbying for state protection in poor years (Andreoni, Kaziboni, and Roberts 2021). Other basic metals producers in aluminium and stainless steel are also operations of TNCs, including BHP. These were established in the 1990s based on substantial investment incentives and the very cheap electricity at the time and are now legacy investments struggling to be competitive with much higher electricity prices (­Chapter 26 in this volume). A different picture is presented by the ultra-​ capital-​ intensive basic chemicals industries (Bell, Monaco, and Mondliwa 2021; Chapter 24 in this volume). These are dominated by Sasol (26th on the JSE in 2020), which is the largest producer of liquid fuels (from coal and gas) and has continued to benefit from a very favourable regulatory regime for fuels (Mondliwa and Roberts 2019). In contrast with ISCOR, Sasol has maintained and expanded its backwards integration into mineral feedstock, adding natural gas piped from Mozambique (Mondliwa and Roberts 2017). This base has enabled it to withstand losses from badly judged international investments. AECI (spun out from Anglo, at position 88 on the JSE) and Omnia which produce explosives, fertilizer, and other chemicals, source feedstock from Sasol and, in the form of explosives and water treatment chemicals, supply into mining. These companies have continued to invest, build capabilities and linkages, supplying markets across southern Africa from which base they have outwardly internationalized.

358    Pamela Mondliwa and Simon Roberts The second group of companies is exemplified by relatively labour-​ absorbing manufacturing of diversified metal products, and machinery and equipment that have not performed well overall (Andreoni et al. 2021; Bell et al. 2021; ­Chapter 24 in this volume). This includes engineering and metals businesses divested by Anglo, such as Boart Longyear and Dorbyl, where technical capabilities were lost (Mohamed 2020). While there are some relative successes in mining machinery, these are isolated examples of capabilities when set against the hollowing-​out of capabilities and the loss of dense networks of local linkages after the unbundling (Andreoni, Kaziboni, and Roberts 2021; Zalk 2017). By comparison, the third group of companies including AECI in chemicals, along with large businesses in paper and sugar, reflect large-​scale industries with two or three incumbent companies in entrenched positions, which have maintained and sometimes grown their positions in their respective industries. These included Sappi (from Sanlam), Mondi (from Anglo American) in paper and pulp; and Tongaat-​Hulett (from Anglo American) and Illovo (now owned by Associated British Foods) in sugar.5 These companies have been able to leverage their positions deriving from historic state support and industrial policies. In alcoholic beverages and cigarettes, SAB (now ABInBev) and BATSA (from the Anglo and Remgro stables) have literally held quasi-​monopolies over their products in South Africa and, at the same time, are the largest producers in the world.6 In spirits and other alcoholic beverages Distell (part of Remgro) is the largest producer in South Africa. Other large industrial companies in the top 100 listed firms include a range of food producers, led by Tiger Brands, AVI, and Oceana in the top 100, with RCL and Astral, the two largest poultry producers, just outside the list at positions 104 and 107 respectively. Food represents a sector where urbanization and growing local demand coupled with linkages into agriculture, packaging, logistics, and retail have been the basis for strong corporate groupings including in Remgro. Two food and logistics companies, Bidvest and Bidcorp (spun off from Bidvest in 2016),7 are also in the top 100, along with the major supermarket chains (Shoprite, PicknPay, Spar, and Woolworths). Other industrials in the top 100 companies are in a mixture of logistics and sales in the form of Barloworld, Imperial, Super Group, and diversified holding company KAP Industrial Holdings, with businesses across industrial production in auto components, bedding, timber products, and polymer chemicals.

17.3.1.2 Finance Finance companies in the JSE top 100 include banks, insurance, and property funds. Financial services can enable economic activity through intermediating funds and providing insurance and other services. However, in South Africa, there have been high 5  In sugar, Africa’s biggest producer, Illovo, was acquired by Associated British Food in 2016 (after a majority stake was bought in 2006). Illovo had been acquired from Tate & Lyle by CG Smith in 1977. 6  See, for example, estimates by Euromonitor. 7  Bidcorp has the bulk of its business in Europe and Australasia.

Corporate Structure, Industrial Development, and Structural    359 levels of mergers and acquisitions activity (including as FDI inflows and outflows), and trading in the stock exchange continued to grow while not enabling broader economic growth.8 This is in line with international experience that shows that capital flows are highly unstable and are not associated with productive investment (UNCTAD 2014). Outflows also continued to increase, consistent with the shifting emphasis by local firms to investing outside the country, at the same time as international investors continued to acquire local companies (see Bell et al. 2017; Bosiu et al. 2017; Mohamed 2020). Capital account liberalization allowed South African corporations to move capital abroad on a grand scale, both legally and illegally (see Mohamed and Finnof 2005; Ashman et al. 2011; UNCTAD 2020). As well as acting as the major financial centre in Africa, South Africa also has real estate funds with investments around the world. For example, Growthpoint Properties has assets in South Africa, Eastern Europe, Australia, and the United Kingdom, and Nepi Rockcastle is the owner of shopping centres in Central and Eastern Europe. Other companies in the top 100 include Capital & Counties which has properties in central London, and Sirius Real Estate has business parks, offices, and industrial complexes in Germany. These are effectively vehicles for South Africans to invest outside the country. They reflect the fact that low investment rates in South Africa are not related to low savings but poor prospects, given the low growth rates, weak demand, and lack of dynamism in the real economy.

17.3.1.3 Other Services In other services, we pick out three notable trends: the growth of retail chains, including the further supermarketization of the South African economy (das Nair 2019a); the growth in the value of private hospital and insurance groups, and the importance of media platforms and telecoms. The spread of shopping malls and formal retail space across urban areas in South Africa has seen the expansion of supermarkets and other retail groups to peri-​urban areas. This spread has been much more substantial than the regional expansion of South African supermarket chains to other countries in the continent (das Nair 2019a). The supermarket chains have developed formats and acquired smaller local chains to target lower-​income consumers (das Nair 2019b). These formats include U-​Save, Boxer, and Savemore stores of Shoprite, Pick n Pay, and Spar respectively. The large chains effectively govern consumer grocery spending and are the route to market for producers across southern Africa (das Nair 2017; das Nair, Chisoro, and Ziba 2018). The three main private hospital groups, Mediclinic, Life, and Netcare, grew rapidly in the first decade of democracy as they acquired independent hospitals (Competition Commission 2019). However, the largest medical scheme administrator, Discovery Health, is larger than them all at position 25 on the JSE. Discovery has built on its

8 

These developments in South Africa are assessed in detail in other chapters in this volume (see for example Mncube and Theron, C ­ hapter 25; Karwowski, C ­ hapter 47).

360    Pamela Mondliwa and Simon Roberts position and information base in medical insurance to be a major platform extending across financial services and has also internationalized. With its Vitality product (which it has launched internationally) Discovery collects data on exercise and eating habits to offer insurance, along with discounts on leisure and other products. With the advent of wearable technology to monitor people’s movement and vital signs, coupled with digitalization of purchases through the payments system, Discovery has been able to reward people on their decisions to, for example, purchase healthier foods in major supermarket chains. In 2017, it obtained a banking licence and started operating in 2019, marketing itself as the ‘world’s first behavioural bank’.9 While the two main mobile phone companies, Vodacom and MTN, are also very significant in their own right, at positions 10 and 22 on the JSE in 2020, it is the evolution of Multichoice/​Naspers which is most remarkable in the media and communication space. From the monopoly in satellite television under apartheid, through investments in newspapers in the Media24 group, Naspers has extended into digital platforms with the Takealot e-​commerce platform (Goga, Paelo, and Nyamwena 2019) and the investment in China’s Tencent which was spun off into the separately listed Prosus in 2019. Prosus was the largest stock on the JSE in 2020, by some margin. Discovery and Naspers demonstrate the importance of data, rather than physical capital, in building competitive capabilities in multi-​sided platforms (Roberts 2020a).

17.3.1.4 The Example of Remgro While Anglo American unbundled in the first decade of democracy, the second largest family conglomerate, the Rupert family’s Remgro Ltd continued to be a diversified conglomerate business (see also Goldstein 2010).10 It has invested and grown its activities, predominantly focused on the South African economy. Remgro was established in 1948 as a tobacco company, then called Rembrandt Group Ltd by Dr Anton Rupert. It was listed on the JSE in 1956. The Rupert family maintains a significant share of the voting rights and Johann Rupert, eldest son of Anton, is chairman of Remgro, along with Richemont, their Swiss luxury goods business. Remgro’s diversified nature, reflected in interests across food, beverages, technology, health-care, and logistics sectors, includes substantial stakes in RCL Foods, Unilever, Distell, DFA, Seacom, Grindrod, and Mediclinic. The company also set up a separate venture and capital growth subsidiary Invenfin, whose priority areas of investment are the food and technology sectors. Invenfin invests in companies that have acquired some traction in their markets with intellectual property and capabilities which set them to grow in line with global trends. It reflects a continued willingness to make longer-​term investment bets and to develop integrated businesses along value chains. The networks and integration across and along value chains is illustrated by RCL Foods. This company is involved in several food segments including poultry, sugar, milling,

9 

10 

https://​www.discovery.co.za/​bank/​bank-​healthier [Accessed on 2 October 2020]. This draws heavily from Mondliwa et al. (2017).

Corporate Structure, Industrial Development, and Structural    361 animal feed, and baking. RCL Foods is also integrated into distribution, while Remgro is invested in Grindrod in infrastructure and logistics. Remgro’s linkages to Export Trading Group (ETG) through its investment in the Pembani Remgro Infrastructure Fund also provide it with a critical link into regional agricultural markets. ETG is involved with purchasing from producers and trading agricultural products internationally, with a well-​established footprint in East and Southern Africa (Bosiu and Vilakazi 2020).

17.3.2 Internationalization The largest South African companies are all extremely international in scope. Far from being the result of substantial inward foreign direct investment, these are associated with outward expansions leveraging capital built up from domestic operations coupled with core capabilities. In some cases, the largest companies globally in a given industry have a South African business core. This is the case in beer and cigarettes, as highlighted above. It is also true in gold, platinum, coal, and diamond mining. In paper Mondi and Sappi are among the world’s largest paper manufacturers. The significance of TNCs in South Africa’s economy is in line with global trends which have seen individual corporations controlling resources (at least in monetary terms) and having security, intelligence, and public relations operations larger than many states, as well as huge lobbying ability including campaign donations (UNCTAD 2017; Zingales 2017).11 This is not a new phenomenon in South Africa, however. The largest South African conglomerates, led by Anglo American and Richemont/​Remgro, had always been internationalized even while being identified as South African. Anglo American had international assets in its Luxembourg-​registered holding company, Minorco, while Richemont is Swiss based (Mohamed 2020; Goldstein 2010). The internationalization was spurred by apartheid and sanctions. This meant the Oppenheimer and Rupert families built their offshore operations in Luxembourgand Swiss-registered businesses.12 Glencore also made substantial profits trading with South Africa, prior to 1994 and in contravention of international sanctions.13 With the end of apartheid there was a major restructuring of the minerals-​based conglomerate groupings which led to large international companies that evolved through subsequent mergers, such as in the case of BHP (which had South Africa resources company Gencor as one of its components). 11  The significance of large global corporations is not new, as illustrated by the East India Company of the United Kingdom whose influence over politics saw a fifteen-​year initial monopoly right to trade in a range of goods including tea, lasting for 233 years (Zingales 2017). 12  Although Anglo American had been highly internationalized even before this, for example, with major US investments (Innes 1984). 13  Glencore was set up by Marc Rich (as Marc Rich + Co). One of Marc Rich + Co’s companies Minoil sold oil to South Africa in 1979 at prices around one-third higher than prevailing spot prices (Ammann 2009: 191–​2). Glencore merged with Xstrata in 2013, which had substantial interest in coal mining, and has acquired the Chevron refinery in Cape Town.

362    Pamela Mondliwa and Simon Roberts In the post-​apartheid era large local businesses have been able to expand internationally. While some companies have been very successful, others have used profits earned in the South African market for investments which proved unwise. The successful expansions include: Bidvest, whose company Bid Corporation is one of the two largest food service wholesalers in the United Kingdom, with operations across Europe and in Australasia; SAB, which acquired brewers around the world before merging with Miller; Discovery Health; and MTN. Less successful outward internationalizations included Netcare’s acquisition of the UK’s largest hospital group, GHG, which has performed poorly. The major supermarket chains also made outward acquisitions and found trading much more challenging than in the southern African region. Examples include Woolworths’ acquisition of Australia’s David Jones, and Shoprite’s investments in Tanzania and India from which it later exited. Sasol has made a series of major international investments which have lost money, effectively being subsidized by South African profits (Mondliwa and Roberts 2019). In terms of integration into industrial TNCs, the results are mixed. In auto, the major global OEMs have re​integrated with their South African businesses. However, this has not supported the development of strong capabilities backwards into components manufacturing, in part, because of policies favouring the OEMs (Barnes, Black, and Monaco 2021). In steel, ISCOR was acquired by Mittal Steel, subsequently ArcelorMittal and, after initial investments in restructuring, instead of stronger local capabilities and value chain linkages, there have been substantial outflows of profits (Zalk 2017; Andreoni, Kaziboni, and Roberts 2021).

17.3.3 The Role of Policy in Aligning Firm Strategies with Industrial Development Goals Economic policies, laws, and institutions play an important role in orienting the strategies of large firms for economic development. These institutions not only set the rules of the game but can create incentives for firms to invest in productive activities, including promoting non-​price rivalry to ensure continued investment in capabilities. In 1994, South Africa had large conglomerates constituting a strong industrial base and there was an opportunity to use policy and institutions to re-​orient the strategies of these firms to support industrial development. These firms had already amassed significant resources, capabilities, and economies of scale and scope in key intermediate industries that could be leveraged for South Africa’s industrial development (Fine and Rustomjee 1996; Zalk 2017). The opportunity for diversified industrial development provided by the established base was foregone. South Africa’s 1994 compromises influenced by the lobbying of big businesses were premised on the understanding that efficient capital allocation and higher fixed investment required for growth would best be achieved by liberalizing markets and addressing market failures (Michie and Padayachee 2019; van Niekerk and Padayachee 2019; Mondliwa and Roberts 2021). This reduced the role of policy to that of removing ‘market distortions’ rather than shaping the development path.

Corporate Structure, Industrial Development, and Structural    363 The liberalization policies influenced the changes that are observed above in the corporate structure. South Africa’s approach to capital-​account liberalization supported international expansion by the major conglomerates rather than raising capital for the domestic market (Zalk 2021). This included the overseas listings pushed for by Anglo American, Old Mutual, SAB, and Gencor (as Billiton) (Chabane et al. 2006). The internationalization of the firms taken together with the growth in the importance of institutional investors and the shareholder-​value movement led to conglomerates shedding non-​core assets to increase dividend pay-​outs. This meant that profits were diverted away from productive investment in the economy. South Africa’s approach to trade liberalization had largely benefited those firms that were already competitive at the time of opening the economy (Driver 2019) and undermined the development of capabilities in downstream industries due to premature exposure to international competition (Mondliwa and Roberts 2021). One of the roles of industrial policy is to support firms to reach levels of international competitiveness. South Africa did not, in fact, have an overarching industrial policy until 2007. Though there was a range of incentives to promote investment, exports, and technological improvements, these were largely soft measures and ultimately most benefited the same industries that were already internationally competitive (Black and Roberts 2009; Black and Hasson 2016). There has also been limited success in designing, monitoring, and enforcing conditionalities, one of the levers for orienting firm strategies and decisions (Mondliwa and Roberts 2019). Rather than policies working together to engage with the power of large corporations to encourage dynamic efficiency and productive capabilities, there has been a tension between industrial and competition policy. South Africa’s competition law focused on exchange rather than production, and static rather than dynamic efficiency (Mondliwa, Roberts, and Ponte 2020). As a result, despite successful enforcement by the competition authorities for over twenty years against cartels and mergers, there has been limited success in promoting rivalry (Roberts 2020b, 2020c). A focus on dynamic efficiency implies taking into account investment and the development of productive capabilities within and across firms, which generates competition (Blaug 2001; Budzinski and Beigi 2015; Mondliwa et al. 2021). The latter approach has been adopted in countries such as Germany, Japan, and Korea where the ideal was ‘optimal competition’, a balance between cooperation and competition between firms with the objective of long-​term growth.

17.4  Corporate Structure and the Implications for Structural Change Concerns about market concentration have been growing around the world, at least since the financial crisis, and salient with a number of recent studies pointing to highly

364    Pamela Mondliwa and Simon Roberts concentrated markets and their implications.14 South Africa has been facing such challenges since before the first democratic elections where the outcomes in many sectors and markets are determined by the corporate strategies of a handful of large firms. Where policy is successful in orienting these strategies for long-​term productivity growth, this concentration could be beneficial to the economy. However, there is danger of abuse of market positions as warned by Lamoreaux’s (2019) ‘problem of bigness’. It is no comfort to find that, with the global concentration of economic power, the world appears to be becoming more like South Africa. The ‘business friendly’ environment, which was aggressively promoted to South Africa in the 1990s by international financial institutions and international businesses, led to major corporate restructuring but not to a decline in market concentration or the importance of large firms. Instead of focusing on business confidence we need to grapple with the implications of the central role of large and internationalized businesses to understand how the configurations of economic power impact on the development of dynamic capabilities within and across firms (Pisano 2017; Mondliwa et al. 2021). In South Africa, the apartheid state left a small number of family-​controlled conglomerates astride the economy alongside powerful state-​owned, and former state-​owned, corporations. In 2020, the country’s economic outcomes remained in the hands of decision- m ​ akers at a few corporations. This is a set of corporations in which there have been dramatic changes alongside striking patterns of continuity. In the policy arena, the large firms lobbied strongly for their central role in driving investment and the need for far-​reaching liberalization to spur productivity while not tackling market concentration head on. The claims by some that the concentration reflected their efficiency does not square with the continued poor productivity performance and low investment in the economy as a whole. And, the subsequently revealed prevalence of cartels in South Africa indicated the insider relationships which had continued to govern markets (Muzata et al. 2017; Roberts 2020c). Large firms have also lobbied and strategized to undermine rivals, as would be expected (Makhaya and Roberts 2013). The main changes have been the hollowing-​out of manufacturing industry and the growth in services, including financial services, alongside the ongoing internationalization of large businesses with South African origins (Andreoni et al. 2021). The financial services growth was especially marked in the 2000s with the global commodities boom, as short-​term capital flows flooded into the JSE and the resource-​based companies listed on it. This was not linked to investment in capabilities in South Africa—​in people and technological capabilities—​and the hangover from the subsequent crash in 2008 continued for another decade and more. The legacy was compounded by capture of state institutions for the direction of rents (Bhorat et al. 2017). Overall, there is a common thread in that regulations and industrial policies have largely failed to grapple 14  Such as Wu (2020), Díez et al. (2018), and with reference to the United States, see De Loecker and Eeckhout (2017) and Autor et al. (2019); Philippon (2019), and in the United Kingdom, Aquilante et al. (2019) and Haldane et al. (2018).

Corporate Structure, Industrial Development, and Structural    365 with the entrenched power of the large and increasingly internationalized companies. This includes where companies have maintained productive bases in industries such as chemicals, paper, and various food products. The challenges of a coordinated set of policies, to engage with the power of the incumbents to channel it towards a diversified industrial base, were greatly increased by the fragmentation of the state under President Zuma (Andreoni et al. 2021b; Mondliwa and Roberts 2021). Other services have also grown, led by telecoms, retail and wholesale, healthcare, and hospitality associated with tourism. In many of these services the growth reflects their extension more widely across the economy with the take-​up of mobile telecommunications and the spread of shopping malls. The dynamism is most evident in platforms which have been built from traditional bases of financial strength and capabilities. Naspers leveraged its positions of market power in satellite television and newspapers into digital platforms, notably through an investment in Tencent. Discovery has evolved from medical-​scheme administration and life insurance into behavioural financial services with world-​leading capabilities underpinning entry into international markets. More prosaically, Bidvest in logistics and food service has grown market-​ leading positions on the part of its BidCorp business outside South Africa. The digitalization of economy activity, which cuts across boundaries between sectors, implies that engaging with large firms with dynamic network and scale economies is essential for countries’ development paths (Andreoni, Barnes, Black, and Sturgeon 2021). This requires South Africa to draw out the implications from the central role of large corporations if it is to design effective and coordinated regulatory, competition, and industrial policies.

17.5 Appendix 1 Company

Sector

Mkt cap billion

Agg Sectors 2020

Prosus N.V. /​Naspers

Software and Computer Services Beverages

2,502

1

8

84

94

1,590

Other services Industrial

2

2

6

3

Tobacco

1,398

Industrial

3

1

Software and Computer Services Mining Personal Goods

1,267

4

781 605

Other services Mining Retail

565 565 330

Mining Mining Mining

7 8 9

Anheuser-​Busch Inbev SA British American Tob Plc Naspers Ltd -​N-​

BHP Group Plc Compagnie Fin Richemont SA Anglo American Plc Mining Glencore Plc Mining Anglo American Plat Ltd Mining

5 6

2012

3 6

2002 1994

Richemont

2 3

4 1 Non-​listed 12 4

-​

5 1

–​

(continued)

366    Pamela Mondliwa and Simon Roberts Company

Sector

Mkt cap billion

Agg Sectors 2020

Vodacom Group Ltd

Mobile Telecommunica­ tions Banks

231

Other services

10

11

225

11

13

11

18

Mining Mining Banks

194 188 181

12 13 14

20 16 9

9 7 12

16 21 11

Industrial Metals and Mining Forestry and Paper Mining Mining

164

Financial services Mining Mining Financial services Industrial

15

10

Anglo

160 135 126

Industrial Mining Mining

16 17 18

40

Anglo

17

10

Mining Life Insurance

124 120

19 20

Spun off from BHP 22 19 –​

Banks

112

21

52

Mobile Telecommunica­ tions Mining Food and Drug Retailers Life Insurance

109

Mining Financial services Financial services Other services

22

5

91 90

Mining Retail

23 24

86

Financial services Industrial Retail

25

67 Spun off from Bidvest 37 65 –​

26 27

7 19

5 60

8 109

28

14

20

41

Firstrand Ltd Gold Fields Ltd Anglogold Ashanti Ltd Standard Bank Group Ltd Kumba Iron Ore Ltd Mondi Plc Sibanye Stillwater Ltd Impala Platinum Hlgs Ltd South32 Ltd Sanlamtd Capitec Bank Hldgs Ltd Mtn Group Ltd Northam Platinum Ltd Bid Corporation Ltd Discovery Ltd Sasol Ltd Shoprite Holdings Ltd

Chemicals Food and Drug Retailers Banks

86 85

General Financial

59

Harmony Gm Co Ltd Aspen Pharmacare Hldgs Quilter Plc

Food and Drug Retailers Mining Pharmaceuticals and Biotech General Financial

Nedbank Group Ltd

Banks

52

Bidvest Ltd Old Mutual Ltd

General Industrials Life Insurance

50 48

Remgro Ltd

General Financial

48

Absa Group Ltd Reinet Investments S.C.A Clicks Group Ltd

75

2012

2002 1994

18

43

–​

57

Financial services Financial services Retail

30

69

90

145

56 56

Mining Industrial

31 32

31 25

21 72

121 –​

53

Financial services Financial services Industrial Financial services Financial services

33 34

Old Mutual assoc UK business 18 15 24

35 36

24 15

28 8

100 –​

37

23

14

10

29

(continued)

Corporate Structure, Industrial Development, and Structural    367 Company

Sector

Mkt cap billion

Agg Sectors 2020

2012

2002 1994

Exxaro Resources Ltd Rand Merchant Inv Hldgs African Rainbow Min Ltd Pepkor Holdings Ltd Nepi Rockcastle Plc

Mining General Financial

47 46

38 39

21

24

Mining

45

Mining Financial services Mining

40

29

–​

–​

General Retailers Real Estate Investmt and Services Health Care Equip and Services Media

44 44

Retail Financial services

41 42

44

43

45

76

–​

Real Estate Investment Trusts General Retailers

38

45

33

–​

–​

38

Other services Other services Financial services Retail

46

28

59

–​

Mediclinic Int Plc Multichoice Group Ltd Growthpoint Prop Ltd Woolworths Holdings Ltd Tiger Brands Ltd The Spar Group Ltd

39

44

36 35

Industrial Retail

47 48

26 54

31 –​

30 –​

Mr Price Group Ltd Ninety One Plc

Food Producers Food and Drug Retailers General Retailers General Financial

34 28

49 50

46

–​

–​

Santam Ltd

Nonlife Insurance

27.8

51

56

50

130

The Foschini Group General Retailers Ltd Life Healthc Grp Hldgs Health Care Ltd Equipment and Services Avi Ltd Food Producers Pick N Pay Stores Ltd Food and Drug Retailers Momentum Met Hldgs Life Insurance Ltd Capital&Counties Real Estate Prop Plc Investment Trusts (REITs) Investec Plc Banks

26.6

Retail Financial services Financial services Retail

52

38

96

47

25.7

Other services

53

41

52

70

25.3 23.5

Industrial Retail

54 55

59 50

58 44

28 103

22.4

Financial services Financial services

56

39

57

57

20.2

58

Vivo Energy Plc Netcare Ltd

General Retailers Health Care Equipment and Services Mining

19.9 19.2

Financial services Retail Other services

51

–​

18.4

Mining Beverages

18.2 16.8

75

117

Royal Bafokeng Platinum Ltd Drd Gold Ltd Distell Group Hldgs Ltd

22.0

59 60

51

Mining

61

77

Mining Industrial

62 63

61

(continued)

368    Pamela Mondliwa and Simon Roberts Company

Sector

Mkt cap billion

Agg Sectors 2020

Sirius Real Estate Ltd

Real Estate Investment and Services Life Insurance

16.4

Financial services

64

16.0

General Retailers Food and Drug Retailers Real Estate Investment Trusts (REITs) General Retailers Real Estate Investment Trusts (REITs) Real Estate Investment Trusts (REITs) General Financial

15.9 15.7

Financial services Retail Retail

Liberty Holdings Ltd Italtile Ltd Dis-​Chem Pharmacies Ltd Resilient Reit Ltd Truworths Int Ltd Fortress Reit Ltd A Redefine Properties Ltd Transaction Capital Ltd Sappi Ltd Coronation Fund Mngrs Ltd Textainer Group Hldgs Ltd Ninety One Ltd

42

17

4*

66 67

97

–​

–​

68

15.3 15.1

Retail Financial services

69 70

30

80

–​

14.3

Financial services

71

48

–​

–​

14.2

Financial services Industrial Financial services Industrial

72 62 82

13

17

Financial services Other services

76

Industrial Mining Financial services Financial services

78 79 80

47

25

33

Financial services Financial services

82

Other services Financial services Financial services

84

14.0 13.7

Industrial Transportation General Financial

13.7 13.4

12.9 11.5 10.4

Equites Prop Fund Ltd

Real Estate Investment Trusts (REITs) General Financial

10.3

9.7

Allied Electronics Corp Investec Ltd

Real Estate Investment and Services Software and Computer Services Banks

Psg Konsult Ltd

General Financial

9.1

Globe Trade Centre S.A.

65

Financial services

Fixed Line Telecommunica­ tions Barloworld Ltd General Industrials Pan African Resource Plc Mining Jse Ltd General Financial

Psg Group Ltd

2002 1994

15.4

Forestry and Paper General Financial

Telkom Sa Soc Ltd

2012

13.0

10.3

9.6 9.5

73 74 75

77

92

81

83

85 86 (continued)

Corporate Structure, Industrial Development, and Structural    369 Company

Sector

Mkt cap billion

Agg Sectors 2020

Cartrack Holdings Ltd

Support Services

9.0

Other services Industrial Financial services

Aeci Ltd Lighthouse Capital Ltd

Chemicals Real Estate Investment and Services Oceana Group Ltd Food Producers Super Group Ltd Industrial Transportation Motus Holdings Ltd General Retailers Stenprop Limited Real Estate Investment Trusts (REITs) Imperial Logistics Ltd Industrial Transportation Investec Property Fund Real Estate Ltd Investment Trusts (REITs) Mas Real Estate Inc. Real Estate Investment and Services Rdi Reit Plc Real Estate Investment Trusts (REITs) Adcock Ingram Hldgs Pharmaceuticals Ltd and Biotechnology Massmart Holdings Ltd General Retailers Kap Industrial Hldgs Ltd General Industrials

8.9 8.7

2012

2002 1994

87 88 89

70

88

51

99 107

85

–​

36

30

80

81

–​

8.1 7.7

Industrial Industrial

90 91

7.7 7.7

Retail Financial services

92 93

7.5

Industrial

94

7.5

Financial services

95

7.3

Financial services

96

7.3

Financial services

97

7.0

Industrial

98

75

6.9 6.8

Retail Industrial

99 100

32

Note: 2020 data downloaded on 18 September 2020; other years are as of 31 December.

References Ammann, Daniel. 2009. The King of Oil. New York: St Martin’s Griffin. Amsden, Alice Hoffenberg. 2001. The Rise of the ‘Rest’:  Challenges to the West from Late-​ Industrializing Economies. New York: Oxford University Press. Amsden, Alice Hoffenberg, and Hikino Tashiki. 1994. ‘Project execution capability, organizational know-​how and conglomerate corporate growth in late industrialization’, Industrial and Corporate Change, 3(1): 111–​47. Andreoni, Antonio and Simon Roberts. 2020. ‘Governing data and digital platforms in middle income countries:  Regulations, competition and industrial policies, with sectoral case studies from South Africa’. Digital Pathways at Oxford. Paper Series no. 5, Blavatnik School of Government, Oxford University. Andreoni, Antonio, and Fiona Tregenna. 2020. ‘Escaping the middle-​income technology trap:  A comparative analysis of industrial policies in China, Brazil and South Africa’, Structural Change and Economic Dynamics, 54: 324–​40.

370    Pamela Mondliwa and Simon Roberts Andreoni, Antonio, and Fiona Tregenna. 2021. ‘The middle-​income technology trap and premature deindustrialization’, in Antonio Andreoni, Pamela Mondliwa, Simon Roberts, and Fiona Tregenna (eds) Structural Transformation in South Africa:  Globalization and the Challenges of Inclusive Industrial Development in a Middle-​income Country. Oxford: Oxford University Press. Andreoni, Antonio, Justin Barnes, Anthony Black, and Timothy Sturgeon. 2021. ‘Digitalization, industrialization, and skills development: Opportunities and challenges for middle-​income countries’, in Antonio Andreoni, Pamela Mondliwa, Simon Roberts, and Fiona Tregenna (eds) Structural Transformation in South Africa: Globalization and the Challenges of Inclusive Industrial Development in a Middle-​income Country. Oxford: Oxford University Press. Andreoni, Antonio, Lauralyn Kaziboni, and Simon Roberts. 2021. ‘Mining opportunities: transformations, power and governance in the metals, machinery and mining equipment value chains in South Africa’, in Antonio Andreoni, Pamela Mondliwa, Simon Roberts, and Fiona Tregenna (eds) Structural Transformation in South Africa: Globalization and the Challenges of Inclusive Industrial Development in a Middle-​income Country. Oxford: Oxford University Press. Andreoni, Antonio, Pamela Mondliwa, Simon Roberts, and Fiona Tregenna. 2021a. ‘Framing structural transformation in South Africa and beyond’, in Antonio Andreoni, Pamela Mondliwa, Simon Roberts, and Fiona Tregenna (eds) Structural Transformation in South Africa: Globalization and the Challenges of Inclusive Industrial Development in a Middle-​ income Country. Oxford: Oxford University Press. Andreoni, Antonio, Pamela Mondliwa, Simon Roberts, and Fiona Tregenna. 2021b. ‘Towards a new industrial policy for structural transformation’, in Antonio Andreoni, Pamela Mondliwa, Simon Roberts, and Fiona Tregenna (eds) Structural Transformation in South Africa: Globalization and the Challenges of Inclusive Industrial Development in a Middle-​ income Country. Oxford: Oxford University Press. Andreoni, Antonio, Nishal Robb, and Sophie van Huellen. 2021. ‘Profitability without investment: How financialization undermines structural transformation in South Africa,’ in Antonio Andreoni, Pamela Mondliwa, Simon Roberts, and Fiona Tregenna (eds) Structural Transformation in South Africa:  Globalization and the Challenges of Inclusive Industrial Development in a Middle-​income Country. Oxford: Oxford University Press. Aquilante, Tommaso, Shiv Chowla, Nikola Dacic, Andrew Haldane, Riccardo Masolo, Patrick Schneider, Martin Seneca, and Srdan Tatomir. 2019. ‘Market power and monetary policy’. Staff Working Paper no. 798, Bank of England. Arrow, Kenneth. 1962. ‘Allocation of resources for invention’, in The Rate and Direction of Inventive Activity:  Economic and Social Factors, National Bureau of Economic Research. Princeton, NJ: Princeton University Press. Ashman, Samantha, Ben Fine, and Susan Newman. 2011. ‘The crisis in South Africa: Neoliberalism, financialization and uneven and combined development’, Socialist Register, 47: 174–​95. Autor, David, David Dorn, Lawrence F. Katz, Christina Patterson, and John van Reenen, 2019. ‘The fall of the labor share and the rise of superstar firms’, Quarterly Journal of Economics, 135(2): 645–​709. Barnes, Justin, Anthony Black, and Lorenza Monaco. 2021. ‘Government policy in multinational-​dominated global value chains: Structural transformation within the South African automotive industry’, in Antonio Andreoni, Pamela Mondliwa, Simon Roberts, and Fiona Tregenna (eds) Structural Transformation in South Africa: Globalization and the

Corporate Structure, Industrial Development, and Structural    371 Challenges of Inclusive Industrial Development in a Middle-​income Country. Oxford: Oxford University Press. Bell, Jason, Sumayya Goga, Pamela Mondliwa, and Simon Roberts. 2018. ‘Structural transformation in South Africa: Moving towards a smart, open economy for all’. CCRED Working Paper no. 9/​2018, Johannesburg. Bell, Jason, Lorenza Monaco, and Pamela Mondliwa. 2021. ‘Leveraging linkages for diversification: Building production capabilities in the plastic industries in South Africa and Thailand’, in Antonio Andreoni, Pamela Mondliwa, Simon Roberts, and Fiona Tregenna (eds) Structural Transformation in South Africa:  Globalization and the Challenges of Inclusive Industrial Development in a Middle-​ income Country. Oxford:  Oxford University Press. Bell, Jason, Nicholas Nhundu, Anthea Paelo, Mmamoletji Thosago, and Thando Vilakazi. 2017. ‘Growth and strategies of large and leading firms: Metals, machinery and equipment sector assessment’. CCRED Working Paper no. 18/​2017, Johannesburg. Bhorat, Haroon, Mbongiseni Buthelezi, Ivor Chipkin, Sikhulekile Duma, Lumkile Mondi, Camaren Peter, Mzukisi Qobo, Mark Swilling, and Hannah Friedenstein, 2017. Betrayal of the Promise: How South Africa is being Stolen. Stellenbosch: State Capacity Research Project. Black, Anthony, and Reviva Hasson. 2016. ‘Capital-​intensive industrialisation, comparative advantage and industrial policy’, in Anthony Black (ed.) Towards Employment-​intensive Growth in South Africa. Cape Town: Juta and Company (Pty) Ltd. Black, Anthony, and Simon Roberts. 2009. ‘The evolution and impact of industrial and competition policies’, in Janine Aron, Brian Kahn, and Geeta Kingdon (eds) South African Economic Policy under Democracy. Oxford: Oxford University Press. Blaug, Mark. 2001. ‘Is competition such a good thing? Static efficiency versus dynamic efficiency’, Review of Industrial Organization, 19(1): 37–​48. Bloom, Nicholas, Erik Brynjolfsson, Lucia Foster, Ron Jarmin, Megha Patnaik, Itay Saporta-​ Eksten, and John van Reenen. 2019. ‘What drives differences in management?’, American Economic Review, 109(5): 1648–​83. Bloom, Nicholas, John van Reenen, and Heidi Williams. 2019. ‘A toolkit of policies to promote innovation’, Journal of Economic Perspectives, 33(3): 163–​84. Bosiu, Teboho, and Thando Vilakazi. 2020. ‘Competition and inclusive regional economic growth in food production: Barriers to entry and the role of African multinational corporations’. WIDER Working Paper no. 2020/​88, Helsinki. Bosiu, Teboho, Nicholas Nhundu, Anthea Paelo, Mmamoletji Thosago, and Thando Vilakazi. 2017. ‘Growth and strategies of large and lead Firms: Top 50 JSE firms’. CCRED Working Paper no. 2017/​17, Johannesburg. Buccirossi, Paolo, Lorenzo Ciari, Tomaso Duso, Giancarlo Spagnolo, and Cristiana Vitale. 2013. ‘Competition policy and productivity growth:  An empirical assessment’, Review of Economics and Statistics, 95(4): 1324–​36. Budzinski, Oliver, and Maryam Beigi. 2015. ‘Generating instead of protecting competition’, in Michal Gal, Mor Bakhoum, Josef Drexl, Eleanor Fox, and David Gerber (eds) Economic Characteristics of Developing Jurisdictions: Their Implications for Competition Law. Cheltenham: Edward Elgar. Buthelezi, Thembalethu, Thando Mtani, and Liberty Mncube. 2019. ‘The extent of market concentration in South Africa’s product markets’, Journal of Antitrust Enforcement, 7(3): 352–​64. Chabane, Neo, Simon Roberts, and Andrea Goldstein. 2006. ‘The changing face and strategies of big business in South Africa: More than a decade of political democracy’, Industrial and Corporate Change, 15(3): 549–​77.

372    Pamela Mondliwa and Simon Roberts Chandler, Alfred D. 2001. Inventing the Electronic Century: The Epic Story of the Consumer Electronics and Computer Industries. Cambridge, MA: Harvard University Press. Chandler, Alfred D., Franco Amatori, and Takashi Hikino. 1997. Big Business and the Wealth of Nations. Cambridge: Cambridge University Press. Competition Commission. 2019. Healthcare Market Inquiry. Pretoria:  Competition Commission. Dallas, Mark P., Stefano Ponte, and Timothy J. Sturgeon. 2019. ‘Power in global value chains’, Review of International Political Economy, 26(4): 666–​94. das Nair, Reena. 2017. ‘The internationalisation of supermarkets and the nature of competitive rivalry in retailing in southern Africa’, Development Southern Africa, 35(3): 315–​33. das Nair, Reena. 2019a. ‘The spread and internationalisation of South African retail chains and the implications of market power’, International Review of Applied Economics, 33(1): 30–​50. das Nair, Reena. 2019b. ‘The internationalisation of supermarkets in Southern Africa, the nature of competitive rivalry between grocery retailers and the implications for local suppliers’. Unpublished PhD dissertation, University of Johannesburg. das Nair, Reena, Shingie Chisoro, and Francis Ziba. 2018. ‘Supermarkets’ procurement strategies and implications for local suppliers in South Africa, Botswana, Zambia and Zimbabwe’, Development Southern Africa, 35(3): 334–​50. Dauda, Seidu, Sara Nyman, and Aalia Cassim. 2019. ‘Product market competition, productivity, and jobs: The case of South Africa’. World Bank Policy Research Working Paper no. 9084, Washington, DC. De Loecker, Jan, and Jan Eeckhout. 2018. ‘Global market power’. NBER Working Paper no. w24768, Cambridge, MA. Díez, Federico, Daniel Leigh, and Suchanan Tambunlertchai. 2018. ‘Global market power and its macroeconomic implications’. IMF Working Paper no. 18/​137, Washington, DC. Driver, Ciaran. 2019. ‘Trade liberalization and South African manufacturing: Looking back with data’. UNU-​WIDER Working Paper no. 2019/​30, Helsinki. Fine, Ben, and Zavareh Rustomjee. 1996. The Political Economy of South Africa: From Minerals-​ energy Complex to Industrialisation. London: Hurst & Co. Publishers. Gereffi, Gary, and Joonkoo Lee. 2016. ‘Economic and social upgrading in global value chains and industrial clusters: Why governance matters’, Journal of Business Ethics, 133: 25–​38. Goga, Sha’ista, Anthea Paelo, and Julius Nyamwena. 2019. ‘Online retailing in South Africa: An overview’. CCRED Working Paper no. 2019/​2, Johannesburg. Goldstein, Andrea. 2010. ‘Business Groups in South Africa’, in Asli M. Colpan, Takashi Hikino, and James R. Lincoln (eds) The Oxford Handbook of Business Groups. Oxford:  Oxford University Press. Haldane, Andrew, Tommaso Aquilante, Patrick Schneider, Martin Seneca, Shiv Chowla, Nikola Dacic, Riccardo Masolo, and Srdan Tatomir. 2018. ‘Market power and monetary policy’. Speech at the Federal Reserve Bank of Kansas City Economic Policy Symposium, Jackson Hole, Wyoming (Vol. 24). Lall, Sanjaya. 1993 ‘Introduction: Transnational corporations and economic development’, in Sanjaya Lall (ed.) Transnational Corporations and Economic Development. London: Routledge. Lamoreaux, Naomi. 2019. ‘The problem of bigness: From Standard Oil to Google’, Journal of Economic Perspectives, 33(3): 94–​117.

Corporate Structure, Industrial Development, and Structural    373 Lazonick, William, 2010. ‘The Chandlerian Corporation and the Theory of Innovative Enterprise’, Industrial and Corporate Change, 19: 317–​49. Makhaya, Getrude, and Simon Roberts. 2013. ‘Expectations and outcomes—​considering competition and corporate power in South Africa under democracy’, Review of African Political Economy, 138: 556‒71. Mathis, Jerome, and Wilfried Sand-​Zantman. 2014. ‘Competition and investment:  What do we know from the literature?’. Institut d’Economie Industrielle Working Paper, Universite Toulouse. McMillan, Margaret, Dani Rodrik, and Inigo Verduzco-​Gallo. 2014. ‘Globalization, structural change and productivity growth with an update on Africa’, World Development, 63: 11–​32. Michie, Jonathan, and Vishnu Padayachee. 2019. ‘South African business in the transition to democracy’, International Review of Applied Economics, 33(1): 1–​10. Mohamed, Seeraj. 2020. ‘Anglo-​American corporation and corporate restructuring in post-​ apartheid South Africa’, International Review of Applied Economics, 34(4): 1–​17. Mohamed, Seeraj, and Kade Finnof. 2005. ‘Capital flight from South Africa, 1980–​2000’, in Gerald A. Epstein (ed.) Capital Flight and Capital Controls in Developing Countries. Cheltenham: Edward Elgar. Mondliwa, Pamela, and Simon Roberts. 2017. ‘Economic benefits of Mozambican gas for Sasol and South African Government’. Centre for Competition Regulation and Economic Development Working Paper no. 2017/​23, Johannesburg. Mondliwa, Pamela, and Simon Roberts. 2019. ‘From a developmental to a regulatory state? Sasol and the conundrum of continued state support’, International Review of Applied Economics, 33(1): 11–​29. Mondliwa, Pamela, and Simon Roberts. 2020. ‘Black economic empowerment and barriers to entry’, in Thando Vilakazi, Sumayya Goga, and Simon Roberts (eds) Opening the South African Economy: Barriers to Entry and Competition. Cape Town: HSRC Press. Mondliwa, Pamela, and Simon Roberts. 2021. ‘The political economy of structural transformation:  Political settlements and industrial policy in South Africa’, in Antonio Andreoni, Pamela Mondliwa, Simon Roberts, and Fiona Tregenna (eds) Structural Transformation in South Africa: Globalization and the Challenges of Inclusive Industrial Development in a Middle-​income Country. Oxford: Oxford University Press. Mondliwa, Pamela, Sumayya Goga, and Simon Roberts. 2021. ‘Competition, productive capabilities and structural transformation in South Africa’, European Journal of Development Research (forthcoming). Mondliwa, Pamela, Nicholas Nhundu, Anthea Paelo, Mmamoletji Thosago, and Thando Vilakazi. 2017. ‘Remgro Ltd Company Assessment’. Industrial Development Research Programme Research Report, Centre for Competition Regulation and Economic Development, Johannesburg. Mondliwa, Pamela, Simon Roberts, and Stefano Ponte. 2020. ‘Competition and power in global value chains’, Competition & Change. doi.org/​10.1177/​1024529420975154. Muzata, Tapera, Simon Roberts, and Thando Vilakazi. 2017. ‘Penalties and settlement for South African cartels:  An economic review’, in Jonathan Klaaren, Simon Roberts, and Imraan Valodia (eds) Competition Law and Economic Regulation. Johannesburg:  Wits University Press. Padayachee, Vishnu, and Robert van Niekerk. 2019. Shadow of Liberation: Contestation and Compromise in the Economic and Social Policy of the African National Congress, 1943–​1996. Johannesburg: Wits University Press.

374    Pamela Mondliwa and Simon Roberts Philippon, Thomas. 2019. The Great Reversal:  How America Gave Up on Free Markets. Cambridge, MA: Harvard University Press. Pisano, Gary. 2017. ‘Towards a prescriptive theory of dynamic capabilities: Connecting strategic choice, learning and competition’, Industrial and Corporate Change, 26(5): 747–​62. Ponte, Stefano, Simon Roberts, and Lance van Sittert. 2007. ‘Black economic empowerment, business and the state in South Africa’, Development and Change, 38(5): 933–​55. Roberts, Simon. 2020a. ‘Enterprises and industrial policy: Firm-​based perspectives’, in Arkebe Oqubay, Christopher Cramer, Ha-​Joon Chang, Richard Kozul-​Wright (eds) The Oxford Handbook of Industrial Policy. Oxford: Oxford University Press. Roberts, Simon. 2020b. ‘Assessing the record of competition law enforcement in opening up the economy’, in Thando Vilakazi, Sumayya Goga, and Simon Roberts (eds) Opening the South African Economy: Barriers to Entry and Competition. Cape Town: HSRC Press. Roberts, Simon. 2020c. ‘Cartel enforcement—​critical reflections from the South African experience’, in Deborah Healey, Michael Jacobs, and Rhonda L. Smith (eds) Research Handbook on Methods and Models of Competition Law. Cheltenham: Edward Elgar. Sutton, John. 2012. Competing in Capabilities:  The Globalization Process. Oxford:  Oxford University Press. Thakoor, Vimal. 2020. ‘Market power, growth, and inclusion: The South African experience’. IMF Working Paper no. WP/​20/​206, Washington, DC. UNCTAD. 2014. Trade and Development Report:  Global Governance and Policy Space for Development. Geneva: UNCTAD. UNCTAD. 2017. ‘ Beyond austerity: Towards a global new deal’, in Trade and Development Report 2017. Geneva: UNCTAD. UNCTAD. 2020. Trade and Development Report:  From Global Pandemic to Prosperity for All: Avoiding Another Lost Decade. Geneva: UNCTAD. Vilakazi, Thando, and Teboho Bosiu. 2021. ‘Black economic empowerment, barriers to entry, and economic transformation in South Africa’, in Antonio Andreoni, Pamela Mondliwa, Simon Roberts, and Fiona Tregenna (eds) Structural Transformation in South Africa: Globalization and the Challenges of Inclusive Industrial Development in a Middle-​ income Country. Oxford: Oxford University Press. World Bank. 2014. ‘South Africa economic update: Focus on export competitiveness’. South Africa Economic Update no. 5, Washington, DC. World Bank. 2016. ‘South Africa economic update: Promoting faster growth and poverty alleviation through competition’. South Africa Economic Update no. 8, Washington, DC. Wu, Tim. 2020. The Curse of Bigness. London: Atlantic Books. Zalk, Nimrod. 2017. ‘The things we lost in the fire: The political economy of post-​apartheid restructuring of the South African steel and engineering sectors’. PhD thesis, University of London, School of Oriental and African Studies. Zalk, Nimrod. 2021. ‘Structural change in South Africa: An historical sectoral perspective’, in Antonio Andreoni, Pamela Mondliwa, Simon Roberts, and Fiona Tregenna (eds) Structural Transformation in South Africa:  Globalization and the Challenges of Inclusive Industrial Development in a Middle-​income Country. Oxford: Oxford University Press. Zingales, Luigi. 2017. ‘Towards a political theory of the firm’, Journal of Economic Perspectives, 31(3): 113–​30.

Chapter 18

Valu e Chai ns a nd In dustrial Dev e l opme nt i n Sou th A fri c a Mike Morris, Justin Barnes, and David Kaplan

18.1 Introduction The process of globalization has resulted in production being increasingly organized within global value chains (GVCs). Lead firms, typically located in developed countries, acquire inputs from diverse global locations. Suppliers may simply supply intermediate inputs whereby the last stages of assembly are undertaken by the lead firms or they may supply the finished product with the lead firms undertaking activities such as marketing, branding, advertising, and customization. GVCs are recognized as being important because they drive the organization of industrialization activities (encompassing manufacturing and related services, agri-​ processing, minerals, and energy) on a global and national scale. GVCs are essentially about global and regional linkages, and hence exports and imports play an important role in fostering industrial employment. These global and regional value chains are themselves driven by powerful lead firms which exercise power (in the form of structured governance) over chain activities. It is the dynamics of governance that is crucial for understanding access to value chains, upgrading within them to build capabilities and skills, and linkages (forward, backward, and horizontal) to extend value-added productive activities more widely and more deeply across an economy. The GVC literature stresses that upgrading processes are shaped by the type of value chain in which firms are inserted, and in particular by the governance structure of chains, as determined by lead firms. These structures influence the flow and allocation of activities and resources within chains, and hence, firms’ prospects of entry and

376    Mike Morris, Justin Barnes, and David Kaplan upgrading and the distribution of rewards and risks along chains (Gereffi et al. 2001; Gereffi et al. 2005; Kaplinsky and Morris 2001). Currently, engagement in GVCs is heavily concentrated in the regional blocs of East Asia, Europe, and North America; Africa’s participation is very limited (Baldwin 2012; UNCTAD 2013; African Economic Outlook 2014). However, while dwarfed by other regional trading blocs in developed and developing regions, Africa’s imports and engagement in GVCs has been growing more rapidly. As a result, Africa captures a small but growing share of trade in GVCs (Foster-​McGregor et al. 2015). Engagement in GVCs has many potential advantages for developing countries. No longer required to create entire new industries, developing countries can specialize in particular tasks and provide intermediate inputs to GVCs for final assembly and market customization. Engagement in GVCs and participation in global markets has the potential for technology transfer, enhancing efficiencies and management on the part of participating firms. This, in turn, may have positive spillover effects on other local firms. Notwithstanding the positive potential of engagement in GVCs, countries may become trapped in low value-added segments of the GVC in which there is little potential for technology transfer and consequently positive spillover effects. A critical issue therefore is whether engagement in GVCs allows firms, over time, to enhance productivity and to become competitive in higher value-added processes and products. A very large part of Africa’s participation in GVCs consists of firms that are located upstream and that provide primary products or, less frequently, very simple manufactures to firms located further downstream. These production activities, more particularly the production of primary products, are low value added and low skilled with very little potential for technology or skill upgrading. Engagement in GVCs by these firms consequently has very limited positive spillovers on other local firms. By contrast, downstream participation by African firms is very limited and has shown only marginal signs of increasing over the last two decades. As noted by Foster-​McGregor et al. (2015: 6), ‘Downstream participation in GVCs by Africa is found to be relatively low, and more importantly has shown little sign of increasing since 1995.’ The net effect is that the enhanced participation in GVCs has had only a limited impact on transforming production in African economies. However, there are some significant exceptions. Some African countries have succeeded in developing significant production capacities downstream. A number of African countries have reported that more than 50 per cent of their total GVC involvement is located in downstream activities. These countries include Mauritius, Botswana, Ethiopia, Kenya, and Tanzania (Foster-​McGregor et al. 2015). The Southern African region is more heavily integrated into GVCs than the rest of the continent. Moreover, participation in GVCs has increased significantly over the last decade (AUC/​OECD 2019). However, the report noted that ‘Southern Africa’s participation in global value chains remains peripheral’. It notes that South Africa is a partial exception as it is more integrated into GVCs than any other African country, and provides a potential entry point for other countries in the region to engage in GVCs. South Africa

Value Chains and Industrial Development in South Africa    377 is hence pivotal to the region. It ‘is the region’s natural gateway into global value chains’ (AUC/​OECD 2019: 137). The further development of regional value chains focused on South Africa in products that are destined for global markets represents the best opportunity for countries in the region to deepen their engagement in GVCs. The AUC/​OECD report recommended that regional value chains should be developed so as to ‘piggy-​back on South Africa’s participation in GVCs’ (AUC/​OECD 2019: 120–​1). The report singled out five value chains in the region that have the greatest potential for further development—​ automotive, textiles and apparel, meat, agribusiness, and minerals. This chapter is concerned with the first two of these high potential value chains (viz. textiles and apparel, and automotive). A significant part of South Africa’s participation in GVCs is in downstream production. Moreover, much of South Africa’s engagement in GVCs is located in high value, and skill intensive, activities that offer at least the potential for significant technological advancement and positive spillovers to the rest of the economy. Upgrading within GVCs is therefore central to South Africa’s further industrial and economic development. The automotive and apparel manufacturing sectors have the most developed GVC linkages in South Africa, but in very different ways. The chapter will therefore analyse them as two case studies to understand the different dynamics exhibited by industry in South Africa linking into GVCs. These two case studies form the substantive part of the chapter, analysing their recent trends, current dynamics, and prospects. The South African automotive industry is deeply engaged in GVCs and is undoubtedly the sector that currently, through this engagement, offers the greatest potential for technological advancement, increasing skill intensity and upgrading. Automotive is a scale and skill intensive industry requiring very considerable foreign investment and advanced management practices. Moreover, the capabilities developed in the South African automotive sector are core to the network of globally exported products. These same technological capabilities are located in a product space that strongly suggests that these same technological capabilities can be fairly readily deployed to the production of other sophisticated products (AUC/​OECD 2019: 125).1 The South African automotive sector is predominantly export oriented, but it is also highly import intensive. It is an example of a medium technology industry that requires relatively high levels of capabilities and skills. The foreign-owned assemblers located in South Africa drive the value chain according to strict adherence to lean manufacturing principles. They determine the levels of local sourcing in domestically assembled vehicles that are then sold locally, regionally, and internationally. The locally based component sector plays an important role in terms of GVC linkage development, as well as providing technology spillovers into other sectors. However, the South African

1 

The automotive industry is centrally situated within the map of economic complexity as developed by Harvard University (2019).

378    Mike Morris, Justin Barnes, and David Kaplan automotive component industry still exhibits relatively thin levels of value added and independent design and product development capabilities. South Africa’s apparel industry is almost wholly domestic market oriented, sourcing both globally as well as through extended regional value chains in sub-​Saharan Africa. The sector is an example of a low technology, labour intensive industry. A few Johannesburg Stock Exchange (JSE) listed large retail groups drive the sector and determine what is manufactured in South Africa, elsewhere in the Southern African region, and from Asia. It is these firms that determine the different dynamics prevalent within local supplier activities, and hence opportunities for upgrading and development. The sector is characterized by lower levels of managerial capabilities and worker skills. However, more recently the retail lead firms are moving to develop Quick Response (QR) based value chain capabilities and are therefore either developing local suppliers or building their own vertical design, product development, and manufacturing capabilities. The major challenge the industry faces is raising its capabilities and worker skill levels to meet these new value chain requirements while also extending the supplier base to increase value addition (and by implication employment) in the economy. The production of textiles and apparel is widespread. All the countries in the Southern African region have some production with a concentration on Lesotho and South Africa. The South African textile and apparel industry is focused almost exclusively on the domestic market. Textiles and apparel in SSA (in Lesotho, Swaziland, Mauritius, and Madagascar) by contrast, is entirely export focused towards the US, EU, and South African high income markets. Lesotho has developed basic sewing production capacities in apparel and in wool and cotton-​related products (AUC/​OECD 2019). However, by contrast with automotive, the productive capacities in textiles and apparel are far less easily redeployed into other goods. The further industrialization and development of South Africa and of the countries in the Southern African region will depend heavily on further developing their engagement in GVCs and simultaneously upgrading their capacities into higher valued and more skill and intensive activities in these two key sectors. The chapter focuses on the two case studies to elucidate key GVC themes: The South African automotive sector as an example of export oriented GVC engagement and the South African clothing and textiles sector as an example of import oriented GVC engagement. These two case studies permit a more detailed interrogation of the GVC implications for South African industrial development.

18.2 The South African Automotive Industry The South African automotive industry has a long domestic history. The first local vehicles were assembled by Ford in 1924, while the first automotive component

Value Chains and Industrial Development in South Africa    379 manufacturers were established around 1930 (Barnes 2013). As demand for vehicles grew from the 1950s, the assembly of vehicles using imported kits proliferated. The low local content of the wide range of assembled vehicles limited the development of the components industry and placed substantial pressure on the country’s balance of payments. In response, following an Import Substitution Industrialization (ISI) approach, the South African government in 1961 introduced the first of its automotive local content programmes (LCPs). These LCPs were scaled up over time, but up until 1989 were exclusively focused on substituting imported vehicles with locally assembled equivalents for a growing domestic market that, in the early 1980s, was recognized as one of the most important developing economy markets internationally. The intensification of the LCPs over time reduced the proliferation of locally assembled vehicles and increased levels of local content in South African vehicles, but at a substantial cost to the South African economy. South African-assembled vehicle standards were low by comparative international standards, and many models were assembled for extended periods to amortize investments in a low volume, but highly protected domestic market environment (Barnes, Kaplinsky, and Morris 2004). Aside from the importation of a few ultra-​luxury vehicles, the domestic market was essentially closed to vehicle imports. Due to sanctions, most vehicle assembly operations were locally owned, with multinational vehicle assemblers choosing to have their vehicles assembled in South Africa under licence, rather than through direct ownership. The exceptions to this were the German luxury vehicle assemblers who maintained either full (BMW, Volkswagen/​Audi) or partial (Mercedes-​Benz) ownership of their South African operations. In 1989, the South African government introduced the sixth phase of the LCP. In some ways this represented the zenith of the LCPs, with local content needing to reach 60 per cent of the value of locally assembled vehicles. However, it also represented an important shift in government policy. For the first time local vehicle assemblers could calculate the local content in vehicle and component exports as part of their 60 per cent domestic local content requirement (Barnes et al. 2004). The South African automotive industry was ‘nudged’ towards an export orientation. The sixth phase of the LCP remained in place until 1995. This period coincided with two momentous historical moments—​one domestic and one international. Domestically, South Africa was slowly emerging from the sanctions era and being re-​engaged internationally. The ANC government in waiting had initiated dialogues on economic policy amongst its social partners, and the multinational vehicle assemblers had re-​engaged with their licensed operations in South Africa to explore post-​apartheid production and market opportunities. Internationally, the global automotive industry was undergoing its own cathartic change. Seminal work from the Institute for Motor Vehicle Productivity (IMVP) had highlighted how uncompetitive Western vehicle manufacturers were in relation to vehicle producers that had embraced lean production systems (Womack et al. 1990). China had initiated engagements with multinational vehicle assemblers to establish Chinese operations, initiating the start of unprecedented market and production growth in the world’s most populous country.

380    Mike Morris, Justin Barnes, and David Kaplan Vehicle production began shifting towards being organized at regional or global levels under the control of ever fewer, and ever larger, lead vehicle assemblers (Sturgeon and Florida 1999). It is within this context that the new South African government terminated the LCPs in 1995 and introduced a comprehensive Trade Related Investment Measure (TRIM) in the form of the Motor Industry Development Programme (MIDP). The MIDP initiated fundamental change in the South African automotive industry (Barnes and Black 2014; Barnes et al. 2004). It immediately reduced vehicle tariffs from 115 per cent to 65 per cent, reduced component tariffs, removed all local content requirements, and introduced a very generous export incentive that essentially allowed vehicle assemblers (and component manufacturers) to earn import credits for exports, thereby allowing them to import products free of duty provided they exported sufficient values of automotive products. This critical shift in government policy resulted in the lead vehicle assemblers progressively acquiring their South African-licensed operations. Nissan, Toyota, Ford, and General Motors joined Mercedes-Benz, Volkswagen, and BMW in taking direct control of their South African operations and strategically repositioning them to take advantage of the MIDP. After seven decades of domestic market orientation, the South African automotive industry was re-​oriented to become fully part of the vehicle assembler GVCs (Barnes and Morris 2008). In accordance with the GVC framework, the number of vehicle platforms manufactured in South Africa plummeted, the average volume assembled per platform increased significantly, and both imports and exports surged (Barnes and Black 2014). A major challenge over the next fifteen years related to the need to upgrade the South African operations in alignment with the vehicle assemblers’ global standards. Driven by a relentless lean production focus, South African assembly and component operations were exhaustively benchmarked against global standards and forced to upgrade their capabilities. This has been extensively documented through the activities of the South African Automotive Benchmarking Club, with major competitiveness improvements noted within the South African automotive value chain (Barnes and Black 2014). However, the period of GVC engagement also coincided with the gradual ‘thinning out’ of local content in South African production (Kaplinsky 2013). From local content of around 60 per cent in each assembled South African vehicle in the early 1990s, the figure was only 46.6 per cent in the final year of the MIDP in 2012. The impact of the MIDP was scrutinized throughout its lifespan, with some commenting that the programme led the automotive industry to be excessively export oriented (Black 2009), too import dependent (Barnes 2013), or was too generous and hence too great a cost for the South African state (Flatters 2005; Flatters and Netshitomboni 2007). However, Barnes and Black (2014) argue that the programme was on balance, a success, with the economic gains associated with the automotive industry’s growth and development far outweighing its fiscal costs. Moreover, the automotive industry performed substantially better than the balance of the domestic manufacturing

Value Chains and Industrial Development in South Africa    381 sector and significantly improved its competitiveness, ensuring South Africans had access to modern, safe vehicles at globally comparative prices (Barnes 2017). The MIDP was replaced by the Automotive Production and Development Programme (APDP) in 2013 and will continue until at least 2026 as part of the national government’s new South African Automotive Masterplan (SAAM), which runs until 2035. The APDP replaced the MIDP because of pressure from the World Trade Organization (WTO), and an acceptance on the part of the national government that the MIDP contravened the WTO’s Agreement on Subsidies and Countervailing Measures. The APDP shifted the government’s support from automotive exports to production, irrespective of market focus. However, it kept the fundamental premise of incentivizing firms through the ability to earn import rebates, and therefore essentially maintained the industry’s established set of export linkages within the increasingly global automotive GVC framework. Despite the poor performance of the South African market since the early 2010s, the APDP has been credited with maintaining South Africa’s vehicle production base, while simultaneously improving the industry’s trade balance. The comparatively stronger performance of vehicle production relative to domestic market vehicle sales since 2014 is presented in Figure 18.1. As highlighted, apart from a brief period in the early 2000s, domestic market sales have tended to be higher than local production, suggesting an important recent shift in the position of the South African automotive industry. Vehicle production has remained more buoyant than domestic sales because approximately 70 per cent of all vehicles produced in South Africa are exported. In fact, several locally assembled models such as the BMW X3 and the Mercedes-Benz C-​Class are almost exclusively exported, and mostly to the European Union. As a result, over 40 per cent of all vehicle production in South Africa is for the EU market.

800 000 700 000

Number of Vechicles

600 000 500 000 400 000 300 000 200 000

Domestic Vehicle Production

2018

2017

2016

2015

2014

2013

2011

2012

2010

2009

2007

2008

2006

2005

2004

2003

2001

2002

2000

1999

1998

1997

1996

0

1995

100 000

Domestic Vehicle Sales

Figure 18.1  South African light vehicle domestic market performance versus production, 1995 to 2018 Source: Calculated from National Association of Automotive Manufacturers of South African (NAAMSA) annual vehicle sales and production publications.

382    Mike Morris, Justin Barnes, and David Kaplan Despite the automotive industry’s long-​standing production in South Africa, and the comparative successes of the MIDP and APDP, the domestic industry remains a marginal player globally, contributing only 0.64 per cent of global vehicle output in 2018.2 Not only is the country’s total production output of around 600,000 units insignificant globally, it has increased only slightly since 2005, as highlighted in Figure 18.1. Vehicle production also remains highly fragmented, making it difficult for the automotive components industry to secure the scale economies required to compete with leading competitors, resulting in local content levels within South African vehicles declining to 39.2 per cent in 2017 (Monaco et  al. 2018), from 47 per cent in 2012 (Barnes and Black 2014). Despite its challenges, the automotive industry remains at the forefront of the national government’s industrialization strategy. It is one of the few domestic manufacturing sectors that has recorded real growth over the last decade, and accounts for an important proportion of total manufacturing output. The industry’s comparative resilience, its established foundations, employment of 112,000 South Africans, export contribution, managerial capabilities, worker skill building, and its recognized technology multipliers, have positioned it as a core, strategic domestic industrial sector (Barnes and Black 2017). It is this position that ensures the government’s continued support for the industry. Despite substantial government support, neither the MIDP nor the APDP has substantially shifted the position of the South African automotive industry as a second tier player globally. While this perspective varies considerably between vehicle assemblers, there is currently no domestic plant that is the primary source for a vehicle platform globally. Even the largest volume models in South Africa, such as the Ford Ranger, Toyota Hilux, Volkswagen Polo, and Mercedes-Benz C-​Class (all above 80,000 units per annum), represent small production volumes relative to larger plants elsewhere. The result is the low local content exhibited amongst South African assemblers, placing substantial pressure on the automotive components industry. An important deficiency in respect of the industry’s present position relates to its poor recent domestic market performance, the extent of imports into the domestic market, and deteriorating regional market conditions. Domestic vehicle sales had contracted significantly pre the COVID-​19 pandemic since the record level achieved back in 2006, with imports comprising over 70 per cent of the domestic car market in 2018. The local market is far from having sufficient demand to attract local assembly exclusively for domestic market supply. A modern assembly plant requires at least 80,000 units of a platform to justify production, and yet the top selling model in South Africa only achieves sales of around 50,000 units (the VW Polo/​Vivo). There is also limited demand for South African vehicles within sub-​Saharan African markets (Barnes et  al. 2019). The short-term prognosis for the regional market is muted, as consumers within this market have the additional choice of purchasing

2 

Calculations derived from OICA production data for 2018 (https://​www.oica.net).

Value Chains and Industrial Development in South Africa    383 pre-​owned vehicle imports at low prices from developed economy markets. These vehicles are moreover sold at highly discounted rates, essentially escaping end-​of-​life vehicle scrapping costs in advanced economies. Coupled with negative domestic market conditions, the local automotive industry is consequently in a difficult strategic position, with vehicle production tied to exports into distant developed economies, most notably the European Union and the United States. These exports are supported by AGOA, the EU-​South African Economic Partnership Agreement, and incentives such as the APDP that compensate for the industry’s substantial cost disadvantages relative to competitors such as Thailand (Barnes et al. 2017). South African automotive production is, therefore, being driven less by local or regional market factors, which underpin the competitive advantage being secured by almost all the country’s more successful competitor economies. Notwithstanding the competitive pressures confronting the domestic industry, the base vehicle ownership profile of South Africa (and regionally) suggests major growth opportunities through to 2035—​provided economic growth increases (to above 3 per cent per annum) and the industry’s base competitiveness continues to improve (Barnes and Black 2017). The SAAM has the central objective of ensuring the local automotive industry increases its production to 1 per cent of global output by 2035 (or potentially 1.4 million units of production) and increases local content in domestically assembled vehicles to 60 per cent (from a 38 per cent base). This would double total employment in the automotive value chain (to 224,000 jobs) even after factoring in major productivity improvements over the period. In doing so, the industry would substantially increase its contribution to the South African economy through to 2035 (Barnes and Black 2017). The automotive GVC is, however, undergoing a pronounced set of changes, which could fundamentally disrupt global production and markets for vehicles. The challenges facing the industry and the development of an appropriate policy is being driven by established GVC pressures relating to efficiencies and competitive factor costs; and the simultaneous emergence of a set of major technological and socio-​economic changes which are transforming the global automotive industry. Three major emerging GVC drivers appear to be shaping the potential of the South African automotive industry and need to be considered. First, developed country vehicle fleet fuel economy requirements are becoming increasingly onerous. The European Union has, for example, significantly tightened the fuel efficiency standards that need to be achieved in its market. From 2021, the new vehicle fleets sold into the European Union by vehicle assemblers need to achieve an average carbon emission per-​kilometre level of 95 grams (from 130 grams). This translates into an average fuel efficiency requirement of 4.1 litres per 100 kilometres for a petrol car and 3.6 litres for a diesel equivalent.3 This is a level which most South African assembled vehicles cannot attain. This type of legislation is driving the development of high technology, smaller displacement internal combustion engines (ICEs) requiring

3 

https://​ec.europa.eu/​clima/​policies/​transport/​vehicles/​cars_​en, accessed 13 November 2019.

384    Mike Morris, Justin Barnes, and David Kaplan extremely clean diesel and petrol that is presently unavailable in South Africa. It is simultaneously driving the rapid evolution of alternative, competing engine technologies, such as battery electric vehicles (BEVs) and hydrogen fuel cell-based vehicles. These new technologies are likely to substantially increase their market share in major developed economies over the next set of model changes, which take place on a six to eight year cycle. Many major cities and several countries have also announced the imminent banning of ICE vehicles, and along with other environmental legislation, the EU, US, and Chinese markets are likely to look fundamentally different over the next two model changes (twelve to sixteen years). A Bloomberg report has predicted that 5 per cent of global vehicle sales would comprise BEVs by 2022, and then rise rapidly to 35 per cent by 2040 (Bloomberg New Energy Finance 2016). How the South African market will be directly affected is unclear with hardly any alternative energy vehicles sold domestically, while sub-​1,000cc ICE vehicles also remain only a small (albeit growing) part of the domestic market. What is clear is that the ability of South African vehicle assemblers to sell larger ICE vehicles into developed economy markets such as the European Union will come under significant pressure. Second, the development of new materials has the potential to displace standard automotive materials, such as steel and plastics, which presently comprise the majority of a vehicle’s material base. This may fundamentally shift the profile of local production. As vehicles become lighter, and functionally more advanced, what new materials will dominate vehicle production? Will these be sourced in South Africa? What role will material composites and nano-​technology play in respect of automotive materials use, and could this displace ladder-​chassis technologies that presently underpin the production of light commercial vehicles in South Africa? As rapidly advancing green manufacturing requirements shape what vehicles and components are produced, as well as what materials and manufacturing processes are used, it is unclear whether these requirements will open or close off production opportunities in South Africa. What green production capabilities will South African manufacturers need to master to ensure continued supply into increasingly environmentally conscious developed economy markets? Third, infotainment, vehicle connectivity, and passive and active vehicle safety developments are fundamentally altering the nature of the vehicle driving experience, and the associated functionality of vehicles. How will this change the nature and cost profile of vehicle production? As additional safety features are developed, with the ultimate objective of dramatically reducing road deaths, what are the consequences for vehicle production? How will South African manufacturers adjust to these emerging requirements, particularly if EU and US market safety requirements diverge dramatically from South African and other developing economy requirements? Tied to this is the profoundly disruptive potential of autonomous vehicles and their consequences for vehicle use and ownership. Forecasts vary but according to the Black Rock Institute (2017), fully autonomous vehicles (AVs) will take off rapidly from 2025 with 75 per cent adoption by 2035. This forecast may be too optimistic, but even a 20 per cent to 30 per cent adoption rate will profoundly affect South African production.

Value Chains and Industrial Development in South Africa    385 How will South African automotive production be impacted by the emergence of autonomous vehicles? How will global, regional, and local vehicle use change? Will vehicle demand contract? Will it support the transition away from private vehicle ownership to IOT-​enabled mobility services? What happens to local production if major developed economies evolve into mass mobility markets serviced by AVs controlled by ride-​ hailing service providers, as opposed to the present model of private vehicle ownership? For example, the Boston Consulting Group (2016) suggests four possibilities: the first being private ownership of AVs alongside conventional vehicles; and the second being a takeover by AVs. In the third scenario, urban transport is dominated by ‘robo-​taxis’; and in the fourth scenario urban transport becomes dominated by the ‘ride-​sharing revolution’. The first scenario has a negligible effect on the number of vehicles sold but the third and fourth scenarios have dramatic effects with the number of vehicles sold reduced by 46 per cent and 59 per cent respectively. These emerging GVC drivers could fundamentally re-​shape the future of the South African automotive industry. It has already led to one casualty, with General Motors (GM) exiting the country in 2017 and selling off its domestic assets to Isuzu. The reasons for GM’s exit were directly tied to its parent company’s decision to exit all markets dominated by legacy (i.e. ICE) technologies requiring new capital investment, and to concentrate global resources on the development of BEVs and autonomous vehicles. These emerging GVC drivers do not yet feature in the domestic or regional market. While the MIDP (1995 to 2012) and APDP (2013 to present) have clearly built on the preceding LCPs and supported the development of a substantially more competitive automotive industry through participation in the automotive GVC, the industry remains vulnerable. The increasingly export-oriented industry is facing major structural change within the automotive GVC and is likely to face major export headwinds in future, at the same time as the domestic and regional markets regress. These export headwinds are recognized in the SAAM, with increasing policy support from 2021 for emerging green and autonomous vehicle technologies. The SAAM also recognizes the associated need for technology reskilling and digital infrastructure development as the GVC drivers impact the industry.

18.3 The South African Clothing and Textiles Industry As in many other sectors, production and trade in the apparel sector is organized in GVCs where production and assembly into final products is carried out via inter-​firm networks on a global scale. A large part of apparel production remains labour intensive, has low start-​up and fixed costs, and requires simple technology. These characteristics encouraged the move to low-cost locations in developing countries exporting into high-income markets and driven by intense inter-​ firm/​ country competition. The

386    Mike Morris, Justin Barnes, and David Kaplan ‘intangible’ GVC activities (product development, design, textile input sourcing, logistics and distribution, branding, and retail) are controlled by a combination of lead firms, intermediaries, and large transnational suppliers. Upgrading strategies are extremely important for suppliers to sustain and improve their positions in apparel value chains. Sourcing decisions are frequently motivated by labour cost differentials but in addition to the classic criteria of costs, quality, and reliability, other criteria are increasingly shaping sourcing decisions (Gereffi and Frederick 2010; Staritz 2011), for example, lead times and flexibility, which is related to the shift to lean retailing and QR production where buyers defray the inventory risks associated with supplying apparel to fast changing, volatile, and uncertain consumer markets by replenishing items on their shelves in very short cycles and minimizing inventories (Abernathy et al. 2006). This ensures more cost-effective forms of supply chain management and reduces the complexity of their supply chains. Price is still an important market requirement but new dynamics favour additional factors such as speed to market and flexibility through small volume runs to achieve this. Until the mid-​1990s, the South African garment and textile sector was locked into import substituting industrialization (ISI), with firms protected by an almost impenetrable thicket of targeted import quotas and high, product-​specific tariffs. In 1994, the government initiated a radical garment tariff phase-​down agreement which saw the elimination of import quotas, movement to a more uniform tariff structure, and a major reduction in nominal tariffs. By 2001, tariffs on textiles were down to 28 per cent and tariffs on garments down to 40 per cent, both from over 100 per cent. In 2001, the US Africa Growth and Opportunity Act (AGOA), allowed South African apparel producers duty- and quota- free access to the US market which met particular rules of origin and US retailer buyers flew in to seek potential new sites for orders. Simultaneously there was a rapid depreciation of the exchange rate from R6.94 per $1 in 2000 to R11.61 in January 2002. Local apparel manufacturers used this to sign numerous export orders to US retailers, seeking larger profits than supplying the domestic market. Total exports of apparel at nominal prices jumped dramatically from R471m in 1995 to R1,901m in 2001 and R2,590m in 2002 (Morris and Einhorn 2008; Morris and Levy 2017). Many manufacturers did not have sufficient capacity to supply both export and domestic markets and reneged on their domestic orders. Hence South African retailers went offshore to import Chinese apparel. By 2004, the appreciation of the Rand/​US$ exchange rate (to R5.73) turned the entire scenario around, creating easier import access, crippling exports, and forcing many local manufacturers to renege on their US export orders. In dollar terms, exports collapsed from $231.8m in 2003 to $24m in 2007, and then even further to only $6m by 2012. Local apparel manufacturers sought in vain to return to supplying the South African domestic retailers, but the market configuration had irretrievably altered. Apparel imports (mainly from China) grew from $192m in 2000 to $755m by 2005 and $1,534m in 2011. Large-scale imports of apparel from China (and later from other apparel producing countries) into South Africa and the value of apparel imports, and that of local apparel production for the domestic market, diverged radically. Local apparel

Value Chains and Industrial Development in South Africa    387 production decreased from 76 per cent of domestic demand in 2005 to 60 per cent in 2011, whilst the share accorded to imports increased from 25 per cent to 40 per cent respectively (Morris and Barnes 2014). As a result of these dynamics employment in firms registered with the national bargaining council dropped dramatically—​from 97,960 in 2003 to 52,656 in 2013 and then by 2016 rose again to 68,757. Although the bargaining council data base provides the only reliable available quantitative data on firm and employment numbers, it also underestimated the actual number of firms and employees working in the industry.4 There were, and remain in existence, many small unregistered firms intersecting into the apparel value chain as informal suppliers. These informal enterprises are highly differentiated in scale, scope, and performance, run by ex-​formal sector skilled workers, mostly competing on price, and feeding into design houses and larger apparel firms, and thereby indirectly into the retail chains. Their exact number, levels of differentiation, and workers employed is unknown. These GVC dynamics continued to shape the apparel and textile sector for the following decade. The industry was caught in a set of contradictions between the strategic positions of its import substituting industrialization and the globalization requirements of the future which paralysed its ability to respond consistently. South African domestic apparel enterprises were caught in a contradiction between being inefficiently and ineffectively set up for import substituting industrialization but having to play on the field of globalization which was characterized by import threats from rising Asian producers operating under entirely different competitive production platforms. The South African government was itself locked into a contradictory policy response—​radically reducing the sector’s protective barriers but unable to significantly assist enough domestic firms to raise their systemic competitiveness and align with the demands of retail buyers driving their value chains. Finally, the unions were caught by their historically successful strategy of maintaining high wages, blocking piece-​rate wage systems, and restricting shift work which rendered productivity hamstrung in the face of global competitor countries. What followed was a series of sometimes contradictory policy responses trying to restrict Chinese imports and force retail chains to procure locally, rather than tackle the underlying problems of weak competitiveness and dysfunctional value chain alignment. When these had palpably failed, after 2010 the Department of Trade and Industry put in place a narrow but focused set of industrial policy interventions under the Clothing and Textiles Competitiveness Programme (CTCP) which comprised two components, the Production Incentive (PI) and the clothing and textiles Competitiveness Improvement Programmes (CIP), to assist firms to become more competitive. The CIP also offers incentives to National Clusters, with South African registered firms receiving financial support on a yearly graduated basis. CIP grants are targeted at interventions that

4 

See Edwards and Morris (2007) for the methodological problems in the official apparel employment statistics.

388    Mike Morris, Justin Barnes, and David Kaplan improve competitiveness, including upgrades and expansion of capital equipment, increasing productivity, enhancing employee skills, improving products and processes, and reducing costs. The industry was also caught in an industrial relations bargaining trap based on a process of rigidifying labour markets and the introduction of a nationally centralized bargaining council system. This drove a strategy of raising wages and closing the gap between the metro and non-​metro wage levels, resisting management attempts to dilute historically inherited operational systems, and pursuing firms deemed to be non-​compliant in terms of wage and other benefits requirements. The latter included government rendering firms deemed to be non-​compliant ineligible for various industry support and training schemes. By the second decade of the twenty-first century the industry displayed very divergent pathways. On the trade front this divergence in retail market supply between domestic suppliers and imports has become a permanent feature of the apparel trade landscape. The government attempt to control Chinese imports not only failed but it also led to import divergence as retail buyers discovered other supply sources. By 2017, South Africa imported over 54 per cent of its apparel from China, followed by Mauritius, Madagascar, and Lesotho with contributions of 7.5 per cent, 6.2 per cent, and 6.2 per cent respectively (Trade Map 2017). South African apparel imports have fluctuated over the period but averaged high annual growth rates of 6.8 per cent from 2007 to 2016. Data from 2017 estimate South African retail purchases for clothing at Rand 43.0 billion. Import purchases are substantially greater than local purchases, with the import estimates for clothing calculated at 53.9 per cent (Barnes and Hartogh 2018). The most striking aspect of this process of import divergence was the rapid growth of imports from the rest of southern Africa—​Lesotho, Swaziland, Mauritius, and Madagascar—​using the duty-​free preferential access protocols of the South African Customs Union (SACU) and South African Development Community (SADC) to enter the domestic market. This has given rise to the growth of regional value chains from Lesotho, Swaziland, Mauritius, and Madagascar feeding into the South African end market, starting tentatively in 2006 and accelerating after 2010 (Table 18.1). Tariff-​free Table 18.1: SSA apparel exports to South Africa (in US$000) 2006

2008

Swaziland

Madagascar Tanzania SSA

2012

2014

2016

2017

Share

0

57,178

100,173

133,186

152,085

189,289

35%

15

44,464

63,305

93,359

114,556

128,519

24%

21,253

47,251

69,249

147,724

130,186

108,605

115,173

21%

72

7,046

18,238

62,182

84,812

91,600

85,592

16%

382

545

2,281

2,257

8,724

6,659

8,371

2%

51,975 79,677 234,084

394,340

460,511

484,213

540,533

100%

Lesotho Mauritius

2010

Source: UN COMTRADE, South African SARS data.

Value Chains and Industrial Development in South Africa    389 access via SACU has been crucial in Lesotho and Swaziland, whilst Mauritius and Madagascar have used SADC protocols to enter the South African market on favourable tariff terms. Imports from Lesotho and Swaziland have been a direct result of South African clothing manufacturers relocating entire plants to the neighbouring countries. In the case of Swaziland, by 2017 apparel exports to South Africa had completely supplanted its relationship to US buyers. It is also important to note that officially recorded imports do not capture the de facto impact of foreign apparel circulating in the domestic economy. The borders are porous, customs officials insufficiently trained to monitor false product declarations, and organized crime syndicates operate freely. Hence the actual quality and value of imports seeping into South Africa from illegal operators and sources is anecdotally recognized as being much higher than the official data reveal. Calculating the opportunity costs associated with the very high levels of importing into the South African clothing and textile value chain is extremely difficult. First, many imported products cannot be manufactured in South Africa. Second, the pricing of imports into South Africa is a major area of dispute. Industry stakeholders, and previous industry research, strongly suggest that the value of such imports is substantially lowered to reduce ad valorem-​based tariff payments. Third, a sizeable amount of imports into South Africa may not be declared at all. Barnes and Hartogh (2018) attempted to calculate the gross value added (GVA) and employment loss impacts on the local industry of clothing and textile imports destined for South African retail consumption, providing two estimations: a narrow loss based on declared import values, and a broader loss based on the assumption that under-​invoicing results in a one-third discounting of imports. Both calculations assume that all imports can be manufactured in South Africa, which is obviously problematic. Leaving aside the problems with the assumptions, they estimate for the clothing industry GVA losses of between Rand 10.7 billion (assuming import values are correct) to Rand 16 billion (assuming significant under-​invoicing is occurring), and lost employment opportunities ranging from 80,519 to 120,779 sectoral jobs. However, given the problematic assumptions that have had to be made, the point is not to focus on the estimates, but rather to see them as displaying the major GVC trends manifested in the relationship between clothing imports and domestic production which has created a bifurcated clothing and textile industry landscape. This bifurcated value chain landscape has had a divergent impact on firm level growth paths. Most of the average firms meandered along in a fairly uncompetitive manner. However, those firms that qualified and received CTCP support have seen major improvements in economic, efficiency, and competitiveness performance indicators. The IDC surveyed a population of 516 CTCP beneficiary companies in 2017, and secured data from 148 recipients who were able to provide data for the period 2011 to 2016 (Table 18.2). At the top end of the clothing industry production spectrum there have been a small number of firms that have utilized the support measures in conjunction with a couple of the clothing retail lead firms to achieve value chain alignment. They have created

390    Mike Morris, Justin Barnes, and David Kaplan Table 18.2: Clothing and textiles key performance indicator data Indicator

Units

2011

2016

Change

MVA

Rm

4,793

7,706

60.8%

Employment

Jobs

27,239

31,810

16.8%

Sales

Rm

10,683

19,846

85.8%

Production

Units

778m

962m

23.7%

MVA as a % of sales

%

49.0

47.7

–​2.7%

MVA per employee

R’000

214.5

262.3

22.3%

Absenteeism rates

%

5.4%

4.7%

12.0%

Customer return rates

%

3.6

2.2

38.2%

On-​time and in-​full delivery to customers

%

88.1

94.4

7.2%

Internal rework rates

%

5.1

3.4

34.4%

Source: Barnes and Hartogh (2018).

individually driven retail supply chain clusters, with suppliers and buyers working closely together to embrace upgrading challenges. The aim has been to support the adoption of Quick Response (QR) platforms in order to compete with cheaper imports. QR leveraged off the main geographical advantage that local apparel manufacturers had in taking of their localness and introducing short production cycles to provide retailers with speed and flexibility of supply. It also required retailers shifting from buying cheap goods on long lead times, maintaining large stocks and having massive sales at season end, to moving to a new retailing model based on minimizing inventories and increasing their returns through repeatedly turning over stock at full price within the year. At the other end of the production spectrum, those apparel manufacturers operating at the bottom to lower-​middle end of the market have responded in a different manner. Some relocated to Lesotho and Swaziland to escape the restrictive labour market conditions and wage systems, and exported back into the South African market (Morris et  al. 2011, 2016). In Newcastle and Ladysmith, a number of Chinese-owned firms, caught by the tendency to close the gap between metro and non-​metro wages, simply ignored the bargaining council wages and paid a wage that local workers would accept to secure employment (Nattrass and Seekings 2014, whilst other small firms simply hid beneath the radar in the informal economy. A strength and weakness analysis undertaken in 2017 of firms all along the clothing and textile value chain reveal the following. Strengths include quick response retailing developments; labour availability; institutional support for the supply chain; and local market size and potential. Weaknesses encompass weak human capital and associated skills; outdated production processes and equipment; local supply chain capability deficiencies; labour market issues; management deficiencies; inappropriate government policies and interventions; and corruption (Barnes and Hartogh 2018).

Value Chains and Industrial Development in South Africa    391 Although this summarizes the strengths and weaknesses of the aggregate of firms in the clothing and textile industry it does not take into account the divergent nature of the growth paths prevalent in the industry. In many respects its highly bifurcated nature is shaped by the fact that it is inserted in a domestically driven GVC which is heavily impacted on, and influenced by, global imports.

18.4 Conclusion The two case studies highlight the inescapable impact of GVCs on South Africa’s manufacturing sector. Whether exporting into advanced developed economy markets, as per the automotive industry, or supplying primarily into the South African market, as per the clothing and textiles industry, the specialist capabilities developed within GVCs are clearly framing the operating space for locally based operations. This has partly been driven by the country’s reduction of import tariffs over the last two decades, and partly by the continued growth and development of multinational-​led business models that have extended and deepened the reach of GVCs. Manufacturers in South Africa no longer have the space to develop their capabilities in a domestic market, which provides some level of protection from the cold winds of international competition. Firms essentially need to be born export competitive, as this is the operating standard that needs to be attained whether supplying the domestic or export markets. The consequence of this for established manufacturers has been profound, as evident from the two case studies. Firms have upgraded their capabilities and succeeded in competing against their international counterparts, upgraded and still failed, re-​located their operations regionally, or simply wound their operations down over time. While the pockets of firms in the first category have continued to provide government and other industry stakeholders hope of a manufacturing renaissance in South Africa, the reality is that far more firms have fallen into the latter three categories. This is evident in the capital investment data and gross fixed capital formation data for South African manufacturers. These firms have moreover signalled to foreign investors and local entrepreneurs alike the challenges of manufacturing in South Africa, leading to reduced FDI and the limited establishment of new locally owned firms. It is perhaps unsurprising then that it is the South African automotive industry that has led the competitiveness drive within the broader manufacturing sector. The local subsidiaries of the vehicle assembly lead firms were forced into upgrading their capabilities to compete, and as heads of supply chains they have similarly driven their suppliers to improve their competitiveness. Coupled with supportive policy from government, these firms were largely able to craft out positions for their South African operations within their GVCs. They secured export contracts and invested in their capabilities. Fundamentally, however, they did so because of their sunk investments in South Africa and the initial scale and sophistication of their local operations. This

392    Mike Morris, Justin Barnes, and David Kaplan gave them a base to upgrade and to develop their capabilities—​as they were integrated into GVCs. The clothing and textiles industry did not receive the same type of support from the lead firms in their GVC, the domestic retailers, and as such typically decided not to invest in upgrading their capabilities. Consequently, most ceased operating, or slowly reduced their operating footprint in accordance with their diminishing product and production capabilities. This perpetuated a cycle of diminishing capabilities, firm closures, and increasing imports. However, more recently this has begun to change. As South African retailers respond to the entry of international retailers in the domestic market, and expand internationally, they have increasingly recognized the benefits of having agile, flexible, local supply chains, and have begun investing in their supply chains. While this may neither completely reverse the decline of the clothing and textiles industry nor shift most retail purchasing back on shore, it does suggest the emergence of a new lead firm-led business model that holds considerable potential in rebuilding at least a part of the domestic clothing and textiles value chain. This potential is recognized in the national government’s recently announced South African Retail-​Clothing, Textiles, Footwear and Leather Value Chain Masterplan, which focuses on the competitive advantage domestic retailers receive when purchasing from local manufacturers on short lead times. The masterplan includes a commitment from South African retailers to increase their local purchases to 65 per cent of total procurement. There is, however, a proviso that the government supports the development of competitiveness capabilities within the domestic supply chain and that there is a demonstrable improvement in the competitiveness capabilities of local manufacturers (encompassing their cost, quality, speed, and flexibility capabilities). How do regional value chains fit into the complex intersection of domestic and global value chain drivers? Again, this appears to be context specific. The automotive GVC is dominated by multinational vehicle assemblers that organize their global production footprint on the basis of a global strategy, while the clothing and textiles value chain is dominated by JSE-​listed retailers looking to optimize their sourcing models to compete more effectively domestically. How do regional value chains support the effectiveness of their primary strategies? Thus far, regional purchasing by South African retailers has opened up substantial opportunities for suppliers in Lesotho, Swaziland, Mauritius, and Madagascar to further extend and build their capabilities. It is still too early to predict how these regional value chains will develop and whether local clothing firms will take advantage of backward linkages into the region. As regards the automotive industry the operational demands and requirements of lead firms would appear to open up only limited opportunities for regional suppliers except in particular niche operations. Ultimately it lies in the hands of the lead automotive firms in terms of locating some of the assembly operations and sales activities as they attempt to spread into other key countries in sub-​Saharan Africa. How an economy like South Africa positions itself within these opportunities is dependent on the strategic positioning of lead firms within GVCs (and the role they articulate for their South African operations and suppliers); the strength of the domestic economy; narrow sector-​specific industrial policies that frame the development of the

Value Chains and Industrial Development in South Africa    393 country’s leading value chains; and the broader transversal government policies that either enhance or undermine South Africa’s industrialization potential and hence its position as the ‘natural gateway into Africa’.

References Abernathy, Frederick H., Anthony Volpe, and David Weil. 2006. ‘The future of the apparel and textile industries: Prospects and choices for public and private actors’, Environment and Planning A, 38(12): 2207–​32. African Economic Outlook. 2014. Global Value Chains and Africa’s Industrialisation. African Development Bank, Organisation for Economic Co-​operation and Development, United Nations Development Programme. AUC/​OECD. 2019. Africa’s Development Dynamics, 2019: Achieving Productive Transformation. Addis Ababa: AUC; Paris: OECD. Baldwin, Richard. 2012. ‘Global supply chains: Why they emerged, why they matter, and where they are going’. CEPR Discussion Paper no. 9103. Barnes, Justin. 2013. ‘Capital structure of the South African automotive industry: Historical perspectives and development implications’, Transformation, 81(2): 236–​59. Barnes, Justin. 2017. ‘The automotive GVC:  Policy implications for developing economies’, in J. Keane and R. Baimbill-​Johnson (eds) Future Fragmentation Processes:  Effectively Engaging with the Ascendency of Global Value Chains. London: The Commonwealth, 133–​43. Barnes, Justin, and Anthony Black. 2014. ‘The motor industry development programme 1995–​ 2012: What have we learned?’. International Conference on Manufacturing-​Led Growth for Employment and Equality in South Africa, Johannesburg, South Africa, 20–​21 May. Barnes, Justin, and Anthony Black. 2017. ‘Developing a South African automotive masterplan to 2035 in the context of global value chain drivers:  Lessons for second-​tier automotive economies’. Conference Paper, GERPISA Colloquium, 14–​16 June, Paris. Barnes, Justin, and Mike Morris. 2008. ‘Staying alive in the global automotive industry: What can developing economies learn from South Africa about linking into global automotive value chains?’, European Journal of Development Research, 20(1): 31–​55. Barnes, Justin, and Tamryn Hartogh. 2018. ‘Status quo of the South African retail-​CTFL value chain’. Report 1 of 4 of the South African Retail-​CTFL Masterplan Project, B&M Analysts for the Department of Trade and Industry, Durban. Barnes, Justin, Anthony Black, and Kriengkrai Techakanont. 2017. ‘Industrial policy, multinational strategy, and domestic capability: A comparative analysis of the development of South Africa’s and Thailand’s automotive industry’, European Journal of Development Research, 29: 37–​53. Barnes, Justin, Alec Erwin, and Faizel Ismail. 2019. ‘Realising the potential of the Sub-​Saharan African automotive market: The importance of establishing a sub-​continental Automotive Pact’. Report compiled for Trade and Industrial Policy Strategies and the Association of African Automotive Manufacturers, 31 July. Barnes, Justin, Raphael Kaplinsky, and Mike Morris. 2004. ‘Industrial policy in developing economies: Developing dynamic comparative advantage in the South African automobile sector’, Competition and Change, 8: 153–​72. Black, Anthony. 2009. ‘Location, automotive policy and multinational strategy: The position of South Africa in the global automotive industry since 1995’, Growth and Change, 40(3).

394    Mike Morris, Justin Barnes, and David Kaplan Black Rock Institute. 2017. ‘Future of the vehicle: Winners and losers, from cars and cameras to chips’. Global Insights, April. Bloomberg New Energy Finance. 2016. ‘Advanced transport:  Research note’. Bloomberg Finance, 25 February. Boston Consulting Group. 2016. ‘Self-​driving vehicles, robo-​taxis, and the urban mobility revolution’. [online] http://​www.automotivebusiness.com.br/​abinteligencia/​pdf/​BCG_​ SelfDriving.pdf. Edwards, Lawrence, and Mike Morris. 2007. ‘Undressing the numbers: The employment effect of import quotas on clothing and textiles’, Journal of Development Perspectives, 2(2): 121–​40. Flatters, Frank. 2005. ‘The economics of the MIDP and the South African motor industry’. Paper presented at the TIPS/​NEDLAC South Africa Trade and Poverty Programme Policy Workshop, Johannesburg. Flatters, Frank, and Nnzeni Netshitomboni. 2007. ‘Trade and poverty in South Africa: The Motor Industry Development Programme’, Studies in Economics and Econometrics, 31(2). Foster-​McGregor, Neil, Florian Kaulich, and Robert Stehrer. 2015. ‘Global value chains in Africa’. Research, Statistics and Industrial Policy Branch Working Paper no. 4/​2015, UNIDO, Vienna. Gereffi, Gary, and Stacey Frederick. 2010. ‘The global apparel value chain, trade and the crisis: Challenges and opportunities for developing countries’, in O. Cattaneo, G. Gereffi, and C. Staritz (eds) Global Value Chains in a Post-​crisis World. Washington, DC: World Bank. Gereffi, Gary, John Humphrey, Raphael Kaplinsky, and Timothy Sturgeon. 2001. ‘Introduction: Globalisation, value chains and development’, IDS Bulletin, 32(3): 1–​8. Gereffi, Gary, John Humphrey, and Timothy Sturgeon. 2005. ‘The governance of global value chains’, Review of International Political Economy, 12(1): 78–​104. Harvard University. 2019. Atlas of Economic Complexity. https://​atlas.cid.harvard.edu/​. ITC Trade Map. 2017. Data and Statistics. http://​www.trademap.org/​index.aspx. Kaplinsky, Raphael. 2013. ‘Global value chains, where they came from, where they are going and why this is important’. IKD Working Paper no. 68. Kaplinsky, Raphael, and Mike Morris. 2001. A Handbook for Value Chain Research [online] https://​www.globalvaluechains.org/​tools.html. Monaco, L., Justin Barnes, and Anthony Black. 2018. ‘Multinational strategy, structural change and supply chain development in the South African auto industry’. Conference Paper, GERPISA Colloquium, 11–​14 June, Sao Paulo. Morris, Mike, and Justin Barnes. 2014. ‘The challenges to reversing the decline of the apparel sector in South Africa’. International Conference on Manufacturing-​ Led Growth for Employment and Equality in South Africa, Johannesburg, 20–​21 May. Morris, Mike, and Gil Einhorn. 2008. ‘Globalisation, welfare and competitiveness: The impacts of Chinese imports on the South African clothing and textile industry’, Competition and Change, 12(4). Morris, Mike, and Brian Levy. 2017. ‘The limits of co-​operation in a divided society: The political economy of South Africa’s garment and textile industry’, in Anthony Black (ed.) Towards Employment Intensive Growth in South Africa. Cape Town: University of Cape Town Press. Morris, Mike, Leonhard Plank, and Cornelia Staritz. 2016. ‘Regionalism, end markets and ownership matter: Shifting dynamics in the apparel export industry in sub-​Saharan Africa’, Environment and Planning A, 48(7): 1244–​65.

Value Chains and Industrial Development in South Africa    395 Morris, Mike, Cornelia Staritz, and Justin Barnes. 2011. ‘Value chain dynamics, local embeddedness, and upgrading in the clothing sectors of Lesotho and Swaziland’, International Journal of Technological Learning, Innovation and Development, 4, 1/​2/​3: 96–​119. Nattrass, Nicoli, and Jeremy Seekings. 2014. ‘Job destruction in Newcastle: Minimum wage-​ setting and low wage employment in the South African clothing industry’, Transformation, 84: 1–​30. Staritz, Cornelia. 2011. Making the Cut? Low-​income Countries and the Global Clothing Value Chain in a Post-​quota and Post-​crisis World. A  World Bank Study. Washington, DC: World Bank. Sturgeon, Timothy, and Richard Florida. 1999. ‘The world that changed the machine: Globalization and jobs in the automotive industry’. Final report to the Alfred P. Sloan Foundation, International Motor Vehicle Program. UNCTAD. 2013. World Investment Report. UNCTAD: Geneva. Womack, James, Daniel Jones, and Daniel Roos. 1990. The Machine That Changed the World. New York: Rawson Associates.

Chapter 19

S ou thern A fri c a n Regional Valu e C ha i ns and Integ rat i on Reena das Nair

19.1 Introduction Global trade has changed substantially since the early twentieth century, with the share of developing countries in global production and trade steadily rising. The contribution of the ‘Global South’1 to world trade reached around 50 per cent in the 2000s, with almost half of global manufacturing export value emerging from the Global South (Horner and Nadvi 2018). Within this, ‘south–​south’ exports grew faster than total world exports between 1995 and 2016 (at annual average rates of 13 per cent and 8 per cent respectively), amounting to US$4 trillion in 2016. The growing south–​south trade is driven in part by increasing final demand and consumption, spurred by urbanization and rising incomes. Foreign direct investment (FDI) from developing countries also increased from around US$110 billion in 2005 to US$381 billion in 2017 (around 30 per cent of global flows) (UNCTAD 2018). This steep growth in south–​south trade has been concentrated in Asia (UNCTAD 2018), but the significance of intra-​regional trade in Africa is increasing, although from low levels. Intra-​African exports were around 10 per cent of total African exports in 1995, and grew to 17 per cent in 2017, with varying degrees of intra-​regional trade and integration within the different Regional Economic Communities (RECs) (Songwe 2019). The creation of a single market for goods and services, free movement of people, and expanded intra-​Africa trade as part of the African Continental Free Trade Agreement

1  Classified

by the World Bank as low-​or middle-​income countries in Africa, Asia, Oceania, Latin America, and the Caribbean.

Southern African Regional Value Chains and Integration    397 (AfCFTA) signed in 2018, is expected to greatly boost trade and integration in Africa, and to shape regional value chains (RVCs) by opening up markets. A value chain refers to the full range of activities and value-​addition firms engage in to bring products or services from conception to end use (Gereffi and Fernandez-​ Stark 2016). RVCs have suppliers and buyers operating within one region (for example, within Africa or within RECs). Conceptually similar, global value chains (GVCs) refer to activities undertaken by geographically dispersed players across continents. RVCs can boost regional trade, investment, and development through linkages and capability upgrading, and therefore are an integral component of regional integration. Global trends are fuelling growing interest in RVCs. RVCs are emerging not only due to the slowing down of GVCs in global trade and investment, but are also spurred by new production technologies that allow for less fragmentation and more local production, and the ‘backshoring’ of production activities to home countries (Dachs and Pahl 2019). RVCs are further becoming important given growing uncertainty in global markets and trade conflicts, as well as possibilities for greater value-​addition and structural transformation for developing countries than offered by GVCs in some industries (Andreoni and Boys 2020). The regional integration agenda in Africa is not a new one and was set in motion by the establishment of the Organisation of African Unity in 1963. The African Economic Community (AEC) sought to create a customs union, a single market, and an economic and monetary union. In 2002, the African Union (AU) was established to reignite the integration agenda (UNCTAD 2018). Yet, relative to the developments in Europe, North America, Asia, and Latin America, there has been limited progress in integrating the African continent through stimulating intra-​regional trade. The RECs were positioned as the building blocks for the AEC. They differ in structure and mandate, and have developed at different paces, with eight AU-​recognized RECs playing key roles in pursuing regional integration (UNCTAD 2018). The levels of intra-​regional trade and integration vary widely in the RECs, with the Southern African Development Community (SADC) and the East African Community (EAC) achieving the highest levels of intra-​regional exports, at 15 per cent and 19 per cent on average, respectively, over 1995–​2019 (UNCTAD trade data). South Africa is a member of SADC comprising sixteen member states and is also a member of the Southern African Customs Union (SACU), with four other members. Notwithstanding relatively more progress of SADC than the other RECs in terms of trade, few RVCs have developed in which significant value-​addition is from within the region. Drawing from past and ongoing research in selected agro-​processing value chains, through case studies, this chapter examines why RVCs have not developed as much as they could have. The chapter uses a GVC framework to unpack the nuances that affect participation, investment, and upgrading in these RVCs, and explores factors at firm and value-​chain level, such as governance structures, upgrading pathways, and the extension of strategies of South African firms with market power as they spread in the region. Political economy dynamics, which are critical in understanding the outcomes seen, are evaluated. The chapter further briefly assesses the record of regional integration

398   Reena das Nair in southern Africa against key stated objectives of SADC and SACU, highlighting the skewed position of South Africa in the region, and contributing to understanding the slow development of RVCs. Agro-​processing value chains were selected given their general importance for RVCs and regional integration in southern Africa in several respects. First, many countries in Africa are net importers of processed foods. Food prices in sub-​Saharan Africa have been found to be 30–​40 per cent higher than in comparable countries when differences in income are accounted for (Allen 2018). However, spurred by the growth of agricultural productivity and technological developments, there is potential to produce more food products, replace imports, and drive down prices. This has important implications for regional food security. Second, there are other positive economic and social impacts of developing food RVCs. While traditional ‘smokestack’ manufacturing activities were touted to be the path to structural transformation in developing countries, there is growing recognition of the high levels of industrial sophistication in producing ‘agro-​ industrial’ products. Termed ‘industries without smokestacks’, these have the potential to increase value-​addition per worker, grow trade, and transform and grow economies in Africa (Newfarmer, Page, and Tarp 2018). Third, processed food value chains have forward and backward linkages with important multiplier effects, and opportunities for inclusive participation. Developing food RVCs not only benefits farmers and rural communities, but also benefits input suppliers further up the value chain (e.g. fertilizers, seed, chemical, farm equipment) and ancillary industries (e.g. packaging, processing equipment, cold chain, logistics). Last, RVCs in food in southern Africa are important, given climate change. With low rainfall and drought in the southern ​most parts of the continent, but with good rains in other parts in the same periods, the choice of location for agricultural activities is critical. Food production will increasingly come under pressure and hard decisions on where to produce different foods need to be made. Developing RVCs in food could mitigate these risks for the region (Paremoer 2018; Roberts 2019). Within agro-​processing, case studies in the animal feed-​to-​poultry and sugar-​to-​ confectionery value chains are undertaken, along with the closely linked downstream retail or supermarket level. These value chains were selected because they illustrate key dynamics that affect upgrading and participation. The investments, coordination, governance structures, and market power in the hands of a few lead South African firms in these value chains affect development, with implications for regional integration. The chapter illustrates the importance of tailored and coordinated interventions and investments by firms and governments at a regional level to develop RVCs. National policymakers have yet to appreciate the potential that RVCs have to drive inclusive growth, and few structured interventions have been designed with a regional development agenda in mind. Section 19.2 of this chapter sets out the framework used for assessment. In section 19.3 case studies in the selected agro-​processing and retail RVCs are undertaken. In section 19.4 the record of regional integration in SADC and SACU against stated objectives is briefly discussed. Section 19.5 concludes with policy measures to build RVCs and promote greater regional integration.

Southern African Regional Value Chains and Integration    399

19.2  A GVC Framework to Understand RVC Dynamics The GVC framework has long been used to evaluate inter-​firm linkages and vertical relations, and to assess governance and the impact of the power of lead players on upgrading in value chains (Kaplinsky and Morris 2014; Gereffi and Fernandez-​Stark 2016). The framework allows for practical policy actions to develop a value chain, redistribute rents or upgrade certain actors, particularly in developing countries. It can be used to identify opportunities and bottlenecks for upgrading capabilities at particular levels. The related Global Production Networks (GPN) framework adds the significance of institutions, embedded network relationships, and state and non-​chain actors in explaining outcomes (Coe and Yeung 2015). Governance and upgrading are particularly relevant in GVCs. Governance is the process of exercising control along the value chain through specification of what product should be supplied, what quantity, at what price, and how it should be produced. It refers to authority and power relationships that determine the allocation and flow of resources and rents within a value chain. Lead firms with this power are often ‘gatekeepers’ in value chains (Gereffi and Lee 2014). In early GVC studies, governance was assessed in one of two ways, whether value chains were producer or buyer driven. Producer-​driven value chains are typically found in industries with high technological, capital, and proprietary requirements. Buyer-​driven value chains, for instance those driven by large retailers, are typically found where market information, product design, marketing, and advertising costs set barriers to entry (Humphrey and Schmitz 2002). The GVC literature has since progressed to more nuanced governance categories based on informational complexity, supplier capabilities, codification of information, and switching costs. Under these criteria, governance of value chains can be classified—​in order of increasing level of control over suppliers—​as market, modular, relational, captive, or hierarchical governance at different nodes (Gereffi, Humphrey, and Sturgeon 2005). There is also growing recognition that the exertion of power in value chains is not always limited to a lead firm, or to dyadic, bargaining relationships between buyers and sellers. There are often other dimensions of power exercised by various stakeholders (Dallas, Ponte, and Sturgeon 2019). Governance affects upgrading, which refers to firms or countries maintaining or improving their positions within value chains. There are different types of upgrading—​ process upgrading (transformation of inputs into outputs more efficiently by reorganizing the production system or introducing superior technology); product upgrading (moving into more sophisticated products); functional upgrading (acquiring new functions to increase overall skill content of activities); and intersectoral upgrading (moving into new activities) (Humphrey and Schmitz 2000). For the previously dominant south–​north direction of trade, the GVC and GPN frameworks provided useful tools to assess the implications for Global South actors

400   Reena das Nair supplying goods to lead firms in the north. The ‘Rise of the South’ however requires an adapted lens through which the impact on firms, industries, workers, and consumers can be understood (Pickles, Barrientos, and Knorringa 2016; Horner and Nadvi 2018). As noted, global trends are shaping growing interest in RVCs, shifting the policy focus towards developing RVCs as pathways to industrialization, rather than trying to upgrade to plug into GVCs. Advances in technology, global uncertainty, not least due to COVID-​19 and trade conflicts, all contribute to the importance of RVCs. While GVC and GPN principles apply to RVCs, there are important nuances in the assessments of RVCs. A fundamental premise for RVCs as a development tool is that increased market size and easier market access spurs production and facilitates industrialization (Paremoer 2018). Entry barriers into RVCs may be lower than into GVCs because they are likely to be less tightly governed or controlled, or because there is often less pressure from consumers, governments, and civil society (Kaplinsky and Farooki 2011). Firms can therefore participate with lower levels of product development, marketing, and distribution capabilities than required in GVCs. Lower standards requirements can also present more surmountable entry barriers for small and medium-​ sized enterprises (SMEs). Similarities in cultures, tastes, and preferences within regions also lower barriers, making it easier to adapt to host-​market demands. There are obvious geographic benefits in terms of shorter distances. With suppliers in closer proximity to final consumption, transport and related costs are potentially lower. RVCs can also act as ‘stepping-​stones’ for firms to enter GVCs once they develop capabilities to compete internationally (Keane 2015). However, as noted, participation in GVCs may not necessarily be the desired goal, and as argued below for the case of agro-​processing, developing RVCs is more desirable than participating in GVCs. The political, social, and economic realities in different countries in a region affect regional development. Uneven starting points in levels of development create tensions between neighbours, and national development tends to take priority over regional development in SADC. Powerful lobby groups and dominant players further influence policies, affecting who has access to resources and who can extract rents in value chains, including through lobbying. The pattern of FDI in RECs further impacts RVC development. Section 19.3 highlights how a few large South African firms dominate regional markets in ownership, production, and investment. At the upstream level, large South African players dominate input markets like fertilizers and seed, while at the food processing level, South African producers have large shares in several product markets. At the retail level, a handful of South African supermarket chains dominate. While opening markets presents upgrading opportunities for regional players to participate in value chains, and for the transfer of information, capabilities, and technology, the dominance of a few large firms with market power raises concerns of anticompetitive behaviour extending to the region. Regional competition body, the COMESA Competition Commission, and the SADC declaration on Regional Cooperation in Competition and Consumer Policies, serve to address competition problems with regional impacts. Yet, few cross-​border competition transgressions have been investigated and prosecuted in the region, even

Southern African Regional Value Chains and Integration    401 when the same firms are found implicated in anticompetitive conduct in one country and have operations in, or export into, neighbouring countries (Kaira 2015). Cartel conduct, abuses of dominance, and the exertion of buyer power in one country can spill into other countries through the extension of business strategies. This creates barriers to entry, distorts trade, and undermines RVCs and integration. As discussed in section 19.3, anticompetitive behaviour by firms with market power has been a feature in the selected agro-​processing value chains.

19.3  RVCs in Agro-​processing Developing RVCs in food can make the SADC region less reliant on imports, and stronger capabilities and competition can reduce food prices and insecurity. The development of agro-​processing industries can also drive industrialization (see section 19.4). Governance structures and upgrading of capabilities that affect the development of RVCs are value-​chain specific. To illustrate these concepts, case studies in the animal feed-​to-​poultry and the sugar-​to-​confectionery value chains are presented in section 19.3.1. A critical level of these value chains, the downstream retail level, is evaluated in section 19.3.2. A few lead South African firms with market power operate in these value chains in SADC. The extension of strategies of these firms affects participation and upgrading. Even if not physically located in the region, large South African food processors significantly impact the region through exports. South Africa was a large net exporter of food and beverage products to SADC between 1995 and 2019 (see Figure 19.1b). This has been spurred by South African supermarket chains growing their store networks and procuring products through imports from processors in South Africa (section 19.3.2). There is limited reciprocal imports of food products through the supermarkets from the region into South Africa. The potential in two food value chains, animal feed-​to-​poultry and sugar-​to-​confectionery, to develop into RVCs is discussed below.

19.3.1 RVCs in Selected Agro-​processing Chains 19.3.1.1 Animal Feed-​to-​Poultry Growing incomes and urbanization have driven the demand for poultry as a key source of affordable animal protein. Producing around 1.7 million tonnes, contributing to almost 17 per cent of total agricultural value and employing around 112,000 people, the poultry industry in South Africa is important (Goga and Bosiu 2019). Demand in South Africa has increasingly been met by imports, which in 2017, accounted for 24 per cent of demand. Consumers in Europe and South America prefer white meat in breast portions, and historically, the less-preferred bone-​in portions were ‘dumped’ into South

402   Reena das Nair (a) Proportions of diversified exports by destination 60% 50% 40% 30% 20% 10%

Africa Europe

SADC Oceania

Americas Antarctica

2019

2018

2017

2016

2015

2014

2013

2011

2012

2010

2009

2008

2007

2006

2005

2004

2003

2001

2002

2000

1999

1998

1997

1996

1995

0%

Asia Not allocated

(b) Diversified exports by industry 1500

500 0

1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019

US Dollar, millions

1000

Food, beverages and tobacco Textiles, clothing and leather Wood and of products of wood and cork, except furniture Other non-metalic mineral products Electrical machinery and apparatus Radio, television and communication equipment and apparatus and medical, precision and optical instruments Transport equipment

Figure 19.1  South Africa’s exports of diversified manufactured goods to SADC Source: Quantec. Notes: 1. Diversified exports exclude basic metals, coke and petroleum, and basic chemicals. 2. In 2010 an adjustment was made to the data to capture South Africa’s exports to SACU countries. The figures in previous years are therefore understated, and the adjustment accounts for a substantial proportion of the jump of exports in 2010 (Roberts and Arndt 2018).

Africa by poultry producers in these regions (ITAC 2015). Growing imports of bone-​in chicken portions triggered interventions in South Africa (the latest being the Poultry Master Plan) to improve competitiveness and grow local production and exports (South African Poultry Sector Master Plan 2019).

Southern African Regional Value Chains and Integration    403 There is a high degree of vertical integration in the poultry industry. Four large firms, RCL Foods (South African), Astral (South African), Country Bird Holdings (CBH, South African), and Irvine’s Africa (Zimbabwean), either own or control key inputs of animal feed and the licences for breeding stock in the region. Two firms have exclusive national breeding licences for South Africa (Astral and RCL), while the other two have exclusive multi-​territorial breeding licences for the most widely used breeds in the region. Control of breeding licences in turn controls critical and profitable levels of the value chain—​the breeding and sale of parent stock and day-​old chicks—​in which high rents are made (Ncube 2018; Ncube, Roberts, and Zengeni 2016). These firms have invested in the region, expanding their operations mainly through acquisitions (Goga and Bosiu 2019). By controlling key inputs, they control entry and terms of participation. Poultry producers are forced to buy breeding stock, and in some instances animal feed, from them, and therefore they act as gatekeepers to the value chain. Forms of hierarchical governance by these large players are evident in relationships through this control. Vertical integration and control have acted as barriers to entry, and abuses of positions of market power have led to investigations by the Competition Commission of South Africa. There have been cases on abuse of dominance and cartel conduct at different levels of the value chain. For example, when CBH entered into a joint venture (JV) with Astral in South Africa, JV conditions precluded it from sourcing breeding stock from any other player through an exclusive supply agreement. This foreclosed CBH from access to other, possibly more competitively priced, inputs. Further, JV conditions prevented CBH at the time from entering into the upstream level itself for breeding stock in competition to Astral (Bosiu, das Nair, and Paelo 2017). After Astral reached a settlement with the Commission, CBH was able to introduce a new bird breed in South Africa and has since grown significantly in the region (Ncube 2018). Captive forms of governance are also evident with respect to relationships with contract growers. In South Africa, 60–​70 per cent of poultry production of the larger players is outsourced to contract growers. These third-​party growers are often forced to exclusively use the lead firm’s breeds and animal feed (Ncube 2018). The concentrated nature of the value chain has further facilitated cartel conduct in South Africa. The Competition Commission reached a settlement in 2010 following a leniency application implicating lead firms in price fixing of fresh chicken (Bosiu, das Nair, and Paelo 2017). There has also been collusion at the animal feed level (Ncube, Roberts, and Zengeni 2016). Such conduct raises costs for new entrants and maintains the power of incumbents in RVCs. There are opportunities to add value along the chain in different countries in the region that are not being exploited. These opportunities can lead to a more regionally competitive poultry value chain, potentially displacing deep-​sea imports. South Africa also imports significant volumes of soybean oilcake to make animal feed. There has been considerable investment in soybean crushing capacity in South Africa, with the Department of Trade, Industry and Competition pursuing an import-​substitution strategy to increase local crushing/​processing (Paremoer 2018). The outcomes of this investment are seen in the change of South Africa’s net import to net export status of

404   Reena das Nair residues (which contain soybean meal: Quantec data). Even though soybean crushing capacity has increased significantly, South Africa still needs to import soybean products from Brazil, Argentina, and Paraguay to satisfy demand. South Africa does not have suitable agro-​ecological and climatic conditions to grow more soybeans to meet the local demand of 2 million tonnes per annum (Paremoer 2018). Further, climate change is going to put even more pressure on South Africa’s ability to grow soybeans. However, other countries in the region like Zambia can supply the region with adequate support and at least partly replace imported soybeans. This requires offtake commitments from animal feed processors, and collaborative and coordinated investments by lead firms at a regional level to build capabilities in Zambia to increase scale, quality (which has been a concern), and cost-​competitiveness to make sourcing from Zambia attractive. It requires interventions to ensure competitive transport costs and climate-​smart investments in the chain (Paremoer 2018). Regional linkages are thus important for a competitive poultry value chain with respect to key inputs. The South Africa Poultry Master Plan concluded in 2019 focuses on protecting and promoting the national industry and driving local demand. It aims to promote investments in both large- and small-​scale, particularly Black-​owned, farming, and to improve feed costs, and meet safety and veterinary requirements. It also aims to increase breast meat exports to the European Union, which South Africa has been unable to do, given its inability to meet sanitary and phytosanitary standards (SPS). The Master Plan, however, does not have a regional focus. It does not engage with the sector under SADC regional industrialization objectives (section 19.4), nor consider climate change, which limits the ability of South Africa to produce cheap feed. It also does not consider the quid pro quo and goodwill gains that could result from the reciprocal access to end markets for South African poultry products through South African supermarkets in the region. There have been tensions in the region as South Africa is viewed as ‘extracting’ from the region without contributing to its development. As discussed below, this has resulted in protectionist measures by countries in the region against imports from South Africa. The poultry Master Plan is a short-​term response to rising deep-​sea imports in South Africa, and not a long-​term view on how the value chain can be developed sustainably to benefit the whole region. This is partly reflective of political economy dynamics, where the lead South African poultry producers and industry associations have lobbying power to protect the national industry and jobs. Regional industrial development requires coordinated investments by the lead firms with operations across the region, supported by governments, to build scale and capabilities (Ncube 2018).

