A Multidimensional Economic Assessment of Africa [1st ed.] 9789811545092, 9789811545108

Citation preview

Frontiers in African Business Research

Jacob Wood Olivier Habimana   Editors

A Multidimensional Economic Assessment of Africa

Frontiers in African Business Research Series Editor Almas Heshmati, Jönköping International Business School, Jönköping, Sweden

This book series publishes monographs and edited volumes devoted to studies on entrepreneurship, innovation, as well as business development and managementrelated issues in Africa. Volumes cover in-depth analyses of individual countries, regions, cases, and comparative studies. They include both a specific and a general focus on the latest advances of the various aspects of entrepreneurship, innovation, business development, management and the policies that set the business environment. It provides a platform for researchers globally to carry out rigorous analyses, to promote, share, and discuss issues, findings and perspectives in various areas of business development, management, finance, human resources, technology, and the implementation of policies and strategies of the African continent. Frontiers in African Business Research allows for a deeper appreciation of the various issues around African business development with high quality and peer reviewed contributions. Volumes published in the series are important reading for academicians, consultants, business professionals, entrepreneurs, managers, as well as policy makers, interested in the private sector development of the African continent.

More information about this series at http://www.springer.com/series/13889

Jacob Wood Olivier Habimana •

Editors

A Multidimensional Economic Assessment of Africa

123

Editors Jacob Wood JCU Singapore Business School, James Cook University Singapore, Singapore

Olivier Habimana College of Business and Economics University of Rwanda Kigali, Rwanda

ISSN 2367-1033 ISSN 2367-1041 (electronic) Frontiers in African Business Research ISBN 978-981-15-4509-2 ISBN 978-981-15-4510-8 (eBook) https://doi.org/10.1007/978-981-15-4510-8 © Springer Nature Singapore Pte Ltd. 2020 This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar methodology now known or hereafter developed. The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication does not imply, even in the absence of a specific statement, that such names are exempt from the relevant protective laws and regulations and therefore free for general use. The publisher, the authors and the editors are safe to assume that the advice and information in this book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or the editors give a warranty, expressed or implied, with respect to the material contained herein or for any errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional claims in published maps and institutional affiliations. This Springer imprint is published by the registered company Springer Nature Singapore Pte Ltd. The registered company address is: 152 Beach Road, #21-01/04 Gateway East, Singapore 189721, Singapore

Acknowledgements

The chapters in this edited volume were presented at the 4th Eastern Africa Business and Economic Watch that took place from 12 to 14 June 2019 in Kigali, Rwanda. The conference was organized jointly by University of Rwanda, College of Business and Economics and Jönköping University, International Business School. Jacob Wood Olivier Habimana

v

Contents

1

An Overview: A Multidimensional Economic Assessment of Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . Jacob Wood and Olivier Habimana

Part I 2

Agriculture and Livestock

Determinants of Smallholders’ Export-Oriented Cash Crop Production Decisions in Ethiopia: A Case of the Sesame Sector . . . Gemechis Mersha Debela, Engdawork Assefa, Dawit Diriba Guta, and Sosina Bezu

3

Spatial Integration of Livestock Markets in Ethiopia . . . . . . . . . . . Gutu Gutema

4

New Urban Consumption Patterns and Local Agriculture: Application to the Bukavu HORECA Sector (DRC) . . . . . . . . . . . . Angélique Ciza Neema and Lebailly Philippe

Part II

1

21

41

67

Consumption, Poverty and Inequality 87

5

Multidimensional Inequality in Ethiopia . . . . . . . . . . . . . . . . . . . . . Getu Tigre

6

Factors Influencing Consumers’ Preferences for Cement Products: Case of Cement Brands in Ethiopia . . . . . . . . . . . . . . . . 115 Yared Tadesse and Workneh Kassa Tessema

7

The Value of Roads in Rural Household Consumption . . . . . . . . . 137 Marshal Negussie Simie

vii

viii

Contents

Part III

Financial Services, Employment and Corporate Governance

8

Corporate Social Responsibility Practices and Motivations in a Least Developed Country . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 159 Yohannes Workeaferahu Elifneh

9

Psychological Determinants of Employees’ Intentions to Retire: A Case of Public Universities in Kenya . . . . . . . . . . . . . . . . . . . . . . 181 Lucy Jepchoge Rono and Ester Agasha

10 Implementation of Corporate Governance Strategy: An Overview of Africa . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 199 Ambrose Kipruto Chepkwei 11 Efficiency of Commercial Banks in Ghana . . . . . . . . . . . . . . . . . . . 217 Eric Baah Nyannor and Isaac Osei Mensah 12 Cooperative Societies as a Distribution Channel for Insurance Services in Kenya: A Situational Analysis . . . . . . . . . . . . . . . . . . . . 233 Robert Kuloba, Esther Gicheru, and Silas Maiyo Part IV

Economic Integration, International Trade and FDI

13 The FDI-Domestic Investment Nexus in SSA . . . . . . . . . . . . . . . . . 253 Yemane Michael 14 China’s Unprecedent Move and Its Repercussion on African Economies: Empirical Evidence from Ethiopia . . . . . . . . . . . . . . . . 283 Zerayehu Sime Eshete 15 Does Free Trade and Institutional Quality Affect the Economic Community of the West African Trading Bloc? . . . . . . . . . . . . . . . 305 Luqman Olanrewaju Afolabi

Editors and Contributors

About the Editors Jacob Wood is currently the Associate Dean of Research for the College of Business, Law and Governance in JCU Australia. In addition to this, he is also the Director of the Centre for International Trade and Business in Asia (CITBA) and an Associate Professor of International Business at JCU Singapore. Prior to joining JCU, he spent more than 10 years’ working in South Korea where he has held roles as an Assistant Professor at Korea University of Technology and Education and then more recently as an Assistant Professor of Asia Business at Chungnam National University. He specializes in the fields of international trade negotiation and the effect of non-tariff barriers on international trade flows. In addition to this, he published a series of international business studies on social business adoption, employee engagement and the important role it plays within a HRM context. His research has been published in a range of prominent international journal publications including the Journal of Cleaner Production, Risk Management, Singapore Economic Review, Sustainability, Scientometrics, Technology Analysis and Strategic Management, Global Business Review, Business Information Review, Social Sciences, and the Journal of Asian-Pacific Economic Literature, among others. Olivier Habimana is currently a Lecturer of Economics and Econometrics at the University of Rwanda, College of Business and Economics. He has a Ph.D. in Economics from Jönköping University (2018). His main research interests lie in international macroeconomics and applied time series econometrics. His research has been published in Computational Economics, Defence and Peace Economics, Journal of Economic Asymmetries, Journal of Economic Integration, and the International Journal of Finance & Economics, among others. He is a member of the Rwanda Academy of Sciences.

ix

x

Editors and Contributors

Contributors Luqman Olanrewaju Afolabi Faculty of Economics and management Science, Department of Economics and Statistics, Kabale University Uganda, Kabale Municipality, Uganda Ester Agasha Department of Finance, Makerere University Business School, Kampala, Uganda Engdawork Assefa College of Development Studies, Center for Environment and Development Studies, Addis Ababa University, Addis Ababa, Ethiopia Sosina Bezu Chr. Michelsen Institute, Bergen, Norway Ambrose Kipruto Chepkwei Jomo Kenyatta University of Agriculture and Technology, Nairobi, Kenya Gemechis Mersha Debela College of Development Studies, Center for Environment and Development Studies, Addis Ababa University, Addis Ababa, Ethiopia Yohannes Workeaferahu Elifneh Department of Management, Addis Ababa University, Addis Ababa, Ethiopia Zerayehu Sime Eshete Addis Ababa University, Addis Ababa, Ethiopia Esther Gicheru Cooperative University of Kenya, Nairobi, Kenya Dawit Diriba Guta College of Development Studies, Center for Environment and Development Studies, Addis Ababa University, Addis Ababa, Ethiopia Gutu Gutema Department of Economics, Addis Ababa University, Addis Ababa, Ethiopia; Jonkoping International Business School, JIBS, Jönköping, Sweden Olivier Habimana College of Business and Economics, University of Rwanda, Kigali, Rwanda Robert Kuloba Insurance Regulatory Authority, Nairobi, Kenya Silas Maiyo Cooperative University of Kenya, Nairobi, Kenya Isaac Osei Mensah Bank of Ghana, Accra, Ghana Yemane Michael Department of Economics, College of Business and Economics, Mekelle University, Tigrai, Ethiopia Angélique Ciza Neema Faculté des Sciences Economiques et de Gestion, Université Evangélique en Afrique, Bukavu, Democratic Republic of Congo; Unité d’Economie et Développement Rural de Gembloux Agro-Bio Tech, Université de Liège, Gembloux, Belgium

Editors and Contributors

xi

Eric Baah Nyannor Zenith Bank (Ghana) Limited, Accra, Ghana Lebailly Philippe Unité d’Economie et Développement Rural de Gembloux Agro-Bio Tech, Université de Liège, Gembloux, Belgium Lucy Jepchoge Rono Department of Accounting and Finance, Moi University, Eldoret, Kenya Marshal Negussie Simie Department of Economics, Addis Ababa University, Addis Ababa, Ethiopia Yared Tadesse Express Management Consulting PLC, Addis Ababa, Ethiopia Workneh Kassa Tessema Department of Management, Addis Ababa University, Addis Ababa, Ethiopia Getu Tigre Department of Economics, Addis Ababa University, Addis Ababa, Ethiopia Jacob Wood Associate Professor, JCU Singapore Business School, James Cook University, Singapore, Singapore

Abbreviations

AE AR ATA BE BMI CC CCE CCEMG CCEP CCEPMG CCO CD CIC COP CR CSA CSA CSA CSR DCCE DCCEP DEA DFE D-H DHS DI DRC ECM FAO FDI

Adult Equivalence Autoregressive Agricultural Transformation Agency Between Effects Body Mass Index Contingency Coefficient Common Correlated Effects estimators Common Correlated Effects of the Mean Group estimator Common Correlated Effects Pooled estimator Common Correlated Effects Pooled Mean Group Congolese Control Office Cross-sectional Dependence Cooperative Insurance Company Certificate of Proficiency Corporate Responsibility Central Statistics Authority Central Statistical Authority Central Statistical Agency Corporate Social Responsibility Dynamic Common Correlated Effects estimators Dynamic Common Correlated Effects Pooled estimator Data Envelopment Analysis Dynamic Fixed Effect Double-Hurdle Demographic and Health Survey Domestic Investment Democratic Republic of Congo Error Correction Model Food and Agricultural Organization Foreign Direct Investment

xiii

xiv

FE FGDs FINSAP FINSSP GDP GDP GE GFCF GLS GMM HCES HDI HIES HORECA ICMF ICT IRA ISO JKUAT LDC LP-HC LSMII MDI MG MII ML MLE MMUST MNCs MoFED NGOs NHIF OLS PA PEL PLWD PMG POLS PP RE SACCO SASRA SDGs SFA SNNP

Abbreviations

Fixed Effect Focus Group Discussions Financial Sector Structural Adjustment Program Financial Sector Strategic Plan Gross Domestic Production Gross Domestic Product Generalized Entropy Gross Fixed Capital Formation Generalized Least Squares Generalized Method of Moments Household Consumption and Expenditure Survey Human Development Index Household Income and Expenditure Survey Hotel Restaurant Café International Cooperative and Mutual Finance Federation Information, Communication Technology Insurance Regulatory Authority International Organization for Standardization Jomo Kenyatta University of Science and Technology Least Developed Countries Low Priority—High Consequence Living Standard Multidimensional Inequality Index Multidimensional Inequality Mean Group Multidimensional Inequality Index Maximum Likelihood Maximum Likelihood Estimation Masinde Muliro University of Science and Technology Multinational Corporations Ministry of Finance and Economic Development Non-Governmental Organizations National Hospital Insurance Fund Ordinary Least Squares Population Averaged Patrice Emery Lumumba Persons Living With Disability Pooled Mean Group Pooled Ordinary Least Square Polypropylene Random Effect Savings Cooperative Sacco Society’s Regulatory Authority Sustainable Development Goals Stochastic Frontier Analysis South Nations and Nationalities People

Abbreviations

SPSS SSA TLUs TTB UN UN UNCTAD UNDP USA VIF VIF WIDE

xv

Statistical Package for Social Sciences Sub-Saharan Africa Tropical Livestock Units The Trust Bank United Nation United Nations United Nations Conference on Trade and Development United Nation Development Program United States of America Variance Inflating Factor Variance Inflation Factor World Inequality Database on Education

Chapter 1

An Overview: A Multidimensional Economic Assessment of Africa Jacob Wood and Olivier Habimana

Abstract In the face of weak growth globally, there is still an opportunity for Africa to further build upon the developmental successes it has enjoyed in recent decades. This introductory chapter provides an overview of the key areas of analysis located in the book, in particular an assessment of the African agricultural sector, the issues of poverty and inequality, banking and financial services, as well as opportunities for further international trade development in the region. The chapter highlights both areas of economic strength and vulnerability as well also providing an in depth assessment of various policy platforms that have been implemented in the region. Keyword Economic development · International trade · Inequality · Agriculture · Financial services

1.1 An Introduction Africa is at a critical juncture in its development trajectory. Transitioning the region to a level of greater economic independence and growth is very much a work in progress with the continent making steady progress in building the critical ingredients for sustainable and resilient societies. In recent times, economic growth in Africa has been stable moderating from 3.4% in 2017 to 3.2% in 2018, however, within the region, East Africa remains the fastest growing, at 6.1 to 6.2% over the same period (UNECA 2019a). These results reflect stronger domestic demand and increased trade between Africa and growing global markets, particularly in Asia. In addition to

J. Wood (B) Associate Professor, JCU Singapore Business School, James Cook University, Singapore, Singapore e-mail: [email protected] O. Habimana College of Business and Economics, University of Rwanda, Kigali, Rwanda e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 J. Wood and O. Habimana (eds.), A Multidimensional Economic Assessment of Africa, Frontiers in African Business Research, https://doi.org/10.1007/978-981-15-4510-8_1

1

2

J. Wood and O. Habimana

this, improving agricultural commodity prices have remained a key determinant of economic growth in Africa. Africa has also made significant progress in healthcare, education and social welfare support. Moreover, poverty levels, another important developmental metric, has also been in retreat, with poverty rates falling from 54.3% in 1990 to 36% in 2016 (UNECA 2019a). Nonetheless, the ability to grow in an inclusive manner has remained relatively elusive with income inequality remaining high at 0.44 on the Gini coefficient index (UNECA 2019b). A key mechanism to facilitate economic development in the future is a stable fiscal policy agenda. Stability in this regard would help Africa achieve its sustainable development goals (SDGs) that the region signed up to as part of global 2030 SDG. With the SDG deadline a decade away, the onus is on African governments to implement the right kind of reformative policy agenda that can help to increase government revenue and Gross Domestic Product (GDP) performance. A key developmental goal for the region is the African Union’s Agenda 2063: the Africa we want, which details its economic blueprint and master plan for transforming Africa into the global powerhouse in the future. The Agenda provides the continent’s strategic framework that aims to deliver on its goal for inclusive and sustainable development and is a concrete manifestation of the pan-African drive for unity, self-determination, freedom, progress and collective prosperity (African Union 2019). In the face of weak governance and growth globally, there is still a good opportunity for countries in Africa to build on not only their traditional industrial capabilities, but also path the way for positive developments in international trade and in the way governments tackle poverty and inequality. In the global context, Africa remains a vitally important continent with a rich abundance of natural resources and a rapidly growing population. This chapter provides an overview of the four sections examined in this book, these include: (1) agriculture and livestock; (2) consumption, poverty, and inequality; (3) financial services, employment and corporate governance; (4) economic integration, international trade and FDI. From an international trade perspective, free trade presents a myriad of opportunities for the region. In this introductory chapter, we outline the important role it can play. We also look at the role that sound trade policy can play as a mechanism for transformative and inclusive growth in Africa. In addition to this, we also look at the important role the agricultural sector plays in Africa. This chapter also examines the issue of inequality and poverty in Africa. Finally, an assessment of the financial services sector is also provided.

1.2 Agriculture in Africa African agricultural development has flourished over the last 30 years. Contrary to popular belief, agricultural production has grown significant with its total value increasing by 160%, to a level that is comparable to that of South America (NEPAD

1 An Overview: A Multidimensional Economic Assessment of Africa

3

% 60 50 40 30 20

Chad

Mali

Ethiopia

Malawi

Kenya

Tanzania

Sudan

Uganda

Mozambique

DRC

Ghana

Nigeria

Zambia

Angola

Zimbabwe

South Africa

India

Botswana

China

Brazil

Rest of the World

Argentina

Russia

Chile

Mexico

Australia

European Union

Canada

0

United States

10

Sub-Saharan Africa

Fig. 1.1 Agriculture as a share of total GDP in 2014. Source World Bank

2015). Despite these improvements, there has been little improvement in the production factors of labor and land with more production only being achieved by simply mobilizing more people to cultivate an increasing amount of land. A technique that does nothing to improve yields. In addition to this, the region has also experienced significant demographic change, with the African population doubling overall and tripling in urban areas over the last 30 years. However, despite this population growth cereal production has only increased by a factor of 1.8 (NEPAD 2015). As a consequence of these developments Africa has moved from being self-sufficient to becoming a net importer of cereals. Nonetheless, the Agricultural sector remains a major sector in most African economies (see Fig. 1.1). After decades of stagnation within the agricultural sector, it is now experiencing a significant surge in growth. A key driver in this regard has been the crop sector which accounts for almost 85% of total production value over the 1990–2013 period (see Fig. 1.2). This share varies across the SSA region from 53% in Southern Africa to more than 90% in Western Africa (OECD-FAO 2016). Across Africa, the five biggest crops make up almost half of the total production. More specifically, rice (Eastern and Western Africa), potatoes (Eastern and Central Africa) cassava (Western and Eastern Africa) and plantains (Eastern and Central Africa). While in Southern African, fruit and vegetables make up a large part (see Fig. 1.3). In terms of livestock, poultry contributes a large proportion of the total production value for livestock, with its level of importance ranging from 12% in Eastern Africa to 45% in Central Africa and Southern Africa Also of note are figures for Central Africa, in which the livestock production value is smaller than any of the other three regions, whereby game meat accounts for 35% of livestock value (Fig. 1.4) (OECD-FAO 2016). In addition to livestock, fisheries and aquaculture also make a multifaceted contribution to national economies in SSA. The region has vast fish resources, in marine and inland waters, and is characterized by diverse fishing

4

J. Wood and O. Habimana Agriculture growth rate

Crops index

Livestock index

Agriculture index

Index (2005=100)

%

140

12 10

120

8 100

6

80

4

60

2 0

40

-2 20 0

-4 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012 2013

-6

Fig. 1.2 Gross agricultural production value in Sub-Saharan Africa. Source World Bank

Cereals

Other Cash Crops

Oilseeds

Roots and tubers

Pulses

Fruit and vegetables

% 100 90 80 70 60 50 40 30 20 10 0

Eastern Africa

Central Africa

Southern Africa

Western Africa

Fig. 1.3 Crop mix across the Sub-Saharan Africa Region. Source World Bank

communities. The sector offers food security to many as well as an important source of employment and foreign revenue. However, it does face many challenges including mismanagement of fishing stocks, weak and uncoordinated institutions, ineffective legal frameworks, lack of adequate infrastructure and climate change (OECD-FAO 2016). So what is driving development and growth within the agricultural sector is SSA? In this regard, domestic and foreign research and development (R&D) investments will play a very important role in developing the capabilities of the agricultural industry in Africa. R&D Investments are normally seen as an important mechanism behind driving productivity growth through new knowledge and innovation (Arlene 2010; Rahman and Salim 2013). Such a belief is well-established within a sub-Saharan

1 An Overview: A Multidimensional Economic Assessment of Africa Other

Poultry

Pork

Sheep and Goat

5 Beef

Dairy

% 100 90 80 70 60 50 40 30 20 10 0

Eastern Africa

Central Africa

Southern Africa

Western Africa

Fig. 1.4 Livestock mix across the Sub-Saharan African region. Source World Bank

African (SSA) context (see Masters et al. 1998; Beintema and Stads 2011) where the agricultural industry is a key contributor to household income and GDP. Furthermore, there is also significant academic support for the notion that conventional inputs do not contribute greatly to agricultural productivity growth (Schultz 1956; Fan et al. 2004). There exists a plethora of studies that have examined productivity and performance levels within the African agricultural sector. In this instance, both country-level and cross-country analyses have been conducted in which a myriad of methodological approaches, variable choices and samples have been adopted. Initial efforts provide evidence of poor aggregate productivity performance across the 1960s and 70s (Nkamleu 2004), while other studies by Alene, (2010) and Fulginiti et al. (2004) highlight the positive productivity gains that have occurred since the 1980s. There is also sufficient empirical evidence to suggest that technological innovation has been a major contributing factor to these gains (Alene 2010). Overall, the literature in this area acknowledges the important role that R&D has played within agricultural productivity in Africa. Moreover, the returns of R&D investment in SSA has been quite high (Maredia et al. 2000). While Beintema and Stads (2011) showed that R&D investments in the agricultural sector also play an integral role in not only reducing poverty levels but also the region’s ability to adapt to climate change. More recently Adetutu and Ajayi (2019) showed that total agricultural productivity in SSA is strongly influenced by domestic and foreign R&D spending in the Agricultural sector. In terms of growth itself, the agricultural sector has been driven by expansion and intensification of cropping systems (Brink and Eva 2009; NEPAD 2015), which is in contrast to South America which was driven by significant improvements in labor productivity. Given that SSA is land abundant, continued expansion is no a cause for concern, however, rural SSA is highly heterogeneous with much of the

6

J. Wood and O. Habimana

land underutilized. Moreover, a significant proportion of its rural population reside in smallholder farming enterprises that are in fact densely populated (Jayne et al. 2014). These levels of population density can lead to other problems with many SSA studies highlighting soil degradation arising from unsustainable farming practices as a particular concern, particularly in places like Malawi and Kenya (Drechsel et al. 2001; Tittonell and Giller 2012). In these instances a lack of crop rotation leads to a depletion of organic carbon levels, which in turn makes the soil less responsive to fertilizer application. Despite these difficulties progress is being made through the establishment of a policy agenda in terms of short-term objectives that are more closely aligned to the regions’ long term goals (FAO 2015). Other policies have been implemented to address the lack of stability in the policy making sphere with many studies highlighting the impact that poor governance has on the agricultural sector in Africa (Willis 1990; Seini 2002; Kimenyi et al. 2017). In addition to this, fertilizer subsidy programs have been utilized in Zambia and Malawi, achieved some productivity gains, although the jury is still out as to whether or not progress can be maintained over the long-term (Jayne and Rashid 2013).

1.3 Poverty and Inequality in Africa SSA lags behind other developing nations in a number of key human development indicators (UNDP 2010). A key factor in this regard, is poverty with SSA facing the highest regional poverty rate with 422 million Africans living below the poverty line of $1.90 per day, representing 70% of the world’s poorest (Hamel et al. 2019). However, progress is being made, with more Africans now escaping poverty than are falling (or being born) below the poverty line (Hamel et al. 2019). Despite this progress several trends still persist. Poverty is more common in young families, for example, as asset ownership is lower and dependency ratios are generally higher (Handley et al. 2009). For other families in SSA, poverty is very much embedded in and is experienced for most if not all of one’s life and is often passed on from one generation to the next (CPRC 2004). With poverty very much entrenched in many parts of SSA there are a number of important socio-economic drivers and maintainers of poverty in the region. Key factors in this regard include, harvest failure, market failure and volatility, conflict and health shocks (Handley et al. 2009). Given SSA’s geography, agro-ecology, climate change vulnerabilities and inefficient agricultural technologies, harvest failure is a key risk for rural households (Sinha and Lipton 1999). Harvest failure did not only affect crop-dependent families but also wider sections of the rural community and the stability of the nation as a whole. The 2001–2003 food crisis was a good example of its potential impact, coupled with heavy rains that destroyed a large proportion of the regions maize crop (Wiggins 2005). While harvest failure was a key factor here, institutional weaknesses, political factors, donor policies and economic inequalities were also important (Booth et al. 2006). Another key driver of poverty has been market failure and market volatility. This is because the poorer

1 An Overview: A Multidimensional Economic Assessment of Africa

7

segments of African society do not possess the physical and human capital assets required to insulate themselves from market downturns. Moreover, inadequate institutional and infrastructural linkages mean that the markets are also poorly integrated. While market volatility is driven by international economic shifts or more localized market failures such as significant increases in the price of commodities like tea and sugar which hurt Ugandan consumers in the 1990s or the collapse in coffee prices which impacted producers in the region (CPRC 2004). Armed conflict is also a driver of poverty with over 40% of the world’s armed conflicts occurring in SSA in recent times (Ploughshares 2007). While destroying public infrastructure, private property, killing countless lives, and undermining law, these conflicts have also displaced millions of people which creates significant financial burdens for host countries (Goodhand 2001). Health shocks are also a key factor with the statistics for the region causing ongoing great concern. The infant mortality rate for children under the age of 5 remains high as does the number of woman that die from complications from pregnancy and childbirth (UNDPI 2007). In addition to this, is the growing HIV/AIDS pandemic which has significantly reduced life expectancy in SSA. Moreover, families affected by HIV/AIDS earn less and are more vulnerable to market downturns (Harvey 2004). Inequality is also another important issue, with Africa being the second most unequal continent next to Latin America (Ravallion and Chen 2012). High levels of inequality is by no means and new thing and has persisted in SSA for a long time (Bigsten 2014). For an assessment of the Gini coefficient for 29 SSA countries, Cornia and Martorano (2016) found that for 13 of these economies the level of inequality fell between 1991 and 2011. While in contrast, in seven countries the Gini Index rose during this period. Of the 29 countries, Ethiopia had the lowest Gini index at 33.6 (2011), while the Gini indices for Botswana (2009) and South Africa (2011) were the highest, at 68.6 and 65.0 respectively. The reasons behind the changes in economic inequality differ between countries. The overall fall in equality before 1980 reflects growth in the types of economic opportunities that are available as well as social mobility in post-colonial Africa, which was brought about by greater investment and broader participation in education (Adesina 2016). In the post-1980 period, most of the gains were made through deregulation and economic liberalization. In contrast the 2000s saw differing results with countries such as Botswana and Ghana experiencing increases in inequality levels. Other important considerations surrounding inequality in SSA include improvements in the education sector. Between 1999 and 2011, SSA showed improvements in several education indicators, including net enrolment ratio in primary school (58 to 77%) to gender parity (0.85 to 0.93), while the youth literacy rate also increased (Adesina 2016). Withincountry indicators also vary widely by gender, by location (urban, rural) and by income group. In order to overcome these differences it is important that governments appropriately invest in education and significant improvements have been made in this regard in recent times. In particular, school fees have been abolished and school feeding programs have been introduced. Despite this large differences in attendance and achievement exist within countries. For example, in 2013, a child

8

J. Wood and O. Habimana

from the poorest quintile in Nigeria was more than 20 times more likely to not be attending school than a child from the most affluent quintile. While in Niger only 15% of girls’ complete primary school compared to 30% of boys. This is in stark contrast to South Africa where 95% for girls and 91% for boys complete primary school education. Health is another area in which SSA has greatly improved. Like education, there are also significant differences at the national level in health (WHO 2014). For example, with life expectancy, there are large variations, ranging from 74 years in Mauritius to 46 years in Sierra Leone. According to the WHO (2014), the main factors influencing health outcomes are wealth, the rural/urban divide, and mothers’ education level. Immunization coverage is also a major concern with Nigeria in particular showing great disparities with the poorest quintile eight times less likely to be immunized than the wealthiest quintile. Health care expenditure show inequalities exist along wealth lines. In South Africa in 2015, total health-care expenditure made up 8.5% of GDP. Of this, less than half (4.1%) was public health-care spending, which covered 84% of the population. While the remaining portion of the GDP that was committed to health care (4.4%, or more than half) was used by the 16% of the population who have private health insurance (Adesina 2016). Finally, the issue of gender inequity is also an important factor when assessing inequality in SSA. The World Economic Forum Report (2014) analyzed the issue of gender in four areas: educational achievement, political empowerment, health and survival, and economic participation. The results showed that from an overall perspective, Burundi, Rwanda and South Africa are the three best-performing countries on the overall gender-gap index, while Chad, Côte d’Ivoire and Mali are the worst. In 2015, Rwanda was the highest ranked nation for political empowerment in SSA, which was due to the fact that it has the highest percentage of female legislators in the world with 63.8% of its members of parliament being women (International Parliamentary Union 2016).

1.4 Banking and Financial Services Sector The banking system dominates the financial landscape in SSA. The banking sector accounts for the biggest share of assets in most countries in the region, with the exceptions being middle income countries, such as Lesotho, Namibia, Swaziland, and South Africa in which nonbank assets account for more than 50% of financial sector assets (IMF 2016a). Foreign-owned subsidiaries make up the majority of banking assets across all country groups, particularly in some of the smaller less developed economies such as Guinea, Guinea-Bissau, Madagascar, São Tomé and Príncipe, while the contribution of foreign branches is minor (IMF 2016b). Moreover, state-owned banks’ assets are key players in Ethiopia, Rwanda, Seychelles and Sierra Leone. Within the nonbank financial sector, pension funds make up a significant proportion of the systems assets. While SSA stock exchanges are generally

1 An Overview: A Multidimensional Economic Assessment of Africa

9

underdeveloped and illiquid and are present in less than 60% of countries in the region (IMF 2016b). Homegrown SSA banks are playing an integral part in the development of the region. Their growth has also been quite profound with the scale of their operations now larger than that of other American and European Banks, which has retrenched their operations in the aftermath of the global financial crisis. The broad reach of the SSA banks means that out of the six Pan African Banks (PABs) operating in the region, all have a branch in at least 10 countries, while some are based in more than 30 countries. While their obvious strengths lie in traditional bank intermediation, they also have significant dealings in capital markets, insurance, pensions, micro-finance, and even non-financial transactions (IMF 2016a). The growth in these banking operations has seen increased competition for loan and deposits, which has in turn led to the creation of new service offerings such as mobile banking and web-based technologies. These developments have led to changes in how they capture funding. As Stijns (2015) notes, this has meant that many PABs have higher profitability and cost-income ratios. The growth in PABs also creates several challenges, such as those that relate to their cross-sectoral and cross-border exposures. This can then lead to increases in the risk of spillovers from other segments (i.e. non-bank activities) and exposure to exogenous financial shocks. There are also supervisory problems with reporting standards differing across the region. Moreover, the implementation of Basel accords has not been done in a uniformed manner, with higher standards only adopted in Kenya, Malawi. Mauritius, Mozambique and South Africa (IMF 2016b). Crisis management measures are also poorly implemented with some deposit insurance schemes remaining underfunded and able to cover 80% of retail deposits in systemic banks (IMF 2016b). Moreover, the state-owned SSA banks often have no clear goals or performance measures (Beck and Maimbo 2013). The diversity of stages in financial market development remain a key area of needed improvement for SSA, which significant differences existing across the region (see Table 1.1). The banking markets are also characterized by high lending rates and low deposit rates with interest rates across the region trending downwards. However, profitability is high because banks tend to invest in risk-free government bonds which tend to pay high yields (European Investment Bank 2018). The development of the banking sector in SSA has also coincided with the increasing potential of Islamic finance. Islamic banking practices have proven to be an effective way of overcoming institutional barriers and promoting further economic development in the region (IMF 2016a). In simplistic terms, Islamic finance adheres to a set of ethical principles and laws (Shari’ah). According to the IMF (2016a), Islamic finance products can be classified in three main categories: (1) debt-like financing through sales or deferred payments, (2) equity-like financing through profit-and-loss sharing, and (3) financial services. While only making up 15% of total banking assets, there is obvious potential for significant growth moving in SSA. In studies into its potential impact, Imam and Kpodar (2015) and Kammer et al. (2015) have found that Islamic Banking is conducive to economic growth and financial inclusion in lowand middle-income countries, including in sub-Saharan Africa.

10

J. Wood and O. Habimana

Table 1.1 Financial Soundness Indicators in selected SSA Banking Markets Country

Date

Capital to assets

NPLs to total gross loans

Return on assets

Return on equity

Angola

2015A1

8.42

10.61

2.11

21.6

Botswana

2017Q4

8.7

5.28

1.93

16.28

Burundi

2017Q3

12.21

13.61

2.9

20.51

Cameroon

2018M2

Central African Republic

2017M11

18.9

22.65

8.63

45.85

Chad

2107M11

11.99

25.1

1.87

17.22

Congo, Republic of

2017M11

17.79

13.4

8.33

52.52

Equatorial Guinea

2017M11

15.46

27.33

5.32

35.2

Gabon

2017M11

9.15

10.04

13.04

134.64

Gambia

2016Q2

15.04

7.6

3.46

20.62

Ghana

2017A1

13.11

21.59

4.29

18.58

Guinea

2017Q4

11.86

10.68

2.05

16.74

Kenya

2017Q4

15.28

10.08

3.24

21.38

Lesotho

2018Q1

12.12

4.16

3.92

31.06

Madagascar

2018Q1

10.72

7.9

4.5

43.16

Malawi

2017Q4

15.6

8.55

5.06

31.11

Mauritius

2017Q4

10.08

7.03

1.55

15.16

Nambia

2017A1

11.73

2.49

3.02

28.01

Nigeria

2017Q3

7.43

15.13

2.43

20.54

Rwanda

2017Q1

14.28

7.73

2.71

14.88

Seychelles

2017M1

12.11

7.03

3.37

27.85

South Africa

2018M2

8.77

3.10

1.70

19.78

Tanzania

2018Q1

12.72

11.13

1.71

11.37

Uganda

2017Q4

13.81

5.51

3.85

22.84

Zambia

2017Q4

11.23

11.98

2.64

22.68

8.028

12.14

Source European Investment Bank (2018)

The Islamic finance principles of risk-sharing and asset-based financing help to promote macroeconomic and financial stability through more effective risk management practices from both the banks themselves and their customers (IMF 2016a). In addition to this, Islamic finance principles also adhere to the financing needs of small and medium business, which helps to promote inclusive growth. Despite such benefits, Islamic finance poses a number of regulatory, supervision, and monetary policy challenges owing to the ways it carries out its transactions. For example,

1 An Overview: A Multidimensional Economic Assessment of Africa

11

such supervisory frameworks need to consider Islamic finance specificities such as profit-sharing investment accounts and Shari’ah governance (IMF 2016a).

1.5 On the Road to Free Trade in Africa On the global stage, Africa is a marginal player in the global trade of goods. When compared to the rest of the world, total trade from Africa to the rest of the world averaged $760.5 Billion USD in current prices over the 2015–2017 period. An amount that is not significantly larger than Oceania’s $481 billion, but very much smaller than the $4,109.1 billion from Europe, $5,139.6 billion from the United States (US) and $6,801.4 billion from Asia (UNCTAD 2019). At the heart of African exports in sub-Saharan Africa are petroleum, gold, coal, and natural gas. More generally, Table 1.2 documents the leading exports by sector as listed by the World Integrated Trade Solution database. In recent times, the African economy has had to grapple with the 2008 global recession and its aftermath as well as political instabilities in countries like Zimbabwe. Despite such setbacks, Africa has taken significant steps towards integrating its regional market of 55 (only 54 have become signatories to date, with only Eritrea still to join) African Countries through the establishment of the Africa Continental Free Trade Agreement (AfCFTA). The agreement would initially require member states to remove tariffs from some 90% of goods and in doing so allow free access to commodities, goods, and services across the continent. It would also allow for the free movement of business travelers and investments, and create a continental customs union that not only streamlines trade but also attracts long-term regional and foreign investment (Cloete 2019). The agreement would also provide significant job growth and help to build on intra-Africa trade flows, which has been historically low. In 2017, intra-Africa exports made up only 16.6% of total exports, significantly less than its European (68%) and Asia (59%) counterparts. The AfCFTA provides an opportunity to turn around these intra-Africa trade figures, which can help countries in the region grow. The African Union is currently implementing the AfCFTA with it being inforce across 27 countries, with the agreement slated to take effect in June 2020 (Manders 2019). The benefits of international trade to the African economy cannot be overemphasized. Trade is considered a key driver of economic growth and development as it helps to integrate economies, generate foreign exchange, increase technological transfer, generate higher efficiency levels among firms due to greater competition, increase employment opportunities, and alleviate poverty (Krueger 1980; Frankel and Romer 1999; Agrawal 2015; Wood 2017; Sakyi et al. 2018; Heshmati et al. 2019). The removal of trade barriers is also an integral part in helping to drive international trade flows (Wood et al. 2017a, b, 2019a, b) which is a key part of what AfCFTA is seeking to achieve. However, much of what AfCFTA is trying to achieve depends on how the trade itself is facilitated. In this instance, trade facilitation refers to the simplification and harmonization of international trade procedures in an effort

12

J. Wood and O. Habimana

Table 1.2 Sub-Saharan Africa Exports by Sector (2017) Product group

HS Code (2-digit)

Export (US$ Thousand)

Export product share (%)

All products

1–99

363,088,831.02

100

Mach and elec

84, 85

74,762,228.70

20.59

Transportation

86, 87, 88, 89

46,504,546.82

12.81

Fuels

27

41,610,497.04

11.46

Chemicals

28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38

37,477,264.71

10.32

Metals

72, 73, 74, 75, 76, 78, 79, 80, 81, 82, 83

23,483,290.21

6.47

Textiles and clothing

50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63

21,230,783.85

5.85

Vegetable

06, 07, 08, 09, 10, 11, 12, 13, 14, 15

20,685,565.96

5.7

Food products

16, 17, 18, 19, 20, 21, 22, 23, 24

20,588,295.39

5.67

Miscellaneous

90, 91, 92, 93, 94, 95, 96, 97, 98, 99

20,047,084.17

5.52

Plastic or rubber

39, 40

15,781,397.40

4.35

Stone and glass

68, 69, 70, 71

11,513,492.76

3.17

Animal

01, 02, 03, 04, 05

10,433,677.28

2.87

Wood

44, 45, 46, 47, 48, 49

8,492,671.03

2.34

Footwear

64, 65, 66, 67

5,108,560.20

1.41

Minerals

25, 26

3,953,637.52

1.09

Hides and skins

41, 42, 43

1,415,837.98

0.39

Source World Integrated Solution Database

to reduce transaction costs and other arrangements, such as the professionalism of customs authorities, synchronization of various standards and conformity to international or regional regulations (Buyonge and Kireeva 2008: Sakyi et al. 2018). The benefits that can be achieved through an alignment and harmonization of procedures in Africa has the potential to exceed the cost savings achieved from the elimination of tariffs (Portugal-Perez and Wilson 2009). From an economic development perspective, trade facilitation could lead to better social welfare outcomes in Africa (Narayanan et al. 2016) as it can enhance the competitiveness of firms operating in the market (Sakyi et al. 2018). The literature in this area shows that trade facilitation helps the low- and middle-income countries decrease poverty and inequality, and increases in per capita GDP (Viet 2015). While Seck (2016) reported a positive impact of trade facilitation on both African imports and exports. Sakyi et al. (2018) showed in their Africa centric study that effective trade facilitation reforms, which look to improve infrastructure, institutional bodies, and

1 An Overview: A Multidimensional Economic Assessment of Africa

13

market efficiency, are associated with social welfare improvements. Other economic development studies on Africa have also paid particular attention to the aid for trade issue and there has been much research conducted on the topic. At a basic level, research has sought to address one fundamental question; does aid for trade improve trade performance and export expansion? More specifically, Karingi and Leyaro (2009) examined the effects of aid for trade on export diversification and other growth performance measures. From their analysis they showed that aid for trade addressed capacity constraints, reduced trade costs, promoted export diversification and as a consequence improved Africa’s competitiveness in the world economy. More recently, Lemi (2017) found that aid for trade flow allocated to trade facilitation had unintended effects on both donor and African countries. Moreover, aid for trade did not lower importing costs as Cali and te Velde (2011) had earlier shown, nor did it have a positive impact on exports as Seck (2016) found. In fact, Lemi found that trade for aid actually helped to reduce export flows to Organization for Economic Cooperation and Development (OECD) countries. With an era of free trade moving closer for the African region and with many studies highlighting the economic development gains that can be achieved through trade facilitation and trade for aid initiatives, the future for the region is looking brighter. However, it is still imperative from a governance and geopolitical perspective that countries continue to reduce corruption and in doing so implement more transparent means of bureaucracy.

1.6 Concluding Observations SSA is a region undergoing significant change. In order to outline areas of this change, this Chapter has given an overview of the four key segments detailed in this book: (1) agriculture and livestock; (2) consumption, poverty, and inequality; (3) financial services, employment and corporate governance; (4) economic integration, international trade and FDI. Our introductory assessment of the Agricultural sector showed that the industry has flourished in recent times. At the heart of this growth has been significant investments in domestic and foreign R&D which have played a very important role in developing the capabilities of the agricultural industry in Africa. Growth has also been driven by the expansion and intensification of cropping systems. However, there is much still to be done to address the problems that are associated with a lack of crop rotation which leads to a depletion of organic carbon levels, which in turn makes the soil less responsive to fertilizer application. Nonetheless, new policy measures such as fertilizer subsidy schemes are having a positive impact. From a poverty and inequality perspective, SSA lags behind other developing nations in a number of key human development indicators. Poverty is very much entrenched in many parts of SSA, with harvest failure, market failure and volatility, conflict and health shocks playing a key role. The issue of inequality is also another important problem, with Africa being the second most unequal continent next to Latin

14

J. Wood and O. Habimana

America. Progress is being made, with deregulation and economic liberalization and improvements in the education sector being instrumental in reducing the number of people living below the poverty line. Access to proper health care has also been important although the rural/urban divide, mothers’ education level and a lack of an effective immunization coverage causing concerns in several SSA countries. Finally, greater gender equality has also been achieved across Africa, with progress being made in the areas of educational attainment, political empowerment, health and survival, and economic participation. The banking system dominates the financial landscape in SSA. Within this environment, homegrown SSA banks play an integral role. In recent times, their growth has been such that they now have larger operations than those of their American and European Banking counterparts. These PABs also have significant reach across the region with most of the major players having operations in more than 30 African countries. While offering traditional banking service they are also heavily involved in capital markets, insurance, pensions, micro-finance, and even non-financial transactions which has in turn led to the creation of new service offerings such as mobile banking and web-based technologies. Moreover, increased interest in Islamic financial practices is also changing the financial landscape in SSA. With growth in international trade flows driving economic growth and development, countries in the region are better placed to integrate their economies, generate foreign exchange, increase technological transfer, generate higher efficiency levels among firms due to greater competition, increase employment opportunities, and alleviate poverty. The establishment of the Africa Continental Free Trade Agreement (AfCFTA) is another positive development with member states having to remove tariffs from some 90% of goods. Such a development would not only allow free access to commodities, goods, and services across the continent but it would also provide a mechanism that facilitates the free movement of business travelers and investments, and create a continental customs union that not only streamlines trade but also attracts long-term regional and foreign investment.

References Adesina, J. O. (2016). Chapter 18: Inequality in sub-Saharan Africa: dimensions and drivers, in World Social Science Report 2016: Challenging inequalities, pathways to a just world. United Nations. Retrieved December 5, 2019 from: https://unesdoc.unesco.org/ark:/48223/ pf0000245946. Adetutu, M. O., and Ajayi, V. (2019). The Impact of Domestic and Foreign R&D on Agricultural Productivity in sub-Saharan Africa. World Development 125: 1–13. Africa Union. (2019). Agenda 2063: The Africa we want. Africa Union. Retrieved online November 13, 2019 from: https://au.int/en/agenda2063/overview. Agrawal, P. (2015). The role of exports in India’s economic growth. Journal of International Trade and Economic Development 24 (6), 835–859. Alene, A. D. (2010). Productivity growth and the effects of R&D in African agriculture. Agricultural Economics, 41 (3–4), 223–238.

1 An Overview: A Multidimensional Economic Assessment of Africa

15

Beck, T., and Maimbo, S. M. (2013). Financial Sector Development in Africa: Opportunities and Challenges. World Bank, Washington. Beintema, N. M., and Stads, G-J. (2011). African agricultural R&D in the new millennium: Progress for some, challenges for many. IFPRI, food policy report. Washington, DC: International Food Policy Research Institute. Bigsten, A. (2014). Dimensions of income inequality in Africa, WIDER Working Paper 2014/050. Booth, D., Cammack, D., Hannigan, J., Kanyongolo, J., Mataure, M., and Ngwira, N. (2006). Drivers of change and development in Malawi, ODI Working Paper 261. London: Overseas Development Institute. Brink. A. B., and Eva, H. D. (2009). Monitoring 25 years of land cover change dynamics in Africa: A sample based remote sensing approach, Applied Geography, 29 (4): 501–512. Buyonge, C., and Kireeva, I. (2008). Trade facilitation in Africa: challenges and possible solutions. World Customs Journal, 2 (1), 41–54. Cali, M., and te Velde, D. W. (2011). Does aid for trade really improve trade performance? World Development, 39 (5), 725–740. Chronic Poverty Research Centre. (CPRC, 2004). The Chronic Poverty Report 2004–05. Manchester. Chronic Poverty Research Centre. Retrieved online December 5, 2019 from: http://www. chronicpoverty.org/resources/cprc_report_2004-2005_contents.html. Cloete, K. (2019, Sept 6). Africa’s new free trade area is promising, yet full of hurdles. World Economic Forum. Retrieved online November 13, 2019 from: https://www.weforum.org/agenda/ 2019/09/africa-just-launched-the-world-s-largest-free-trade-area/. Cornia, G. A., and Martorano, B. (2016). Building the integrated database and the seven sins of inequality measurement in Sub-Saharan Africa, UNDP Africa Policy Notes 2016-02, UNDP Regional Bureau for Africa. New York: UN Development Program. Drechsel, P., Gyiele, L., Kunze, D., and Cofie, O. (2001). Population Density, Soil Nutrient Depletion, and Economic Growth in Sub-Saharan Africa. Ecological Economics, 38: 251–258. European Investment Bank. (2018). Banking in Africa: Delivering on Financial Inclusion, Supporting Financial Stability. European Investment Bank. Retrieved online December 6, 2019 from: https://www.eib.org/attachments/efs/economic_report_banking_africa_2018_en.pdf. Fan, S., Zhang, X., and Rao, N. (2004). Public expenditure, growth and poverty reduction in rural Uganda. DSG discussion paper no. 4. Washington, DC: IFPRI. Food and Agriculture Organization. (FAO, 2015). The State of Food Insecurity in the World 2015. Meeting the 2015 international hunger targets: taking stock of uneven progress. Food and Agriculture Organization Publications, Rome. Frankel, J. A., and Romer, D. (1999). Does trade cause growth? American Economic Review 89 (3), 379–399. Fulginiti, L. E., Perrin, R. K., and Yu, B. (2004). Institutions and agricultural productivity in SubSaharan Africa. Agricultural Economics 31 (2–3), 169–180. Goodhand, J. (2001). Violent conflict, poverty and chronic poverty, CPRC Working Paper 6. Manchester: Chronic Poverty Research Centre. Handley, G., Higgins, K., Sharma, B., Bird, K., and Cammack, D. (2009). Poverty and Poverty Reduction in sub-Saharan Africa: An overview of the issues. Overseas Development Institute Working Paper 299. London. Hamel, K., Tong, B., and Hofer, M. (2019). Future Development: Poverty in Africa is now falling— but not fast enough. Brookings. Retrieved online December 5, 2019 from: https://www.brookings. edu/blog/future-development/2019/03/28/poverty-in-africa-is-now-falling-but-not-fast-enough/. Harvey, P. (2004). HIV/AIDS and humanitarian action, HPPG Briefing Paper No. 14, http://www. odi.org.uk/hpg/papers/hpgbrief14.pdf. Heshmati, A., Kim, J., and Wood, J. (2019). A Survey of Inclusive Growth Policy. Economies. 7(3), 65, https://doi.org/10.3390/economies7030065. Imam, P., and Kpodar, K. (2015). Is Islamic Banking Good for Growth? IMF Working Paper 15/81, International Monetary Fund, Washington.

16

J. Wood and O. Habimana

International Monetary Fund. (IMF, 2016a). Financial Development in Sub-Saharan Africa: Promoting Inclusive and Sustainable Growth. IMF. Retrieved online December 6, 2019 from: https:// www.imf.org/external/pubs/ft/dp/2016/afr1605.pdf. International Monetary Fund. (IMF, 2016b). West African Economic and Monetary Union (WAEMU): Staff Report on Common Policies of Member Countries. IMF Country Report 16/96, International Monetary Fund, Washington, March. International Parliamentary Union. (2016). Women in national parliaments. Retrieved online December 6, 2019 from http://www.ipu.org/wmn-e/classif.htm. Jayne, T. S., Headey, D., and Chamberlin, J. (2014). Land Pressures, the Evolution of Farming Systems, and Development Strategies in Africa: A Synthesis. Food Policy, 48: 1–17. Jayne T. S., and Rashid, S. (2013). Input Subsidy Programs in Sub-Saharan Africa: A Synthesis of Recent Evidence. Agricultural Economics, 44(6): 1–16. Kammer, A., Norat, M., Pinon, M., Prasad A., Towe, C., and Zeidane, Z. (2015). Islamic Finance: Opportunities, Challenges, and Policy Options. IMF Staff Discussion Note. SDN/15/05. IMF. Retrieved online December 11, 2019 from: https://www.imf.org/external/pubs/ft/sdn/2015/ sdn1505.pdf. Karingi, S., and Leyaro, V. (2009). Monitoring aid for trade in Africa: an assessment of the effectiveness of the aid for trade. African Trade Policy Centre, Work in Progress, No. 83. UN Economic Commission for Africa. Kimenyi, M., Adibe, J., Djire, M., Jirgi, A. J.Kergna, A., Deressa, T. T., Pugliese, J. E., and Westbury, A. (2017). The Impact of Conflict and Political Instability on Agricultural Investments in Mali and Nigeria. Africa Growth Initiative Working Paper 17. Brookings. Retrieved online December 4, 2019 from: https://www.brookings.edu/wp-content/uploads/2016/06/14–07-22-Impact-ofConflict-MaliNigeria_FINAL.pdf. Krueger, B. A. (1980). Trade policy as an input to development. American Economic Review 70 (2), 288–292. Manders, S. (2019). Expert Analysis: Africa’s Free Trade Area—Where are we now? Global Trade Review. Retrieved online November 19, 2019 from: https://www.gtreview.com/news/africa/ expert-analysis-africas-free-trade-area-where-are-we-now/. Maredia, M. K., Byerlee, D., and Pee, P. (2000). Impacts of food crop improvement research: Evidence from sub-Saharan Africa. Food Policy, 25(5), 531–559. Masters, W. A., Bedingar, T., and Oehmke, J. F. (1998). The impact of agricultural researching Africa: Aggregate and case study evidence. Agricultural Economics, 19 (1–2), 81–86. Narayanan, B., Sharma, S., and Razzaque, M., (2016). Trade facilitation in the commonwealth: an economic analysis. Margin Journal of Applied Economic Research 10 (3), 305–336. https://doi. org/10.1177/0973801016645218. New Partnership for Africa Development. (NEPAD, 2015). Agriculture in Africa: Transformation and outlook. NEPAD. Retrieved online November 14, 2019 from: https://www.nepad.org/caadp/ publication/agriculture-africa. Nkamleu, G. B. (2004). Productivity growth, technical progress and efficiency change in African agriculture. African Development Review, 16 (1), 203–222. Organization for Economic Co-operation and Development—Food and Agriculture Organization (OECD-FAO, 2016). Agricultural Outlook 2016–2025. FAO-OECD. Ploughshares. (2007). Human Development and Armed Conflict. Retrieved online December 5, 2019 from: http://www.ploughshares.ca/imagesarticles/ACR07/humdev06.pdf. Portugal-Perez, A., and Wilson, J. S. (2009). Why trade facilitation matters to Africa. World Trade Review, 8 (3), 379. Rahman, S., and Salim, R. (2013). Six decades of total factor productivity change and sources of growth in Bangladesh agriculture (1948–2008). Journal of Agricultural Economics, 64 (2), 275–294. Ravallion, M., and Chen S. (2012). Monitoring Inequality. Mimeo. Retrieved online December 5 from: https://blogs.worldbank.org/developmenttalk/files/developmenttalk/monitoring_ inequality_table_1_pdf.