19.3.1.2 Sugar-​to-​Confectionery Like poultry, there is potential to develop the sugar-​to-​confectionery value chain in SADC. South Africa, Zambia, Malawi, Mauritius, Eswatini, and Mozambique are large net exporters of sugar, reflective of low production costs. Yet, the SADC region is a net importer of sugar confectionery products. Confectionery production (biscuits, sweets,

Southern African Regional Value Chains and Integration    405 chocolates, etc.) can grow significantly in the region, making it less import-​dependent (das Nair, Nkhonjera, and Ziba 2017). Global sugar production is highly protected and heavily subsidized. In SADC, there is a network of bilateral and multilateral agreements that control traded volumes. While there is a Sugar Cooperation Agreement incorporated into the SADC Trade Protocol with the ostensible objective to promote regional growth and competitiveness, the complex interlace of national, bilateral, and multilateral arrangements appears to have the opposite effect. Collectively, these arrangements serve to protect only sugar producers and a handful of large, powerful millers rather than to advance a regional industrial agenda (das Nair, Nkhonjera, and Ziba 2017). In Zambia, the prices for industrial and household sugar sold by quasi-​monopolist miller, Zambia Sugar, a subsidiary of South African Illovo Sugar,2 were the subject of an investigation by the Competition and Consumer Protection Commission. The investigation in 2017 found that Zambia Sugar abused its dominance by engaging in price discrimination and charging unfair prices. This negatively impacted industrial users of sugar, making them uncompetitive in confectionery production, despite Zambia being a low-​cost, net exporter of sugar. The Zambian household sugar market was further shielded from cheaper imports through Vitamin A fortification requirements, essentially acting as a non-​tariff barrier (NTB) and entrenching Zambia Sugar’s dominant position (das Nair et al. 2017; Black et al. 2020). Zambian confectionery producers have the capabilities to export more to the region, especially if they can source competitively priced sugar and access supermarket shelves. In South Africa, following deregulation of agricultural markets, the sugar industry was the only industry to remain regulated under the Sugar Act (1978), in conjunction with the Sugar Industry Agreement (2000). Bulk sugar exports from South Africa are managed by the Sugar Export Corporation, in effect controlling local available volumes. The Sugar Act and Sugar Industry Agreement collectively have been used to determine a ‘division of proceeds’ between sugarcane growers and millers (dominated by three large players), with the purpose of ensuring fair and equitable returns. Over the years, the sugar industry has also been able to lobby for import duties to protect millers and growers from cheap imports. The benefits of these pricing arrangements and protectionism, however, do not trickle downstream to the confectionery industry (das Nair et al. 2017). The South Africa Sugar Master Plan of 2020 aims to protect jobs, rural livelihoods, and businesses, as well as create diversified revenue streams for producers. The extent to which it promotes the competitiveness of downstream confectionery producers is unclear, with the emphasis again being on the upstream industry’s role as a critical employer, including for rural communities. There are also no indications that it takes a wider regional integration and development view. This case study highlights how a

2 

Subsidiary of Associated British Foods.

406   Reena das Nair combination of trade policy, NTBs, and market power of lead firms affects regional trade and the emergence of RVCs.

19.3.2 The Growth and Spread of Supermarket Chains— ​a Driver of Food RVCs Supermarket chains are a growing route to market for suppliers of processed foods. Rising levels of urbanization have spurred the growth and spread of supermarket chains since the mid-​1990s. In South Africa, the end of apartheid saw the main chains expanding within urban areas through locating in shopping malls and centres, but also spreading to peri-​urban and rural areas. Along with the geographic spread, their formats and offerings also proliferated to target consumers in all income groups. This extended their control in the grocery retail landscape in South Africa (das Nair 2019)  and contributed to the ‘supermarket revolution’ or ‘supermarketization’, which refers to the global trend of retail food sales increasingly going through supermarkets (Reardon, Timmer, and Berdegué 2004). While early predictions were that the supermarket revolution in Africa would follow trends similar to those seen in Latin America, where transnational chains entered and proliferated rapidly (Traill 2006; Reardon, Timmer, and Berdegué 2004), this was not the case in Africa. Instead, there was a ‘regionalization’ of supermarket chains, with South African chains spreading in southern Africa and Kenyan chains in East Africa. The main South African supermarket chains, Shoprite, Pick n Pay, SPAR, Woolworths, Fruit and Veg City (Food Lovers’ Market) and Game (Massmart/​Walmart), all spread into the region. In 2020 shoprite, the largest, had 2,829 stores in fifteen countries (Shoprite 2020), while Pick n Pay, the next biggest, had 1,925 stores across seven African countries (Pick n Pay 2020). The growing store numbers of the South African chains has been a key driver of processed food products exported from South Africa into the region. Supermarket procurement strategies, which mainly involve centralized procurement from suppliers large enough to meet volume and other requirements, mean that large suppliers located in South Africa supply the region through supermarket distribution centres or direct exports (das Nair 2019). Over and above legal requirements, the supermarket chains often set private standards.3 This has spurred important upgrading of some food suppliers to meet legal and private supermarket standards. Studies have highlighted that suppliers have engaged in product and process upgrading, for instance, improving the quality of existing products; producing more sophisticated product lines;4 upgrading

3  Such as Hazard Analysis Critical Control Points, International Organisation for Standardisation and Food Safety System Certification accreditations and packaging requirements. 4  For example, new flavoured yoghurt and milk drinks, crumbed chicken, pre-​cooked and flavoured maize meal (das Nair et al. 2018).

Southern African Regional Value Chains and Integration    407 plant, machinery, and equipment; investing in better quality management systems and technologies; and acquiring higher hygiene and food safety standards. Other forms of upgrading have included investments in distribution and storage facilities. Functional and intersectoral upgrading has also been seen where food processors started producing their own packaging (das Nair, Chisoro, and Ziba 2018). While supermarket requirements have driven such upgrading in some suppliers, it has also led to the exclusion of other suppliers, particularly SMEs and suppliers from other countries in the region who do not have the capabilities or resources to meet requirements. SMEs are often forced to sell through alternative, sometimes more precarious, routes to market. An outcome has been that high levels of concentration in food processing markets are perpetuated, with mostly only larger players being able to upgrade. In South Africa, the ten largest food and beverages enterprises account for almost 40 per cent of income in the sector. Concentration levels are even higher within specific product markets (StatsSA 2017). The five main retailers account for 64 per cent of the national South African market in terms of sales, giving them considerable power (Competition Commission South Africa 2019). The spread of the main supermarkets into peri-​urban and rural geographies through tailored formats has also displaced independent and informal retailers like ‘spaza’ shops, contributing to rising concentration. While the degree of competition from other forms of grocery retail (independent retailers, informal retailers, and wet markets) differs widely in different countries in SADC, the South African supermarket chains have significant control over the suppliers that sell through them and shape many food value chains. This control is not just through procurement and standards requirements. Studies have highlighted the implications of uneven bargaining power between supermarket chains and suppliers in southern Africa. Supermarkets are able to exert buyer power by unilaterally demanding trading terms from suppliers, including forcing payment of a range of fees.5 Long payment periods negotiated to the benefit of the supermarket chain place pressure on supplier cash flow. It is estimated that these costs collectively take off around 15 per cent of the price that suppliers get, squeezing their margins, and if not justified by true costs to supply the supermarket chains, can be a means by which supermarkets extract rents in value chains (das Nair et al. 2018). Concerns about high concentration levels, the market power of supermarkets, and other features of the grocery retail sector triggered a Grocery Retail Market Inquiry (GRMI) by the Competition Commission of South Africa in 2015. In addition to other areas of inquiry, buyer power of the national retail chains was evaluated and recommendations were made in 2019 to finalize regulations and guidelines for the new Buyer Power provisions in the amended Competition Act to protect SME suppliers (Competition Commission South Africa 2019). This was subsequently finalized in May 2020. 5  Listing/​ support

fees, rebates, advertising allowances, promotion fees, settlement discounts, new store openings fees, etc. (das Nair et al. 2018).

408   Reena das Nair The GRMI also highlighted the powerful role that supermarket chains can play in the development of suppliers through supplier development programmes (SDPs), contributing to upgrading of supplier capabilities. The Inquiry gave recommendations that supermarkets should scale- ​up their SDPs to support SMEs, including through binding and escalating industry targets for procurement, and for a proportion of turnover to be spent on SME development (Competition Commission South Africa 2019). However, given the jurisdiction of the South African competition authority, these interventions only apply at a national level. Previous studies have also evaluated the importance of supermarket-​driven SDPs. As the final link to customers, supermarkets collect important information on consumer trends and purchasing patterns which are critical for suppliers. In South Africa, supermarket SDPs have evolved to provide support to qualifying suppliers to assist in their upgrading. Support has included technical assistance to meet minimum food safety and quality standards; access to infrastructure; training and mentorship; business development and management; and market-​readiness support. The programmes also generally include offtake and shelf space commitments, and preferential trading terms such as faster payments (das Nair and Landani 2020). However, in other countries in the region, the South African supermarket chains have few structured SDPs to develop processed food supplier capabilities. The lack of capabilities and capacity of local suppliers is often cited as a reason for extensive imports from South Africa (Ziba and Phiri 2017). Yet, supermarket local procurement initiatives and support measures in the region are mostly limited to fresh produce, with little support for suppliers of value-​added products. Support has mainly been reactive to pressures from governments to source more locally (das Nair and Landani 2020). Coupled with limited support from respective national governments, suppliers outside South Africa are unable to use the supermarket chains to sell via their store networks in the region. Procurement practices by the supermarket chains mainly through imports from South Africa thus displace locally manufactured food in host countries, leading to deindustrialization of their food processing industries. The persistent one-​way trade (Table 19.1) has resulted in mistrust and animosity towards South African FDI in the region and has led to protectionist measures being imposed to promote national industries. For instance, in Botswana, Zambia, and Zimbabwe there have been bans on poultry, maize meal, and cooking oil imports, while duties have been levied on other foods to protect local suppliers (das Nair et al. 2018; Black et al. 2020). An example of a national-​level intervention which combines incentives to develop suppliers and a voluntary code of conduct governing supermarket behaviour is the Namibian Retail Charter of 2016. Some South African supermarkets operating in Namibia have signed up for the charter committing to procure locally to boost local manufacturing and to develop local suppliers. The Namibian government spearheaded the development of the Charter and provided complementary support in terms of investments in infrastructure such as barcoding facilities. The initiative has seen some results with increased local procurement of certain food products (das Nair and Landani

Southern African Regional Value Chains and Integration    409 Table 19.1: Country share of intra-​SADC and total trade

Angola

SADC imports as a % of total imports

SADC exports as a % of total exports

2000

2000

2019

2019

6

5

7

5

Botswana

77

69

11

14

Comoros

13

8

1

12

DRC

15

28

29

40

Eswatini

94

76

76

79

Lesotho

73

81

28

40

Madagascar

6

10

3

6

Malawi

53

25

18

19

Mauritius

17

11

7

20

Mozambique

52

27

36

19

Namibia

87

63

32

35

Seychelles

14

8

3

3

South Africa

1

6

15

23

Tanzania

13

6

7

25

Zambia

69

45

39

21

Zimbabwe

49

79

17

73

Source: IMF Direction of Trade Statistics (2000).

2020). For supermarket chains to be a catalyst for regional industrialization and integration by driving RVCs, and to tackle regional level supply side constraints, such programmes need to be extended to suppliers in the region in a coordinated manner. Building suppliers in the region benefits supermarket chains in the long run. A more closely located and more competitive supplier base improves efficiencies, reduces the carbon footprint, and reduces direct logistics, warehousing, and transport costs. It also reduces reliance on imported food products and exposure to exchange rate risks. As noted, climate change will put pressure on food systems, and there will be an increasing need to diversify and develop stronger regional supplier bases to ensure shelves remain stocked. Supermarkets also benefit as lead firms when they participate in RVCs. By tailoring their business models to meet different challenges, they upgrade their own capabilities. For instance, Shoprite, as the first chain to expand into Africa, had to respond to major infrastructural challenges in the region, such as weak telecommunications. It had to invest in installing satellite link-​ups, systems for ordering stock, and global tracking systems for trucks. Shoprite was also the first company in the region to communicate electronically with customs authorities in some countries (Shoprite 2001).

410   Reena das Nair

19.4  Record of Regional Integration in Southern Africa In this section, the record of regional integration in southern Africa against key objectives of SADC and SACU is briefly assessed, further highlighting factors that affect regional integration and RVC development. It presents a high-​level overview of outcomes with respect to economic objectives and is not intended to be an exhaustive account of the successes and failures of SADC and SACU. Assessments of South Africa’s trade, including to the region and its position with respect to the rest of Africa, is undertaken in Chapters 20 and 21 in this volume, respectively. The regional integration agenda is central in the SADC Treaty of 1992, aiming to create free movement of people, capital, labour, goods and services, and reducing barriers within the region (Tralac 2013). SADC’s stated vision, mission, and objectives further speak to its commitment to deeper cooperation and regional integration.6 Similarly, SACU aims to promote regional integration and development, industrial and economic diversification, intra-​regional trade and investment, and global competitiveness.7 Regional integration has always been at the heart of the economic objectives of SADC and SACU. However, a lack of focus with multiple themes and priorities pursued simultaneously through a proliferation of plans, a lack of capacity, limited human and capital resources for effective implementation, and vastly different levels of development of member state economies, have meant that many (often over-​ambitious) targets have not been met and desired outcomes not achieved. To provide a more concrete implementation plan, the Regional Indicative Strategic Development Plan (RISDP) was approved by the SADC Summit in August 2003. A fifteen-​year plan (2005–​20), the RISDP defined priority intervention areas, policies, strategies, and activities necessary to achieve integration. It outlined the following targets for a trade-​led regional integration agenda—​the establishment of a Free Trade Area (FTA) by 2008; a SADC customs union by 2010; a common market by 2015; a monetary union and central bank by 2016; and a common regional currency and economic union by 2018. While the objectives of RISDP were not legally binding on member states, as part of the SADC Trade Protocol of 1996 (amended in 2010), the following legally binding trade objectives continued to be pursued by SADC: the gradual elimination of tariffs; adoption of common rules of origin; harmonization of customs rules and procedures; attainment of internationally acceptable standards, quality, accreditation, and metrology; harmonization of SPS measures; elimination of NTBs; and the liberalization of trade in services. The Protocol on Trade in Services was further signed 6  https://​www.sadc.int/​about-​sadc/​overview/​; https://​www.sadc.int/​about-​sadc/​overview/​sadc-​vision/​, accessed 1 December 2020. 7  https://​www.sacu.int/​show.php?id=395, accessed: 1 December 2020.

Southern African Regional Value Chains and Integration    411 in August 2012. In August 2020, the SADC Vision 2050 and the extension of the RISDP to 2030 were approved by the 40th Ordinary Summit (Gore 2020). Both aim to take forward the regional integration agenda building on previous commitments with renewed emphasis on industrial development and market integration, infrastructure, and human capital development, and linking to the AfCFTA (Tralac n.d.). Progress has been made in some areas in the first RISDP and on certain trade protocols. For instance, the phasing-​down of tariffs resulted in achieving the 85 per cent duty-​free threshold needed for the establishment of an FTA by 2008 (Tralac 2012). However, not all member states have yet implemented the FTA. The Democratic Republic of the Congo, Comoros, and Angola are yet to do so. The ambitions for a common market by 2015, a monetary union by 2016, and a single currency by 2018 did not materialize, and few countries met the other targets, like the macroeconomic targets on inflation of less than 3 per cent by 2017/​18 (UNECA et al. 2019). But several targets were not met. As Tralac (2013) argued, the RISDP was premised upon a linear integration model, which required member states sharing sovereignty, strong common legal institutions overseeing integration, and strong political leadership. With widely differing economic structures and levels of development, SADC was not ready for this model. The transition from the FTA to a customs union requires agreement on common external tariffs, amongst other things, and the divergences between economies make this difficult. The focus in 2010 shifted back to creating a better-​functioning FTA (Tralac 2013). Some progress has been made in reducing NTBs as part of this, but there is still significant work needed. Key NTBs identified for intervention included customs clearance processes, trade document requirements, rules of origin interpretations, standards harmonization, and other costs of clearance of goods in transit. Competitiveness in RVCs hinges on reduced transaction costs and smooth cross-​ border linkages (Tralac 2013). NTBs, like cumbersome border controls, regulations, and corruption, make regional trade difficult and expensive. Customs and border delays significantly raise trade costs. Estimates are that a one-​day delay in inland transit times can reduce export values by about 7 per cent, costing around US$400 per day for a stationary truck at border posts. The delays have more damaging effects on perishable food and time-​sensitive products (Vilakazi and Paelo 2017). Although not an SADC initiative, the AU’s Programme for Infrastructure Development in Africa and the associated Priority Action Plan aim to reduce these costs. One Stop Border Posts (OSBPs) were key projects to enhance interconnectivity of markets and regional integration. OSBPs are frameworks, facilities, and procedures to enable goods, people, and vehicles to stop in a single facility to legally transit borders. In SADC, the Chirundu OSBP between Zambia and Zimbabwe is the first fully functional OSBP in Africa (One Stop Border Posts Sourcebook 2016). SADC also has the Regional Infrastructure Development Master Plan to guide the implementation of coordinated, integrated, efficient, trans-​ boundary infrastructure networks in energy, transport, tourism, ICT, meteorology, and water (Ngwawi 2018). Even with all these initiatives, there are significant challenges with regards to infrastructure in the region.

412   Reena das Nair Another NTB is anticompetitive conduct. As noted, despite there being regional competition institutions, cartels which undermine RVCs and integration have not been investigated regionally. In the agricultural sector, for example, fertilizer cartels have been prosecuted individually by national competition authorities in South Africa and Zambia (Vilakazi and Roberts 2019). Even though the same players implicated are present in, or export to, other SADC countries, whether cartel conduct had impacted other countries has not been investigated. Cartel rents provide incentives for incumbent firms to protect their markets from new entry, undermining RVCs and trade. Other areas, such as rules of origin, are highly contentious, with concerns that they are too complex and onerous. The initial target of 2008 for the review of rules of origin was not met (Tralac 2012). Restrictive rules-​of-​origin requirements in agro-​processing, and in clothing and textiles (around double transformation requirements) still need to be addressed (Andreoni and Boys 2020). Initiatives are underway in different sectors to simplify these and this is also an area covered in Phase 1 of the AfCFTA (Hartzenberg n.d.). Some NTBs are a direct result of allowed exemptions. In SACU, for instance, Article 26 of the 2002 agreement provides for special and differential treatment for the protection of infant industries in all member states, except South Africa (UNCTAD 2005). While this allows space for members to pursue their own development strategies, it also restricts the emergence of RVCs, particularly in the food sector (Edwards and Lawrence 2008). The harmonization of standards and SPS measures have seen some progress through the setting up of regional structures. The SADC Accreditation Service was set up in 2009 in Botswana, with testing laboratories in other countries. Other factors affecting integration, such as the protocol on finance and investment, were in force by 2010, but lack of technical support and capacity has slowed implementation (Tralac 2012). Inter-​regional initiatives like the Tripartite FTA (COMESA-​EAC-​SADC) declaration signed in 2011, have been riddled with delays and complexities. Comprising twenty-​six countries, the agreement aimed to deepen integration through the harmonization of policies and programmes across the three RECs on trade, customs, and infrastructure development.8 But with different trade arrangements and rules about co-​existence of multiple existing agreements, the process has been messy (Tralac 2013). By February 2020, only eight countries had ratified, with a remaining six needed to meet the threshold.9 Closely related to the development of RVCs and regional integration is the SADC Industrialization Strategy and Roadmap (2015–​63) adopted in 2015 to promote the 8  https://​w ww.sadc.int/​about-​s adc/​continental-​interregional-​integration/​tripartite-​cooperation/​, accessed 5 December 2020. 9  https://​www.comesa.int/​implementation-​of-​the-​tripartite-​free-​trade-​area-​now-​in-​sight/​, accessed 5 December 2020.

Southern African Regional Value Chains and Integration    413 ‘structural transformation of the SADC region by way of industrialization, modernization, upgrading and closer regional integration’ (SADC Industrialization Strategy and Roadmap 2015:  1). It is anchored on three pillars:  economic transformation, enhancing competitiveness, and deeper regional integration, and focuses on three priority growth paths for accelerated industrialization: agriculture-​led growth (including agro-​ processing), natural resource-​ led growth (minerals beneficiation and processing), and value-​chain linkages. The Roadmap recognizes some of the challenges in agro-​processing discussed in section 19.3, such as the role of supermarket chains and the difficulties faced by SME suppliers in upgrading and diversifying. It recognizes the importance of modernizing production systems to bring them up to international standards. The responsibility is with member states to implement the strategy. RVC development, despite this Roadmap, has been poor. As noted, substantial public-​and private-​sector coordination and resources are needed, and there is a large funding gap in the region. RVC development within the SACU countries is also weak. A greater emphasis in SACU has been on plugging into GVCs, rather than in developing RVCs, as highlighted in Chapter 18 in this volume. On trade, as noted, SADC has amongst the highest intra-​regional exports of the RECs with 15 per cent on average over 1995–​2019, with intra-​regional imports at 19 per cent (UNCTAD trade data). But there is considerable heterogeneity in this trade and the trade pattern is highly skewed (Table 19.1). South Africa dominates exports to the region but hardly imports from it (Black et al. 2020). The share of exports from South Africa to SADC grew from 15 per cent in 2000 to 23 per cent of total exports in 2019, while the share of imports from SADC was just 1 per cent of total imports in 2000, growing to 6 per cent in 2019. However, given South Africa’s size, it is still an important trading partner for the other countries. As noted, uneven trade patterns have resulted in antagonism towards South African FDI and exports in the region, leading to protectionist measures. The skewed nature of trade in the region has also increased other costs. Large volumes of exports leaving South Africa and little reciprocal imports mean that trucks leaving South Africa full of goods return empty, given the unavailability of return loads. This results in inbound freight rates to South Africa being double the outbound rates, further disadvantaging producers in the region trying to sell into South Africa (Ncube, Roberts, and Vilakazi 2015). The composition of SADC trade is noteworthy. Intra-​SADC trade is characterized by more diversified manufactured goods than exports to the rest of the world (Black et  al. 2019). Greater exports of manufactured products can spur industrialization by providing opportunities to upgrade, further motivating for the development of RVCs in sectors with potential. For South Africa, SADC is a growing destination for diversified manufactured exports (Figure 19.1a), particularly for food and beverage products, and transport equipment (Figure 19.1b). Accompanying this trade has been growth in trade of services in the form of FDI by South African firms in retail, banking, insurance, transport, and business support services (Arndt and Roberts 2018).

414   Reena das Nair

19.5 Conclusion Changing global trends have led to greater attention being paid to how RVCs can be a vehicle for industrialization and growth. In southern Africa, the record of RVC development and regional integration has been poor, despite numerous protocols, strategies, and plans as part of SADC and SACU mandates. Using a GVC framework and taking into account regional dynamics, the case studies in agro-​processing and retail value chains highlighted the opportunities available to develop specific value chains, but also showed the serious challenges and bottlenecks that prevent development. Desired outcomes will not be achieved if the same approach to regional integration continues to be taken going forward. The AfCFTA presents an opportunity to approach regional development differently and pragmatically. While there are cross-​cutting issues to be fixed, such as border controls and weak infrastructure, value-​chain-specific interventions are necessary to develop RVCs. This requires an understanding of power dynamics and the role of large firms in specific value chains with potential. With advanced capabilities in several sectors, the spread of South African lead firms or their exports to the region has created a skewed playing field in which trade and investment have been one-​way. This has contributed to deindustrialization in other countries, creating tensions and triggering protectionist responses, further undermining the regional development vision. Developing RVCs requires a coordinated approach to investment, competition, and industrial growth in the region to systematically, and in a coordinated manner, build capabilities and reduce barriers. It requires long-​term compacts with lead firms in value chains through public–​private partnerships. Development finance institutions also play an important role in making funding affordable, accessible, agile, and patient. Dealing with entrenched power in value chains requires coordination of national and regional competition authorities. Phase 2 of the AfCFTA involves the development of protocols on competition policy. This has to include commitments to curb anticompetitive behaviour of large lead firms across borders. Regional competition bodies need to work closely with national authorities to investigate firms that have been implicated in anticompetitive conduct in one country, and which have activities in others. Cooperation in investigations and processes to deal with such conduct needs to be developed. Certain provisions, like the prohibition of the abuse of buyer power in South Africa, that are not part of competition legislations of other countries may need to be extended to the region. Developing RVCs requires identifying mutually beneficial outcomes. This will necessarily create winners and losers in specific value chains in the short term, but the premise is that, overall, the region as a whole is better off in the longer term. While South Africa exports certain value-​added products to the region, it has to commit to import products which other countries can produce more efficiently and sustainably. This has to filter down to the procurement practices of lead firms on the ground. ‘Local’ content

Southern African Regional Value Chains and Integration    415 policies or local sourcing requirements should take on a regional view so that these are in line with a regional development vision. As South African FDI grows in the region, there needs to be commitment to sourcing more from the region and contributing to the development of supplier capabilities. Complementary investments and commitments are needed from governments, finance institutions, international agencies, and donors to develop the public infrastructure, facilities, and skills needed to build capacity. In retail value chains, for instance, a voluntary ‘Regional Charter’ could be developed, building on the Namibian model that encompasses commitments by supermarkets on supplier development and a code of conduct when dealing with regional suppliers. Formalizing and strengthening such programmes can facilitate upgrading and contribute to levelling trade flows. Supporting alternative routes to market, like independent retailers, wholesalers, and buyer groups at a regional level will also provide suppliers with outside options, increasing their bargaining power with supermarkets. Technology can further be strongly leveraged to facilitate regional linkages. Expanding market access and payments can be facilitated through adoption of e-​ commerce and mobile money platforms. Blockchain technologies are also important for traceability requirements which is important for rules of origin. This requires an enabling regulatory framework at a regional level. Past regional plans and strategies have failed to develop strong RVCs because they remain high-​level plans with little direction on specifics and practical implementation. These plans are often too wide or too ambitious, aiming to include all member states in all interventions, with limited private-​sector involvement. In practice, development of a selected RVC may only require commitments from two or three countries. This chapter also showed that policymakers continue to make policies that only serve national interests. The political economy dynamics need to be carefully managed to ensure that powerful players do not lobby to protect their own interests, detracting from a true regional agenda. This will not be easy given competing interests and pressure on governments to protect local industries, but necessary if we are serious about regional development.

References Allen, Thomas. 2018. ‘Food prices must drop in Africa: How can this be achieved?’. OECD Development Matters [online] https://​oecd-​development-​matters.org/​2018/​02/​09/​food-​ prices-​must-​drop-​in-​africa-​how-​can-​this-​be-​achieved/​. Andreoni, Antonio, and Julian Boys. 2020. ‘Promoting industrialization in Africa:  Global versus regional value chains’. [blog] https://​www.globaldev.blog/​blog/​ promoting-​industrialization-​africa-​global-​versus-​regional-​value-​chains. Arndt, Channing, and Simon Roberts. 2018. ‘Key issues in regional growth and integration in Southern Africa’, Development Southern Africa, 35(3): 297–​314. Black, Anthony, Lawrence Edwards, Ruth Gorven, and Willard Mapulanga. 2020. ‘Agro-​ processing, value chains, and regional integration in Southern Africa’. WIDER Working Paper no. 2020/​36, Helsinki.

416   Reena das Nair Black, Anthony, Lawrence Edwards, Faizel Ismail, Brian Makundi, and Mike Morris. 2019. ‘Spreading the gains? Prospects and policies for the development of regional value chains in southern Africa’. SA-​TIED Working Paper no. 38, UNU WIDER, Helsinki. Bosiu, Teboho, Reena das Nair, and Anthea Paelo. 2017. ‘The global food value chain and competition law and policy in BRICS countries: Insights from selected value chains in South Africa’. CCRED Working Paper no. 2017/​21, Johannesburg. Coe, Neil, and Henry Wai-​chung Yeung. 2015. ‘Toward a dynamic theory of global production networks’, Economic Geography, 91(1): 29–​58. Competition Commission of South Africa. 2019. ‘Grocery Retail Market Inquiry Report’. Competition Commission of South Africa, Pretoria. Dachs, Bernhard, and Stefan Pahl. 2019. ‘Are global value chains in decline?’. [blog] Global Investments and Local Development, The London School of Economics and Political Science. https://​blogs.lse.ac.uk/​g ild/​2019/​07/​26/​are-​g lobal-​value-​chains-​in-​decline/​#:~:text= New%20evidence%20suggests%20that%20the,since%20the%20global%20financial%20 crisis. Dallas, Mark, Stefano Ponte, and Timothy Sturgeon. 2019. ‘Power in global value chains’, Review of International Political Economy, 26(4): 666–​94. das Nair, Reena. 2019. ‘The spread and internationalisation of South African retail chains and the implications of market power’, The International Review of Applied Economics, 33(1): 30–​50. das Nair, Reena, and Namhla Landani. 2020. ‘New approaches to supermarket supplier development programmes in Southern Africa’, Development Southern Africa, 38(1): 4–​20. das Nair, Reena, Shingie Chisoro, and Francis Ziba. 2018. ‘Supermarkets’ procurement strategies and implications for local suppliers in South Africa, Botswana, Zambia and Zimbabwe’, Development Southern Africa, 35(3): 334–​50. das Nair, Reena, Maria Nkhonjera, and Francis Ziba. 2017. ‘Growth and development in the sugar to confectionery value chain’. CCRED Working Paper no. 2017/​16, Johannesburg. Edwards, Lawrence, and Robert Lawrence. 2008. ‘SACU tariff policies: Where should they go from here?’. CID Faculty Working Paper no. 169, Massachusetts. Gereffi, Gary, and Katerina Fernandez-​Stark. 2016. Global Value Chain Analysis:  A Primer. Durham, NC: Duke University. Gereffi, Gary, and Joonkoo Lee. 2014. ‘Economic and social upgrading in global value chains and industrial clusters: Why governance matters’, Journal of Business Ethics, 133: 25–​38. Gereffi, Gary, John Humphrey, and Timothy Sturgeon. 2005. ‘The governance of global value chains’, Review of International Political Economy, 12: 78–​104. Goga, Sumayya, and Teboho Bosiu. 2019. ‘Governance of poultry value chains—​a comparative perspective on developing capabilities in South Africa and Brazil’. CCRED Working Paper no. 2019/​10, Johannesburg. Gore, Innocent. 2020. ‘SADC approves Vision 2050, RISDP 2020–​30’. SANF 20 no. 34, The Southern African Research and Documentation Centre. Hartzenberg, Trudi. n.d. ‘The African Continental Free Trade Area Agreement—​what is expected of LDCs in terms of trade liberalisation?’. [online] https://​www.un.org/​ldcportal/​ afcfta-​what-​is-​expected-​of-​ldcs-​in-​terms-​of-​trade-​liberalisation-​by-​trudi-​hartzenberg/​. Horner, Rory, and Khalid Nadvi. 2018. ‘Global value chains and the rise of the Global South: Unpacking twenty-​first century polycentric trade’, Global Networks, 18(2): 207–​37. Humphrey, John, and Hubert Schmitz. 2000. ‘Governance and upgrading: Linking industrial cluster and global value chain research’. IDS Working Paper no. 120, University of Sussex.

Southern African Regional Value Chains and Integration    417 Humphrey, John, and Hubert Schmitz. 2002. ‘How does insertion in global value chains affect upgrading in industrial clusters?’, Regional Studies, 36(9): 1017–​27. ITAC (International Trade Administration Commission of South Africa). 2015. ‘Investigation into the alleged dumping of frozen bone-​in portions of fowls of the species Gallus domesticus: Final Determination’. [online] http://​www.itac.org.za/​upload/​document_​files/​ 20150306125607_​Report-​No-​492.pdf. Kaira, Thula. 2015. ‘A cartel in South Africa is a cartel in a neighbouring country: Why has the successful cartel leniency policy in South Africa not resulted into automatic cartel confessions in economically interdependent neighbouring countries?’. Annual Competition and Economic Regulation Week 2015, Victoria Falls, Zimbabwe. Kaplinsky, Raphael, and Masuma Farooki. 2011. ‘What are the implications for global value chains when the market shifts from the north to the south?’, International Journal of Technological Learning, Innovation and Development, 4(1): 13–​38. Kaplinsky, Raphael, and Mike Morris. 2014. ‘Developing industrial clusters and supply chains to support diversification and sustainable development of exports in Africa’. Report for the AfrEximBank, Cape Town: University of Cape Town. Keane, Jodie. 2015. ‘Firms and value chains in Southern Africa’. World Bank Working Paper, London. Ncube, Phumzile. 2018. ‘The southern African poultry value chain:  Corporate strategies, investments and agro-​industrial policies’, Development Southern Africa, 35(3): 369–​87. Ncube, Phumzile, Simon Roberts, and Thando Vilakazi. 2015. ‘Study of competition in the road freight sector in the SADC region: Case study of fertilizer transport and trading in Zambia, Tanzania and Malawi’. CCRED Working Paper no. 3/​2015, Johannesburg. Ncube, Phumzile, Simon Roberts, and Tatenda Zengeni. 2016. ‘Development of the animal feed to poultry value chain across Botswana, South Africa, and Zimbabwe’. Working Paper no. 2016/​2, UNU-​WIDER, Helsinki. Newfarmer, Richard, John Page, and Finn Tarp. 2018. ‘Industries without smokestacks and structural transformation in Africa:  Overview’, in Richard Newfarmer, John Page, and Finn Tarp (eds) Industries without Smokestacks:  Industrialisation in Africa Reconsidered. Oxford: Oxford University Press. Ngwawi, Joseph. 2018. ‘Realigning infrastructure with industrialisation SADC assesses progress’. SANF 18 no. 16, The Southern African Research and Documentation Centre. One Stop Border Posts Sourcebook. 2016. 2nd edition, Tralac. [online] https://​www. tralac.org/​ n ews/​ article/​ 1 1453-​ one-​ s top-​ b order-​ p ost-​ o sbp-​ s ourcebook-​ 2 nd-​ e dition. html#:~:text=The%20OSBP%20concept%20refers%20to,and%20enter%20the%20 adjoining%20state. Paremoer, Tamara. 2018. ‘Regional value chains: Exploring linkages and opportunities in the agro-​processing sector across five SADC countries’. CCRED Working Paper no.  4/​2018, Johannesburg. Pick n Pay. 2020. ‘Integrated annual report’. [online] https://​www.picknpayinvestor.co.za/​ downloads/​annual-​report/​2020/​pick-​n-​pay-​iar.pdf. Pickles, John, Stephanie Barrientos, and Peter Knorringa. 2016. ‘New end markets, supermarket expansion and shifting social standards’, Environment and Planning A, 48: 1284–​301. Reardon, Thomas, Peter Timmer, and Julio Berdegué. 2004. ‘The rapid rise of supermarkets in developing countries: Induced organisational, institutional, and technological change in agrifood systems’, e-​Journal of Agricultural and Development Economics, 1(2): 168–​83.

418   Reena das Nair Roberts, Simon. 2019. ‘(Re)shaping markets for inclusive economic activity: Competition and industrial policies relating to food production in Southern Africa’, in Ravi Kanbur, Akbar Noman, and Joseph E. Stiglitz (eds) The Quality of Growth in Africa. New York: Columbia University Press. SADC. 2015. Industrialisation Strategy and Roadmap 2015–​2063. [online] https://​www.sadc.int/​ files/​2014/​6114/​9721/​Repriting_​Final_​Strategy_​for_​translation_​051015.pdf. Shoprite. 2001. ‘Integrated annual report’. [online] https://​www.shopriteholdings.co.za/​content/​dam/​MediaPortal/​documents/​shoprite-​holdings/​integrated-​report/​2001/​AR2001_​ Shoprite.pdf. Shoprite. 2020. ‘Integrated annual report’. [online] https://​www.shopriteholdings.co.za/​group. html. Songwe, Vera. 2019. ‘Intra-​African trade: A path to economic diversification and inclusion’. Boosting Trade and Investment: A New Agenda for Regional and International Engagement. Brookings Institution, Washington, DC. South African Poultry Sector Master Plan. 2019. Department of Trade, Industry and Competition (DTIC) and Department of Agriculture, Land Reform and Rural Development (DALLRD) [online] https://​www.dalrrd.gov.za/​docs/​media/​SA%20Poultry%20Sector%20 Master%20Plan%201.pdf. StatsSA (Statistics South Africa). 2017. ‘Manufacturing industry:  Financial, 2017’. Report no. 30-​02-​03 (2017), StatsSA, Pretoria. Traill, Bruce. 2006. ‘The rapid rise of supermarkets’, Development Policy Review, 24(2): 163–​74. Tralac. n.d. ‘SADC discusses Vision 2050, RISDP 2020-​30’. [online] https://​www.tralac.org/​ news/​article/​14852-​sadc-​discusses-​vision-​2050-​risdp-​2020-​30.html. Tralac. 2012. ‘The regional indicative strategic development plan: SADC’s trade-​led integration agenda. How is SADC doing?’. Trade Brief no. S12TB02/​2012, Stellenbosch. Tralac. 2013. ‘The review of the SADC RISDP matters’. Policy Brief, Stellenbosch. UNCTAD. 2005. ‘The Southern African custom union (SACU) regional cooperation framework on competition policy and unfair trade practices’. UNCTAD/​DITC/​CLP/​2005/​3, Geneva. UNCTAD. 2018. ‘Forging a path beyond borders: The Global South’. UNCTAD/​OSG/​2018/​1, Geneva. UNECA, African Union Commission and African Development Bank. 2019. ‘Assessing regional integration in Africa IX: Next steps for the African Continental Free Trade Area’. Addis Ababa. [online].https://​repository.uneca.org/​bitstream/​handle/​10855/​42218/​ b11963189.pdf?sequence=1&isAllowed=y. Vilakazi, Thando, and Anthea Paelo. 2017. ‘Understanding intra-​ regional transport:  Competition in road transportation between Malawi, Mozambique, South Africa, Zambia, and Zimbabwe’. UNU-​WIDER Working Paper no. 2017/​46, Helsinki. Vilakazi, Thando, and Simon Roberts. 2019. ‘Cartels as “fraud”? Insights from collusion in southern and East Africa in the fertiliser and cement industries’, Review of African Political Economy, 46(161): 369–​86. Ziba, Francis, and Mwanda Phiri. 2017. ‘The expansion of regional supermarket chains: Implications for local suppliers in Zambia’. UNU-​WIDER Working Paper no. 2017/​ 58, Zambia Institute for Policy Analysis and Research.

Chapter 20

Sou t h Africa’s E c onomi c Role in A fri c a Mills Soko and Mzukisi Qobo

20.1 Introduction The demise of apartheid paved the way for the re-​admission of South Africa into the global community after decades of international ostracism. Since the end of apartheid, South Africa’s role on the African continent has been far-​reaching and prominent. South Africa has spearheaded the creation and strengthening of institutions dealing with development, peace, and security—​notably the New Partnership for Africa’s Development (NEPAD), the African Union (AU)—​as well as initiatives to deepen trade integration. Private corporations and other non-​state actors have also independently ventured into the continent for reasons driven by their corporate strategies and mainly for market expansion reasons. This chapter discusses South Africa’s role in Africa, focusing on political and economic relations in the post-​apartheid period. It argues that South Africa’s expansion into Africa has mostly been a product of a series of deliberate internal and foreign policy decisions implemented by consecutive post-​apartheid governments. By prioritizing the ‘African Agenda’, South Africa has sought to harness its representation at the helm of continental institutions into meaningful foreign policy outcomes, including economic diplomacy objectives. South Africa has largely succeeded in fulfilling the goals outlined in its foreign policy and African Agenda. Nevertheless its actions, as the discussion on the implementation of the Protocol on Trade’s rules of origin in the SADC region shows, have also undermined the economic interests of neighbouring countries. The degree to which South Africa will be able to sustain its foreign policy performance in Africa in the future will depend on how far it weighs its domestic policy pressures with its continental obligations, on how it responds to and accommodates the economic concerns of its regional neighbours, and on how it manages African perceptions of the country. In this chapter, we are more interested in the nexus between politics and trade. We use a political economy lens to critique South Africa’s regional economic strategy

420    Mills Soko and Mzukisi Qobo with particular reference to its role in the Southern African Development Community (SADC), the Southern African Customs Union (SACU), and the broader African continent. The chapter has two sections. The first discusses South Africa’s political and diplomatic relations in Africa, focusing on the country’s African Agenda, its role in the evolution and consolidation of NEPAD, its relationship with the AU, and its contribution as a provider of aid to Africa. The second section explores the country’s economic relations on the continent with specific reference to the trade integration process in the SADC (focusing specifically on the impact of the SADC Protocol on Trade), trade and investment trends in the region, the changing nature of SACU, as well as South Africa’s prospects in the African Continental Free Trade Area (AfCFTA).

20.2  Political and Diplomatic Relations 20.2.1 South Africa’s African Agenda Several forces have shaped South Africa’s relations with the African continent and its adoption of the African Agenda in the post-​1994 period. Africa’s economic underdevelopment has been uppermost in South Africa’s strategic thinking about the region (see South African Government 1996; Dirco 2011: 20–​3, 2019: 8–​9). The continent’s underdevelopment has been exacerbated by many years of political instability and economic mismanagement. South Africa’s relative economic dominance on the African continent, the history of the ruling party, the African National Congress (ANC), as a liberation movement whose anti-​apartheid struggle was actively supported by several African states, as well as the changing global context in which the country has operated, have all played an essential role in shaping South Africa’s regional engagements (Gelb 2001: 3). South Africa’s vision of the African Agenda is informed primarily by its desire to shape the African continent’s development agenda. The African Agenda is about ‘charting a new strategic path in order to effect a turn-​around in the continent’s economy, politics, governance and development orientation’ (Landsberg and Kondlo 2007: 1). Following the end of apartheid, the newly installed ANC government set out a foreign policy that underscored the country’s democratic credentials, African identity, and global ambitions (Habib 2009). In sum, South Africa centred its foreign policy on the African continent. In a seminal article that Mandela wrote for the Foreign Affairs journal, he stated: ‘South Africa cannot escape its African destiny. If we do not devote our energies to this continent, we too could fall victim to the forces that brought ruin to its various parts’ (Mandela 1993: 89). In this statement rests the normative grounding of the ANC government’s policy posture towards Africa. It is an approach that was ‘premised on the

South Africa’s Economic Role in Africa    421 belief that South Africa’s fortunes and future were indissolubly bound with those of the African continent’ (Sidiropoulos 2008). It represented an unambiguous assertion that South Africa viewed itself as an African country, not as part of the ‘Western bloc’, as was the case during the apartheid era (see Nolutshungu 1994: 129–​36). Under the Thabo Mbeki presidency, South Africa played an essential part in recasting the security dialogue in Africa. The desire to infuse foreign policy with an ethical grounding extended to the region. This activism on the part of South Africa was apparent, for instance, in the SADC Protocol on Politics, Defence, and Security Cooperation. The protocol spelt out a security agenda that straddled politico-​military threats as well as non-​military threats, including inter-​state war; human rights abuses; war crimes; crimes against humanity; humanitarian and natural disasters; and genocide (Hammerstad 2005: 4). The African Agenda’s core goals include strengthening the AU and its institutions, promoting regional integration within the SADC, advancing peace and security in Africa, and contributing towards post-​conflict rebuilding and development (Landsberg 2009: 1–​2). South Africa was central to the founding of the AU and in negotiating NEPAD. In essence, the African Agenda is about South Africa’s role in the AU, NEPAD, and the SADC. This is detailed in three key pillars of engagement with Africa, outlined in the South African government’s Strategic Plan for 2006–​09. As Maloka (2019:  48) pointed out, these pillars are:  strengthening Africa’s multilateral institutions, especially the AU and SADC; supporting the implementation of NEPAD, Africa’s socio-​economic development programme; and bolstering bilateral political and socio-​economic relations through effective structures for dialogue and co-​ operation. The African Agenda recognizes that promoting stability and sound governance across Africa is crucial for overall success. Before initiating the African Agenda, South Africa’s policy had been conceptualized through the lens of the outcomes of the Conference on Security, Stability, Development, and Cooperation in Africa. Through the African Agenda, South Africa has sought to position itself as both a middle power and champion of multilateralism in the global system. Following the end of apartheid, the South African government set out to craft a foreign policy anchored on strong ethical foundations. To dispense with the legacy of apartheid and improve its international image, the government integrated the country into the global order, and it became an active champion of multilateralism (Habib 2009). South Africa’s unwavering support for multilateralism stemmed partly from a ‘desire to play a role as a bridge-​builder between the advanced industrial countries and developing countries, and to be seen as an honest broker in international affairs’ (Qobo 2012: 4). The Mbeki administration, in particular, evinced an assertive multilateralist approach. Its focus was on collaborating with other countries to craft collective solutions to global and regional problems. The pursuit by the ANC government of national goals multilaterally has enabled the country to capitalize on its moral and political authority derived from its nascent democratic order, while also championing Africa’s integration into the global system on terms that benefit the continent. Thus, foreign policy became more aimed at boosting South Africa’s international status and using multilateral bodies to uphold human rights and democratic global governance (Qobo 2012: 4), In this context, the

422    Mills Soko and Mzukisi Qobo lingering apartheid-​era policy of destabilization in the Southern African region gave way to dialogue and mediation. The new policy, which the Mbeki government sought to export to the rest of Africa, focused on developing political solutions to conflicts and championing initiatives designed to reduce regional insecurity (Mda 2004: 138).

20.2.2 NEPAD NEPAD represents the policy and institutional expression of the African Agenda. Adopted in 2001 and ratified in 2002, NEPAD was established to ‘promote accelerated growth and sustainable development, eradicate widespread and severe poverty, and halt the marginalisation of Africa in the globalisation process’ (NEPAD Secretariat 2001). NEPAD’s four key objectives are to eliminate poverty, promote sustainable growth and development, foster the African continent’s integration into the world economy, and accelerate the empowerment of women. To this end, the programme has focused on six strategic themes: namely agriculture and food security; climate change and natural resource management; regional integration and infrastructure; human development; economic and corporate governance; as well as cross-​cutting issues including gender, information communication technologies, and capacity building (NEPAD Secretariat 2001). Programmatically, NEPAD has concentrated on, among other things, building democracy; promoting regional cooperation and integration; building capacity and human development with a focus on health, education, science, and technology as well as skills development; increasing domestic savings and investments; growing intra-​ African trade and broadening access to markets of industrialized nations; generating foreign direct investment; stimulating diversification of production and exports, especially in relation to agro-​industries, manufacturing, mining, mineral beneficiation, and tourism; and building and upgrading infrastructure, including information and communication technology, energy, transport, and water and sanitation (NEPAD Secretariat 2001). Yet NEPAD has been castigated by African scholars and civil society groups. There are two major concerns that critics have raised about this programme. The first is that NEPAD is a top-​down-approach, elite-​driven initiative. According to Landsberg (2007:  205), it failed to sufficiently acknowledge the ‘experiences, knowledge and demands of African people’. The second criticism has underlined the programme’s fixation with external engagements, especially with the G8 countries. In its early years of development, NEPAD exhibited over ​dependence on external financial support for its implementation. This is in contrast to the earlier plans such as the Lagos Plan of Action (1980), the African Alternative Framework for Structural Adjustment Programmes (1989), and the Arusha Charter (1990) all of which called for collective self-​reliance and popular participation. NEPAD has recorded some notable successes but also spawned failures. The United Nations on Conference on Trade and Development (2012: 3–​4) has highlighted the resurgence of Africa’s growth in the first nine years of the promulgation of NEPAD, during

South Africa’s Economic Role in Africa    423 which the continent registered an impressive average increase of 5 per cent compared to the period between 1990 and 1999 that saw a lacklustre performance at a 2.7 per cent average growth rate. There was also growth in real output per capita from 0 per cent to 2.6 per cent over this period. The agricultural sector recorded progress. African nations embraced the Comprehensive Africa Agriculture Development Programme, which helped the continent pay close attention to agriculture and stimulate higher agricultural productivity and output (United Nations 2011a). NEPAD has also succeeded in facilitating the implementation of infrastructure projects on the African continent. For instance, the Programme for Infrastructure Development in Africa, unveiled at the AU’s 2010 summit in Kampala, outlined a comprehensive strategy for developing regional and continental infrastructure in Africa (Khati 2006). Furthermore, NEPAD helped to project Africa to the global stage and garnered international support for the continent. As a result of NEPAD, the Group of Eight (G8) countries launched its Africa Action Plan in June 2002. More concretely, external development funding through Official Development Assistance increased from $21.4 billion in 2002 to $47.9 billion in 2010, with bilateral support from the Development Assistance Committee of the Organisation for Economic Co-​operation and Development growing from $13.4 billion to $29.3 billion during the same period (UNCTAD 2012: 3). The NEPAD plan was pivotal in shaping the dialogue platform between the G8 and African leaders. As a result, a select group of African leaders became regular attendees of the G8 summits. They would later be invited to also attend the G20 meetings. These engagements between African leaders and their Western counterparts elevated Africa on the global stage and ensured that the continent’s concerns were on the global agenda. Another positive contribution NEPAD has made is in improving economic and political governance in Africa, as illustrated by the introduction of economic reforms in tax administration, access to credit, enforcement of contracts, as well as electoral issues and processes (United Nations 2011b). The United Nations (UN) also turned its attention on Africa and put in measures that improved coordination and coherence of development support for Africa. Further, Africa earned UN endorsement for its NEPAD programme, and this was done through the UN General Assembly resolution 57/​7 of 4 November 2002. This vote of confidence was not an insignificant development: it lent greater recognition of NEPAD as a socio-​economic programme developed by Africans for Africans, and a mechanism for development engagement between Africa and the rest of the developed world. Notwithstanding the progress made to date in executing the programmes and activities of NEPAD, the African continent still has a long way to go before it can accomplish its important goals of eradicating poverty and reversing its marginal status in the world economy. This weakness can be attributed to several factors that have hobbled the continent’s progress, including capacity constraints; low levels of human and financial resources; poor participation by local actors in the process; as well as the paucity of measurable yardsticks for evaluating progress (Economic Commission for Africa 2007).

424    Mills Soko and Mzukisi Qobo

20.2.3 African Union Prior to 1994, South Africa’s engagements with the Organisation of African Unity (OAU)—​the forerunner to the AU—​were carried out through its two key liberation movements: the ANC and the Pan Africanist Congress. The OAU was instrumental in lending international legitimacy to the anti-​apartheid struggle, but also in leading international efforts towards bringing about a peaceful negotiated settlement in South Africa (Maloka 2019: 41). A key tenet of the country’s relations with the outside world has been the belief that the country’s external policy should mirror the interests of the African continent. During Mbeki’s first term as president between 1999 and 2004, the ‘African Renaissance’ vision was distilled into the African Agenda. Mbeki’s notion of the African Renaissance, first propounded in his celebrated ‘I Am an African’ speech to the South African national legislature in May 1996, rested on the belief that Africans must address the problems and ills afflicting the continent in order to attain cultural, scientific, and economic revival (Mbeki 1996). South Africa was influential in the creation of the AU. The AU was set up in 2002 in the port city of Durban. As host of the AU’s inaugural summit, South Africa became the first African state to chair the continental body. The conversion of the OAU into the AU dovetailed with the founding of NEPAD and the African Peer Review Mechanism (APRM). NEPAD was Mbeki’s initiative that was born outside the OAU. Other African leaders later supported it, and its programme was adopted by the OAU in 2001. Although the African Renaissance constituted the conceptual basis of Mbeki’s African Agenda, NEPAD was the foremost implementation organ. South Africa has been committed to consolidating the continental body and its constituent sub-​structures. Since the founding of the AU, South Africa has been one of the top five budget contributors to the AU budget. Every year, South Africa contributes in the region of R125–​200 million to the AU budget, accounting for 15 per cent of the organization’s total budget (Besharati 2013). As the first chair of the AU, South Africa invested extensive diplomatic energy and financial resources into making the continental body effectively operational (Sidiropoulos 2008). For example, the cabinet cluster on International Cooperation, Trade, and Security established an African Renaissance committee to prioritize the AU, NEPAD, SADC, and SACU. From the outset, Mbeki worked hard to ensure that ‘the secretariats of NEPAD, the AU Parliament and the APRM were headquartered in South Africa’ (Maloka 2019: 49). The South African corporate sector was also drafted in to support the African Agenda. This occurred at two levels: first, through the NEPAD Business Group and, second, through South Africa’s state-​owned enterprises including the Development Bank of Southern Africa, Transnet, Telkom, and the Industrial Development Corporation (Maloka 2019: 51). South Africa’s involvement as a central actor in these initiatives provided a stage for the Mbeki administration to foster a progressive African agenda constructed around the ‘principles of integration, peace and security, democratic governance and economic growth’ (Landsberg 2007: 195).

South Africa’s Economic Role in Africa    425 South Africa’s relationship with the AU, and previously the OAU, since the dawn of democracy has not been without pitfalls. First, there were tensions between South Africa and other African countries over the future shape of NEPAD, and whether it was appropriate to integrate this into the AU. Some countries felt strongly that the NEPAD secretariat should be moved from South Africa to Addis Ababa. Second, Nigeria’s opposition to Dr Nkosazana Dlamini-​Zuma’s bid for appointment as chair of the AU Commission fractured the relations between the two countries. The decision by South Africa in 2012 to nominate her to lead the AU Commission was the result of the government’s new secondment policy to encourage South Africans to apply for posts in African regional institutions and the UN. Third, the cost of membership of the AU has been highly prohibitive for a developing country like South Africa, even though the most industrialized one in Africa. In 2009, the cost of membership stood at R200 million, dropping to R125 million in 2012. Additionally, AU operations have been constricted by the challenges of implementing the African Agenda, compounded by geo political dynamics and competing agendas of some states (Maloka 2019: 54–​5). In recent years, the relationship between the AU and South Africa has been strained by other issues. South Africa has been accused of acting in an inconsistent manner in relation to how it takes positions within the AU and the UN Security Council. In 2011, the country expressed opposition to intervention in Libya in the AU, while it voted in the Security Council for Resolution 1973 which approved international intercession. This resulted in the ouster of Muammar Gaddafi and the subsequent implosion of the Libyan state. As one commentator noted: ‘South Africa has two platforms for projecting power. One is the AU, and one is the UN, and at times these roles are contradictory’ (Schoeman, Kefale, and Alden 2017). The relationship has also been fractured by xenophobic violence perpetrated by South Africans against migrants and refugees from other African countries. Between 2000 and 2008, at least sixty-​seven people perished under a harrowing wave of violence that was attributed to xenophobia in South Africa. In 2015 , xenophobic violence, which started in the KwaZulu-​Natal province, erupted nationwide, resulting in at least seven deaths and displacing about five thousand migrants and refugees (Pambazuka News 2015). In September 2019, South Africa experienced another wave of attacks against African foreign nationals. Reacting to the attacks, several African countries, notably Nigeria, moved quickly to evacuate their citizens from the country. The attacks elicited a sharp rebuke from the AU, with the AU Commission chair Moussa Faki Mahamat denouncing them and expressing the AU’s willingness to work with the South African government to deal with the problem.

20.2.4 South Africa as a Provider of Aid to Africa Historically, South Africa has been a major supplier of aid to the African continent. During the apartheid era, the Department of Foreign Affairs used the Economic Cooperation Promotion Loan Fund (set up in 1968 and reviewed in 1986) to provide financial aid to a number of African countries—​including Gabon, Zaire, Cote d’Ivoire,

426    Mills Soko and Mzukisi Qobo Malawi, Equatorial Guinea, Comoros, Swaziland, and Lesotho—​as a strategy to galvanize political backing for South Africa’s positions in multilateral institutions such as the UN. The Fund was deployed as part of military tactics that underpinned South Africa’s foreign policy goals against the backdrop of the Cold War and decolonization in parts of the Southern African region. Moreover, the apartheid government utilized this Fund to counter international isolation that was the consequence of economic sanctions against South Africa (Besharati 2013: 17). South Africa’s aid policy has been motivated by a mixture of political and economic factors in the post-​apartheid era. Partly, it has been inspired by the internationalist agenda that has been pursued by the ANC based on the historical links it had with the world during the liberation years. It stemmed from the ANC’s commitment to repay African countries who supported the liberation struggle in South Africa. It has also been driven by a desire to contribute towards economic and political stability in Africa. Additionally, South Africa has sought to use foreign aid as a tool of ‘soft power’ to mobilize the support of African countries for its foreign policy ambitions and positions in multilateral bodies such as the AU and UN. Some observers and economic actors within South Africa have, in recent years, propounded the idea that South Africa should tie development assistance in the region to trade and investment agreements (Besharati 2013: 25). According to a study by Braude, Thandrayan, and Sidiropolous (2008), there are no precise figures for South African development aid. This is due to deficiencies in reporting, a consequence of there being no ‘systematic database that tracks South African development assistance’ (Braude, Thandrayan, and Sidiropolous 2008: 13). Even so, a report published by the National Treasury in 2006 showed that total transfers by the National Treasury and other government agencies to African nations grew by nearly 26 per cent from 2002 to 2004 (R9.5 billion to R15.2 billion). The African Renaissance Fund (ARF) emerged as the most coherent and well-​defined South African development assistance programme. Disbursements from the ARF grew from R50 million in 2003/​04 to R100 million in 2005/​06. The ARF was created in 2000 under Mbeki’s administration to replace the Economic Cooperation Promotion Loan Fund, which was birthed under apartheid. The ARF’s normative thrust was Mbeki’s African Renaissance doctrine, and was put together mainly as a mechanism to ‘establish partnerships, demonstrate solidarity and support the economic empowerment of Africa’ (Braude, Thandrayan, and Sidiropolous 2008: 19). Broadly, the ARF was designed to finance activities in the fields of democracy, social and economic development, conflict resolution, technical cooperation, capacity building, as well as humanitarian and disaster relief. Some of its recipients include farm projects in Zimbabwe, cultural programmes in Mali, humanitarian aid in Somalia, water dams in Lesotho, and support for the African Cup of Nations football tournament. The Fund has also been utilized to support peace initiatives in Burundi, the DRC, and Comoros, and to bolster public administrative capabilities in Sudan. In some cases, it was also used to repay the debts of Gabon, Comoros, Mozambique, Lesotho, the Central African Republic, and Malawi (Braude, Thandrayan, and Sidiropolous 2008: 19).

South Africa’s Economic Role in Africa    427

20.3 Economic Relations 20.3.1 South Africa and its Neighbours: Tensions in Regional Trade Relations The Southern African region has always been the centrepiece of South Africa’s external engagements. The ANC’s Reconstruction and Development Programme (1994: 119) articulated the governing party’s approach to the region. It contended that South Africa’s economic progress was bound up with the fortunes of its region. South Africa is, without a doubt, a dominant player in the region. It was thus unsurprising that the country’s entry into SADC was cheered as a spur for regional prosperity. This entry coincided with important shifts in the global system. The end of the Cold War and, subsequently, the demise of apartheid redefined the regional political landscape. South Africa was drawn into regional arrangements as a key partner, given its large economy and institutional strengths. The establishment of SADC in 1992 happened under these changing circumstances. As it evolved, SADC signalled a shift from old, inward-​looking forms of integration. The new framework of integration in SADC sought to embrace an open, deeper, and outward-​oriented model of economic integration. SADC member states expected this new era of regional integration to equalize economic gains by narrowing trade imbalances and delivering benefits in the form of infrastructure development and cross-​border value chains. SADC countries did not view trade as an end in itself, but as a tool of economic progress, and South Africa was expected to carry poorer countries on its shoulders as the regional economic powerhouse (Alden and Soko 2005).

20.3.2 The SADC Protocol on Trade 20.3.2.1 The Evolution of the Trade Protocol The SADC Trade Protocol was finalized and signed in 1996 amid major political shifts in the region. As noted previously, it heralded new forms of open regionalism and reconstituted the terms of engagement between countries in the region. Trade and investment moved to the centre of these new relationships. The adoption of the protocol was also inspired by the ambition on the part of SADC countries to achieve economic growth and development and facilitate their insertion into the world economy on beneficial terms. The protocol was designed to liberalize intra-​regional trade in goods and services, with the eventual goal of establishing a free trade area within the SADC region through a process of phasing down tariffs asymmetrically. In September 2000, four years after the initial signing, the protocol came into force. Since its ratification, the implementation of the trade protocol has moved at a glacial pace.

428    Mills Soko and Mzukisi Qobo The smaller SADC states could not carry out deeper trade liberalization on their own due to lack of structural complementarities; these countries produced similar products and have historically competed in the same EU markets. The process of liberalization was thus bound to be tortuous. South Africa’s lethargic leadership was unhelpful, especially in the early phases of the SADC Protocol on Trade. South Africa’s attention was also divided during this time as its resources were devoted to launching the EU-​ South Africa Trade and Development Cooperation Agreement. It took a while to conclude the SADC Protocol on Trade, and it was only implemented from 2000, six years after it was initially signed. The implementation of the agreement was bogged down by many obstacles that amplified the complexity of yoking countries of varying economic weights. At the beginning of 2004, SADC countries adopted a Regional Indicative Strategic Development Programme (RISDP) to complement the SADC Protocol on Trade. This programme set unrealistic targets for achieving a SADC customs union by 2010, a missed deadline. The SADC Secretariat noted the attainment of a Free Trade Agreement (FTA) in 2008 ‘as a step towards achieving a Customs Union and subsequently a Common Market’ (SADC Secretariat 2004). This objective, the RISDP stated, would be achieved ‘on the basis of a greater commitment to the implementation of the protocol on trade, appropriately designed rules of origin, and greater harmonization of customs rules and procedures, including standards’ (SADC Secretariat 2004). The RISDP outlines ambitious targets for realizing deeper integration: conclusion of talks for the SADC customs union in 2010; conclusion of negotiations for the SADC common market in 2015; diversification of industrial exports in 2015; and setting up of an SADC monetary union in 2016 (SADC Secretariat 2004). These unrealistic targets were set despite the known challenges with the slow pace of implementing the trade protocol. The lethargy in enacting the protocol is attributable to a number of factors. First, although SADC countries were aware of the challenges of implementing the protocol, they did not have a credible plan to meet its ambitious targets. Second, the RISDP did not consider problems on the ground, such as the preponderance of overlapping and multiple integration schemes. Third, this regional strategy overlooked tensions that could be induced by the impact of the new SACU Agreement on the SADC region. It also did not take into full account the very complex negotiations between SADC member states and the European Union under the aegis of the Economic Partnership Agreements (EPAs). The SADC Protocol on Trade has undergone various amendments to address the aforesaid challenges. Some of these amendments have been on technical issues related to non-​tariff barriers, setting parameters for trade in sugar in the region, limitations on second-​hand goods, and defining the dispute settlement mechanism. Even though the protocol established the principle of asymmetry, designed to accommodate less-​ developed member states’ special concerns regarding the pace of liberalization, restrictive rules of origin prevented these countries from taking full advantage of South Africa’s market. According to the asymmetry principle, South Africa and SACU countries were expected to set the pace on tariff liberalization and offer more benefits to the rest of the SADC countries. It was also expected that within three years of the

South Africa’s Economic Role in Africa    429 entering into force of the protocol, SACU countries would provide duty-​free access to over two-​thirds of total tariff lines. Most SADC countries depend on revenues generated through tariffs at the border for their fiscal stability.

20.3.2.2 The Effects of the Rules of Origin Rules of origin determine whether a product can be regarded as emanating from outside or from within the FTA. Put simply, it determines the nationality of a product, and assigns it a level of benefit within the integrating area. Rules of origin delineate criteria for determining products that are eligible for duty-​free market access or specified tariff levels. The absence of rules of origin creates a strong incentive for trade deflection. The requirement for value-​addition or ‘substantial transformation’ of a product in a member country tends to be very strong in determining the rules of origin (Krueger 1993: 5–​6). In the beginning, the SADC rules of origin were designed to be flexible, requiring only 35 per cent of local content in the traded product. However, these were later changed as a result of pressure from South Africa and other SACU member countries. These countries argued for complex and much tighter rules to protect import-​competing sectors in auto, textiles and clothing, and sugar. South Africa advanced two reasons: one related to the need to curtail trans-​shipment, where third countries use a country with the lowest tariff rate to re-​export their product to the high-​tariff region; the other argument was that more stringent rules of origin would foster industrial development in the region. According to a World Bank report in 2005: ‘Specifying rules of origin on a product by product basis offers opportunities for sectoral interests to influence the specification of the rules in a protectionist way’ (World Bank 2005: 70). And these were the effects of the kind of rules of origin preferred by South Africa. These rules created variation across product lines. For instance, in some products, what is needed is a simple change of tariff heading, while in other products a change of tariff chapter is required. Still, in others, there was a detailed specification of requirements for value-​addition for those products that would qualify for preferential treatment. This variation generated confusion for customs officials in the region who had limited technical capacity. Given the prickly political dynamics in the region, the prevailing trade arrangement in SADC could not promote harmonious relations. On the contrary, it compounded the existing tensions. Some of the damaging effects of rules of origin highlighted by Flatters (2002: 34) include the following: • Restrictive rules deprive producers of access to raw materials or intermediate products from low-​cost international sources. This has a cost-​raising effect on the consumers as high-​cost regional producers protected by a tariff would be preferred. In this regard, consumers are deprived of access to cheaper sources of inputs and producers are shielded from competition. • Downstream industries bear the burden of high costs of input, which certainly goes against exactly what rules of origin are intended for: encouraging industrial development. In essence, they have a limiting effect on industry competitiveness.

430    Mills Soko and Mzukisi Qobo In their current form, SADC rules of origin reveal the depth of protectionist interests in South Africa. The General Agreement on Tariffs and Trade’s provisions on the rules of origin set out broad principles on how these are to be applied. They state that rules of origin should facilitate, and should not create unnecessary barriers, to the flow of international trade. They should be implemented in an impartial, transparent, and consistent fashion. This contradicts what has happened in SADC, which has very complex and restrictive rules of origin similar to those applied in preferential agreements with highly industrialized countries (World Trade Organization 1986: 211). Besides these complex rules of origin, other intricate problems exist in the region, and their roots lie in moribund institutions and dysfunctional economies. These problems include a lack of political will to implement the protocol and overlapping memberships and competition between various regional integration schemes. Other countries in the region had a legitimate expectation that South Africa would offer more concessions in giving them greater access to its market, and that it would take a giant leap in liberalizing its tariffs. However, this is not what happened. During the negotiations on the SADC Trade Protocol, South Africa insisted on stringent rules of origin. It adopted an inflexible stance that inflamed the political sensitivities of its regional counterparts, potentially eroding the reservoir of political influence the country could hope to exercise in the future. South Africa has, for example, evinced a strong commitment to protecting import-​competing interests at all cost. There have also been issues related to standards. South Africa insisted that other SADC countries should meet the technical standards set by the South African Bureau of Standards, even though there was a lack of requisite infrastructure in most of these countries to comply with some of the standards. This insistence created significant non-​tariff trade barriers. Other countries also presented problems for the implementation of the SADC Protocol on Trade. In the middle of the implementation of the Protocol, Zambia imposed restrictions on Zimbabwean milk. It insisted that milk originating from Zimbabwe to Zambia should be labelled in Zambia’s local languages, something that has a cost-​raising effect on imports. South Africa was also at one point slapped with protectionist measures that went against the grain of the SADC Protocol on Trade. Early in the negotiations, Botswana imposed restrictions on South African brown bread and milk imports, ostensibly for infant industry protection. Wholesalers in Namibia prevailed on government to enforce a ban on imports of horticultural products from South Africa to protect the domestic industry. Nonetheless, South Africa’s overwhelming dominance in regional trade has presented difficulties in terms of the distribution of costs and benefits, while it has also posed a threat to the economies of the other SADC countries. This has encouraged a slide towards protectionism and anti-​liberalization in the region. McCarthy, for instance, has averred that ‘South Africa’s competitive position and trade situation within the region is characterized by a mercantilist pattern of one-​way trade which cannot form the basis for region-​wide economic growth and development’ (McCarthy 1996). Compounding this mercantilist trade pattern has been the creation of the SADC FTA in an industrially

South Africa’s Economic Role in Africa    431 polarized region with the potential to produce a skewed pattern of distribution of costs and benefits. As de la Rocha (2003: 10) cautioned: ‘Asymmetric distribution can also be caused by agglomeration effects. Industry and investments tend to concentrate on the more developed member in the regional block once the trade barriers are removed.’ Other likely negative outcomes of the SADC FTA observed by commentators include diversion of trade to South Africa’s large manufacturing base, redistribution of revenues from poor countries in the direction of South Africa, and further deepening of the ‘core–periphery’ dichotomy within the region (Holden 2003: 164). Such effects are not politically sustainable in southern Africa. Apart from South Africa’s mercantile trade policies and complex rules of origin, there have been other impediments to trade in the SADC region. First, there have been difficulties associated with trade facilitation and non-​tariff measures. These have hindered the cross-​border flow of goods. Although some measures have been put in place to ease trade flows—​including pre-​border clearance and other transit issues at the border, training of officials in risk management, harmonization of computer systems, development of single clearance documents, and standards and quality innovations—​ difficulties still abound, and progress has been slow. Second, owing to a lack of technical capacity, customs officials took longer to implement new tariff rates. Third, progress in most SADC countries has been stunted by uncompetitive industries, tiny markets, undiversified economies, inadequate infrastructure, poor regulatory frameworks, underdeveloped financial and capital markets, lack of modern technology, and unskilled personnel (Tekere 2012). Fourth, another problem afflicting SADC countries is their enormous trade dependence on traditional European markets. This dependence has hindered product diversification in these countries’ export baskets and undermined intra-​regional integration.

20.3.3 The Changing Nature of SACU 20.3.3.1 The Democratization of SACU The end of apartheid in South Africa created the conditions for the democratization of SACU. The history of SACU demonstrates that the southern African region has not always been treated respectfully by South Africa. Originally established in 1910, SACU—​made up of South Africa and the BLNS states (Botswana, Lesotho, Namibia, and Swaziland)—​is the oldest customs union in the world and the deepest form of integration in Africa. The SACU Agreement of 1910 provided for the governance of the BLNS states as sub-​regions of the South African economy. It created a common market anchored by a CET decided by South Africa, a common pool of currencies tied to the South African Rand, as well as integrated administrative structures and infrastructure. The agreement provided for the unrestricted exchange and flow of industrial goods and for the disbursement of customs union revenue through a formula that specified a fixed share for each of the countries (McCarthy 1998: 75).

432    Mills Soko and Mzukisi Qobo Nonetheless, South Africa enacted policies that undercut the economic advancement of the BLNS countries. The advent in South Africa in 1925 of an import substitution strategy, undergirded by high tariff walls, to promote industrial development had an adverse impact on the BLNS economies. It encouraged trade diversion in the region as the BLNS economies were compelled to purchase high-​cost South African goods. It also led to a decline in the general level of customs income as a share of GDP. Moreover, industrial growth behind high tariff barriers led to lopsided industrial development concentrated in South Africa (Gibb 1997: 75). The push for revising the SACU arrangement did not originate wholly from the BLNS countries. South Africa was also interested in transforming the existing arrangement. Following prolonged negotiations, the five SACU member countries concluded in October 2002 an extensive, revised SACU Agreement. The 2002 SACU Agreement—​ which came into effect in July 2004—​brought about significant changes. It created new democratic institutions for SACU, hailed as a victory for balanced regional integration. Moreover, the agreement stipulates that SACU countries ought to execute trade relations and negotiations with third parties as a single entity. It prescribes that no SACU member state is allowed to conduct negotiations, or enter into new preferential trade deals, with third parties, or change existing agreements without the consent of other member states (Soko 2008). Of great significance, the agreement substantially revised the revenue-​sharing formula among the SACU economies: the new formula is made up of a customs component, an excise component, and a development component. In this regard, customs revenues would be disbursed according to intra-​SACU imports, implying that South Africa would provide compensation to the other SACU member nations for the trade advantages—​the so-​called polarization effects—​it derived as a result of the customs union (Soko 2008).

20.3.3.2 SACU in the Post-​Brexit Era Until the United Kingdom resolved to leave the European Union, it conducted trade with South Africa under an EPA between the European Union and SADC countries (Tralac 2018). The EU-​SADC EPA entered into force on 10 October 2016. The United Kingdom’s departure from the European Union at the beginning of 2021, meant it would no longer benefit from trading arrangements concluded by the European Union with third parties, including the EU-​SADC EPA. Parallel to the existing EPA agreement, SACU countries plus Mozambique (SACUM) struck an agreement with the United Kingdom in 2019 to adopt the provisions of the EU-​SADC EPA into a new two-​way trade agreement, in order to give certainty to consumers and exporters. The roll ​over was calculated to ensure continuity in trade ties after the United Kingdom had left the European Union (Department of Trade and Industry 2020). Some of the provisions that the two parties adopted included ‘rules for trade in goods, preferential tariff rates on all sides, trade remedies, technical standards for health and safety for agricultural and industrial products, protection of South African and UK’s geographical indications, and dispute settlement’ (Department of Trade and Industry 2020). Other technical issues included ‘tariff-​rate quotas, the sourcing of inputs from

South Africa’s Economic Role in Africa    433 across the EU region into production for export; the treatment of bilateral safeguard measures; other transitional arrangements; geographical indication; and the built-​in agenda’ (Patel 2019). In light of the advanced nature of its economy, South Africa does not enjoy preferences for its fish, aluminium, and agricultural exports. A new trade clause was inserted that permitted SACUM countries and the United Kingdom to cumulate and utilize EU inputs for production to export to each other’s markets. This guaranteed the continuity of integrated value chains across South Africa, the European Union and the United Kingdom, especially in the car industry (Department of Trade and Industry 2020). The SACUM–UK agreement makes provision for transitional arrangements that facilitate the transfer to it of some trading terms from the EU-​SADC EPA. These include, among others, customs issues in respect of the protections of geographical indications. The United Kingdom, for its part, is required to provide ‘adequate time for exporters from SACUM nations to adjust to new technical rules for industrial goods in cases where those diverge from EU regulations’ (Soko 2020). A deal was also agreed on a built-​in agenda to deal with areas of interest in the future that could not be addressed during the negotiations, including ‘market access issues, regional cumulation, export taxes, technical barriers to trade, geographical indications, and electronic certification’ (Soko 2020).

20.3.4 Trade and Investment Trends South Africa’s trade with the African continent is characterized by the export of industrial goods and the import of mostly primary-​resource-​based goods (Fundira 2017). This is reflected in the composition of the country’s trade exports and imports on the continent (see Figures 20.1 and 20.2). Between 2010 and 2016, South Africa’s trade with

Mineral Products

19.21%

Machinery

16.08%

Products Iron Steel 10.21% Chemicals

Vehicles Aircraft Vessels

9.12%

Vegetables

Plastics Rubber

5.53% Toys Live Animals Sport

Apparel

9.87% Prepared Foodstuffs 9.52%

4.82%

1.99%

1.94%

Textiles

Wood Pulp Paper

Figure 20.1  Composition of South Africa’s exports in Africa, 2019 Source: SARS (2020).

434    Mills Soko and Mzukisi Qobo

Wood Pulp Paper

Machinery

Photographic Medical Equipment 5.87%

15.24% Precious Metal

24.35% Vehicles Aircraft Vessels

Works of Art 4.93%

11.18%

Mineral Products 22.13%

10.29%

Figure 20.2  South Africa’s imports composition in Africa, 2019 Source: SARS (2020).

the continent was skewed heavily towards East and Southern Africa—​the so-​called ESA region. The ESA region accounted for 88 per cent and 90 per cent of South Africa’s African exports and imports respectively (Fundira 2017). Both SACU and SADC account for the majority of intra-​African trade, dominated considerably by South Africa (see Figure 20.3). In 2017, South Africa accounted for 71 per cent of intra-​SACU exports (see Figure 20.3). SACU remains the most crucial trading arrangement for South Africa (Makokera-​Grant and Makokera 2020) and it provides important benefits for the country, including savings on import duties and transportation costs for South African businesses.

0

81.5 53.4 EAC

Rest of Africa 2010–2012 Intra-regional economic community 2010–2012

ECOWAS

52.2 47.8

49.0 IGAD

48.2

51.0

50.4

17.7

ECCAS

49.6

56.7

82.3

46.6

43.3 48.3 33.9

44.7 CEN-SAD COMESA

66.1

5

51.7

10

55.3

15

59.5 40.5

20

61.8 38.2

25

58.4 52.6

30

41.6

35

84.9 15.1

40

47.4

45

18.5

50

SADC

51.8 AMU

Rest of Africa 2014–2016 Intra-regional economic community 2014–2016

Figure  20.3 Intra-​ regional economic community trade in Africa, 2010–​ 12 and 2014–​ 16 (billions of dollars and percentage of total African trade) Source: UNCTAD (2019).

South Africa’s Economic Role in Africa    435 Since 1994, there has been a surge in outward foreign direct investment (FDI) by South African firms to the African continent. South Africa is one of the major investment suppliers to the continent (Kiss 2017) and the bulk of its FDI has been destined for the SADC region (Draper, Kiratu, and Samuel, 2010). In 2014, South African investments in Africa totalled $5.6 billion—​with the major investments in telecommunications, mining, and retail. At the start of the decade, South African companies led a surge in cross-​border greenfield projects, which rose by 18 per cent between 2009 and 2013. Although South Africa’s FDI inflows declined by 15 per cent in 2019, the country continued to be the biggest capital exporter in Africa (UNCTAD 2020). Almost half of Africa’s large firms are from South Africa, with companies such as Bidvest, Anglo Gold Ashanti, MTN, Shoprite, Standard Bank, Aspen Pharmacare, and Naspers being among the prominent investors on the continent (Leke, Chironga, and Desvaux 2018).

20.3.5 The African Continental Free Trade Area South Africa ratified the AfCFTA in February 2019. The AfCFTA was formally launched at the 12th Extraordinary Summit of the AU in Niamey, Niger on 7 July 2019. It is designed to establish a continent-​wide market for goods and services in Africa, generate opportunities to grow intra-​Africa trade, create jobs, and develop power infrastructure. It brings together fifty-​four African countries with a combined gross domestic product in excess of $3.4 trillion (Tralac 2020). It is one of the flagship projects of the AU’s Agenda 2063, which seeks to tackle Africa’s developmental problems, including youth unemployment, women’s empowerment, infrastructure development, and industrialization. Upon its implementation, the AfCFTA will cover a market of over 1.2 billion people and will be the world’s biggest free trade area since the founding of the World Trade Organization. It has the potential to bolster intra-​African trade by 52.3 per cent. Currently, only 18 per cent of Africa’s exports go to other African countries. This is in contrast to Europe and Asia, where intra-​continental trade is respectively at 70 per cent and 60 per cent (Napoli 2021). Trading under the AfCFTA Agreement had initially been scheduled to start on 1 July 2020, but it only began on 1 January 2021 due to delays caused by the outbreak of the global coronavirus pandemic. Even though the AfCFTA is unlikely to transform the trade patterns of individual countries in the short term, it will nonetheless help to unlock opportunities to expand intra-​Africa trade and generate spillover effects realized through creating dynamic markets of complementary goods and services. In 2019, 14.4 per cent of African exports went to other African countries, a small proportion compared with the 52 per cent in intra-​Asian trade and 73 per cent between European nations in the same year (Pilling 2020). While the AfCFTA is expected to boost intra-​African trade, it is also projected to lay the basis for ultimately creating a continental customs union (African Union 2018). That ambition may take many decades to realize. Despite its great promise, the AfCFTA is likely to be faced with

436    Mills Soko and Mzukisi Qobo similar challenges to those experienced in the implementation of the SADC Protocol on Trade. The reality is that most countries on the African continent have been hamstrung by, among others, poor leadership, a lack of political will, economic mismanagement, an excessive reliance on tariffs for state revenues, weak institutions, deficient infrastructure, and overlapping memberships in regional integration schemes. Addressing these problems is vital to fostering successful regional integration and placing the African continent on an upward trajectory.

20.4 Conclusion This chapter has examined South Africa’s role in Africa since 1994. Africa constitutes the foremost priority in South Africa’s foreign policy. The South African government regards the country’s long-​range national destiny as indivisibly tied to that of the wider African region. South Africa’s foray into Africa has been a result of intentional internal and external foreign policy decisions enacted by consecutive ANC administrations. South Africa’s vision of the African Agenda has been informed by its desire to shape the African continent’s development agenda. NEPAD epitomizes the policy and institutional expression of the African Agenda. It has produced clear successes. These include remarkable growth rates, higher agricultural productivity, accelerated infrastructure development, increased external development funding, and improved dialogue between African leaders and their G8 counterparts. Yet NEPAD has also been beset with failures and problems, including financial, technical, and human capacity constraints. South Africa was influential in the establishment of the AU and has invested extensive diplomatic energy and financial resources into making the continental body effectively operational. Furthermore, South Africa has been a major supplier of aid to the African continent. It has sought to use foreign aid as a tool of ‘soft power’ to mobilize the support of African countries for its foreign policy ambitions. However, the country’s relationship with the AU has not been without problems. It has been fractured by, among others, tensions over the future shape of NEPAD as well as the furore that marred Dr Nkosazana Dlamini-Zuma’s ascendancy to the leadership of the AU Commission. Moreover, the cost of AU membership has become highly exorbitant even by the standards of the most industrialized country on the African continent, faced with the dilemma of balancing its domestic concerns with its continental obligations. The incidents of xenophobic violence against African migrants and refugees have undermined South Africa’s foreign policy and damaged its international reputation. They have limited the country’s ability to project its ‘soft power’ and sowed doubts about whether it has the moral authority to lead Africa. The South African government ought to do more to protect African migrants and refugees, enforce the rule of law, and improve public-​order policing. Importantly, it must address the underlying causes of xenophobia. South Africa must be mindful of, and attuned to, how its role and influence is

South Africa’s Economic Role in Africa    437 perceived on the African continent. This is necessary if the country is to salvage its credibility and become accepted as a continental leader. The southern African region has been the bedrock of South Africa’s external engagements. The adoption of the SADC Protocol on Trade was inspired by the ambition on the part of SADC countries to achieve economic growth and development and facilitate their insertion into the world economy on favourable terms. Given its economic preponderance in the SADC region, other countries expected South Africa to be more accommodative and accelerate its tariff liberalization. However, as the issue of rules of origin has shown, South Africa has at times embraced mercantilist trade practices that have inflamed political sensitivities in the region. South Africa needs to do more to act in ways that balance its national interests with its regional commitments. Nevertheless, South Africa has acted differently in respect of SACU. By signing and ratifying the 2002 SACU Agreement, the country created the conditions for the democratization of the customs union. Not only did it give birth to new democratic institutions for SACU, it overhauled its negotiating procedures and recast its revenue-​sharing formula. The transfer of the terms of the SADC-​EU EPA into the SACUM-​UK EPA helped to forestall trade disruption and ensure continuity in trade ties after the United Kingdom’s departure from the European Union. This augurs well for the future of trade relations of the customs union and the United Kingdom. The AfCFTA, if successfully implemented, has enormous potential to create a continent-​wide market for goods and services, generate opportunities to grow intra-​Africa trade, and deliver jobs and power infrastructure development. Notwithstanding its considerable promise, nonetheless, the AfCFTA is likely to be faced with similar challenges to those experienced in the implementation of the SADC Protocol on Trade. South Africa has largely succeeded in fulfilling the goals outlined in its foreign policy and African Agenda. Nonetheless its actions, as the discussion on the implementation of the Protocol on Trade’s rules of origin in the SADC region has demonstrated, have also undermined the economic interests of its regional neighbours. The degree to which South Africa will be able to sustain its foreign policy performance in Africa in the future will depend on how far it weighs its domestic policy pressures with its continental obligations, on how it responds to and accommodates the economic concerns of its regional neighbours, and on how it manages African perceptions of the country.

References African National Congress. 1994. The Reconstruction and Development Programme. Johannesburg: Umanyano Publications. African Union. 2018. ‘Agreement establishing the African Continental Free Trade Area’. [online] https://​au.int/​sites/​default/​files/​treaties/​36437-​treaty-​consolidated_​text_​on_​cfta_​-​_​ en.pdf. Alden, Chris, and Mills Soko. 2005. ‘South Africa’s economic relations with Africa: Hegemony and its discontents’, Journal of Modern African Studies, 43(3): 367–​92.

438    Mills Soko and Mzukisi Qobo Besharati, Alessandro Neissan. 2013. South African Partnership Development Agency (SADPA): Strategic Aid or Development Packages for Africa? Braamfontein: SAIIA. Braude, Wolfe, Pearl Thandrayan, and Elizabeth Sidiropolous. 2008. Emerging Donors in International Development Assistance: The South African Case. Braamfontein: SAIIA. de la Rocha, Manuel. 2003. ‘The Cotonou Agreement and its implications for the regional trade agenda in Eastern and Southern Africa’. World Bank Policy Research Working Paper, June, Washington, DC. Department of International Relations and Cooperation. 2011. ‘Building a better world: The diplomacy of Ubuntu’. White Paper on South Africa’s foreign policy, 13 May [online] https://​ www.gov.za/​sites/​default/​files/​gcis_​document/​201409/​foreignpolicy0.pdf. Department of International Relations and Cooperation. 2019. ‘Foreign policy review: A strategic reflection and critical appraisal of the orientation and implementation of South Africa’s foreign policy’. Media Briefing on the Ministerial Panel Report, 17 April [online] http://​ www.dirco.gov.za/​docs/​2019/​foreign_​policy_​review_​report0417.pdf. Department of Trade and Industry. 2020. ‘Economic Partnership Agreement (EPA) between the Southern African Customs Union (SACU) and Mozambique (together SACUM)-​ United Kingdom (UK)’. Presentation to WESGRO Brexit Readiness Seminar 15 January. UK Department of Trade and Industry, London. Draper, Peter, Sheila Kiratu, and Cezanne Samuel. 2010. ‘The role of South African FDI in Southern Africa’. [online] https://​www.files.ethz.ch/​isn/​120589/​2010-​08e.pdf. Economic Commission for Africa. 2007. ‘Challenges and prospects in the implementation of NEPAD’. Report by the RCM-​Africa Secretariat. Flatters, Frank. 2002. ‘SADC rules of origin:  Undermining regional free trade’. A  paper prepared for TIPS Forum, South Africa, 9–​11 September. Fundira, Taku. 2017. ‘South Africa’s trade relationship with Africa 2010–​ 2016’. Tralac, Stellenbosch [online] https://​www.tralac.org/​resources/​infographic/​12045-​south-​africa-​s-​ trade-​relationship-​with-​africa.html. Gelb, Stephen. 2001. ‘South Africa’s role and importance in Africa and for the development of the African Agenda’. The Edge Institute, Braamfontein. Gibb, Richard. 1997. ‘Regional integration in post-​apartheid Southern Africa:  The case of renegotiating the Southern African Customs Union’, Journal of Southern African Studies, 23(1): 67–​86. Habib, Adam. 2009. ‘South Africa’s foreign policy:  Hegemonic aspirations, neoliberal orientations and global transformation’, South African Journal of International Affairs, 16(2): 143–​59. Hammerstad, Anne. 2005. ‘People, states and regions’, in Anne Hammerstad (ed.) People, States and Regions. Braamfontein: SAIIA. Holden, Merle. 2003. ‘Southern African economic integration’, The World Economy, 17 February. Khati, Thekiso. 2006. ‘The achievements and challenges of the New Partnership for Africa’s Development’, Africa Insight, 36(2):25–​34. Kiss, Judit Vekei. 2017. ‘South Africa—a re-​emerging player in outward FDI’. IWE Working Paper no. 235, Institute for World Economics, Centre for Economic and Regional Studies, Hungarian Academy of Sciences. Krueger, Anne Osborne. 1993. ‘Free trade agreements as protectionist devices: Rules of origin’. NBER Working Paper no. 4352, April, Cambridge, MA.

South Africa’s Economic Role in Africa    439 Landsberg, Chris. 2007. ‘South Africa and the making of the African Union and NEPAD:  Mbeki’s “progressive African agenda”’, in A. Adebajo, A. Adedeji, and C. Landsberg (eds) South Africa in Africa: The Post-​Apartheid Era. Scottsville: University of KwaZulu-​Natal  Press. Landsberg, Chris. 2009. ‘South Africa’s “African Agenda”: Challenges of policy and implementation’. Report prepared for the Presidency Review Project. Landsberg, Chris, and Kwandiwe Kondlo. 2007. ‘South Africa and the “African Agenda”’, Policy, Issues and Actors, Centre for Policy Studies, 20(13): 1–​14. Leke, Acha, Mutsa Chironga, and Georges Desvaux. 2018. ‘Africa’s overlooked business revolution’. [online] https://​www.mckinsey.com/​featured-​insights/​middle-​east-​and-​africa/​ africas-​overlooked-​business-​revolution. Maloka, Eddy. 2019. When Foreign Becomes Domestic:  The Interplay of National interests, Pan-​ Africanism and Internationalism in South Africa’s Foreign Policy. Sandton:  Ssali Publishing House. Mandela, Nelson. 1993. ‘South Africa’s future foreign policy’, Foreign Affairs, 72(5): 86–​97. Grant-​Makokera, Catherine, and Mary-​Beth Makorera. 2020. ‘South Africa’s trade policy post Covid-​19’. [online] https://​saiia.org.za/​research/​south-​africas-​trade-​policy-​post-​covid-​19/​ #. Mbeki, Thabo. 1996. ‘I am an African’. Address of the President of South Africa, Thabo Mbeki, at a Joint Sitting of the Houses of Parliament on the Occasion of the 10th Anniversary of the Adoption of the Constitution of the Republic of South Africa, Cape Town, 8 May [online] http://​www.dirco.gov.za/​docs/​speeches/​2006/​mbek0508.htm. McCarthy, Colin. 1996. ‘Regional integration: Part of the solution or part of the problem?’. in Stephen Ellis (ed.) Africa Now. Oxford: James Currey. McCarthy, Colin. 1998. ‘South African trade and industrial policy in a regional context:  Economic challenges and policies for the future’, in Lennart Petersson (ed.) Post-​ Apartheid Southern Africa—​ Economic Challenges and Policies for the Future. London: Routledge. Mda, Nomazulu. 2004. ‘South Africa’s role in conflict resolution in Southern Africa: Prospects for cooperation with the US’, in SA Yearbook of International Affairs 2003/​ 04. Braamfontein: SAIIA. Napoli, Antonella. 2021. ‘AfCFTA to start, Africa is the largest free trade area on the planet’, Focus on Africa, 1 January [online] https://​www.focusonafrica.info/​en/​afcfta-​to-​start-​africa-​ is-​the-​largest-​free-​trade-​area-​on-​the-​planet/​. NEPAD Secretariat. 2001. ‘New Partnership for Africa’s Development’. Midrand: NEPAD. Nolutshungu, Sam Clement. 1994. ‘South Africa’s position in world politics’, in Vincent Maphai (ed.) South Africa—​The Challenge of Change. Harare: Sapes Books. Pambazuka News. 2015. ‘Open letter to the AU on resurgence of xenophobic violence in South Africa’. Presented on the occasion of the AU Summit in Sandton, Johannesburg, June [online] https://​www.pambazuka.org/​governance/​ open-​letter-​au-​resurgence-​xenophobic-​violence-​south-​africa. Patel, Ebrahim. 2019. ‘Statement by the minister of trade and industry on the conclusion of an agreement with the UK addressing the trading relationship in the event of a no-​deal Brexit’. 17 September. Department of Trade and Industry: Pretoria. Pilling, David. 2020. ‘African countries not ready to implement free trade from January’, Financial Times, 29 December.

440    Mills Soko and Mzukisi Qobo Qobo, Mzukisi. 2012. ‘South Africa’s foreign policy: An assessment of the draft White Paper and the ANC’s policy document on international relations’. South African Foreign Policy Initiative Brief no. 2, May, Braamfontein. SADC Secretariat. 2004. Regional Indicative Strategic Development Plan 2004. Gaborone: SADC. https://​www.sadc.int. Schoeman, Maxi, Asnake Kefale, and Chris Alden. 2017. ‘It’s time South Africa tuned into Africa’s views about its role in Africa’, The Conversation Africa, 24 January. Sidiropoulos, Elizabeth. 2008. ‘South African foreign policy in the post-​Mbeki period’, South African Journal of International Affairs, 15(2): 107–​20. Soko, Mills. 2008. ‘Building regional integration in Southern Africa: Southern African customs union as a driving force?’. South African Journal of International Affairs, 15(1): 55–​69. Soko, Mills. 2020. ‘Here’s why South Africa and its neighbours are anxious about EU and UK post-​Brexit trade talks’, The Conversation Africa, 18 September. South African Government. 1996. ‘Foreign policy for South Africa: Discussion document’. [online] https://​www.gov.za/​documents/​foreign-​policy-​south-​africa-​discussion-​document-​ 0?gclid=EAIaIQobChMI3OOMy8zr7gIVd4BQBh0aZADyEAAYASAAEgL3iPD_​BwE. Tekere, M. 2001. ‘Can trade arrangements reduce poverty in Southern Africa?’. Article adapted for Bridges from a paper presented at the Southern Africa Regional Poverty Network Discussions, 26 April. [online] https://​sarpn.org/​RegionalPovertyPapers/​April2001/​ rppApril2001b.pdf. Tralac. 2018. ‘Economic Partnership Agreement between the European Union and Southern African Development Community Group’. FAQs no.  2 [online] https://​www.tralac.org/​ documents/​resources/​faqs/​2049-​sadc-​eu-​epa-​faqs-​july-​2018/​file.html. Tralac. 2020. ‘The African Continental Free Trade Area: A Tralac guide’. [online] https://​www. tralac.org/​publications/​article/​13997-​african-​continental-​free-​trade-​area-​a-​tralac-​guide. html. United Nations. 2011a. ‘New Partnership for Africa’s development: Ninth consolidated progress report on implementation and international support’. Report of the Secretary-​General, A/​ 66/​202, 28 July. United Nations. 2011b. ‘NEPAD:  Building foundations for a new Africa’, Africa Renewal, December. United Nations Conference on Trade and Development. 2012. ‘The New Partnership for Africa’s development:  Performance, challenges and the role of UNCTAD’. Note by the UNCTAD Secretariat. Geneva. United Nations Conference on Trade and Development. 2020. ‘Investment flows in Africa set to drop 25% to 40% in 2020’. [online] 16 June https://​unctad.org/​news/​ investment-​flows-​africa-​set-​drop-​25-​40-​2020. World Bank. 2005. Global Economic Prospects:  Trade, Regionalism, and Development. Washington, DC: World Bank. World Trade Organization. 1986. The Legal Texts:  The Results of the Uruguay Round of Multilateral Trade Negotiations. Cambridge: Cambridge University Press.

Chapter 21

Sou th Af ri c a ’ s Internationa l   T ra de Lawrence Edwards

21.1 Introduction South Africa faces an economic growth and employment crisis. Gross Domestic Product (GDP) grew at a lacklustre average of 1.67 per cent per year from 2010 to 2019, down from 3.5 per cent per year over the prior post-​apartheid period 1994 to 2009.1 The implication has been a failure of the economy to generate sufficient employment opportunities to absorb new entrants into the labour force, resulting in sustained high levels of unemployment. One reason for this weak growth is South Africa’s failure to enhance its long-​run export performance, both in absolute terms and relative to its middle-​income peers (Hausman and Klinger 2008). Export volumes per capita grew at an annualized rate of only 0.45 per cent from 1960 to 2019 compared to over 3 per cent in other resource-​abundant economies such as Argentina, Australia, Brazil, Chile, and Malaysia. The implication has been that South Africa’s share of world merchandise exports fell from 1.56 per cent in 1962 to 0.4 per cent in 2019. There is no shortage of policy recommendations to enhance export performance in South Africa. The Integrated National Export Strategy (DTI 2016), the National Development Plan (NDP) Vision 2030 (National Planning Commission 2012), and the Department of Trade and Industry’s Industrial Policy Action Plans all see raising exports and diversifying the export bundle as central policy objectives. Under the NDP, the objective is to increase export volumes by 6 per cent per annum by 2030, with non-​ traditional exports growing by 10 per cent a year. 1 This chapter draws on data from several sources including the South African Reserve Bank, UN Comtrade data via the World Integrated Trade Systems, Statistics South Africa, World Trade Organization database and the World Bank Development Indicator database. Unless otherwise specified, all data values presented in the chapter are drawn from these sources.

442   Lawrence Edwards Enhancing export growth is also imperative from a macroeconomic and growth perspective. South Africa’s macro-​competitiveness—​the ability of the economy to generate sufficient foreign exchange through trade to support the growth process (Thirlwall 1979)—​has diminished (Bell et al. 2002). Whereas during the 1960s a current account deficit to GDP ratio of 2.1 per cent was associated with a two-​year average GDP growth rate of 6 per cent, over the 2015–​18 period, a higher deficit ratio of 3.4 per cent was associated with a growth rate of only 0.94 per cent per annum. The implication of this decline in macro-​competitiveness is that a much higher export growth is required to sustain a growth recovery. Should export growth not recover, South Africa faces the prospect that higher economic growth will be choked off by a lack of foreign exchange, as has occurred frequently in the past.2 This chapter uses South Africa’s integration in the global economy as a lens to understand the dynamics behind South Africa’s economic performance. It first presents the historical context leading up to South Africa’s current situation, commencing from the country’s position as primarily a gold exporter pursuing an import substitution industrialization strategy, to its transition to a more open economy with the ending of sanctions and tariff liberalization from the early 1990s. The focus then shifts to a more critical assessment of South Africa’s trade performance in the post-​apartheid period. The post-​apartheid period is characterized by shifts in government policy on international trade—​multilateral liberalization from 1994 to 2000, preferential tariff reform from 2000 and sector-​driven industrial policy from around 2007—​and dramatic changes in the global trading order—​the rise of China from 2001, and the emergence of global value chains in driving participation by firms in international trade. Key questions debated in the literature will be considered: How has trade liberalization contributed to structural change in the composition of trade, including ‘premature deindustrialization’ of the South African economy? What is the impact of the rise of China on trade, industrial production, and employment? Has trade liberalization shifted South Africa onto a new export-​led growth path? To illustrate these relationships, the chapter draws on new insights based on research using disaggregated product-​level trade data and firm-​level data. The product-​level trade data provides a deeper understanding of the product and destination/​origin dynamics that lie behind South Africa’s growth in international trade. The firm-​level data provides a more nuanced understanding of the heterogeneity behind aggregate trade flows. Analysis using this data opens up new insights on participation by South African firms in international trade posing new challenges to policymakers. The remainder of the chapter is structured as follows. Section 21.2 presents an overview of South Africa’s historical trade performance leading up to 1994. Section 21.3 discusses South Africa’s post-​1994 programme of tariff reform and provides an overview

2 Acute balance-​ of-​payment crises in response to foreign exchange outflows following the 1976 Soweto riots and the debt crisis in 1985 led to the imposition of emergency measures, including surcharges and import controls, to curtail imports.

South Africa’s International Trade    443 of trade performance. Section 21.4 then analyses the relationship between trade, production, and employment in manufacturing. Section 21.5 concludes the chapter.

21.2  South African Trade Performance up to 1994 South Africa’s historical growth process and industrial development from the late 1800s to 1994 is defined by its participation in international trade. As covered in ­Chapters 2 and 3 of this volume, the discovery and export of gold and to a lesser extent diamonds, transitioned the economy onto an export-​led growth path (Feinstein 2005). Gold exports rose rapidly from close to zero in the early 1880s to 60 per cent of total exports by 1905–​09 and then 72 per cent by 1935 (Feinstein 2005: 101). Exports were further boosted by the discovery of new gold reefs in the Transvaal and the Orange Free State from 1939, with output of gold rising from 11 million fine ounces in 1945–​48 to an all-​time peak of 32.2 million fine ounces in 1970 (Feinstein 2005). The associated trends in trade volumes of goods and services and real GDP (indexed such that 2010 = 100) are illustrated in Figure 21.1. From 1946 to 1970, export volumes of goods and services, driven by gold, grew by 7.5 per cent per annum. Real GDP also grew strongly rising by 4.6 per cent per year in the 1950s and 5.3 per cent in the 1960s. In addition to its direct impact on the economy during this period, the production of gold generated large rents that served as a source of income and taxes and royalties to government revenue that financed investment in infrastructure (Feinstein 2005). Gold

Trade volumes (goods and services) and real GDP, 2010 = 100 140

120 100 80 60 40

0

1946 1949 1952 1955 1958 1961 1964 1967 1970 1973 1976 1979 1982 1985 1988 1991 1994 1997 2000 2003 2006 2009 2012 2015 2018

20

Export volumes

Import volumes

GDP

Figure 21.1  Trade volumes (goods and services) and real GDP, 2010 = 100 Source: Own calculations using data obtained from the National Income and Production Accounts data provided by the South African Reserve Bank. GDP is measured in market prices. Trade volumes include goods and non-​factor services. Income receipts and payments are not included.

444   Lawrence Edwards exports played a vital role in sustaining the country’s balance of payments and earning of foreign exchange to offset the rise in imports associated with the high economic growth. The extraction of gold attracted foreign capital to South Africa, leading to the establishment of large mining houses that would later play a central role in the industrialization process through diversification of their investments and provision of technical and managerial skills for manufacturing and other sectors (Feinstein 2005: 109). Further gold production supported and enabled the process of industrialization, serving as a source of demand for locally produced manufactured consumer and intermediate goods, a process encouraged by the Import Substitution Industrialization (ISI) policies following the enactment of the 1925 Customs Tariff and Excise Duties Amendment Act (no. 36). The ISI policies initially focused on protecting semi-​durable consumer goods, but from the 1930s protection extended to industrial products such as cement and steel, and then from the 1950s into the upstream chemicals industry (Bell 1993; Fallon and Pereira de Silva 1994). The Second World War provided a further stimulus to local industry with domestic manufacturers responding to opportunities created by shortages of imported products and new demands arising from South Africa’s war effort (Feinstein 2005: 123). In addition, active state intervention shaped the growth and composition of the manufacturing sector through the establishment of the Electricity Supply Commission (Eskom) in 1923, the Iron and Steel Industrial Corporation (ISCOR) in 1928 and then later the South African Coal, Oil and Gas Corporation (SASOL) in 1950, giving rise to what Fine and Rustomjee (1996) termed the mineral-​energy complex. These policies drove a rapid increase in manufacturing as a share of GDP from 8.1 per cent in 1926/​27 to 23 per cent by the 1970s, as well as a decline in manufactured imports as a share of domestic production (101.4 per cent in 1936/​37 to 30.3 per cent in 1967/​68) (Marais 1981: 36). The impact is evident in Figure 21.1 where import volumes grew slowly relative to GDP in the 1950s. While exports of manufactured goods also rose in value and as a share of exports (see Table 21.1), the increase was driven by the growing production capacity in the sector rather than a shift towards export-​orientation (Holden 1990). Gold, and other mining products, continued to dominate exports, and manufacturing exports remained heavily oriented towards resource-​based products (mainly minerals-​ based). Concerns regarding the continued dominance of gold in exports led to the Reynders Commission in 1969 that recommended export promotion policies to redress some of the anti-​export bias of the existing tariff policies and stimulate manufacturing export growth (Ratcliffe 1975). The period 1971/​72 to 1994 signals dramatic shifts in South Africa’s trade and economic performance, as can be seen in Figure  21.1. GDP growth fell from the highs of the 1960s, with the country entering into a recession over the period 1990 to 1992. Growth in real manufacturing gross value added also slowed during the 1970s, with the sector’s share of real gross value added reaching a peak of 17.8 per cent in 1981 (23 per cent to 24 per cent if using nominal values). Real manufacturing value added at the end of 1993 was actually lower than it was in 1981. The manufacturing share of non-​ agricultural employment also began to fall from its 1981 peak of 23 per cent. The process

South Africa’s International Trade    445 Table 21.1: South African export composition by technology classification (values and share structure) 1962

1970

1980

1990

2000

2010

2019

Total Value Trade (US dollar billions)

2.0

3.2

24.4

20.4

30.3

79.1

77.2

Annual growth in decade

5.8

20.4

–​1.8

4.0

9.6

–​0.2

5.8

Gold

44.0

36.7

53.3

31.7

13.2

10.3

6.0

Primary, excl. gold

33.7

37.5

25.2

36.5

33.4

30.2

29.7

Total Manufacturing

22.2

25.7

21.5

31.8

53.3

59.5

64.3

Resource-​based

84.0

69.3

60.8

50.2

35.7

40.7

43.2

Low technology

6.3

7.2

13.8

15.9

16.6

9.8

6.9

Medium technology

9.1

22.0

17.6

29.1

41.1

45.5

46.3

High technology

0.6

1.5

7.8

4.8

6.6

4.0

3.6

91.8

86.0

75.8

73.8

55.2

60.0

58.7

8.2

14.0

24.2

26.2

44.8

40.0

41.3

Share total value (%)

Share in manufacturing exports (%)

Commodity manufacturing Non-​commodity manufacturing

Source: Own calculations using Feenstra et al. (2005) World Trade Flow database from 1962 to 2000, and UN Comtrade data from 2001. Notes: Categories are based on the Lall (2000) Technology Classification of exports. Excludes special transactions that range from 0 to 3% of trade values. 1960s covers the period 1962–​70. Commodity manufacturing includes Resource-​based manufactures plus Process industries (Synthetic fibres, chemicals and paints, fertilizers, plastics, iron, pipes/​tubes).

of deindustrialization, or what Rodrik (2016) refers to as ‘premature deindustrialization’, in South Africa had commenced at a level of GDP per capita and manufacturing share of employment that was far lower than was the experience in the advanced and newly industrialized countries.3 Trade performance from the 1971/​72 period was also poor. Export volumes plateaued in absolute terms and relative to GDP, and only started a sustained recovery, driven by manufactures, after a massive depreciation of the real exchange rate from 1984. By the end of 1993, the volume of goods and services exports was only 26 per cent higher than in 1971. Import volumes also stagnated during this period (Figure 21.1) driven lower by 3 

South Africa achieved a peak manufacturing share of total employment of 17 per cent at 6,500 US dollars per capita (1990 PPP dollars), compared to advanced economies where employment shares peaked at 25–​33 per cent at approximately 14,000 US PPP dollars per capita (Rodrik 2016). Tregenna (2016) adopts a more nuanced view of deindustrialization that takes into account whether the level of manufacturing employment and GDP have also started to diminish. Even in this case, deindustrialization appears to have occurred from 1980 as real volumes of manufacturing collapsed, and the sector’s share (real and nominal) of total GDP reached a peak. Manufacturing employment levels, however, continued to rise, only starting to decline from the late 1980s.

446   Lawrence Edwards weaker economic growth and the intermittent use of import surcharges to deal with persistent balance-​of-​payments crises. Import surcharges were first used from 1977 to 1979 in response to the cessation of capital inflows after the Soweto riots in 1976, from 1982 to 1983 following the collapse in the gold price, and then again in 1985 after a sovereign debt crisis ensued following the ‘Rubicon Speech’ by the then President Botha (Cassim et al. 2009). These surcharges reached up to 60 per cent for luxury items. Like exports, by 1993 import volumes were only 28 per cent higher than they were in 1972. This stagnation of South African trade is even more striking when compared to trends in global trade volumes that grew by 140 per cent over the period 1972 to 1993. The principal reason for the poor growth in aggregate export volumes was the decline in gold production that close to halved over the period 1970 to 1994 (Feinstein 2005: 206). Non-​gold exports of goods and services actually increased moderately. Manufacturing exports, for example, grew sharply following the depreciation of the Rand in 1985, driving up aggregate export volumes (Figure 21.1). However, this growth was primarily driven by vent-​for-​surplus sales in response to excess capacity following the domestic recession and not new investment in export-​oriented production facilities (Fallon and Pereira de Silva 1994). Further, the growth in non-​gold exports was insufficient to fully transition South Africa onto a manufacturing export-​led growth path, unlike what was happening in Malaysia at this time (Hausmann and Klinger 2008). The implication was that the export bundle by 1994 was still heavily dependent on gold and other commodities. Several reasons underpin this lack of transition into a manufacturing export-​led growth path during the 1980s. Bell et al. (2002) argue that the composition of South African exports was primarily an outcome of macroeconomic forces, including economic growth, and external forces rather than a lack of an ‘export culture’, or the ‘anti-​export bias’ created by protection (Bell 1993, 1997; Bell et al. 2002). In particular, South Africa experienced its own version of the Dutch Disease effect in response to the unexpected boom in the gold price from just under 200 US dollars in 1978 to a peak of 850 US dollars in 1980. The effect was a sharp appreciation (38 per cent) of the real effective exchange rate from 1978 to 1983 that undercut the relative profitability of manufacturing exports, particularly non-​commodity exports. Global economic stagnation and world economic crises (e.g. oil crises) of the 1970s further reduced demand for South African manufactured goods. These explanations, however, discount the impact of domestic policies and the severe structural and political problems that undermined the competitiveness of the manufacturing sector. Although South Africa had already initiated several tariff reforms during the 1970s and 1980s with the tariffication of import quotas and removal of import licences, the tariff structure remained complex, and protection high (Belli et al. 1993). The late 1980s saw import barriers rise in response to the imposition of import surcharges, and the award of tariff protection to businesses struggling during the economic downturn (Holden 1992; Bell 1993; Cassim et al. 2009).4 High effective protection 4  For a detailed overview of South Africa’s tariff policies, including the political economy of tariff policy, see Holden (1992), Bell (1993), Bell (1997) Casale and Holden (2002), Edwards (2005), Edwards (2011) and Hirsch and Hines (2005).

South Africa’s International Trade    447 rates and tariffs on imported intermediate inputs meant that production for the domestic market continued to be incentivized as opposed to sales for the export market (Fallon and Pereira de Silva 1994). The anti-​export bias arising from tariff protection was higher for non-​commodity manufactures given their greater dependence on intermediate inputs in production. The consequence was that tariff protection reinforced the commodity-​intensity of the manufacturing export bundle (Edwards and Lawrence 2008a, 2008b). Finally, although new export incentives were introduced under the General Export Incentive Scheme (GEIS) in 1990 to offset some of the anti-​export bias, they primarily benefited the large capital-​intensive manufacturers of intermediate products such as paper, steel, and basic chemicals, who received huge tax-​free windfalls on products they would have exported anyway (Hirsch and Hines 2005). The transition into manufacturing exports was also impeded by inherent structural constraints associated with South Africa’s mining and apartheid policies. The migrant labour system was inconducive to the creation of manufacturing jobs, where resident labour was required. Job reservations and education policies diminished the supply of relatively skilled labour required in manufacturing production (Feinstein 2005). These constraints revealed themselves in rising capital intensity of production and low total factor productivity growth (0.05 per cent per annum from 1972 to 1983) (Belli et al. 1993). Rising wages without concomitant increases in productivity further increased production costs, negatively affecting manufacturing export performance (Edwards and Golub 2004). The competitive basis for manufacturing to launch into exporting had not yet been achieved. Externally imposed trade sanctions during the 1980s further constrained South African export performance, although the extent of the impact is debated as the trade sanctions were narrowly targeted (agricultural goods, uranium, coal, oil, and iron and steel) and widely circumvented by South African firms (Lipton 1988; Manby 1992; Evenett 2002). The most profound effect of the sanctions on manufacturing exports may actually be the investment ban and the pressure placed on foreign firms to disinvest from the South African economy in the late 1980s. This constrained access to foreign direct investment at a key time when this was needed to integrate South African manufacturing firms into global value chains that were to emerge as the driving force behind manufacturing trade.

21.3  Trade Liberalization from 1994 The democratic election of 1994 coincided with a dramatic shift in South Africa’s policies towards an open trade regime achieved through a tariff liberalization programme negotiated during the Uruguay Round of the GATT/​WTO.5 The tariff structure

5  While this section refers to South African tariffs, South Africa is part of the Southern African Customs Union (SACU) and consequently adopts the common external tariff of SACU. However, South Africa de facto makes tariff determinations on behalf of the other members.

448   Lawrence Edwards was rationalized and import weighted tariff rates on manufactured goods, inclusive of surcharges, fell from 19.9 per cent in 1994 to 9.6 per cent in 2000 (Cassim et al. 2009). Although contested in the literature (Fedderke and Vase 2001; Cassim 2003; Rangasamy and Harmse 2003), the reductions in nominal tariffs led to even larger reductions in effective protection from 43.3 per cent to 14.9 per cent for manufacturing over the period 1994–​2000 (Edwards 2005). Openness of the economy was further enhanced through the removal of import surcharges and the ending of trade and investment sanctions. What drove the dramatic reductions in tariff protection? One explanation is that South Africa participated as a ‘developed’ country during the Uruguay negotiations and was therefore required to adopt large and rapid cuts in protection. Despite this, there was broad political support for tariff reform by business, government, and labour, as well as the African National Congress, who through its alliance partner the Congress of South African Trade Unions (COSATU), actively participated in the consultations leading up to the offer (Hirsch and Hines 2005). Bell (1997) argues that their interest in binding tariff reductions was to signal a clear departure in economic policies from the apartheid regime, induce industrial efficiency through the curbing of domestic monopoly power, and lower consumer prices. The apartheid regime, in turn, was interested in removing levers of industrial policy from the future government. Actual tariff cuts on clothing and motor vehicles went beyond what was required under South Africa’s offer to the WTO, reflecting significant unilateral trade liberalization (Bell 1997). Domestic policy reforms further enhanced the shift towards an open trade regime. This includes the Growth, Employment, and Redistribution (GEAR) macroeconomic policy in 1996 that aimed to transform South Africa into a ‘competitive, outward orientated economy’ (Republic of South Africa 1996), and the deregulation of agricultural marketing and control boards established under the Agricultural Marketing Act of 1968. The opening up of the economy, therefore, primarily arose from a domestic policy programme, as opposed to being externally imposed. While multilateral tariff liberalization largely came to a halt from 2000, aggregate protection continued to fall through preferential trade agreements. During the 1990s and early 2000s, the policy objective was to conclude free trade areas (FTAs) that covered substantially all trade. These include the Southern African Development Community (SADC) Free Trade Protocol (from 2000); the South Africa-​European Union (EU) Trade, Development and Cooperation Agreement (TDCA) (from 2000), which was replaced by the SADC Economic Partnership Agreement in 2016; and a free trade agreement with the European Free Trade Association (EFTA) in 2008. The effect of these agreements was a reduction in tariffs by 2018 to 3.36 per cent for EFTA members, 2.37 per cent for EU members, and close to zero for SADC members.6 In the case of SADC, however, overly complex and restrictive rules of origin requirement preclude many SADC countries from accessing the South African market at these preferential tariff rates (Brenton et al. 2005).

6 

Based on calculations using the ad valorem components of SACU tariff rates at the 8-​digit level of the Harmonized System classification.

South Africa’s International Trade    449 In 2007, the then named Department of Trade and Industry launched the National Industrial Policy Framework (DTI 2007), which was followed in 2010 by the Trade Policy and Strategy Framework (TPSF) (DTI 2010). These policy documents articulate a critical assessment of the impact of tariff liberalization on the South African economy and outlined a new approach towards tariff determinations that were to be ‘conducted on a case-​by-​case basis, taking into account the specific circumstances of the sector involved’ (DTI 2010: 3). Any tariff reductions were to be concentrated in upstream intermediate input industries, while protection on downstream industries was to be largely preserved. In addition, the TPSF outlined a narrower and more ‘strategic’ objective with respect to trade and investment agreements. Less emphasis was to be placed on negotiating comprehensive free trade agreements, with the focus shifting to partial scope agreements that cover a narrower range of products. Trade with the region was also to be emphasized. Finally, a more defensive stance towards multilateral tariff liberalization under the World Trade Organization was adopted. An important strength of the new approach towards tariff policy is that it emphasizes greater coherence between the use of tariff instruments and industrial policies. While tariffs provide protection to domestic industries, complementary targeted industrial policies are often required to alleviate the fundamental supply constraints giving rise to the firm’s competitiveness problems. Nevertheless, South Africa’s current trade policy faces several severe limitations. The policy is heavily focused on domestic concerns and has the danger of placing South Africa at a disadvantage as its exporters seek access to the growing emerging economies (Edwards and Lawrence 2012). On the multilateral liberalization front, two of South Africa’s explicit objectives are to enhance market access to developed countries, as well as to eliminate industrial countries’ subsidies and support to agriculture. Yet, tariff barriers imposed by emerging economies on South African exports far exceed those by developed economies where in several cases South Africa already has preferential access (e.g. the EU, the United States under the African Growth and Opportunity Act, and Japan through the Generalized System of Preferences). The shift in focus towards partial scope trade agreements, as reflected in the agreement with the Southern Common Market (MERCOSUR) (from 2016) and a proposed agreement with India, also do not give South Africa the same market access benefits that would be achieved under a comprehensive free trade area (Edwards and Lawrence 2012). In contrast, the prospects for improved market access into the region are stronger. SACU members commenced negotiating the establishment of a free trade area between the Common Market for Eastern and Southern Africa (COMESA), the East African Community (EAC) and the SADC. This process has been surpassed by negotiations around establishing the more extensive African Continental Free Trade Area (AfCFTA) that commenced from 2021. Subject to the finalization of tariff phase-downs, sensitivity lists, and rules of origin requirements, the AfCFTA agreement has the prospect of providing South African firms with improved market access into the large regional economies of Nigeria, Egypt, and Kenya.

450   Lawrence Edwards A second concern with South Africa’s trade policy is that the commitment to deal with tariffs on a case-​by-​case basis is unlikely to resolve the complexities and inefficiencies in the current tariff structure that reflect the outcomes of historical policies. For example, Edwards and Lawrence (2008b) find high levels of differentiation in tariff rates within narrow industry groupings and argue that the tariff structure is costly in terms of supporting employment, and, given relatively high tariffs on consumer goods, is regressive in its impact on income distribution. An approach that sets individual tariffs differentially is also prone to be captured by politically connected and economically powerful vested interest groups. See, for example, Chapter  17 in this volume, which discusses how corporate strategies of large lead firms dictate policy outcomes. Edwards and Lawrence (2008b) recommend a simpler tariff structure comprising a few tariff bands, with explicit processes and rules regarding exemptions for industrial policy purposes. A simpler tariff structure would also facilitate the conclusion of regional trade agreements, in particular customs unions that are the objective of the SADC agreement as well as the AfCFTA.7 The defensive approach by the Department of Trade, Industry and Competition towards tariff setting has been evident in active use of tariffs to protect domestic industries, including wearing apparel, food products (chicken, wheat, sugar), machinery (top-​ loaded washing machines), and iron and steel products (tubes, pipes, hollow profiles), amongst others. In several cases, the tariff increases are part of a package of incentives for, and commitments by retail outlets to source locally and producers to invest in productive capacity (e.g. the Retail-​Clothing, Textile, Footwear and Leather Master Plan) reflecting the sectoral approach to industrial policy. A shift towards a more protective stance is also reflected in the 2017 amendments to the Preferential Procurement Policy Framework Act, which require all organs of the state to purchase designated products locally. As of early 2020, twenty-​seven products with local content thresholds ranging from 30 per cent to 100 per cent have been designated for local procurement. While localization may have the capacity to raise demand for targeted domestic products, its limitation is that, as a programme, it is unable to deal with structural impediments to supply that cut across all industries, or serve as an instrument to enhance competitive exports.

21.4  Trade Performance from 1994 Post-​1994 trends in South African trade flows have been studied extensively using aggregate and industry-​level data. Increasingly, the focus has shifted towards analysing product-​and firm-​level data as this has become available. A  common

7  For

analysis emphasizing the importance of sub sector specific trade and industrial policies, see Roberts (2000) and DTI (2010).

South Africa’s International Trade    451 theme in much of the literature has been the extent to which exports have grown and diversified (Roberts 2000; Hausmann and Klinger 2008; Purfield et al. 2014; Bhorat et al. 2019). The aggregate trends in trade flows post-​1994 are shown in Figure 21.1. Export and import volumes rose strongly relative to GDP from 1994, but faltered during the 1997 Asian and 1998 Russian financial crises, and then again in the early 2000s. From 2004 to 2007, exports grew very strongly, driven by the commodity boom, but imports rose even more sharply leading to a current account deficit of 5.5 per cent of GDP in 2008. Towards the end of 2008, trade volumes collapsed in response to the global financial crisis. While import growth recovered relatively quickly, growth in export volumes has been tepid, even lagging the weak GDP growth. By 2019, export volumes were only 3 per cent higher than they were in 2008 and comprised a lower share of real GDP (29.5 per cent vs 32.5 per cent), whereas import volumes were 17 per cent higher, leading to continued pressure on the current account. Behind these aggregate trends lie important changes in the geographic, industry, product, and firm composition of trade flows. These aspects are discussed in the following sub ​sections.

21.4.1 Diversification of Exports—​Industry, Product, and Firm Dynamics Table 21.1 shows evidence of broad changes in the industry composition of South Africa’s export bundle from 1990. The importance of gold continued to decline with manufacturing emerging as the dominant source of exports, reaching 64 per cent of goods exports in 2019. The industry composition of manufacturing exports also saw some diversification, with the share of medium-​technology products rising from 29 per cent in 1990 to 46.3 per cent in 2019. The principal determinants of this increase are exports of motor vehicles and machinery and equipment. Motor vehicles and other transport equipment alone made up 28.5 per cent of the total value of manufacturing exports in 2019. Much of this growth can be attributed to the export and investment incentives provided under the Motor Industry Development Programme (1995–​2012) and the Automotive Production Development Programme (APDP) from 2012 (Black 2011; Madani and Mas-​Guix 2011). Export growth, however, was not only concentrated in the vehicle industry. Using data for manufacturing at the 3-​digit Standard Industrial Classification (SIC) level reveals that exports rose as a share of sales in almost all industries, reflecting the broad-​ based re-​orientation of production towards exports.8 These broad-​based outcomes, in 8  Exports as a share of total sales rose in thirty-​six of the forty-​four industries from 1995 to 2010, with an average increase of 7.6 percentage points. Very large increases (over 20 percentage points) were experienced in machinery, electrical equipment, and motor vehicle industries.

452   Lawrence Edwards part, reflect the outcome of reductions in the anti-​export bias arising from lower import tariffs. Edwards and Lawrence (2008a), for example, calculate that reductions in the implicit export tax from tariff liberalization from 1988 to 2003 was equivalent to an improvement in export profitability of 34 per cent for commodity manufacturing and a much higher 60 per cent for non-​commodity manufacturing. Further, their estimates reveal that a significant portion of manufacturing export growth and diversification into non-​commodity products from 1990 to 2002 can be attributed to reductions in the anti-​ export bias from trade liberalization.9 Drilling down to the firm level also reveals widespread participation in exporting, irrespective of the industry classification (Matthee et al. 2018), as well as a close association between exporting and importing (Edwards et al. 2018, 2020). Manufacturing exporters that import are also more productive, are less likely to exit from exporting, have higher average export values (R14.4 million vs. R2.2 million), export more products per destination (9.4 vs. 7.6), and to more destinations per product (2 vs. 1.4) compared to exporters that do not import (Edwards et al. 2018). The implication is that improved access to imported intermediate inputs and the backward integration of exporters into global intermediate input supply chains that was made possible by tariff liberalization have played a prominent role in driving South Africa’s post-​1994 export performance. For a further disaggregated perspective, Table 21.2 presents several measures of concentration, competitiveness, and export sophistication based on disaggregated product-​level export data (the 3-​or 4-​digit level of the Standard International Trade Classification or the 6-​digit level of the Harmonized System [HS]) for South Africa. These indicators provide further evidence of changes in South Africa’s export bundle. South Africa exports a wide range of products with an export presence in over 90 per cent of possible 6-​digit HS lines in all years. The range of countries it exports to high and increased from 194 in 1995 to 218 in 2019. The number of product varieties (defined as 3-​digit SITC product-​destination combinations) rose from 11,633 to 18,345 over the period, reflecting a diversification of South Africa’s export bundle. More than half of this increase took place from 1995 to 2000 when tariffs fell most strongly. According to the Revealed Comparative Advantage measures, the number of 4-​digit SITC products that South Africa is revealed to be competitive in, rose from 179 to 253 from 1995 to 2000, but this number then declined to 177 by 2019. The concentration of exports, however, has risen, as is reflected by the rising share of total exports accounted for by the top five destinations (36.9 per cent to 43.2 per cent) 9  Other studies that find a statistically significant positive relationship between aggregate or industry-​ level exports and tariff reductions include Tsikata (1999), Alves and Edwards (2006), and Edwards and Lawrence (2008a). These findings contrast with the argument by Trevor Bell who, in several articles, argued that the performance and composition of exports was primarily an outcome of macroeconomic forces, including economic growth and changes in the real exchange rate associated with international commodity prices (particularly gold) rather than a lack of an ‘export culture’, or the ‘anti-​export bias’ created by protection (Bell 1993, 1997; Bell et  al. 2002). The results also contrast with the argument made by the DTI (2010: xiii) that trade liberalization reinforced export specialization in resource-​based products that reflect South Africa’s ‘static comparative advantage’.

South Africa’s International Trade    453 and the top five products (at 3-​digit level, excluding gold) (32.3 per cent to 43.5 per cent) over the period 1995 to 2019 (Table 21.2). Concentration levels are even higher when using trade transaction data. Purfield et al. (2014: 21) use South African customs and excise transaction data for 20,000 firms from 2001 to 2012 and find that the top 5 per cent of South Africa’s exporting firms account for more than 90 per cent of its exports—​a comparatively high share compared to many other emerging economies. Relatively low entry rates compared to many other emerging economies (Purfield et al. 2014; Edwards et al. 2018), as well as a large degree of churn amongst smaller exporters (Matthee et al. 2018),  imply that new entrants fail to make a substantive impact on export concentration. As a consequence, the contribution of new exporters and products to export growth remains low (and falling) (Purfield et al. 2014; Matthee et al. 2016a). The firm-​level data therefore point to a lack of dynamics in the firm composition of South African exports that impedes the diversification of South Africa’s export bundle. Table 21.2: Indicators of export concentration and complexity 1995

2000

2010

2019

Number 6-​digit HS products exported (as % total possible lines)

4,696 (93%)

4,690 (94%)

4,551 (94%)

4,451 (95%)

Number destinations

194

208

212

218

Number 3-​digit STIC product-​destinations

11,633

15,310

17,616

18,345

Share top 5 destinations

36.89

39.48

39.19

43.22

Share top 5 products, excl. gold

32.3

34.8

41.2

43.5

Number of products with RCA

179

253

195

177

Economic Complexity Index

0.31

0.27

0.13

-​0.02

Complexity Outlook Index

1.24

2.25

1.60

1.37

World market share: merchandise exports (%)

0.58

0.50

0.60

0.48

World market share: manufacturing exports (%)

0.36

0.34

0.47

0.38

Measures of concentration

Measures of competitiveness and complexity

Source: Own calculations using data from Un Comtrade via World Integrated Trade Solution, Harvard Growth Lab’s ATLAS of Economic Complexity (https://​atlas.cid.harvard.edu/​) and the World Trade Organization (https://​data.wto.org/​). Notes: RCA denotes Revealed Comparative Advantage and is based on 4-​digit SITC Rev. 2 data. A value greater than 1 implies that the share of the product in South African exports exceeds the share of that product in world exports. The Complexity Outlook Index (COI) is a measure of how many complex products are near a country’s current set of capabilities. A high COI value reflects an abundance of nearby, complex products that rely on similar capabilities as those present in current production. The Economic Complexity Index is a measure of the knowledge in a society as expressed in the products it makes. The economic complexity of a country is calculated based on the diversity of exports a country produces and their ubiquity, or the number of the countries able to produce them (and those countries’ complexity). The higher the index, the greater the economic complexity.

454   Lawrence Edwards South Africa also appears to differ from other emerging economies where the concentration of exports is truncated by too few large exporters (Fernandes et al. 2016). In contrast, the relatively high export concentration amongst firms in South Africa points to the presence of a ‘missing’ middle. More firm-​level research is required to understand the dynamics behind this, but this feature of trade is consistent with the arguments presented in Chapters 17 and 25 of this volume that market power and concentration in South Africa have inhibited the emergence of competitive small and mid-​sized firms. Despite some diversification of the export bundle, resource-​based products remain a salient feature of the country’s export profile, accounting for 43.2 per cent of manufacturing exports in 2019 (Table 21.1). If mining and agricultural sector exports are included, their combined value makes up 63 per cent of the total value of South Africa exports of merchandise goods. Given South Africa’s abundance in natural resources, resource-​based products are expected to be a prominent feature of South Africa’s export bundle. Nevertheless, several factors worked to amplify the contribution of resource-​ based products to South African exports post-​1994. Industrial policy, particularly in the 1990s, continued to incentivize large-​scale capital-​intensive projects in minerals-​intensive industries such as the non-​ferrous metals and basic iron and steel sub s​ ectors. These incentives took the form of an accelerated depreciation allowance (e.g. to Columbus Stainless Steel and the Saldanha Steel plant), tax relief under the Strategic Industrial Projects programme, artificially low electricity prices by Eskom, and investment support for large-​scale mineral beneficiation projects by the state-​owned Industrial Development Corporation. For fuller discussions on these incentives and industrial policy in South Africa, see ­Chapters 17, 18, and 24 in this volume. As will be discussed in more detail later, the rapid growth of China and its demand for resources in the 2000s reinforced this trend. The resultant commodity boom re-​orientated South African exports towards resource-​based and primary products, while Chinese exports crowded out South African manufacturing exports in third markets (Edwards and Jenkins 2014, 2015b). Chinese competition also displaced domestic manufacturing production, undermining the supply base (Edwards and Jenkins 2015b). This initiated a particular form of deindustrialization—​a shift of the economy into extractive industries, rather than into services (Imbs 2013).10 A further contributing factor is market power by entrenched lead firms in the upstream petrochemical and metals industries. This has resulted in above-​competitive price levels for intermediate inputs, thus undermining the competitiveness of downstream industries (see ­Chapter 25 in this volume). One important caveat to the above analysis is services trade that is not shown in Table 21.1. According to South African Reserve Bank data, the growth in exports of services exceeded that of goods from the mid-​1980s, with their share in the value of total exports rising from 9 per cent in 1985 to 14 per cent in 2019 (with a peak of 18 per cent in 2003). 10  Based on South African Reserve Bank data, the share of mining in nominal gross value added rose from 7.4 per cent in 2000 to 9.2 per cent in 2008. The share of mining in real GDP, however, fell from 13 per cent to 9.4 per cent over this period, reflecting a very weak real output response in mining to the commodity boom.

South Africa’s International Trade    455 The bulk of these services exports comprise travel and transport services (over 70 per cent share), but with the rise in trade through global value chains, the share of business and financial services has risen in importance. The contribution of services to exports is even more important if indirect linkages are taken into account. In value-added terms, services account for close to 40 per cent of South Africa’s exports, indicating the intensive dependence of exporters on services inputs.11

21.4.2 Geographical Diversification A further avenue for diversification is through the expansion of exports into new markets. At the aggregate level, there have been substantial changes in the geographical composition of South African exports. In 1995, the United Kingdom was the primary export destination accounting for 11 per cent of South Africa’s export value (excluding gold, platinum, and unspecified products), but by 2019 it had fallen out of the top five destinations. China, which ranked as the eighteenth most important destination in 1995, emerged as South Africa’s top export destination from 2009/​10, accounting for 15 per cent of the country’s exports of goods in 2019. For manufacturing goods, diversification into Africa, driven initially by the ending of sanctions and from 2001 by the reductions in tariffs under the SADC FTA together with relatively strong economic growth in the continent, has been a major contributor to export growth. The share of SADC countries (excluding exports to other SACU members) in total South African exports rose from 7.1 per cent in 1994 to 13 per cent in 2019. If South African exports to other SACU members are included, the SADC share is a high 26 per cent in 2019. These exports are ‘desirable’ from an industrialization perspective as they are strongly oriented towards manufactured goods. The implication is that, including SACU exports, Africa accounts for around half of South Africa’s non-​mineral exports (Purfield et al. 2014). In addition to being a source of demand for manufactured goods, this trade has been accompanied by rapid growth in services alongside significant foreign direct investment (FDI) from South African countries in sectors such as retail, banking, insurance, transport, and business support services (Arndt and Roberts 2018). The establishment of the South African retail chains in the region has been particularly effective as a conduit for South African goods to enter into the African markets. For a fuller discussion on South African retail chains in the region, see ­Chapter 19 in this volume. The product-​and firm-​level growth dynamics behind South African export trade to the region also differs from those of the rest of the world. Using South African Revenue Services (SARS) administrative data for, on average, 29,000 firms (5,700 exporters) per year from 2010 to 2013, Matthee et al. (2018) find that firms exporting to SADC and other African countries export a smaller proportion of their output, export less sophisticated

11  Own

oecd.org.

calculations drawing on the OECD Trade in Value Added data obtained from https://​stats.

456   Lawrence Edwards products, are more capital‐intensive, pay lower wages, and present no productivity premium between them and domestic‐orientated firms. Using transaction data over the earlier 2001 to 2012 period, Purfield et al. (2014) find that exports to sub-​Saharan Africa are less concentrated (top 1 per cent account for 46 per cent of export value, vs. 80–​ 85 per cent for BRICS and EU), and the average export spell of products is shorter, the value of new exporters lower, and growth of surviving exporters also lower. However, the survival rates of firms exporting to Africa is higher than other regions. The African market will continue to play an important role in driving the growth and diversification of South African exports, considering its rising population and the proposed implementation of the African Continental Free Trade Agreement. The expectations are high. As President Ramaphosa declared in his acceptance statement on assuming the chair of the African Union for 2020 on 9 February 2020, AfCFTA is expected to ‘reignite industrialisation and pave the way for Africa’s integration into the global economy as a player of considerable scale’.12 However, the nature of the products exported and ad hoc use of the African market by South African firm—​exporting when opportunities come rather than seeking them out (Purfield et al. 2014)—​suggests more moderate expectations may be required of the potential of trade with Africa to serve as a springboard for South African industrialization and entry into highly competitive global markets.

21.4.3 Structural Transformation at the Product Level Other research on exports has focused on the implications of structural shifts in the product composition of South African exports for growth and factor usage (Hausmann and Klinger 2008; Purfield et al. 2014; Bhorat et al. 2019). For example, Purfield et al. (2014) use product-​level indicators of revealed factor intensity to show that South Africa’s exports are concentrated in products with human capital and physical capital intensities beyond the country’s endowments (see also Matthee et al. 2016b). This finding corresponds with that of Alleyne and Subramanian (2001) who use industry data during the 1990s to show that South Africa is paradoxically revealed through trade to be relatively capital abundant and a net exporter of capital-​intensive goods. Similarly, at the firm level, compared to domestic-​oriented firms, manufacturing exporters are larger, more productive, pay higher wages, and are more capital-​and skill-​intensive (Matthee et al. 2016a, 2018: 104; Edwards et al. 2018). The product and firm composition behind South Africa’s exports thus gives rise to a mismatch between the intensity of factors demanded in exports and the endowments available. The implication of this finding is that continued export growth under the current structure is unlikely to absorb less-​skilled labour, which accounts for much of South Africa’s unemployed. Hausman and Klinger (2008) and Bhorat et al. (2019) present an alternative approach to analysing South Africa’s structural transformation and path dependence by locating 12 

http://​w ww.thepresidency.gov.za/​speeches/​acceptance-​statement-​president-​c yril-​ramaphosa-​ assuming-​chair-​african-​union-​2020.

South Africa’s International Trade    457 its exports within a product space representation of the relatedness of products developed by Hidalgo et al. (2007). Within this network are core areas where products are proximate, plentiful, and easy for economies to transition into, and peripheral areas where products (typically primary products) are more distant with fewer connections. Rapid development is associated with transition into the dense part of the network (Hidalgo et al. 2009). Bhorat et al. (2019) use graphical visualizations of the product space to illustrate key features of South Africa’s export structure. Products where South Africa has a comparative advantage are rooted in commodities (platinum, iron ore and concentrates, coal, gold and diamonds), horticulture (citrus, apples, potatoes, sugars) and agro-​processing (juices, sugar products, edible products, eggs, jams, etc.) that are largely in the periphery of the product space and have weak links into the dense part of the network. Only a few products, such as passenger vehicles, filtering and purifying machinery, construction and mining machinery, and pumps for liquids, are located within the dense part of the network. Using similar diagrams, Hausmann and Klinger (2008) illustrate that there was some re-​orientation of the export basket towards the centre from 1995 following trade liberalization, but the transition was weak relative to countries such as Malaysia that developed capabilities in the export of electronics-​related goods from the 1980s. Bhorat et al. (2019) extend this analysis to compare 1995 with 2015 and also find evidence of weak structural transformation. They argue that South Africa actually experienced declines in the complexity of its export bundle (Economic Complexity Index) and proximity to unexploited product diversification opportunities (Opportunity Value Index) over this period, but a closer look at these measures presented in Table 21.2 shows that the decline takes place after 2000. The period 1995 to 2000, is characterized by rising or stable indicators of Economic Complexity and Opportunity Value. Nevertheless, the overall implication drawn from the analyses is that South Africa’s endowments in natural resources, combined with its location on the periphery of the product space, have impeded structural transformation. The post-​1994 period thus presents a mixed picture with respect to growth and diversification of South Africa’s export bundle. The overall picture is one in which there was relatively strong growth and diversification of exports from 1995 to 2000, with stagnation or reversal in some of these gains subsequently. The following section looks at the implications of these trends for output and employment in manufacturing.

21.5  Trade, Labour, and Deindustrialization The liberalization of the economy has had far-​reaching implications for production, productivity, prices, trade, and employment, amongst others, and has initiated

458   Lawrence Edwards considerable debate on the merits of the policy. This is to be anticipated as the impact of trade liberalization does not fall equally across all in society. A comprehensive evaluation of the impacts of liberalization on the economy is beyond the scope of this chapter. Rather, this section updates the prior reviews by Edwards (2006) and Cassim et  al. (2009) and presents new data to analyse the relationship between trade, production, and employment. While the post-​1994 period saw increases in trade volumes and export-​orientation, these trends corresponded with declines in manufacturing employment both in absolute levels and as shares of total employment. Depending on the data used, total employment in manufacturing fell from 1.43 million in 1994 to 1.3 million in 2000, and then further to 1.2 million in 2019.13 Manufacturing’s share of non-​agricultural employment also fell from between 17 and 19 per cent in 1994 to under 12 per cent by 2019. Some of the decline in manufacturing employment can be attributed to outsourcing-​ type reallocation and reclassification of services such as cleaning and security (Tregenna 2010). Nevertheless, the relatively poor manufacturing employment growth raises concerns that trade liberalization accelerated the process of deindustrialization that had commenced from the early 1980s. To study the relationship between trade, production, and employment over the period 1992 to 2019, Table 21.3 presents a Chenery-​style decomposition of output growth and employment growth in manufacturing into Final Demand, Exports, Import Penetration, and Technology.14 The decomposition extends the work of Edwards and Jenkins (2015a) and therefore also isolates how China affects output and employment. While these decompositions face several limitations (Edwards and Jenkins 2015a), they are informative in providing a broad overview of structural changes in demand, trade, and technology in the economy and how these relate to output and employment growth. Several key features regarding South Africa’s post-​1994 growth in manufacturing can be identified from Table 21.3. Output growth, which initially grew by a high 26.5 per cent from 1992 to 2001, declined sharply over the following two periods 2001–​10 and 2010–​ 19. The dominant driver of this trend is domestic demand which accounts for most of the aggregate growth in the sector. Growth in exports also contributed significantly towards raising output, particularly in the first period 1992–​2001 where it raised output growth by 17 per cent. However, export growth only raised employment by 15.9 per cent over the full period—​less than half its impact on output. This reflects the capital-​intensity of 13 The

data for 1994 and 2000 are obtained from Statistics South Africa Survey of Employment and Earnings (P0271), while the data for 2019 are obtained from the Quarterly Employment Statistics (P0277). 14  The decomposition method is described as follows: gross output (X) is expressed as X = dD + E, where d is the ratio of domestically produced goods to total demand, D is total demand (inclusive of imports) and E is exports. This relationship can be decomposed into changes in demand (ΔD), export expansion (ΔE) and import penetration (ΔdD) as follows: ΔX = dΔD+ΔdD+ΔE. This method can be extended to study change in employment by decomposing total employment (N) into nΔX+ΔnX, where n is employment per unit output X. The final term, ΔnX, is an indicator of how technological change affects employment.

South Africa’s International Trade    459 Table 21.3: Contribution to output and employment growth in manufacturing as share initial total output or employment, 1992–​2019 1992–​ 2001

2001–​ 2010–​ 1992–​2019 10 19 Total

Capital-​ Ultra-​L Medium intensive L-​intensive intensive

Decomposition of output Growth of Domestic Demand

19.8

17.1

9.4

54.6

32.7

59.7

61.1

Increased exports

17.0

3.4

6.8

30.8

6.9

16.9

47.0

  (of which exports to China)

0.3

0.6

0.3

1.4

0.2

0.6

2.3

Increased import penetration

–​10.2

–​9.4

–​6.9

–​31.9

–​25.7

–​29.0

–​35.7

–​1.2

–​5.5

–​4.4

–​14.3

–​16.1

–​18.8

–​11.5

6.8

–​6.0

–​0.2

–​1.1

–​18.8

–​12.1

11.3

  (of which imports from China) Net trade % Change in Output

26.5

11.1

9.2

53.5

13.9

47.5

72.4

142.1

74.9

69.6

286.6

16.1

64.1

206.4

Growth of Domestic Demand

12.8

17.9

7.5

34.3

23.1

43.7

41.5

Increased exports

Change Output (Rmill) Decomposition of employment

13.0

0.8

2.7

15.9

8.2

13.4

28.3

  (of which exports to China)

0.2

0.3

0.3

0.7

0.2

0.4

1.7

Increased import penetration

–​8.3

–​9.0

–​8.7

–​23.2

–​20.8

–​23.8

–​25.9

  (of which imports from China)

–​1.5

–​7.0

–​5.3

–​11.9

–​12.9

–​11.0

–​11.3

4.6

–​8.2

–​5.9

–​7.3

–​12.6

–​10.4

2.4

–​31.4

–​14.5

2.5

–​41.8

–​45.0

–​36.9

–​41.5

Net trade Productivity % Change in Employment

–​13.9

–​4.8

4.1

–​14.7

–​34.5

–​3.6

2.5

Change Employment (‘000)

–​197.6

–​58.6

48.0

–​208.3

–​205.8

–​13.4

11.0

Source: Own calculations extending Jenkins and Edwards (2015a) using trade data obtained from UN Comtrade via World Integrated Trade Systems, and employment and output data obtained from Statistics South Africa. Notes: Based on forty-​four manufacturing industries at the 3-​digit level of the SIC. Ultra-​labour-​intensive and capital-​intensive groups comprises of the top and bottom third, respectively, of industries according to average sales per worker from 1992–​2019. Medium-​labour-​intensive makes up the remaining industries. Output and trade data are deflated using industry producer price indices obtained from Statistics South Africa. The calculations assume common deflators for output and trade values within each industry.

South Africa’s export bundle as well as the much stronger contribution of exports from capital-​intensive industries to growth in output (47 per cent) and employment (28.3 per cent) compared to exports from ultra-​labour-​intensive industries (6.9 per cent and 8.2 per cent, respectively). The contribution of exports to output growth over the full period, however, was completely offset by rising import penetration that resulted in a small negative 1.1

460   Lawrence Edwards per cent net trade impact. Looking within manufacturing, net trade reduced output and employment growth in the ultra-​labour-​intensive and medium-​labour-​intensive industries but contributed positively toward growth in capital-​intensive industries. This structural shift in net trade towards capital-​intensive and skill-​intensive sectors is also found in the factor-​content analyses of Bell and Cattaneo (1997) and Edwards (2001a). Looking across the sub-​periods, net trade had a positive impact on output and employment growth from 1992 to 2001 when tariff liberalization was at its most intense, as is also found by Edwards (2001a, 2001b), Jenkins (2008), Dunne and Edwards (2007), and Edwards and Jenkins (2015a). The contribution of net trade only turned negative from 2001 as import penetration continued to rise very strongly and export growth slowed. China is a central factor behind this change. Following its entry into the World Trade Organization in 2001 and its rapid economic growth, commodity prices boomed, and Chinese exports grew sharply. This had direct and indirect effects on the South African economy. Indirectly, the commodity price boom improved South Africa’s terms of trade, initiating a rise in commodity exports and improvements in South Africa’s growth that stimulated demand for manufactured imports. The real effective exchange rate appreciated, depressing growth in non-​commodity manufactured exports that declined as a share of trade (see Table 21.1). More directly, imports from China rose rapidly and by 2009 China had become South Africa’s most important trading partner, surpassing the United States as a destination for exports, and Germany as a source of imports. Although increased imports from China partly replaced imports from other countries, Edwards and Jenkins (2015a) find that most of the increase in Chinese penetration of the market was at the expense of local production. The decompositions in Table 21.3 show that import penetration by Chinese goods reduced output (employment) by 5.5 per cent (7 per cent) from 2001 to 2010 and by 7 per cent (5.3 per cent) from 2010 to 2019, with the strongest impact in the ultra-​labour-​intensive sectors. In contrast to the impact of imports, exports of manufactures to China did not add significantly to industrial growth. The net export effect of exports to China may actually be lower than presented in Table 21.3, as Edwards and Jenkins (2014, 2015b) estimate that Chinese competition crowded out South African exports, particularly in African markets, with the strongest effects in medium-​and low-​ technology products. As found in the other similar decomposition studies (Edwards 2001a; Jenkins 2008; Dunne and Edwards 2007; Edwards and Jenkins 2015a), the primary source of employment losses are productivity improvements within industries that lowered manufacturing employment by 41.8 per cent (or half a million jobs) from 1992 to 2019. However, some of this productivity improvement is itself explained by international trade, as is found at the aggregate level by Jonsson and Subramanian (2000) and Fedderke (2002). Productivity improvements associated with international trade are also found at the firm level. Using SARS administered firm data from 2010 to 2013, Matthee et  al. (2018) estimate total factor productivity premiums of 5 per cent for exporters compared to firms only selling to the home market. This rises to 8–​10 per cent

South Africa’s International Trade    461 for firms that export outside of Africa and is also higher for multi-​destination exporters. Edwards et al. (2020) also use the SARS administered data and show how the use of imports by manufacturing firms raises productivity through access to a wider range of inputs and technology transfer. These results suggest that compositional shifts towards exporters and increased use of imported intermediate inputs in response to liberalization or competition from China account for some of the job losses through productivity growth shown in Table 21.3. Trade may also have contributed towards the persistence in wage inequality that has been a dominant contributor to the rising Gini coefficient in South Africa (Wittenberg 2017) (see also C ­ hapter 9 in this volume, which provides a fuller discussion on inequality in South Africa). South African manufacturing exporters pay higher wages and employ relatively skilled labour compared to non-​exporters (Matthee et al. 2018). The within-​ firm dispersion of wages is also higher amongst exporters (Matthee et al. 2017). Growth in manufacturing exports will thus have raised the relative demand for skilled and high wage workers, placing additional pressure on wage inequality. One limitation of the above research is that it ignores potentially important regional variations in the impact of trade reform. Erten et al. (2019) and Lepelle (2020) respond to this shortcoming by focusing on the local labour market impacts of trade liberalization. Using a panel of cross-​sectional household data from 1994 to 2004, Erten et al. (2019) find that workers employed in districts facing larger tariff reductions experienced a significant decline in employment, driven primarily by a decline in manufacturing sector employment. The displaced workers tend to exit the labour force entirely and access government transfers, rather than transition into other sectors or migrate to less affected regions. These effects were stronger in regions with relatively high union and unemployment rates. Lepelle (2020), using more aggregated regional data over a longer period, from 1996 to 2011, finds similar results, but shows that declines in employment were concentrated amongst females. These papers are insightful in that they highlight how labour market rigidities to the relocation of workers across industries or locations, combined with differences in industry composition across regions, result in distinct local labour market outcomes from trade liberalization. There is much scope for further research in this area, including a study of how growth in exports impacted on local-​level employment.

21.6 Conclusion South Africa faces an export predicament. The country has experienced a long-​run decline in its export performance relative to its peers. Its macro-​competitiveness has also diminished, implying that an acceleration in economic growth will not be possible without substantially improving export growth. This chapter analyses South Africa’s participation in international trade as a way to better understand the dynamics behind South Africa’s economic performance.

462   Lawrence Edwards A key constraint to South Africa’s growth potential that was already apparent from the 1970s is its failure to transition out of gold exports into an export-​led growth path driven by manufacturing or an alternative commodity. The chapter highlights several early reasons that underpin this failure—​commodity dependence and booms, structural constraints and isolation from international markets through the imposition of sanctions, and domestic policies that constrained access to imports. The 1990s saw a dramatic opening up of the economy through trade liberalization, and a rise in both exports and imports as a share of output and consumption. However, this did not lead to a sustained transition of the economy onto an export-​led growth path, as was envisaged under the GEAR macroeconomic framework. While tariff liberalization reduced the anti-​export bias and stimulated export growth and diversification, these impacts were not sustained. An active industrial policy, implemented from 2007, also failed to make a significant mark on exports outside of motor vehicles. Trade liberalization also exposed South African manufacturing firms to increased competition, leading to employment losses, particularly in labour-​intensive industries, such as wearing apparel. The continued vulnerability of the South African manufacturing industry, which had not yet transitioned onto a competitive export-​oriented growth path, was further exposed by China’s export-​led boom from 2001. Although not discussed in detail in the chapter, South Africa appears to have fallen into a lower-​level growth equilibrium following the financial crisis in 2007/​8. Export and output growth in manufacturing and the rest of the economy has been very weak. In contrast, imports have continued to rise, placing continued pressure on South Africa’s balance of payments. The decline in commodity prices after 2008 and weaker global growth in trade volumes explain some of this performance. A slow recovery in global trade following the COVID-​19 pandemic will place further external demand pressures on South African exports. However, deeper and more structural supply factors are the primary constraints to an export growth recovery. The lack of supply dynamism is reflected in the firm export data that reveal low levels of entry, declining product ranges, high and rising levels of concentration (Purfield et al. 2014), and a diminished responsiveness of exports to the exchange rate (Hlatshwayo and Saxegaard 2016; Edwards and Hlatshwayo 2020). Adverse supply conditions include high trade costs linked to freight, transport, and port services provided by state-​owned enterprises (Purfield et al. 2014); repetitive electricity shortages; policy uncertainty (Hlatshwayo and Saxegaard 2016); and lack of competitiveness in the domestic market (Purfield et al. 2014), amongst others. As manufacturing production has become more skill-​intensive, globally and locally, the education system has struggled to increase the supply of more skilled labour, a constraint that was also prevalent during apartheid. Without resolving these supply constraints, it is unlikely that export growth will improve substantially. The AfCFTA presents an exciting opportunity to expand and diversify South African manufacturing exports. With a growing large population, the continent presents an important source of future demand for South African manufactured goods. However, South Africa’s highly unbalanced trade with Africa, given its low imports from the

South Africa’s International Trade    463 region, poses a threat to the successful implementation of the trade agreement. Looking at other sectors, services trade, including maintenance and support services to the region, business process outsourcing (e.g. call centres), and tourism are identified as key opportunities to expand exports (Arndt and Roberts 2018). Realizing these outcomes will require active intervention by the state to improve domestic market conditions and conclude international institutional agreements (e.g. trade, investment, and services agreements) that govern trade.

References Alleyne, Trevor, and Arvind Subramanian. 2001. ‘What does South Africa’s pattern of trade say about its labor market?’. IMF Working Paper no. 01/​148, Washington, DC. Anand, Rahul, Roberto Perrelli, and Boyang Zhang. 2016. ‘South Africa’s exports performance: Any role for structural factors?’. IMF Working Paper no. 16/​24, Washington, DC. Arndt, Channing, and Simon Roberts. 2018. ‘Key issues in regional growth and integration in Southern Africa’, Development Southern Africa, 35(3): 297–​314. Bell, Trevor. 1993. ‘Should South Africa further liberalise its foreign trade?’, in Merle Lipton and Charles Simkins (eds) State and Market in Post-​ Apartheid South Africa. Johannesburg: Witwatersrand University Press. Bell, Trevor. 1997. ‘Trade policy’, in Jonathan Michie and Vishnu Padayachee (eds) The Political Economy of South Africa’s Transition. London: Dryden Press. Bell, Trevor, and Nicolette Cattaneo. 1997. ‘Foreign trade and employment in South African manufacturing industry’. International Labour Office, Geneva. Bell, Trevor, Greg Farrell, and Rashad Cassim. 2002. ‘Competitiveness, international trade and finance in a minerals-​rich economy’, in José María Fanelli and Rohinton Medhora (eds) Finance and Competitiveness in Developing Countries, Vol. 19: 181. London: Routledge. Belli, Pedro, Michael Finger, and Amparo Ballivian. 1993. ‘South Africa:  A review of trade policies’. World Bank Informal Discussion Papers on Aspects of the South African Economy, no. 4, World Bank, Southern Africa Department. Bhorat, Haroon, Arabo Ewinyu, Kezia Lilenstein, Christopher Rooney, François Steenkamp, and Amy Thornton. 2019. ‘Economic complexity and employment expansion: The case of South Africa’. Development Policy Research Unit, University of Cape Town. Black, Anthony. 2011. ‘Trade liberalization, technical change and firm-​level restructuring in the South African automotive component sector’, International Journal of Institutions and Economies, 3(2): 173–​202. Brenton, Paul, Frank Flatters, and Paul Kalenga. 2005. ‘Rules of origin and SADC: The case for change in the mid-​term review of the trade protocol’. World Bank Africa Region Working Paper Series no. 83, Washington, DC. Casale, Daniela, and Merle Holden. 2002. ‘Endogenous protection in a trade liberalizing economy: The case of South Africa’, Contemporary Economic Policy, 20(4): 479–​89. Cassim, Rashad. 2003. ‘The pace, nature and impact of trade policy in South Africa in the 1990s’, South African Journal of Economic History, 18(1–​2): 76–​95. Cassim, Rashad, Lawrence Edwards, and Dirk Van Seventer. 2009. ‘Trade policy since Democracy’, in Janine Aron, Brian Kahn, and Geeta Kingdon (eds) South African Economic Policy under Democracy. Oxford: Oxford University Press.

464   Lawrence Edwards Department of Trade and Industry. 2007. ‘A national industrial policy framework’. Department of Trade and Industry, Pretoria. Department of Trade and Industry. 2010. ‘South African trade policy and strategy framework’. Department of Trade and Industry, Pretoria. Department of Trade and Industry. 2016. ‘Integrated national export strategy (Export 2030)’. Department of Trade and Industry, Pretoria. Dunne, Paul and Lawrence Edwards. 2007. ‘Trade, enterprise production and employment’, Journal for Studies in Economics and Econometrics, 31(2): 49–​68. Edwards, Lawrence. 2001a. ‘Globalisation and the skills bias of occupational employment in South Africa’, South African Journal of Economics, 69(1): 40–​7 1. Edwards, Lawrence. 2001b. ‘Trade and the structure of South African production, 1984–​97’, Development Southern Africa, 18(4): 471–​91. Edwards, Lawrence. 2005. ‘Has South Africa liberalised its trade?’, South African Journal of Economics, 73(4): 754–​75. Edwards, Lawrence. 2006. ‘Trade liberalisation and labour demand in South Africa during the 1990s’, in Haroon Bhorat and Ravi Kanbur (eds) Poverty and Policy in the Post-​Apartheid South Africa. Pretoria: Human Science Research Council Press. Edwards, Lawrence and Phil Alves. 2006. ‘South Africa’s export performance: Determinants of export supply’, South African Journal of Economics, 74(3): 473–​500. Edwards, Lawrence, and Stephen Golub. 2004. ‘South Africa’s international cost competitiveness and exports in manufacturing’, World Development, 32(8): 1323–​39. Edwards, Lawrence, and Ayanda Hlatshwayo. 2020. ‘Exchange rates and firm export performance in South Africa’. SA-​TIED Working Paper no. 89, UNU-​WIDER, Helsinki. Edwards, Lawrence, and Rhys Jenkins. 2014. ‘The margins of export competition:  A new approach to evaluating the impact of China on South African exports to sub-​Saharan Africa’, Journal of Policy Modeling, 36: 132–​50. Edwards, Lawrence, and Rhys Jenkins. 2015a. ‘The impact of Chinese import penetration on the South African manufacturing sector’, The Journal of Development Studies, 51(4): 447–​63. Edwards, Lawrence, and Rhys Jenkins. 2015b, Is China “crowding out” South African exports of manufactures’, European Journal of Development Research, 27(5): 903–​20. Edwards, Lawrence, and Robert Lawrence. 2008a. ‘SACU tariff policies: Where should they go from here?’. Centre for International Development Working Paper no. 169, Kennedy School of Government, Harvard University. Edwards, Lawrence, and Robert Lawrence. 2008b. ‘South African trade policy matters: Trade performance and trade policy’. Economics of Transition, 16(4): 585–​608. Edwards, Lawrence, and Robert Lawrence. 2012. ‘A strategic view of South African trade policy in relation to the future global trading environment’, South African Journal of International Affairs, 19(3): 277–​98. Edwards, Lawrence, Marco Sanfilippo, and Asha Sundaram. 2018. ‘Importing and firm export performance: New evidence from South Africa’, South African Journal of Economics, 86: 79–​95. Edwards, Lawrence, Marco Sanfilippo, and Asha Sundaram. 2020. ‘Importing and productivity: An analysis of South African manufacturing firms’, Review of Industrial Organization, 57(2): 411–​32. Erten, Bilge, Jessica Leight, and Fiona Tregenna. 2019. ‘Trade liberalization and local labor market adjustment in South Africa’, Journal of International Economics, 118: 448–​67.

South Africa’s International Trade    465 Evenett, Simon J. 2002. ‘The impact of economic sanctions on South African exports’, Scottish Journal of Political Economy, 49(5): 557–​73. Fallon, Peter, and Luiz Pereira da Silva. 1994. ‘South Africa:  Economic performance and policies’. World Bank Discussion Paper no.  7.  The Southern Africa Department, World Bank, Washington, DC. Fedderke, Johannes. 2002. ‘Technology, human capital and growth: Evidence from a middle income country case study applying dynamic heterogeneous panel analysis’. Royal Economic Society Annual Conference 2002, Royal Economic Society. Fedderke, Johannes, and Prabhat Vaze. 2001. ‘The nature of South Africa’s trade patterns by economic sector, and the extent of trade liberalization during the course of the 1990s’, South African Journal of Economics, 69(3): 436–​73. Feenstra, Robert, Robert Lipsey, Haiyan Deng, Alyson Ma, and Hengyong Mo. 2005. ‘World trade flows: 1962–​2000’. NBER Working Paper no. 11040, Cambridge, MA. Feinstein, Charles H. 2005. An Economic History of South Africa: Conquest, Discrimination, and Development. Cambridge: Cambridge University Press. Fernandes, Ana, Caroline Freund, and Martha Pierola. 2016. ‘Export behavior, country size and stage of development:  Evidence from the Exporter Dynamics Database’, Journal of Development Economics, 119: 121–​37. Fine, Ben, and Zavareh Rustomjee. 1996. The Political Economy of South Africa: From Minerals-​ Energy Complex to Industrialisation. London: C. Hurst and Co Publishers Ltd. Hausmann, Ricardo, and Bailey Klinger. 2008. ‘South Africa’s export predicament’, Economics of Transition, 16(4): 609–​37. Hidalgo, César A., Ricardo Hausmann, and Partha Sarathi Dasgupta. 2009. ‘The building blocks of economic complexity’. Proceedings of the National Academy of Sciences of the United States of America, 106(26): 10570–​75. Hidalgo, César A., Bailey Klinger, Albert-​Laszlo Barabási, and Ricardo Hausmann. 2007. ‘The product space conditions the development of nations’, Science, 317(5837): 482–​7. Hirsch, Alan, and Sally Hines. 2005. Season of Hope: Economic Reform under Mandela and Mbeki. Ottawa: International Development Research Centre. Hlatshwayo, Sandile, and Magnus Saxegaard. 2016. ‘The consequences of policy uncertainty: Disconnects and dilutions in the South African real effective exchange rate–​export relationship’. International Monetary Fund, Washington, DC. Holden, Merle. 1990. ‘The growth of exports and manufacturing in South Africa from 1947 to 1987’, Development Southern Africa, 7(3): 363–​6. Holden, Merle. 1992. The Structure and Incidence of Protection in South Africa. Johannesburg: Leading Issues in South African Microeconomics, Southern Book Publishers. Imbs, Jean. 2013. ‘The premature deindustrialization of South Africa’, in Joseph Stiglitz, Justin Yifu Lin, and Ebrahim Patel (eds) The Industrial Policy Revolution II. London:  Palgrave Macmillan. Jenkins, Rhys. 2008. ‘Trade, technology and employment in South Africa’, The Journal of Development Studies, 44(1): 60–​79. Jonsson, Gunnar, and Arvind Subramanian. 2001. ‘Dynamic gains from trade: Evidence from South Africa’, International Monetary Fund Staff Papers, 48(1): 197–​224. Lall, Sanjaya. 2000. ‘The technological structure and performance of developing country manufactured exports, 1985–​98’, Oxford Development Studies, 28(3): 337–​69. Lepelle, Refilwe. 2002. ‘Gendered effects of tariff liberalisation in post-​apartheid South Africa: A spatial analysis’. PhD thesis, University of Cape Town, South Africa.

466   Lawrence Edwards Lipton, Merle. 1988. ‘Sanctions and South Africa:  The dynamics of economic isolation’. Economist Intelligence Unit, London. Madani, Dorsati, and Natalia Mas-​Guix. 2011. ‘The impact of export tax incentives on export performance:  Evidence from the automotive sector in South Africa’. World Bank, Washington, DC. Manby, Bronwen. 1992. ‘South Africa: The impact of sanctions’, Journal of International Affairs, 46(1): 193–​217. Marais, G. 1981. ‘Structural changes in manufacturing industry 1916 to 1975’, South African Journal of Economics, 49(1): 26–​45. Matthee, Marianne, Thomas Farole, Tasha Naughtin, and Neil Rankin. 2016a. ‘South African exporters and the global crisis: Intensive margin shock, extensive margin hangover’, South African Journal of Economics, 84(2): 183–​98. Matthee, Marianne, Ernst Idsardi, and Waldo Krugell. 2016b. ‘Can South Africa sustain and diversify its exports?’, South African Journal of Economic and Management Sciences, 19(2): 249–​63. Matthee, Marianne, Neil Rankin, and Carli Bezuidenhout. 2017. ‘Labour demand and the distribution of wages in South African manufacturing exporters’. WIDER Working Paper no. 2017/​11, Helsinki. Matthee, Marianne, Neil Rankin, Tasha Webb, and Carli Bezuidenhout. 2018. ‘Understanding manufactured exporters at the firm level:  New insights from using SARS administrative data’, South African Journal of Economics, 86: 96–​119. National Planning Commission. 2012. ‘National Development Plan 2030’. Pretoria, South Africa: Government Printers. Purfield, Catriona Mary, Thomas Farole, and Fernando Im. 2014. ‘South Africa Economic Update:  Focus on export competitiveness’. South Africa economic update issue no. 5. Washington, DC. Rangasamy, Logan, and Chris Harmse. 2003. ‘The extent of trade liberalisation in the 1990s: Revisited’, South African Journal of Economics, 71(4): 705–​28. Ratcliffe, Anne E. 1975. ‘Export policy in perspective’, South African Journal of Economics, 43(1): 45–​55. Republic of South Africa. 1996. ‘Growth, employment and redistribution: A macroeconomic strategy’. Pretoria: South African Government Printers. Roberts, Simon. 2000. ‘Understanding the effects of trade policy reform: The case of South Africa’, South African Journal of Economics, 68(4): 270–​81. Rodrik, Dani. 2016. ‘Premature deindustrialization’, Journal of Economic Growth, 2(1): 1–​33. Thirlwall, Antony. 1979. ‘The balance of payments constraint as an explanation of international growth rate differences’. Banca Nazionale Del Lavoro Quarterly Review, 32(128): 45–​53. Tregenna, Fiona. 2010. ‘How significant is the intersectoral outsourcing of employment in South Africa?’, Industrial and Corporate Change, 19(5): 1427–​57. Tregenna, Fiona. 2016. ‘Deindustrialization and premature deindustrialization’, in Erik Reinert, Jayati Ghosh, and Rainer Kattel (eds) Handbook of Alternative Theories of Economic Development. Cheltenham: Edward Elgar Publishing. Tsikata, Yvonne M. 1999. ‘Liberalization and trade performance in South Africa’. World Bank, Southern Africa Department. Wittenberg, Martin. 2017. ‘Wages and wage inequality in South Africa 1994–​2011:  Part  2–​ Inequality measurement and trends’, South African Journal of Economics, 85(2): 298–​318.

Chapter 22

Innovation a nd t echnol o gica l c ha ng e in Sou th A fri c a Erika Kraemer-​M bula and Rasigan Maharajh

22.1 Introduction The year 1994 saw the fall of the apartheid regime in South Africa and the beginning of a new democratic era. Since then, the country has embarked on major government reform, although the social structures built up during the stage of racial discrimination have proven to be difficult to eradicate. Science and technology (S&T) were central to the apartheid regime, manifest in the pursuit of big scientific missions and large-​scale technology projects that fulfilled various purposes—​including the symbolic portrayal of power and national superiority, as well as demonstration of self-​sufficiency by building capabilities in key technological areas. However, the repressive and segregationist rationale embedded in apartheid’s S&T system posed a huge obstacle to developing broad-​ based human capabilities essential to sustain an industrial and economic development agenda (Scerri 1998). Thus, the South African S&T system pre-​1994 has been described as weak, fragmented, and lacking coherence (IDRC 1993). South Africa inaugurated its new democratic era with an S&T system in disarray and incoherent governance structures. The current S&T system was severely impacted by exclusion, subjugation, oppression, and exploitation ranging across approximately three and a half centuries. This chapter explores the main achievements and remaining challenges in the contemporary South African STI system, more concretely focusing on the period between the two White Papers in 1996 and 2019. Moreover, it discusses the main shifts in policy emphasis

468    Erika Kraemer-Mbula and Rasigan Maharajh (intents) of these two policy/​institutional developments in connection to the STI system performance. This chapter argues that the dominant discourses in the innovation and technology development literature in South Africa have been, and largely remain narrowly constrained by the overarching neoclassical orientation of macroeconomic policy. The general neo-​liberal reforms that followed the original White Paper of 1996 enabled the maintenance of the STI system whilst seeking increased participation of the private sector premised upon the fallacy that improved business performance would trickle down into an improved quality of life for all (cf. RSA 2002). However, little evidence exists of either of these two objectives being achieved. South Africa’s adoption of the systems of innovation approach to development offered a progressive alternative to the globally ascendant market-​biased approaches and their export-​led growth prescriptions of the mid-​1990s. Without the wider congruence afforded by complimentary policy approaches, the full utility and value of the NSI was lost and each separate ministry and government department sought ‘siloed’ S&T outputs and outcomes, often failing to promote innovation impacts. The bureaucratic capture of the policy discourse, and its subsumption under the fiscal and monetary policy diktats of the National Treasury further entrenched this exclusion and relegation. The chapter starts with a review of the evolution of STI policy in South Africa, anchoring the contemporary developments between the two White Papers (in 1996 and 2019) in the inherited features of the apartheid regime in section 22.2. Section 22.3 describes the performance of the South African STI system from a quantitative perspective. This section also discusses some of the limitations of these indicators. In response to such limitations, section 22.4 reflects on evolutionary economics and what it offers to the analysis of South African STI policy; and section 22.5 concludes.

22.2  Evolution of STI Policy in South Africa The history of STI policy in South Africa has been shaped by its tumultuous political and economic history, as it transited from colonial occupation to segregation through apartheid to its first democratic government. Several scholars have explored the complex evolution of STI policy in South Africa from colonial to apartheid to present times, through various lenses (Dubow 1995; Kaplan 1995, 1996, 2008; Kahn 2006, 2013; Maharajh 2011; Marais 2000; Marais and Pienaar 2010; Mouton 2006; and Scerri 1998, 2009, among others). Scholarly contributions have been complemented with multiple system-​wide reviews, such as the 2007 Organisation for Economic Cooperation and Development (OECD) Review of Innovation Policy (OECD 2007)—​which was informed by an earlier local perspective on the system conducted by the National Advisory Council on Innovation (NACI 2006); the 2012 Ministerial Review Committee report on the STI

Innovation and technological change in South Africa    469 landscape (DST 2012); the Academy of Science of South Africa Review of the State of the STI system in South Africa (ASSAf 2013); and the 2017 report of the STI Institutional Landscape (STIIL) Review Panel, commissioned by the then Department of Science and Technology. These contributions have helped identify gaps and policy dimensions that could influence the STI system’s further evolution. The STI policy environment post-​1994 reflected the urgency to break from the earlier dispensation, resulting in a drastic shift in policy priorities and overall understanding of the objectives of the STI system. It is, therefore, useful to anchor the contemporary analysis of STI in what was inherited from the earlier period. The South African policy and institutional environment preceding the democratic era was driven by the principles of racial segregation and dominance tightly preserved by the apartheid regime. Hence the policy choices and options of the time reflected the need to secure the viability and sustainability of such a worldview. In his account of South Africa’s STI policy since the first national S&T planning framework in 1916, Scerri (2009) reflects on the wide reach of policy dimensions, affected by what we now relate to the functioning of the STI system, including labour, industrial development, education, and science-​related policies, as the apartheid regime gradually forced more and more sectors of the economy to comply with its discriminatory rules. Maharajh (2011) provides a detailed account of the political economy, power struggles, and social tensions driving the STI system under racial capitalism and apartheid. While government valued scientific knowledge and research—​heavily subsidizing scientific research in defence, energy, and areas related to industrial expansion—​it also put pressure on the scientific community to segregate along racial lines (UNESCO 1967). Similarly, large-​ scale technological projects served to expand the apartheid state’s apparatus and display its power (Edwards and Hecht 2010). In fact, the apartheid STI policy has been described as mission-​oriented, by establishing capacities in strategic areas, such as atomic energy and armaments (Kaplan 1996). However, these big missions appeared to be divorced from other areas relevant to S&T policy, such as industrial development;1 and there was little concern about their impact on building capabilities for the S&T system (Kaplan 1996). Moreover, there was no unified vision across government, with the management of science and technology divided between the Department of Education (responsible for science) and the Department of Trade and Industry (responsible for technology). In his description of the dynamics of ‘racial capitalism’, Maharajh (2011) asserts that most policy measures were aimed at maintaining a stable and steady availability of ‘cheap Black labour’, effectively inhibiting the aspirations of the vast majority of South Africans. Notably, the educational inequalities created through policies such as the Bantu Education Act of 1952, limited the educational potential of most of the population, posing severe limitations to economic development and social equality (on education, see C ­ hapter 33 by Branson and Lam in this volume). The advocates of these policies expected that these limitations would be overcome through accelerated mechanization,

1 

See Chapter 24 in this volume by Anthony Black on industrial policy.

470    Erika Kraemer-Mbula and Rasigan Maharajh the development and adoption of labour-​saving technologies, and the importation of skills through the immigration of whites into the country (Scerri 1998). Persistent low levels of human capabilities and skills eventually shook the foundations of economic development, rendering the country internationally uncompetitive. But more importantly, such distortions left a lasting scar in South Africa’s socio-​economic context and remain extremely difficult to repair to date. To inform the establishment of a new regime, the new democratic government commissioned a review of the state of S&T in South Africa, sponsored by the Canadian IDRC. The report confirmed that the S&T system inherited from decades of apartheid was fragmented and uncoordinated, it did not serve the interests of all South Africans, and it was ineffective and inefficient (IDRC 1993). The democratic transition provided an opportunity to make radical changes to a distorted and fragmented system inherited from a racialized system. It raised new economic, political, and social imperatives, which widened the discourse about the objectives of S&T. From discussions amongst the mass democratic movement, through to the negotiations on a future dispensation for the post-​apartheid South Africa, a broader preoccupation with national development emerged and also centred on the role of S&T. As raised by Kaplan, there was a ‘need to develop an S&T system which simultaneously supports the emergence of an internationally competitive business sector and the enhanced provision of infrastructure, such as housing, clean water and domestic electricity’ (Kaplan 1995: 1). The need for a more coordinated view of S&T crystallized in the establishment of the first Department of Arts, Culture, Science, and Technology (DACST) in 1994, with the role of overseeing and coordinating policy for the entire S&T system. Following its formation, DACST initiated a process of policy formulation with a Green Paper that, after subsequent consultation, led to the inaugural White Paper on Science and Technology in 1996. The White Paper entitled ‘Preparing for the 21st Century’, was adopted by the Cabinet in 1996, and suggested a radical departure from the approaches utilized by the apartheid regime. It took the concept of national innovation system (NSI) as an organizing framework; establishing broad goals of the system, identifying the requirements that the NSI should meet and also the articulation of the main stakeholders in the system (RSA 1996)–​those being the business sector, the higher​education sector, the science councils, government departmental research institutes, and NGOs. This focus brought a renewed interest in the commercial viability of scientific knowledge, and hence the collaboration between scientific research, the private sector, and government organizations. Such policy focus on stimulating innovation through networking and collaboration did, however, recognize the context of extreme inequality, a legacy of exclusion, endemic unemployment, and widespread poverty that characterized the lives of the majority of South Africans. While the development of the White Paper of Science and Technology maintained strong resonance with the Reconstruction and Development Programme of the newly elected government, its adoption coincided with the introduction of the Growth, Employment, and Redistribution (GEAR) strategy in 1997 (RSA 1996). It had two

Innovation and technological change in South Africa    471 interrelated but distinct aims: to create economic growth and enhanced participation in the economy; and to achieve this growth through the innovative social development of the population (DST 2012: 9). Such an expansive and inclusive approach was, however, stymied by the austerity of GEAR, which reduced government expenditures thereby curtailing many of the more radical elements of S&T transformation. Following the development of the broad policy framework through the White Paper, DACST was split into two separate entities: The Department of Arts and Culture (DAC) and the Department of Science and Technology (DST). The newly established DST published the second foundational policy document: the National Research and Development Strategy (NRDS) in 2002, which aimed to address the deficiencies of South Africa’s S&T system by highlighting the low investments in R&D—​particularly by the business sector; the declining and ageing scientific population—​skewed by gender and race; the lack of an intellectual property (IP) framework; and generally the lack of a coordinated governance system for the NSI, related to institutional fragmentation (see critical review by Kaplan 2004). The NRDS also was supported by the National Key Research and Technology Infrastructure Strategy, launched in 2004 in response to earlier concerns raised by the findings of a National Research and Technology Audit in 1998, which described South Africa’s infrastructure for research and technology development as old, and not enabling South African researchers to compete effectively internationally.2 More specifically, the NRDS argued that ‘the total capacity of the system is about one-​third to one-​half the size it should be to form the basis of a competitive knowledge-​based economy for South Africa in the medium to long term’ (RSA 2002: 40). Similarly, ‘this new R&D Strategy depends on doubling government investment in science and technology over the next three years, with more gradual increases thereafter. This would raise the national investment to somewhat over 1 per cent, not yet as large as many of our competitors, but enough to signal an appropriate, comprehensive and sustainable strategy for the knowledge economy’ (RSA 2002: 17). The earlier focus on the S&T system, explicit in the NRDS (RSA 2002), later shifted towards a preoccupation for the commercialization of research outputs, materializing in the Ten-​Year Innovation Plan (TYIP) (RSA 2008). The TYIP (2008–​18) had an explicit focus on innovation as a route for economic progress, aiming to support the production and dissemination of knowledge through innovation and entrepreneurship. The Plan identified five Grand Challenges—the bioeconomy, space, energy, climate change, and social dynamics—as key focus areas that would allow South Africa to advance in emerging technologies by leveraging its natural resources. Broad reviews of the STI system (OECD 2007; DST 2012; ASSAf 2013; and DST 2017) generally concur that despite the quantitative improvements in the STI outputs since 1994, the innovation system appears to have failed to meet the needs of the majority as the quality of the changes has not radically transformed the economy, nor

2 

It emphasized that only 10 per cent of the country’s equipment base at the time could be considered as state-​of-​the-​art, with the remainder being largely outdated.

472    Erika Kraemer-Mbula and Rasigan Maharajh significantly altered social relations consistent with the massive political reconstruction engendered by the democratic breakthrough. See two illustrative statements below: 1. South Africa’s NSI is making insufficient contribution to poverty reduction and wider inclusion in the mainstream economy. (OECD 2007) 2. The state’s investment in innovation has been biased towards ‘big science’ and inadequate focus had been placed on requirements for meeting the social development priorities. (DST 2012)

Thus, South Africa remains an exclusionary and dualistic economy with one of the highest and most persistent inequality rates in the world. The centrality of the skills shortages in South Africa was confirmed by the OECD review which characterized human resource development as ‘perhaps the issue that will be central to all other aspects of the development of the STI system over the next decade’ (OECD 2007: 87). This concern arose from the large gap generated by the combination of slow growth in the supply of university graduates capable of undertaking research, and the growing demand for design and engineering skills generated by the increased rate of investment across the economy (OECD 2007: 7). A critique of the high-​skills argument is presented by Kraak et al. (2006), who argue that the national skills problem is not located only in the high-​skill end of the spectrum, but also in terms of the intermediate and low-​skill needs. The report from a ministerial review committee tasked to review the innovation landscape (DST 2012), provided extensive recommendations to revive what it called a ‘fatigued’ system of education policy changes and reforms. The Ministerial Review Committee recommends (amongst others) the revitalization of technical colleges, preferential funding schemes for the development of strategic skills, and measures to improve the academic job market by opening up opportunities in public and research enterprises (DST 2012). Twenty-​three years after adopting the first White Paper on Science and Technology in 1996, a new White Paper on Science, Technology and Innovation (STI) was endorsed by Cabinet in March 2019 (RSA 2019). The new White Paper sets out to advance beyond the initiatives established in its predecessor while prioritizing higher and inclusive economic growth; speeding up socio-​economic development; promoting environmental sustainability; improving government’s service delivery and decision-​making; and enhancing the efficiency of institutions. While the new White Paper of 2019 acknowledges some quantitative improvements (e.g. a three fold increase in publications, significant growth in the participation of Black people and women in the research and development workforce, and a rise in doctoral graduation rates), several challenges that were recognized in the 1996 White Paper are again reiterated—​such as the severe deficits in the human resource base and its insufficient transformation, inadequate funding for STI activities, and lack of policy coordination across different government agencies. The new White Paper accounts for global dynamics, marked by population trends, fast technological developments in the digital space, changes in geopolitics, as well as the impending climate crises (RSA 2019)—On climate change, see Chapter 16 of this volume. See also

Innovation and technological change in South Africa    473 Chapter 23 on the fourth industrial revolution in this volume. In terms of governance of the STI system, the new White Paper makes concrete proposals for integrated STI policy planning and implementation, such as establishing a ministerial STI structure and well-​ functioning ‘core’ policy nexuses to harmonize and coordinate implementation plans.

22.3  Performance of the South African STI System: A Quantitative Perspective The last twenty-​five years have witnessed various reforms and revisions of the institutional landscape related to STI, in the form of new policies and organizations supporting, funding, and performing research, innovation, and technological advances. The previous section indicated that STI policy has evolved into a comprehensive and ambitious transformative agenda. One of its main priorities is to address the needs and demands of the majority of citizens. This section assesses to what extent such policy aspirations have manifested in improvements in STI performance. South Africa’s STI system is often described as one of the most advanced on the continent (e.g. UNESCO 2015; AUDA-​NEPAD 2019). The assessment of its STI performance heavily relies on a set of standard indicators, which portray specific dimensions of the system. This suite of indicators has gained relevance in decision-​making at many levels. Their development and use have been primarily guided by the perceived need to capture data that can be internationally comparable. They include (a) publications and R&D efforts to measure the capacity for knowledge generation; (b) patents and innovation data to measure innovativeness and technological sophistication; and (c) human resources for STI to measure human capabilities in the system. Below, we describe the STI system through the lens of these indicators, raising concerns about an over-​reliance on quantitative measures to effectively assess the strengths, weaknesses, and broad opportunities for STI to drive the ambitious, transformative agenda formulated in STI policy.

22.3.1 South Africa STI Trends and Figures 22.3.1.1 Measures of Knowledge Generation South Africa has positioned itself as the biggest producer of publications on the African continent—​accounting for 77.4 per cent of the publications from SADC countries (NACI 2020). It contributed to 25,372 articles in ISI-​listed journals in 2018, more than twice those published in 2009 (10,774)3—see Figure 22.1. Its share of world output has 3  According

to Mouton et al. (2019), this increase in figures must be viewed cautiously due to the prevalence of predatory journal publications.

474    Erika Kraemer-Mbula and Rasigan Maharajh 30,000 0.73%

25,000 20,000

0.50% 0.51%

0.57%

0.62% 0.62%

0.77%

0.90% 0.80% 0.81% 0.80%

0.65%

0.70% 0.60% 0.50%

15,000

0.40% 0.30%

10,000

0.20%

5,000

0.10%

0

0.00% 2009

2010

2011

2012

2013

2014

South African Publications

2015

2016

2017

2018

World Ratio

Figure 22.1  South African publications and world ratio (count) Source: NACI (2020).

also doubled from 0.4 per cent in 2000 to 0.81 per cent in 2018. While its world ranking in terms of the number of publications has improved (from position number 34 in 2000 to 28 in 2016), it is still surpassed by all other BRICS countries (China in position 2, India 7, Brazil 13, and Russia 15 in 2016) (Mouton et al. 2019). Publications concentrate in the area of natural sciences (34 per cent), social science and humanities (38 per cent), and medical and health sciences (20 per cent), with these areas of specialization having remained relatively stable over the past two decades (Mouton et al. 2019). Scientific publications emerge mainly from five research-​intensive universities (Cape Town, Witwatersrand, Pretoria, KwaZulu-​Natal, and Stellenbosch), which account for about two-​thirds of all publications in the country. Networking and linkages in knowledge production have also been on the rise, primarily through collaborations in scientific publications with international partners outside Africa (mainly universities in Europe and the United States). Overall, international collaboration has increased from 34 per cent in 2000 to 52 per cent in 2016 (Mouton et al. 2019). The use of publication data to explore the strengths and weaknesses of the science system in South Africa has gained traction over the last two decades. Multiple scholars have generated a large body of scientometric studies (Mouton et al. 2019; Kahn 2011; Pouris 2012; among many others). These efforts culminated in the establishment of the DST/​NRF Centre of Excellence in Scientometrics, and the Science, Technology and Innovation Policy (SciSTIP) in 2014, which has been influential in shaping policy interventions focused on promoting scientific productivity of South African science and research institutions. Turning to R&D data, the mainstream neoclassical literature adheres to a linear view of technological progress, following a discrete path from basic and applied research to

Innovation and technological change in South Africa    475 technological development and eventually to innovation. This view regards R&D as the main driver of innovation (focused on technological innovation).4 In line with this perception, R&D investment and R&D intensity have become key indicators to monitor resources devoted to STI worldwide, including in South Africa. In South Africa, the systematic collection of R&D indicators through R&D surveys is conducted by CeSTII of the Human Science Research Council (HSRC).5 Since the first surveys started in 1991/​92 (by the then Foundation for Research and Development) to the present, the collection of R&D data has sought to follow the guidelines provided by the Frascati Manual (OECD).6 The R&D surveys measure inputs into the conduct of R&D (namely people, equipment, and funding) and cover the business, government (including the science councils), higher-​education, and non-​profit sectors. The statistics are used to develop science policy, to set government R&D priorities and funding levels, and for monitoring and benchmarking purposes. According to the last R&D survey in 2017/​18, R&D intensity (measured as South Africa’s gross domestic expenditure on R&D (GERD)7 as a percentage of GDP) was 0.83 per cent, which remains considerably below the ambitious 1.5 per cent national policy target. This percentage has remained more or less static for most of the past fifteen years. On this basis, studies such as Mouton et al. (2019) have stressed South Africa’s poor research funding performance. This becomes more evident when South Africa is compared with the average in upper-​middle-​income countries, which stands at 1.46 per cent (GERD as a percentage of GDP) (NACI 2020). Moreover, while these countries saw a steady increase between 2008 and 2013, South Africa experienced a decline in the same period. Although South Africa has the highest R&D intensity in Africa, it stands behind China, Brazil, and Russia among the BRICS countries. Most R&D in South Africa is funded by the government (46.7 per cent of total R&D) in 2017/​18, and the business sector is the second-​largest funder. In terms of performance, business enterprise R&D spending (BERD) is typically considered an important indicator of business commitment to innovation. In South Africa, the business sector is the largest performer of R&D in 2017/​18 (representing 41 per cent of GERD), followed by the higher-​education and science councils. However, although BERD has had an increasing trend from 2012 to 2018 in nominal values, its share of the total R&D expenditure has been in decline—see Figure 22.2. To support business R&D, in 2006 the government introduced the R&D tax incentive programme, which gives a 150 per cent tax deduction for expenditure on eligible R&D. The introduction of such an instrument had 4  In contrast, evolutionary economists generally consider R&D as just one input to innovation. From an evolutionary economics perspective, only a small, albeit significant, proportion of all innovations arise from R&D. 5  A summary of the results can be found at http://​www.hsrc.ac.za. 6  Some authors (Blankley and Kaplan 1997)  note the inconsistency of the R&D data series as the survey has been conducted by different agencies with differing methodological gaps in time. 7  GERD is an aggregated measure of in-​ house R&D expenditure performed domestically in five sectors, namely government, science councils, higher education institutions, the business sector, and the not-​for-​profit sector.

476    Erika Kraemer-Mbula and Rasigan Maharajh 18,000 16,000 55.9% 57.7% 58.6% 14,000

70% 53.2%

49.7%

60% 47.1%

44.3% 45.9% 45.3%

10,000

42.7% 41.4% 41.0% 50% 40%

8,000

30%

12,000

6,000

20%

4,000

10%

2,000 0

20 06 /2 00 20 7 07 /2 00 8 20 08 /2 00 9 20 09 /2 01 0 20 10 /2 01 1 20 11 /2 01 2 20 12 /2 01 3 20 13 /2 01 4 20 14 /2 01 5 20 15 /2 01 6 20 16 /2 01 20 7 17 /2 01 8

0%

BERD (ZAR millions)

BERD/GERD

Figure 22.2  South African business expenditure on R&D (BERD) and the BERD/​GERD ratio, 2006–​18 Source: Molotja et al. (2019) and data from the National R&D Survey (HSRC 2020).

been suggested in earlier literature by authors such as Blankley and Kahn (2005) and Kaplan (2001). The outputs from the South African R&D survey have been widely analysed, resulting in a rich literature using these data to raise several issues. Concurring with the limitations raised by the NRDS in 2002, Walwyn (2008) highlights the importance of a higher BERD for South Africa’s economy to shift from its dependence on resource-​based industries towards more knowledge-​intensive economic activities. Walwyn contends that this is particularly important for sectors that face intense international competition such as food, metal products, and petrochemicals, which do not display high levels of research intensity. In contrast, the largest proportion of BERD investment in 2017/​18 is in ‘financial, intermediation, real estate and business services’, accounting for 48.8 per cent of BERD. The manufacturing sector accounted for 28.2 per cent of BERD in 2017/​ 18 (HSRC 2020). Molotja, Parker, and Mudavanhu (2019) find that R&D performance in the business sector is concentrated in a few industries and a relatively small number of large firms. Their study finds that most small and medium-​sized enterprises invest in R&D projects only sporadically, for no more than two consequential years.

22.3.1.2 Measures of Technological Sophistication and Innovativeness The use of patent statistics as a measure of technological advance rests on the assumption that they reflect the inventive activities in technological innovation. Despite being an outlier on the continent in patenting activity (AUDA-​NEPAD 2019), the number of South African patents applications and patents granted has remained stagnant throughout 1994–​2019, as shown in Figure 22.3.

Innovation and technological change in South Africa    477 4,000 3,500 3,000 2,500 2,000 1,500 1,000

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

1999

1998

1997

1996

1995

0

1994

500

Patent applications (direct and PCT national phase entries) Patent grants (direct and PCT national phase entries)

Figure  22.3  South African patent applications and patents granted, by country of origin, 1994–​2019 Source: WIPO IP Statistics Data Center (https://​www3.wipo.int/​ipstats).

This contrasts with the rapid increase in patenting activity in upper-​middle-​income countries in the decade 2009–​2018, with patent applications per million more than tripling (from 180 in 2009 to 634 in 2018); while in South Africa, there has been a tendency for patent applications per million population to decline during the same period (from thirty-​nine in 2009 to thirty-​two in 2018) (NACI 2020). The level of innovativeness in the economy is also assessed using the National Innovation Survey data.8 According to the latest Innovation Survey available, innovation is pervasive across all sectors of the South African economy, with more than two-​thirds (69.9 per cent) of South African businesses reporting being active innovators in 2014/​ 16. The engineering and tech, manufacturing, and trade sectors reported the highest incidence of innovation among all sectors of activity (HSRC 2020).9 Among those firms that innovate, four types of innovation featured almost in equal shares: product innovation (48.2 per cent), organizational innovation (42.0 per cent), marketing innovation (41.7 per cent), and process innovation (34.6 per cent). The literature generally assumes that innovation in developing countries tends to be incremental (e.g. improvements to products and services), rather than radically new (Kraemer-​Mbula and Wamae 2010). The innovation survey confirms this assumption since 80.5 per cent of innovative 8  Innovation surveys based on the Oslo Manual are embedded in an innovation systems framework, highlighting the importance of linkages and collaboration across organizations. The surveys started in 2004 collected by the Centre for Science, Technology and Innovation Indicators (CeSTII) at the Human Sciences Research Council (HSRC). 9  The 2014–​16 Business Innovation Survey is the sixth such survey undertaken in South Africa. The results were released in July 2020.

478    Erika Kraemer-Mbula and Rasigan Maharajh business turnover was generated by goods and services that were unchanged or marginally modified. Guided by the third edition of the Oslo Manual, the South African innovation survey attempts to address the systemic dimension of innovation. One such systemic dimension relates to the interactions and collaborations across actors in the system, resulting in knowledge flows conducive to innovation. South Africa’s innovation system displays low collaboration levels, with only 20.8 per cent of innovation-​active enterprises indicating any form of collaboration in their innovation development in 2014/​16 (HSRC 2020). Less is known about the various effects that innovation activities have on firm performance, employment, and broader socio-​economic aspects. Few studies have emerged using the results from the Innovation Survey in South Africa. Some exceptions are Sithole and Buchana (2021), who assess the impact of innovation on employment at the firm level; and Machaka et al. (2019a) and Machaka et al. (2019b) who use the responses from small South African firms in the Innovation Survey conducted from 2010 to 2012 to explore the factors that influence innovation activities in small manufacturing sector firms, as well as the complementarity between various innovation knowledge sources.

22.3.1.3 Measures of Human Capabilities The shortage of skills and human resources is one of the most cited constraints in South Africa’s research and innovation system (OECD 2007; DST 2012). The provision of sufficient graduates and researchers, especially at the tertiary level are referred to as enablers’ in the main policy documents in South Africa (NACI 2020). Human resources in S&T are often proxied by the STI workforce, or R&D personnel—​ consisting of permanent and temporary members of the R&D system. From this lens, there has been a notable upward trend in South Africa’s R&D human resources between 2001/​02 and 2017/​18, whether measured in personnel headcount or full-​time equivalents (FTEs) (HSRC 2020)—see Figure 22.4. However, this positive trend has to be read with caution, since methodological changes in the collection of R&D personnel data in the 2016/​17 R&D survey may have contributed to the increase in the headcount and number of FTEs observed (HSRC 2020). More broadly, Pogue (2007) points out the limitations of assessing the human resources in S&T by counting R&D personnel, since a more accurate assessment would require not only capturing those that are active in the system, but also those who are qualified but not currently employed in the system. The social composition of the research system has undergone slow, although steady, transformation over the past decade. The percentage of female researchers has increased to reach 45.3 per cent of the total number of researchers in 2017. The number of African researchers increased from 6,595 in 2008 to 10,815 in 2017, a vital trajectory to address the distortions caused by apartheid’s racialized system. However, Blankley (1994) re-​ directs our attention to the broader ‘abyss in African school education in South Africa’, confirmed by Kahn (2006), who states that the growth of the stock of high-​level skills depends upon the flow that higher education receives from the school system—on education, ​also see ­Chapter 33 by Branson and Lam in this volume.

Innovation and technological change in South Africa    479 90000 80000 70000 60000 50000 40000 30000 20000 10000

20

01 20 /02 02 /0 20 3 03 / 20 04 04 / 20 05 05 / 20 06 06 / 20 07 07 / 20 08 08 / 20 09 09 / 20 10 10 / 20 11 11 / 20 12 12 / 20 13 13 / 20 14 14 / 20 15 15 / 20 16 16 / 20 17 17 /1 8

0

R&D personnel headcount

R&D personnel FTE

Figure 22.4  R&D personnel, 2001–​18 Source: HSRC R&D surveys.

22.3.2 Limitations of Current Indicators While available and widely used indicators provide a useful overview of the system, they have certain limitations worth mentioning. Discussing their limitations does not proclaim that they lack value or that they do not capture specific dimensions of the innovation process. Instead, it cautions against exclusive reliance on them since they offer a partial representation of the innovation system, which requires further qualitative contextualization.10

22.3.2.1 Knowledge Generation Data on publications may give us an indication of the knowledge that is generated within the scientific community; however, it tends to exclude other essential sources of knowledge produced by actors outside the community and those based on tacit knowledge (doing, using, and interacting) (Jensen et al. 2007). Criticisms around the costs associated with access to scientific publications, the oligopoly of publishers, and incentivization schemes premised upon publication in accredited journals abound (e.g. Muller 2017 and Tomaselli 2018, among others). These scholars point to the perverse consequences of overdetermining publications as proxies of knowledge creation and especially the commodification of the process and products themselves.

10 

A similar point has been raised recently by Kruss et al. (2021).

480    Erika Kraemer-Mbula and Rasigan Maharajh Turning to the R&D indicators, the literature using R&D data has highlighted some of its limitations, as it does not fully capture (1) sectoral and firm specificities, nor (2) the activities of micro, small, and medium enterprises. Moreover, early work by Scerri (1990) cautions against the unquestioned use of R&D variables in the context of developing countries due to the high proportion of imported technologies, making R&D statistics an unsatisfactory estimate of a country’s ability to innovate new processes and services. More broadly, the dangers of relying almost exclusively on R&D data to assess the innovative performance and technological advance of an STI system have been stressed by scholars and practitioners (see Gault 2013).

22.3.2.2 Technological Advancement While patents are commonly used as a proxy for technological advancement, there are several limitations, which include: (a) many inventions are not patented, (b) many patents have no commercial value, (c) not all innovation requires intellectual property protection, and (d) there are other strategies beyond patents to protect intellectual property (such as, secrecy or industrial designs).

22.3.2.3 Human Capabilities Some studies in South Africa have addressed the dissonance of speaking of capabilities whilst pursuing narrower human capital goals. Maharajh and Motala (2016) warn us about the narrow and reductive assumptions underlying the human capital theory, arguing that ‘linear representations of education’s role in socio-​economic development are both unhelpful and misleading since there is [ . . . ] a wide range and complex array of factors that are constitutive of such a relationship’ (2016: 29). The authors draw attention to the qualitative attributes of education and training systems as well as the singularly important historical issues of racist discriminatory practice, gender and rurality, social fragmentation, and the power of capitalist concentration in conceptualizing human capabilities and shaping adequate policy responses. In this vein, Maharajh and Scerri (2011) also highlight the importance of aligning measures of human capabilities with broader developmental goals. While quantitative indicators provide a useful overview of the STI, there is also a unanimous recognition that the South African STI system has had limited impact in improving the lives of the majority. In response to such critical limitations, a key question remains: Do quantitative indicators give us a full picture of the strengths, weaknesses, and opportunities offered by knowledge, technology, and innovation dynamics in South Africa? We argue that addressing the severe and pressing challenges affecting the country, as well as taking advantage of present and future opportunities, requires complementary lenses to look at the STI system and explore possibilities for STI policy. It requires an appreciation for the qualitative factors affecting change, which are embedded in, and profoundly shaped by, culture, history, and institutions. This lens is offered by evolutionary economics specifically, utilizing a progressive reading of the neo-​Schumpeterian literature on innovation. This notion is further discussed in the next section.

Innovation and technological change in South Africa    481

22.4  Evolutionary Economics, National Systems of Innovation, and South Africa’s Contemporary STI Policy Challenges The relationship between science, technology, and innovation has been the subject of debate amongst various schools of economics across previous centuries. It is, therefore, useful to locate neo-​Schumpeterian approaches within the broader history of economic thought and contemporary policy debates. As noted by Roncaglia, ‘underlying the debate there are quite often, hidden from sight but still very significant, different approaches to economics, and not only different opinions on policy’ (2017: ix). Besides the neo-​Schumpeterian approach being regarded as a heterodoxy, the epistemic lacuna in the history of economic thought is also important. The voices and experiences of those in the Global South often only find representation through writing published in the Global North (Ndhlovu and Khalema 2015). The challenges of decoloniality and decolonizing discourses on the economy, and the field of economics more generally, requires much more attention from the current cadre of economics scholars. Other chapters in this volume also carry forward theoretical roots that emerge from across the various schools of economic thinking since the ‘marginalist revolution’ abandoned the classical approach, and rather shifted focus ‘to a new approach based on a subjective theory of value and the analytical notion of marginal utility’ (Roncaglia 2017: 144).11 As noted earlier, the choice of a national system of innovation (NSI) approach to the reform of South Africa’s S&T sector was determined by the country’s Cabinet in its adoption of the 1996 White Paper on Science and Technology. In tracing the origins of the NSI policy framework, Sharif argued that ‘it would be difficult to overemphasize the extent to which the [NSI] concept originated as part of a direct attack on modern mainstream economics’ (2006:  753). In a more recent review, Golichenko endorses Sharif ’s perspective and states more boldly that the NSI conceptualization arose ‘due to the dissatisfaction of a number of economists in the neoclassical mainstream of economic theory and the inadequacy of the interpretations of the role of technology, knowledge, and innovations in economic development within the standard mainstream approaches’ (2016: 464). It is therefore important to locate the NSI framework within a selected account of the history of economic thought to better understand its counterfactual value. 11  This chapter does not have the latitude to expand more on the long history of the emergence of the neo-​Schumpeterian approach and readers are encouraged to engage with Erik Reinert’s The Other Canon Foundation (othercanon.org), Fagerberg et al. (2013), and an anniversary working paper of IERI (2014) for more extensive treatment.

482    Erika Kraemer-Mbula and Rasigan Maharajh The NSI concept finds its origins in classical political economists—​from Adam Smith, to Ricardo and Marx—​whose reflections on the science, technology, and innovation, were seen as necessarily bound within their historical context. In line with this, the work of Schumpeter highlighted that innovation does not occur in a vacuum. Schumpeter’s contribution, according to Metcalfe, posed at least three major challenges to the neoclassical framing of general equilibria and these were: ‘the impossibility of predicting its evolution ex ante even when the general rules of its functioning are understood; the irreversible effects of the growth of knowledge and the impossibility of placing an economy in equilibrium if knowledge is not in equilibrium; and the inevitable link between individuality and personal knowledge such that socially situated individuals matter vitally to the evolution of the system’ (Metcalfe 2009: 55–​6). Richard R. Nelson, a primary contributor to neo-​Schumpeterian theory, describes, together with Sidney Winter, the neo-​Schumpeterian approach as accounting for ‘growth, profits, and capital formation [which] are all generated largely by innovation’ (Nelson and Winter 1977: 22). They highlight the importance of context in shaping the selection process of innovation by stating that ‘firms also differ in the probability that they will create an innovation, or adopt better technology used by others, over a given time period. Not all innovations are superior to existing technology, so the selection process is a key part of the model. Better technology, when it is created, is spread through the system both by expansion of the innovating firm and by imitation’ (Nelson and Winter 1977: 22). As argued by Helena Lastres, a wider array of neo-​Schumpeterian scholars, including Giovanni Dosi, Nathan Rosenberg, Luc Soete, Carlota Perez, and Christopher Freeman, amongst others, contributed since the 1970s to the task of ‘develop[ing] analytical instruments and broader and more complex policy guidelines than those offered by traditional economic theory’ (Lastres 2017:  2). It was not just these individual researchers, but the broader institutional array that they represented, including new research centres based in universities and the range of graduate students that they mentored and who would come to populate a much more diverse and representative array of evolutionary economists now distributed across world systems. Christopher Freeman (1921–​2010), Richard Nelson (1930–​), and Bengt-​Åke Lundvall (1941–​) play a major role in providing and popularizing a shared perspective around the concept of a national system of innovation (NSI) for evolutionary economists. Besides the NSI framework, critical concepts such as techno-​economic paradigms, socio-​institutional structures, and path dependencies were generated and empirically tested. Whilst this rich and complex intellectual history of economic thought underpins South Africa’s adoption of the NSI perspective in seeking to reform its science and technology system and orientate it towards the generation of innovation as an outcome, little engagement with this scholarly tradition is discernible in the local literature until after the democratic breakthrough of 1994. For Mario Scerri, ‘the openness of the systems of innovation approach lends itself to a wide range of ideological positioning, which can range from Marxian to neoliberal ideological underpinnings without violating the fundamental premises of the contra-​neoclassical evolutionary school. While the oppositional positioning of the systems of innovation approach is against mainstream

Innovation and technological change in South Africa    483 neoclassical economics, specifically in its core assumption of full information sets, this cannot be used to infer its ideological slant, since both the neo-​liberal and Marxian schools of thought have little in common with the neoclassical paradigm and much more with the institutional basis of the systems of innovation approach’ (2017: 31–​2). Drawing upon the econometric work by Fedderke, Kaplan argued that ‘the economy shed labour such that labour made a negative contribution to growth. With low levels of investment, capital made a much smaller, albeit positive, contribution to growth. [Total Factor Productivity] TFP growth, by contrast, became the major source of growth’ (2008: 1). Kaplan reflects on the experience of a National Treasury-​appointed international panel which ‘examined South Africa’s policies for growth and specifically its ambitions to achieve both a more ambitious growth target and simultaneously more inclusive employment generating growth. A number of proposals were made by the panel . . . However, the panel made no study of or recommendations in respect of innovation or technological change more broadly—​despite this having been the principal contributor to growth’ (2008: 2). This remains in stark contrast to the National Development Plan (NDP), which mentions innovation at least 108 times in 489 pages (RSA 2012). According to the NDP, ‘overall, South Africa’s global competitiveness needs to be improved, and the system of innovation has a key role to play. It is the principal tool for creating new knowledge, applying knowledge in production processes, and disseminating knowledge through teaching and research collaboration’ (RSA 2012: 326). Thus, and whilst various economic policy frameworks that were deployed to advance post-​apartheid reforms often include or at least support discrete science and technology initiatives, they do not necessarily embrace the critique of neoclassical economics that was advanced by neo-​Schumpeterian evolutionary perspectives. The failure to achieve South Africa’s putative target of 1 per cent of ‘total intramural expenditure on R&D performed in the national territory during a specific reference period’ as a proportion of GDP and captured through the indicator of gross domestic expenditure on R&D (GERD)12 provides at least a quantifiable verifiable objective that has not obtained since 1996. It is well known that South Africa remains one of the most unequal societies in the world. Entrenched inequalities also manifest in the structure and functioning of the innovation system. Inclusion as an ‘end’ has recently gained renewed policy attention under the 2019 White Paper, raising a renewed interest in inclusive innovation in South Africa, both in the literature as in policy practice. Conceptually, inclusive innovation emerges from an evolutionary economy foundation. Paying attention to innovations ‘below the radar’ (Kaplinsky et al. 2009), means recognizing the importance of poverty alleviation and equality in order to have sustainable, long-​term economic development (Heeks et al. 2014). In South Africa, a body of literature has started to emerge (Phiri et al. 2016; Hart et al. 2020, Petersen and Kruss 2019; Petersen et al. 2018, and others) not only expressing discomfort with an exclusive

12 

Definition adopted from OECD (2015: 372).

484    Erika Kraemer-Mbula and Rasigan Maharajh approach to innovation to serve global competitiveness and firm profitability but also discussing new avenues of activating an inclusive innovation focus from a policy perspective. An early comparative study and systemic perspective by Kraemer- Mbula and Wunsch-Vincent (2016) captures the importance of capturing the innovation activities not only ‘for’ but ‘by’ marginalized communities, such as those in the informal economy.13 These contributions complement the ongoing work done by the CeSTII capturing innovation in informal communities and piloting the development of new indicators. The DSI is drafting a strategy to promote innovation for inclusive development (IID). It defines IID as ‘innovation that addresses the triple challenge of inequality, poverty and unemployment and enables all sectors of society, particularly the marginalized poor, informal sector actors and indigenous knowledge holders to participate in creating and actualizing innovation opportunities as well as equitably sharing in the benefits of development’ (DST 2016: 11). These efforts open up new opportunities for the transformation of South Africa’s innovation system, and at the same time raise new challenges in terms of measurement, institutional development, alignment of policy networks (Petersen and Kruss 2019), and integration of innovation policies with much broader transformative social policies (Phiri et al. 2016).

22.5 Conclusion The South African STI system has undergone significant changes across the span of history. In this chapter, we argue that the dominance of orthodox macroeconomic frameworks has retarded the flourishing of a heterodox-​inspired systems of innovation approach and that South Africa’s adoption of the systems of innovation approach to development offered a progressive alternative to the more fundamentalist neoclassical market-​biased approaches, its export-​led growth orientation, and a belief in trickle-​down theories. The bureaucratic capture of the policy discourse, the adoption of austerity frameworks, and the persistence of rampant corporate corruption has further reduced the progressive possibilities and opportunities arising from the innovation systems and evolutionary economics in South Africa. This chapter offers an overview of the South African STI system and its governance through policy in the post-​apartheid period. The overview reveals that many of the distortions and inequalities created by the apartheid regime have persisted over time, despite the shifts in policy orientation. As noted by Scerri and Maharajh (2018) ‘building a robust and sustainable system of innovation in South Africa remains possible, and indeed, still offers a plausible strategy for creatively destroying an inequitable past, tools for confronting contemporary contradictions, and advancing foresight as a means of

13 

See ­Chapter 35 in this volume by Rogan and Skinner on the informal sector.

Innovation and technological change in South Africa    485 collectively defining a better future life for all’ (2018: 317). The chapter, therefore, argues that an evolutionary approach offers the lens needed to capture the more qualitative and institutional dimensions of the STI system for it to drive transformative change.

References ASSAf. 2013. ‘The state of the science, technology and innovation (STI) system in South Africa’. Academy of Science of South Africa [online] http://​hdl.handle.net/​20.500.11911/​79. AUDA-​NEPAD. 2019. African Innovation Outlook III. Johannesburg:  African Union Development Agency. Blankley, William. 1994. ‘The abyss in African school education in South Africa’, South African Journal of Science, 90(2): 54. Blankley, W., and Kaplan, D. (1997) Innovation Patterns in South African Manufacturing Firms: Report on the survey of innovative activity in South African manufacturing firms, ISP/FRD. DST. 2012. ‘Ministerial Review Committee on the science, technology and innovation landscape in South Africa’. Pretoria. DST. 2016. ‘Innovation for inclusive development: Draft strategic framework’. Department of Science and Technology, Pretoria. DST. 2017. ‘Ministerial Review Panel. Ke Nako: A review of the South African science, technology and innovation institutional landscape’. Pretoria. Dubow, Saul. 1995. Scientific Racism in Modern South Africa. Cambridge:  Cambridge University Press. Edwards, Paul N., and Gabrielle Hecht. 2010. ‘History and the technopolitics of identity: The case of apartheid South Africa’, Journal of Southern African Studies, 36(3): 619–​39. Gault, Fred (ed.). 2013. Handbook of Innovation Indicators and Measurement. Cheltenham: Edward Elgar Publishing. Golichenko, Oleg. 2016. ‘The national innovation system’, Problems of Economic Transition, 58(5): 463–​81. Hart, Tim, Irma Booyens, and Sikhulumile Sinyolo. 2020. ‘Innovation for development in South Africa:  Experiences with basic service technologies in distressed municipalities’, Forum for Development Studies, 47(1): 23–​47. Heeks, Richard, Christopher Foster, and Yanuar Nugroho. 2014. ‘New Models of Inclusive Innovation for Development’, Innovation and Development, 4(2): 175–​85. HSRC. 2020. South African National Survey of Research and Experimental Development Main Report: 2017/​18. Tshwane: Human Sciences Research Council. IDRC. 1993. ‘Towards a science and technology policy for a democratic South Africa’. IDRC Mission Report July, International Development Research Centre, Ottawa. IERI. 2014. ‘Revisiting some of the theoretical and policy aspects of innovation and development’. IERI 10th Anniversary Working Paper, Tshwane. Jensen, Morten Berg, Björn Johnson, Edward Lorenz, and Bengt Åke Lundvall. 2007. ‘Forms of knowledge and modes of innovation’, Research Policy, 36(5): 680–​93. Kahn, Michael. 2006. ‘The South African national system of innovation: From constructed crisis to constructed advantage?’, Science and Public Policy, 33(2): 125–​36. Kahn, Michael. 2011. ‘A bibliometric analysis of South Africa’s scientific outputs: Some trends and implications’, South African Journal of Science, 107(1–​2): 1–​6.

486    Erika Kraemer-Mbula and Rasigan Maharajh Kahn, Michael. 2013. ‘Rhetoric and change in innovation policy:  The case of South Africa’, Science, Technology and Society, 18(2): 189–​211. Kaplan, David. 1995. ‘The state of science and technology in South Africa: New priorities, new policies’. DPRU Working Paper no. 95/​001, University of Cape Town. Kaplan, David. 1996. ‘Science and technology in a democratic South Africa—​New challenges and new policy directions’. Science, Technology and Society, 1(1): 101–​27. Kaplan, David. 2001. ‘Rethinking government support for business sector R&D in South Africa: The case for tax incentives’, South African Journal of Economics, 69(1): 72–​92. Kaplan, David. 2004. ‘South Africa’s national research and development strategy: A review’, Science, Technology and Society, 9(2): 273–​94. Kaplan, David. 2008. ‘Science and technology policy in South Africa: Past performance and proposals for the future’, Science, Technology and Society, 13(1): 95–​122. Kaplinsky, Raphael, Joanna Chataway, Norman Clark, Rebecca Hanlin, Dinar Kale, Lois Muraguri, Theo Papaioannou, Peter Robbins, and Watu Wamae. 2009. ‘Below the radar:  What does innovation in emerging economies have to offer other low-​income economies?’, International Journal of Technology Management and Sustainable Development, 8(3): 177–​97. Kraak, Andre, Hugh Lauder, Phillip Brown, and David Ashton. 2006. Debating High Skills and Joined-​up Policy. Cape Town: HSRC Publishers. Kraemer-​Mbula, Erika, and Watu Wamae (eds). 2010. Innovation and the Development Agenda. Paris: OECD/​IDRC. Kraemer-​Mbula, Erika, and Sacha Wunsch-​Vincent (eds). 2016. The Informal Economy in Developing Nations. Cambridge: Cambridge University Press. Kruss, Glenda, Moses Sithole, and Yasser Buchana. 2021. ‘Towards an indicator of R&D and human development’, Development Southern Africa, 38(2): 248–​63. Lastres, Helena M.  M. 2017. ‘Development, innovation, sustainability and policies:  Chris Freeman’s legacy’. Globelics Working Paper no. 2017-​02, Ankara. Machaka, Ronald, Nontombi P. Marule, Ogundiran Soumonni, Moses Mefika Sithole, and Precious Mudavanhu. 2019a. ‘Complementarity between innovation knowledge sources in small manufacturing sector firms: Evidence from the South African Business Innovation Survey’. [online] https://​ssrn.com/​abstract=3530099 or http://​dx.doi.org/​10.2139/​ ssrn.3530099. Machaka, Ronald, Ogundiran Soumonni, Precious Mudavanhu, and Moses Mefika Sithole. 2019b. ‘Determinants of innovation in small manufacturing firms: Evidence from the South African Business Innovation Survey’. [online] https://​ssrn.com/​abstract=3530068 or http://​ dx.doi.org/​10.2139/​ssrn.3530068. Maharajh, Rasigan. 2011. Innovating beyond Racial Capitalism:  A Contribution towards the Analysis of the Political Economy of Post-​Apartheid South Africa. Lund: Lund University. Maharajh, Rasigan, and Enver Motala. 2016. ‘Re-​forming human capabilities utilising science, technology, and innovation policy: The case of South Africa’. Submission to the National Advisory Council on Innovation, Tshwane. Maharajh, Rasigan, and Mario Scerri. 2011. ‘Economic growth and human development challenges for science, technology and innovation in Africa’. African Innovation Outlook, African Union, Addis Ababa. Marais, Hendrik C. 2000. Perspectives on Science Policy in South Africa. Menlo Park: Network Publishers.

Innovation and technological change in South Africa    487 Marais, Hendrik C. and Magdal Pienaar. 2010. ‘Evolution of the South African science, technology and innovation system 1994–​2010:  An exploration’, African Journal of Science, Technology, Innovation and Development, 2(3): 82–​109. Metcalfe, Stanley J. 2009. ‘Marshall and Schumpeter:  Evolution and the institutions of capitalism’, in Andreas Pyka, Uwe Cantner, Alfred Greiner, and Thomas Kuhn (eds) Recent Advances in Neo-​Schumpeterian Economics Essays in Honour of Horst Hanusch. Cheltenham: Edward Elgar Publishing. Molotja, Neo, Saahier Parker, and Precious Mudavanhu. 2019. ‘Patterns of investing into business R&D in South Africa’, Foresight and STI Governance, 13(3): 51–​60. Mouton, Johann. 2006. ‘Science for transformation:  Research agendas and priorities in South Africa’, in Louk Box and Rutger Engelhard (eds) Science and Technology Policy for Development: Dialogues at the Interface, 89–​106. London: Anthem Press. Mouton, Johann, with Isabel Basson, Jaco Blanckenberg, Nelius Boshoff, Heidi Prozesky, Herman Redelinghuys, Rein Treptow, Milandré van Lill, and Marthie van Niekerk. 2019. ‘The state of the South African research enterprise’. DST-​NRF Centre of Excellence in Scientometrics and Science, Technology and Innovation Policy, Stellenbosch University, Stellenbosch. Muller, Seán M. 2017. ‘Academics as rent-​seekers: Distorted incentives in higher education, with reference to the South African case’, International Journal of Educational Development, 52: 58–​67. NACI. 2006. The South African National System of Innovation:  Structures, Policies and Performance. Pretoria: NACI. NACI. 2020. ‘The 2020 South African science, technology and innovation indicators report’. Pretoria: NACI. Ndhlovu, Tidings P., and Nene Ernest Khalema. 2015. ‘Southern Africa’, in V. Barnett (ed.) Routledge Handbook of the History of Global Economic Thought. Oxford: Routledge. Nelson, Richard R., and Sidney G. Winter. 1977. ‘In search of a useful theory of innovation’, Research Policy, 6(1): 37–​76. OECD. 2007. Review of the South African Innovation Policy. Paris: OECD. Petersen, Il-​haam, and Glenda Kruss. 2019. ‘Promoting alignment between innovation policy and inclusive development in South Africa’, Development Southern Africa, 36(3): 351–​75. Petersen, Il-​haam, Glenda Kruss, Micahel Gastrow, and Patson C. Nalivata. 2018. ‘Innovation capacity‐building and inclusive development in informal settings: A comparative analysis of two interactive learning spaces in South Africa and Malawi’, Journal of International Development, 30(5): 865–​85. Phiri, Madalitso Zililo, Neo Molotja, Hlamulo Makelane, Takura Kupamupindi, and Catherine Ndinda. 2016. ‘Inclusive innovation and inequality in South Africa: A case for transformative social policy’, Innovation and Development, 6(1): 123–​39. Pogue, Thomas E. 2007. Mobility of Human Resources and Systems of Innovation: A Review of Literature. Cape Town : HSRC Press. Pouris, Anastassios. 2012. ‘Scientometric research in South Africa and successful policy instruments’, Scientometrics, 91(2): 317–​25. Roncaglia, Alessandro. 2017. A Brief History of Economic Thought. New  York:  Cambridge University Press. RSA. 1996. ‘White Paper on science and technology’. Department of Arts, Culture, Science and Technology, Pretoria.

488    Erika Kraemer-Mbula and Rasigan Maharajh RSA. 2002. National Research and Development Strategy’. Department of Science and Technology, Tshwane. RSA. 2008. ‘Innovation towards a knowledge-​based economy: A ten-​year plan for South Africa (2008–​2018)’. Department of Science and Technology, Tshwane. RSA. 2019. ‘White Paper on science, technology, and innovation’. Department of Science and Technology, Tshwane. Scerri, Mario. 1990. ‘R&D and the international competitiveness of the South African manufacturing sector’, South African Journal of Economics, 58(3): 212–​28. Scerri, Mario. 1998. ‘The parameters of science and technology policy formulation in South Africa’, African Development Review, 10(1): 73–​89. Scerri, Mario. 2009. The Evolution of the South African System of Innovation since 1916. Cambridge: Cambridge Scholars Publishing. Scerri, Mario, and Rasigan Maharajh. 2018. ‘The policy environment and policy options for the South African system of innovation’, in M. Scerri (ed.) The Emergence of Systems of Innovation in South(ern) Africa:  Long Histories and Contemporary Debates. Johannesburg: Mapungubwe Institute for Strategic Reflection, New African Publishers. Sharif, Naubahar. 2006. ‘Emergence and development of the National Innovation Systems concept’, Research Policy, 35(5): 745–​66. Sithole, Moses M., and Yasser Buchana. 2021. ‘Effects of innovation activities on employment growth in upper-​middle-​income countries with high unemployment rates’, Development Southern Africa, 38(3): 371–​390. Tomaselli, Keyan G. 2018. ‘Perverse incentives and the political economy of South African academic journal publishing’, SAJS, 114(11/​12): 1–​6. UNESCO. 1967. ‘Apartheid in South Africa’, The UNESCO Courier. March. UNESCO. 2015. UNESCO Science Report: Towards 2030. Paris: UNESCO. Walwyn, David. 2008. ‘A target for South Africa’s business expenditure on research and development based on industry structure’, South African Journal of Science, 104(9–​10): 340–​4.

Chapter 23

S ou th Afri c a a nd the Fourt h Industrial Revolu t i on Bhaso Ndzendze and Tshilidzi Marwala

23.1 Introduction The Fourth Industrial Revolution (4IR) refers to the oncoming ubiquity of technology, which will redefine society. As an industrial phenomenon, the 4IR is widely predicted (based on present trends) to transform what have been manual-​labour-​dominated sectors. Due to advances in robotics, sensors, and artificial intelligence, factories, farms, and mines are seen as being likely to be run autonomously, with little human input in terms of manual labour. Furthermore, advances in 3D printing are likely to eliminate the manufacturing process as it has been for centuries through mass transformation of the manufacturing process from being subtractive to being additive. From past developments and present trends, it is clear that South Africa’s path towards the 4IR entails household, private-​sector, and government incentives with prospects for new forms of employment, a resuscitation and diversification of the South African economy in a future-​proof manner, and more efficient service delivery. Risks also persist, however. Among others, these include growing unemployment, encroachment on decent wages, and pressure on the fiscus. The aim of taking advantage of the 4IR while minimizing its risks by government, businesses, and individuals, highlights its own imperatives in terms of urgent reforms in governance to address gaps in education, connectivity, infrastructure, energy, and inequality in access (along the lines of gender, wealth, and between rural and urban areas) which pose major challenges. In the first section, the chapter offers an overview of the concept of an industrial revolution, and briefly reviews past industrial revolutions. We then discuss more recent global industrial trends around 4IR technologies becoming more ubiquitous and converging with one another. The second section reviews the implications for the

490    Bhaso Ndzendze and Tshilidzi Marwala South African economy. In this regard, we look at the country’s economic prospects in the 4IR based on recent trends, academic and policy literature, and historical experiences with the first three industrial revolutions. In this section we ponder the implications for a decent wage, also conduct sectoral analyses of industries and the extent to which they have already been impacted by the 4IR, and prospects for further change under the 4IR. We also take stock of the adoption of various 4IR technologies by government, business, and consumers in comparative perspective. This entails a comparative analysis of South Africa’s performance vis-​à-​vis peer and globally leading countries, and proceeds to a discussion of the country’s 4IR aspirations at the highest level, with emphasis placed on the 2020 published recommendations of the Presidential Commission on the Fourth Industrial Revolution before turning to an analysis of potential obstacles and challenges ahead in the country’s publicized vision of an inclusive 4IR future. In the final section, the chapter concludes with a summary of the key issues covered in the chapter.

23.2  The Fourth Industrial Revolution Popularized in the English-​speaking world by British economic historian Arnold Toynbee (1852–​83)1 in his posthumously published lectures at the University of Oxford (1880–​81), the term ‘industrial revolution’ denotes the concept of a major shift in productive processes. While deindustrialization indicates a debilitation or collapse of industries due to underuse, industrial revolution refers to the introduction of new machines and systems of production which result in greater output and efficiency (Van Creveld 2010:  16). This is inevitably accompanied by social changes as well as new patterns of migration and class structures that emerge given the deep shifts to the world of work. Such a phenomenon began with the first industrial revolution (nominally referred to as the Industrial Revolution) in mid-​eighteenth-​century England. Before the Industrial Revolution, the industry was mostly manual, rural, and small scale rather than mechanized, urban, and large scale. As such, men and women moved across industries with relative ease; ‘there was much coming and going between manufacture and agriculture; and mines, furnaces, and cottage workshops suspended their activities in the summer and early autumn so that the workers could help with the harvests’ (Ashton 1967: 49). Why did the Industrial Revolution take place at the time it did? While the Industrial Revolution led to social change, it was itself the product of social antecedents; while there had been some ingenuity in the early eighteenth century, industry ‘had to wait until the idea of progress—​as an ideal and as a process at work in society—​spread from the minds of the few to those of the many’ (Ashton 1967: 57). This

1  To

distinguish from his historian-​philosopher and international relations scholar nephew Arnold J. Toynbee.

South Africa and the Fourth Industrial Revolution    491 was signified by the growth of capital in large enough quantities, which made it possible to construct large buildings and appliances. The process was all-​encompassing, with not only the invention of new gadgets, but complimentary innovations in agriculture, transport, trade, and finance which surged in concert. As T. S. Ashton puts it in The Industrial Revolution, 1760–​1830, the process was characterized by the application of science to industry, and ‘with a more intensive and extensive use of capital’ (1967: 142). For this reason, the Industrial Revolution is to be seen as a movement instead of a period of time or a singular event. A set of new, higher-​order innovations that marked a second industrial age are noted to have begun around the 1870s and proceeded until the middle of the twentieth century. The second industrial revolution (2IR) has come to also be known as the technological revolution. It was a period of rapid industrial development and growth for pre-​existing industries from the first industrial revolution, brought about by several key factors, primarily in countries such as Britain, Germany, France, Italy, Japan, and the United States. These inventions include mass electrification (with the emergence of new sources of energy such as electricity, petroleum, and oil); the expansion of long-​haul networks of railroads; the development of chemical synthesis that introduced synthetic fabrics, dyes, and fertilizers used in agriculture; widespread telecommunication over long distances with the electrical telegraph; and the assembly line (Lambrechts, Sinha, and Marwala 2021). The third industrial revolution (3IR) took shape in the 1960s with the advent of semiconductors and computing and got a further boost with the arrival of the Internet in the 1990s (Schwab 2017: 7). First postulated by the German government in its 2011 ‘Industrie 4.0’ (Industry 4.0) national strategic initiative, the observation that the world is in the process of entering a fourth industrial revolution has become a commonly held view among policymakers and economists. Its origins as an aspirational and projected future have resulted in it being seen as less of a pattern. Thus, a disputed concept (as we shall review in the next section), it nonetheless entails some defining features which represent new modes of production, consumption, and provision of services. The defining technologies of the 4IR, which have a production and service delivery component, include artificial intelligence (AI), 3D printing (also known as additive manufacturing), new materials (such as graphene), the Internet of Things (IoT), and blockchain. Within AI, the specialized areas of robotics, machine learning (the fastest growing area of AI with the claim to the majority of new patents and funding), and natural language processing (NLP) stand to accelerate the nature of automation in manufacturing, transform the process of invention, and erode human relevance in customer care respectively. Among these, it has been suggested that there is nothing in them which precludes human–​machine interaction. For its part, 5G is prognosticated (given its early deployment so far) to be a key infrastructural technology which will enable the realization of smart homes, self-​ driving cars, and smart cities (Guerrera and Cheein 2020). The convergences of these technologies thus provide the basis of the observations and predictions about the future ubiquity of these 4IR technologies. Thus, many identify that the exceptional and growing computational power and Big Data will become the distinguishing factor of

492    Bhaso Ndzendze and Tshilidzi Marwala the current age of AI from previous ‘AI winters’ in which technology experienced major setbacks in funding (roughly from 1974 to 1980, and again from 1987 to 1993). These technologies have already changed the manufacturing process, with countries such as China, South Korea, Japan, and the United States seeing rapid growth in smart factories in the last ten years. These are factories that consist of relatively lower numbers of personnel as they take advantage of automation. What most distinguishes them, however, is their ability to effectively communicate with one another in unison through sensors that lead to a cyber–​physical integration which are the bases of the IoT. In a 2019 estimate by PricewaterhouseCoopers, Machine learning (ML) alone was noted to represent about $2 trillion of today’s global economy and was predicted to reach some $16 trillion by 2030 (roughly 10 per cent of current world GDP) given growth in patent applications—​being the third fastest-​growing category of all patents granted and representing nearly 60 per cent of all new investment in AI. New avenues of AI are led today by the private sector. These include the likes of Amazon (advantaged by owning one the world’s largest data storage facilities, AWS); Google (through DeepMind and the reinforcement learning-​ based AlphaZero); Huawei (through Ascend 910 and Mindspore, which aims to uncover a new frontier in open access, AI-​based app development); IBM (through Watson); and Samsung (through a self-​professed focus on the IoT through Bixby and other facets of natural language processing and device-​driven AI).

23.3  The Fourth Industrial Revolution and the South African Economy Different scholars have problematized the degree to which the country, and the continent it is located in, ought to even consider the 4IR as relevant to them. In their article in Technology Analysis and Strategic Management, Ayetimi and Burgess ask, for example, ‘to what extent is the 4IR relevant to sub-​Sahara Africa where there is a large informal economy, limited public infrastructure, where technical skills levels are low, and advanced technology can be found in only a few sectors that are dominated by foreign multinational companies and staffed by expatriate workers?’ (Ayentimi and Burgess 2019: 641). Sutherland firmly asserts that ‘4IR is not the result of careful historical analysis, rather it is a flag to rally and a rhetorical device for those trying to create particular economic and commercial futures, hoping to ride waves of Schumpeterian economic disruption caused by “extreme automation and extreme connectivity” ’, asserting that ‘this has been combined with strong lobbying by manufacturers and the WEF, seeking to persuade governments to change their policies to support the deployment of 4IR technologies and to mitigate their adverse socio-​economic effects. The consequences of these neo-liberal efforts potentially destroy jobs, depress wages and increase inequality’ (2020: 233). A similar line of argument is adopted in c­ hapter 5 of the NPC’s 2020 draft report ‘Digital Futures: South Africa’s Digital Readiness for the Fourth Industrial Revolution’.

South Africa and the Fourth Industrial Revolution    493 However, in all of these texts there is a recognition that the onset of the roll-out and adoption of the 4IR is an inevitability precisely because of its alignment to powerful interests in the form of large corporations of the Global North and the East. Indeed, these critiques are cognizant of the 4IR and understand it instead as a harbinger of inequality, rather than an imagined or overhyped phenomenon. Moreover, some go so far as to see it as presenting opportunity for South Africa and the broader continent. For example, a 2019 paper notes that ‘the 4IR presents an important and valuable opportunity to drive social and economic growth and leverage development across nation states, regardless of their location or state of development—​and this applies particularly to the sub-​Saharan African region’ (Ayentimi and Burgess 2019: 641). Others have included suggestions that South Africa’s lack of a steady supply of electricity, among others, could be an opportunity to leapfrog into alternative sources of energy, with some highlighting that 5G is in any case set to bypass the current forms of connectivity which are lagging. Indeed, the latter point is commonly made with much emphasis on the notion that if the country is not an active participant, it could find itself retrospectively fitting foreign-​ developed technologies into its domestic contexts. Moreover, many see opportunities for more efficient governance, with the 4IR being potentially utilized for better service delivery and for the realization of smart cities and smart villages. For example, 4IR technologies such as AI and blockchain could be utilized to curtail the loss of agricultural produce in transit, as high as 90 per cent in some African countries to below 20 per cent, and thus assist small-​scale rural farmers. On the other hand, sociologist Grace Khunou argues that 4IR technologies can bridge the fatherhood gap which has come about in South Africa with the advent of the migrant labour system and other forms of absenteeism through enhanced connectivity, though she highlights the lack of affordability for the many (Khunou 2018). Ayentimi and Burgess ultimately argue that its global origins mean that ‘the 4IR has the potential to impact on all industries and all nations, regardless of their location or state of development’ (Ayentimi and Burgess 2019: 646). On the other hand, in their 2019 article (‘Review of Preparedness of Rural African Communities Nexus Formal Education in the Fourth Industrial Revolution’) in the South African Review of Sociology, Uleanya and Ke find that South Africa is ‘not preparing for the envisaged industrial revolution like other western nations; rather, the focus remains on issues such as decoloniality, decolonization and glocalisation’ (p. 91). History indeed casts a long shadow and this is to be expected, given that South Africa’s industrial prospects are shaped by its historical experience with industrialization. This is briefly considered below. Emanating from such a historical review are the related and overlapping issues of domestic innovation (and thus innovation policy, including the relative merits of openness compared to protectionism), inclusivity, economies of scale, and human capacity development. The general industrial history of the country is summarized by Carien du Plessis and Martin Plaut as follows: ‘from the beginning of colonialism in the seventeenth century until late in the nineteenth century, the world regarded South Africa as not much more than an outpost on the way to somewhere more useful’ (du Plessis and Plaut 2019: 111). It

494    Bhaso Ndzendze and Tshilidzi Marwala was regarded as ‘a place of agricultural output’ (and conflicts). This underwent a transformation in the late 1800s with the discovery of diamonds and gold in Kimberley and the Witwatersrand basin. Subsequently, the country, united after 1910, saw roughly four distinct phases. These entailed basic industrialization after the First World War, secondary industrialization after the Second World War, ‘growth and then stagnation under apartheid’, and the developmental policies under the African National Congress which have seen mixed results typified by growing unemployment. From 1912 to 1939, the country saw value-​added private manufacturing rise from £8 million to £53.8 million. This was in large part due to industrialization policies put in place in the interwar period (1919–​39) during which the country’s governments had enacted the establishment of the Electricity Supply Commission (Eskom) in 1923, the privately-​owned AE&CI (African Explosives and Chemical Industries) in 1924, and the now-​privatized ISCOR in 1928. The 1932 Carnegie Commission of Investigation on the Poor White Question in South Africa identified the existence of 300,000 poor whites mostly in the Afrikaner community (out of 1.5 million Afrikaners). The government sought to close this gap by creating jobs through the expansion of secondary industries. Thus, whereas in 1919 agriculture and industry were roughly equal in terms of output, by 1939 industrial output was about two-​and-​a-​half times that of agricultural output. This was greatly aided by the tariff regime, put in place in 1925. The onset of the Second World War also presented an opportunity for further industrialization and manufacturing. As the British South Africans, who were the country’s commercial base, went to fight in the Second World War, they left open vacancies which were filled by the increasingly urbanizing African population and manufacturers responded favourably to the opportunities created by wartime shortages of imported products and the demands of the war economy (du Plessis and Plaut 2019: 112). This saw the volume of outputs double, with the number of people employed in private manufacturing growing from 440,000 in 1948 to 1.16 million in 1971. In the post-​Second-​World-​War period, the country saw continued growth (by an average annual rate of 4 per cent until the 1970s, and by as much as 5.8 per cent in the 1960s [Freund 2019: 191]) and diversification (though not to the same levels as other contemporaneous ‘miracles’ such as Germany and Japan). However, South Africa’s inward orientation and semi-​isolated global status were its downfall: it caused a limited reach in the international market, for example more advanced military weaponry was being produced elsewhere, all while domestic demand for technology was confined to the well-​off white minority as a result of the country’s apartheid policy. With neither extensive international nor domestic markets, the country’s negative incentives depressed the expansion of whatever capacity for innovation might otherwise have been commercially exploited (as evidenced by the innovative oil-​from-​coal techniques pioneered by SASOL in 1950, the tellurometer in 1954, and nuclear medicine from 1965 [Majozi and Marwala 2020]): In general, South African manufacturing firms had no reputation for genuine product innovation. Engineers were known for their ability to modify products

South Africa and the Fourth Industrial Revolution    495 and processes, and South African manufactured products were and continue to be made under license to intellectual property rights holders in the advanced industrial economies. This is not surprising as the nature of South Africa’s protectionist regime encouraged licensing and copying for the domestic market, not world class innovation. (Hirsch 2005: 148)

Du Plessis and Plaut similarly observe that: Displacement of more complex intermediate and capital goods was severely constrained by several factors, including the small scale of South Africa’s domestic market, the lack of necessary skills and technological capabilities, and the inability to raise protection of these industries sufficiently to displace imports without doing great damage to other domestic industries. The desperate poverty of the black urban population also effectively closed off another possibility touted by some: ‘inward industrialisation’. (du Plessis and Plaut 2019: 115)

How much can South Africa’s economy be said to be meeting the prerequisites of the 4IR or already observing large-​scale use of 4IR technologies and what are the implications for the fiscus and decent wages? In line with this, what has been the policy response of the government? These questions are reviewed below.

23.3.1 Decent Wages, Industrial Maturity, and the Fiscus: Economic Effects and Implications of the 4IR in South Africa We have noted the thinly spread benefits of second and third industrial revolutions manifest in the problems of policy. These are compounded by globalization and the incapacity of South African firms to compete without the protectionist policies of the apartheid era. This has been noted to be a major cause of inequality during apartheid and of growing unemployment in the post-​apartheid era. There is, therefore, cause for some concern regarding the country’s capability to meet certain prerequisites in order to be viable in the 4IR. The notion that a country needs to meet certain prerequisites holds substantial purchase in the literature (Hantraisa, Allin, Kritikos, Sogomonjan, Anand, Livingstone, Williams, and Innes 2020: 2). South Africa’s own National Planning Commission argues that: There is a core set of indicators that all organisations require, and all have identified the need for demand-​side data (via nationally representative surveys) to supplement administrative supply-​side data and the limited set of ICT indicators from the census and the annual national household survey conducted by Stats-SA. Historically, all data has been collected on an ad hoc basis when resources could be secured. This needs to be regularised, standardised and institutionalised and framed within the context of an open-​data policy that safeguards privacy rights and makes anonymised

496    Bhaso Ndzendze and Tshilidzi Marwala data enable the free flow of information required for more effective planning by government and service delivery entities, for private use by entrepreneurial and innovative enterprise. (NPC 2020: v)

A mainstay among these criticisms is the lack of a stable supply of electricity, given the country’s chronic shortages known as load-​shedding. This criticism is justified: electricity is crucial for the AI value chain, as it entails the use of energy to cool data centres (with some 1–​1.5 per cent of global electricity consumption used to cool computers the world over [Energy Innovatio 2020]). Added to this is the observation that a considerable (albeit shrinking) portion of the country (43.7 per cent of the population in 2020) does not have access to the Internet; the country’s higher-​than-​average connectivity costs (at 148 out of 228 countries on mobile data prices); and the rural–​urban divide in digitization, for example only about 1.2 per cent of households in the rural areas of Eastern Cape have access to the Internet. Other provinces are not much better. For example, the KZN has a rural Internet penetration rate of 1.1 per cent, followed by North West (0.9 per cent), and Limpopo (0.5 per cent) according to the most recently available General Household Survey (StatsSA 2015). Most rural regions are usually covered by one of the two dominant service providers—​Vodacom and MTN—​and are not the beneficiaries of competitive service offerings due to the absence of Cell C and Telkom, who are typically less expensive. ‘As in the case of the mobile network, competition in the fibre market is largely in the municipal areas and on the main transmission routes, with most residential areas remaining without fibre connectivity to the home, and with competition only on the main intercity transmission routes’ (NPC 2020: iv). This raises issues about fairness, however; rural customers are essentially paying the same prices as their urban counterparts for less service delivery. In this regard, the benefits of the 4IR such as faster transmission at the behest of 5G, as well as access to learning opportunities, are substantially stacked in favour of urban dwellers. In this regard, prerequisites are nuanced across regions within the same country. Other realities in the country, however, mean that the 4IR stands to not only ‘miss’ those in disadvantaged areas, but also subject others, particularly workers, to vulnerabilities.

23.3.1.1 Impact on Decent Wages The role of technology as a long-​term labour substitute has long been a concern. In the early twentieth century prominent economists, including John Maynard Keynes (1962: 358), were pondering the implications of automation. A WEF Report, The Future of Jobs, has stated that the coming years will see new work roles emerging, some work roles remaining stable, and many becoming redundant. A 2018-​published study by the Department of Arts and Culture, Nelson Mandela Metropolitan University and the South African Cultural Observatory noted that ‘by 2030 over 2 billion jobs as we know them today will have disappeared, freeing up talent for many and new 4IR fledgling industries, fundamentally changing the nature of work’, and estimates that ‘60 per cent of jobs that will exist towards the end of 2030, have not yet been conceived of or invented

South Africa and the Fourth Industrial Revolution    497 yet’ (Adendorff, Lutshaba, and Shelverp 2018: 2–​3). This comes in a context in which the push for the minimum wage has been increasingly vocal. The National Minimum Wage for South Africa Research Initiative suggests that ‘in South Africa, the level of economy-​ wide output would be 2.1% higher with a national minimum wage (beginning at levels between R3 500 and R4 600) and the average GDP growth rate is projected to be 2.8%–​ 2.9% instead of 2.4% without a national minimum wage’ (2016: iii). One of the distinct trends according to the available data is the growth in average incomes in the manufacturing industry in South Africa. Using the 2017–​20 period as a snapshot, we find that the country has seen consistent growth in wages, which have been in keeping with the inflation rates of 5.27 per cent, 4.62 per cent, and 4.13 per cent.2 However, the onset of COVID-​19 has caused a decline in manufacturing wages. A major force behind this rise in incomes has been collective bargaining by workers (of which 32 per cent are low earners [Isaacs 2016: iii]), which is threatened by automation. Automation will not only mean fewer workers, but also decreased bargaining power for the few remaining workers. Thus, government will be faced with the option of either allowing companies to operate relatively unimpeded (through relaxed labour laws) or prioritizing workers, possibly to the detriment of retaining (foreign and domestic) corporations. Global trends indicate that the former is more likely, given that ‘the 4IR is rewriting the rules of manufacturing because low-​cost labor is not an effective strategy for attracting manufacturing investment as the cost of automation plummets,’ and ‘the 4IR facilitates the start of a trend toward reshoring manufacturing back to the rich world’ (Lee, Wong, Intarakumner, and Limapornvanich 2020: 409). Doing otherwise would possibly lead to a reduced fiscus. In section 23.3.1.2, we note that among the principal concerns for directors is what they see as a lack of skilled workers in the country. This is true in various industries, as we shall see, with corporations looking to cut personnel costs and pursue the alternatives offered by the 4IR technologies. In Figure 23.1 we compare the income levels in industries related to the 4IR (in black, first six bars) and those which are not (in grey, last nine bars). Those in black are in the robotics industry in particular. We note that the lowest paid among the sample (design engineers) earn substantially higher than the lower-​paid non-​4IR employees (retail store manager) and indeed above the average in the non-​4IR careers. There is a high barrier to entry into these jobs, however, as these are highly trained specialists, requiring tertiary education (itself with very high cost inputs). To be sure, there are areas of the 4IR which have a low barrier to entry and which require little formal education. One of these is annotation. This refers to the labelling of the physical objects in the physical world into a digital interface so that they may be used in algorithms by AI. This is a tedious but necessary task which requires punctilious human inputs. Engineers and company routinely outsource this work to low-​income

2 

Trading Economics (2020), ‘South Africa average monthly wages in manufacturing’ [online] https://​ tradingeconomics.com/​south-​africa/​wages-​in-​manufacturing (last accessed 15 September 2020).