1 An Overview: A Multidimensional Economic Assessment of Africa

17

Sakyi, D., Bonuedi, I., and Opoku, E. E. O. (2018). Trade Facilitation and Social Welfare in Africa. Journal of African Trade, 5, 35–53. Schultz, T. W. (1956). Reflections on agricultural production, output and supply. Journal of farm Economics, 38 (3), 748–762. Seck, A. (2016). Trade facilitation and trade participation: are sub-Saharan African firms different? Journal of African Trade 3 (1–2), 23–39. Seini, A. W. (2002). Political Instability and Agricultural Policy Dynamics. Law and Politics in Africa, Asia, and Latin America, 35 (2): 414–430. Sinha, S., and Lipton, M. (1999) Damaging Fluctuations, Risk and Poverty: A Review. World Bank. Stijns, J. -P. (2015). Recent Trends in Banking in Sub-Saharan Africa: From Financing to Investment. European Investment Bank, Luxembourg. Tittonell, P., and Giller, K. (2012). When Yield Gaps Are Poverty Traps: The Paradigm of Ecological Intensification in African Smallholder Agriculture. Field Crops Research, 143 (1): 76–90. United Nations Conference on Trade and Development. (UNCTAD, 2019). Economic Development in Africa Report 2019: Made in Africa—Rules of origin for enhanced intra-Africa trade. UNCTAD. Retrieved online November 13, 2019 from: https://unctad.org/en/pages/ PublicationWebflyer.aspx?publicationid=2463. United Nations Development Program. (UNDP, 2010). Human Development Report 2009. Technical Report. UNDP. United Nations Department of Public Information. (UNDPI, 2007). Africa Progress. UNDPI. Retrieved online December 5, 2019 from: http://www.un.org/millenniumgoals/docs/ MDGafrica07.pdf. United Nations Economic Commission for Africa. (UNECA, 2019). Fiscal Policy for Financing Sustainable Development in Africa. UNECA. Retrieved online November 13, 2019 from: https:// www.uneca.org/publications/economic-report-africa-2019. Viet, N. C. (2015). The impact of trade facilitation on poverty and inequality: evidence from lowand middle-income countries. Journal of International Trade and Economic Development, 24 (3), 315–340. https://doi.org/10.1080/09638199.2014.898315. Wiggins, S. (2005). Smallholder farming in Africa: Stasis and dynamics. In D. Belshaw and Livingstone A. L. (eds.), Renewing Development in Sub-Saharan AfricaPolicy, Performance and Prospects, (pp. 101–120). London, UK: Taylor & Francis. World Health Organization. (WHO, 2014). The Health of the People: What Works. The African Regional Health Report 2014. Brazzaville, WHO Regional Office for Africa. Willis, G. W. (1990). The Effects of Political Instability on Economic Growth: A Case for SubSaharan Africa. Retrieved online December 4, 2019 from: https://corescholar.libraries.wright. edu/econ_student/82. Wood, J. (2017). An Industrial Sector Analysis of the Factors Effecting FTA Negotiation Outcomes. Global Business Review. 18 (4) 1–16. Wood, J., Wu, J., Li, Y., and Jang, H. (2017a). The Economic Impact of SPS Measures on Agricultural Exports to China: An Empirical Analysis Using the PPML Method. Social Sciences. 6 (51), 1–18. Wood, J., Wu, J., Li, Y., and Kim, J. (2017b). TBT and SPS Impacts on Korean exports to China: empirical analysis using the PPML method. Asian-Pacific Economic Literature. 31 (2), 96–114. Wood, J., Wu, J., Li, Y., and Kim, J. (2019a). The Impact of TBT and SPS Measures on Japanese and Korean Exports to China. Sustainability. 11 (21), 6141. Wood, J., Li, Y., and Wu, J. (2019b). An Analysis of the Trends and Determinants of Intra-Industry Trade Between China and Asia Pacific Economic Cooperation Member Countries. Singapore Economic Review. https://doi.org/10.1142/S0217590819500486. World Economic Forum. (2014). Global Gender Gap Report 2014. Retrieved online December 6, 2019 from: http://reports.weforum.org/global-gender-gapreport-2014/.

Part I

Agriculture and Livestock

Chapter 2

Determinants of Smallholders’ Export-Oriented Cash Crop Production Decisions in Ethiopia: A Case of the Sesame Sector Gemechis Mersha Debela, Engdawork Assefa, Dawit Diriba Guta, and Sosina Bezu Abstract Though Ethiopia is one of the major sesame producing countries in the world, studies have shown that its sesame sector is performing far below its full potential. This study examines the factors that impact smallholders’ decisions about sesame production and the intensity of their production in the western part of Ethiopia using cross-sectional data collected from 400 sampled households. The study uses the double-hurdle model to empirically identify the factors that affect households’ separate decision-making processes regarding their participation in sesame production (using the probit model) and the intensity of their production (using a truncated model). The results show that decisions about opting for sesame production are significantly constrained by farmers’ resource endowments and market information. Smallholders with more education, land, food availability, and access to credit are more likely to plant sesame. Among the sesame producers, the level of production is also positively and significantly correlated with number of oxen household owns, the family’s access to food for the whole year, access to credit and off-farm activities, access to market price information, and selling channels. The direction of this influence and the significance of some of the explanatory variables vary between the two stages of production decisions. Consequently, it is imperative to devise policies that enhance farmers’ food security in areas which have the potential for growing cash crops so as to enable them to fully engage in the production of market-oriented cash crops. Interventions that promote farmers’ access to credit, alternative income G. M. Debela (B) · E. Assefa · D. D. Guta College of Development Studies, Center for Environment and Development Studies, Addis Ababa University, Addis Ababa, Ethiopia e-mail: [email protected] E. Assefa e-mail: [email protected] D. D. Guta e-mail: [email protected] S. Bezu Chr. Michelsen Institute, Bergen, Norway e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 J. Wood and O. Habimana (eds.), A Multidimensional Economic Assessment of Africa, Frontiers in African Business Research, https://doi.org/10.1007/978-981-15-4510-8_2

21

22

G. M. Debela et al.

earning sources, establishing updated sesame market information flows, and appropriate selling channels are also useful. These will be of paramount importance in commercializing and shifting rural farming to high value crops like sesame that have high export potential. Keywords Sesame production · Determinants · Double-hurdle model · Ethiopia Codes Q12 · C21

2.1 Introduction Except Africa, all other developing regions in the world have achieved the Millennium Development Goal of reducing poverty by half between 1990 and 2015 (UN 2015). It is argued that since most of Africa’s poor depend largely on agriculture for their livelihoods (Carmody et al. 2015), improving the productivity, profitability, and sustainability of the agricultural sector is the main pathway out of poverty for the continent (Asfaw et al. 2012; Christiaensen et al. 2011; Dawson et al. 2016). The agricultural sector’s growth in Africa has been slow (Diao et al. 2012). In particular, agricultural productivity in sub-Saharan Africa (SSA) remains stagnant (Tittonell and Giller 2013). Over the past four decades, growth in agricultural productivity in SSA averaged only 2.4% while productivity in the rest of the developing world improved by 4% (Dzanku et al. 2015). Like many sub-Saharan African countries, agricultural growth holds the key to economic growth and development in Ethiopia where agriculture accounts for 40.2% of the GDP, 80% of the employment, and 70% of the country’s export earnings (UNDP, 2015). About 85% of Ethiopia’s population lives in rural areas and depends on agriculture for necessities and as a source of livelihood (Negatu et al. 2016). Ethiopia’s agricultural sector recorded remarkably rapid growth in the last decade and was also the major driver of poverty reduction (World Bank 2015). The agriculture sector was also dominated by smallholders with about 12 million smallholder farming households accounting for an estimated 95% of the agricultural production and 85% of all employment creation (FAO 2015). Hence, this sector’s performance largely determines the fate of the country’s economy. Ethiopia is one of the major sesame producing countries in the world, which was ranked number 5 in production after Myanmar, India, China, and Sudan till 2010 (FAO 2015). But it gave its rank to Tanzania in 2011 mainly because of a decline in its area of production and the consequent decline in its production (ATA 2015). The growth achieved in production was mainly explained by extension activities and although there was an increase in average yields from 2000 to 2012, they were still considered low as compared to sesame production’s full potential (FAO 2015). Despite the country’s immense potential to increase its production and productivity, government policies are targeted at promoting commercial agriculture and despite a significant increase in international demand for sesame, studies show that a number

2 Determinants of Smallholders’ Export-Oriented Cash Crop …

23

of challenges have hampered the development of this sector (ATA 2015). As a result, Ethiopia’s sesame sector is performing below its full potential (FAO 2015). Studies on the Ethiopian sesame sector are dominated by an analysis of the sector’s value chain which indicates that both production and marketing of sesame face many constraints. Though there is high potential, sesame production is characterized by low productivity which has recently been showing declining trends (ATA 2015). The main constraints are: low use of improved seeds, fertilizers and cultivars; biotic stress and lack of knowledge about adequate post-harvest crop management; poor improved input supply systems and insufficient extension services; and little research support for increasing yields (ATA 2015; Coates et al. 2011; FAO 2015; Meijerink 2014; SID-Consult 2010; Sorsa 2009, Wijnands et al. 2007). Studies further indicate that the sesame sector faces marketing constraints and is losing out on opportunities of gaining high prices due to poor quality, poor organization of the sesame value chain, and underdeveloped market information systems. Most of the studies have assessed general production and the trade set-up of the sesame sector by mainly focusing on the marketing aspects. By focusing on common sesame production and marketing related problems which are mainly external to individual farm households, these studies have overlooked factors affecting the production, productivity, and marketing decisions at the individual household level. Methodologically, they rely on general descriptive assessments rather than on rigorous statistic and econometric analyses. Hence, it is important to identify and analyze household specific factors which are hampering households from participating in sesame production and achieving the highest levels of production. This needs a robust analysis by identifying specific agro-ecologically feasible areas for growing sesame. Given this background, this study addresses the following research question: what are the main factors influencing and determining smallholder farmers’ decisions to participate in sesame production and the intensity/level of sesame production? It analyzes household specific factors influencing smallholders’ production decisions and level of participation in sesame production in the East Wellega Zone for the production period 2017. The general objective of the study is analyzing smallholder sesame producers’ production decisions and performance. Specifically, it analyzes the factors that influence farmers’ decisions to participate in sesame production and examines the determinants of the level of sesame production.

2.2 Methodology 2.2.1 Description of the Study Area This study was undertaken in the west of Ethiopia, in the East Wellega Zone in Oromia region. The zone is geographically located between 9° 31’ 9” North latitude and 36° 45’ 27” East longitude. Based on the 2007 Census conducted by the Central

24

G. M. Debela et al.

Fig. 2.1 Location of the study area in Oromia regional state

Statistics Authority (CSA), this zone had a total population of 1,213,503 (606,379 men and 607,124 women) with an area of 12,579.77 square kilometers and a population density of 96.46. There are 18 districts in this zone and Sasiga and Gida Ayana districts were selected for this study. According to data from the Rural and Agricultural Development Office the total population of Sasiga woreda was 80,814 (41,326 men and 39,488 women); 2,573 or 3.18% of its population was urban dwellers. A survey of the land in this woreda showed that it had 11.9% arable or cultivable land, 2.8% pasture land, 1.6% forests, and the remaining 83.7% was swampy and marshy or otherwise unusable. On the other hand, the total population of Gida Ayana district was about 142,408 of which 66,918 (47%) were women and 75,490 (53%) was men. A survey of the land in this woreda showed that 65.7% of the land was arable or cultivable (61% was under annual crops), 22.8% was pasture land, 8.7% was forests, and the remaining 2.8% was unusable. Sesame and khat are two important cash crops in this woreda (see Fig. 2.1).

2.2.2 Data and Methods of Data Collection This research is primarily based on primary data generated through a cross-sectional survey during the 2017 production season. Semi-structured questionnaires were used

2 Determinants of Smallholders’ Export-Oriented Cash Crop …

25

for collecting primary data on decisions about crop production. Both closed and open-ended semi-structured questionnaires were used for generating the information. Additionally, key informant interviews were done to collect more information on some pertinent issues. Further, with checklists/unstructured questions, discussions and focus-group discussions (FGDs) were also held with selected sesame producing farmers to get some general information about the study area.

2.2.3 Sampling Procedure A three-stage sampling technique was used to draw an appropriate sample for the study. In the first stage, Sasiga and Gida Ayana woredas were purposively selected from the top five sesame growing woredas in the East Wellega Zone based on information obtained from the Zonal Agricultural Office. Second, the study included 50% of the total sesame growing kebeles from each of the two selected woredas using the simple random sampling method. The selection of an equal percentage was a feature provided by the probability proportional to the size sampling design. This gives all kebeles in the population an equal probability of being selected for the sample. Based on these criteria, five kebeles were selected from each woreda randomly. Finally, 400 farm households were randomly selected from lists of names of household heads in the kebeles using a computer-generated random number table. The total sample was distributed among the different sample kebeles based on the probability proportional to their total size.

2.3 Data Analysis 2.3.1 Participation in Sesame Production (Probit Model) The recommended approach which was applied in this study is the double-hurdle (D-H) model. This model assumes that the farmers face two hurdles in any agricultural decision making process (Cragg 1971; Humphreys 2010; Sanchez 2005). Accordingly, the decision to participate in an activity is taken before taking a decision regarding the level of participation in the activity. The double-hurdle model was chosen because it allows making a distinction between the determinants of production participation and the level of participation in sesame production in two separate stages. This model estimation procedure involves running a probit regression for identifying factors that affect the decision to participate in the activity using all the sample population in the first stage, and a truncated regression model of the participating households to analyze the extent of participation in the second stage. To derive the likelihood function, we started with the first stage (production decision) where the households were identified according to whether they were producers

26

G. M. Debela et al.

or not using a probit analysis. To do so, let Pi denote a binary indicator function taking the value ‘1’ if farmers participated in sesame production in the 2017 production year and ‘0’ otherwise. Further, let Li denote the land allocated to sesame produced in the specified production year. Then we can derive the likelihood function for the standard double-hurdle model as:       Y − χ2 β 1 ×φ ln ϕ(χ1 γ ) × ln L = G1=1 σu σu .   

     + ln ϕ χ2 β σu × (1 − φ(χ1 γ )) + ln 1 − ϕ χ2 β σu G2=1

G3=1

(2.1) where ϕ and φ are the probability density and cumulative distribution function of the standard normal variable respectively; G1, G2, and G3 are indicator functions showing whether a given observation belongs to group one, two, or three respectively: households producing sesame, households wanting to produce but reporting zero production, and households choosing not to produce. Equation (2.1) can be estimated using the maximum likelihood (ML) techniques, which give consistent estimates of the parameters. If ui and ei are independent, the ML function can be separated into a probit and a truncated normal regression model.

2.4 Empirical Model’s Specifications Based on this, the linear probit model can be specified as: P(Yi = 1) = β0 + βi X i + ε

(2.2)

where P is the probability of an individual farm household participating in sesame production, βi is the vector of parameters to be estimated, Xi is the vector of exogenous explanatory variables expected to influence the participation decision’s probability, and ε is the error term. The probit model specifies the functional relationship between the probability of participating in sesame production and the list of various explanatory variables thought to influence this decision. These factors can be either continuous or discrete explanatory variables. Therefore, the reduced functional relationship between the binary dependent variable (producing sesame or not) and a list of explanatory variables for the empirical analysis can be specified using the basic probit model’s specifications as: Pr (P) = β0 + β1 X 1i + β2 X 2i + . . . + β14 X 14i εi

(2.3)

where Pr is the probability of an individual household participating in sesame production represented by (P = 1), βi’s are the coefficients to be estimated, and εi is the error

2 Determinants of Smallholders’ Export-Oriented Cash Crop …

27

term. The lists of X1i to X14i include age of household head, sex of household head, education level of the household, family size in adult equivalence, number of oxen owned, total land owned, annual income earned from livestock sale, access to family food, access to credit, cooperative membership, access to farm extension services, access to non-farm income, distance from household’s home to the extension service center, and distance between a household’s home and the nearest marketplace.

2.4.1 Level of Participation in Sesame Production (Truncated Model) In the second stage of the double-hurdle model we examined factors affecting the level of sesame production, conditional on participation decisions using the truncated regression analysis. This involves the truncated regression that can be specified as: L = L ∗ i f L ∗ > 0 and Y = 1 L = 0 other wise

(2.4)

From this, we can specify the reduced form of the truncation model as: L = β0 + βi Z i + υi

(2.5)

where L is the size of the land allocated to sesame production in hectares, L* is the latent variable which indicates that the land size is greater than zero, βi is the vector of parameters to be estimated, Zi is the vector of exogenous explanatory variables, and υi is the error term. The empirical model used in this study assumes that the total quantity of sesame produced is a linear function of continuous and dummy explanatory variables and is specified as: L = β0 + β1 X 1i + β2 X 2i + . . . + β16 X 16i + νi

(2.6)

where L is the size of land allocated to sesame production in hectares in the 2017 production year, βi’s are the coefficients to be estimated, and υi is the error term. The lists of X1i to X16i include sex of household head, education level of the household, family size in adult equivalence, number of oxen owned, total land owned, annual income earned from livestock sales, food availability for the family, access to credit, cooperative membership, access to farm extension services, access to nonfarm income, years of sesame farming, access to market price information, sesame selling channels, distance from household’s home to the extension service center, and distance between household’s home and the nearest marketplace.

28

G. M. Debela et al.

2.5 Statistical and Specification Tests Before executing the final model regressions, all the hypothesized explanatory variables were checked for the existence of statistical problems such as multicollinearity problems. The variance-inflating factor (VIF) technique is commonly used for this and we also used this to detect a multicollinearity problem among the continuous explanatory variables (Gujarati 2004). Mathematically, VIF for individual explanatory variables (Xi) can be computed as: VIF (Xi) = 1/(1 − R2)

(2.7)

where R2 is the coefficient of correlation among the explanatory variables. According to Gujarati (2004), the larger the value of VIF the more the collinearity will be among one or more of the model’s explanatory variables. As a thumb rule, if the VIF of a variable exceeds 10, which will happen if a multiple R-square exceeds 0.90, that variable is said be highly collinear. We used the contingency coefficient (CC) method to detect the degree of association among discrete explanatory variables (Healy 1984). The discrete/dummy variables are said to be collinear if the value of CC is greater than 0.75. Mathematically: CC =

X2 n + X2

(2.8)

where CC is the contingency coefficient, n is the sample size, and X2 is the chi square value.

2.6 Definitions of Explanatory Variables and Working Hypothesis The first dependent variable in this study is production participation, which is a dichotomous variable taking the value 1 if a household participates in sesame production and 0 otherwise. This lets us analyze the main determinants of households’ sesame production participation decisions using the probit regression. The second dependent variable is the size of land allocated for sesame production in hectares in the production year 2017. We used this variable to capture the factors that influenced the intensity of sesame produced or level of sesame production participation after a household decided to produce sesame by using a truncated regression. As shown in Table 2.A1 in the Appendix, we have included personal, household, socioeconomic, crop-specific, institutional, marketing, and geographic location variables in our analysis. The selection of these variables is guided by existing empirical literature (for example, Ashfaq et al. 2008; Burke 2009; FAO 2012; Fetien et al. 2009; Mishra and

2 Determinants of Smallholders’ Export-Oriented Cash Crop …

29

El-Osta 2002; Sichoongwe 2014; Weiss and Briglauer 2000; Wondimagegn et al. 2011). It is expected that as age (a personal characteristic of the household) increases, farmers would acquire knowledge and experience through continuous learning which will help them to actively participate in the production of market-oriented cash crops. However, previous studies show that elderly farmers look at farming as a way of life while young farmers look at farming as a business opportunity for their families’ sustenance (FAO 2012). As other studies have also confirmed, since young farmers do farming as a business following a profit oriented approach, it is expected that the younger farmers will diversify more than their elder counterparts (Ashfaq et al. 2008; Mishra and El-Osta 2002; Sichoongwe 2014). Previous studies have also shown that the sex of the household head has a positive effect on diversification (Fetien et al. 2009; Wondimagegn et al. 2011). These studies show that male-headed households are more likely to participate in sesame production as compared to female-headed households. Education of the household head is another personal characteristic which can positively affect farmers’ participation in market-oriented crop diversification (Ibrahim et al. 2009; Sichoongwe 2014). Previous studies have also shown that when it comes to personal characteristics of households, the larger the household size (active family labor), the more likely it is to diversify from food crops to cash crops (Benin et al. 2004; Wondimagegn et al. 2011). Coming to socioeconomic variables, the number of oxen owned by a household is expected to positively enhance the probability of participating in producing sesame. Previous studies have also shown that farmers’ landholding size is another variable which has a positive and significant relationship with participation in agricultural production (Poudel et al. 2012; Wondimagegn et al. 2011). Access to off-farm activities is another variable which can have either of two effects on farmers’ participation in sesame production and their level of production. Mishra and El-Osta (2002) and Weiss and Briglauer (2000) report a negative effect. Among the off-farm activities that farmers engage into earn annual income is selling livestock. Wondimagegn et al. (2011) show that off-farm income sources negatively impacted farmers’ farm involvement. Another important economic factor which is expected to highly influence farmers shift to cash crop production is the availability (access) of food for the family for the whole year. It is also argued that smallholder farmers in developing countries participate in the production of cash crops only if they can produce enough of family food. Thus, if farmers have potential and experience in producing sufficient family food for the whole year, it is expected that they are more likely to participate in the production of cash crops such as sesame. Farm-specific characteristics are also the main determinants of farmers’ involvement in the production of sesame. It is expected that farmers’ who have participated in the production of sesame for long years will have experience of its production and were likely to produce more in the survey year. Farmers’ access to credit is one of the main institutional determinants. Those who have more access to credit services are expected to produce market-oriented cash crops like sesame. As also shown by Bruke (2009), availability of credit is an important determining factor in all stages of a farmer’s production and marketing

30

G. M. Debela et al.

decisions. Being members of cooperatives is another institutional factor which can positively influence farmers’ cash crop production since it is a place where they get access to necessary inputs, market information, and selling their produce at better prices. An important marketing variable is farmers’ access to updated price information about the sesame market. It was hypothesized that access to price information positively affected the income earned from sesame sales and thus from the level of sesame production. Another variable is the sesame selling channel that either sells directly to the buyers or sells through brokers. We hypothesized that the existence of middlemen in the selling channel will negatively affect farmers’ earnings and hence their level of production. Further, distance from the extension service center is a geographical factor which can impact farmers’ production behavior. According to Wondimagegn et al.’s (2011) findings, the larger the number of contacts that a farmer has with an extension agent, the more likely he is to engage in more production. Fetien et al. (2009) also support this hypothesis. The same analogy also works for farmers’ distance to the nearest market center.

2.7 Results and Discussion 2.7.1 Descriptive Statistics Tables 2.1 and 2.2 present the summary statistics for the dummy and continuous variables. As can be seen in the tables, about 93% were male-headed households. On average, the sample respondents had been engaged in farming for 11 years and about 23% of the sample household heads did not have formal schooling. The mean Table 2.1 Association between determinants (discrete variables) of sesame production decisions Variables

Categories

Producers (n)

Non-producers (n)

Chi-square value

Sex

Female

24

5

22.140a

Credit access

Yes

153

46

8.341a

Coop membership

Yes

189

65

5.761a

Farm extension

Yes

173

62

3.090c

Income non-farm

Yes

60

35

2.999c

Food availability

Yes

251

66

10.431b

Access to price info

Yes

195

76

1.170ns

Selling channels

Via brokers

150

99

31.764c

Note a,b,c, ns significant at 1, 5, 10% and not significant Source Computed from survey data

2 Determinants of Smallholders’ Export-Oriented Cash Crop …

31

Table 2.2 Association between determinants (continuous variables) of sesame production decisions Producers

Non-Producers

Mean

SD

Mean

SD

t-value

Age

41.539

12.093

41.313

10.375

0.188 ns

Education

4.555

3.575

4.008

3.567

3.399***

Family size in AE

9.21

3.676

4.886

4.241

10.313***

Number of Oxen

2.103

1.526

1.807

1.58

1.758*

Land total

4.410

2.959

3.209

2.523

3.863***

Year Sesame

10.934

7.393

11.437

8.503

−0.594ns

Income livestock

7.244

8.030

6.847

8.124

2.451**b

Distance extension

2.285

4.251

2.229

2.930

0.131ns

Distance market

2.506

2.092

2.323

2.047

0.608ns

Note ***, **, * significant at 1, 5, and 10%; ns not significant Source Computed from survey data

family size of the sample households measured in adult equivalence (AE)1 was 7.1. On average, respondents owned 1.96 oxen and they owned 4.26 units of livestock measured in tropical livestock units (TLUs) which are equivalent to 0.81 TLU per adult equivalent. The total size of the land owned by the sample respondents ranged from 0.25 to 24 hectares while the average size was 3.81 hectares. Nearly half the total respondents had access to any form of credit, 66% of them had cooperative membership, and 24% were engaged in non/off-farm activities. Sixty percent of the respondents had access to food for the whole year and 59% of them had access to farm extension services. The summary statistics in Tables 2.1 and 2.2 also show the differences in characteristics between sesame producing and non-producing households. The chi-square analysis shows that a larger proportion of sesame producing households were members of cooperatives; had access to market price information; and had food available for the whole year. The results further show that greater proportions of sesame producing households were male-headed, had access to extension advisory services and credit, and had no off-farm activities, as compared to their counterparts. The independent t-test’s results show that there was a significant mean difference between sesame producing and non-producing households with respect to the active family labor force, size of total land owned, and productivity of sesame.

1 Family

size is calculated by converting difference in age and sex of family members using the conversion factor given in Table 2.A2 in the Appendix while Table 2.A3 in the Appendix gives the conversion factor used for computing TLU.

32

G. M. Debela et al.

2.8 Econometric Analysis 2.8.1 Determinants of Sesame Production Participation Decisions (Probit Regression) A probit model was used to identify potential explanatory variables affecting households’ sesame production participation decisions. Before doing the analysis, multicollinearity was checked using VIF for continuous variables and CC for dummy variables. The calculated VIF values were all less than 10 (the cut-off point) and contingency coefficients were less than 0.75 (the cut-off point) which indicates that multicollinearity is not a serious problem. Table 2.3 provides the parameter estimates and average marginal effects respectively of the probit model’s results. As expected, the education level of the household head had a positive and significant (p < 0.05) relationship with the probability of the household participating in sesame production. Govereh and Jayne (2003), also found that the education level Table 2.3 Probit regression of factors determining the probability of participating in sesame production Producers

Coef.

St.Err

t-value

Sex Age

Marginal effects

0.125

0.352

0.35

0.038

−0.004

0.007

−0.56

−0.001

Education

0.045

0.022

2.02**

0.014

Family size in AE

0.048

0.033

1.46

0.015

Number of Oxen

0.073

0.054

1.35

0.022

Land total

0.100

0.031

3.19*

0.031

Income livestock

−0.056

0.094

−0.59

−0.017

Credit access

0.407

0.150

2.72*

0.125

Coop membership

0.266

0.175

1.51

0.082

Farm extension

0.171

0.150

1.14

0.053

Income non-farm

−0.429

0.169

−2.53**

−0.131

Food availability

0.365

0.178

2.05**

0.112 −0.013

Distance extension

−0.041

0.017

−2.38**

Distance market

0.048

0.030

1.60

0.015

_cons

−0.795

0.507

−-1.57

0.038

Mean dependent var

0.702

SD dependent var

0.458

Pseudo r-squared

0.115

Number of obs

399.000

Chi-square

55.770

Prob > chi2

0.000

Akaike crit. (AIC)

460.508

Bayesian crit. (BIC)

520.342

Note *,** Significant at 1, 5, and 10% probability level, respectively Source Computed from survey data

2 Determinants of Smallholders’ Export-Oriented Cash Crop …

33

of the household head was the most critical determinant of smallholders’ decisions to produce cotton in Zimbabwe. Among the socioeconomic variables, the total size of the land owned by a household positively and significantly impacted the probability of its taking a decision to participate in sesame production (p < 0.01). For each additional hectare of land owned, households were 3.1% more likely to participate in sesame production. This is reasonable because larger farms are not only wealthier but also have a higher capacity to expand agricultural production that in turn enables them to produce more cash crops. This result is also consistent with Poulton et al. (2001) who suggest that land is an important factor which influences farmers’ decisions to produce a cash crop. Availability of food all through the year for the family influenced the households’ probability of participating in sesame production positively and significantly (P < 0.05) as households with access to food for the whole year were 11.2% more likely to participate in sesame production as compared to those who did not have food available throughout the year. This is probably because farmers first want to secure food for their families and participate in cash crop production only if they have potential and experience in producing sufficient family food for the whole year. This is in line with Jayne (1994), who attests that the food security condition is one possible factor that limits smallholder farmers’ cash crop production. Agreeing with the a priori assumption, access to credit for sesame production had a positive and significant (p < 0.01) effect on the probability of participating in sesame production. A household that had access to credit for sesame production was 12.5% more likely to participate in sesame production as compared to a household that had no access to credit. This suggests that farmers’ access to credit for sesame production is of paramount importance in their decisions to participate in the production of sesame as they can get finance for agricultural inputs like improved seeds and fertilizers. In line with our result, Burke’s (2009), findings also suggest that credit is an important determinant in all stages of a farmer’s production and marketing decisions. This finding is also confirmed by Lukanu et al. (2004).

2.8.2 Determinants of Intensity of Sesame Production (Truncated Regression) A truncated model was used to identify potential explanatory variables affecting households’ sesame production intensity or level captured via size of land allocated to sesame production. The results show that there was no multicollinearity and series association problem among the variables. Table 2.4 gives the parameter’s estimates and the average marginal effects of the truncated model’s results. As hypothesized, the number of oxen owned by a household had a positive effect and significantly determined the level of participation in sesame production where a household with one more ox was 7.6% more likely to increase the level of sesame produced. Since an ox is the most important means of land cultivation and asset in

34

G. M. Debela et al.

Table 2.4 Truncated Regression of determinants of intensity of sesame production Land sesame Sex

Coef. 1.649

St.Err

t-value

Marginal effects

0.631

2.61***

0.024

0.383

Education

0.103

0.060

1.73*

Family size in (AE)

0.074

0.054

1.38

0.017

Number of oxen

0.325

0.117

2.78***

0.076

1.96**

Land total

0.114

0.058

Year Sesame

−0.045

0.028

−1.56

−0.010

0.026

Income livestock

−0.042

0.030

−1.41

−0.010

Food availability

1.045

0.500

2.09**

Credit access

0.994

0.438

2.27**

0.231

Coop membership

0.716

0.556

1.29

0.166

Access price info

1.062

0.503

2.11**

0.247

Farm extension

0.464

0.428

1.09

0.108

Income non-farm

0.948

0.446

2.13**

0.220

0.243

−1.274

0.707

−1.80*

0.049

0.044

1.12

0.011

Distance market

−0.088

0.074

−1.19

−0.020

_cons

−7.065

2.276

−3.10***

/sigma

1.525

0.215

7.10***

Mean dependent var

0.902

SD dependent var

Selling channels Distance extension

Number of obs Prob > chi2

281.000 0.042

Chi-square Akaike crit. (AIC)

−0.296

0.914 26.983 465.213

Note ***,**,* Significant at 1, 5, and 10% probability level, respectively Source Computed from survey data

poor rural areas in Ethiopia, the number of oxen available to a household positively enhances the level of crop production. The size of total land owned by a household impacted the level of sesame production positively and significantly. This is reasonable because larger farms are not only wealthier but also have a higher capacity to expand agricultural production that in turn enables them to produce more cash crops. Edriss and Simtowe (2002), also show that the increment of land allocated to crops increases production and output. Availability of food throughout the year for the family was found to have a positive and significant impact on households’ level of sesame production. This is probably because farmers first want to secure food for their families and they participate in cash crop production only if they have potential and experience in producing sufficient food for their families for the whole year. Boughton et al. (2007), also argue that the main challenge and constraint for smallholder farmers’ participation in cash crop production is low productivity in food crops and their market failures. Access to non/off-farm activities also had a positive and significant influence (P

2 Determinants of Smallholders’ Export-Oriented Cash Crop …

35

< 0.05) on households’ level of sesame production. This can be explained by the fact that if the farmers have access to alternative income generating activities which enable them to purchase more inputs they are more likely to participate in sesame production. Further, the regression results show that access to credit for sesame production was one of the determining factors that influenced the level of sesame production positively and significantly (p < 0.05). Households’ access to market price information positively and significantly (p < 0.05) impacted their level of sesame production. This is reasonable because having access to price information builds farmers’ confidence and expectations in the market situation which in turn will encourage them to produce more output. This result is in line with the findings of Cadot et al. (2006), who showed that information communication facilities were a major determinant of participating in cash crop production. The regression results further show that the sesame selling channels had a negative and significant effect on the level of sesame production. This shows that those households who sell sesame directly to buyers allocate less land to sesame than those who sell their output via brokers. This is because a broker can help farmers sell sesame at higher prices which encourages them to produce more.

2.9 Conclusion and Policy Implications This study analyzed the determinants that influence smallholders’ probability of participating in sesame production–one cash crop that has export potential—and its production levels based on data collected from the western part of Ethiopia. Using the double-hurdle model, the study empirically identified the factors affecting farmers’ decisions regarding participating in sesame production and their level of production. Our study proves that households’ who have potential and experience in producing sufficient food for their families lead in intensifying sesame production. Farmers’ access to credit for sesame production is a critical factor that helps determine the level of participation in sesame production. Though access to off-farm activities initially reduces farmers’ probability of participating in sesame production, later it increases the size of the land allocated for sesame production. This possibly indicates that as farmers earn more from non-farm activities, there is a higher probability of their purchasing farm inputs and food for their families easily enabling them to intensify sesame production by allocating more land to it. The regression results show that farmers’ access to price information and type of sesame selling channels they use have significant positive and negative effects on the level of sesame production respectively. Based on the findings we make some recommendations for improving farmers’ participation in the production of the market oriented cash crop sesame. To improve farmers’ participation in the production of sesame, more emphasis should be given to insuring farmers’ food security throughout the year. Farmers’ participation in sesame production and their level of participation can also be improved by improving their access to sesame specific credit mainly from formal sources. The positive

36

G. M. Debela et al.

impact of access to off-farm activities on sesame production indicates that the government’s efforts at diversifying rural livelihoods should be promoted to enhance commercialization of rural farming. Having observed the vital role that access to price information and getting proper sesame selling channels play, having effective sesame marketing systems in place will encourage households’ to allocate more land to sesame production. In this regard, what is needed are interventions to solve inputoutput hurdles in the sesame market mainly through access to updated market price information and better selling channels. We also recommend a further study on a detailed sesame marketing system to identify marketing factors that work against the well-being of farmers.

Appendices Description of explanatory variables and hypotheses Description of explanatory variables and hypotheses Dependent variable

Description

Producers

Sesame production participation decisions; 1 if household head has participated in sesame production and 0 otherwise

Land Sesame

Land allocated for sesame cultivation in hectares

Expected effect

Independent variables Age

Age of household head

“+”, “−”

Sex

Gender of the household head; 1 if male and 0otherwise

“+”, “−”

Education

Educational status of household head; years of formal schooling and 0 if not attended

“+”

Active family labor (AE)

Family labor force in adult equivalent

“+”, “−”

Number of oxen

Number of oxen that the household owns

“+”

Land total

Size of total land owned by the household in hectares

“+”

Income livestock

Annual income the household head earned by selling livestock

“+”, “−”

Year sesame

Number of years since the household head started sesame production

“+”

Credit access

Access to credit for sesame production; 1 if the household head has access 0 otherwise

“+”

Coop membership

Cooperative membership; 1 if the household head is a member and 0 otherwise

“+”

Farm extension

Access to farm extension (FTC)

“+” (continued)

2 Determinants of Smallholders’ Export-Oriented Cash Crop …

37

(continued) Description of explanatory variables and hypotheses Dependent variable

Description

Expected effect

Income non-farm

Participation in off/non-farm activities; 1 if the headearns an income by participating in off/non-farm activities and 0 otherwise

“−”

Food availability

Availability of food all over the year; 1 if the household head has food available for the whole year and 0 otherwise

“+”

Access price info

Access to market price information; 1 if the household head has access and 0 otherwise

“+”

Selling channels

Sesame selling channels; 1 if sold directly and 0 if sold via brokers

“+”, “−”

Distance extension

Distance from the respondent’s residence to thefarm extension center measured in kilometers

“−”

Distance market

Distance from the respondent’s residence to thenearest market measured in kilometers

“−”

Source Own Selection based on literatures

Conversion factor for computing adult equivalent Age group (Years)

Adult equivalent Male

Female

50

1

0.7

Source Storck et al. (1991)

Conversion factors used for estimating tropical livestock unit (TLU) equivalent Animal category

TLU

Calf

0.25

Donkey (young)

0.35

Weaned calf

0.34

Camel

1.25

Heifer

0.75

Sheep and goat (adult)

0.13

Cow and ox

1.00

Sheep and goat (young)

0.06 (continued)

38

G. M. Debela et al.

(continued) Animal category

TLU

Horse

1.10

Chicken

0.013

Donkey (adult)

0.70

Source Storck et al. (1991)

References Asfaw, S., B. Shiferaw, F. Simtowe, and L. Lipper. (2012). Impact of modern agricultural technologies on smallholder welfare: evidence from Tanzania and Ethiopia. Food Policy, 37(3):283–295. Ashfaq, M., S. Hassan, Z.M. Naseer, A. Baig, and J. Asma. (2008). Factors affecting farm diversification in rice-wheat. Pakistan Journal of Agricultural Sciences, 45(3): 45–47. ATA. (2015). Sesame Value Chain Development Strategy (Working Document 2015–2019). Addis Ababa, Ethiopia. Benin, S., B. Smale, M. Pender, J. Gebremedin Berhanu, and M. S. Ehui. (2004). Determinants of cereal crop diversity on farms in the Ethiopian highlands. Journal of Agricultural Economics, 31(2–3): 197–208. Boughton, D., D. Mather, C. Barrett, R. Benfica, A. Abdula, D. Tschirley, and B. Cunguara. (2007). Market Participation by Rural Households in a Low-Income Country: An Asset-Based Approach Applied to Mozambique. Faith and Economics, 50: 64–101. Burke, W. (2009). Triple hurdle model of smallholder production and market participation in Kenya’s dairy sector. MSc thesis. Cadot, O., L. Dutoit, and M. Olarreaga (2006). How Costly Is It for Poor Farmers to Lift Themselves Out of Subsistence? World Bank Policy Research Working Paper 3881. Carmody, P., J. Burrell, E. Oreglia, I. Management, A. Benson, T. Jafry, et al. (2015). New realities, new challenges : new opportunities for tomorrow’s generation. Vol. 21, Information Technology for Development, pp. 177–192. Christiaensen, L., L. Demery, and J. Kuhl.(2011). The (evolving) role of agriculture in poverty reduction: an empirical perspective. J Dev Econ, 96(2):239–254. Coates, M., R. Kitchen, G. Kebbell, C. Vignon, C. Guillemain, and R. Hofmeister. (2011). Financing Agricultural Value Chains in Africa. Focus on Coffee and Sesame in Ethiopia. GIZ Bonn and Eschborn, Germany. Cragg, J. (1971). Some statistical models for limited dependent variables with application to the demand for durable goods. Econometrica, 39: 829–844. Diao, X., J. Thurlow, S. Benin, and S. Fan (2012). Strategies and Priorities for African Agriculture: Economywide Perspectives from Country Studies. International Food Policy Research Institute (IFPRI research monograph). Dawson, N., A. Martin, and T. Sikor. (2016). Green revolution in Sub-Saharan Africa: implications of imposed innovation for the wellbeing of rural smallholders. World Dev, 78: 204–218. Dzanku, F. M., M. Jirström, and H. Marstorp. (2015). Yield gap-based poverty gaps in rural SubSaharan Africa. World Dev, 67: 336–362. Edriss, A. K., and F. Simtowe. (2002).Technical Efficiency in Groundnut Production in Malawi: An Application of a Frontier Production Function. UNISWA Research Journal of Agriculture, Science and Techn’gy, 6 (1): 45–50. Food and Agriculture Organization of the United Nations. (2012). Sustainable Crop Production Intensification. Twenty-third Session. Rome: FAO.

2 Determinants of Smallholders’ Export-Oriented Cash Crop …

39

FAO. (2015). Analysis of price incentives for Sesame seed in Ethiopia, 2005–2012. Technical notes series, MAFAP, by Kuma Worako, T., MasAparisi, A., Lanos, B., Rome. Fetien, A., A. Bjornstad, and M. Smale. (2009). Measuring on farm diversity and determinants of barley diversity in Tigray, northern Ethiopia. Momona Ethiopia Journal of Science, 1 (2): 44–66. Govereh, J., and T. Jayne. (2003). Cash cropping and food crop productivity: synergies or tradeoffs? Agricultural Economics, 28 (1): 39–50. Gujarati, D. (2004). Basic Econometrics (4th ed.). New York: McGraw-Hill Co. Science. Healy, F. J. (1984). Statistics: A Tool for Social Research. California: Wads Worth Publishing Company. Humpheys, R. B. (2010). Dealing With Zeros in Economic Data, December 14, 2010. available at http://www.ualberta.ca/~bhumphre/class/zeros_v1.pdf. Ibrahim, H., S. A. Rahman, E. E. Envulus, and S. O. Oyewole. (2009). Income and crop diversification among farming households in a rural area of north central Nigeria. Journal of Tropical Agriculture, Food and Environment Extension, 8 (2): 84–89. Jayne, T. (1994). Do High Food Marketing Costs Constrain Cash Crop Production? Evidence from Zimbabwe. Economic Development and Cultural Change, 42 (2): 387–402. Lukanu, G., Green, M., Greenfield, P., and Worth, S. (2004). Farmers’ cash crop cultivation decisions in Southern Niassa province, Mozambique. Development Southern Africa, 21: 3, 531–554. Meijerink, G. (2014). Farmers, traders and a commodity exchange: institutional change in Ethiopian sesame markets. Wageningen: Wageningen University. Mishra, A., and H. El-Osta (2002). Risk management through enterprise diversification: A farm level analysis. Paper presented at AAEA meetings in Long Beach, CA, July 28–31. Negatu, B., H. Kromhout, Y. Mekonnen, and R. Vermeulen. (2016). Use of chemical pesticides in Ethiopia: a cross-sectional comparative study on knowledge, attitude and practice of farmers and farm workers in three farming systems. Ann Occup Hyg, 60 (5): 551–566. Poudel, S., H. Basavaraja, L. Kunnal, S. Mahajanashetti, and A. Bhat. (2012). Crop diversification in Karnataka: An economic analysis. Dharwad, Karnataka: Department of Agricultural Economics, University of Agricultural Sciences. Poulton, C., R. Al-Hassan, G. Cadish, C. Reddy, and L. Smith. (2001). The cash crop vs. food crop debate. Crop post harvest program. Issue Paper 3. Sanchez, V. (2005). The Determinants of Rural Non-Farm Employment and Incomes in Bolivia. MSc thesis, Michigan State University. Sichoongwe, K., M. Laqrene, D. Ng’ng’ola, and G. Temb. (2014). The Determinants and Extent of Crop Diversification among Smallholder Farmers: A case study of Southern Province, Zambia. Washington DC: Malawi Strategy Support Program, Working Papers 05. SID-Consult-Support Integrated Development. (2010). Market Assessment and Value Chain Analysis in Benishangul Gumuz Regional State, Ethiopia Final Report. Sorsa, Debela Gelalcha (2009). Sesame trade arrangements, costs and risks in Ethiopia: A baseline survey. VC4PD Research Papers. Storck, H., Bezahih Emana, Berhanu Adnew, A. Borowiecki, and Shimelis Wolde Hawariat. (1991). Farming systems and farm management practices of smallholders in the Hararge highlands. Farming systems and resource Economics in the tropics. Wisssenschafts, Germany. The World Bank. (2015). 4th Economic Update: Overcoming Constraints in the Manufacturing Sector. Washington, DC: The World Bank. Tittonell, P., and K. E. Giller. (2013). When yield gaps are poverty traps: the paradigm of ecological intensification in African smallholder agriculture. Field Crop Res, 143: 76–90. UN (United Nations). (2015). The millennium development goals report 2015. New York: United Nations. Weiss, C. R., and W. Briglauer. (2000). Determinants and Dynamics of Farm Diversification. Working paper EWP 0002. Department of Food Economics and Consumption Studies, University of Keil.

40

G. M. Debela et al.

Wijnands, J. J. Biersteker, and R. Hiel. (2007). Oilseeds business opportunities in Ethiopia. The Netherlands: Ministry of Agriculture, Nature and Food Quality. Wondimagegn, M., B. Fufa, and J. Haji. (2011). Pattern, trend and determinants of crop diversification: empirical evidence from smallholders in eastern Ethiopia. Journal of Economics and Sustainable Development, 2 (8): 78–89.

Chapter 3

Spatial Integration of Livestock Markets in Ethiopia Gutu Gutema

Abstract Integration of livestock markets is important for stabilizing livestock prices, allocating resources, and alleviating market imperfections to improve market efficiency. It has been shown that market integration has a considerable effect on the successful design of agricultural price stabilization policies and in improving food security and the welfare of producers and consumers particularly in highly diverse and vulnerable countries like Ethiopia. Using a time varying threshold autoregressive model, this study contributes to literature by studying the spatial livestock market integration in Ethiopia. The results show that although there is a statistically significant improvement in market integration over the studied time period (2001– 2018), regional livestock markets in Ethiopia are not fully integrated with the central market (Addis Ababa). This affects (based on economic theory), improvements by producers, consumers, and traders’ welfare which can be improved by investments. However, the study also finds that the transaction costs for all livestock types reduced over the study period, but the deviation from the equilibrium was still far from short. Policy options that aim to reduce bottlenecks in livestock market integration like policies that lead to a reduction in transaction costs, should be implemented to improve integration in the livestock market. Keywords Market integration · Threshold autoregression model · Livestock · Ethiopia JEL Codes Q11 · Q13 · C22

G. Gutema (B) Department of Economics, Addis Ababa University, Addis Ababa, Ethiopia e-mail: [email protected] Jonkoping International Business School, JIBS, Jönköping, Sweden © Springer Nature Singapore Pte Ltd. 2020 J. Wood and O. Habimana (eds.), A Multidimensional Economic Assessment of Africa, Frontiers in African Business Research, https://doi.org/10.1007/978-981-15-4510-8_3

41

42

G. Gutema

3.1 Introduction Ethiopia is one of the fastest growing economies in the world. The economy has experienced a high and steady GDP growth, averaging between 8–11% annually for more than a decade. This growth can be largely attributed to the agricultural sector, while the government is also pushing the manufacturing and service sectors (The World Bank 2018). However, a closer look at the agricultural sector, particularly the livestock sub-sector in Ethiopia shows that there are structural problems in the sector, particularly in the market for livestock (Gebre-Selassie and Bekele 2012; Yonas 2006) as there are significant spatial variations in prices (Solomon and Authority 2003). The agriculture sector performs multiple functions in the Ethiopian economy as it provides food, raw material for industries, inputs for crop production, serves as a source of collateral, and employment. More than 70% of the country’s population is employed in the agriculture sector. This implies that agriculture and livestock are the main economic pillars of the rural economy and the overall economic growth in the country. Despite its significant economic contributions, the agriculture sector is almost entirely dominated by subsistence, small-scale, and resource deficient farmers using low technologies. Livestock is one of the least developed sub-sectors of the agriculture sector in Ethiopia. Crop production dominates and relative to the growth in real crop output which averaged 8.8% during 2004–2013, the growth in livestock output was slower, at around 6% (NBE 2014) during the same period. Moreover, crop and livestock accounted for an average 68 and 22% of real agricultural output. Even if the contribution of crop has increased over time, the contribution of livestock output has declined in almost every year of the studied period (2001–2018). Despite its poor performance, Ethiopia has a large livestock population. Specifically, the Ethiopian Central Statistical Agency reports that the country has the largest livestock population and the highest number of draft animals (like the donkeys, horses, and mules) in Africa (CSA 2018a), implying that the sub-sector has important possibilities for future agricultural growth because a large livestock population provides households with the possibility of improving their nutritional status, asset value, sources of employment, soil fertility, and agricultural traction while at the same time providing them with the possibility of consuming and selling by-products (milk, cheese, eggs, and butter) (Bettencourt et al. 2015; Rawlins et al. 2014). In addition to economic value, livestock plays a significant role in the social and cultural values of Ethiopian society. Even though the contribution of livestock to smallholder farmers’ livelihood is significant, the production system is not adequately market oriented, that is, the marketing aspect of the livestock and livestock by-products sub-sectors is relatively neglected and commercially weak. Therefore, there is a problem of market integration of the livestock market in the country. Recently due to economic growth with an increasing population the demand for livestock and livestock by-products like meat and dairy products increased. Supplying adequate and quality livestock and livestock by-products to meet this demand may make a positive contribution to the country’s economic growth. But by its nature

3 Spatial Integration of Livestock Markets in Ethiopia

43

livestock and livestock by-products are bulky or/and perishable and mostly their production and consumption are concentrated in different locations. Due to this and other reasons, the livestock and livestock by-products markets require a careful and dynamic analysis to facilitate market integration by removing the bottlenecks (Negassa et al. 2012). In addition, improved spatial market integration in livestock is more important in developing economies than in developed economies because of low developed market institutions and market inefficiencies in developing countries. The successful designing of agricultural price stabilization policies and proper functioning of the market are affected by the degree of market integration (Fackler and Goodwin 2001). Market integration makes a great contribution to the welfare of producers and consumers as well as for food security and growth in a diverse and highly vulnerable country like Ethiopia (Krishna 2004). Market integration is also expected to ensure more rapid and effective price adjustments between markets with the help of market reforms (Goletti and Babu 1994). On the contrary, if the market is not properly integrated or the prices are not properly transmitted, localized scarcities and surpluses can hurt both consumers and producers and lead to an increase in price volatility (Goletti et al. 1995). It can also constrain sustainable agricultural development and aggravate income inequalities. Therefore, significant attention needs to be paid to market integration in developing countries due to its important policy implications (Dercon 1995; Negassa and Jayne 1998; Van Campenhout 2007). Analyzing the market for agricultural products like grains and livestock and its integration has important policy implications for food security and poverty reduction. Most studies done on spatial price transmission in Africa have mainly focused on grain markets, for example, Dadi et al. (1992) on the maize market in western Ethiopia; Goletti and Babu (1994) on the maize market in Malawi; Negassa and Jayne (1998) on the grain market in Ethiopia; Abdulai (2000) on the maize market in Ghana; Alemu and Biacuana (2006) on the maize market in Mozambique; Van Campenhout (2007) on the maize market in Tanzania; Alemu and Van Schalkwyk (2008) on the maize market in Mozambique; Langyintuo (2010) on the grain market in Ghana; Tamru (2013) on the cereal market in Ethiopia; Kouyaté et al. (2016) on the rice market in western Africa; and Gitau and Meyer (2018) on the maize market in Kenya. However, very few studies have been done on livestock price transmission in Africa (Fafchamps and Gavian 1996 on the livestock market in Niger; and Bizimana et al. 2015 on the cattle market in Mali) with the result that little is known how livestock price transmissions take place between markets in Africa including in Ethiopia. In Ethiopia, as far as we are aware, there is no published research on livestock price integration till now. Focusing on livestock price transmission influences producers and other players in the livestock value chain since it will affect their marketing decisions. Moreover, some of the studies conducted so far on market integration in Africa suffer a methodological pitfall. They use methods like price series correlation, cointegration, and causality tests which do not take transaction costs into account. Therefore, the purpose of this chapter is studying the extent of spatial integration using a method that considers transaction costs in Ethiopian livestock markets. The rest of the chapter is organized as follows. Section 3.2 gives a brief literature review on the theory of market integration and the livestock market in Ethiopia.