498    Bhaso Ndzendze and Tshilidzi Marwala Design Engineer

284000 318000

Project Engineer Production Manager Robotics Researcher

600000 597000 605000

Quality Manager Software Engineer

R731,000

Supply Chain Management E-Commerce Software Development City Government

509000 402000 380000

Government Employees National Average Salary

377,000 268644 204072 162,000 162000

Average Manufacturing Salary (Anuualised) Database Developer Quality Assurance/Quality Control Inspector

108,000 90000

Personal Assistant Retail Store Manager

62000 0

100000

200000

300000

400000

500000

600000

700000

800000

Figure 23.1  4IR-​related and non-​related industrial wages in South Africa (annual, in ZAR) Source: Trading Economics, ‘South Africa Average Monthly Wages in Manufacturing’ (2020) https://​tradingeconomics. com/​south-​africa/​wages-​in-​manufacturing (last accessed 15 September 2020). Chart by authors. Data sourced from Salary Data and Career Research Center (South Africa) and the South African Department of Labour.

countries precisely due to this tedium. This could be a possible job creator. However, there are three drawbacks in this regard. The first is that the process requires a smartphone with connectivity. Second, it favours those who are English speaking. Third, it is not sustainable in the long run as the process is often project-​based and works towards completion; as AI becomes more intelligent through ML, the extent of human annotation may also be less required as AI increasingly learns unsupervised.

23.3.1.2 Industrial Maturity Many South African firms have engaged in strategies pertaining to the 4IR. Our review of the literature depicts an uneven landscape. Statistics South Africa (StatsSA) observed that in 2019 manufacturing constituted slightly over 13.53 per cent of the country’s GDP (or its fourth largest industry). Studies into this have posited that this had previously been higher but was declining due to trade openness and the subsequent impact of imports. Hirsch interjects with some circumspection, as he regards the extent to which the country engages in substantial manufacturing as overstated. He highlights that a large part of South Africa’s ‘manufactured’ products are basic metals, such as iron, steel, and aluminium as well as basic chemicals, wood pulp, and paper that represent manufacturing only in the narrowest sense according to international classifications. The processes rather entail final stages in the extraction of minerals than preparatory beneficiation, and South Africa is reliant more on the advantage brought on by its climate and abundance of natural resources ‘than on acquired skills or expertise’ (Hirsch 2005: 117). Noticeably, as recently as 2019, the top five sources of employment in the manufacturing sector are basic metals, fabricated products, and machinery equipment (at 22 per cent), followed by food, beverages, and tobacco (at 20 per cent), and coke, refined petroleum, and nuclear fuel (at 11.29 per cent), as well as wood and wood

South Africa and the Fourth Industrial Revolution    499 products (11.29 per cent), and transport equipment (9.83 per cent). Overall, within the country’s manufacturing sector, food and beverages constituted a share of 26 per cent, followed by petroleum and chemical products (at 24 per cent), basic iron and steel (at 19 per cent), wood products (at 11 per cent), and textiles (at 3 per cent). Advanced and semi-​advanced manufacturing constituted slightly over 11 per cent, with motor vehicles (including parts and accessories) at 7 per cent, followed by electrical machinery and communications and professional equipment at 2 per cent each. In the 2010–​20 period, imports as a percentage of domestic manufacturing sales have hovered above 59.86 per cent and reached a peak of 70.47 per cent in 2013. Automation remains a minimal concern for directors, however, indicating a favourable view of this automated future. Rather, they emphasize the lack of specialized skill among South African workers. The Director Sentiment report, an annual survey of directors in the private sector (n = 475) active since 2016, notes in its 2019 edition that the key concerns for directors are shortage of skilled labour in the country, poor infrastructure, and policy uncertainty. Only 3 per cent of those surveyed stated that they were most concerned by automation, while another low 3 per cent expressed concern over high costs of IT infrastructure when expressing the main business challenges currently facing their industry (Institute of Directors South Africa 2019: 22). As seen, high costs of data connectivity mostly impact rural dwellers. Promisingly, ‘the communications market has grown significantly in this uncertain environment, with tens of billions of Rands annually in private investments in the extension of fibre networks and upgrading of mobile networks to support the roll-​out of data services. Despite the introduction of a horizontal licensing regime over a decade ago, the market remains structured around several integrated network and services operators. MTN and Vodacom, in particular, dominate the mobile telecommunications market, with a total market share of 78 per cent’ (NPC 2020: ii). In July 2020, MTN launched 5G coverage in 100 sites across multiple areas in Johannesburg, including Randburg, Bryanston, Fourways, Lonehill, and Fairland (its HQ), another in Bloemfontein (at the University of the Free State) and one in Cape Town (in Bloubergstrand) while Rain was the first network to launch 5G in South Africa. Rain is differentiated from MTN in its use of fixed-​wireless applications instead of mobile (McLeod 2020). On the other hand, calls for allocation of spectrum by the Independent Communications Authority of South Africa (ICASA) remain vocal, with ICASA having scheduled an auction for such an allocation in late 2020.

23.3.1.3 Sectoral Analysis In the main, there is no universally adopted measure of whether a country meets a hypothetical 4IR threshold or scale and indeed such an index does not yet exist although the most widely regarded is a 10 per cent critical mass threshold for every technology. But one of the findings of the South African Presidential Commission on the Fourth Industrial Revolution (PC4IR) is the country’s relative lack of measurement despite much capacity for data-​gathering by government agencies and private-​sector players. Moreover, some discernible proxies exist. Below we briefly review patterns in some key

500    Bhaso Ndzendze and Tshilidzi Marwala sectors before turning to studying trends that may be termed 4IR indicators among peer countries.

23.3.1.3.1 e-​Commerce While e-​commerce is not necessarily exemplary of the 4IR (Amazon, for example, was established in the 1990s), some aspects of its architecture are complimentary with numerous 4IR technologies such as blockchain (with payments via cryptocurrencies and tokens, for example), the IoT (especially the ability to track one’s order in real-​time), and Big Data. The latter of the three, in particular, indicates the accumulation of data which can be used to build digital profiles of users and thus inform recommendation algorithms. This digitization of South African social and business life indicates a long-​ term shift away from a one-​size-​fits-​all model of general retail towards individually customized retail, with implications for jobs and future skills requirements. Retail represents some 15 per cent of South Africa’s GDP. South Africa’s online retail market was around Rand 14 billion in 2019, representing some 1.4 per cent of its total retail and 18.43 million e-​commerce users (Pillay 2019). Thus, the e-​commerce market in South Africa has a high growth potential. The top three purchases consist of clothing, books, and beauty products. A  2019 study indicates that when South Africans are making online purchases, they often end up spending more than they commonly do in brick-​and-​mortar stores. Crucial to this is the incentive of free deliveries after reaching a certain price threshold (Davis 2019). Moreover, the majority of these online purchases are on customers’ mobile phones: As much as 18% (out of 29%) of South African internet users bought something online via mobile phone in the past month (We Are Social, 2018), so having a mobile-​ friendly online store is important. Since many South Africans are using their mobile phones to shop online, Visa has realised that one must make it easy for them to shop online on a small mobile screen and hold the phone in one hand. (Davis 2019)

‘The three most popular online shopping categories for South African consumers who shop online were clothing/​apparel (53 per cent), entertainment/​education (digital/​ downloadable) (51 per cent), and event tickets (51 per cent)’ (Davis 2019). When shopping online, consumers mostly prefer local e-​commerce platforms like Takealot (South Africa’s largest online retailer with over 10,000 parcel dispatches per hour [Malinga 2019]) and Bidorbuy. Broadly, 84 per cent of purchases are from these local platforms (Mkhosi 2017: 2). The most popular platform for international purchases is Amazon. In a nod to the growth potential of e-​commerce for small and medium enterprises in the country, the government introduced the Electronic Communications and Transactions Act in 2002 in order, among other objectives, to facilitate and regulate electronic communications and transactions and promote human capacity development (skills) in this budding sector as well as prevent forms of abuse of information systems (Republic of South Africa 2002:  1). Nevertheless, the government is still

South Africa and the Fourth Industrial Revolution    501 challenged in the regulation and taxation of externalized funds due to the difficulty of tracking financial transactions. In this regard, ‘one of the responses emanating from the South African Reserve Bank is incentivising merchants to adopt 3D Secure (under the auspices of the Payment Association of South Africa; a division of the SARB)’ (Mkhosi 2017: 2).

23.3.1.3.2 Banking The banking sector in South Africa exhibits contradictory trends towards the 4IR: digital inequality among its customers on one hand and incentives for cost-​cutting through digitization on the other. Recently, South African banks have been cutting jobs in pursuit of lowering costs and keeping up with slow economic growth in the country and new competition in the industry from branchless (digital) competitors such as Bank Zero, TymeBank, and Discovery. In 2019, the SARB and the Intergovernmental Fintech Working Group initiated a programme to look into the appropriateness of policies and regulatory regimes as a result of the fintech innovation (Marwala 2020: 120). One of the key issues identified concerns the effects of income inequality and what these effects may mean for the increasing digitization of banking. A similar conclusion was drawn by the Centre of Excellence in Financial Services: The country’s significant potential for digital innovation must be considered alongside concerns of whether this will be exclusionary, and whether the transformation will enhance or diminish domestic value creation.

Additionally, the report observes, digital banking still serves a ‘niche, relatively affluent and financially savvy consumer market,’ in spite of the growing penetration of smartphones. The country is still held back by considerable financial illiteracy (Marwala 2020: 120–​1). At the same time, however, the gloomy prospects for banking, even before COVID-​19, has made investors weary of the sector such that banks’ valuations declined by between 15 per cent to 20 per cent in 2019. As a consequence, banks will have to adapt. Fourth industrial revolution technologies, particularly AI, hold substantial promise (Marwala 2020: 121). In this regard, a 2019 PricewaterhouseCoopers report on the major banks in the country notes that ‘staff costs continue to comprise the majority of overall group costs, reflecting both the inflationary environment as well as the demand for critical talent in response to increasing specialization in the areas of risk, compliance and IT’. According to sectoral analysis, by the year 2030 AI technologies will see banks shed operating costs by some 22 per cent. A notable trend in South Africa has seen increases in IT expenditure, ‘as the banks grow their direct investments in their applications and systems infrastructure towards digitising their platforms’ (PWC 2019: 1). A transformation is underway in South African banking.

23.3.1.3.3 Construction Given the decline seen in the industry in the years since 2017 (StatsSA 2019 Q4) and across the world (with a McKinsey Global Institute study noting that it was twice as

502    Bhaso Ndzendze and Tshilidzi Marwala expensive in 2017 to construct a building than in 1970), a 2018 University of Johannesburg study found that construction professionals are increasingly willing to adopt robotics and construction automation. The study came to the conclusion that automation and the introduction of robots in the industry ‘would have positive effects on the delivery of the construction project by increasing quality of the construction product, enhancing supervision, improving working conditions, cost-​effectiveness and it will also reduce construction accidents if adopted’. On the other hand, we note that the use of robotics will replace some manual labour, for example workers engaged in tasks such as excavation will be affected by the deployment of robots. The report concludes that ‘workers will need to be retrained to upgrade their skills to avoid replacement by automation and in fact to be the operators of the machineries. Training and education may be required for the construction professional that are threatened by the implementation of automation to become familiar with the use of new technology.’ A recent start-​up, Build Robotics, has carved a niche of about Rand 1.3 billion worth of signed contracts as of 2019, with autonomous excavators that have lidar (a remote sensor method), GPS, and Wi-​Fi to map and navigate surroundings (Business Insider 2019).

23.3.1.3.4 Mining Mining conglomerates have been retrenching workers due to a trade-​off between profitability and ‘bloated’ workforces. There is some debate in this industry as to the desirability of 4IR technologies. Some note that automation, and technologically advanced machinery, can emerge as an alternative to human labour; although start-​ up costs are high, the running costs of machinery are far cheaper than wage-​earning workers. Recently, the mining sector in South Africa has been facing challenges associated with cutting back large numbers of workers. The platinum sector has felt the highest impact. In 2018, Implats, the world’s second-​largest platinum miner, stated that some thirteen thousand employees stood to lose their jobs within the following three years. Lonmin is also set to retrench roughly the same number of workers. The same trend is noted in the gold sector. In late 2018, Gold Fields retrenched up to 1,560 employees (or 30 per cent of its workforce) at its South Deep mine in Gauteng. In some important ways, however, the mining sector is fundamentally challenged in the extent to which it can adopt 4IR technologies. Particularly given the high temperatures (at around 50oC), humidity (which causes the technologies to perform more slowly and their materials to rust), and vibration levels in underground conditions, semiconductor devices, such as an AI-​powered robot, cannot operate efficiently or for very long (Marwala 2020: 54).

23.3.1.4 Policy Imperatives and the Fiscus: The South African Government and the 4IR State involvement is one of the key accelerators of industrial change. This has been evident in past industrial revolutions and elements of it are evident in the current onset of the 4IR. This was demonstrated, for example, by the developmental experience of Japan, the ‘four Asian tigers’, and China and their resulting East Asian miracle (Lee,

South Africa and the Fourth Industrial Revolution    503 Wong, Intarakumner, and Limapornvanich 2020: 408). Similarly, the United States and China have close ties to technology companies, who in turn provide military and economic advantage. The early discourse on the 4IR was led by the government of Germany under the umbrella of ‘Industrie 4.0’ which was subsequently underpinned by the ‘Action Plan High-​Tech Strategy 2020’, launched by the German Federal Government in 2010 to bring together the private sector, labour, research institutes, and even political organizations. This is especially crucial given the unchartered ethical, security, and legal territory presented by smart factories and its AI-​saturated environment (Banthien 2020). COVID-​ 19 has generally accelerated the adoption of 4IR technologies by governments world-wide. In a study for the journal Contemporary Social Science, Hantrais et  al. gather evidence from different areas about the impacts of COVID-​19 and show how ‘the pandemic supported changes in data collection techniques and dissemination practices for official statistics, and how seemingly insuperable obstacles to the implementation of e-​health treatments were largely overcome . . . the pandemic accelerated the uptake of digital solutions’ (Hantraisa, Allin, Kritikos, Sogomonjan, Anand, Livingstone, Williams, and Innes 2020: 1–​2). The South African government has similarly conceptualized its own role in the 4IR. The government of South Africa reviews the country’s readiness for 4IR and its recommendations for going forward are reviewed below before proceeding to a peer analysis.

23.3.1.4.1 PC4IR Recommendations and Potential Hurdles President Cyril Ramaphosa established the Presidential Commission on the Fourth Industrial Revolution (PC4IR) in 2018, which appointed commissioners in May of 2019. The PC4IR presented its report in August of 2020, following two years of work characterized by hundreds of consultative sessions and stakeholder engagement traversing various aspects of the South African economy, and research undertaken to diagnose the country’s readiness for the 4IR and to identify opportunities as well as mitigate risks. In its diagnosis, the Commission’s report argues that: The high-​technology industries in which the country is seriously underperforming are those of artificial intelligence, Blockchain, virtual/​augmented reality simulation environments, automatic data-​processing machines, electrical and electronic goods, biotechnologies, storage/​transmission, advanced materials, advanced sensor platforms as well as medicinal products and pharmaceuticals. (2020: 98)

On the basis of these identified gaps, the PC4IR has made eight core recommendations. These include investing in human capital, establishing an AI institute, establishing a platform for advanced manufacturing and new materials, securing and availing data to enable innovation, incentivizing future industries, platforms and applications of 4IR technologies, building 4IR infrastructure, reviewing, amending, or creating policy and legislation, and establishing a 4IR Strategy Implementation Coordination Council in the Presidency.

504    Bhaso Ndzendze and Tshilidzi Marwala In the March 2019 Skills Development Summit, the Department of Higher Education and Training (DHET) noted that the 4IR ‘would have a significant impact on the future skills that South Africa requires, as well as how the country prepares to meet that skills demand through retraining people for future jobs’. Encouragingly, the DHET responded by supporting large-​scale youth entrepreneurship programmes that took in some 1.1 million people in twenty-​one Setas (sector education and training authorities) across the country, funded by some Rand 63 billion accumulated in the preceding five years from the Skills Development Levy—​an additional 330,000 learners were financed by the National Skills Fund. There has been no impact assessment on these programmes, however. Indeed the DHET, through deputy minister Buti Manamela, admitted that these initiatives were not extensive enough. In terms of employment, the deputy minister noted that the number of learners who were successfully absorbed into full-​time work following their education was higher in modes such as apprenticeships, learnerships, and internships while formal unemployment continued to rise. The government has also expressed awareness of the 4IR pioneers who have left the country, in addition to widespread concern that the PC4IR’s recommendations may not materialize (Jenkinson 2020). These have been informed by past experience. The National Integrated ICT Policy has long observed that for there to be a ‘vibrant and inclusive knowledge economy’ in the country there needs to be affordable access to communication (equity); increased accessibility of services, devices, infrastructure, and content to all citizens (accessibility), and proper data governance ensured (user protections). In turn, the NPC terms these ‘preconditions of an equitable digital economy and society’ (2020: ii). The issue of allocation of spectrum has already been mentioned. It is indeed exemplary of the characteristic delay in digital policy by the government. Through SA Connect, the government sought to realize 90 per cent broadband access in the country by 2020 and 100 per cent by 2030 as part of the NDP and placed priority on connecting all schools, health centres, post offices, and Thusong Centres. ‘However, progress with the big broadband push has been limited and characterized by various uncoordinated initiatives’ (Mzekandaba 2019). This 2012 initiative was followed by the adoption of the National Integrated ICT White Paper in 2016. The policy paper emphasized private-​sector investment in the roll-out of seamless ‘critical infrastructure and services required for a modern economy’ (RSA 2016). The implementation of this plan was delayed, however. At the same time, Connect SA’s initially ambitious goal was scaled down to become a connectivity project led by Broadband Infraco (termed an undercapitalized SOE) and with the Universal Service and Access Agency of South Africa providing limited connectivity (NPC 2020: ii). The NPC review places politics (especially the rapid turnover of ministers and directors general and the splitting of the Communications Ministry into two departments) at the centre of these failures, delays, and shifting targets. With data extracted from Antonysamy (2019), PricewaterhouseCoopers (2019), Standard Bank (2020), Mzakandaba (2020), Presidential Commission on the 4IR (2020), Brothwell (2020), and Traxcn (2020) we take stock of the prevalence of adoption

South Africa and the Fourth Industrial Revolution    505 of various 4IR technologies in South Africa up to 2020. What is evident from the analysis is that the 4IR—​as a vision and as a set of technologies that can lead to efficiency—​ has seen uneven adoption, with more adoption towards AI and the IoT, and low adoption of blockchain and 3D printing, and with more leadership by the private sector and, by extension, consumers, than by government. Additionally, Big Data generated from South Africa has been accumulated and commercialized mainly by foreign multinational corporations such as Oracle, Uber, Google, Facebook, and their affiliates (e.g. Instagram and WhatsApp). However, we can also deduce that the 4IR is engendering increasing interest from government, business, and consumers, but that they are yet to cross the nominally used 10 per cent adoption threshold set by industry watchers (see WEF’s Deep Shift report, 2015). In their survey-​driven study published in the journal Political Research Exchange, Dermont and Weisstanner (2020: 1) find that ‘technology entrepreneurs have endorsed a universal basic income (UBI) as a remedy against disruptions of the work force due to automation.’ Noting the same three years earlier, Vegter inferred that this is perhaps driven by a desire to create an insulating effect by Big Tech and automating corporations as ‘those who create automation technology want to avert a popular or regulatory backlash to what they’re selling’ (2017). On the other hand, David Autor of MIT argues that texts on automation which emphasize the future loss of employment ‘ignore the strong complementarities that increase productivity, raise earnings and augment demand for skilled labour’ (Autor 2015: 1). Frey and Osborne provide a quantifiable estimate of the number of jobs they anticipate as susceptible to automation: roughly 47 per cent (Frey and Osborne 2013: 1). This could provide a guide for policymakers and allow for targeted interventions. Disadvantageously, these studies are conducted in the European Union and the United States and data are scant on South Africa and the rest of Africa. This is an important area for future research. While the government of South Africa (driven by Minister of Social Development Lindiwe Zulu) has recently revived considerations of providing a UBI, such debates (which have their origins as early as Mandela’s presidency) have been revived by COVID-​19 and not by automation. Originally planning to publish implementation plans in October 2020, the Ministry of Social Development has deferred such discussions. Globally, not a single country pays its citizens an unconditional UBI but it has been observed that ‘the economic crisis caused by the coronavirus has put the idea back on the table, even in fiscally conservative countries’ (Toyana 2020).

23.3.1.4.2 Peer Country Analysis The World Economic Forum’s Global Competitiveness Index (GCI), which annually tracks changes in countries’ performances across twelve pillars which contain a further 114 sub-​pillars, is notable for its utility in comparisons. Indeed, its underlying methodology rests on the assumption that it is not a country’s policies and their results in isolation which matter, but the policies being taken by other countries as well and thus the relative placement of one country vis-​à-​vis almost two hundred others. The country’s performance compared to its self-​identified peer countries, the BRIC countries, shows that South Africa is in fact outperforming them in terms of the WEF’s operationalized

506    Bhaso Ndzendze and Tshilidzi Marwala capacity for innovation. However, the country is unable to match these countries for scale in terms of exploiting and commercializing patents compared to the BRICS, while also being unable to match the more established industrialized countries in terms of private-​sector investment as well as government expenditure on R&D. Moreover, they are more advantaged by their scale in line with Krugman’s theory. In 2020, the head of the Office of Digital Advantage at the CSIR, Akhona Damane, estimated that ‘South Africa spent 10 percent of its GDP on ICT goods and services, most of which are imported’ (see Jenkinson 2020). The relatively smaller members, such as Brazil and Russia, are able to compensate with their comparative strengths in energy, which is crucial for the operation of hyperscale data centres (Carnegie Endowment 2019). While Russia relies on its oil abundance, Brazil on the other hand achieved energy self-​sufficiency in 2007 from renewable energy resources and from the exploitation of offshore minerals since 2016. Globally, leading states in terms of innovation also tend to invest more in R&D. Germany, the United States, Japan, Israel, and South Korea have a gross expenditure on R&D of more than 3 per cent of national GDP. The comparable national expenditure figure in South Africa stands at about 0.7 per cent as of 2019, which falls short of the government’s own target of an increase to 1.5 per cent.

23.4  Conclusions and Prospects It is clear from the preceding analytical reviews that South Africa’s path towards the 4IR is real, and driven by consumer, private-​sector, and government incentives. Moreover, it entails prospects for new forms of employment, a resuscitation and diversification of the South African economy in a future-​proof manner, and service delivery. On the other hand, such aspirations highlight their own imperatives in terms of the urgency of reforms in governance to address gaps in education, connectivity, infrastructure, energy, and inequality in access (along the lines of gender, wealth, and between rural and urban), and curbing the prospects of a brain drain, given the rapid advancement of other countries towards the 4IR. This has been given further urgency by the COVID-​ 19 pandemic, which has accelerated uptake of technologies and systems. Policy options over the next decade will increasingly be shaped by trends in automation as companies increasingly look to automation to be viable (and withstand competition by other, foreign companies), and as the government seeks to remain fiscally viable.

References Adendorff, Christian, Unathi Lutshaba, and Amy Shelver. 2018. ‘Policy implications of the 4th Industrial Revolution for the cultural and creative economy’. Department of Arts and Culture, Nelson Mandela Metropolitan University and the South African Cultural Observatory.

South Africa and the Fourth Industrial Revolution    507 African National Congress. 1992. ‘Ready to govern: ANC policy guidelines for a democratic South Africa’. Adopted at the National Conference on 28–​31 May 1992. Antonysamy, Felix. 2019. ‘Boosting blockchain adoption in SA’, Business Report, 8 January [online] https://​www.iol.co.za/​business-​report/​opinion/​boosting-​blockchain-​adoption​in-​sa-​18741416. Ashton, T. S. 1967. The Industrial Revolution, 1760–​1830. London: Oxford University Press. Autor, David H. 2015. ‘Why are there still so many jobs? The history and future of workplace automation,’ Journal of Economic Perspectives, 29(3): 3–​30. Ayentimi, Desmond Tutu, and John Burgess. 2019. ‘Is the Fourth Industrial Revolution relevant to sub-​Sahara Africa?’, Technology Analysis and Strategic Management, 31(6): 641–​52. Banthien, Henning. 2020. ‘Implementation of an Industry 4.0 strategy: The German Plattform Industrie 4.0’, European Commission, March. [online] https://​ec.europa.eu/​ digital-​single-​market/​en/​blog/​implementation-​industry-​40-​strategy-​german-​plattform​industrie-​40. Brothwell, Ryan. 2020. ‘Gauteng’s R500 billion plan to build a mega-​province—​including a new university and smart cities’, Business Tech., 8 March. [online] https://​businesstech.co.za/​ news/​business/​379477/​gautengs-​r500-​billion-​plan-​to-​build-​a-​mega-​province-​including-​a-​ new-​university-​and-​smart-​cities/​. Business Insider. 2019. ‘This start-​up construction company wants to tune-​up excavators with self-​driving tech’, Business Insider, April. [online] https://​www.businessinsider.co.za/​ built-​robotics-​brings-​self-​driving-​to-​construction-​2019-​4. Carnegie Endowment (Charlotte Stanton, Vivien Lung, Nancy Hanzhuo Zhang, Minori Ito, Steve Weber, and Katherine Charlet). 2019 ‘What the Machine Learning Value Chain Means for Geopolitics’. Carnegie Endowment for International Peace. https://​carnegieendowment. org/​2019/​08/​05/​what-​machine-​learning-​value-​chain-​means-​for-​geopolitics-​pub-​79631. Conigliaro, Andy, Joshua Elman, Jeremy Schreiber, and Tony Small. 1996. History of American Railroads. The Stanford University Danger of Corporate Monopoly Project, Stanford, CA. Davis, Louise. 2019. ‘What does the current South African e-​commerce landscape look like?’. Flicker Leap [online] https://​flickerleap.com/​south-​african-​ecommerce-​landscape/​. Dermont, Clau, and Weisstanner, David. 2020. ‘Automation and the future of the welfare state:  Basic income as a response to technological change?’, Political Research Exchange, 2(1): 1–​11. du Plessis, Carien, and Martin Plant. 2019. Understanding South Africa. Auckland Park: Jacana. Energy Innovation. 2020. ‘How much energy do data centers really use?’. Energy Innovation [online] https://​energyinnovation.org/​2020/​03/​17/​how-​much-​energy-​do-​data-​centers-​ really-​use/​. Fine, Ben, and Zavareh Rustomjee. 1996. The Political Economy of South Africa: From Minerals-​ Energy Complex to Industrialisation. London: C. Hurst & Co. Frank, Andre Gunder. 2005. ‘Africans and the Industrial Revolution in England:  A study in international trade and economic development (review)’, Journal of World History, 16(2): 232–​5. Freund, Bill. 2019. Twentieth-​century South Africa:  A Developmental History. Cambridge: Cambridge University Press. Frey, Carl Benedikt, and Michael A. Osborne. 2013. The Future of Employment: How Susceptible are Jobs to Computerization?, Oxford Martin School, September. Grossman, G. M. and E. Helpman. 1991. lnnovation and Growth in the Global Economy. Cambridge, MA: MIT Press.

508    Bhaso Ndzendze and Tshilidzi Marwala Guerrera, Leonardo, and Fernando Cheein. 2020. ‘The role of 5G technologies: Challenges in smart cities and intelligent transportation systems’, Sustainability, 12(1): 1–​15. Hantrais, Linda, Paul Allin, Mihalis Kritikos, Melita Sogomonjan, Prathivadi B. Anand, Sonia Livingstone, Mark Williams, and Martin Innes. 2020. ‘Covid-​19 and the digital revolution’, Contemporary Social Science, 16(2): 256–​70. Harari, Yuval Noah. 2011. Sapiens: A Brief History of Humankind. London: Vintage. Hirsch, Alan. 2005. Season of Hope: Economic Reform under Mandela and Mbeki. Scottsville, KZN: UKZN Press; Ottawa: International Development Research Centre. Hutschenreiter, G., Y. M. Kaniovski, and A. V. Kryazhimskiy. 1995. ‘Endogenous growth, absorptive capacities and international R&D spillovers’. IIASA Working Paper, Laxenburg, Austria [online] https://​core.ac.uk/​download/​pdf/​529463. Inikori, Joseph E. 2002. Africans and the Industrial Revolution in England:  A Study in International Trade and Economic Development. Cambridge: Cambridge University Press. Institute of Directors South Africa. 2019. Directors’ Sentiment Index Report:  4th Edition. Institute of Directors South Africa. Isaacs, Gilad. 2016. ‘A national minimum wage for South Africa’. National Minimum Wage for South Africa Research Initiative [online] http://​nationalminimumwage.co.za/​wp-​content/​ uploads/​2016/​08/​NMW-​RI-​Research-​Summary-​Web-​Final.pdf. Jenkinson, Gareth. 2020. ‘South Africa looks to blockchain to reduce unemployment’, Cointelegraph, 13 March [online] https://​cointelegraph.com/​news/​south-​africa-​looks-​to​blockchain-​to-​reduce-​unemployment. Keynes, J. M. 1962. Essays in Persuasion. New York: W. W. Norton & Co. Khunou, G. 2018. ‘The Fourth Industrial Revolution and absence of fathers’, Sunday Independent. https://​www.iol.co.za/​sundayindependent/​analysis/​the-​fourth-​industrialrevolution-​and-​absence-​of-​fathers-​15771396. Kleingeld, Christo. 2003. ‘A South African railway history’, MWeb [online] http://​mysite. mweb.co.za/​residents/​grela/​transnet.html#:~:text=The%20first%20railway%20line%20 in,opened%20on%2026%20June%201860. Lambrechts, Wynand, Saurabh Sinha, and Tshilidzi Marwala. 2021. ‘The Global South and Industry 4.0: Historical development and future trajectories’, in D. Monyae and B. Ndzendze (eds), The BRICS Order: Assertive or Complementing the West? 249–81. London: Palgrave Macmillan. Lee, Keun, Chan-​Yuan Wong, Patarapong Intarakumner, and Chaiyatorn Limapornvanich. 2020. ‘Is the Fourth Industrial Revolution a window of opportunity for upgrading or reinforcing the middle-​income trap? Asian model of development in Southeast Asia’, Journal of Economic Policy Reform, 23(4): 408–​25. Majozi, Thokozani, and Tshilidzi Marwala. 2020. ‘Investing in innovation will reduce unemployment and poverty’, Mail & Guardian [online] https://​mg.co.za/​education/​2020-​ 08-​ 3 0-​ i nvesting-​ i n-​ i nnovation-​ w ill-​ r educe- ​ u nemployment- ​ a nd- ​ p overty/ ​ ? fbclid= IwAR1CBPUe-​w7lpiRWOMjRYbG8cY5n7k5jGwrQiqp1JijtV7UoweBVmpU0ptk. Malinga, Sibahle. 2019. ‘Inside Takealot’s upgraded warehouse and distribution centre’, IT Web, 19 November [online] https://​www.itweb.co.za/​content/​dgp45MaGlOdvX9l8. Marwala, Tshilidzi. 2020. Closing the Gap:  The Fourth Industrial Revolution in Africa. Johannesburg: Macmillan. Mayer, Claude-​Hélène. 2020. ‘Key concepts for managing organisations and employees turning towards the Fourth Industrial Revolution’, International Review of Psychiatry, 32(7–​8): 673–​84.

South Africa and the Fourth Industrial Revolution    509 McLeod, Duncan. 2020. ‘This is where you can get MTN 5G coverage’, Tech Central, 30 June [online] https://​techcentral.co.za/​this-​is-​where-​you-​can-​get-​mtn-​5g-​coverage/​99203/​. Mkhosi, Percy. 2017. ‘National report on e-​commerce development in South Africa’. Working Paper no. 18/​2017, United Nations Industrial Development Organization, Vienna. https://​ www.unido.org/​sites/​default/​files/​2017-​10/​WP_​18_​2017.pdf. Mzekandaba, Simnikiwe. 2019. ‘SA Connect: Another year, another moving target’, IT Web [online] https://​www.itweb.co.za/​content/​lwrKxv3JRQGqmg1o. Mzekandaba, Simnikiwe. 2020. ‘SA’s 5G-​ready smart city in the works, says Ramaphosa’, IT Web [online] https://​www.itweb.co.za/​content/​KA3WwMdlPnzMrydZ. NPC. 2020. ‘Digital futures:  South Africa’s digital readiness for the Fourth Industrial Revolution’. Draft report, NPC. Pillay, Semona. 2019. ‘The Fourth Industrial Revolution in retail and the low-​income consumer’, Biz Community [online] https://​www.bizcommunity.com/​Article/​196/​185/​191397.html. Porter, Michael E. 1990. ‘The competitive advantage of nations’, Harvard Business Review [online] https://​hbr.org/​1990/​03/​the-​competitive-​advantage-​of-​nations. Presidential Commission on the Fourth Industrial Revolution. 2020. Diagnostic Report of the Presidential Commission on the Fourth Industrial Revolution. Pretoria:  Presidency of the Republic of South Africa. PricewaterhouseCoopers. 2019. ‘Major banks analysis’, PricewaterhouseCoopers [online] https://​www.pwc.co.za/​en/​press-​room/​major-​banks-​analysis-​-​-​september-​2019.html. Republic of South Africa. 2002. ‘Electronic Communications and Transactions Act, 2002’, Republic of South Africa. Roy, Abhijit. 2020. ‘The Fourth Industrial Revolution’, Journal of International Consumer Marketing, 32(3): 268–​70. RSA. 2016. National Integrated ICT White Paper. Pretoria: Government Gazette. Schwab, Klaus. 2017. The Fourth Industrial Revolution. Geneva: Portfolio and Penguin. Smith, Adam. 2003 (1776). The Wealth of Nations. New York: Bantam Classic. Standard Bank (Nicole Crampton). 2020. ‘How to start a 3D printing business’, Standard Bank [online] https://​bizconnect.standardbank.co.za/​start/​business-​options/​how-​to-​start-​a-​3d-​ printing-​business.aspx. StatsSA. 2015. General Household Survey. Statistics South Africa, Pretoria, 17 May [online] http://​www.statssa.gov.za/​publications/​P0318/​P03182014.pdf. Sutherland, Ewan. 2020. ‘The Fourth Industrial Revolution: The case of South Africa’, Politikon, 47(2): 233–​52. Toyana, Mfuneko. 2020. ‘S. Africa will not have universal basic income grant this year, minister says’, Reuters. https://​www.reuters.com/​article/​ozabs-​uk-​safrica-​economyuniversalincome-​idAFKBN24X3L4-​OZABS. Toynbee, Arnold. 2015 (1881). The Industrial Revolution. Boston, MA: The Beacon Press. Traxcn. 2020. ‘Blockchain start-​ups in South Africa’, Traxcn [online] https://​tracxn.com/​explore/​Blockchain-​Startups-​in-​South-​Africa. Uleanya, Chinaza, and Yu Ke. 2019. ‘Review of preparedness of rural African communities nexus formal education in the Fourth Industrial Revolution’, South African Review of Sociology, 50(3–​4), 91–​103. Van Creveld, Martin. 2010. Technology and War. New York: Simon & Schuster. World Economic Forum (Global Agenda Council on the Future of Software and Society). 2015. Deep Shift—​Technology Tipping Points and Societal Impact. Geneva: World Economic Forum.

Chapter 24

I ndu stria l P ol i c y in Sou th   A fri c a Anthony Black

24.1 Introduction The manufacturing sector is widely seen to be an engine of growth and has received much attention from policymakers as well as direct support. At the time of democratization in South Africa, the sector faced a series of major problems. Industrial development in the pre-​1994 era had been shaped by interventionist policies, focused on the strategic concerns of the apartheid state, especially with regard to heavy industry. The manufacturing sector was protected and generally uncompetitive. Since then it has been through major phases of restructuring and become much more globalized. But manufacturing has performed extremely poorly. Its share of output, investment, and employment have all declined sharply since 1994 and this has been in an underperforming economy. In 2019, manufacturing accounted for just 13.2 per cent of GDP, this share having declined from 21.0 per cent in 1994. The sector has shed employment at an alarming rate and continues to run a large trade deficit. In short, far from being an ‘engine of growth’ it has contributed to South Africa’s economic malaise. This presents a major dilemma for policymakers although, of course, manufacturing sector outcomes do not just depend on industrial policy. There are a number of more important external influences, not least the overall health of the economy. Globally, industrial policy in its various guises has gained greater acceptance, including among institutions such as the World Bank (Wade 2011). In a 2019 IMF Working Paper, Cherif and Hasanov make the case for what they term ‘technology and innovation policy’ (TIP) based on the experience of the ‘Asian Miracles’. According to them, the three key ingredients are (i) state intervention to address market failures that limit the development of domestic firms in sophisticated industries; (ii) export orientation; and (iii) competition both internationally and domestically (Cherif and Hasanov 2019: 6). Industrial policy is once again central in economic policy thinking. While one

Industrial Policy in South Africa    511 may not agree with Weiss (2011: 1) that ‘industrial policy has attracted considerable controversy in the development context, possibly out of all proportion to its potential impact (either positive or negative)’ the mode of implementation and impact in different settings has certainly continued to be subject to much debate. A starting point is the Kaldorian ‘laws’ about the special characteristics of manufacturing. This argument sees manufacturing as an ‘engine of growth’ with dynamic effects that induce productivity growth in both manufacturing and non-​ manufacturing sectors.1 Even if these characteristics did not apply, there would still be a case for industrial policy. Tradable services are growing rapidly and can also be developed through industrial policy. Celebrated industrial policy successes, such as Korea and Taiwan, have received much attention (Amsden 1992; Wade 1990). Within Africa, such success stories are more difficult to find. Mauritius is one and perhaps Ethiopia (Oqubay 2015) although in the case of the latter, it is quite early days. But there have, of course, been more dismal outcomes, for instance in a number of Latin American countries. In fact, it appears that industrial policy is more complicated in resource-​rich countries, which frequently seem to misuse their endowments, not least on inefficient efforts to industrialize. This leads us to another important issue—​the objectives of industrial policy in relation to comparative advantage and the degree to which industrial policy should be used to develop sectors in advance of a country’s existing comparative advantage.2 This chapter examines the evolution of South African industrial policy and assesses its impact on the manufacturing sector. The elements mentioned above—​the importance of manufacturing, industrial policy in resource-​rich contexts, and the use of industrial policy to shift comparative advantage, loom large in the South African debate and are central to any analysis. Section 24.2 sets the context in relation to the prospects for middle-​income countries. It also situates South Africa internationally in terms of the performance of the sector. Section 24.3 examines the evolution of policy. Outcomes are examined in section 24.4 with special attention paid to three sectors (heavy industry, automotive, and light manufacturing, especially apparel) as well as one ‘transversal’ sector (regional policy and special economic zones). Section 24.6 concludes with some proposals for a way forward.

24.2  The International Context The international context since the 1990s has been characterized by trade liberalization and globalization. The World Trade Organization was established, tariffs were reduced,

1 

2 

See Kaldor (1966), Thirlwall (1983), and UNIDO (2017). See, for example, the debate between Justin Lin and Ha-​Joon Chang (2009).

512   Anthony Black dozens of regional trade agreements were put in place, and China emerged as the workshop of the world. Many developing countries, especially outside of Asia, were not able to compete and experienced significant deindustrialization. More recently, globalization, as measured by the growth in trade and foreign direct investment, has stalled as a result of a number of factors. First, tariffs are no longer declining. In fact, protectionist pressures are building and have been given significant impetus by the US-​China ‘trade war’. COVID-​19 has further disrupted supply chains but the bigger threat is that the resulting economic slowdown could exacerbate protectionist pressures across the globe. In most upper-​middle-​income countries, manufacturing has been declining as a share of GDP (Table 24.1). There is evidence that deindustrialization appears to have happened at an earlier stage or ‘prematurely’ (Dasgupta et  al. 2006; Rodrik 2016; Tregenna 2009). Rodrik (2016), for example, shows that the share of manufacturing in GDP in developing countries peaked at significantly lower income levels than

Table 24.1: Economic performance of selected middle-​income countries, 1994–​2019 Middle-​ income countries

South Africa

Brazil

Mexico

Thailand

Tunisia

Turkey

Income per capita, 2019 ($)

12,999

15,259

20,411

19,228

11,201

27,875 12,135

Average annual growth in GDP 1994–​2019 (%)

2.7

2.5

2.4

3.7

3.5

4.4

5.0

Average annual 2.1 growth in manufacturing value added (%)

0.8

2.3

3.9

3.2

5.0

–​

Manufacturing value 19.3 added as ratio of GDP, (%): 1994

23.2

17.4

25.9

18.5

22.1

–​

Manufacturing value 11.8 added as ratio of GDP, (%): 2019

9.4

17.3

25.3

14.3

19.0

18.8

Average growth in total exports (%)

2.9

4.6

6.3

6.1

6.7

7.8

4.9

Total exports as ratio of GDP, 2019 (%)

29.9

14.3

39.1

59.7

49.8

31.6

24.7

Source: Calculated from World Bank: World Development Indicators. Notes: Income per capita is measured as GDP per capita for 2019, US$, at PPP exchange rates.

Industrial Policy in South Africa    513 was the case for the developed world. This perspective has been questioned in the sub-​Saharan African context by Nguimkeu and Zeufack (2019) who argue for a greater emphasis on Dutch-​ Disease-​type factors in understanding the African manufacturing experience. Rodrik (2016) argues that technological progress is a convincing explanation of the situation in the developed world because employment deindustrialization has been much more pronounced than output deindustrialization. But in developing countries where the decline in output share has been relatively larger this is a less convincing explanation. Instead, globalization and trade liberalization have been bigger drivers of premature deindustrialization in the developing world. Many countries lacked comparative advantage and the demise of protection led to the retreat of manufacturing (Rodrik 2016). Indeed, what has occurred may not be so much a case of premature deindustrialization but rather the concentration of manufacturing in a smaller number of countries, most notably China (Haraguchi et al. 2017). Table 24.1 presents basic growth parameters for a set of upper-​middle-​income countries. South Africa has performed exceptionally poorly. Manufacturing value-​ added grew at an average rate of only 2.1 per cent per annum from 1994 to 2019. The share of manufacturing in GDP fell in all countries. South Africa experienced the greatest decline, apart from Brazil, with the share of manufacturing in GDP declining by no less than 7.5 percentage points over this period, in spite of the fact that the economy was growing only slowly. Low export growth has been a factor as has growing import penetration.

24.3  The Evolution of Industrial Policy After 1994, the new democratic government embarked on a programme to develop the industrial sector and introduced a stream of new initiatives and policies. The main elements of industrial policy are summarized in Table 24.2. Zalk (2014: 327) divides post-​apartheid industrial policy into two phases; with a ‘shift from neoclassical-​based to structural-​based reforms’ taking place from 2007. While this sharp distinction is probably overdrawn, especially in terms of actual policy impact, the earlier period did include significant trade liberalization and privatization. Trade liberalization began well before 1994, but after 1994 there were further tariff reductions and the elimination of most quantitative controls. Effective rates of protection on manufactured goods had a weighted average of 35 per cent in 1984. These declined to 12.9 per cent in 2000 and then to 9.5 per cent in 2006 (Edwards and Lawrence 2006; Black and Roberts 2009). These measures were complemented by a host of supply-​side measures to promote exports, investment, and skills upgrading (Table 24.2). Special loan programmes were

514   Anthony Black Table 24.2: A chronology of industrial policy 1991

Early 1990s 1995 1995–​6

1995–​8

1996

1998 1999 2002

2003 2005 2007 2010 2011

2013 2014 2015 2018 2019

Introduction of Section 37E of the Income Tax Act (accelerated depreciation for large-​ scale projects aimed mainly at the export market) Introduction of Regional Industrial Development Programme (RIDP) (incentives for industrial expansion in peripheral regions) Introduction of Support Programme of Industry Innovation (SPII) and Technology and Human Resources for Industry Programme (THRIP) Introduction of Motor Industry Development Programme (MIDP) Establishment of small business support agencies: Centre for Small Business Promotion, Ntsika Enterprise Promotion Agency, Khula Enterprise Finance, and the National Small Business Council Introduction of supply-​side incentives including Competitiveness Fund (to promote competitiveness particularly by SMMEs); Short-​Term Export Finance Guarantee Facility (aimed at SMME exporters); Life Scheme (IDC low-​interest financing to export-​oriented projects); Duty Credit Scheme (to promote exports by offering import rebate certificates to exporters of clothing and textiles); Sectoral Partnership Fund (to promote groups of firms to cooperate to address common problems); Workplace Challenge (to improve productivity via joint training of workers and managers) Cancellation of Regional Industrial Development Programme (RIDP) Tax holiday scheme introduced Spatial Development Initiatives (to align public infrastructure provision with Private-​sector investment on a regional basis) Competition Act introduced End of tax holiday scheme Establishment of new competition authorities Announcement of Integrated Manufacturing Strategy with emphasis on knowledge and technology Introduction of Strategic Investment Programme (SIP) and Critical Infrastructure Programme Advanced Manufacturing Technology Strategy Formation of Small Enterprise Development Agency (SEDA) Establishment of Apex Fund to support loans to micro-​businesses Announcement of National Industrial Policy Framework (NIPF) and Industrial Policy Action Plan (IPAP) Launch of Clothing and Textiles Competitiveness Programme (CTCP) Special Economic Zones Bill Amendment of Preferential Procurement Regulations to enable the DTI to designate products for local procurement by government. Introduction of Automotive Production and Development Programme (APDP) Industrial Policy Action Plan 2014/​15–​2016/​17 Introduction of Black Industrialists Scheme (BIS) to support majority black-​owned manufacturing firms. The BIS offers funding for medium-​sized investments. Industrial Policy Action Plan 2018/​19–​2020/​21 Launch of South African Automotive Masterplan (SAAM) Launch of South African Clothing, Textiles, Footwear and Leather Masterplan

Sources: Hirsch (2005); Black and Roberts (2008: 215); the DTI (2018).

Industrial Policy in South Africa    515 put in place to develop small firms. A further objective was to promote ‘knowledge-​ intensive’ production and advanced technology (Black and Roberts 2009). These were highlighted in the Department of Science and Technology’s Advanced Manufacturing Technology Strategy (NACI/​DST 2003). The net impact of this plethora of new measures is far from clear. The objective was to counteract the previous government’s support for large-​scale capital-​intensive industries and the legacy of poor productivity, and to facilitate the development of non-​traditional manufactured exports (Hanival and Hirsch 1998; Joffe et al. 1995). However, this has only happened to a limited degree (see section 24.4). While the stated objective of policy has been to encourage higher value-​added activities, labour-​intensive activities and smaller firms, in practice the weight of support continued to be focused, at least in the early stages, on larger-​scale capital-​intensive activities. Government also embarked on a series of regional industrial policies in the form of industrial development zones (IDZs), spatial development initiatives (SDIs), and special economic zones (SEZs), the latter being introduced in 2011. The year 2007 saw the introduction of explicit industrial policy in the form of the National Industrial Policy Framework (NIPF). The NIPF was aimed at diversification beyond commodities, intensification of the industrialization process, the promotion of labour-​absorbing industrialization, greater black participation, and the development of marginalized regions. It was accompanied by the first iteration of a series of Industrial Policy Action Plans (IPAPs), which developed more targeted approaches in relation, for example, to industrial financing, procurement, trade policies, skills development, and innovation (Zalk 2014). In addition, more selective strategies at the sector level were embarked upon. These included further iterations of the Motor Industry Development Programme (MIDP) and the introduction in 2013 of the Automotive Production and Development Programme (APDP). In the clothing sector, which was hard hit by reduced protection, the Clothing and Textiles Competitiveness Programme (CTCP) was introduced. These strategies are discussed further in section 24.4. There was a host of other measures as well. These included public–​ private partnerships. In the case of mining capital equipment, the Mandela Mining Precinct was established, as a partnership between the Chamber of Mines and government to undertake research and to develop suppliers. A growing sector is business process services (BPS), supported by the BPS Incentive Programme (DTI 2018). There was a growing emphasis on black economic empowerment (BEE) and, in 2015, the Black Industrialists Scheme (BIS) was introduced. A more recent development has been the focus on ‘masterplans’ initially in the automotive and garment sectors. These have been based on consultative processes with industry and other stakeholders to forge a common vision and set ambitious objectives.

516   Anthony Black

24.4 Industrial Performance As indicated above, the manufacturing sector has performed poorly. Trade liberalization, which started to take effect before 1994, led to rapid import penetration across a wide range of sectors. The increase in import penetration ratios in major sectors is indicated in Table 24.3. Imports of vehicles and parts rose rapidly in the automotive sector although that was partly offset by rising exports. Imports since 1994 have increased most rapidly in the labour-​intensive and ultra-​labour-​intensive categories (Mercer 2019).3 In the important clothing sector, import penetration increased from 8.1 per cent in 1990 to 58.3 per cent in 2016 (Table 24.3). The export of manufactures has increased but at a relatively pedestrian pace. The basic pattern of reliance on mining exports has not changed very much. There has been some diversification of total merchandise exports towards manufactured goods although this is mainly due to the expansion in motor vehicle exports, which benefited from large-​ scale assistance. With the rapid increase in import penetration and modest growth in exports, manufacturing has continued to run a large trade deficit, which amounted to nearly Rand 300 billion in 2017 (DTI 2018). Diversified exports have expanded more rapidly to the rest of Africa. The continent accounted for 38.8 per cent of manufactured exports in 2017 and 25.3 per cent of total exports, with the latter share showing a large increase from just 17.9 per cent in 1996 (DTI 2018). This is an important shift as the basket of exports to the rest of the continent consists of mainly processed manufactured goods with the four most important categories comprising non-​electrical machinery, processed food, chemical products (e.g. pharmaceuticals), and motor vehicles parts and accessories (DTI 2018: 15). Indeed, the large trade surplus that South Africa has with the region in manufactured goods may present a stumbling block to further integration (Black et al. 2021).

Table 24.3: Import penetration ratios in key sectors Import penetration ratios (%) 1972

1990

2016

32.0

24.2

65.1

8.0

7.6

17.0

Machinery

48.8

62.5

92.2

Clothing

16.4

8.1

58.3

Motor vehicles Food

Source: Adapted from Mercer (2019).

3 

See Table 24.3 for the classification of these sectors.

Industrial Policy in South Africa    517 The impact on employment has been severe. Manufacturing employment peaked at 1.79 million in 1981 and declined precipitously to just 1.22 million in 2019. The impact on employment of weak manufacturing growth has been compounded by rapidly declining employment intensity. In 2010 prices, employment per million rand of manufacturing output declined from 3.1 in 1972 to 2.2 in 1990 and 0.8 in 2016 (Mercer 2019). The employment intensity of manufacturing has in fact declined much more rapidly than the economy as a whole and reflects, in part, skill-​biased technical change across all sectors but also important shifts between sectors. Table 24.4 provides detailed employment data organized according to factor intensity. Overall, employment has declined by 6.8 per cent over the period although it needs to be noted that major employment losses took place during the 1980s and 1990s. The picture is mixed across factor-​intensity categories. Most notable has been the further decline in the ultra-​labour-​intensive sector. The period from 2007, saw a shift to more intensive industrial policy being implemented through a series of IPAPs. But this coincided with a number of very negative developments. The global financial crisis and associated recession was a major external factor. Within South Africa, the onset of ‘state capture’ under the Zuma presidency contributed to weakness in key institutions and state-​owned enterprises (Zalk 2014). The most damaging was the constraints on electricity output where poor planning and a lack of maintenance led to rolling blackouts coupled with rapidly escalating electricity tariffs. But there were other constraints as well. High port charges and the lack of rail capacity impeded exports. Problems in the commuter rail system added to the difficulties of workers travelling long distances to work. In many areas, dysfunctional local government inhibited firm operation and expansion. The IPAPs have a sectoral focus as well as a set of transversal focus areas. The transversal focus areas refer to cross-​cutting areas of policy implementation. In IPAP 2018/​ 19–​2020/​21, they comprise public procurement and local content; industrial financing; developmental trade policies; African integration and industrial development; special economic zones; and innovation and technology (DTI 2018). In the following sections we focus on industrial policy in three sectors (heavy industry, automotive, and garments) and one transversal focus area (special economic zones).

24.4.1 Heavy Industry The minerals-​energy complex4 that developed in the apartheid era was capital intensive and depended on low cost, subsidized coal-​based electricity. As a result, it was also highly emission intensive (OECD 2013). The share of capital-​intensive manufactures in total manufactured exports increased from 27.9 per cent in 1970 to 49.4 per cent by 1990 although it has declined somewhat since then.

4 

See Fine and Rustomjee (1996).

518   Anthony Black Table 24.4: Manufacturing employment by sub s​ ector, 2000–​19, categorized according to factor intensity

Industry Coke and petroleum products

Factor-​ intensity

12,201

0.93

21,308 19,704

Other chemicals Beverages

Basic chemicals

Ultra-​ capital-​ intensive (UCI)

Annual Annual average average employment employment growth growth Employment, Share Employment, Share rate (%), rate (%), 2000 (%) 2000 2019 (%) 2019 (2000–​09) (2010–​19)

Non-​ferrous metals

TOTAL

UCI

Glass

Capital-​ intensive (CI)

23,063

1.89

10.75

–​1.56

1.63

23,511

1.93

–​0.37

1.58

1.51

14,939

1.22

0.94

–​3.15

36,315

2.78

69,846

5.73

3.66

3.75

40,125

3.07

34,096

2.80

–​2.48

0.79

129,654

9.91

165,454

13.56

1.65

1.07

12,028

0.92

8,394

0.69

1.27

–​4.17

24,159

1.85

33,069

2.71

–​5.12

–​3.98

Basic iron and steel

48,265

3.69

32,381

2.65

0.76

–​4.40

Non-​metallic minerals

54,758

4.19

51,489

4.22

–​0.23

0.08

Motor vehicles

111,598

8.53

90,786

7.44

–​2.06

–​0.09

2,766

0.21

2,738

0.22

–​0.96

1.00

253,573

19.38

218,858

17.94

–​0.48

–​0.98

18,296

1.40

11,617

0.95

–​2.77

–​1.71

62,034

4.74

33,134

2.72

–​2.24

–​3.81

Paper and paper products

Tobacco TOTAL

CI

Rubber

Intermediate-​ capital-​ intensive (ICI)

Other manufacturing Food

195,257

14.93

209,499

17.17

–​0.86

1.63

Printing and publishing

51,330

3.92

53,001

4.35

1.21

–​0.73

Other transport equipment

13,578

1.04

17,487

1.43

2.67

0.68

340,495

26.03

324,738

26.62

–​0.76

0.29

TOTAL

Continued

Industrial Policy in South Africa    519 Table 24.4: (Continued )

Industry Machinery Textiles

Factor-​ intensity Labour-​ intensive (LI)

Annual Annual average average employment employment growth growth Employment, Share Employment, Share rate (%), rate (%), 2000 (%) 2000 2019 (%) 2019 (2000–​09) (2010–​19) 74,980

5.73

113,745

9.32

3.41

1.26 –​3.30

52,039

3.98

25,781

2.11

–​3.80

Plastics

51,738

3.96

49,176

4.03

–​2.72

2.07

TV, radio, and communication equipment

21,961

1.68

12,240

1.00

–​5.05

–​0.84

Electrical machinery

40,243

3.08

49,334

4.04

–​0.75

3.22

Professional and scientific equipment

15,402

1.18

17,111

1.40

1.13

0.29

100,926

7.72

103,938

8.52

0.75

–​0.22

Metal products Wood TOTAL

LI

Leather and Ultra-​ leather products labour-​ intensive Footwear (ULI) Furniture Clothing Total Grand Total

ULI

51,638

3.95

47,236

3.87

–​1.97

1.11

408,927

31.26

418,560

34.31

–​0.34

0.60

8,656

0.66

4,808

0.39

–​4.72

–​0.89

18,254

1.40

8,585

0.70

–​7.32

–​0.08

45,727

3.50

35,090

2.88

–​2.96

0.64

102,825

7.86

43,700

3.58

–​5.63

–​3.11

175,463

13.41

92,182

7.56

–​5.11

–​1.56

–​0.82

0.08

1,308,111

100

1,219,792

100

Source: Derived from Quantec data. Notes: The categorization according to factor intensity draws on Mercer (2019). Food includes: meat, fish, fruit; dairy products; grain mill products; and other food products. Textiles include: other textile products. Motor vehicles include: parts and accessories. Clothing includes: knitted, crocheted articles; and wearing apparel. Wood includes: sawmilling and products of wood. Electrical machinery includes: electric motors, generators, transformers; electricity distribution and control apparatus; insulated wire and cables; and other electrical equipment. Metal products include: structural metal products; and other fabricated metal products. Radio, television and communication equipment includes: household appliances. Machinery includes: general purpose machinery; and special purpose machinery. Professional and scientific equipment includes: office, accounting, computing machinery.

Revealed comparative advantage in heavy industry was partly a distortion resulting from very substantial direct and indirect state support to the sector. This included subsidized electricity and other forms of government support. For example, aluminium production was based entirely on using low-​priced electricity to process imported bauxite. Historically, cheap electricity has been a function not just of abundant coal

520   Anthony Black resources, but also the extraordinary electricity pricing policy. Heavy over ​investment in electricity capacity in the 1970s and early 1980s by Eskom, led government to set extremely low tariffs to attract huge investments in a series of metal processing plants. And with capacity running out, agreements were still being reached in 2007 with Alcan for an aluminium smelter at Coega, at highly favourable electricity prices (Black and Roberts 2009). The proposed Rand 21 billion smelter, which would have exported most of its output in primary form, was also in line for huge investment and tax allowances. Given later developments, it is perhaps fortunate that emerging constraints on South Africa’s generation capacity led to the cancellation of the project in 2009. The same issues apply with regard to Eskom’s preferential pricing arrangements with other large, energy-​intensive industries, such as BHP Billiton’s aluminium smelters. While the clearly stated objective of industrial policy is to restructure the economy to promote growth and jobs, some of the very substantial support programmes provided by government have reinforced rather than altered the industrial development path, especially in the first decade of the transition. An accelerated depreciation allowance under the 37E incentive was given to major resource-​based projects in the 1990s such as Columbus Stainless Steel and the now-​mothballed Saldanha Steel plant. From 2002 to 2005, the Strategic Industrial Projects (SIP) programme provided tax relief equivalent to Rand 7.7 billion for large capital-​intensive projects, mostly in sectors such as steel, ferroalloys, aluminium, and basic chemicals (Black and Roberts 2009). Another related DTI initiative was the funding of mega projects (defined as more than Rand 1 billion) and industrial development zones. State support for such projects is multifaceted, including infrastructure support, industrial subsidies, cheap land and water, as well as preferential electricity tariffs. These developments have generally been aimed at large-​scale capital-​intensive and energy-​intensive projects (Black 2010). Further direct state support for heavy industry has been provided by the state-​owned Industrial Development Corporation (IDC), which has played an important role in influencing economic growth in accordance with government’s strategic objectives. The IDC supports firms by providing equity finance and loans, frequently at concessional rates. Historically, it has funded large-​scale mineral beneficiation projects and has been closely associated with the parastatals as well as with large private-​sector conglomerates. The IDC has more recently increased the emphasis on labour-​intensive sectors such as tourism, agriculture, and smaller-​scale enterprises (Black and Hasson 2016). The long history of artificially low electricity prices has led the economy to its current predicament where electricity supply is inadequate and prices are rising sharply. Rising prices and supply interruptions have impacted very seriously on this sector even though a number of the larger operations continue to receive special pricing deals. Since 2008, electricity constraints, and more recently over-​capacity in the global steel industry, have placed a brake on heavy industry development and led to the abandonment of a number of major proposed investments and the closure of foundries and other metal processing plants.5

5 

The most recent major closure was that of Saldanha Steel, which was announced in 2019.

Industrial Policy in South Africa    521 Also, the market power of large upstream producers in sectors such as steel and chemicals (e.g. Arcelor-​Mittal and Sasol) profoundly disadvantaged more labour-​ intensive downstream production (Roberts and Rustomjee 2009). The lack of competition has enabled upstream producers to use import parity pricing, so that domestic fabricators of metal and plastic products have derived no advantage from low domestic production costs of key inputs such as steel and basic chemicals. In certain cases, government policy has exacerbated this situation, for instance in the imposition of a 10 per cent tariff on steel in 2015, which imposed significant costs on downstream fabricators.

24.4.2 The Automotive Sector The automotive sector has been the recipient of intensive industrial policy support and is frequently cited as a policy success story (Hirsch 2008; Barnes et al. 2004) although the policy has also come in for criticism.6 The sector has experienced a rapid increase in exports with its share of manufactured exports increasing from 4.3 per cent in 1995 to over 20 per cent by 2019. But this is more a story of rapid international integration than developing competitiveness. While exports have risen rapidly so have imports and South Africa remains a net importer of automotive products. The Motor Industry Development Programme (MIDP), introduced in 1995, had very specific industrial policy objectives. In the early 1990s, as a result of high protection, there was a proliferation of makes and models being produced in low volumes in South Africa. In turn, this made it impossible for component suppliers to achieve anything close to minimum efficient scale. The MIDP therefore aimed to increase the volume and scale of production through a greater level of specialization in terms of both vehicle models and components. This could be achieved by earning import credits on exports of locally produced vehicles and components. These credits could be used to import either additional models for sale in the domestic market, or components required in vehicle assembly (Barnes et al. 2021). Initially, this strategy was not wholly successful as a number of vehicle assemblers opted in the short term for the lower-​risk option of generating large-​scale exports of components, such as catalytic converters, in order to offset these duty liabilities. Exports of catalytic converters which contain platinum-​group metals rocketed from R389  million in 1995 to Rand 21.8 billion in 2016 after which they declined slightly (AIEC 2020). Most light vehicle producers have now secured major ongoing vehicle export contracts and have invested accordingly to upgrade assembly plants. However, large export volumes have not led to an increase in local content. And, as a result of the stagnating domestic market and poor investment climate, the government is on a weak footing when bargaining with multinationals to achieve some degree of reciprocity, for instance via higher levels of local content (Barnes et al. 2021). The result is that local

6 

See, for example, Flatters and Netshitomboni (2007) and Kaplan (2019).

522   Anthony Black content has actually slightly declined with some hollowing-​out of second-​and third-​tier component suppliers. In comparing the South African and Thai automotive industries, Barnes et al. (2017) argue that while exports have expanded rapidly from both countries, global integration has turned out less favourably for the South African automotive industry. South Africa’s high input costs into manufacturing, particularly for skilled labour (artisans, technicians, and managers), have limited its competitiveness. As is well known, outcomes in the training of artisans have been extremely poor. Indeed, a striking feature about the labour market in South Africa is not so much that wages of production workers are higher than competitors (although in many cases they are), but the high costs of managers and skilled staff. This conclusion is supported by earlier studies. For example, a 2007 World Bank study found that unskilled workers in South Africa earned slightly less than in Poland but somewhat more than in Brazil (Clarke et  al. 2007). However, managers’ wages were 2.5 and 3 times higher than in Poland and Brazil respectively, and wages of professional and skilled employees in South Africa were also much higher than in the other two countries (Black and Hasson 2016). Other costs are also higher in South Africa. Port charges, a critical factor in the trade-​ intensive automotive industry, are notoriously high (DTI 2018). As indicated above, electricity prices have risen sharply along with the ongoing problem of load-​shedding. Mediocre economic growth and the failure to develop a low-​cost manufacturing environment have clearly limited South Africa’s attractiveness for foreign direct investment. Thailand, in contrast, has attracted much higher levels of investment which has deepened the component supply base. Although both countries have seen very rapid export expansion, Thailand provides a genuine export platform. Export growth has far exceeded the expansion of imports and Thailand runs a large automotive trade surplus. In South Africa, investment in exports has in part been driven by the desire to earn import rebate credits, which support the import of vehicles and components. The automotive industry is considered to be a flagship of South African industrial policy and there have been positive achievements in modernizing and upgrading the industry. However, limited progress with deepening the supply chain is indicative of the weaknesses in manufacturing, and the same factors which militate against real competitiveness in the sector are obstacles to downstream manufacturing in South Africa.

24.4.3 Light Manufacturing: Clothing Light manufacturing has performed poorly. The ultra-​labour-​intensive (ULI) category comprising clothing, leather, footwear, and furniture has been in decline since the early 1990s and from 2000 to 2019 employment declined a further 47.4 per cent (Table 24.3). The key garment sector saw employment fall from the already low figure of 102,825 to only 43,700 over this period. The footwear industry was also severely affected. As per capita incomes rise, it is to be expected that the relative share of labour-​intensive sectors will decline as these industries shift to lower-​wage economies. But the decline has

Industrial Policy in South Africa    523 Table 24.5: Share of manufacturing value added in textiles and clothing, 1996–​2018 1996

2000

2004

2008

2010

2014

2018

South Africa

7.8

4.9

4.5

3.0

1.8

1.8

1.8

Brazil

8.1

6.9

5.9

5.6

6.6

6.6

6.8

Malaysia

4.6

4.1

2.8

1.9

1.9

1.7

2.1

Mexico

4.4

3.9

3.5

2.8

3.4

3.0

2.7

Thailand

8.6

12.4

7.9

8.8

6.4

6.0

6.0

Tunisia

34.0

35.1

29.6

23.3

21.9

19.2

17.6

Turkey

17.2

15.7

21.5

15.5

16.5

17.7

16.4

Source: World Bank, World Development Indicators.

been particularly rapid in South Africa, in spite of the fact that per capita incomes have hardly risen. A key contributor to the decline in labour-​intensive employment has been the poor performance of the textiles and clothing sectors. Their share of manufacturing output declined from 7.8 per cent in 1990 to just 1.8 per cent in 2014 but has stabilized since then (Table 24.4). While other middle-​income comparator countries have also seen declines in the sector, none has been as rapid as in South Africa (Table 24.5). This decline has particularly affected some poorer regions of the country. For example, in the Eastern Cape the textiles, clothing, and leather sector was the largest manufacturing employer with 26,000 workers in 1996, equal to 19.8 per cent of the manufacturing labour force in the province. Already by 2012 employment had declined to just 10,700 workers, just 11.8 per cent of a much smaller manufacturing labour force (Kaplan et al. 2014: 25). The contrast with comparator countries is striking. Malaysia, Mexico, Thailand, Tunisia, and Turkey all have at least two of their top five export products in the ultra-​labour-​intensive or labour-​intensive categories. Mexico’s main exports are cars and oil, but it exports huge volumes of consumer electronics and machinery and office equipment. Turkey exported $14.7 billion of apparel in 2016 and Malaysia’s main export is consumer electronics, a labour-​intensive sector. Apparel is Tunisia’s major export with revenues of $2.1 billion in 2016 (Black et al. 2016). South Africa’s poor position in light manufacturing compared to other upper-​middle-​ income countries is all the more striking if account is taken of its exceptionally high poverty rates. Although South Africa is an upper-​middle-​income country, a large sector of the population has characteristics approximating that of a lower-​middle-​income country. Apart from Tunisia, the comparator countries in Table 24.4 all have higher per capita incomes on a PPP basis and much lower poverty levels than South Africa. It is difficult to see how this sector would have survived trade liberalization unscathed, but it was not helped by labour market policies, such as the extension of bargaining

524   Anthony Black council wage settlements to non-​parties especially in non-​metropolitan areas (Nattrass and Seekings 2016). The decline of ‘decentralized industries’ in and around the former Bantustans, as incentives were withdrawn, was a further factor. Also important has been the lack of government enthusiasm for mass employment in labour-​intensive (low-​ wage) manufacturing. While ‘decent work’ may be the ideal, the actual outcome has been declining employment. Measures to rescue the industry have included the Clothing and Textiles Competitiveness Programme (2010) which has had some positive outcomes among beneficiary firms. But more marginal (non-​metropolitan, lower-​wage) producers are excluded. The recently announced masterplan will only provide support to ‘legal’ firms. There has also been a tilt to greater protection with pressure placed on major retailers to increase domestic sourcing. This may indeed be warranted to limit imports from Asia, but if policymakers are serious about regional integration, they need to ensure that growing clothing imports from the rest of Africa are not impeded.

24.4.4 Regional Policies and Special Economic Zones In order to understand the development trajectory of regional policy, and SEZs in particular, it is necessary to take account of the history of spatial planning under apartheid. To provide some kind of economic basis for the imposition of ‘grand apartheid’ from the late 1950s, there was a series of regional development interventions to encourage manufacturing firms to locate in designated growth points in peripheral economic regions within, or close to, the Bantustans (Wellings and Black 1986). These growth points offered incentives, including wage subsidies, specifically designed to attract labour-​intensive firms (Tomlinson and Addleson 1987). While the programme was an ideologically driven failure it did lead to significant relocation and investment in labour-​intensive industries. By 1984, employment in the ‘decentralization points’ accounted for 21.7 per cent of total manufacturing employment, up from 8.8 per cent in 1972 (Wellings and Black 1986: 13). With the withdrawal of incentives many of these decentralized industries collapsed, contributing to the decline in manufacturing employment. After 1994, the democratic government embarked on a range of interventions with a regional development dimension (Nel and Rogerson 2014). The objective of spatial development initiatives (SDIs) was to unlock development along regional and transport corridors (Jourdan 1998), such as in the case of the Maputo Corridor, which developed the link between Gauteng and the Mozambican port (Yang et al. 2020; Rogerson 2001). From 2000, a series of IDZs were established close to major transport hubs, the largest of which was Coega, outside Port Elizabeth but other sites included East London, Richards Bay, and OR Tambo International Airport (ORTIDZ). The main objectives were to promote export-​orientated manufacturing and attract foreign investment (Yang et al. 2020; Centre for Development and Enterprise 2012; DTI 2015; Jourdan 1998). The lack of special incentives (Nyakabawo 2014) and poor planning and

Industrial Policy in South Africa    525 marketing produced limited results (Yang et al. 2020) and it was decided to convert the programme into an SEZ programme (DTI 2012). SEZs can aim at improving industry competitiveness via good-​quality infrastructure, establishing clusters, addressing economic distortions, or experimenting with reform. The actual rationale in South Africa is not very clear. The Special Economic Zones Bill of 2011 provided upgraded incentives in established and new zones. Some of the aforementioned existing zones are well located close to urban areas but others are more remote and clearly have locational disadvantages. A  number of the SEZs are located in the former Bantustans and some are far from established nodes of economic activity. The location of SEZs in backward regions with limited skills and weak infrastructure and linkages is likely to retard their ability to attract investment (Farole and Sharp 2017). There has been little to show for the colossal investments that have been made. Coega is perhaps the most controversial because of its huge scale and underutilized capacity, but overall performance has been unspectacular in terms of investment and employment. The DTI’s 2015/​16 SEZ Performance Analysis Bulletin states that of the 73,000 jobs created in the IDZs by that stage, as many as 88 per cent were in construction or ‘indirect’ jobs, with 8,500 workers employed in manufacturing and zone service industries (Yang et al. 2020). Since then, there has been some increase in the rate of new investment. According to the Coega Development Corporation’s 2018/​19 Annual Report (Coega Development Corporation, 2019) the five-​year target for operational jobs was exceeded. But the target of only 8,902 jobs was rather modest (Black and Yang 2021). Cluster development has been limited. For example, high-​profile investments by Chinese multinationals, such as FAW and BAIC at Coega and Hisense in Atlantis, use very little in the way of domestic parts.

24.5  Conclusion: What Role for Industrial Policy? The manufacturing sector has performed poorly with its share of output declining in a low-​growth economy. To make matters worse the employment intensity of the sector has declined quite rapidly. The reality is that South Africa is currently not a very competitive manufacturing location as indicated by the pedestrian performance of manufactured exports, and evident lack of progress with downstream manufacturing. The historic bias of incentives towards heavy industry is one problem and has been damaging, not only for employment but also for growth. The apartheid-​era legacy of limited skills development is another. Unfortunately, this deficit has not been decisively addressed since 1994. At the dawn of the democratic era, the pattern of manufactured exports reflected the distorted pattern of development manifested in the ‘minerals-​energy complex’ with

526   Anthony Black ‘revealed comparative advantage’ in steel and basic metals, in part the result of heavy state support and artificially cheap energy. South Africa’s extraordinary economic structure comprising a high level of resource dependence, a capital-​intensive export profile, and massive structural unemployment has created a complex adjustment puzzle which is far from being resolved. In the democratic era, the manufacturing sector has been buffeted by a series of adverse events. The first was trade liberalization. Opening the economy was supposed to unleash labour-​absorbing export growth but this did not materialize. The one major sector which did achieve rapid export growth was the automotive sector, assisted by large-​scale subsidies. Given the generally weak export response and the fact that exports are more capital intensive than imports, it is not surprising that employment declined. The second shock was to the mainstay of heavy industry resulting from infrastructure failures (e.g. unstable electricity supply and sharply rising prices) and the gradual withdrawal of incentives. While electricity prices had been artificially low and the large incentives for heavy industry were inappropriate, these changes have imposed huge adjustment costs on the sector, which has suffered from falling investment and plant closures. A third and related shock has been the flatlining of mineral exports (in dollar terms). Commodity prices of course play an important role but damage has been inflicted by regulatory failure and lack of infrastructure capacity. The lack of mining export growth impacted manufacturing in two ways; directly because it is a major source of demand for engineering products, but more importantly because it is a major driver of domestic consumption. Fourth, there has been the decline of labour-​intensive manufacturing. This has occurred in a high-​unemployment economy which is growing slowly relative to its middle-​income peers. Import competition has been a major factor but the state has also failed to mobilize the investments required to take advantage of South Africa’s huge unemployed-​labour resources. The objectives of industrial policy have been quite diverse, ranging from beneficiation to small business support, as well as labour-​absorbing industry and the promotion of the ‘knowledge economy’. Industrial policy has sought to shift industrial development onto a different trajectory, but this has proved extraordinarily difficult and has met with limited success (Roberts and Rustomjee 2009). So what is the way forward? While industrial policy is usually defined as a set of selective interventions to promote industrial upgrading, a more appropriate conception may be to improve economy-​wide efficiency. Massive structural unemployment is without doubt the most glaring inefficiency facing the economy. The bulk of our unemployed labour are unskilled or semi-​skilled and cannot easily be absorbed into sophisticated industries. They could, however, be absorbed into labour-​intensive activities, including in manufacturing (Black 2010). In a diverse, middle-​income economy it is not appropriate to have a single industrial policy objective. But South Africa still faces important choices and the crisis confronting

Industrial Policy in South Africa    527 the country calls for a shift in emphasis. There are a number of options available. South Africa is a mineral-​rich economy, and it may be possible, even sensible, to have an industrial policy which promoted capital-​intensive, resource-​based exports with employment being generated elsewhere in services, (protected) manufacturing for the domestic market, or agriculture. Or industrial policy could target more advanced, leading sectors which may lead to little direct employment growth, but which would generate the export expansion required to finance development with employment being created in the protected sectors of the domestic economy (Black 2010). There are also sectors, such as agro-​processing, which are not particularly labour demanding but where the indirect employment multipliers (into agriculture) are very significant. But there remains an important potential role for industrial policy to support more employment-​intensive growth. This would also complement efforts to reduce carbon emissions and develop a greener development path. To move to a more labour–absorbing growth path, South Africa will need to actively intervene to create competitive advantage in more labour-​demanding economic activities. Competition in light manufacturing is a reality and South Africa could do much better than it has been doing. But what of objectives such as moving up the value chain, technological upgrading, and promoting the digital economy?7 These activities can and should be supported and may be complementary to labour-​absorbing growth in some ways. But, in the end, they constitute a partial development strategy in the South African context because a large section of the (unemployed) labour force is not equipped with the skills to be employed in these sectors. It is clear that manufacturing growth already exhibits a strong skills bias which has raised wages for skilled workers and failed to create large-​scale employment in the unskilled or semi-​skilled categories. Skills upgrading is essential but currently skills are generally more highly priced than those of our middle-​income competitors. It can also be argued that higher employment and the growth in labour-​absorbing activities is the best way of promoting upgrading and the acquisition of human capital (Black et al. 2016). A key question is whether the incentive structure can be re-​shaped to facilitate employment creation much more strongly. This means providing appropriate infrastructure and investments to improve competitive capabilities in more labour-​ demanding activities. It could also mean support for small firms and training, particularly at a basic level and an examination of the regulatory environment. Certain labour-​market rigidities do need to be addressed and there is a role for targeted import protection as well. Incentives should subsidize labour and training rather than capital investment, electricity, and infrastructure for capital-​intensive firms (Black and Nasson 2016). A key challenge facing South African policymakers is to mobilize the potential of an under-​employed and poorly skilled workforce. Industrial policy has to play its part in this crucial endeavour.

7 

See Andreoni et al. (2021).

528   Anthony Black

References Amsden, Alice. 1992. Asia’s Next Giant:  South Korea and Late Industrialization. Oxford: University Press. Andreoni, Antonio, Justin Barnes, Anthony Black, and Timothy Sturgeon. 2021. ‘Digitalisation, industrialisation and skills development: Opportunities and challenges for middle-​income countries’, in Antonio Andreoni, Pamela P. Mondliwa, Simon Roberts, and Fiona Tregenna (eds) Structural Transformation in South Africa:  Sectors, Politics and Global Challenges. Oxford: Oxford University Press. Automotive Industry Export Council (AIEC). 2020. Automotive Export Manual. Pretoria: AIEC. Barnes, Justin, Anthony Black, and Lorenza Monaco. 2021. ‘Government policy in multinational-​dominated global value chains: Structural transformation within the South African automotive industry’, in Antonio Andreoni, P. Mondliwa, Simon Roberts, and Fiona Tregenna (eds) Structural Transformation in South Africa: Sectors, Politics and Global Challenges. Oxford: Oxford University Press. Barnes, Justin, Anthony Black, and Kriengkrai Techakanont. 2017. ‘Industrial policy, multinational strategy and domestic capability: A comparative analysis of the development of South Africa’s and Thailand’s automotive industries’, The European Journal of Development Research, 29(1): 37–​53. Barnes, Justin, Raphael Kaplinsky, and Mike Morris. 2004. ‘Industrial policy in developing economies: Developing dynamic comparative advantage in the South African automobile sector’, Competition and Change, 8(2): 153–​72. Black, Anthony. 2010. ‘Tilting the playing field: Labour absorbing growth and the role of industrial policy’. University of Cape Town. CSSR Working Paper no. 279. Black, Anthony 2010 Stephanie Craig and J Paul Dunne. 2016. Capital intensity, industrial policy and employment in the South African manufacturing sector. University of Cape Town. SALDRU. REDI3x3 Working Paper no.23 Black, Anthony, and Reviva Hasson. 2016. ‘Capital intensive industrialisation and comparative advantage: Can South Africa do better in labour demanding manufacturing?’, in A. Black (ed.) Towards Employment Intensive Growth in South Africa. Cape Town: UCT Press. Black, Anthony, and Simon Roberts. 2009. ‘The evolution and impact of industrial and competition policies’, in J. Aron, (eds) South African Economic Policy Under Democracy. Oxford: Oxford University Press. Black, Anthony, and Chongsheng Yang. 2021. ‘South Africa’s special economic zones as destinations for Chinese investment: Problems and possibilities’, in Chris Alden and Yu-​ Shan Wu (eds) China and South Africa. Basingstoke: Palgrave. Black, Anthony, Lawrence Edwards, Faizel Ismail, Brian Makundi, and Mike Morris. 2021. ‘The role of regional value chains in fostering regional integration in Southern Africa’, Development Southern Africa, 38(1): 39–​56. Centre for Development and Enterprise. 2016. ‘An EPZ for the Nelson Mandela Metro’. Growth Series Report no. 7, Johannesburg. Cherif, Reda, and Fuad Hasanov. 2019. ‘The return of the policy that shall not be named: Principles of industrial policy’. IMF Working Paper no. WP/​19/​74, Washington, DC. Clarke, George, James Habyarimana, Mike Ingram, David E. Kaplan, and Vijaya Ramachandran. 2007. An Assessment of the Investment Climate in South Africa. Washington, DC: World Bank.

Industrial Policy in South Africa    529 Coega Development Corporation. 2019. ‘Integrated annual report 2018/​2019’. Dasgupta, Sukti, and Ajit Singh. 2006. ‘Manufacturing, services and premature deindustrialization in developing countries: A Kaldorian analysis’. WIDER Research Paper, Helsinki. Department of Trade and Industry (DTI). 2012. ‘Policy on the development of special economic zones in South Africa’. Department of Trade and Industry, Pretoria. Department of Trade and Industry (DTI). 2016. ‘2014/​15 SEZ performance analysis bulletin’. Department of Trade and Industry, Pretoria. Department of Trade and Industry (DTI). 2018. ‘Industrial policy action plan’. Department of Trade and Industry, Pretoria. Edwards, Lawrence, and Robert Lawrence. 2006. ‘South African trade policy matters: Trade performance and trade policy’. CID Working Paper no. 135, Harvard University. Farole, Thomas, and Mark Sharp. 2017. ‘Spatial industrial policy, special economic zones and cities in South Africa’. Unpublished manuscript. Washington, DC: World Bank. Fine, Ben, and Zavareh Rustomjee. 1996. The Political Economy of South Africa: From Minerals-​ Energy Complex to Industrialisation. London: C. Hurst and Co. Publishers. Flatters, Frank ,and Nnzeni Netshitomboni. 2007. ‘Trade and poverty in South Africa: The motor industry development programme’, Studies in Economics and Econometrics, 31(2): 143–​60. Hanival, Stephen, and Alan Hirsch. 1998. ‘Industrial policy and programmes in South Africa’. Paper presented at TIPS Forum, Johannesburg. Haraguchi, Nobuya, Charles Fang, Chin Chang, and Eveline Smeets. 2017. ‘The importance of manufacturing in economic development:  Has this changed?’, World Development, 93: 293–​315. Hirsch, Alan. 2008. Season of Hope:  Economic Reform under Mandela and Mbeki’. Ottawa: IDRC . Joffe, Avril. 1995. ‘Improving manufacturing performance in South Africa: The report of the Industrial Strategy Project’. Ottawa: IDRC . Jourdan, Paul. 1998. ‘Spatial development initiatives (SDIs):  The official view’, Development Southern Africa, 15(5): 717–​25. Kaldor, N. 1966. Causes of the Slow Rate of Economic Growth of the United Kingdom:  An Inaugural Lecture. Cambridge: Cambridge University Press. Kaplan, David. 2019. ‘South Africa’s industrial policy: Time for a review and a rethink’. Centre for Development and Enterprise, Johannesburg. Kaplan, David, Mike Morris, and Lucy Martin. 2014. ‘Identifying and developing sustainable interventions to promote non-​automotive industries in the Eastern Cape’. Report prepared for National Treasury, Pretoria. Lin, Justin, and Ha‐Joon Chang. 2009. ‘Should industrial policy in developing countries conform to comparative advantage or defy it? A debate between Justin Lin and Ha‐Joon Chang’, Development Policy Review, 27(5): 483–​502. Mercer, Sean. 2019. ‘Making manufacturing work: An investigation of employment outcomes in South African manufacturing’. Unpublished Master of Commerce dissertation, University of Cape Town. NACI/​DST. 2003. ‘Advanced manufacturing technology strategy’. NACI/​DST, Pretoria. Nattrass, Nicoli, and Jeremy Seekings. 2016. ‘Institutions, wage differentiation and the structure of employment in South Africa’, in Anthony Black (ed.) Towards Employment Intensive Growth in South Africa. Cape Town: University of Cape Town Press. Nel, Etienne L., and Christian M. Rogerson. 2014. ‘Re-​spatializing development: Reflections from South Africa’s recent re-​engagement with planning for special economic zones’, Urbani Izziv, 25: 24–​35.

530   Anthony Black Nguimkeu, Pierre, and Albert G. Zeufack. 2019. ‘Manufacturing in structural change in Africa’. Policy Research Working Paper no. 8992, World Bank, Washington, DC. Nyakabawo, Wendy. 2014. ‘The geographic designation of special economic zones’. TIPS Working Paper, Pretoria. OECD. 2013. ‘Economic growth in South Africa: Getting to the right shade of green’, in OECD Economic Surveys: South Africa 2013. Paris: OECD Publishing. Oqubay, Arkebe. 2015. Made in Africa:  Industrial Policy in Ethiopia. Oxford:  Oxford University Press. Roberts, Simon, and Zavareh Rustomjee. 2009. ‘Industrial policy under democracy:  Apartheid’s grown-​up infant industries? Iscor and Sasol’, Transformation:  Critical Perspectives on Southern Africa, 71(1): 50–​75. Rodrik, Dani. 2016. ‘Premature deindustrialization’, Journal of Economic Growth, 21(1): 1–​33. Rogerson, Christian. 2001. ‘Spatial development initiatives in Southern Africa: The Maputo development corridor’, Tijdschrift voor economische en sociale geografie, 92(3): 324–​46. Thirlwall, A. P. 1983. ‘A plain man’s guide to Kaldor’s growth laws’, Journal of Post Keynesian Economics, 5: 345–​58. Tomlinson, Richard, and Mark Addleson. 1987. Regional Restructuring under Apartheid: Urban and Regional Policies in Contemporary South Africa. Johannesburg: Ravan Press South Africa. Tregenna, Fiona. 2009. ‘Characterising deindustrialisation:  An analysis of changes in manufacturing employment and output internationally’, Cambridge Journal of Economics, 33(3): 433–​66. UNIDO. 2017. ‘Structural change for inclusive and sustainable industrial development’. UNIDO, Vienna. Wade, Robert. 1990. Governing the Market: Economic Theory and the Role of Government in East Asian Industrialization. Princeton, NJ: Princeton University Press. Wade, Robert. 2011. ‘Why Justin Lin’s door-​opening argument matters for development economics’, Global Policy, 2(1): 115–​16. Weiss, John. 2011. ‘Industrial policy in the twenty-​first century’. Working Paper, no 2011/​55, UNU-​WIDER, Helsinki. Wellings, Paul, and Anthony Black. 1986. ‘Industrial decentralization under apartheid: The relocation of industry to the South African periphery’, World Development, 14(1): 1–​38. Yang, Chongsheng, Wang Yong, Anthony Black, and Chen Meiying. 2020. ‘Evaluating special economic zones in South Africa:  History, performance, and challenges’, in Niblock Tim, Guang Yang, and Yan Zhou (eds) Area Studies:  New Realities, New Conceptions. Beijing: China Social Sciences Press. Zalk, Nimrod. 2014. ‘Industrial policy in a harsh climate: The case of South Africa’, in Salazar-​ Xirinachs et al. (eds). Transforming Economies: Making Industrial Policy Work for Growth, Jobs and Development, 327–56. Geneva: International Labour Organisation.

Chapter 25

C ompetition P ol i c y in Sou th A fri c a Liberty Mncube and Nicola Theron

25.1 Introduction Competition policy encompasses the aim of promoting and maintaining competition in the marketplace as well as, promoting other government policies that allow competitive marketplaces to develop in a way that is not detrimental to society.1 The key challenge is to get the balance right. In this chapter we will critically examine whether the emphasis in South Africa on creating a more equal society, by also including public interest goals such as employment protection, economic development, export promotion, has diluted the impact of competition policy. It is useful to note that the early criticism about including public interest goals in the Competition Act warned that this will unnecessarily complicate the task of the competition authorities.2 Globally, the roots of competition policy are old. In its contemporary form, competition policy responds to the universal pursuit of economic power, in which those who have market power seek more profits and wealth at the expense of those who do not. Although the competition laws in each country differ widely, these laws share the protection of competition in markets as one common goal.3 Each country’s competition law is designed to respond to distorting or oppressive economic power given the socio-​ economic context. Competition policy has two main instruments, competition law and competition advocacy. Competition law is about enforcing competition rules against anti-​competitive 1 

See Motta (2004). Competition policy is ‘the set of policies and laws which ensure that competition in the marketplace is not restricted in such a way as to reduce economic welfare’. 2  See Reekie (1999). 3  For a review of competition policy in the European Union and the United States, see Motta (2004), ­chapter 1.

532    Liberty Mncube and Nicola Theron market conduct. It focuses on anti-​competitive practices by private or public companies to ensure that markets are competitive. Competition advocacy is about promoting less anti-​competitive means of achieving the goals of other government policies. In other words, competition advocacy is about achieving voluntary compliance with competition law in the design of other government policies.4 This chapter is about competition law and competition advocacy. Competition law enforcement in South Africa has been characterized throughout by change and controversy. Since the 1998 Competition Act was promulgated and the Competition Authorities commenced their duties on 1 September 1999, the application of competition law has become much more active, with many high profile cases in key industries being adjudicated by the competition authorities. Competition law has become an important feature of the economic landscape in South Africa, being one of the instruments used by government to move the country to greater equity and greater efficiency. This chapter will commence, in section 25.2, with a discussion of the context and history of competition law (including competition law’s origins and evolution over time). In section 25.3, we discuss the multiple goals of competition law. In South Africa, as elsewhere, competition law has adapted to unique national experiences. In other words, asserting relevance, South Africa chose to put both efficiency and equity goals as the focus of its competition law. The focus on these twin goals has arguably been strengthened in the amendments to the Competition Act of 2018. In section 25.4, we explain the substance of competition enforcement, using precedent-​setting competition cases to illustrate the application in practice. In sections 25.5 and 25.6, we discuss the interaction between competition and regulated sectors as well as competition and state-​ owned companies. In section 25.7, we briefly reflect on the limits of competition law and how competition advocacy is used by the competition authorities to create awareness about competition law and policy. In section 25.8 we conclude.

25.2  Context and Evolution of South Africa’s Competition Laws Although stemming from a basic need to confront the abuse of market power, the form and intensity of competition laws are shaped by a country’s experience. South Africa is a small open economy. Geographically, it is isolated from global markets. For much of its history, it has been an unequal, exclusionary, and racially skewed society. Economic policies in the nineteenth and twentieth centuries were shaped by a great dependence on industries built around extracting raw materials, including gold, iron, diamonds, and 4 Other government policies with a potential to significantly affect competition include, among others, industrial, consumer protection, standards, intellectual property rights (IPRs), trade and investment policy.

Competition Policy in South Africa    533 other minerals.5 State protection was used to exclude black-​owned firms from effectively participating in markets. Throughout the twentieth century, racial discrimination and state protection were combined to keep the vast majority of South Africans from participating in the economy. Product markets and ownership of capital remained highly concentrated. For example, in the years of apartheid (1948–​94), the economy was dominated by a few diversified conglomerates, predominantly in mining, finance, and industry.6 South Africa’s first competition law statute, the Regulation of Monopolistic Conditions Act, was adopted in 1955.7 It identified several anti-​competitive practices but none of them was prohibited, in and of themselves. It was very cautious. The framework of analysis was whether an anti-​competitive practice was in the public interest, but public interest was not defined. The minister of trade and industry decided who and what was investigated, including what remedies, if any, would be imposed. Over the twenty years between 1955 and 1975, only eighteen investigations were ordered by the minister and only in one case was action taken.8 The 1955 Regulation of Monopolistic Conditions Act was replaced by the Maintenance and Promotion of Competition Act of 1979. It established the Competition Board and introduced actions against anti-​competitive mergers, but the framework of analysis remained whether conduct or a merger was in the public interest. Again, public interest remained undefined. But the Competition Board could now decide to investigate anti-​competitive practices on its own. Between 1979 and 1998, the Competition Board produced seventy-​five formal reports. Post-​1994 mistrust of concentrated markets and large white-​owned firms, who had benefited from the past privileges, left South Africans with something of a dilemma. Economic power concentrated in a few white hands was offensive. Against this background, the first democratic government gave high priority to the project of redressing economic imbalances corresponding to concentrated markets and racial divisions. The Competition Act of 1998 (Competition Act) became effective in September 1999. The application of the Competition Act is the responsibility of independent institutions, the Competition Commission (Commission), the Competition Tribunal (Tribunal), and the Competition Appeal Court (CAC). The Commission mainly investigates and prosecutes. The Tribunal largely adjudicates and reviews the decisions of the Commission, while the CAC hears appeals and reviews the decisions of the Tribunal. In February 2019, the Competition Amendment Act (2018) was signed into law.9 Up until the Competition Amendment Act (2018), the public interest goals were only textually knitted into the merger control and the exemption provisions of the Competition Act.10 The Amendment Act maintains the architecture of the Competition 5 

OECD (2003). OECD (2003). 7  Lewis (2012). 8  OECD (2003). 9  The main focus of the Competition Amendment Act is economic transformation. 10  Fox (2018). 6 

534    Liberty Mncube and Nicola Theron Act while aligning the operations with the stated purpose, particularly with regards to public interest goals.

25.3  Multiple Goals of Competition Law As mentioned earlier, many countries have included multiple goals (in addition to efficiency) in their Competition Law regimes. The economics profession has demonstrated as early as the eighteenth century, that competition in markets increases efficiency, maximizes a society’s resources and allocates them to their best use. The idea that active competition policy can strengthen markets and efficiency is not disputed. The question, however, is whether this should be the single goal of competition enforcement, to the exclusion of all other equity and justice goals. The answer to this question relies mainly on whether competition law is seen as an instrument to rectify huge concentration of economic power and inequality. In the case of South Africa this choice has been made, and South Africa has chosen to include multiple policy goals.11 The motivations of competition law are recorded in the preamble of the Competition Act.12 The preamble details relevant contextual facts (apartheid and other discriminatory practices of the past resulted in excessive concentration of the economy).13 The preamble labels restrictions on free and full participation in the economy by all South Africans as ‘unjust’. Importantly, the preamble suggests the kind of outcomes that South Africa seeks to achieve through its competition law. It explains that a credible competition law and institutions are essential for an efficient economy. It explains further that an efficient economy balancing the interests of workers, owners, and consumers, and focused on development, will benefit all South Africans. It is important to point out that South Africa is not unique in its emphasis on public interest goals as part of competition law. In the European Union, there are also multiple factors considered as part of competition law, such as the promotion of market integration and environmental reasons.14 The critical question is whether the inclusion of multiple goals has detracted from the economic-​efficiency rationale. The choice of a competition law goal or goals drives the outcomes. In South Africa, the principal objective of competition law is stated as ‘to promote and maintain competition’. This suggests that competitive markets, as an alternative to more intrusive 11 

South Africa’s first democratic government intentionally made competition regulation its favoured means of regulating private and public companies in the public interest. 12  In the United States, for example, the political motivations of the Sherman Act reflect the concerns for small businesses and farmers who blamed the trusts for many economic misfortunes. See, generally, Gotts (2019) and Hovenkamp (1988). 13  Buthelezi et al. (2019) point out that many of South Africa’s key sectors remain highly concentrated. 14  Motta (2004), c ­ hapter 1.

Competition Policy in South Africa    535 government regulation or control, should allocate resources. The principal objective is accompanied by six goals aimed at creating an inclusive economy. The first goal is about economic efficiency (recognizing the total welfare standard). The second goal is concerned with competitive prices and choices for consumers (recognizing consumer welfare and consumer choice standards). The other four goals are concerned with equity (or public interest). The first public interest goal is to promote employment. The second public interest goal is to expand the opportunities to participate in world markets. The third public interest goal is to ensure equitable opportunities for small and medium enterprises (SMEs) to participate in the economy, and the fourth public interest goal is to promote a greater spread of ownership, in particular to increase ownership stakes of historically disadvantaged persons. The principal objective ‘to promote and maintain competition’ is an intermediate objective, while economic development, capturing all six goals, is a final aim. Each society elects value choices that motivate its competition law and reflect the problems that are important for its economy. Some critics of multiple goals point out apprehensions about predictability15 and administrability16 of enforcement decisions. Even if the apprehensions are appropriate, which is an open question, such concerns reflect a misunderstanding of multiple goals. The fact that competition law explicitly encompasses both efficiency and public interest goals requires implementation coherence. This helps avoid tensions and duplication of efforts, or indefensible softening of the competition process as the principal driver of economic development. The Competition Act specifies what qualifies as public interest to address the problem of discretionary bounds.17 In merger control, for example, a complete competition analysis is done first and separately. The competition analysis is followed by a public interest analysis. This separation makes it clear what a competition analysis requires and what a public interest analysis requires. Because of this separation, it is possible to observe trade-​offs.18 To third parties, the separation makes it possible to predict competition enforcement outcomes. The separation has, in our view, advanced the cause of both competition and public interest. Lewis (2012), correctly, points out that the exclusion of public interest goals, for example in protecting jobs, ensuring an equal opportunity for small businesses to participate, and empowering the historically excluded and disadvantaged population, would have meant no competition law in South Africa. The design of South Africa’s 15 

Predictability looks at whether decisions are likely to be consistent in comparable cases and therefore means that outcomes will be predictable. 16  Administrability considers whether firms and competition authorities can implement standards in an analytically cost-​effective manner. 17  See First and Fox (2015). 18  For example, in Tepco, after finding no competitive harm from an acquisition, the Tribunal was able to assess the claim that the acquired black enterprise’s trademark should be separately preserved and used by the acquiring company. It found that to do so would be costly and the principal would impose such costs on prospective acquiring companies that the value of black businesses would be reduced. Shell South Africa (Pty) Ltd and Tepco Petroleum (Pty) Ltd (66/​LM/​Oct01) [2002] ZACT 13 (2002).

536    Liberty Mncube and Nicola Theron competition law allows it to fit the socio-​economic characteristics of South Africa. This enhances its credibility and legitimacy.

25.4  Confronting the Exercise of Anti-​competitive Market Power Enforcing competition law involves enforcing prohibitions related to anti-​competitive practices and regulating mergers. Anti-​competitive practices may involve horizontal restraints, vertical restraints, or abuse of dominance, and are backward-​looking in nature. We briefly review competition law enforcement instruments (including forward-​ looking merger control), below.

25.4.1 Merger Control Merger thresholds classify mergers into large, intermediate, or small categories. This classification is useful in determining the process of approval by the competition authorities. For mergers categorized as large mergers, the Commission makes a recommendation to the Tribunal on whether a merger should be approved (with or without conditions) or prohibited. Small mergers do not need to be notified to the Commission, but the Commission may sometimes request notification of a small merger if it considers that the merger may lead to a substantial lessening of competition or may not be in the public interest. If the Commission decides to prohibit or decides to conditionally approve an intermediate merger, any party to the merger may request the Tribunal to reconsider this decision. Following the Tribunal’s decision, any party to the merger may request the CAC to reconsider this decision. If one considers the period 1999 to 2019, only twelve mergers were prohibited by the Tribunal, out of 1,537 cases decided. The vast majority of the mergers have been approved (or conditionally approved). Over the period 1999 to 2020, in only two of the five instances wherein the CAC was asked to consider the Tribunal prohibition ((1) the proposed Mondi Ltd and Kohler Cores and Tubes (a division of Kohler Packaging Limited) merger, and (2) the proposed Imerys South Africa (Pty) Ltd and Andalusite Resources (Pty) Ltd merger)), has the CAC agreed with the decision of the Tribunal to prohibit a merger. In another three instances, the CAC overturned the Tribunal’s decision to prohibit the mergers.19

19  See e.g. the following cases: (1) Medicross Healthcare Group (Pty) Ltd and Prime Cure (Pty) Ltd (Tribunal case number 55/​CAC/​Sept05); (2) Schumann Sasol (South Africa) (Pty) Ltd and Price’s Daelite (Pty) Ltd (Tribunal case number 10/​CAC/​Aug01); and (3) Pioneer Hi-​Bred International and Another v Competition Commission (Tribunal case number 113/​CAC/​Nov11).

Competition Policy in South Africa    537 The majority of mergers raises no substantial lessening of competition and are routinely approved without conditions. However, in many contested merger cases, the Commission is forced to litigate cases that are clearly anti-​competitive. In an ideal world, the merging parties would not even consider requesting approval of an anti-​ competitive merger in the first place, or would abandon the request for approval when the Commission makes its concerns known that the merger will result in a likely substantial lessoning of competition or a negative public interest. The fact that the Commission must litigate anti-​competitive mergers points to a limitation in the current framework.20 The Competition Act prohibits mergers that are likely to substantially prevent or lessen competition unless outweighed by efficiency gains or justified on certain public interest grounds. The substantial lessening of competition (SLC) test focuses on increased market power as the competitive harm to be prevented. Market power, in the context of merger analysis, may be defined as the ability to increase prices profitably, reduce quality, reduce innovation, or reduce consumer choice from pre-​merger levels for a significant period. This may arise through the individual decisions of the merged firm and its competitors or through coordinated behaviour. Merger analysis is predictive and requires the employment of an appropriate counterfactual.21 In merger cases the assessment of the relevant counterfactual is an essential part of the analysis. This involves a comparison of market outcomes that would prevail without the merger. In many cases, competition authorities usually take the status quo as the counterfactual to be compared with the scenario that is likely to prevail post-​merger. The difference between the two scenarios informs the threshold question—​whether the merger would lead to a likely substantial lessening of competition. It is possible for a merger to increase efficiency while competition is lessened. Efficiencies are balanced against the competitive detriment and the combined effects assessed to see if, overall, consumers benefit (or total welfare increases). The key case for guidance on efficiencies is the Trident/​Dorbyl merger.22 Merging parties have to prove that efficiencies will outweigh the anti-​competitive effects. In the assessment of whether the claimed efficiencies will outweigh the likely anti-​competitive effects of the merger, competition authorities consider whether the claimed efficiencies (1) constitute real efficiencies; (2) are verifiable (i.e. are capable of measurement); and (3) benefit consumers. Further, efficiencies are expected to be timely, likely, and sufficient to outweigh the likely substantial lessening of competition.23 Efficiency claims must also be merger specific and therefore be a direct consequence of the merger.

20 

Grimbeek et al. (2013) point out that the Commission is less likely to approve mergers that they link to markets that are less contestable. 21 See Life Healthcare Group (Pty) Ltd and Joint Medical Holdings Ltd (Case No: 74/​LM/​Sep11). 22  Tribunal’s Decision: Case no: 89/​LM/​Oct00. 23  See CAC decision in Pioneer/​Pannar: 113/​CAC/​Nov11, page 22.

538    Liberty Mncube and Nicola Theron A merger that is not likely to give rise to a substantial lessening of competition may still be prohibited on substantial public interest grounds.24 The public interest grounds are limited and include the effect on: promotion of a particular industrial sector or region; promotion of a greater spread of ownership, in particular to increase the levels of ownership by historically disadvantaged persons and workers in firms in the market; promotion of employment; promotion of the ability of small businesses or firms controlled by historically disadvantaged persons to effectively enter into, participate in, or expand in the market; and promotion of the ability of national industries to compete globally. Remedies may be unilaterally imposed on merging parties by competition authorities or agreed jointly with the merging parties for mergers which substantially lessen competition or negatively affect public interest. Measures that aim to restore or maintain the competitive structure of the market are classified as structural remedies. Ongoing measures that are intended to control or limit the behaviour of merger parties are classified as behavioural remedies. The purpose of remedies is to address possible anti-​competitive effects of a merger or negative public interest effects. For example, in the Wal-​Mart Inc and Massmart Holdings Ltd merger, following concerns that the retrenchments of 503 employees by a division of Massmart in June 2010 occurred because of the merger, the CAC ordered that the employees retrenched by Massmart be re-​employed. In relation to concerns about the effect of the merger on local suppliers to Massmart, the CAC ordered the capital amount of the fund to be used to support local procurement and to be increased to R200 million. Further, that fund were to be spent over a period of five years. While very few mergers have been prohibited outright, conditional approvals have increased over time. Many of the conditional approvals related to public interest concerns. For example, a fixed period where no retrenchments are allowed. In a country with very high unemployment rates, a strong focus on employment retention is probably justified, especially if dealt with via a condition and not an outright prohibition. It is too early to comment on the application of the Competition Amendment Act (2018) and its increased focus on public interest criteria in merger control, but this is a developing field which researchers will continue to monitor closely.

25.4.2 Abuse of Dominance The abuse-​of-​dominance provision prohibits dominant firms from engaging in conduct that anti-​competitively excludes rivals (takes advantage of market power to prevent rivals’ access to markets) or that exploits consumers (takes advantage of market 24  For example, in proposed merger between Anglo American Holdings Ltd and Kumba Resources Ltd, after the Tribunal concluded that the merger was unlikely to result in any substantial lessening of competition, it stated that it must nevertheless evaluate whether the merger can be prohibited on public interest grounds (Tribunal case 46/​LM/​Jun02 para 137 to 139).

Competition Policy in South Africa    539 power to charge excessive prices or discriminate among customers). Enforcing abuse-​ of-​dominance provisions, in South Africa and in other parts of the world, is one of the most complicated and contentious areas of competition policy in part because of the evidential burden of proving such abuses.25 Over the period 1999 to 2020, the Tribunal has adjudicated twenty-​one litigated cases of abuse of dominance (excluding settlements). Of the twenty-​one adjudicated cases, the Tribunal dismissed seven cases, finding no contravention. The Tribunal found a contravention on fourteen cases. Ten of the fourteen cases in which the Tribunal found a contravention were appealed to the CAC. In four cases, the CAC overturned the decision of the Tribunal. The CAC remitted one case back to the Tribunal for a thorough evaluation. In the remaining five cases, the CAC found abuse of dominance. The investigation and litigation involving abuse-​of-​dominance cases have been protracted, in most cases exceeding five years. This means that even where the complainants have been successful, the difficulty and cost of those successes illustrate that the system places less emphasis on the benefits of competition and concerns of increased market power. A firm is dominant in a market if it has a market share of 45 per cent or more. If a firm has a market share of between 35 and 45 per cent it is assumed to be dominant unless the firm can show that it does not have market power. A firm with a market share below 35 per cent, is assumed not to be dominant unless it has market power. The Competition Act defines market power as the power of a firm to control prices, exclude competition or to behave to an appreciable extent independently of its competitors, customers, or suppliers. A dominant position is not in itself a contravention of competition law, it is only the abuse of a dominant position that is prohibited. A firm that has a dominant position in a market has a special responsibility to ensure that its conduct does not result in an anti-​competitive effect that outweighs the pro-​competitive justification. Examples of behaviour that would amount to abuse of a dominant position include charging excessive prices, price discriminating, engaging in exclusionary conduct, and refusing a competitor access to an essential facility when it is economically feasible to do so. The first prohibited abuse-​of-​dominance practice is charging an excessive price to the detriment of consumers or customers by a dominant firm. An excessive price is determined by reference to whether the price concerned is higher than a competitive price, and whether such difference is unreasonable. A competitive price is determined by considering benchmarks such as the cost of producing that good or service (plus a

25 To

illustrate, between 2003 and 2014, Europe’s enforcement decisions in abuse of dominance accounted for only 20 per cent of all enforcement actions compared to 48 per cent of cartel enforcement. Other jurisdictions such as the United States, focus less on this area of enforcement, in part due to the dominant view that market exploitation by dominant firms attracts opportunities for new entrants and erodes market power. Roberts (2012) examines the record of abuse-​of-​dominance cases in South Africa after more than a decade of the new competition regime. He points out that the South African experience raises the question of whether conduct by a dominant firm, which has different dimensions, can be readily pigeon-​holed in the way anticipated by the legislation.

540    Liberty Mncube and Nicola Theron reasonable mark-​up) and the prices charged (by the respondent or relevant comparator firm) in similar but competitive markets. The Harmony Gold Mining Company Ltd v. Mittal Steel South Africa Ltd case was South Africa’s first adjudicated case dealing with allegations of excessive pricing. Harmony Gold lodged a complaint against Mittal Steel, alleging that Mittal (a near monopoly in the steel market) was selling its flat steel products at an excessive price. Mittal was found to have priced domestic sales of flat steel at import parity prices, whilst selling flat steel products for export markets at the much lower export parity price. According to Harmony Gold, the export parity price constituted the economic value of the product. The price on domestic sales of flat steel products was excessive when compared to the export parity price. The Tribunal found that Mittal was ‘super-​dominant’ and that its import parity pricing was evidence of excessive pricing. On appeal, the Tribunal’s decision was set aside by the CAC and remitted back to the Tribunal.26 The matter was finally settled out of court before a final determination was made. The next notable excessive-​pricing case was the Competition Commission v Sasol Chemical Industries Ltd (the Sasol case), Sasol Chemicals (successor to the state-​owned firm and beneficiary of state privileges), was alleged to have charged excessive prices for its propylene and polypropylene products. The Tribunal found Sasol guilty of charging excessive prices, considering Sasol’s history and past privileges. On appeal, the CAC reversed the decision of the Tribunal. The CAC found that the prices in question did not contravene the Competition Act. In addition to the above litigated cases, the Tribunal has confirmed more than forty cases involving allegations of excessive prices. Most of these cases relate to excessive-​ pricing or price-​gouging concerns in the aftermath of the COVID-​19 crisis and the subsequent declaration of a state of National Disaster.27 To date, the most famous settlement arising from a concern about excessive prices is the Hazel Tau settlement in 2002. The settlement concerned allegations that the pharmaceutical companies had priced excessively for their patented drugs to treat patients during the dreadful HIV/​AIDS epidemic. The price of the drugs was alleged to have been excessive and was detrimental to patients. The Tribunal and the CAC did not get the chance to hear arguments on the case. The pharmaceutical companies settled the case and avoided opening their books to reveal their costs. There are several benefits of settlements in competition cases.28 The Commission saves resources that they would otherwise need to prosecute a competition case. The Tribunal and the CAC save resources related to hearings and producing fully reasoned decisions. For respondent firms, the benefits include a reduced fine and the 26  The

CAC stated that the Tribunal’s approach was ‘fundamentally flawed’ and that the Tribunal’s structural test and notion of ‘super-​dominance’ found no support in the Competition Act. The matter was referred back to the Tribunal to re-​evaluate the evidence in line with the requirements of the Competition Act. 27 See https://​ccle.sun.ac.za/​excessive-​pricing-​covid-​19/​ for a useful list of resources relating to COVID-​19 excessive pricing cases. 28  The settlement procedure is not limited to abuse-​of-​dominance cases, it also relates to horizontal and vertical restraints. Indeed, many cartels have been resolved through a settlement process.

Competition Policy in South Africa    541 ability to avoid a protracted, costly litigation that can distract management and generate negative publicity. The second prohibited practice relates to a dominant firm which refuses a competitor access to an essential facility when it is economically feasible to do so. The Competition Act defines an essential facility as a resource or infrastructure that cannot reasonably be duplicated and without access to which competitors cannot reasonably provide goods or services to their customers. An abuse of dominance relating to refusing a competitor access to an essential facility cannot be countervailed by efficiency gains. For example, in the Competition Commission v Telkom SA SOC Ltd case, the Tribunal found Telkom to have abused its dominant position in the telecommunications market by refusing to supply essential access facilities to independent value-​added network services (VANS) providers. The conduct took place between 1999 and 2004, a period during which Telkom was a monopoly provider. Telkom admitted that the facilities in question comprised an essential facility and that it was economically feasible for it to supply these facilities. Telkom’s only justification for its conduct was that, in its view, the Value Added Network Service Providers (VANS) provided services illegally and that this justified its refusal. The Tribunal concluded that Telkom’s conduct was contrary to the interests of both the VANS providers and their customers. VANS and their customers relied on Telkom for network services. Further, that it was unnecessary to show anti-​competitive harm in cases relating to refusing a competitor access to an essential facility. When different consumers are charged different prices for the same good and it costs the firm the same amount to produce and serve these consumers, this is called price discrimination. There are two grounds for prohibiting price discrimination. First, price discrimination perpetrated by a dominant firm is considered an abuse when the conduct has the effect of substantially lessening competition and the sale relates to goods or services of similar grade and quality to different buyers in an equivalent transaction. The price discrimination must be related to prices charged; discounts, rebates, or credit given or allowed; or payment method and terms of payment of goods and services. An interesting case of price discrimination that dealt with the application of the substantial lesson of competition test is the Nationwide Poles v Sasol (Oil) (Pty) Ltd case. The Tribunal found that Sasol had engaged in unlawful price discrimination in the sale of creosote. The complainant, Nationwide Poles, was a small player. Sasol was accused of offering discounts to larger customers which it did not extend to small customers. On appeal, the Tribunal was overturned by the CAC which did not find harm to competition, although Nationwide Poles as a small customer had suffered harm. Second, as an outcome of the Competition Amendment Act in 2018, price discrimination provisions also prohibit dominant firms from price discriminating when they sell to small and medium-sized businesses and to firms owned or controlled by historically disadvantaged persons if the effect is to impede the ability of such firms to participate effectively. Introduced in 2018, it is a bit too early to gauge the effect of the public interest test and how it will be enforced.29 29 

Competition Amendment Act 2018.

542    Liberty Mncube and Nicola Theron The Competition Act also prohibits a firm from abusing its dominant position through an exclusionary act. An exclusionary practice is defined as one in which a dominant firm impedes or prevents another firm from entering into or expanding within a market.30 Put differently, exclusionary practices include exclusive contracts, pricing strategies, and other actions of dominant firms that deter entry of new rivals, force current rivals to exit, or restrict them to niche markets. Exclusionary practices come in a wide range of varieties, many of which are separately categorized in the Competition Act, including engaging in margin squeeze, predation, and exclusive dealing. A dominant firm can violate the abuse-​of-​dominance provisions of the Competition Act if it engages in predation. Predation involves temporarily charging prices below an appropriate specified cost benchmark (such as average variable costs, average avoidable costs). Predation can harm competition and consumers: for example, harm could occur if a dominant firm priced below the appropriate cost to marginalize and force a rival to exit and then, following the exit of the rival, increase prices to supra-​competitive levels for a significant period. An interesting case on this prohibition is the Media 24 (Pty) Ltd v Competition Commission of South Africa case where the CAC overturned the Tribunal’s decision. The Tribunal had concluded that Media 24 had contravened the Competition Act by engaging in an exclusionary practice that involved selling a loss-​ making community newspaper below its average total cost (ATC). The Tribunal found that Media 24 engaged in exclusionary conduct similar to predatory pricing conduct when it sold below ATC with the purpose of excluding its rival from the market (this intention was captured in its strategy documents). The CAC held that intention is irrelevant and that pricing below ATC but above average variable cost is not exclusionary. An exclusive contract is a contract between a firm and its buyer whereby a buyer commits not to make any purchase from a competing firm. In some circumstances, exclusive contracts can be used by a dominant firm to foreclose a market, in fact, to monopolize a market and thereby prevent the dominant firm’s rivals from competing effectively. An interesting case involving exclusive contracts is the Competition Commission v Computicket (Pty) Ltd case where the Tribunal found Computicket to have contravened the Competition Act. Computicket had imposed exclusive agreements on its inventory providers (buyers). On appeal, the CAC agreed with the Tribunal and found Computicket’s exclusive contracts exclusionary. A dominant firm may offer rebates on all units of a single product conditioned upon the level of purchases. The rebates may be conditioned upon the quantity of product purchased or on the percentage of needs. In some instances, these rebates may be a strategy to foreclose the market and force rivals to exit the market. An interesting case on the anti-​competitive use of rebates is the Nationwide v South African Airways and Competition Commission v South Africa Airways cases where the Tribunal in both cases found that SAA’s incentive schemes, in which commissions were paid to travel agents to

30 

‘Exclusionary act’ is defined in the Competition Amendment Act (2018) as: ‘ “exclusionary act” means an act that impedes or prevents a firm from entering into, participating in or expanding within a market’.

Competition Policy in South Africa    543 incentivize them to book their clients onto SAA’s flights instead of rival airlines, such as Comair and Nationwide, were exclusionary. Up until this point, we have been discussing the exercise of market power by a firm participating in the market as a seller. The Competition Act also recognizes that market power can be exercised by a firm either as a seller or as a buyer. The buyer power provisions prohibit dominant buyers in certain sectors from requiring or imposing unfair prices or trading conditions on firms that are small and medium-sized businesses, or firms owned or controlled by historically disadvantaged persons. Introduced in the Competition Amendment Act in 2018, it is a bit too early to gauge the effect of the buyer power provisions and how they will be enforced.31

25.4.3 Horizontal and Vertical Agreements The most egregious infringements in competition law relate to collusion. Collusion between competitors is presumed to distort competition by allowing firms to exercise market power that they would otherwise not have and is not afforded an efficiency defence. Collusion is defined as an agreement between firms (explicit collusion) or a concerted practice (tacit collusion) by firms in a horizontal relationship (or competitors) involving price fixing,32 allocating markets,33 and collusive tendering.34 Mncube and Grimbeek (2016) discuss the prevalence and persistence of cartels in South Africa. They point out that South Africa’s first democratic administration took significant steps to liberalize many of the formerly price-​regulated markets. Deregulation and liberalization led to the break-​up of regulated cartels, but liberalization may have inadvertently, by increasing competition in formerly protected markets, also increased the incentives for firms to participate in cartels. Detecting collusion is the biggest challenge to enforcing cartel prohibitions. Unsurprisingly, firms seek to hide their involvement in collusive arrangements. The most successful tool that has been used to uncover cartels is the corporate leniency policy (CLP).35 CLP programmes grant complete or partial exemption from prosecution for firms that collaborate with the competition authorities. A CLP was first introduced in South Africa by the Commission on 6 February 2004 and modified in 2008. Very few cartels were investigated and prosecuted under the Competition Act before the

31 

Competition Amendment Act 2018. An example of a price-​fixing case is the Competition Commission v Pioneer Foods (Pty) Ltd case in which bread producers agreed to fix the prices of bread, maize meal, and wheat flour. 33  An example of a market-​allocation case is the Competition Commission v Pioneer Fishing (Pty) Ltd case where firms agreed to allocate the supply of horse mackerel in different territories such as Limpopo, Mpumalanga, and the North West Provinces. 34  An example of a collusive tendering case is the uncovering of bid-​rigging in the construction industry, which led to the fast-​track settlement process in February 2011. 35  See Competition Commission, Corporate Leniency Policy section 3.1 (2004). 32 

544    Liberty Mncube and Nicola Theron introduction of the CLP. The secretive nature of cartels, which makes it difficult to both detect and investigate cartels, is perhaps one way to explain the above observation. The CLP is meant to destabilize cartels by encouraging firms, of their own accord, to defect and report their behaviour to the Commission. The CLP allows a cartel member to receive immunity from prosecution before the Tribunal and from an administrative fine. The cartel member is required to disclose all relevant information and evidence relating to the workings of the cartel in return. Granting of immunity is a continuous process. It only concludes when a final determination is made by the Tribunal or if the decision is appealed, or a final judgement is made by the CAC. Detecting collusive arrangements is combined with sanctions to reduce the extent of collusive practices in the economy. Deterrence is the primary motivation of sanctions imposed on firms found guilty of participating in cartels. Firms found guilty of collusion are subject to an administrative penalty of up to 10 per cent of the firm’s annual turnover in the preceding financial year. The Tribunal considers the following aggravating factors when it determines what should be the level of an appropriate penalty: (1) the nature, duration, gravity, and extent of the cartel conduct; (2) whether the cartel conduct has resulted in any loss or damage; (3) the behaviour of the firm; (4) the market circumstances in which the cartel conduct took place; (5) the level of profit derived by the firm from the cartel conduct; (6) the degree to which the firm has co​operated with the Commission and the Tribunal; and (7) whether the firm has previously been found in contravention of the Competition Act. In addition to fines, competition authorities can also order structural and behavioural remedies. The Competition Act allows for the possibility that any person in a position of having management authority within the firm could face criminal charges if they either cause the firm to participate in collusive conduct, or knowingly accept collusive conduct. The penalty for a person found guilty could include a fine with a monetary value of up to R500,000 or a ten-​year maximum prison sentence, or both may be imposed. The possibility of imprisonment provides a helpful supplement in achieving deterrence because of the inadequacy of fines and remedies. The Competition Act makes a distinction between collusive arrangements from all other horizontal arrangements that may have the effect of substantially preventing or lessening competition in the market but could be justified on efficiency reasons. Such horizontal arrangements may include information exchange agreements, standard-​ setting agreements, cross-​licensing agreements, and joint ventures.36 Competition law also separates restrictive practices involving firms in a vertical relationship from those in a horizontal relationship. Vertical restraints are conditions and

36  An example of a case of a restrictive horizontal agreement (but not a collusion allegation) is the Netstar case. In this case, the Tribunal found that three vehicle tracking firms and the Vehicle Security Association had contravened the Competition Act by setting standards for anti-​theft devices in the market for stolen vehicle recovery, which created barriers to entry, prevented rivals from entering or expanding in the market and denied consumers the opportunity to benefit from lower prices, greater choice, and innovation.

Competition Policy in South Africa    545 restrictions on trade imposed by firms that are in a vertical relationship and serve two motives.37 On the one hand, they may be used to enhance market power. For example, reducing intra-​brand competition and inter-​brand competition, consumers unable to exploit alternative sources may face high prices. Forcing distributors to resell goods at minimum prices reduces intra-​brand competition and increases the opportunities for collusion. On the other hand, vertical restraints may be important to the realization of efficiencies. The Competition Act prohibits an agreement between firms in a vertical relationship if it has the effect of substantially lessening competition in a market, unless the agreement can be justified by efficiency considerations which outweigh the anti-​competitive effect. The complainant has the burden to prove an anti-​competitive vertical agreement. The respondent is permitted an efficiency defence to rebut the complainant. There have been very few cases litigated involving only concerns about anti-​competitive vertical agreements. A key question in many vertical-​restraint allegations has been: who has market power in a vertical relationship? Answering this question has been important in analysing anti-​competitive effects. For example, several abuse-​of-​dominance cases have also captured concerns relating to anti-​competitive vertical restraints. Only the practice involving a supplier prescribing to a downstream reseller the minimum price at which a good can be sold is prohibited by itself without any efficiency justification.38

25.4.4 Exemptions The system of prohibitions described above relating to abuse of dominance, horizontal, and vertical restraints is balanced by an arrangement for exemptions. There are three groups of exemptions that can be granted by the Commission: (1) public interest exemptions; (2)  intellectual property exemptions; and (3)  professional association exemptions. The Commission is required to grant an exemption for a stated term. On public interest exemptions, grounds for exemption include (1) maintenance or promotion of exports; (2)  promotion of entry into, participation in, or expansion within a market of medium-sized businesses and small businesses or firms controlled by historically disadvantaged persons; (3) changing capacity to stop decline in an industry; (4) economic development, growth, transformation, or stability in a designated industry; and (5) promote 37 Examples

of different forms of vertical restraints include, resale price maintenance, exclusive contracts, exclusive territories, slotting allowances, and tying and bundling. 38  To establish a minimum resell price maintenance case, the complainant is required to establish (1) a minimum resale price; (2) the implementation of the practice of minimum resale price maintenance; and that there are ways in place to enforce or maintain the practice. An example is the Federal Mogul Aftermarket Southern Africa (Pty) Ltd v Competition Commission case. Federal Mogul, a wholesale distributor of a range of motor vehicle components, imposed on its distributors a minimum price at which they were compelled to sell its products. The Tribunal found that the practice of minimum resale price maintenance had been established and imposed a fine on Federal Mogul. On appeal, the Competition Court upheld the Tribunal’s decision.

546    Liberty Mncube and Nicola Theron employment or industrial expansion. First, the Commission is required to establish whether the restrictive practice is essential in order to achieve the public interest objective. Second, the Commission is required to establish whether the agreement or practice will actually contribute to the objective. To date, no application relating to vertical-​restraints prohibitions or abuse-​of-​dominance prohibitions has been granted by the Commission.39 The exemptions granted have all related to horizontal-​restraint prohibitions. The Competition Act also makes provision for professional associations to apply for exemption for their rules if the rules relate to restrictive horizontal or vertical practices. Intellectual property (IP) is an important element of innovation and competition policy. IP law tolerates the creation of market power, while competition law responds to the abuse of market power. The IP exemptions relate to exempting agreements or practices involving the exercise of IP and is key to policy coherence. Very few exemptions have been granted since 2000. One example is the National Hospital Network (NHN) exemption. The NHN is a non-​profit company. It is a co-​ operative venture that is controlled by its members. Its members are a group of independent private hospitals. This group operates medical establishments such as day clinics, sub-​acute facilities, and psychiatric facilities. These members are competitors in the provision of private health-​care services. On 1 November 2018, the Commission conditionally granted the NHN a five-​year exemption commencing from 1 November 2018 to 31 October 2023. The NHN sought the exemption in order to engage in collusive conduct on the basis that the exemption would promote the ability of small businesses or firms controlled or owned by historically disadvantaged persons to be competitive. The 2018 exemption was not the first exemption granted to the NHN. The Commission had previously granted the NHN three exemptions on the premise that the exemptions were required for the objective of promoting the ability of small businesses or firms controlled or owned by historically disadvantaged persons to be competitive. The first exemption was granted in June 2006 for a period of five years. The second exemption was granted in May 2010 for an additional five-​year period. The third exemption was granted in October 2014 for a four-​year period. The system of exemptions has helped alleviate tensions between competition policy and other economic development policies, in particular industrial policy.40

25.4.5 Market Inquiries Market-​inquiry provisions were only introduced in the Competition Act in 2013 and allow the Commission to analyse the state of competition in a particular market rather than the conduct of individual firms in a market. The market-​inquiry provisions allow the Commission, for example, to analyse and address alleged structural concerns 39 

Correct as at 30 January 2021. For a detailed examination of industrial policy in South Africa, see ­Chapter 24 in this volume by Anthony Black. 40 

Competition Policy in South Africa    547 in a market. The Commission can conduct a market inquiry if any feature or combination of features in a market adversely affects competition in that market. A feature of a market includes: (1) the structure of the market, including levels of concentration and barriers to entry in a market; (2) The outcomes observed in the market, such as ownership, prices, innovation, employment, and the ability of national industries to compete in international markets; and (3) the conduct in that or any related market. In a market inquiry, the Commission is required to decide whether any feature, including structure and levels of concentration, impede, restrict, or distort competition. The Commission is required to consider the impact of the adverse effect on competition on small and medium-sized businesses, or firms controlled or owned by historically disadvantaged persons when making its decision. The Commission may, in relation to each adverse effect on competition, take action to remedy the adverse effect on competition. The Commission can also make recommendations for a change of policy, legislations, and regulations as well as recommendations to other regulatory authorities. South Africa’s economy is laden with barriers to entry and its markets are highly concentrated.41 Market inquiries represent a useful tool for competition authorities to develop a better understanding of markets. Our view is that it is a good idea to study the markets and see what can be done without losing the organic efficiencies of integration. We note, however, that the recommendations from recently concluded inquiries, such as the LP Gas Market Inquiry, the Healthcare Market Inquiry, the Retail Market Inquiry, and the Data Services Market Inquiry (DSMI), had rather muted recommendations, after years spent on investigating specific market features. The recommendations from these inquiries have to be seen in the light of the fact that these industries usually have multiple sector regulators. In the case of the DSMI, the Inquiry by the Commission, coincided with a separate Inquiry by the sector regulator, ICASA, the so-​called Broadband Inquiry. In fact, some of the recommendations of the DSMI were that ICASA should further investigate certain aspects and findings. Similarly, the Healthcare Market Inquiry made certain findings that related to the role of the statutory body, the Council for Medical Schemes (CMS), and its failure to regulate certain aspects of health-​care markets properly. Issues around joint jurisdiction have not been settled in the application of Competition Law in South Africa.

25.5  Competition and State-​owned Companies The Commission is empowered to consider complaints against state-​owned companies where a state-​owned company participates in a market as a firm providing goods or 41 

Roberts (2017) studies barriers to entry in different markets in South Africa to consider the nature and extent of these barriers and the implications for competition policy.

548    Liberty Mncube and Nicola Theron services. In the AEC Electronics (Pty) Ltd v The Department of Minerals and Energy42 case, the Tribunal concluded that competition authorities cannot review the exercise of state power by state functionaries. This suggests that competition authorities cannot intervene in markets where state-​owned companies benefit from market-​distorting decisions made by government or regulatory bodies. Competition authorities can intervene in complaints relating to the conduct of state-​owned companies and have done so, on several occasions. Below, we offer a brief review of some of the interventions involving state-​owned companies. First, we deal with the case of Telkom. Telkom is a vertically integrated provider of telecommunications services in South Africa. The South African government holds the largest shareholding. Telkom is subject to regulation by the Independent Communications Authority of South Africa (ICASA). Telkom’s conduct has been the subject of several complaints. The complaints relate to allegations that Telkom is abusing its dominance. As an outcome of these complaints, Telkom has been found guilty of engaging in exclusionary conduct. It has also admitted to engaging in several abuse practices, including margin squeeze, in a settlement with the Commission which was confirmed by the Tribunal in 2013.43 The settlement remedies included the implementation of functional separation between Telkom’s retail and wholesale divisions along with the implementation of a transparent transfer pricing programme to ensure non-​ discriminatory service provision by Telkom to its retail business and rivals. Second, there is the case of Eskom. Eskom is a state-​owned power utility, responsible for generating 96 per cent of the country’s electricity requirements. Eskom is subject to regulation by the National Energy Regulator of South Africa (NERSA). Eskom enjoys a near monopoly in generating electricity. It is a vertically integrated firm. Its operations include generation and distribution and it owns the transmission network. Eskom accounts for about 60 per cent of distribution; the remainder is distributed by municipalities. Over the years, the Commission has received several complaints relating to Eskom’s refusal to sign purchasing agreements with Independent Power Producers (IPPs). To date, the Commission has not intervened on concerns about Eskom’s abuse of dominance. Although there is clear evidence that it should.44 Third, another state-​owned company with a specific history of competition issues is SAA. SAA (South African Airlines) is South Africa’s national airline. SAA is wholly owned by the government. Over the years, SAA’s conduct has been the subject of several complaints. The complaints have related to allegations that it is abusing its dominance, it is engaged in cartel behaviour, as well as allegations that state support (bail​outs) to SAA is distorting competition in the airline industry. SAA has been found guilty of engaging in abuse-​of-​dominance and collusion conduct, not once but twice for each offence. SAA has been a serial repeat offender. Financial penalties do not seem to deter a desire to engage in anti-​competitive behaviour at SAA. 42 

AEC Electronics (Pty) Ltd v Department of Minerals and Energy 48/​CR/​Jun09. Tribunal case number: 016865. 44  Correct as at 30 January 2021. 43 

Competition Policy in South Africa    549 While competition authorities have treated state-​owned companies and private firms equally, focusing on the conduct of the firms rather than the identity, important challenges remain. For example, state support of failing state-​owned companies is contrary to the essential principle of competition law and policy that all firms, regardless of state ownership, should compete on merit.

25.6  Competition and Regulated Sectors In regulated sectors, the Competition Act applies even where the competition issues in question may be seen as falling within the purview of a sector regulator. When the Competition Act first came into effect, it excluded conduct subject to public regulation but was amended in 2000 to make provision for concurrent jurisdiction with sector regulators. Over the period 1999 to 2020, the Commission has asserted jurisdiction in many cases involving concurrency. At the same time, it has also signed memoranda of understanding with sector regulators spelling out how concurrent jurisdiction would be managed and how coordination would operate.45 An interesting case relating to concurrency arose in February 2009. The Commission referred an abuse-​of-​dominance case against Telkom. Telkom did not answer the Commission’s case, preferring to challenge the Commission’s decision. Telkom argued that the Commission failed to adhere to the memorandum of understanding it had signed with ICASA. Telkom argued that the agreement required ICASA to lead, given ICASA’s sector knowledge. The Supreme Court of Appeal ultimately decided the case. The Supreme Court of Appeal affirmed the competition authorities as the institutions mandated to enforce the Competition Act but did not make a finding on whether the memorandum of understanding precluded the Commission from investigating and referring the complaint to the Tribunal. Mergers in the banking sector hold a special position in relation to the Competition Act. The regulation of banks is largely the responsibility of the South African Reserve Bank. The Banks Act of 1990 grants the minister of finance and/​or the registrar of banks jurisdiction over bank mergers, subject to certain provisions. The Competition Act was amended in 2001 to reflect the special status held by mergers in the banking sector. Competition authorities cannot decide on mergers involving banks if the merging parties have obtained the permission from the minister of finance, as per the Banks

45  It has signed a memorandum of understanding with ICASA, NERSA, the Ports regulator, the Council for Medical Schemes, the National Gambling Commission, and the National Liquor Authority, among others.

550    Liberty Mncube and Nicola Theron Act, or the minister of finance decides that the bank merger concerned is in the public interest.

25.7 Competition Advocacy Competition authorities cannot interfere with policies adopted and implemented by government; they are not empowered to strike down government policies, regulations, and legislation even if these policies, regulations, and laws lead to a substantial prevention or lessening of competition. In response to restrictions on competition imposed by government, competition authorities can only advocate the benefits of competition to society, as well as to decision-​makers. Competition advocacy refers to activities conducted by competition authorities associated with the promotion of a competitive marketplace by means of non-​enforcement mechanisms. Competition advocacy can be an effective tool to enhance compliance with the Competition Act, as well as creating a competition culture. From the beginning, the Commission has seen the value of advocacy, policy engagements, and education. Each year, it launches a number of initiatives including information campaigns aimed at communicating the work of the competition authorities, programmes to respond to and influence government policies and draft legislation, workshops with key stakeholders on the benefits of competition policy, and drafting guidelines on how it will interpret important sections of the law. In other words, every year the Commission engages with key stakeholders in order to promote voluntary compliance with the Competition Act, both in the public and the private sector. Competition issues can arise during government policy formulation and implementation. To date, the Commission has taken seriously its duty to sensitize policymakers on possible competition law synergies and/​or trade-​offs which arise from certain policy measures. It should continue advocating on behalf of the competition introduced by new entrants, building strategic partnerships with government, business, consumers, and labour so that it can promote the work of the competition authorities and create a competition culture. It should continue undertaking strategic market inquiries to increase transparency in markets and identify competition distortions. It should continue safeguarding opportunities to participate in markets by advocating against prescriptive regulatory regimes.

25.8 Conclusion After more than twenty years of the new competition regime in South Africa, many markets still do not function well. Many sectors of the economy are characterized by high barriers to entry, concentrated markets, presence of (current and former)

Competition Policy in South Africa    551 state-​owned companies,46 informal markets, and regulations that limit competition. Corruption remains a persistent and stubborn problem. To the basket of problems and challenges, add the problem of inequality of wealth and economic opportunity, which is also widespread. Access to competitive and inclusive markets stands side by side with access to food, health, shelter, education, environment, and infrastructure, among others, as critical tools to combat South Africa’s economic challenges. Over the period 1999 to 2020, South Africa has built a credible competition law and effective structures to administer the law. It did not happen instantaneously. There has been a great deal of learning by doing. The enforcement of competition law has done much to bring better and more affordable goods and services to all South Africans and to create an environment that encourages economic participation by all South Africans and increases the competitiveness of South African business. Yet the goal of restructuring South Africa’s economic order by controlling private and public enterprise in the public interest to create an efficient, inclusive competitive economy focused on development, remains elusive.

References Buthelezi, Thembalethu, Thando Mtani, and Liberty Mncube. 2019. ‘The extent of market concentration in South Africa’s product markets’, Journal of Antitrust Enforcement, 7(3): 352–​64. Competition Amendment Act, 2018. Government Gazette. Act No. 18 of 2018. First, Harry, and Eleanor Fox, 2015. ‘Philadelphia National Bank, globalization, and the public interest’, Antitrust Law Journal, 80: 350–​1. Fox, Eleanor. 2018. ‘Competition policy at the intersection of equity and efficiency:  The developed and developing worlds’, The Antitrust Bulletin, 63: 3–​6. Gotts, Ilene Knable, 2019. ‘Back to the future:  Should the “consumer welfare” standard be replaced in U.S. M&A antitrust enforcement?’, Antitrust Report,  1–​2. Grimbeek, Sunel, Steven Koch, and Richard Grimbeek. 2013.‘The consistency of merger decisions at the South African Competition Commission’, South African Journal of Economics, 81(4): 561–​80. Hovenkamp, Herbert. 1988. ‘Antitrust’s protected classes’, 88 Michigan Law Review,  1–​48. Lewis, David. 2012. Thieves at the Dinner Table:  Enforcing the Competition Act:  A Personal Account. Jacarana Media. Mncube Liberty, and S. Grimbeek. 2016. ‘A history of collusion: The persistence of cartels in South Africa’, in Frederic Jenny and Yannis Katsoulacos (eds) Competition Law Enforcement in the BRICS and in Developing Countries. Springer: International Law and Economics. Motta, Massimo. 2004. Competition Policy:  Theory and Practice. Cambridge:  Cambridge University Press.

46  State-​owned companies benefit from privileges (state protection) that entrench dominant market positions which hinders rather than promotes competition and private-​sector development. In some markets where state-​owned companies are present, distortions of competition both for and in the market are exacerbated by regulatory capture.

552    Liberty Mncube and Nicola Theron OECD. 2003. ‘South Africa: Peer review of competition law and policy’. [online] http://​www. oecd.org/​southafrica/​34823812.pdf. Reekie, W. Duncan. 1999. ‘The Competition Act, 1998: An economic perspective’, South African Journal of Economics, 67(2): 257–​88. Roberts, Simon. 2012. ‘Administrability and business certainty in abuse of dominance enforcement: An economist’s review of the South African record’, World Competition, 35(2): 273–​300. Roberts, Simon. 2017. ‘Barriers to entry and implications for competition policy’, in T. Bonakele, E. Fox, and Liberty Mncube (eds) Competition Policy for the New Era: Insights from the BRICS Countries. Oxford: Oxford University Press.

Chapter 26

Regul ation of Net work Indu stri e s in Sou th A fri c a James Hodge and Tamara Paremoer

26.1 Introduction Network industries provide essential infrastructure for citizens and the economy, spanning utilities such as communications, energy, and transport (das Nair and Roberts 2017). Their key role in the economy and the belief that they exhibit natural monopoly features1 along with network effects,2 resulted in them being the subject of public provision historically. However, thinking changed in the 1980s and 1990s resulting in a wave of privatization and liberalization of network industries globally. These reforms sought to tap into private investment to develop the infrastructure and create competition where feasible to improve efficiency and innovation. This was accompanied by the establishment of economic regulators and toolkits with the twin aims of facilitating new entry and regulating market power in the interim (Kim and Horn 1999). The key challenges facing economic regulators of utilities are finding ways to constrain the monopoly power of incumbents (often through price regulation), ensuring universal coverage where rolling out the network is socially desirable but comes at a high cost, and incentivizing operational efficiency often through introducing competition at the network nodes where it is possible to do so. These regulatory questions are particularly challenging in developing countries because network industries often

1  Network industries are characterized by interdependent demand, increasing returns to scale, and negligible incremental cost associated with producing the marginal unit of output. 2  External economies in consumption mean that the utility derived from a service with network effects increases as more users join the system (Rohlfs 1974).

554    James Hodge and Tamara Paremoer provide essential services, meaning that decisions around service provision and pricing are thus politically important (Economides 2004). In South Africa, growing debt and fiscal constraints, along with losses at inefficient public enterprises, saw a policy shift to liberalization in the twilight of the apartheid era with the publication of the 1987 White Paper on privatization and deregulation (Gumede et al. 2016). This resulted in the corporatization of most network industry SOEs and opening competition in areas such as aviation and telecommunications. Given the hangover of debt and fiscal constraints to the new democratic government, this policy direction was perpetuated. However, the new democratic government also had to deal with the wide racial disparities in access to essential utilities, given the apartheid-​era focus on delivery to the white minority. This required a focus on ensuring investment to achieve universal access in services such as electricity and telecommunications. Where the international experience was predominantly positive, such as tele­ communications and airlines, liberalization proceeded with regulators overseeing monopoly components and facilitating the introduction of competition. In other areas, such as electricity, rail, and ports, reform stalled following a shift in policy thinking in 2002 to the Asian model of using SOEs and infrastructure to support the ‘developmental state’ (see ­Chapter  27 in this volume on SOEs). In these cases, regulators were left overseeing state monopolies with broad mandates that sometimes conflicted with regulatory principles of cost-​recovery and removal of cross-​ subsidies (Meyiwa and Chasomeris 2016). Government conflicts of interest resulted in a weakened regulatory oversight that could not prevent mismanagement and inefficiencies in core network infrastructure such as electricity and transport, as outlined in C ­ hapter 27 in this volume. This chapter explores the regulatory experience in telecommunications, electricity, and transport.

26.2 Telecommunications Telkom operated as a state telecommunications monopoly during the apartheid period. Following international trends, the South African telecommunications sector began its liberalization path in the early 1990s with the opening of the value-​added network services (VANS) and customer premises equipment in 1993, followed by the licensing of two mobile operators in 1994 (Horwitz 1997; Hodge 1999; Hawthorne et al. 2016). The mobile licences included a requirement to cover 70 per cent of the population within four years and a price cap on tariffs, restricting price increases to consumer price inflation (CPI) less a productivity factor set at zero per cent.3 At that stage, mobile phone 3  Price cap regulation involves setting the maximum aggregate increase in prices annually at CPI less a productivity factor. The productivity factor represents the expected efficiency improvements that could be achieved, expressed in percentage terms. An increase in productivity would warrant a lower price increase than cost inflation. The price cap applies to the basket of services to enable optimal relative tariff design by the operator.

Regulation of Network Industries in South Africa    555 service was considered to have a limited market and the absence of a regulator meant that regulation in line with best practice occurred through licensing. Broad consultation over reform for the fixed-​line sector was left to the new government, which embarked on a White Paper process and ultimately passed the Telecommunications Act 103 of 1996.

26.2.1 Fixed line The policy for fixed line was driven by the over riding priority of network expansion into underserviced areas to address the disparity in access. There would be a partial privatization of 30 per cent of Telkom to raise capital for investment and bring in management expertise to modernize the network and improve efficiency. Telkom would also be granted a five-​year exclusivity period to enable it to subsidize the roll-​out of 2.71 million lines in underserviced areas and to rebalance tariffs (PMG 2000). Telkom would be subject to a price cap during this period to prevent the abuse of monopoly power. The five-​year transition period would then be followed by gradual liberalization, with the introduction of a second national operator (SNO). The White Paper on Telecommunications Policy of 1996 proposed an independent regulator, the South African Telecommunications Regulatory Authority (SATRA), to oversee the monopoly period and gradual liberalization. While early drafts of the Telecommunications Act kept to this principle, the final version shifted this power to the minister. This shift has been put down to lobbying by the strategic equity partner (SEP) in Telkom and the government’s reluctance to cede control over the delivery of key goals in the Reconstruction and Development Plan (RDP) to a regulator (Horwitz and Currie 2007). The minister would set tariffs for the first three years, invite and issue major licences under gradual liberalization, and approve all SATRA regulations. This created problematic conflicts of interest given government shareholding and a desire to maximize the enterprise value (Cohen 2003). The minister set the initial price cap productivity factor for the first three years at a low 1.5 per cent with a maximum 20 per cent movement in any single tariff under rate rebalancing. The productivity factor used contrasted with actual labour productivity achieved of 10.8 per cent for the five-​year period (Hodge 2004). Whilst the regulator, at that point the Independent Communications Authority of South Africa (ICASA) following the merger of SATRA with the Independent Broadcasting Authority, could set the productivity factor for the next period, it was sabotaged from doing so effectively. The licence conditions provided scope for Telkom to delay providing cost of accounts (COA) required to inform the tariff regulation, and a lack of funding meant ICASA lacked the skills and resources necessary to undertake a vigorous exercise. ICASA ultimately proposed a 5 per cent productivity factor but the minister refused to approve the regulations and instead proposed a 1.5 per cent factor (ICASA 2005). In the meantime, Telkom pressed ahead with its own zero per cent productivity factor (Horwitz and Currie 2007). Retail regulation ultimately fell away following the introduction of n SNO in 2006.

556    James Hodge and Tamara Paremoer The lack of independent regulatory oversight enabled pricing closer to monopoly levels (Hodge 2004). This, combined with aggressive rate rebalancing that saw high residential price increases, ultimately undermined the universal service roll-​out as some 2m of the 2.7m lines were disconnected due to the unaffordable tariffs. Two other universal access initiatives also failed. The 1996 Telecommunications Act established the Universal Service Agency (USA) which would administer the operator contributions to the Universal Service Fund (USF). However, the USA, later renamed USAASA,4 lacked capacity and was undermined by a dual reporting mandate to the minister and ICASA. It is estimated that the first ten years saw it spend only a third of the USF funds, and predominantly on implementing telecentres against its own mandate (Lewis 2013). In 2001, there were amendments to allow under-​serviced area licensees to operate in areas with a fixed-​line teledensity of under 5 per cent.5 However, the lack of financial and regulatory support along with the growth of mobile prepaid service meant that by 2007 none of the original licensees were operational and the initiative was formally scrapped in 2009 (Lewis 2013). The liberalization of VANS in 1993 saw a proliferation of service providers in Internet access and virtual private networks (VPNs). VPNs were popular as it allowed businesses to share expensive Telkom infrastructure in the core VPN network. Under the SEP management aiming to maximize its returns over the exclusivity period, Telkom unsuccessfully sought to have the Internet declared an exclusive service and prevent the resale of infrastructure by VPN providers. Telkom did, however, use its position as exclusive infrastructure provider to place the VANs in a margin squeeze by entering into competition with them at retail rates below wholesale levels (Competition Tribunal 2012). Telkom also at one point stopped supplying new lines to VANS and even discontinued services to AT&T (Horwitz and Currie 2007). Whilst ICASA sought to make facilityleasing regulations in 2000, these were initially passed but then reversed by the minister. However, the lack of COA for Telkom also meant the proposed cost-​based pricing could not be determined or enforced. The monopolistic pricing of international bandwidth, through the SAT-​3 undersea cable required for Internet access, came to epitomize the infrastructure wholesale problem, with South Africa quickly slipping down the global rankings for Internet access (Esselaar et al. 2006). The South African VANS Association (SAVA) ultimately took their complaint to the Competition Commission in 2004 which referred a case of margin squeeze and excessive pricing of wholesale infrastructure. The Competition Tribunal decided in the Competition Commission’s favour in August 2012, fining Telkom R449  million (Competition Tribunal 2012). The Commission subsequently referred a further case for the period following this. At this point Telkom, under new management, sought to settle with the Commission. The result was a far-​reaching settlement agreement which sought to address high wholesale prices through almost Rand 1 billion in price cuts and margin

4  5 

Universal Service and Access Authority of South Africa. Telecommunications Amendment Act No. 64 of 2001.

Regulation of Network Industries in South Africa    557 squeeze through the functional separation of the wholesale and retail businesses along with transfer and retail pricing rules based on regulatory best practice (Competition Tribunal 2013). The settlement effectively achieved a degree of wholesale regulation that the sector regulator had failed to put in place. The shift in thinking to a developmental state and the use of SOEs to expand infrastructure became evident in the SNO process (Gillwald 2007). Instead of selling the communications infrastructure of Eskom (Esitel) and Transnet (Transtel) to the SNO, government took a 30 per cent stake in the SNO, Neotel, in exchange for the metro fibre assets, and created a new state entity, Infraco, which would lease the national transmission fibre network to Neotel for five years. The result was that by the end of the managed liberalization period, government would have a sizeable stake in all three infrastructure providers and the largest mobile network operator (MNO), Vodacom, through Telkom. This reduced the incentives for government to drive a real reform agenda and may account for the lack of resources and independence of ICASA. Whereas the amended ICASA Act in 2006 gave many of the powers back to the regulator, one form of control was replaced by another as the ministry was empowered to appoint councillors and determine its budget (Esselaar et al. 2006; Horwitz and Currie 2007). In 2005, the Telecommunications Act was replaced by the Electronic Communications Act No. 36 of 2005 (ECA), which sought to bring the legislation in line with increased convergence and next generation networks built on Internet Protocol (IP). There was a shift from licensing vertically integrated operators to horizontal licensing of network services and pure services separately (Esselaar et al. 2006). The ECA permitted the regulation of facilities leasing but required that all other economic regulation be justified by ICASA through a section 67 market review process. That only permitted regulation of firms with significant market power (SMP) in markets found to be ineffectively competitive (Sibinda 2008; Granville and Irvine 2015). This would put the brakes on much-​ needed wholesale regulation in both the fixed-​line and mobile markets. Local loop unbundling at Telkom was the subject of a policy decision in 2007 but even this was eventually shelved after being opposed by the Portfolio Committee of Communications in Parliament and a new minister which felt it would harm Telkom (Hawthorne 2015). Similarly, self-​provisioning by VANS was initially envisaged but the policy directives were withdrawn at the eleventh hour (Gillwald 2007). However, the mobile operators were given the right to self-​provision fixed infrastructure due to being awarded a technology-​neutral electronic communications network services (ECNS) licence. The only upside to a lack of wholesale regulation of fixed line by ICASA was that it incentivized infrastructure roll-​out by other operators once they secured the legal right to do so. In international connectivity, the SNO would operate a landing station for the new SAFE cable on which the SNO shareholder, VSNL, held a capacity share. Soon thereafter came Seacom (2009), EaSSY (2010), and WACS (2012) with the MNOs taking capacity share initially, given their ECNS licences. The MNOs also invested in national transmission networks along with the SNO, leveraging off their customer traffic to secure demand for the new networks. Dark Fibre Africa (DFA) saw an opportunity to service the MNOs jointly for metro fibre, developing a dark fibre network which was

558    James Hodge and Tamara Paremoer ‘lit’ by MNOs under their ECNS licences even if DFA did not have one. It is only many years after the passing of the ECA that ECNS licences were awarded to new entrants (Abrahams 2011). In fixed line this has mostly resulted in entry into fibre to the home/​ business (FTTH/​ FTTB) networks, with national transmission still dominated by Telkom, Liquid Telecom (previously Neotel), Broadband Infraco (formerly Infraco), and the mobile operators.

26.2.2 Mobile Market The 1996 Telecommunications Act envisaged the licensing of two more MNOs after five years. ICASA ultimately issued only one licence to Cell C which began operating in late 2001 (Cohen 2003). As a new network starting off with no coverage, and up against rivals with national coverage and a sizeable subscriber base, Cell C would have to rely on national roaming and facilities leasing for infrastructure roll-​out in its formative years to compete (Atiyas et al. 2017). In addition, interconnection rates would largely shape their effective retail rates as their own subscribers would be making off-​net calls primarily. The failure to regulate wholesale access to level the playing field for new entrants was an error that was to be repeated in the mobile sector with severe consequences for competition. In 2020, Vodacom and MTN collectively still accounted for approximately 80 per cent of mobile revenue (Competition Commission 2019). The initial mobile interconnection rates of R0.20 (peak) and R0.10 (off-​peak) were increased six f​old in 1999 by the two incumbents, with built-​in CPI or R0.02 annual escalations. This precipitated the 1999 ICASA interconnection guidelines being approved by the minister and the imminent licensing of a third operator. As the economic literature on interconnection identifies, high interconnection rates are favoured by established incumbents because it facilitates high profits without colluding and it acts as a barrier to new entrants who face a greater proportion of off-​net calls which attract the higher price. This enables the incumbents to discount on-​net calls to create club effects to attract subscribers (Hodge 2004). The inbuilt escalations also obviated the need for new agreements in the future, which ICASA would only be empowered to oversee. Whilst the ICASA regulations stipulated a Long Run Incremental Cost (LRIC) approach for major operators, it lacked the COA to undertake a LRIC assessment and would need to declare the incumbents as major operators first. As a result, the new interconnection regulations found no application under the 1996 Act. The facilities leasing regulations (ICASA 2000) of ICASA suffered from the same enforcement issues. As a result, Cell C was severely hampered by excessively high interconnection rates as well as high roaming and facilities leasing fees. Not only were the roaming charges high, but the lack of seamless handover undermined network quality. No rate review of the mobile licence price caps was undertaken in the mistaken belief that a duopoly would deliver competitive rather than cooperative pricing (Hodge 2004). It is only after the ECA was signed into law in 2006 that wholesale regulation was back on the agenda. A wholesale interconnection review was undertaken in 2007 but

Regulation of Network Industries in South Africa    559 legal challenges over the interpretation and requirements of the new section 67 market reviews delayed implementation for years (ICASA 2007). It was revived when Telkom decided to sell its share in Vodacom in 2009 and enter the mobile market in 2010. Parliamentary pressure resulted in a deal with the MNOs to reduce interconnection rates from R1.25 to 40c (peak) over a three-​year period. Following the initial glide path, ICASA engaged in a further regulatory process to reduce interconnection fees based on actual costing models. This saw a further immediate reduction to 20c (peak) and a glide path thereafter. This reduction in interconnection fees saw prepaid call prices reduce by over 40 per cent (Hawthorne 2018). ICASA moved to regulate access to facilities under the 2009 regulations, promoting infrastructure sharing for scarce metro and high sites. However, complaints have persisted from the newer networks of frustrating access and high prices where it is granted. The large asymmetry in site numbers between operators reduces the bargaining power of the smaller networks and reduces the incentive of Vodacom and MTN to provide widespread access on equitable terms (Competition Commission 2019). Better deals are done between the two leading MNOs as they bring similar site volumes to the table. Roaming agreements suffer from the same asymmetrical bargaining power issues as facilities leasing but are not regulated. Only MTN and Vodacom have national coverage which makes the smaller networks dependent upon them to offer a national service (Tzarevski 2019). One of the biggest regulatory and policy failures has been in respect of spectrum assignment. The International Telecommunications Union (ITU) seeks to get agreement on spectrum assignment to different uses. In the early 2000s the ITU identified a potential digital dividend for mobile and broadband services by shifting television from analogue to digital (ITU 2012). The so-​called digital migration was to take place by 2015 globally. This high demand spectrum (HDS) was valuable as it lay in the sub-​1GHz range, which provided broader coverage per cell and better building penetration. Despite starting the policy process in 2007 with an objective of meeting the ITU deadline, South Africa only looks set to license HDS to mobile operators in 2021. Delays have largely resulted from litigation amongst broadcasters and policy flip-​flops on the inclusion or not of encryption on the subsidized set-​top boxes. However, the policy approach to the licensing of spectrum has also been equally litigated, with ICASA first releasing an invitation to apply (ITA) for HDS in 2010, which was withdrawn while the minister made policy directives. This was followed by another ITA in 2011, also withdrawn, and a further one in 2015, also withdrawn to await policy directives. The latest ITA was released in October 2020 (ICASA 2020). The failure to release HDS has created some opportunities for fixed-​line operators with bandwidth in the 1800MHz, 2300MHz, and 3500MHz bands to offer fixed-​ wireless broadband. Telkom’s own mobile service has suffered from the lack of sub-​ 1GHz spectrum, given its late entry, but the fixed-​line business has launched Long Term Evolution (LTE) wireless networks in urban areas as a fibre substitute, using 60MHz in the 2300MHz band. This has allowed it to contest subscription-​based data-​rich mobile packages (Competition Commission 2019). Others with 1800MHz and 3500MHz

560    James Hodge and Tamara Paremoer spectrum included Wireless Business Solutions (WBS) and Liquid Telecom (formerly Neotel, the SNO). Desperate for spectrum, Vodacom made a failed bid for Neotel in 2014 to access its spectrum after the deal was contested before the Competition Tribunal and ICASA. Vodacom was then forced to strike a deal with RAIN, a consortium that acquired WBS, whereby RAIN would be able to have deep passive sharing on Vodacom facilities in exchange for Vodacom getting a roaming agreement on this new RAIN network. A similar deal has subsequently been done with Liquid Telecom. These deals have enabled RAIN and Liquid Telecom to launch their own mobile and fixed-​wireless services to the market with the remaining capacity not contracted, expanding the number of players. However, they remain small and metro focused. MTN has also struck a deal with the financially vulnerable Cell C whereby Cell C will convert its roaming agreement to a similar deep passive sharing of the MTN network in exchange for capacity from its spectrum. These deals provide the basis for new entry but they also cement the leadership of Vodacom and MTN because they are the only networks able to offer widespread site access for the spectrum holders, leaving them with an effective spectrum advantage. In 2017, the Competition Commission launched a Data Services Market Inquiry (DSMI) into high data prices. This was followed shortly thereafter by a section 67 market review by ICASA into mobile data prices (ICASA 2019). The Commission released its final report in December 2019, finding that the mobile market was uncompetitive and that data prices of Vodacom/​MTN were excessive and anti-​poor in structure, given the vast differences in the price per MB for smaller prepaid bundles compared to postpaid bundles (Competition Commission 2019). Vodacom and MTN elected to enter settlement agreements with the Commission, agreeing to reduce monthly prepaid tariffs by 35–​50 per cent and to offer extensive zero-​rating of educational and government websites along with some daily free data (Competition Tribunal 2020a and 2020b). The DSMI report also recommended wholesale regulation of roaming and facilities leasing, as well as legislative changes to enhance regulatory oversight, including wholesale revision of section 67. It also recommended the promotion of fibre broadband and free public Wi-​ Fi to lower income areas to address inequality of address and create some constraints on mobile pricing (Competition Commission 2019). These are being pursued through the ICASA market review and proposed amendments to the ECA by the new Department of Communications and Digital Technology (DCDT).

26.2.3 Telecommunications Conclusion A common perception amongst industry analysts is that ICASA has failed to regulate the telecommunications market effectively, permitting a situation of continued mobile dominance and even areas of fixed-​line dominance twenty-​five years after the liberalization process began. However, the regulator has also been hindered by a lack of independence and resources by political design, as the state policy of ‘managed liberalization’ was itself derailed by a shift to supporting state-​owned enterprises to deliver on

Regulation of Network Industries in South Africa    561 infrastructure mandates. Ironically, it has primarily been the private investment in mobile infrastructure that has succeeded in delivering on the universal service mandate but a lack of mobile regulation has resulted in high prices and inequitable access. The better resourced and independent Competition Commission has stepped in strategically in both fixed line and mobile to provide impetus to sector reform through enforcement and settlements. The 2020 to 2025 period will be critical as the mobile market moves towards 5G services and fixed-​line firms establish widespread FTTH and backhaul fibre to support both FTTH and 5G networks.

26.3 Electricity Globally, the 1990s saw increasing debates regarding the restructuring of the electricity industry. Until then, regulatory interventions commonly focused on price regulation, specifically preventing exploitative prices, preventing ‘cream-​skimming’ in service provision, and ensuring that fees are levied in such a way that otherwise uneconomical services would continue to be provided. The first focus was the transmission grid, where there was general agreement that access to the electricity grid should be open and non-​discriminatory. The two primary approaches to achieving this were vertical separation of the transmission grid from production/​distribution with independent grid managers regulating access to the grid or retaining vertically integrated energy utilities but introducing neutral transmission operators (OECD 20019). New Zealand and the Netherlands followed the first route while countries like France and Germany followed the second. A  second regulatory question that emerged was the appropriate compensatory mechanisms for utilities with universal service obligations. The general approach adopted was the establishment of competitively neutral funds that would compensate companies with universal service obligations (Tirole 2017). In South Africa, the primary challenge facing policymakers in the early 1990s was the vastly unequal access to electricity due to the racist policies of the apartheid state. To contextualize the challenges and the interests at play, it is useful to provide a brief background to the evolution of South Africa’s electricity landscape and regulatory framework. This discussion will also contextualize the politically powerful position that the energy utility, Eskom, had to influence and challenge policy.

26.3.1 Eskom and the History of Energy Regulation in South Africa Much of South Africa’s early economic landscape was shaped by the discovery of diamonds in Kimberley in 1867 and gold in the Witwatersrand in 1886, which led to

562    James Hodge and Tamara Paremoer the agglomeration of services in certain nodes. The unequal provision was worsened by the racialized allocation of state resources under formal apartheid. The same is true for the electricity sector, with Kimberley being the first African city to install electric streetlights in 1882. In about 1891, an electricity reticulation system was established in Johannesburg and private provision of electricity continued to grow to meet the demands of the mines. Prior to the establishment of Eskom, South Africa’s state-​owned electricity utility, electricity provision was fragmented and decentralized, with generation and distribution facilities operated by municipalities and private companies (Ramokgopa and Pietersen 2007). The passage of the Electricity Act of 1922 paved the way for a regulated electricity industry and the emergence of a state-​owned utility, Eskom (or Escom at the time), tasked with the provision of cheap and abundant electricity, particularly to the mining and heavy industrial sectors (Ramokgopa and Pietersen 2007). A process of incorporating regional transmission networks into a national grid ensued, with Eskom assuming control over generation and transmission and local authorities retaining some influence over distribution, with the exception that Eskom directly supplied mines and heavy industry. Eskom only extended its reach in distribution later under the electrification drive in the early 1990s, as the country started its transition to democracy and extended basic services to black communities in rural and urban areas. In the early 1990s, Eskom embarked on a massive electrification drive connecting 1.75 million, mostly black, households by 1999 and 7.8 million households by 2004. Most of these households were in poorer rural and semi-​formal urban township areas, thus expanding Eskom’s distribution activities beyond heavy industrial users to domestic and light industrial use. The utility emerged as a truly vertically integrated monopoly in the 1990s with a notable presence from generation, through transmission to distribution and occupying an important and influential political space as distributor to heavy users and newly connected black households. The prominent role of Eskom has shaped the way it has positioned itself in response to moves to open the sector to greater competition. At present, Eskom is responsible for about 90 per cent of the electricity generated in South Africa. It faces limited competition in generation from independent power producers (IPPs) and some municipalities,6 who collectively generate less than 10 per cent of South Africa’s electricity requirements. Eskom also is the sole transmission licensee (DME 2008)  and accounts for about 60 per cent of distribution; the rest is distributed by municipalities (Baker and Phillips 2019). Eskom operates thirty power stations with a nominal generating capacity of 45,561 MW. The power utility is also building new power stations and high-​voltage power lines to meet South Africa’s growing energy demand. The capacity expansion programme is expected to be completed in 2022 (GCIS 2019).

6 

Municipalities produced 26 per cent in the 1950s; see Marquard (2006).

Regulation of Network Industries in South Africa    563

26.3.2 The Regulatory Landscape Prior to the transition to democracy, Eskom assumed almost sole responsibility for power planning in consultation with large power users (Eberhard, Kolker, and Leigland 2014 and Marquard 2006). Post-​apartheid, the responsibility for electricity regulation is split between two ministries and an independent sector regulator. The Department of Mineral Resources and Energy is responsible for energy policy while the Department of Public Enterprises is the principal shareholder of the electricity utility. Power-​planning resides with the minister of energy who is tasked with producing an energy plan (known as the Integrated Resource Plan or IRP) and issues determinations on the size and sources of new energy. The National Energy Regulator of South Africa (NERSA) is responsible for economic regulation and issues licences in line with ministerial determinations. The National Energy Regulator (NER), which preceded NERSA, was established in 1995 and was envisaged to be a strong, independent regulator that would act as a broker between the powerful utility, users, and government. It was initially headed by a former Eskom CEO and though it started with a strong dual focus on both policy and economic regulation focus, the policy focus shifted to government as it developed capacity in power-​planning and electricity-​system-​structuring and NERSA’s influence in this regard waned (Eberhard 2004 and Marquard 2006). The primary tool available to NERSA to discipline the monopoly utility is thus the determination of tariff methodologies and pricing frameworks to evaluate tariff applications from licensees. NERSA’s approach to tariff determination is reflective of the competing priorities generally facing economic regulators as well as the particularly wide disparity in income within South Africa. Its Electricity Pricing Policy tries to strike a balance between the objectives of social equity, efficiency, and viability of the utility by ensuring that it receives an appropriate rate of return to allow it to access finance at reasonable rates and encourage necessary investment. The equity objective is reflected in, amongst others, differential tariffs and the principle of cross-​subsidization with lower tariffs for low-​ income consumers and cost-​reflective tariffs for others (DME 2008). An assessment of average electricity prices since 2003 (see Figure 26.1) shows that average prices have been trending steadily upwards though the rate of increase has declined from the steep increases in 2007/​08, associated with a period of significant constraints in electricity supply as well as the start of Eskom’s new build programme. The chart also points to a differential pricing system, with commercial users paying the lowest tariffs and residential and agricultural customers facing the highest absolute prices. The rate of price increases amongst various users has converged to around 5 per cent per annum by 2018/​19. Over the same period, Eskom’s generation performance and its ability to meet the country’s energy demand has declined from a peak of supplying 97 per cent of the country’s energy production in 2009 to 90 per cent in 2020 (Statistics South Africa 2020).

564    James Hodge and Tamara Paremoer

Figure 26.1  Average electricity prices (c/​kWH) Source: Eskom https://​www.eskom.co.za.

In 2007–​08, when South Africa first started experiencing regular rolling blackouts (or ‘load-​shedding’ in local parlance), the constrained energy system was explained by a combination of policy failure, inadequate systems planning, insufficient investment, and ageing power stations. These challenges partly arose from the fact that the country was still grappling with the optimal design of its electricity system to increase competitiveness and attract private-​sector investment. During this time, Eskom was prevented from adding generation capacity (Eberhard 2008). The complexity of the debates that ensued (as well as lobbying by the incumbent utility) contributed to indecision on new investment. The result was consistent high utilization of an ageing generation fleet with reduced reserve margins, large maintenance backlogs and increasingly frequent breakdowns. These challenges in generation, difficulty in containing cost escalations, and the increasing inefficiency of Eskom raises questions about the effectiveness of regulation to promote efficiency. However, a closer look at Eskom’s most recent revenue application (the multi-​year price determination for 2019/​20–​2021/​22 or MYPD4) process shows the challenges facing the regulator in driving the objectives of equity and efficiency. In the contested MYPD4 tariff application for the three-​year period from 2019/​20, NERSA conducted a critical review of the utility’s sales forecast, noting that Eskom’s forecast did not sufficiently consider users’ price elasticity of demand and the historical declining trends of sales volume in light of increasing prices and decreasing reliability of energy availability. NERSA also cautioned against building inefficiency into the tariff determination process and declined to approve some of the most costly and inefficient generation options included by Eskom in its application. The NERSA decision

Regulation of Network Industries in South Africa    565 highlights that Eskom has failed to arrest declining performance of its plants and that its energy availability factor (EAF) had decreased from 78 per cent to around 65.9 per cent. NERSA insisted that an EAF of at least 71.5 per cent be used in the tariff application, increasing to 73.5 per cent in 2022, to encourage the utility to improve the efficiency of operations and to avoid building costly generation options into its production plan (NERSA 2019b). Further, in calculating Eskom’s allowable revenue, NERSA removed about Rand 69 billion in state support over the tariff period, arguing that the tariff determination process already allows an efficient utility to cover its debt-​service obligations. Setting off the cash injection against the utility’s allowable return on assets effectively constrained tariff increases in the first year to 9.41 per cent (against the 22.59 per cent applied for). Eskom successfully challenged the decision in the courts, though the decision is being appealed by the regulator (Eskom 2019). In the context of these challenges, and the continued poor performance of Eskom, the effectiveness of the fragmented regulatory approach and reliance on price regulation to discipline the energy utility have not delivered the desired results. In the following section we review the introduction of competition in generation where reform has been slow. Eskom has been able to position itself to resist fundamental reform and has, in relation to the introduction of independent power producers (IPPs), been able to act contrary to state policy.

26.3.3 System Design and Introducing Competition in Generation Since the 1990s, government policy has proposed the unbundling of the vertically integrated Eskom monopoly, introducing competition in generation/​distribution and ensuring independent management of the transmission grid. The policy position was set out in the democratic government’s first White Paper on energy policy released in 1998 (discussed below).

26.3.3.1 Distribution In 1998, distribution was highly fragmented with more than four hundred distributors, many of which were not financially viable. The White Paper proposed the consolidation of some of these distributors into independent, financially viable regional energy distributors that would benefit from economies of scale and skills.

26.3.3.2 Generation Though the White Paper is clear that electricity generation would be opened to competition from IPPs, it is cautious about the pace of restructuring of the energy sector. The White Paper is clear that the drive to electrify underserved communities would be the foremost government priority and that decisions on restructuring would be delayed until the bulk of the electrification programme was completed.

566    James Hodge and Tamara Paremoer

26.3.3.3 Transmission The White Paper proposes that Eskom be split into separate generation and transmission companies and that the transmission grid would be operated on open and non-​ discriminatory terms with transparent tariffs and full disclosure of cost and pricing information to the sector regulator. The principles of open and fair access were already incorporated into Eskom’s transmission licence at the time, in preparation of liberalization of the energy sector. The policy positions articulated in the 1998 White Paper were thus in line with leading practice and principles about restructuring of energy markets for improved efficiency at the time. Implementing these reforms could potentially have been transformative in preventing the decline in the financial and operational performance of Eskom (Kessides 2020). However, many critical decisions about the restructuring of the energy sector were not implemented. In 2019, twelve years after the publication of the White Paper, the government has again committed to unbundling Eskom (DPE 2019). The current unbundling represents a relatively weak form of vertical separation as Eskom will be unbundled into three subsidiary entities managed by the current Eskom management team and subject to current governance processes (Kessides 2020). The possibility of splitting generation into several subsidiaries to introduce intra-​company competition has also been mooted (DPE 2019). Concerns have been raised about the sufficiency of such incremental structural reforms in reinvigorating performance and improving the efficiency and operational performance of the entity, while noting that there are some positive aspects of creating subsidiary entities as it retains coordination efficiencies and reduced transaction costs.

26.3.4 Renewable Energy In 2009, shortly after the first rolling blackouts exposed a concerning lack of generation capacity in the country, the government started evaluating the feasibility of feed-​ in tariffs for renewable energy. This proposal was eventually replaced by a system of competitive tenders for renewable energy production with Eskom as the sole purchaser (Eberhard et al. 2014). The programme, known as the Renewable Energy Independent Power Producer Procurement Programme (REIPPP) was launched in August 2011. At the time, the IRP estimated that about 17,800 MW of energy (or 9 per cent of energy produced at the time) would be generated from new renewable-​energy sources over a period of twenty years from 2010 to 2030. The first round of projects (REIPPP Round 1) came on board relatively quickly, two years after the approval of the first round of bids. Authors have praised the flexibility, transparency, and technical capacity of the REIPPP programme, which operated as an ad hoc unit within the Department of Energy and drew on expertise from both the public and private sectors (Eberhard et al. 2014). The REIPPP process also exhibited flexibility in its approach to bid design and pricing. The first round of projects was contracted at relatively high prices, but lessons from each

Regulation of Network Industries in South Africa    567 round have been built into successive REIPPP rounds and prices have reduced significantly. Eskom, as the sole transmission licensee, would procure the entire offtake at contracted prices and would face no costs in this regard, as all costs would be recovered by tariffs on users. Arguably, the biggest challenge to the success of the REIPPP programme came in 2016 when Eskom refused to sign outstanding power purchase agreements (PPAs) with Independent Power Producers (IPPs) that won earlier bid windows despite a ministerial determination requiring it to do so. The reasons cited by Eskom at the time were that it had surplus generating capacity and the costs of the agreements were higher than the cost at which Eskom could generate electricity itself. The IPPs turned to NERSA and the Competition Commission for assistance in gaining access to its transmission grid. The impasse was resolved in 2018, when intervention by newly appointed ministers of public enterprises and energy resulted in the PPAs being signed. During this time, renewable energy forums expressed concern at the level of power the utility had to resist legal and political instruction to enter contracts in line with government policy (Odendaal 2016).

26.3.5 Energy Conclusion Despite long-​standing policy positions that advocate for liberalization and competition, the pace of reforms in the electricity sector has been slow. This is at least partly explained by the political importance of the energy utility, the ability of the insider-​incumbent to respond to and lobby in the face of proposed policy changes and to undermine policy changes that threaten its monopoly. Marquard (2006) writes that Eskom responded to the introduction of the concept of unbundling in the White Paper by commissioning paid research to play up the value of a single utility model. Similarly, Steyn (2013) notes that in response to policy direction that contemplated unbundling Eskom into separate entities for generation, transmission, and distribution, Eskom created three divisions of the same names. Later, when the introduction of an independent system operator seemed imminent with the submission of the Independent Systems and Market Operator (private) Bill to parliament in 2012, Eskom created a ‘Systems and Market Operator Division’. It proposed that this unit could perform all the functions envisaged for the independent systems operator. This points to the strategic advantages that important SOEs have to lobby government, particularly when faced with the prospect of greater competition.

26.4 Transportation In line with the shift to privatization and deregulation in the 1980s, the 1986 De Villiers Commission recommended the corporatization of transport services under a single

568    James Hodge and Tamara Paremoer entity. This resulted in the creation of Transnet through the Legal Succession Act of 1989 (Baloyi 2014). The liberalization of aviation (1990) and road freight transport (1988) proceeded immediately based on the recommendations of the 1987 White Paper on privatization and deregulation, whilst ports and rail were left to the new democratic government. The White Paper on transport policy provided a broad vision which was to be supplemented by modal strategic plans.7 That vision was for the Department to focus more on policy and strategy, and to increasingly outsource to professional agencies the oversight of safety/​quality whilst regulating monopolies to prohibit excessive tariffs (White Paper 1996). Economic regulation of a ports authority was envisaged for the future, but most of the immediate focus was on safety and quality regulation (Teljeur et al. 2003). This resulted in the creation of the Civil Aviation Authority, the SA Maritime Safety Authority, the National Railway Safety Regulator, and the South African National Roads Agency (SANRAL). SANRAL would also be responsible for the development of road infrastructure funded by a fuel levy and the determination of tolls on road concessions. The Department of Public Enterprises (DPE) has some oversight as the shareholder, but this is not a substitute for economic regulation as that mandate is primarily to ensure financial viability and alignment of their operation to government policy (Baloyi 2014).

26.4.1 Aviation South African Airways (SAA) was given a legal right under the 1949 International Air Services Act to monopolize the major domestic trunk and international routes, with Comair (1940s) and later BopAir and LinkAir (1970s) servicing smaller feeder routes. The Domestic Air Transport Policy (1990) commercialized SAA and permitted domestic competition. This saw the entry of FliteStar in 1991, which competed head to head with SAA but faced persistent anti-​competitive conduct due to SAA’s control of airports and landing slots, until it exited in 1994. This resulted in the move to separate airports and navigation and placed them under new entities, namely the Airports Company Limited (renamed ACSA) and the Air Transport Navigation Services (ATNS) in 1993 (Ellison 1992). The Airport Companies Act also established the Regulatory Committee, a part-​ time regulator, to set the tariffs for these two natural monopolies (1993). Following the separation, another five airlines entered within two years and low-​cost carriers entered in the early 2000s. As a result of the competition, there was no requirement to regulate SAA tariffs and conduct outside of competition law. SAA did briefly get a EP in the form of Swiss Air, but the shares were repurchased by the government after Swiss Air went bankrupt following the collapse of international travel after 9/​11 terror attacks. The Regulatory Committee makes use of a price cap for both ACSA and ATNS to drive operational efficiencies whilst permitting a fair return on assets. The tariff permissions are

7 

National Transport Policy White Paper 1996.

Regulation of Network Industries in South Africa    569 for a five-​year period and are based on supporting business plans and demand projections. This ensures that future infrastructure investments are anticipated and enabled by tariff adjustments, such as negative productivity factors (Teljeur et  al. 2003). All tariff permissions must be approved by the minister. Aviation regulation is generally considered to have been successful with major airports all upgraded for the 2010 World Cup and operating relatively efficiently. ACSA itself briefly had an SEP in the form of Aeroporti di Roma but the Public Investment Corporation (PIC) bought their share in 2005. ACSA has also successfully bid to operate airports in Mumbai and Sao Paulo, along with the six international and five local South African airports. In contrast, SAA as an airline has been less successful, being subject to mismanagement and corruption that finally placed it in business rescue in 2019 (see Chapter 27 in this volume).

26.4.2 Ports The primary difficulty in reforming ports (and rail) was the massive cross-​subsidization taking place within Transnet itself, whereby port charges supported rail investment. As a result, port charges were high by global standards but this windfall did not translate into greater port investment needed to relieve congestion and promote efficiency (Teljeur et al. 2003). The other regulatory problem was that Transnet not only operated the port infrastructure, but also provided port terminal services in competition with private concessionaires. This created a conflict of interest as Transnet is both referee and player, securing the most profitable container and automobile terminal concessions for itself (Trade and Industrial Policy Strategies 2014). The Department of Transport (DoT) issued the National Commercial Ports Policy document in 2002, which set out the strategy for this mode. It recommended the separation of Portnet into two entities, namely Transnet National Ports Authority (TNPA) and Transnet Port Terminals (TPT), which occurred that same year. It also recommended the establishment of a port authority and a regulator to oversee it in line with the 1996 White Paper. The National Ports Act of 2005 established the ports regulator and the legislative basis for TNPA to transfer its assets to a new entity outside of Transnet, namely the National Ports Authority (NPA). However, the transfer of assets to the NPA is based on a date set by the shareholding minister, and fifteen years later it still had not occurred (NPA Act No. 12 of 2005). This is because of the continued cross-​subsidization that occurs within Transnet and which would be precluded as soon as ports were transferred. Whilst TNPA is operationally separate from TPT, the common shareholding means that complaints of unfair treatment by private concessionaires persist with primarily competition law able to address such conduct (CAC 2017) even though the ports regulator can hear complaints on some aspects of the service. The ports regulator was operational in 2009 and proceeded to regulate tariffs from 2011. The tariff methodology is to set tariff increases to achieve a required revenue (RR) which covers operational costs and a fair return on assets (i.e. a cost-​plus methodology). It is generally accepted that this methodology is ill suited to ports, given that it provides

570    James Hodge and Tamara Paremoer no incentives to improve operational efficiency (Meyiwa and Chasomeris 2016). The RR is also heavily influenced by the regulatory asset base and the permitted return on those assets, with criticisms that the asset base is inflated and investments are recouped over short periods, which artificially inflate tariffs (Gumede and Chasomeris 2015). The ports regulator has persistently not granted the increases requested by TNPA and has also sought to make targeted tariff adjustments to promote export competitiveness. For instance, export container tariffs were reduced by 43 per cent and those on vehicle exports by 21 per cent. There is also a tariff discount provided, based on the extent of local beneficiation (das Nair and Roberts 2017). However, the challenge is to gradually bring tariffs down from the high levels found by the ports regulator in its April 2012 global comparison, whilst promoting investment to improve operational efficiency.

26.4.3 Rail Rail had been used for employment creation during the apartheid era and had suffered from a lack of investment despite being protected from road freight. However, the dissatisfaction with the service and inability to invest due to fiscal constraints resulted in the liberalization of road freight transport in the late 1980s (Havenga et al. 2014). This has resulted in a massive shift of general freight from rail onto road, largely blamed on the fact that road usage charges do not reflect the costs of road development and those charges are also primarily funded by non-​freight users. This has resulted in the poor performance of the general freight business (GFB) of Transnet Rail and the need for even greater levels of cross-​subsidization from both TNPA and from the more profitable dedicated ore lines (Truen 2008). The ore lines remain profitable due to the inability to switch to road, and include the Coalink lines from the Mpumalanga coal fields to Richards Bay and the Iron Ore lines from Sishen to Saldhana Bay. Despite generating most of the profit, these subsidize investment in the GFB which undermines their own capacity and efficiency (Havenga et al. 2014). The National Freight Logistics Strategy 2005 sought to set out the detailed modal strategy for rail. The policy continued to promote the Anglo-​Saxon reform approach of separating rail infrastructure from rail services, allowing competition for services at commercial rates and ensuring the rail infrastructure was profitable. However, this was not well accepted at a time when policy thinking had shifted to the Asian developmental infrastructure approach, which sought to achieve social objectives from infrastructure SOEs rather than just ensuring their profitability. The same policy perspective made it into the Green Paper on Rail Policy in 2011, but this was subsequently withdrawn and replaced in 2013 with one that espoused an investment-​led reform rather than institutional reform (Havenga et al. 2014). Around the same time, the DoT started a review and update to the 1996 White Paper, which recommended the creation of a single transport economic regulator that would then cover rail infrastructure too. In late 2019, the Economic Regulation of Transport Bill was tabled before Parliament, which seeks to put in place the proposed single regulator. The Bill cites as its purpose

Regulation of Network Industries in South Africa    571 the consolidation of economic regulation under a single framework through the establishment of a Transport Economic Regulator (TER) with appeals to be heard by a Transport Economic Council (TEC). As such, the Bill makes provision for the transfer of existing regulatory functions to this new entity. The Objects of the Bill expressly recognizes that logistics costs are unacceptably high, conditions do not exist for efficient infrastructure services, and the modes are dominated by SOEs with market power. It also expressly recognizes that regulation is absent for rail despite oversight in other areas. There is a shift back towards the potential of private service concessions on the rail infrastructure which seems to have been driven by fiscal constraints for investment following the Zuma era. The Bill requires that the minister identify sectors that require regulation based on the existence of a firm with market power or an essential facility. The Bill continues to state that once the minister determines that this applies to rail, then the TER must determine the costs of access, including the use of infrastructure to run trains, but extending to interconnection and even investment in infrastructure if it is not being provided.

26.4.4 Transportation Conclusion Economic regulation has been successful in aviation in part due to consensus around liberalization, well-​established global regulatory models, and a focus purely on tariffs for the essential facilities rather than mediating competition. Regulation in ports and rail has been held up by the complex web of cross-​subsidization and the shift in policy towards a more developmental agenda for infrastructure SOEs. However, the lack of economic growth, fiscal constraints, and huge debt from transport SOEs coming out of the Zuma era has forced government thinking back to a reformist agenda as is evident in the most recent ERT Bill. If enacted in its current form, it will permit the regulator to impose much more discipline on SOEs, given that the mandate includes the requirement for regulated entities to provide development plans, the setting of service standards, and for oversight of complaints related to a wide range of issues including favouring their subsidiaries, impeding intermodal connection, and failure to provide services on a FRAND basis.

26.5 Conclusion State conflicts of interest and the lack of a clear commitment to liberalization has directly impacted on the success of network industry regulation. The initial momentum towards liberalization in the 1990s, driven by global policy thinking, stalled in the early 2000s as policy thinking shifted to using SOEs to deliver on development objectives. In essence, this reflects a lack of faith that effective regulation of private enterprises can deliver on important socio-​economic development mandates.

572    James Hodge and Tamara Paremoer Where liberalization through private enterprise entry had already occurred, or had been committed to in the 1990s (aviation and telecommunications), regulatory oversight has proceeded with pockets of success (airports, interconnection) amongst some weaknesses in addressing wholesale regulation of dominant companies (mobile). However, where liberalization reforms stalled, either regulation was abandoned entirely (rail) or the regulators were left to simply preside over the tariffs of SOE monopolies (electricity, ports). This outcome is not conducive to effective regulation due to the conflicts of interest of the state, a fact exploited by SOEs themselves through direct lobbying. After all, regulation is primarily designed to oversee private enterprises in the context of liberalization where such conflicts do not exist. Until there is a commitment to liberalization as a policy, economic regulation in these sectors will not be effective in driving operational efficiency and lowering tariffs. The SOE crisis that has emerged from the ten years to 2017 (see ­Chapter 27 in this volume) may provide some impetus for such a commitment.

References Abrahams, Lucienne. 2011. ‘South Africa’s e-​development still a futuristic task’, Journal of the Knowledge Economy, 2: 145–​7 1. Atiyas, Izak, Brian Levy, and Michael Walton. 2017. ‘Rent creation and rent containment: The political economy of telecommunications in Mexico, South Africa’. [online] https://​ openknowledge.worldbank.org/​bitstream/​handle/​10986/​26202/​112874-​WP-​PUBLIC-​ WDR17-​BP-​Political-​economy-​of-​telecoms.pdf?sequence=1&isAllowed=y. Baker, Lucy, and Jon Phillips. 2019. ‘Tensions in the transition: The politics of electricity distribution in South Africa’, Environment and Planning C: Politics and Space, 37(1): 177–​96. Baloyi, Basani. 2014. ‘The role of South Africa’s freight rail regulatory framework in General Freight’s sluggish growth performance’, Johannesburg [online] https://​www.competition. org.za/​working-​papers. Cohen, Tracy. 2003. ‘Rethinking (reluctant) capture: South African telecommunications and the impact of regulation’, Journal of African Law, 47(1): 65–​87. Competition Appeal Court. 2017. Siyakhuphuka Investment Holdings vs Transnet SOC and Transnet Ports Authority. Case no. 158/​CAC/​Nov17. Pretoria. Competition Commission South Africa. 2019. ‘Data services market inquiry final report’. Pretoria [online] http://​www.compcom.co.za/​wp-​content/​uploads/​2019/​12/​DSMI-​Non-​ Confidential-​Report-​002.pdf. Competition Tribunal South Africa. 2012. Competition Commission vs Telkom SA Ltd. Case no. 11/​CR/​Feb04. Pretoria. Competition Tribunal South Africa. 2013. Competition Commission vs Telkom SOC. Case no. 016865. Pretoria. Competition Tribunal South Africa. 2020a. Consent Agreement, Competition Commission and Vodacom (Pty) Ltd. Case no. CO166Mar20. Pretoria. Competition Tribunal South Africa. 2020b. Consent Agreement, Competition Commission and Mobile Telephone Networks (MTN) (Pty) Ltd. Case no. CO006Apr20. Pretoria. das Nair, Reena, and Simon Roberts. 2017. ‘Competition and regulation interface in energy, telecommunications and transport in South Africa’, in Jonathan Klaaren, Simon Roberts,

Regulation of Network Industries in South Africa    573 and Imraan Valodia (eds) Competition Law and Economic Regulation: Addressing Market Power in Southern Africa. Johannesburg: Wits University Press. Department of Minerals and Energy. 2008. ‘Electricity pricing policy (EPP) of the South African electricity supply industry’. Pretoria. Department of Public Enterprises. 2019. ‘Roadmap for Eskom in a reformed electricity supply industry’. Pretoria. Eberhard, Anton. 2004. ‘The political economy of power sector reform in South Africa’ [online]. https://​www.gsb.uct.ac.za/​files/​StanfordPSREberhardSep2004final.pdf. Eberhard, Anton. 2008. ‘Moving beyond blame: What is required to shore up Eskom’. [online] https://​www.gsb.uct.ac.za/​files/​MovingBeyondBlame.pdf. Eberhard, Anton, Joel Kolker, and James Leigland. 2014. ‘South Africa’s renewable energy IPP procurement program: Success factors and lessons’. [online] https://​openknowledge. worldbank.org/​handle/​10986/​20039?locale-​attribute=es. Economides, Nicholas. 2004. ‘Competition policy in network industries:  An introduction’. NET Institute Working Paper no. 04/​24, New York. Ellison, Anthony P. 1992. ‘South Africa’s transport policies’, Journal of Transport Economics and Policy, 26(3): 313–​18. Eskom. 2019. ‘NERSA’s reasons for decision on Eskom’s revenue application raise more questions than answers’. [online] https://​www.eskom.co.za/​news/​Pages/​2019Oct11.aspx. Esselaar, Steve, Alison Gillwald, and Christoph Stork. 2006. ‘South African Telecommunications Sector Performance Review 2006’. [online] https://​www.researchgate. net/​publication/​228683809_​S outh_​African_​Telecommunications_​S ectorPerformance_​ Review_​2006. Gillwald, Alison. 2007. ‘Between two stools: Broadband policy in South Africa’, The Southern African Journal of Information and Communication, 8: 53–​77. Government Communication and Information System. 2019. ‘South Africa Yearbook 2018/​19 Energy’, Pretoria [online] https://​www.gcis.gov.za/​sites/​default/​files/​docs/​resourcecentre/​ yearbook/​yb1919-​8-​Energy.pdf. Granville, Lara, and Heather Irvine. 2015. ‘The impact of regulation on competition in telecommunications and piped gas’, The African Journal of Information and Communication, 14: 38–​49. Groenewald, Yolandi. 2017. ‘Kubayi: Eskom must sign IPP agreements by end of October’. [online] https://​www.news24.com/​fin24/​Economy/​Eskom/​breaking-​eskom-​must-​sign-​ipp​agreements-​by-​end-​of-​october-​20170901. Gumede, Nyawo, Kwame Asmah-​Andoh, and Humayun Kabir. 2016. ‘The restructuring and privatisation trajectory in South Africa: Tracing the historical public policy origins, purpose, methods and guidelines in the pre-​1994 epoch’, Journal of Governance and Regulation, 5(1): 31–​6. Gumede, Sanele, and Mihalis Chasomeris. 2015. ‘South Africa’s port pricing methodology and financing investments’. [online] http://​2015.essa.org.za/​. Havenga, Jan H., Zane P. Simpson, and Anneke De Bod. 2014. ‘South Africa’s freight rail reform: A demand-​driven perspective’, Journal of Transport and Supply Chain Management, 8(1): 1–​7. Hawthorne, Ryan. 2015. ‘Economic regulation of the telecommunications sector in South Africa 2009–​2014’, The African Journal of Information and Communication, 14: 20–​37. Hawthorne, Ryan. 2018. ‘The effects of lower mobile termination rates in South Africa’, Telecommunications Policy, 42(5): 374–​85.

574    James Hodge and Tamara Paremoer Hawthorne, Ryan, Pamela Mondliwa, Tamara Paremoer, and Genna Robb. 2016. ‘Competition, barriers to entry and inclusive growth: Telecommunications sector study’. [online] https://​ www.competition.org.za/​working-​papers. Hodge, James. 1999. ‘The state of the telecommunications industry in South Africa and the potential costs/​benefits of liberalisation’. [online] http://​www.tips.org.za/​files/​284.pdf. Hodge, James. 2004. ‘Liberalising communication services in South Africa’, Development Southern Africa, 17(3): 373–​87. Horwitz, Robert B. 1997. ‘Telecommunications policy in the new South Africa: Participatory politics and sectoral reform’, Communicatio, 23(2): 63–​78. Horwitz, Robert B. and Willie Currie. 2007. ‘Another instance where privatization trumped liberalization: The politics of telecommunications reform in South Africa—​a ten-​year retrospective’, Telecommunications Policy, 31: 445–​62. ICASA. 2000. ‘Facilities leasing guidelines 2000’. Gauteng. ICASA. 2005. ‘ICASA annual report 2005’. Gauteng. ICASA. 2007. ‘ICASA general notice: Interconnection regulations’. Gauteng. ICASA. 2010. ‘Electronic communications facility leasing regulations 2010’. Gauteng. ICASA. 2019. ‘Discussion document on the market inquiry into mobile broadband services in South Africa’. Gauteng. ICASA. 2020. ‘ITA for the radio frequency spectrum licences for IMT spectrum bands’. Gauteng. ITU. 2012. ‘Digital dividend insights for spectrum decisions’. [online http://​www.itu.int/​ITU-​ D/​tech/​digital_​broadcasting/​Reports/​DigitalDividend.pdf. Kessides, Ioannis N. 2020. ‘The decline and fall of Eskom: A South African tragedy’. [online] https://​www.thegwpf.org/​content/​uploads/​2020/​06/​Decline-​Fall-​Eskom.pdf. Kim, Ran S. and Albrecht Horn. 1999. ‘Regulation policies concerning natural monopolies in developing and transition economies’. DESA Discussion Paper no. 8, United Nations. https://​www.un.org/​esa/​esa99dp8.pdf. Lewis, Charley. 2013. ‘Universal access and service interventions in South Africa: Best practice, poor impact’, The African Journal of Information and Communication, 13: 95–​107. Marquad, Andrew. 2006. ‘The origins and development of South African energy policy’. PhD thesis, University of Cape Town. http://​www.erc.uct.ac.za/​sites/​default/​files/​image_​tool/​ images/​119/​Papers-​2006/​06Marquard-​PhD_​Thesis.pdf. Meyiwa, Ayanda, and Mihalis Chasomeris. 2016. ‘Restructuring port governance in South Africa’, Journal of Economic and Financial Sciences, 9(3): 854–​73. Mhlanga, Oswald, and J. N. Steyn. 2016. ‘The aviation industry in South Africa: A historical overview’, African Journal of Hospitality, Tourism and Leisure, 5(4): 1–​13. NERSA. 2019a. ‘NERSA annual report 2018/​2019’. Pretoria. NERSA. 2019b. ‘NERSA decision: Eskom’s MYPD4 revenue application for the 2019/​20–​2021/​ 22 financial year’. Pretoria. Odendaal, Natasha. 2016. ‘Sarec too raises IPP delay concerns’, Engineering News [online] https://​sawea.org.za/​wp-​content/​uploads/​2016/​11/​SAREC_​engineeringnews.co_​.za_​9-​ Nov-​2016.pdf. Organisation for Economic Co-​Operation and Development (OECD). 2019. ‘Restructuring Public Utilities for Competition’, OECD [online]. https://​www.oecd.org/​competition/​ sectors/​19635977.pdf Parliamentary Monitoring Group. 2000. ‘Telkom’s monopoly’. [online] https://​pmg.org.za/​ committee-​meeting/​9951/​.

Regulation of Network Industries in South Africa    575 Ramokgopa, Bolden, and Edwina Pietersen. 2007. ‘Tariff history 2002–​2007’. Eskom [online] https://​www.eskom.co.za/​CustomerCare/​TariffsAndCharges/​Documents/​Tariff_​History_​ 2007_​_​_​Edited_​Draft.pdf. Rohlfs, Jeffrey. 1974. ‘A theory of interdependent demand for a communications service’, Bell Journal of Economics, 5(1): 16–​37. Sibinda, Gershon. 2008. ‘Regulatory environment analysis in the South African telecommunications industry’, South African Journal of Economics, 76(2): 212–​27. Statistics South Africa. 2020. ‘Electricity generated and available for distribution (Preliminary)’. Statistical Release P4141, 30 September. Steyn, Elsa. 2013. ‘Dawn of a competitive electricity sector for South Africa: The Independent System and Market Operator Bill B 9-​2012—​Context, content and comment’. [online] http://​www.saflii.org.za/​za/​journals/​DEJURE/​2013/​23.pdf. Teljeur, Ethel, Alison Gillwald, G. Steyn, and D. Storer. 2003. ‘Regulatory frameworks:  Impact and efficacy, volume II:  Detailed sectoral reports’. TIPS [online] https:// ​ w ww.tips.org.za/ ​ r esearch- ​ a rchive/ ​ a nnual- ​ f orum- ​ p apers/ ​ 2 003/ ​ i tem/​ 364-​regulatory-​frameworks-​impact-​and-​efficacy-​volume-​ii-​detailed-​sectoral-​reports. Tirole, Jean. 2017. Economics for the Common Good. Princeton, NJ: Princeton University Press. Trade and Industrial Policy Strategies. 2014. ‘Review of regulation in the ports sector’, TIPS [online] http://​www.tips.org.za/​files/​ccred-​edd-​recbp_​regulation_​in_​the_​ports_​sector_​-​_​ farr_​levin.pdf. Truen, Sarah. 2008. ‘Regulation of state-​owned enterprises Transnet freight rail case study’. [online] https://​www.researchgate.net/​publication/​228286424_​Regulation_​of_​State_​ Owned_​Enterprises_​-​_​Transnet_​Freight_​Rail_​Case_​Study. Tzarevski, Angelo. 2019. ‘South Africa:  Cell C and MTN conclude extended roaming agreement’. Baker McKensie [online] https://​www.bakermckenzie.com/​en/​insight/​ publications/​2019/​12/​cell-​c-​and-​mtn-​conclude-​extended-​roaming-​agreement.

Chapter 27

State-​O wned E nt e rpri se s in Sou th   A fri c a Mark Swilling and Nina Callaghan

27.1 Introduction By 2019, twenty-​five of the largest state-owned enterprises (SOEs) accounted for a fifth of South Africa’s capital stock, a seventh of annual total investment in the South African economy, and around 1 per cent of employment (Makgetla 2020). They are, therefore, large-​scale institutions responsible for managing capital-​intensive infrastructures and investments that play a key role in the national economy. Contrary to those who depict South Africa as an ‘outlier’ when it comes to SOEs (usually via a narrow neoclassical lens), internationally, SOEs are not in decline despite a global wave of privatization (Nem Singh and Chen 2018). There are fifty-​eight listed and 1,617 unlisted SOEs in OECD countries that collectively have assets equal to 32 per cent of GDP. Chinese SOEs account for 67 per cent of the market value of all domestic listed enterprises. In Latin America, on average SOE output accounts for 15 per cent of GDP per country. Like other industrializing economies, SOEs have played a crucial role in the evolution of the South African political economy. Following Bowman’s summary of the literature, SOEs have traditionally played a number of roles, including the lowering of the cost of doing business by providing low-​cost infrastructures, investing in high-​risk knowledge and capital-​intensive industries, providing finance (usually via development finance institutions) to priority industries that private banks stay away from, and actively supporting a particular fraction of the business class that are deemed politically and strategically significant by the governing elite (Bowman 2020). SOEs in South Africa have played all four roles during the pre-​and post-​1994 eras. The promise of 1994 was premised on the assumption that the new democratic state would integrate South Africa into a worldwide ‘norm’ that Hart and Padayachee refer to as ‘national capitalism [which] is the modern synthesis of nation-​states and industrial capitalism, the institutional attempt to manage money, markets and

State-Owned Enterprises in South Africa    577 accumulation through central bureaucracy for the benefit of a cultural community of national citizens’ (Hart and Padayachee 2013:  57). As with all colonial-​enclave economies premised on cheap labour and resources, the options are to replicate this model disguised by nationalist rhetoric (the target of the Fanonist critique), successfully build an inclusive ‘national capitalist’ alternative (what, in theory, ‘development’ is supposed to be about), or introduce a post-​capitalist alternative (ranging from the ‘non-​capitalist road’ alternative of the mid-​twentieth century, the Soviet-​style statist options, through to some of the current left-​wing social–democratic or even ‘developmental state’ alternatives). South Africa opted for the second alternative, albeit a more conservative version than what was envisaged in the Macroeconomic Research Group (MERG) report that was rejected by the ANC leadership in 1993 (Padayachee and van Niekerk 2019). As Hart and Padayachee argue, South Africa has attempted to build ‘national capitalism’ on two occasions:  the first was between the World Wars, but culminated in apartheid after the Second World War where racism prevented the full realization of the ‘national capitalist’ model; and the second was after 1994 when the promise of a prosperous inclusive democracy was thwarted by state interventions that did little to dismantle income and asset inequality in order to fully enfranchise the black African majority. Following Feinstein (2005), Hart and Padayachee argue that the ‘national capitalism’ South Africa aspired to build after 1994 could only have been achieved if ‘the home market is stimulated by equalising incomes across classes; a national system of education ensures the development of a skilled labour force in support of industrialisation; citizenship is extended to the workplace (unions, bargaining, etc.); and the government cares for the health, welfare and housing of all the people’ (quoted in Hart and Padayachee 2013: 60). Needless to say, this combination of developmental outcomes was not achieved. Given the focus of this chapter, what is significant is that SOEs were central to the first attempt at building a (racially exclusive) ‘national capitalism’. They were not, however, a central focus of the second attempt during the immediate post-​1994 era. They did, however, become central during the state-​capture years after 2009 because it was their procurement spend that was used ostensibly to build a ‘black industrial class’. In reality, the SOEs were, of course, looted by a Zuma-​centred power elite that came to power via a silent coup (Bhorat et al. 2017). It will be argued that the post-​COVID-​19 era is South Africa’s third opportunity to build a fully inclusive form of ‘national capitalism’ driven by a post-​minerals-​energy-​ complex (MEC) industrialization programme. However, this is bound to fail if an appropriate strategy for transforming and re-​deploying the SOEs is not put in place. This will mean going beyond the current obsession with ‘fixing the SOEs’. A much greater ambition and clearer vision is required. This chapter contributes to this discussion by engaging with the National Planning Commission’s (NPC) recommendations for restructuring the SOE sector. However, to lay the foundation for this, it is necessary to discuss the governance of SOEs during three eras: pre-​1994, post-​1994, and the state-​ capture years between 2009 and 2018.

578    Mark Swilling and Nina Callaghan Given the renewed popularity of state-​led industrial development in Africa (Nem Singh and Chen 2018), developmental-​state theory is now widely used to make sense of current trajectories. However, for reasons that are unclear, developmental-​state theory has largely ignored the role of SOEs, despite SOEs playing increasingly important economic roles (Nem Singh and Chen 2018). Developmental-​state narratives are attractive because they tend to idealize the Asian Tigers’ experience by depicting developmental bureaucracies as autonomous, rational, relatively free of rent-​seeking, unconstrained by clientelism, and institutionally well resourced. Research becomes a search for these conditions or not, and policy is about putting them in place. However, as Whitfield argues, this risks searching ‘for an ideal type of governance that never existed in the first place’ (Bowman 2020:7) The alternative is to be found in the so-​called political settlements literature which goes beyond the formal roles and characteristics of state institutions. Similar to Swilling’s conceptualization of the composition of the ‘polity’ in developmental states drawing on ‘policy regime’ theory (Swilling 2020: 195–​226), the political settlements literature argues that ‘understanding why widely adopted industrial policies succeeded in some instances but not others requires extending the analytical frame to encompass the wider distribution of, and struggle for, organizational power and rents among social groups within and beyond the state’ (Bowman 2020: 8). What makes the difference between success and failure is not whether the state is ‘autonomous’ or ‘developmental’, but whether the balance of forces within and outside the state allows for the emergence of ‘pockets of effectiveness’ to flourish (Bowman 2020:  8). This can happen in generally corrupt or relatively uncorrupted polities, depending on the context. It is obviously easier to achieve in relatively uncorrupted polities. The upshot is that a more nuanced understanding of the institutional context of each SOE is required to inform policy and governance of SOEs. In particular, it creates the space for interventions that do not depend entirely on first ‘fixing everything’ before anything is done. In short, this chapter is informed by three questions: What is the role of SOEs in driving industrialization within a ‘national capitalism’ framework, historically and in the present? What were the relational dynamics and contestations that shaped and undermined these roles? This provides the basis for the third: What role should SOEs play in the third attempt to build an inclusive ‘national capitalism’ in South Africa?

27.2  SOEs during the Apartheid Era The apartheid state enjoyed its most economically successful era between 1948 and the late 1960s. It was during this time that the axis upon which the economy turned, was clearly defined as the MEC. Powerful alliances between finance, mining, energy, and its associated industries locked in the key political, economic, and energy arrangements of apartheid South Africa, glued together with cheap coal and cheap black labour (Fine and Rustomjee 1996). SOEs like Iscor, Eskom, the South African Coal and Oil Company (Sasol), and other

State-Owned Enterprises in South Africa    579 chemical and mineral beneficiation industries, grew to dominate the MEC landscape. Fine and Rustomjee (1996) estimate domestic investment in the steel and chemical sectors at nearly 50 per cent of total investment in the economy in the 1970s, investments directed mainly by another SOE, the Industrial Development Corporation (IDC). SOEs served to buttress the MEC, allowing economic advantages for a core set of industries and institutions, to the detriment of the growth of other industrial sectors and more diverse actors (Fine and Rustomjee 1996). State-​directed industrialization provided sheltered employment for the white working class who also enjoyed the added benefits of housing and food subsidies, pensions, and networked infrastructure for water, electricity, sanitation, and public transport around which a host of SOEs proliferated. Not only did SOEs create the resilient arrangements for white capital and citizen privilege, they also became apartheid’s lifeline during a period of tightening international sanctions and disinvestment during the 1980s (Schneider 2018) providing fuel, arms, minerals, energy, and food. SOEs were the levers to power and racial dominance during apartheid, helping to embed structural inequality in South Africa’s economy and create the obdurate conditions for the MEC.

27.3  SOEs in Early Democratic South Africa The democratically elected government of 1994 was tasked to unravel over three hundred years of structural inequality where the gains of industry were divided between Afrikaner capital and foreign capital (Clark 2014; Hart and Padayachee 2013; Schneider 2018). Their answer to this overwhelming task was to open the doors to the market, favouring privatization and liberal trade. Neoliberalism is characterized by a rejection of the interventionist state that is seen as a hindrance to economic and political freedom. State intervention is seen to throttle competition, which in turn, is understood to be an equalizing force and sets limits on rent-​seeking. State institutions are further cast as lethargic, inefficient bureaucracies that are easily corruptible to serve narrow interests (Mazzucato 2015). After focusing on political–constitutional transition during the lead-​up to 1994, the African National Congress (ANC) came to power without a clear-​cut economic policy framework (Padayachee and van Niekerk 2019). Nevertheless, President Nelson Mandela introduced the first general developmental policy of democratic South Africa, the Reconstruction and Development Programme (RDP), between 1994 and 1996. The RDP ushered in the country’s biggest social welfare initiatives—​social grants, housing, health, electrification, and school feeding schemes. Acknowledging the skewed benefit that SOEs delivered during apartheid, the government intended to use them for redress, as a way to fund transformation.

580    Mark Swilling and Nina Callaghan Between 1994 and 1999, many SOEs were operating at a loss and became a drain on the public purse. Privatization of SOEs, the transfer of state ownership rights to private shareholders, was seen as a remedy for poor economic performance. This transfer shifts the responsibility of delivering public goods and services from the state to the private sector. It implies a steep social cost to poor consumers as prices are determined by market mechanisms and regulation and not by a social contract (Parker and Saal 2003). The RDP was abandoned in favour of the Growth, Employment and Redistribution Strategy (GEAR) implemented between 1997 and 2005. By then, several lesser SOEs had been restructured, an approach that was intensified during GEAR as a way to attract foreign investment. The economic policy came in for severe criticism for being too market friendly while subjecting the poor to punishing austerity (Mosala, Venter, and Bain 2017). Market-​related tariffs, privatization, and flexible and relaxed exchange controls were clearly motivated by neo-liberal ideas. This may have resulted in fiscal gains, but the poor black majority were not given broad-​based access to the economy. In this configuration, SOEs were not envisioned as solving democratic South Africa’s most pressing problems. Nor were they strategically positioned as part of a nation-​ building project to build an inclusive national capitalism premised on an enlarged internal market coupled to an expanded manufacturing base. Even though GDP growth was achieved, it did not translate into greater rates of employment, well-​being, or restorative justice. Inequality and poverty actually rose during GEAR’s implementation. Fine (2010) points to relaxed regulations as one of the main reasons for capital flight, with an estimated 20 per cent of GDP leaving the country during the post-​apartheid era up to 2012. After a programme of restructuring SOEs during the GEAR years to set them up for privatization, the idea of a wholesale sell-​off of SOEs was abandoned after 2002 following the adoption of the developmental-​state narrative at the 2002 ANC Policy Conference. SOEs, together with the Public Investment Corporation (PIC), came to be seen as key building blocks of a developmental state. However, the focus was on state intervention in the economy and not the complex requirements to initiate an industrialization process (Greenberg 2006; Smith 2010). The Accelerated and Shared Growth Initiative (ASGISA) was implemented from 2006 to 2009, including large investments in physical and social infrastructure, much of which was executed via SOEs. The Rand 787 billion infrastructure expenditure plan (9.7 per cent of GDP) was set to be implemented between 2009 and 2012. The plan was considered by government as a way of rectifying poor services in black communities and to provide a buffer against the impact of the global recession following the economic crisis of 2008. Major SOEs like Eskom, Transnet (transport and ports), Sanral (roads), Infraco (broadband), Telkom, SARCC (roads), and Acsa (airports) were set for substantial inflows of funds to finance counter-​cyclical infrastructure investments by the SOEs. This shift from the GEAR-​oriented privatization strategies of the late 1990s, to the ‘developmental state’ approach to SOEs between 2002 and 2009, paved the way for what followed during former President Jacob Zuma’s era from 2009 onwards. Whereas the SOEs and the PIC became Mbeki’s delivery vehicle for infrastructure-​led growth after 2002, for Zuma SOEs were the honeypot at the centre of the state-​capture project. In

State-Owned Enterprises in South Africa    581 his speech at the historic 2007 ANC Elective Conference, Mbeki admitted that Black Economic Empowerment had failed because it depended on the white-​owned corporate sector to create opportunities for black business. This reinforced internal ANC policy thinking that maybe the future lies not in further dependence on the corporate sector, but on using what were called ‘our assets’ to transform the racial composition of the economy.

27.4  Using Procurement to Build a Black Industrial Class During the Zuma era, the focus shifted to the procurement spend of SOEs. The stated rationale was building a black industrial class. While a laudable and much-​needed goal, given the dominance of white capital, this switch in practice had two consequences. First, it reinforced a questionable assumption that capital-​intensive investments in large-​scale infrastructures (i.e. ‘capital deepening’) leads to the type of growth and development that is needed. Capital-​intensive investments have a poor Rand-​to-​ job ratio (Rodrik 2006). Second, it prepared the way for state capture as the shadow-​ state networks came to broker the deal-​making process that a focus on procurement spend opened up. Within this context the rackets thrived. The conditions were ripe for an assertive power elite to repurpose state institutions in the name of addressing the contradictions of the Mbeki era. The solution Zuma’s policy group offered was therefore unsurprising:  heavy dependence on the use of the procurement systems of the SOEs. As articulated by the Department of Trade and Industry (DTI) from 2014 onwards, SOE procurement spend held the key to building a ‘black industrial class’ (Department of Trade and Industry 2014). In reality, this meant re-​purposing the SOEs to become the primary mechanisms for rent-​seeking at the interface between the constitutional and shadow states (Bhorat et al. 2017). This became the strategic focus of the power elite that formed around Zuma. To facilitate this, they needed brokers to help by-​pass regulatory controls and shift money around (through local and international financial circuits) to finance deals, as well as the transformation of the ANC into a compliant legitimating political machine. The Gupta and BOSASA networks emerged as the anointed brokers of this expanding rent-​seeking system. They were the grand masters of the dark arts of racketeering. They actively and systematically established an interlinked set of rackets across the apexes of key SOEs (Bhorat et al. 2017; Swilling, Foley, and Callaghan 2021). Starting as a revolt against Thabo Mbeki, though not yet associated with a clear ideology, the Polokwane moment in 2007 gave rise to a new power elite that used the language of ‘radical economic transformation’ to bind together the new Zuma-​centred political coalition. Claiming to speak for ‘ordinary people’, those who are not well educated, who don’t speak English well, who live in shacks, small towns, and rural areas, and who are excluded from the economy and the formal institutions of the state, this

582    Mark Swilling and Nina Callaghan narrative gave rise to a politics profoundly mistrustful of the formal ‘rules of the game’, whether of the constitution or legislation. The formal rules are rigged, this position proclaimed, in favour of whites and urban elites and against ordinary people. Radical economic transformation was thus presented as a programme that required changing—​ and frequently breaking—​the rules, even those of the Constitution. The argument was compelling when viewed from the perspective of those who did not benefit significantly during the Mbeki era. More significantly, it was a licence to unleash rent-​seeking rackets on a grand scale. And at the centre of it all was an eminently believable lie: breaking the rules is a necessary condition for building a robust black industrial class to displace the dominance of ‘white monopoly capital’. The SOEs—​and their respective procurement spends—​were at the very centre of this political strategy.

27.5  Current State of Play Excluding the PIC, the SOEs directly managed over R2 trillion worth of assets by 2019. The PIC’s investments on behalf of the Government Employees Pension Fund (GEPF) equals the total assets of all the SOEs put together, i.e. R2 trillion. This is also equal to two-​thirds of GDP. To put this into perspective, all the other development finance institutions (DFIs) put together (i.e. DBSA, IDC, Land Bank, NEF, etc.) manage investments worth only 5 per cent of GDP (Makgetla 2020). Figure 27.1 presents a breakdown of the SOEs according to share of total capital assets. Eskom, Transnet, and Sanral are by far the largest, with the Industrial Development Corporation (IDC) and Development Bank of South Africa (DBSA) the largest DFIs. In June 2020, the National Planning Commission (NPC) released a report entitled ‘The Contribution of SOEs to Vision 2030: Case studies of Eskom, Transnet and PRASA’ (NPC 2020c). The report concludes that four factors have contributed to ‘chronic underperformance of some SOEs, and in some cases, near collapse’. These four are: ‘years of uncertain policy expectations, precarious funding strategies, poor institutional accountability and poor governance, and political interference’. One key consequence has been the failure of these SOEs to deliver large-​scale capital projects within approved budgets, thus contributing in a major way to the current fiscal crisis of the South African state—​ what the report gently refers to as ‘macro-​fiscal stress’ (NPC 2020c: 9). In a PowerPoint presentation entitled ‘Infrastructure SOEs Contribution to NDP Vision 2020’ presented at the NPC-​GTAC webinar on 30 July 2020, some extreme cases of cost over-​runs, late delivery, and poor value for money were presented, including: • Gauteng Freeway Improvement Project—budget: Rand 11.4 billion; actual: Rand 17.4 billion • Gautrain Rapid Rail Link System—budget: Rand 6.8 billion; actual: Rand 6.8 billion • Eskom’s Ingual Pumped Storage Scheme—budget: Rand 8.89 billion; actual: Rand 25.9 billion

State-Owned Enterprises in South Africa    583 SAA 1% 13 others 2%

SAPO 1% ACSA 1% TCTA 2% CEF 2% Land Bank 2% Telkom 3% Prasa 4%

Eskom 36% DBSA 4% IDC 7%

Transnet 17%

Sanral 18%

Figure 27.1  Assets by SOC and DFI (excluding PIC), in billions of Rand, 2019 Source: Makgetla (2020).

• Transnet’s New Multi-​product Pipeline—budget: Rand 12.7 billion; actual: Rand 30.4 billion • Eskom’s Medupi: budget—Rand 70 billion; actual: Rand 208 billion • Eskom’s Kusile: budget—Rand 80 billion; actual: Rand 239 billion • PRASA—spent Rand 80 billion on new coaches, but 50 per cent are in operation while quality of service has declined resulting in a significant reduction in the customer base. The NPC compares this to the SIPS14 new universities construction project and the 6,422MW of power procured by IPPs for Rand 210 billion—​neither of which cost government a cent and delivered on budget. Unsurprisingly, the underperformance of SOEs resulted in reduced infrastructure spending by the SOEs, which impacted overall on all public-​sector institutions. According to the NPC’s presentation to the NPC-​GTAC webinar, SOE investment in infrastructure has been in decline since 2014, which contributes to the overall decline in public-sector infrastructure spending and, therefore, gross fixed capital formation (GFCF). As discussed below, the post-​2002 policy switch resulted in a significant jump in SOE infrastructure investments for a seven-​year period starting in 2007/​08, after which it declined to a point in 2018/​19 nearly equal to where it was in 2007/​08 (NPC 2020a). As argued elsewhere, although Zuma became president in 2009, it was only from 2014 onwards that all the necessary conditions were in place for grand-​scale looting (Bhorat et al. 2017).

584    Mark Swilling and Nina Callaghan By late 2020, the bond markets were becoming increasingly jittery about the potential of a major SOE debt default to trigger a sovereign debt crisis. In the 3 August edition of Credit Risk, the bulletin published by Rand Merchant Bank (RMB) Global Markets Research, an article entitled ‘The Systemic Risk of SOE Funding’, raises disturbing questions. Without referring explicitly to the possibility of an SOE debt default triggering a sovereign debt crisis, it is argued that the Land Bank default in 2020 and the SAA fiasco have deeply unsettled the bond markets. The article concludes, rather suggestively, as follows: Although the Department of Public Enterprises (DPE) is unlikely to approve applying business rescue proceedings given its experiences with SAA, one can’t rule out a creditor applying for this via the courts—similar to SAA Express. The difference between this potential situation, and that of either Land Bank or SAA, is that Denel has JSE-​listed government-​guaranteed debt, which would become due and payable under business rescue. The question remains, which fund manager holding government-​guaranteed paper would be brave enough to make that call, knowing full well SA’s fiscal situation? That said, these fund managers too have fiduciary duties to follow. The alternative is National Treasury support. However, given the fiscal issues the country is facing, how would/​could they attempt to do so? (Rushton 2020)

SOEs provide infrastructures that benefit the mines, energy, and refineries. At the same time, except mainly for Transnet and Telkom, the SOEs have become an increasingly serious burden for the sovereign. Eskom’s debt, in particular, has become a major threat, having climbed to nearly 12 per cent of national debt and nearly 2 per cent of the national budget. Of the major SOEs, Eskom and Sanral are loss makers, while Telkom and Transnet remain profitable (Rushton 2020). Eskom is also the recipient of the largest chunk of government-​guaranteed debt (Rushton 2020). Given that government adds together actual transfers and guarantees, it is clear that Eskom dominates the picture. An updated and detailed overview of guarantees is provided in Table 27.1 from the National Treasury. Overall, actual exposure has increased from Rand 327.3 billion to Rand 385.3 billion. Once again, the dominance of the Eskom guarantee is very clear. If SOEs were playing an appropriate counter-​cyclical role, one would expect them to be in good-​enough standing to attract increased bank credit during hard times. This has not been the case. Although total credit extended by banks to SOEs levelled off at Rand 52 billion in 2020, there has been a steady decline since 2014, mirroring the overall trends discussed in this chapter, both financial and institutional, as state capture hollowed out these institutions. The declining year-​on-​year growth in bank credit to SOEs reflects declining confidence in these institutions at exactly the time the opposite was required. Bank credit for DFIs, however, has not declined, and some DFIs still successfully hold bond auctions.

State-Owned Enterprises in South Africa    585 Table 27.1: Government guarantee exposure 2017/​18

2018/​19

2019/​20

R billion

Guarantee

Exposure

Guarantee

Exposure

Guarantee

Exposure

Public institutions of which:

469.8

327.3

487,7

368.1

484.4

385.3

Eskom

350.0

250.6

350.0

285.6

350.0

297.4

SANRAL

38.9

30.4

38.9

39.5

37.9

39.9

Trans-​Caledon Tunnel Authority

25.7

18.9

43.0

14.3

43.0

13.5

South African Airways

19.1

11.1

19.1

15.3

19.1

17.3

Land and Agricultural Bank of South Africa

9.6

3.8

9.6

1.0

9.6

0.9

12.2

4.1

11.4

4.3

10.0

4.6

South African Post Office

4.2

0.4

2.9

–​

–​

–​

Transnet

3.5

3.8

3.5

3.8

3.5

3.8

Denel

2.4

2.4

3.4

3.4

6.9

6.9

South African Express

1.1

0.9

2.8

0.2

1.9

0.2

Industrial Development Corporation

0.4

0.1

0.5

0.1

0.5

0.1

Development Bank of Southern Africa

South African Reserve Bank

–​

–​

Independent 200.2 power producers Public–​private partnerships

9.6

0.3

–​

–​

–​

122.2

200.2

146.9

200.2

161.4

9.6

10.5

10.5

8.7

8.7

Source: National Treasury (2019).

So far, we have a high-​level view of the state of SOEs in South Africa’s political economy. It is worthwhile taking a deeper dive into individual SOEs to understand the internal dynamics that reflect the kind of maladies explained in the NPC report. How did uncertain policy, political interference, poor governance, and accountability actually play out inside these SOEs? This chapter now turns to case studies of Eskom and Transnet.

586    Mark Swilling and Nina Callaghan

27.6  Eskom Case Study Eskom was reasonably well run during the immediate post-​apartheid period. It won the Financial Times’ prestigious global power company of the year award in 2001 for providing the world’s lowest-​cost electricity. In 2020, the rating agencies declared Eskom as a major systemic risk for the South African economy and the Special Investigating Unit (SIU) informed parliament that over 5,000 employees are being investigated for possible criminal behaviour (Paton 2020). The Eskom case reveals that the weakening of state-​owned institutions has deeper roots than state capture during the years 2009 and 2018. Jaglin and Dubresson (2016: 6) note that ‘Eskom’s crises, are institutional in nature and arise above all from the nature of the relationship between Eskom and the state’. It is, in fact, a succession of misalignments and readjustments that has left Eskom unable to deliver securely on its electricity mandate as defined in the Electricity Act (Act 4 of 2006). The De Villiers Commission was established in 1983 to ‘examine Eskom and electricity pricing throughout South Africa’ (Conradie and Messerschmidt 2000: 244). Its far-​ reaching recommendations ushered in a new techno-​political regime during the period 1987 and 2001 during which Eskom was incorporated as a company and the National Energy Regulator (NER) was established. It went from being self-​regulating (in its parastatal phase) to agreeing to politically directed price compacts to reduce the real price of electricity between 1985 and 2000 by 20 per cent. Two decades of liberalization of the industry led to the first coal-​power-​system emergency in 2008. The steady increase in demand and the ‘sweat the assets approach’ in the 1990s and 2000s meant higher volumes of coal were required to support the higher generation load factors. This, in addition to lower expenditure on future fuel supplies, meant that coal system buffers were eventually eroded over a two-​year period, leading to the eventual collapse in coal security and the January 2008 load-​shedding crisis (NERSA 2008). In 1999, the government and the NER decided against allowing Eskom to start development on the next base-​load power station, which was critical for the country’s supply adequacy. This decision created severe energy insecurity within the next decade, and South Africa is yet to recover from the economic reverberations this triggered. While the utility remained state owned, the relationship dynamic between the utility and government shifted critically, with the latter now playing the roles of energy policymaker, economic regulator, environmental regulator, shareholder, and appointer of the board and executive leadership, with the utility becoming subordinate in the relationship yet accountable for the management of the contradictory requirements from these different spheres of government. The three new power stations (Ingula, Medupi, and Kusile) were eventually commissioned in 2005. This massive expansion was in addition to the requirement that Eskom maintain peak performance from existing operating plants to meet demand.

State-Owned Enterprises in South Africa    587 With support from the shareholder to respond urgently to the energy shortages (Mbeki 2008), Eskom, with limited skills and project development capacity, proceeded with the new build programme and internalized all the mega-​project risks (engineering, procurement, construction) to gain traction on the build programme. Figure 27.2 illustrates the events of a tumultuous decade after the 2008 emergency for Eskom. The viability of the organization declined as costs increased and revenues declined amid frequent leadership changes. The decision to support the build programme with government guarantees, instead of a larger equity injection in proportion to the size of the capital investment, created a high-​risk condition for the sovereign, which has subsequently materialized. The sovereign’s declining credit profile, due to the extensive support it has provided to Eskom, has created a systemic risk for the financial system. From the start of the build programme, the National Energy Regulator of South Africa (NERSA) attempted to protect electricity consumers from tariff spikes. That culminated in a decade-​long series of enormous revenue shortfalls, increased levels of borrowing, and finally a company liquidity crisis in 2018 and a national fiscal crisis in 2019. From 2015, Eskom made a concerted effort to transform the ownership of the existing forty-​year coal-​supply agreements from the diversified global mining companies to new Broad-​Based Black Economic Empowerment entrants. This purchasing strategy, in addition to the sustained under ​investment in the tied collieries, contributed to another decline of the coal stockpiles, leading to a second coal emergency in a decade in November 2018. This purchasing strategy led to a steep increase in coal costs as high volumes had to be trucked over longer distances. By 2019, the board revoked this decision and reverted to the previous coal strategy by re-​establishing the tightly co-​ ordinated coal supply and logistics chain between dedicated coal mines and power stations (Eskom 2019). From 2013 to 2018, customer arrears increased drastically, due mostly to a decline in municipal payment levels. The contributing factors to this revenue loss was an inability to exercise legal credit control processes and the limited results of government-​led processes to address municipal government payment performance. From 2014, the ethical tone of successive CEOs and the broader leadership declined. This was reflected in direct instructions that subordinates over r​ ide governance controls and intervene in the procurement and supply chains of Eskom (Corruption Watch 2019, 2020). Deteriorating employee relations, the weakened role and authority of the board, together with a complete loss of business agility and an inability to control costs and respond to operational crises led to the second major coal stockpile depletion in November 2018 and deep load-​shedding in March 2019. Jaglin and Dubresson (2016: 6, citing Hecht) note that Eskom is fundamentally viewed as a techno-​political object through which technology and politics are entangled for ‘the strategic practice of designing or using technology to constitute, embody or enact political goals.’ In this context, Eskom was transformed into a corporatized state-​owned entity that lost the ability to influence its business environment effectively. It remains constrained by the following dynamics:

Figure 27.2  A tumultuous decade leading up to the second coal-​power-​system emergency in 2018 and the liquidity insolvency in 2019 Source: Erica Johnson in Mark Swilling, Robyn Foley and Nina Callaghan, 2021 (in refs) 

State-Owned Enterprises in South Africa    589 • the contradictions of state policy which, in turn, ultimately results in the weakening of the balance sheet • the inability to execute any decisions in its own interest as the interests of its stakeholders and shareholders dominate decision-​making • because of path-​dependent behaviours resistant to change, key stakeholders have the ability to enact forms of soft violence on the state-​owned entity that remains a constant and imminent threat. In Parliament in September 2019, former Eskom chairman Jabu Mabuza said the company could only support debt of Rand 200 billion. However, at the end of March 2019, Eskom had debt of Rand 440.6 billion. During the year to the end of March 2019, Eskom paid interest of Rand 30.2 billion on its debt, which was equivalent to 16.8 per cent of its revenues. An analysis of Eskom’s income statement shows that its costs were about 115.6 per cent of revenues in 2019. Therefore, the gap between revenues and costs is almost the same as the interest payments. So, Eskom keeps borrowing to pay off its debt. With debt of Rand 200 billion (instead of Rand 440.6 billion) during the year to March 2019, Eskom would have slashed its interest payments by Rand 16.5 billion (or 9.2 per cent) of revenues to Rand 13.7 billion, assuming an average interest cost of 6.85 per cent and approval by NERSA of a more cost-​reflective tariff. None of this, of course, materialized. The result is that Eskom’s debt ballooned to Rand 480 billion by the middle of 2020, with no strategy in place to resolve the problem. Just over Rand 200 billion of this debt matures within the next three years. At the time of writing (November 2020) it is clear that if nothing drastic is done, it is conceivable that an Eskom payment default could trigger a sovereign debt crisis.

27.7  Transnet Case Study The case of Transnet is an example of one of the major nodes of organized corruption in South Africa’s recent history. One of Transnet’s focus areas during the 2010s was to increase its freight transport operations capacity as a response to an expected growth in demand. What followed was Transnet’s capital expansion programme, the Market Demand Strategy (MDS), which kicked off in 2012 and had its sights set on the expansion of rail, port, and pipeline infrastructure. As a Rand 300 billion capital investment plan (with Rand 420 billion budgeted overall for operations) over seven years, Transnet’s new procurement strategy attracted the attention of the rent-​seekers in the emerging Zuma-​centred power elite, who pocketed billions. With the implementation of the MDS, Transnet’s annual capital expenditure shot up from Rand 28.3 billion in 2012 to Rand 36.4 billion in 2015, the largest amount ever spent by Transnet in a twelve-​month period (DPRU 2020). However, spending began to falter in 2016 as the projected economic growth failed to materialize and doubts about the financial viability, and foundational assumptions of the MDS, resurfaced. Transnet was also included in the Public Protector’s ‘State of Capture’ report in November 2016.

590    Mark Swilling and Nina Callaghan Capital spending decreased in the following years and by 2018, the last year of the strategy, half of the MDS budget remained unspent (DPRU 2020). Evidence suggests that the first major move to enable access to substantial rents at Transnet was the appointment of Malusi Gigaba as the new minister of public enterprises in 2010 (DPRU 2020). Gigaba oversaw the MDS, Transnet management, as well as the top positions which he quickly filled with members of his shadow-​state network. In 2011 Brian Molefe—​ known for his close ties with the Gupta family—​was appointed as Transnet’s CEO. This was followed by the reappointment of Siyabonga Gama as head of Transnet Freight Rail (DPRU 2020) after being dismissed for awarding irregular contracts. Anoj Singh, Transnet’s chief financial officer (CFO) from 2009 to 2015 oversaw the MDS and other Transnet financial operations. These appointments were followed by the awarding of a contract to ZPMC towards the end of 2011 (DPRU 2020). Of the R650-​million contract for seven cranes, R95.6 million was transferred to a Gupta shell company as a kickback payment. By 2012, Iqbal Sharma—​another well-​known associate of the Guptas—​was appointed as the head of the newly formed Board of Acquisitions and Disposals Committee (BADC) (DPRU 2020), formed to oversee large infrastructure spending at Transnet. Sharma was also connected to Salim Essa—​an important member of the state-​capture network—​which resulted in the latter being paid billions of Rands in kickbacks from Transnet deals between 2012 and 2016 (DPRU 2020). This practice of absorbing illicit kickback payments into the estimated total cost of large contracts at Transnet occurred for several years, with the Guptas being primary beneficiaries. Procurement corruption, while a problem at many SOEs, was unique at Transnet because of its scale and level of sophistication. The infiltration extended into nearly complete control of Transnet, becoming more brazen over time as procurement procedures and anti-​corruption mechanisms were routinely over​ridden. The most brazen of the procurement contracts awarded was for over a thousand locomotives at Transnet from 2012 to 2014. Two Chinese companies were awarded the contracts totalling almost Rand 25 billion and kickbacks to the Gupta network made up 25 per cent of the total contract (DPRU 2020). The process of confinement was the preferred strategy to ensure specific firms were awarded the contracts as it bypassed competitive tender procedures. Transnet later awarded four other contracts for a further 1,064 locomotives. The BADC approved the overall procurement budget even though it had increased from Rand 38.1 billion to Rand 54.5 billion (a 43 per cent increase that likely incorporated the kickback payments (DPRU 2020)). Through analysis of the correspondence contained within the #GuptaLeaks, the DPRU estimates that between 2013 and 2018 Transnet spent over Rand 37 billion on deals linked with corruption, nearly Rand 8 billion of which was paid to firms linked to the Guptas in direct kickbacks. Evidence shows Transnet executives worked to facilitate the corruption. Lack of delivery on contracts, unfulfilled contractual agreements around local content which specified that components of the trains had to be manufactured locally, and

State-Owned Enterprises in South Africa    591 the dismantling of systems of control and oversight, are some of the legacies of state capture that occurred at this key South African SOE.

27.8  SOE Governance and the Third Bid for an Inclusive ‘National Capitalism’ The national debate about the future of SOEs instigated by the NPC investigation into the governance of SOEs comes at a crucial historical moment within the wider political economy of the Global South. After three decades of privatization of SOEs at the end of the last century, more and more SOEs are emerging across the Global South as part of the so-​called new developmentalism. According to Singh and Chen: After 2000, another variant of state-​led networks in newly industrialised countries began to emerge—​competitive SOEs that can foster efficient allocation of state resources to achieve developmental objectives. In the post-​Washington Consensus era, convergence between market competition and re-​incorporation of SOEs concretely offered a new developmental approach for emerging markets. Importantly, new developmentalism shifts the analytical focus from state–​social alliances towards the remoulding of state–​state relations. (Nem Singh and Chen 2018:  1079, emphasis added)

The strategic focus since 2000 has been less about how SOEs relate to business, and more about how states create appropriate institutional relations with and for SOEs that enable them to be competitive, developmental, and accountable at the same time. The two case studies confirm this argument—​they clearly show that it was the relations between the SOE and the relevant minister and/​or department that went wrong. In many other Global South jurisdictions, the lessons are being learnt and appropriate governance structures put in place. In short, using the language of this chapter, many industrializing countries in the Global South are redeploying SOEs to foster more inclusive growth-​oriented modes of ‘national capitalism’. As Singh and Chen observe: ‘Crucially, many countries achieved economic success with a large SOE sector’ (Nem Singh and Chen 2018: 1083). It is arguable that the COVID-​19 crisis has starkly revealed the contradictions of South African economic policy and practice. In a perfect storm, the fiscal challenges created during the state-​capture years have been exacerbated by the impact of the ‘pancession’ on tax revenues, fiscal spending, and debt levels. As we have argued elsewhere, there needs to be a fundamental break from past economic policy paradigms (Moleko and Swilling 2020). The ‘New Wine into New Wine Skins’ report (Moleko and Swilling 2020) recommends a policy package that effectively envisages what would be South Africa’s third bid to create a fully-​inclusive ‘national capitalism’ at a time within the Global South when similar strategic initiatives are underway.

592    Mark Swilling and Nina Callaghan Substantiated by extensive research by the CSIR, the ‘New Wine’ report puts an energy transition at the very centre of such a bid—​from an MEC-​centred economy premised on coal-​fired power to an economy driven by a new vast decentralized and distributed renewable-​ energy infrastructure. Although the Presidential Economic Advisory Council (PEAC) may not share the economic assumptions of the ‘New Wine in New Wineskins’ report, its October 2020 report does echo the report’s energy approach (PEAC 2020): What used to be a choice is now mandatory. Those countries not adapting to a green transition will find themselves behind and excluded. They will be behind on the innovation curve, the cost curve, will suffer from stranded assets and will face increasing barriers to markets that have accelerated their own transitions.

The energy transition is ‘mandatory’ rather than a ‘choice’ because coal-​fired power stations must close by prescribed dates. They cannot be replaced by more coal-​fired power stations because nearly all major international and local funders have decided to withdraw from coal assets. Renewables are now cheaper than fossil fuels by far. In short, there is a structural techno-​industrial and financial imperative that prohibits the continuation of coal-​fired electricity. Just like the steam engine made the industrial revolution possible, so renewables will create a post-​MEC industrial complex (Swilling 2020b, 2020a, 2019). But all depends on whether Eskom—​South Africa’s biggest and most troubled SOE—​can enable the transition. Both the PEAC report and the NPC reports cited earlier clearly emphasize that none of the above is possible without fundamentally re-​organizing the governance of the SOEs. Clearly influenced by the post-​neo-liberal orientation of Mariana Mazzucato (a member of the PEAC), the PEAC report refers to ‘mission-​driven’ public investments via the SOEs. The NPC gives substance to this pragmatic approach by proposing a range of ‘organizational reforms’ and more drastic ‘structural interventions’ along a continuum of options ranging from retention of 100 per cent public ownership through to full privatization (NPC 2020b). The NPC warns against ideological approaches (‘nationalization versus privatization’), preferring rather a contextual analysis of each case to arrive at a pragmatic conclusion. The NPC makes two important recommendations. The first is to narrow down and standardize the mandates of the SOEs. Specifically, these should be: ‘(1) delivery of a core economic mandate on commercial terms; (2) strong governance standards; and (3) financial sustainability’ (NPC 2020b) The second is to better coordinate the governance of SOEs, including considering as a serious option the introduction of a new state holding company that would own all—​or most—​of the SOEs, thus streamlining and aligning policy and governance of SOEs, especially board appointments, and limiting political interference. In recognition of the need to address the governance challenge posed by the SOEs, on 11 June 2020, President Cyril Ramaphosa appointed the Presidential State-​owned Enterprise Council (PSEC) with the following mandate:

State-Owned Enterprises in South Africa    593 The Council’s mandate includes strengthening the framework governing SOEs, including the introduction of an overarching Act governing SOEs and the determination of an appropriate shareholder ownership model. The council will also ensure that SOE-​specific interventions are implemented to stabilize companies through the strengthening of their governance, addressing their immediate liquidity challenges and implementing agreed turnaround strategies. Furthermore, the council will review business models, capital structure and sources of financing for SOEs and will monitor and mitigate risks.

This is a crucial intervention and in tone and inclination is similar to the NPC approach, but both are somewhat at odds with the ‘mission-​driven’ approach emphasized in the PEAC report: by comparison, the more narrow technocratic framing of the problem by the PSEC and NPC is noteworthy. Consistent with a common trend since 2018, except for the PEAC report, the focus of the Presidency and DPE seems to be on ‘fixing the SOEs’. The root problem, however, is political: the absence, since 1994, of a clear developmental vision for the SOEs and how they should be managed accordingly. The passing of a new SOE Governance Act and the creation of a new ownership model should be about creating a massive institutional capacity and capital base for counter-​cyclical investing plus a wide range of new investment instruments to promote diversification. This must include a reassessment of the PIC’s mandate. Given the enormous size of the PIC, it cannot be allowed to pursue pro-​cyclical and ‘safe’ investments that do little to diversify the economy. Government must impose a clear public mandate on the PIC as part of the overall re-​organization of the SOE sector. The underlying cause of what the NPC refers to as the ‘near collapse’ of the SOEs can be traced back to a political leadership that made a succession of specific strategic and tactical decisions over a period of two and a half decades that resulted in the current state of affairs. It is time they took full responsibility for this. Some of these decisions were purely ideological, such as the decision to ignore Eskom’s requests from the late 1990s to build more power stations because the political leadership was deluded by the neo-​ liberal belief that new power stations could be built by the private sector in a low-​cost environment. Other decisions were explicitly about rent extraction, including the political decisions during the Zuma era that resulted in the appointment of ‘crony boards’ and politically loyal CEOs to facilitate large-​scale looting, aided and abetted by some of South Africa’s leading consulting, legal, and accounting firms. Whatever the reason, the brutal reality is clear: the current crisis of the SOEs is caused by decisions made by the political leadership, specifically members of the South African Cabinet. This has much to do with an ever-​shifting mix of ideological misapprehensions and rent-​seeking behaviours that began during the apartheid era and continued into the post-​1994 era. A clear consistent strategic vision of the role of SOEs in national economic development was absent during the Mandela, Mbeki, and Zuma presidencies. The NPC recognizes this centrality of SOEs in a national project when it defines SOEs as ‘engines of economic development’. Unfortunately, as demonstrated in

594    Mark Swilling and Nina Callaghan the case studies and financial analysis presented in this chapter, the reality is very different. According to both the NPC (2020) and Makgetla (2020), the key problem with SOEs is that their developmental mandates and KPIs are unclear, with a narrow focus on outputs rather than outcomes while overall governance is in disarray. Of the thirteen SOEs that were examined closely by Makgetla (2020), only four had clear mandates. Table 27.2 provides an analysis of the executive and policy responsibility for thirteen SOEs. Not only is it clear that responsibility for SOEs is highly dispersed across many different entities, but there are SOEs that are subject to a split between policy responsibility and executive responsibility. Eskom is a perfect case in point, and a key reason Eskom cannot resolve its problems. The Department of Mineral Resources and Energy (DMRE) is responsible for energy policy (in particular the IRP and NERSA), whereas the Department of Public Enterprises (DPE) is responsible for the governance of Eskom as an institution. However, Eskom cannot resolve its governance and financial problems by closing down coal-​fired power plants and ramping up renewables

Table 27.2: Oversight of priority SOEs and DFIs Other departments affected

Other spheres affected

DOE

Economic departments; DEA

Municipalities

NDOT

Same

Economic departments; DEA

Municipalities and provinces

Transnet

DPE

NDOT

The DTI, DMR, DAFF

Municipalities and provinces

Prasa

NDOT

Same

DCOG, DTI

Gauteng metros, Cape Town

SAA

NT

NDOT

Tourism

Municipalities with airports

CEF

DOE

Same

The DTI

Denel

DPE

DTI

Defence, DIRCO

SABC

DOC

Same

GCIS

SAPO

DTPS

Same

DSD

PIC

NT

Same

Economic departments

IDC

EDD

Same

The DTI, DSBD (oversees sefa)

DBSA

NT

Same

DCOG

Municipalities

Land Bank

NT

DAFF

DRDLR

Provincial agriculture

Company

Shareholder

Policy department

Eskom

DPE

Sanral

Source: Makgetla (2020).

Provinces and municipalities

State-Owned Enterprises in South Africa    595 (as required by DPE) if energy policy and implementation is driven by the coal-​and nuclear-​oriented  DMRE. The ‘political settlements’ literature cited earlier argues for a focus on state–​state relations. Table 27.2, which provides a snapshot of the incoherencies, clearly confirms this contention. It also helps to contextualize one of the best-​kept secrets in the public sector, namely the fractious nature of the relationships between SOE executives on the one hand and the senior officials and ministerial advisors of the oversight ministries (e.g. DPE, NT, and NDOT) on the other. The latter resent the substantially higher salaries earned by the former, and compensate by abusing their authority (especially when it comes to approving salary increases and bonuses). These conflicts—​some of which have lasted for years—​undermine the collegiality required to jointly tackle massive strategic challenges.