44

G. Gutema

Section 3.3 describes the estimation techniques and data used for the study and Sect. 3.4 presents the empirical findings. Section 3.5 gives the conclusion and policy implications of the findings.

3.2 Livestock Marketing in Ethiopia—Background In Ethiopia, the number of sedentary farmers and pastoralists participating in livestock marketing is low. Large pastoralists and large farmers in particular participate little in livestock marketing because typically, livestock and livestock by-products are primarily consumed within the household. Sedentary farmers and pastoralists occasionally supply a few less productive animals to the market especially when they need cash for food, clothes, and medical and educational expenses (Gebremadiun et al. 2007). They are also forced to supply their livestock to the market unwillingly when there is a drought which puts the lives of their animals at risk by directly affecting the availability of feed. This shows that livestock by-production and supply is not a strategically important marketing option. Due to sporadic supply of animals to the market, traders are unable to rely on stable volumes of animals which in turn lowers their market power. In addition, less working capital is another constraint for small scale livestock traders and farmers. Providing credit-based systems following appropriate approaches and methodologies for both livestock traders and farmers are very important for promoting livestock development (Gebremarium et al. 2010). The provision of credit for livestock development like purchasing livestock, feed, and health services plays an important role in encouraging new investments in the sector. However, in Ethiopia the participation of commercial banks in providing credit is very limited compared to most other countries. They provide credit in a situation where the government provides a financial guarantee against loss of animals or low repayment rates. The high-interest rate on credit for livestock development does not encourage large investments in the livestock sector. Also, the focus of the available credit for livestock in Ethiopia is on short term activities, with short re-payment times such as fattening animals. By their nature, investments in the livestock sector are considered a high-risk activity. The risks involve a perception of nature which can affect farmers and traders’ production and marketing decisions both directly and indirectly. Natural disasters like droughts, disease outbreaks, and policy related risks too can result in production losses (Hazel 1971). In Ethiopian livestock markets, there is also a risk of losses because of gaps in market information and due to the fluctuating nature of demand and prices. Therefore, accessibility to insurance for livestock development which minimizes the risks against asset losses plays an important role in encouraging investments in new technologies and managing difficult problems like droughts and disease outbreaks. However, in Ethiopia the culture of participating in formal livestock insurance is not common and has received little attention. Most of the time, the response to

3 Spatial Integration of Livestock Markets in Ethiopia

45

the loss of livestock because of a natural disaster is through the provision of food aid and compensation for the loss of livestock by the government and NGOs. This shows that the insurance industry is underdeveloped in Ethiopia. Therefore, rather than always depending on aid and compensation for natural calamities, it would be better if livestock insurance is developed following the success stories of India, Thailand, Malaysia, Philippines, and Indonesia (Hanke and Barkmann 2017). In general, marketing processes for livestock and livestock by-products in Ethiopia are severely underdeveloped. The major problems are a traditional livestock production system which is not market-oriented, weak marketing systems, and poor infrastructure (like roads, marketplaces, livestock routes, quarantine, and abettors). Poor financial services (like credit and insurance), and the presence of illegal cross-border trade at a much lower price compared to markets in the country are some of the other major constraints (Hurissa and Eshetu 2002). The absence of a standardized unit of transaction in the livestock market is also another complicating limitation. Commonly, exchange bargaining is based on the weighing scale and visual assessment of the livestock’s body condition. When a trader collects the small ruminants based on the weighing scale from regional markets to transport to the central market, he is not sure of his profit margins. Due to this and other reasons the trader wants to avoid risks which would result in a loss for the producer. This discourages farmers to supply more animals to the market (Legese et al. 2008).

3.3 Literature Review 3.3.1 Market Integration In economics, the concept of market integration is at the center of a welfare analysis. It is an alternative approach for stabilizing prices, allocating resources, and correcting market imperfections. Literature identifies three different but inter-related types of market integration: spatial integration (concerned with the integration of geographically separated markets trading in homogenous products), vertical integration (concerned with the pass-through of a commodity price across stages of its marketing chain, that is, producer-wholesale-retail-consumer), and cross-commodity integration (related to integration between two commodities) (Sunga 2017). The focus of this study is spatial market integration. There is no consensus in literature on spatial market integration (Fackler and Goodwin 2001). Spatial market integration is the process by which price interdependence occurs (Faminow and Benson 1990) and is a situation in which prices of a commodity in spatially separated markets move together and price signals and information are transmitted smoothly across the markets (Fackler and Goodwin 2001; Goletti et al. 1995). Information on market integration provides specific evidence on the competitiveness of the market, the effectiveness of trade (Carter and Hamilton 1989), and the efficiency of pricing (Buccola 1983). Markets that are not integrated may convey

46

G. Gutema

inaccurate price information that might distort producers’ marketing decisions and contribute to inefficient product movements (Goodwin and Schroeder 1991). This distorted price signal may also lead to inefficient resource allocations which will end in limited sustainable development and aggravated income inequalities (Mushtaq et al. 2008). The market can be integrated via trade as well as through information or by a state trading institution that fixes the prices in the absence of trade (Fackler and Goodwin 2001). Markets play an important role in bringing together demand and supply across actors who are in different geographical locations. Following price signals regions with excess supply in well-integrated markets, can transfer the output to regions with excess demand (Cirera and Arndt 2008). Based on such a tradability view, trade flows are enough to reveal spatial market integration but do not necessarily imply price equalization (Barrett 2008). This argument supports the thinking that for the markets to be spatially integrated there must be physical trade between the two geographically separated markets besides their sharing similar long-run prices (Gonzalez-Rivera and Helfand 2001). According to this thinking, trade is taken as a key mechanism for market integration. In contrast, is the argument that supports trade being enough of a condition for market integration but not a necessary condition (Fackler and Goodwin 2001) because two markets can be well-integrated by information or by having a state trading institution that fixes prices in response to any price shocks. In such cases, price shocks can be transmitted even in the absence of trade. It is also believed that spatial market integration happens when there are no positive marginal profits for profit-making traders. This happens when the price differential for a homogeneous product equals the transportation costs of transferring the product between the markets which makes agents indifferent to trading and reaching a competitive equilibrium (Barrett 2001; Barrett and Li 2002; Miljkovic and Paul 2003). This equilibrium has the property that if trade takes place between any two regions, the price in the importing region equals the price in the exporting region plus the unit transaction cost1 incurred for moving between the two regions (Baulch 1997). If these conditions are satisfied then we say the markets are spatially integrated (Ravallion 1986). This study follows Fackler and Goodwin (2001), who argue that the market can be integrated with and without a physically flow of products. A study of spatial market integration helps estimate the speed of price adjustment and the level of market integration between central and peripheral markets which is important for policy recommendations and interventions (Dercon 1995). This chapter adopts a time-varying TAR model to address some of the limitations of the other models and to better capture the existence and degree of spatial integration

1 Transaction

costs often include expenditure on fuel, time, and effort to coordinate shipments and pick-up of transported commodities and synchronization between the buyers and sellers.

3 Spatial Integration of Livestock Markets in Ethiopia

47

of livestock markets in Ethiopia. However, this method of spatial market integration is not free of limitations.2 By construction, a time-varying threshold autoregressive model can capture nonlinearity in the price series due to non-constant transaction costs and reversals in the direction of trade flows. This helps estimate the average speed of price adjustments among the paired markets, half-life, and the threshold value without directly relying on transaction costs which we may not find.

3.4 Methodology 3.4.1 Threshold Autoregressive Model Consider two markets i and j located in different regions that trade a homogenous j commodity where the price of a given good at time t is pti and pt respectively. The price difference between market ‘i’ which is the market under investigation and market ‘j’ which is the price in a reference market at time ‘t’ is given as: ij

dt

j

= pti − pt

(3.1)

To estimate how the price difference in the previous periods between two markets ij ij is transmitted to the current period, that is, the effect of dt−1 on dt following the autoregressive model: ij

dt ij

ij

ij

= δdt−1 − εt

(3.2)

ij

where dt = dt − dt−1 , is the speed of adjustment measuring the rate at which the price difference in the previous period is corrected to achieve price equilibrium between the market under investigation and the reference market; and εt is error  term which is assumed to be normally distributed, εt ∼ N 0, σ 2 The speed of adjustment is the base for calculating the half-life. Deviation in a half-live which is , represents the time needed for one-half of a deviation calculated as h = 1n(0.5) 1n(δ) from the equilibrium to be eliminated. The autoregressive form of order one in Eq. 3.2 ignores the costs of the transaction. But transaction costs are expected to influence price adjustment. Therefore, to consider the existence of transaction costs the threshold autoregressive model is estimated as an extension of AR (1):

2 Since

it is essentially bivariate by its construction it does not consider a case for trade flow from one market to another market via a third market. In addition, the transaction costs in this method are estimated as a linear function of time.

48

G. Gutema

ij

dt

⎧ ij ij ⎪ ⎨ δout dt−1 + εt if dt−1 > θij ij ij = δin dt−1 + εt if −θ ≤ dt−1 ≤ θij ⎪ ⎩ δ dij + ε if dij < −θij out t−1 t t−1

(3.3)

where δin is the adjustment inside the band (that is, the speed of the adjustment parameter when the price margin is below the threshold); δout is the adjustment outside the band (speed of the adjustment parameter when the absolute value of the price margin is greater than the threshold); and θij is the threshold value which is assumed to be the cost of transfer between market ‘i’ and ‘j’. This threshold value categorizes the trade mechanism into three regimes (Baulch 1997): (i) the price ij difference between the two markets is above the transaction costs dt−1 > θij (ii) ij

the absolute value of the price difference is above transaction costs dt−1 < −θij and difference between the two markets is below transaction costs  (iii) the price ij |dt−1 | < θij Given these three categorizations, there is a price adjustment only in the first and second scenarios, that is, in the two outer regimes there is profitable trade that needs to be fully exploited by profit-making traders in moving commodities from market ‘i’ to ‘j’ or vice versa. But in the third scenario, it is unprofitable for a trader to engage in trade because the costs of transfer are more than the price difference between the market pairs. In this case the markets are assumed to be integrated and the trade is efficient (Van Campenhout 2007). Theory assumes that there is no adjustment inside the band which is formed by transaction costs (θij ), that is, δin = 0 Therefore, Eq. 3.3 can be re-written as:

ij

dt

⎧ ij ij ⎪ ⎨ δout dt−1 + εt if dt−1 > θij ij = εt if −θ ≤ dt−1 ≤ θij ⎪ ⎩ δ dij + ε if dij < −θij out t−1 t t−1

(3.4)

The standard threshold autoregressive model, Eq. 3.4 assumes that the transaction costs are constant over time. But transaction costs can vary depending on season, quality of roads, fuel prices, type of car, and other factors. Hence, considering the time trend in estimating the threshold value is very essential. There is high seasonality in animal source food consumption in Ethiopia due to religious beliefs. Animal sourced food consumption in Ethiopia is a religious practice. An estimated 43% of the population in Ethiopia is Orthodox Christians (CSA 2009) and their religion is characterized by important constraints on all animal sourced food during fasting periods. Throughout the fasting periods (56 days before Easter and 40 days before Christmas) no animal sourced food can be consumed. There are also other shorter fasting periods during the year. This results in significant reductions in the consumption of animal sourced food during the fasting periods (FVI-Idele 2016). This reduction in demand for animal sourced food items can have an impact on transaction costs. For these and other reasons, Van Compenhout (2007), extended the standard threshold autoregressive model which assumes that constant transaction costs include a time trend

3 Spatial Integration of Livestock Markets in Ethiopia

49

in both the adjustment parameter and the transaction costs in the model. Transaction costs are modelled as a simple function of time by θt = θ0 + (θT (T−) θ0 ) t. Therefore, Eq. 3.4 can be re-specified as:

ij

dt

⎧  ij ij ij ⎪ ⎨ δout dt−1 + δuot tdt−1 + εt if dt−1 > θij ,  ij ij ij ij ij = δin dt−1 + δin tdt−1 + εt if −θt ≤ dt−1 ≤ θt , ⎪ ⎩ δ dij + δ tdij + ε if dij < −θij , out t−1 t uot t−1 t−1

(3.5)



where δuot denotes the time varying speed of the price adjustment parameter towards ij the band; θt denotes the time varying threshold variable; and ‘t’ represents the time ij ij variable which runs from 0 to T. Based on Eq. 3.5, θij = θ0 at t = 0, and θij = θT at t = T. The threshold value is estimated through a grid search like the standard threshold autoregressive model. Similarly, in both the standard and time varying threshold autoregressive models’ adjustments in the inner band are assumed to be random, that is, there are no price adjustments within the band since transaction costs of moving commodities from market ‘i’ to ‘j’ or vice versa are greater than previous price differences between the two markets.

3.4.2 Data The price data used in this study was collected from CSA’s Retail Price Survey (CSA 2018b). This is monthly data collected in 119 urban retail markets in all regions of Ethiopia. Our analysis covers the period July 2001 to June 2018. The number of markets in each region is approximately proportional to the region’s share of the total urban population to ensure enough degree of national representativeness. The retail price data that we use was collected from major outlets in selected urban marketplaces by assigned enumerators (CSA 2018b). The retail prices of goods and services were mainly collected from traders and sometimes also from consumers at the time of purchase. For each item, a maximum of three price quotations were collected from three retailers on the same day. The enumerators collected the monthly retail prices of goods and services by interviewing the retailers and by measuring the weight of the items on the spot for about 15 days (starting from the 1st day through to the 15th day). The enumerators were well-trained and permanently assigned to every marketplace and lived there. They had a good chance of familiarizing themselves with the market and being on friendly terms with traders, owners of establishments, and households. Thus, it can be expected that the respondents were cooperative and provided reliable information. This helped us minimize exaggeration or underestimation of prices albeit an interview method of data collection was used rather than actual purchases and recording of prices (CSA 2018b).

50

G. Gutema

The livestock types covered in this retail price survey are cattle (heifer, cow, bull, and oxen), small ruminants (sheep and goats), and poultry (hens and cocks). We analyzed markets for small ruminants (sheep and goats), cattle (bulls and oxen), and cocks from poultry. Capital Addis Ababa was taken as the reference market in our analysis. The criteria for choosing the central market is very important and has its own policy implications. In some studies, the choice of a reference market is based on the net producing region (Alderman, 1991; Amikuzuno 2009; Shively 1996) or the net consumption market (Mensah Bonsu et al. 2011). Shock originating markets are used as reference markets since they influence other local or regional markets. Based on this we selected Addis Ababa’s market as a reference or central market by considering this market as a net consumer market which had the ability to influence other regional markets. In addition, the Addis Ababa market has location advantages which can influence other local markets since it is located in the center of the country (all markets). After identifying the reference markets, we selected the regional markets for investigation for each livestock type based on the following criteria. First, markets with few missing3 values were selected. Second, based on the test of price stationarity, the regional market in which prices were stationary4 at level were ignored. Finally, we considered the markets’ geographical location so that the figures represented national levels. This was done to include all corners of the country to have a national level understanding. The number of markets selected for each livestock item varied based on these criteria. Thus, based on these three criteria for sheep markets we selected 22 regional markets; for goats, we selected 25 regional markets; for oxen, we selected 18 regional markets; for bulls, we selected 16 regional markets; and for cocks we selected 15 regional markets which were paired with the Addis Ababa market. Finally, we used retail price data to estimate the extended TAR model with a time trend to analyze speed of adjustment and transaction costs with the reference market. To control for the effect of inflation on the estimation results, following Van Campenhout (2007), we used the price difference relative to the average price levels of the two paired markets.

3 Most of the districts have missing observations due to animals from a category not being presented

for sale or because the enumerators failed to collect animal prices for that type of livestock during that month. 4 The individual price series (regional market prices) are integrated of order one, I (1). But the difference between prices at two locations (the margin) is stationary, so we estimated a modified TAR model on the margin.

3 Spatial Integration of Livestock Markets in Ethiopia

51

3.5 Results 3.5.1 Descriptive Statistics Table 3.1 provides the average retail prices of all livestock types for all regions and their mean comparison with the national and Addis Ababa markets (which is considered a reference market in this study). This section summarizes the average retail prices for all livestock types at the national level and regional and Addis Ababa markets. Table 3.1 shows significant price variations by region. When we compare the average retail prices in the reference market (Addis Ababa) and the national market with other regional markets we observe significant variations from region to region. For example, the price of goats is approximately 50% lower in Amhara, Oromia, Somalia, Benishangul-Gumuz, and the Southern Nation, Nationality, and People (SNNP) markets as compared to Addis Ababa. This is because these regions have high levels of livestock production and are endowed with a large number of livestock. There are also similar variations for other livestock types. Further, the average prices in Addis Ababa for all livestock types are consistently above the national mean prices. This is probably because in the regional markets incidents of weather chocks, diseases, and religious compulsions (that is, the season) influence the supply of and demand for livestock and livestock by-products; but since Addis Ababa is a consumer city there is always demand and supply of livestock and livestock by-products. In addition, in Addis Ababa we find exporters and processors of livestock by-products, many hotels and restaurants, and better consumption demand patterns due to better incomes. This shows that Addis Ababa is a major consumption area for livestock and livestock by-products.

3.5.2 Econometric Results We estimated the time varying threshold autoregressive model for all sample livestock types as in Eq. 3.5. The estimation results and their interpretations for selected livestock types are given in the next sub-section.

3.6 Shoat Markets Tables 3.2 and 3.3 give the estimated results of the time-varying TAR models for the sheep and goat markets respectively. For a spatial market integration analysis, 22 and 25 regional markets were paired with the Addis Ababa market for sheep and goats respectively. The results of the estimation show that out of the 22 and 25 regional markets considered for sheep and goats, 45 and 56% of regional markets were significantly integrated with the reference market (Addis Ababa) at the beginning

piece

piece

Hen (indigenous)

Cock (indigenous)

56

44

3289

2124

2406

1667

385

426

Tigray

%

%

Oxen (4 years or more)

Hen (indigenous)

Cock (indigenous)

81

94

98

64

82

80

75

70

87

92

59

72

69

66

64

63

52

41

3061

1864

2082

1474

355

331

Afar

75

80

67

79

77

73

53

67

45

35

3471

2033

2318

1630

290

353

Am hara

Source Authors’ computation using CSA (2018b) retail price data

%

%

Bull (2 to 4 years)

%

%

Cow (4 years or more)

Goat (10 to 15 kg)

Heifer (2 to 4 yrs)

%

%

Sheep (10 to 15 kg)

Comparison with reference market Mean (%)

piece

piece

piece

Cow (4 years or more)

Oxen (4 years or more)

piece

Heifer (2 to 4 yrs)

Bull (2 to 4 years)

piece

piece

Goat (10 to 15 kg)

Unit

Sheep (10 to 15 kg)

Mean prices (Birr)

Livestock type

84

70

72

77

71

68

53

65

51

31

3730

1983

2146

1523

295

341

Oro miya

66

55

94

75

91

70

54

55

39

24

4859

1933

2721

1562

300

293

So mali

97

87

67

72

85

77

51

68

58

39

3473

1870

2562

1714

283

358

B. Gumuz

Table 3.1 Average retail price of livestock by region for the period July 2001 to June 2018

83

65

64

64

65

57

49

57

50

29

3323

1643

1958

1277

273

301

SNNP

109

86

63

62

75

66

69

86

66

38

3283

1606

2249

1469

380

457

Gam bella

64

49

111

124

114

88

88

113

38

22

5740

3196

3418

1964

485

596

Harari

100

100

100

100

100

100

100

100

60

44

5177

2581

3003

2233

552

528

Addis Ababa

71

55

104

94

80

91

64

64

43

25

5361

2430

2390

2031

354

337

Dire Dawa

84

76

79

82

82

75

65

74

51

34

4070

2115

2478

1686

359

393

Total

52 G. Gutema

3 Spatial Integration of Livestock Markets in Ethiopia

53

Table 3.2 Spatial integration of the Sheep markets in Ethiopia (July 2001 to June 2018) Market pair Addis Ababa-Maychew

Extended TAR model 1

2

ρ

ρ’(t)

Half-life

N

13.6

5.1

−0.0364

−0.0023***

18.6

186

10.3

185

4.3

189

2.1

190

14.7

197

3.3

190

17.0

191

7.9

191

22.8

185

2.9

197

4.20

181

8.2

193

15.6

193

3.5

191

8.2

195

2.2

184

8.8

187

(0.0570)

(0.0007)

Addis Ababa-Dubite

18.4

11.8

−0. 0649 (0.0543)

(0.0008)

Addis Ababa-Kobo

10.3

7.5

−0.1481**

−0.0017*

(0.0733)

(0.0009)

Addis Ababa-Moxa

10.6

7.7

−0.2728*** (0.0854)

Addis Ababa-Dembeca

11.0

3.2

−0.0459

8.1

7.8

−0.1846**

(0.0478) Addis Ababa-Amde wark (Umera) Addis Ababa-Kemise

(0.0791) 12.5

9.5

−0.0399 (0.0464)

Addis Ababa-Ghimbi

11.8

11.8

−0.0830*

15

−0. 0299

(0.0426) Addis Ababa-Dembi Dolo

9.8

(0.0557) Addis Ababa-Nekemte

12.2

4.5

−0.2118*** (0.0725)

Addis Ababa-Jimma

8.7

2.2

−0. 0163

17.8

14.5

−0.0803*

(0.0577) Addis Ababa-Shashamane

(0.0436) Addis Ababa-Asebe Teferi

15.6

7.3

−0.0432 (0.0577)

Addis Ababa-Bedessa

10.5

4.3

−0.1784***

Addis Ababa-Mambuk

6.0

8.7

−0.0807

(0.0608) (0.0638) Addis Ababa-Mender7

9.3

3.4

−0.2612*** (0.0925)

Addis Ababa-Shone

23.4

4.1

−0.0757** (0.0357)

−0.0023***

−0.0002 (0.0009) −0.0012* (0.0006) −0.0002 (0.0007) −0.0026*** (0.0009) −0.0007 (0.0006) −0.0026*** (0.0010) −0.0009 (0.0008) −0.0047*** (0.0008) −0.0006 (0.0005) −0.0022** (0.0008) −0.0020** (0.0008) −0.0071*** (0.0015) −0.0009 (0.0009) −0.0006 (0.0006)

(continued)

54

G. Gutema

Table 3.2 (continued) Market pair Addis Ababa-Alaba

Extended TAR model 1

2

ρ

ρ’(t)

Half-life

N

21.9

4.5

−0.0340

−0.0029***

20.0

159

7.5

188

13.5

193

26.1

185

11.5

176

(0.0455) Addis Ababa-Hagera Selam (Hulla)

18.1

12.5

−0.0872**

Addis Ababa-Mizan Teferi

9.9

6.6

−0.0500

(0.0375) (0.0345)

Addis Ababa-Bestecire (Dawro)

22.7

Addis Ababa-Melka Jabdu

11.6

20.0

−0.0261 (0.0564)

8.6

−0.0584 (0.0532)

(0.0009) −0.0002 (0.0004) −0.0012** (0.0005) −0.0027*** (0.0009) −0.0011* (0.0006)

Note Half-life expressed in months. All models are estimated without a constant. ρ, 1 and 2 denote the adjustment parameter on the lagged price difference, threshold value at the beginning of the sample period, and threshold value at the end of the sample period expressed as the percentage of mean price in the two markets respectively. Standard errors in the brackets. As starting values for the thresholds, at least 20% of the observations were either within or outside the band formed by the thresholds. *** indicates significance at the 1% level, ** indicates significance at the 5% level, and * indicates significance at the 10% level

of the sample period but at the end of the sample period the number of regional markets that were integrated with the reference market had increased. Accordingly, at the end of the sample period, the number of regional markets integrated with the reference market increased to 64 and 72% for sheep and goat markets respectively or an improvement of 18.1% for sheep and 16% for goat markets. These results confirm that there was an improvement in shoat markets’ integration with the reference market over time, but the markets were not fully integrated with the reference market, hence providing the traders an opportunity to engage in profitable trade. Table 3.2 shows that there is evidence of a statistically significant increase in the speed of price adjustments in 12 of the 22 market pairs for sheep and 11 of the 25 market pairs for goats. The average speed of price adjustments for these integrated markets is –0.158 and –0.153 at the beginning; and, –0.003 and –0.002 at the end of the sample period for sheep and goat markets respectively. The speed of price adjustments is approximately similar for both sheep and goat markets. This shows that the rate at which the price difference between the regional and central market dies out is almost similar for both sheep and goat markets due to comparable consumer preferences. Turning to transaction costs for the trade routes (see Tables 3.2 and 3.3) out of the 22 trade routes for sheep markets the transaction costs decreased for 19 trade routes, increased for two trade routes, and remained the same for one trade route. Similarly, for the goat markets, the transaction costs reduced for 19 trade routes, increased for one trade route, and remained the same for five trade routes. These results show that

3 Spatial Integration of Livestock Markets in Ethiopia

55

Table 3.3 Spatial integration of Goat markets in Ethiopia (July 2001 to June 2018) Market pair

Extended TAR model 1

2

ρ

ρ’(t)

Half-life

N

Addis Ababa-Endebaguna (WestTigray)

12.3

10.4

−0.1766*

−0.0028**

3.5

194

(0.0905)

(0.0011)

Addis Ababa-Awash 7 kile

12.3

10.7

2.6

196

Addis Ababa-Dessie

18.3

6.0

10.8

195

3.5

192

12

196

10.9

192

5

194

5.4

196

4.8

196

5.2

184

13.7

182

5.2

184

18.4

182

6

182

7.9

181

4.6

180

24.9

188

12

186

−0.2264*** (0.1278) −0.0620 (0.0682)

Addis Ababa-Debremarkos

18.4

12.5

−0.1790** (0.0765)

Addis Ababa-Chagni

13.8

13.8

−0.0528

Addis Ababa-Kemise

16.0

12.3

−0.0612

(0.1278) (0.0562) Addis Ababa-Nekemte

20.0

7.6

−0.1217*

−0.0005 (0.0009) −0.0035** (0.0015) −0.0016** (0.0007) −0.0005 −0.0015*

14.8

−0.1187*

Addis Ababa-Bambassi

23.1

18.0

−0.1325**

(0.0715) (0.0607) 7.8

(0.0008)

(0.0007)

18.6

21.7

(0.0015) −0.0017**

(0.140) Addis Ababa-Asalla

Addis Ababa-Dubite

−0.0017*

−0.1242** (0.0502)

(0.0008) −0.0034*** (0.0012) −0.0009 (0.0007)

Addis Ababa-Debretabor

18.0

18.0

−0.0490 (0.1274)

(0.0008)

Addis Ababa-Moxa

21.1

21.1

−0.1247*

−0.0022**

(0.0737)

(0.0011)

Addis Ababa-Asebe Teferi

21.6

2.1

−0.0369 (0.0355)

Addis Ababa-Bedessa

18.9

18.9

−0.1038

Addis Ababa-Baroda (East Hararghe)

13.1

13.2

−0.0836

Addis Ababa-Negelle

25.0

(0.0676) (0.0649) 23.4

−0.1375*** (0.0497)

Addis Ababa-Erer

20.4

8.6

−0.0273

Addis Ababa-Hartishki (Kebri Beyah)

20.7

19.4

−0.0524

(0.0358) (0.0513)

−0.0034***

−0.0009* (0.0005) −0.0035** (0.0016) −0.0018** (0.0009) −0.0005 (0.0010) −0.0009** (0.0004) −0.0028** (0.0012)

(continued)

56

G. Gutema

Table 3.3 (continued) Market pair

Extended TAR model 1

Addis Ababa-Mambuk

14.1

2 9.1

ρ

ρ’(t)

Half-life

N

−0.1228*

−0.0016*

5.2

188

(0.0704)

(0.0009) 3.7

188

7.0

182

7.3

184

3.1

183

11.3

182

2.7

180

Addis Ababa-Asossa

17.2

16.6

−0.1669***

Addis Ababa-Kemeshi

23.1

13.2

−0.0939

(0.0512)

Addis Ababa-Hosaena

15.7

15.7

(0.0003)

−0.0901**

−0.0018**

(0.0435)

(0.0009)

18.2

14.1

−0.2000***

Addis Ababa-Benchi

27.9

15.3

−0.0590

(0.0594) (0.0375) 14.2

4.5

(0.0007) −0.0002***

(0.0341)

Addis Ababa-Awassa

Addis Ababa-Kefira Gebiya

−0.0003

−0.2199*** (0.0651)

−0.0004 (0.0011) −0.0015** (0.0005) −0.00008 (0.0007)

Note Half-life expressed in months. All models are estimated without a constant. ρ, 1 , and 2 denote the adjustment parameter on the lagged price difference, threshold value at the beginning of the sample period, and threshold value at the end of the sample period expressed as the percentage of mean price in the two markets respectively. Standard errors in brackets. As starting values for the thresholds, at least 20% of the observations were either within or outside the band formed by the thresholds. *** indicates significance at the 1% level, ** indicates significance at the 5% level, and * indicates significance at the 10% level

there were probably infrastructural improvements in the regions over the study period, but they were unevenly distributed. An increase in transaction costs for some trade routes might be due to poor infrastructure like roads and information technologies in some parts of the region. The results in Tables 3.2 and 3.3 also show that on average the transaction costs between the market pairs were at least 13.35% for sheep and 18.55% for goat markets above the inter-market price margins at the beginning of the period, whereas, they were 8.21% for sheep and 13.08% for goat markets above the inter-market price margins at the end of the study period. This shows that on average the estimated transaction costs, expressed as a percentage of mean prices reduced for both sheep and goat markets over the sample period by 38.5 and 29.5% respectively. A look at the estimated half-life price adjustments shows that the time required to eliminate 50% of the positive or negative deviations from the long run equilibrium provided unexploited trade opportunities that persisted a little longer for goat markets than for sheep markets. The half-life of the price adjustments for the sheep and goat markets is on average 10.6 months and 7.8 months respectively. This result shows that shoat market prices require a longer period to adjust to the long run equilibrium prices than cattle and poultry markets (see Table 3.A1 in the Appendix).

3 Spatial Integration of Livestock Markets in Ethiopia

57

3.7 Cattle Markets Tables 3.4 and 3.5 present the estimated results of the time-varying TAR model for oxen and bull markets respectively. In this section, we also compare regional market integration of oxen and bull markets with the central market. Eighteen and 16 regional markets were paired with the central market of which 50% oxen and 56% bull markets were significantly integrated at the beginning of the sample period. But at the end of the sample period, the level of market integration improved to 72 and 62% for oxen and bull markets respectively. Thus, this result shows that the level of integration increased for oxen and bull markets over the sample period. In both oxen and bull markets, there was also a statistically significant increase in the speed of price adjustments. The average price adjustment for the integrated oxen markets was −0.083 at the beginning and −0.003 at the end of the sample period with an average transaction cost of 8.76 and 6.84% of the mean paired market price for the respective periods. This shows that the transaction costs for oxen trade routes on average reduced by 38.5% in the study period whereas, the average price adjustment for the significantly integrated bull markets was −0.241 at the beginning and −0.004 at the end of the study period. The average transaction costs for bull markets was 9.91% at the beginning and 7.42% at the end of the study period. The average transaction costs for integrated bull markets also declined by 25% over the sample period (2001−2018). Hence, there is significant evidence of a reduction in average transaction costs for the cattle market over the study period. The reduction in transaction costs might also be a possible effect of improvements in infrastructure over the last few decades in the country. Tables 3.4 and 3.5 also report that the average time required to eliminate half of the deviations from the long run equilibrium for those significantly integrated oxen and bull markets is about 5.8 months and 4.6 months respectively. This implies that even though there is evidence of price adjustments compared to the long run equilibrium there is no immediate market efficient adjustment of the prices.

3.8 Poultry Markets Like the shoat and cattle sectors, the economic contributions of the poultry sector are also important. Poultry makes significant contributions to poverty alleviation by supplying a surplus for sale to generate some cash and providing good nutrition to the low-income group. Marketing channels for poultry are either farmers selling to consumers directly or selling to retail traders who take them the large urban markets. To study the extent of market integration in the cock markets over the study period, Table 3.6 gives the estimated results of the time-varying TAR model. Looking at the results of the market integration given in Table 3.6, out of the 15 regional cock markets paired with Addis Ababa about 66% at the beginning and 93% at the end of the study period were significantly integrated.

58

G. Gutema

Table 3.4 Spatial integration of Oxen markets in Ethiopia (July 2001 to June 2018) Market pair

Extended TAR model 1

Addis Ababa-Melka Werer

10.6

2 5.0

P

ρ’(t)

Half-life

N

−0.1825***

−0.0010

3.4

181

6.2

89

2.7

195

9.0

189

2.2

192

12.2

191

2.4

191

3.1

197

5.1

197

6.5

197

2.7

195

15

195

2.4

197

5.6

197

8.4

195

5.3

189

6.3

185

(0.0632) Addis Ababa-Hager Mariam

8.2

8.8

−0.1047

Addis Ababa-Teppi

9.6

8.6

−0.2244

(0.0739) (0.1492)

Addis Ababa-Jimma

9.3

4.5

−0.0734 (0.0769)

Addis Ababa-Jijiga

3.5

6.5

−0.2662

Addis Ababa-Mekele

9.2

8.8

−0.0549

(0.1661) (0.0863) Addis Ababa-Endebaguna (West Tigray)

11.0

8.0

−0.2429** (0.0958)

Addis Ababa-Kemisie

5.5

7.9

−0.1960***

Addis Ababa-Bati

7.1

7.1

−0.1253

(0.0730) (0.0964) Addis Ababa-Asalla

7.7

10.1

−0.0999 −0.2234**

Addis Ababa-Asossa

9.0

7.4

−0.0425

(0.0908) (0.0593) −0.2426*** (0.0590) Addis Ababa-Shoa Benchi

11.9

4.5

−0.1147***

8.0

8.0

−0.0786

(0.0419) Addis Ababa-Bahirdar Addis Ababa-Gonder Addis Ababa-Alaba

8.0 14.6

2.8 7.6

(0.0032) −0.0029** (0.0011) −0.0036*** (0.0010) −0.0045** (0.0018) −0.0007 (0.0016) −0.0024** (0.0011) −0.0054*** (0.0020) −0.0039** −0.0040**

7.0

2.6

(0.0022) −0.0068***

(0.0015)

5.3

13.2

−0.0083***

(0.0769) Addis Ababa-Asaba Teferi

Addis Ababa-Yirga cafe

(0.0008)

(0.0017) −0.0034*** (0.0012) −0.0011 (0.0012) −0.0002 (0.0005) −0.0056**

(0.1166)

(0.0027)

−0.1221*

−0.0020*

(0.0627)

(0.0011)

−0.1033* (0.0060)

−0.0013 (0.0010)

(continued)

3 Spatial Integration of Livestock Markets in Ethiopia

59

Table 3.4 (continued) Market pair

Extended TAR model 1

Addis Ababa-Shoa Robit

6.6

2 6.6

P

ρ’(t)

Half-life

N

−0.1465**

−0.0026**

4.3

184

(0.0730)

(0.0011)

Note Half-life expressed in months. All models are estimated without a constant. ρ, 1 , and 2 denote the adjustment parameter on the lagged price difference, threshold value at the beginning of the sample period, and threshold value at the end of the sample period expressed as the percentage of mean price in the two markets respectively. Standard errors in brackets. As starting values for the thresholds, at least 20% of the observations were either within or outside the band formed by the thresholds. *** indicates significance at the 1% level, ** indicates significance at the 5% level, and * indicates significance at the 10% level

The speed of price adjustments for cock markets significantly increased for 33% of the trade routes over the study period. The average price adjustment for the significantly integrated cock markets with the Addis Ababa market was −0.27 at the beginning and −0.003 at the end of the study period. Over the study period, transaction costs for markets integrated with the Addis Ababa market reduced by more than 50% on average. A look at the estimated average half-life of price adjustments for the integrated cock markets shows that prices need about 5.4 months to correct half of the deviations from the long run price equilibrium.

3.9 Conclusions and Recommendations This study tested the spatial livestock markets’ integration in Ethiopia using the extended threshold autoregressive model (TAR). The monthly livestock retail price data used in the study was obtained from the Central Statistical Agency and spanned the period July 2001 to June 2018 for all regional markets in Ethiopia. The chapter also analyzed the change and degree of spatial regional livestock markets’ integration with the central market (Addis Ababa). Overall, the results from the time-varying TAR model show that the level of livestock market integration over the study period significantly increased except for a few market pairs. For example, out of the 22 regional markets considered for sheep, 45% at the beginning and 63% at the end of the sample period were statistically integrated with the Addis Ababa market. Similarly, out of the 25 selected regional goat markets 56% at the beginning and 72% at the end were statistically integrated with the Addis Ababa market. Turning to the oxen and bull markets, about 42 and 56% of these markets at the beginning and 73 and 62% at the end of the study period were also statistically significantly integrated to the central market out of 18 and 16 regional markets respectively. Finally, out of the 15 regional cock markets paired with Addis Ababa about 66% at the beginning and 93% at the end of the study period were significantly integrated. The results show that it was more common for markets

60

G. Gutema

Table 3.5 Spatial integration of Bull markets in Ethiopia (July 2001 to June 2018) Market pair Addis Ababa-Melka Warari

Extended TAR model 1

2

P

ρ’(t)

8.7

3.1

−0.1268*

−0.0035***

(0.0610)

(0.00123)

Addis Ababa-Dessie

8.5

8.0

−0.175*

Addis Ababa-Aksum

7.5

9.9

−0.1470

(0.0987) (0.0991) Addis Ababa-Nekemte

13.1

6.2

−0.2276*** (0.0597)

Addis Ababa-Bedelle

8.9

7.1

−0.4192***

Addis Ababa-Jimma

10.8

10.3

−0.3229*

(0.0902) (0.1863) Addis Ababa-Hartishek

4.3

9.3

−0.0489 (0.0574)

Addis Ababa-Shone

17.0

8.9

−0.1381**

Addis Ababa-Teppi

11

11

−0.1197**

(0.0557) (0.0512) Addis Ababa-Bonga

10.6

11.2

−0.0734 (0.0497)

Addis Ababa-Awassa

7.3

4.3

−0.2079***

Addis Ababa-Jijiga

8.2

8.2

−0.4334***

(0.0719) (0.1179) 10.3

−0.0654

Addis Ababa-Hagere Mariam

9.7

Addis Ababa-Negelle

5.4

6.3

−0.2135

Addis Ababa-Bahirdar

8.7

8.7

−0.3165***

(0.0683) (0.0669) (0.0950)

Addis Ababa-Fiche

14.3

6.6

−0.0891 (0.0744)

−0.00446**

Half-life

N

5.1

175

3.6

157

4.3

134

2.6

175

1.2

176

1.7

173

13.8

181

4.6

179

5.4

179

9.0

156

2.9

119

1.2

184

10.2

169

2.8

174

1.8

175

7.4

177

(0.00183) −0.0053*** (0.0015) −0.0002 (0.0006) 0.0009 (0.00107) −0.0016 (0.0017) −0.00343*** (0.000998) −0.0026** (0.0012) −0.0033*** (0.0010) −0.0053*** (0.0012) −0.00037 (0.00076) −0.0011 (0.0015) −0.0043*** (0.0011) −0.0017*** (0.0008) −0.0037** (0.0016) −0.0039** (0.0017)

Note Half-life expressed in months. All models are estimated without a constant. ρ, 1 , and 2 denote the adjustment parameter on the lagged price difference, threshold value at the beginning of the sample period, and threshold value at the end of the sample period expressed as the percentage of mean price in the two markets respectively. Standard errors in brackets. As starting values for the thresholds, at least 20% of the observations were either within or outside the band formed by the thresholds. *** indicates significance at the 1% level, ** indicates significance at the 5% level, and * indicates significance at the 10% level

3 Spatial Integration of Livestock Markets in Ethiopia

61

Table 3.6 Spatial integration of Cock markets in Ethiopia (July 2001 to June 2018) Market Pair

Extended TAR model 1

2

ρ

ρ’(t)

Addis Ababa-Endabaguna (West Tigray)

12.5

5.5

−0.221**

−0.00282*

(0.0934)

(0.00151)

Addis Ababa-Dessie

14.8

4.8

−0.165**

−0.00218*

(0.0733)

(0.00122)

Addis Ababa-Bati

18.5

7.3

−0.242***

−0.000382

(0.0798)

(0.00120)

Addis Ababa-Woldiya

28.3

25.6

−0.0551 (0.0547)

Addis Ababa-Kobo

28.9

18.2

−0.0590

Addis Ababa-Karat

22.5

4.7

−0.0707

(0.0633)

Addis Ababa-Chagini

15.8

3.5

−0.00340***

(0.0674)

(0.00126) −0.00202*

−0.172**

Addis Ababa-Wolkite

13.7

3.4

−0.136*

(0.0740) (0.0706) 9.2

−0.187 (0.168)

−0.00508***

−0.100* (0.0581)

(0.000971)

Addis Ababa-Fishe

8.2

3.5

−0.313***

−0.00371**

(0.0961)

(0.00155)

−0.00361***

−0.523***

0.00173**

(0.0871)

(0.000748)

Addis Ababa-Agaro

8

8

−0.0692

Addis Ababa-Shashamane

14.8

5.8

−0.155*

(0.0570) (0.0924)

11.4 9.5 3.3 3.7 4.7 3.3

(0.00155)

11.3

6.3

12.2

(0.00130)

20.4

14.03

2.5

(0.00119) −0.00364***

Addis Ababa-Negelle

Addis Ababa-Hagere Mariam

3.8

(0.00120) −0.00250***

−0.189***

4.6

5.5

−0.00361***

(0.000897)

24.8

2.8

(0.00121)

(0.0503)

Addis Ababa-Gidole

Addis Ababa-Metu

−0.00423***

Half-life

−0.00345***

6.6 1.8 0.9 9.6

(0.00102) −0.00238**

4.1

(0.00120)

Note Half-life expressed in months. All models are estimated without a constant. ρ, 1 , and 2 denote the adjustment parameter on the lagged price difference, threshold value at the beginning of the sample period, and threshold value at the end of the sample period expressed as the percentage of mean price in the two markets respectively. Standard errors in brackets. As starting values for the thresholds, at least 20% of the observations were either within or outside the band formed by the thresholds. *** indicates significance at the 1% level, ** indicates significance at the 5% level, and * indicates significance at the 10% level

62

G. Gutema

to be integrated at the end than at the beginning of the study period. This might be due to large investments in infrastructural improvements by the government, which may have helped smoothen trade among the market pairs. The study also showed that the speed of price adjustments or the rate at which the price difference disappeared for several trade routes improved for the livestock types over the study period. Integrated market pairs, on average, took a minimum of five and a maximum of 10 months to eliminate half the price differences. According to existing studies the estimated time required to reduce half of the deviations from the long run price equilibrium is longer for livestock than for cereal markets in Ethiopia. For example, the total average half-life required is 3.75 months for the major cereal markets (see Tamru 2013), and 6.89 months for livestock. On average, transaction costs (that is, the threshold) declined over the study period for all livestock types. But the reductions in transaction costs for all market pairs were not equal. This might be due to the unevenly distributed nature of infrastructure in Ethiopia such as road networks, telecommunications, and transport services. Under normal circumstances, if markets in the same country are segmented with limited price transmission or integration, many economic policies become less effective (Cirera and Arndt 2008). This implies that economic policies without price transmission or integration do not work. Therefore, government officials can use these research findings and further investigate why some markets are not integrated; and why the time that is required to reduce or eliminate price differences is so long even for integrated markets. For instance, poor roads result in high transaction costs due to faster depreciation of vehicles, higher maintenance costs, higher fuel consumption, high tirereplacement costs, and loss of time due to low speed. These affect not only the livestock sector but also the economy in general (Shi et al. 2017). The quality of roads and the means of communication matter in marketing livestock and livestock by-products. Therefore, investments in infrastructure is one among of the important policy actions that the government needs to consider. In general, the results of our study show that all regional livestock markets are not fully integrated to the center and long periods are required to eliminate the price differences even for integrated markets. To achieve the objective of integration there is also a need for more investments by the government for promoting information sharing and developing communication networks. Lastly, more research in this area must be conducted using an alternative approach like a multivariate analysis in Ethiopia using level II and level II studies.

Appendix See Table 3.A1.

10

45.5

– 0.158

Number of regional markets integrated with the center

Percent integrated

Average speed of Adjustment for the Integrated market

13.35

Source Authors’ computations (2019)

10.6

8.21

86.4

Reduced (%)

Threshold (% of price) for the Integrated market

4.5

Half-Life (months)

9.1

Same (%)

– 0.003

63.6

14

22

Increased (%)

Threshold values over the study period

22

Total Pairs

18.55

7.87

– 0.153

56.0

14

25

Goat July 2001

July 2001

June 2018

Sheep

13.08

76

20

4

– 0.002

72.0

18

25

June 2018

8.34

5.41

– 0.083

50

9

18

July 2001

Oxen

6.41

55.56

16.67

27.78

– 0.003

72.2

13

18

June 2018

9.91

4.68

– 0.241

56.3

9

16

July 2001

Bull

Table 3.A1 Summary of price adjustment, Threshold and Half-life for the all livestock types (July 2001 to June 2018)

7.42

50.00

18.75

31.25

– 0.004

62.5

10

16

June 2018

Cock

16.45

5.45

– 0.27

66.7

10

15

July 2001

8.27

86.67

6.67

6.67

– 0.003

93.3

14

15

June 2018

3 Spatial Integration of Livestock Markets in Ethiopia 63

64

G. Gutema

References Abdulai, A. (2000). Spatial price transmission and asymmetry in the Ghanaian maize market. Journal of Development Economics, 63:327−349. Alemu, Z. G. and H. D. Van Schalkwyk. (2008). Market integration in Mozambican maize markets. African Books Collective. Organization for Social Science and Research in Eastern and Southern Africa. Alemu, Z. G. and G. R. Biacuana. (2006).‘Measuring market integration in Mozambican maize markets: A threshold vector error correction approach, In Contributed paper presented at the International Association of Agricultural Economists Conference, Gold Cost, Austria. Amikuzuno, J. (2009). Spatial price transmission and market integration in agricultural markets after liberalization in Ghana: Evidence from fresh tomato markets, Doctoral dissertation. Barrett, C. B. (2001). Measuring integration and efficiency in international agricultural markets. Review of Agricultural Economics, 23(1): 19−32. Barrett, C. B. and J. R. Li. (2002). Distinguishing between equilibrium and integration in spatial price analysis. American Journal of Agricultural Economics, 84(2): 292−307. Barrett, C. B. (2008). ‘Spatial Market Integration’, In S.N. Durlauf and L.E. Blume (eds). The New Palgrave Dictionary of Economics (vol. 7). London: Palgrave Macmillan, pp. 752−754. Bettencourt, E. M. V., M. Tilman, V. Narciso, M. L. D. S. Carvalho and P. D. D. S. Henriques. (2015). The livestock roles in the wellbeing of rural communities of Timor-Leste. Revista de Economia e Sociologia Rural, 53: 63−80. Baulch, B. (1997). Transfer costs, spatial arbitrage and testing for food market integration. American Journal of Agriculture economics, 79: 477−487. Buccola, S. T. (1983). Risk preferences and short-run pricing efficiency. American Journal of Agricultural Economics, 65 (3): 587−591. Carter, C. A. and N. A. Hamilton. (1989). Wheat inputs and the law of one price. Agribusiness, 5(5): 489−496. Cirera, X. and C. Arndt. (2008). Measuring the impact of road rehabilitation on spatial market efficiency in maize markets in Mozambique. Agricultural Economics, 39(1): 17−28. CSA (Central Statistical Agency). (2009). Population Census of Ethiopia 2007. Addis Ababa: Federal Democratic Republic of Ethiopia. CSA (Central Statistical Agency). (2018a). Agricultural Sample Surveys, Report on Livestock And Livestock Characteristics. Addis Ababa: Statistical Bulletin. CSA (Central Statistical Agency). (2018b). Retail Price Survey. Addis Ababa: Federal Democratic Republic of Ethiopia. Dadi, L., A. Negassa and S. Franzel. (1992). Marketing maize and tef in western Ethiopia: Implications for policies following market liberalization. Food Policy, 17 (3): 201−213. Dercon, S. (1995). On market integration and liberalisation: method and application to Ethiopia. The Journal of Development Studies, 32(1): 112−143. Fackler, P. L. and B. K. Goodwin. (2001). Spatial price analysis. Handbook of agricultural economics, 1: 971−1024. Fafchamps, M. and S. Gavian. (1996). The spatial integration of livestock markets in Niger. Journal of African Economies, 5(3): 366−405. Faminow, M. D. and B. L. Benson (1990). Integration of spatial markets. American Journal of Agricultural Economics, 72(1): 49−62. FVI-Idele (France Vétérinaire Internationale—Institute de l’Elevage). (2016). Complementary feasibility study for the relocation and modernization of Addis Ababa Abattoirs. Study prepared for Addis Ababa Abattoirs Enterprise (AAAE). Paris, France. Gebremariam, S., S. Amare, D. Baker and A. Solomon. (2010). Diagnostic study of live cattle and beef production and marketing: Constraints and opportunities for enhancing the system. Gebremedhin, B., D. Hoekstra and S. Jemaneh. (2007). Heading towards commercialization? The case of live animal marketing in Ethiopia. IPMS Working Paper 5. Nairobi (Kenya): ILRI.

3 Spatial Integration of Livestock Markets in Ethiopia

65

Gebre-Selassie, A. and T. Bekele. (2012). A review of Ethiopian agriculture: Roles, policy and small-scale farming systems. C. Bell & J. Prammer (Researchers), C. Eder, D. Kyd-Rebenburg, & J. Prammer (Eds.), Global growing casebook: Insights into African agriculture, 36−65. Gitau, R. and F. Meyer. (2018). Spatial market integration in the era of high food prices. A case of surplus and deficit maize markets in Kenya. Agrekon, 57(3−4): 251–265. Goletti, F. and S. Babu. (1994). Market liberalization and integration of maize markets in Malawi. Agricultural Economics, 11: 311–324. Goletti, F., R. Ahmed and N. Farid. (1995). Structural determinants of market integration: The case of rice markets in Bangladesh. The Developing Economies, 33(2): 196–198. Gonzalez-Rivera, G. and S. M. Helfand. (2001). The extent, pattern, and degree of market integration: A multivariate approach for the Brazilian rice market. American Journal of Agricultural Economics, 83(3): 576–592. Goodwin, B. K. and T. C. Schroeder. (1991). Cointegration tests and spatial price linkages in regional cattle markets. American Journal of Agricultural Economics, 73 (2): 452–464. Hänke, H. and J. Barkmann. (2017). Insurance function of livestock, Farmers coping capacity with crop failure in southwestern Madagascar. World Development, 96: 264–275. Hazel, C. R. (1971). Sensitivity of striped bass and stickleback to ammonia in relation to temperature and salinity. Calif. Fish Game, 57, 154–161. Hurissa, B. and G. Eshetu. (2002). Challenges and opportunities of livestock trade in Ethiopia. Paper presented at the 10th Annual Conference of Ethiopian Society of Animal Production (ESAP), Addis Ababa, Ethiopia, August 22–24. Bizimana, J. C., J. P. Angerer, D. A. Bessler and F. Keita. (2015). Cattle markets integration and price discovery: the case of Mali. The Journal of Development Studies, 51(3): 319–334. Kouyaté, C., S. von Cramon-Taubadel and I. Fofana. (2016). Proximity and price co-movement in West African rice markets. African Journal of Agricultural and Resource Economics, 11: 167–182. Krishna, A. (2004). Escaping poverty and becoming poor: who gains, who loses, and why? World development, 32(1), 121–136. Langyintuo, A. S. (2010). Grain distribution in Ghana under imperfectly competitive market conditions, In AAAE Third Conference/AEASA 48th Conference, Cape Town, South Africa. Legese, G., H. Teklewold, D. Alemu and A. Negassa. (2008). Live animal and meat export value chains for selected areas in Ethiopia: Constraints and opportunities for enhancing meat exports. ILRI Improving Market Opportunities Discussion Paper 12. Nairobi, Kenya: ILRI. Mensah-Bonsu, A., A. Agyeiwaa-Afrane and J. K. Kuwornu. (2011). Efficiency of the plantain marketing system in Ghana: A co-integration analysis. Journal of Development and Agricultural Economics, 3(12): 593–601. Miljkovic, D. and R. Paul. (2003). Agricultural trade in North America: Trade creation, regionalism and regionalisation. Australian journal of agricultural and resource economics, 47(3): 349–366. Mushtaq, K., A. Gafoor and M. Dad. (2008). Apple Market Integration: Implications for Sustainable Agricultural Development. The Lahore Journal of Economics, 13 (1): 129–138. NBE (National Bank of Ethiopia). (2014). Annual Report. Addis Ababa. Negassa, A. and T. S. Jayne. (1998). Vertical and Spatial Integration of Grain Markets in Ethiopia: Implications for Grain Markets (No. 55598). Michigan State University, Department of Agricultural, Food, and Resource Economics. Negassa, A., S. Rashid, B. Gebremedhin and A. Kennedy. (2012). Livestock production and marketing. Food and agriculture in Ethiopia. Progress and policy challenges: 159–190. Ravallion, M. (1986). Testing market integration. American Journal of Agricultural Economics, 68 (1): 102–109. Rawlins, R., S. Pimkina, C. B. Barrett, S. Pedersen and B. Wydick (2014). Got milk? The impact of Heifer International’s livestock donation programs in Rwanda on nutritional outcomes. Food Policy, 44: 202–213.

66

G. Gutema

Sunga, C. (2017). An analysis of spatial market integration: a case of Zambian dry bean markets connected by informal trade to Tanzania and the Democratic Republic of Congo, Doctoral dissertation, Stellenbosch: Stellenbosch University. Tamru, S. (2013). Spatial integration of cereal markets in Ethiopia. Ethiopia Strategy Support Program-Ethiopian Development Research Institute. Shively, G. E. (1996). Food price variability and economic reform: An ARCH approach for Ghana. American Journal of Agricultural Economics, 78(1): 126–136. Shi, Y., S. Guo and P. Sun. (2017). The role of infrastructure in China’s regional economic growth. Journal of Asian Economics, 49, 26–41. Solomon, A. and E. L. M. Authority. (2003). Livestock marketing in Ethiopia: a review of structure, performance, and development initiatives (Vol. 52). ILRI (aka ILCA and ILRAD). The World Bank. (2018). Retrieved from: https://www.worldbank.org/en/country/ethiopia/ overview. Van Campenhout, B. (2007). Modeling trends in the food market integration: Method and an application to Tanzanian maize markets. Food policy, 32 (1): 112–127. Yonas, K. (2006). Post-1991 Agricultural Polices: The Role of National Extension Program in Addressing the Problem of Food Security. Ethiopia: Politics, Policy Making and Rural Development, Department of PSIR, AAU, Addis Ababa.

Chapter 4

New Urban Consumption Patterns and Local Agriculture: Application to the Bukavu HORECA Sector (DRC) Angélique Ciza Neema and Lebailly Philippe

Abstract The rapid urbanization experienced by all developing countries is recognized as a key factor in the current evolution of food consumption. In Bukavu, consumption patterns are evolving and diversifying into out-of-home catering in a new restauration mode. The purpose of this study is analyzing whether this new consumption pattern can constitute an opportunity for local agricultural products. The study surveyed 45 sampled restaurants in Bukavu city. A survey questionnaire based on food types, offers for consumers, frequency of consumption, and income was submitted to the restaurateurs. Two main types of restaurants were identified, of which 71% were ‘Malewa’ and 29% were modern restaurants. In all these restaurants both local and imported products were generally used. The survey results show that in Malewa restaurants type 58% of the income from sales was used for purchasing local products and only 18% was used for purchasing imported products and in the modern restaurants type 38% of the sales income was used for purchasing local products against 31% of the income used for purchasing imported products. Hence, there is an opportunity for local agricultural producers as they can find new remunerative outlets in this new consumption mode. Keywords New consumption patterns · Out-of-home catering · Modern restaurants · Malewa · Local products JEL codes Q1 · Q110

A. C. Neema (B) Faculté des Sciences Economiques et de Gestion, Université Evangélique en Afrique, Bukavu Sud-Kivu,, Democratic Republic of Congo e-mail: [email protected]; [email protected] A. C. Neema · L. Philippe Unité d’Economie et Développement Rural de Gembloux Agro-Bio Tech, Université de Liège, Belge, Passage des Déportés, 2, 5030 Gembloux, Belgium e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 J. Wood and O. Habimana (eds.), A Multidimensional Economic Assessment of Africa, Frontiers in African Business Research, https://doi.org/10.1007/978-981-15-4510-8_4

67

68

A. C. Neema and L. Philippe

4.1 Introduction It is widely recognized in the Democratic Republic of Congo (DRC) that a succession of wars, looting, robberies, insecurities, and population displacements have made the position of the rural people more precarious and made it difficult for them to practice agriculture, which has aggravated the already very difficult living conditions of the country’s population. After a long decline, which started with the ‘Zairianization’ policy initiated in 1973, the 1996–2002 conflict led to the collapse of agriculture. In principle, with the restoration of peace and the return of the displaced to their original land, there should have been some improvements in the agricultural sector. However, the vast majority of the rural population in the entire country still practices subsistence farming, characterized by extremely low productivity, exorbitant marketing costs, and almost inaccessible markets. However, increasing urbanization represents an opportunity for what is commonly referred to as peri-urban agriculture. However, to cater to this new market, farmers’ offers have to be adapted to meet the expectations of new urban consumers. In DRC in general and in the province of South Kivu in particular, victims of several repeated wars have experienced a population explosion leading to large migrations to urban areas, triggering very rapid urbanization which can be seen throughout the city of Bukavu (Casinga et al. 2017). With this urbanization, we are witnessing a proliferation of different restaurants and supermarkets which are importing their raw material (Vwima et al. 2013). In addition, with urban growth being marked by population growth, there is a significant change in agricultural production accompanied by a multitude of diversified products in the market to meet a new demand for eating food outside the home. Hence, the promotion of local products by restaurants owners which were once reserved for certain segments of the population, are also accessible to low-income households; so, access to local products is not only for a certain particular category but for all social strata of the population (Grain de sel 2012). It is worth emphasizing that the ultra-rapid urbanization of the past 10 years has resulted in the coexistence of two interdependent economic sectors: ‘modern’ and ‘informal’ (OCDE/BAD 1993). The informal sector has developed considerably because of its dynamism, ability to adapt to market requirements and to the cultural context, and flexibility with the result that restaurants in this sector are operating without regulations or culinary norms, as their objective is earning money in return for services offered. On the other hand, clients are concerned about satisfying their food needs, especially since they live at a considerable distance from their workplace or school and home (Bendech et al. 2000). As in most African cities, out-of-home consumption of food is picking up pace even in Bukavu but the thorny question that remains is whether local production can find an opportunity in this new form of food consumption. This study’s aim is identifying and presenting new urban consumption patterns and analyzing the capacity of local agriculture to meet the demand of the catering sector in Bukavu city.

4 New Urban Consumption Patterns and Local Agriculture …

69

4.2 Methodology 4.2.1 Spatial Localization Covering an area of 60 km2 and located in the east of DRC, Bukavu has Lake Kivu to its north which constitutes its limit, the territory of Kabare in the north, Ruzizi river in the east which hosts the waters of Lake Kivu, and Nyamuhinga river on the west. River Nyamuhinga is located in Bagira commune and separated from Rwanda by Lake Kivu (1460 m) and river Ruzizi. Bukavu is administratively sub-divided into three communes with respective areas of 37.6 km2 in the Bagira commune; 12.3 km2 in the Ibanda commune; and 10.1 km2 in Kadutu (Faye 2012; Sadiki et al. 2010; Vwima and Lebailly 2013) (see Fig. 4.1). Vwima (2014), highlighted the importance of food being supplied from Rusizi district (Rwanda) for Bukavu’s (Democratic Republic of Congo) food security. Due to the informal nature of this supply, this chapters characterized its magnitude, drivers, and consequences as well as suggestions on how to develop policies that are likely to improve this supply and facilitate regional integration. Bukavu is a major food market for Rusizi district’s (Rwanda) food products. This, however, is not confirmed by official import statistics of the Congolese Control Office (CCO) which, except for rice, underestimates the border trade for a large portion of food commodities that flow through.

Fig. 4.1 The three communes of Bukavu city

70

A. C. Neema and L. Philippe

The significant flow of food supplies from Rusizi district makes a major contribution to the livelihoods of thousands of families in Bukavu. It emerges that all products supplied by Rusizi district by are also reselling to households. To achieve this, investments need to be channeled to the agricultural sector. This will also help revive the agriculture sector in South Kivu so that it reaches the threshold of selfsufficiency, and even creates a surplus for exports. This regional trade remains a short-term solution for food security in the city and the province. In the long term, it will be better to stimulate this production and promoting the local food trade.

4.2.2 Methods and Techniques This study uses a qualitative research approach based on surveys, interviews, and direct observations. It also reviews relevant literature as the first step. Given the importance of concepts related to out-of-home catering, the literature review includes several areas including food consumption, eating habits, food security, agriculture, and nutrition. The study also does a statistical analysis to analyze and interpret the data so as to link it to the existing realities. The study considered a sample of 45 restaurateurs (15 restaurateurs each in the communes of Bagira, Ibanda and Kadutu which constitute Bukavu) to assess their demand in the food markets. What offers they make to their consumers was also part of the information that was collected. Based on the type of restaurant, using a comparative method made it possible to observe the food habits in Bukavu and also compare them to food practices and habits in the three communes separately. In each commune, all neighborhoods on both sides of the avenue were taken into account. An avenue is an open road in a neighborhood because the city understands the problem of urbanization. Thus, not having a list of all restaurants in the city of Bukavu for random sampling (which may require drawing up a specific list), we were still able to do non-random sampling or exploratory sampling, that is, the researcher set a sampling step that he respected in case he found himself in front of a population concentrated in one place. Thus, the sampling rate considered for our research was three restaurants in each of the three communes. In addition, due to the disparities between the different municipalities, and given the number of restaurants in the different municipalities, while respecting the sampling rate that was decided, we also decided to standardize the sample size to 15 restaurants by municipality. In addition, the survey was supplemented by interviews to get the most information to answer our research question. The restaurants in the three communes were also observed through direct interactions between consumers and service providers (in this case the managers of the restaurants). We observed the ways in which transactions and exchanges between restaurateurs and consumers took place, or we used the participant observation approach. The researcher was initially a stranger to the phenomenon that he/she was studying, but during observations, he/she was no longer just a spectator and

4 New Urban Consumption Patterns and Local Agriculture …

71

became an actor and participated in the development of the phenomenon that he/she was studying (Loubet des Bayle 2000). Our research made it necessary for us to integrate in the life of the population of the area of study to become aware of different eating habits and to get to know new consumption patterns in Bukavu as well as their opportunities for local agriculture. In addition, interviews and questionnaire survey techniques were useful for data collection. The interview as a process of scientific investigation aims at knowing directly what the subjects’ surveyed think, feel, desire, know, or do; this technique uses verbal communication to collect information for a specific purpose (Dépelteau 2000; Mulumbati 1980). Specific questions were formulated to allow us to highlight a link between our study’s variables. Semi-structured interviews were then retained as part of our approach. For the interviews, we had a series of relatively open questions, which we asked the interviewees (Campenhoudt et al. 2017). These interviews were conducted individually with the various food service providers of out-of-home food items. After the interviews we triangulated the collected information to draw broader lessons. Due to the topographical complexity of Bukavu and the sample size used, the survey, which enabled the collection of information from the restaurant owners lasted 55 days (March 15 to May 10, 2018 (Table 4.1). The survey was administered every day between 10 a.m and 5 p.m. and the administration of each questionnaire on average lasted 60 min depending on the availability of the respondent. Table 4.1 gives the dates, municipalities, and restaurants covered by our survey.

4.2.3 Statistical Analysis To analyze the data from the field survey, we used the statistical analysis software. The collected information was encoded in the SPSS version 20 software and then it was exported to STATA 14, a data analysis software that allowed us to relate the variables of the study. This linking of certain variables allowed us to establish a link that may exist between them to meet the objectives of our study. Subsequently, the Ms Excel software allowed us to generate some tables and figures for both qualitative and quantitative data. Based on the comparative method we also compared the revenue generated by the restaurateurs according to whether it was a Malewa or Modern restaurant in Kadutu, Bagira, and Ibanda communes. In addition, the cross-tabulated table linked the dishes offered by the restaurants and the recipes of these dishes. The aim of this study was identifying the types of products commonly used in restaurants (local and imported products). We also identified the reasons that led to food consumption outside the home. Besides asking consumers about this we also asked the restaurateurs for their opinions based on this new trend that they had observed in recent years in Bukavu. This variable allowed us to draw conclusions about the new trend of out-of-home catering and eating.

72

A. C. Neema and L. Philippe

Table 4.1 Details of the survey Dates

Municipalities

Neighborhood

Types of restaurants Malewa

15 March 2018

BAGIRA

Quarter D

Modern

La casa di brayana —–

16 March 2018

Cahi

Maison jéremie —–

17 March 2018

Cahi

Chez donado —–

18 March 2018

Cahi

Restaurant agapé —–

29 March 2018

Quarter A —–

30 March 2018

Quarter B

La simplicité nganda resto Resto chez anicet

—– 1 April 2018

Brasserie

Chez Coccili —–

3 April 2018

Center

Zahwire zanahwire —–

3 April 2018

Center

Resto karibu rafiki —–

4 April 2018

Center

Resto afya bora —–

5 April 2018

Center —–

6 April 2018

Brasserie

Resto chez membre Chez ABC Plage

—– 14 April 2018

Quarter C

15 April 2018

—–

—–

Resto WhatsApp la chaleur

—–

Nganda resto aux coins de sage

2

13

Quarter B

16 April 2018

Center

n1 19 April 2018

Restaurant Chez Taté

IBANDA

Hippodrome

Elisabeth —–

23 April 2018

PEL

New Delicia

26 April 2018

Nyawera

Lingo pizza

29 April 2018

PEL

Maman kinja

2 April 2018

PEL

Mont kahuzi

—– —– —– —– (continued)

4 New Urban Consumption Patterns and Local Agriculture …

73

Table 4.1 (continued) Dates

Municipalities

Neighborhood

Types of restaurants Malewa

9 April 2018

PEL

10 April 2018

Nyalukemba

Modern

La ripaille café —– New carrefour —–

10 April 2018

Nyawera

Keneth house —–

10 April 2018

Av. gouverneur

Fast food —–

11 April 2018

Av. gouverneur

Kabanda

11 April 2018

Av. gouverneur

Roby guest

12 April 2018

Nyalukemba

—– —– —– 12 April 2018

PEL

Jet 7 Chez Muzungu

Aux Elysées —–

12 April 2018

PEL

Chez da Espé —–

13 April 2018

PEL

Saint Luc —–

n2 16 April 2018

11 KADUTU

Mosala

4 Chez da Bishi

—– 17 April 2018

Mosala

Restaurant —–

17 April 2018

Kibonge

Restaurant Fanki —–

18 April 2018

Nyamugo

Chez nyaba deux —–

19 April 2018

Nyamugo

Chez shudja —–

20 April 2018

Cinema

Chez da bintu —–

22 April 2018

Kasali

Restaurant victoria —–

23 April 2018

Industrial

La différence —–

6 May 2018

Mosala

Resto chez papix —– (continued)

74

A. C. Neema and L. Philippe

Table 4.1 (continued) Dates

Municipalities

Neighborhood

Types of restaurants Malewa

8 May 2018

Kasali

Modern Chez kibibi

—– 9 May 2018

Kasali

Restaurant bandal —–

10 May 2018

Nyamugo

Chez da linda —–

10 May 2018

Nyamugo

Chez binja —–

n3

0

15

N = n1 + n2 + n3 = 45

13

32

Source Authors’ survey (2018)

4.3 Results 4.3.1 Distinctions Between the Catering Sectors Our study found that there were 32 restaurants or 71% Malewa restaurants known by the vulgar name ‘Reko’, in each of the communes. They were much more frequented and continued to grow in popularity and secondly, 13 or 29% of the restaurants were Modern in the three communes considered, which imitated the western model and were frequented by the higher social classes. Further, Malewa restaurants existed in the informal sector and so could not be identified by the public administration.

4.3.2 Sources of Supply for the Out-of-Home Catering Sector As Kouassi et al. (2006) maintain, an analysis of the supply of food products is not limited to the local level as it also includes imports and food aid. From our study, it became clear that out-of-home catering in Bukavu used local and imported products in their dishes (see Fig. 4.2). However, 19% of Malewa restaurants exclusively used imported products in their menus while 15% of the Modern restaurants did so. Local products, that is, products originating in the interiors of the Democratic Republic of Congo were also used by the restaurants (59% Malewa and 38% Modern restaurants). Twenty-two percent Malewa restaurants and 46% Modern restaurants used both local and imported products. This finding demonstrates the valuation, integration, and diversification of local and imported products in the out-of-home catering segment.

4 New Urban Consumption Patterns and Local Agriculture …

75

70 60 50 40 30 20 10 0 Only local products

Only imported products

Local and imported products

Malewa

59

19

22

Moderne

38

15

46

Fig. 4.2 Product diversity in restaurants. Source Authors’ survey (2018)

4.3.3 Perceptions of Restaurateurs About the Dishes Offered The demand for a particular food product was determined by the frequency of its consumption while the level of demand was judged on a scale considering the frequency of consumption of various dishes offered by the restaurants. Table 4.2 shows that the demand for products offered in all the categories was very high due to strong preferences of consumers to order food from them. The table also shows that although the products were offered by the restaurants, they did not influence consumer choice even when some dishes in their menus had very little or no demand; the restaurants, however, kept these items on their menus. This shows that consumers were free to choose whatever food items they wanted and the restaurants had to supply them to meet these expectations; this is a strategy followed by the entire food sector.

4.3.4 Overview of Revenue Generated by Products and the Catering Sector Congolese culture and tradition remain the main elements guiding restaurant owners in the choice of products that they offer. Table 4.2 shows that all food products are offered by restaurants based on the food habits of Bukavu’s residents. It can also be seen in Table 4.2 that some products are served only by Modern restaurants and not Malewa restaurants.

76

A. C. Neema and L. Philippe

Table 4.2 Demand for dishes offered Products

Very high demand

High demand

Lower demand

Very few requests

Not requested

Total

Goat meat

33

2

1

3

1

40

Pork meat

28

5

1

0

0

33

Beef meat

21

0

3

4

4

32

Chicken

6

0

1

3

5

15

Fresh fish

16

0

1

5

2

24

Fresh mukeke

7

0

0

0

0

7

Captain’s net

1

3

1

0

0

4

Shrimps (prawns)

2

0

0

0

0

2

Chawarma/pizzas

1

0

0

1

0

2

Cassava or maize bread

25

1

0

3

3

32

Vegetables

40

1

0

2

2

45

Potato/Fries

19

2

1

3

3

28

Green bananas

3

0

1

1

1

6

Plantain bananas

20

5

0

0

0

20

Kwanga

0

0

1

0

0

1

Rice

12

3

1

1

2

19

Beans

17

5

2

2

0

26

Source Authors’ survey (2018)

The total receipts generated by the restaurants are given in Table 4.3. It can be seen in the table that the Modern restaurants charged more for their products that because of the social class of their customers as compared to maquis or Malewa restaurants. In addition, there are a number of identical products (both main dishes and accompaniments) on the menus of both types of restaurants.

4.3.5 New Modes of Food Consumption in Bukavu City: An Opportunity for Local Agriculture A comparison of these two new ways of food consumption in Bukavu, shows that the out-of-home catering sector is an outlet for local producers.When a Malewa restaurant sells a daily recipe for $ 250 by including any product category, and since supplies are made weekly, its sales are valued at $ 1,752 per week. From these weekly sales, the restaurateur allocates 1,024 US dollars to the supply of local products (potatoes, fresh vegetables, beef, goat, pork, zingaro, corn flour, flour of cassava, and sorghum, Lake Mukeke, and fries) and US $ 314 to the supply of imported products (rice, peas, fish (tilapia), corn flour, and beef). This shows that

Vegetarian PLATES

Meats

6

Fresh fish

Fish

3

Chawarma

5

3

Pizza

Bananas- Green peas

2

Sausage

5

6

Rabbit meat

5

7

Chicken

Potatoes-green peas

6

Pork meat

Rice-beans

5

Shrimps (prawns)

Beef meat

7

Captain’s net

6

7

Fresh mukeke

Goat meat

6

8

Fretins

Price (in dollars)

Modern

Dishes + Accompanimentsa

Products

2

1

1

1

2

4

2

9

18

6

13

4

4

3

3

8

Quantity

10

5

5

3

6

8

12

63

108

30

78

28

28

24

18

48

Revenue earned per day ($ per day)

Table 4.3 Revenue generated by Malewa and Modern restaurants in the study area

Potatoes- beans

Potatoes-Green peas

Rice-beans

–

–

–

–

Zingaro

Pork meat

Beef meat

Groat meat

–

Salted fish

Fresh mukeke

Fretins

Fresh fish

Dishes + Accompanimentsb

1.6

1.6

1.2

–

–

–

–

1.5

2.3

2.1

2

–

1.3

1.5

1.6

2.1

Price (in dollars)

Malewa

18

8

30

–

–

–

–

4

18

8

15

–

5

18

15

10

Quantity

(continued)

28.8

12.8

36

–

–

–

–

6

41.4

16.8

30

–

6.5

27

24

21

Revenue earned ($ per day)

4 New Urban Consumption Patterns and Local Agriculture … 77

Salads - egg

Dishes + Accompanimentsa

5

Price (in dollars)

Modern

2

Quantity

– Daily Recipe

474

Dishes + Accompanimentsb

10

Revenue earned per day ($ per day) –

Price (in dollars)

Malewa

–

Quantity

250.3

–

Revenue earned ($ per day)

in Modern restaurants include banana fries, potato fries, rice with sauce, peas, foufou, vegetables (sombe, amarantes, bishusha, choux, and aubergines), and potatoes b Accompaniments in Malewa restaurants include green bananas, rice, beans, peas, foufou, vegetables (sombe, bishogolo, amarantes, bishusha, and choux), and potatoes Source Authors’ survey (2018)

a Accompaniments

Daily Recipe

Products

Table 4.3 (continued)

78 A. C. Neema and L. Philippe

4 New Urban Consumption Patterns and Local Agriculture …

79

58.5% of the sales by Malewa restaurants are directed towards the supply of local products and 18% towards the supply of imported products. The remaining 28% of a Malewa restaurant’s revenue is allocated to other expenses related to the managerial functioning of the catering business. The Modern restaurants have added additional products as compared to Malewa restaurants in their menus thus earning more. When a Modern restaurant has weekly earnings of US $ 3,318, it allocates 31% or 1,018 US dollars to the supply of imported products and 38% to the supply of local products (1,245 US dollars) per week. These supplies include local plants such as potatoes, plantains, peas, flour (cassava and sorghum), fresh vegetables (sombe, amarantes, bishusha, cabbage), and animal products like fish and Mukeke Lake, fretins or sambaza, and goat meat, pork, and chicken. The result is that restaurants in Bukavu buy agricultural products and value local products by devoting a larger part of their revenues to them.

4.3.6 Reasons for Out-of-Home Eating Restaurateurs in Bukavu discussed some reasons for people opting for new patterns of food consumption in Bukavu’s towns. Table 4.4 shows the main reasons for this new trend from the point of view of restaurant owners. Several studies have shown that more people are working away from home, and have a workplace located outside their residential towns (Baccaïni et al. 2007; Muteba 2014). The average distance between home and the workplace has also increased and this does not allow people to consume at home (domicile). As shown by our results, in Bukavu the distance from the workplace to home is a major reason for the popularity of the out-of-home catering sector (44% importance). This sector allows workers to eat, save time, and get back to work quickly. This is followed by the quality of food served by these restaurants (40%); the availability Table 4.4 Reasons for out-of-home food consumption The reasons for food consumption outside the home

Total (N = 45)

Percentages

Diverse dishes

13

29

Distance from work or school to home

20

44

Cooking time

12

27

Food quality

18

40

Conviviality (meeting others)

7

16

Availability and price of food

17

38

Income level

6

13

Others reasons

3

7

Source Authors’ survey (2018)

80

A. C. Neema and L. Philippe

and price of food (38%); and also, the time associated with cooking food. Out-ofhome food consumption makes it possible to eat products that people do not have time to cook at home; the cost of the dishes also probably encourages people to eat out-of-home. The popularity of this new trend is also explained by other reasons including comfort and the renowned name of the restaurant, the welcoming atmosphere created by the restaurant’s staff, internet connection which benefits customers, and sometimes customers avoid family problems during working hours; also, the matrimonial status of a person (for example, for bachelors who haven’t time to prepare food) obliges them to consume food outside the home. Note also that the high level of a person’s income (13%) also pushes him to eat outside because with the money at his disposal he can find something he wishes to consume at whatever price as he can afford it. Unlike home-based meals, whose composition depends on the family’s decision, out-of-home restaurants shift the group’s food choices (that is, the choice of the whole family) to individual choices as every member is free to choose what he wants to eat. Taste of food is also an important motivating variable for eating out frequently, because the lack of taste that leads to a poor interest in certain dishes and the good quality of the food are synonymous with a balanced diet, pleasure and wellbeing, which is therefore a determining factor in food consumption outside the home (Magali, nd). Hence, out-of-home food consumption needs to meet consumers’ expectations, including comfort (space, equipment, service, lighting), the freedom to choose dishes, conviviality in the restaurant, and the quality of food (Karsten et al., 2015 cited by Magali, op cit.).

4.4 Discussion The results of our study are supported by those obtained by previous research and provide a multidimensional understanding of eating habits and the popularity of out-of-home eating in the research area. Agence (2017), concluded that local products were favored in Modern restaurants and their share in out-of-home food was estimated at 2.9% of the purchase value of the food. Ségolène and Christine (2014), showed that plant origin products (vegetables and fruits) and animal origin products (meat and poultry) were regularly supplied with local products in out-of-home catering. This is in line with what has been found in the South Kivu province in the Democratic Republic of Congo where the integration of products of plant and animal origin has been done in the out-of-home catering sector in Bukavu. In his research in the city of Kinshasa, Muteba (2014), found that with the economic crisis, small makeshift restaurants called Malewa developed near professional circles and in working-class neighborhoods where people ate, especially when they

4 New Urban Consumption Patterns and Local Agriculture …

81

found themselves at a considerable distance from the workplace and home. This reason was also supported by our research, according to which distance from home was a major reason for eating out. An analysis of our findings showed that in consumption patterns in Bukavu, Malewa restaurants form a part of the new consumption patterns and are seen as a sector of activity that has emerged with population growth due to urbanization. The same is true in Abidjan (Côte d’Ivoire) (Akindes 1991), where this sector is strongly influenced by the economic recession and constitutes a strategy of integration and initiative in urban economic life. Akindes also shows that the catering sector, through the new modes developed in recent years, has become more important and constitutes an essential link in production, distribution, and food consumption. These new consumption patterns in Abidjan are also marked by a strong demand for local products due to cultural preferences. Demand for products of plant origin (rice, tubers, plantains, and yam) remains high in this country. The sauce that accompanies the dishes made from products of plant origin is also made from local products, usually vegetables. As our study shows, new food consumption patterns present a great opportunity for local agriculture in the Democratic Republic of Congo. For Magali (nd), this encourages not only the acquisition of basic food products but also shows the population’s enthusiasm in appreciating and consuming local products through outof-home food consumption. This also allows everyone, regardless of income levels, to access products that are nutrient-rich and thus contributes to an increase in demand from local producers. This provides sufficient proof that the catering sector can play a major role in the development of local agricultural products.

4.5 Conclusion The constant evolution of the catering sector and the multitude of people adopting new consumption patterns led us to reflect on this issue of food consumption outside the home. A survey was conducted among 45 restaurateurs using a standardized sampling of 15 restaurant owners each in three communes to evaluate the products as well as the food items offered to consumers. The categorization of the out-of-home catering sector in Bukavu showed different modes of food consumption, including Malewa restaurants, Modern restaurants, and sidewalk restaurants; these, however, were not the focus of our study. Bukavu’s restaurants source locally and directly from nearby retailers by developing certain strategies such as direct purchase of food products, bidding by themselves or through suppliers, direct delivery by suppliers, and placing orders with safe suppliers. In addition, whether it is a Malewa or Modern restaurant, a significant proportion of the revenue generated is used for purchasing local products. With the development of these new modes of food consumption manifested by the out-of-home catering sector, it can be concluded that these have a significant

82

A. C. Neema and L. Philippe

impact on the level of local agricultural production. In both types of restaurants, the integration, valorization, and supply of local products encourages local farmers to produce more. The predominance of local agricultural products especially fresh vegetables and those of animal origin in what the restaurants offer shows the place and importance given to local products by restaurateurs in Bukavu. This finding invalidates the hypothesis that the catering sector concentrates more on imported, processed, and ready-to-eat products and confirms the hypothesis that the new mode of out-of-home catering values local agricultural production and is an outlet for local agricultural products. Agricultural development must not only be a question of supply because to be sustainable, it also has to cater to demand. It is no longer enough to produce because for agricultural products to sell well to make money they have to match consumer demand. With rapid urbanization, consumption patterns in cities are changing from those in the villages. This research was conducted to better understand the expectations and demands of restaurants and consumers in the cities. Although local production in the Democratic Republic of Congo is not sufficient to meet the needs of the population and imports are essential, we recommend that the government of South Kivu province support the out-of-home catering sector, which is necessary and important for the survival of people and for improvements in the national economy because this sector contributes enormously to the economic and social development of the province and of the entire nation; local production should be encouraged by protecting producers from imported products sold in the local markets while improving agricultural feeder roads for better transport of local products to markets.

References Agence, B. (2017). Observatory of organic products in out-of-home catering, Organic agency consumption barometer, CSA Research. Akindes, F. (1991). Urbanization and development of the informal food sector in Côte d’Ivoire: the example of Abidjan, School of Higher Studies in Social Sciences (S.H.S.S.S), Paris. Baccaïni, B., F. Semecurbe, and G. Thomas. (2007). Home-to-work travel amplified by periurbanization, Insee Première, n°1129, INSEE. Bendech, V., M. Chauliac, P. Gerbouin Rérolle, N. Kante, and D. J. M. Malvy. (2000). The challenges of food consumption in urban areas in Bamako. Cahier santé publique, 12, (1): 45–63. Campenhoudt, L. V., J. Marquet, and R. Quivy, (2017). Research manual in Social Science, Dunod, 5ème édition, p. 379. Casinga, C. M., C. A. Neema, C. M. Kajibwami, N. L. Nabahungu, and B. P. Mambani. (2017). Effect of Soil Moisture Regimes on Seed Iron and Zinc Concentration of Biofortified Bean Genotypes against Malnutrition in Sud-Kivu Highlands. Journal of Agricultural Science, 9 (12): 241–252. Dépelteau, F. (2000). The process of research in the social sciences and humanities: from the initial question to the communication of results. Bruxelles, De Bœck Université, coll. Méthodes en Sciences Humaines, p. 417. Faye, M. (2012). Environmental and Social Management Framework (ESMF). Urban Development Project (UCOP). Project Coordination Unit in the DRC.

4 New Urban Consumption Patterns and Local Agriculture …

83

Grain de sel. (2012). Valorization of local products: in the face of challenges, a diversity of solutions, GRET, revue d’inter-réseau, n° 58, développement rural. Retrieved online at: www.Interreseaux.org. Kouassi, B., G. Sirpe, and A. Gogue. (2006). Agricultural trade and sustainable food security in Central West Africa, Paris: Karthala. Loubet des Bayle, J. L. (2000). Introduction to social science methods, Paris-Montréal: L’Harmattan, Éditeur, p. 366. Magali, R. (nd). Good food strategies: towards a sustainable food system in the Brussels-Capital Region, Bruxelles environnement, 1000 Bruxelles. Mulumbati, N. (1980). Manual of General sociology, Ed. Africa, Lubumbashi. Muteba, D. (2014). Characterization of household food consumption patterns in Kinshasa: Analysis of the interrelationships between lifestyles and eating habits (PhD thesis). Liege-University Gembloux-Agro-Bio Tech, Belgium, p. 179. OCDE/BAD. (1993). Elements of a forward-looking economic vision: West Africa on the 10-year horizon. SAH/D (93) 411, document de travail; 3: p. 61. Sadiki, N., V. Ine, M. Jan, O. André, O. Pierre, D. Kalegamire, and C. Bahati. (2010). Development of the city of Bukavu and vulnerability mapping, D.R. Congo, Annales Sci. & Sci. U.O.B. Appl. Vol. 2. Ségolène, D., and A. Christine. (2014). Demand for local collective catering products: what are the links with local supply in an industrial agricultural region? Le cas de l’Ile-de-France, Géocarrefour [On line], 89/1–2 | 2014, put online on 23 December 2014, accessed on 12 April 2019. Vwima, S. (2014). The role of the border trade of food products with Rwanda in provisioning the urban households of Bukavu town, PhD thesis. Vwima, S., and P. H. Lebailly. (2013). Food supply to the city of Bukavu: flows from the interior of South Kivu, Rwanda and North Kivu, Peri-urban territories: development, challenges and prospects in the countries of the South, International symposium—ULg-Gembloux 19 December 2013.

Part II

Consumption, Poverty and Inequality

Chapter 5

Multidimensional Inequality in Ethiopia Getu Tigre

Abstract Ethiopia has consumption inequalities which are higher in urban areas as compared to the rural areas in the country. There are considerable differences in regional consumption inequalities between different regions of the country. This study uses a multistage multidimensional inequality analysis for multidimensional poverty indicators. Even though multidimensional poverty is high, multidimensional inequalities are quite low in Ethiopia. The inequalities in the multidimensional indicators decrease over the wealth quintiles and living standards contribute the most to multidimensional inequalities. There are inequalities in landholding patterns in Ethiopia and these inequalities differ across regions and wealth quintiles. Reducing inequalities between socioeconomic groups will have a big impact on reducing poverty as compared to reducing inequalities within groups as between-group elasticity is greater than within-group elasticity. A gender based decomposition of inequalities shows that within-group inequalities, the marginal impact of inequalities, and the marginal impact of poverty are greater than the between-group components. Between regions inequalities are greater than within region inequalities, so there are significant differences in the inequalities between regions in Ethiopia which need to be considered. Parents’ education has a positive impact on children’s education. Mothers’ education contributes more than fathers’ education to both sons’ and daughters’ education. In countries like Ethiopia where girls are marginalized, educating daughters (tomorrow’s mothers) has a positive intergenerational inequality reducing effect. Keywords Inequality · Education · Ethiopia · Multidimensional · Elasticity JEL codes I31 · C21 · C43 · D63

G. Tigre (B) Department of Economics, Addis Ababa University, Addis Ababa, Ethiopia e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 J. Wood and O. Habimana (eds.), A Multidimensional Economic Assessment of Africa, Frontiers in African Business Research, https://doi.org/10.1007/978-981-15-4510-8_5

87

88

G. Tigre

5.1 Introduction The recent increase in income inequalities in a number of developed and developing countries and their potential impact on an economy have made inequality a central issue in economics. Many African nations are growing but the growth is not equally distributed across socioeconomic groups and regions. While poverty reduction is of wide interest and a top policy priority in Ethiopia, inequalities are less frequently addressed. Concerns about inequalities are essential because high levels of inequalities pose a serious threat to progress and social stability. Consensus exists that high levels of inequalities are a social injustice. Conventional inequality indices are based on the assumption that individuals and groups of individuals can be ranked according to specific characteristics such as income and it is a straightforward exercise. To measure income correctly a precise record of income is essential which is very unlikely in developing nations like Ethiopia (Vida et al. 2008). Consumption is generally considered a more appropriate measure of well-being than income, especially in poor countries where the main concern is the fulfillment of basic needs (Idrees and Ahmad 2010). Gross and net incomes are different in progressive tax systems and in this case consumption is better than gross income as a measure of well-being. Hence, it is important to measure inequalities in consumption expenditure including food, housing, education, and health expenditure. The superiority of consumption in measuring well-being is attributed to its close correspondence with individuals’ basic needs and its low dispersion. Human well-being is multidimensional and any analysis of inequalities should take the many dimensions of well-being into account (Bourguignon and Morrisson 2002). Measuring inequalities and their various dimensions is, however, not an easy task. Ranking individuals along educational, health, and other non-monetary attributes is a complex exercise. This complexity is associated with not the differences between distributions but also because of possible correlations between the various attributes of welfare. This study examines inequalities in different dimensions of well-being focusing on three important dimensions of life: standard of living, health, and education. This study uses two approaches: inequality in each dimension and inequality in the combined indicators of well-being. Inequality in each dimension indicates the extent of inequalities in each indicator and can help the government and other actors to take some corrective measures to reduce inequalities in that dimension. It also helps identify areas of intervention to reduce existing multidimensional inequality. A combined inequality considers the correlation between indicators and gives us the combined effect of inequalities on well-being which helps compare households and regions in the country. Inequality has been analyzed in Ethiopia (Gelow 2009; Kedir et al. 2014; MoFED 2014; Tesfaye and Mulberge 2014; Woldehanna et al. 2008). These studies use inequality indices to discuss inequalities in income and consumption expenditure in Ethiopia and show the existing levels of inequalities and highlight areas of interventions to reduce the existing income inequalities in the country. However, these

5 Multidimensional Inequality in Ethiopia

89

studies focus on inequalities in income and consumption expenditure among households and overlook other dimensions of inequalities such as health and education. This is an observed gap in this area. This study uses a multistage multidimensional inequality analysis which combines different dimensions of well-being. The general objective of this study is assessing the extent of one-dimensional and multidimensional inequalities in Ethiopia and highlighting areas of interventions to reduce existing multidimensional inequalities in the country, its regions, and within income groups. The specific objectives of the study are: Examining the extent of one-dimensional and multidimensional inequalities in the country, between regions, and income groups. Identifying areas of interventions to reduce the existing multidimensional inequalities and multidimensional components and indicators of inequalities. Identifying the determinants of intergenerational multidimensional inequalities.

The rest of the chapter is organized as follows. Section 2 does a review of theoretical and empirical literature. Section 3 discusses the data used for the analysis and one-dimensional and multidimensional inequality measures used in the analysis. Section 4 discusses the results. Section 5 gives a conclusion and Sect. 6 gives some recommendations based on the theoretical literature reviewed and the study’s empirical findings.

5.2 Literature Review 5.2.1 Unidimensional Measures Inequalities are a development challenge both in rich and poor counties and are clearly an important issue which requires much more attention. The Oxford Dictionary of Economics defines inequality as differences in the distribution of economic stocks or flows among economic agents. Scott and Marshall (2009), define inequality as unequal rewards or opportunities for different individuals or households within a group or groups in a society. In this definition, unequal rewards refer to outcomes or achievements while unequal opportunities are the freedom to obtain alternative outcomes. Inequalities weaken the poverty reducing capability of economic growth and there is an increasing pressure on governments to address inequalities. Increased inequalities can also lead to dissatisfaction, social unrest, and violence and hinder the growth process in a country and worsen insecurity (Ostry et al. 2014). Measuring inequality has received much attention both in theoretical and empirical research. Literature on inequality measures has increased since the publication of the Lorenz curve and the Gini index in the early 1900s. An interesting measurement of the inequality theory was also proposed in Theil (1967, 1979). The extension of Theil’s inequality decomposition technique provides a useful tool for examining

90

G. Tigre

the contributions of different sub-groups to total inequality. Another measurement of inequality is Atkinson’s (1970) index which generates multifaceted inequality calculations. The argument of using a monetary attribute in an inequality analysis shows that there are reasons to believe that economic conditions drive other aspects of living standards and that a monetary indicator does tell us what we need to know about wider aspects of well-being. But monetary inequalities are ambiguous when households have different characteristics (Maasoumi 1999). It should also be mentioned that a change in consumption inequalities may be a result of bad outcomes in other welfare dimensions. In addition, the well-being of a household might have dimensions that cannot be purchased (Duclos et al. 2001).

5.2.2 Multidimensional Measures of Inequality Stiglitz et al. (2009), argue that individual well-being is multidimensional. If we want to take the multidimensionality of individual well-being seriously, we need to incorporate its various dimensions explicitly into an analysis of inequality and consider their correlation. A multidimensional perspective on measuring inequality and considering inequalities at different levels of aggregation and time horizons using both qualitative and quantitative techniques is likely to be particularly valuable. In doing so, one should account for the inter-relationships and correlation between the different dimensions in measuring and analyzing inequalities (Heshmati 2014). The most popular multidimensional inequality index of well-being is the Human Development Index (HDI) summarizing the performance of countries in three dimensions of well-being: standard of living, health, and education. Two alternate approaches for measuring multidimensional inequality can also be distinguished: the normative approach and the two-stage approach (Bosmans et al. 2015). The normative approach was developed in the unidimensional setting by Atkinson (1970), and extended to the multidimensional setting by Kolm (1977). In the normative approach, inequality measures are derived from social welfare functions and inequality is defined as social welfare gains that can be obtained by optimally redistributing the available goods. The two-stage approach was pioneered by Maasoumi (1986), in which the first stage associates a well-being level to the bundle of goods of each individual. Once we get the well-being level of each household the second stage simply uses unidimensional inequality. The Lorenz zonoid was introduced as a mdimensional generalization of the standard Lorenz curve by Koshevoy and Mosler (1996). A multidimensional Gini coefficient can be derived from Lorenz zonoid (Decancq and Lugo 2012). An alternative strategy by Anderson (2004), proposed a multidimensional distance measure to measure the pairwise distance between the vectors of the outcomes. Both approaches represent a mathematical or a geometrical extension of the one-dimensional Gini coefficient. However, they lack normative content in the sense that inequality cannot be readily interpreted in terms of welfare losses. Hence, this study regards them as less attractive for measuring inequalities in well-being (Decancq and Lugo 2012).

5 Multidimensional Inequality in Ethiopia

91

In the two-stage approach two indices for measuring multidimensional inequality are derived, and there is aggregation across dimensions and across individuals. The sequencing of aggregations is essential in terms of the underlying principles. In the first approach aggregation is first done across individuals and then across dimensions which is not sensitive to the correlation between the dimensions. In the second approach we first aggregate across dimensions and then across individuals which can be sensitive to a correlation between dimensions. To obtain a correlation sensitive rank dependent inequality index one must be willing to give a large weight to the bottom of the distribution (Decancq and Lugo 2012). There is no single best measure of inequality; each measure has its own advantages and limitations. An exhaustive measure of inequality is possible only when different methods and indices are combined and applied (Vida and Jonas 2008). Following Weymark (2006), there are a number of basic properties that a multidimensional inequality index should satisfy. These can be grouped into two sets of axioms. The first focuses on those properties that are not concerned with the distributional sensitivity of the inequality measures. These non-distributional axioms which are straightforward generalizations of their unidimensional counterparts include continuity, anonymity, normalization, replication invariance, scale invariance, and decomposability. While inequalities can be considered or measured in many different dimensions, the two key measures are inequality of outcomes and inequality of opportunities. Inequality of outcomes refers to the inequality of income, wealth, education, and health status while inequality of opportunity refers to the choice offered to an individual or a household. The notion of equal opportunities is that success in life reflects a person’s choices, efforts, and talent and not a person’s background defined by circumstances as race, place of birth, and family origin. According to Roemer (2014), opportunities have to be equitable and opportunity equality is an attractive notion of justice. An individual’s outcome or achievements is a mix of opportunities afforded to the individual and the choices she/he makes. Currently there is an argument that inequality of opportunities has to be given due attention and therefore equality of opportunities is getting the attention of policymakers. Equality of opportunities aims to level the playing field so that circumstances such as gender, birth place, and ethnicity which are beyond the control of an individual do not influence one’s life chances. An equal-opportunity policy should aim at providing everyone with the same opportunities to achieve or enjoy excellent outcomes. Though it is more difficult to measure inequality of opportunities, ensuring individuals have equal opportunities (equality of opportunity) is a policy goal for achieving equal outcomes.

5.2.3 Empirical Evidence There are considerable income inequalities in Ethiopia across regional and ethnolinguistic groups (Kedir et al. 2014; Tesfaye et al. 2014). A study of rural poverty in Ethiopia by Gelow (2009), using a fixed effect model showed that a change in

92 Table 5.1 Consumption inequality trends in Ethiopia as measured by the Gini coefficient

G. Tigre Residence

1995/1996

1999/2000

2004/2005

Rural

0.27

0.26

0.26

Urban

0.34

0.38

0.44

National

0.29

0.28

0.30

Source Woldehanna et al. (2008)

inequalities significantly affected the poverty gap. Inequality emphasizes on dispersion across agents or households and how the welfare cake (for example, GDP) is shared among the population. Some recent income inequality analyses point out that there are higher inequalities in urban than in rural areas in Ethiopia (see Table 5.1). According to Woldehanna et al. (2008) consumption inequalities in Ethiopia which were 0.29 in 1995–1996 increased to 0.30 in 2004–2005; there was a significant increase in inequalities in urban areas and the increase in inequalities at the national level was mainly because of a substantial increase in inequalities in urban areas. According to Fentaw (2016), consumption inequalities in Ethiopia were 0.38 in 2010–2011. Further, in estimates of inequalities in South Wollo zone, the highest inequalities were observed in major towns, followed by emerging towns in the region but small inequalities were observed in small towns. Similarly, the Gini coefficient estimates of urban inequalities at 0.37 were greater than in rural areas (0.27) (MoFED 2014). However, urban inequalities declined over time and the gap between urban and rural inequalities narrowed. Cain et al. (2012), showed that the fact that urban inequalities were higher than those in rural areas, urban biased policies in developing countries exacerbate inequalities between urban and rural areas. An inequality analysis across different regions of the country is essential to highlight inequality reducing intervention areas. Gakidou et al. (2000) and Ribero and Nunez (1999), state that sustainable development cannot be achieved without significant investments in human capital; education and healthcare are key elements in developing human capital. There have been general improvements in the health status in Ethiopia (Tranvag et al. 2013), and health inequalities have declined over time. According to Gallardo et al. (2017), inequalities in health status depend on the education level of the mother, household income, gender, and place of residence.

5.2.4 Components of Multidimensional Inequality Multidimensional inequality encompasses many dimensions and indicators. Sustainable development cannot be achieved without significant investments in human capital (Gakidou et al. 2000). Health and education inequalities together with other living standard indicators have resulted in an increasing interest in a multidimensional inequality analysis (Checchi 2000).

5 Multidimensional Inequality in Ethiopia

93

5.3 Education Ethiopia has quite high illiteracy rates on average, especially in rural areas because of less access to education (CSA 2016). Educational inequalities can be examined using indicators that capture different levels of education (Bigotta et al. 2014), for example, the percentage of individuals who have attained a particular level of education, or the number of years of education attained (Meschi et al. 2014; Morrisson et al. 2013). Education is a key to attaining sustainable development goals by 2030. For education to have a positive impact on growth and development, it is necessary to ensure equality of learning. The World Inequality Database on Education (WIDE) highlighted that since 2010 less than 25% children in rural areas in 24 of the 52 countries studied have had an opportunity to complete a pre-primary program. Less than 50% of the poorest children in 40 out of the 93 countries studied have completed primary school and less than 50% of young people in 57 out of 127 countries have completed upper secondary school. In Ethiopia, the percentage of women with no education decreased from 66% in 2005 to 48% in 2016 while the percentage of men with no education also declined from 43% in 2005 to 28% in 2016 (CSA 2016). Education is an important factor influencing an individual attitudes and opportunities. All household members benefit from having an educated person in the household as education is assumed to have positive externalities for other family members and society. Children’s education is influenced by parents’ education because families are the building block of a society and are the primary institution of learning for growing children. Parents’ education and experience play an important role in shaping children’s future (McLanahan et al. 2008). Children with educated families are better educated than those with uneducated families. Econometrically this can be estimated using: G E(0) = G E(1) =

n y¯ 1 log n i=1 yi

(1)

n yi 1  yi log n i=1 y¯ y¯

where yi j is educational attainment of child i in household j; e j is parental educational attainment of the household head j; β is the intergenerational education coefficient; xi j is a vector of household characteristics including gender and religion; and υi j is the error term.

5.4 Health Like wealth, health too is not equally distributed among individuals or households. Health inequalities have been a main challenge for public health policies. The World

94

G. Tigre

Health Organization (WHO) has targeted the reduction of health inequalities. Health inequalities should be a major concern for government policies in all countries, particularly among the most disadvantaged. Inequalities in health often reinforce inequalities in other dimensions of life such as income and education (The World Bank 2005). In Ethiopia, 1 in 15 children dies before reaching age 5; 7 in 10 of the deaths occur during infancy (CSA 2016). The World Bank research also adds that child mortality varied across regions in Ethiopia. Under-5 mortality ranged from 39 deaths per 1,000 live births in Addis Ababa to 125 deaths per 1,000 live births in Afar. Malnutrition among children and adults is one of the widely used health indicators in multidimensional poverty and inequality measures. Stunting (low height-for-age) is a sign of chronic under-nutrition that reflects failure to receive adequate nutrition over a long period and can also lead to recurrent and chronic illnesses. The nutritional status of adults can be measured using the body mass index (BMI). The post-2015 Sustainable Development Goals (SDGs) are based on the central notion of addressing health inequalities in all countries by promoting universal health coverage for people of all ages. There is consensus that health inequalities are not self-correcting and require interventions (policies and programs) to change. In Africa, there is little evidence of success in improving equality of health outcomes (Stephen and David 2007). Using Ethiopian DHS data, Tranvag et al. (2013), showed that there was unequal distribution of healthcare centers in Ethiopia. Their research also pointed out those inequalities in length of life within wealth quintiles was considerably larger than between them.

5.5 Living Standards The term living standards is used for expressing the conditions in which a person or a nation lives. Living standards are represented by access to electricity, clean drinking water, improved sanitation, floor material, cooking fuel, and asset ownership. Electricity is an important indicator of living standards. It can be used for light, cooking, and electronic devices. Electronic equipment enhances productivity and time spent on leisure activities. A household is deprived in electricity if it has no access to it. Life is impossible in the absence of water. Contaminated groundwater is a source of many diseases like diarrhoea leading to many deaths. A household is deprived in water if it has no access to drinking water or a safe drinking water source is more than 30 min walking distance. Sanitation is directly related to hygiene and access to improved sanitation is vital for a healthy life. Cooking fuel is important and related to the use of biomass in rural areas and also in some urban areas. Biomass fuel causes environmental degradation. Use of appropriate cooking fuel indirectly provides environmental sustainability. Asset ownership is related to access to information (TV, radio, and mobile phone), access to easy mobility (bicycle, car, and motorbike), and assets for livelihood (livestock and agricultural land). In poorer countries, the dimension of standard of living contributes the most to multidimensional poverty. Empirical research shows that in Ethiopia the living standard indicators contributed

5 Multidimensional Inequality in Ethiopia

95

Table 5.2 Households’ access to different living standards indicators Living standards indictors

Rural (%)

Urban (%)

Total (%)

Access to improved sources of water

57

97

–

Access to electricity

8

93

25

Use of improved toilet facilities

4

16

6

Source CSA (2016)

more than 46% to multidimensional poverty (Alkire et al. 2011; Getu 2018). There are also rural-urban variations in access to some of the living standard indicators in Ethiopia (Table 5.2).