27.9 Conclusion To conclude, given their historical role prior to 1994 and their sheer size and impact (in terms of institutional capacity, assets, and capital deployment), a logical expectation after 1994 is that the post-​apartheid democratic state would have strategically deployed the SOEs (including the PIC) in ways that purposively directed capital investments to diversify the economy, drive development, instigate socio-​technical innovations, and deploy counter-​cyclical interventions. Instead, after 1994, blinded by the neo-​liberal promise inscribed into the GEAR policy, despite rhetorical commitments to an inclusive ‘national capitalism’, SOEs were regarded as fiscally burdensome and less significant than the private sector. A key set of instruments for building an inclusive ‘national capitalism’ was therefore neglected, a strategic miscalculation consistent with the ANC’s rejection of the economic policy framework proposed by the MERG on the eve of the 1994 founding election. This changed slightly after the adoption of the developmental state narrative in 2002. However, little was done to rationalize the overall governance of SOEs and to clarify their respective mandates. The unintended end result of these strategic blunders was that the developmental-​state narrative created the legitimating rationale for masking state capture during the Zuma years. Since 2018, the Ramaphosa government has emphasized the need to ‘fix the SOEs’. This is all well and good, but this does not resolve the real underlying cause of the problem: the absence of a clear strategic political vision for SOEs as key drivers of a national economic development programme to overcome South Africa’s key structural economic weakness, namely the inherited enclave nature of the economy built on the colonial-​type foundations of the mineral-​energy complex. For the first time ever, by triggering a large-​scale energy transition, South Africa has an opportunity to transcend the MEC and the enclavism that it reproduces in favour of a more inclusive and more sustainable ‘national capitalism’. However, this third bid for an inclusive ‘national capitalism’ will depend on the thorough and complete restructuring of SOE governance in line with international trends

596    Mark Swilling and Nina Callaghan in developing countries since 2000. As argued by the PEAC, this ‘mission-​oriented’ approach should become the strategic vision for the SOEs in a post-​COVID, green, and inclusive economic recovery strategy.

References Bhorat, Haroon, Mbongiseni Buthelezi, Ivor Chipkin, Sikhulekile Duma, Lumkile Mondi, Camaren Peter, Mzukisi Qobo, Mark Swilling, and Hannah Friedenstein. 2017. Betrayal of the Promise: How South Africa Is Being Stolen. Stellenbosch and Johannesburg: Centre for Complex Systems in Transition and Public Affairs Research Institute. Bowman, Andrew. 2020. ‘Parastatals and economic transformation in South Africa: The political economy of the Eskom crisis’, African Affairs, 119(476): 1–​37. Clark, Nancy L. 2014. ‘Structured inequality:  Historical realities of the post-​ apartheid economy’, Ufahamu: A Journal of African Studies, 38(1): 93–​118. Conradie, Steve R and L. J. M. Messerschmidt. 2000. A Symphony of Power: The Eskom Story. Johannesburg: Chris van Rensburg Publications. Corruption Watch. 2019. ‘Zondo Commission—​Eskom pressed to sign up for dodgy R25bn loan agreement’, Corruption Watch, 1 March [online] https://​www.corruptionwatch.org. za/​wp-​content/​uploads/​2019/​03/​Zondo-​Commission-​–​-​Eskom-​pressed-​to-​sign-​up-​for-​ dodgy-​R25bn-​loan-​agreement.pdf. Corruption Watch. 2020. ‘Zondo Commission—​Eskom’s Gupta contract under scrutiny’, Corruption Watch, 23 January [online] https://​www.corruptionwatch.org.za/​wp-​content/​ uploads/​2018/​11/​23Jan2020-​Update2-​Zondo-​Commission-​Eskoms-​Gupta-​contract-​under-​ scrutiny.pdf. Department of Trade and Industry. 2014. ‘The DTI to create 100 black industrialists in three years’. South African Government, Pretoria. DPRU. 2021. ‘State capture and the economics of corruption:  The case of Transnet’, in Mark Swilling, Robyn Foley, and Nina Callaghan (eds) Anatomy of State Capture. Stellenbosch: Sunmedia and Centre for Complex Systems in Transition. Eskom. 2019. ‘Eskom integrated report 2019’. Eskom. Feinstein, Charles H. 2005. An Economic History of South Africa: Conquest, Discrimination, and Development. Cambridge: Cambridge University Press. Fine, Ben. 2010. “Can South Africa be a developmental state’, in O. Edigheji (ed.) Constructing a Democratic Developmental State in South Africa, Potentials and Challenges. Cape Town: HSRC Press. Fine, Ben, and Zavareh Rustomjee. 1996. The Political Economy of South Africa: From Minerals-​ Energy Complex to Industrialization. Boulder, CO: Westview Press. Greenberg, Stephen. 2006. The State, Privatisation and the Public Sector in South Africa. Harare, Zimbabwe: Southern African Peoples’ Solidarity Network. Hart, Keith, and Vishnu Padayachee. 2013. ‘A history of South African capitalism in national and global perspective’, Transformation: Critical Perspectives on Southern Africa, 81(1): 55–​85. Jaglin, Sylvy, and Alain Dubresson. 2016. ESKOM: Electricity and Technopolitics in South Africa. South Africa: Juta and Company (Pty) Ltd. Makgetla, Neva. 2020. ‘The crisis at the state-​owned enterprises’, Trade and Industrial Policy Strategies, Policy Brief 1/​2020.

State-Owned Enterprises in South Africa    597 Mazzucato, Mariana. 2015. ‘Building the entrepreneurial state:  A new framework for envisioning and evaluating a mission-​orientated public sector’. Levy Economics Institute of Bard College Working Paper no. 824, New York. Mbeki, Thabo. 2008. ‘State of the nation address of the president of South Africa, Thabo Mbeki: Joint sitting of Parliament’. South African Government, Pretoria. Moleko, Nthabiseng, and Mark Swilling. 2020. New Wine into New Wine Skins: An Alternative Economic Strategy for South Africa’s Reconstruction. Stellenbosch:  Stellenbosch Business School and Centre for Complex Systems in Transition. Mosala, Seshupo, Jan Venter, and Edward Bain. 2017. ‘South Africa’s economic transformation since 1994: What influence has the national democratic revolution (NDR) had?’, The Review of Black Political Economy, 44(3): 327–​40. Nem Singh, Jewellord, and Geoffrey C. Chen. 2018. ‘State-​owned enterprises and the political economy of state–​state relations in the developing world’, Third World Quarterly, 39(6): 1077–​97. NERSA. 2008. ‘Inquiry into the national electricity supply shortage and load shedding’. Report by the Energy Regulator. NPC. 2020a. ‘Infrastructure SOEs contribution to NDP Vision 2030’. The Presidency, Pretoria [online] https://​www.gtac.gov.za/​Publications/​Infrastructure SOEs contribution to NDP Vision 2030.pdf. NPC. 2020b. ‘Linking state-​owned enterprises to economic transformation and inclusive growth’. The Presidency, Pretoria [online] https://​www.nationalplanningcommission.org. za/​assets/​Documents/​The role of state-​owned enterprises in achieving economic transformation and inclusive growth. Paper 4 Synthesis Report.pdf. NPC. 2020c. ‘The contribution of SOEs to Vision 2030: Case studies of Eskom, Transnet and PRASA’. The Presidency, Pretoria [online] https://​www.nationalplanningcommission.org. za/​assets/​Documents/​NPC Position Paper on The Contribution of SOEs to Vision_​2030.pdf. Padayachee, Vishnu, and Robert van Niekerk. 2019. Shadow of Liberation: Contestation and Compromise in the Economic and Social Policy of the African National Congress, 1943–​1996. Johannesburg: Wits University Press. Parker, David, and David Saal. 2003. International Handbook on Privatization. Cheltenham: Edward Elgar Publishing. Paton, Carol. 2020. ‘SIU red-​flags about 13% of Eskom staff ’, Business Live, 14 October. PEAC. 2020. ‘Briefing notes on key policy questions for SA’s economic recovery October 2020’, The Presidency, Union Buildings, Pretoria, October. Rodrik, Dani. 2006. Understanding South Africa’s Economic Puzzles. Cambridge, MA: John F. Kennedy School of Government, Harvard University. Rushton, Kate. 2020. ‘The systemic risk of SOE funding’, RMB Global Markets Research Credit Insight’ August: 1–​55. Schneider, Geoffrey E. 2018. ‘The post-​apartheid development debacle in South Africa: How mainstream economics and the vested interests preserved apartheid economic structures’, Journal of Economic Issues, 52(2): 306–​22. Smith, Adrian. 2010. ‘Civil society in sustainable energy transitions’, in G. Verbong and D. Loorbach (eds) Governing the Energy Transition:  Reality, Illusion, or Necessity. New York: Routledge. StatsSA. 2019. ‘Gross domestic product (GDP), 2nd quarter 2019’. [media presentation], Pretoria.

598    Mark Swilling and Nina Callaghan Swilling, Mark. 2020a. ‘Investing in renewables to replace ageing coal-​fired power stations is a no-​brainer’, Daily Maverick, 8 June. Swilling, Mark. 2020b. ‘Load shedding will cripple our economic recovery—​we must bring renewables onstream fast’, Daily Maverick, 30 September. Swilling, Mark. 2020. The Age of Sustainability: Just Transitions in a Complex World. London and New York: Routledge Earthscan. Swilling, Mark. 2019. ‘Zexit, a nuclear deal and Russia’s strong-​arm persuasion’, Daily Maverick, 8 January. Swilling, Mark, Robyn Foley, and Nina Callaghan (eds). 2021. The Anatomy of State Capture. Stellenbosch: Sunmedia and Centre for Complex Systems in Transition.

Chapter 28

Bl ack Ec onomi c Emp owerm e nt i n Sou th Af ri c a Thando Vilakazi

28.1 Introduction The Black Economic Empowerment (BEE) policy has been one of the most direct attempts by the African National Congress (ANC)-​led government to achieve racial transformation in the South African economy. BEE was at first defined broadly as ‘an integrated and coherent socio-​economic process that directly contributes to the economic transformation of South Africa and brings about significant increases in the number of black people who manage, own and control the country’s economy, as well as significant decreases in income inequalities’ (DTI 2003).1 The chapter ambitiously focuses on distilling contemporary perspectives on BEE and juxtaposing the BEE programme and its outcomes with a wider understanding of ‘meaningful economic inclusion’ and participation—​in essence, introducing a discussion about the substance of inclusion. It seeks primarily to offer a pragmatic, critical view of how BEE’s implementation transpired, and the pathways forward. The chapter first reviews the state of affairs with respect to black economic empowerment in South Africa and considers key outcomes and understandings of BEE as a tool for addressing constraints to economic participation of black people in the economy since the democratic transition in 1994. For this, the chapter draws on the vast body of literature and reports that have considered aspects of empowerment in different sectors and industries. Much has been written and said about BEE, and the literature contains many

1  In terms of the amended BEE legislation of 2013, black people are defined as Africans, Coloureds and Indians who are citizens of South Africa by birth or descent; or who became citizens by naturalization in different defined parameters.

600   Thando Vilakazi very valuable resources. The chapter critically engages with the main cross-​cutting insights and contributions in the literature in order to identify the gaps that exist in BEE policy in the context of achieving meaningful economic inclusion, and explores the likely pathways in the evolution of BEE policy going forward. The chapter does this by first reviewing briefly the initial interpretations and intent behind BEE and how it evolved from the 1990s. This is a helpful exercise as it highlights an important point for the chapter—​that there was no clear policy from 1994 and so black empowerment took on the narrow (ownership-​focused), least-​disruptive form that it did up to 2000, in the interests of both white business in maintaining their position as they diversified assets internationally and disposed of non-​core assets, and the government in being cautious not to do anything too drastic with its policies that would deter investors. In section 28.2 we consider how this precarious starting point shaped the policy approach that emerged for achieving black economic empowerment as embodied in the 2003 legislation, before focusing on key outcomes and challenges encountered in implementing this framework from 2004 in section 28.3. The latter are analysed through comparing recent evidence on empowerment outcomes against some of the policy objectives of BEE but also a broader lens that considers the more significant challenge of overcoming the systemic ways in which entrants remain excluded from the economy, including very high barriers to entry in various sectors. As a core contribution, we add to the debates an interpretation of what meaningful inclusion could look like as a future pathway for South Africa, highlighting that looking only at quantitative outcomes of BEE misses key issues in terms of the types of barriers that sustain exclusion and concentration. We find that a broader, integrated approach is required, which includes different policies such as for competition and procurement and other measures to facilitate entry and participation in the economy. We find that although BEE takes us in the right direction in this regard, it has not done enough and we suggest priorities for the pathway forward that include a focus on the nature and impact of high barriers to the participation of black businesses, the importance of funding as part of a broader set of interventions, the importance of parallel social and economic policies, and the strategic role that procurement and enterprise development can play in the long term.

28.2  The Evolution of Black Economic Empowerment Policy in South Africa To ‘locate the empowerment project as part and parcel of the transformation of South African society’ (BEECom 2001, Prologue by Cyril Ramaphosa, BEE Commission chairman).

This was one of the central mandates of the BEE Commission which was appointed by the umbrella body for black business organizations, the Black Business Council

Black Economic Empowerment in South Africa    601 (BBC), in 1998. The Commission was tasked with deriving a clear strategy and definition for BEE and finding avenues for the voice of black business to be united. This statement is significant for our analysis in this chapter, when considering how the BEE Commission report would come to shape the trajectory of BEE policy in South Africa in fundamental ways. First, BEE policy was not well located or spelt out in great detail in the Reconstruction and Development Programme (RDP), the ANC’s primary manifesto for transforming the society as it took leadership of the country in 1994, nor the Department of Finance’s Growth, Employment and Redistribution (GEAR) economic policy of 1996. This omission is somewhat surprising, given the extensive history of economic exclusion of black South Africans which seemed to beg for concerted and immediate redress once political freedom was achieved in 1994. While the need to empower black South Africans through training, sector education, small business development, and public procurement had been established in policy (including in the RDP), the RDP (and the ANC’s Ready to Govern before that) had relatively limited detail putting forward black empowerment as a central policy strategy, and certainly not in the form that it eventually took from 2003 (Hirsch 2005). The Department of Trade and Industry (DTI) would later state that the government did not wish to define BEE too broadly so as to be tantamount to the entire economic strategy of the country (DTI 2003), suggesting that BEE had been, and would always be, subservient to a broader programme of economic policy and socio-​economic transformation. Second, it marked a critical shift in thinking about empowerment, brought about by an important grouping of black businesspeople towards a more ‘broad’ interpretation of BEE, no doubt a response to the popular concerns across various quarters that the first five years of empowerment had been skewed towards the enrichment of a few. There was a belief expressed at the Black Management Forum national conference in 1997 that the poor outcomes arose from this lack of control and vision for BEE and disenchantment with a process driven by ‘white institutions’, ultimately resolving that ‘black people should direct and take charge of a new vision for BEE, a process that, until then, had been conceptualised, controlled and driven by the private sector’ (BEECom 2001: 1). Third, and following from the above, the statement suggests that empowerment as a ‘project’ was to be viewed as only one part of a wider process of transformation of the society. This points to the often-​understated fact that various other processes and policies of government were meant to contribute together to achieving (racial) transformation in the society as a whole, and BEE would normatively play only a contributory rather than a central role. In other words, policies in education, health-care, labour and competition law or for state-​owned enterprises, for example, would also carry the burden of socio-​economic and racial transformation of the society, ostensibly with positive impacts on poverty reduction, human development, and economic growth and inclusion. In this way, the policy was arguably a way to correct a market failure relating to redistribution and a tendency to concentration in the context of South Africa’s emerging political settlement and economic structure. We return to these issues in the analysis to follow.

602   Thando Vilakazi The reality that the BBC was grappling with in 1998 was that there was no clear strategy and policy on the part of the still very new government for attaining black economic (rather than political) empowerment. Others have argued that there was in fact a narrow strategy in place to achieve black economic empowerment which was not embodied expressly in ANC and government policy documents but driven by leading established white businesspeople in agreement with the incoming ruling elite (Moeletsi Mbeki quoted in Davis 2020). That is, as part of the various bargains and trade-​offs negotiated with business in the lead up to the elections in 1994, and thereafter, white businesses had already recognized the risks that they faced in not putting something on the table in terms of inclusion of black businesses, managers, and businesspeople in the mainstream economy. There was also an opportunity to garner favour with the political elite and shape government policies in exchange for making some overt strides towards the inclusion of blacks in the upper echelons of business leadership (Bracking 2019). As such, they embarked on a process of transferring ownership of certain business interests to black South Africans (many from within the ANC’s ranks), particularly those such as Cyril Ramaphosa (later president of South Africa), Tokyo Sexwale and various others who had gone into the private sector to build a base of black capital. This first period of black empowerment, between 1994 and 2000, is often referred to as the first phase of BEE, characterized in the main by deals involving so-​called politically connected individuals—​but, critically, driven by established white business interests including their advisors and financiers, notionally on their terms, and not black people, as described by Ruel Khosa in 1999 (Ruel Khosa quoted in Hirsch 2005; 220). The process is well described in the contributions of various authors, and effectively begins in 1993 when Sanlam (a major insurance conglomerate) sold 10 per cent of Metropolitan Life (life assurer) to Methold, a consortium of black businesses and businesspeople led by Dr Ntatho Montlana (later renamed New Africa Investments Limited (NAIL)) (Gqubule 2006; Hirsch 2005). Some of the black businesspeople that served as executive directors of NAIL, which went on to acquire several businesses in different industries, would come to be labelled as the new BEE elite, berated for self-​enrichment and unduly hoarding the gains of the empowerment process (Southall 2007; Freund 2007; Bracking 2019). This concern, which has dominated the media, public perception, and writing about the empowerment process in South Africa (Bracking 2019), is a binary view that seems to obscure some key issues. As BEE was not as yet clearly spelt out in policy throughout the 1990s, it can be said to have taken on the form (in the first phase) that was perhaps the most expected outcome—​that some white capitalists would seek strategically to offer minority ownership of existing businesses, and that only some black ‘capitalists without capital’ would have the resources and strategic and socio-​political positioning to enter into these relatively expensive arrangements at first (Southall 2007; Andreasson 2006). In other words, there was logically no way in the absence of a clear policy that the first phase of capital reform could have been broad-​based and all encompassing, especially in a context where there was also still significant distrust and socio-​political tension between black and white South Africans, and where black businesspeople had no capital.

Black Economic Empowerment in South Africa    603 It is not clear from literature and media why the ‘enrichment’ of a small group of leading black businesspeople was viewed as necessarily problematic in the context of an entire economy in which capital remained concentrated largely in the hands of around six white-​owned conglomerate and family business groups (Fine and Rustomjee 1996; Hirsch 2005; Lewis 1995). On this point, Duma Gqubule, a leading author of the BEECom report and subsequent scholarly contributions on BEE, puts it this way in his book with reference to multiple similar BEE deals that took place after the NAIL transaction in 1993: ‘The real story is that hundreds of capital reform initiatives have delivered negligible equity to black companies. The key issue is not that a few people have been enriched through capital reform, but that so few have been empowered’ (Gqubule 2006: 35). Indeed, by the mid-​2000s very little progress had in fact been made on ownership or presence of black people on the boards of leading firms (Southall 2007), and black representation on boards of Johannesburg Stock Exchange-​listed (JSE) companies was 43 per cent in 2019, below the target of 50 per cent (BBBEE 2019). Arguably, even some of those that had been enriched had not been empowered, and not many had been enriched to begin with. In a broader sense, although poverty levels are lower than in 1994, income inequality in South Africa remained amongst the highest in the world (for a full discussion on inequality, see C ­ hapter 9 in this volume) and unemployment amongst the majority black population has been persistently high, as addressed more comprehensively in ­Chapter 7 of this volume. In addition, inequalities in terms of the distribution of wealth are significant and reinforced by poor design and implementation of the broad set of policies meant to achieve transformation, of which BEE is only one part—​Orthofer (2016: 23) finds that ‘wealth is much more unequally distributed than incomes. One percent of the South African population owns at least half of all wealth, the top decile together owns more than 90–​95 percent’; and inequality in terms of assets and access to finance by race is significant (Chatterjee et al. 2020) and shows limited progress in terms of black participation and transformation. The ability of BEE, which also targets income inequality in its objectives, to contribute significantly in addressing inequality and other social challenges stems at least in part from the balance of power in the broader South African political settlement in which economic policies have favoured and entrenched incumbent white business interests (Chabane et al. 2006). The continued framing in the public discourse of BEE in terms of ownership and control of existing businesses reflects the prevailing contests in the political settlement over the division of existing rents and the strong position and influence still held by leading business groups including through lobbying for policies that favour their interests. In the context of liberalization and a largely orthodox economic policy landscape, as has existed in South Africa, the approach adopted for BEE policy has arguably reinforced rather than radically disrupted the concentration of ownership and economic control of the economy historically held by large conglomerate and family groups. Furthermore, key policy debates have not been resolved, such as on land (re)distribution as a key potential pillar for black wealth creation following outcomes of the national land audit in 2017, which showed ownership of agricultural holdings and farms,

604   Thando Vilakazi for example, being 72 per cent owned by white South Africans (see RSA 2007; and, for a fuller discussion of land reform issues, see ­Chapter 12 in this volume). The issue of land ownership, which is extensive and beyond the scope of this chapter, also relates to the inability of blacks to secure finance for business ventures; and outcomes have also not been good in terms of critical issues for empowering a wider majority of marginalized people in society, such as improving education and health sector outcomes, labour supply quality improvements, employment equity, and skills development (see ­Chapter 31 in this volume for a discussion of changing labour market dynamics), and making greater strides in terms of ensuring that finance is reoriented as a tool to drive development and inclusion, as discussed below and in ­Chapters 46 and 47 of this volume. It seems reasonable also to have expected that those black businesspeople and companies that had been involved in different deals from the mid-​1990s would in time use the accumulated wealth to become industrialists and business owners in their own enterprises, with the hope that they would empower others in turn. Indeed, some of the ‘big-​6’ BEE consortia2 that dominated the early deal-​making before 2003, about 72 per cent of value, are still significant players in different economic sectors, such as finance and mining, and have since built up and leveraged strong capital bases of their own as we discuss below (see Andreasson 2006: 313; and Bowman 2019 on ARM). Some have done so despite a context of deindustrialization and financialization of the economy (Bell et al. 2018), which means a declining contribution of manufacturing value in the economy as a whole. This is not to say that there have not been those who have sought to benefit unduly from the BEE process, or use corrupt practices (Bracking 2019). However, this is the case with most policies and in many countries (Khan 2001, 2010), including in Malaysia, which is often cited as a comparator for successful affirmative action policies, where there are also concerns about fronting, corruption, and extractive practices (Lee 2015). It could be argued that barriers to accessing capital and resources may be so significant and exclusionary, that there are few alternative pathways for capital accumulation but to leverage blackness and government resources. A sympathetic view of the enrichment issue is that the process of inclusion was not going to happen by itself and had to start somewhere, but the stock market crash of 1998, the subsequent drying-​up of deals, and the poor economic performance of various deals that had been concluded, made it clear towards the end of the 1990s that it was time for a coherent policy framework. The DTI recognized ultimately that although the government had instituted various laws cutting across different spheres of society to engender holistic transformation,3 progress had been slow since 1994 and racial transformation of ownership and control in the business sector was especially stubborn. The process of black empowerment, which had been approached loosely through various pathways at least in the public sector, had ‘lacked focus and an overarching strategic framework’ (DTI 2003: 10). 2 

In 2003, African Rainbow Minerals (ARM), Mvelaphanda, Shanduka, Safika, Kagiso, and Tiso. These included legislations for employment equity, land rights and tenure, a national empowerment fund, preferential procurement, and other sector-​specific laws. 3 

Black Economic Empowerment in South Africa    605 The BEECom report of 2001, which was received by then President Thabo Mbeki, informed a highly consultative policy development process led by the DTI, which culminated in the publication of South Africa’s first black economic empowerment strategy and law in 2003 (DTI 2003). Since the passing of this law, most deals, which by 2020 numbered in the thousands, have been broad-​based in some form in terms of beneficiaries, with involvement of a wider range of beneficiaries, such as ownership shares for employees and women, and emergence of various development trusts, and community and education schemes (Patel and Graham 2012; Intellidex 2015). The main proposals put forward in the early 2000s were on the definition of ‘broad-​based’ empowerment, key tenets of a legal framework for BEE, a proposal to institute a government BEE Commission to oversee black empowerment processes, and suggestions on key criteria and measurement parameters for broad based empowerment (BEECom 2001). The Broad Based BEE Act 53 of 2003 (‘BEE Act 2003’) was signed into law by President Mbeki in January 2004. Prior to this, black empowerment interventions had existed in various forms since the democratic transition in 1994, but without a central coordinating policy framework. As part of this second phase (from 2000), various sectors entered into sector-​specific empowerment charters starting with the Liquid Fuels Charter in 2001 followed by mining (we do not discuss these in this chapter, however there are very useful accounts that focus on specific sectors, such as Bowman (2019)), and promulgation of the Act was also accompanied by the creation of guidelines and associated Codes of Good Practice, a balanced scorecard system for ‘gauging success’, and the establishment of an Advisory Council on BEE to advise government and review progress (DTI 2003: 5; BEE Act 2003, section 4).

28.2.1 BEE Legislation and Practice The BEE Act defined broad-​based black economic empowerment as follows (BEE Act 2003, section 1):



The economic empowerment of all black people including women, workers, youth, people with disabilities and people living in rural areas through diverse but integrated socio-​economic strategies that include, but are not limited to—​ (a) increasing the number of black people that manage, own, and control enterprises and productive assets; (b) facilitating ownership and management of enterprises and productive assets by communities, workers, cooperatives, and other collective enterprises; (c) human resource and skills development; (d) achieving equitable representation in all occupational categories and levels in the workforce; (e) preferential procurement; and (f) investment in enterprises that are owned or managed by black people.

606   Thando Vilakazi This policy approach is considered to be ‘broad-​based’ because it focused in its objectives on various aspects of economic participation (ownership, management, control of productive assets, investment, skills and capabilities, procurement, access to infrastructure, and workplace inclusion) by various groups in society that had been previously marginalized including black women and youth, and those living in the rural communities, amongst others (see BEE Act 2003, section 2). While the government’s approach to BEE since the 1990s had never defined empowerment as only being about ownership, in practice and in the public discourse the issues centred on the perceived empowerment of a few black businesspeople (largely male) in the first phase, which explains the increased emphasis on a broader approach that emerged. Under the new strategy, ownership became one of seven main criteria against which the empowerment credentials of businesses in South Africa would be assessed—​the others being management representation, employment equity, skills development, preferential procurement, enterprise development, and corporate social investment. These criteria, as measured in terms of the clearly defined Codes of Good Practice and scorecard system, underpinned the implementation of BEE from 2004 onwards. Against the scorecard, companies would receive classifications from empowerment verification agencies designating their BEE ‘level’ as determined by their scores on the seven different criteria. Some authors have taken the view that the creation of the scorecard system, as embodied in the Codes, and the related BEE verification agencies which would be empowered to assess performance and issue certificates of BEE compliance to companies, would turn BEE into a ‘technical exercise’ that was likely to lead to a focus on superficial compliance rather than the broader objectives of the policy (Ponte et al. 2007: 944). Companies were rated by various agencies and given scores across points levels with Level 1 being a well-​performing firm. Indeed, some authors published roadmaps for how large firms could achieve compliance using different approaches to scoring on the ownership criteria in the main, and how best to identify strategic BEE partners so as to manage certain risks and get partners that would still add some value to the business (Sartorius and Botha 2008). However, it is hard to see how else the task of enlisting established white-​owned businesses to start the process of incorporating black people and businesses in their enterprises and value chains could have happened without a relatively simplified codification system (to deal with a very complex set of problems) which at least provided certainty for firms. Moral suasion would not work, and nor did the ‘rainbow nation’ ethos. A more important question is whether the system, as it was designed in the early 2000s, could drive the economy towards the targeted outcomes. We consider the detail of this question in the section to follow. However, we can already interpret from the revisions of BEE policy mechanisms, which led to an amendment of the law in 2013, that some things did not work as well as anticipated. The BBBEE Amendment Act (46 of 2013) (‘BEE Act 2013’) was signed by the president in January 2014, and became law in October 2014, some ten years since the first act came into force. It sought to strengthen the previous legislation in several important

Black Economic Empowerment in South Africa    607 ways and remained in effect as the prevailing BEE legislation at the time of writing. The amendments provided for, amongst other things, measures to enhance compliance by public entities and monitoring and evaluation; to provide for incentive schemes to support black-​owned and managed businesses; to provide for various penalties and offences such as those to do with fronting practices by businesses; and to establish the BBBEE Commission to deal with compliance with the legislation and enhance awareness (BEE Act 2013). Greater emphasis was also placed in the amendments on increasing access to finance and incentives for black-owned businesses of various sizes. In many ways, the 2013 Act was seen as a step by the government towards more stern enforcement of BEE rules (DTI 2014; Bracking 2019). The changes emphasized the fact that government needed to address lack of compliance by state-​owned companies (which previously had an option not to comply), and low compliance in the private sector because the system was previously based on voluntary compliance by businesses (PMG 2015). It also signalled the government’s intention to focus more on other obstacles facing black businesses beyond access to capital, such as access to markets and supply chains of established companies (PMG 2015). The new Act also led to the 2015 revisions of the Codes, primarily amending the balance between different criteria through reducing the generic scorecard from having seven scoring elements to five; namely, ownership (weight 25, up from 20), management control (15), skills development (20), enterprise and supplier development (ESD) (40), and socio-​economic development and sector specific contributions (5); and setting minimum requirements of 40 per cent of the available score on three ‘priority elements’ being ownership, skills development, and ESD (RSA 2013). The ESD element combines two previous elements from the seven-​element system, being preferential procurement and enterprise development (of suppliers), and the management control element combines employment equity and management control (BBBEE Commission 2017). There are various criteria and targets within each element for firms to be able to score points and obtain a BEE Level rating from 1 to 8 (there are various exemptions for small and micro enterprises—​PMG 2015). The revisions to the codes also included provisions for multinational companies to circumvent having to transfer ownership to black South Africans by contributing to equity-​equivalent programmes approved by the DTI (RSA 2013; Bracking 2019).4 It is this latest revision of the BEE legislation and implementation that provides the entry point for our analysis of the outcomes of BEE. What have been the outcomes and what pathways exist for enhancing the racial transformation of South Africa’s economy in future?

4  Includes initiatives or investments that ‘promote and advance enterprise and supplier development, research and development and critical and core skills’ (see CDH at: https://​www.cliffedekkerhofmeyr. com/​en/​news/​publications/​2014/​bee/​bee-​alert-​3-​november-​bee-​amendment-​act-​and-​draft-​codes-​of-​ good-​practice.html and https://​infrastructurenews.co.za/​2014/​03/​24/​new-​codes-​in-​bee/​).

608   Thando Vilakazi

28.3  Progress with Black Economic Empowerment: Evaluating Outcomes and Substance More than anything, BEE laws put the ordinarily uncomfortable subject of transforming the race of capital in South Africa firmly into the boardroom and corridors of businesses, and society as a whole. It is a pervasive theme in the discourse of many government departments and business platforms, including amongst the largest listed firms. As noted above, this level of discourse about race and its economic implications in the country is likely not to have existed absent a concerted effort to drive home the message about scorecards, affirmative action in the workplace, and maintaining good BEE levels. The Malaysian experience tells us that changing the culture and societal norms of the marginalized majority and elite interests is especially important in driving racial transformation (see foreword by former prime minister of Malaysia in Gqubule 2006). These strides, which are arguably more important than outcomes on the measures themselves, have been attained seemingly in the face of stern opposition and concern with BEE policy and its impact on economic efficiencies. It is easy to underestimate the scale of this task and how long it should be expected to take, especially in undoing the damage of many decades of apartheid policies. While it is not the task of this chapter to estimate how long it will take South Africa to achieve even mildly palatable levels of racial transformation, it is of some significance that Malaysia’s affirmative policies, as part of its New Economic Plan, were put in place in 1971 (Gqubule 2006), some fifty years ago by 2021, and that the process to transform that society, although there has been significant economic progress, is still ongoing and very imperfect (Lee 2015). This lens is helpful in that it allows us to reflect on BEE policy in context, and with a healthy dose of (critical) realism. This approach is taken not to excuse poor outcomes or to be sympathetic to policymakers. Instead, it is helpful in the context of this volume because it enables us to reflect on both the successes and limitations of the policy when considered in context, and to focus on capabilities built in the BEE policy machinery that can be leveraged for future progress. The assessment that follows is structured according to key thematic areas, which are by no means comprehensive of all issues to do with BEE policy in South Africa, but are critical in our view from the perspective of driving meaningful economic participation as a goal of the policy. These include BEE and its ability to engage with barriers to entry in the economy; funding as a critical enabler for BEE; the importance of the transformation project cutting across various policy areas and not just BEE; and critical issues around supplier development, procurement, and recent evidence on emerging black manufacturing firms.

Black Economic Empowerment in South Africa    609

28.3.1 Barriers to Meaningful Participation of Black Enterprises The BEE Act of 2003 and the accompanying strategy referred to the idea of engendering effective and meaningful participation of black people in the economy, without defining the term. However, this framing suggests that there is a desirable substance to inclusion that the policy seeks to achieve, and black companies and individuals simply getting a foot in the door is not enough. In other words, there is an important aspect of the quality of inclusion of black businesses and businesspeople, and not only their quantitative inclusion, that matters. This substantive understanding of participation is also envisaged in different ways in other legislation, such as in the amended Competition Act of 1998 (amended 2019) which added a definition of participation as ‘the ability of or opportunity for firms to sustain themselves in the market’ (Competition Act of 1998, as amended, section 1). In recent studies of competition and industrial development in the South African economy, meaningful participation has been referred to as the ability of (smaller and/​or black-​ owned) firms to enter markets and grow their businesses, and to become effective rivals insofar as they are able to compete with incumbent firms to influence market outcomes (Vilakazi, Goga, and Roberts 2020). In the context of BEE, the draft mining charter of 2018 is the most comprehensive in referring to various aspects of meaningful participation including BEE shareholders having (voting) rights in meetings and decision-​ making to do with key aspects of the business; their ability to leverage their interests to invest in other projects; access to unencumbered net value of their interests; access to a proportion of dividends declared; and, being recognized as clearly identifiable partners in the form of BEE entrepreneurs, communities, and employees (Deloitte 2018). In the context of BEE policy, and future policy pathways, it is important that there is some clarity on what is desired in terms of outcomes. This is critical as it sets the appropriate benchmark for measuring success or failure of the programme. The government in the legislation and the mining charter seems to aim for participation that is more comprehensive in terms of the participation of black South Africans. The approach is akin to the broader definitions of inclusive growth:5 Inclusive growth is both an outcome and a process, which ensures that everyone can participate in the growth process, both in terms of decision-​making, for organizing growth progression as well as in participating in the growth itself (and earning income). On the other hand, it goes some way towards ensuring that everyone equitably shares the benefits of growth. Inclusive growth implies participation and benefit sharing. Participation without benefit sharing will make growth unjust and sharing benefits without participation will make it a welfare outcome. (Emphasis added)

5 See https://​www.redi3x3.org/​inclusive-​growth#_​ftn1 [Accessed: 9 October 2020].

610   Thando Vilakazi The focus on both the process and the sharing of benefits is a useful starting point. It links to BEE policy in that the policy seeks to address the barriers to participation in the process (albeit inadequately as argued below), as well as highlighting the importance of ownership and control so that black South Africans can share in the benefits, including through employment and reduction of income inequalities. The implication is that there needs to be a very good understanding of factors that can inhibit participation and/​or benefit sharing. The economic barriers which keep out rivals in various sectors of the South African economy are significant (Vilakazi et al. 2020). They include a range of often mutually reinforcing factors that make it particularly difficult for new entrants to participate and grow. Barriers affect all potential entrants but are higher for many black businesses because they are often also excluded on the basis of race and not being welcomed into value chains and business networks, as well as in terms of access to capital owing to a lack of collateral and a historical asset base to leverage (Bosiu et al. 2020; Vilakazi et al. 2020). This compounds the effect of various other barriers which include anticompetitive strategic behaviour by incumbent firms to foreclose rivals; high switching costs and long-​term arrangements which limit the ability to access customers; legacy contractual agreements between established businesses and asymmetric bargaining power of smaller rivals and suppliers; challenges in accessing strategic routes to market and value chains; poor access to finance from both private- and public-sector funders; denial of access to essential infrastructures; regulatory provisions which are lobbied for by large firms and often skewed against disruptive entrants; and various structural features of key industries such as banking, finance, telecommunications, and emerging technology markets where network economies and first-​mover advantages mean that the odds are stacked against entrants in terms of expansion and effective participation as rivals (Vilakazi et al. 2020; Bosiu et al. 2020). Through the 2013 amendments, BEE laws and codes have shifted to increase the emphasis on funding and other non-​financial support for black companies (PMG 2015). However, the systemic nature and sector specificity of some of the barriers listed above suggest that BEE policy tools on their own are limited in their ability to address the constraints. For example, collusion or exclusionary practices by incumbents cannot be addressed through BEE, however in some industries it is precisely the fact that these practices exist that means black rivals are kept out, sometimes unknowingly. Addressing barriers of this nature requires competition enforcement and industrial policies in support of entrants, that extend beyond what BEE can accomplish within its legal remit. The above concern speaks to a core limitation of the approach taken on BEE. Although the scorecard mechanism provides incentives for firms to improve procurement from black-​owned suppliers, for example, it largely leaves it up to the companies themselves to determine the nature and terms of engagement between those black suppliers and the procuring firms. That is, in focusing on the score it is agnostic as to what terms of trade those suppliers are offered for example, or the forms of patronage and (racial, gender, and other) biases that mean one supplier or individual is chosen over another. It has been shown in recent research that smaller suppliers, including black-​owned firms, often

Black Economic Empowerment in South Africa    611 experience unfavourable terms, and are often disadvantaged in terms of the asymmetric bargaining relationships they have with powerful buyers (CCSA 2019; Chisoro-​Dube and das Nair 2020). Many black businesses repeat a concern that they simply do not feel trusted by white counterparts, and so they are denied opportunities even where they offer sophisticated or competitive products or services (Bosiu et al. 2020). BEE laws were not designed to address some of the more specific barriers which really keep black firms out. However, a coordination with other policy areas, such as competition law and industrial development strategies, seems to have been envisaged although not necessarily implemented in practice (DTI 2003). There are also crucial links to funding, perhaps the most pervasive constraint for black enterprises, which affect both the ability to enter markets organically and grow and the ability to acquire interests in existing enterprises through BEE initiatives.

28.3.2 Deals and Funding Are Only Part of the Story By 2018, the value of BEE deals concluded in South Africa since 1994 amounted to more than Rand 600 billion (~US$ 50 billion, 2018) (Gqubule 2018). It has been a very expensive process, although it has also generated at least some value for its beneficiaries—​an Intellidex study released in 2015 found that the top 100 companies on the JSE alone had generated value for beneficiaries from BEE deals (net asset value after debt and other financial obligations were deducted) of approximately Rand 317 billion (Intellidex 2015). At the heart of the issue is that many black South Africans still do not have access to finance to support capital accumulation, starting at the level of a highly asymmetric distribution of household wealth (Chatterjee et al. 2020). Furthermore, proposed changes to BEE regulations are likely to foreclose a large proportion of potential BEE deal-​ making funding in finance and mining, which have underpinned large proportions of transformation spend and value since the 1990s (Gqubule 2018). These are also still strategic sectors in the context of the economy as a whole (Fine and Rustomjee 1996; Bell et al. 2018). Specifically, the ‘once empowered always empowered’ or continuing consequences principle (reinforced through DTI pronouncements and court judgments in 2018) means that firms are not required to top up their empowerment credentials after a black empowerment partner exits a partnership, and they retain their status (Bowman 2019; Gqubule 2018). This issue has been especially critical in mining in the context of heavily contested conditions of the mining sector charter—​the government had proposed that it would require companies to first raise their empowerment credentials again before applying for any new licences (Bowman 2019). Industrial financing of deals and firms in the BEE process has been extensive from the 1990s including in various NAIL transactions (DTI 2003; Gqubule 2006). The Industrial Development Corporation (IDC) has been the primary development finance institution (DFI) in this regard. However, the IDC has lagged in terms of its focus on financing entry of new black industrialists and riskier ventures linked with diversification of the economy (Goga, Bosiu, and Bell 2019). In other words, the bias seems to have been to

612   Thando Vilakazi support BEE deals and investments in existing sectors and incumbent enterprises (Goga et al. 2019). In many ways this is a function of the IDC’s mandate as South Africa’s largest DFI, and its funding constraints. Most relevant for this chapter is that the scale of funding required to continue to drive transformation through BEE is significant and requires a durable commitment from DFIs and no doubt private financiers. The challenge is that this cannot be guaranteed and depends also on conditions in global and domestic capital markets. It is precarious that the fate of black empowerment is almost solely dependent on the outcomes in volatile financial markets, although this is somewhat unavoidable, given the risks of the value of transactions being completely eroded, as we saw in the late 1990s with the stock market crash and more recently in the global financial crisis of 2007/​08. The above suggests that while BEE ownership deals are important because they can bring about access to decision-​making, wealth, and economic control fairly rapidly, they can only be part of the picture if racial transformation of the economy is to be durable. Longer-term dynamism and structural transformation is more likely to come from supporting rivals that can challenge incumbent players. As of 2020, there are no significant black-​owned businesses listed in the top tiers of the Johannesburg Stock Exchange (JSE), and certainly very few firms in the entire JSE that are black companies that have been started and grown by black South Africans since the 1990s (see Bosiu et al. 2017a). The exact level of black ownership of companies on the JSE is disputed, at around 3 per cent if one considers direct black ownership, the JSE estimates 23 per cent (which is not significant in any event, and certainly not proportional to racial distributions in the population) when accounting for foreign ownership; and institutional holdings such as pension funds, and there are others still (National Treasury 2017). Average black ownership has improved in some sectors (Vilakazi and Ponte 2020), but overall, across the economy (based on reporting companies only), black ownership was only 29 per cent in 2019 and some sectors like agriculture have black ownership as low as 12 per cent (BBBEE Commission 2019), which also corresponds with the severe inequalities in land ownership in agriculture mentioned above (RSA 2007). These are disturbing outcomes, which lend some support to the popular view that BEE has been a disaster in South Africa (see, for example, Saba 2018). Arguably, long-​term racial transformation, social stability, and economic dynamism will require both new rivals and greater participation in existing enterprises by black South Africans. It is important to note that BBBEE policies and scoring criteria actually do not only focus on ownership, contrary to the view in the public discourse and reporting, and a fair reading of the criteria shows that the four other elements do command a significant share of scores, although not all are considered priority elements. However, there has been a focus in practice and public discourse on ownership scores as the primary and most obvious barometer of transformation, and companies have used ownership changes as one of their main compliance strategies, which has meant that outcomes have generally been slow across other areas. The more significant issue here is that the incorporation of the black business elite into the commanding heights of the economy has, in the context of South Africa’s evolving political settlement, also served to align

Black Economic Empowerment in South Africa    613 interests of black and white elites and mitigated some of the potential tensions with the ruling elite and those black groups in society with significant holding power (which is not necessarily equivalent to electoral outcomes) that could seek more radical shifts in the distribution of economic power (MacDonald 2006; Ponte et al. 2007; Gqubule 2006; Andreasson 2006; Mondliwa and Roberts 2020; Vilakazi and Ponte 2020). In addition, a narrow focus on these scores in analysis has not been helpful and there are more interesting perspectives emerging on important related issues such as changing the structure of the economy, addressing market power and concentration, and the lack of investment in the economy even by so-​called transformed large firms (Bosiu et al. 2017b). Companies have leant on ownership transfers as a key strategy for achieving BEE score targets, most evident in the outcomes observed in concentrated sectors such as industrial fisheries (Vilakazi and Ponte 2020). The reality is that for many large firms it turned out to be less disruptive to business processes to diversify ownership of the business slightly to a carefully selected minority BEE partner (Sartorius and Botha 2008), than to have to completely change to procuring inputs from unfamiliar black-​ owned suppliers, or introduce untrusted or untested black managers or employees in key positions linked to the competitive and productive strategy of the business. Indeed, in the first decade of BEE the majority of high-​level appointments of black South Africans in established businesses were in relatively non-​core areas such as personnel management and transformation portfolios (Southall 2007). Management cultures and practices are still biased against black individuals (Bergh and Hoobler 2019), and it is likely that the same skewing is true for how funding is allocated, even by public institutions (Goga et al. 2019; Bosiu et al. 2020).

28.3.3 Procurement, Enterprise and Supplier Development Are Strategic Priorities Opening up access to markets through private and public-sector procurement and enterprise development and supplier development has been prioritized because of its potential to develop organic black-​owned businesses that build capabilities and an asset base over time. This lever is especially important in the context of the discussion about economic barriers above. The use of state procurement as a lever to empower black enterprises has not been successful (Crompton and Kaziboni 2020). Very few state-​owned enterprises (SOEs) comply and enforcement of rules about sourcing from black suppliers has not been effective at all (BBBEE Commissioner as quoted in Saba 2018; and, BBBEE Commission 2019). Black-​owned manufacturing firms have expressed the same concerns about the ineffectiveness and difficulty of accessing state procurement processes, and the lack of coordination between organs of state (Bosiu et al. 2020). This is problematic to the extent that it is hard to see when the issues will be addressed—​approximately 90

614   Thando Vilakazi per cent of state procurement is done by a handful of the largest SOEs such as Eskom (power utility), PRASA (passenger rail), PetroSA in petroleum products, and Transnet (various logistics and supply chain infrastructure) (PMG 2015)—​some of which are very troubled organizations, including significant links to corrupt activities. There have been significant challenges in terms of corruption, misappropriation, and inadequacy of policy tools employed, such as in procurement by Transnet in rail and under the Black Industrialists Policy which aims to develop black manufacturing firms (Crompton and Kaziboni 2020; Bosiu et al. 2020). It is striking that after around two decades of the implementation of BEE in its contemporary form, the BBBEE Commission has in 2019 still had to recommend what seems to be an obvious point that government itself needs to improve its implementation of empowerment priorities, stating that ‘organs of state and public entities must align their requirements and criteria for licencing, grants, incentives, public private partnership (PPP) and procurement to include broad-​based black economic empowerment compliance as it is mandatory under section 10 of the B-​ BBEE Act’ (BBBEE Commission 2019: 97). The parallel issue with respect to the private sector relates to enterprise and supplier development, which also incorporates measures of procuring from black-owned enterprises. There has been some progress. For example, BBBEE Commission data shows that in 2019 firms in all sectors of the economy, other than property, achieved scores of more than 50 per cent on average of the ESD target score in the scorecard (BBBEE Commission 2019). While some sectors such as construction (92 per cent score) lend themselves more to developing black suppliers, it is evident that raising scores across the board in this area should be an absolute priority for structural transformation of the economy, particularly where black businesses are involved in supplying intermediate inputs to manufacturing and high-​value-​added manufactured products. However, the problem has been that for certain value chains procurement from black businesses has only meant sourcing ancillary cleaning and transport services (Vilakazi and Ponte 2020), which are important in the context of job creation but less so in terms of building diverse capabilities in shifting to the production of high-​value-​ added manufactured goods—​a potential driver of long-​term growth. At issue is whether a broader form of inclusion can be fostered in which strong, medium-​sized firms can emerge, survive, and build capabilities over time to compete effectively with established players. Their ability to do this is critical as it can also mean progressively reduced dependence on various forms of state support, and potential for local sourcing and supplier development within their own supply chains to take place as the recent evidence on so-​called black industrialists is starting to show (Bosiu et al. 2020). In this regard, there is other recent evidence of a progressive shift by large firms away from short-​term corporate social investment initiatives with black suppliers, to more sustainable models of supplier development, which include elements of funding and opening up access to markets (das Nair and Landani 2020). The impacts of the progression are yet to be analysed in detail. Ultimately, driving inclusion of black-​owned enterprises is about recognizing that goals of equity can in some cases be achieved alongside market efficiency objectives

Black Economic Empowerment in South Africa    615 (Lewis 1995; Atkinson 2015). Furthermore, growth achieved without inclusion will over time raise social tensions which will render the growth path and political settlement unsustainable and unstable. In the prevailing policy context, this means that getting BEE policy and implementation right, along with corresponding policies to build the capabilities of black people and businesses, is critical.

28.3.4 Driving Transformation across All Spheres of Society Is Critical Although BEE was thought of as one part of the broader set of national policies to transform South African society, it has necessarily taken on far greater significance. This is because BEE speaks to the fundamental alignment of holding power and benefits in the political settlement of South Africa because it relates to who controls resources, access, and capital, and how that aligns to the political and other sources of power and influence of various groups in society (Khan 2010). The perceived and actual misalignment of holding power and benefits in the prevailing South African political settlement, which is not about a settlement ‘event’ such as the 1994 elections but a consideration about the wider balance of forces in the society over time (Khan 2018), has fuelled much of the contemporary discourse about radical economic transformation, state capture, and the continued marginalization of the black majority in society (Bhorat et al. 2017; Von Holdt 2019). The race (and gender) defined asymmetries and marginalization run deep and in all spheres of the society (Von Holdt 2019; Terreblanche 2002; MacDonald 2006), such that they require tremendous and sustained coordinated effort and an unbearably long time to disentangle and address. In this regard, it was appropriate that various policy documents in the 1990s set out very comprehensive objectives for racial transformation across diverse spheres of the society, including education and training, small business development, health-care and housing, labour laws, employment equity and empowerment of women and youth, and competition and economic regulation. All of these policies have direct and indirect links with economic outcomes, but few go straight to the heart of the issue of changing the ownership and control of capital in the manner that BEE does. However, they are critical insofar as they affect aggregate outcomes in the society that, in turn, shape medium-​and long-​term economic outcomes. The challenge is that outcomes in these different critical areas have been poor, although there has been substantial progress. For example, education and health-care system outcomes, as key determinants of long-​term growth, have by most accounts not been good since 1994 as discussed in Chapter 33 on the economics of education and ­Chapter 39 on health-care in this volume. South Africa has not been as successful as hoped in achieving equity and empowerment of previously marginalized groups in the workplace, or addressing wage and gender-​based income and wealth inequalities, whether defined on race or non-​race parameters as discussed above. The net effect is

616   Thando Vilakazi that in key areas, such as economic sector relevant training and education, for example, black South Africans including the youth still lag in terms of skills (see ­Chapter 31 on labour market trends and Chapter 32 on youth employment in this volume). The counterweight has been to encourage companies to contribute to skills development and training through increasing the weight of this element in the scorecard system, but this places the responsibility and burden on firms that are likely not to fill this gap comprehensively, if at all, not least because they may not be able to appropriate fully the benefits of training a particular employee who may progress and leave the company. In other economic policy areas, failures of coordination have meant that structural transformation of the economy has not occurred and there is no clear path for achieving racial transformation alongside industrial development and structural change (Bell et al. 2018). Strategies to increase the quantity and quality of employment are critical as well as part of a broader conception of black empowerment (Gqubule 2006)—​BEE policy in the formal definition adopted in 2003 seeks also to address income inequality and a key channel for doing so is through wage employment and ensuring related rights in terms of conditions of employment and labour protections. BEE policy initially sought to contribute to addressing these issues primarily through scores given for achieving (narrowly) specified goals on the racial composition of employment in different tiers of company structures (employment equity and skills development provisions), which were surprisingly de-​emphasized in the 2015 revisions to the codes in not being recognized as priority elements. However, a plausible rationale for this shift is that, other things equal, various aspects of labour empowerment are encompassed more comprehensively in other laws (DTI 2003) (also true of other areas such as state procurement), and it seems unlikely that the problem of exclusionary labour markets would be solved by making BEE policy the primary vehicle for this aspect of societal transformation. To illustrate the point above with a simple example, while education and sector skills development policies could drive empowerment of blacks through improving their skills in general terms, BEE was arguably meant to work in a complementary manner by ensuring that businesses absorbed these black individuals into their businesses and gave them opportunities including in management and more firm-​specific training. Presently, it seems there are obvious gaps in implementation and reporting on BEE, and various BBBEE Commission reports only deal with broad quantitative shifts without critical engagement with the reasons for specific outcomes (see BBBEE Commission 2017, 2019). For example, a key gap seems to be the fact that there is no clear way for identifying specific reasons why companies have performed so poorly on black management control, which encompasses employment equity and management involvement (average score 39 per cent in 2019, down from 45 per cent in 2018 (BBBEE Commission 2019)), whether this is because there is an underlying systemic deficiency in the quality and quantity of specific skills supplied by the education and training system, or other factors including institutional racism and norms, and how those can be addressed to improve the BEE scorecard outcomes. In other words, BEE policy should have a broad, cross-​cutting ambition in that it seeks to change multiple aspects of the South African

Black Economic Empowerment in South Africa    617 business environment more directly and comprehensively than most other economic policies; however, this is not to say that BEE policy should effectively become the de facto transformation policy in South Africa. Perhaps one of the reasons for the more skewed and narrower critiques of BEE in literature and public discourse is this problem—​that BEE policy as the proverbial ‘straw man’ is understood very broadly, often without due acknowledgment, that it was ostensibly only part of a wider nexus of policies that were also meant to drive transformation. To make the focus of BEE interventions to be about more than deracialization of existing structures, it is critical that its analytical, policy, and practical linkages to other areas such as labour, competition, health, and industrial policies are better understood. In this regard, policy coordination is critical but absent in key areas. For example, the links between various aspects of economic policy, let alone other spheres, are not well formed. Competition laws and industrial policies would be the natural companions for BEE policy, to the extent that these policies can help to address various barriers to entry such as funding, anticompetitive behaviour by incumbents, and other market access constraints. However, there is no clear link and coordination of strategies in each of these areas, despite a common understanding in the National Development Plan that high concentration and high barriers to entry are pervasive constraints to dynamism and participation in the economy (Vilakazi et al. 2020). It is perhaps this failure of coordination and implementation across different policy areas that has also led to the 2019 amendments of the competition laws, where there has been an efficient institution leading enforcement, to expand the remit of the regulators on issues of economic inclusion and participation, particularly of firms owned by historically disadvantaged South Africans.

28.4 Conclusion The chapter takes on the ambitious task of engaging with the nuances of policy formation, implementation, and outcomes with respect to black economic empowerment in South Africa. Part of the reason that the task is a challenging one is that much has been written and said about BEE, most of which concludes that there has been an outright policy failure. Indeed, the outcomes are not good by various measures. The chapter distinguishes itself by offering a slightly different, pragmatic view of the progress made, and not made, in the two decades since 2000, recognizing as some authors do that BEE is a policy imperative in South Africa for economic and socio-​political reasons, and it is not going away. As such, in the context of this volume, it more relevant to also cast an eye to key issues and pathways for advancing the policy agenda going forward. The chapter also adds to the debate the key question about the substance of inclusion and what meaningful participation could mean, suggesting ultimately that a deeper form of racial transformation of the economy, rather than simply leaving the door ajar for black people, is required and desirable.

618   Thando Vilakazi Understanding the ways in which the existing BEE framework can be built on to drive this form of transformation requires understanding that the forms of barriers that prevent meaningful participation by black South Africans are varied and significant (and need to become the focus of policies across the board). The chapter canvasses some of these barriers, why they may be more acute for black businesses, and why empowering black businesses through an emphasis on funding, procurement, and enterprise and supplier development is especially important because it creates muchneeded new black businesses and can help to reduce concentration in the economy. Compared to ownership transfers that do not significantly change the economic incentives and orientation of predominantly white-​owned businesses, this appears to be a more desirable form of transformation for achieving long-​term economic dynamism and growth. By focusing on these issues, the chapter accomplishes at least two things. First, through drawing together what we already know about the evolution and effectiveness of BEE, a critical, pragmatic perspective can be formed about the effectiveness of the policy in the context of other policies. Second, it points towards a new thinking about BEE, focused on enhancing the quality and not only the quantity of participation by integrating a nuanced understanding of the barriers that keep black businesses and people out of the economy. A broader, integrated policy approach is required that incorporates other policies and measures to facilitate entry and participation in the economy, such as competition and procurement interventions. We find that although BEE has taken South Africa in the right direction towards inclusion, not enough has been achieved with the policy in its contemporary form, and we suggest priorities for the pathway forward that include a focus on the nature and impact of high barriers to the participation of black businesses, the importance of funding as part of a broader set of interventions, the importance of parallel social and economic policies, and the strategic role that procurement and enterprise development can play in the long term. A narrow focus on BEE alone will not bring economic freedom.

References Andreasson, Stefan. 2006. ‘The African national congress and its critics:  “Predatory liberalism”, black empowerment and intra-​alliance tensions in post-​apartheid South Africa’, Democratization, 13(2): 303–​22. Atkinson, Anthony. 2015. Inequality:  What Can Be done? Cambridge, MA:  Harvard University Press. BBBEE Commission. 2019. ‘National status and trends on Broad-​Based Black Economic Empowerment’. Report [online] https://​www.bbbeecommission.co.za/​wp-​content/​ uploads/​2019/​06/​National-​Status-​Trends-​on-​Black-​Economic-​Empowerment-​Report-​31-​ March-​2019-​Approved-​FINAL.pdf.

Black Economic Empowerment in South Africa    619 Bell, Jason, Sumayya Goga, Pamela Mondliwa, and Simon Roberts. 2018. ‘Towards a smart, open economy for all’, New Agenda: South African Journal of Social and Economic Policy, 70: 6–​11. Bergh, Carin and Jenny M. Hoobler. 2019. ‘Implicit racial bias in South Africa:  How far have manager-​employee relations come in “The Rainbow Nation”?’, African Journal of Management, 4(4): 447–​68. Bhorat, Haroon, Mbongiseni Buthelezi, Ivor Chipkin, Sikhulekile Duma, Lumkile Mondi, Camaren Peter, Mzukisi Qobo, Mark Swilling, and Hannah Friedenstein. 2017. Betrayal of the Promise: How South Africa Is Being Stolen. Stellenbosch: State Capacity Research Project. Black Economic Empowerment Commission. 2001. ‘Report’. Johannesburg: Skotaville Press. [online] http://​www.kznhealth.gov.za/​TED/​commission.pdf. Bosiu, Teboho, Sumayya Goga, and Simon Roberts. 2017b. ‘Concentration, profits and investment:  Let’s focus on the structure of the economy, not “cash hoarding”’. Industrial Development Think Tank Policy Briefing Paper no. 1, CCRED, Johannesburg [online] https://​ static1.squarespace.com/​static/​52246331e4b0a46e5f1b8ce5/​t/​5a717fffec212dcf0f02bea0/​ 1517387779116/​Policy+Brief+1.pdf. Bosiu, Teboho, Nicholas Nhundu, Anthea Paelo, Mmamoletji Thosago, and Thando Vilakazi. 2017a. ‘Growth and strategies of large and lead firms: Top 50 JSE firms’. CCRED Working Paper no. 2017/​17, Johannesburg. Bosiu, Teboho, Grace Nsomba, and Thando Vilakazi. 2020. ‘South Africa’s black industrialists scheme:  Evaluating programme design, performance and outcomes’. CCRED Working Paper no.  1/​ 2020, Johannesburg [online] https://​static1.squarespace.com/​static/​ 52246331e4b0a46e5f1b8ce5/​t/​5eaa7589b0caa85f5c4d4234/​1588229526930/​CCRED+Report_​ BIS+Economic+Evaluation_​CCRED+Working+Paper+1-​2020.pdf. Bowman, Andrew. 2019. ‘Black economic empowerment policy and state-​business relations in South Africa: The case of mining’, Review of African Political Economy, 46(160): 223–​45. Bracking, Sarah. 2019. ‘Black economic empowerment policy in Durban, eThekwini, South Africa:  Economic justice, economic fraud and “leaving money on the table”, Review of African Political Economy, 46(161): 415–​41. Broad-​Based Black Economic Empowerment Act, No. 53 of 2003 (BEE Act 2003) [online] https://​www.gov.za/​sites/​default/​files/​gcis_​document/​201409/​a53-​030.pdf. Broad-​Based Black Economic Empowerment Amendment Act, No. 46 of 2013 (BEE Act 2013) [online] https://​www.gov.za/​sites/​default/​files/​gcis_​document/​201409/​37271act46of2013. pdf. Broad-​ Based Black Economic Empowerment Commission (BBBEE Commission). 2017. ‘National state of transformation and trend analysis’. National Report [online] https://​ bbbeecommission.co.za/​wp-​content/​uploads/​2017/​05/​B-​BBEE-​C ommission-​National-​ State-​of-​Transformation-​and-​Trend-​Analysis-​National-​Report-​Booklet.pdf. Chabane, Neo, Andrea Goldstein, and Simon Roberts. 2006. ‘The changing face and strategies of big business in South Africa: More than a decade of political democracy’, Industrial and Corporate Change, 15(3): 549–​77. Chatterjee, Aroop, Léo Czajka, and Amory Gethin. 2020. ‘Estimating the distribution of household wealth in South Africa’. Southern Centre for Inequality Studies Working Paper no. 3 [online] https://​www.wits.ac.za/​scis/​publications/​working-​papers/​. Chisoro-​Dube, Shingie, and Reena das Nair. 2020. ‘Confronting entry barriers in South Africa’s grocery retail sector’, in Thando Vilakazi, Sumayya Goga, and Simon Roberts (eds) Opening the South African Economy: Barriers to Entry and Competition. Cape Town: HSRC Press.

620   Thando Vilakazi Competition Commission of South Africa (CCSA). 2019. Final Report: Grocery Retail Market Inquiry. CCSA, Pretoria. Crompton, Rod, and Lauralyn Kaziboni. 2020. ‘Lost opportunities? Barriers to entry and Transnet’s procurement of 1,064 locomotives’, in Thando Vilakazi, Sumayya Goga, and Simon Roberts (eds) Opening the South African Economy: Barriers to Entry and Competition. Cape Town: HSRC Press. Das Nair, Reena, and Namhla Landani. 2020. ‘New approaches to supermarket supplier development programmes in Southern Africa’, Development Southern Africa, doi:  10.1080/​ 0376835X.2020.1780565. Davis, Rebecca. 2020. ‘Moeletsi Mbeki:  Big business, not the ANC, created BEE’, Daily Maverick, 28 August [online] https://​www.dailymaverick.co.za/​article/​2020-​08-​28-​ moeletsi-​mbeki-​big-​business-​not-​the-​anc-​created-​bee/​. Deloitte. 2018. ‘Unpacking the 2018 draft mining charter: Commentary and key insight’, [online] https://​www2.deloitte.com/​content/​dam/​Deloitte/​za/​Documents/​energy-​resources/​ Deloitte_​2018-​Draft-​Mining-​Charter_​070818.pdf. DTI (Department of Trade and Industry). 2003. ‘South Africa’s economic transformation: A strategy for Broad-​Based Black Economic Empowerment’. [online] http://​www.thedtic.gov. za/​wp-​content/​uploads/​bee-​strategy.pdf. DTI (Department of Trade and Industry). 2014. ‘Trade and industry on proclamation of Broad-​Based Black Economic Empowerment Amendment Act’, 26 October [online] https://​ www.gov.za/​proclamation-​broad-​based-​black-​economic-​empowerment-​amendment-​act-​ no-​46-​2013. Fine, Ben, and Zavareh Rustomjee. 1996. South Africa’s Political Economy:  From Minerals-​ Energy Complex to Industrialisation. London: C. Hurst & Co. Publishers. Freund, Bill. 2007. ‘South Africa: The end of apartheid and the emergence of the “BEE Elite”’, Review of African Political Economy, 34(114): 661–​78. Goga, Sumayya, Teboho Bosiu, and Jason Bell. 2019. ‘Linking IDC finance to structural transformation and inclusivity in post-​apartheid South Africa’, Development Southern Africa, 36(6): 821–​38. Gqubule, Duma. 2006. Making Mistakes, Righting Wrongs:  Insights into Black Economic Empowerment. Johannesburg: Jonathan Ball Publishers. Gqubule, Duma. 2018. ‘New sectoral charters threaten transformation and empowerment’, BEE News, 4 December [online] https://​www.bee.co.za/​post/​2018/​12/​04/​opinion-​duma-​ gqubule-​new-​sectoral-​charters-​threaten-​transformation-​and-​empowerment. Hirsch, Alan. 2005. Season of Hope:  Economic Reform under Mandela and Mbeki. Scotsville: University of KwaZulu-​Natal Press. Intellidex. 2015. ‘The value of BEE deals’. Research Report, June [online] https://​www. intellidex.co.za/​wp-​content/​uploads/​2015/​06/​Intellidex-​report-​The-​Value-​of-​BEE-​Deals. pdf. Khan, Mushtaq. 2001. ‘The new political economy of corruption’, in Ben Fine, C. Lapavitsas, and J. Pincus (eds) Development Policy in the Twenty-​first Century:  Beyond the Post-​ Washington Consensus. London: Routledge. Khan, Mushtaq. 2010. ‘Political settlements and the governance of growth-​ enhancing institutions’. [online] https://​eprints.soas.ac.uk/​9968/​1/​Political_​Settlements_​internet.pdf. Khan, Mushtaq. 2018. ‘Power, pacts and political settlements: A reply to Tim Kelsall’, African Affairs, 117(469): 670–​94.

Black Economic Empowerment in South Africa    621 Lee, Hwok-​Aun. 2015. ‘Affirmative action in Malaysia and South Africa: Contrasting structures, continuing pursuits’, Journal of Asian and African Studies, 50(5): 615–​34. Lewis, David. 1995. ‘Markets, ownership and manufacturing performance’, in Avril Joffe, David Kaplan, Raphael Kaplinsky, and David Lewis (eds) Improving Manufacturing Performance in South Africa: Report of the Industrial Strategy Project. Cape Town: UCT Press. MacDonald, Michael. 2006. Why Race Matters in South Africa. Scotsville:  University of KwaZulu-​Natal  Press. Mondliwa, Pamela, and Simon Roberts. 2020. ‘Black economic empowerment and barriers to entry’, in Thando Vilakazi, Sumayya Goga, and Simon Roberts (eds) Opening the South African Economy: Barriers to Entry and Competition. Cape Town: HSRC Press. National Treasury of the Republic of South Africa. 2017. ‘Ownership of JSE-​listed companies’. Research Report [online] http://​www.treasury.gov.za/​comm_​media/​press/​2017/​ 2017100301%20Ownership%20monitor%20-​%20Sept%202017.pdf. Orthofer, Anna. 2016. ‘Wealth inequality in South Africa: Insights from survey and tax data’. REDI3X3 Working Paper no.  15, University of Cape Town [online] https://​www.redi3x3. org/​sites/​default/​files/​Orthofer%202016%20REDI3x3%20Working%20Paper%2015%20-​ %20Wealth%20inequality.pdf. Patel, Leila, and Lauren Graham. 2012. ‘How broad-​based is broad-​based black economic empowerment?’, Development Southern Africa, 29(2): 193–​207. PMG (Parliamentary Monitoring Group). 2015. ‘Broad-​Based Black Economic Empowerment Codes of Good Practice’. Department of Trade and Industry briefing, 9 June [online] https://​ pmg.org.za/​committee-​meeting/​21049/​. Republic of South Africa. 2013. Government Gazette Notice No. 36928, 11 October [online] https:// ​ b bbeecommission.co.za/ ​ w p- ​ c ontent/ ​ uploads/ ​ 2 017/ ​ 1 2/ ​ P hase- ​ 1 - ​ 3 6928_ ​ 1 1- ​ 1 0_​ TradeIndCV01_​3a.pdf. RSA (Republic of South Africa). 2007. ‘Land Audit Report—​Phase II: Private land ownership by race, gender and nationality’. [online] https://​www.gov.za/​sites/​default/​files/​gcis_​document/​201802/​landauditreport13feb2018.pdf. Saba, Athandiwe. 2018. ‘Has BEE been a dismal failure?’, Mail & Guardian, 31 August [online] https://​mg.co.za/​article/​2018-​08-​31-​00-​has-​bee-​been-​a-​dismal-​failure/​. Sartorius, Kurt, and Gerhard Botha. 2008. ‘Black economic empowerment ownership initiatives:  A Johannesburg Stock Exchange perspective’, Development Southern Africa, 25(4): 437–​53. Southall, Roger. 2007. ‘Ten propositions about black economic empowerment in South Africa’, Review of African Political Economy, 34(111): 67–​84. Terreblanche, Sampie. 2002. A History of Inequality in South Africa, 1652–​ 2002. Scotsville: University of KwaZulu-​Natal Press. Vilakazi, Thando, and Stefano Ponte. 2020. ‘The political economy of competition, regulation and transformation: Black Economic Empowerment (BEE) and quota allocations in South African industrial fisheries’. Joint CCRED and CBDS Working Paper [online] https://​ static1.squarespace.com/​static/​52246331e4b0a46e5f1b8ce5/​t/​5f5b78afd3a08d138e1c2899/​ 1599830240744/​3f70d0_​775a7443c7834c76885b378179b21ad2.pdf. Vilakazi, Thando, Sumayya Goga, and Simon Roberts (eds). 2020. Opening the South African Economy: Barriers to Entry and Competition. Cape Town: HSRC Press. Von Holdt, Karl. 2019. ‘The political economy of corruption:  Elite-​ formation, factions and violence’. Working Paper no.  10, Society, Work and Politics Institute, University of Witwatersrand, Johannesburg.

Chapter 29

E ntrepreneursh i p a nd SMMEs in Sou t h   A fri c a Boris Urban

29.1 Introduction Entrepreneurship and small, medium, and micro enterprises (SMMEs) have been recognized as imperatives for national economic development globally and in South Africa, which can contribute to macroeconomic growth, poverty alleviation, and employment creation (ILO 2019; Stam and Bosma 2015; Venter and Urban 2015). The central argument guiding this chapter is that since 1994, although the South African government has consistently and publicly supported SMMEs and repeatedly categorized small businesses as key to unlocking inclusive growth of the economy, in reality, this sector does not match the rhetoric. Currently in South Africa, entrepreneurs and SMMEs face a multitude of obstacles and the country’s established-​business ownership rate is considerably below the average for the overall African region and for other developing countries (Bosma et al. 2020; Urban 2020). In addressing the main argument, this chapter starts in section 29.2 with a critical overview of the complexity of describing entrepreneurship and SMMEs due to the heterogeneity of entrepreneurship and SMMEs in South Africa, which range from informal, survivalist enterprises to formal, high-​growth enterprises. Reflecting on the growing scholarship on entrepreneurship as a field of study, it is observed that the discourse is moving away from considering entrepreneurs as a relatively homogeneous assemblage of actors to a set of typologies. This section draws on a diverse literature for conceptual approaches and international comparisons, while maintaining the focus on South Africa. While recognizing shortcomings related to sourcing reliable and representative SMME data in South Africa, different statistics from multiple sources are analysed in section 29.3. These figures indicate this sector is dominated by mostly micro and small informal enterprises, which do not contribute significantly to the GDP, nor generate

Entrepreneurship and SMMEs in South Africa    623 much productive employment. Historically, the small business sector in South Africa has been marginalized and, as a result, there is a prominent political focus on micro enterprises in the informal sector, and on the other hand, large firms dominate the industrial space, while the ‘missing middle’ typified by small and medium enterprises is largely neglected (Bhorat et al. 2018; Urban 2020). Section 29.4 describes and critically evaluates the entrepreneurial ecosystem in South Africa, where several major factors inhibiting the start-​up and growth of SMMEs are analysed. It is clear from the analyses that a number of structural, institutional, and policy factors obstruct SMME growth and development in South Africa. This section ends by emphasizing that an effective entrepreneurial ecosystem is best arranged as a set of interdependent actors and factors, coordinated in such a way that they enable productive entrepreneurship. Section 29.5 is focused on policy critique and future considerations. Notwithstanding the limited efficacy of entrepreneurship policy, governments around the world continue to introduce entrepreneurship and SMME policy. In South Africa, policymakers fail to recognize that while state-​sponsored entrepreneurship interventions in some instances may have positive results, there are ‘opportunity costs’ associated with such policies and may have the unintended consequence of squandering resources and fostering non-​ productive entrepreneurship. The chapter ends with a concluding section where, in the final analysis, it is ascertained that SMMEs in South Africa, as a result of a mix of structural and policy caveats, do not live up to their potential as the key to unlocking growth of the economy. While this chapter focuses on economic aspects, the content is based on a more holistic perspective of entrepreneurship, which draws on literature from other disciplines through diverse analytical perspectives; consequently, publications in a wide range of journals have been cited. Moreover, although this chapter can be read and cited as a stand-​alone, authoritative reference on entrepreneurship and SMMEs in the South African economy, many of the themes and analytical perspectives in this chapter are directly related to other chapters in this handbook. In particular, these include the chapters on the informal sector in South Africa by Michael Rogan and Caroline Skinner (­Chapter 35) as well as Erika Kraemer-​Mbula and Rasigan Maharajh’s chapter on innovation and technological change in South Africa (­Chapter 22).

29.2  An Overview of Entrepreneurship and SMMEs During the apartheid era, black South Africans were largely prohibited from owning property, preventing them from using their property as a form of collateral, which had a negative effect on their ability to start businesses. During this era, not only was there an absence of small businesses in the dominant sectors of the economy, but also very little

624   Boris Urban attention was paid to small enterprise promotion in public policy. Since the dawn of democracy in 1994, South Africa has faced numerous economic, political, and social challenges, particularly in terms of extensive unemployment and inequality (SEDA 2020; Urban 2015). Responding to these challenges, the government has prioritized small business and entrepreneurial development (Rogerson 2004). South Africa’s policy framework for entrepreneurship, in the context of SMMEs, originated in its White Paper on the National Strategy for the Development and Promotion of Small Business (RSA 1996). At the same time, entrepreneurship development is part of a sizeable industry in South Africa, which has witnessed a proliferation of incubators, innovation and tech hubs, accelerators, and start-​up boot camps, most of which seem to have adopted a US Silicon-​Valley-​business-type model. These entities provide a somewhat clichéd approach to start-​ups, characterized by sleek websites and extensive public relations endeavours, but offer little by way of demonstrating what measurable impact they are having on increasing overall entrepreneurial activity and employment rates in South Africa (Urban 2020).

29.2.1 Conceptualizing Entrepreneurship and SMMEs Notwithstanding the conceivable significance of entrepreneurship for economic development, definitional controversies persist. Indeed, there is no shortage of definitions of entrepreneurship and an explanation for such assortment of definitions is that researchers tend to define entrepreneurs based on the premises of their own background or disciplines (Bygrave 1989). Economists tend to accentuate the classic models of economic behaviour and innovation, while management experts emphasize the resourcefulness and organizing competencies of entrepreneurs (Shane and Venkataraman 2000; Venter and Urban 2015). Schumpeter (1934) arguably launched the field of entrepreneurship, not only by associating entrepreneurs with innovation, but also by demonstrating the importance of entrepreneurs in ‘creative destruction’ and hence economic development. Kirzner’s (1997: 62) definition is focused on the entrepreneur as someone who ‘facilitates adjustment to change by spotting opportunities for profitable arbitrage’, while Knight (1921) identified the entrepreneur as someone who deals with the uncertainty involved in the exploitation of opportunities. More recently, Shane and Venkataraman (2000:  220) suggest research on entrepreneurship focuses on ‘the central question of the entrepreneur in terms of why, when and how some people and not others discover and exploit opportunities’. The complexity of defining entrepreneurship is further due to the variety of contexts in which it occurs, as distinguished by: (1) level (e.g. individual, team, firm, community); (2) type of organization (family, small business, corporate, franchise); or (3) geographic location (Kloepfer and Castrogiovanni 2018). Furthermore, a standard, internationally recognized definition of SMMEs, or micro​, small​, and medium-​sized

Entrepreneurship and SMMEs in South Africa    625 enterprises (MSMEs), or small and medium-​ sized enterprises (SMEs) does not exist. The variations in classifications differ due to specific country legislation, where countries use different threshold conventions, and in particular, because the enterprise dimension ‘small’ and ‘medium’ are often qualified in accordance with the size of the domestic economy (OECD 2009). According to the ‘revised definition for SMMEs in South Africa, as contained in the amended Schedule 1 in terms of the National Small Enterprise Act of 1996, the size or class of enterprise/​enterprise category is defined using two proxies: total full-​time equivalent (FTE) of paid employees, and total annual turnover’ (RSA 2019: 111). The proxies used to define the size depend on the sector or sub​sector within which the enterprise falls. For instance, the classification of SMMEs for the manufacturing sector are:  (1) medium (51–​250 full-​time employees (FTE); ≤ 210 million); (2) small (11–​50 FTE; ≤ 50 million); and (3) micro (0–​10 FTE; ≤ 15 million) (RSA 2019). While the prescribed SMME definitions with strict size classifications are useful as they are simple to grasp and facilitate statistical analyses, such generalized measures do have several limitations. In many instances, these SMME classifications tend to conceal the real nature of entrepreneurship, where the number of individuals involved is highly dependent on the sector of the business. For instance, while ten people may constitute a very small manufacturing enterprise the same number is a sizeable number of individuals for a medium-​sized specialized consultancy practice (Venter and Urban 2015). Self-​employment, business ownership rates, or new venture creation, as defined by the series of Global Entrepreneurship Monitor (GEM) reports, suffer from the same problems as they typically refer to the level and/​or the dynamics of entrepreneurship and identify the percentage of the working-​age population that is engaged or willing to engage in entrepreneurial activity (Bosma et al. 2020). This is a shortcoming, since ‘with emphasis on national representativeness, the GEM consortium has had to trade sample size for the breadth and depth of empirical measures’ (Bergmann et al. 2014: 251).

29.2.2 Different Typologies of Entrepreneurial Activity The central argument driving this chapter is elucidated with the recognition that entrepreneurship is a complex phenomenon and judging from some of the fundamental restrictions in defining entrepreneurship and SMMEs, it seems that merely to seek out and focus on some concept of the ‘average’ entrepreneur, or the ‘typical’ enterprise is ineffectual (Bygrave 1989). Researchers show that entrepreneurship, as a field of study, is moving away from considering individual entrepreneurs as a relatively homogeneous set of actors to a set of typologies (Kloepfer and Castrogiovanni 2018). Recognizing that entrepreneurship and SMMEs are difficult to define precisely and there seem to be fundamental differences in theory and practice related to these terms, several juxtapositions are briefly analysed to emphasize the heterogeneous nature of entrepreneurship:

626   Boris Urban Productive vs. unproductive entrepreneurship: Baumol (1990) illustrates, by using historical evidence, that entrepreneurs respond to the incentive structures underpinning an economy. Institutional factors provide incentives for rent-​seeking entrepreneurial activities (such as crime and corruption) versus socially productive entrepreneurial activities (such as the establishment of new enterprises). In this regard, entrepreneurial activity is viewed according to the outcomes it delivers and is based on the composition of incentives in a specific environment. High-​ growth entrepreneurship vs. replicative entrepreneurship: High-growth ventures, which are only a tiny subset of all ventures, contribute disproportionately to innovation, while replicative entrepreneurs produce or sell goods and services already present in the marketplace (Venter and Urban 2015). The vast majority of all entrepreneurs in developing markets appear to be of the replicative variety, where the Schumpeterian view that entrepreneurship is a major engine of economic growth has not been proven empirically for developing countries (Naudé 2008). Local vs. systemic entrepreneurship:  Local entrepreneurship is ‘socially productive entrepreneurial activity that is limited to a small number of market transactions, does not entail a complex division of labour, nor involve a deep accumulation of capital, and primarily rests on personal and informal relations’. On the other hand, the term ‘systemic entrepreneurship refers to socially productive entrepreneurial activities that go beyond the local level and takes place through the establishment of organizational structures that enable the exploitation of opportunities beyond the initial local level’ (Sautet 2013: 392–​3). Whereas the distinction between local and systemic is not clear-​cut, it is important to recognize that in some places in the world, the possibility for local entrepreneurship to become systemic does not exist due to a lack of any real opportunities in the environment (Kujinga and Urban 2017). Opportunity-​ motivated vs. necessity-​ driven entrepreneurs: Since its inception, the GEM has distinguished between opportunity and necessity as primary motivations for entrepreneurial activity (Bosma et al. 2020; Bowmaker-​Falconer and Herrington 2020). However, it is widely acknowledged that this dichotomy may not fully reflect the nuances in motivations for founding a business (Urban 2015; Williams and Kedir 2018). Consequently, the GEM has now changed how they measure motives for starting a business. The latest GEM results indicated that there is substantial variation in motivations across economies, sometimes between neighbours, and some commonalities between vastly different economies (Bosma et al. 2020).

29.2.3 Alternative Perspectives on Entrepreneurship Social entrepreneurship:  Several researchers and governments are looking at entrepreneurship having not only an economic purpose or Schumpeterian rationale, in terms of entrepreneurs driving innovation and stimulating structural changes in an economy,

Entrepreneurship and SMMEs in South Africa    627 but also increasingly recognizing a social component to entrepreneurship (Bosma and Levie 2009; Kujinga and Urban 2017; Nicholls 2006). Social entrepreneurship has been proposed as a way for developing new, sustainable models of social-​sector value and systemic impact, based on satisfying local needs, which address a range of social issues in innovative and creative ways (Nicholls 2006). Social entrepreneurship reflects diverse initiatives and activities that might manifest through philanthropic efforts, non-​profit organizations (NPOs), non-​governmental organizations (NGOs), and organizational social initiatives such as corporate social responsibility (CSR) projects. Social entrepreneurship has direct relevance to South Africa in terms of its transformative role in reducing severe inequalities (Venter and Urban 2015). A growing number of studies indicate that the social enterprise sector across sub-​Saharan Africa is still highly influenced by local, international, and bilateral political and economic decisions that often challenge the success of social entrepreneurship initiatives (Kujinga and Urban 2017). While the social enterprise sector in South Africa is difficult to estimate, a survey of social enterprises in South Africa indicates that the majority of social enterprises do not depend on grants and donations, and are generally small, with only 12 per cent generating an income of over R1 million. The vast majority employ between one and 50 people, and serve fewer than a hundred beneficiaries a month (Myers et al. 2017). A  more established measure of social entrepreneurship activity (SEA) (Bosma and Levie 2009) shows that early-​stage social enterprise activity in South Africa is about the same as the average rate reported across all forty-​nine countries. Despite the proliferation of social entrepreneurship in South Africa, the survival of many social enterprises and NGOs remains questionable and the legislation and policies created to empower previously disadvantaged groups through social enterprises has at best achieved mixed results (Venter and Urban 2015). Sustainable or green entrepreneurship:  The early 1990s saw an increasing focus on the market opportunities offered by the sustainability agenda, predominantly from an environmental perspective, which is often termed ‘green entrepreneurship or ecopreneuring’ (Marks and Hidden 2017). There is a wide diversity of initiatives that may be considered instances of sustainability entrepreneurship. For instance, Honey Care Africa, while focused on providing jobs and economic opportunities, has created a new model of bee farming by emphasizing the environmental benefits of bee farming, with bees providing vital pollination services to surrounding ecosystems (Urban 2015; Venter and Urban 2015). Research on sustainable practices in South Africa is only emerging but shows that many SMMEs do not as yet see sustainable practices as a competitive advantage, except in the instances where valued clients or customers impose it upon them (Marks and Hidden 2017). Based on the changing trends facing the world today, there seems to be a noticeable change regarding how economic and business activity is becoming more deeply rooted in social and environmental frameworks, and where so-​called impact investors are seeking to ‘invest for a purpose’ and look for some form of social return (Nicholls 2006; Urban 2015).

628   Boris Urban

29.3  The Entrepreneurship and SMME Landscape in South Africa Several reports are generated every year, which highlight a number of wide-​ranging factors facing SMMEs, and generally focus on the macro-​and micro-​ economic environments. Recognizing this widely available descriptive work, only select descriptives are presented and analysed in this section. Furthermore, it must be acknowledged that there is restricted established descriptive research providing reliable and valid information and trends on SMMEs in South Africa (Bhorat et al. 2018; SEDA 2020). As a result of such deficiencies related to sourcing reliable and representative SMME data in South Africa, the statistics in this section instead reflect multiple sources that are often contradictory. Moreover, it must be understood that each data source or study tends to use different approaches to classify SMMEs in South Africa, which, when recognizing the inherent variation in entrepreneurial activity as discussed in the preceding section, makes comparability across datasets problematic.

29.3.1 Number of SMMEs in South Africa According to a baseline study conducted by the Small Business Institute (SBI), the number of formal SMMEs in South Africa was approximately 260,000 firms in 2016. In 2016, there were 176,333 firms defined as micro (66 per cent of the total), 68,494 small firms (26 per cent of the total), and 17,397 medium ones (6.5 per cent of the total) (SBI 2018). Contrastingly, the ‘Annual Review of Small Businesses and Cooperatives report (DSBD 2017) indicated the total number of people that identify as self-​employed in 2016 (regardless of the business registration status) was approximately 2.2 million’ (SEDA 2020: 32). An annual report published by the Small Enterprise Development Agency specified that the ‘total number of SMMEs amounted to approximately 2.6 million in the third quarter of the 2018 financial year, with formal SMMEs making up approximately 29 per cent of all SMMEs. The data further indicated that the number of SMMEs has increased by 13.6 per cent year on year for the 2017 to 2018 period’ (SEDA 2020: 33). The SMME Quarterly Update, Third Quarter 2019, reported that there were 2,653,424 SMMEs, of which 779,297 were formal SMMEs and 1,791,431 informal SMMEs, together providing 11,592,677 jobs (SEDA 2019; StatsSA 2019). In the year up to the ‘third quarter of 2019, the number of SMMEs grew by around 97,000 against a backdrop where the available jobs declined by around 100,000’ (SEDA 2019: 15).

29.3.2 Formal vs. Informal SMMEs In South Africa, the ‘share of SMMEs operating in the informal sector stood at 68 per cent in the third quarter of 2019, with the share operating in the formal sector

Entrepreneurship and SMMEs in South Africa    629 at 29 per cent. These ratios have been very stable since 2010’ (SEDA 2019: 16). It has been noted that SMMEs in South Africa tend to possess qualities of survivalist firms, especially those in the smaller size categories. The ‘largest group of SMME owners are own-​account workers (businesses with no employees), constituting about 63 per cent of the self-​employed’ (Bohart et al. 2018: 11). Moreover, own-​account workers’ ‘characteristics indicate that SMMEs, as opposed to large firms, present more self-​ employment opportunities for workers with fewer labour market opportunities, such as females, young workers, Africans, and less educated workers. Together, the high proportion of young business owners (less experienced), low levels of education, and low median wages for SMME owners indicate that SMME owners have relatively low skills levels’ (Bhorat et al. 2018: 11). Consequently, the informal sector is a relatively small source of secondary employment in South Africa, which is often an overlooked fact in terms of the South African SMME landscape (Bhorat et al. 2018; ILO 2019). ‘South Africa has a relatively small rate of informality, with only 29 percent of individuals in informal employment. This is less than half the average informality rate for sub-​Saharan Africa, and well below the informality rate of many developing regions’ (Bhorat et  al. 2018:  17). Reflecting on the heterogeneous nature of the informal sector in South Africa, it seems for some, informal entrepreneurs are seen as survivalists, pushed into this realm by their inability to find formal employment, and view informal entrepreneurship as unregulated, low paid, and an insecure kind of survival-​driven self-​employment. On the other hand, others seem to regard informal entrepreneurship as a sensible economic approach pursued by entrepreneurs whose outlook is suppressed by state-​imposed institutional restrictions, and view informal entrepreneurship as the road to advancement (Venter and Urban 2015; Williams and Kedir 2018).

29.3.3 Demographics Factors of SMME Owners This sub s​ ection considers two key characteristics of any given population that may have a significant influence on the level of entrepreneurial activity, namely age and gender (Bosma et al. 2020).

29.3.3.1 Age Different age groups have different levels of entrepreneurial activity, and according to the GEM report (Bosma et al. 2020), economies in all geographic regions show bell-​ shaped age distributions with the highest entrepreneurship rates generally occurring among 25-​to 34-​year-​olds. Plausible reasons for these age groups being so prevalent in entrepreneurial activity is that individuals in early-​to mid-​career trajectories ‘have had time to develop their knowledge and skills through education as well as through work experience’ (Venter and Urban 2015: 234). In South Africa, ‘entrepreneurial activity has almost doubled from 7.5 per cent in 2017 to 14.3 per cent in 2019 in the age group 45 to 54 years, but has decreased in the age bracket 35 to 44 years’ (Bowmaker-​Falconer and Herrington 2020: 15).

630   Boris Urban

29.3.3.2 Gender According to the GEM report, the majority of countries continued to have higher male levels of entrepreneurial activity, than female levels in 2019 (Bosma et al. 2020). In South Africa, the ratio of male to female entrepreneurial activity has changed from 1.52 (12.5 male:  8.2 female entrepreneurs) in 2017, to 1.14 (10.9 male:  9.6 female entrepreneurs) in 2019, indicating that female entrepreneurship is on the rise (Bowmaker-​Falconer and Herrington 2020). However, the majority of female entrepreneurs in South Africa operates within the crafts, hawking, personal services, and retail sectors, where low technology is utilized in these undifferentiated businesses and they tend to have relatively lower revenue and employ less staff, than those owned by men (SEDA 2019). Several studies confirm that female entrepreneurs are confronted with greater obstacles than men are when starting small enterprises. Factors influencing such discrepancies include discrimination and bias, lower educational and business experience levels, limited capital and assets, fewer business-​oriented networks and support mechanisms, and differentials in societal norms and gender expectations (Bosma et al. 2020; Venter and Urban 2015).

29.3.4 SMMEs Sector and their Geographic Location South African SMMEs ‘trade predominantly in the wholesale and retail sector, which represents almost half (46.1 per cent) of all early stage entrepreneurship activity (down from 50.4 per cent in 2015), with the manufacturing sector growing significantly since 2015 (3.6 per cent to 13.1 per cent)’ (Bowmaker-​Falconer and Herrington 2020: 8). The overtrading in the wholesale and retail sector by SMMEs is often attributed to significant barriers to entry in terms of skills and capital as well as the price competitive nature of these sectors, which increases the likelihood of business failures. This situation is further exacerbated due to the oligopolistic structure of the retailer sector where dominant retailers with high buying power purchase at highly competitive rates, creating barriers, and eroding any competitive advantage for SMMEs (Bhorat et  al. 2018; SEDA 2020). Nearly a third of ‘SMMEs operated in Gauteng in the third quarter of 2019, followed by close to 16 per cent in KwaZulu-​Natal and 13 per cent in the Western Cape. Growth in the Western Cape was a staggering 37 per cent over this period, with the province gaining more than 96,000 SMMEs. This gain coincides with an even greater number of job losses (109,000 jobs) in the formal sector of the Western Cape economy (StatsSA 2019). Gauteng experienced the largest decline in number of SMMEs (close to 24,000; 2.7 per cent decrease), with close to 29,000 formal sector jobs being created. The Northern Cape suffered the greatest proportional decline (30 per cent), probably as the severe drought there forced many farmers and agriculturally linked SMMEs out of business’ (SEDA 2019: 21).

29.3.5 Different Types of Entrepreneurial Activity A key GEM indicator is the early-​stage TEA rate in a country, which measures the number of individuals who are considered ‘nascent entrepreneurs (individuals who

Entrepreneurship and SMMEs in South Africa    631 Table 29.1: Types of entrepreneurial activity, by region and select country as percentage of adult population New Nascent business entrepreneurship ownership rate rate

Early-​stage total entrepreneurial activity (TEA) rate*

Established business ownership rate

Region

Country

Africa

Egypt

5.0

1.8

6.7

1.5

Madagascar

8.4

11.4

19.5

20.2

Morocco

7.3

4.4

11.4

7.9

South Africa

7.3

3.7

10.8

3.5

Asia and Oceania

China

5.3

3.6

8.7

9.3

India

9.4

5.9

15.0

11.9

South America

Brazil

8.1

15.8

23.3

16.2

Europe

Russia

4.6

4.8

9.3

5.1

United Kingdom

6.5

3.1

9.3

8.2

North America South Africa’s rank /​50 countries

United States

11.8 21/​50

5.9 31/​50

17.4

10.6

25/​50

44/​50

Source: Adapted from Bosma et al. (2020) and Bowmaker-​Falconer and Herrington (2020). Note: *Read as ‘Early-​stage entrepreneurial activity rate (TEA rate) was 10.8 per cent of South Africa adult population in 2019’.

have committed resources to starting a business but have not yet paid salaries or wages for more than three months)’ (Bosma et  al. 2020:  26). TEA also measures the ‘new business ownership rate (owners of a new business that is less than 42 months old)’ (Bosma et al. 2020: 26). South Africa has relatively low scores across the GEM measures of entrepreneurial activity when compared to other countries. Table 29.1 shows the scores for these different types of entrepreneurial activity measures by different regions where select countries were nominated to show how South Africa compares to other African countries (Egypt, Madagascar, Morocco), BRIC nations (Brazil, Russia, India, China), and developed countries (the United Kingdom, the United States). In the African region, which includes a limited sample of four countries, South Africa’s TEA rate, at 10.8 per cent (population aged between 18 and 64 engaged in various levels of entrepreneurial activities), was below the average of 12.1 per cent for the African region in 2019. In addition, South Africa was ‘below the 8.3 per cent average for the African region at 3.5 per cent of established-​business ownership, which measures the percentage of the adult population who are currently owner-​managers of an established business, which has paid salaries or wages, or any other form of payment for more than 42 months’ (Bosma et al. 2020: 19). Relative to other BRIC nations, while China (8.7

632   Boris Urban per cent) and Russia (9.3 per cent) lag behind South Africa (10.8 per cent) in terms of the TEA rate, this advantage disappears once the established-​business ownership rate is similarly compared with both China (9.3 per cent) and Russia (5.1 per cent), which both surpass South Africa (3.5 per cent). Similarly, the United Kingdom (8.2 per cent) and the United States (10.6 per cent) have higher established-​business ownership rates. South Africa’s low established-​business ownership rate is worrying as it implies that most businesses are not surviving beyond 42 months. There can be little doubt that this points to difficulties in transitioning from an early-​stage entrepreneurial business into established businesses. Table 29.1 also shows South Africa’s rank out of the fifty countries surveyed in the GEM study. While it has a relatively competitive rating in terms of nascent entrepreneurship (21/​50), this dwindles considerably when established-​business ownership (44/​50) is measured (Bosma et al. 2020).

29.3.6 Employment Provided by SMMEs Contrasting findings are reported with regard to employment provided by SMMEs in South Africa. According to the SBI, while constituting 98.5 per cent of formal firms in South Africa, the SME sector only contributed 28 per cent to employment (SBI 2018). Findings by Bhorat et al. (2018) using 2013 data on labour market dynamics in South Africa, show that 58 per cent of workers are employed in SMMEs, and most of these are found in the SMME category of ten to forty-​nine employees, considered to be small businesses. According to the ‘Quarterly Labour Force Survey (StatsSA 2019) in the third quarter of 2019, the SMME sector provided 70 per cent of all jobs in South Africa (11.6 of 16.6 million), up from 61 per cent (10.1 of 16.6 million) a year ago’ (SEDA 2019: 16). Internationally, there has been growing recognition of the role of self-​employment and micro-​enterprises in driving employment (see ILO 2019). Drawing on a database from household, labour, and other statistical surveys conducted in a representative set of ninety-​nine countries between 2009 and 2018, the ILO (2019) report shows that in many countries, the self-​employed and micro-​enterprises make up more than 50 per cent of total employment. The regions with the highest employment share of self-​employment also reveal the highest employment share of the informal sector and the highest share of employment in agriculture (ILO 2019). Based on the findings it seems as if some form of entrepreneurial activity can certainly contribute towards poverty alleviation, particularly where employment opportunities in the formal sector are limited.

29.3.7 Contribution of SMMEs to South Africa’s GDP Sources differ in the stated contribution of SMMEs to South Africa’s GDP, where, according to Statistics South Africa, the ‘SMME sector contributes approximately 42 per cent to South Africa’s GDP, while others estimate that SMMEs contribute an estimated 45 to 50 per cent of GDP’ (DTI 2004, cited in Bohart et al. 2018: 4). According

Entrepreneurship and SMMEs in South Africa    633 to the annual financial statistics, the contribution to GDP by formal SMEs outside of agriculture, estimated as post-​tax profit plus employment costs, is calculated at 33 per cent of the total for formal private enterprises. Most estimates put the share of the informal sector in the South African GDP at around 6 per cent. Since the public sector contributes a fifth of the GDP, that would mean that small business as a whole, including medium-​sized enterprises, contributed just under a third of the non-​agricultural GDP (REB 2019). The ‘DSBD, in its Annual Review of Small Businesses and Cooperatives, using both StatsSA and Quarterly Labour Force Survey (QLFS)’ data indicate a 17 per cent increase in micro enterprises’ contribution to GDP. There is a reduction of 13 per cent for small and very small enterprises, and a 6 per cent increase for medium and large enterprises (SEDA 2020: 35). Critically, it has been argued, ‘slow economic growth in itself may cause the wrong allocation of ability and entrepreneurship. It is well known that when economic growth is low and employment opportunities in the formal sector are scarce, that self-​employment will rise, and that this increase will include a large proportion of people with low levels of entrepreneurial ability’ (Naudé 2008: 104). Under this scenario, the ‘quality of the entrepreneurial pool in a country worsens from both the inflow of low-​ability entrepreneurs as well as the outflow of high-​ability entrepreneurs. This will lead to further restrictions from the side of credit markets, in the form of higher interest and/​or collateral requirements, and the consequence is that poor countries may be caught in a self-​reinforcing entrepreneurial development trap’ (Naudé 2008: 107). As the per capita income of the country increases, industrialization and economies of scale allow larger and established firms to take advantage of the increasing demand of growing markets and to increase their relative role in the economy (Urban 2020). These larger firms, because they provide the necessary employment in the form of more stable jobs, tend to result in a reduction of entrepreneurial activity (Audretsch et al. 2007). Furthermore, there is some evidence that entrepreneurial activity rises again as income per capita increases and countries with the highest levels of GDP show increasing early-​ stage entrepreneurial activity, particularly in terms of opportunity-​driven activities. This explanation of entrepreneurial activity is often stylized as the ‘U-​Curve’ (Bosma et al. 2020).

29.3.8 Overall Impressions The South African entrepreneurship and SMME landscape is largely dominated by micro and small informal enterprises. Many of these informal micro enterprises tend to be survivalist firms, which do not contribute significantly to the GDP, nor generate many productive employment opportunities. Compared to informal business, formal small and medium enterprises typically have more capital and skills, use more advanced technologies, and employ more people. However, SMMEs as a whole remain underdeveloped by international standards and South Africa has a low established-​business ownership rate, which is disturbing as it indicates that most businesses are not surviving in the long-​term.

634   Boris Urban

29.4  The Entrepreneurial Ecosystem in South Africa Widespread research shows that a foremost precondition for a productive SMME sector in any country is the presence of an ‘enabling environment’. For instance, the series of GEM reports (Bosma et al. 2020), the Global Entrepreneurship and Development Institute (GEDI) (Acs and Szerb 2010), as well as the Global Competitiveness Index (GCI) (WEF 2020), measure ‘entrepreneurial environment conditions which focus on institutions, infrastructure, macroeconomic environment, health and primary education, higher education and training, good market efficiency, labour market efficiency, financial market development, technological readiness, market size, business sophistication, and innovation’ (World Bank 2019: 612). Global Performance Indicators (GPIs), such as the aforementioned, rate and rank states against one another, and deliberately package information to influence the perceptions of states and publics and the decisions of economic actors (Doshi et  al. 2019). The Ease of Doing Business (EDB) index, a widely used GPI and adopted in many reports (see Global Entrepreneurship Monitor (GEM) studies), is regularly used to compare the ease of doing business across different economies (WEF 2020; World Bank 2019). Despite its dominance, the EDB indicator occupies a contested space and faces criticisms about its accuracy and validity. Since it was launched in 2003, the EDB has attracted criticism for its anti-​regulation bias. Inspired by the ‘Index of Economic Freedom’ at the conservative Heritage Foundation, the report has encouraged countries to take part in the ‘deregulation experience’ including reductions in employment protection, lower social-​ security contributions and lesser corporate taxation. Doshi et al. (2019) demonstrate how the EDB ranking system affects policy through bureaucratic, transnational, and domestic political channels. According to these authors, by benchmarking and especially by ranking, the World Bank intentionally exerts competitive social pressure on states to deregulate. Furthermore, unions and the ILO have criticized the EDB for neglecting the consequences of business deregulation for workers. The World Bank eventually removed labour-​related components from the Index. Even so, the EDB is at odds with the changing focus of the World Bank, with reports such as the ‘Balancing Regulations to Promote Jobs’ manual. Perhaps, as Doshi et al. (2019) suggest, the case of the World Bank’s EBD Index is a reminder that widespread comparative quantification simply reinforces global power structures. Notwithstanding there is no universal agreement about the definition of the entrepreneurial ecosystem, and the causal links within the system (Stam and Bosma 2015), a broad range of factors essential to the entrepreneurship problem in South Africa (Urban 2015), are analysed in this section, specifically factors where there has been convergence of research and policy findings. This section then ends by demonstrating the importance of interactions between different elements of an entrepreneurial ecosystem.

Entrepreneurship and SMMEs in South Africa    635

29.4.1 Education Education is widely accepted as having a positive relationship with running a successful business and enterprise growth (Marvel et al. 2016). In all countries, an educated population with the requisite knowledge, skills, and capacity for innovation has proven vital to driving competitiveness, productivity, and sustainable growth (Stam and Bosma 2015). Human capital in the form of ‘entrepreneurial skills and competencies are central to the successful establishment and performance of SMMEs, where extensive research shows positive links between various human capital factors and entrepreneurial success’ (Marvel et al. 2016: 611). Historically, the series of GEM reports emphasize that the education system in South Africa has failed entrepreneurs, where from ‘2017 to 2019, there was a significant drop in education completion beyond the primary education level. This finding is extremely concerning as primary education level amongst total early-​stage entrepreneurs is vital for any developing nation’ (Bowmaker-​Falconer and Herrington 2020: 17). Moreover, the lack of quality education in South Africa and subsequent poor coverage of science, technology, engineering, and medicine (STEM) education, results in the majority of youth entrepreneurs converged in low-​technology sectors (SEDA 2020).

29.4.2 Access to Finance For many SMMEs, access to finance is hindered by a range of demand-​and supply-​ side obstacles, such as skills shortages, poor management practices, and workforce training limiting their productivity and innovation (OECD 2009). Several reports reveal that a dearth of adequate finance is a critical barrier to SMME growth in South Africa (SEDA 2020). Similar to many developing and emerging nations, start-​up funding in South Africa ‘often comes from personal savings or money from family, where the youth, women, and people in townships and rural areas are likely to be particularly disadvantaged when it comes to accessing financing for their start-​ups’ since they also lack collateral and have excessive outstanding debt (Ndou and Urban 2019: 5). Furthermore, as many as two-​thirds of applications for bank credit submitted by new SMEs in South Africa are rejected, according to SEDA (2020). Additionally, South Africa has little tradition in business angel investment activity, despite being a key funding option for entrepreneurs worldwide. In South Africa, a significant amount of the capital available from venture capitalists is dedicated to the well-​developed late-​ stage investment market, and finance for seed and start-​up phases remains insufficient (Ndou and Urban 2019). Notwithstanding this access to funding obstacles, alternative financing/​funding options are becoming more prevalent in the South African entrepreneurial ecosystem, where a ‘blended financing model’ in the form of the Small Business Innovation Fund aims to lower financial costs for entrepreneurs with a Rand 1 billion budget for the 2019/​20 period (Bowmaker-​Falconer and Herrington 2020: 46).

636   Boris Urban

29.4.3 Access to Markets Many SMMEs in South Africa face restricted market access and opportunities to participate in market exchanges and networks (Venter and Urban 2015). Moreover, restricted market access for SMMEs is generally cited due to the concentrated nature of the South African economy, which is dominated by large private companies (Bhorat et al. 2018) and monopolies are prevalent in the retail and wholesale, energy supply, telecommunications, financial services, and transport sectors (Bowmaker-​Falconer and Herrington 2020). These ‘big players’ often hold a majority of the market share, with established supplier relationships prohibiting SMMEs from taking advantage of any real market opportunities. Such hindrance to gain entry into the market is evident insofar as most small enterprises in South Africa are predominantly found in trade and business services and the informal sector is overwhelmingly concentrated in the retail sector. Such limited entrepreneurial activities prohibit SMMEs from climbing up the value chain and they remain suppressed and uncompetitive (Ndou and Urban 2019).

29.4.5 Access to Technology Technology is a powerful instrument to advance entrepreneurial activity in any country, where the use of the e-​commerce, Internet, and smartphones has seen exponential growth in both developed and developing countries (WEF 2020). In South Africa a lack of suitable skills, weak supporting infrastructure, inadequate policy and leadership, as well as cultural barriers have been cited as obstacles that prevent the potential of technology and the digital economy to be fully realized (Bosma et al. 2020; SEDA 2020). Moreover, research shows that the role of ICT in fostering financial inclusion and growth in developing countries is not very promising, as a very small percentage of SMMEs in both the formal and informal sector are able to access and utilize technology to position themselves successfully in a competitive landscape (WEF 2020). This phenomenon has been noted as the ‘digital divide’, which clearly disadvantages women and youth in rural areas and the informal economy in South Africa (SEDA 2020).

29.4.6 Access to Support Networks In the field of entrepreneurship, social networks as a form of social capital have emerged as an important contextual factor, as typified by the social structure within which the actor is located (Venter and Urban 2015). Social exchanges typically occur between families, kin, and other investors to help new firms get started, while educational establishments and media encourage entrepreneurship through promoting role models and highlighting entrepreneurial events. For instance, the South African Business Angel Network, as well as a host of government and private incubator organizations,

Entrepreneurship and SMMEs in South Africa    637 such as the Awethu Project and the Tshimologong Digital Innovation Precinct, support entrepreneurs in South Africa (Venter and Urban 2015). Typically, entrepreneurs and SMMEs might use social capital in the form of networking to mitigate risks and overcome challenges, specifically in the informal economy. However, studies indicate that SMME owners cite a lack of networks in South Africa that can provide support in terms of opportunities, funding avenues, markets, and information sources (Ndou and Urban 2019). While networks may exist, low levels of awareness, and the ability of SMMEs to access and utilize such networks is limited in South Africa (DSBD 2017; SEDA 2020). Research also shows that most networks’ members are either friends or family members, and networking is largely unstructured and coincidental in nature, where often, because of inadequate systems within government departments, SMMEs are subject to or complicit in corruption (SEDA 2019). The result is that nascent entrepreneurs are reluctant to deal with government departments or agencies since they anticipate corruption and cronyism (Ndou and Urban 2019).

29.4.7 Regulatory and Political Environment The regulatory environment includes existing national laws, rules, policies, and regulatory bodies which will permit certain behaviours while restricting others. In South Africa, government often creates regulatory burdens for entrepreneurs due to ideological differences or inertia, inefficient government bureaucracy (red tape), and as a result of a lack of coordinated focus and policy instability (Bowmaker-​Falconer and Herrington 2020; DSBD 2017). The cost of regulatory compliance is extremely high for entrepreneurs and SMMEs who have to deal with restrictive employment laws, which are one of the biggest regulatory obstacles they face (Ndou and Urban 2019). Moreover, small business owners report that the current regulatory environment is more appropriate and advantages big business as opposed to small business (DSBD 2017), where poor service delivery puts an unnecessary burden on already very scarce management resources, which constrains SMMEs’ performance and growth (SEDA 2020). Furthermore, starting a business requires extensive ‘infrastructure; specifically transport networks, communication systems, provision of electricity, and other utilities. This is particularly important for SMMEs, for whom the costs and time spent using infrastructure are a relatively high proportion of income’ (Bohart et al. 2018: 47). Poor physical and ICT infrastructure prevents SMMEs from operating efficiently and accessing international markets. Additionally, in several instances, red tape (which is a code word for rent-​seeking activities, such as nepotism and corruption) concerning infrastructure and regulations results in unproductive entrepreneurship. Crime has also been frequently cited as a matter of concern for SMMEs specifically, as it disproportionately affects SMMEs relative to large firms (Bohart et al. 2018). Crime is not only an unproductive economic activity but any incentive to reinvest in a business is eroded once a small business owner has been a victim of crime (DSBD 2017).

638   Boris Urban

29.4.8 An Enabling Entrepreneurial Ecosystem Scholarly research shows that the broader institutional environment together with the more immediate business environment affects the level of entrepreneurial activity of a country. In any context, the entrepreneurial ecosystem needs to be analysed as a reciprocal process between an individual entrepreneur and the external environment, with the entrepreneur and his/​her environment interacting and influencing each other (Stam and Bosma 2015). Two decades of empirical research have generated a great number of studies demonstrating that entrepreneurial activity depends on interactions between three components:  individuals, organizations, and institutions (Bosma et al. 2020). While an effective entrepreneurial ecosystem is best designed as a ‘set of interdependent actors and factors coordinated in such a way that they enable productive entrepreneurship’ (Stam and Bosma 2015: 331), an important point is to recognize how the wider ecosystem structure of the South African economy affects entrepreneurship. Several structural factors impede the South African entrepreneurial ecosystem. These include issues such as a poor-​performing economy in terms of GDP growth and an economy dominated by large firms. Moreover, South Africa’s dual economy excludes a high proportion of the population from participating in the formal economy, which is exacerbated by inadequate education and energy infrastructure, and an unusually low share of employers and self-​employed people in the labour force, which leads to persistent poverty and inequality (REB 2017, 2019; Urban 2020). These wider ecosystem conditions highlight that even though ecosystem elements can support each other, they can never completely replace one another. In this sense, the elements of the broader economy and entrepreneurial ecosystem interact in complex and specific ways that lead to unique configurations of different regional and local ecosystems (Bosma et al. 2020). Larger firms are also part of the ecosystem and large firms in terms of knowledge spillovers influence SMMEs, their access to networks, and opportunities to collaborate with other players in the ecosystem. Upstream and downstream linkages with larger companies are vital for SMMEs and the benefits of such inter-​firm linkages can be crucial for the competitiveness of entire supply chains, at both local and global levels. In this respect, close interdependence and coordination between large and small businesses represents an important source of value and competitiveness along supply chains (OECD 2009). Such reciprocity between large firms and SMMEs is important in South Africa as large firms dominate employment growth (in terms of both the number of jobs added and growth rates in the 2011 to 2016 period) (SBI 2018). Moreover, large companies often have procurement systems constructed to acquire goods on a scale beyond the scope of small businesses (REB 2017). Since SMMEs are often embedded in local ecosystems, which represent their primary source of knowledge, skills, finance, business opportunities, and networks, it is also important to consider factors affecting framework conditions at the local level, and how interventions developed at the national level are suited to local conditions (OECD 2009;

Entrepreneurship and SMMEs in South Africa    639 Urban 2015). For instance, instead of focusing on one or two isolated factors, such as start-​up finance, which tends not to have the anticipated effect as a stand-​alone intervention, a synergistic combination of several relevant local ecosystem factors is required for any intervention to have a positive effect on entrepreneurship in a specific region. Studies conducted in the South African context demonstrate how interactions between various institutional factors and different motivational and behavioural outcomes such as self-​efficacy have a positive effect on small businesses (Urban 2015). Recognizing that the entire entrepreneurial process unfolds because ‘individuals act and are motivated to pursue opportunities’ (Shane and Venkataraman 2000), concentrating on the ‘actor’ in the entrepreneurial ecosystem is imperative, since to be an agent is to make things happen intentionally by one’s own actions (Bygrave 1989; Venter and Urban 2015; Welter and Smallbone 2011). Agency in the form of an entrepreneurial mindset and behaviours has been shown to interact positively with the institutional environment and have an impact on entrepreneurship activity (Audretsch et al. 2007; Krueger et al. 2000). In this sense, the environment is ‘not something that can be taken as a given, but instead is enacted by entrepreneurs. Environments are neither certain nor uncertain in themselves, but thinking makes them so—​action develops in a duality between agency and structure’ (Welter and Smallbone 2011: 110).

29.5  Policy Critique and Future Considerations Over the past twenty years, the South African government has launched a number of policy initiatives designed to support SMMEs, including a new Ministry of Small Business Development established in early 2014. The National Development Plan (NDP), as a fundamental socio-​economic strategy is designed to eliminate poverty and reduce inequality in South Africa by 2030. This broad-​spanning and large-​scale investment project aims to develop the economic infrastructure, education, social welfare, science, and innovativeness of the country (Venter and Urban 2015). Additionally, the small business policy framework is propelled primarily by the Small Business Development Act which has as its objective the stimulation of growth and development of the South African SMME sector. The Department of Trade and Industry (DTI) in South Africa also has a number of schemes to support the funding and advancement of SMMEs, namely in the form of the Small Enterprise Development Agency (SEDA), which provides business development and support services for small enterprises through its national network. Part of the central dilemma driving this chapter relates to the lack of implementation of these policies and quality of support services that vary considerably and where in many areas, the business advisors are not properly trained nor do they have the practical skills and experience required to give advice to small business entrepreneurs (Bowmaker-​Falconer and Herrington 2020).

640   Boris Urban Notwithstanding the unproven efficacy and limited empirical results of entrepreneurship policy, governments around the world continue to introduce entrepreneurship and SMME policy instruments, even though they are separate entities (Acs et al. 2016). Whereas entrepreneurship policy is focused on the ‘pre-​start, the start-​up and post-​ start-​up phases’ of the entrepreneurial process and aims primarily to encourage more individuals to consider entrepreneurship as a career option, SMME policy, in contrast, is focused on growing the ‘existing population’ of SMMEs by developing interventions and support measures to encourage their sustainability (Acs et al. 2016; Stam and Bosma 2015). This differentiation seems to be lost on South African policymakers. A missing part of the policy puzzle in South Africa is the failure by government to recognize that while state-​ sponsored entrepreneurship interventions in some instances may have positive results, there are ‘opportunity costs’ associated with such policies and actions which may be unexpected and have the unintended consequence of fostering non-​productive entrepreneurship (Baumol 1990; Shane 2009). Policy actions and programmes aimed at start-​ups often subsidize low-​growth firms that add little to employment and economic development. In this regard Shane (2009: 141) declares, ‘policy makers believe a dangerous myth, insofar as they think that start-​ups are a “magic bullet” that will transform depressed economic regions, generate innovation, create jobs, and conduct all sorts of other economic wizardry’. According to Shane (2009), this belief is flawed because the typical start-​up is not innovative, creates few jobs, and generates little wealth. Certainly, to obtain more economic growth by having more start-​ups, these new ventures would need to be more innovative and productive than existing companies are, but generally, this is not the case (Acs et al. 2016; Shane 2009). In many instances policymakers find themselves picking winners from among the population of nascent entrepreneurs (Acs et al. 2016), which inadvertently stimulates a disproportionate number of people to start new businesses in competitive industries with lower barriers to entry and high rates of failure (Shane 2009). In other words, a policy intended to foster productive entrepreneurship may ‘perversely’ suppress viable entrepreneurial ventures while inadvertently supporting ideas and enterprises that are non-​viable or accepted by the market (Acs et al. 2016; Shane 2009). In reality, most governments and policymakers do not have a convincing record of picking winners, principally as the economic value of any offering must itself be ascertained through the trial-​and-​error process of entrepreneurship (Shane 2009; Shane and Venkataraman 2000). While offering prescriptive advice is not the point of this chapter, it may be useful to stress the many voices which call for evidence-​based interventions to improve and enhance SMME policies (Acs et al. 2016; Rogerson 2004; Shane 2009; Stam and Bosma 2015; Urban 2020). In South Africa, policymakers and SMME development agencies have been criticized for not providing a ‘level playing field’ for SMMEs, which could be achieved by strengthening competition legislation to lower barriers to entry for new operators in banking, telecoms, and other sectors where large companies dominate. However, one single policy cannot affect entrepreneurial activity as there is a broad range of issues underlying the entrepreneurship problem in South Africa. In which

Entrepreneurship and SMMEs in South Africa    641 case prioritizing interventions at a macro-​level, combined with the political determination to introduce a new system of regulation, taxation, and incentivization, are necessary to save small businesses and help reduce their vulnerability, especially as a post-​COVID strategy. Perhaps a more ordered approach could make a significant difference to SMME policy if it focused on separating direct versus indirect support to SMMEs by developing reliable infrastructure (physical and digital), ensuring a region-​ specific mix of small business-​ friendly regulation, while also developing strategies to mitigate potential business costs of monopolies, crime, and corruption. Similiar to other countries, policymakers may encourage entrepreneurship through a low-​regulation route and simultaneously adopt a high-​support route (in the form of mentors, information, training, or finance) in which the number and severity of obstructive regulations are minimized to enable growth (Stam and Bosma 2015; Ndou and Urban 2019). Recognizing the heterogeneity of the South African SMME landscape, and to avoid a ‘one-size-fits-all’ approach, a dynamic evolutionary ecosystem approach is needed to develop an overall coordinated enabling environment which makes it possible to construct and compare different regional entrepreneurial ecosystems in their evolution and performance over time (Audretsch et al. 2007; Ndou and Urban 2019). In this regard, a policy framework is required that encompasses broad policies that influence SMMEs as well as specific targeted high-growth entrepreneurship policies (OECD 2009). These policy interventions often span the boundaries of ministries and government agencies, as well as levels of government, making it difficult to coordinate in the current South African public-​sector milieu characterized by political intrigue and pointless party posturing. In the final analysis, public-​sector integrity and transparency, coupled with public administration efficiency and quality of public-​services offerings are essential for a vigorous SMME sector (OECD 2009; Urban 2015).

29.6 Conclusion This chapter highlights the complexity of understanding entrepreneurship and SMMEs in South Africa due to the heterogeneity in entrepreneurial activity and the variety of contexts in which it occurs. It is argued that although the South African government has publicly and consistently supported SMMEs, in reality, entrepreneurs and SMMEs face a multitude of obstacles and the country’s established-​business ownership rate is still far below the average for other developing countries. Notwithstanding the lack of consistent and accurate data, a descriptive and analytical overview highlights that the South African SMME landscape is dominated by mostly micro and small informal enterprises, which do not contribute significantly to the GDP, nor generate many productive employment opportunities. Additionally, in analysing the entrepreneurial ecosystem in South Africa, a number of structural, institutional, and policy issues are exposed as major hurdles inhibiting the start-​up and growth of SMMEs.

642   Boris Urban It is further established that documenting SMME activity rates has become a major focus of research and in the process some of the wider issues addressed by theory have disappeared (Urban 2015). Providing more emphasis on prior research and theories surrounding entrepreneurship as a field of study is advocated, rather than simply constructing a highly descriptive statistical annual profiling of the SMME activity rates and ecosystem factors. Realizing that entrepreneurship and SMMEs are difficult to define, as there are fundamental differences in theory and practice related to these terms, the point is made that the entrepreneurship process is premised upon opportunity recognition and innovations, which is influenced by the configuration of incentives in a specific environment. Historically, in South Africa the SMME sector has been marginalized in favour of big business and currently there is a political attachment to micro enterprises in the informal sector. Consequently, it is not entirely surprising that productive, systemic, high-​ growth entrepreneurial activity is largely missing and most SMME activity is survivalist in nature. Accordingly, if fostering an enabling entrepreneurial environment is desirable for policymakers, the importance of coordinating and achieving synergistic interactions between the ‘agent in the entrepreneurial drama’ and a wider set of ecosystem factors in the economy is pivotal. Analogous to the entrepreneurship process, translating policy into actual entrepreneurial activity depends not only on the formulation of policy but rather more so on the quality of its execution. While it is almost clichéd to note that South Africa requires less policy formulation and more effective policy implementation, in the entrepreneurship context this observation is critical, as the economic value of any offering must itself be realized through the navigating of the entrepreneurship process. In addressing the principal problem steering this chapter, it is shown that government officialdom as well as corruption both in the public and private sectors stifle entrepreneurial activity, where in several instances SMME policy results in unproductive entrepreneurship. In the final analysis, entrepreneurship in South Africa is a highly constrained activity that does not live up to its potential as the key to unlocking inclusive growth of the economy, as many ideologues would have us believe.

References Acs, Zoltan and László Szerb. 2010. ‘The Global Entrepreneurship and Development Index (GEDI)’. Paper presented at the Summer Conference 2010 on Opening up Innovation: Strategy, Organization and Technology, June, Imperial College, London. Acs, Zoltan, Thomas Astebro, David Audretsch, and David Robinson. 2016. ‘Public policy to promote entrepreneurship: A call to arms’, Small Business Economics, 47(1): 35–​51. Audretsch, David, Isabel Grilo, and Roy Thurik. 2007. Handbook of Research on Entrepreneurship Policy. Cheltenham: Edward Elgar. Baumol, William. 1990. ‘Entrepreneurship: Productive, unproductive and destructive’, Journal of Political Economy, 98(5): 893–​921.

Entrepreneurship and SMMEs in South Africa    643 Bergmann, Heiko, Susan Mueller, and Thomas Schrettle. 2014. ‘The use of global entrepreneurship monitor data in academic research: A critical inventory and future potentials’, Journal of Entrepreneurial Venturing, 6(3): 242–​76. Bhorat, Haroon, Zaakhir Asmal, Kezia Lilenstein, and Kirsten Van Der Zee. 2018. ‘SMMEs in South Africa: Understanding the Constraints on Growth and Performance’. Development Policy Research Unit (DPRU) Working Paper 2018, 2 July. University of Cape Town. Bosma, Niels, and Jonathan Levie. 2009. ‘Global entrepreneurship monitor 2009 global report’. Global Entrepreneurship Research Association, Babson College, Babson, MA. Bosma, Niels, Stephen Hill, Aileen Ionescu-​Somers, Donna Kelley, Jonathan Levie, and Anna Tarnawa. 2020. ‘Global entrepreneurship monitor 2019/​2020 global report’. Global Entrepreneurship Research Association, London Business School. Bowmaker-​Falconer, Angus and Mike Herrington. 2020. ‘Global entrepreneurship monitor South Africa (GEM SA) 2019/​2020 report’. Global Entrepreneurship Research Association, Stellenbosch University, Cape Town. Bygrave, William. 1989. ‘The entrepreneurship paradigm (I & II):  Chaos and catastrophes among quantum jumps’, Entrepreneurship Theory and Practice, 14(1): 7–​26. Doshi, Rush, Judith Kelley, and Beth A. Simmons. 2019. ‘The power of ranking: The ease of doing business indicator and global regulatory behavior’, International Organization, 73(3): 1–​33. DSBD (Department of Small Business Development). 2017. ‘Small business and cooperatives in South Africa: Annual review 2016/​2017, 2019.’ DSBD’. Pretoria. ILO (International Labour Organization). 2019. ‘Small matters: Global evidence on the contribution to employment by the self-​employed, micro-​enterprises and SMEs’. Publications Production Unit (PRODOC) of the ILO, Geneva. Kirzner, Israel. 1997. ‘Entrepreneurial discovery and the competitive market process:  An Austrian approach’, Journal of Economic Literature, 35(1): 60–​85. Kloepfer, Kathryn, and Gary Castrogiovanni. 2018. ‘Entrepreneurship: Venture creation sub processes, subdomains, and interfaces’, International Entrepreneurship and Management Journal, 14(4): 681–​96. Knight, Frank Hyneman. 1921. Risk, Uncertainty and Profit. New  York:  Houghton Mifflin (Chicago, IL: University of Chicago Press (1971)). Krueger, Norris, Michael Reilly, and Alan Carsrud. 2000. ‘Competing models of entrepreneurial intentions’, Journal of Business Venturing, 15(3): 411–​32. Kujinga, Leanne, and Boris Urban. 2017. ‘The institutional environment and social entrepreneurship intentions’, International Journal of Entrepreneurial Behaviour and Research, 23(4): 638–​55. Marks, Jonathan, and Karen Hidden. 2017. ‘SMMEs and the green economy: Muddy waters and murky futures’, Gordon Institute of Business Science, University of Pretoria. Marvel, Matthew, Justin Davis, and Curtis Sproul. 2016. ‘Human capital and entrepreneurship research: A critical review and future directions’, Entrepreneurship Theory and Practice, 40(3): 599–​626. Myres, Kerrin, Anastacia Mamabolo, Nyasha Mugudza, and Lauren Jankelowitz. 2017. ‘Survey of social enterprises in South Africa’. Entrepreneur Development Academy, Gordon Institute of Business Science, University of Pretoria. Naudé, Wim. 2008. ‘Entrepreneurship in the Field of Development Economics’, in Boris Urban (ed.) Frontiers of Entrepreneurship. Advanced Entrepreneurship Higher Education Series. Berlin: Springer.

644   Boris Urban Ndou, Brian, and Boris Urban. 2019. ‘Informal entrepreneurship: A focus on South African township entrepreneurs’, Journal of Developmental Entrepreneurship, 24(4): 1–​19. Nicholls, Alex. 2006. Social Entrepreneurship:  New Models of Sustainable Social Change. Oxford: Oxford University Press. OECD (Organisation for Economic Co-​operation and Development). 2009. ‘Co-​operation and development, organisation for economic, measuring entrepreneurship:  A collection of indicators’, 2009 edition. Ewing Marion Kauffman Foundation Research Paper [online] http://​dx.doi.org/​10.2139/​ssrn.1581491. REB (Real Economy Bulletin). 2017. ‘Special edition: The state of small business in South Africa’. [online] https://​www.tips.org.za. REB (Real Economy Bulletin). 2019. ‘Special edition: The state of South Africa’. [online] https://​ www.tips.org.za. Rogerson, Christian, 2004. ‘The impact of the South African government’s SMME programmes: A ten-​year review (1994–​2003)’, Development Southern Africa, 21(5): 765–​84. RSA (Republic of South Africa). 1996. ‘Government Gazette No. 17612. Act No. 102, 1996 National Small Business Act. 1996’. 27 November. Pretoria: Government of SA. RSA (Republic of South Africa). 2019. Government Gazette No. 42304. ‘Department of Small Business Development, Revised Schedule 1 of the National Definition of Small Enterprise’. Pretoria: Government of SA. Sautet, Frederic. 2013. ‘Local and systemic entrepreneurship: Solving the puzzle of entrepreneurship and economic development’, Entrepreneurship Theory and Practice, 37(2): 387–​402. SBI (Small Business Institute). 2018. ‘The number of formal micro, small and medium businesses in South Africa: Preliminary findings of stage 1 of the baseline study of small businesses in South Africa’. Johannesburg. Schumpeter, Joseph Alois, 1934. The Theory of Economic Development. Cambridge, MA: Harvard University Press. SEDA (Small Enterprise Development Agency). 2019. ‘Small Enterprise Development Agency. SMME quarterly update: 3rd quarter 2019’. SEDA, Pretoria. SEDA (Small Enterprise Development Agency). 2020. ‘Small Enterprise Development Agency and the Department of Small Business Development: Women and youth owned SMMEs. Integrated report’. Urban-​Econ Development Economists, Brooklyn, Pretoria. Shane, Scott. 2009. ‘Why encouraging more people to become entrepreneurs is bad public policy’, Small Business Economics, 33(2): 141–​9. Shane, Scott, and Sankaran, Venkataraman. 2000. ‘The promise of entrepreneurship as a field of research’, Academy of Management Review, 25(2): 217–​35. Stam, Erik, and Niels, Bosma. 2015. ‘Growing entrepreneurial economies: Entrepreneurship and regional development’, in T. Baker and F. Welter (eds) The Routledge Companion to Entrepreneurship. London: Routledge. StatsSA (Statistics South Africa). 2019. Quarterly Labour Force Survey. Pretoria:  Statistics South Africa. Urban, Boris. 2015. ‘Indicators of entrepreneurial activity: Untangling the relationship between economic and social factors driving entrepreneurship’, in Boris Urban (ed.) Perspectives in Entrepreneurship:  A Research Companion. Advanced Entrepreneurship Higher Education Series. Cape Town: Pearson HE Publishers. Urban, Boris. 2020. ‘The interplay between institutional factors and experience as determinants of employment growth in South African SMEs’, International Journal of Business and Globalisation, 24(1): 1–​18.

Entrepreneurship and SMMEs in South Africa    645 Venter, Robert, and Boris Urban. 2015. Entrepreneurship Theory in Practice, 3rd ed. Cape Town: Oxford Southern Africa. WEF (World Economic Forum). 2020. ‘The global competitiveness report 2019’, edited by K. Schwab. Geneva: WEF. Welter, Friederike, and David Smallbone. 2011. ‘Institutional perspectives on entrepreneurial behaviour in challenging environments’, Journal of Small Business Management, 49(1):107–​25. Williams, Colin, and Abbi Kedir. 2018. ‘Explaining cross-​national variations in the prevalence of informal sector entrepreneurship:  Lessons from a survey of 142 countries’, Journal of Developmental Entrepreneurship, 23(1): 1–​12. World Bank, 2019. ‘Doing business 2019: Training for reform’. World Bank, Washington, DC [online] https://​www.doingbusiness.org/​en/​reports/​global-​reports/​doing-​business-​2019.

Chapter 30

Urbaniz at i on, Ag gl om erat i on, a nd E c onomic Dev e l op me nt in Sou th   A fri c a Ivan Turok

30.1 Introduction Around the world, economists of various persuasions have been developing a deeper understanding of the importance of geography for economic growth and development (Porter 2003; Krugman 2011; Glaeser 2011; Collier and Venables 2017). An initial interest in transport costs has broadened into the positive effects of spatial concentration on productivity and innovation. Proximity between firms and households in cities is believed to generate economies of scale and foster interactions which promote learning, improve efficiency, stimulate enterprise, and raise investment returns. Although disentangling cause–​effect relationships is difficult within these ‘agglomeration economies’, there is increasing empirical endorsement, including from new digital and satellite data available at granular resolutions (Duranton and Puga 2020; Jones et al. 2020). This body of knowledge also indicates that the process of urbanization is generally beneficial for drawing dispersed populations and resources closer together, creating more productive jobs, and lifting people out of poverty (Spence et al. 2009; OECD/​EC 2020). It suggests that dense, compact forms of urban development are most advantageous, along with transport and digital connectivity within and between cities (Lall et al. 2017; Ahlfeldt and Pietrostefani 2019). Belief in the significance of agglomeration economies has begun to influence national economic policies and infrastructure plans (Pike et al. 2017). There is a growing sense that cities need to be treated as distinct economic units because this is where the wealth of nations is increasingly created (Glaeser and Joshi-​Ghani 2015). Tackling the

Urbanization, Agglomeration, and Economic Development  647 congestion, higher costs, pollution, and other negative externalities in burgeoning cities can therefore have disproportionate benefits for national prosperity. The uplift in land and property values arising from well-​configured urban development can also generate valuable additional tax revenues for reinvestment in essential infrastructure to render the whole process cumulative and self-​sustaining (Ingram and Hong 2012). These propositions have been embraced by international organizations such as the United Nations (2020), OECD (2014, 2018), and World Bank (2009, 2013). They have endorsed a new emphasis on cities in global thinking on sustainable development, both to harness the progressive potential of rapid urbanization underway across Africa and Asia, and to mitigate the risks that it will prove dysfunctional and degrade the environment (Turok and McGranahan 2013; United Nations 2016). Successful outcomes are believed to depend on governments establishing capable institutions to guide the process because market mechanisms cannot organize urban development effectively or provide sufficient public goods and services to leave no one behind (Collier and Venables 2015; Scott and Storper 2015). Economists in South Africa have shown limited interest in the spatial economy, or in the contribution cities make to development. With some exceptions, political elites and economic policymakers have also been ambivalent with the result that building better cities is not viewed as an economic priority (SACN 2016; Duminy et al. 2020). There is no common vision or shared objectives across government towards urban development, nor even an agreed approach towards using vacant state-​owned urban land effectively. Instead of integrative place-​based policies, the government has emphasized spatially blind measures, such as industrial sector masterplans and universal access to public services and housing. The latter are intended to redress historic injustices through a rights-​based framework that is largely indifferent to geography. Hesitancy about urbanization neglects the global experience of cities as dynamic production systems that create opportunities to foster all-​round progress. Agglomeration principles are also relevant to the structural constraints to growth and inclusion arising from South Africa’s large territory and the fractured form of its cities and towns (National Treasury 2018). The separation between places of work (or production) and living (or social reproduction) hampers economic and social progress in many situations, especially dormitory townships and the former homelands (Todes and Turok 2018; World Bank 2018). It renders urban labour markets less efficient, marginalizes poor black communities, and adds to the costs of transport for commuters and service delivery for taxpayers (SACN 2016; National Treasury 2018). There is little sign that these spatial scars of apartheid are disappearing, either through spontaneous economic processes or through government spatial plans that repeatedly espouse urban integration and spatial transformation (Gardner 2018; Turok 2018; McKenna 2020). The purpose of this chapter is to assess the contribution of cities to economic development in South Africa. It also examines distinctive features of the country’s urban trajectory. The discussion challenges some aspects of agglomeration theory and some attitudes towards urbanization. It argues that the economic benefits of cities are not automatic but instead depend on a conducive context. The way urban growth is shaped

648   Ivan Turok and supported through investment in public infrastructure and inclusive institutions is particularly important. The chapter argues that South Africa’s spatial divides impede inclusive growth, and that more could be done about urban consolidation. The structure is as follows. Section 30.2 sets the context with a brief history of the country’s urban growth. The rest of the chapter is organized around three fundamental features of the urban morphology, namely the triangular relationship between the location of firms, households, and the transport system. Section 30.3 discusses how agglomeration affects the performance of firms and industries. Section 30.4 considers how households are accommodated in expanding cities. Section 30.5 examines how the transport system connects urban areas. Section 30.6 concludes.

30.2  South Africa’s History of Urbanization and Urban Policy South Africa emerged as a resource-​based economy in the nineteenth century, with a particular settlement pattern shaped by that reality. Urbanization was dominated for many decades by the need for cheap migrant labour to mine gold, diamonds, and other minerals, which meant that coercion was a long-​standing feature (Wilson 1972). The rural–urban transition also differed from the pattern in many other countries where structural transformation involved rising agricultural productivity and even more productive urban industrialization. This parallel process fuelled a generalized increase in living standards and widespread investment in infrastructure. Agriculture never developed to its full potential in South Africa, and the same is arguably true about manufacturing. As the twentieth century progressed, many mining towns developed complementary manufacturing functions, including engineering, steel, and chemicals. Many of these were established by the largest mining corporations in order to control their intermediate inputs and profit from emerging opportunities, even branching into various financial services (Harrison and Zack 2012). These conglomerates benefited from significant internal economies of scale and scope, while booming cities elsewhere in the world were characterized more by external economies associated with independent firms driving up productivity through competition (Jacobs 1984; Glaeser 2011). A pattern of concentrated ownership and oligopoly was established in South Africa which became an enduring feature of many economic sectors (Fine and Rustomjee 1996). South Africa’s fastest growing economic centres emerged in and around Gauteng, which soon became the country’s leading metropolitan region (Harrison and Zack 2012). Coastal towns also prospered by functioning as international entrepÔts and regional service centres (Turok 2014). They enabled the export of minerals from the interior, allowed essential inputs to be imported, and coordinated the processing of agricultural goods for domestic consumption and export. Therefore, geographical advantages conferred by natural features were uppermost in fashioning South Africa’s

Urbanization, Agglomeration, and Economic Development  649 spatial economy for many decades. Natural resources remain important in determining whether many towns grow or decline (especially those specializing in mining and tourism), although the economics of agglomeration are more important in moulding the major cities. An essential message is that the country’s wealth has been largely created in cities and towns. Colonial and apartheid governments played instrumental roles in urban development, typically to enforce racial separation and probably hindering economic dynamism in the process. They were responding out of fear to accelerating urbanization during the first half of the twentieth century, caused by the mining boom and industrialization. Less than a fifth of the country’s population (18 per cent) lived in urban areas in 1911, which doubled to 35 per cent in 1951 (Turok 2014). Ideological efforts to suppress black urbanization intensified between the 1950s and 1980s (World Bank 2018; Duminy et al. 2020). This conflicted with the interests of the thriving urban economy, which required a more stable, skilled, and enlarged workforce over time. The stringent restrictions on permanent urban residence gave rise to a detrimental form of (circular) migration between rural and urban areas (Bank et al. 2020). Meanwhile, callous controls on where racial groups could live within cities created a fractured urban structure with poor black communities removed to sterile townships on the periphery. Local government was weak, fragmented along racial lines, and mainly responsible for basic infrastructure services and regulating the development of land.

30.2.1 Urbanization Post-​apartheid During the transition from apartheid the rate of urbanization increased following the removal of migration controls. This sealed the country’s position as one of the most urbanized on the continent, with two-​thirds of the total population now living in urban areas (UNDESA 2018). Figure 30.1 shows the increasing importance of large cities and the relative decline of rural areas. This trajectory mirrors global trends, but it has not been embraced by decision-​makers post-​apartheid (Turok et al. 2021a). In the absence of a national policy to prepare for urban growth, municipalities have battled to keep pace with the demands for public services and shelter, resulting in swelling informal settlements and escalating social protests (Scheba and Turok 2020). Such events focus public attention on the downsides of urbanization, while the less visible benefits are overlooked. Many land occupations occur on hazardous sites and perpetuate the haphazard structure of cities. Little has been done to plan ahead for urbanization by making serviced land available in advance of human settlement (SACN 2014; Turok 2016a; Gardner 2018).

30.2.2 Urban Policies and Politics Local government consolidation during the 1990s signalled a promising future, as the eight largest cities were equipped with more robust administrations than the previous

650   Ivan Turok 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% 1975 1980 1985 1990 1995 2000 2005 2010 2015 2020 2025 2030 Urban centres

Urban clusters

Rural

Figure 30.1  South Africa’s population distribution, 1975–​2030 Source: European Commission Global Human Settlements Data, https://​ghsl.jrc.ec.europa.eu/​. Note: Urban centres are cities with over fifty thousand inhabitants in contiguous dense areas; urban clusters are towns with over five thousand inhabitants in contiguous semi-​dense areas; rural areas consist mostly of low-​density areas.

patchwork. Single-​tier metropolitan authorities were given wide territorial boundaries to reflect their functional urban areas (including surrounding commuting zones) and to prevent leapfrog development into neighbouring localities. The metros incorporated the outlying townships and affluent suburbs to permit effective redistribution, with vital powers to raise property taxes for reinvestment in public services. They were required to prepare Integrated Development Plans to cover the five-​year electoral cycle, in conjunction with other spheres of government to ensure alignment of plans and budgets. Metro responsibilities were broader than their predecessors, although housing and public transport were assigned to the provinces (SACN 2014). Subsequent problems have stemmed more from harmful political practices and institutional cultures than from deficient structures. Undue political interference, unstable leadership, and excessive central regulation have impaired the developmental agenda originally envisaged, aggravated by inappropriate senior appointments, mismanagement, and a loss of specialist expertise (Palmer et al. 2017; van Ryneveld 2018; Olver 2020; World Bank 2018). Many townships have benefited from improved public services, although the fundamental course of urban development continues to bear a strong imprint of the apartheid past, reflecting inertia in the built environment and the durability of fixed assets. Deep inequalities in labour market earnings and stark neighbourhood contrasts also make it practically difficult to reshape settlement patterns and integrate diverse communities. Cities continue to be encumbered by low residential densities in well-​located areas, poor transport systems, and high mobility costs (SACN 2016; National Treasury 2018;

Urbanization, Agglomeration, and Economic Development  651 World Bank 2018). Hierarchical, command-​and-​control attitudes, and bureaucratic behaviours within government constitute another form of path dependency. Top-​down, sectoral policies take precedence over cross-​cutting spatial plans and the building of stronger relationships with civil-​society organizations, which impedes trust and accountability (Todes and Turok 2018; Duminy et al. 2020; McKenna 2020). This, in turn, hampers the metros’ ability to respond to community concerns and initiatives, including engaging with informal activities. The strength of functional silos within government also inhibits the metros’ capacity to coordinate infrastructure investments and to plan in an integrated way according to the unique economic assets and priorities of each city.

30.2.3 Urban–​Rural Tensions An intangible legacy of the past is the ambivalent attitude of many decision-​makers towards urbanization, and perhaps even a fear of the city (Turok et al. 2021a). The political economy of urban areas is inevitably more complicated and contested than rural areas because of the diverse interests involved. Meanwhile, the provinces have become more powerful than envisaged in the Constitution, as the main political parties have adopted federal structures along provincial lines. The ruling party originally envisaged strong metros to expedite socio-​economic development and democratic accountability, but provincial leaders have tended to usurp their authority. The rural provinces have acquired disproportionate influence through their superior ability to mobilize party membership and popular votes, reflecting the widespread dependence of jobs and livelihoods on state patronage in these regions (van Ryneveld 2020). A risky and unhealthy misalignment has arisen between the locus of power within the party and the location of the economy in the big cities. It has aggravated the underlying contradiction at the heart of government between the desire of politicians to spend public resources (preferably on popular necessities) and the need to help generate those resources through supporting (and then taxing) productive activities. During the Zuma era there was very little backing for cities as strategic economic units worthy of special attention (Duminy et al. 2020). The narrative of an urban–​rural divide prevailed, highlighting countryside hardships. The emphasis on rural areas in land reform policy (covering redistribution, restitution, and tenure security) is a good example, ignoring the centrality of urban land reform to building more prosperous and equitable cities (PAPLRA 2019). Persistent equivocation over urbanization is reflected in sparse economic data at the city level, and meagre state funding for systematic or sustained urban research. Provincial economic data are more comprehensive, yet these are mostly administrative units rather than functional areas. The national statistics agency does not even have a robust definition of urban areas; it continues to use an old administrative categorization. Data deficiencies help to explain the limited awareness of spatial dynamics among the country’s economists, despite flourishing interest elsewhere in the world. One exception is the Cities Support Programme (CSP) located within the National Treasury. Since 2012, the CSP has sought to improve metro capabilities to carve out their

652   Ivan Turok own agendas by introducing new skills, knowledge, and decision-​making systems. These encourage city plans and developments that are more integrated, inclusive, and resource efficient (Turok 2016a; Duminy et al. 2020). There are some positive initiatives in the pipeline, although few changes on the ground as yet.

30.3  Spatial Concentration and the Performance of Firms 30.3.1 Theoretical Advantages of Agglomeration There is a substantial body of knowledge on the economic advantages of agglomeration stretching back to Alfred Marshall. It has grown to encompass multiple perspectives rather than constituting a singular theoretical framework. One interpretation stems from the so-​called New Economic Geography (NEG) (Krugman 2011)  and another from established economic geography (e.g. Storper 2011). The essential ideas in these and other approaches proceed as follows. Firms and workers that depend on each other for inputs and outputs cluster together for ease of access. They interact in many different ways, creating a dynamic system with many synergies and cumulative effects (Glaeser 2011; Storper 2013). Dense networked cities promote entrepreneurship, creativity, business dynamism, knowledge spillovers, larger markets, and faster growth (Jacobs 1984; Glaeser and Xiong 2017). Firms specialize in particular products or tasks, which amplifies their capabilities and productivity. Many cities also benefit from specializing in particular functions (‘localization economies’) and exchanging outputs with other places (Lall et al. 2017; Iammarino et al. 2019). ‘Tradable’ goods and services are crucial because their growth is not constrained by local demand and they often generate sizeable multiplier effects. NEG agglomeration theory emphasizes increasing returns to scale as the main causal mechanism. This gives metropolitan areas exceptional advantages over smaller cities (World Bank 2009). Major cities offer businesses deep labour pools, a large choice of suppliers and customers, and extensive professional services and communications infrastructure. This suits the biggest firms and generates the greatest returns on public and private investment. Researchers have successfully quantified these benefits using econometric techniques, finding that doubling the city size raises productivity by between 2 and 5 per cent (Melo et al. 2009; Graham and Gibbons 2019). Furthermore, the benefits of proximity to the main hubs of economic activity decay rapidly with distance, which suggests that low density sprawl undermines economic vitality by lengthening commuting times and limiting business interactions (Rice et al. 2006). NEG theory postulates that national living standards benefit from concentrating activity in the largest cities and that wider disparities with other regions are the price to pay for these productivity gains (Glaeser 2011). Governments should not try to diffuse

Urbanization, Agglomeration, and Economic Development  653 activity to smaller cities and towns by investing in place-​based policies (World Bank 2009). Agglomeration economies should provide sufficient benefits for aggregate growth and national income to compensate places left behind and leave everyone better off. The theory also proposes that people be encouraged to migrate from poorer regions to the large cities. In other words, the solution to geographical inequalities is regional integration and factor mobility (World Bank 2009). Improved transport links will diminish the spatial frictions of distance and division, reduce the surplus rural population, and fuel the metropolitan growth machine. More expansive perspectives challenge these ideas as reductionist and oversimplified, especially the argument that bigger is necessarily better (Iammarino et al. 2019; Sunley et al. 2020). NEG portrays agglomeration economies as having universal applicability and an almost law-​like quality, regardless of the context. Alternative approaches view agglomeration as a potent force, but neither immutable not deterministic. The prosperity of cities is driven by various factors besides their size and density (including their economic structure and social diversity), implying that second-​order cities can also thrive. Social and political institutions are crucial in mediating agglomeration economies—​either helping or hindering their realization (Feldman and Storper 2018). For example, there are stronger traditions of business collaboration, information sharing, and mutual learning in some cities than in others (Storper et al. 2015). Countries with a long history of racial division and mistrust are bound to face a greater hindrance in this respect. Diseconomies of scale cannot be ignored either, such as congestion, overloaded infrastructure, and social discord. Local public institutions can alleviate certain growth pressures through far-​sighted spatial planning and active urban management (OECD 2018; Turok 2017). The crucial point is that agglomeration processes don’t operate in the same way in different places, because formal policies, rules, and organizing systems differ, along with informal norms and conventions. Political economy frameworks may help to make sense of these dynamics, including whether the plans of city governments are endorsed or undermined by private interests (such as property developers) and by regional and national spheres of government (Olver 2020). The relevance of agglomeration processes to informality is generally neglected in established theories, but also needs to be factored in. Built-​environment institutions are particularly important in countries experiencing rapid urbanization (Collier and Venables 2015, 2017; World Bank 2013; Turok 2016b). They have the potential to perform multiple functions which strengthen agglomeration economies and limit opportunistic rent-​seeking, wasted resources, and speculative landholding to create artificial scarcity. Municipalities can signal the city’s future growth trajectory to developers through their land-​use plans, zoning schemes, and building regulations. Plans need alignment with government infrastructure investments to give them teeth and credibility. They make possible the efficient configuration of employment nodes, residential areas, and transport connections, which adds enduring value to the occupiers of land by ensuring activities complement each other and avoid erratic development.

654   Ivan Turok City governments perform additional functions in pursuit of this ‘urban premium’ (Turok 2016c). Sensitive design and management of the public realm (from educational facilities and transport systems to public open spaces) is important for human interactions to be harmonious, inclusive, and cooperative, rather than estranged, discordant, and exclusionary (Gieryn 2000). Carefully calibrated regulatory procedures reduce the burden of compliance and costly delays in developing and refurbishing buildings—​essential for the adaptive re​use of property in dynamic economies (Turok et al. 2021b). Many of the benefits of agglomeration are capitalized in land values, so the capacity to understand the incentives and functioning of the property market, and to raise property taxes, is essential to achieve desired outcomes and to fund expanded municipal services. Otherwise much of the value of well-​structured cities accrues to property owners, who become unreasonably wealthy (Collier and Venables 2015; Olver 2020).

30.3.2 South Africa’s Urban Experience The paucity of reliable economic data on city-​level output, per capita income, and labour productivity obstructs systematic research. Statistical deficiencies also blunt policy responses to the distinct economic problems and potential of the major cities. Estimates suggest that South African cities have generally performed better than towns and rural areas, although not well by international standards. The metros, in particular, make a disproportionate contribution to the national economy. They generate 57 per cent of national output and 50 per cent of all formal and informal employment, with 40 per cent of the country’s population (National Treasury 2018). This differential is reflected in the employment rate. Over 52 per cent of adults in the Gauteng metros have a job (despite high in-​migration), 45 per cent in the coastal metros, 43 per cent in the secondary cities, 41 per cent in the commercial farming areas and only 21 per cent in the former homelands (Turok 2018). South Africa’s major cities have outpaced the rest of the country for some time. The rate of economic growth of Gauteng and Cape Town—​the two largest agglomerations—​ averaged 3.5 per cent between 1993 and 2016, compared with 2.7 per cent nationally (National Treasury 2018). Cities also make a disproportionate contribution to public finances, the main source of which is personal income tax (PIT). The share of national PIT paid by the three relatively urbanized provinces of Gauteng, Western Cape, and KwaZulu-​Natal is 75 per cent, with the other six provinces paying the remainder. Gauteng alone pays 46 per cent, with more economic activity, more taxpayers, and more productive, higher-​skilled jobs paying higher salaries and taxes (Turok 2018). The presence in Gauteng of government departments and the headquarters of many state-​owned enterprises is relevant too. Nearly a third (29 per cent) of Gauteng’s jobs are professional, technical, or managerial, compared with only one-​seventh (14 per cent) in the Northern Cape. What’s more, Gauteng residents pay more than twice as much PIT on average as people elsewhere, indicating the higher average earnings and employment levels.

Urbanization, Agglomeration, and Economic Development  655 Whether rural residents become better off by moving to cities is crucial for urbanization. Visagie and Turok (2020a) used a unique dataset that tracks the progress of 30,000 individuals over time through repeat surveys. The decision to migrate pays off for most people, contrary to the popular view that migrants are unskilled and can’t compete for jobs, leading to discontent and despair. Two-​thirds of rural-​to-​urban migrants who were poor in 2008 managed to exit poverty by 2014, adding up to approximately 385,000 people nationwide. These people succeeded in finding some kind of job, even if casual and low paid. Their progress was all the more surprising considering the depressed state of the labour market over this period and the low level of social mobility in the country (Statistics South Africa 2019). Using their own initiative to escape poverty by moving to the city has been one of the few opportunities for advancement open to people living in the countryside.

30.3.3 The Experience of Different Industries The importance of tradable goods and services was outlined earlier. South Africa’s cities have developed distinctive economic capabilities over time that have helped to create jobs and raise incomes. These specializations are poorly understood. Manufacturing has been most important since the contraction in gold mining. The metros have 2.5 times as many manufacturing jobs per resident as the rest of South Africa (National Treasury 2018). The main manufactured exports are motor vehicles, iron and steel, machinery, chemicals, and processed food (World Bank 2018). Compared with many other middle-​ income countries, South Africa’s exports have under performed for years (Bhorat et al. 2019). Urban infrastructure limitations are partly responsible, including inefficient harbours, port logistics, shortfalls in electricity supply, and the relatively high cost of urban operating environments (National Treasury 2018; World Bank 2018). Lack of international competitiveness is one of the reasons why every major city in South Africa has experienced deindustrialization in recent decades. Local exporters have struggled to expand in external markets and foreign imports have displaced local production, leading to business contractions and closures. Manufacturing was particularly important in the cities of Durban, Ekurhuleni (East Rand), and Port Elizabeth, which is why deindustrialization has hit them hardest, with little diversification into other tradable sectors. The overall level of employment in Durban and Port Elizabeth slumped by 15 per cent between 2008 and 2017, and in Ekurhuleni by nearly 10 per cent (SACN 2016). The other large metros were less badly affected. Total employment in Johannesburg increased by 5 per cent over the same period, Cape Town by 8 per cent and Tshwane (Pretoria) by 12 per cent. This is meagre considering the underlying population growth. The 1995–​2008 period was better for South Africa’s cities than the subsequent decade, reflecting the national economic malaise since the Great Recession and the end of the commodities boom (SACN 2016). Poor manufacturing performance in some other comparable countries has been offset by stronger growth in tradable services, covering finance and insurance, business

656   Ivan Turok services, digital technologies, creative industries, tourism, health, and higher education (Hoekman and de Velde 2017; Newfarmer et al. 2019). High-​productivity services seem to benefit more from agglomeration than manufacturing because of the importance of high-​level skills and knowledge spillovers, that is, human interaction in exchanging information and ideas (Graham and Gibbons 2019; Glaeser 2011). There has been some growth in most of these sectors in South African cities, albeit more modest than in other countries, suggesting that South Africa has few globally competitive service firms (Bhorat et al. 2019). Export growth has been stronger for traditional services, such as transport and tourism, than for advanced services. The main source of services growth in Gauteng has been business and financial services, the output of which expanded by a sizeable 7.4 per cent per annum between 1995 and 2017, compared with 3.8 per cent in KwaZulu-​Natal, 2 per cent in the Western Cape, and 2 per cent in the Eastern Cape (Visagie and Turok 2020b). These jobs are predominantly white collar and less accessible than manufacturing to workers without secondary education. Most of these services have a large non-​tradable component, so they probably do not generate the same local spinoffs as manufacturing or mining. The strong growth of financial services has raised broader questions about the phenomenon of financialization and its potential drawbacks for local and national economic development (Fine and Rustomjee 1996; Karwowski et al. 2018). The fact remains that the finance sector generates a striking 40 per cent of total corporate income tax in South Africa, so its contribution to public finances, at least, cannot be neglected. In most cities there has been moderate growth in transport, retail, personal services, private security, and public services. Their productivity is generally rather low and they serve predominantly local markets, so they play little role in spurring broader growth. Recognizing tourism as a tradable sector, the government sought to boost this by hosting the World Cup in 2010. It used the event as a catalyst for investments in public transport, football stadia, and precinct upgrades. There were some benefits to construction and related sectors, but they proved short-​lived in the absence of successor events (Bhorat et al. 2019). Insufficient efforts were made to sustain tourism demand, and most of the stadia have become white elephants. Meanwhile individual cities have had some success in promoting business tourism by building international convention centres. Higher education is another sector with export potential. South African cities have reputable universities that could attract more students from the sub-​continent.

30.3.4 Towards More Granular Analysis Recent work has begun to disaggregate the sectoral analysis more systematically to identify specific strengths of each city using location quotients (Visagie and Turok 2020b). Johannesburg’s economy has become skewed towards financial services, with nearly twice as many jobs as expected, given its size. Finance generated over thirty thousand additional jobs for the city between 2008 and 2017. Johannesburg is the headquarters

Urbanization, Agglomeration, and Economic Development  657 of the country’s major banks, insurance companies, and stock exchange. Most are clustered, along with related professional services (legal, accountancy, IT, and management consultancy) in Sandton, South Africa’s premier commercial precinct located in the northern suburbs, with the costliest real estate in the country. The international component of financial services is illustrated by the five main banks having Rand 600 billion worth of assets tied up in other African countries, representing 12 per cent of their total assets (Thompson and Donnelly 2020). Standard Bank and Absa have the biggest footprints on the continent. The former generates about 31 per cent of its earnings from the rest of Africa and has 3.8 million customers there (Turok and Visagie 2020) The telecommunications sector is similar in some respects to finance. Two South African corporations have expanded beyond the domestic market and become important service providers elsewhere on the continent. There are few other South African corporates producing high-​value tradable services that have become successful exporters (Bhorat et al. 2019). Fast-​growing African cities offer possibilities in sectors such as engineering, real estate, infrastructure, and design. SA companies have proven capabilities in these fields, but seem to have lacked the ambition and/​or appropriate strategies to succeed (Turok and Visagie 2020). Global consolidation risks South Africa’s leading business service firms being acquired by foreign multinationals and ending up as outposts in their international networks. It may not be far-​fetched to envision a SA government committed to city-​building and forming strategic alliances with private-​ sector and NGO expertise to offer a competitive package of professional urban services to enable other African countries to plan and manage urbanization better. Sandton’s experience indicates how agglomeration processes can operate at a localized scale. It also illustrates the significance of active and deliberate management of dense urban spaces. Many of Sandton’s corporate offices moved there from downtown Johannesburg during the 1990s. Similar decentralization trends occurred in Pretoria, Durban, and other smaller cities (Turok et al. 2021b). The shift was prompted by tumultuous changes during the political transition, spurred by many black people moving from the townships to be closer to jobs. There was a positive aspect to people claiming spaces from which they had been excluded. Yet, the proliferation of street hawkers, minibus taxis, and hustlers provoked an exodus of white businesses to the suburbs. Their loss of confidence in the inner city was compounded by municipal service breakdowns and flouting of bylaws while local government was being reorganized. Civic leaders ignored the inherent value of the central city and failed to respond to the opportunity created by thousands of working people eager to live nearby. Instead, hundreds of buildings were literally abandoned and assets written-​off as occupiers and owners fled to suburban nodes, attracted by private-​property interests and rival municipalities. In Durban, for example, the dominant landowner Tongaat Hulett orchestrated a major greenfield commercial precinct in Umhlanga, which has hastened the central business district’s (CBD)’ decline since the 1990s. McKenna (2020) cites a range of other examples of ‘mega-​projects’ in Gauteng that are likely to have similar effects. Tighter planning controls around Cape Town and more responsive institutions governing its

658   Ivan Turok CBD mostly avoided the same fate and actually created a more attractive and upmarket place to invest, work, live, study, and visit. Nevertheless, the sizeable incentives to bypass or short-​circuit the city’s plans and regulations mean that property developers continue to cultivate privileged relationships with political leaders and senior officials in order to get their projects approved (Olver 2020). Where they have succeeded, the results have perpetuated socio-​spatial divisions and encroached upon environmentally sensitive areas and public open spaces. Elsewhere, the level and timing of CBD disinvestment varied from city to city, but signs of renewal emerged within a decade (Turok et al. 2021b). Small entrepreneurial developers observed the pent-​up demand to live centrally and began to convert run-​ down buildings into affordable housing and related uses. They targeted novel markets shunned by the major developers, such as lower-​paid workers, freelancers, students, and young black professionals (CDE 2020). Some organized street cleaning, security services, and landscaping to tackle the physical degradation and decay in their precincts. These initiatives are bringing renewed energy and investor interest to unexpected places. They illustrate that intense settlement stresses the urban environment and needs careful local oversight, including maintaining the infrastructure and governing the public realm. Government indifference towards the interrelated processes of urbanization and densification is a recipe for social division, instability, and the destruction of value.

30.4  Housing and Urban Density Although housing is not a tradable sector, it needs to be planned in relation to the labour market and other urban assets. Urbanization is less productive or inclusive if people moving to cities cannot find affordable places to live that are accessible to workplaces. Maximizing workers’ choice of employers and vice versa improves the fit in terms of skills and aptitudes. Labour-​market matching is a cornerstone of agglomeration, leading to higher productivity, more job stability, and raised incomes. Dispersed cities with long travel-​to-​work distances tend to exclude less-​skilled job-​seekers. Since housing is the largest user of urban land, its location and density have a major bearing on the city’s morphology. Flexible access to a diverse residential stock (including rental options) is important for people to obtain more suitable accommodation as their family circumstances or earnings change. The property market tends to distribute households across the city according to their ability to buy into neighbourhoods with different quality attributes, and property developers reinforce these segmented and segregated patterns. Without state support, the poor will end up in the least desirable places with restricted access to opportunities. For all these reasons, housing needs to be conceived of as intrinsic to a broader vision of compact, integrated, and variegated settlements, and institutionalized accordingly. These are the kinds of places where social and economic outcomes improve over time (Jacobs 1984).

Urbanization, Agglomeration, and Economic Development  659

30.4.1 The Government’s Mass Housing Programme Urban housing was one of the government’s top priorities in 1994. This commitment stemmed from the previous regime denying black people the right to own property in urban areas and calling a halt to housebuilding in the 1980s in an attempt to curtail urbanization. This resulted in overcrowded township housing and mushrooming squatter settlements in places where people could evade eviction. Consequently, decent housing became a prominent demand for social redress and dignity, and a fundamental right inscribed in the Constitution. Yet, a narrow policy response has limited housing’s contribution to building more functional cities that can ensure economic advancement as well as social justice. The government’s chosen approach gave it exceptional control over the quantity and type of housing delivered, with a bold target to produce one million ‘RDP’ (Reconstruction and Development Programme) houses within five years. State-​led provision avoided the need to involve the private sector and to engage with the wider housing system to achieve broader gains. However, accepting full responsibility for solving the problem let the banks, private developers, and even employers off the hook (Savage 2014). RDP housing was conceived as something separate from other residential property—​a form of social compensation instead of the first step on the ladder of an integrated property market. The houses were designed as free-standing starter units, with little funding set aside for the land (Gardner 2018). Qualifying households earning less than R3,500 a month (about half the population) were given these units on condition they did not sell them for eight years. Homeownership was the preoccupation, with little attention to the benefits of renting in urban settings. High expectations were converted into accelerated delivery from approximately 60,000 units in 1994/​95 to about 230,000 five years later. By 2020, over 3 million units had been built altogether, amounting to over 40 per cent of the country’s total housing stock (McKenna 2020). This is a notable achievement, with the sheer scale dominating other formal housebuilding. Households have benefited from living in solid and safer structures with internal services. Yet, housing has been treated essentially as a consumption good, with little consideration for how it might help address the underlying problem families face of economic hardship, and how housing can help to build better cities. The delivery imperative has also been used to justify not devolving housing responsibilities to the metros, where implementing a more holistic approach would be more feasible. The housing programme has not improved people’s skills or connected them to the labour market. Most projects have been large-​scale, monofunctional, and in marginal locations where the whole community is poor and detached from social networks offering useful information about job vacancies (SACN 2014; Turok 2016a; McKenna 2020). The desire to accelerate production meant using large greenfield sites that could be delivered quickly. Many were on the outskirts of existing townships and far from jobs and amenities, thereby entrenching segregation, inequality, and low-​density sprawl

660   Ivan Turok (Gardner 2018; World Bank 2018; CDE 2020). Many households have found themselves trapped in isolated places, unable to sell their homes legally because of the eight-​year rule and a 1.1 million backlog of transferring title deeds. Potential buyers also struggle to get bank mortgages in these areas, so resale prices are sluggish and fall further and further behind suburban houses prices (Savage 2014; Marais et al. 2020). Mass housing has been a classic silo intervention from the centre: restricted in scope, insulated from economic reasoning, and out of sync with city needs. It contradicts the logic of agglomeration, whereby households benefit from the opportunities created by dense economic networks, diverse social connections, and well-​connected infrastructure systems. The programme has also been susceptible to fraudulent tendering, corruption, and substandard construction (GTAC 2016; SACN 2014). Olver (2020) tells the alarming story of how weak checks and balances enabled a cabal of regional politicians and building contractors in Port Elizabeth to misappropriate sizeable housing resources over a lengthy period—​a pattern repeated elsewhere. These are among the reasons why the rate of delivery has fallen throughout the country since 2010, and the housing backlog has risen (Gardner 2018; McKenna 2020). Unfulfilled promises have triggered social unrest, with community protests and land occupations regularly obstructing construction projects.

30.4.2 Backyard Rental Housing Meanwhile, an alternative approach to low-​cost housing has been evolving from the ground up in response to these weaknesses. It involves the adaptive reuse of existing settlements to create higher densities, greater diversity, and more economic opportunities (Scheba and Turok 2020; CDE 2020). Poor homeowners and small-​scale developers have been deploying their modest resources to meet the demand for affordable housing by constructing rental units in their backyards (Gardner 2018). This generates a regular income for their families and demonstrates a viable market for such housing in the accessible parts of townships and inner suburbs. Backyard dwellings are also helping to create local jobs and skills during construction and subsequent maintenance. These features help to explain why backyarding has taken over as the fastest growing housing segment, expanding three​fold in Gauteng over the last fifteen years (Hamann et al. 2018), and with considerable potential for further growth. The burgeoning backyard sector is more responsive to the locational needs of households than mass housing. The supply of decent rental units matches the requirements of many young working people who want flexibility and who can’t afford to purchase their own property, or who don’t qualify for free houses (CDE 2020; Scheba and Turok 2020). A cohort of emerging developers and homeowners are using their own ingenuity and common sense to construct improved accommodation. This has progressed over time from shacks to solid brick and mortar structures with internal ablutions (Turok and Borel-​Saladin 2016). These provide more secure and healthier living conditions for people who can afford a basic rent, but much less than typical

Urbanization, Agglomeration, and Economic Development  661 formal market rentals. There has been further evolution in some localities to two-​storey rental units and small blocks of flats on infill sites. Several private finance firms have acknowledged the potential by providing loans and practical assistance to help people construct these buildings in return for a share of the rent. Otherwise they rely on unsecured personal loans and borrowing from family and friends. Yet, the very success of backyarding is creating other challenges that require attention, including overloaded infrastructure and safety hazards. When ten or more families occupy sites designed for single families, there are heightened risks of electricity breakdowns, water and sewage system failures, and the spread of fires. Most dwellings do not comply with municipal bylaws, zoning schemes, or building regulations because these procedures are too onerous. The government doesn’t know what to do about this, or about backyard housing generally (Scheba and Turok 2020). Public officials assume they can govern these places through formal rules and standards that are just too burdensome. The regulations were designed in different circumstances and aren’t realistic or appropriate in the current context (CDE 2020; McKenna 2020). Some officials try to enforce compliance but are obliged to retreat in the face of community resistance and threats. Most just look the other way and disregard everyday occurrences of non-​compliance. Widespread indifference and neglect risk a vicious cycle of vulnerability, ungovernability, and slum formation as rising population densities overload public services, infrastructure falls into disrepair, public spaces deteriorate, public trust dwindles, and the ability of government to enforce any rules at all is eroded. Municipalities already collect very little from property taxes or service fees in these areas, despite some landlords now earning quite substantial rents. A more favourable scenario would involve the negotiation of some kind of social compact whereby public authorities simplify the regulatory framework, offer constructive assistance, and upgrade their infrastructure services in return for people complying with essential standards and beginning to contribute to local taxes. A purposeful, hands-​on strategy to retrofit, rehabilitate, and maintain townships in this way would create the conditions for more liveable and prosperous neighbourhoods into the future. With guidance and ingenuity, one could also envisage the creation of mixed-​ use high streets and vibrant precincts accommodating all kinds of retail activities, personal services, and workshops on the ground floor of two-​or three-​storey residential units (Charman et al. 2020). External support, including bank financing, could help some backyard developers to upscale and become more significant providers of affordable housing with a bigger impact on the backlog. This scenario requires municipalities to engage more deliberately with the existing processes underway on the ground. Streamlining regulations and offering advice with building design and structural standards would improve the quality of dwellings and provide safeguards for the health and safety of residents. Formal recognition of these new rental properties would also increase their long-​term resale value and bring them into the tax system. National government could help with the regulatory reforms and dedicated funding to expand the bulk infrastructure.

662   Ivan Turok

30.5  Transport and Urban Connectivity An efficient and affordable transport system is vital to connect jobs and housing in ways that maximize the choices available to workers and firms. Spatial mobility also fosters social mobility by expanding job-​seeker horizons and opportunities. The speed, safety, and reliability of transport networks are relevant, including how easily commuters can transfer between different modes. Every city has a different built form and history of transport investment, so devolving policy levers and operational controls to local authorities seems sensible to improve transport planning and system responsiveness. Yet this is complicated in practice, especially with an incoherent transport network in place and limited existing institutional capabilities available. Public transport plays a particularly important role in South African cities because of their disjointed built form and long distances between most townships and economic nodes (Turok 2014). Most poor black households face very lengthy commutes and high travel costs, typically consuming between 20-​40 per cent of their incomes (van Ryneveld 2018). Cities are also burdened by the legacy of a transport system designed to control where people lived and how they travelled. Transport infrastructure was used to separate different groups and marginalize black communities in isolated locations. The system for white communities was dominated by building roads for private car travel. Key decisions were centralized within national and provincial government, where the transport sector was well resourced with strong technical (engineering) capacity to pursue its mandate. There was little competence for transport planning at the city level, confounding any simple objective of devolution. Transport was identified as one of five priority areas for socio-​economic development by the 1994 government, although policy did not change greatly in practice. The historic emphasis on private road-​based transport continued, illustrated by the massive Gauteng Freeway Improvement Project, despite doubts about roadbuilding as a lasting solution to traffic congestion (van Ryneveld 2018). Meanwhile, investment in the commuter rail network continued to be neglected. The Constitution stated that public transport was an overlapping responsibility of the three spheres of government. In practice this seems to have frustrated the formulation of an alternative, more transformative strategy. The most significant initiative in the democratic era was the boost to public transport linked to the 2010 World Cup. It was driven from the centre by the government giving large capital grants to the metros to implement bus rapid transit (BRT) systems, the operating costs of which were to be met by local ratepayers. This was the first major injection of public transport funds to city authorities in the country’s history (van Ryneveld 2018). The idea was imported from Latin America by consultants and intended to become the backbone of each city’s public transport network, absorbing and replacing minibus-​ taxis. However, within a few years it became apparent that BRT was ill suited to SA’s low

Urbanization, Agglomeration, and Economic Development  663 densities and long travel distances. Commuting patterns are tidal rather than two-​way, with little seat turnover during each trip. Consequently, ticket revenues are far lower as a share of operating costs than in cities elsewhere. This has created serious financial burdens for municipalities and provoked doubts about whether BRT can be sustained into the future (Duminy et al. 2020). The whole experience raises question-​marks about the government foisting an inappropriate system onto the cities, and about the competence of the metros in relation to transport planning (van Ryneveld 2018). BRT has become an additional stand-​alone service and achieved little by way of reforming other parts of the transport system. This is reminiscent of the RDP housing programme in standing separate from the wider property market. Current urban mobility arrangements face three additional challenges. First, institutional responsibilities and funding streams are fragmented, which is a serious hindrance to integrating the transport network. Commuter rail is managed by a national agency, which has suffered from mismanagement and corruption for many years (Chipkin and Swilling 2018). The service has deteriorated to the point where it barely functions in several cities, causing increased road congestion. Over the years, there has been no apparent effort to encourage residential development around the stations on land owned by the agency, which would increase ridership and generate income for reinvestment in the service. The bus system is managed by provincial governments based on subsidies provided to private companies. The service is widely considered to be inefficient and the vehicle fleet is outdated (Turok 2014). The 2009 National Land Transport Act advocated transferring public transport functions to the metros, but progress has been very slow (van Ryneveld 2018). Inadequate technical capacity is an obvious obstacle, along with resistance from vested interests and lack of clarity on how this should happen (Duminy et al. 2020). Devolution poses serious financial risks to the metros, although it also offers the potential to transform everyday travel experiences for millions of people and to create better cities. Second, urban transport planning is poorly articulated with land-​use and built environment decisions. This is partly because the transport functions are dispersed across agencies, but also because the transport sector operates in a silo with its own regulations, procedures, professional skillsets, and traditions (Turok 2014). City governments lack the policy instruments and know-​how to coordinate transport investments with housing decisions, spatial plans, land-​use controls, and infrastructure projects. The Cities Support Programme has endeavoured to strengthen municipal capacity in transport planning, which may in due course facilitate devolution (Duminy et al. 2020). To shift inherited settlement patterns and densify the built environment will require more determined efforts to develop infill sites (many of which are publicly owned), to intensify activity in the inner cities and along transport corridors, and to convert and redevelop low-​rise buildings in the inner suburbs for residential purposes (Turok 2016a). Third, public transport is dominated in practice by minibus-​taxis, which traditionally have fallen outside the scope of transport plans and subsidies. Two-​thirds of all public transport trips for work and study use minibus-​taxis—​twice as many as trains and buses combined (van Ryneveld 2018). The taxi industry is more flexible and responsive to

664   Ivan Turok commuter needs than the formal transport system. It also generates many employment opportunities and operates with very little government funding. However, taxis have exploitative labour practices, a poor safety record, and do not operate scheduled services (Charman et  al. 2020). Operating licences are regulated by provincial governments, so the metros tend to exclude taxis from their transport plans. Nevertheless, there is gradual acceptance across government that taxis need to be incorporated into plans for a future hybrid public transport system which would be more flexible and cost-​effective than the current infrastructure-​heavy rail and BRT services (World Bank 2018; Duminy et al. 2020). The transition will be difficult to manage because the industry is currently largely informal, often unruly, and socially powerful, which fits uncomfortably alongside the rule-​bound procedures of the state. Purposeful leadership with an appetite for negotiation and creative problem-​solving will be required to integrate taxis into a superior transport network. Strengthening the role of civil-​society organizations and community groups would help to promote the interests of commuters and not just those of more powerful stakeholders.

30.6 Conclusion There are compelling arguments and increasing evidence from around the world that agglomeration promotes economic dynamism and social progress. The economies of scale and frequent business and human interactions foster learning, expedite labour market matching, and enable the sharing of resources and infrastructure. Yet, large dense cities also generate adverse effects, including congestion, overcrowding, and social discord. These drawbacks tend to be more conspicuous than the benefits and exert a bigger influence on the attitudes of decision-​makers. Consequently, many public officials are sceptical about urbanization and do not appreciate the need—​or lack the appetite—​to actively support, guide, and manage the process. Ambivalence towards urbanization is problematic because it inhibits a common vision and coordinated strategy across government towards building cities. Without anticipatory investment in urban land and infrastructure and dynamic shaping of the public domain, large-​scale urban growth is likely to be detrimental to the economy, society, and environment. A constructive approach that prepares the ground and deliberately engages with the processes at work stands a much better chance of realizing the potential of urbanization to raise productivity, boost private investment, and generate employment. Increasing the capabilities of city governments responsible for planning, regulating, and framing the built environment can make a major contribution to successful urban development. The post-​apartheid government has taken a more neutral stance towards urbanization than the previous regimes. Robust metropolitan authorities were created to manage city growth, but without some of the essential policy levers in land, housing, and transport to densify well-​located areas and improve the spatial form. Efforts to empower the metros

Urbanization, Agglomeration, and Economic Development  665 and hold them accountable for leading the way in national economic development lost momentum after the 1990s. By treating different places even-​handedly, national government has avoided inflaming sensitivities surrounding territorial issues. However, relatively little positive action has been taken to overcome the spatial distortions of apartheid and to harness the developmental possibilities of better structured cities. Policies towards housing and transport have inadvertently perpetuated inefficient and unfair urban forms. This has aggravated the racialized reproduction of poverty and inequality that dates back more than a century. Some of the dilemmas faced are illustrated by the growth of urban informality—​in the labour market, housing, transport, and other sectors. On the one hand, informal modes of provision show signs of dynamism, creativity, and adaptation to the requirements of poor communities. Grassroots initiatives, energy, and capacity are being mobilized to meet everyday needs, exposing weaknesses in many formal policies in the process. On the other hand, many of those involved as workers, households, and service users lack important safeguards and protections because of governance limitations in many informal systems. In view of the capacity constraints facing the state, informal practices need to be recognized and ways found to strengthen and regularize them. This requires a shift in approach from regulatory compliance and top-​down delivery methods, towards more of an enabling, capacity-​building role. Municipalities need to get more closely involved in understanding developments on the ground and working together with civil ​society actors and interests on joint projects to chart the way forward. Experimenting with different forms of practical assistance for informal enterprises and new governance arrangements is likely to be important, along with more research, evaluation, and evidence collection.

References Ahlfeldt, Gabriel, and Elisabetta Pietrostefani. 2019. ‘The economic effects of density’, Journal of Urban Economics, 111: 93–​107. Bank, Leslie, Dorrit Posel, and Francis Wilson (eds). 2020. Migrant Labour after Apartheid. Cape Town: HSRC Press. Bhorat, Haroon, Christopher Rooney, and Francois Steenkamp. 2019. ‘Understanding and characterizing the services sector in South Africa’, in R. Newfarmer et  al. (eds) Industries without Smokestacks:  Industrialisation in Africa Reconsidered. Oxford:  Oxford University Press. Centre for Development and Enterprise (CDE). 2020. Building Better Cities. Johannesburg: CDE. Charman, Andrew, Leif Peterson, and Thireshen Govender. 2020. Township Economy: People, Spaces and Practices. Cape Town: HSRC Press. Chipkin, Ivor, and Mark Swilling. 2018. Shadow State:  The Politics of State Capture. Johannesburg: Wits University Press. Collier, Paul, and Anthony Venables. 2015. ‘Housing and urbanisation in Africa: Unleashing a formal market process’, in Edward Glaeser and Abha Joshi-​Ghani (eds) The Urban Imperative: Towards Competitive Cities. Oxford: Oxford University Press.

666   Ivan Turok Collier, Paul, and Anthony Venables. 2017. ‘Urbanisation in developing countries:  The assessment’, Oxford Review of Economic Policy, 33(3): 355–​72. Duminy, James, Susan Parnell, and Mercy Brown-​Luthango. 2020. Supporting City Futures: The Cities Support Programme and the Urban Challenge in South Africa. Cape Town: African Centre for Cities. Duranton, Gilles, and Diego Puga. 2020. ‘The economics of urban density’. NBER Working Paper no. 27215, Cambridge, MA. Feldman, Maryann, and Michael Storper. 2018. ‘Economic growth and economic development: Geographical dimensions, definition and disparities’, in Gordon Clark et al. (eds) The New Oxford Handbook of Economic Geography. Oxford: Oxford University Press. Fine, Ben, and Zavareh Rustomjee. 1996. The Political Economy of South Africa. Johannesburg: Wits University Press. Gardner, David. 2018. ‘South African urbanisation review: Analysis of the human settlement programme and subsidy instruments’. National Treasury, Cities Support Programme, Pretoria [online] www.csp.treasury.gov.za/​Resource%20_​Centre/​Conferences/​Pages/​CSP-​ Tools.aspx. Gieryn, Thomas. 2000. ‘A space for place in sociology’, Annual Review of Sociology, 26: 463–​96. Glaeser, Edward. 2011. Triumph of the City. London: Penguin. Glaeser, Edward, and Abha Joshi-​Ghani (eds) 2015. The Urban Imperative: Towards Competitive Cities. Oxford: Oxford University Press. Glaeser, Edward, and Wentao Xiong. 2017. ‘Urban productivity in the developing world’, Oxford Review of Economic Policy, 33(3): 373–​404. Graham, Daniel, and Stephen Gibbons. 2019. ‘Quantifying wider economic impacts of agglomeration for transport appraisal: Existing evidence and future directions’, Economics of Transportation, 19: 100–​21. GTAC. 2016. Programmatic and costing comparison of the housing programmes. Pretoria: Government Technical Advisory Centre. www.gtac.gov.za. Hamann, Christian, Thembani Mkhize, and Graeme Gotz. 2018. ‘Backyard and informal dwellings (2001–​2016)’, Gauteng City Region Observatory Map of the Month [online] www. gcro.ac.za. Harrison, Philip, and Tanya Zack. 2012. ‘The power of mining:  The fall of gold and rise of Johannesburg’, Journal of Contemporary African Studies, 30(4): 551–​70. Hoekman, Bernard, and Dirk Willem te Velde. 2017. Trade in Services and Economic Transformation. London: Overseas Development Institute. Iammarino, Simona, Andrés Rodriguez-​Pose, and Michael Storper. 2019. ‘Regional inequality in Europe:  Evidence, theory and policy implications’, Journal of Economic Geography, 19(2): 273–​98. Ingram, Graham, and Yu Hung Hong (eds). 2012. Value Capture and Land Policies. Cambridge, MA: Lincoln Institute for Land Policy. Jacobs, Jane. 1984. Cities and the Wealth of Nations:  Principles of Economic Life. New  York: Random House. Jones, Patricia, Paul Collier, and Dewi Spijkerman. 2020. ‘Cities as engines of growth: Evidence from a new global sample of cities’, Journal of Applied Business and Economics, 22(2): 158–​88. Karwowski, Ewa, Ben Fine, and Samantha Ashman. 2018. ‘Introduction to the special section “Financialisation in South Africa”, Competition and Change, 22(4): 383–​7. Krugman, Paul. 2011. ‘The new economic geography, now middle-​aged’, Regional Studies, 45(1): 1–​7.

Urbanization, Agglomeration, and Economic Development  667 Lall, Somik, Vernon Henderson, and Anthony J. Venables. 2017. Africa’s Cities: Opening Doors to the World. Washington, DC: World Bank. Marais, Lochner, Bankole Awuzie, Jan Cloete, Mark Napier, and Erica Greyling. 2000. ‘A tale of two markets: Unequal access to private property in a South African city’, Tijdschrift voor Economische en Sociale Geografie, 111(1): 80–​98. McKenna, Leila. 2020. Epilogue: Integration and Spatial Transformation of South African Cities. Johannesburg: Centre for Affordable Housing Finance in Africa. Melo, Patricia, Daniel Graham, and Robert Noland. 2009. ‘A meta-​analysis of estimates of urban agglomeration economies’, Regional Science and Urban Economics, 39(3): 332–​42. National Treasury. 2018. ‘South African urbanisation review:  Managing urbanisation to achieve inclusive growth, overall report’. National Treasury, Cities Support Programme, Pretoria [online] www.csp.treasury.gov.za/​Resource%20_​Centre/​Conferences/​Pages/​CSP-​ Tools.aspx. Newfarmer, Richard, John Page, and Finn Tarp (eds) 2019. Industries without Smokestacks: Industrialisation in Africa Reconsidered. Oxford: Oxford University Press. OECD (Organisation for Economic Cooperation and Development). 2014. OECD Regional Outlook 2014: Regions and Cities: Where Policies and People Meet. Paris: OECD Publishing. OECD. 2018. Rethinking Urban Sprawl. Paris: OECD Publishing. OECD/​European Commission. 2020. Cities in the World: A New Perspective on Urbanisation. Paris: OECD Publishing. Olver, Crispian. 2020. ‘Power, institutions and rents in two South African cities’, Area Development and Policy [online] https://​doi.org/​10.1080/​23792949.2020.1793680. Palmer, Ian, Nishendra Moodley, and Susan Parnell. 2017. Building a Capable State: Service Delivery in Post-​Apartheid South Africa. London: Zed Press. Pike, Andy, Andres Rodriguez-​Pose, and John Tomaney. 2017. Local and Regional Development. London: Routledge. Porter, Michael. 2003. ‘The economic performance of regions’, Regional Studies, 37: 549–​78. Presidential Advisory Panel on Land Reform and Agriculture (PAPLRA). 2019. Final Report [online] https://​w ww.gov.za/​d ocuments/​f inal-​report-​presidential-​a dvisory-​p anel​land-​reform-​and-​agriculture-​28-​jul-​2019-​0000. Rice, Patricia, Anthony Venables, and Eleonora Patacchini. 2006. ‘Spatial determinants of productivity:  Analysis for the regions of Great Britain’, Regional Science and Urban Economics, 36(6): 727–​52. Savage, David. 2014. ‘Public financing for housing’, Haroon Bhorat et  al. (eds) The Oxford Companion to the Economics of South Africa. Oxford: Oxford University Press. Scheba, Andreas, and Ivan Turok. 2020. ‘Informal rental housing in the South: Dynamic but neglected’, Environment and Urbanisation [online] doi.org/​10.1177/​0956247819895958. Scott, Allen, and Michael Storper. 2015. ‘The nature of cities: The scope and limits of urban theory’, International Journal of Urban and Regional Research, 39(1): 1–​16. South African Cities Network (SACN). 2014. From Housing to Human Settlements: Evolving Perspectives. Johannesburg: SACN. South African Cities Network (SACN) 2016. State of the Cities Report 2016. Johannesburg: SACN [online] https://​www.sacities.net/​state-​of-​cities-​reports-​2/.​ Spence, Michael, Patricia Annez, and Robert Buckley (eds) 2009. Urbanisation and Growth. Washington, DC: World Bank. Statistics South Africa. 2019. ‘Inequality trends in South Africa: A multidimensional diagnostic of inequality’, Report no. 03-​10-​19, Pretoria.

668   Ivan Turok Storper, Michael. 2011. ‘From retro to avant-​garde: A commentary on Paul Krugman’s “The new economic geography, now middle-​aged”, Regional Studies, 45(1): 9–​15. Storper, Michael. 2013. Keys to the City: How Economics, Institutions, Social Interaction, and Politics Shape Development. Princeton, NJ: Princeton University Press. Storper, Michael, Thomas Kemeny, Naji Makarem, and Taner Osman. 2015. The Rise and Fall of Urban Economies: Lessons from San Francisco and Los Angeles. Palo Alto, CA: Stanford University Press. Sunley, Peter, Ron Martin, Ben Gardiner, and Andy Pike. 2020. ‘In search of the skilled city: Skills and the occupational evolution of British cities’, Urban Studies, 57(1): 109–​33. Thompson, Warren, and Lynley Donnelly. 2020. ‘Prudential authority undaunted by challenges facing SA banks in Africa’, Business Day, 24 April. Todes, Alison, and Ivan Turok. 2018. ‘Spatial inequalities and policies in South Africa: Place-​ based or people-​centred?’, Progress in Planning, 123: 1–​32. Turok, Ivan. 2014. ‘South Africa’s tortured urbanisation and the complications of reconstruction’, in George Martine and Gordon McGranahan (eds) Urban Growth in Emerging Economies: Lessons from the BRICS. London: Routledge. Turok, Ivan. 2016a. ‘South Africa’s new urban agenda: Transformation of compensation?’, Local Economy, 31(1): 9–​27. Turok, Ivan. 2016b. ‘Getting urbanisation to work in Africa:  The role of the urban land-​ infrastructure-​finance nexus’, Area Development and Policy, 1(1): 30–​47. Turok, Ivan. 2016c. ‘Housing and the urban premium’, Habitat International, 54: 234–​40. Turok, Ivan. 2017. ‘Urbanisation and development: Reinforcing the foundations’, in Gautam Bhan, Smita Srinivas, and Vanessa Watson (eds) The Routledge Companion to Planning in the Global South. Oxford: Routledge. Turok, Ivan. 2018. ‘Worlds apart:  Spatial inequalities in South Africa’, in Mike Smith (ed.) Confronting Inequality: The South African Crisis. Johannesburg: Jacana Media. Turok, Ivan, and Jackie Borel-​Saladin. 2016. ‘Backyard shacks, informality and the urban housing crisis in South Africa:  Stopgap or prototype solution?’, Housing Studies, 31(4): 384–​409. Turok, Ivan, and Gordon McGranahan. 2013. ‘Urbanisation and economic growth:  The arguments and evidence for Africa and Asia’, Environment and Urbanisation, 25(2): 465–​82. Turok, Ivan, and Justin Visagie. 2020. ‘Building malls or metros? South Africa’s exports of tradable urban services to the rest of Africa’. UNU-​Wider Working Paper no. 2020/​94, Helsinki. Turok, Ivan, Andreas Scheba, and Justin Visagie. 2021a. ‘Questioning urban pessimism:  A decade of HSRC research on cities’, in Crain Soudien, Sharlene Swartz, and Greg Houston (eds) A History of the HSRC in South Africa. Cape Town: HSRC Press. Turok, Ivan, Leanne Seeliger, and Justin Visagie. 2021b. ‘Restoring the core: Central city decline and transformation in the South’, Progress in Planning, 144(6): 100434. United Nations. 2016. New Urban Agenda. Resolution 71/​256, adopted by the General Assembly on 23 December, New York. United Nations. 2020. ‘Shaping the trends of our time.’ Report of the UN Economist Network, New York. United Nations, Department of Economic and Social Affairs (UNDESA). 2018. World Urbanization Prospects: The 2018 Revision. New York: UNDESA, Population Division. van Ryneveld, Philip. 2018. ‘South African urbanisation review:  Urban transport analysis’. National Treasury, Cities Support Programme, Pretoria [online] www.csp.treasury.gov.za/​ Resource%20_​Centre/​Conferences/​Pages/​CSP-​Tools.aspx.

Urbanization, Agglomeration, and Economic Development  669 van Ryneveld, Philip. 2020. ‘This is why control over provincial government is so vital to the ANC’, Business Live, 7 December [online] https://​www.businesslive.co.za/​bd/​opinion/​2020-​ 12-​07-​this-​is-​why-​control-​over-​provincial-​government-​is-​so-​vital-​to-​the-​anc/​. Visagie, Justin, and Ivan Turok. 2020a. ‘Rural-​urban migration as a means of getting ahead’, in Leslie Bank et al. (eds) Migrant Labour after Apartheid. Cape Town: HSRC Press. Visagie, Justin, and Ivan Turok. 2020b. ‘The economic landscape of South African cities:  Diversity across space, sector and skills’. [online] https://​www.sacities.net/​wp-​ content/​uploads/​2020/​03/​Urban-​Economies-​Papers.pdf. Wilson, Francis. 1972. Labour in the South African Gold Mines 1911–​1969. Cambridge: Cambridge University Press. World Bank. 2009. Reshaping Economic Geography. Washington, DC: World Bank. World Bank. 2013. Planning, Connecting, and Financing Cities—​ Now. Washington, DC: World Bank. World Bank. 2018. An Incomplete Transition:  Overcoming the Legacy of Exclusion in South Africa. Washington, DC: World Bank.

PA RT   I V

T H E L A B OU R M A R K E T, DI ST R I BU T ION , A N D S O C IA L   P OL IC Y

Chapter 31

Changing Dyna mi c s in the Sou th A fri c a n L ab our Ma rket Haroon Bhorat, Ben Stanwix, and Amy Thornton

31.1 Introduction At the centre of many of South Africa’s major economic problems sits the country’s labour market, where the most striking feature is a narrow unemployment rate that is approaching 30 per cent and has remained above 20 per cent for the last twenty-​ five years (StatsSA 2019). Unemployment in South Africa is a long-​run structural challenge—​with deep historical origins that are evident in the structure, level, and nature of economic growth. The large numbers of people who cannot find work, and thus earn zero incomes, underpin rates of household poverty that remain stubbornly high and have not shifted dramatically since the end of apartheid. Similarly, extreme levels of income inequality persist, driven primarily by growing labour market inequality, which is estimated to account for between 85 and 91 per cent of total income inequality in the country (Leibbrandt et al. 2010, 2012). In the face of such enduring unemployment, poverty, and inequality then, what are some of the major changes that have taken place in the labour market over the last two decades, and how can these help to make sense of South Africa’s tepid post-​apartheid economic growth and development trajectory? In this chapter we draw on existing research, and our own analysis of household survey data, to examine three broad areas of labour market transition. First, we document a trend of de-​industrialization and skills-​biased change. This involves quantifying

674    Haroon Bhorat, Ben Stanwix, and Amy Thornton a relatively rapid shift in the structural composition of employment and output, where we observe an expansion of employment in skilled occupations, and growth that is concentrated almost exclusively in the tertiary sectors. Second, we analyse changing labour market returns and highlight a U-​shaped pattern of wage growth driven by real gains at the top and, to a lesser extent, the bottom of the distribution, with negative wage growth for workers in the middle. Wage growth at the bottom of the distribution is likely owed to minimum wages, but other explanations for this U-​shaped pattern relate to occupational shifts and changes in the returns to education. We also allude to the rise of the public sector and a public-sector union class that commands a considerable wage premium—​a fundamentally new feature of the domestic labour market. Finally, a series of policy interventions that include progressive labour market regulation, the introduction of minimum wages, and several active labour market policies, have reshaped the regulatory structure of the labour market in important ways. Yet, despite these efforts to improve employment security and wages for vulnerable workers, evidence suggests that the violation of labour market regulations is widespread, and that high unemployment, declining union membership in the private sector, and limited enforcement, undermine the bargaining power of workers in South Africa.

31.2 Data The analytical work in this chapter makes use of nationally representative household survey data from Statistics South Africa (StatsSA) for the 2000–​19 period. Specifically, for 2000–​07 we use the Labour Force Survey (LFS) which was a cross-​sectional, household survey, conducted biannually. From 2008 to the present, we use the Quarterly Labour Force Survey (QLFS), which replaced the quarterly LFS. Both surveys contain individual-​and household-​level data on individuals aged fifteen years or older and include a wide array of demographic and labour market information. Earnings data for the QLFS series is released once a year in the form of the Labour Market Dynamics data (LMD). These three datasets are combined in the Post-​Apartheid Labour Market Series (PALMS) version 3.3, a stacked cross-​sectional dataset created by DataFirst at the University of Cape Town (UCT). The PALMS is a harmonized dataset of microdata from sixty-​nine household surveys conducted by StatsSA between 1994 and 2019 and allows for a more sophisticated treatment of key labour market variables, especially when using survey weights to assess changes over time. As made clear by Wittenberg (2017a, 2017b) there are important structural breaks in the available household survey data that is collected in the PALMS. This is particularly problematic for information on wages, and as such we choose to begin the majority of our analysis from the year 2000 onwards, where the data are relatively more reliable. Note also that the earnings data in this version of PALMS ends in 2017, which was the latest release of the LMD data when the version was compiled.

Changing Dynamics in the South African Labour Market    675

31.3  Trends in Output, Employment, and Earnings Over the last two decades there has been a structural shift in the South African economy, which mirrors trends observed in the rest of the world. This pattern of structural transformation involves a shift away from the primary and secondary production that has long characterized South Africa’s economic base, toward services-​related activities in the tertiary sector (Nattrass and Seekings 2011; Andreoni and Tregenna 2018). For much of the twentieth century, agriculture, mining, and later manufacturing, were the backbone of the country’s economy (Feinstein 2005). Together these sectors accounted for about half of both value a​ dded to GDP and employment in the 1960s (Bhorat et al. 2020a). This began to shift in the late 1980s as the protectionism that had defined the apartheid economy was reduced, and eventually dismantled during a period of rapid trade liberalization during the 1990s (Edwards and Lawrence 2008). These changes, in combination with industry-​specific factors and global trends, led to a steady decline in the contribution of the primary and secondary sectors to South Africa’s GDP and in turn, to the country’s economic growth trajectory. Since 2000, a variety of low-​and high-​productivity services sectors have risen to prominence as the main engines of both GDP and employment growth in South Africa. To document these shifts Figure 31.1, below, plots GDP and employment growth between 2000 and 2019, by main sector. The x-​axis measures contribution to GDP growth and the y-​axis employment growth. Each bubble is weighted by the size of employment in 2019, and sectors that lie above the 45-​degree line have seen labour-​ absorptive growth—that is, employment growth has exceeded output growth. As the figure shows, the only significant sectors in this category are Community, Social, and Personal Services (CSP) and Financial and Business Services (FIN). The Utilities (UTIL) sector also reports positive growth, but only accounts for a small share of total employment and is primarily comprised of parastatals. Of the five sectors exhibiting positive value ​added to GDP and employment, four are services related—​CSP, FIN, Transport (TRNPRT), and Wholesale and Retail Trade (WRT). The only other sector in this category is Construction (CONST), which itself contains several services-​based sub​ sectors. In contrast, the aggregate decline of Manufacturing (MANU), Mining (MIN), and Agriculture (AGRI) is evident, characterized by a combination of negative employment growth and marginal contributions to GDP. For the once-​dominant mining sector, growth in employment and value ​added to GDP has been zero, on average, across the period. The sharp rise of the services sector at the aggregate level has come as a consequence of two separate trends. First, the rapid growth of financial and business services, which has absorbed a large share of the economy’s scarce, high-​skilled individuals. And second, a growth in low-​paid and relatively low-​skilled service jobs, such as security guard work, care work, or casual manual work, which have grown at least partly because these types

676    Haroon Bhorat, Ben Stanwix, and Amy Thornton

Annual average growth rate of employment (%)

6 FIN 4

UTIL

CONST CSP

2

TRNPRT

WRT

0

MANU

MIN

AGRI –2 –1

0

1

2

3

4

5

6

Annual average growth of value added to GDP (%)

Figure 31.1  Value a​ dded and employment growth by sector, 2000–​19 Source: GDP data from the South African Reserve Bank; employment data from PALMS v3.3. Notes: AGRI = Agriculture; MANU = Manufacturing; MIN = Mining; WRT = Wholesale and Retail Trade; TRNPRT = Transport; UTIL = Utilities; CSP = Community, Social, Personal Services; FIN = Financial Services; CONST = Construction. Employment data weighted using survey weights for a sample of all employed, excluding self-​ employed agricultural workers.

of services cannot be offshored or outsourced to mechanized production systems. What we observe then, in contrast to the so-​called Asian Tiger model of manufacturing-​led growth, is that South Africa has followed a pattern of what could be seen as ‘premature deindustrialization’ (Palma 2005; Rodrik 2016). The manufacturing sector has never emerged as a major driver of growth in South Africa—​and instead what we observe is that a diverse range of services has become dominant (Andreoni and Tregenna 2018; Bhorat et al. 2020a). To crystallize this point we look more closely at employment data for the main industry and occupation codes in South Africa between 2000 and 2019, in Table 31.1 and Table 31.2, respectively. Industries are categorized into the standard three groupings (primary, secondary, and tertiary), and occupations are divided by skill level. Looking at Table 31.1, we estimate that about 4.9 million new jobs were created over the period, and 4.3 million of these (or 89 per cent) were in the tertiary sector. More specifically, the Finance and CSP sectors each accounted for about a third of the total increase, making them by far the most important sources of employment growth. Notably, these employment increases took place in a period when the primary sectors shed close to 300,000 jobs. There are also important intra-​sectoral changes that took place, which are not obvious from the aggregate figures. For example, much of the increase in financial services employment is due to a growth in contract work, formally called Temporary Employment Services (TES), which we discuss in more detail below. In addition, growth of CSP jobs

Changing Dynamics in the South African Labour Market    677 Table 31.1: Sectoral employment distribution and change, 2000–​19 Sectors

Employment Growth 2000–​19

Employment shares (%)

Share of Change

Absolute (000s)

Relative

2000

2000–​19

Primary

–​296.62

–​0.20

12.86

7.22

–​0.06

Agriculture

–​149.33

–​0.16

8.12

4.79

–​0.03

Mining

2019

–​147.29

–​0.27

4.74

2.43

–​0.03

Secondary

919.59

0.38

20.85

20.19

0.19

Manufacturing

179.33

0.11

14.15

11.01

0.04

56.83

0.59

0.84

0.93

0.01

Utilities Construction

683.42

1.01

5.86

8.25

0.14

4,392.93

0.58

65.53

72.57

0.89

Trade

870.27

0.34

22.35

20.94

0.18

Transport

423.93

0.71

5.16

6.18

0.09

Finance

1,653.27

1.74

8.2

15.78

0.34

CSP

1,491.35

0.70

18.35

21.91

0.30

–​45.89

–​0.03

11.47

7.76

–​0.01

4,933.55

0.43

100

100

1.00

Tertiary

Domestic services Total

Source: Own calculations using PALMS v3.3. Notes: Adjusted using sampling weights for a sample of all employed excluding self-​employed agricultural workers.

has been bolstered by an expanding public sector, which in 2019 accounted for around 20 per cent of total employment. The expansion of the public sector has played a central role in driving up total union membership, and the public-sector wage bill, in South Africa over this period. We observe then that the growth in services has occurred across different portions of the wage and skill distribution, with implications for how services-​ led growth impacts on both wage and household inequality. Changes at the occupational level require a slightly more nuanced analysis since the patterns are less clearly defined than at the sectoral level. The data show that four occupations—​managerial, clerk, service work, and elementary occupations—​account for 4.2 million, or 85 per cent, of all jobs generated over the 2000–​19 period. What this suggests is that to some extent across the skill categories growth is happening at the poles, where both highly skilled work and elementary occupations have grown relatively quickly. At the upper end of the skill spectrum, growth in managerial occupations has been assisted by the robust expansion of employment in the public sector and the financial and business services sector. Within elementary occupations, it is manual work in mining, construction, and manufacturing that has grown.

678    Haroon Bhorat, Ben Stanwix, and Amy Thornton Table 31.2: Occupational employment distribution and change, 2000–​19 Occupations

High Skill

Employment Growth 2000–​19

Employment shares (%)

Share of Change

Absolute (000s)

Relative

2000

2019

2000–​19

1,255.94

1.08

10.07

14.67

0.25

Managers

912.11

1.46

5.42

9.33

0.18

Professionals

343.84

0.64

4.65

5.34

0.07

Mid–​High Skill

323.65

0.28

10.03

8.99

0.07

Technicians

323.65

0.28

10.03

8.99

0.07

2,066.79

0.36

49.95

47.54

0.42

645.38

0.59

9.49

10.57

0.13

Mid–​Low Skill Clerks Service

1,239.56

0.84

12.75

16.45

0.25

Skilled agriculture

–​328.71

–​0.92

3.11

0.18

–​0.07

Trades workers

372.40

0.23

13.86

11.97

0.08

Operators

138.17

0.11

10.74

8.37

0.03

Low Skill

1,334.63

0.39

29.52

28.78

0.27

Elementary

1,440.50

0.63

19.87

22.66

0.29

–​105.87

–​0.09

9.65

6.12

–​0.02

4,933.55

0.43

100

100

1.00

Domestic workers Total

Source: Own calculations using PALMS v3.3. Notes: Adjusted using sampling weights on a sample of all employed excluding self-​employed agricultural workers.

However, employment has also expanded for services workers and clerks—​both situated toward the middle of the wage distribution. Most service work jobs offer returns that are slightly below the median, while clerks occupy the ‘bottom’ of the top end of the distribution, around the 80–​85 percentile (Bhorat et al. forthcoming). The growth in service work is strongly related to the expansion in personal care occupations for women and protective service occupations for men, which are classified under the CSP sectoral classification. Clerical work is largely a female vocation in South Africa, and growth here is related to the success of the financial and business services sector, as well as wholesale and retail trade (Bhorat et al. forthcoming). Finally, a new dynamic that has accompanied the dramatic shift toward services and away from primary and secondary sector employment has been the increasing casualization of labour (Budlender 2013; Bhorat et al. 2016). Accurately identifying this trend empirically is notoriously difficult because South Africa’s labour market surveys are not designed to measure casual or outsourced work and, as a result, accurate data does not exist. However, labour economists have reached some consensus about a sub-​industry

Changing Dynamics in the South African Labour Market    679 classification in the survey data that can be used as a blunt instrument to trace the expansion of casual work.1 Individuals working for temporary employment agencies, or labour brokers, are classified under this code, but the consistency of this identification is not easy to evaluate so it remains a preliminary exercise. Nevertheless, using this approach, it is possible locate large shares of low-​skilled service workers in jobs that are commonly linked to outsourcing, such as security guards, and office and hotel cleaners, among others. We estimate that the Business NEC sub s​ ector included a total of 1.2 million workers in 2019, having trebled in size since 2000. The sector’s share of employment grew from 2.9 per cent of total employment to 7.7 per cent over the period. As alluded to above, more than 45 per cent of those employed within this category are protective services workers, and an additional 20 per cent are cleaners in hotels or offices. The remaining 30 per cent of workers are from a wide range of occupational categories, none with a share contributing more than 3 per cent of the total. This growing trend of outsourced, temporary work is concerning inasmuch as work of this nature is more precarious, less well-​protected, and generally leaves workers more vulnerable (Budlender 2013). The data suggest that most workers associated with this type of employment are low-​paid, low-​skilled, service workers, making an important connection between casualization and the shift toward services in South Africa. Conventionally, ‘tertiarization’ is associated with advanced economies getting richer, or at least, a higher share of output from services is regularly linked to higher per capita income (Eichengreen and Gupta 2013). The premature deindustrialization that appears to be taking place in South Africa, by contrast, suggests two stylized facts. First, the decline of manufacturing and the rise of services as a share of GDP has meant that the services sector—​in its multiple forms across the productivity spectrum—​has generated the overwhelming majority of all jobs in South African since 2000. Second, this rise in services has led to a possible bifurcation in employment trends—​where services-​driven employment growth has created a more rapid expansion in well-​regulated high-​skilled high-​productivity jobs, but also in increasingly precarious low-​productivity jobs that have grown partly because they cannot be outsourced or replaced by machine or computer technology.

31.3.1 Earnings At the same time as the South African economy has shifted toward services, there have been substantial changes to the national wage structure, which are partly connected to this shift but also influenced by a range of other non-​market factors. To characterize a key feature of these changes, Figure 31.2 plots the average annual growth rate of wages,

1 

Specifically, this relies on the three-​digit Standard Industrial Classification category of ‘Business Not Elsewhere Classified’ (Business NEC, code 889).

680    Haroon Bhorat, Ben Stanwix, and Amy Thornton

Annual average growth rate of wages (%)

3

2

1

0

-1 0

10

20

30

40

50 60 Wage percentile

70

80

90

100

Figure 31.2  Annual average growth rate of employee earnings, 2000–​17 Source: Own calculations using PALMS v3.3. Notes: Adjusted using bracket weights. Annual average growth rate of wages calculated by averaging the year-​to-​year change in wages at each wage percentile for all years between 2000 and 2017.

across the wage distribution, between 2000 and 2017; the latest year of available earnings data in our dataset. The figure shows that returns have grown consistently to the top end of the distribution, and this skills-​biased trend in earnings growth is consistent with the relative labour demand shifts observed above. A range of highly skilled occupations within the services sector has thus seen rapid employment growth and a significant increase in real wages. At the bottom end of the distribution, we also observe positive wage growth over the period. Here the effects of an expanding sectoral minimum wage regime that legislated consistent real wage increases over the period appears to have had a measurable impact. Perhaps most striking, however, is that wages toward the middle of the distribution have experienced the weakest growth. Indeed, we observe negative real growth between the 40th and 65th percentiles. This U-​shaped pattern of wage growth fits with an established narrative of ‘wage polarization’—​a trend that characterized wage growth in many advanced economies in the early to late 2000s, associated with rising inequality (Autor et al. 2003; Goos et al. 2014). This pattern can be at least partially explained by the ‘routine-​biased technical change’ hypothesis, which links job demand to how easily a job can be either automated or replaced by a machine or computer (Firpo et al. 2011). Jobs that involve more routine tasks are thus at higher risk of replacement and are also more commonly located around the middle of the earnings distribution. In South Africa, jobs with a significant component of tasks that can be identified as routine are found in occupational categories such as machine operators and bookkeepers, but also a variety of ‘white-​collar’ administrative jobs that have been

Changing Dynamics in the South African Labour Market    681 replaced to a greater or lesser degree by computers and other mechanized technological processes. At the bottom end of the wage distribution, and in addition to the non-​routine elements of some low-​wage work, there is another factor that protects employment from mechanization and offshoring. This is the degree to which the work must be done on site. Consider, for example, personal care workers and security service workers, which together account for about 738,000 jobs, or 4.5 per cent of total employment, in 2019. This work cannot, or can only to a limited extent, be replaced by computer technology or offshored, and much of it requires a significant on-​site component. These two features protect the existence of such jobs in the middle-​and lower-​skilled occupational categories (Bhorat et al. forthcoming). It is perhaps no coincidence then that in aggregate the sectors and occupations that are expanding are primarily those involving non-​routine tasks, and those with key elements that can only be done on site. While these explanations are plausible, it is unlikely that the shape we observe in Figure 31.2 is driven primarily by routine-​biased technical change, as it has been conceptualized in industrialized countries. There remain a number of other country-​ specific reasons that help to explain the hollowing out of wage growth in the middle of the distribution, not related to the task content of jobs. These include the patterns of structural transformation described above, as well as the changing role of education, the shifting structure of union membership, and the role of minimum wages. We turn our attention to each of these issues below.

31.3.1.1 Returns to Education Over the post-​apartheid period the educational composition of the employed has transformed dramatically, with significant implications for labour market returns. The roll-​out of free access to public schooling has resulted in South Africa boasting some of the highest enrolment rates for its income level (Spaull 2013). The share of employed individuals who achieved a high school qualification rose from 20.8 per cent in 2000 to 33.4 per cent in 2019. By 2019, slightly more than one-​fifth (22.8 per cent) of those employed had some form of post-​secondary educational qualification.2 Critically, quality has not kept pace and one effect of increased enrolment has been to undermine returns to high school education relative to other qualification levels. We observe that those with the lowest and the highest levels of education have experienced the greatest improvement in returns, whilst those with high school and incomplete high school education have seen their returns stagnate for most of the two decades since 2000. The observed growth in returns for the least educated is almost certainly related to many of these workers occupying low-​paid but minimum-​wage-covered jobs, which would have benefited from regularly legislated real wage increases (Bhorat et al. 2020c).

2 

This includes diplomas and certificate-​type qualifications, which can refer to vocational training qualifications.

682    Haroon Bhorat, Ben Stanwix, and Amy Thornton It is clear though that returns have mainly accrued to the most highly skilled. In this regard, Branson et al. (2012) describe what they call a ‘skills twist’, where the swelling of high school graduates, in combination with a deterioration of the quality of the high school certificate, has benefited those able to study further. As such, returns have consistently increased for the most highly educated individuals in the labour market. This has interacted positively with the growth of the financial and business services sector which offers many well-​paying jobs to these highly qualified graduates. Together then we see that sectoral, occupational, and educational trends have come together to ensure concentration of returns to the most skilled. In contrast, a number of factors have combined to undermine returns for high school graduates. First, deteriorating quality in the schooling system has meant that employers have become less convinced about the strength of the signal sent by a high school certificate. And second, the increased supply of high school graduates has dovetailed with the marginal employment and output performance in the secondary sectors described above, where many such graduates would previously have found work. In effect then, a large group of similarly educated (but possibly less-​skilled) individuals are competing for a dwindling pool of blue-​collar jobs and together these factors have served to erode wage growth in the middle of the distribution (Bhorat et al. 2020c).

31.3.1.2 Union Wage Premiums and the Public Sector An important, and growing, segment of the labour market where we observe significant employment and earnings growth is in the public sector, and in particular for public-sector union members. Indeed, union membership in South Africa is increasingly connected to the rise of the public sector, which has shifted the composition of union membership away from the private sector. While the number of private-sector union members grew from 1.6 million in 2000 to 2.1 million in 2019, the private-sector unionization rate has fallen from around one-​third in the 1990s, to less than one-​fifth in 2019 (Kerr and Wittenberg 2019). In contrast, the proportion of the public sector that is unionized ballooned in the 1990s, and since the early 2000s has consistently exceeded 60 per cent. The result is that in 2019 the public sector had a larger total number of union members than the private sector, despite total employment in the public sector only accounting for 20 per cent of total employment. The proportional decline in the share of private-sector union membership is directly linked to employment shifts above, where employment in highly unionized sectors, such as mining, have decreased significantly, whilst casualization of the labour force has increased. This has had implications for the relationship between union membership, public-sector employment, and wages. We illustrate this basic relationship in Figure 31.3 below. The figure presents a plot of wage distributions for union and non-​union members in both the public and private sector. It is clear that across the income distribution, on average unionized workers earn more than non-​unionized workers, with the wage distribution of public-sector unionized workers being furthest to the right. This segmentation is cemented by the modes of the non-​union wage distribution, where these are significantly to the left of the modes of the

Changing Dynamics in the South African Labour Market    683 .6 .5 .4 .3 .2 .1 0 0

5 Public & Union Public & Non-Union

10

15

Non-Public & Union Non-Public & Non-Union

Figure 31.3  Wage distributions by union status and public/​private sector, 2017 Source: Own calculations using PALMS v3.3.

unionized workers’ wage distributions. More detailed econometric work that estimates these wage premia suggests that for unionized workers, the public-sector wage premium is 20.7 per cent (Kerr and Wittenberg 2017). Certainly an important feature of the current labour market then is the wage wedge that now exists between unionized public-sector workers, and other formal sector workers. In summary, the data on earnings suggest that two main sub-​groups in the economy have been able to accrue most of the returns in the labour market. First, those who are highly skilled and who remain a relatively small proportion of the employed. Second, a public-sector union class, which has grown substantially, and suggests that securing government employment commands a significant wage premium. The majority of the labour market, with lower skill levels, have faced continued contractions in the traditionally labour-​absorbing sectors where they would typically find employment. While for many at the bottom of the earnings distribution, the main source of wage growth has likely been due to above-​inflation increases in sectoral minimum wages.

31.4 Policy Interventions 31.4.1 Minimum Wages In addition to the aggregate wage and employment shifts discussed above there are a series of policy interventions that have also played an important role in shaping the South African labour market over the post-​apartheid period. One of the most influential has been the introduction of sectoral minimum wage laws. Starting in 1999, a series

684    Haroon Bhorat, Ben Stanwix, and Amy Thornton of minimum wage laws were gradually introduced in low-​wage sectors with limited union coverage, beginning with contract cleaners, domestic workers, and farm workers, and expanding to cover the wholesale and retail sector, private security, as well as the taxi and hospitality industries. By 2007, these new regulations covered approximately five million workers, and among the poorest 20 per cent of workers, 80 per cent were covered by a minimum wage (DPRU 2016). Crucially, these sectoral minimum wages were increased annually, at rates above inflation, and econometric evidence shows that they significantly raised the average earnings of covered workers. These positive wage effects did not have any observable negative employment effects in general, with the exception of agriculture where job losses did result. At the aggregate, however, consistent real minimum wage growth does appear to have played a role in shaping returns at the bottom end of the earnings distribution (Bhorat et al. 2020). In 2019, minimum wage coverage in South Africa was expanded further through the introduction of a National Minimum Wage (NMW). This new minimum wage is set close to the median of the national wage distribution and as such it applies to almost half of all workers in South Africa. The legislated minimum wage rate of R20 per hour is also a substantial increase on the previous sectoral rates that it supersedes, and in Figure 31.4 we calculate the proportion of employees by sector who earn less than the NMW in 2017. The data show that while on average 46 per cent of workers in South Africa earn less than the legislated hourly rate, there is substantial sectoral variation around this figure. Yet regardless of these sectoral nuances a wage floor that covers half of all workers is extensive when compared internationally. As one example, in the United Kingdom national minimum wage coverage is below 7 per cent (Low Pay Commission 2018). The

.4

Agriculture

Wholesale & Retail

Construction

Transport

Financial Services

Community

Manufacturing

Average

Private households

0

Electricity

.2

Mining

Sub-Minimum Wages (%)

.6

Figure 31.4  Proportion of workers earning below the national minimum wage by sector, 2017 Source: Own calculations using PALMS v3.3.

Changing Dynamics in the South African Labour Market    685 new national minimum wage, and the considerable wage increases required in certain sectors, thus constitute an important break from the arrangements of the past. On the assumption that there are no major negative employment impacts, amidst reasonable compliance with the law, the policy would generate significant additional wage gains for several million low-​wage workers.

31.4.1.1 Labour Market Regulations and Non-​compliance Beyond minimum wages a relatively comprehensive set of laws governing workers’ rights were created in the years immediately following the end of apartheid, and these are set out in different sections of legislation. The Basic Conditions of Employment Act of 1997, for example, details a set of specific employment regulations that apply to low-​ wage workers. These are complemented by minimum wage policies, and together provide a relatively thorough wage and non-​wage legislative foundation that aims to protect vulnerable workers. There is, however, a significant gap between the regulations that exist on paper and the reality that low-​wage workers face. Indeed, the measured incidence of non-​ compliance across the suite of existing labour regulations shows that a substantial proportion of workers do not enjoy the regulatory protections afforded to them by law (Bhorat et al. 2020). As noted above, one of the most significant areas of regulatory violation is earnings, but, violation of non-​wage employment regulations is also widespread. Around 30 per cent of all workers have no access to paid vacation or sick leave, 5.1 million employees report that they are not registered for unemployment insurance, and 2.2 million workers work longer hours than is legally permitted. Overall then we observe that violation levels across both wage and non-​wage legislation is very high on average, but that this varies considerably by a range of demographic and structural factors. The underlying determinants of non-​compliance are not yet well understood, but there are several important factors worth noting that do impact on compliance. In particular, areas with higher rates of unemployment are clearly associated with more extensive underpayment of wages. This points to the fact that in areas with higher unemployment, workers have less bargaining power, making them more vulnerable to wage exploitation. Bhorat et al. (2020) expand on this observation and note that it holds for both wage-​related and non-​wage-related violations. More detailed econometric work also suggests that violation levels are strongly influenced by both firm size and union membership status; violation at the firm level is lower among larger firms and at the individual level is lower for union members (Bhorat et al. 2017).

31.4.2 Active Labour Market Policies At a macro level the combination of low economic growth and skills-​biased structural change has, along with other factors, helped to reinforce high levels of unemployment. In an effort to combat this, the state has turned increasingly to direct job creation efforts,

686    Haroon Bhorat, Ben Stanwix, and Amy Thornton and introduced four relatively large-​scale Active Labour Market Policies (ALMPs) over the last two decades. The first of these is the Expanded Public Works Program (EPWP), introduced in 2004, which is a labour-​intensive demand-​side policy providing temporary work placements for the unemployed. This was followed, in 2010, by the Community Works Program (CWP), which is a sub-​programme of the EPWP and offers employment with the more focused intention of improving community welfare. In many ways the CWP is South Africa’s version of a classic employment guarantee scheme and, for example, includes the creation of jobs for community care workers and teaching assistants. In 2014, the government launched its first wage subsidy scheme—​ the Employment Tax Incentive (ETI)—​which aims to reduce youth unemployment by lowering the cost of hiring younger workers. Finally, while these three programmes are all designed to address long-​term structural problems, in 2009 the government introduced a job re-​training scheme (known as the Training Layoff Scheme) to address the short-​term adverse employment impacts of the 2009 global financial crisis. To gain insight into the comparative efficiency of each of these schemes Figure 31.5 estimates the cost per beneficiary in each case. The EPWP, or public employment scheme, has by far the highest cost per beneficiary, at an average of US$1,810 per work opportunity created. It also provides a large number of work opportunities—​around 780,000 in 2016/​17. The CWP, by contrast is less than a third of the cost, at $562 per work opportunity. This is likely due to the lower administrative burden of administering the CWP, relative to the suite of programmes available under the public employment scheme. In quantitative terms, the job re-​training scheme appears to be the least 1200 000

1 800 1000 000

Cost per Benenficiary (US$)

1 600 1 400

800 000

1 200 1 000

600 000

800 400 000

600 400

200 000

200 0 EPWP Cost per Beneficiary

CWP

Lay off Scheme

Wage Subsidy

Jobs Supported/Created in Most Recent Year

2 000

-

Jobs Supported/Created in Most Recent Year

Figure 31.5  Cost per beneficiary and jobs created by ALMPs in South Africa Source: Bhorat et al. (2019).

Changing Dynamics in the South African Labour Market    687 successful of the ALMPs, at a relatively high cost per beneficiary (US$980) with very few workers retrained under the scheme. This is probably a result of the short time span for which the scheme was implemented. The most successful of the schemes appears to be the wage subsidy, or ETI, which has thus far reached the largest number of beneficiaries at the lowest per capita cost. In total, the ETI supported over one million workers in 2015/​16, at a cost per beneficiary of US$211. However, it is not clear from a descriptive analysis how many of these supported jobs would have been created regardless of the subsidy, although early econometric evidence appears to support the notion that the subsidy has had a positive, albeit small, net employment effect.

31.5 Conclusion The aim of this chapter has been to describe some of the major developments that have shaped the South African labour market over the last twenty years. Perhaps the biggest aggregate shift we observe is the decline of the primary sectors as drivers of employment and economic growth, the stagnation of the secondary sectors, and an increasingly central role being played by the tertiary sector. This trend of services-​led growth is one of premature deindustrialization and has taken place alongside occupational shifts that have favoured the most highly skilled individuals. We note that without any mitigating factors, an increasingly services-​oriented economy means that job prospects for most people in the labour force, who have limited formal training and skills, are not good. Skills-​biased shifts reinforce the problem of structural unemployment in South Africa, where there is a clear mismatch between the skills required in the labour market and those possessed by the majority of unemployed individuals. Labour market returns have also changed in important ways. Aggregate wage growth has been the highest for those at the top of the income distribution, and for those with high levels of formal education, in line with the occupational trends. We observe that returns to education are decreasing for those with only high school qualifications, which includes many individuals in the middle of the distribution. Alongside the effects of structural transformation this may go some way to explaining the hollowing out of wage growth for those in the middle of the income distribution. For low-​wage earners there is a more positive story of relatively small but positive real earnings growth, where it appears minimum wage legislation has to some extent accounted for this trend. Key policy interventions have contributed to a changing labour market. New labour legislation introduced in the post-​apartheid period established a suite of legislative protections for low-​wage workers in particular, including sector-​specific minimum wages. A new NMW, introduced in 2019, builds on this pre-​existing sectoral system but is set relatively high on the national wage distribution, requiring wage increases for close to half of all workers. Critically, despite the well-​developed labour market regulations that provide employment and wage security for workers on paper, the reality is that

688    Haroon Bhorat, Ben Stanwix, and Amy Thornton regulatory non-​compliance is widespread. A final set of interventions worth noting are four relatively ambitious active labour market policies that, in different ways, attempt to decrease unemployment. We suggest that while the results of these programmes are difficult to assess with accuracy, at present, existing evidence suggests that their effects have been rather limited.

References Andreoni, Antonio, and Fiona Tregenna. 2018. ‘Stuck in the middle: Premature deindustrialisation and industrial policy’. CCRED Working Paper no. 11/​2018, Johannesburg [online] https://​ssrn.com/​abstract=3269739. Autor, David, Frank Levy, and Richard Murnane. 2003. ‘The skill content of recent technological change:  An empirical exploration,’ The Quarterly Journal of Economics, 118(4): 1279–​333. Bhorat, Haroon, Aalia Cassim, and Derek Yu. 2016. ‘Temporary employment services in South Africa: Assessing the industry’s economic contribution’. Labour Market Intelligence Partnership: Report no. 28, Human Sciences Research Council and Department of Higher Education and Training, South Africa. Bhorat, Haroon, Kezia Lilenstein, Morné Oosthuizen, Francois Steenkamp, and Amy Thornton. 2020a. ‘Structural transformation, inequality, and inclusive growth in South Africa’, in Kunal Sen, Andy Sumner, and Arief Yusuf (eds) The Developers Dilemma. Helsinki: UNU-​WIDER. Bhorat, Haroon, Kezia Lilenstein, Morné Oosthuizen, and Amy Thornton. (forthcoming). ‘The evolution of earnings and employment in South Africa:  Occupations, tasks and gender, 2000–​2015’, in Kunal Sen (ed.) The Changing Nature of Work and Inequality. Helsinki: UNU-​WIDER. Bhorat, Haroon, Kezia Lilenstein, Morné Oosthuizen, and Amy Thornton. 2020c. ‘Wage polarization in a high-​inequality emerging economy: The case of South Africa’. UNU-​WIDER Working Paper no. 55/​2020 [online] https://​doi.org/​10.35188/​UNU-​WIDER/​2020/​812-​2. Bhorat, Haroon, Ben Stanwix, and Amy Thornton. 2020. ‘Measuring multi-​dimensional labour law violation with an application to South Africa’, British Journal of Industrial Relations [online] https://​doi.org/​10.1111/​bjir.12580. Branson, Nicola, Julia Garlick, David Lam, and Murray Leibbrandt. 2012. ‘Education and inequality: The South African case’. SALDRU Working Paper no. 75, University of Cape Town. Budlender, Deborah. 2013. ‘Private employment agencies in South Africa’. Sector Working Paper no.  291, ILO, Geneva [online] ilo.ch/​wcmsp5/​groups/​public/​—​eddialogue/​ −−−sector/​documents/​publication/​wcms231438.pdf. DPRU (Development Policy Research Unit). 2016. ‘Investigating the feasibility of a national minimum wage for South Africa’. Working Paper no. 201601, DPRU, University of Cape Town. Edwards, Lawrence, and Robert Lawrence. 2008. ‘South African trade policy matters trade performance and trade policy’, Economics of Transition, 16(4): 585–​608. Eichengreen, Barry, and Poonam Gupta. 2013. ‘The two waves of service-​sector growth’, Oxford Economic Papers, 65(1): 96–​123.

Changing Dynamics in the South African Labour Market    689 Feinstein, Charles, 2005. An Economic History of South Africa: Conquest, Discrimination, and Development. Cambridge: Cambridge University Press. Firpo, Sergio, Nicole Fortin, and Thomas Lemieux. 2011. ‘Occupational tasks and changes in the wage structure’. IZA Discussion Paper no. 5542, Bonn, Germany. Goos, Maarten, Alan Manning, and Anna Salomons. 2014. ‘Explaining job polarization:  Routine-​biased technological change and offshoring’, American Economic Review, 104(8): 2509–​26. Kerr, Andrew, and Martin Wittenberg. 2017. ‘Public sector wages and employment in South Africa’. REDI 3x3 Working Paper, Pretoria. Kerr, Andrew, and Martin Wittenberg. 2019. ‘Earnings and employment microdata in South Africa’. UNU-​WIDER Working Paper no. 2019/​47, Helsinki. Leibbrandt, Murray, Ingrid Woolard, Arden Finn, and Jonathan Argent. 2010. ‘Trends in South African income distribution and poverty since the fall of Apartheid’, OECD Social, Employment and Migration Working Papers, No. 101, OECD Publishing, Paris [online] http://​dx.doi.org/​10.1787/​5kmms0t7p1ms-​en. Leibbrandt, Murray, Ingrid Woolard, Arden Finn, and Jonathan Argent. 2012. ‘Describing and decomposing post-​apartheid income inequality in South Africa’, Development Southern Africa, 29: 19–​34. LPC (Low Pay Commission). 2018. ‘National minimum wage’. Low Pay Commission Report [online] https://​assets.publishing.service.gov.uk/​government/​uploads/​system/​uploads/​ attachment_​data/​file/​759271/​National_​Minimum_​Wage_​-​_​Low_​Pay_​Commission_​2018_​ Report.pdf. Nattrass, Nicoli, and Jeremy Seekings. 2011. ‘The economy and poverty in the twentieth century’, in Robert Ross, Anne Mager, and Bill Nasson (eds) The Cambridge History of South Africa. Cambridge: Cambridge University Press. Palma, José Gabriel. 2005. ‘Four sources of de-​industrialisation and a new concept of the Dutch Disease’, Beyond Reforms:  Structural Dynamics and Macroeconomic Vulnerability, 3(5): 71–​116. Rodrik, Dani. 2016. ‘Premature deindustrialization’, Journal of Economic Growth, 21(1): 1–​33. Spaull, Nicholas. 2013. ‘South Africa’s education crisis:  The quality of education in South Africa 1994–​2011’. Report commissioned by the Centre for Development and Enterprise, Johannesburg. StatsSA (Statistics South Africa) 2019. Quarterly Labour Force Survey 2019 Q1–​Q4. Pretoria. Wittenberg, Martin. 2017a. ‘Wages and wage inequality in South Africa 1994–​2011: Part 1–​Wage measurement and trends’, South African Journal of Economics, 85(2): 279–​97. Wittenberg, Martin. 2017b. ‘Wages and wage inequality in South Africa 1994–​2011:  Part  2–​ Inequality measurement and trends’, South African Journal of Economics, 85(2): 298–​318.

Chapter 32

T he You th L ab ou r Ma rket in Sou th   A fri c a Cecil Mlatsheni

32.1 Introduction The current generation of youth in South Africa has arguably the greatest opportunities of any past generation; however, they are also confronted with many challenges. In general, a key objective of a young person is to achieve independence, financial and otherwise. Central to achieving this objective is the labour market. However, the reality is that many young people in South Africa experience difficulty entering the labour market; this is evident from the labour market statistics that will be discussed below. The breakdown in the transition from schooling to work often occurs amidst threats of poverty, compromised health (both physical and mental), teen pregnancy, substance abuse, gangsterism, and crime (Mlatsheni 2012). As such, causality and interrelationships between these challenges and unemployment manifest in complex ways. Youth in South Africa is defined as those individuals between the ages of 15 and 35. This relatively wide age range, by international standards, was appropriate at the time it was set, given the disruption to education as a result of youth activism, which was a feature of the pre-​democratic era, and consequent late starting of schooling and slow progression through the schooling system (Everatt and Sisulu 1992; Truscott 1993; Van Zyl Slabbert 1994; Anderson, Case, and Lam 2001). The standard ILO definition of youth for labour market purposes has an upper bound of 24. Nevertheless, the 15–​ 35 age range is arguably still relevant for South Africa as many young people are unable to gain financial independence well into their thirties. However, cognisance has to be given to the non-​homogeneity of youth, for instance, 15–​24-​year-​olds may face challenges of transitioning from schooling to work while 30–​35-​year-​olds may struggle with establishing a career path to facilitate wealth accumulation. This chapter begins by outlining some of the key features of the South African youth labour market, citing recent statistics and those at notable points in the economy’s

The Youth Labour Market in South Africa    691 progression. At times reference is made to the 15–​24 age category for international comparison. Following that, an account is given of why youth struggle to find jobs in South Africa, after which some key policies and interventions that have been implemented to address youth unemployment are discussed. This is followed by a brief conclusion.

32.2  Key Features of the South African Youth Labour Market The overall unemployment rate has hovered around a quarter of the labour force since 1994. The youth unemployment rate for the 15–​24-​year-​old searching unemployed, for the sake of international comparability, has generally been greater than 50 per cent, whereas it has been around 40 per cent by the South African youth-​age definition of 15–​35. By comparison, the world average youth unemployment rate (aged 15–​24) in 2020 is 13.8 per cent and for Africa, it is 10.5 per cent (ILO 2020). The relatively low youth unemployment rate for Africa is likely a by-​product of the difficulty of capturing labour market performance where there is a high degree of informality (ILO 2020). Furthermore, 36 per cent of youth are searching unemployed while 15 per cent are discouraged. Racial and gender disparities in access to work are entrenched features of the South African labour market. Recent statistics reflect that African youth (aged 15–​35) have the worst unemployment rate compared to the other race groups at 57 per cent, followed by coloured youth at 41 per cent, Indian youth at 27 per cent, and white youth at 15 per cent. Young women (aged 15–​35) experience a higher unemployment rate of 57 per cent while that of young men is 49 per cent (QLFS 2020: Q1). African women have the worst unemployment rate at 62 per cent followed by African men at 52 per cent. Coloured women have an unemployment rate of 40 per cent and coloured men have an unemployment rate of 42 per cent, while white women and men display unemployment rates of 14 per cent and 16 per cent respectively. For the sake of international comparison, the unemployment rate for searching unemployed 15–​24-​year-​old males is 55 per cent and for females it is 63 per cent, while comparable statistics for males in Africa reflect an average unemployment rate of 10.4 per cent and for females 11 per cent (ILO 2020). Youth unemployment was on a path of improvement during the boom years of 2003–​ 08, reaching a low in 2008, but has steadily increased ever since. Currently youth unemployment is at 53 per cent by the 15–​35 age classification. Economic growth has also not approached the admirable levels achieved in the mid-​2000s. The relationship between economic growth and employment creation is indeed well established. The implication of this is that efforts to significantly reduce youth unemployment are being hamstrung by sluggish economic growth. However, as discussed later, there is scope for employment creation, though on a smaller scale, as reflected by successes achieved by intermediaries and public works programmes, for example. There is also urgent need for

692   Cecil Mlatsheni these and other micro-​level interventions, given the adverse effects of unemployment on youth. From 2008 to 2020 youth (aged 15–​35) experienced an increase in the unemployment rate of 14 percentage points, from 38 per cent to 53 per cent by the broad definition, while for non-​youth (aged 36–​64) an increase of 10 percentage points was observed, from 17 per cent to 27 per cent. The general trend has been one of worsening unemployment and youth have been hardest hit. Regarding the nature of work that youth acquire, 75 per cent work in the formal sector, 20 per cent work in the informal sector, while 5 per cent work in private households. The distribution is fairly similar between youth and non-​youth (aged 36–​ 64) in that 72 per cent of non-​youth work in the formal sector, 18 per cent work in the informal sector, and 10 per cent work in private households. In addition, 50 per cent of youth have permanent contracts whereas 67 per cent of non-​youth have jobs of a permanent nature. Half of the youth are working in permanent jobs whereas two-​thirds of non-​youth hold permanent jobs. Close to 20 per cent of youth are employed in jobs of limited duration. A sizeable 30 per cent of youth and 22 per cent of non-​youth are employed in jobs of unspecified duration, which spells uncertainty for these workers as these jobs could end at any time. Taking the preceding statistics together one can sum up that just under half of youth wanting to work have jobs and that of those, half have security of employment. The disproportionate effect of unemployment on youth is exacerbated by duration of unemployment. The long-​term unemployment rate for the working-​age population, measured as a percentage of total unemployed, was 57 per cent in 2008 at the end of the economic boom and had increased to 72 per cent by 2020. For youth (aged 15–​35) long-​ term unemployment is 70 per cent and for the aged 36–​64 cohort it is 74 per cent. By way of comparison, for select countries, the long-​term unemployment rate in 2019 was 11.5 per cent in Colombia, 9.9 per cent in Namibia, 12.9 per cent in Costa Rica, 12.7 per cent in the United States, 25.1 per cent in the United Kingdom, 38 per cent in Germany, 37.8 per cent in Spain, 57 per cent in Italy, and 70 per cent in Greece (OECD 2021). Long unemployment duration could lead to discouragement such that affected individuals end up disengaging with the labour market and studies (Mlatsheni 2012). What is worse is that 57 per cent of youth (aged 15–​35) and 44 per cent of youth (aged 25–​35) have never worked for pay. In more developed countries such as the United States and Canada, the unemployed in times of increased joblessness increase uptake of training. Those who are youth tend to stay longer in school and they extend the period of living with their parents (Topel and Ward 1992; OECD 2010a). These options are not always available in the developing-​country context, such as that of South Africa. There is also long-​standing evidence that lack of viable options often encourages criminal activity even in developed countries. In France, crime rates increase and in the United States the probability of incarceration increases as joblessness increases (Topel and Ward 1992). Labour market discouragement is a serious concern, as duration of unemployment has negative consequences on the self-​esteem of job-​seekers as well as on the likelihood

The Youth Labour Market in South Africa    693 of an employer hiring them. The negative consequences of long-​term unemployment are well documented (Hammarstrom and Janlert 1997). A  meta-​analysis across 237 cross-​sectional and 87 longitudinal studies found that mental health worsens with job loss and improves when a person obtains a job (Paul and Moser 2009). However, unemployment is not always a cause of mental of illnesses, rather mental illnesses can lead to unemployment (Blakely et al. 2003).

32.3  Why Youth Struggle to Find Jobs 32.3.1 Insufficient Employment Opportunities Availability of employment opportunities is the chief determinant of the level of the overall unemployment rate. Economic conditions and the structure of the labour market have an important bearing on employment opportunities. Ultimately aggregate demand is the main macroeconomic determinant of youth unemployment (O’Higgins 2001). When an economy is in a slump or is not growing at a fast enough pace, employment creation tends to be dampened. Under such circumstances, labour market interventions aimed at increasing employment generally fail to have significant impact. The period between 2003 and 2008 provides evidence of the importance of aggregate demand:  two million jobs were created in the South African economy during this period when economic growth averaged 4.9 per cent (National Treasury 2011). A number of economic shocks rocked the global economy, such that the South African economy was also adversely affected. The economy was already beginning to slow down at the time of the 2008–​09 recession and it subsequently went into recession. The negative effects of the global economic shocks were dampened by prudent macroeconomic policies. Nevertheless, South Africa lost close to 1 million jobs during the 2008–​09 recession (IMF 2011). In addition to this, by 2011 the ratio of employment to the working-​ age population had declined from around 45 per cent in 2008 to 40 per cent in 2011, implying that employment had remained below the pre-​crisis level. The labour market has therefore remained in a relatively depressed state. The effects of the 2008–​09 global recession served to reverse any employment gains that had been made in the earlier period of peak economic growth. It is only in 2013 that employment levels returned, for the first time, to the levels of 2008. The lack of sustained economic growth has led to the overall unemployment rate consistently hovering around 25 per cent. In addition to economic growth, the structure of that growth and the structure of the economy are important determinants of job creation. The South African economy has been characterized as exhibiting skills mismatch, where the skills in demand are out of alignment with the skills sets of the unemployed (Bhorat et al. 2013). In addition, the skills shortage may be a consequence, rather than a cause, of the structure of our growth (Hausmann 2008).

694   Cecil Mlatsheni It is arguable that poor economic performance is as much an obstacle to non-​youth employment as it is to youth employment because a fall in aggregate demand would lead to a general increase in the number of lay-​offs and a fall in new hires. However, to the extent that the nature of the employer–​employee relationship is different for youth and non-​youth, the influence of economic performance on employment would be different for youth and adults. The opportunity cost to employers of firing youth is lower than that of firing adults, partly because youth are less generally likely to be unionized or protected by legislation (Rees 1986) and past investment in them by the employer is likely to have been lower.

32.3.2 Education Global evidence confirms that generally the higher the level of educational attainment of youth, the higher the probability of them finding employment (World Bank 2007). In the case of South Africa, the strongest effects are observed post matric (van der Berg and van Broekhuizen 2016). However, as the discussion below reveals, even youth who have completed secondary schooling often struggle to find employment and this is an enduring feature of the labour market. This should send a signal to youth to acquire as much education as possible in order to improve employment prospects. Indeed, there is evidence that youth pick up on this signal as a significant number of those who drop out of secondary schooling anticipate returning to their studies and even aim to achieve a tertiary-​level qualification (De Lannoy 2008). In reality, about half of youth who start school proceed to complete grade 12 (Spaull 2015). The reasons for dropping out include financial constraints, grade repetition, and teen pregnancy (Branson et al. 2015), all of which can be linked to socio-​economic status. Many youths who exit the schooling system prior to completing matric ultimately enter the labour market in a manner that could be described as premature, given that unemployment plagues even youth with post-​matric tertiary qualifications. These findings suggest that there is a problem with youth’s work-​readiness upon entering the labour market at all education levels. Analysis of the Quarterly Labour Force Survey (QLFS 2020: Q1) reveals that, of the youth (aged 15–​24) who are no longer pursuing studies, 46 per cent have not completed grade 12, while 42 per cent have completed grade 12 but no further. This set of statistics highlights two challenges that have a bearing on the employability of youth:  one is completing secondary schooling and the other is progressing from secondary schooling to tertiary education. Studies indicate that youth see the value of education as a means to improve their career prospects (Graham 2012; Swarts et al. 2012), however, in addition to the challenges that lead to dropping out of secondary schooling mentioned above, poor quality of education and poor schooling environments can result in weak school-​ leaving grades, and thus limit access to higher education. Teacher absenteeism, resource shortages, and infiltration of violent behaviour are some factors that mar the learning environment (Swartz et  al. 2012). In addition, poor performance by learners in secondary school is partly rooted in learning gaps developed in primary school (Mlatsheni

The Youth Labour Market in South Africa    695 2006; Spaull 2015). Besides these constraints, some youth may opt for earlier entry into the labour market and thus a low-​paying, mediocre job because of immediate pressure to supplement family income, more especially when there are younger siblings in need of support. The South African education system offers a technical and vocational training avenue both for youths who complete secondary education and those who do not. Learners are able to pursue studies through the Technical Vocational Education and Training (TVET) system and the National Certificate (Vocational) (NC(V)). However, the reach of these initiatives is limited in that only 8 per cent of 15–​24-​year-​old youth attend university or college (Branson et al. 2015). Funding, learner preparedness, and quality of teaching and learning are among the significant limiting factors in extending coverage to a larger proportion of youth in need of the TVET line of study. The unemployment rate for youth (aged 15–​35) with grade 10 completion only is 60 per cent, for those with grade 11 completion it is 63 per cent, while for those who have completed matric it is 51 per cent (QLFS 2020: Q1). It is clear from these statistics and also borne out by multivariate analysis (Mlatsheni and Ranchhod 2017) that premature entry into the labour market, in terms of qualifications and work-​readiness, is characterized by large-​scale joblessness. Furthermore, failure to satisfactorily complete secondary schooling and to advance to further studies affects later productivity while the foregone earnings and lack of skill accumulation may make it difficult to escape poverty. The benefits of schooling are recognized worldwide, as in 2007 international evidence was that across sixty-​one developing countries the average return per year of schooling was 7.3 per cent for men and 9.8 per cent for women (World Bank 2007).