5.6 Data 5.6.1 Data This research uses the Ethiopian Demographic and Health Survey (DHS) data. DHS is a comprehensive cross-section dataset that consists of samples from all regions in Ethiopia. This study mainly uses the most recent DHS data from the survey conducted in 2016. Since the DHS data has no income variable, we used the Household Consumption and Expenditure Survey’s (HCES) data for the unidimensional income inequality analysis. It is a survey which contains a nationally representative sample to characterize important aspects of households’ socioeconomic conditions. In Ethiopia, like in many poor countries where the main concern is the fulfillment of basic needs, it is more important to measure inequalities in the consumption expenditure as income data is not easily available and if available it is not reliable. The unit of analysis is a household; a household has common resources and takes decisions that affect almost all the members of a household.

5.7 Methodology The analysis of inequality has a long history but there is no agreement on how best to measure inequality with the result that a number of inequality measures exist, and different combinations of these are used by different studies. Our analysis uses unidimensional and multidimensional measures of inequality.

96

G. Tigre

5.7.1 Unidimensional Measures of Inequality There are different measures of unidimensional inequality. There is no consensus about using a single inequality measure in all the cases. All unidimensional inequality measures have their own advantages and limitations. We discuss some of the most commonly used inequality measures to decide which measure provides the best fit for our data. The Gini coefficient is the most widely used measure of inequality in empirical literature which measures the extent to which the distribution deviates from equal distribution. The Gini coefficient was developed by Italian statistician Corrado Gini. The Gini coefficient varies from 0 which indicates complete equality to 1 when all the income of a country is owned by one individual (complete inequality). The Gini index also facilitates direct comparisons with any quantitative variables which describe two or more populations, regardless of their sizes. It can, therefore, be used easily for comparing inequalities between groups, regions or countries. The Gini coefficient satisfies the important principles of anonymity, scale independence, population independence, and the transfer principle. Researchers working with the Gini coefficient need to be aware that it is most sensitive to inequalities in the middle part of the income spectrum, but in some cases researchers have valid reasons to emphasize inequalities in the top or bottom of the spectrum. Because of these and other limitations, besides the Gini coefficient, entropy measures (for example, the Atkinson and Theil indices) are frequently used inequality measures in empirical literature. British economist Anthony Barnes Atkinson developed the Atkinson index. The Atkinson index is useful for measuring inequality and determines which end of the distribution contributes most to an observed inequality. Atkinson’s inequality measure is a welfare-based measure of inequality. It shows the percentage of total income that society should forego to have more equal distribution of income. This depends on the degree of aversion to inequalities. The inequality aversion parameter ε measures the social utility gained from a complete redistribution of resources. The choice of the Atkinson inequality measure relative to the Gini coefficient is guided by the sub-groups’ consistency and sensitivity to inequality in the lower end of the distribution. If inequalities increase in one sub-group (region, religion, or ethnic group) and remain unchanged in all the other groups, then overall inequalities increase. But the Gini coefficient does not have this property. The Atkinson inequality measure puts more weight on the lower end of the distribution but the Gini coefficient gives equal weight to the entire distribution. The Atkinson coefficient is more appropriate when we are more interested in the lower end of the distribution such as child mortality and illiteracy. The Theil index is a member of the generalized entropy class of inequality measures. The Theil index can be used for individual as well as group data and allows decomposing inequality into within-group and between-group components. The Theil index is a widely used inequality measure because it has the desirable property

5 Multidimensional Inequality in Ethiopia

97

of decomposability. If the population is divided into many groups, overall inequality can be expressed as the sum of two terms capturing within and between-group inequalities. The within-group inequality index indicates how much inequality is due to variations between individuals in each of these groups, whereas the between-group inequality index quantifies inequality due to differences in the average incomes of the groups. The Theil index has many desirable properties such as population replication invariance, translation invariance, and scales invariance. The index has a wider range of scalar variations and is bounded to 0 and infinity: the closer to zero, the lower the inequality. This index cannot directly compare populations of different sizes or group structures. In a situation where the population sizes of the groups being considered are different, the difference in the index among the regions or groups may be partially because of the differences in population size. The generalized entropy index is one of the most widely used measures of inequality, all of the generalized entropy (GE) class of inequality measures can be expressed in terms of the following general formula:   n   1 y α 1 −1 G E(α) = 2 α − α n i=1 y¯

(2)

where GE(α) is the generalized entropy index, the value of GE ranges from 0 to ∞, zero represents perfect equality and the larger its value the higher the inequality. The parameter α (α ≥ 0) represents the weights given to distances between incomes or other values at different parts of the distribution. The most common values of α are 0, 1, and 2. When α = 0 more weight is given to distances at the lower end of the distribution, that is, GE is more sensitive to changes at this end of the distribution. If α = 1, equal weights are given across the distribution, while α = 2 gives more weight to distances between incomes at the higher end of the distribution. This decomposition is usually applied only to the generalized entropy GE (0) because the arithmetic can be complex for some inequality measures; it can also be shown that the generalized entropy measure with GE(0) and GE(1) become two of Theil’s measures of inequality: n y 1  log n i=1 yi

(3)

n yi 1  yi log n i=1 y y

(4)

G E(0) = G E(1) =

98

G. Tigre

5.7.2 Multidimensional Measurement of Inequality Well-being is a multidimensional concept and recognition of this fact has resulted in an increase in interest in distribution analyses of education and health (Bakare 2012). Education status and access to education are not equally distributed across countries or population groups. Given the impact of education and health on economic development, inequalities in education and health status represent a loss in aggregate welfare. Any inequality measure of well-being should take this multidimensionality explicitly into account. Education, living standards, and health status are multivariate distributions that make the traditional univariate measures of inequality such as the Gini coefficient and concentration index less attractive. Most inequality measures are unidimensional; however, there are also multidimensional measures of inequality.

5.7.3 Frameworks Measuring Multidimensional Inequality The first step in measuring multidimensional inequality is identifying the indicators of well-being and the second step is measuring these inequality indicators for each person or household based on the unit of analysis. Let us assume that the domains of well-being have been identified and households’ achievements in all the dimensions are measured and comparable. Suppose there are n individuals and there are j relevant n× j dimensions of well-being. Each distribution matrix X in R++ represents a particular distribution of the outcomes for n individuals in j dimensions: ⎡

x11 ⎢x ⎢ 21 ⎢ X = ⎢ x31 ⎢ ⎣ . xn1

x12 x22 . . xn2

x13 . . . .

. . . . .

⎤ x1 j x2 j ⎥ ⎥ ⎥ . ⎥ ⎥ . ⎦ xn j

(5)

A row of matrix X refers to the outcomes of one individual and a column refers to j the outcome of one dimension. The set of bundles is B = R++ and the set of the entire n distribution is D = B . Dimensions can be measured in an interpersonal comparable way; some standard dimensions considered include income, living standards, health, and education. Let x ij be the non-negative achievement of an individual or household i in dimension j and let the achievement vector x = (x i1 , x i2 , …, x im ) summarize these achievements across all m dimensions for individual i. The distribution matrix can be compared by making use of a social evaluation function that maps a positive n × j distribution matrix to the positive real line. The social welfare function is W : D → R. The value W(x) is the social welfare level associated with distribution x in D. This study uses aggregation first across dimensions and then across individuals. This is preferred because it takes into account the correlation or interdependence

5 Multidimensional Inequality in Ethiopia

99

between dimensions which is very important in an analysis of multidimensional inequality. This multidimensional inequality index is different from other indices. For example, this multidimensional inequality index is different from the Human Development Index (HDI). Health indicators in HDI are life expectancy at birth but health indictors in this inequality measure are child mortality rate and child and adult malnutrition. Similarly, the living standard indicators of HDI are gross national income per capita but living standard indicators in this multidimensional inequality index are access to electricity, water, sanitation, and asset ownership of households (such as land, car, and refrigerator).

5.7.4 The Araar (2009) Multidimensional Inequality Index This research uses the most recent multidimensional inequality index- the Araar et al. (2009) multidimensional inequality index which satisfies a fundamental set of desired properties. The Araar multidimensional inequality index for the k-dimension of well-being can be formulated as: I =

K 

ϕk [λk Ik + (1 − λk )Ck ]

(6)

i=1

where I is the Araar multidimensional inequality index, k is dimensions considered in the multidimensional inequality analysis, and ϕk is the weight attributed to each dimension. The parameter λk controls the sensitivity of the index to the intercorrelation between the dimensions of well-being; Ik is the relative inequality index of component k; and Ck is the absolute concentration index of component k. The index has a more flexible functional form in multi aspects of social preferences. It satisfies the main desirable properties and allows establishing a complete order for social welfare. The index has components which can be understood and is easily interpretable considering its functional form. Moreover, this index is multi-level decomposable by components or dimensions, and by the uni and multidimensional forms of inequality. MDI is quite sensitive to the choice of the parameter λ. Araar et al. (2009), state that the nature of the components used in the analysis determines the size of this parameter. If the components are perfect substitutes for the other set of components, it is appropriate to set the λ to zero. But if the components are a perfect complement then λ will converge to one. Setting λ = 0.5 probably leads to reasonable values in the multidimensional inequality measure. In this multidimensional inequality index we used a two-stage approach. In the first stage, we considered inequalities in living standards and then health and education as the other indicators of well-being and finally we estimated the multidimensional inequality index.

100

G. Tigre

5.7.5 Inequality Decomposition Inequality measures are often decomposed by population sub-groups such as regions, gender, or rural-urban groups to assess the extent and contribution of each group to total inequality. Inequalities within and between groups can also be used for assessing the major contributors to inequalities. In this inequality estimation, the overall observed inequality can be decomposed into within (W), between (B), and overlapping (L) components. The Gini decomposition can be formulated as (Heshmati 2004): Gini = Wi + Bi + L i =

n 

1  (y j − yi )Pi P j + L i μ i=1 j>i n

Gini i Pi πi +

i=1

n

(7)

where Ginii is the Gini coefficient of group I; Pi is the population share of the group (rural-urban, male-female or regions); πi is the income share of the total income in the region or group; μ is the mean income; and yi is the mean income of group i. We decomposed our analysis into rural-urban, gender based, and within and between regions inequalities (Table 5.3). Studies point out that rural-urban inequalities are a main contributor to total inequality and these inequalities are attributed to education and household residency (Charles 2011; Deaton et al. 2002). Table 5.3 Per capita consumption inequalities in of households using the Gini coefficient and Atkinson inequality indices Gini index

Atkinson index ε = 0.1

ε = 0.5

ε=1

Ethiopia

0.385

0.026

0.121

0.221

Urban

0.369

0.024

0.111

0.205

Rural

0.295

0.015

0.072

0.138

Tigray

0.402

0.028

0.131

0.237

Afar

0.360

0.023

0.104

0.188

Amhara

0.415

0.030

0.136

0.247

Oromia

0.358

0.023

0.105

0.195

Somali

0.310

0.019

0.084

0.154

Benishangul

0.375

0.024

0.112

0.206

Regions of the country

SNNP

0.375

0.025

0.115

0.211

Gambela

0.370

0.024

0.111

0.202

Harari

0.337

0.020

0.098

0.197

Addis Ababa

0.366

0.023

0.108

0.199

Dire Dawa

0.382

0.026

0.119

0.212

Source Author’s calculations using HICE data

5 Multidimensional Inequality in Ethiopia

101

5.7.6 Inequality Dominance A stochastic dominance is a useful and simple probabilistic concept that can be used for assessing inequalities. The most familiar graphical tool for examining inequalities in income or consumption is the Lorenz curve which is a plot of the cumulative fraction of the population starting from the poorest (on the X-axis) against the cumulative fraction of income or consumption (on the Y-axis). We are interested in the notion of Lorenz curve (L) which is given by: p

L = ∫ F −1 (q)dq, f or p ∈ [0, 1]

(8)

0

where F is the probability distribution function, a Lorenz curve that is closest to the 45 degree line is more equitable than a Lorenz curve which is far from the 45 degree line. In inequality sense distribution, F is preferred for the distribution G if and only if: L F ( p) > L G ( p) f or p ∈ [0, 1]

(9)

If we consider two countries, the Lorenz curve of country 2 will (first order) stochastically dominate that of country 1 if the Lorenz curve of country 2 is closer to the 45 degree line and everywhere above the curve of country 1 which implies that there are less inequalities in country 2 as compared to country 1.

5.8 Results and Discussion This study’s per capita consumption inequality analysis using HICE data and Gini and Atkinson indices showed that consumption per capita inequalities are quite high in Ethiopia (Gini = 0.385 and Atkinson index = 0.221, with epsilon = 1) (Table 5.3). Consumption per capita inequalities are higher in urban than in rural areas as per both the indices. There are regional consumption per capita inequality differences between regions (Table 5.3). Inequality of multidimensional inequality indicators decreases over the wealth quintiles in general; however, inequalities in asset ownership increase over the last wealth quintiles (Table 5.4). Education develops human capital and increases the productive capacity of labor. Education inequalities are high in Ethiopia in general and in Afar, Somali, and Amhara regions in particular (Table 5.5). Less educational inequalities are observed in Addis Ababa and Dera Dawa regions which are urban areas as compared to other regions. These differences may be a reflection of the differences in access to education across regions because of poor

102

G. Tigre

Table 5.4 The Gini index of multidimensional inequality indicators across different income groups Indicators

Poorest Quartile 1

Poorer Quartile 2

Middle Quartile 3

Richer Quartile 4

Richest Quartile 5

Overall Quartiles

Electricity access

0.988*** (0.002)

0.948*** (0.005)

0.904*** (0.007)

0.790*** (0.010)

0.070*** (0.004)

0.641*** (0.004)

Sanitation

0.983*** (0.002)

0.961*** (0.004)

0.947*** (0.005)

0.930*** (0.006)

0.781** (0.006)

0.902*** (0.002)

Water

0.837*** (0.006)

0.776*** (0.009)

0.748*** (0.010)

0.682*** (0.011)

0.146*** (0.005)

0.568*** (0.004)

Cooking fuel

0.999*** (0.000)

0.997*** (0.001)

0.997*** (0.001)

0.989*** (0.002)

0.678*** (0.006)

0.890*** (0.003)

Floor

0.990*** (0.001)

0.978*** (0.003)

0.936*** (0.006)

0.856*** (0.008)

0.202*** (0.006)

0.702*** (0.004)

Asset

0.648*** (0.007)

0.541*** (0.011)

0.405*** (0.011)

0.259*** (0.010)

0.450*** (0.007)

0.493*** (0.004)

Child mortality

0.002*** (0.000)

0.002** (0.001)

0.001 (0.001)

0.001 (0.001)

0.000 (0.000)

0.001 (0.000)

Nutrition

0.262*** (0.007)

0.212*** (0.009)

0.202*** (0.010)

0.199*** (0.009)

0.134*** (0.005)

0.197*** (0.003)

education

0.581*** (0.009)

0.426*** (0.010)

0.356*** (0.013)

0.292*** (0.015)

0.180*** (0.007)

0.422*** (0.005)

Child school attendance

0.493*** (0.007)

0.421*** (0.011)

0.402*** (0.011)

0.359*** (0.011)

0.159*** (0.005)

0.344*** (0.004)

Note * P < 0.1, ** P < 0.5, and *** P < 0.01 Source Author’s calculations using DHS (2016) data Table 5.5 Educational inequalities across regions of the country Regions in the country

Gini coefficient

Generalized entropy measure of inequality GE(0)

GE(1)

GE(2) 0.473

Tigray

0.371

0.167

0.174

Afar

0.644

0.311

0.358

1.851

Amhara

0.469

0.225

0.253

0.867

Oromia

0.413

0.194

0.188

0.483

Somali

0.573

0.220

0.170

0.545

Bensihangul

0.392

0.173

0.185

0.541

SNNP

0.366

0.155

0.137

0.301

Gambela

0.303

0.110

0.104

0.245

Harari

0.329

0.143

0.108

0.167

Addis Ababa

0.146

0.042

0.035

0.036

Dire Dawa

0.299

0.153

0.139

0.264

Ethiopia

0.422

0.185

0.175

0.479

Source Author’s calculations using DHS data

Cumulative proportion of years of education

5 Multidimensional Inequality in Ethiopia

103

1

.8

.6

.4

.2

0 0

.2

.4

.6

.8

1

Cumulative proportion of samples 45 line

Population

poorest

poorer

middle

richer

richest

Fig. 5.1 Lorenze curves based on years of schooling for different income quintiles

infrastructural development in the rural areas as compared to the urban areas. Educational inequalities also differ across wealth quintiles, high education inequalities are observed within the poorest households and less educational inequalities are observed within the richest households (Fig. 5.1). Health care is very important for productive activities. Health inequalities are less than living standard and education inequalities in Ethiopia (see Table 5.6) In rural Ethiopia, production and accumulation of wealth are highly associated with agricultural activities which in turn are related to inequalities in landholding (Charles 2011). High landholding inequalities are observed in SNNP (Gini = 0.603) followed by Tigray (0.578) and Somali (0.563) (Table 5.7). SNNP is known to be the most densely populated region in the country and small landholdings are expected here. Less agricultural landholding inequalities are observed in Amhara, Gambela, and Benshiangul regions. Gambela and Benshiangul regions are known to be less densely populated with less landholding inequalities (Gini = 0.42) next to Amhara region (Table 5.7). Agricultural landholding inequalities differ across wealth quintiles. Landholding inequalities are the highest for poor rural households compared to middle and rich rural households, especially in the higher percentiles (Fig. 5.2). Living standards contribute the most to multidimensional inequality except for some regions (Afar, Somali, and Addis Ababa) (Table 5.9). High contribution of living standards to multidimensional inequality inspired us to analyze inequalities in living standards. The living standard multidimensional inequality index (LSMII) showed that inequalities in living standards were very high in Ethiopia (0.642). The

104

G. Tigre

Table 5.6 Health inequalities across regions Regions in the country

Gini coefficient

Generalized entropy measure of inequality GE(0)

GE(1)

GE(2)

Tigray

0.197

0.118

0.096

0.086

Afar

0.213

0.126

0.107

0.095

Amhara

0.137

0.085

0.066

0.054

Oromia

0.139

0.086

0.067

0.055

Somali

0.147

0.091

0.071

0.058

Bensihangul

0.115

0.072

0.055

0.044

SNNP

0.093

0.058

0.044

0.034

Gambela

0.187

0.110

0.091

0.080

Harari

0.108

0.067

0.051

0.041

Addis Ababa

0.110

0.068

0.052

0.042

Dire Dawa

0.141

0.087

0.068

0.056

Ethiopia

0.145

0.089

0.070

0.058

Source Author’s calculations using DHS data

Table 5.7 Agricultural landholding inequalities of rural households across regions Regions

Gini

Generalized entropy measure of inequality GE(0)

GE(1)

GE(2)

Tigray

0.578

0.627

1.014

6.338

Afar

0.557

0.577

0.881

3.628

Amhara

0.393

0.273

0.290

0.468

Oromia

0.562

0.577

0.794

3.150

Somali

0.563

0.569

0.815

2.718

Bensihangul

0.417

0.307

0.318

0.458

SNNP

0.603

0.694

1.130

7.700

Gambela

0.424

0.337

0.349

0.563

Harari

0.556

0.556

0.773

2.732

Dire Dawa

0.434

0.348

0.354

0.528

Ethiopia

0.543

0.548

0.762

3.511

Source Author’s calculations using DHS data

inequalities were higher in rural areas (0.747) as compared to urban areas (0.342) (Table 5.8). Of the living standard’s indicators considered in the analysis, cooking fuel and sanitation contributed the most to LSMII (Table 5.8). Multidimensional poverty is quite high in Ethiopia (Getu 2018) but multidimensional inequality index (MII) is low (0.301) in Ethiopia (see Table 5.9).

5 Multidimensional Inequality in Ethiopia

105

1

.8 middel

cumulative land

.6 rich

.4 poor

.2

0 0

.2

.4

.6

.8

1

Percentiles (p)

Fig. 5.2 Lorenz curves of agricultural land holding of rural households Table 5.8 The living standards multidimensional inequality indix (LSMII) and contribution of each indicator to LSMII LSMII

Contribution of each indicator to LSMII (percentage) Electriccity

Sanitation

Water

Cooking fuel

Floor

Asset

Total

Ethiopia

0.642

16.15

20.79

13.62

22.29

17.60

9.56

100

Urban

0.342

3.20

33.22

5.49

28.82

9.38

19.89

100

Rural

0.747

18.63

18.60

14.71

19.46

18.87

9.73

100

Regions of the country Tigray

0.668

16.01

19.69

13.89

22.14

19.81

8.45

100 100

Afar

0.762

16.78

19.72

16.51

19.43

18.05

9.51

Amhara

0.765

17.03

19.80

14.17

19.77

19.03

10.20

100

Oromia

0.711

18.37

19.45

14.27

20.69

18.70

8.52

100

Somali

0.732

18.30

18.19

16.54

20.78

16.99

9.20

100

Benishangul

0.718

17.25

20.18

11.44

22.25

19.23

9.64

100

SNNP

0.732

18.35

17.17

15.63

21.73

18.30

8.81

100

Gambela

0.666

16.49

20.61

8.65

23.88

19.28

11.09

100

Harari

0.421

7.21

28.87

10.68

27.39

12.19

13.66

100

Addis Ababa

0.26

0.29

45.16

2.99

21.37

2.65

27.55

100

Dire Dawa

0.425

9.12

25.32

10.47

27.43

10.22

17.44

100

Source Author’s calculations using DHS data

106

G. Tigre

Table 5.9 The multidimensional inequality indix (MII) and contribution of each dimension to MII MII

Contribution of dimensions to MII (percentage) Living standared

Health

Education

Total

Ethiopia

0.301

51.61

6.09

42.30

100

Urban

0.168

51.01

11.08

37.91

100

Rural

0.303

50.61

6.11

43.12

100

Regions of the country Tigray

0.291

52.75

7.41

39.83

100

Afar

0.340

40.53

8.88

50.60

100

Amhara

0.343

53.38

4.70

41.92

100

Oromia

0.294

49.42

5.88

44.71

100

Somali

0.330

41.95

6.25

51.80

100

Benishangul

0.272

51.02

5.75

43.23

100

SNNP

0.286

56.06

3.77

40.17

100

Gambela

0.244

52.87

11.13

35.99

100

Harari

0.278

48.75

6.84

44.41

100

Addis Ababa

0.078

39.25

16.04

44.71

100

Dire Dawa

0.377

61.06

7.22

31.72

100

Source Author’s calculations using DHS data

Access to electricity and water contributed less to LSMII in urban areas. But the contribution of assets to LSMII was less (9.75%) in rural areas. Policies aimed at reducing inequalities in living standards should focus on cooking fuel, access to electricity, floor material, and sanitation. Multidimensional inequality indices are quite sensitive to the choice of parameter λ. By giving parameter λ different values (Table 5.10) the estimated results differ considerably for different values of λ (for example, λ = 0.1, 0.3, 0.5, 0.7, 0.9) but they follow the same pattern in each range. In other words, as the value of λ increases, the multidimensional inequality index also increases for all regions considered in the index. This range represents the most applicable range for the multidimensional inequality index (Table 5.10). Table 5.10 Multidimensional Inequality Estimates—Robustness check Multidimensional inequality index (MDII)

Selected regions of the country Tigray

Amhara

Oromia

SNNP

Addis Ababa

MDII (λ = 0.1)

0.250

0.306

0.258

0.256

0.030

MDII (λ = 0.3)

0.259

0.316

0.268

0.265

0.034

MDII (λ = 0.5)

0.268

0.325

0.278

0.274

0.037

MDII (λ = 0.7)

0.277

0.335

0.287

0.282

0.041

MDII (λ = 0.9)

0.286

0.344

0.297

0.291

0.044

Source The Author

5 Multidimensional Inequality in Ethiopia

107

The results of the consumption decomposition show that the incidence of consumption poverty is higher in rural than in urban Ethiopia (Table 5.11). But inequalities among urban households are greater than in rural households. Within-group consumption inequalities, as calculated by the Gini coefficient, dominate the betweengroups inequalities (rural-urban) (Table 5.11). This shows that if the government or policymakers were to target consumption differences within groups, this could help in Table 5.11 Poverty (FGT) and inequality (Gini) indices and marginal impacts and elasticities (by groups) Groups

Population share

Poverty (FGT)

Inequality (Gini)

Marginal impact on inequality

Marginal impact on poverty

Elasticity

Urban

0.764

0.175

0.361

0.282

0.412

1.966

Rural

0.236

0.666

0.316

0.033

−0.000

−0.019

Within

–

–

0.252

0.316

0.411

1.757

Between

–

–

0.118

0.082

0.115

1.905

Total

1.000

0.291

0.393

0.393

0.491

1.685

Gender based decomposition Male-headed hh

0.811

0.315

0.399

0.318

0.388

1.642

Female-headed hh

0.189

0.186

0.355

0.068

0.096

1.871

Within

–

–

0.268

0.387

0.484

1.683

Between

–

–

0.028

0.003

0.005

2.448

Total

1.000

0.291

0.393

0.393

0.491

1.685 1.606

Region based decomposition Tigray

0.069

0.319

0.418

0.032

0.038

Afar

0.038

0.196

0.361

0.014

0.020

2.000

Amhara

0.163

0.276

0.382

0.061

0.075

1.652

Oromia

0.226

0.330

0.393

0.084

0.113

1.811

Somali

0.039

0.255

0.370

0.013

0.015

1.553

Benishangul

0.049

0.284

0.389

0.019

0.022

1.581

SNNP

0.155

0.412

0.420

0.061

0.060

1.340

Gambela

0.053

0.360

0.407

0.021

0.025

1.642

Harari

0.020

0.195

0.352

0.007

0.011

2.214

Addis Ababa

0.168

0.150

0.355

0.062

0.084

1.823

Dire Dawa

0.020

0.305

0.384

0.007

0.011

2.139

Within

–

–

0.055

0.380

0.476

1.685

Between

–

–

0.061

0.007

0.013

2.416

Total

1.000

0.291

0.393

0.393

0.491

1.685

Source Author’s calculations using HICE data

108

G. Tigre

reducing overall consumption inequalities more than if they target consumption differences between groups. Urban households will benefit more because the marginal impact on inequalities is higher for urban households than their rural counterparts. Reducing inequalities between groups (rural-urban) will have more impact on reducing poverty than reducing inequalities within groups (households) as between-group elasticity is greater than within-group elasticity (Table 5.11). Gender based decomposition results (Table 5.11) show that the incidence of consumption poverty is high for male-headed households as compared to female-headed households. In addition, inequalities among male-headed households are greater than those in female-headed households. If income is used, the results might be different. Research also indicates that men earn more than women as there is gender based discrimination in developing countries. Men earn more than women but women manage consumption expenditure better than men. Within-group inequalities, the marginal impact of inequalities and the marginal impact of the poverty component register larger values than between-group components (Table 5.11). Therefore, reducing the average number of deprived households among male-headed and female-headed households will reduce overall deprivation more than reducing deprivation between these two groups. Region based decomposition results show that between-group inequalities are greater than within-group inequalities. This means that between regions inequalities are more than within region inequalities, so there are differences in inequality between regions in Ethiopia which need to be considered. There are educational inequalities in Ethiopia and various factors contribute to these. One of the factors assumed to affect children’s level of education is parents’ level of education. Educated parents have better knowledge and understanding of the benefits of education and would like to educate their children as compared to uneducated parents. Our analysis showed that parents’ education had a positive impact on children’s education. Mothers’ education contributed more both to sons’ and daughters’ education than fathers’ education, other factors being controlled for (Table 5.12). This result is consistent with the saying that educating a mother (woman) is educating a family. Therefore, educating daughters (tomorrow’s mothers) has more positive intergenerational inequality reducing effects than educating sons.

5.9 Conclusions Consumption inequalities are high in Ethiopia and these are higher in urban as compared to rural areas in the country. The results of a regional comparison show that there are regional consumption per capita inequality differences between regions which need to be carefully considered both by federal and regional governments. Inequalities in multidimensional indicators over the wealth quintiles are quite high in Ethiopia but the inequalities of these indicators decrease over the wealth quintiles. Educational inequalities in rural areas are higher than that in urban areas. These differences may be because of poor infrastructure development in rural areas as compared

5 Multidimensional Inequality in Ethiopia Table 5.12 Education regression results (by gender)

109

Variables

Son’s education

Daughter’s education

Children (Total samples)

Father education

0.520**

0.539***

0.529***

Mother education

0.551***

0.572***

0.561***

Family size

0.044***

0.027

0.037***

Children under 5

−0.209***

– 0.271***

−0.237***

Child age

0.217***

0.234***

0.223***

Have a bank account

0.219***

0.143*

0.187***

Rural(urban is base)

−0.429***

– 0.491***

−0.469***

Head’s age group 30–39

0.518***

0.428***

0.477***

40–49

0.715***

0.698***

0.712***

50–59

0.864***

0.857***

0.868***

60–69

0.759***

0.750***

0.759***

70+

0.560***

0.484**

0.522***

Regions (Tigry-base) Afar

– 0.523***

– 0.813***

– 0.667***

Amhara

– 0.619***

– 0.512***

– 0.577***

Oromia

– 0.542***

– 0.770***

– 0.663***

Somali

– 0.175

– 0.565***

– 0.365***

Benishangul

– 0.320***

– 0.594***

– 0.455***

SNNP

– 0.308***

– 0.762***

– 0.518***

Gambela

0.207

– 0.691***

– 0.221**

Harari

– 0.250

– 0.864***

– 0.537***

Addis Ababa

– 0.213

– 1.214***

– 0.668***

Dire Dawa

– 0.182

– 0.641***

– 0.399***

_cons

0.166

0.545***

0.373***

Note * P < 0.1, ** P < 0.5, and *** P < 0.01 Source Author’s calculations using DHS data

to urban areas. Developing schools in different areas and at different levels and other infrastructure are very important for improving access to education thus reducing existing educational inequalities. Rural areas and poor households require special attention for reducing the existing educational inequalities in the country. A large proportion of the Ethiopian population lives in rural areas and is engaged in agriculture. Landholding inequalities are the highest for poor rural households compared to

110

G. Tigre

middle and rich rural households. Agricultural transformation is thus important for improving smallholder farmers’ productivity and for developing manufacturing and other industries to create jobs for these rural smallholder agricultural households. There are considerable living standard multidimensional inequalities in Ethiopia. If we want to reduce living standard inequalities, cooking fuel and sanitation facilities have to be improved. Policies aimed at reducing living standard inequalities should focus on cooking fuel, access to electricity, and sanitation and floor material. Federal and regional governments’ aim of reducing consumption inequalities should focus on within-group inequalities than on between-group inequalities. Reducing inequalities between groups (rural-urban) will have more impact in reducing poverty than reducing inequalities within groups (households) as between-group elasticity is greater than within-group elasticity. Consumption poverty and inequalities are high for male-headed households as compared to female- headed households. Within-group inequalities, the marginal impact of inequalities, and the marginal impact of poverty component register larger values than between-group components. Therefore, reducing the average number of deprived households among male-headed and female-headed households will reduce overall deprivation more than reducing deprivation between these two groups. Region based decomposition results show that between-group inequalities are more than within-group inequalities which means that inequalities are higher between regions than within regions; there are differences in inequalities between regions in Ethiopia which require careful consideration. Parental education has a positive impact on children’s education and educated mothers have a stronger influence on their children’s education. In Ethiopia because of social and cultural reasons girls are marginalized and have less access to education than boys. Giving girls’ more access to education and encouraging them to get educated will have more intergenerational educational and multidimensional inequality reducing effects.

5.10 Recommendations Based on a theoretical review of inequality literature and an empirical analysis of inequality, the following recommendations are made: Attention should be paid to improving living standards in Ethiopia in general and among poor and rural people in particular for reducing the existing living standard inequalities in the country. Land inequalities are high in rural Ethiopia and increasing land productivity through professional support systems and modernizing the farming system will contribute to reducing the existing inequalities in the country. Parallel to this, increasing industrialization in urban areas for creating job opportunities both for urban and rural growing populations could absorb the excess labor force in the agricultural sector. In a country like Ethiopia where girls are marginalized because of social and cultural reasons increasing girls’ access to education will reduce intergenerational differences in inequality.

5 Multidimensional Inequality in Ethiopia

111

There should be a combined effort to reduce poverty, vulnerability to poverty, and inequalities. Reducing inequalities in the country will enhance poverty and vulnerability to poverty reduction efforts. Inequality reduction also significantly affects the poverty gap. Poverty, vulnerability to poverty, and inequality research has become extremely important for Ethiopia. There are some difficulties in measuring these problems: (1) availability and reliability of up to date panel and cross-sectional data, and (2) the consistency and comparability of such data over time. Despite these measurement difficulties and data problems, poverty, vulnerability to poverty, and inequality measures remain useful for gaining some understanding of the severity of the problem and forwarding possible recommendations.

References Alkire, S., and J. Foster. (2011). Counting and multidimensional poverty measurement. Journal of Public Economics, 95(7–8): 476–487. Anderson, G. (2004). Indices and tests for multidimensional inequality: multivariate generalizations of the Gini coefficient and Kolmogorov–Smirnov two sample test. University of Toronto, Department of Economics. Araar, A. (2009). The Hybrid Multidimensional Index of Inequality. Cahiers de recherche 0945. Retrieved from: http://ideas.repec.org/p/lvl/lacicr/0945.html. Atkinson, A. B. (1970). On the measurement of inequality. Journal of Economic Theory, 2(3): 244–263. Bakare, A. S. (2012). Measuring the income inequality in Nigeria : the Lorenz Curve and Gini Co-efficient approach. American Journal of Economics, 2(1): 47–52. Bigotta, M., J. Krishnakumar, and U. Rani (2014). Further results on the regression based approach to inequality decomposition with evidence from India. Empirical Economics: 1–34. Bosmans, K., K. Decanq, and E. Ooghe. (2015). What do normative indices of multidimensional inequality really mean? Journal of Public Economics, 130: 94–104. Bourguignon, F., and C. Morrisson. (2002). Inequality among world citizens: 1820–1992. American Economic Review, 92(4): 727–744. Cain, J., R. Hasan, and R. Magsombol (2012). Inequality and Poverty in Asia. Poverty, Inequality, and Inclusive Growth in Asia. J. Zhuang, Anthem Press and ADB: 35–85. Charles, J. (2011). Understanding income inequality: concept, causes and measurement. International Journal of Economics and Management, 1(3): 17–28. Checchi, D. (2000). Does Educational Achievement Help to Explain Income Inequality? UNU/WIDER Working Paper no. 208, UNU-WIDER, Helsinki. CSA. (2016). Ethiopia Demographic and Health Survey 2016. Addis Ababa, Ethiopia, and Rockville, Maryland, USA: CSA. Deaton, A., and J. Dreze. (2002). Poverty and Inequality in India: A Re-Examination. Economic and Political Weekly, 37(36): 3729–3748. Decancq, K., and M. A. Lugo. (2012). Inequality of wellbeing: a multidimensional approach. Economica, 79(316): 721–746. Duclos, J., D. Sahn, and S. D. Younger. (2001). Robust Multidimensional Poverty Comparisons. Working Paper 0115, Université Laval, Département d’Économie, Québec. Fentaw, A., and A. Rao. (2016). Determinants of income inequality in urban Ethiopia: A study of South Wollo Administrative zone, Amhara National Regional State. International Journal of Applied Research, 2(1): 550–563.

112

G. Tigre

Gakidou, E., C. Murray, and J. Frenk. (2000). Defining and Measuring Health Inequality: An Approach Based on the Distribution of Health Expectancy. Bulletin of the World Health Organisation, 78(1): 42–54. Gallardo, K., L. Varas, and M. Gallardo. (2017). Inequality of opportunity in health: evidence from Chile. Rev Saude Publica, 51: 110. Gelow, F. (2009). The Relationship between poverty, inequality and Growth in the rural Ethiopia: micro ev-idence. Paper prepared for presentation at the International Association of Agricultural Economists Confer-ence, Beijing, China, August 16–22. Getu, T. (2018). Multidimensional Poverty and Its Dynamics in Ethiopia, in A. Heshmati and H. Yoon (eds.). Economic Growth and Development in Ethiopia. Perspectives on Development in the Middle East and North Africa (MENA) Region. Singapore: Springer, 161–195. Heshmati, A. (2004). The World Distribution of Income and Income Inequality. MTT Economic Research, Agrifood Research Finland. Discussion paper 2004/9. Heshmati, A. (2014). Income versus consumption inequality in Korea: evaluating stochastic dominance ranking by various householde attributes. Asian Economic Journal, 28(4): 413–436. Idrees, M., and E. Ahmad (2010). Measurement and Decomposition of Consumption Inequality in Pakistan, 2(Winter), 97–112. Kedir, A., and A. Oterova. (2014). Parental Investments on Girls, Post-marital residence and motives for private transfers, econometric evidence from Ethiopia, mimeo. Kolm, S. C. (1977). Multidimensional egalitarianism. The Quarterly Journal of Economics, 91(1): 1–13. Koshevoy, G., and K. Mosler. (1996). The Lorenz Zonoid of a multivariate distribution. Journal of the American Statistical Association, 91(434): 873–882. Maasoumi, E. (1986). The measurement and decomposition of multidimensional inequality, Econometrica, 54(4): 991–997. Maasoumi, E. (1999). Multidimensioned approaches to welfare analysis, in J. Silber (ed.), Handbook of income inequality measurement. London: Kluwer Academic. McLanahan, S., and P. Christine. (2008). Family structure and the reproduction of inequalities. Annual Review of Sociology, 34(1): 257–276. Meschi, E., and F. Scervini. (2014). Expansion of schooling and educational inequality in Europe: the educational Kuznets curve revisited. Oxford Economic Papers, 66(3): 660–680. MoFED. (2014). Development and Poverty in Ethiopia—1995/96–2010/11. Addis Ababa: Ministry of Fi-nance and Economic Development. Morrisson, C., and F. Murtin. (2013). The Kuznets curve of human capital inequality: 1870–2010. Journal of Economic Inequality, 11(3): 283–301. Ostry, J., A. Berg, and C. Tsangarides. (2014). Redistribution, inequality and growth, IMF Staff Discussion Note, Washington, DC. Ribero, R., and J. Nunez. (1999). Productivity and the Household Investment in Health—The Case of Colombia. Inter-American Development Bank, Working Paper R-354. Roemer, J. (2014). Economic development as opportunity equalization. World Bank Economic Review, 28(2): 189–209. Scott, J., and G. Marshall. (2009). Oxford Dictionary of Economics. Oxford: Oxford University press. Stephen, D., and E. David (2007). Inequality and poverty in Africa in an Era of golobalizaion: looking beyond income to health and education. World Institute for Development Economic Research. Stiglitz, J. E., A. Sen, and J. P. Filtoussi. (2009). Report by the Commission on the Measurement of Economic Performance and Social Progress, mimeo. Tesfay, N., and L. Malmberg. (2014). Horizontal inequalities in childrens educational outcomes in Ethiopia. International Journal of Educational Development, 39: 110–120. The World Bank. (2005). World Development Report 2006: Equity and Development. Washington, DC: The World Bank and Oxford University Press. Theil, H. (1967). Economics and Information Theory. Amsterdam: North Holland.

5 Multidimensional Inequality in Ethiopia

113

Theil, H. (1979). World income inequality and its components. Economics Letters, 2: 99–102. Tranvag, E., M. Ali, and O. Norheim. (2013). Health inequalities in Ethiopia: modeling inequalities in length of life within and between population groups. International Journal for equality in health, 12(52). Vida, C., and C. Jonas. (2008). Economic ineqaulity: measurment and the analysis, Economic and Rural Development, 4(1): 7–14. Weymark, J. A. (2006). The normative approach to the measurement of multidimensional inequality, in F. Farina and E. Savaglio (eds.). Inequality and Economic Integration. London: Routledge, 133–161. Woldehanna, T., J. Hoddinott, and S. Dercon. (2008). Poverty and inequality in Ethipia: 1995/96 – 2004/05. Addis Ababa: Department of Economics, Addis Ababa University.

Chapter 6

Factors Influencing Consumers’ Preferences for Cement Products: Case of Cement Brands in Ethiopia Yared Tadesse and Workneh Kassa Tessema

Abstract Understanding factors that influence consumer preferences for a product is vital for marketers for developing appropriate marketing strategies and winning over competition. But measuring the impact of each factor on consumer preferences is difficult due to consumers’ trade-off between attributes and other factors. The main objective of this study is identifying the factors that affect consumer preferences for different cement brands in Ethiopia. The study uses the conjoint analysis technique to analyze consumer preferences for cement products. This is a multivariate technique that is used for understanding consumer preferences for a product from a bundle of the product’s attributes and other factors. The study also uses the random sampling technique to select respondents from individual home builders as consumers of different brands. The sample size is 100. The variables are preference (dependent variable), price, quality, packaging, service, external influencers, advertising, and socio-demographic and economic factors like age, sex, and income. The findings of the study show that service, price, quality, income level, and gender of the respondents significantly influence consumer preferences for cement products. Among product attributes service is the main attribute that influences consumer preferences followed by price and quality. In terms of profile preferences, the product profile with lower than average market price, being as per the standard quality, using paper bags for packaging, and good service have the highest utility value. This indicates that consumers prefer a specific bundle of attributes in cement. Thus, marketers of cement product need to carefully organize attributes that are preferred more by consumers. Keywords Consumer preference · Cement product · Conjoint analysis · Ethiopia JEL Codes M11 · M31 · M38 · M16 · L74 · D11 · C12 Y. Tadesse Express Management Consulting PLC, Addis Ababa, Ethiopia e-mail: [email protected] W. K. Tessema (B) Department of Management, Addis Ababa University, Addis Ababa, Ethiopia e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 J. Wood and O. Habimana (eds.), A Multidimensional Economic Assessment of Africa, Frontiers in African Business Research, https://doi.org/10.1007/978-981-15-4510-8_6

115

116

Y. Tadesse and W. K. Tessema

6.1 Introduction With growing competition among different cement brands, cement companies in Ethiopia have to understand the factors that influence consumer preferences so that they can develop appropriate marketing strategies to win over competition. Availability of various branded cement products in the market also creates complications for consumers in their preferences and purchase decision processes (for example, Maity 2014). Thus, a correct understanding of the needs and preferences of customers by marketers is crucial for maneuvering the competitive environment and targeting customer segments (Brassington and Pettitt 2003). Consumer preferences for products involve the ranking of goods and services according to how much benefit the products provide them (for example, Udomkun et al. 2018). As a consequence, consumer preferences for a product are seen as a reflection of their inner views and perceptions (Cao and Ramani 2010). When deciding to purchase a product, consumers evaluate the attributes of all the competing products before making a final choice (Basavaraj et al. 2015).Consumers have expectations from branded cement products with regard to quality, instant drying, price, packaging, distribution, and delivery methods (Amutha and Vinayak 2015). Since different brands of cement products are increasingly available in the market, consumers might give due attention to certain features while making their preferences. This study analyzes the factors that influence consumer preferences for cement products. To understand the factors and product attributes that influence consumer preferences, the study uses the conjoint analysis method (for example, Green and Srinivasan 1990).

6.1.1 Background of the Cement Industry in Ethiopia Cement production has a long history in Ethiopia where the first cement plant, Dire Dawa Cement and Lime Factory, was built by the Italians in Dire Dawa in 1936 (National Cement Share Company 2012). Its capacity at that time was 90 tons of cement per day and the production was mainly used for constructing infrastructure. Later, in 1964 and 1965 Masawa (then part of Ethiopia) and Addis Ababa cement factories were inaugurated respectively with individual production capacity of 200 tons per day (National Cement 2011). After the 1960s, it took Ethiopia more than 20 years to add cement industries to its economy and two more cement industries, Mugger cement and Mossobo were inaugurated in 1983 and 1997 respectively. Following the national economic reforms in 2001, the development of all the sectors picked up pace with a corresponding national economic growth rate of more than 10% (African Development Bank Group 2014). For example, the annual report of the Ethiopian Ministry of Finance and Economic Development (2010–11) showed an average growth rate of 11.4% for the last decade. The construction sector accounted for a 13.6% growth rate and this drastic economic growth raised demand for cement in the country (MoFED 2010). Due to the increase in demand for cement the country

6 Factors Influencing Consumers’ Preferences for Cement Products …

117

had to import and in 2011 one million tons of cement were imported (Bloomberg 2011). The boom in the demand for cement products also attracted new cement producers to invest in the sector and existing cement factories to go in for further expansion. As per the country’s Ministry of Industry report, 12 million tons of cement was produced in the 2012–13 budget year but the actual annual demand in the same year was 5.4 million tons (All Africa 2014). Even though this difference had a positive impact on consumers by bringing down market prices from 480 birr per 100 kg to 200 birr per 100 kg (Capital Ethiopia 2012), it posed a serious challenge for cement producers, as competition among them intensified. Today, cement industries in Ethiopia have started exporting cement to neighboring countries to compensate for the demand gap in the local market (Walta Information Center 2014). By 2013–14 Ethiopia had got more than US dollar 9 million from cement exports (Walta Information Center 2014). Cement industries also developed different strategies to win over competition in the local market and started finding out consumer preferences. Currently, there are about 18 cement industries in the country with an aggregate production capacity of 12.12 million tons per annum (Capital Ethiopia 2012). Most of the industries are located within 100 km around Addis Ababa, except Mossobo in the north and three plants in Dire Dawa.

6.1.2 Problem Statement Following the country’s economic growth and government policies many foreign and local companies joined the Ethiopian cement industry and this brought an unexpected gap in production and demand. This gap was a challenge for cement industries and competition among them became very stiff (The Reporter 2015). Due to this, aggressive promotional campaigns on electronic media about the industry’s offerings, repetitive price discount schemes, change in packaging material, and introduction of different delivery schemes became a common practice in the cement industry to find our consumer preferences (The Reporter 2015). These different marketing efforts by cement companies were aimed at influencing buyers. During the purchase process, consumers form their preferences based on the best combination of attributes by evaluating different products (Felipe et al. 2012).In today’s highly competitive markets, doing marketing activities without identifying the factors and product attributes that influence consumer preferences may not be effective in meeting their objectives (for example, Kotler et al. 2009). In a rapidly changing market, demand for a product is often influenced by consumer preferences (Fornell 1992). Thus, a major concern for many companies is identifying and managing consumers’ demands and preferences for different products (for example, Kotler et al. 2009). Therefore, identifying the major factors and attributes that affect consumer preferences and buying decisions in the cement industry is helpful in developing appropriate marketing strategies for companies which are entering the Ethiopian cement market.

118

Y. Tadesse and W. K. Tessema

However, even though some research on cement markets in other countries exists (Kumar 2013; Maity 2014; Murugan and Ganapathi 2012), these studies do not address all the factors fully. For example, Kumar (2013), focuses on cement industries’ marketing mix strategies and Maity (2014), focuses on the impact of external influences on demand for cement. Maity (2014), found that masons and engineers had a significant impact on the cement purchasing behavior of individual home builders and non-trade customers. Murugan and Ganapathi (2012), studied the cement market by comparing consumer preferences for selected brands. Murugan and Ganapathi’s (2012) findings show that price, quality, instant drying, availability, weight, packaging, credit facility, door delivery, and nearness (proximity) influenced consumer preferences. A study conducted in Ethiopia by Sisay (2012) focused mainly on market structure, conduct, and performance of cement industries. It showed that advertising and market share had a direct positive relationship as industries with a higher market share had more advertising expenditure. But the study fails to show how advertising influenced consumer preferences for cements products. Thus, so far there are no empirical studies on the factors that influence consumer preferences for cement products in Ethiopia. Therefore, this research on the Ethiopian cement market identifies: (a) factors that influence consumer preferences for cement products, (b) product attributes that have a major influence on consumer preferences for different cement brands, and (c) product attributes that influence consumer preferences for cement products. The next sections of this study are organized as follows. Section 2 presents a literature review which is followed by the methodology of the study in Sect. 3. The empirical results and discussion of the study are presented in Sect. 4. Finally Sect. 5 gives the conclusion and recommendations of the study.

6.2 Literature Review 6.2.1 Consumer Preference According to Cao and Ramani (2010), a consumer’s preference for a product can be viewed as a reflection of his or her inner world. Hence, it is consumers’ attitudes and perceptions of a product or company which determine their preferences. Consumer preferences involve the ranking of goods and services according to how much benefit they provide. According to Nithiyananth and Arunpandian (2012), consumer preferences are the process of how consumers choose a commodity and service in relation to its attributes like taste and price, and individual factors like individual choices and incomes. Consumers evaluate the perceived benefits of a product before they take a decision to purchase it (Kotler et al. 2009). Kotler et al. (2009) also included consumption costs, opportunity costs of other offerings, and switching costs from other brands. Consumers value these and other added benefits when making a purchasing decision (Kotler et al. 2009).

6 Factors Influencing Consumers’ Preferences for Cement Products …

119

If consumers like a product, the company can exist for years and sell the product in millions. On the other hand, if consumers do not like its product a company can close down quickly. Therefore, it is important for companies to understand consumers’ needs when marketing their products (Pullman et al. 2002). Through knowledge about consumer preferences and their values, companies can design and implement appropriate marketing strategies to increase customer satisfaction, loyalty, and retention to enhance their competitive positions. Today it is impossible to remain cost competitive and offer every feature desired by consumers (Pullman et al. 2002) because the influence of various environmental, social, personal, and marketing factors, consumer preferences have a trade-off with a bundle of attributes that a product comes with (for example, Kotler et al. 2009). Issues that consumers consider during their purchase decisions make an analysis of consumer preferences for a product difficult as one cannot simply take each attribute to be weighted by consumers for marketers (Kanetkar 2005).

6.2.1.1

Brand Awareness and Preferences

The American Marketing Association (2018:1), defines a brand as “a name, term, sign, symbol or design, or a combination of them intended to encourage prospective customers to differentiate a producer’s product (s) from those of competitors.” Raza (2011) also defines a brand as a name, term, design, symbol, or any other feature that identifies one seller’s goods or services as distinct from others. A brand also encompasses a set of expectations associated with a product or service which typically arise in the minds of the buyers (Ferree 2011). Consumers find different features in a brand after using the product. These features affect consumer behavior such as joy of planning a purchase and the real purchase (Raza 2011). Brand awareness is an important indicator of consumers’ knowledge about a brand, the strength of a brand’s presence in consumers’ minds, and how easily that knowledge can be retrieved from memory while taking purchase decisions (for example, Dhurup et al. 2014). Brand awareness is the probability that consumers will easily recognize the existence and availability of a company’s product or service (Dhurup et al. 2014). It is generally assumed that higher brand awareness brings brand equity to the company (for example, Dhurup et al. 2014; Kotler et al. 2009).

6.2.1.2

Factors Influencing Preferences for Cement

Chimboza and Mutandwa (2007), identified four factors as key determinants of product choice: promotion, price and availability of a product, attractive packaging, and product quality. According to Prakash (2011), quality, price, and advertising influence customers switching from one brand to another. A study on Dangote cement by Nartey (2013) found that the main factors considered by consumers before deciding to buy Dangote cement’s products were cost, quality, and accessibility of the product. Murugan and Ganapathi (2012), identified nine factors for measuring consumer

120

Y. Tadesse and W. K. Tessema

preferences for different cement brands: price, quality, instant drying, availability, weight, packaging, credit facility, door delivery, and nearness (proximity). Existing research (see Blum 2007; Ghafran et al. 2014; Kumar 2013; Maity 2014; Rajaguru and Matanda 2006) has shown that product quality, packaging, price, promotion, convenience, and accessibility are factors that influence consumers’ preferences for different cement brands.