32.3.3 Geographical Isolation, Cost of Job Search, and Perceptions of Opportunity Another reason why many youths do not actively search for work is the cost of accessing areas that could potentially provide employment. This factor predominantly affects black youth, as townships are often situated far from business centres. The formal exposition of the effects of distance on labour market outcomes was first given by Kain (1968) as the ‘spatial mismatch hypothesis’ in relation to the United States. In the US context, the location of jobs was becoming increasingly decentralized and poor minority households (mainly black) were being left behind in central cities through constraints in housing choices. These developments decreased the employment prospects of the individuals concerned through lesser job access and earnings. The South African situation is similar in that individuals from poor households (mainly black) are far from where jobs are located. However, the difference in South Africa is that it is the jobs that have historically been concentrated in the cities while the location of poor households has been on the outskirts. The cost of a single trip from a township to seek work in a city or industrial centre is significant for an unemployed

696   Cecil Mlatsheni person. Graham et  al. (2019) estimated the average monthly transport cost for job search amongst youth to be R280, roughly 10 per cent of the effective minimum wage calculated via sectoral determinations. Furthermore, data from the March 2005 Labour Force Survey reveal that of the youth who did not seek employment or start their own business, 49 per cent said it was because there were no jobs in their area, while 23 per cent said that they lacked the money necessary to look for work. Taken together, over 70 per cent of non-​searchers indicated that their location constrained them from looking for work, as lack of money to search is brought about by lack of proximity to employment sources. This phenomenon has persisted, as in 2020 the QLFS indicates that 72 per cent of youth cite lack of jobs in the area and 5 per cent cite lack of finance for transport as reasons for not searching for work. Perceptions of the likelihood of finding employment become an important issue when the cost of job search is binding. It is logical that youth perceptions of the labour market have a strong effect on the effort they make to search actively for employment. International evidence reflects that such perceptions are formed in part by the characteristics of the neighbourhoods in which youth reside (Case and Katz 1991). In a study of poor neighbourhoods in Boston, Case and Katz (1991) find that neighbourhood peers significantly influence youth behaviour, including the propensity to work. Furthermore, a neighbourhood can negatively influence labour market outcomes through absence of positive role models, lack of informal job contacts, and the presence of disruptive forces. There exists evidence of negative effects of perceptions on job search in the South African literature as well (Wittenberg 2002). Wittenberg (2002), using a number of household datasets, finds that the proportion of the unemployed engaged in active search hardly ever reaches 50 per cent, a factor which suggests that very often discouragement is based on perceptions of opportunity rather than personal experience. In addition, the disappointment that often accompanies failure to find employment can manifest as a type of paralysis that prevents any structured active job search (Graham and Mlatsheni 2015). It follows from the above-​mentioned literature that where perceptions of the labour market are overly pessimistic, the inadequate search attempt that results then feeds back and contributes to further entrenching youth unemployment.

32.3.4 Networks and Social Capital Obtaining finance for prolonged active job search is a challenge for many young people and it follows that they would rely on their networks to help them find work. Although most do not claim to use networks as their main search method, it is nevertheless the most common method of obtaining employment amongst youth (Schoer and Leibbrandt 2006; Ingle and Mlatsheni 2016). Indeed, having an employed friend or relative, either residing in the same household or a different one, has a significant effect on employability of youth. The NIDS data indicate, for example, that of youth who were unemployed in one wave and employed in the next, about half were informed

The Youth Labour Market in South Africa    697 of the job by a friend or relative in a different household (Ingle and Mlatsheni 2016). This is a long-​standing feature of the functioning of the youth labour market. Evidence from Khayelitsha/​Mitchell’s Plain Survey (KMPS) conducted in Cape Town in 2000 also indicates that more than 55 per cent of the respondents obtained their current job through friends and relatives (Schoer and Leibbrandt 2006). Information asymmetry between young job-​seekers and potential employers is a further impediment that is mitigated by use of networks. Employers often recruit via networks (Bernstein 2014; De Lannoy et al. 2020). Intermediaries such as the Harambee Youth Employment Accelerator and the Digital Jobs Fund programme, operating within the landscape of the South African youth labour market, have achieved results that suggest the existence of challenges of job matching and signalling of attributes by youth (De Lannoy et al. 2020). Furthermore, these intermediaries are able to provide mentoring and training that fills some of the soft skills deficiencies that may be present in unemployed youth. In addition, randomized control trials conducted with South African youth lend further support to this notion. Specifically, these studies find that assessing work-​seekers and giving them results of their skills sets to share with employers increases call-​backs and employment, also standardized reference letters improve employer response rates to job applications by youth (Abel et al. 2020). In the absence of these measures, to limit information asymmetry employers often rely on their employees to recommend workers to fill vacancies.

32.4  An Account of Key Policies Implemented to Address Youth Unemployment in South Africa Significant reduction in the unemployment rate will come through economic growth. However, prospects of achieving the requisite level of growth are bleak in the short to medium term. With long-​term unemployment at close to 40 per cent, what is needed is policy to keep unemployed youth engaged with the labour market to mitigate the ill effects of prolonged unemployment outlined earlier. Active labour-​market policies must be central as they can improve job matching, fill the gap when employers and workers underinvest in training, and mitigate the effects of recession by providing temporary employment or creating incentives for employers to hire (World Bank 2012). Even when active labour-​market policies do not result in significant job creation, they are not a waste because they play an important role in ensuring that youth remain in touch with the labour market (OECD 2010a). The most vulnerable and marginalized group of youth is arguably that which is not in employment, education, or training (the NEETs). Youth who find themselves in these circumstances are not acquiring human capital in the form of studies or training, nor are they gaining any on-​the-​job experience. This category of youth needs to be targeted

698   Cecil Mlatsheni for policy intervention as some of them have quit their studies prematurely (and the mechanisms behind these decisions have to be understood), while the majority lack skills that would give them an edge in the labour market. Furthermore, these idle youth generally come from disadvantaged backgrounds. In a sense, intervention at this stage is tantamount to an exercise in damage control because, as the literature confirms, the greatest rewards result from early and sustained interventions (Garces et al. 2000). With a given range of policies in place to assist youth, policymakers have to recognize the importance of getting as many of the targeted youth to use the resources available to them to mitigate the ill effects of unemployment.

32.4.1 Public Training Policies The National Skills Development Strategy (NSDS) gives direction with regard to training. The National Skills Fund (NSF) and the Sector Education Training Authorities (SETAs) are responsible for implementing the NSDS. Training is carried out by national, provincial, and local governments with the private sector partnering in most cases (De Lannoy et al. 2020). The range of training that is provided in general can be classified as either apprenticeships, learnerships, internships, or other work-​integrated learning, service programmes, or solely skills training. The apprenticeships that have been available to youth in South Africa involve skills training and practical training through work-​integrated learning. For example, youth can train in trades as plumbers, mechanics, and electricians. The learnerships are usually focused on non-​trade jobs with the participant being an employed individual with a matric qualification. The internships are generally targeted at youth with some post-​ secondary schooling qualifications (De Lannoy et al. 2020). There is some evidence that learnerships and apprenticeships have had a positive effect on job placements, with most participants transitioning into employment that lasts up to five years (Kruss et al. 2012). Participants in learnership programmes, in particular, report an increase in soft skills but not technical, numeracy, or literacy skills while apprenticeship participants show an increase in technical and soft skills (Kruss et al. 2012). Other research indicates that learnership participants enjoy higher wages than non-​participants and that they enjoy greater job satisfaction in the short term but that effect weakens with time (Rankin et al. 2014). Furthermore, playing an arguably significant role as far as the provision of training in intermediate-level technical and vocational skills is concerned, are the technical and vocational education and training colleges (TVET). However, according to the White Paper for Post-​School Education and Training, even though numbers of students enrolled in TVET colleges have increased over the years, enrolment is still lower than that of universities, whereas the reverse has to be the case in order to effectively plug the gap in mid-​level skills (Department of Higher Education and Training 2014). In addition, the TVET offerings of programme and qualifications mix (PQM) is reported to be complex to administer, difficult for students to understand, and often

The Youth Labour Market in South Africa    699 poorly administered. Many lecturers do not have the necessary workplace experience that would enable them to deliver effective vocational programmes (Department of Higher Education and Training 2014). Furthermore, technical skills acquisition and apprenticeships need to be bolstered. An inadequate level of apprenticeships impacts on youth disproportionately because youth make up the bulk of the unemployed and unskilled. Furthermore, learnerships are more suited to workers employed in the formal economy, whereas the most vulnerable youth are either unemployed or engaged in survivalist micro-​enterprises. In addition, the learnership contract requires a willing employer, but many employers have traditionally been discouraged by the heavy administrative burden involved in the process. However, the NSDS has undergone a revamp and the Department of Higher Education and Training (DHET) has implemented reforms to make the system more effective and accountable. Global evidence highlights the importance of aligning skills taught with labour demand (World Bank 2012). A number of other factors have been found to be useful in ensuring increased effectiveness of public training programmes: (1) tight targeting of participants; (2) keeping the scale of programmes relatively small; (3) training resulting in a qualification that is recognized and valued by the market; and (4) having a strong on-​the-​job training component and therefore strong links with local employers. There is evidence that the outlook of youth involved in training programmes is improving (World Bank 2012), which would be beneficial in the South African context of the high duration of unemployment and the subsequent discouragement. However, there can be substitution and deadweight losses associated with training. In some instances, individuals who are hired after training would have been employed anyway. Therefore, evaluation should indicate whether the hired workers substituted others, or whether they were hired because the training signalled higher productivity to employers. Furthermore, global evidence indicates that public training agencies are often too slow to respond to the changing needs of firms and job s​ eekers. To counter this, where feasible, public training funds could be directed to private and non-​profit training providers (intermediaries) on a competitive basis. Performance-​based tendering can create incentives for more relevant training.

32.4.2 Job-​Search Assistance In South Africa passive job search may be due to financial constraints that are brought about by the high costs of actively searching for a job, accompanied by the low probability of finding employment. As discussed earlier, this is a factor that predominantly affects black youth, as townships are often situated far from business centres. Indeed, the labour-​force surveys indicate that the majority of non-​searchers indicated that their location constrained them from looking for work. Often youth search for jobs in manner that is not efficient, such as submitting CVs to firms that are not looking to hire and applying for widely advertised positions in which they compete with a multitude of other applicants (Graham et al. 2019).

700   Cecil Mlatsheni The National Youth Development Agency (NYDA) provides a number of avenues to assist job search and improve matching. The schemes include the Graduate Development Programme (GDP) and Job Protection Programme (JPP) to help jobless graduates and matriculants to find jobs; the National Youth Service allows youth to gain work experience while providing community service; the Jobs and Opportunity Seekers and graduate database aim to link jobless youth with employment opportunities; and Youth Advisory Centres (YACs), which are walk-​in centres in communities, where youth can access all NYDA resources, including career counselling. However, many high-​and middle-​income countries have overhauled their job-​search services with a move towards private provision as opposed to public provision (World Bank 2012). In South Africa intermediaries are achieving success in this sphere. The challenge is scaling the existing useful job-​search measures, such as active placement methods, raising the motivation of the unemployed, as well as encouraging and monitoring job-​search behaviour. There is scope for the NYDA to play a coordinating role in the quest to scale useful interventions. With a policy of job-​search assistance there is, of course, an underlying assumption that there exists a significant degree of frictional unemployment and matching constraints within the economy.

32.4.3 Subsidies to Private-​Sector Employment South Africa has in place the Employment Tax Incentive (ETI) as a subsidy that is directed at incentivizing the employment of youth. Since the beginning of 2014 employers can claim reduction in taxation for each qualifying youth hired (aged 18–​29). The aim of the ETI is to lessen the cost and risk to employers of hiring youth. Employers are encouraged to incorporate a training component for the youth they hire. Early evaluations of the ETI, at six months and twelve months into the programme, reflected no effect on youth employment and that the new hires claimed by the ETI would have occurred even in the absence of the programme (Ranchhod and Finn 2015). Evaluations at two to three years of the programme’s implementation also report no significant impact on youth employment and allocative inefficiency in that employment created would have occurred in any event (Ebrahim et al. 2017). The degree of allocative inefficiency of the ETI is estimated to be 57 per cent (National Treasury 2011). However, a modest net increase in employment of youth was reported for small firms with less than fifty employees and those with less than 200 employees (Ebrahim et al. 2017; Rankin and Chaterjee 2016). Subsidies to private-​sector employment can be directed either at the employer or the worker. Employer-​side subsidies, like the ETI, provide financial incentives to firms to hire workers by reducing hiring and employment costs. The structure of the subsidies varies greatly across countries because they are likely to be targeted at different specific groups in each country. The subsidies can be in the form of reimbursing a firm for a fraction of the wages of the covered workers or training costs or a one-​off bonus. The subsidies usually carry a stipulation that employment must last a certain minimum period. Employer-​side subsidies can be either targeted or untargeted, but targeted

The Youth Labour Market in South Africa    701 subsidies are more common, and indeed found to be relatively more successful. A disadvantage of untargeted subsidies is that they could lead to deadweight losses and a substitution by individuals who would have found employment anyway. However, in some instances, high deadweight and substitution effects may be tolerated because part of the objectives of the policy would have been to rearrange the queue of job-​seekers, such that the individuals who struggle most to find employment move further forward in the job queue. In addition, the real costs of wage subsidies are often not easy to calculate (World Bank 2012).

32.4.4 Direct Job Creation in the Public Sector The Expanded Public Works Programme (EPWP) is a key medium-​to long-​term strategy by the South African government to create labour-​intensive jobs in order to ease the plight of the unemployed. It includes a training and skills development component to facilitate easier transition into self-​employment and formal employment. The policy was initially formulated as the National Public Works Programme (NPWP) in 1998. It had limited success and fell short of set targets in its early phase. In 2004 it was redesigned as the EPWP with expanded targets (Mayer et al. 2011). The target for the first phase of the programme (2004–​09) was to provide 1 million jobs, the second phase (2009–​14) had a target of 4.5 million jobs, while a target of 6 million jobs was set for the third phase (2014–​19). Focus on reaching more youth through the programme has intensified as of the beginning of 2020. Among the most recent stated goals by the Department of Public Works and Infrastructure is the desire to create 5 million jobs in the five-​year period to 2024 and that 55 per cent of those be allocated to youth. Evaluations of the EPWP indicate the following: that the employment opportunities created have been shorter than the targeted duration, the poverty alleviating effects of the programme are limited, and the EPWP has limited effect on employability of participants outside of the programme (Department of Public Works 2015). The Community Work Programme (CWP), rolled out in 2008, is an additional public employment programme that is designed to provide an employment safety net. The CWP is implemented at the local level and is designed to enrol a maximum of 1,000 people per site. It offers two days of work a week, eight days a month, or 100 days a year (Philip 2013). The CWP differs from the EPWP in that it is community based and relies on community members to identify specific developmental projects. The scope of the projects encompasses early childhood development, school support, food gardens, community safety, and promoting arts, culture, and recreation (Philip 2013). The CWP has met its goal of providing an employment guarantee of regular part-​time employment and where training has occurred it has been successful such that some participants are able to use the CWP training to develop a career path in the care sector (Philip 2013). Internationally, there has been a marked move away from public-​sector job-​creation programmes in favour of other active measures because of the disappointing results achieved in terms of helping unemployed people get permanent jobs in the labour

702   Cecil Mlatsheni market (Martin and Grubb 2001; OECD 2010b). It is becoming increasingly clear, though, that public-​works programmes have a better chance of becoming jobs ladders when they also offer additional technical and life-​skills support. However, there is still a great deal of debate around the use of these programmes because they can fulfil objectives other than just the creation of permanent jobs. Public-​ sector, employment-​creation programmes can be used to help the most disadvantaged unemployed to maintain contact with the labour market, especially in times of weak aggregate demand and scarcity of vacancies (Fay 1996). Furthermore, in countries such as South Africa and India these programmes have attracted women mainly, providing temporary earnings opportunities to many poor households. However, the number of days of work offered has not been enough to significantly lessen poverty (OECD 2010b).

32.4.5 Support for Self-​employment Promotion of entrepreneurship—​and especially small, medium and micro enterprises (SMMEs)—​is a concrete intervention strategy in the face of high levels of unemployment. The Quarterly Labour Force Survey (2020:  Q1) reveals that 5 per cent of the employed are entrepreneurs employing at least one worker, among youth the comparable statistic is 3 per cent, while for those over the age of 34 the figure is 6 per cent. In addition, 8 per cent of employed youth are own account workers, while 11 per cent of non-​youth (aged 36–​64) fall into this category. These statistics suggest scope for increased entrepreneurial activity by youth. There have been ongoing initiatives to promote the development and sustainability of SMMEs in South Africa. The Small Business Act was passed in 1995 and subsequently organizations such as the Small Business Council and Ntsika Enterprise Promotion Agency were established, as well as the Small Enterprise Development Agency in 2004. The EPWP is also charged with providing SMME support. However, despite these initiatives, SMMEs struggle to thrive. Success as an entrepreneur depends on a person’s intrinsic entrepreneurial ability, availability of investment capital, risk-​absorption capacity, financial management skills, enterprise development, and—​very importantly—​market accessibility (Lazear 2005; Oosterbeek et al. 2010). It seems clear from this as well as the analysis of formal-​sector firm data (Kerr et al. 2013) that entrepreneurship is not easy and even when it happens it does not necessarily have large employment multipliers.

32.5 Conclusion Failure of the labour market to absorb young entrants as they transition from schooling to work has an ill effect on youth and on society in general. Policy must therefore ensure that youth remain in touch with the labour market and that long-​term unemployment is minimized. The most significant impact on youth unemployment will come

The Youth Labour Market in South Africa    703 from healthy economic growth that will result from investment. Government has to create an enabling environment and reduce corruption. Nevertheless, intermediaries operating on the supply side are having some success in teasing out ‘veiled’ demand for labour by reducing the risk of hiring youth that arises from information asymmetry and lack of work-​readiness. Such interventions within the youth labour market need to be supported via public–​private partnerships and scaled up. Government has to continue offering opportunities via programmes such as the EPWP and CWP and follow through on the planned infrastructure investment strategy. In addition, efforts to shift the education system towards a dual system, such as the famed German system, should be intensified. The German approach fosters a tradition of vocational orientation within the secondary school education system. Involvement of the private sector and community (in this case workers and unions) in the content and certification of vocational training, as well as the content and conditions governing on-​ the-​job training, is a key component of this system. These measures will ensure a labour force with a higher and more relevant skills base, ready to capitalize in the event of an upswing in the economy.

References Abel, Martin, Rulof Burger, and Patrizio Piraino. 2020. ‘The value of reference letters: Experimental evidence from South Africa’, American Economic Journal: Applied Economics, 12(3): 40–​7 1. Anderson, Kermyt G., Anne Case, and David Lam. 2001. ‘Causes and consequences of schooling outcomes in South Africa:  Evidence from survey data’, Social Dynamics, 27(1): 37–​59. Case, Anne, and Lawrence Katz. 1991.’ The company you keep: The effects of family and neighborhood on disadvantaged youths.’ NBER Working Paper No. 3705. Bernstein, Ann. 2014. ‘South Africa’s key challenges: Tough choices and new directions’, The Annals of the American Academy of Political and Social Science, 652: 20–​47. Bhorat, Haroon, Ravi Kanbur, and Natasha Mayet. 2013. ‘The impact of sectoral minimum wage laws on employment, wages, and hours of work in South Africa’, IZA Journal of Labor and Development, 2(1). https://​doi.org/​10.1186/​2193-​9020-​2-​1. Blakely, Tony, Sunny Collings, and June Atkinson. 2003. ‘Unemployment and suicide: Evidence for a causal association?’, Journal of Epidemiology and Community Health, 57(8): 594–​600. Branson, Nicola, Clare Hofmeyr, Joy Papier, and Seamus Needham. 2015. ‘Post-​school education: Broadening alternative pathways from school to work’, in Ariane De Lannoy, Sharlene Swartz, Lori Lake, and Charmaine Smith (eds) South African Child Gauge 2015. Cape Town: Children’s Institute, University of Cape Town. De Lannoy, Ariane. 2008. ‘Educational decision-​making in an era of AIDS’, PhD thesis, University of Cape Town. De Lannoy, Ariane, Lauren Graham, Leila Patel, and Murray Leibbrandt. 2020.’ Why is youth unemployment so intractable in South Africa? A synthesis of evidence at the micro-​level’, Journal of Applied Youth Studies, 3(2): 115–​31. Department of Higher Education and Training. 2014. White Paper for Post-​School Education and Training. Pretoria.

704   Cecil Mlatsheni Department of Public Works. 2015. Department of Public Works 2014/​15 Annual Report. Pretoria. Ebrahim, Amina, Murray Leibbrandt, and Vimal Ranchhod. 2017. ‘The effects of the employment tax incentive on South African employment’. UNU-​WIDER Working Paper no. 2017/​ 5, Helsinki. Everatt, David, and Elinor Sisulu. 1992. Black Youth in Crisis. Braamfontein: Ravan Press. Fay, Robert. 1996. ‘Enhancing the effectiveness of active labour market policies:  Evidence from programme evaluations in OECD countries’. OECD Labour Market and Social Policy Occasional Papers, No. 18. Garces, Eliana, Duncan Thomas, and Janet Currie. 2000. ‘Longer-​term effects of head start’, The American Economic Review, 92(4): 999–​1012. Graham, Lauren. 2012. Understanding Risk in the Everyday Identity-​work of Young People on the East Rand of Johannesburg. Johannesburg: University of Johannesburg. Graham, Lauren, and Cecil Mlatsheni. 2015. ‘Youth unemployment in South Africa:  Understanding the challenge and working on solutions’, in Ariane De Lannoy, Sharlene Swartz, Lori Lake, and Charmaine Smith (eds) South African Child Gauge 2015. Cape Town: Children’s Institute, University of Cape Town. Graham, Lauren, Leila Patel, Gina Chowa, Zoheb Khan, Rainier Masa, Senzelwe Mthembu, and Leilana Williams. 2019. ‘Developing youth assets for employability’. Centre for Social Development in Africa, University of Johannesburg. Hammarstrom, Anne, and Urban Janlert. 1997. ‘Nervous and depressive symptoms in a longitudinal study of youth unemployment—​selection or exposure?’, Journal of Adolescence, 20(3): 293–​305. Hausmann, Ricardo. 2008. ‘Final recommendations of the international panel on ASGISA’. CID Working Paper no. 161, May, Harvard University. Ingle, Kim, and Cecil Mlatsheni. 2016. ‘The extent of churn in the South African youth labour market: Evidence from NIDS 2008–​2015’. NIDS Discussion Paper no. 17/​2016, University of Cape Town. International Labour Office. 2020. Global Employment Trends for Youth 2020: Technology and the Future of Jobs. Geneva: ILO. International Monetary Fund. 2011. South Africa: 2011 Article IV Consultation—​Staff Report; Staff Supplement. IMF Country Report No. 11/​258. Kain, John. 1968. ‘Housing segregation, negro employment, and metropolitan decentralization’, Quarterly Journal of Economics, 82: 175–​97. Kerr, Andrew, Martin Wittenberg, and Jairo Arrow. 2013.’ Job creation and destruction in South Africa’. SALDRU Working Paper no. 92, University of Cape Town. Kruss, Glenda, Angelique Wildschut, Dean Janse van Rensburg, Mariette Visser, Genevieve Haupt, and Joan Roodt. 2012. ‘Impact assessment of National Skills Development Strategy II: Developing skills and capabilities through the learnership and apprenticeship pathway systems’. Research report, Human Sciences Research Council and Development Policy Research Unit, University of Cape Town. Lazear, Edward. 2005. ‘Entrepreneurship’, Journal of Labour Economics, 23(4): 649–​80. Martin, John, and David Grubb. 2001. ‘What works and for whom:  A review of OECD countries’ experiences with active labour market policies’, Swedish Economic Policy Review, 8(2): 205–​58.

The Youth Labour Market in South Africa    705 Mayer, M. J., Saguna Gordhan, Rachel Manxeba, Colleen Hughes, Penny Foley, Carmel Maroc, and Matthew Nell. 2011. ‘Towards a youth employment strategy for South Africa’. DBSA Working Paper no. 28, Halfway House, Development Bank of Southern Africa, Johannesburg. Mlatsheni, Cecil. 2006. ‘The youth labour market: What does it take to succeed?’, in Susan Brown (ed.) Money and Morality. Cape Town: Institute for Justice and Reconciliation. Mlatsheni, Cecil. 2012. ‘The challenges unemployment imposes on youth’, in Helene Perold, Nico Cloete, and Joy Papier (eds) Shaping the Future of South Africa’s Youth:  Rethinking Post-​ School Education and Skills Training. Cape Town:  Centre for Higher Education Transformation/​African  Minds. Mlatsheni, Cecil, and Vimal Ranchhod. 2017. ‘Youth labour market dynamics in South Africa: Evidence from NIDS 1-​2-​3’. REDI 3x3 Working Paper no. 39, Southern Africa Labour and Development Research Unit, University of Cape Town. National Treasury. 2011. ‘Confronting youth unemployment: Policy options for South Africa’. Discussion Paper, Pretoria. O’Higgins, Niall. 2001. Youth Unemployment and Employment Policy:  A Global Perspective. Geneva: International Labour Organisation. Oosterbeek, Hessel, Mirjam van Praag, and Auke Ijsselstein. 2010. ‘The impact of entrepreneurship education on entrepreneurship skills and motivation’, European Economic Review, 54(3): 442–​54. Organisation for Economic Co-​operation and Development (OECD). 2010a. Off to a Good Start? Jobs for Youth. Paris: OECD Publishing. Organisation for Economic Co-​ operation and Development (OECD). 2010b. Tackling Inequalities in Brazil, China, India and South Africa: The Role of Labour Market and Social Policies. Paris: OECD Publishing. Organisation for Economic Co-​operation and Development (OECD). 2021. ‘Long-​term unemployment rate.’ [online] https://​data.oecd.org/​unemp/​long-​term-​unemployment-​rate.htm. Paul, Karsten, and Klaus Moser. 2009. ‘Unemployment impairs mental health: Meta-​analyses’, Journal of Vocational Behavior, 74(3): 264–​82. Philip, Kate. 2013. ‘The Community Work Programme: Building a society that works’. ILO Working Paper no. 149, Geneva. Ranchhod, Vimal, and Arden Finn. 2015. ‘Estimating the effects of South Africa’s youth employment tax incentive—​An update’. SALDRU Working Paper no.  152, University of Cape Town. Rankin, Neil, and Aroop Chaterjee. 2016. ‘Estimating the impact of the Employment Tax Incentive’. Presented at the Growth and Development Policy: New Data, New Approaches, and New Evidence, Pretoria. Rankin, Neil, Gareth Roberts, and Volker Schoer. 2014. ‘The success of learnerships? Lessons from South Africa’s training and education programme’. UNU-​WIDER Working Paper no. 2014/​068, Helsinki. Rees, Albert. 1986. ‘An essay on youth joblessness’, Journal of Economic Literature, 34: 613–​28. Schoer, Volker, and Murray Leibbrandt. 2006. ‘Determinants of job search strategies: Evidence from the Khayelitsha/​ Mitchells Plain survey’, South African Journal of Economics, 74(4): 702–​24. Spaull, Nic. 2015. ‘Schooling in South Africa: How low quality education becomes a poverty trap’, in Ariane De Lannoy, Sharlene Swartz, Lori Lake, and Charmaine Smith (eds) The South African Child Gauge 2015. Cape Town: Children’s Institute, University of Cape Town.

706   Cecil Mlatsheni Swartz, Sharlene, James Harding, and Ariane De Lannoy. 2012. ‘Ikasi style and the quiet violence of dreams: A critique of youth belonging in post-​apartheid South Africa’, Comparative Education, 48(1): 27–​40. Topel, Robert, and Michael Ward. 1992. ‘Job mobility and the careers of young men’, Quarterly Journal of Economics, 107(2): 439–​79. Truscott, Kate. 1993. ‘Youth education and work: The need for an integrated policy and action approach’. CASE, University of Witwatersrand. van Broekhuizen, Hendrik, and Servaas van der Berg. 2016. ‘How high is graduate unemployment in South Africa? A much-​needed update’. [online] http://​www.econ3x3.org/​article/​ how-​high-​graduate-​unemployment-​south-​africa-​much-​needed update. van Zyl Slabbert, Frederik. 1994. Youth in the New South Africa. Pretoria: HSRC Press. Wittenberg, Martin. 2002. ‘Job Search in South Africa:  A nonparametric analysis’, South African Journal of Economics, 70(8): 1163–​1197. World Bank. 2007. ‘World Development Report 2007: Development and the next generation’. World Bank, Washington, DC. World Bank. 2012. ‘World development report 2013: Jobs’. World Bank, Washington, DC.

Chapter 33

The Ec onomi c s of Educat i on i n Sou th Af ri c a Nicola Branson and David Lam

33.1 Introduction The quality of education provision in South Africa is strongly stratified by race, geographical location, and income. Where a child is born and the income level of their parents overwhelmingly predicts their educational trajectory (Finn et al. 2016) and subsequent welfare as an adult (Patrinos 2015). Inequalities in education are therefore an important link in understanding the relationship between inequality, unemployment, and economic growth in South Africa. Much work has been done in the post-​apartheid years to build a unified education system in South Africa. Education provision in South Africa is guided by national policies including the Constitution of the Republic of South Africa, 1996, which recognizes basic and further education as a right for all South Africans that the state must make ‘progressively available and accessible’ (Section 29(1)b). In this chapter we show that while there have been substantial improvements in educational attainment in South Africa, with narrowing racial gaps in the years of education attained, there continue to be large disparities in education outcomes, leading to differences in the proportion of South Africans completing secondary school or studying further by race. Low post-​secondary enrolment largely reflects limited, and starkly unequal, levels of learning in primary and secondary schools. Credit constraints, while secondary to academic performance, mean that some low-​income students who are eligible to enrol in university do not enrol, or enrol in college rather than university.1 Furthermore we note that the college system, while in 1  In the South African system, the term ‘college’ refers to public Technical and Vocational Education and Training (TVET) and private Further Education and Training (FET) colleges that offer vocational, occupational, and artisan training, and does not include universities.

708    Nicola Branson and David Lam principle able to cater to the large share of learners leaving the schooling system without completing matric, has relatively low levels of enrolment and has seen limited growth since 2015, even though increased technical and vocational education and training (TVET) enrolment is stated as a national priority. Finally, we turn to the labour market to look at the impact of education on employment and earnings for working-​age adults. South Africa has very high returns to post-​secondary education by world standards, and employment rates are much higher for post-​secondary graduates than for those with less education. Higher education is becoming increasingly important in the link between education and inequality.

33.2  Education, Race, and Class in South Africa Education provision in South Africa was segregated by race and location even before the Bantu Education Act of 1953 was promulgated under the apartheid government (Fiske and Ladd 2004). Under apartheid there were different education systems for each race group,2 with education for Africans purposefully inferior to that of whites in terms of enrolment requirements,3 curriculum, resources, teacher qualifications, and performance evaluation (Pillay 1984; Fiske and Ladd 2004). For example, in 1974/​75, per capita spending on African learners was only 6.7 per cent of per capita spending on whites, with coloured and Indian per capita expenditure at 20.7 per cent and 28.3 per cent of white’s, respectively (Pillay 1984). The enrolment rate4 in 1970 was 54 per cent for Africans, compared to 96 per cent for whites, 84 per cent for Indians, and 75 per cent for coloureds. Since 1994, government education policy has focused on creating a coordinated system to redress these inherited educational inequalities (for example, the South African Schools Act (1996); Further Education and Training Act (1998); Employment of Educators Act (1998); National Norms and Standards for School Funding (2006); and the South African Qualifications Authority Act No. 58 of 1995). Two and a half decades into democracy, education inequalities along racial and spatial lines persist. Leibbrandt and Díaz Pabón (­Chapter 9 in this volume) emphasize that a ‘detailed understanding of the intersectionality of race and class and gender and space’ is required in thinking about inequalities in South Africa. In this chapter, we unpack

2 

There were seventeen separate education departments, eleven for Africans, four for Whites (one per province), one for Indians, and one for Coloureds. 3  Compulsory education was up until age 15 for Whites, Indians, and Coloureds, and four years for Africans (Pillay 1984). 4  Ratio of enrolled to population aged 5–​19 years.

The Economics of Education in South Africa    709 inequalities in education and frequently use the apartheid racial classifications to disaggregate the national picture. The disaggregation is driven by the availability of measures in the national datasets. As we will show, however, this simple classification, while blurring slightly at the edges with time, continues to paint a stark picture of inequalities in educational attainment in South Africa.

33.3  Education Provision in South Africa Following South Africa’s transition to democracy, the South African Qualifications Authority (SAQA) Act of 1995 initiated the establishment of the National Qualification Framework (NQF).5 The objective of the NQF was to create a single system covering all education and training, including basic education, higher education, and vocational and occupational training. The NQF has ten levels, with qualifications in three sub-​frameworks:  (1) General and Further Education and Training Qualifications, (2) Higher Education Qualifications, and (3) Occupational Qualifications.

33.3.1 General and Further Education and Training Qualifications Sub-​framework: NQF Levels 1–​4 Compulsory education consists of schooling for learners aged 7–​15, or the completion of grade 9, whichever occurs first (South African Schools Act 1996). Compulsory education is divided into the foundation phase (grades R–​3), intermediate phase (grades 4–​6), and senior phase (grades 7–​9), all at NQF level 1. Although compulsory education is from age 7, grade R, introduced in 2001, targets learners turning age 5 by 30 June in the year of attendance. The National Development Plan (NDP) goal for 2030 is two years of pre-​primary education, but while enrolment in grade R is close to universal (94 per cent in 2018), only around three-​quarters of learners enrol in a pre-​grade R year (van der Berg et al. 2020). Upon completing grade 9, learners may continue to the further education and training phase in the schooling system and work through grades 10, 11, and 12, which align with NQF levels 2, 3, and 4, respectively. The first national standardized examination taken within the public-​school stream is in grade 12, commonly referred to as the matric examinations. Learners passing the matric examinations are awarded a National

5 

This section draws heavily from Branson, Culligan, and Ingle (2020).

710    Nicola Branson and David Lam Senior Certificate (NSC) at NQF level 4. A learner can pass the matric examinations with a bachelor pass (previously called matric exemption), diploma pass, or higher certificate pass. The type of NSC pass determines the type of post-​secondary institution the learner is eligible to apply to. Alternatively, learners may opt to leave the schooling system at age 15 and pursue further education and training via the college system. The college system includes publicly funded TVET colleges that offer vocational, occupational, and artisan training and community colleges (CET) that focus on providing youth and adults who did not complete school with basic literacy, numeracy, and other skills required for employment. Learners receive qualifications based on their courses of study. Learners who take the NC(V) route obtain a National Certificate (Vocational) at NQF level 4, the vocational equivalent of the NSC.

33.3.2 Higher Education Qualifications Sub-​framework: NQF Levels 5–​10 Post-​ secondary education, aimed toward equipping students with skills for the economy, falls under the Higher Education Qualifications sub-​ framework, with qualifications from NQF levels 5 to 10. NQF levels 5 to 7 are undergraduate qualifications and NQF levels 8 to 10 are postgraduate qualifications. These qualifications are predominantly provided by public and private universities, but students can also obtain qualifications at NQF level 5 and 6 from public TVET and private FET colleges.

33.3.3 The Occupational Qualifications Sub-​framework: NQF Levels 1 to 8 The Occupational Qualifications sub-​framework facilitates the provision of occupational education. Occupational qualifications are envisaged to address the needs of the labour market and therefore tend to be occupation-specific. The qualification certifies that the individual has the theoretical knowledge, practical skills, and workplace experience associated with a trade, occupation, or profession. Occupational education is provided by skills development providers that are accredited by the Quality Council for Trades and Occupation and include learnerships, internships, and skills programmes. These providers include, but are not limited to, TVET colleges, private colleges, companies, universities of technology, and non-​governmental organizations (NGOs). With this mapping of the education and training system in South Africa, section 33.4 describes trends in educational attainment for different cohorts.

The Economics of Education in South Africa    711

33.4  Trends in Educational Attainment and Educational Inequality South Africa has experienced substantial improvements in education in recent decades, with racial and gender gaps in educational attainment narrowing and an overall decline in schooling inequality. Figure 33.1 shows mean years of education (highest completed year of formal education) by year of birth and population group in the 2016 Community Survey (CS). A striking feature of Figure 33.1 is that African education was rising rapidly before the end of apartheid. Africans born in 1991 (age 25 in 2016) completed 10.8 years of education on average, 5.8 years more than those born in 1950. The gap in completed years of education between whites and Africans fell from 7.1 years to 1.9 years between the 1950 cohort and the 1991 cohort. The pace of increase has been considerably slower since the end of apartheid, however, with average schooling of Africans rising only 0.3 years between the 1980 cohort and the 1991 cohort, and with the gap between whites and Africans only falling from 2.5 years to 1.9 years. The gap between coloured and African schooling was 2.3 years in the 1951 cohort, but was eliminated by the 1975 cohort and has

Mean years of education by year of birth and population group 13 12

Mean years of schooling

11 10 9 8 7 6 5 4 1950

1955

1960 All

1965 African

1970 1975 Year of Birth Coloured

1980 Indian

1985

1900

White

Figure 33.1  Mean years of education by year of birth and population group Source: Own Calculation using Community Survey 2016; three-​year moving averages.

712    Nicola Branson and David Lam remained closed. The gap in mean education attainment between those living in urban versus traditional and farm areas has steadily decreased across the cohorts (not shown), from over four years in the 1950 cohort to a difference of less than a year for those born in 1991. Figure 33.2 shows the trend in the proportion completing grade 12 (matric), a critical education milestone. The proportion of Africans completing grade 12 increased from 8 per cent in the 1950 cohort to 53 per cent in the 1991 cohort. The white advantage over Africans in secondary completion rates was 64 percentage points in the 1950 cohort, falling to 33 percentage points by the 1991 cohort. One disappointing pattern in Figures 33.1 and 33.2 is the limited progress made since the end of apartheid. Mean completed education of Africans who started school in 1994 (born in 1987) is only 0.7 grades higher than those who were age 20 in 1994. The gap between African and coloured education has almost been eliminated, but both groups continue to have about 1.9  years lower educational attainment than whites. While African and coloured students are progressing well into secondary school, there has been less improvement in the share reaching grade 12 and passing the grade 12 matriculation exam. As shown in Figure 33.2, the proportion of South Africans completing secondary school has been rising at a fairly slow rate for those born since 1975. The proportion of Africans completing grade 9 has risen to over 80 per cent, but the proportion Proportion completing secondary by year of birth and population group 1

Proportion with grade 12 and above

.9 .8 .7 .6 .5 .4 .3 .2 .1 0 1950

1955

1960 All

1965 African

1970 1975 Year of Birth Coloured

1980 Indian

1985

1900

White

Figure 33.2  Proportion completing secondary by year of birth and population group Source: Own calculations using Community Survey 2016; three-​year moving averages.

The Economics of Education in South Africa    713 Proportion completing post-secondary (of those with 12+ years of education) By year of birth and population group

Proportion completing post-secondary (of those with 12+ years of ed)

.6 .5 .4 .3 .2 .1 0 1950

1955

1960

All

1965

African

1970 Year of Birth Coloured

1975

Indian

1980

1985

1900

White

Figure 33.3  Proportion completing post-​secondary (of those with twelve+ years of education) by year of birth and population group Source: Own calculations using Community Survey 2016; three-​year moving averages.

of those 9th graders who go on to complete grade 12 has been declining (around 50 per cent in recent cohorts). Figure 33.3 shows that the proportion of secondary school graduates who go on to complete some kind of post-​secondary education6 has also declined steeply for African and coloured cohorts, widening the racial gap in post-​secondary educational attainment for younger cohorts. About 40 per cent of Africans and whites born before 1955 who completed 12 years of education went on to complete post-​secondary education, with a rate of around 30 per cent for coloureds and Indians. This rate has steadily declined for the African and coloured cohorts born subsequently; only 20 per cent of those born in 1990 who completed 12  years of education converted this into a post-​secondary qualification. Figure 33.1 showed that this decline coincides with growth in the share of Africans and coloureds completing secondary education; a five f​old increase is evident between the 1950s and 1975 cohorts. The proportion of Indians converting grade 12 to a post-​school qualification remained stable however, even with similar, if not larger, increases in grade 12 attainment. 6 

Post-​secondary education includes all qualifications at NQF level 5 and above, including diplomas and certificates requiring matric as well as university degrees.

714    Nicola Branson and David Lam

33.4.1 Enrolment, Grade Advancement, and Grade Repetition Although Africans in recent cohorts end up with 1.9 fewer years of schooling than whites, this is not mainly due to Africans dropping out of school. As shown by Branson and Lam (2010), grade repetition plays a major role. This is seen in Table 33.1, which shows how the racial gap in education emerges as youth progress through school. The first four columns show that school enrolment is almost universal for all population groups between ages 6 and 15, after which schooling is no longer compulsory. While coloured youth begin dropping out at around age 15, enrolment for Africans remains high, with 87 per cent enrolled at age 17 and 74 per cent enrolled at age 18, almost identical to whites. In spite of similar enrolment rates, educational attainment for Africans drops progressively further behind whites as youth advance through school. By age 15, Africans are more than half a grade behind whites, and by age 18 the gap is 1.1 grades. Grade repetition plays a key role. The last three columns of Table 33.1 show that 6–​7 per cent of coloured and African students are repeating a grade until age 14. Repetition rates increase substantially in secondary school, with 24 per cent of African and 23 per cent of coloured 18-​year-​olds enrolled in school repeating their current grade. These high rates of grade repetition at ages 16 onwards are consistent with a strategy called ‘weeding’ (Van der Berg et al. 2019), where weaker students in grades 10 and 11 who are thought unlikely to pass the matriculation exam are purposefully held back in an attempt to achieve high matric pass rates for the school. Grade repetition varies dramatically across socio-​ economic levels. Using the National Income Dynamics Study (NIDS), Branson et al. (2014) find that 30 per cent of low-​income students in grade 11 in 2008 repeated at least one grade in the next two years. This compared to only 8 per cent of high-​income students. Lam et al. (2011), using longitudinal data from the Cape Area Panel Study (CAPS), found that grade repetition was badly targeted, especially in poor schools. A CAPS-​administered literacy and numeracy exam was much less correlated with grade advancement for Africans than it was for other groups, suggesting that predominantly Black schools do a poor job deciding which students to hold back. This evidence of poor evaluation is consistent with van der Berg and Shepherd’s (2010) finding of little correlation between continuous assessment marks provided by teachers and the externally evaluated matriculation exam. Van der Berg et al. (2019) estimate that grade repetition costs around 8 per cent of the national budget for basic education and note that policies promulgated to limit grade repetition (Burger et al. 2015; Kika and Kotze 2019) have had little long-​term impact. For African youth the combination of high enrolment rates into the late teens and frequent grade repetition creates a situation in which Africans spend 1.5 more years going to school than whites, but end up with a grade less schooling (Branson and Lam 2010). Columns 1–​8 in Table 33.1 show that Africans and whites have similar enrolment rates at ages 18–​20, combining enrolment rates in school and college/​ university. Africans are mainly enrolled in secondary school, however, while whites

Table 33.1: Enrolment, educational attainment, and grade repetition by age Enrolled in school

Enrolled in college or university Coloured

Indian

Repeated previous year (of those enrolled in school)*

Years of education

Age

African

Coloured

Indian

White African

5

90%

77%

80%

82%

0%

0%

0%

0%

0.1

0.1

0.1

0.1

1%

2%

0%

6

96%

94%

91%

93%

0%

0%

0%

0%

0.6

0.4

0.6

0.4

3%

2%

1%

7

97%

96%

96%

96%

0%

0%

0%

0%

1.3

1.1

1.4

1.1

5%

3%

1%

8

97%

96%

95%

96%

0%

0%

0%

0%

2.0

1.9

2.1

1.9

6%

4%

4%

9

97%

96%

96%

97%

0%

0%

0%

0%

3.6

3.5

3.7

3.5

7%

5%

1%

10

97%

97%

94%

97%

0%

0%

0%

0%

4.2

4.0

4.4

4.1

6%

8%

0%

11

97%

95%

95%

96%

0%

0%

0%

0%

4.8

4.7

5.2

4.9

7%

11%

3%

12

97%

95%

93%

97%

0.0%

0.0%

0.0%

0.0%

5.6

5.6

6.1

5.7

7%

5%

2%

13

96%

95%

92%

95%

0.0%

0.1%

0.1%

0.1%

6.6

6.5

7.1

6.7

6%

5%

4%

14

96%

92%

94%

95%

0.1%

0.2%

0.3%

0.2%

7.4

7.5

8.2

7.8

7%

5%

0%

15

95%

89%

93%

95%

0.1%

0.6%

0.6%

0.5%

8.2

8.2

9.1

8.8

11%

14%

1%

16

92%

82%

90%

93%

0.4%

1.2%

0.5%

1.0%

9.0

9.1

9.9

9.7

15%

10%

3%

17

85%

70%

70%

86%

1.5%

1.9%

9.6

9.8

10.6

10.5

16%

13%

4%

18

66%

33%

27%

36%

8%

9%

30%

28%

10.1

10.2

11.4

11.2

24%

23%

3%

19

47%

16%

10%

10%

13%

12%

42%

44%

10.3

10.5

11.6

11.7

28%

22%

0%

20

31%

7%

4%

4%

16%

13%

43%

49%

10.5

10.6

11.7

11.8

30%

27%

0%

21

19%

3%

3%

3%

16%

12%

36%

47%

10.6

10.7

11.9

12.0

34%

13%

0%

22

12%

2%

2%

2%

14%

10%

34%

36%

10.7

10.7

11.9

12.3

33%

23%

0%

23

7%

1%

1%

2%

12%

7%

21%

27%

10.7

10.7

11.9

12.4

33%

67%

0%

24

5%

1%

2%

1%

9%

5%

18%

20%

10.8

10.7

12.0

12.7

37%

100%

0%

25

0%

0%

0%

0%

12%

5%

18%

17%

10.8

10.7

12.1

12.7

35%

0%

0%

3.4%

White African

2.2%

Coloured

Indian

White African

Coloured

White

Sources: Data sourced from Community Survey 2016 for enrolment and years of education; GHS 2017/​18 for repetition as this variable is not in the Community Survey. Indian repetition rates are excluded due to small sample sizes in GHS.

716    Nicola Branson and David Lam transition into tertiary education. Many will be surprised that the African enrolment rate of 60 per cent (47 per cent + 13 per cent) is higher than the 54 per cent enrolment rate of whites at age 19. But 78 per cent of enrolled Africans are still trying to finish secondary school, while 82 per cent of enrolled whites are attending post-​secondary institutions. Table 33.1 highlights low enrolment in the college and university system for learners under age 18. The percentage of 15-​and 16-​year-​olds enrolling in college or university is below 1.2 per cent in all population groups. This is corroborated by Papier (2009) who notes that the college system is not typically used as an alternative pathway through FET for learners, but rather used by those scholars who have already at least attempted grade 12. The distribution of post-​secondary enrolment across university and college differs across race groups, with African and coloured youth more likely than white and Indian youth to enrol in college than university. For example, 64 per cent of Africans between the age of 20 and 25 who are enrolled out of the school system are enrolled in college,7 compared to 18 per cent of white, 24 per cent of Indian and 50 per cent of coloured youth at these ages (own calculations, Community Survey 2016). Figure 33.4 displays post-​secondary qualification attainment by NQF level and year of birth. The figure shows a rapidly increasing share of post-​secondary qualifications at NQF level 5 (N4–​6 and certificates) for those born from 1970 onwards, with a decline in the share in NQF levels 7 and above. NQF level 5 qualifications are predominantly obtained in college, while qualifications at NQF level 7 and above can only be obtained from universities. These changes have implications for the distribution of skills in the labour market. Gender differences in education in South Africa have moved in the direction of a female advantage. Girls progress through school faster than boys in all population groups and end up with higher educational attainment (Anderson et al. 2001; Lam et al. 2011). Branson et al. (2014), using national longitudinal data, show that close to 40 per cent of boys in grades 1–​4 in 2009 repeated at least one grade over the next two years, about double the level for girls. Boys are also more likely than girls to drop out in secondary school. This selection results in an academically stronger group of boys reaching grade 12 and writing the NSC (Spaull and Makaluza 2019), evident in higher average pass rates for boys (82.8 per cent compared to 80.1 per cent) (DBE 2020b). While South Africa has made large strides in improving average educational attainment, the rate of tertiary education completion among Africans and coloureds who reach matric has not kept pace. With a labour market that increasingly favours those with tertiary education, failure to close racial gaps in education at the completed secondary and tertiary education level plays an important part in perpetuating the cycle of inequality. We examine this in section 33.7 below.

7 

College here includes TVET, CET, and private colleges.

The Economics of Education in South Africa    717

Proportion of post-secondary qualifications

.5

.4

.3

.2

.1

0 1950

1955 NQF level 5

1960

1965

NQF level 6

1970 Year of Birth NQF level 7

1975

1980

NQF level 8

1985

1990

NQF level 9 or 10

Figure 33.4  Proportion of post-​secondary qualifications by level and year of birth Source: Own calculations using Community Survey 2016; three-​year moving averages.

33.5  Education Funding and Private-​user Fees Around 15.1 million South Africans were enrolled in some form of education in 2019,8 representing around 26 per cent of the population or a gross enrolment ratio of 59 per cent using the population aged 5–​29 as the denominator (DBE 2020a; DHET 2020a). Seventy per cent of enrolment is in compulsory education (grades R–​9), with a further 19 per cent of enrolment at NQF levels 2–​4. Enrolment in post-​secondary education (NQF level 5 and higher) represents only 9 per cent of total enrolment. Most primary and secondary schools (hereafter schools) in South Africa are publicly funded. In 2019, 92 per cent of the approximately 25,000 schools in South Africa were classified as public, with the remaining 8 per cent classified as independent (i.e. private schools) (DBE 2020a). Furthermore, 95 per cent of learners attend public schools, close 8  Information in this section draws from the most recent published institutional reports—​the DBE school realities report of learners in 2019 and the DHET Statistics for Post-​school Education and Training in South Africa, 2018.

718    Nicola Branson and David Lam to 12.5 million learners. Enrolments at NQF levels 2–​4 (grades 10–​12 or NC(V) 2–​4) are concentrated in the schooling system, with only 12.7 per cent of learners at this level enrolled in colleges (DBE 2020a; DHET 2020a). Similarly, the share of students enrolled in NQF 5–​7 (post-​secondary) education at colleges is relatively low (26 per cent), with most students at that level enrolling in universities. Provision of post-​secondary education, particularly undergraduate enrolment, is predominantly provided by public universities (77 per cent). Provision of education in the South African system varies across institutions on several dimensions, many originating from inequalities in the past. One dimension is differences in funds available and the ratio of public to private funding (typically user fees) inputs.

33.5.1 The National Education Budget In 2018/​19, the education departments spent about Rand 335 billion on education, around 6.8 per cent of GDP), 5 per cent on basic education and 1.8 per cent on post-​ secondary education and training (Department of National Treasury, 2019). Seventy-​ three per cent of the education budget went into the basic education schooling system, 3 per cent to TVET colleges, 0.7 per cent to CET colleges, and 10 per cent to universities. A further 6.8 per cent of the education budget was allocated to the National Student Financial Aid Scheme (NSFAS), which funds students from families with annual household incomes below R350,000 enrolled in TVETs and universities. The remaining 5.7 per cent of the education budget went to skills development institutions, including the National Skills Fund, the National Skills Authority, and twenty-​one Sector Education Training Authorities (SETAs).

33.5.2 Education Funding and User Fees User fees are an important dimension of education funding. Public schools are divided into quintiles based on the socio-​economic level of the school’s neighbourhood. Schools in quintiles 1–​3 are classified as no-​fee schools and are prohibited from charging fees. Schools in quintiles 4 and 5 may charge fees determined by School Governing Bodies and voted on by parents (SASA, section 21). The National Norms and Standards for School Funding (DBE 2006, amended last in 2014) allocates non-​personnel public expenditure to schools on a pro-​poor basis; 60 per cent of non-​personnel resources are allocated to the poorest 40 per cent of learners, such that lower quintile schools receive higher allocations per learner. Personnel expenditure allocations, however, are based on the number of posts at the school and the education and experience levels of the teachers and principal. Motala and Carel (2019) show that schools in quintiles 4 and 5 tend to attract more highly qualified teachers and principals, and therefore command a higher share of the personnel expenditure budget. They show that with 80 per cent of

The Economics of Education in South Africa    719 the education budget assigned to personnel expenditure, this results in a total learner allocation of public funds that is pro-​rich and not redistributive. Furthermore, fee-​paying schools have discretionary funds from fees that can be used to improve school resources and employ additional teachers and support staff (Motala and Carel 2019). University income is comprised of public funds, student tuition fees, and third-​stream income.9 Student fees accounted for about a third of university incomes in 2018, up from around a quarter in 2000 (DHET 2020b). The increased reliance on fees was largely in response to decreasing state funding, which fell from 49 per cent of total university income in 2000 to 39 per cent in 2015 (DHET 2020b) and was a catalyst for student protests between 2014 and 2017. Protests culminated in the announcement by former president Zuma of ‘fee-​free education’ for families with incomes below R350,000 via the NSFAS. The NSFAS is a public entity under the Department of Higher Education and Training that provides bursaries to academically eligible students from households with incomes below R350,000 and further subsidizes the NC(V) or NATED/​Report 191 (N1 to N6) courses at 80 per cent of the total programme cost at TVET colleges (DHET 2020b). Prior to 2018, funding for students was provided as a loan, part of which could be converted to a bursary upon good academic performance. The number of students funded via NSFAS grew from 25,574 in 1994 to 586,763 in 2018, with reports that this growth was not sufficient to cover all eligible students in need of funding (Barberton et  al. 2016). In addition, the programme was criticized for burdening already socio-​ economically disadvantaged students with debt, poor administration that resulted in late payment, and for providing insufficient loan amounts to cover the full cost of studying (FCS). The 2018 NSFAS policy change addressed these issues by committing to providing bursaries covering the FCS to all first-​time entering students from families with household income below R350,000. Debates about whether this commitment is financially sustainable continue in public and policy circles. The tuition fee component of university income has therefore always been partly state funded via the NSFAS, but the ratio of public to privately funded tuition fees has grown substantially over the past two decades, with large variations in reliance on state-​funded tuition fees across institutions (Barberton et al. 2016).

33.6  Academic Performance and Inequality within the System Figure 33.1 showed improvements in average educational attainment in South Africa. Yet, these improvements conceal low levels of cognitive learning and stark inequalities in learner performance. International tests with standardized measures show that 9 

Third-​stream income includes donations, endowments, investment income, and money generated through research and entrepreneurial activities (DHET 2019).

720    Nicola Branson and David Lam cognitive learning in most South African schools is extremely poor by international standards, and very unequal. This has implications for performance on the national matriculation exam and limits post-​secondary options and subsequent labour market outcomes, perpetuating the cycle of inequality.

33.6.1 Performance on Standardized Tests South Africa scores near the bottom on international tests such as TIMSS and PIRLS, and performs poorly even when compared to other African countries. In the 2007 SACMEQ study10 of grade 6 learners, South Africa ranked 10th in reading and 8th in mathematics out of fifteen countries, worse than low-​income countries such as Kenya, Tanzania, Swaziland, and Zimbabwe. Twenty-​seven per cent of South African 6th graders were classified as functionally illiterate and 40 per cent as functionally innumerate (Moloi and Chetty 2010). Learner outcomes, as measured by international tests, have improved over time, albeit from a very low base (Gustafsson 2020; Van der Berg and Gustafsson 2019)11. Gustafsson (2020) shows that once the 2011 PIRLS test is calibrated to be comparable with the 2006 and 2016 tests, average grade 4 reading scores increased from 250 in 2006 to 295 in 2011 and to 320 in 2016. TIMSS grade 9 mathematics scores increased from 285 in 2003 to 350 in 2011 and 370 in 2016 (Gustafsson 2020). Improvements on the NSC are also evident. Van der Berg and Gustafsson (2019) show that the number of African matriculants achieving high-​level mathematics scores12 increased by 65 per cent between 2002 and 2016, an increase higher than the increase in the number of Africans writing matric.

33.6.2 Inequality in Performance across Schools Performance across South African schools is extremely unequal. Van der Berg and Gustafsson (2019) examine inequalities in performance on international tests within the school system. In 47 per cent of schools, not a single learner reached the TIMSS Intermediate International Benchmark of 475. This is almost double that of the second most unequally performing country, Saudi Arabia. Furthermore, in the next 33 per cent of schools, while at least one learner meets the intermediate benchmark, none meet the high international benchmark (550). Similarly, they show that in the SACMEQ 2007

10 

Spaull and Pretorius (2019) note that the quality and comparability of the SAQMEQ 2013 data has not yet been established. We therefore defer to the 2007 data. 11  The TIMSS and PIRLS low international benchmark is 400, representing that students have a basic knowledge of mathematics and literacy. Average scores for South Africa are below this level in both TIMSS and PIRLS. 12  Mathematics scores above an annual threshold determined by the top performing 20 per cent of White learners.

The Economics of Education in South Africa    721 study, South Africa has the second-​to-​highest level of inequality in performance after Mauritius, out of fifteen countries included. To analyse the nature of inequality in performance across schools, Van der Berg and Gustafsson (2019) regress SACMEQ 2007 test scores on the average socio-​economic status of the school and its square. They find that close to half of the inequality in performance across South African schools is explained by the average wealth/​socio-​ economic status of the school, double that apparent in Mauritius and argue that once school socio-​economic characteristics are controlled for, individual socio-​economic variables explain very little further variation in inequality. The school a learner attends will therefore overwhelmingly affect their performance, and this is largely determined by where a learner is born.13

33.6.3 Factors Associated with Performance Differentials International studies show that quality early childhood development (ECD) can play an important part in levelling the playing field for learners entering primary education (see van Huizen and Plantenga (2018) for a review). The South African national government recognizes this in policy; the 2030 NDP targets include universal access to two years of pre-​primary education and general ECD services for children aged 0–​3 and, in 2015, the Integrated Early Childhood Development Policy, a multi-​sectoral comprehensive package of ECD services, came into effect (Republic of South Africa 2015). However, implementation of ECD falls short of policy intentions and provision varies across space, income level, and race in terms of access, length of exposure, quality of provision, and state funding (Hall et al. 2017; Gustafsson 2017; Ashley-​Cooper et al. 2019). For example, in 2018, 66 per cent of children aged 3–​5 living in the poorest 20 per cent of households were attending an early learning programme, compared to 82 per cent of children living in the richest 20 per cent (own calculations, Statistics South Africa 2019). As a result, current ECD provision typically further dis-​equalizes opportunities for children from poorer backgrounds (van der Berg et al. 2013). Teacher subject matter and pedagogical knowledge have been shown to be lacking in South Africa (Taylor and Taylor 2013; Hoadley 2016). Consequently, learners in many South African schools struggle to acquire the foundational numeracy and literacy skills for subsequent learning. Most foundation teachers in South Africa are not taught to teach ‘reading for meaning’, a task the Funda Wande: Teaching Reading for Meaning project has set out to address (Spaull 2016). The right to learn in any of the eleven official South African languages (RSA 1996)14 adds an additional complexity, as teachers have limited opportunities to learn how to teach reading, particularly in African languages. 13 

Eighty-four per cent of learners attended their closest school (Hall 2019). Language in Education Policy of 1997 qualifies, however, that this right can only be realized where reasonably practical, defined as when there are forty (thirty-​five) or more learners within a primary (secondary) grade requesting the language of instruction. 14  The

722    Nicola Branson and David Lam The prioritization of improving learner outcomes is exemplified in the NDP target that schools will have better educational outcomes and every 10-​year-​old child will be able to read for meaning by 2030. State and donor interventions have thus focused on developing and evaluating the impact of structured pedagogical interventions (Cilliers et al. 2020) that provide curriculum-​aligned learning materials such as graded readers (not readily available in all African languages), lesson plans, and teacher professional development, found to be successful in improving foundation numeracy and literacy in international research (e.g. Popova et al. 2018).

33.6.4 Access and Performance at University and College Differences in cognitive learning during primary and secondary school limit choices for post-​secondary skill attainment. Figure 33.3 showed that the proportion of Africans and coloureds converting their grade 12 into a post-​secondary qualification has decreased steadily over the past four decades, while the rate has remained constant for Indians and whites. Furthermore, Africans who continue their studies after school are more likely to enrol in college than university. Performance on the NSC determines eligibility to study at university and is unequal across schools. In 2019, for example, 37–​40 per cent of learners in quintile 1–​3 schools and 45 per cent in quintile 4 schools who passed the NSC achieved a bachelor’s pass, compared to 61 per cent in quintile 5 and 58 per cent in independent schools (DBE 2020b). Van Broekhuizen et al. (2019) show the strong link between NSC performance—​both in average score and whether the learners took mathematics and physical science—​and university access. Only 9.2 per cent of learners from quintile 1 schools who took the 2008 NSC exam enrolled in a public university within six years, compared to 45.2 per cent of learners from quintile 5 schools. However, restricting the sample to those who passed the NSC with a bachelor’s pass, the differences were reduced, with 63 per cent of quintile 1 and 70 per cent of quintile 5 learners enrolling in the public university system within six years. Controlling further for NSC score and subjects taken does not remove the difference completely, however, as learners from poor schools were less likely to enrol at university. Two other studies using longitudinal data support the finding that academic performance is the main factor determining post-​secondary access—​for enrolment in both college and university—​though credit constraints continue to play a part (Lam et al. 2013; Branson and Kahn 2019). Furthermore, Branson and Kahn (2019) find that the relationship between enrolment in college and test scores represents an inverted U-​shape for low-​and high-​ income learners but increases linearly for middle-​income students. These findings are consistent with the NSFAS funding policy prior to 2018. High performing students from low-​income households had access to NSFAS funds, and therefore

The Economics of Education in South Africa    723 could choose university over college, while those in middle-​income households did not, and therefore appear to have opted for the less-​expensive college option. The NSFAS model from 2018 is designed to alleviate the financial constraint from students’ decisions about whether, when, and where to enrol in post-​secondary education. Completion rates within the university sector have improved and dropout rates have decreased over the past decade but continue to vary by mode (contact versus distance), gender, qualification programme, and race (DHET 2019). For example, tracking the 2008 entrance cohort for ten years, the DHET (2019) show that the dropout rates for contact students was 27.7 per cent, compared to 70 per cent for distance-​learning students. Women outperform men in all qualifications, and white and Indian students outperform their African and coloured counterparts in all qualifications (DHET 2019). Racial differences mainly reflect differences in academic performance on entering university, however. Van Broekhiuzen et  al. (2019) show that once school quintile, NSC score, and whether the learner took mathematics and physical science subjects are taken into account, African students are less likely than white students to drop out and are equally likely to complete their qualification in six years. White students retain certain advantages, however. For example, they are more likely to enrol in degree programmes—​the qualification with the highest labour market rewards—​and are more likely to complete their qualifications in four years. This advantage may reflect differences in educational preparedness not accounted for in the model, or other challenges discussed in the literature (Branson et al. 2015). Beyond academics, the background socio-​ economic characteristics of students enrolled in the TVET college system are similar to learners who attempt matric but do not enrol in education thereafter (Branson and Kahn 2019). University students, on the other hand, come from, on average, more advantaged backgrounds, have parents with higher educational attainment, and reside in households with higher income levels during their matric year. The NDP goal to increase enrolment in TVET colleges to 2.5 million students by 2030 illustrates that government recognizes the TVET colleges as providers of an alternative education pathway for students. However, although enrolment in TVET colleges more than doubled between 2010 and 2015, there has been a decline in enrolment since (DHET 2020a). Many of the public colleges have encountered management hurdles, limited teaching capacity, and funding constraints (DHET 2013). In addition, colleges have, to date, maintained a reputation as being institutions that provide inferior education to those who do not have the academic marks to attend university. Confusing course structures and application processes, lack of student support, low course completion rates, and the disconnect between course content and the skills demanded in the labour market have not helped to improve the status of South African public colleges.

724    Nicola Branson and David Lam

33.7  The Relationship between Education, Employment, Earnings, and Inequality The previous sections have analysed some of the important economic determinants of education outcomes. In particular, we have seen how inequality in income and resources is associated with large disparities in educational attainment and the quality of education received. In this section, we look at how inequality in education in turn affects inequality in employment and earnings when learners enter the labour market. We begin by looking at the distribution of education in the working-​age population. We then look at the relationship between education and labour market outcomes. Last, we discuss some of the implications for income inequality in South Africa and its transmission across generations.

33.7.1 The Distribution of Education in the Working-​age Population The improvements in education across birth cohorts, documented in Figures 33.1–​33.3, imply that the distribution of education in the working-​age population has also been increasing. Figure 33.5 shows, for 1993 to 2018, mean years of education of the working-​ age population (defined as ages 25–​59 to allow most people to complete their education) and the proportion within each category of completed education. Mean years of completed education (highest grade completed) in the working-​age population has risen steadily from 7.7 years in 1993 to 10.5 years in 2018. There have been significant declines in the proportion of the working-​age population with no schooling (13 per cent to 3 per cent) and the proportion with one to seven years (30 per cent to 12 per cent), and significant increases in the proportion with grade 12 (14 per cent to 31 per cent) and beyond grade 12 (9 per cent to 16 per cent). The proportion with incomplete secondary, grades 8–​11, has remained relatively constant, as movement into the category from lower grades has replaced movement out of the category into higher grades.

33.7.2 The Relationship between Education and Employment As documented in other chapters in this volume (Mlatsheni, ­Chapter 32; Heintz and Naidoo, ­Chapter 7) South Africa has some of the highest unemployment rates in the world. The probability of being employed is strongly associated with education. Figure

The Economics of Education in South Africa    725 Education distribution of working-age population (25–29) by year 12

10

Mean

11

Proportion

9 8 .4 .3 .2 .1 0 1993 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015 2017 Year Mean years of Education Grade 12

Beyond Grade 12

Incomplete Secondary (8–11) Grades 1–7

No Schooling

Figure 33.5  Education distribution of working-​age population (aged 25–​59) by year Source: Own calculations using Post Apartheid Labour Market Series PALMS and General Household Survey; three-​year moving averages.

33.6, below, shows the proportion of the population employed by years of completed education for Africans and whites aged 25–​59 in 1994/​95 and 2017/​18.15 The figure shows that less than 50 per cent of working-​age Africans with under eight years of completed schooling were working in both 1994/​95 and 2017/​18. The proportions working at these lower levels of schooling are surprisingly constant in the two periods.16 This constancy masks some changes when looking at men and women separately (not shown). For example, for African men with under eight years of schooling, the proportion employed fell over this 24-​year period from about 60 per cent to 50 per cent. For African women with under eight years of schooling, the proportion employed increased from about 25 per cent to 35 per cent. As seen in Figure 33.6, employment probabilities remain relatively flat for Africans until around grade 12, when they begin to increase substantially. Whites

15  We pool two years to increase sample size, given small numbers of observations for some years of education. 16  We do not show White employment rates below eight years of schooling due to the very small sample sizes.

726    Nicola Branson and David Lam Relationship between schooling and employment, 1994–95 and 2017–18, Ages 25–59 1 .9 .8

Proportion working

.7 .6 .5 .4 .3 .2 .1 0 0

1

2

3

1994–95 White

4

5

6 7 8 9 Highest grade completed

2017–18 White

10

11

1994–95 African

12

13

14

15

2017–18 African

Figure 33.6  Relationship between schooling and employment, 1994/​95 and 2017/​18, ages 25–​59 Source: Own calculations using PALMS. Notes: Highest grade completed represents school grades between grades R/​0 and 12, with NQF levels 5 and 6 represented by highest grades 13 and 14, and NQF levels 7–​10 (university) classified as grade 15. NC(V) 2–​4 and NATED 1–​3 are classified as grades 10–​12.

have relatively similar employment rates as Africans at grades 8 and 9, but white employment increases faster in the incomplete secondary grades, with whites having significantly higher employment at grade 12. The African schooling–​employment gradient is very steep above grade 12, with Africans who have completed university (shown as grade 15) having almost the same high employment rates as whites, around 90 per cent. Indeed, there has been a lively debate (see Van Broekhuizen (2016) for a summary) about the level of graduate unemployment in South Africa. As is evident from Figure 33.6, however, the problem has been much exaggerated and was partly a result of people with pre-​NQF level 5 certifications being classified as graduates. While white labour force participants have an employment advantage at schooling levels around grades 11 and 12, the racial differences at most grades are not enormous. Although in 2018, the overall employment rate for Africans was 54 per cent, compared to 76 per cent for whites, whites’ higher education levels are more important in explaining this difference than are racial differences in employment rates at specific levels of education.

The Economics of Education in South Africa    727

33.7.3 The Relationship between Education and Earnings In addition to the strong relationship between education and employment shown in Figure 33.6, there is a strong relationship between education and earnings for those who have jobs. High rates of return to education in South Africa have been the focus of an extensive literature, with particular focus on the high returns to post-​secondary education, some of the highest in the world (Mwabu and Schultz 1996, 2000; Keswell and Poswell 2004; Branson and Leibbrandt 2013; Salisbury 2016; Patrinos and Psacharopoulos 2020). Figure 33.7 shows the relationship between earnings and highest grade completed for Africans and whites in 1994/​95 and 2017/​18. The figure is based on earnings regressions in which the natural logarithm of earnings is regressed on dummy variables for each year of completed schooling, age, age squared, a female dummy, and province dummies, with separate regressions for Africans and whites in each period. As with standard log-​earnings regressions, a 0.1 point difference in log earnings translates into approximately a 10 per cent difference in earnings. We do not present estimates for

Log earnings minus log earnings of Africans with zero schooling

Relationship between schooling and earnings, 1994–95 and 2017–18, Ages 25–59 2.5

2

1.5

1

.5

0 0

1

2

3

1994–95 White

4

5

6 7 8 9 Highest grade completed

2017–18 White

10

11

1994–95 African

12

13

14

15

2017–18 African

Figure 33.7  Relationship between schooling and earnings, 1994/​95 and 2017/​18, ages 25–​59 Source: PALMS and General Household Survey. Notes: Highest grade completed represents school grades between grades R/​0 and 12, with NQF levels 5 and 6 represented by highest grades 13 and 14, and NQF levels 7–​10 (university) classified as grade 15. NC(V) 2–​4 and NATED 1–​3 are classified as grades 10–​12.

728    Nicola Branson and David Lam whites below nine years of schooling due to small cell sizes. The figure shows predicted log earnings at each schooling level relative to predicted log earnings for Africans with zero schooling in that year (only individuals with positive earnings are included). For example, in 1994/​95, Africans who had completed grade 10 had 0.96 higher mean log earnings than Africans with zero schooling. Exponentiating this log difference, this translates into 2.6 times higher earnings for Africans with grade 10. Whites had 1.6 higher mean log earnings than Africans with zero schooling, or 5.1 times higher earnings. Whites with grade 10 education earned 1.9 times higher earnings than Africans with grade 10 in 1994/​95. Figure 33.7 shows the strong relationship between education and earnings in both periods, with high returns for both population groups. For Africans in 1994/​95, each year of additional schooling increased earnings by around 15 per cent for grades between 4 and 11, very high by international standards (Patrinos and Psacharopoulos 2020), with even higher returns for grade 12 and above. Although whites had an earnings advantage over Africans at every grade level in 1994/​95, the education–​earnings gradient was somewhat flatter for whites. An important feature of Figure 33.7 is the decline over time in returns to schooling for Africans through grade 11. This has been observed for some time in previous studies (Moll 1996; Keswell and Poswell 2004; Salisbury 2016). The return to an additional year of schooling in the grades 5–​11 range fell from rates of around 15 per cent in 1994/​95 to rates of 5–​7 per cent in most grades by 2017/​18. While the rate of return to these grades declined over time, the rate of return to grade 12 and higher increased substantially. The earnings advantage of university-​educated Africans (shown as grade 15) over grade-​12-​educated Africans rose from an already high 2.1 times higher earnings in 1994/​95 to 3.4 times higher earnings in 2017/​18. The increasing returns to tertiary education are an important part of the education–​earnings relationship in South Africa and have been observed in the United States and most other countries (Patrinos and Psacharopoulos 2020). This increase in returns to schooling at the highest levels of schooling has been an important contributor to increasing earnings inequality in many of these countries (e.g. Juhn et al. 1993; Lam and Levison 1991), and has had a dis-​equalizing effect in South Africa as well (Lam et al. 2015; Branson et al. 2012). Comparing the African and white lines in Figure 33.7, there is a large earnings gap at every level of education. In contrast to the data on employment in Figure 33.6, large gaps in earnings at each education level are an important factor explaining the overall African–​white earnings gap, over and above the higher levels of education among whites. The earnings gap is smaller at the highest levels of education, however, and there is evidence of a narrowing of the gap over time. Looking at those with university education, whites had 47 per cent higher earnings than Africans in 1994/​ 95, but this had decreased to 16 per cent in 2017/​18, controlling for age, gender, and province.

The Economics of Education in South Africa    729

33.8  Conclusion: Education and the Intergenerational Transmission of Inequality Linking this chapter with themes found in many of the other chapters in this volume, it is clear that education plays a central role in explaining South Africa’s high levels of inequality in employment and income. This is the result of both inequality in education outcomes, as documented throughout this chapter, and impacts of education on employment and income that are among the strongest in the world. While the increases in education in the working-​age population, shown in Figure 33.5, are an important sign of progress, they have also had the effect of pushing the population into the range of education where the earnings–​education gradient is the steepest. This change can have a dis-​equalizing effect, even though the improvements in education, in and of themselves, may have had an equalizing effect. The large increase in returns to post-​secondary education, shown in Figure 33.7, have also had a significant dis-​equalizing effect. Lam (2020) shows that many countries in the world have had declining inequality in education accompanied by rising inequality in earnings, with the increase in returns to post-​ secondary education playing a major role. South Africa is following this pattern, with the slow increase in post-​secondary education contributing to the high and rising rates of return. Inequality in income among parents strongly translates into inequality in education outcomes for the next generation. The large disparities we have documented in grade attainment and performance are heavily influenced by the disparities in access to resources among parents and the school a child subsequently attends. Continued policy emphasis on improving learner acquisition of foundational numeracy and literacy skills, particularly in quintile 1–​3 schools, should therefore remain a key priority. To date, careful evaluation work has shown that there is potential, through carefully crafted interventions, to improve the classroom learning environment. The next crucial step is to implement these interventions nationally and monitor the quality with which they are integrated into classroom practices. The importance of developing learning materials and pedagogical knowledge and procedure in African languages has also been raised as a formidable challenge to the South African schooling system. A  review of the language policy, with the possible adoption of a few languages that can be taught well may be the way forward. The distribution of state funds within the education sector provides concrete insight into policy priority. The commitment of resources to ECD and the college sector are not as strong as they should be, given the importance of these two sectors in the South African education landscape and policy discussion. While rigorous empirical evidence of the importance of ECD in South Africa is still limited, international literature

730    Nicola Branson and David Lam highlights the fact that quality ECD can contribute significantly to levelling the playing field for children from unequal socio-​economic backgrounds. At the other end of the education spectrum, the current reality is that 40 per cent of learners will not complete matric or will not be eligible to study at university. The college system has the potential to provide these learners with further education that could improve their labour market and life outcomes. Current college-​programme provision falls far short of this, however. Commitment of state resources focused on improving the quality of education provided in these institutions needs to be on a par with that in basic schooling and university education. South Africa continues to be in a cycle where space, income, and racial inequalities determine the quality of education an individual receives, reproducing inequalities in employment and earnings across generations. The critical role of education in determining intergenerational welfare, makes addressing inequality in education opportunities of fundamental importance to South Africa.

References Anderson, Kermyt, Anne Case, and David Lam. 2001. ‘Causes and consequences of schooling outcomes in South Africa: Evidence from survey data’, Social Dynamics, 27(1): 1–​23. Ashley-​Cooper, Michaela, Lauren-​Jayne van Niekerk, and Eric Atmore. 2019. ‘Early childhood development in South Africa:  Inequality and opportunity’, in Nic Spaull and Jonathan Jansen (eds) South African Schooling:  The Enigma of Inequality. Switzerland:  Springer Nature. Barberton, Conrad, Jonathan Carter, Carmen Abdoll, Fiona Lewis, and Charles Sheppard. 2016. Performance and Expenditure Review:  National Student Financial Aid Scheme (NSFAS)’. Cornerstone Economic Research. Branson, Nicola and Amy Kahn. 2019. ‘The post matriculation enrolment decision: Do public TVET colleges provide students with a viable alternative?’, in Mike Rogan (ed.) Post-​School Education and the Labour Market in South Africa. Cape Town: HSRC Press. Branson, Nicola, and David Lam. 2010. ‘Education inequality in South Africa: Evidence from the National Income Dynamics Study’. Studies in Economics and Econometrics, 34(3): 85–​109. Branson, Nicola, and Murray Leibbrandt. 2013. ‘Educational attainment and labour market outcomes in South Africa, 1994–​2010’. OECD Economics Department Working Paper no. 1022, Paris. Branson, Nicola, Samantha Culligan, and Kim Ingle. 2020. ‘Developing Siyaphambili:  A stronger South African nation website: Moving towards a unified goal to combat inequality and unemployment’. SALRDU Report no. 20/​01, University of Cape Town. Branson, Nicola, Julia Garlick, David Lam, and Murray Leibbrandt. 2012. ‘Education and inequality:  The South African case’. SALDRU Working Paper no.  75, University of Cape Town. Branson, Nicola, Clare Hofmeyr, and David Lam. 2014. ‘Progress through school and the determinants of school dropout in South Africa’, Development Southern Africa, 31(1): 106–​26. Branson, Nicola, Clare Hofmeyr, Joy Papier, and Seemus Needham. 2015. ‘Post-​school education: Broadening alternative pathways from school to work’, in A. De Lannoy et al. (eds) South African Child Gauge. Cape Town: Children’s Institute, University of Cape Town.

The Economics of Education in South Africa    731 Burger, Rulof, Servaas van der Berg, and Dieter von Fintel. 2015. ‘The unintended consequences of education policies on South African participation and unemployment’, South African Journal of Economics, 83(1): 74–​100. Cilliers, Jacobus, Brahm Fleisch, Cas Prinsloo, and Stephen Taylor. 2020. ‘How to improve teaching practice? An experimental comparison of centralized training and in-​classroom coaching’, Journal of Human Resources, 55(3): 926–​62. Department of Basic Education (DBE). 2006. ‘National norms and standards for school funding (NNSSF)’. Government Notice, 31 August, Pretoria. Department of Basic Education (DBE). 2020a. ‘School realities, 2019’, January, Pretoria [online] https://​www.education.gov.za/​Portals/​0/​Documents/​Reports/​School%20Realities%20 2019%20Final%20.pdf?ver=2020-​02-​07-​101051-​330. Department of Basic Education (DBE). 2020b. ‘Report on the 2019 National Senior Certificate Examination’. January, Pretoria [online] https://​www.education.gov. za/​ P ortals/​ 0 /​ D ocuments/​ R eports/​ 2 019%20NSC%20Examination%20Report. pdf?ver=2020-​01-​07-​155811-​230. Department of Higher Education and Training (DHET). 2013. White Paper for Post-​School Education and Training: Building an Expanded, Effective and Integrated Post-​School System. Pretoria. Department of Higher Education and Training (DHET). 2019. ‘Post-​school education and training monitor: Macro-​indicator trends’. March, Pretoria. Department of Higher Education and Training (DHET). 2020a. ‘Statistics for post-​school education and training in South Africa: 2018’. March, Pretoria. Department of Higher Education and Training (DHET). 2020b. ‘Amended rules and guidelines for the administration and management of the Department of Higher Education and Training Technical and Vocational Education and Training College Bursary Scheme for 2020’. Pretoria. Department of National Treasury. 2019. 2018/​ 2019 National Treasury Annual Report. September, Pretoria. Finn, Arden, Murray Leibbrandt, and Vimal Ranchhod. 2016. ‘Patterns of persistence:  Intergenerational mobility and education in South Africa’. SALDRU Working Paper no. 175 and National Income Dynamic Study (NIDS) Discussion Paper no. 2016/​2, University of Cape Town. Fiske, Edward, and Helen Ladd. 2004. Elusive Equity:  Education Reform in Post-​apartheid South Africa. Washington, DC: Brookings Institution Press. Gustafsson, Martin. 2017. ‘Enrolments, staffing, financing and the quality of service delivery in early childhood institutions’. Unpublished. Gustafsson, Martin. 2020. ‘A revised PIRLS 2011 to 2016 trend for South Africa and the importance of analysing the underlying microdata’. RESEP Working Paper Series no. WP02/​2020, Stellenbosch University. Hall, Katherine. 2019. ‘Child count: Statistics on children in South Africa: Children living far from school indicator’. Children’s Institute, University of Cape Town. Hall, Katherine, Winnie Sambu, Lizette Berry, Sonja Giese, and Colin Almeleh. 2017. ‘South African early childhood review 2017’. Children’s Institute, University of Cape Town and Ilifa Labantwana, Cape Town. Hoadley, Ursula. 2016. ‘A review of the research literature on teaching and learning in the foundation phase in South Africa’. Research on Socioeconomic Policy (ReSEP) Working Paper no. 05/​16, Stellenbosch University.

732    Nicola Branson and David Lam Juhn, Chinhui, Kevin M. Murphy, and Brooks Pierce. 1993. ‘Wage inequality and the rise in returns to skill’, Journal of Political Economy, 101(3): 410–​42. Kerr, Andrew, David Lam, and Martin Wittenberg. 2019. Post-​Apartheid Labour Market Series 1993–​2019 [dataset] Version 3.3. DataFirst, Cape Town. Keswell, Malcolm, and Laura Poswell. 2004. ‘Returns to education in South Africa: A retrospective sensitivity analysis of the available evidence’, South African Journal of Economics, 72(4): 834–​60. Kika, Jesal, and Janeli Kotze. 2019. ‘Unpacking grade repetition patterns in light of the progression policy in the further education and training Phase’. SALDRU Working Paper no. 243 and NIDS Discussion Paper no. 2019/​10, SALDRU, University of Cape Town. Lam, David. 2020. ‘Why has income inequality increased while education inequality has decreased in many developing countries?’, in Sisay Asefa and Wei-​Chiao Huang (eds) The Political Economy of Inequality: U.S. and Global Dimensions. Kalamazoo, MI: W.E. Upjohn Institute for Employment Research. Lam, David, and Deborah Levison. 1991. ‘Declining inequality in schooling in Brazil and its effect on inequality in earnings’, Journal of Development Economics, 37(1–​2): 199–​225. Lam, David, Cally Ardington, Nicola Branson, and Murray Leibbrandt. 2013. ‘Credit constraints and the racial gap in post-​secondary education in South Africa’. SALDRU Working Paper no. 111, University of Cape Town. Lam, David, Cally Ardington, and Murray Leibbrandt. 2011. ‘Schooling as a lottery:  Racial differences in school advancement in urban South Africa’, Journal of Development Economics, 95(2): 121–​36. Lam, David, Arden Finn, and Murray Leibbrandt. 2015. ‘Schooling inequality, returns to schooling, and earnings inequality:  Evidence from Brazil and South Africa’. WIDER Working Paper Series wp-​2015-​050, World Institute for Development Economic Research. Moll, Peter G. 1996. ‘The collapse of primary schooling returns in South Africa, 1960–​90’, Oxford Bulletin of Economics and Statistics, 58(1): 185–​209. Moloi, Meshack, and Mark Chetty. 2010. ‘The SACMEQ III project in South Africa: A study of the conditions of schooling and the quality of education: South Africa country report’. Department of Basic Education, Pretoria. Motala, Shireen, and David Carel. 2019. ‘Educational funding and equity in South African schools’, in Nic Spaull and Jonathan Jansen (eds) South African Schooling: The Enigma of Inequality. Switzerland: Springer Nature. Mwabu, Germano, and Paul Schultz. 1996. ‘Education returns across quantiles of the wage function:  Alternative explanations for returns to education by race in South Africa’, American Economic Review, 86(2): 335–​39. Mwabu, Germano, and Paul Schultz. 2000. ‘Wage premiums for education and location of South African workers, by gender and race’, Economic Development and Cultural Change, 48(2): 307–​34. Papier, Joy. 2009. ‘Getting the right learners into the right programmes: An investigation into factors that contributed to the poor performance of FET college learners in NCV 2 and NCV 3 programmes in 2007 and 2008–​–​reasons and recommendations’. FETI, University of Western Cape. Patrinos, Harry Anthony, and George Psacharopoulos. 2020. ‘Returns to education in developing countries’, in Steve Bradley and Colin Green (eds) The Economics of Education, 2nd ed. London and Cambridge, MA: Academic Press.

The Economics of Education in South Africa    733 Pillay, Pundy. 1984. ‘The development and underdevelopment of education in South Africa’. Carnegie Conference Paper no. 95, Cape Town, 13–​19 April. Piraino, Patrizio. 2015. ‘Intergenerational earnings and inequality of opportunity in South Africa’, World Development, 67: 396–​405. Popova, Anna, David Evans, Mary Breeding, and Violeta Arancibia. 2018. ‘Teacher professional development around the world:  The gap between evidence and practice’. World Bank Policy Research Working Paper (August) [online] https://​doi.org/​10.1596/​ 1813-​9450-​8572. Republic of South Africa. 1995. ‘South African Qualifications Authority Act No. 58 of 1995’. Government Gazette, 8 September. Republic of South Africa. 1996a. ‘Constitution of the Republic of South Africa, Act 108 of 1996’. Republic of South Africa. 1996b. ‘South African School Act No. 84 of 1996’. Government Gazette, 15 November, Pretoria. Republic of South Africa. 1998. ‘Further Education and Training Act No. 98 of 1998’. Government Notice, 2 November, Pretoria. Republic of South Africa. 2011. ‘Employment of Educators Act No. 76 of 1998’. Pretoria: Government Gazette, 19 September. Republic of South Africa. 2015. ‘National integrated early childhood development policy’. Government Printers, 9 December, Pretoria. Salisbury, Taylor. 2016. ‘Education and inequality in South Africa: Returns to schooling in the post-​apartheid era’, International Journal of Educational Development, 46: 43–​52. Spaull, Nic. 2016. ‘Learning to read and reading to learn’. Research on Socioeconomic Policy (RESEP) Policy Brief, Stellenbosch. Spaull, Nic, and Nwabisa Makaluza. 2019. ‘Girls do better:  The pro-​female gender gap in learning outcomes in South Africa 1995–​2018’, Agenda, 33(4): 11–​28. Spaull, Nic, and Elizabeth Pretorius. (2019). ‘Still falling at the first hurdle: Examining early grade reading in South Africa, in Nic Spaull and Jonathan Jansen. (eds) South African Schooling: The Enigma of Inequality. Cham, Switzerland: Springer Nature. Statistics South Africa. 2017. Community Survey 2016 [dataset]. Version 1.0.0. Pretoria: StatsSA [producer]; Cape Town: DataFirst [distributor] Statistics South Africa. 2019. General Household Survey 2009–​ 2018 [dataset]. Version 1. Pretoria: StatsSA [producer]; Cape Town: DataFirst [distributor] Taylor, Nick, and Stephen Taylor. 2013. ‘Teacher knowledge and professional habitus’, in Nick Taylor, Servaas van den Berg, and T. Mabogoane (eds) Creating Effective Schools: Report of South Africa’s National Schools Effectiveness Study. Cape Town: Pearson. van Broekhuizen, Hendrik. 2016. ‘Graduate unemployment and higher education institutions in South Africa’. Bureau for Economic Research and Stellenbosch Economic Working Paper no. 08/​16, Stellenbosch. van Broekhuizen, Hendrik, Servaas van der Berg, and Heleen Hofmeyr. 2019. ‘From grade 12 into and through university: Higher education access and outcomes for the 2008 national matric cohort’, in Mike Rogan (ed.) Post-​School Education and the Labour Market in South Africa. Cape Town: HSRC Press. van der Berg, Servaas, and Martin Gustafsson. 2019. ‘Educational outcomes in post-​apartheid South Africa: Signs of progress despite great inequality’, in Nic Spaull and Jonathan Jansen (eds) South African Schooling: The Enigma of Inequality. Switzerland: Springer Nature.

734    Nicola Branson and David Lam van der Berg, Servaas, and Deborah Shepherd. 2010 ‘Signalling performance: An analysis of continuous assessment and matriculation examination marks in South African schools’. Umalusi, May, Pretoria [online] https://​www.umalusi.org.za/​docs/​research/​2009/​signaling_ ​performance.pdf. van Huizen, Thomas, and Janneke Plantenga. 2018. ‘Do children benefit from universal early childhood education and care? A  meta-​analysis of evidence from natural experiments’, Economics of Education Review, 66: 206–​22. van der Berg, Servaas, Elizabeth Girdwood, Debra Shepherd, Chris van Wyk, John Kruger, and Janeli Viljoen. 2013. ‘The impact of the introduction of grade R on learning outcomes’. Report to Department of Basic Education and Department of Performance Monitoring and Evaluation in the Presidency, University of Stellenbosch. van der Berg, Servaas, Martin Gustafsson, and Kholekile Malindi. 2020. ‘Education and skills for the economy and links to labour markets in South Africa’. Research on Socio-​Economic Policy 7, Department of Economics, Stellenbosch University. van der Berg, Servaas, Gabrielle Wills, Rebecca Selkirk, Charles Adams, and Chris van Wyk. 2019. ‘The cost of repetition in South Africa’. Stellenbosch Economic Working Paper no. 13/​ 2019, August.

Chapter 34

Gender and Work i n Sou th Af ri c a Daniela Casale, Dorrit Posel, and Jacqueline Mosomi

34.1 Introduction Historically, access to paid employment and economic resources in South Africa was sharply differentiated by both race and gender. Women’s participation in the labour market was affected not only by legal restrictions on where people of colour could live and work, and the types of work they could do, but further by ‘internal structures of control’ (Walker 1990: 178) in the household and the community. These included gender ideology, women’s economic dependence and social pressure that enabled fathers, husbands, and chiefs to reinforce women’s traditional roles in reproduction and household labour. The transition to democracy saw the removal of many of the structural constraints that inhibited the labour-​force participation of both African women and men, and it ushered in a very progressive constitution that emphasizes dignity, justice, and equality. The South African government further introduced a range of protective labour and equal opportunity legislation intended to redress both race and gender inequality. This includes the Basic Conditions of Employment Act (1997), the Employment Equity Act (1998), and minimum wage legislation in low-​wage employment (for example, for contract cleaners in 1999, domestic workers in 2002, agricultural workers in 2003, and nationally in 2018). Marking a formal policy commitment to reducing gender inequality, South Africa has also been a signatory to numerous regional and international gender protocols (Posel and Casale 2019). In this chapter, we draw on existing research and our own analysis of national micro-​data from 1994 to 2019, to trace changes in women’s participation in both paid and unpaid work. We show that progress over the post-​apartheid period has been uneven. Whilst we describe important and significant advances, for example in a narrowing of the average gender wage gap, and the employment of more women in high-​skilled occupation types, we show also

736    Daniela Casale, Dorrit Posel, and Jacqueline Mosomi that gender divisions of labour in the home and in the paid economy persist. We first review and detail participation in the post-​apartheid labour market from the perspective of gender, and then describe gender differences in household labour and care obligations. At the time of writing, South Africa had emerged from a very strict national lockdown in response to the COVID-​19 pandemic, and we consider how this crisis illuminates the vulnerability of women in the labour market, the importance and the demands of work undertaken in the home, and women’s primary responsibility for this work.

34.2  Gender Differences in the Labour Market Global research on long-​term labour market trends has documented a rise in women’s labour-​force participation relative to men’s, and particularly during the 1980s,1 with women’s share of employment increasing as a result. However, in most countries, men remain considerably more likely than women to be labour-​force participants, and many of the jobs that women have gained have been in precarious or non-​standard forms of employment—​the types of jobs that have also been on the rise in recent decades (ILO 2018). Standing’s (1989, 1999) well-​known reference to the ‘feminization’ of the labour force therefore held an intended double-​meaning: women were increasingly likely to participate in the workforce; and, in a fast-​changing global economy, the work available was increasingly of a ‘flexible’ or ‘informal’ nature, attributes often associated with female labour. In this section, we show that South Africa mostly conforms to these global trends, although significant gains have also been realized. We highlight some of the more distinctive features of women’s rising labour-​force participation, including the very large gap between the number of women who want to work and who find work, and the intersection of gender with race disadvantage in the labour market. Although labour market data collected by South Africa’s statistical agency prior to 1994 were not nationally representative and overlooked large parts of the population living in the former homelands, there is some limited evidence from the pre-​democracy period that women’s participation in the labour market was increasing from around the middle of the twentieth century already (Posel and Todes 1995; Standing et al. 1996). Estimates based on the Population Census, for example, suggested that in 1960 women accounted for 23 per cent of the labour force; by 1985 this figure had risen to 36 per cent; and by 1991 it had grown further to 41 per cent (Standing et al. 1996: 60). In 1994, Statistics South Africa (StatsSA) started collecting nationally representative data that allowed a more rigorous analysis of labour market dynamics, first with the October Household Surveys (OHS) from 1994 to 1999, then with the biannual Labour 1 

In recent decades, however, labour-​force participation rates globally have slowly declined for both men and women, but with a slower decline for women (ILO 2018).

Gender and Work in South Africa    737 Force Surveys (LFS) of 2000 to 2007, and finally with the Quarterly Labour Force Surveys (QLFS) from 2008 onwards. In the early years of post-​apartheid data collection (1994 to 2000 especially), sampling and survey design features changed quite often, unfortunately compromising comparability of some of the key statistics over time (see Mosomi 2018 for more details). Some of this ‘noise’ is evident in the data series we present here, and we refer to this where relevant. However, to ensure that we use the most comparable datasets and definitions available given these limitations, we rely on the Post-​Apartheid Labour Market Series (PALMS) from 1994 to 2019 for our empirical analysis.2

34.2.1 Broad Trends in Labour-​force Participation, Employment, and Unemployment Consistent with the trend of the pre-​democracy period, women’s labour-​force participation continued to increase after 1994. Despite some volatility in the early period (likely due to sampling issues and survey design), Figure 34.1 shows a general upward trend in the percentage of the working-​age population wanting, and searching for, work, with the growth in labour-​force participation rates considerably larger for women than men. Female labour-​ force participation rose from about 40 per cent in 1994 to 54 per cent in 2019, while male labour-​force participation rose from 60 per cent in 1994 to 67 per cent in 2019 (see also selected years of data in Table 34.1). Mirroring global trends, women’s share of the labour force therefore grew over the period, from 42 per cent in 1994 to 46 per cent in 2019. Much of the increase in women’s labour-​force participation reflects changes in African women’s labour supply (Mosomi 2019a). For instance, while African women’s labour-​force participation rates rose from 36 to 53 per cent between 1994 and 2019, the increase was much smaller (although from a higher base) for white women, from 55 to 59 per cent.3 The ‘feminization’ of the labour force in post-​apartheid South Africa derives partly from the removal of apartheid laws that inhibited the labour supply of African women (and men) and their ability to migrate to places of employment (Posel 2014). It also likely reflects (and reinforces) several other societal shifts (Casale and Posel 2002; Ntuli and Wittenberg 2013; Mosomi 2019a). Social norms about women’s work have been changing, bolstered by the raft of new labour legislation introduced in the post-​apartheid years to encourage the hiring of women. Women’s levels of education have also been rising (and now surpass those of men on average) (Spaull and Makaluza 2019), while

2 

The PALMS dataset is created and updated by a team at DataFirst at the University of Cape Town, who collate the labour market data from all of the above-​mentioned surveys conducted by StatsSA and package them in a comparable and reliable form for public use (Kerr et al. 2019). At the time of writing, the data were available up until Quarter 2 of 2019. 3  The key labour market statistics shown in Table 34.1 are disaggregated further by race in Supplementary Table S.1 and the occupational distributions by race are shown in Supplementary Figure S.1, available in the longer working paper version of this chapter, which can be accessed at DOI:10.13140/​ RG.2.2.32400.84483.

738    Daniela Casale, Dorrit Posel, and Jacqueline Mosomi

Labour-force participation rate (%)

70

60

50

40

30 1995

2000

2005

2010

2015

2020

Year Financial crisis

Female

Male

Figure 34.1  Labour-force participation rates by gender, 1994–​2019 Source: Own calculations from PALMS V3.3. Notes: Labour-​force participation is defined as the percentage of the working-​age population (aged 15–​65) that is employed or looking for work (reflecting the official definition of unemployment). The vertical line at 2000 marks the changeover from the OHS to the LFS, and the vertical line at 2007 is the changeover from the LFS to the QLFS.

fertility rates have fallen (Moultrie and Timaeus 2003). Collectively these factors would have raised the opportunity cost of remaining in the home rather than joining the paid workforce. But many women have likely also been ‘pushed’ into the labour market out of need (Casale and Posel 2002). Women have been less able to share the costs of household maintenance and childcare obligations with a partner, as marriage rates have declined substantially over the period (particularly among Africans), and the percentage of households that rely on income earned only by women has risen apace (Posel 2014; Posel and Hall, ­Chapter 37 in this volume). Research taking stock of these changes in labour-​market participation during the first decade of democracy concluded that they were not all positive, and women’s increased participation in the paid economy had not ‘bought’ women very much in terms of job security and earnings (Casale 2004; Casale and Posel 2005; Posel 2014). As evidenced in countries throughout the world, much of the increase in women’s share of employment during the first post-​apartheid decade derived from women’s work in low-​skilled jobs and self-​employment in the informal sector, typically poorly paid and precarious work.4 4  A portion of the increase in informal activities between 1994 and 2001 would very likely have been a result of changes in survey design and the improved collection of data on more marginal employment. However, given the low absorptive capacity in the formal labour market, some of this increase no doubt reflected real changes in employment (Casale and Posel 2002).

Table 34.1: Key labour market statistics by gender, selected years from 1994–​2019 All

1994

2001

2007

2011

2017

2019

Female Male

Female Male

Female Male

Female Male

Female Male

Female Male

Working-​age population (000s)

12,586 11,450 14,979 13,498 16,535 15,316 17,775 16,676 19,264 18,320 19,710 18,846

Strict labour force (000s)

5,054

6,843

7,476

8,538

8,245

8,686

10,426 10,376 12,299 10,575 12,631

Broad labour force (000s)

6,405

7,696

9,502

9,662

10,484 11,135 9,959

11,423 11,763 13,366 12,121 13,846

Strict LFP rate (%)

40.2

59.8

49.9

63.3

49.9

64.4

48.9

62.5

53.9

67.1

53.7

67.0

Broad LFP rate (%)

50.9

67.2

63.4

71.6

63.4

72.7

56.0

68.5

61.1

73.0

61.5

73.5

Strict unemployed (000s)

1,271

1,164

2,437

2,214

2,125

1,965

2,383

2,383

3,100

3,195

3,330

3,385

Broad unemployed (000s)

2,621

2,017

4,463

3,337

4,363

3,242

3,656

3,380

4,487

4,262

4,876

4,600

Strict unemployment rate (%)

25.1

17.0

32.6

25.9

25.8

19.9

27.4

22.9

29.9

26.0

31.5

26.8

Broad unemployment rate (%)

40.9

26.2

47.0

34.5

41.6

29.1

36.7

29.6

38.1

31.9

40.2

33.2

Employment rate (%)

30.1

49.6

33.6

46.9

37.0

51.5

35.5

48.2

37.8

49.7

36.8

49.1

Employed (000s)

3,784

5,679

5,039

6,324

6,120

7,893

6,303

8,043

7,276

9,105

7,245

9,246

—​Paid employed/​employees

3,451

5,209

4,115

5,349

5,131

6,694

5,444

6,750

6,405

7,756

6,273

7,563

—Self-​employed, registered

34

154

111

280

158

375

120

373

112

377

123

390

—​Self-​employed, unregistered

286

285

725

640

782

780

671

867

708

940

776

1,260

Mean monthly earnings (rands)

4,372

7,103

5,237

8,036

6,589

9,620

7,475

10,268 7,088

9,757

–​

–​

Median monthly earnings (rands)

3,073

4,429

2,492

4,486

3,261

4,707

3,481

4 826

4,371

–​

–​

9,858

3,026

Source: Own calculations from PALMS V3.3. Notes: The working-​age population consists of individuals aged 15–​65. The official/​strict unemployment rate includes only the searching unemployed. The broad or expanded unemployment rate also includes the non-​searching unemployed. Correspondingly, the strict labour force includes the employed and the searching unemployed, while the broad labour force includes the employed, the searching unemployed, and the non-​searching unemployed. Self-​employed registered refers to individuals operating businesses registered for VAT. Earnings data in PALMS are only available up until 2017 and shown here in 2017 prices.

740    Daniela Casale, Dorrit Posel, and Jacqueline Mosomi Furthermore, in South Africa, a large part of women’s rising labour supply translated into unemployment, with particularly high rates among African women. Using the longer series of data covering the full post-​apartheid period, we find similar trends with respect to unemployment and women’s growing representation in low-​skilled work, but we also highlight some of the gains made over the period. Figure 34.2 (left frame) shows that the employment rate for women (the percentage of working-​age women with employment) rose from 30 to 37 per cent between 1994 and 2019, while men’s employment rate remained largely unchanged at 50 per cent (see also Table 34.1). Women’s share of employment therefore increased over the period, from 40 to 44 per cent. However, as Figure 34.2 (right frame) clearly illustrates, many female (and male) labour-​force participants did not find work and joined the growing ranks of the searching unemployed. While there was some decline in unemployment rates in the economic growth years of the 2000s (pre-​financial crisis), the overall trend has been upward: from 1994 to 2019, the percentage of female labour-​force participants who were unemployed and searching for work rose from 24 to 31 per cent, while the male unemployment rate increased from 17 to 27 per cent (Table 34.1). These unemployment rates are considerably higher than in most other countries for which there are data. Yet, unemployment rates in South Africa are even higher still, and the gender gap in unemployment larger, if we use the broad or expanded definition of unemployment, which includes the non-​searching unemployed. By 2019, the broad

Employment rate

60

Unemployment rate

35

Unemployment rate (%)

Employment rate (%)

30 50

40

25

20

15 30 10 1995

2000

2005 2010 Year

2015

2020

Financial crisis

1995

Female

2000

2005 2010 Year

2015

2020

Male

Figure 34.2  Employment and unemployment rates by gender, 1994–​2019 Source: Own calculations from PALMS V3.3. Notes: The employment rate is the percentage of the working-​age population (aged 15–​65) that is employed. The unemployment rate only includes the unemployed who searched for work.

Gender and Work in South Africa    741 unemployment rate for women was 40 per cent compared to 33 per cent for men (Table 34.1). Childcare constraints, living further away from centres of employment, and fewer resources to fund costly job search, likely contribute to why relatively fewer unemployed women than men report actively searching for work (Posel et al. 2014), and they also help to explain why racial differences among women remain stark. In 2019, the broad unemployment rate among African women was 45 per cent, compared to only 11 per cent among white women. In absolute terms, approximately 5.7 million additional women entered the labour force, broadly defined (i.e. the searching and non-​searching unemployed) from 1994 to 2019, and of these, roughly 2.3 million joined the unemployed, while the other 3.4 million found work. The large majority of the new jobs recorded for women (2.8 million) were in wage employment, with only 600,000 in self-​employment, mostly (500,000) in the informal sector (or in unregistered businesses). Approximately 6.1 million more men joined the labour force over the period, with 3.5 million men finding employment. Much of this also reflected wage employment, but men were more likely than women to enter self-​employment, which increased by 1.2 million (although also mostly in the informal sector). Ideally, we would want to track which of the jobs in wage employment were formal versus informal. Unfortunately, there is no easy way to classify employees consistently over the entire period. One option is to analyse who has a written contract with their employer, information consistently available since 2001. The data suggest that for both women and men with wage employment, the percentage with a written contract has increased since 2001, possibly due to the strengthening of labour legislation over the period and particularly with respect to domestic workers. By 2019, similar percentages of men and women in wage employment—​about 80 per cent—​reported having a written contract with their employer (own calculations based on QLFS data).5 Gender differences in self-​employment likely reflect different constraints that women and men face when starting their own businesses, whether in the formal or informal sector. Consistent with the labour force statistics, Global Entrepreneurship Monitor’s latest report on South Africa found women less likely than men to be involved in early-​ stage entrepreneurial activity, with only seven women entrepreneurs for every ten male entrepreneurs. They also found women more likely than men to open a business out of necessity rather than choice or opportunity (34 per cent of female compared to 18 per cent of male entrepreneurs) (Global Entrepreneurship Monitor 2018). In addition, South Africa performs quite poorly on global rankings when measured both in terms of female business ownership (ranking 39th out of fifty-​eight countries, with only 22 per cent of all business owners being women) and on a composite score measuring whether country conditions are conducive to narrowing the gender gap in entrepreneurship (35th out of fifty-​eight countries, down from earlier years) (Mastercard 2019).

5 

Further analysis of trends in the informal sector and in informal employment in South Africa is available in Rogan and Skinner, C ­ hapter 35 in this volume.

742    Daniela Casale, Dorrit Posel, and Jacqueline Mosomi Barriers identified as affecting women entrepreneurs specifically include limited access to formal credit; difficulties accessing markets for their goods; less developed networks compared to men; fewer role models; lower self-​confidence and a greater fear of failure; and difficulties juggling work and childcare responsibilities (Global Entrepreneurship Monitor 2018; Mastercard 2019; Rogan and Alfers 2019).6

34.2.2 Occupational Segregation Further insight into women’s changing participation in the labour market is provided by tracking occupation of employment. Job segregation along gender and racial lines has been a long-​standing feature of the South African labour market (Maconachie 1989; Casale 2004; Gradin 2018; Espi et al. 2019). Table 34.2 shows that in 1994, women were far more likely than men to be employed in jobs classified as ‘low-​skilled’ according to the International Standard Classification of Occupations (ISCO)—​38 per cent of employed women compared to 29 per cent of employed men. One in every five employed women was a domestic worker (a category on its own given the size and distinct features of this occupation in South Africa), and 18 per cent worked in other elementary occupations; while only 4 per cent of employed men were domestic workers and a quarter worked in elementary occupations. By 2019, the percentage of both employed women and men who were working in low-​skilled occupations fell to 35 per cent and 23 per cent respectively, consistent with a general trend towards more skilled employment in South Africa. However, the decline among women was driven particularly by changes in paid domestic work. For women, the number employed in domestic work increased in absolute terms (207,000) but by relatively less than the overall increase in female employment, producing a considerable fall in the percentage of employed women in domestic work, to 13 per cent. Nonetheless, domestic work became even more feminized over the period because, by 2019, there were also almost no men employed in this category.

6  A topic which we have not covered in this chapter is the scourge of gender-​based violence in South Africa. A KPMG study in 2014 estimated that the direct and indirect effects of gender-​based violence cost South Africa between Rand 28.4 billion and Rand 42.4 billion per year, or between 0.9 per cent and 1.3 per cent of GDP. To our knowledge, however, there has been no published research in the economics literature in South Africa which tries to measure the relationship between gender-​based violence (or its threat) and women’s participation in the labour market. Undoubtedly, the threat of gender-​based violence will influence women’s ability to move freely and to work unusual hours or in less frequented spaces (see Rogan and Alfers (2019) for some anecdotal evidence on self-​employed women in the informal sector), but there will also be more insidious effects which will constrain women’s ability to participate in the labour market, their productivity when at work, and the conditions under which they work. It is not surprising that in a 2016 Gallup World Poll of 142 countries, the two most commonly cited constraints to labour-​force participation were the ‘struggle to balance work and family’ and ‘abuse, harassment or discrimination’ (ILO 2017). This very important topic warrants much more attention in the South African economics literature than it has received to date.

Table 34.2: Occupational distributions by gender, 1994 and 2019 Occupation

Change 1994–​2019 Female (000s)

Change 1994 distribution 1994–​2019 Male (000s)

Legislators, senior officials, and managers

377

Professionals

150

Technical and associate professionals

374

Clerks

545

86

Service workers and shop and market workers

855

866

–​6

Craft and related-​trade workers Plant and machine operators and assemblers Elementary occupations Domestic workers Total

% of employed males

3

7

113

8

294

11

% of employed females

% of employed males

1994 Share of occupation female (%)

2019 Share of occupation female (%)

7

12

22

31

5

6

4

50

52

7

11

8

51

53

19

7

17

5

64

72

12

10

18

15

45

48

–​43

0

1

0

0

15

18

55

792

4

17

3

19

15

11

–​15

264

5

17

2

13

16

12

921

699

18

25

22

23

32

43

207

–​165

20

4

13

0

79

96

3,463

3 565

100

100

100

100

40

44

Skilled agricultural and fishery workers

659

% of employed females

2019 distribution

Source: Own calculations from PALMS V3.3. Note: Sample contains all employed individuals aged 15–​65.

744    Daniela Casale, Dorrit Posel, and Jacqueline Mosomi However, from 1994 to 2019, women’s employment in other elementary occupations rose in both absolute and relative terms (by 920,000 jobs or four percentage points). For men, the increase was smaller (700,000 jobs), and the percentage of employed men in elementary occupations fell to 23 per cent. As a result, women’s share in elementary occupations grew from 32 to 43 per cent over the period (final two columns of Table 34.2). At the upper end of the occupational distribution, both women and men became more likely to find employment in ‘high-​skilled’ work (classified as the top two occupational categories in Table 34.2), with the percentage of the employed working in these top occupations increasing from 11 to 13 per cent for women, and from 12 to 16 per cent for men. The percentage of women and men working as professionals declined over the period, so changes in the percentages of high-​skilled workers were driven by increased employment in the category ‘legislators, senior officials, and managers’ specifically. Because the relative growth here was greater for women (off a smaller base) than for men, women’s share of this top category rose from 22 to 31 per cent. Nonetheless, in 2019 women remained heavily under-​represented in this work relative to their share of total employment (44 per cent). These broad occupational patterns and trends are further segregated by race. The most extreme instances are in the low-​skilled categories. Less than 1 per cent of employed white women were in domestic work and less than 2 per cent were in other elementary occupations in 2019, compared to 16 and 26 per cent respectively of employed African women. In contrast, 17 per cent of employed white women were in the professional occupations and 23 per cent in the top managerial category in 2019, whereas the corresponding figures for employed African women were just 5 and 4 per cent respectively. Therefore, while women have made progress in accessing higher-​skill jobs since 1994, for African women specifically, this progress has been slow.

34.2.3 Earnings Differences in the types of employment women access affect their earnings potential. Figure 34.3 shows that women’s average monthly earnings remained well below those of men’s over the entire period. However, there was a marked decline in the gender gap at the mean over time. The ratio of female to male mean earnings grew from 0.616 to 0.726 between 1994 and 2017 (the latest year for which earnings data are available in PALMS—​ see Table 34.1). However, if we consider median wages or wages at the 50th percentile (also shown in Figure 34.3), the gender gap in earnings has hardly changed: the ratio of female to male median earnings was 0.694 in 1994 and 0.692 in 2017 (Table 34.1).7 This finding 7  Also evident from Figure 34.3 is how mean earnings diverge from median earnings over the period for both men and women, indicative of rising wage inequality in post-​apartheid South Africa (see Bhorat et al., C ­ hapter 31 in this volume for more discussion on wage polarization in the post-​apartheid period).

Gender and Work in South Africa    745

Monthly earnings (2015 Rands)

15000

10000

5000

0 1994

1998

2002

Financial crisis

2006 Year

2010

Mean female

Median female

2014

2018

Mean male

Median male

Figure 34.3  Real monthly earnings by gender for the employed (mean and median), 1994–​2017 Source: Own calculations from PALMS V3.3. Notes: Sample contains all employed (self-​and wage-​employed) aged 15–​65. Earnings reported in brackets weighted by bracket weights. The vertical line at 2000 marks the changeover from the OHS to the LFS, and the vertical line at 2007 is the changeover from the LFS to the QLFS.

is consistent with other research examining the gender wage gap in South Africa in the post-​apartheid period. Mosomi (2019b), in a detailed analysis of the gap across the earnings distribution, shows how the fall in the gender gap at the mean was driven by a decline at the bottom (10th percentile), and to a lesser degree, at the top (90th percentile) of the earnings distribution, while the gap at the median was largely stagnant. The falling gender gap at the lower end of the earnings distribution likely reflects the introduction of minimum wage legislation which would have benefited low-​skilled African women especially. At the upper end of the distribution, employment equity and equal pay legislation (coupled with a rising demand for tertiary-​educated workers) may have benefited higher-​skilled women relatively more (Mosomi 2019b). Nonetheless, among the employed, a substantial gender gap in earnings remains. A consistent finding from the South African literature that explores the reasons for the gender wage gap is that a large part remains unexplained by observable characteristics (see, among others, Grün 2004; Casale and Posel 2011; Bhorat and Goga 2013; Mosomi 2018, 2019b). In other words, differences in the individual, household, and job characteristics of women and men, as captured in household and labour-​force surveys, cannot fully explain the gender gap in earnings. Some of the unexplained gap may reflect

746    Daniela Casale, Dorrit Posel, and Jacqueline Mosomi pay discrimination, where women and men receive different returns for similar types of work. But information captured in large surveys is also often not fine-​grained enough to capture some of the nuanced differences in attributes and workplace experiences. For example, surveys used to analyse wage gaps show that among the employed, women on average are more educated than men (especially among the younger cohorts). They are more likely to complete their matric and graduate with a bachelor’s degree, and they tend to outperform (or at least do not do worse than) men in terms of achievement scores (van Broekhuizen and Spaull 2017; Spaull and Makaluza 2019). However, what is not included in many surveys (and therefore econometric analyses of wage gaps) is detail on subject choice or field of study. Education research shows that women remain under-​represented in the more lucrative STEM (science, technology, engineering, and mathematics) fields, and are instead more likely to enrol in the ‘care economy’ subjects such as education, psychology, and the health professions (van Broekhuizen and Spaull 2017; Spaull and Makaluza 2019; Statistics South Africa 2019a). A similar point can be made with respect to the occupational categories described earlier. Wage gap studies using national micro-​data highlight that women’s uneven distribution across the broad occupational categories is an important driver of the wage gap, but what they cannot account for is that even within these broad categories, women do different kinds of jobs from men which are rewarded differently (Casale and Posel 2011). Consider, for example, the professional occupations (a category in which women are over-​represented relative to their share of total employment):  women are much more likely to be teachers and nurses, while men are more likely to be headmasters and doctors. Budlender (2019: 65) argues that even though South Africa has tried to legislate equal pay for equal work in the Employment Equity Act, [i]t is not that women are paid less for the same job, but rather that women are paid less because the jobs that they do are seen as women’s (care) work, and women’s (care) work is under-​valued.’ In addition, even within similar job categories, women might be employed at lower ‘grades’ due to gender bias in hiring and promotion, or due to family commitments and gender norms that hold women back from applying for or accepting higher-​paying, higher-​responsibility jobs. For example, our earlier analysis showed that women are increasingly entering the managerial occupations (despite remaining under-​represented in this category). However, more detailed data from the Johannesburg Stock Exchange (JSE) show that very few women are appointed at the highest echelons of business and the corporate sector. In companies listed on the JSE as at April 2019, only 8.1 per cent of executive directors, 12.8 per cent of chief financial officers, and 3.3 per cent of chief executive officers were women, with female executive directors earning on average 74.5 per cent of what their male counterparts earned (PWC 2019: 32). A further, and enduring, constraint affecting women’s access to the labour market is the additional responsibility they face in the home. As we show in more detail in the next section, women continue to perform the bulk of unpaid childcare and housework in South Africa. While this unpaid labour is valuable in and of itself, it is also vital to sustaining the paid economy (Budlender and Brathaug 2002; Oosthuizen 2018). However, it reduces the time and energy that women can devote to their careers.

Gender and Work in South Africa    747 For a start, it affects women’s ability to pursue higher education. Although women are more likely to graduate with a matric (grade 12) or a bachelor’s degree in South Africa, they remain under-​represented in postgraduate studies (Spaull and Makaluza 2019). This is an area which warrants more research, but it may be related to women nearing peak child-​bearing age. Among the youth aged 18 to 24 years who were not in an educational institution in 2017, a lower percentage of women than men (17.7 versus 21.6 per cent) reported being satisfied with their completed level of study, and a much higher percentage of women than men (20.7 versus 1.3 per cent) reported not studying further due to family commitments (Statistics South Africa 2019a, based on the General Household Survey 2017). Further, unpaid care work affects women’s likelihood of participating in the labour market, their tenure in employment,8 and the returns to their employment, often referred to as the ‘motherhood penalty’ (Magadla et al. 2019).9 The double burden of paid and unpaid labour is perhaps particularly acute in the South African context, as many women looking after children are not married or living with the father of their children and typically cannot rely on men’s (income) support (Posel et al. 2016; see also Posel and Hall, ­Chapter 37 in this volume).

34.3  Gender, Household Labour, and Unpaid Care Work The preceding section described how women’s participation in paid work outside the home has increased considerably in South Africa in recent decades. Nonetheless, women remain primarily responsible for housework and the provision of caring labour in the home. Quantitative evidence for South Africa is captured in a range of national household surveys, and two dedicated Time Use Surveys (TUS) from 2000 and 2010. Consistent with the gender gap in labour-​force participation, women are far more likely than men to specialize in household production as full-​time ‘homemakers’. An analysis of data from the National Income Dynamics Study (NIDS), for example, shows that 98 per cent of all African adults who reported being a full-​time housewife

8 

In addition to women often exiting the labour market during child-​bearing episodes, women are more likely to engage in part-​time work which allows them to juggle employment and childcare responsibilities (Posel and Muller 2008). In 2019, 20 per cent of women compared to 11 per cent of men were in part-​time work (measured as less than 35 hours a week) and 57 per cent of part-​time employment was accounted for by women (authors’ own calculations based on the QLFS 2019). 9  While the causal relationship between fertility and labour market outcomes is very hard to estimate, detailed and careful work by Branson and Byker (2018) found that teen childbearing was related to both lower schooling attainment and lower earnings in young adulthood.

748    Daniela Casale, Dorrit Posel, and Jacqueline Mosomi or homemaker from 2008 to 2017 are female (Posel and Bruce-​Brand 2021). Most of these women are mothers10 of co-​resident children, and they are considerably more likely to live in rural households and households with lower levels of assets, where the demands of domestic work are more labour intensive (see e.g. Rubiano-​Matulevich and Viollaz 2019). As women’s labour-​force participation has increased, so the share of working-​age women (15–​65 years) who report not working because they are full-​time homemakers has declined, from an estimated 16 per cent in 1995 to less than 7 per cent in 2018 (own calculations from the OHS 1995 and NIDS 2018). However, even where women are not housewives or ‘stay-​at-​home mothers’, they spend far more time on household and caring labour than men. This has been clearly illustrated with quantitative data on how people spend their time (Budlender et al. 2001; Charmes 2006; Rubiano-​Matulevich and Viollaz 2019). South Africa is among only a handful of countries in Africa that have conducted time-​use surveys. These surveys collect detailed information on how people (aged 10 years and older) allocate their time both inside and outside the home, over a twenty-​ four-​hour cycle. Data from the most recent survey (the TUS 2010), reported in Table 34.3, show that the modal activity for both working-​age women and men is time spent on personal care (which includes sleeping, eating, and personal hygiene). But the next most common activity among women is household maintenance (which includes housework and household shopping), an activity in which nine out of ten women engaged on a daily basis (compared to seven out of ten men), and on which women spent almost four hours a day on average (compared to less than 100 minutes by men) (Table 34.3, upper frame). The TUS data record considerably less time spent on caring labour compared to household work. An important reason for this finding, which is documented also in other developing countries (Budlender and Lund 2011), is that childcare is often undertaken alongside other activities (such as cooking and cleaning). As the TUS data reported here only capture the primary activity undertaken in each time slot, it is likely that the average time spent on childcare is considerably under ​estimated.11 Nonetheless, a sizeable gender gap is still evident in time allocations to caring labour. The gap arises partly because women in South Africa are more likely than men to live with children (Posel et al. 2016; see also Posel and Hall, ­Chapter 37 in this volume). But even among only those adults who co-​reside with at least one young child, Table 34.3 (middle frame) shows that women are far more likely than men to provide any caring labour, and they spend significantly more time on this activity.

10 

Full-​time homemaking is distinctive in South Africa in that almost half of these women were not married or in a cohabiting union. 11  Average time spent on childcare is considerably larger when this time is calculated only for those who undertake any caring labour (i.e. the conditional mean), but it remains substantially lower than average time spent on housework.

Gender and Work in South Africa    749 Table 34.3: Mean total time and activity participation rates by gender, 2010 Mean total time Women

Activity participation rate Men

Women

Men

All working-​age adults (aged 15–​65) Household maintenance Care of persons

224.2**

97.9

91.6**

69.8

31.7**

4.4

28.7**

5.3

Market work

155.8**

255.5

42.0**

55.0

Personal care

744.2

739.3

Leisure

236.8**

289.0

Other

47.3**

53.9

100.0

100.0

90.6**

92.4

17.7

18.9

Working-​age adults living with at least one child (