6.2.1.3

Product Quality

Quality is an important component of a company’s brand identity (for example, Nartey 2013). As a result, consumers’ preference decisions are highly influenced by the level of information about the product’s qualities associated with certain brands (for example, Soegoto 2018). Quality refers to the degree to which a set of inherent product characteristics fulfill user requirements (ISO 9000, 2015). If a product fulfills consumers’ expectations, they will be pleased and the product will be seen as acceptable or even having high quality (Jahanshahi et al. 2011).Cement is a standardized homogeneous product in which most industries meet the minimum quality standards (Blum 2007). Due to similar raw material inputs and production processes, there is no significant difference in the cement produced across firms (Maity 2014). The cement industry does not have a large product mix and the same applies for various cement companies. The product mix can be classified on the basis of the type of cement available. Quality is still one of the factors which consumers take into account when choosing between different cement brands (Kumar 2013). But when it comes to cement products, different countries have their own standards of physical properties and quality levels that the cement products need to fulfill (Tennis 1999). In Ethiopia, the Ethiopian Standards Agency has set the minimum mandatory standards of cement specifications for each type of cement. No company is allowed to produce and sell cement that is below these standards. Failing to meet the minimum standards for cement product leads to the cancellation of the producer company’s working permits (Ethiopian Standards Agency 2015). As a result, there is no visible or big difference in the quality of cement in the country. But there are some variations in the consistency within the standards.

6.2.1.4

Packaging

Due to the bulky nature of the product and its sensitivity to external factors like rain or humidity it needs to be transported correctly. Once bagged, cement becomes a perishable commodity, so packaging and transportation are key activities. Transportation costs account for approximately 25% of the total costs for cement products (Kumar 2013). According to Kumar (2013), visual attractiveness is not a major concern for cement since it is usually dusty. Thus, unlike other products it does not require any specialized designing of product packaging to attract customers. The major concern

6 Factors Influencing Consumers’ Preferences for Cement Products …

121

is ease of transportation, storage, and minimizing pilferage (Kumar 2013). The packaging should not allow water to enter in as cement hardens when it comes in contact with water. The sacks should be light weight as the product is bulky. Even today the most preferred medium of packaging is bags (Kumar 2013). As per our observations of the Ethiopian cement market, paper bags and PP bags are used for packing 50 kg cement.

6.2.1.5

Price

The price is the amount a customer pays for a product. Price can be perceived in terms of the specific monetary value that a customer attaches to a good or service (Kent and Omar 2003). Buyers’ view of a product’s price as high, low, or fair ultimately affects their willingness to buy the product (Ahmad and Vays 2011). According to Macdonald and Sharp (2000), price becomes a reason for brand choice either by going for the lowest price to avoid a financial risk or the highest price for getting better quality. However, according to Kumar (2013), price in the cement industry is used only as a differentiator between various competing brands. The prices of different brands in the same segment remain more or less similar. Sometimes price also varies with the order placed by a customer based on whether the order placed is in trade or in non-trade. Pricing decisions in the cement industry largely depend on the price of energy and raw material inputs like coal, excise duties and taxes, and general operating profits (Kumar 2013). Generally, an increase or decrease in prices affects all brands in the market.

6.2.1.6

Promotion

Consumers’ buying behavior can be motivated through various promotional channels including promotional techniques like free samples and price discounts (Ghafran et al. 2014). In this respect, promotion plays a big role in creating brand awareness among consumers and it has a positive effect on brand equity (Karunanithy and Sivesan 2013). Sisay’s (2012) study of the cement sector showed that advertising and market share had a direct positive relationship. According to Sisay’s study, industries with a higher market share had higher advertising expenditure. The same study also noted that advertising was capable of creating a differentiated product for cement producers.

6.2.1.7

Convenience

Convenience of a brand has a significant effect on consumers’ preference for a particular product (Lin and Chang 2003). For example, easy access to a brand/product in a store is vital when buying low involvement products (Kotler et al. 2009). Rajaguru and Matanda (2006), show that store attributes such as service quality, convenience of store, and product attributes such as product quality, price, and availability of new

122

Y. Tadesse and W. K. Tessema

products were significant for customer loyalty. Arokiasamy (2012), found that store image and distribution intensity had a significant positive influence on brand loyalty. But according to Schramm-Klein et al. (2008), the issue of convenience varied based on the strength of the brand. In this respect, convenience was not a basic factor that passes through retailing channel for strong brand items, but it does help out weaker brands (Schramm-Klein et al. 2008). In other words, the product brand is less conditional on its retail channel convenience; rather the product characteristics themselves exert a higher influence on brand loyalty (Schramm-Klein et al. 2008).

6.2.1.8

Influence of External Factors on Consumer Preferences

Regarding external influencers of consumer preferences for cement, masons and engineers have a significant impact on the cement purchasing behavior of individual home builders and non-trade customers (Maity 2014). Consumer preferences for cement products are highly influenced by masons, contractors, engineers, distributors, and retailers (Kumar 2013).

6.3 Methodology 6.3.1 Description of the Study Area This study was conducted in three cities of Ethiopia, Adama, Dire Dawa, and Jigjiga. These three cities were selected for the following reasons. First, like many regional cities in Ethiopia, a lot of construction activities are going on here. Secondly, there are different brands of cement in the markets in the three cities. Thirdly, the three cities also have exposure to imported cement products since they are closer to international borders leading to better consumer exposure to different cement products. Thus, it is assumed that consumers have long experience in buying cement and testing different brands in the selected cities.

6.3.2 Data Collection The study used the survey method using a structured questionnaire. The questionnaire was divided into three sections: background, other factors, and conjoint profiles. In the other factors and conjoint sections, data was gathered using a five point Likert type scale. Following Kambewa (2007), for other factors we used (1 = strongly disagree and 5 = strongly agree); and for the conjoint analysis (1 = least preferred and 5 = most preferred).

6 Factors Influencing Consumers’ Preferences for Cement Products …

123

6.3.3 Sampling and Determining the Sample Size A systematic random sample of 100 home builders was selected in expansion villages (villages where construction work was going on during the data collection) in the three cities. Villages that were expanding in the three cities were randomly selected and a sample in each village was taken randomly where the first home in a village was selected through lottery and the next home was selected by n + 2 from those homes which were under construction and had to be completed within six months. The allocation of samples for the cities was computed on the basis of CSA’s (2008) summary report on population Census estimates for the cities; 38 for Adama, 40 for Dire Dawa, and 22 for Jigjiga. All the 100 questionnaires for individual home builders were collected on time. If owners were not available at home during data collection those who were available and their addresses were found in their working areas were considered as a part of the sample.

6.3.4 Data Analysis Techniques For the data analysis, descriptive statistics and the ordinary least squares (OLS) regression were used for identifying the major factors and attributes that affected consumer preferences. A conjoint analysis technique was applied for determining the dependent variable ‘preference’ for product attributes and profiles which were preferred by cement consumers. For other factors not included in the conjoint analysis such as external influences, promotions, and availability, a descriptive statistical tool was used to analyse their impact on preferences for cement. In this study, the full profile conjoint analysis method was selected for analyzing consumer preferences. This method is often recommended when few (up to ten) factors are used (Green and Srinivasan 1978, Hair et al. 1998). A full profile analysis remains the most common form of a conjoint analysis and has the advantage that the respondent evaluates each profile holistically and in the context of all other stimuli. A full profile means evaluating all the attributes or factors at a time by ranking from the most preferred product profile to the least preferred product profile or by assigning scores to each product profile from lesser preference to higher preference. For this research, four attributes and ten levels were identified (see Table 6.1). To identify the possible factors and attributes that affected consumer preference, existing research was referred to (Kumar 2013; Maity 2014; Oyatoye et al. 2013). In addition, the researchers did a reconnaissance study of ten individual home builders and exploratory discussions were held with experts from the national cement industry. As shown in Table 6.1, the total number of profiles or stimuli generated with our list of attributes and levels (four attributes and ten levels) are 3*2*2*3 = 36. But the number of profiles (36) might lead to information overload that will eventually reduce the accuracy of the respondents’ preference evaluations (Kambewa 2007). Moreover, respondents may not provide proper and meaningful evaluations when

124

Y. Tadesse and W. K. Tessema

Table 6.1 Attributes and their levels No.

Attribute

Level

Measurement

1

Price (Oyatoye et al. 2013)

1. Lower than average market price 2. Equal to average market price 3. Higher than average market price

Sales price

2

Quality (Blum 2007; Maity 2014)

1. As per standard 2. Higher than the standard

Setting time, comprehensive strength, and workability

3

Packaging (Kumar 2013)

1. Paper 2. Laminated PP

Strength, lateral usage, and ease of transportation

4

Customer service (Oyatoye et al. 2013)

1. Good 2. Fair 3. Poor

Credit facility, access to information, hospitality of sales people, test and conformation, and warranty

a large number of product profiles are presented. To solve this problem, we used the fractional factorial main effect design (Hair et al. 1998) to make the number of profiles manageable while keeping the orthogonality of the factors. Therefore, we used an orthogonal array design (Hair et al. 1998). Through the design, nine product profiles with different combinations of attribute levels were developed. The number of profiles used for the analysis is more than the minimum requirement (Kuzmanovi´c et al. 2013) at seven profiles (sum of all attribute levels- number of attributes +1). Orthogonality makes the correlation between attributes very minimum (almost zero) for the regression analysis and makes each level appear in equal numbers (Green and Srinivasan 1990). In the full profile conjoint analysis, ranking or scoring of the product profiles can be used (Green and Srinivasan 1990). We selected the purpose scoring method and the five level Likert type scale (1 for least preferred and 5 for most preferred) to capture the preference scores of each respondent for the product profiles generated from the orthogonal array design (Kambewa 2007). A sample product profile with a five level Likert type scale used in the survey is shown in Table 6.2. Table 6.2 Profile numbers Cement ID

Selling price

Quality level

Type of package

Level of customer service

1

Higher than average market price

As per the standard

Laminated PP

Good

Least preferred 1… 2… 3… 4… 5… most preferred

6 Factors Influencing Consumers’ Preferences for Cement Products …

125

To measure the internal validity and predictive accuracy of the conjoint design profiles, we generated two holdout profiles (Kuzmanovic et al. 2013) and incorporated it in the survey but these were not used in the conjoint analysis.The coefficients were estimated through the linear regression model to determine the utility value of each attribute level. The basic conjoint model (for example, Shalini and Msood 2010) used in this research is represented as: U(X) =



αij xij

(6.1)

i = 1, j = 1 where U (X) = Overall utility (importance) of an attribute αij = part-worth utility of the jth level of the ith attribute i = 1, 2……, m, j = 1, 2……, ki xij = 1, if the jth level of the ith attribute is present = 0, otherwise. The coefficients for each level of the attributes were estimated using the SPSS 20 conjoint statistical tool regression syntax by running both the plan and data files. Respondents’ ratings for the profile (stimuli) values form the dependent variable ‘preference.’ The measures of each customer value (the attribute level) are the independent (predictor) variables. The estimated betas associated with the independent variables are the utilities (preference scores) for the levels. The coefficients are the utility estimates (part worth) of the attribute levels.

6.3.5 Data Analysis Method To identify the factors affecting consumer preferences, the OLS regression analysis was done. Preference scores of the profiles assigned by respondents were taken as the dependent variable. Product profiles were units of analysis and these bring the total number of profiles in the analysis to 900 which is computed by multiplying the total number of profiles (9) by the number of respondents (100) (for example, Kambewa 2007). Attribute levels of each attribute are coded with dummy coding, 1 if present in the profile and 0 otherwise. In this coding, the total dummy variables in the analysis became ten. But the levels of all the attributes were not equal in the regression; therefore, one dummy variable from each attribute was dropped to keep the identity matrix. The dummy variables were used as independent variables in the analysis. We used the OLS model specified as: Yim = β0 + β1 X1 im + β2 X2 im + β3 X3 im + β4 X4 im + β5 X5 im + β6 X6 im + β7 X7 im + β8 X8 im + β9 X9 im + εim where

(6.2)

126

Y. Tadesse and W. K. Tessema

Yi = Score assigned by the mth respondents, β0 = intercept, β1 … β8 are the parameters to be estimated, X1 = dummy for equal to average market price, X2 = dummy for lower than average market price, X3 = dummy for higher standard quality, X4 = dummy for laminated Pip package, X5 = dummy for good service, X6 = dummy for poor service, X7 = age of the respondent, X8 = dummy for the sex of the respondent (1 = male, 0 = female), X9 = income level of the respondent, and εim = error term of the equation.

6.3.6 Reliability and Validity Tests Data was examined for any extremes and missing values for the conjoint analysis. Missing values were not found for the product profiles. To test the validity of the conjoint model, we examined Pearson’s correlation coefficient (Green and Srinivasan 1990) and the results were 0.976. For internal validity, Kendall’s tau was examined and the result was 0.889; Kendall’s tau for holdout was 1.000. Both the Pearson’s correlation coefficient and the Kendall’t tau test showed that the model was fit and was able to predict the outcomes (Green and Srinivasan 1990). The data was also tested for multicollinearity. As a rule of thumb if the VIF result is lower than 10, multicollinearity is not a series problem (Gujarati 2004). The mean VIF values showed that there was no multicollinearity in the data. The error terms were identically distributed. So, there was no heteroskedasticity problem in the data. We did the Durbin Watson test and all the values were within the range 1.25 and 2.75 (Shalini and Msood 2010) so there was no auto-collinearity in all the regression models.

6.4 Results and Discussion 6.4.1 Socioeconomic Background of the Respondents Table 6.3 gives the socioeconomic background of the respondents. About 24% of the respondents were in the 18 to 30 years age bracket whereas about 45% were between 31 and 40 years, and about 23% were between 41 and 50 years of age. The remaining 8% were above the age of 50. This means that about 69% of the respondents among the home builders were less than 40 years of age. In terms of sex, about 62% of the respondents were male and the remaining 38% were female. With respect to education levels, about 87% were either university/technical (54%) or secondary school (31%) graduates. About 9% had attended elementary school and about 4% were limited to reading and writing. Most of the respondents (63%) were employed in other organizations (governmental, nongovernmental or private) whereas only 37% were self-employed. For the type of homes that the respondents were building about 52% were building a villa residential

6 Factors Influencing Consumers’ Preferences for Cement Products … Table 6.3 Socioeconomic background of the respondents

Socioeconomic characteristic

127

Frequency (100)

Percentage

18 to 30

24

24.0

31 to 40

45

45.0

41 to 50

23

23.0

Over 50

8

8.0

Total

100

100.0

Male

62

62.0

Female

38

38.0

Total

100

Age

Sex

Occupation Employed

63

63.0

Self-employed

37

37.0

Total

100

100.0

Education Read and write

4

4.0

Elementary school

9

9.0

Secondary school

31

31.0

College and University graduate

56

56.0

Total

100

100

Type of home construction Villa

52

52.0

Ground plus

10

10.0

Other

38

38.0

Total

100

100

Source Authors’ survey (2015)

home, 38% doing other construction (like service, shops, fences), and the remaining 10% were building a ground plus type of home. With respect to the income levels of the home building respondents, their average income was birr 6,531.25 with birr 30,000 and birr 900 being the maximum and minimum income among the respondents.

128

Y. Tadesse and W. K. Tessema

6.4.2 Factors Influencing Consumer Preferences for Cement Products The factors that influenced preferences for cement products were analyzed using conjoint tools and descriptive statistics. The regression analysis showed that price, quality, and service significantly affected consumer preferences for individual home builders but packaging was not significant in influencing consumer preferences (Table 6.4). With respect to price, price equal to average market price (−0.653, p < 0.01), and price lower than average market price (0.787, p < 0.01) significantly affected consumer preferences for cement. This means that a higher price of cement negatively affected consumer preferences whereas individual home building consumers preferred prices which were lower than average. In other words, residential home builders preferred products with lower than the average market price or equal to the average market price to products which were priced higher than the average market price. With respect to the quality of cement, consumers preferred cement with a higher quality (0.210, p < 0.05) than a product with quality levels as per the standard. Poor customer services (−1.35, p < 0.01) negatively affected their preference for the product and they preferred products with good service rather than fair service. But packaging type did not significantly (−0.070, p > 0.1) affect the preferences of home building consumers. This may mean that home building consumers did not give much weight to the packing while making their cement purchase decisions. Table 6.4 Regression results of factors influencing preferences for cement

Variables

Individual home builders Coefficient

t-value

Constant

2.242a

12.75

Equal to average market price

0.653a

8.03

Lower than average market price

0.787a

9.67

Higher standard quality

0.21a

2.98

Laminated PP package

−0.07

−0.99

0.73a

8.97

−1.35a

−16.59

Age

0.046

1.14

Sex

0.135c

1.87

Income

1.33b

1.98

Good service Poor service

F-statistics (df)

F (9, 890), 88.54a

R2 (Adj. R2 )

0.473 (0.469)

a, b,

c

Note and are significant at 1%, 5%, and 10% (two-tailed), respectively Source Authors’ survey (2015)

6 Factors Influencing Consumers’ Preferences for Cement Products …

129

The three control variables (age, sex, and income) were studied to see how they influenced consumer preferences for cement products. In this respect, income (1.33, p < 0.05) and sex (0.135, p < 0.1) of the respondents significantly influenced preferences for cement. Age did not influence consumer preferences (0.046, p > 0.1). The significance of the sex variable in our study shows that given the cement buyer is male influences cement preferences than the buyer being female.

6.4.2.1

Advertising

Respondents were also asked whether they were influenced by the promotional ads in the electronic and print media. About 65% strongly disagreed that they preferred a brand that they had heard about through ads. About 22% strongly agreed that ads influenced their preference. The remaining (13%) were indifferent to the effect of promotional ads on their preferences. The influence of external factors including engineers and distributors as well as advertisements was descriptively analyzed regarding their influence on consumer preferences. Respondents were also asked about the influence of engineers, distributors, retailers, friends, and masons in their preferences. About 78% said they (strongly) agreed with engineers, 46% with distributors and retailers, 78% with friends, and 65% with masons. This finding regarding the influence that engineers, distributors, retailers, friends, and masons have is in line with the findings of Maity (2014) and Kumar (2013), who studied on impact of influencers on cement consumers.

6.4.2.2

Utility Estimates of Attribute Levels

As shown in Table 6.5 there is an inverse relationship between price and utility with respect to cement products, with higher prices corresponding to lower utility (larger negative values mean lower utility). There is a direct relationship between quality and utility, where a higher quality results in a higher utility value. When it comes to service, good service quality had a higher preference (utility level) as compared to a fair and poor service and fair service had more utility value than poor service. Since the utilities are all expressed in a common unit, they can be added together to give the total utility of any combination. For instance, a cement brand’s preference scores (utility scores) of a profile with price equal to average market price; higher than the standard quality; paper bag packaging; and fair service for residential home builders is: −0.787 + 0.42 + 0.035 + 2.08 + 1.348 (Constant) = 3.096. The range of utility values (highest to lowest) for each factor (attribute) provides a measure of how important the factor is to the overall preferences. Factors with greater utility ranges play a more significant role than those with smaller ranges (Gaurav and Anurag 2015).The findings of the present study (See Table 6.6) showed that price and service have greater utility ranges, and hence they are more important than quality and packaging for consumer preferences for cement in Ethiopia. We computed the

130

Y. Tadesse and W. K. Tessema

Table 6.5 Utility values Attribute levels

Residential home builders Utility estimate

Package Price

Paper bag

0.035

Std. error 0.111

Laminated PP

−0.035

0.111

Lower than average market price

−0.393

0.129

Equal to average market price

−0.787

0.257

Higher than average market price

−1.180

0.386

Quality

As per the standard

0.210

0.223

Higher than the standard

0.420

0.445

Service

Poor

1.040

0.129

Fair

2.080

0.257

Good (Constant)

3.120

0.386

1.348

0.482

Source Authors’ survey (2015)

Table 6.6 Relative importance of attributes

Factor/attribute

Relative importance Residential homes

Service

52.600

Price

23.131

Quality

12.633

Package

11.635

Total

100.00

Source Authors’ survey (2015)

relative importance of each factor known as an importance score or value to identify which factor or product attribute was more important for the respondents. The values were computed by taking the utility range for each factor separately and dividing it by the sum of the utility ranges for all the factors. The values thus represent percentages and have the property that they add to 100. The calculations were done separately for each subject, and the results averaged over all the subjects (Table 6.6). Service had a higher relative importance for residential homes (52.6%).Considering the customer service attribute’s levels, good customer service gives more utility value to preference for a product (3.12), followed by fair customer service (2.080); poor customer service (1.04) had the least utility value for residential home builders. Price was the second most important attribute (23.131%) for residential home builders. From the price attribute’s levels, price lower than the average market price(−0.939) had a lesser reducing effect on preference/utility/value and price equal to average market price (−0.787) was second while higher than the average market price (−1.180) had a higher reducing effect on the utility value.

6 Factors Influencing Consumers’ Preferences for Cement Products … Table 6.7 Total utility score and rank of profiles

Profile number

131

Individual home builders Total utility score

Rank

1

3.568

3

2

2.598

6

3

2.991

5

4

1.768

11

5

1.881

10

6

4.241

2

7

3.525

4

8

2.345

8

9

4.425

1

10

2.275

9

11

2.485

7

Source Authors’ survey (2015)

Quality was the third important attribute (12.633%) for cement consumers. From the quality attribute’s levels, higher than the standard gave more utility value (0.420) than quality level as per the standard (0.210). Packaging was the least important attribute (11.635%) for cement consumers. From the packaging attribute’s levels paper bags had a positive utility (0.035) and PP bag (−0.035) had a negative value indicating that paper bags were preferred over PPbags. This finding is different from Kumar’s (2013) finding as he identified PP bags as being preferred more than paper bags. By computing the utility values of each attribute’s levels, the total utility of the product profiles was computed (Table 6.7). As shown in Table 6.7, product profile number 9 (lower than average market price, as per the standard quality, paper bag, and good service) had the highest utility value in all the profiles. Profile 6 came second and profile 7 third. In all the profiles, the service level was either good or fair and in most cases the price was lower than average market or equal to average market price. Therefore, we can argue that service and price attributes are more important than other attributes in consumer preferences for cement products in the Ethiopian market.

6.5 Conclusion and Recommendations 6.5.1 Conclusion The main objective of this study was identifying factors that influenced consumer preferences for cement products. The study identified factors and attributes including service, quality, price, availability, and other external factors that influence consumer

132

Y. Tadesse and W. K. Tessema

preferences for cement products in Ethiopia. Door to door service affected preference the most; and in the external influencers, all influencers (engineers, friends, masons, distributors, and retailers) affected the preferences of the sampled respondents. The study found that during the purchase process, consumers formed their preferences based on the best combination of attributes for cement products. In this respect, consumers preferred a product profile with lower than average market price or equal to the average market price, as per standard quality, paper bag packaging, and good service. In our study homebuilding consumers’ preferences were significantly affected by the service level, price, and quality of the cement products. But the packaging type had no influence on home building consumers’ preferences. A male home builder with a high-income level positively and significantly influenced cement preferences. At the attributes level, service and price were more important than other attributes in influencing consumer preferences for cement. This indicates that while designing marketing strategies to influence consumers for cement products, companies need to focus largely on the service and price attributes of their products. This study did not find packaging type a significant factor influencing home building consumers’ preferences.

6.5.2 Recommendations Supporting marketing decisions based on scientific marketing research will help companies develop strategies for achieving their targets in a better and efficient way. The findings of this research can help companies identify the major factors and product attributes that influence consumer preferences for the products that they are offering. Based on the findings of this study, the following recommendations are made. First, service is the most important attribute influencing consumer preferences for cement products. Therefore, cement producers and suppliers should pay close attention to their customer service and delivery systems to satisfy their customers. Customer service packages such as easy access to information, flexible and easy ordering and delivery processing systems, availability of credit, mode of payment, complaint handling, the personality of the sales team, after sales support, and advice should align with consumers’ interests and preferences. This may require deploying increased transportation vehicles to enhance door to door services for cement products’ transportation to customers, and the vehicles should fit consumers’ different delivery requirements. Second, price is the second most important attribute influencing consumer preferences for cement products. This indicates that while designing marketing strategies to influence consumers, companies need to focus largely on price attributes. Hence, cement companies need to be cost conscious to keep their prices at a level which is acceptable to consumers. In addition, marketing strategies that focus on pricing might provide competitive leverage to cement companies to retain their customer base so as to achieve a sustainable competitive position.

6 Factors Influencing Consumers’ Preferences for Cement Products …

133

Third, providing higher quality at a lower price in most cases is not feasible. Hence, cement industries should keep their quality at the standard level and keep their costs low, rather than improving quality at the expense of lower prices. Fourth, enhancing awareness levels of engineers and masons about specific cement brands can help in encouraging cement consumers to prefer their products. This is because expert based word of mouth publicity (for example, from construction engineers) about specific cement brands influences consumer preferences. Therefore, cement industries should arrange events for creating awareness among these people and create good relationships with them.

References African Development Bank Group (2014). Ethiopian economic outlook. Retrieved from: http:// www.afdb.org/en/countries/east-africa/ethiopia//. Ahmad, T. and N. Vays (2011). The impulse buying behavior of consumes for the FMCG products in Jodhpur: Australian Journal of Basic and Applied Sciences, 5 (11): 1704–1710. All Africa (2014). Ethiopia Leads the East African Cement Market. Retrieved from: http://allafrica. com/. The American Marketing Association (2018).Marketing Essay. Retrieved from: https://www. ukessays.com/essays/marketing/the-american-marketing-association-marketing-essay.php? vref=1. Amutha, K.and P. Vinayak (2015). A study on brand preference of selected cements with special reference to Dindigul districts. International Journal of Multidisciplinary Research and Development, 2(1): 263–267. Arokiasamy A.R. (2012). The Effect of Marketing Mix and Customer Perception on Brand Loyalty: IOSR Journal of Business and Management (IOSRJBM)4 (2):01–11. Basavaraj, G., P. Parthasarathy Rao, A. Lalith,V.G. Lagesh, S. Pokharkarj, K.Gupta, and K.A. Ashok (2015). Understanding Trait Preferences of Farmers for Post-Rainy Sorghum and Pearl Millet in India: A Conjoint Analysis. Indian Journal of Agricultural Economics, 70 (l): 130–143. Bloomberg (2011). Ethiopian cement plants to double capacity amid construction industry boom. Retrieved from: http://www.bloomberg.com/news/2011-05-20/l. Blum U (2007). The East German Cement cartel: Cartel Efficiency and Policy after Economic Transformation. Eastern European Economics, 45(6):5–28. Brassington, F.and S. Pettitt (2003). Principles of Marketing (3rd ed.). England: Pearson Education Ltd. Cao, D., Z. Li, and K. Ramani (2010). Ontology Based Customer Preference Modeling for Concept Generation. Advanced Engineering Informatics. Retrieved from: 0.1016/j.aei.2010.07.007. Capital Ethiopia (2012). Cement prices fall like a ton of bricks. Retrieved from: http://www. capitalethiopia.com/. CSA (2008). Summary and Statistical Report of the 2007 Population and Housing Census. Federal Democratic Republic of Ethiopia, Population Census Commission, December 2008. Addis Ababa, Ethiopia. Chimboza, E. and E. Mutandwa (2007). Measuring the determinants of brand preference in a dairy product market: African Journal of Business Management,1 (9):230–237. Dhurup, M., C. Mafini, andT. Umasi (2014). The Impact of Packaging, Price and Brand Awareness on Brand Loyalty; Evidence from the paint retailing industry: ActaCommercii 14 (1). Retrieved from: http://dx.doi. org/10.4102/ac.v14i1.194.http://www.actacommercii.co.za. Ethiopian Standards (2015). 2015 Catalogue. Retrieved from http://www.ethiostandards.org/other% 20files/2015%20catalogue.pdf.

134

Y. Tadesse and W. K. Tessema

Felipe, J., A. Abdon and U. Kumar (2012). Tracking the Middle-income Trap: What is It, Who Is in It, and Why?, Levy Economics Institute of Bard College, Working Paper No. 715. Retrieved from: http://www.levyinstitute.org/pubs/wp_715.pdf. Ferree, E. (2011). Small Business Branding, Self-design: what is a Brand.Retrieved from: http:// www.elf-design.com/article. Fornell, C. (1992). A National Customer Satisfaction Barometer: The Swedish Experience: Journal of Marketing, 56:6–21. Gaurav, J. and A. Anurag (2015). Explaining the Factors of Brand Loyalty at Retail Store with Special Emphasis on Shoppers Stop. International Journal of Engineering and Management Science,6 (1): 34–39. Ghafran, M.A., R. Muhammad, M. Aamiriqbaland, and K. Arslan (2014). The promotional tools and situational factors’ impact on consumer buying behavior and sales promotion. Journal of Public Administration and Governance, 4 (2):179–201. Green, P.E. and V. Srinivasan (1978). Conjoint Analysis in Consumer Research, Issues and Outlook. Journal of Consumer Research, 5 (2):103–123. Green, P.E. and V. Srinivasan (1990). Conjoint Analysis in Marketing: New Developments with Implications for Research and Practice. Journal of Marketing, 4:3–19. Gujarati, N. D. (2004). Basic Econometrics (4th ed.). New York: The McGraw-Hill companies. Hair, J.F.,R. Anderson, R.I. Tatham, and W.C. Black (1998). Multivariate Data Analysis (5th ed.).Upper Saddle River, New Jersey: Prentice-Hall International. ISO 900 (2015): Quality management systems: Fundamentals and vocabulary. Retrieved from: https://www.iso.org/obp/ui/#iso:std:iso:9000:ed-4:v1:en. Jahanshahi, A.A., G..M.A, A.S.Mirdamadi, K.Nawaser, S.M.S. Khasksar (2011). Study the Effects of Customer Service and Product Quality on Customer Satisfaction and Loyalty. International Journal of Humanities and Social Science, 1 (7):253–260. Kambewa. E.V. (2007). Balancing the People, Profit and Planet Dimensions in International Marketing Channels, PhD thesis, Wageningen University, the Netherlands. Kanetkar, V. (2005). Conjoint and Discrete Choice Designs for Pricing Research. University of Guelph, Guelph, ONT: Department of Consumer Studies. Karunanithy, M. and S. Sivesan (2013). An Empirical Study on the Promotional Mix and Brand Equity: Mobile Service Providers. Industrial Engineering Letters (online) 3 (3). Kent, T. and O. Omar (2003). Retailing. New York: McMillan.Kotler P., L.K., Kelvin, A. Koshy, and M. Jha(2009). Marketing Management(13th ed.). New Jersey: PrenticeHall publishing. Kotler P., K. Keller, M. Brady, Malcolm Goodman, and T. Hansen (2009). Marketing Management. Canada: Pearson Education. Kumar, S. (2013). Marketing Mix Strategies and Indian Cement Sector. International Journal of Emerging Research in Management &Technology, ISSN: 2278-9359: 52–57. Kuzmanovic, M., M.Radosavljevic, and M.Vujosevic (2013). Understanding Student Preferences for Postpaid Mobile Services using Conjoint Analysis. Acta Polytechnica Hungarica,10 (1): 159–176. Lin, M.Y.and L.H.Chang (2003). Determinant of habitual behavior for national and leading brands in chain. Journal of product and brand management, 12 (2): 94–107. Macdonald, E.K. and B.M. Sharp (2000). Brand Awareness Effect on Consumer’s Decision Making for a Common Repeat Purchase Product: A Replication. Journal of business research, 48 (1):5–15. Maity A. (2014). Impact of External Influencer Recommendation in Purchase Behavior Process of Selection of Cement for Construction. International Journal of Business and Management Invention, 3 (8): 01–11. Ministry of Finance and Economic Development (MoFED) (2010). Federal Democratic Republic of Ethiopia. Growth and Transformation Plan 20 10/11 -2014/15. Volume I: Ministry of Finance and Economic Development, November 2010, Addis Ababa. Retrieved from: http://extwprlegs1. fao.org/docs/pdf/eth144893.pdf.

6 Factors Influencing Consumers’ Preferences for Cement Products …

135

Murugan M. and R. Ganapathi (2012). Consumers’ Preference on Various Brands of Cement in Gobichettipalayam Taluk of Erode District. International Journal of scientific research, 1 (5): 89–97. Nartey, S. (2013). An Assessment of Customers’ Perception on Acceptability of Dangote Cement in Tema Metropolis. Kwame Nkrumah University of Science and Technology (unpublished). National Cement Share Company (2012). History of national Cement: Brochure, Dire Dawa, Ethiopia (unpublished). Nithiyananth, D. andS. Arunpandian (2012). A Study on Customer Preference on Samsung Product in Erode District. IOSR Journal of Business and Management (IOSR-JBM)e-ISSN: 2278-487X, p-ISSN: 2319-7668, PP 20–22. Retrieved from: https://www.academia.edu/31745023/A_Study_ on_Customer_Preference_on_Samsung_Product_in_Erode_District. Oyatoye E.O., S.O.Adebiyi and B.B. Amole (2013). An Empirical Study on Consumers Preference for Mobile Telecommunication Attributes in Nigeria. British Journal of Economic Management & Trade, 3 (4): 419–428. Prakash, C. (2011). Consumer Preference to Health Drinks in Tiruvarur Town. Asian Journal of Management Research, 2 (1):420–427. Pullman, M.E., W.L. Moore, and D.G.Wardell (2002). A Comparison of Quality Function Deployment and Conjoint Analysis in New Product Design. Journal of Product Innovation Management, 19 (5):354–364. Rajaguru, R.and J.M.Matanda(2006). Consumer Perception of Store and Product Attributes and its Effect on Customer Loyalty within the Indian Retail Sector. Monash University, (unpublished). Raza, A.A. (2011). Branding and its Role in Profitability of Cement Holding Companies: Iranian cement news agency. Retrieved from: http://www.cementgroup.ir. Schramm-Klein, H., D. Morschett, and B. Swoboda (2008). Verticalization: The Impact of Channel Strategy on Product Brand Loyalty and the Role of Involvement in the Fashion Industry. Advances in Consumer Research, 35: 289–297. Retrieved from: http://www.acrwebsite.org/volumes/13246/ volumes/v35/NA-35. Shalini, N. and H. Msood (2010). An empirical Study of Tourist Preferences Using Conjoint Analysis. International Journal of Business Science and Applied Management, 5 (2):1–16. Sisay, Tomas (2012). Market Structure, Conduct and Performance Paradigm Re-applied to the Ethiopian Cement Industry. Addis Ababa University (unpublished). Soegoto, D.S. (2018). Analysis of eco-friendly preference and eco-friendly product quality; their implications to customer satisfaction. OP Conference Series: Materials Science and Engineering, 434 (1): 1–7. Tennis, P.D. (1999). Portland cement characteristics-1998, Concrete technology today, Portland cement association, 20 (2): 1–3. Retrieved from: https://pdfs.semanticscholar.org/be1b/ 8122b4478faba54239c312242ff066b075e3.pdf. The Reporter (2015). Anomaly in the cement industry. Retrieved from: https://www. thereporterethiopia.com/content/anomalycementindustry. Udomkun, P., J. Ilukor, J. Mockshell, G. Mujawamariya, C. Okafor, R. Bullock, N.L. Nabahungu, and B. Vanlauwe (2018). What are the key factors influencing consumers’ preference and willingness to pay for meat products in Eastern DRC? Food Sci Nutr.: 6:2321–2336. Walta Information Center (2014). Cement export to neighbouring nations generate over $9 million. Retrieved from: http://www.waltainfo.com.

Chapter 7

The Value of Roads in Rural Household Consumption Marshal Negussie Simie

Abstract Investments are made in rural roads in Ethiopia with the belief that they will lead to poverty reduction. In view of this thinking, some important questions that need to be asked are whether and to what extent road construction has facilitated the alleviation of consumptive poverty. Using a three round longitudinal data, this study uses a dynamic non-linear probit model to assess the impact of rural roads on poverty. The model allows us to account for initial conditions and time invariant unobserved heterogeneity. The findings show that the construction of rural roads has a significant and desirable outcome in reducing consumptive poverty. The findings also show that initial conditions matter in estimating the trajectory of poverty dynamics. Keywords Rural roads · Panel data · Consumptive poverty · Dynamic probit · Ethiopia JEL Codes C23 · D12 · H54 · I32 · O18 · P25

7.1 Introduction Ethiopia has made remarkable progress in its economic growth in the past decade, and the incidence of rural consumptive poverty declined from 45.4% in 2000 to 27% in 2016 (The World Bank 2019). Developments in the agricultural and service sectors with heavy public investments in infrastructure have been the center of the country’s growth experience. According to the World Bank (2015), the country has the third highest public investment rate in the world which rose from about 5% in the early 1990s to 18.6% of GDP in 2011. In line with this development, road construction has taken a prominent portfolio in the government’s spending and is an important catalyst of economic development. The country’s road network increased from 26,550 km in 1997 to 99,522 km in 2016. However, this progress is not without its costs as Ethiopia’s infrastructure deficit is the third largest in Africa. M. N. Simie (B) Department of Economics, Addis Ababa University, Addis Ababa, Ethiopia e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 J. Wood and O. Habimana (eds.), A Multidimensional Economic Assessment of Africa, Frontiers in African Business Research, https://doi.org/10.1007/978-981-15-4510-8_7

137

138

M. N. Simie

Given the large investment costs, there is theoretical and policy desire to better understand the returns from investments in rural roads (Stifel et al. 2016). Simultaneously, a large body of research presents changes in poverty as an important measure of the failure or success of specific government policies (Jefferson 2012; Ravallion and Bidani 1994; Sen 1983). The poverty profile is also used as an eligibility criterion for dozens of assistance plans. It can be argued that sound economics and well-designed economic policies can provide critical assistance in the struggle against poverty (Ravallion 1998). Therefore, it is important to gather more empirical evidence on how differences in access to rural roads explains poverty persistence as the goal of reducing the infrastructural gap is maximizing the society’s well-being. Research shows that access to roads and transport services entails access to opportunities which can fuel the growth process (Batten and Karlsson 1996; Stifel et al. 2016). Improved roads can expand farm incomes through easier access to markets and technology, and increased availability of relevant inputs at lower costs (Damania et al. 2017; Qin and Zhang 2016). Roads can help rural households in diversifying their income sources by opening employment opportunities and allowing economic mobility. Diversification and expansion of incomes of the rural poor improves food security and lowers the poverty rate (Aggarwal 2018; Ramessur et al. 2010). In general, researchers have been trying to identify the diversified links through which road construction might affect a society’s welfare. But, instead of looking for these elusive chains, a more policy informative way especially in developing countries is measuring its impact on alleviating poverty. The premise of this study is that better road access is strong enough to pull households out of poverty. However, the appeal of this type of benefit may be at odds with the increasing skepticism about returns to investments in rural roads. Domestically bound rural travel may result in a limited impact on the lives of rural dwellers (Dawson and Barwell 1993; Porter 2007). Rural roads carry less traffic, are harder to maintain, and costlier to construct (Bryceson et al. 2008). Moreover, low commercial surplus coupled with high costs of wheeled transport services can lead to under-use of rural road infrastructure (Bryceson et al. 2008). These imply that the impact on society stems from how the roads are used and not from the construction of roads. This study investigates how important the difference in road access is in explaining poverty persistence over time in the context of Ethiopian rural households. To provide answers to this question the chapter examines the extent of consumptive poverty, the role that access to roads plays in reducing rural households’ absolute poverty, and the factors associated with the chances of moving into or out of poverty. The welfare economics approach (Sen 1979) seeks to measure poverty through household utility, which in turn is usually assumed to be approximated by consumption or income based measures. Income is generally used as a measure of household welfare in developed countries, but it tends to be seriously understated in self-employed or large agricultural populations, whereas consumption expenditure is less under-reported and more closely reflects permanent income in developing countries (Deaton and Zaidi 2002; Haughton and Khandker 2009). Thus, consumption aggregates are used in the poverty analysis as they are more reliable and stable than income measures (Deaton and Zaidi 2002).

7 The Value of Roads in Rural Household Consumption

139

This study provides various dimensions of innovations including its attention on a comparison between households with and without access to roads relying on quasirandom variations which use a straight-line distance between household location and rural roads. The focus is also on consumptive poverty which integrates the different links and objectives that roads might have on households’ economic and social activities. The study also explores possible determinants of poverty besides roads, using richer data and a richer econometric analysis than previously applied in a developing country. The findings have policy implications for Ethiopia, which has been investing and is planning to invest more in rural roads. The rest of the study is organized as follows. Section 7.2 discusses the theoretical and empirical underpinnings of the study. Section 7.3 presents the context and data. Section 7.4 discusses the research models. Section 7.5 presents the empirical results and a discussion. The conclusions are given in Sect. 7.6.

7.2 Literature Review 7.2.1 Theoretical Review Poverty, in an absolute or relative sense, is the presence of deprivation (Jefferson 2012). Since it is a condition that exists in scarcity, it is of intrinsic interest to economists and policymakers. The success or failure of government policies can be measured by examining changes in poverty. It is well known from previous studies that spatial poverty traps are a silent feature of the rural landscape in developing countries (Edmonds and de Veen 1992; Kam et al. 2005). Lack of reliable access to social, economic, and political domains ensures that societies remain in poverty (Edmonds and de Veen 1992). Improvements in transportation are usually considered a compelling developmental intervention which covers 20–40% of the national budgets in developing countries. Simultaneously, multilateral organizations have been devoting more than 20% of the developmental loans for improvements in transportation. The simple and deterministic conclusion by Lugard (1923) has long prevailed in policy frameworks: “the material development of Africa may be summed up in one word, transport.” P. 5. Literature also recognizes the contribution that road development can make in overcoming spatial poverty traps by both widening households’ interactions with their surrounding environment and providing facilities that improve the quality of their lives. Improvements in the quality of transport services and reduction in transport costs because of access to roads can have helpful effects on increasing households’ real incomes and consumption through three inter-related links: first, transport services provide access to markets and other services; second, they have basic consumption and production value; and third, they affect real wealth (Batten and Karlsson 1996).

140

M. N. Simie

Transport as access to other services: When utility is regarded as inherent in things, transport facilitates the acquisition of other goods and services. There is a thinking that the purpose of roads is to serve as a meandering instrument. They are built not to get direct satisfaction out of them, but to transport something thus satisfying wants by moving things into a more suitable place. In this respect, road improvements can be transferred to household welfare in three ways. First, the benefits of improved roads include access to other consumption opportunities and promoting jobs, education, and healthcare for individuals. According to Ramessur et al. (2010) the restrictions on the standard of living of the poor as a result of transport shortages include income, accessibility, and safety issues. The poor make fewer trips and most of their trips are undertaken on foot. Their ability to access services and job opportunities is limited. They are also locationally and vocationally most exposed to personal conflicts. Moreover, the price of transport services relative to other goods and services affects the level of household consumption that can be achieved in the given budget constraint. Transport prices in turn depend on the availability of transport infrastructure and associated modes of transport. Researchers in tropical Africa document the wide practice of “head loading” of logs, water, and crops because of the absence of roads for motorized and non-motorized means of transportation (Batten and Karlsson 1996; Bryceson et al. 2008). Second, adequate transport facilitates efficient operations of markets thus having a favorable impact on prices for households either as producers or consumers. Rural households participate in markets by trading resources, goods, and services. The market determines the prices of the traded goods as it also determines a large part, if not all, of households’ incomes and consumption (Cowell 2006). The number of both buyers and sellers that can make up a single market is determined by the availability of transportation. A dense market cannot typically be reached without easy means of transportation to bring in and carry back large supplies. With improved access to markets, buyers have the option of selecting a wide variety of commodities and sellers find many buyers gathered at one point (Fetter 1904). The transport system through the market system, therefore, has a big role in influencing household welfare. Third, with improved transportation, rural households can better utilize agricultural and non-agricultural opportunities by employing resources more efficiently (Khandker et al. 2009). As improved road access leads to a reduction in transportation costs, the prices of modern inputs are more likely to fall, and farmers can apply more of these inputs to improve farm productivity. In addition, with better road access, farmers can hire skilled labor to take care of specialized agricultural production (Qin and Zhang 2016). Better means of transportation can also improve productivity by reducing the time, energy, and boredom of walking long distances (Ramessur et al. 2010). Transport as consumption and production value: The availability of transport services can be regarded as a measure of well-being, implying that households are poor so long as they do not have access to transport services of the necessary standard (Batten and Karlsson 1996). “Even in the most remote and least developed of inhabited regions, transport in some form is a fundamental part of the daily rhythm of

7 The Value of Roads in Rural Household Consumption

141

life” (Hoyle 1973: p. 9). Besides, where production is taken as the joining of goods and services into right associations with wants, transportation can be the main and archetypal way of raising incomes. Animals, a senior form of being than the more still plants, in the classification of the scientific theory of evolution find food through movement and can have wider place interactions. The existence of animals is featured by the movement they make between places to find food and escape unsuitable habitats. To this power, human beings have added that of transporting things and, thus, adding to their welfare (Fetter 1904). Access to transport adds to wealth: Access to transportation has high power to condition exchange values. For instance, location may determine the value of land as much as its fertility. “A rocky field near a market may be richer, in an industrial sense, than the richest soil far remote, which can be used only at the cost of alienation from society” (Fetter 1904: p. 281). However, all these implied links of investments in transport and poverty reduction, derive not from the mere construction of the transport infrastructure but from its efficient functioning (Batten and Karlsson 1996; Owen 1959). The construction of transport infrastructure only provides an option, its efficient utilization depends on other exogenous factors that cannot be taken for granted in most developing countries. Concerns start from public finance’s focus on rural road construction. This relates to the limited financial capacity of most developing countries to construct and maintain roads (Edmonds and de Veen 1992). In fact, some authors argue that other interventions with less project time and investment costs are more likely to alleviate poverty (Hirschman 1958; Hoyle 1973). Essentially, the construction of transport infrastructure can be politically motivated and may lead to misdirected investments. Moreover, over engineered roads may not necessarily satisfy the needs of the ‘overwhelmingly rural walking world’ of developing countries (Bryceson et al. 2008). The degree of available economic opportunities as a result of road construction, is a direct function of accessible goods and services. The rural population predominantly travels on foot with small loads over short distances. Most journeys are for purposes which do not involve buying or selling and are to meet basic needs such as water, fuel wood, and food. For most of the rural population, even the very simple, non-motorized, and intermediate means of transportation such as carts and wheelbarrows are not affordable (Edmonds and de Veen 1992). The available nonmotorized transportation means are usually forgotten in policymaking and are put out from the market (Bryceson et al. 2008). To be economical, the expenditure on providing roads must be warranted by the density of road traffic, which is less likely to occur in rural populations in developing countries. Similarly, roads alone are not enough stimuli for development because roads are only part of the solution to the problems of rural populations and their effects are context specific (Demenge et al. 2015; Edmonds and de Veen 1992). While advantages tend to concentrate within a 3 km corridor along a road, benefits are captured by the better-off population (Khandker et al. 2009; Porter 2007). Benefits of roads also depend on the existence and efficiency of integrated development projects (Ambel

142

M. N. Simie

et al. 2015). If benefits from improvements in transport are to be realized, households’ choices should be considered, roads need to be accompanied by additional interventions, and access to essential services needs to be guaranteed.

7.2.2 Empirical Review Practically identifying the link between poverty and road construction is important yet complex. Several studies provide evidence on the effects of roads on a wide variety of economic outcomes in rural areas (Aggarwal 2018; Aparna et al. 2017; Asomani-Boateng et al. 2015; Bryceson et al. 2008; Damania et al. 2017; Demenge et al. 2015; Edmonds and de Veen 1992; Khandker et al. 2009; Rammelt and Leung 2017). While the existing literature is too diversified in identifying the different channels at play between well-being and road provisions, their direct effect on poverty seems to be under-researched. Diversified approaches are useful in fleshing out a multidimensional portrait of roads, but lack of focus can make interpretation difficult. A reliable way of judging the effects of road construction can be through measuring their direct effect on poverty. As Ravallion (1998) argues, a “credible measure of poverty can be a powerful instrument for focusing the attention of policy makers on the living conditions of the poor.” P. 1. Given that measuring poverty is difficult in itself, basic questions on the extent of poverty and how poor households respond to their economic and physical environment need to be described. It can be concluded that quantitative evidence is quite scarce on how roads affect consumptive poverty measures in Ethiopia. One major reason is the difficulty in finding before and after data which enables a causal impact evaluation. The intensive investments needed for roads also make it difficult to implement randomized provisions of roads (Aggarwal 2018). With these difficulties, existing literature on transport infrastructure’s effects has largely relied on quasi-random variations which use a straight-line distance between peripheral regions and roads. A notable work in this respect is by Dercon et al. (2009), which examines the impact of access to roads on poverty and consumption growth by using household level panel data from rural Ethiopia. Their GMM estimation shows a significant positive effect of roads on growth in consumption. Their research used community representatives’ responses for identifying the closest town and the quality of the roads leading to that town. The research failed to control for initial condition problems in their poverty estimation, which is identified as something important by other studies (Khandker and Koolwal 2011) in estimating the trajectory of the impact of roads.

7 The Value of Roads in Rural Household Consumption

143

7.3 Models and Estimators 7.3.1 Consumption Aggregates and Poverty Measures Following the more common practice and the guidelines prepared by Deaton and Zaidi (2002), this study uses a consumption-based measure of poverty. Components of consumption expenditure are aggregated into classes of food items, non-food items, and consumer durables. The food consumption sub-aggregate includes food purchased in the marketplace, food that is home produced, and food items received as gifts. The non-food items include consumption of daily use items such as soaps and cleaning supplies, kerosene and petrol, charcoal and miscellaneous personal care items, as well as other less frequently purchased items such as clothing, footwear, kitchen equipment, household textiles, and other household items. Education expenditure is also included as a non-food item. For durables, annual estimates of service values are used as a measure of consumption. As housing and rental markets are not well developed it is difficult to calculate any serious estimate of rental value and hence this is not included in the consumption aggregate. Consumption is expressed in per adult equivalent terms to account for economies of scale and consumption requirements. Adjustments are also made to account for cost of living differences by deflating the nominal food aggregate by the Paasche price index of the money metric approach. Then, the poverty line is constructed using the “cost of basic needs” approach and the price in 2015 as the base. Following past recommendations for Ethiopia, the food poverty line is estimated based on the cost of a reference food bundle that provides 2,200 kcal per adult per day. To this is added the non-food line by following the method set out in Ravallion and Bidani (1994)—by regressing the actual food share of each household by the log of total actual household expenditure in relation to the food poverty line. The poverty status of each household is then delineated by identifying its position relative to the constructed poverty line. Those with a consumption per adult equivalent above or equal to the poverty line are regarded as non-poor and those below the poverty line as poor. Given the constructed information on the poverty line, we constructed the poverty headcount, gap, and severity indices. The poverty headcount index simply measures the proportion of households living in poverty. The poverty gap index adds up the extent to which households on average fall below the poverty line and expresses it as a percentage of the poverty line. The poverty severity index provides the weighted sum of poverty gaps by emphasizing more on the poorest (Foster et al. 1984).

144

M. N. Simie

7.3.2 Regression Model Researchers have long studied the patterns and causes of consumptive poverty and reached several important conclusions. As discussed earlier, poverty is inextricably linked to improvements in transport infrastructure. Poverty can also persist over time, that is, previous experience of poverty makes future poverty more likely. Studies have shown that “state dependence of poverty” is a well-established fact (Alem et al. 2014). But it is interesting to investigate the extent of state dependency when there are shocks in public investments (road access). The most notable, context specific, economic, and human capital assets that might correlate with poverty are land, livestock, and literacy rate. Moreover, researchers agree that household structures are important as they show a possible correlation between a household’s composition and its level of poverty. Gender of the household head might influence household poverty, and more specifically households headed by women are expected to be poorer than those headed by men in the rural context (Haughton and Khandker 2009). The aging process might also affect a household’s poverty status through changing capabilities over time. Using models where the poverty status is regressed on lagged observations is a standard approach for handling longitudinal poverty dependence (Alem et al. 2014). In such a dynamic model, the probability of being poor could depend on whether the household experienced poverty in the previous period, that is, state dependence is allowed through the lagged dependent variable. Even when the coefficient of the lagged dependent variable is not of direct interest, consistent estimates of other parameters can be recovered by allowing dynamism in the underlying process. Consider the following dynamic probit unobserved effect model for the latent dependent variable of poverty status yit∗ , for household i at time period t: 

yit∗ = γ yit−1 + βxit + ci + u it   yit = 1 yit∗ > 0 (i = 1, . . . , N ; t = 2, . . . , T )

(7.1)

where x it is a vector of explanatory variables including road access, ci is household specific unobserved heterogeneity, uit is assumed to have zero expectation, and its variance is σu2 . The parameters γ and β are the coefficients to be estimated. The cross-section unit “N” is large but observed for a small and fixed period “T ”, so asymptotic properties are going to infinity on “N”. A convenient normalization, since y is a binary variable, is σu2 = 1. The probability of being poor for household i at time t, given ci , is:      P yit |xit , yit−1 , αi = Φ γ yit−1 + βxit + ci (2yit − 1)

(7.2)

where Φ is the cumulative distribution function of standard normal distribution. An important issue in this setting is the treatment of the initial condition, since the influence of the initial observations on each subsequent observation cannot safely be ignored when the time dimension is small. The standard random effects model

7 The Value of Roads in Rural Household Consumption

145

assumes ci is uncorrelated with x it , which may be appropriate if the start of the process coincides with the start of the observation period for everyone. But, if poverty status in the initial year is endogenous, a correlation is induced between the error term and the lagged dependent variable, then ignoring it will lead to a bias in the parameter’s estimations. Two popular approaches have been proposed to handle the problem of the initial conditions. Heckman (1981), suggests ‘integrating out’ the individual specific error term from the likelihood function and approximating the joint distribution of the full observed endogenous variable. But this approach relies on computationally intensive methods. An alternative approach by Wooldridge (2005), specifies a density for the unobserved heterogeneity conditioned on the initial dependent variable and explanatory variables. The Wooldridge (2005), estimator has the same likelihood structure as in the standard random effect probit model with the lagged and initial dependent variables among the regressors. Relatively, Wooldridge’s solution has the advantage of simple computational techniques and can be estimated using standard estimation software. The original Wooldridge (2005) model specifies the auxiliary model for ci as: ci = α0 + α1 yi1 + xi α2 + ai ,

(7.3)

  where xi = (xi1 , . . . , xi T ) and ai ∼ Normal 0, σa2 . A more common specification for ci in literature uses the conditional means of the time varying explanatory variables based on all periods. However, using longitudinally averaged variables is found to be biased unless the panel is sufficiently long (Rabe-Hesketh and Skrondal 2013). Thus, the Wooldridge (2005) model with time-varying covariates at each occasion is used in this study of the probability of being in poverty. A balanced panel data is assumed in developing the Wooldridge estimator (Wooldridge 2005). It may also be applied to a sub-set of observations constituting a balanced panel and assuming that the sample dropout is ignorable. Researchers have found that the impact of attrition is small. To check the robustness of the results, estimates derived using both balanced and unbalanced panels are reported.

7.4 Context and Data 7.4.1 Rural Roads in Ethiopia Investments in rural roads are one of the main priorities of the Government of Ethiopia as these are regarded as a mechanism for reducing poverty. In response to this need, the government has embarked on the Universal Rural Road Access Program (URRAP) that sets out connecting all rural villages by all-weather roads. For roads that are designated as “all-weather roads” an improved surface, for example,

146

M. N. Simie

TIGRAY

60

OROMIA

192

AMHARA

285

3,044 SNNPR

488

12,553

9580

LENGTH IN KM

36,211

Lngth (km)

B/GUMUZ

D/DAWA

HARARI

GAMBELA

REGION

Fig. 7.1 Length of URRAP roads completed in the first phase. Source The Ethiopian road authority

gravel or something similar is required and it needs to be passable for motorized traffic in both wet and dry weather. URRAP targets delivering, expanding, and improving the conditions of rural road networks in all regions of the country. The full-fledged implementation of the URRAP plan will ensure that close to 80% of the total rural population has access to roads year-round. In the first implementation period (2010–15), it was planned to construct 71,523 km of all-weather roads throughout the country at an estimated cost of 1 billion dollars, of which 62,413 km of URRAP roads or 87% of the target for the five years was completed (Fig. 7.1). The construction of these roads increased the number of kebeles connected with allweather roads from 6,222 in 2010 to 11,871 in 2015. However, the extent to which these roads contributed to the welfare of society is inadequately documented. Despite substantial improvements and the huge investments in URRAP, a large proportion of the rural population has no access to all-season roads. In 2015, about 24% rural districts in Ethiopia had no access to all weather roads. As a result, the modes of transportation for the rural population were restricted to humans and pack animals transporting only a few agricultural products and inputs from and to the nearest markets. The government planned to construct 90,000 km of URRAP roads in the second implementation period (2015–20). However, the construction of these roads was stopped without giving any official reason.

7.4.2 Data Considering the lack of proper baseline surveys and documentation of basic indicators for each URRAP road, the Ethiopian Roads Authority collected baseline and evaluation studies of these roads over a period of three years; 2015, 2016 and 2017.

7 The Value of Roads in Rural Household Consumption

147

The baseline studies, follow-ups, and evaluations consider both the positive and negative impacts of the roads on the rural population and their livelihoods through the selection of an array of various economic, social, and physical indicators. This study used the three-round longitudinal household survey of the Ethiopian Road Authority. The baseline, follow-up, and end line surveys were done in 2015, 2016 and 2017 for the same households using the same procedures and instruments. The survey included the four regional states of Oromia, Harari, Gambella, and Dire Dawa. Taking into account 25% of the sample proportion for the URRAP districts, 66 districts from Oromia, two districts from Gambella, three form Harari, and five from Dire Dawa were randomly selected using the proportion to size rule. Then, one URRAP road was randomly selected from each district, all villages crossed by the sample URRAP road were listed and two villages were randomly selected. Villages served by other road networks and nearest to the urban centers were omitted from the frame. The total number of sample villages, therefore, was 152 in 76 road networks in 76 districts. Figure 7.2 shows the study area. After this point, households were categorized based on the road influence area threshold of five kilometers from URRAP roads in accordance with literature and the Ethiopian Ministry of Transport’s classification of accessibility. Hence, households within a radius of less than two kilometers were ‘with access’ to a road and those above a radius of five kilometers were ‘without access.’ In doing so, two development groups were targeted, representing those ‘with access’ and ‘without access’ and 304 enumeration areas were selected from the sampled villages. To ensure potential for similarities in each enumeration area the type of crops, major sources of livelihood, agro-ecological zone, and distance from roads were taken into account. Each development group consisted of 20 to 30 households. Finally, 15 households were

Fig. 7.2 Study area. Source The Ethiopian road authority

148

M. N. Simie

Table 7.1 Sample size Indicator Region

Group Total

Year Oromia

Overall attrition rate (%)

2015

2016

2017

3934

3799

3721

Gambella

120

118

120

5.4 –

Harari

240

211

211

12

Dire Dawa

180

172

172

4.4

With access

2406

2311

2271

5.6

Without access

2068

1989

1953

5.6

4474

4300

4224

5.6

Source The Ethiopian road authority

randomly selected from each enumeration area or 60 households per URRAP road and district. In all, 4,560 households were identified for the study (50% of each with and without access). The final sample households interviewed in the baseline survey were 4,474, 4,300, and 4,224 in 2015, 2016, and 2017 respectively. Table 7.1 presents the sample size of the four regions for each round. Two limitations were encountered in using the data. First, the data replaced the dead, ill or transferred household heads with new household heads from the same household. This might not affect household aggregate factors, but it can change the household head’s characteristics such as age and sex. Second, the data was designed to evaluate the impact of roads on poverty which requires data before and after the intervention. But for some observations within the access group the road construction was completed in the initial period and for others it was not completed even in the last round of the panel survey. Thus, households which were classified as the ‘access group’ in the survey design but without completed roads, or roads under construction, were placed in the ‘without access’ group in the data analysis and estimation. As a remedy, models which considered this irregularity were used in the research analysis.

7.4.3 Descriptive Statistics Table 7.2 shows the descriptive statistics of the dependent variable in terms of poverty headcount, gap, and severity in 2015–17. Poverty headcount declined from about 46% in 2015 to 42% in 2017. The pace of poverty reduction in observations with access to road was strong, particularly when compared with those without access. In 2015, both areas were nearly the same in terms of poverty incidence with 45% for areas with access and 46% for areas without access to roads. But they differed strongly through time as the poverty incidence in areas with road access declined dramatically. In 2017, poverty incidence ratio for observations with access to roads

7 The Value of Roads in Rural Household Consumption

149

Table 7.2 Poverty profiles across time and by access groups Welfare measure

Year

Difference (2015 and 2017)

2015

2016

2017

Poverty headcount/incidence/

0.4551

0.4286

0.4221

−0.0330 (0.0106)***

With access

0.4501

0.3775

0.3668

−0.0833 (0.0160)***

Without access

0.4608

0.4546

0.4533

−0.0075 (0.0146)

Poverty gap/depth/

0.1630

0.1681

0.1505

−0.0124 (0.0049)**

With access

0.1615

0.1360

0.1251

−0.0364 (0.0072)***

Without access

0.1647

0.1844

0.1649

0.0002 (0.0067)

Poverty severity

0.0797

0.0882

0.0729

−0.0068 (0.0022)**

With access

0.0789

0.0694

0.0609

−0.0180 (0.0046)***

Without access

0.0807

0.0977

0.0797

−0.0011 (0.0042)

Note ** p < 0.05, *** p < 0.01 Source Own calculations

was about 37% while for those without access it was about 45%. This reduction mainly took place between 2015 and 2016. Moreover, for observations without road access, poverty remained widespread and the poorest did not see any improvements. This is evident from the small reduction in poverty depth and severity from 2015 to 2017. There was even worsening of poverty depth and severity from 2015 to 2016. In contrast, for those with road access there was a continuous and statistically significant decline in both poverty depth and severity in the study period. This indicates that those who were poor in 2016 and 2017 were on average nearer to the poverty line than those who were poor in 2015, implying a shared prosperity for observations with road access. Table 7.3 presents the mean and standard deviation of the explanatory variables for each survey year by disaggregating them in terms of road access. It can be observed that both survey groups had similar characteristics. The mean age of the sample group was around mid-40s dominated by male headed households. The proportion of uneducated household heads was larger with about 69% being unable to read and write. On average households owned low agricultural assets of land and livestock. Except for livestock units, which registered relatively low levels in 2016, similar characteristics were seen across time.

150

M. N. Simie

Table 7.3 Descriptive statistics of the variables by group and over time Variable

With access

Without access

Total

Mean

Std. dev.

Mean

Std. dev.

Mean

Std. dev.

Age of household head

41.987

14.657

43.032

14.393

42.810

14.454

Male household head

0.924

0.265

0.892

0.310

0.899

0.301

Literate household head

0.295

0.456

0.309

0.462

0.306

0.461

Land size in hectares

1.156

1.081

1.474

1.315

1.407

1.276

Tropical livestock unit

3.878

4.637

4.211

4.908

4.141

4.853

Observations

949

2015 round

3525

4474

2016 round Age of household head

43.052

13.697

43.834

Male household head

0.904

0.295

0.890

Literate household head

0.322

0.468

0.332

Land size in hectares

1.449

1.328

1.427

Tropical livestock unit

1.766

2.995

1.909

Observations

1449

14.074

43.571

13.952

0.314

0.0894

0.307

0.471

0.329

0.470

1.252

1.434

1.278

2.995

1.861

2.995

2851

4300

2017 round Age of household head

44.364

14.331

44.917

14.292

44.718

14.307

Male household head

0.891

0.312

0.892

0.311

0.892

0.311

Literate household head

0.318

0.466

0.300

0.459

0.307

0.461

Land size in hectares

1.437

1.265

1.493

1.174

1.473

1.208

Tropical livestock unit

2.876

3.260

3.203

4.046

3.085

3.784

Observations

1524

2700

4224

Summary of all rounds with balanced data Age of household head

43.403

14.212

43.845

14.225

43.711

14.222

Male household head

0.903

0.297

0.894

0.308

0.896

0.305

Literate household head

0.313

0.464

5.776

0.464

0.314

0.464

Land size in hectares

1.375

1.257

1.463

1.251

1.437

1.254

Tropical livestock unit

2.710

3.672

3.173

4.194

3.033

4.048

Observations

3827

8806

12633

Source Own calculations

7.5 Results and Discussion To check whether the changes in poverty incidence can be attributed to roads, the marginal effects from the dynamic probit model for the probability of being in poverty are given in Table 7.4. Columns 1 and 2 present marginal effects of the standard dynamic random effects estimator and Columns 3 and 4 present Wooldridge’s (2005) maximum likelihood estimates. The parentheses give the asymptotic standard errors.

7 The Value of Roads in Rural Household Consumption

151

Table 7.4 Marginal effects of dynamic random effect probit model: poverty correlates Standard dynamic random effects model

Lagged poverty

With access to road

Year 2017

Wooldridge maximum likelihood model

(1)

(2)

(3)

(4)

0.311***

0.265***

−0.009

−0.008

(0.048)

(0.043)

(0.065)

(0.065)

−0.220***

−0.230***

−0.250***

−0.247***

(0.032)

(0.045)

(0.038)

(0.050)

−0.008

0.396**

−0.015

0.435**

(0.029)

(0.146)

(0.030)

0.054***

Age of household head

Age of household head squared/100

Male household head

Literate household head

Land size in hectares

Tropical livestock unit

Access in 2017

Age in 2017

Male in 2017

Literacy in 2017

Land in 2017

Livestock in 2017

(0.007)

(0.007)

−0.052***

−0.054***

(0.007)

(0.007)

0.404***

0.443***

(0.075)

(0.082)

0.036

0.025

(0.048)

(0.052)

−0.082***

−0.090***

(0.019)

(0.021)

−0.020*

−0.020*

(0.008)

(0.009)

−0.015

−0.035

(0.062)

(0.065)

−0.003

−0.003

(0.002)

(0.002)

−0.161

−0.199

(0.099)

(0.104)

0.047

0.090

(0.067)

(0.072)

0.008

0.006

(0.027)

(0.029)

−0.034***

−0.039***

(0.010)

(0.011) 0.354***

Initial poverty status σu2

(0.153) 0.057***

0.305***

(0.057)

(0.056)

−1.747***

−1.730***

−0.721***

−0.829***

(0.188)

(0.324)

(0.297)

(0.176)

Log likelihood

−5596.2

−5481.3

−5573.4

−5464.1

Observations

8422

8422

8422

8422

Note * p < 0.05, ** p < 0.01, *** p < 0.001 Source Own calculations

152

M. N. Simie

The coefficients are interpreted as marginal effects on the likelihood of being in poverty. Negative coefficients indicate a poverty alleviating effect. Columns 1 and 3 provide basic results where lagged poverty status, road access group, initial poverty status in 2015, and time dummies are included as explanatory variables. The marginal effects of access to roads are statistically significant in both standard and Wooldridge’s estimators, and households living in areas without road access are more likely to be deprived. Access to rural roads reduces the likelihood that a household is poor by 25% points. This shows that providing access to roads has a significant effect in reducing poverty, even within a short period of time. However, the result does not tell us why we observe this effect, but two general thoughts can be raised based on previous research. First, Melo et al. (2013) in their meta-analysis of the empirical evidence on the output elasticity of transport infrastructure, show that road development can have a strong impact when implemented in areas working in the primary sector. Given that many of the areas where the rural roads program has been introduced are engaged in the agricultural sector and that they were formerly with very limited road access, its significant effect within the short period of time is acceptable. Spillover effects and contamination are also expected to be minimal within this short period, resulting in a significant difference between the two groups. Second, Bryceson et al. (2008), show that rural road networks’ expansion in Ethiopia is bound to enhance accessibility rather than enhancing mobility. This means that rural local markets, which have links with urban centers, will be attracted to enter deep into the rural areas, and this in turn will enable households to access local markets easily. Columns 1 and 2 show that the coefficients of the lagged dependent variable are statistically significant in the standard dynamic random effect estimator. This implies that the longitudinal poverty dependence is wholly due to state dependence, with a marginal estimate of 31.1%. But the estimate becomes statistically insignificant, with a much lower magnitude and a different direction once the endogeneity of the initial condition is controlled for with Wooldridge’s model (Column 3). This is an interesting result, especially when considering the important role that roads might play in breaking the poverty state dependency curse. Moreover, the initial value of poverty status, given in Wooldridge’s estimator, is significant. This implies that there is a substantial correlation between the unobserved heterogeneity and the initial condition. In fact, the coefficient of initial poverty (0.354) is much larger than the coefficient of lagged poverty (−0.008). Thus, the estimates from Wooldridge’s model suggest that the longitudinal dependence is due to unobserved heterogeneity. To explicitly control for some observed heterogeneity, Columns 2 and 4 include other confounders. Except the lagged dependent variable, the estimates are very similar in the two models. Age appears to have a marginally significant negative linear relationship with the probability of being poor. To assess the possibility of a non-linear relationship, a quadratic component of age is included, and it is statistically significant. Male household heads are more likely to be out of poverty. As expected, the estimated results of both the models imply that accumulating more wealth in terms of land and tropical livestock units, is associated with reduced poverty. The time and variable interactions are also included to allow for a correlation between

7 The Value of Roads in Rural Household Consumption

153

unobserved heterogeneity and observed variables. There is no clear pattern of the coefficients and only tropical livestock units in 2017 were statistically different from zero at the 1% level. Estimates derived using unbalanced panels are reported in Appendix Table 7.5.

7.6 Conclusions Relating road improvements/construction with poverty dynamics could be far more than a descriptive tool as it may also hold the key to escaping spatial poverty traps. Areas with road access have a clear declining trend in poverty headcount, poverty gap, and poverty severity over time. Interestingly, gains are also made by low-income groups. Moreover, access to roads leads to a lower probability of falling into poverty. Households with access to roads have a 25% lower likelihood of being in consumption poverty. This result is robust to the model’s specifications and estimation techniques. Without controlling for the initial condition and unobserved heterogeneity, it is hard to accept any test for state dependency of poverty. The standard random effect estimator considerably overstates a household’s state dependence or a household’s risk of repeated poverty. This model indicates that the risk of falling into poverty in one year is noticeably higher if poverty is observed in the previous year even after controlling for observed heterogeneity. However, the continuing poverty spell is removed when the initial condition and unobserved heterogeneity are controlled for by using Wooldridge’s original auxiliary estimator. The varying results on state dependency coupled with the strong significance of road access on poverty reduction might be an indicator that roads are the key to escaping repeated poverty traps, and it is worth searching for other convincing methods to test its actual effects. It should be noted that a further disaggregation of road quality matters in estimating the trajectory of roads’ impacts. A casual observer travelling widely along these rural roads will be struck by the differences in road quality, even though they are accommodated in the same rural roads program. Thus, a further analysis considering the differences in the quality of these roads will have a potential policy benefit. This study focused on the effects of roads on consumptive poverty within the short-run, but it is worth investigating its effect in the medium and long-run where spillovers and time variant unobserved heterogeneity are expected to exist. Moreover, since roads have multidimensional effects on households’ activities, it will also be interesting to investigate their impact on multidimensional welfare.

Appendix See Table 7.5.

154

M. N. Simie

Table 7.5 Wooldridge maximum likelihood estimates with unbalanced data Wooldridge maximum likelihood model (1) Lagged poverty

With access to road

Initial poverty status

Year 2017

(2)

0.002

0.002

(0.065)

(0.065)

−0.243***

−0.237***

(0.038)

(0.049)

0.355***

0.305***

(0.057)

(0.056)

−0.014

0.420**

(0.030)

(0.152) 0.058***

Age of household head

(0.007) −0.056***

Age of household head squared/100

(0.007) 0.432***

Male household head

(0.081) Literate household head

0.029 (0.051) −0.090***

Land size in hectares

(0.021) −0.020*

Tropical livestock unit

(0.009) −0.042

Access in 2017

(0.065) −0.003

Age in 2017

(0.002) −0.189

Male in 2017

(0.103) Literacy in 2017

0.086 (0.071)

Land in 2017

0.006 (0.029) −0.038***

Livestock in 2017

(0.011) σu2

−0.743***

−0.856***

(0.179)

(0.191)

Log likelihood

−5632.3

−5520.3

Observations

8511

8511

Note * p < 0.05, ** p < 0.01, *** p < 0.001 Source Own calculations

7 The Value of Roads in Rural Household Consumption

155

References Aggarwal, S. (2018). Do rural roads create pathways out of poverty? Evidence from India. Journal of Development Economics, 133: 375–395. Alem, Y. G. Kohlin, and J. Stage (2014). The Persistence of Subjective Poverty in Urban Ethiopia. World Development, 56: 51–61. Retrieved from: ://WOS:000330496000004. https:// doi.org/10.1016/j.worlddev.2013.10.017. Ambel, A. P. Mehta, and B. Yigezu (2015). Multidimensional Poverty in Ethiopia: Changes in Overlapping Deprivations. Washington, DC: The World Bank. Aparna, B. P. Prasad, M. Cheranchery, and B. Maitra (2017). Effect of Development of Rural Roads and Feeder Service on Trip Characteristics of Households: A Case Study. A Journal of the Transportation Research Group of India (TRG), 3(2): 1–10. Asomani-Boateng, R. R. J. Fricano, and F. Adarkwa (2015). Assessing the socio-economic impacts of rural road improvements in Ghana: A case study of Transport Sector Program Support (II). Case Studies on Transport Policy, 3(4): 355–366. Batten, D. F. and C. Karlsson (1996). Infrastructure and the complexity of economic development. Berlin: Springer. Bryceson, D. F. A. Bradbury, and T. Bradbury (2008). Roads to Poverty Reduction? Exploring Rural Roads’ Impact on Mobility in Africa and Asia. Development Policy Review, 26(4): 459–482. Cowell, F. (2006). Microeconomics: principles and analysis. USA. Oxford University Press. Damania, R. C. Berg, J. Russ, Barra, A. Federico, J. Nash, and R. Ali (2017). Agricultural Technology Choice and Transport. American Journal of Agricultural Economics, 99(1): 265–284. Dawson, J. and I. Barwell (1993). Roads are not enough: new perspectives on rural transport planning in developing countries. London: Intermediate Technology Publications Ltd (ITP). Deaton, A. and S. Zaidi (2002). Guidelines for Constructing Consumption Aggregates for Welfare Analysis. Living Standards Measurement Study Working Paper No. 135. Washington, DC: The World Bank. Demenge, J. R. Alba, K. Welle, K. Manjur, A. Addisu, L. Mehta, and K. Woldearegay (2015). Multifunctional Roads: The Potential Effects of Combined Roads and Water Harvesting Infrastructure on Livelihoods and Poverty in Ethiopia. Journal of Infrastructure Development, 7(2): 165–180. Dercon, S. D. O. Gilligan, J. Hoddinott, and T. Woldehanna (2009). The Impact of Agricultural Extension and Roads on Poverty and Consumption Growth in Fifteen Ethiopian Villages. American Journal of Agricultural Economics, 91(4): 1007–1021. Edmonds, G. and J. de Veen (1992). A labour-based approach to roads and rural transport in developing countries. International Labour Review, 131(1992): 95–110. Fetter, F. A. (1904). The principles of economics, with applications to practical problems. New York: The Century co. Foster, J. J. Greer, and E. Thorbecke (1984). A Class of Decomposable Poverty Measures. Econometrica, 52(3), 761–766. Haughton, J. and S. R. Khandker (2009). Handbook on Poverty and Inequality: Washington, DC: The World Bank. Heckman, J. (1981). The incidental parameters problem and the problem of initial conditions in estimating a discrete time-discrete data stochastic proceess. In Structural Analysis of Discrete Data with Econometic Applications, Manski CF, McFadden D (eds). MIT Press: Cambridge, MA; 179–195. Hirschman, A. (1958). The strategy of economic development: Yale University Press. Hoyle, B. (1973). Transport and development. London and Basingstoke: The Macmillan Press Limited. Jefferson, P. N. (2012). The Oxford handbook of the economics of poverty. Oxford: Oxford University Press. Kam, S. P. M. Hossain M. L. Bose, and L. S. Villano (2005). Spatial patterns of rural poverty and their relationship with welfare-influencing factors in Bangladesh. Food Policy, 30(5): 551–567.

156

M. N. Simie

Khandker, S. R. Z. Bakht, and G. Koolwal (2009). The Poverty Impact of Rural Roads: Evidence from Bangladesh. Economic Development and Cultural Change, 57(4): 685–722. Khandker, S. R. and G. B. Koolwal (2011). Estimating the Long-Term Impacts of Rural Roads: A Dynamic Panel Approach. Policy Research Working Paper No. 5867. Washington, DC: The World Bank. Lugard, F. (1923). Dual mandate in British tropical Africa. Edinburgh. William Blackwood and Sons. Melo, P. C. D. J. Graham, and R. Brage-Ardao (2013). The productivity of transport infrastructure investment: A meta-analysis of empirical evidence. Regional Science and Urban Economics, 43(5): 695–706. Owen, W. (1959). Transportation and Economic Development. The American Economic Review, 49(2): 179–187. Porter, G. (2007). Transport planning in sub-Saharan Africa. Progress in Development Studies, 7(3): 251–257. Qin, Y. and X. Zhang (2016). The Road to Specialization in Agricultural Production: Evidence from Rural China. World Development, 77(C): 1–16. Rabe-Hesketh, S. and A. Skrondal (2013). Avoiding biased versions of Wooldridge’s simple solution to the initial conditions problem. Economics Letters, 120(2): 346. Ramessur, S. B. Seetanah, and S. Rojid (2010). Roads and Poverty: New Evidences from Africa. Journal of Poverty, 14(2): 166–182. Rammelt, C. F. and M. W. H. Leung (2017). Tracing the Causal Loops Through Local Perceptions of Rural Road Impacts in Ethiopia. World Development, 95: 1–14. Ravallion, M. (1998). Poverty Lines in Theory and Practice. IDEAS Working Paper Series from RePEc. Ravallion, M. and B. Bidani (1994). How Robust Is a Poverty Profile? The World Bank Economic Review, 8(1): 75. Sen, A. (1979). Personal Utilities and Public Judgements: Or What’s Wrong with Welfare Economics? The Economic Journal: the Quarterly Journal of the Royal Economic Society, 89(355): 537. Sen, A. (1983). Development: Which Way Now? Economic Journal, 93: 745–762. Stifel, D. B. Minten, and B. Koru (2016). Economic Benefits of Rural Feeder Roads: Evidence from Ethiopia. The Journal of Development Studies, 52(9): 1335–1356. The World Bank. (2015). Ethiopia’s Great Run: The Growth Acceleration and How to Pace It. Washington, DC: The World Bank. The World Bank. (2019). Ethiopia Economic Update 7: Special Topic—Poverty and Household Welfare in Ethiopia, 2011–16. In: World Bank, Washington, DC. Wooldridge, J. M. (2005). Simple solutions to the initial conditions problem in dynamic, nonlinear panel data models with unobserved heterogeneity. Journal of Applied Econometrics, 20(1): 39–54.

Part III

Financial Services, Employment and Corporate Governance

Chapter 8

Corporate Social Responsibility Practices and Motivations in a Least Developed Country Yohannes Workeaferahu Elifneh

Abstract Corporate Social Responsibility (CSR) is one of the least explored subjects in the context of the least developed countries (LDCs). Most studies that have been done on CSR predominantly demonstrate CSR practices in the developed world. There is a dearth of academic CSR research based on empirical evidence from the developing world. This chapter presents the results of an empirical study (using qualitative case studies, which is a widely used approach for exploring contemporary issues such as CSR) to show how companies operating in Ethiopia’s brewery sector engage in CSR activities and their motivations for doing this. It discusses three companies operating in the Ethiopian brewery sector using a narrative lens for the analysis. Data was collected mainly through interviews (with managers, employees, members of the local community, and other stakeholders such as government representatives and experts from NGOs). The study found that nascent CSR practices are being carried out by these three companies, which focus on the social dimension of CSR paying lesser attention to their environmental responsibilities. Given the limited number of studies on CSR in developing countries based on empirical evidence, this study contributes to this domain both theoretically and empirically by extending CSR literature to lesser explored regions of the world. Keywords Corporate social responsibility · CSR practices · Multinationals · Least developed countries · Ethiopia JEL Codes M14 · K20 · L20 · Q01

8.1 Introduction Scholars have noted that “in recent decades, a growing number of academics as well as top executives have been allocating a considerable amount of time and resources to CSR” (Cheng et al. 2014: 2). Consequently, this area of research has become relevant Y. W. Elifneh (B) Department of Management, Addis Ababa University, Addis Ababa, Ethiopia e-mail: [email protected]; [email protected] © Springer Nature Singapore Pte Ltd. 2020 J. Wood and O. Habimana (eds.), A Multidimensional Economic Assessment of Africa, Frontiers in African Business Research, https://doi.org/10.1007/978-981-15-4510-8_8

159

160

Y. W. Elifneh

(Dincer and Dincer 2007). “It is undisputed that CSR has achieved a prominent place in management practice and in the academic arena” (Melo and Garrido-Morgado 2012: 1) and it “has become the norm rather than the exception” (Husted and Allen 2011: 13). Today, CSR has found a place in the mainstream modern day corporate world (Bolton and Mattila 2015). Ioannou and Serafeim (2015: 2) explain, “in recent years, a growing number of companies are adopting various CSR initiatives.” This perspective affirms that CSR activities are inevitable in the modern-day business world. CSR is concerned with issues such as “environmental protection, health and safety at work, relations with local communities, human rights, and corruption,” (Proença and Branco 2014: 252). Graafland and Mazereeuw-Van der Duijn Schouten (2012), identified two broad, major motives for CSR activities involving extrinsic (such as financial) and intrinsic motives (ethical and altruistic). So far, there have been no substantial studies and hence there has been no substantial empirical evidence on CSR activities in the developing world. Utting (2003), asserts that it is unfortunate that much of the evidence for and against CSR in the developing world is based on suppositions, anecdotes, and limited examples of best or bad practices. The author argues that there has not been sufficient systematic research on CSR in the developing world. Egri and Ralston (2008: 324) add, “most studied countries in international CR research have been in high economically developed countries in North America, Western Europe, and East Asia; there has been much less CR research in the less developed and transitional economies of the world.” This study illustrates the major CSR practices that some selected companies follow in an LDC context and their motives for undertaking these activities. Studies on CSR in this part of the world are interesting because the nature of CSR in developing countries is different from that in the developed world as CSR is contextualized and locally formed (shaped) in the developing world (Jamali and Karam 2018). Besides a dearth of CSR research in the developing world (Blowfield and Frynas 2005; Egri and Ralston 2008; Gugler and Shi 2009; Jamali and Mirshak 2007; Utting 2003; Visser 2008), CSR scholars talk about the importance of studying what type of CSR practices are undertaken in the developing world and why. For example, Frynas (2006), maintains that CSR activities vary from context to context as CSR practices develop within a particular social context reaffirming that it is necessary to explore CSR activities in various contexts. Similarly, Dobers and Halme (2009: 238)state “societies are different in many respects implies that CSR can have different faces in different societal contexts.” The findings of this study contribute to knowledge and understanding of CSR and its activities by companies operating in a developing country context where there is little/no hard law enforcing the application of CSR activities. It discusses what companies operating in the brewery sector in Ethiopia are doing in terms of CSR in an LDC context. Given the lack of knowledge about CSR in this part of the world, this research contributes to improving our knowledge of CSR activities. The breweries belong to the alcohol industry, which represents a controversial business whose activities targeting consumers dominate its CSR agenda, making

8 Corporate Social Responsibility Practices and Motivations …

161

CSR activities particularly decisive for them (Cai et al. 2012; Jo and Na 2012; Reast et al. 2013). This study uses a case study approach to explore the following research questions: What are the major CSR activities that the case companies undertake in the local context? Why do the companies engage in CSR activities? What are the benefits and challenges of implementing CSR activities in the local context? The rest of the chapter is organized as follows: the next section highlights the theoretical background, which is followed by the research methodology. This will be followed by sections that discuss CSR activities undertaken by the companies, the motives behind undertaking them, and the benefits and challenges in implementing them. This later section will also outline concerns about CSR activities. Finally, the last section gives conclusions and discusses the findings. The study makes both theoretical and empirical contributions. Its theoretical contribution is by way of extending CSR literature to less charted territories while empirically it uses a novel dataset in its empirically anchored methodology which in itself is a valuable contribution. Confirming that research on CSR activities in developing countries is still underdeveloped, Visser (2008: 493) concluded, “Hence, there is an urgent need for further research on CSR in developing countries at the international, regional, national and sectoral levels.”

8.2 Theoretical Background 8.2.1 CSR’s Span The basic idea underlying CSR activities is that organizations are expected to exert efforts that go beyond playing an economic role in society (Robins 2008). Currently, there are organizations that are assuming responsibilities that transcend their economic activities. CSR encompasses numerous issues that relate to an organization’s behavior in its social environment beyond the solely economic dimensions with which companies are customarily associated (Rigoberto Parada Daza 2009). Idowu (2009: 13) states, “it is widely acknowledged that modern corporations have some social responsibility towards society; even the most adamant opponents of CSR agree with this assertion.” Zheng et al. (2015: 392), explain that CSR practices may include “offering environmental products (responsibility toward customers), applying environmentally sound business practices (responsibility toward the local community), offering adequate labor conditions and just wages (responsibility toward employees), and supporting responsibility measures in the supply chain (responsibility toward suppliers).”

162

Y. W. Elifneh

8.2.2 Motivations for Undertaking CSR Activities According to Vogel (2006: 2), there are “many reasons why some companies choose to behave more responsibly or virtuously in the absence of legal requirements. Some are strategic, others are defensive, and still others may be altruistic or public-spirited.” This indicates that even in countries, where there is no a strong legal framework for CSR, such as in this study’s context, companies can still implement CSR activities for reasons other than governmental requirements. The point is that “CSR should make sense from the perspective of the overall competitive strategy of a firm (and the other way around), and should be treated as an integral part of it; not only because this furthers the long-term survival of a firm, but also because this way the moral claims of stakeholders have the best chance of becoming an accepted part of the firm’s decision-making structure and its organizational culture” (Van de Ven and Jeurissen 2005: 300). According to Graafland and Mazereeuw-Van der Duijn Schouten (2012), there are three main motives for undertaking CSR activities: (1) financial as an extrinsic driver, (2) ethical (moral duty), and (3) altruistic (commitment to fostering the well-being of others); the last two are considered intrinsic motives. Garriga and Mele (2004), indicate that the motivation for undertaking CSR activities may arise because of profit making, for addressing regulatory gaps, for serving social demands, or for fulfilling ethical obligations. Thus, motives for CSR activities can be intrinsic or extrinsic. However, CSR experts show that organizations’ motives for undertaking CSR are different in an underdeveloped country setting as compared to developed countries (Gugler and Shi 2009; Visser 2008). According to Campbell (2007: 947), the intention behind undertaking CSR activities needs to be explored contextually, as “the tendency toward socially responsible corporate behavior varies across countries and much more research is required to understand why.” Huang and Watson (2015: 6), add, “research on determinants of CSR is important on its own.” More so, due to the normative nature of CSR, there are no clear guidelines in literature on CSR activities which explain why businesses undertake these activities. Literature on CSR does not provide one single answer to the question of why businesses undertake CSR activities. Proença and Branco (2014: 253) state, “the motivations for engaging in CSR initiatives differ significantly, and to understand why firms engage in CSR activities it is increasingly necessary to integrate different theoretical perspectives.” Literature on CSR indicates that there are various CSR sources/drivers such as traditional marketing motives/a business case for CSR (like making profits) (Beltratti 2005; Carroll and Shabana 2010; Fombrun et al. 2000; Grigore 2011; Iwu-Egwuonwu and Chibuike 2010; Kurucz et al. 2008; Luo and Bhattacharya 2006; Vogel 2006); a company’s internal governance mechanisms (Beltratti 2005; Harjoto and Jo 2011; Hollender 2004; Jamali et al. 2008; Jo and Harjoto 2012; Ntim and Soobaroyen 2013; Shahin and Zairi 2007; Spitzeck 2009); a company’s values (Crook 2005; Hollender 2004; Sun et al. 2010; Thomsen 2004; Vuta et al. 2007); pressures and expectations of the general public, which are usually considered secondary stakeholders (Campbell 2006; Freeman et al. 2006; Frynas 2006; Gjølberg

8 Corporate Social Responsibility Practices and Motivations …

163

2011; Huang and Watson 2015; Ribstein 2005; Ward 2004); the law (domestic regulations) (Bantekas 2004; Campbell 2006; Lambooy 2014; McBarnet 2007; Vogel 2006); and stakeholders, that is primary stakeholders’ demands and interests (Ali et al. 2010; Corbett 2008; Dawkins and Lewis 2003; Freeman et al. 2006; Millon 2010; Pondeville et al. 2013; Proença and Branco 2014; Serafeim 2013).

8.2.3 Benefits and Challenges in Implementing CSR Activities Literature on CSR indicates that both benefits and barriers exist in the implementation of CSR activities (Branco and Rodriques 2007; D’Amato et al. 2009). Some of the benefits of implementing CSR activities include ‘cost and risk reduction,’ ‘achieving operational efficiency,’ ‘achieving competitive advantages so that the business can perform better financially,’ ‘building and defending reputation and legitimacy,’ and ‘promoting innovations and helping improve the welfare of society’ (Kurucz et al. 2008; Zadek 2000). Besides, CSR also has benefits—it “may attract new customers (socially conscious customers, ‘green’ consumers, etc.), increase the companies’ profitability, and enhance their competitiveness, companies may engage in CSR to improve their efficiency and enhance their reputation, brand, and trust” (Flammer 2015: 2551). Further, “some of the benefits of CSR include better corporate and brand reputation, recruitment of superior talent, less employee turnover, and improved risk management” (Hatch and Stephen 2015: 65). Generally, advocates of CSR maintain that engaging in CSR pays off for a business and its stakeholders in terms of improving customer loyalty, enhancing employee loyalty and morale, improved productivity, positive relations with regulators, public relations and marketing advantages, and first-to-market or leadership benefits (Burke and Logsdon 1996). However, barriers/challenges that affect the implementation of CSR activities also exist. A list of CSR challenges is provided in Berad (2011: 104–105) which includes ‘lack of community participation in CSR activities,’ ‘need to build local capacities,’ ‘issues of transparency,’ ‘non-availability of well-organized non-governmental organizations,’ ‘visibility factor,’ ‘narrow perceptions towards CSR initiatives,’ ‘non availability of clear CSR guidelines,’ and ‘lack of consensus on implementing CSR issues.’ Besides, Lund-Thomsen et al. (2016), have identified what they term “key obstacles” in implementing CSR initiatives as lack of willingness among developing countries’ governments to enforce national labor and environmental laws; economic development imperatives often over-ride social or environmental considerations in such countries; businesses take on an irresponsible attitude seeking to obstruct the enforcement of environmental laws or pollution control initiatives often by bribing public officials; and lack of organized unions that promote better work conditions and benefits.

164

Y. W. Elifneh

8.3 Methodology This study uses a qualitative research approach. According to Campbell (1999: 532), “the simplest way to define qualitative research is to note that the results are primarily expressed with words.” Cassell and Symon (2006: 6) maintain that the qualitative methodology has “a long history and tradition within organization and management research and provides powerful tools for research in this area.” It can be used in research projects that focus on organizational studies (Strauss and Corbin 1990). Accordingly, this study follows a qualitative approach with multiple case studies to explore the CSR practices of breweries in Ethiopia. The qualitative approach to inquiry is a comparative multiple case study approach. Multiple qualitative case studies “address the same research question in a number of settings using similar data collection and analysis procedures in each setting” and the evidence generated from multiple case studies is “quite robust” (Herriott and Firestone 1983: 14). The case study approach is suitable for organizational and management studies like this one (Yin 2003a). In general, case study research involves studying an issue explored through one or more cases within a bounded system, that is, a setting or a context (Creswell 2007). Eisenhardt (1989: 532) maintains it is “a research strategy that focuses on understanding the dynamics present within single setting.” Case studies may involve a single case or multiple cases (Yin 2003a). As this empirical study focuses on CSR practices in a specific sector, the brewery sector in Ethiopia, it considers three brewing companies as cases. The case study approach allows us to study the CSR engagements of the companies in a given context holistically. As CSR is a contemporary issue in a local setting, the case studies are an appropriate research design for understanding this contextually contemporary phenomenon (Yin 2003a). This approach provides detailed information and opportunities for an intensive analysis of the CSR practices of the three companies. As the study follows a multiple case study approach, it also facilitates a comparative analysis of the CSR practices of the three companies. This means that the focus of the study is describing, exploring, and understanding the CSR practices of the cases. The study’s aim is arriving at an in-depth and overall picture of the issue as a whole. In doing so, its aim is not to condense its findings into some form of theoretical explanations. The case studies themselves are the result. To make the findings more robust, the research applies a comparative analysis by examining several cases to understand the similarities and differences between them (Baxter and Jack 2008). Its goal is not primarily to summarize and generalize. In this context, Flyvbjerg says (2006: 23): Case stories written like this can neither be briefly recounted nor summarized in a few main results. The case story is itself the result. It is a ‘virtual reality,’ so to speak. For the reader willing to enter this reality and explore it inside and out the payback is meant to be a sensitivity to the issues at hand that cannot be obtained from theory.

Flyvbjerg further reiterates, “often it is not desirable to summarize and generalize case studies. Good studies should be read as narratives in their entirety” (p. 25). Peattie (2001: 260) says, “it is simply that the very value of the case-study, the

8 Corporate Social Responsibility Practices and Motivations …

165

contextual and interpenetrating nature of forces, is lost when one tries to sum up in large and mutually exclusive concepts.” The empirical case studies focus on a study/analysis of the CSR practices of three companies (breweries) in Ethiopia; two of them are multinationals (coded Case #1 and Case #3 for anonymity and confidentiality) and the remaining one is a domestic company/brewery (coded Case #2) with emphasis on what CSR practices they are undertaking, why are they doing so, and what are the benefits and challenges of implementing CSR activities in the local setting. Data was obtained from various sources by way of interviews. Interviews are considered a vital source of information for case studies (Yin 2003a). This study used guided open interviews and the responses stayed open ended enabling the participants to provide as much information as they wanted. This also gave the researcher an opportunity to probe further while ensuring consistency at the same time (EasterbySmith et al. 2008; Turner 2010). This was largely based on the conviction that this approach will allow the participants to describe their experiences in their own words without being forced to follow any framework. Accordingly, the CSR practices of the breweries were found out by interviewing the corresponding managers of the case companies in the study and as explained by their stakeholders. Employees working in the breweries, NGOs, government agencies, and suppliers were also interviewed. The participants were selected purposefully (non-probabilistic sampling), based on their knowledge about the subject and whether they were affected by the breweries’ practices positively or negatively. Such a triangulation of data sources is decisive. This aligns well with what Eisenhardt and Graebner (2007: 28) suggest: “in particular to interview method, the challenge to interview data can be mitigated by data collection approaches that limit bias. The key to this is to use numerous and highly knowledgeable informants who view the focal phenomenon from diverse perspectives. These informants can include actors within and outside the cases, but have some knowledge about the study subject and the cases.” This research used purposive sampling in gathering data from the participants who were knowledgeable about the subject. With regard to the sample size, there are no particular “guidelines to determine a non-probabilistic sample size and instead purposive samples are widely used and the size of the sample depends on the concept of saturation, or the point at which no new information is observed in the data” (Guest et al. 2006: 1). Judged by this criterion, the number of respondents in this research was sufficient for all the sub-categories that were discerned. Forty-four interviews were done, including with six managers from the three companies (two each), nine workers (three workers from each company), 12 experts from various government offices, seven interviews with six different NGOs, nine farmers whom the companies cooperated with, and four academics (educators in universities) from the three largest universities in Ethiopia. The interviews were conducted by the researcher and their duration ranged between 1–2 h. The interviews were tape recorded and the transcriptions of the interviews were done by professional transcribers. Utmost care was taken to ensure that the transcriptions were done accurately by comparing the transcriptions and listening to the recordings repeatedly. Then the collected data was analyzed using a narrative approach. This multiple case

166

Y. W. Elifneh

study research did a narrative analysis, which is one of the most established ways of doing qualitative research (Savenye and Robinson 1996). It uses rich descriptions and narratives. This is in line with the notion that a “narrative analysis as a research tool has become increasingly useful in organizational studies and the method is useful in the analysis of interview data” (Easterby-Smith et al. 2008: 182). Feldman et al. (2004: 147) assert, “considerable scholarship in the field of organizations and management has recognized the importance of the narrative lens.” According to these authors, in most qualitative studies, “we are likely to find our research informants providing us with information by means of narrative” (p. 148); this itself facilitates using a narrative analysis as the information provided in the narrative is valuable (Feldman et al. 2004). In this connection, Smith (2000: 328), states that a narrative analysis allows for a holistic approach by preserving context and particularity as “narratives yield information that may not be available by other methods.” Smith also suggests that in applying a narrative analysis “the researcher may decide to use one or more existing analysis systems, adapt an existing system, or develop a new one,” (p. 331). According to Feldman et al. (2004), there are lots of different ways of doing a narrative analysis as there is no single definitive way of doing it. This is also in line with what Georgakopoulou (2006), suggests that narrative research is often described as a rich and diverse enterprise. Further, according to Riessman (2005), several typologies of a narrative analysis exist so different approaches can be combined for doing a narrative analysis. He also says that a narrative analysis can also be applied in a thematic analysis, in which “narratives are grouped into a similar thematic category” (p. 3), noting that central to the notion of a thematic analysis is identifying all data that relates to the already classified patterns to combine and catalogue related patterns into themes obtained from or supported by literature (Aronson 1995). In brief, Franzosi (1998: 517) stated that narrative is international and it is simply there, like life itself.” Furthermore, case study research too can be very qualitative and narrative in form (Kaarbo and Beasley 1999). Overall, this study is an empirical research based on qualitative data gathered mainly through interviews. The specific goal of the analysis is finding answers to the research questions. Therefore, in doing a comparative (cross-case) analysis of the case studies, it uses qualitative tables. Tables that contain textual data based on a uniform approach are a convenient way of displaying text from across the whole dataset in a way that makes systematic comparisons easier (Gibbs 2008; Yin 2003b).

8 Corporate Social Responsibility Practices and Motivations …

167

8.4 Findings 8.4.1 CSR Practices The findings of this study show that there were common and distinct types of CSR activities that the case companies undertook for their respective stakeholders: shareholders/owners, employees, the local community, customers, farmers, and the environment. Tables 8.1, 8.2 and 8.3 give an overview of the major CSR practices along with the targets (beneficiaries) and the broad areas of CSR practices. The study’s results also show that most of the CSR practices that the companies undertake in the local context focus on social aspects such as education and healthcare; creating employment opportunities for the local population; employees’ development and learning; improving workplace health and safety; medical and insurance coverage for employees, promoting responsible consumption; building the capacities of small scale farmers through contract farming and sponsorships; extending support to disadvantaged communities with scholarship programs; and women’s empowerment, and supporting street children and persons with disabilities.

8.4.2 Motives for CSR (CSR Sources) This study identified marketing motives (the business case for CSR), the companies’ values and internal governance mechanisms, as well as responsiveness to stakeholders’ demands and interests as the main sources/drivers of CSR engagements in the case companies. In particular, Case #1 and Case #2 did not depend on any form of international CSR tools (CSR initiatives such as UN Global Compact). Coupled with the absence of any local CSR law, their values and internal governance mechanisms were the leading sources of their CSR activities. However, in Case #3, international CSR tools or initiatives were partly the reason why the company engaged in CSR activities. This company observed and strove to implement several international CSR tools. In other words, although the study found that there were no strong and organized local pressures from local NGOs, media, trade unions, and civic society campaigners for businesses to do more CSR activities in the local context, unlike the other case companies, the Case #3 company tried to apply some international CSR tools as a CSR source. Regarding the law (domestic CSR regulations) the study found that there was no specific CSR law that required businesses to practice some form of CSR. The case companies were undertaking CSR activities in the local context on a voluntary basis. All the case companies stated that there was no direct CSR law in the local setting that required businesses to implement CSR initiatives. The interviews with law enforcing authorities also confirmed that at present there is no specific CSR policy or law in the country. The country only has scattered laws, which partly relate to CSR such as

168

Y. W. Elifneh

Table 8.1 Overview of the case companies’ CSR practices and broad areas—Case #1 Case company

Major CSR practices

Targets

Broad area

Case #1

Providing necessary information when needed

Shareholders

Transparency

Providing healthcare coverage and accident insurance, occupational safety and health, annual bonus and holiday gifts, transportation service/allowance, training and education opportunities

Employees

Social

Creating employment opportunities for the local population, building (a few) public schools, sponsoring social, cultural, sports, and entertainment events, community dialogues, participating in women’s empowerment projects, supporting persons with disabilities and providing employment opportunities to disabled persons, extending support to street children, and promoting and sponsoring art and art related work

Local community

Social

Promoting responsible consumption by communicating messages, training on hygienic practices, and training on responsible business practices

Customers

Social

No established relationships with the defined group of farmers to source locally and support the productivity of farmers but does the following for farmers: creates awareness about HIV prevention and care in farmers’ communities, motivates farmers to improve productivity by distributing small farming tools to outstanding farmers, and financial contribution to a local agricultural institute to support its research and development programs on improved quality barley seeds

Farmers

Social

Planting trees and/or participating in green and clean initiatives, waste water treatment plants to treat the breweries’ effluents, using part of the waste as fertilizers for local farmers/communities, using by-products such as spent grain to feed cattle in the local community, and collecting carbon (CO2 ) to reuse it as an ingredient in the production process

Environment

Environment

environmental proclamations, labor proclamations, and human rights proclamations that businesses are expected to consult and abide by. However, these proclamations and laws are not enforced effectively. According to Sacconi (2007: 80), “labourmarket laws and environmental regulations establish a general legal framework, they cannot regulate every detail of firms’ decisions. They may lay down compulsory

Shareholders Employees

Local community

Customers Farmers

Environment

Financial reports are provided on a monthly basis

Providing healthcare coverage and accident insurance, occupational safety and health standards and measures, annual bonus, holiday gifts, transportation service/allowance, and training and educational opportunities

Creating employment opportunities for the local population, sponsoring social, cultural, sports, and entertainment events, community dialogues, access to clean water, industry-university partnerships, building houses for the poorest of the poor, taking care of orphans, and executing HIV prevention, care, and support programs

Promoting responsible consumption—communicating messages, sugar free beer, computerized cleaning, and consumer feedback systems

Market access, subsidizing prices of improved quality seeds, agronomical experts who help improve farmers’ productivity, offering early prices to barley growers, and some transportation services to transport harvest produce for barley farmers that it works with

Planting trees and/or participating in green and clean initiatives, waste water treatment plants to treat the breweries’ effluents; using part of the waste as fertilizers by the local farmers/communities, using by-products such as spent grain to feed cattle in the local community and collecting carbon (CO2 ) to reuse it as an ingredient in the production process, germinating and developing tree seedlings on an industrial scale to help support reforestation efforts across the country, erosion control activities, and recycling water for cleaning purposes

Case #2

Targets

Major CSR practices

Case company

Table 8.2 Overview of the case companies’ CSR practices and broad areas—Case #2

Environment

Social and some environmental

Social

Social

Social

Transparency/social

Broad area

8 Corporate Social Responsibility Practices and Motivations … 169

Employees

Local community

Customers Farmers

Environment

Financial and CSR/sustainability reports

Providing healthcare coverage and accident insurance, occupational safety and health standards and measures, annual bonus, holiday gifts, transportation service/allowance, and training and educational opportunities

Creating employment opportunities for locals, sponsoring social, cultural, sports, and entertainment events, community dialogues, access to clean water, sanitation facilities for vulnerable communities, administering scholarship programs to assist disadvantaged local students, and working on road safety activities in the local area

Promoting responsible consumption—communicating messages, educating the public about the dangers of drunk driving, and no promotion activities near schools

Implementing contract farming with local farmers: access to credit, improved quality seeds and agronomical support, training, and business advice; access to markets, premium prices, and some transportation facilities to transport their harvest and distribution packages

Planting trees and/or participating in green and clean initiatives, waste water treatment plants to treat the company’s effluents, using part of the waste as fertilizers by the local farmers/communities, using by-products such as spent grain to feed cattle in the local community, collecting carbon (CO2 ) to reuse it as an ingredient in the production process, and generating bio-fuel as an energy source for cooking purposes

Case #3

Targets Shareholders

Major CSR practices

Case company

Table 8.3 Overview of the case companies’ CSR practices and broad areas—Case #3 Broad area

Environment

Social and some environmental

Social

Social

Social

Transparency/social

170 Y. W. Elifneh

8 Corporate Social Responsibility Practices and Motivations …

171

conditions, but in many settings their application requires interpretation of a ‘grey’ zone; or else compliance with them may not be observable.” With regard to whether the case companies considered stakeholders as a CSR source, the findings show that the case companies engaged in some forms of CSR practices to be responsive to the interests and demands of their respective stakeholders. However, the results also show that although the companies claimed that they did CSR activities to satisfy their stakeholders in a proactive manner, there was no a strong involvement of the stakeholders in the companies’ CSR agendas (decision making processes) in the local context. This is in line with what (Gugler and Shi 2009: 7) add, “stakeholders in developing countries have been object of CSR initiatives rather than active subject in shaping the CSR agenda.” Further, Blowfield and Frynas (2005: 507), note, “the success of CSR initiatives is often linked to stakeholder dialogue and stakeholder engagement.” Overall, commercial and reputational benefits, better acceptance, and an improved image were a part of the extrinsic sources (drivers) of CSR that the study identified as reasons for case companies engaging in CSR activities. On the other hand, the companies’ values of giving back, respecting, and creating shared values were intrinsic motives that served as CSR sources (drivers) in the local context.

8.4.3 Benefits and Challenges in Implementing CSR Based on the interviews with the companies’ managers, the study found that the benefits of implementing CSR in the local setting were improving their image and reputation, improving competitiveness, and helping them achieve achieving better acceptance. The barriers/challenges that the case companies encountered in implementing CSR activities include lack of or limited awareness about CSR in the local setting, overwhelming social problems, and bureaucratic inefficiencies in government offices. Besides this, the absence of a CSR framework at the national level was also identified as an additional barrier in CSR implementation. Further, there were severe criticisms against the companies by the employees, farmers, NGOs, experts from government offices, and academicians, who participated in this study. These represent additional barriers to the successful implementation of CSR. These concerns are presented in the following paragraphs. The main concerns shared by employees include lack of fairness and equity due to discrimination either as a result of favoritism, nepotism and personal bias or because of the segregation of the employees into ‘new blood’ and ‘old blood’ groups. This led to providing preferential treatment to some employees in terms of salaries, benefits, and promotion opportunities. The views of the employees, who criticized this discrimination, indicate that such a divisive approach is iniquitous and has a negative impact morally, psychologically, and financially on the employees, who are unfavorably treated. The other main concern that the employees shared is that there was a wide salary gap between employees and managers. The employees called upon their respective companies to narrow this excessive salary gap. Besides, the findings of

172

Y. W. Elifneh

this study also show that employees were concerned with the management’s lack of openness in considering their views even for those decisions that affected their interests. Employees also complained about lack of participation in CSR; lack of family medical care; work pressures due to heavy workloads; and systematic control and oppression of employees by division managers and supervisors, who in return got attractive salaries and benefits. They also shared grave concerns that relate to violations of their rights to association. This could either be because of a conflict between the management and employees’ associations due to the management’s failure to renew collective agreements or because the management systematically influenced and manipulated the leaders of the employees’ associations by providing attractive benefits to them. Employees’ right to form associations was threatened as the management did not support the employees in having an association. Finally, employees also said that there was no proper handling of their complaints and grievances. Farmers also criticized and expressed concerns about the companies that they worked with. The farmers’ concerns ranged from lack of robust agronomical support and unsatisfactory prices for their barley to lack of recognition as individual farmers and instances of discontinuation of support from at least one of the breweries without explanation, and incidents of a number of livestock being killed due to the waste released by a brewery that contaminated the nearby river. The main concerns expressed by experts from NGOs, government authorities, and academicians stressed that the brewery sector employed aggressive and enticing advertising and promotional campaigns without considering their audiences’ age. They also expressed worries over the use of water resources by the breweries from various sources—underground water, rivers, and spring water. This consumption was not sustainable and the amount of water consumed by the breweries from these sources largely remained unaccounted for. This was further exacerbated by the lack of serious water recycling practices among the breweries because most of the breweries have plants that operate in water distressed areas of the country. Besides, these critics also criticized the brewery sector for its poor product labels and lack of consumer information. They also expressed their worries that alcoholic products made by the breweries contributed to road accidents in the country due to drunk driving. Hence, more needs to be done in creating awareness about responsible consumption. In this connection, critics pointed out that the breweries did not practically follow up the implementation of their responsibility messages and were instead concentrating on competition in the local market. The commentators criticized the breweries for their lack of engagement in serious and robust environment protection activities in the areas that they operate in.

8.5 Discussion and Conclusion According to the results of this study, more weight has been given to the social aspect of CSR as compared to its environmental aspect. Even so, in the face of the enormous social problems that exist, what has been done by these companies

8 Corporate Social Responsibility Practices and Motivations …

173

thus far is not substantial. But their efforts can be replicated by other businesses in the country. There are also important environmental protection practices that are undertaken by the companies in the brewery sector such as waste water treatment, participation in campaigns like those on planting trees and clean and green initiatives, and germinating and developing tree seedlings on an industrial scale. These are meant to support reforestation and afforestation activities throughout the country. As Blowfield and Frynas (2005: 2050) say, “CSR is a work in progress and that, rather than condemning the good for being imperfect, we should examine the processes under way.” However, more work still needs be done with regard to CSR. The brewery sector, which hosts prominent multinational companies, may have to get ready for more robust CSR initiatives. These initiatives could address chronic poverty and associated social problems as well as declining environmental conditions. This conclusion is drawn after considering that the study context—with a population size of over 100 million people—is typical in parts of the developing world, where there is “the bulk of the world’s population and the lion’s share of global social and environmental problems” (Jamali and Neville 2011: 599). Particularly, in light of the results of this study based on interviews (with NGOs, academicians, experts from government offices, employees, and local farmers that the case companies work with); which demand the necessity and importance of businesses making positive contributions to society and nurturing the environment in the territories that they operate in, as well as based on CSR theories which stress that businesses should accept that they must play more than just an economic role in society (Idowu 2009; Rigoberto Parada Daza 2009; Robins 2008), engage in various CSR practices that focus on environmental protection, building relationships with local communities and local suppliers, and promoting human rights (Dahlsrud 2008; Freeman et al. 2004; Joireman et al. 2015; Kotler and Lee 2005; Moir 2001; Proença and Branco 2014; Russo and Tencati 2009; Sen and Bhattacharya 2001; Sprinkle and Maines 2010); provide infrastructure for the local community (Gifford et al. 2010; Tsang et al. 2009) and implement contract farming, which encompasses providing support to small scale farmers that the businesses work with (Eaton and Shepherd 2001); engage and educate customers, ensuring availability of information to the consumers and doing truthful advertising, and not advertising for children (Bokhodir and Iroda 2010; Heslin and Ochoa 2008; Hong and Xiaoli 2010; Porter and Kramer 2006); ensure health and safety at work and preserve employees’ rights through fair treatment, eliminating discrimination with respect to employment and occupation, freedom of association and the effective recognition of the right to collective bargaining (Bauman and Skitka 2012; Crane et al. 2008; Ellerup et al. 2009; Global Compact 2000; ILO 1998; Jain and Jain 2013; Proença and Branco 2014; Ruggie 2008)—the society, where this study was conducted, expects for more serious CSR interventions from businesses including those businesses which operate in the brewery sector in specific areas such as fight against extreme poverty and joblessness; combating alcohol and drug abuse; and (more) integration with farmers so that the businesses commit themselves to the grassroot levels, and work towards addressing prevailing issues of gender-based violence in the smallholder farming communities. They should also

174

Y. W. Elifneh

address the problems related to employees’ segregation into distinct groups such as ‘new blood’ and ‘old blood’ groups which provide preferential treatment and better salaries and benefit packages to one group of employees over others.. The companies should also strive to help fill gaps related to the use of technology/knowledge, skills, and technology transfers to smallholder farmers and local employees; giving due consideration to water resource development activities and refraining from aggravating water shortages through excessive and unsustainable water consumption; and executing robust participation in addressing environmental problems that require reforestation and restoration of degraded areas. They should also embark on community development programs that target children, women, older persons, and persons with disabilities and make serious commitments to promoting responsible consumption and protecting children from beer promotions. They should address the problem of poor consumer information about the products by using improved product labels; work closely with law enforcement offices to ensure that the minimum age required for alcohol consumption is made practical; and work towards addressing the severe problems associated with drunk driving (which is one of the major causes of road accidents in the country). These companies can also establish alliances and partnerships with universities and research institutes and support the development of the education sector by donating books and teaching aids to rural schools. They also need to improve transparency about their activities and social and environmental impacts of the brewery sector.

References Ali, I. K. U. Rehman, S. I. Ali, J. Yousaf, and M. Zia (2010). Corporate social responsibility influences, employee commitment and organizational performance. African Journal of Business Management, 4(12): 2796–2801. Aronson, J. (1995). A pragmatic view of thematic analysis. The Qualitative Report, 2(1): 1–3. Bantekas, I. (2004). Corporate social responsibility in international law. Boston University International Law Journal, 22(309): 312–317. Bauman, C. W. and L. J. Skitka (2012). Corporate social responsibility as a source of employee satisfaction. Research in Organizational Behavior, 32: 63–86. Baxter, P. and S. Jack (2008). Qualitative case study methodology: Study design and implementation for novice researchers. The Qualitative Report, 13(4): 544–559. Beltratti, A. (2005). The complementarity between corporate governance and corporate social responsibility. The Geneva Papers on Risk and Insurance—Issues and Practice, 30(3): 373–386. Berad, N. R. (2011). Corporate social responsibility—Issues and challenges in India. Paper presented at the International Conference on Technology and Business Management, MET Institute of Management, Nasik, March, 28–30. Blowfield, M. and J. G. Frynas (2005). Setting new agendas: Critical perspectives on corporate social responsibility in the developing world. International Affairs, 81(3): 499–513. Bokhodir, A. and K. Iroda (2010). ‘Feasibility of corporate social responsibility activities practiced by SME’s in Uzbekistan: A stakeholders’ perspective’, in G. Aras and D. Crowther (eds), A handbook of corporate governance and social responsibility. Aldershot: Gower Publishing Limited & Ashgate Publishing Company.

8 Corporate Social Responsibility Practices and Motivations …

175

Bolton, L. E. and A. S. Mattila (2015). How does corporate social responsibility affect consumer response to service failure in buyer–seller relationships? Journal of Retailing, 91(1): 140–153. Branco, M. and L. Rodriques (2007). Positioning stakeholder theory within the debate on corporate social responsibility. Electronic Journal of Business Ethics and Organisation Studies, 12: 5–15. Burke, L. and J. M. Logsdon (1996). How corporate social responsibility pays off. Long Range Planning, 29(4): 495–502. Cai, Y. H. Jo, and C. Pan (2012). Doing well while doing bad? CSR in controversial industry sectors. Journal of Business Ethics, 108(4): 467–480. Campbell, J. L. (2006). Institutional analysis and the paradox of corporate social responsibility. American Behavioral Scientist, 49(7): 925–938. Campbell, J. L. (2007). Why would corporations behave in socially responsible ways? An institutional theory of corporate social responsibility. Academy of Management Review, 32(3): 946–967. Campbell, K. (1999). Qualitative research methods for solving workplace problems. Technical Communication, 46: 532–545. Carroll, A. B. and K. M. Shabana (2010). The business case for corporate social responsibility: A review of concepts, research and practice. International Journal of Management Reviews, 12(1): 85–105. Cassell, C. and G. Symon (2006). Taking qualitative methods in organization and management research seriously. Qualitative Research in Organizations and Management, 1(1): 4–12. Cheng, B. I. Ioannou, and G. Serafeim (2014). Corporate social responsibility and access to finance. Strategic Management Journal, 35(1): 1–23. Corbett, A. (2008). Corporate Social Responsibility: Do we have good cause to be sceptical about it? Griffith Law Review, 17(1): 413–432. Crane, A. D. Matten, and L. J. Spence (2008). ‘Corporate social responsibility in a global context’, In A. Crane, D. Matten, and L. J. Spence (eds), Corporate social responsibility: Readings and cases in a global context. Abingdon, UK: Routledge. Creswell, J. W. (2007). Qualitative inquiry and research design: Choosing among five approaches. Thousand oaks, CA: Sage Publications. Crook, C. (2005). The good company. The Economist, 22(2005): 3–18. D’Amato, A. S. Henderson, and S. Florence (2009). Corporate social responsibility and sustainable business: A guide to leadership tasks and functions. Greensboro, NC: Center for Creative Leadership. Dahlsrud, A. (2008). How corporate social responsibility is defined: An analysis of 37 definitions. Corporate Social Responsibility and Environmental Management, 15(1): 1–13. Dawkins, J. and S. Lewis (2003). CSR in stakeholder expectations: And their implication for company strategy. Journal of Business Ethics, 44(2–3): 185–193. Dincer, C. and B. Dincer (2007). Corporate social responsibility: Future prospects in the Turkish context. Social Responsibility Journal, 3(3): 44–49. Dobers, P. and M. Halme (2009). Corporate social responsibility and developing countries. Corporate Social Responsibility and Environmental Management, 16(5): 237–249. Easterby-Smith, M. R. Thorpe, and P. R. Jackson (2008). Management research (3rd ed.). London: Sage. Eaton, C. and A. Shepherd (2001). Contract farming: Partnerships for growth. FAO Agricultural Services Bulletin No. 145. Rome: Food and Agriculture Organization. Egri, C. P. and D. A. Ralston (2008). Corporate responsibility: A review of international management research from 1998 to 2007. Journal of International Management, 14(4): 319–339. Eisenhardt, K. M. (1989). Building theories from case study research. Academy of Management Review, 14(4): 532–550. Eisenhardt, K. M. and M. E. Graebner (2007). Theory building from cases: Opportunities and challenges. Academy of Management Journal, 50(1): 25–32.

176

Y. W. Elifneh

Ellerup Nielsen, A. and C. Thomsen (2009). CSR communication in small and medium-sized enterprises: A study of the attitudes and beliefs of middle managers. Corporate Communications, 14(2): 176–189. Feldman, M. S. K. Sköldberg, R. N. Brown, and D. Horner (2004). Making sense of stories: A rhetorical approach to narrative analysis. Journal of Public Administration Research and Theory, 14(2): 147–170. Flammer, C. (2015). Does corporate social responsibility lead to superior financial performance? A regression discontinuity approach. Management Science, 61(11): 2549–2568. Flyvbjerg, B. (2006). Five misunderstandings about case-study research. Qualitative Inquiry, 12(2): 219–245. Fombrun, C. J. N. A. Gardberg, and M. L. Barnett (2000). Opportunity platforms and safety nets: Corporate citizenship and reputational risk. Business and Society Review, 105(1): 85–106. Franzosi, R. (1998). Narrative analysis-or why (and how) sociologists should be interested in narrative. Annual review of sociology, 24: 517–554. Freeman, R. E. S. R. Velamuri, and B. Moriarty (2006). Company stakeholder responsibility: A new approach to CSR. Business Roundtable Institute for Corporate Ethics. Retrieved from: http:// www.corporate-ethics.org/pdf/csr.pdf. Freeman, R. E. A. C. Wicks, and B. Parmar (2004). Stakeholder theory and “the corporate objective revisited”. Organization Science, 15(3): 364–369. Frynas, J. G. (2006). Introduction: Corporate social responsibility in emerging economies. Journal of Corporate Citizenship, 2006(24): 16–19. Garriga, E. and D. Melé (2004). Corporate social responsibility theories: Mapping the territory. Journal of Business Ethics, 53(1–2): 51–71. Georgakopoulou, A. (2006). Thinking big with small stories in narrative and identity analysis. Narrative inquiry, 16(1): 122–130. Gibbs, G. R. (2008). ‘Analysing qualitative data’, In F. Uwe (ed.), The SAGE qualitative research kit. Wiltshire, UK: Sage. Gifford, B. A. Kestler, and S. Anand (2010). Building local legitimacy into corporate social responsibility: Gold mining firms in developing nations. Journal of World Business, 45(3): 304–311. Gjølberg, M. (2011). Explaining regulatory preferences: CSR, soft law, or hard law? Insights from a survey of Nordic pioneers in CSR. Business and Politics, 13(2): 1–31. Retrieved from: http://www.degruyter.com/downloadpdf/j/bap.2011.13.2/bap.2011.13.2.1351/ bap.2011.13.2.1351.xml. Retrieved on 23 July 2015. Global Compact (2000). UN global compact. Retrieved from: http://www.unglobalcompact.org/. Graafland, J. and C. Mazereeuw-Van der Duijn Schouten (2012). Motives for corporate social responsibility. De Economist, 160(4): 377–396. Grigore, G. F. (2011). Corporate social responsibility and marketing. Developments in Corporate Governance and Responsibility, 2: 41–58. Guest, G. A. Bunce, and L. Jochhnson (2006). How many interviews are enough? An experiment with data saturation and variability. Field Methods, 18(1): 59–82. Gugler, P. and J. Y. Shi (2009). Corporate social responsibility for developing country multinational corporations: Lost war in pertaining global competitiveness? Journal of Business Ethics, 87(1): 3–24. Harjoto, M. A. and H. Jo (2011). Corporate governance and CSR nexus. Journal of Business Ethics, 100(1): 45–67. Hatch, C. D. and S. A. K. Stephen (2015). Gender effects on perceptions of individual and corporate social responsibility. The Journal of Applied Business and Economics, 17(3): 63–71. Herriott, R. E. and W. A. Firestone (1983). Multisite qualitative policy research: Optimizing description and generalizability. Educational Researcher, 12(2): 14–19. Heslin, P. A. and J. D. Ochoa (2008). Understanding and developing strategic corporate social responsibility. Organizational Dynamics, 37(2): 125–144.

8 Corporate Social Responsibility Practices and Motivations …

177

Hollender, J. (2004). What matters most: Corporate values and social responsibility. California Management Review, 46(4): 111–119. Hong, W. and W. Xiaoli (2010). ‘Corporate environmental responsibility from the perspective of systematic and dialectical science’, In G. Aras and D. Crowther (eds), A handbook of corporate governance and social responsibility. Aldershot: Gower Publishing Limited & Ashgate Publishing Company. Huang, X. B. and L. Watson (2015). Corporate social responsibility research in accounting. Journal of Accounting Literature, 34: 1–16. Husted, B. W. and D. B. Allen (2011). Corporate social strategy: Stakeholder engagement and competitive advantage. New York, NY: Cambridge University Press. Idowu, S. O. (2009). T’he United Kingdom of Great Britain and Northern Ireland’, In S. O. Idowu and L. F. Walter (eds), Global practices of corporate social responsibility. Berlin: Springer, pp. 11–35. ILO (1998). ILO declaration on fundamental principles and rights at work. Geneva: International Labour Office. Ioannou, I. and G. Serafeim (2015). The impact of corporate social responsibility on investment recommendations: Analysts’ perceptions and shifting institutional logics. Strategic Management Journal, 36(7): 1053–1081. Iwu-Egwuonwu, D. and R. Chibuike (2010). Does corporate social responsibility (CSR) impact on firm performance? A literature evidence. Retrieved from: http://ssrn.com/abstract=1659586 website: http://dx.doi.org/10.2139/ssrn.1659586. Retrieved on 18 August 2015. Jain, S. and S. Jain (2013). Corporate social responsibility: A need of present and future. OIDA International Journal of Sustainable Development, 6(07): 21–28. Jamali, D. and C. Karam (2018). Corporate social responsibility in developing countries as an emerging field of study. International Journal of Management Reviews, 20(1): 32–61. Jamali, D. and R. Mirshak (2007). Corporate social responsibility (CSR): Theory and practice in a developing country context. Journal of Business Ethics, 72(3): 243–262. Jamali, D. and B. Neville (2011). Convergence versus divergence of CSR in developing countries: An embedded multi-layered institutional lens. Journal of Business Ethics, 102(4): 599–621. Jamali, D. A. M. Safieddine, and M. Rabbath (2008). Corporate governance and corporate social responsibility synergies and interrelationships. Corporate Governance: An International Review, 16(5): 443–459. Jo, H. and M. A. Harjoto (2012). The causal effect of corporate governance on corporate social responsibility. Journal of Business Ethics, 106(1): 53–72. Jo, H. and H. Na (2012). Does CSR reduce firm risk? Evidence from controversial industry sectors. Journal of Business Ethics, 110(4): 441–456. Joireman, J. D. Smith, R. L. Liu, and J. Arthurs (2015). It’s all good: Corporate social responsibility reduces negative and promotes positive responses to service failures among value-aligned customers. Journal of Public Policy & Marketing, 34(1): 32–49. Kaarbo, J. and R. K. Beasley (1999). A practical guide to the comparative case study method in political psychology. Political Psychology, 20(2): 369–391. Kotler, P. and N. Lee (2005). Corporate social responsibility: Doing the most good for your company and your cause. Hoboken, NJ: John Wiley & Sons, Inc. Kurucz, E. B. Colbert, and D. Wheeler (2008). ‘The business case for corporate social responsibility’, In A. Crane, A. McWilliams, D. Matten, J. Moon, and D. Seigel (eds), The Oxford handbook of corporate social responsibility. Oxford: Oxford University Press. Lambooy, T. E. (2014). Legal aspects of corporate social responsibility. Utrecht Journal of International and European Law, 30(78): 1–6. Lund-Thomsen, P. A. Lindgreen, and J. Vanhamme (2016). Industrial and corporate social responsibility in developing countries: What we know, what we do not know, and what we need to know. Journal of Business Ethics, 133(1): 9–24. Luo, X. and C. B. Bhattacharya (2006). Corporate social responsibility, customer satisfaction, and market value. Journal of Marketing, 70(4): 1–18.

178

Y. W. Elifneh

McBarnet, D. (2007). ‘Corporate social responsibility beyond law, through law, for law: The new corporate accountability’, In D. McBarnet, A. Voiculescu, and T. Campbell (eds), The new corporate accountability: Corporate social responsibility and the law. Cambridge: Cambridge University Press, pp. 9–56. Melo, T. and A. Garrido-Morgado (2012). Corporate reputation: A combination of social responsibility and industry. Corporate Social Responsibility and Environmental Management, 19(1): 11–31. Millon, D. (2010). Enlightened shareholder value, social responsibility, and the redefinition of corporate purpose without law. (Washington & Lee Legal Studies Paper No. 2010–11), 1–46. Retrieved from: http://dx.doi.org/10.2139/ssrn.1625750. Retrieved on 20 June 2015. Moir, L. (2001). What do we mean by corporate social responsibility? Corporate Governance, 1(2): 16–22. Ntim, C. G. and T. Soobaroyen (2013). Corporate governance and performance in socially responsible corporations: New empirical insights from a neo-institutional framework. Corporate Governance: An International Review, 21(5): 468–494. Peattie, L. (2001). Theorizing planning: Some comments on Flyvbjerg’s rationality and power. International Planning Studies, 6(3): 257–262. Pondeville, S. V. Swaen and Y. De Rongé (2013). Environmental management control systems: The role of contextual and strategic factors. Management Accounting Research, 24(4): 317–332. Porter, M. and M. Kramer (2006). Strategy and society, the link between competitive advantage and corporate social responsibility. Harvard Business Review, 85(12): 78–92. Proença, J. and M. Branco (2014). Corporate social responsibility practices and motivations in a peripheral country: Two Portuguese illustrative cases. Corporate Governance, 14(2): 252–264. Reast, J. F. Maon, A. Lindgreen, and J. Vanhamme (2013). Legitimacy-seeking organizational strategies in controversial industries: A case study analysis and a bidimensional model. Journal of Business Ethics, 118(1): 139–153. Ribstein, L. E. (2005). Accountability and responsibility in corporate governance. Notre Dame Law Review, 81(4): 1431–1494. Riessman, C. K. (2005). ‘Narrative Analysis’, In M. S. Lewis-Beck, A. Bryman, and T. F. Liao (eds), The SAGE Encyclopedia of Social Science Research Methods. Retrieved from: http://dx. doi.org/10.4135/9781412950589.n611. Rigoberto Parada Daza, J. (2009). A valuation model for corporate social responsibility. Social Responsibility Journal, 5(3): 284–299. Robins, F. (2008). Why corporate social responsibility should be popularised but not imposed. Corporate Governance, 8(3): 330–341. Ruggie, J. (2008). Protect, respect and remedy: A framework for business and human rights. Innovations, 3(2): 189–212. Russo, A. and A. Tencati (2009). Formal vs. informal CSR strategies: Evidence from Italian micro, small, medium-sized, and large firms. Journal of Business Ethics, 85(2): 339–353. Sacconi, L. (2007). A social contract account for CSR as an extended model of corporate governance (II): Compliance, reputation and reciprocity. Journal of Business Ethics, 75(1): 77–96. Savenye, W. C. and R. S. Robinson (1996). Qualitative research issues and methods: An introduction for educational technologists. Handbook of Research for Educational Communications and Technology, 1171–1195. Sen, S. and C. B. Bhattacharya (2001). Does doing good always lead to doing better? Consumer reactions to corporate social responsibility. Journal of Marketing Research, 38(2), 225–243. Serafeim, G. (2013). The role of the corporation in society: An alternative view and opportunities for future research. Retrieved from: SSRN website: http://dx.doi.org/10.2139/ssrn.2270579. Shahin, A. and M. Zairi (2007). Corporate governance as a critical element for driving excellence in corporate social responsibility. International Journal of Quality & Reliability Management, 24(7): 753–770.

8 Corporate Social Responsibility Practices and Motivations …

179

Smith, C. P. (2000). ‘Content analysis and narrative analysis’, In T. Reis and C. Judd (eds), Handbook of research methods in social and personality psychology. Cambridge: Cambridge University Press, pp. 313–335. Spitzeck, H. (2009). The development of governance structures for corporate responsibility. Corporate Governance, 9(4): 495–505. Sprinkle, G. B. and L. A. Maines (2010). The benefits and costs of corporate social responsibility. Business Horizons, 53(5): 445–453. Strauss, A. and J. Corbin (1990). Basics of qualitative research: Techniques and procedures for developing grounded theory (2nd ed.). Newbury Park, CA: Sage. Sun, W. J. Stewart, and D. Pollard (2010). Reframing corporate social responsibility: Lessons from the global financial crisis. Bingley, England: Emerald. Thomsen, S. (2004). Corporate values and corporate governance. Corporate Governance, 4(4): 29–46. Tsang, S. R. Welford, and M. Brown (2009). Reporting on community investment. Corporate Social Responsibility and Environmental Management, 16(3): 123–136. Turner, D. W. (2010). Qualitative interview design: A practical guide for novice investigators. The Qualitative Report, 15(3): 754–760. Utting, P. (2003). Promoting development through corporate social responsibility—Does it work? Global Future, Third Quarter: 11–13. Van de Ven, B. and R. Jeurissen (2005). Competing responsibly. Business Ethics Quarterly: 299–317. Visser, W. (2008). ‘Corporate social responsibility in developing countries’, In A. Crane, A. McWilliams, D. Matten, J. Moon, and D. Siegel (eds), The Oxford Handbook of Corporate Social Responsibility. Oxford: Oxford University Press, pp. 473–479. Vogel, D. (2006). The market for virtue: The potential and limits of corporate social responsibility. Washington, DC: Brookings Institution Press. Vuta, M. I. Verboncu, D. Ioana, S. Aurelia, R. Gherghina, M. Dudian, … M. H. Dobre (2007). The corporate social responsibility practices in Romania in the context of sustainable development principles. Paper presented at the 47th Congress of European Regional Science Association, ERSA Congress, Paris, August 29–September 2007. Retrieved from: http://dx.doi.org/10.2139/ ssrn.1347358. Ward, H. (2004). Public sector roles in strengthening corporate social responsibility: Taking stock. Washington, DC: The World Bank. Yin, R. K. (2003a). Case studies research: Design and methods (2nd ed.). Thousand Oaks, CA: Sage. Yin, R. K. (ed.). (2003b). Case study research design and methods. Applied social research methods series (3rd ed., Vol. 5). Thousand Oaks, CA: Sage. Zadek, S. (2000). Doing good and doing well: Making the business case for corporate citizenship. (Research Report 1282-00-RR). New York: The Conference Board. Zheng, Q. Y. Luo, and V. Maksimov (2015). Achieving legitimacy through corporate social responsibility: The case of emerging economy firms. Journal of World Business, 50(3): 389–403.

Chapter 9

Psychological Determinants of Employees’ Intentions to Retire: A Case of Public Universities in Kenya Lucy Jepchoge Rono and Ester Agasha

Abstract Literature on individuals’ intentions to retire has mainly focused on financial factors, personal factors, and employees’ entrepreneurial intensity. Very little work has been done to explain the role of psychological factors like self-concept, work attachment, established work relationships, and older worker stereotypes on one’s intention to retire. This study focuses on establishing a better understanding of the psychological factors that influence an employee’s intentions to retire in Kenya. The study is grounded in the image theory which posits that retirement is typified as a new beginning; a full period of life where commitment to work is removed and an employee is able to prioritize non-working activities. Applying an explanatory research design, this study uses stratified sampling. The sample size is computed using Cochran’s formula. The sample is 384 employees with a response rate of 87.1%. Data was collected through questionnaires following a drop and pick approach after at least five days from the respondents. The study analyzes the data using descriptive statistics, the Pearson Moment correlation, a factor analysis using the extraction method, and a principal component analysis. The findings show that psychological factors affect an employee’s intentions to retire. The results also show that work attachment is a significant component of the psychological factors that influence intentions to retire. The study provides a way forward, implications of the various policies, and undertaking reforms in retirement decisions in Kenya and other developing nations. Keywords Psychological factors · Intention to retire · Employees · Kenya

L. J. Rono (B) Department of Accounting and Finance, Moi University, Eldoret, Kenya e-mail: [email protected] E. Agasha Department of Finance, Makerere University Business School, Kampala, Uganda e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 J. Wood and O. Habimana (eds.), A Multidimensional Economic Assessment of Africa, Frontiers in African Business Research, https://doi.org/10.1007/978-981-15-4510-8_9

181

182

L. J. Rono and E. Agasha

9.1 Introduction Numerous studies have shown the importance of predictors of retirement. Most of these studies were carried out on civil servants and retirees in developed countries (U.S., Europe, and Australia) (Kakwani and Hinz 2006). Despite growth in retirement systems in those countries, few studies have been done to study workers’ intentions to retire. Most of the studies on retirement decisions give details of the determinants of intentions to continue working. For example, Shacklock and Brunetto (2005), studied the Australian workforce from the perspective of intention to continue working. It does not discuss the factors that influence the decision to retire. Other studies on retirement focus on how to respond to employees’ retirement scenarios. For instance, empirical work has been done analyzing the experiences of managing older workers, managers’ involvement in employees’ retirement decisions, and managers’ attitudes towards retirement; Tversky and Kahneman 1974). Though these studies emphasize the importance of pre-retirement planning and training for employees, they do not give details of the factors that influence their intentions to retire. Scholarly work on factors that influence retirement concentrates on financial factors, personal factors, and entrepreneurial intensity of employees (Bringley and Pedersen 2011; Hershey and Henkens 2014; Higgs et al. 2003; Van Solinge and Henkens 2014). Higgs et al. (2003), studied British civil servants who chose early retirement and concluded that many participants in their study chose early retirement for monetary gains. However, their results do not show any psychological influence on the choice to retire in the British civil service. In a related study, Asch et al. (2005), show that employees’ retirement decisions are motivated by increased pensions though they make very little mention of the motivation to retire based on psychological factors. This chapter conceptualizes employees’ intentions to retire based on psychological factors including self-concept, work attachment, established work relationships, and older worker stereotypes. Today, measuring the psychological aspects of intentions to retire have become more necessary because the context within which employees retire has changed significantly. For example, demographic projections show that by 2022, nearly 30% of the total U.S. workforce will be 55 years or older, up from just less than 13% in 2000. It is estimated that by 2050, 50% of the population in Europe will be aged 50 years and more. In Kenya, more than 3 million people are expected to be older than 55 years by 2025. This means a sizable increase in the number of people who will transition into retirement during the next few decades (Cremer et al. 2008; Kinsella 1992; Toossi 2004; Wang and Shultz 2009). The question that has not been addressed by previous studies is the impact of psychological factors on employees’ intentions to retire. This study develops insights into the relationship between psychological factors—self-concept, work attachment, established work relationships, older worker stereotypes—and employees’ intentions to retire. Existing studies on retirement intentions have used panel data in the developed economies mainly in the United States, Australia, and Europe. This study gathered empirical information in

9 Psychological Determinants of Employees’ Intentions to Retire …

183

a developing country. Hence, it contributes to our knowledge about the theory and practice of managing retirement intentions. This study’s main focus is on the relationship between the psychological determinants of self-concept, work attachment, established work relationships, older worker stereotypes, and employees’ intentions to retire in public universities in Kenya. It has four specific objectives: Examining self-concept as a predictor of employees’ intentions to retire. Assessing work attachment as a predictor of employees’ intentions to retire. Examining established work relationships predicting employees’ intentions to retire. Examining older worker stereotypes in predicting employees’ intentions to retire. The rest of this chapter is structured as follows. The next section provides a theoretical framework on which the background of this chapter is premised. Section 3 discusses past studies on retirement decisions and the gaps therein. Section 4 deliberates the methods employed and the sampling design and the data collection process. Section 5 details the results from the study and provides it’s relatedness to earlier research findings. A summary and conclusion draws from the previous section and elaborates on the key findings and their implications to practice. The final section of this chapter provides recommendations based on the result.

9.2 Theoretical Underpinnings This study was guided by two theories—the image theory and the role theory. According to the image theory, for employees near retirement age, retirement is typified as a ‘new beginning’ meaning a full period of life where there is no commitment to work and an employee is able to prioritize non-working activities (Adams et al. 2002; Talaga and Beehr 1995). This new beginning may also be characterized by an employee taking on a number of roles such as a parent, child, and a member of a community group; retirement also allows him to pursue existing hobbies or new leisure activities. Therefore, for some employees, retirement can be an opportunity of being free of work obligations to undertake leisure activities. However, for others it is a time of anxiety when they have to figure out how to spend the day, anticipate reduced social networks, and have no work structure. Role theory suggests that work roles play an important role in self-identity. Retirement can lead to the loss of this role which may in turn lead to anxiety, stress, depression, and poor adjustment and adaptation (Wong and Earl 2009). In a study on British civil servants, Mein et al. (2000), show that the role loss due to retirement leads to older workers opting to work longer years rather instead of retiring.

184

L. J. Rono and E. Agasha

9.3 Literature Review Traditionally, retirement has been viewed as withdrawal from the work environment in an organization (Feldman 1994). What an employee expects after retirement is important in predicting his decision to retire (Davies and Cartwright 2010). Retirement decisions for individuals vary considerably as their reactions to a situation depend on their values and beliefs. Shimizutani (2011), identified four attitudes about retirement: a transition to the rest of life; continuity; a new beginning; and an imposed disruption. Existing studies on retirement decisions focus on actual reasons for employees retiring including health, financial, entrepreneurial traits, and organizational policies and practices (Mein et al. 2000; Shacklock and Brunetto 2005; Szinovacz and Davey 2005a, b). The orientation of these studies is on early retirement and involuntary retirement as dictated by the environment (Szinovacz and Davey 2005a, b). Some authors postulate that there are three pathways to retirement: early retirement, normal retirement, and late retirement (Hillier and Hodgson 2011; Zaniboni et al. 2010). Most authors who have studied retirement intentions, however, have a general consensus about retirement as a departure from an organizational career including job changes during the working life of an individual (Duflo and Saez 2003). In this chapter, retirement intentions mean leaving the workforce before the retirement date of 60 years in Kenya. Literature on economics, sociology, gerontology, medicine, human resources, organizational behavior, and psychology focuses on predictors of retirement intentions. Most of these studies discuss health and financial factors as predictors of retirement decisions. For example, Higgs et al. (2003) studied British civil servants who chose early retirement and concluded that many participants chose early retirement for monetary gains. Similarly, Mein et al. (2000), found that an individual’s retirement decision depended on the organization’s pension policy, state pension or both (Higgs et al. 2003; Wang and Shultz 2009). In their study on federal civil service workers in the United States, Asch et al. (2005), concluded that increased annual earnings were an incentive for employees’ retirement decisions. This review shows that no scholar has attempted to investigate how an individual’s retirement intentions are influenced by psychological factors (Beehr and Bennett 2007). Hence, this study bridges this gap in existing knowledge. According to psychologists self-concept plays a significant role in employees’ decisions to retire (Taylor and Shore 1995). Research by Van Solinge and Henkens (2007) on 778 Dutch older workers in retirement transition showed that higher scores in work efficiency was associated with greater ease in adjusting to retirement. Taylor and Shore’s (1995), study on 264 older workers showed that low self-concept may lead to avoidance of retirement. Studies on older workers suggest that older workers seek to continue the same thoughts, patterns, and lifestyles after retirement. This suggests that employees who have pursued leisure activities during their work life expect to continue with those activities after retirement. Additionally, adopting new leisure activities, having more social time, and travelling with family and friends may also happen after retirement (Beehr et al. 2000a, b).

9 Psychological Determinants of Employees’ Intentions to Retire …

185

In addition, studies also suggest that attachment to work too can determine an employee’s decision to retire. Work attachment has been widely studied and has different dimensions including the nature of the work such as managerial, clerical, mechanical, and repetitive work like sewing and packing (Mowday et al. 2004; Shacklock and Brunetto, 2005; Smith et al. 1969). The more strenuous the work the more likely that an employee will not enjoy nor derive satisfaction nor perform exceptionally well and therefore the higher the probability of his desire to retire. The work environment is another lens of work attachment. Literature suggests that working under poor conditions in an organization stimulates decisions to retire (Henkens and Tazelaar 1997). Work conditions which lead to job redundancies and reorganizations may lead to employees’ preferring to retire. In an organization with little or no promotional opportunities, training, and mentoring, an employee is likely to decide to retire (Feldman 1994). The more restrictive the employee’s work environment the stronger the likelihood of his thinking about retirement. Literature also shows that employees’ job satisfaction and work attachment include organizational commitment (Adams et al. 2002). Studies on organizational commitment in relation to retirement have yielded mixed results. For example, Ekerdt and De Viney (1993), studied 1,365 non-retired male workers and found that the closer to retirement the older workers came the less favorable their commitments were to the organizations they were working for. On the contrary, in a study of 197 older workers, Beehr et al. (2000a, b), concluded that organizational commitments contributed to 10% of the workers’ decisions to retire. Retirement intentions also include older worker stereotypes like perceiving retirement as end of work, one’s skills as being obsolete, and perceiving one’s age as not fit for working any longer (Loi and Shultz 2007). A study of 101 older workers in California showed that older workers sought employment when they had a positive attitude towards work and sought to be competitive and up to date. Other workers attitude towards retirement in the organization also influences an employee’s intentions to retire (Conway et al. 2014; Duflo and Saez 2003). This implies that older employees with positive retirement attitudes are likely to think of retiring.

9.4 Methodology This study uses a quantitative research design which is explanatory in nature. It gathered information on perceptions and captured structured data through questionnaires with Likert type questions (Creswell 2009). The focus of this study is better understanding the psychological factors that influence employees’ intentions to retire. The study was carried out in the context of public universities in Kenya. The rationale behind this study this is that most of the research on retirement intentions has been done in developed countries, mainly in USA, Europe and most recently in Australia and Japan. The public university set up in Kenya was chosen to offer knowledge in tandem with what answers past studies have for research questions related to retirement issues (Kinsella 1992; Mowday et al. 2004; Shacklock and Brunetto 2005). At

186 Table 9.1 Public Universities in Kenya which formed part of the study

L. J. Rono and E. Agasha Name of employer

Total staff

Administrative staff

Moi University

3148

2173

Kenyatta University

2645

1851

University of Nairobi

2517

1813

Egerton University

2317

1738

JKUAT

2219

1643

Maseno university

1923

1443

MMUST

1713

1251

Total

16482

19487

Source Survey Data (2013)

the time when this study was done there were seven public universities in Kenya: Kenyatta University, Maseno University, Masinde Muliro University of Technology, Jomo Kenyatta University of Agriculture and Technology, Moi University, Egerton University, and the University of Nairobi. In Kenya, public universities were set up between 1970 and 1994. The universities are spread in different counties including three in Nairobi county and one each in Uasingishu, Nakuru, and Kakamega counties. The research focused on administrative staff in public universities in Kenya (Table 9.1).

9.4.1 Sampling Design and Determination of Sample Size The sample size was scientifically determined using Cochran’s sample size formula (Cochran 1977) and was computed as: n=

Z 2 P(1 − P) C2

where Z is 1.96 for the 95% confidence level c = 5% is the margin of error P = 0.50 the percentage that was used in the study 2 Thus n = 1.96 ×0.5(1−0.5) = 384.16. 0.052 To get 384 respondents from the seven public universities, the study apportioned the ratios by the number of administrative staff per university. The result was at least 60 respondents from each of the seven universities. The confidence limit chosen for this research was 95% (α = 0 .05). This choice is in line with past studies on retirement intentions in organizational behavior research (Barlett et al. 2001a, b; Spector 2006). The margin of error chosen was 3% which means the true mean of the 7-point Likert scale is within ±0.21 (Barlett et al. 2001a, b). Stratified random sampling was used in administering the survey (Miller and Salkind 2002). The strata consist of employees

9 Psychological Determinants of Employees’ Intentions to Retire …

187

of each university assigned to their administration units. At the department level, the respondents were randomly selected to answer the questionnaire. The process was applied in the same manner in the seven public universities at the time of the survey. Stratified random sampling was chosen because it gives an equal chance of participation to respondents and reduces variability. The respondents were administrative employees in Kenyan public universities. The researcher reported to the research offices in all the seven public universities to seek permission to carry out the research. The research offices allowed the researcher to administer the questionnaires. The researcher was directed to the main administration block of the university and shown where the department heads of various sections were located. The heads of department in turn gave permission to randomly distribute the questionnaires. In each university, at least 100 questionnaires were administered. The researcher allowed five days to the respondents to complete the questionnaire and went back to collect the completed questionnaires. The process was repeated in all the seven public universities in Kenya. The respondents were asked to tick an appropriate response on a 7-piont Likert scale with [1] = Strongly Disagree [2] = Moderately Disagree [3] = Slightly Disagree [4] = Neither Disagree nor Agree [5] = Slightly Agree [6] = Moderately Agree and [7] = Strongly Agree. Psychological factors were measured by asking employees their attitude towards work and work commitment. This measure has been used by Mowday et al. (2004), and later by Shacklock and Brunetto (2005). The retirement stereotype was measured by five questions asking employees whether aged employees’ skills were still valuable.

9.5 Results 9.5.1 Descriptive The study examined the psychological factors encompassing an individual’s selfconcept, established relationships at the work place, older worker stereotypes, and work attachment. For all the three statements regarding self-concept, a mean of 5.2, 5.3 and 5.7 respectively was achieved. This result implies that the respondents tended to strongly agree with the statement and hence probably had a positive self-concept and may believe in the roles that they played at work as well as the decisions they took for the tasks they were assigned. This result also implies that the employees were most likely to prolong their stay at their work place in an environment where they could not only create relationships but also build favorable relationships. A descriptive analysis of older worker stereotyping revealed that 78.3% of the respondents believed that retirement did not necessarily mean the end of working. This result shows that the respondents believed that retirement was a new beginning (Table 9.2). We also did a descriptive analysis of work attachment. The results showed that most of the respondents were attached to their work and that they would do whatever

188

L. J. Rono and E. Agasha

Table 9.2 Computed mean and standard deviations for measures of psychological factors Item

Attitudinal statement

Mean

Std. deviation

Min

Max

Self-concept

I have always done the kind of work that I do currently

4.8

2.14

1

7

I have had a change in my job tasks in the recent past

4.1

2.40

1

7

I care about the job tasks I perform

5.7

1.68

1

7

I take decisions about my job tasks

5.3

1.65

1

7

I am confident of taking decisions about my job tasks

5.7

1.52

1

7

My supervisor is confident about my decisions

5.5

1.75

1

7

My co-workers would like me to remain in the organization

5.3

1.83

1

7

My immediate supervisor would like me to continue to work after attaining the mandatory age of retirement

4.9

1.84

1

7

My head of department would support me if I chose to continue working until the mandatory age of retirement

5.2

1.81

1

7

Employees’ skills become outdated

3.7

2.17

1

7

Employees who are aged 50 years and more should retire

3.9

2.22

1

7

An old employee is expected to take his own decision to retire

4.2

3.27

1

7

Older employees are a marginal group

3.4

2.14

1

7

Retirement means end of work

2.8

2.16

1

7

Retirement means a new beginning

4.3

2.40

1

7

Established relationship

Older worker stereotype

Source Survey Data (2013)

9 Psychological Determinants of Employees’ Intentions to Retire …

189

it took to make their institutions succeed. For all the four positive statements with respect to work attachment the study generated means between 4.2 and 5.8 while all the four negative statements on work attachment generated means between 3.4 and 3.9. These results imply that the respondents had a strong attachment to their work (Table 9.3).

9.5.2 Factor Analysis The psychological factors consisted of four constructs: self-concept, work attachment, established work relationships, and older worker stereotypes. Self-concept had six items, work attachment 14 items, established work relationships three items, and older worker stereotypes six items. All the 29 items for psychological factors were tested for reliability using Cronchbach’s alpha. The findings posted a Cronchbach’s alpha value equal to 0.565, an indication that the items were relatively reliable in influencing intentions to retire. A factor analysis was done of all the 29 items using the extraction method of the principal component analysis. The study found nine items with a cumulative percentage of 59.246. The highest percentage of variance was 14.606. This result implies that the nine extracted components together contributed 59.246% of the total influence on intentions to retire. An inspection of the rotated component matrix showed that decision making items such as ‘I take decisions about my job tasks,’ ‘I am confident in taking decisions about my job tasks,’ and ‘my supervisor is confident about my decisions on the job’ had relatively high scores between 0.689 and 0.809. The findings also throw light on the meaning of retirement that is whether it is ‘end of work’ or ‘a new beginning’ with 0.584 and 0.621 values respectively. This finding contradicts the findings of an Australian study carried out to establish the perceived meaning of retirement (Shacklock and Brunetto 2005). My study also found that the extracted components had Eigen values greater than 1. This implies that the extracted components had a relatively strong contribution on psychological factors affecting intentions to retire. A Kaiser-Meyer-Olkin measure of sampling adequacy was equal to 0.730, the Bartlett’s test of sphericity posted an approximate chi-square of 2063.258 with 406 degrees of freedom. These results were statistically significant at p < 0 .05% (Tables 9.4 and 9.5).

9.5.3 Correlation Results A bivariate correlation analysis was done between the dependent and independent variables. The results show that of the four psychological factors comprising of self-concept, work attachment, older worker stereotype, and established work relationships, only established work relationships showed a negative relationship (r = 0.201, p < 0.01) with employees’ intentions to retire.

190

L. J. Rono and E. Agasha

Table 9.3 Computed mean and standard deviations for measures of work attachment Item

Attitudinal statement

Mean

Std. deviation

Min

Max

Work attachment

I am willing to put a great deal of effort beyond what is normally expected to help this organization be successful

5.8

1.64

1

7

I talk about this organization with my friends as a great organization to work for

5.8

1.65

1

7

I feel very little loyalty to this organization

3.4

2.30

1

7

I would accept almost any type of job assignment keep working for this organization

4.2

2.24

1

7

I find my values and the organization’s values to be very similar

4.7

2.11

1

7

I am proud to tell the others that I am proud of this organization

5.3

2.09

1

7

I could just as well be working for a different organization as long as the type of work is similar

4.2

2.21

1

7

it would take very little change in my present circumstances to make me leave this organization

3.8

2.07

1

7

I am extremely glad that I chose this organization to work for over the others that I was considering

5.3

1.88

1

7

There is not too much to be gained by sticking to this organization indefinitely

3.3

2.16

1

7

(continued)

9 Psychological Determinants of Employees’ Intentions to Retire …

191

Table 9.3 (continued) Item

Attitudinal statement

Mean

Std. deviation

Min

Max

Often I find it difficult to agree with this organization’s policies on important matters relating to its employees

4.0

2.20

1

7

I really care about the fate of this organization

5.6

1.84

1

7

For me this is the best possible organization to work for

5.3

1.94

1

7

Deciding to work for this organization was a definite mistake on my part

3.1

2.42

1

7

Source Survey Data (2013)

9.6 Summary and Conclusions 9.6.1 Psychological Factors as Predictors of Intentions to Retire The objective of this study was assessing the psychological factors predicting employees’ retirement intentions. Four sub-constructs for psychological factors were used in the study—self-concept, work attachment, established work relationships, and older worker stereotypes. The findings show that all these sub-constructs are important components of psychological factors affecting an employee’s intention to retire. Self-concept is a vital construct for psychological factors predicting retirement intentions and has been used in existing studies on retirement intentions. The findings of my study are consistent with the image theory which holds that actions that fit with the employers’ goals are important for maintaining a positive and consistent self-identity (Brougham and Walsh 2007). The findings also show that work attachment is an important component of the psychological factors that affect an employee’s intention to retire. This finding is similar to previous research that work attachment has a positive relationship with intentions to retire (Adams et al. 2002; Henkens and Tazelaar 1997). This is explained by the fact that work attachment includes various facets of work such as organizational commitment, nature of work, the work environment, training, promotion, and mentoring. Given the diversity of employees in this study it is likely that it captured all these aspects. Alternatively, it might be possible that the organizations from which all the respondents were drawn have goals that are similar to those of their employees and therefore the employees are highly attached to their organizations.

2.366

2.220

1.987

1.584

1.398

1.269

1.095

1.025

0.997

3

4

5

6

7

8

9

10

3.438

3.536

3.775

4.377

4.821

5.463

6.853

7.657

8.158

14.606

62.684

59.246

55.710

51.935

47.557

42.737

37.273

30.421

22.764

14.606

1.025

1.095

1.269

1.398

1.584

1.987

2.220

2.366

4.236

3.536

3.775

4.377

4.821

5.463

6.853

7.657

8.158

14.606

Extraction method: principal component analysis Kaiser-Meyer-Olkin measure of sampling adequacy = 0.730 Bartlett’s test of sphericity: approx. chi-square = 2063.258 Df = 406 Sig. = 0.000 Source Survey Data (2013)

4.236

2

59.246

55.710

51.935

47.557

42.737

37.273

30.421

22.764

14.606

Cumulative percent

Total

Percent of variance

Extraction sums of squared loadings

Cumulative percent

Total

Percent of variance

Initial eigenvalues

1

Component

Table 9.4 Psychological factors’ total variance explained

1.305

1.595

1.644

1.743

1.893

1.942

2.200

2.349

2.511

Total

4.500

5.499

5.670

6.009

6.526

6.698

7.585

8.101

8.658

Percent of variance

59.246

54.746

49.247

43.577

37.567

31.041

24.344

16.758

8.658

Cumulative percent

Rotation sums of squared loadings

192 L. J. Rono and E. Agasha

2

My supervisor is confident about my decisions on the job

6

My co-workers would like me to remain in the organization

Deciding to work for this organization was a definite mistake on my part

For me this is the best possible organization to work for

I really care about the fate of this organization

Often I find it difficult to agree with this organization’s policies on important matters relating to its employees

0.756

0.689

0.524

It would take very little change in my present circumstances to make me leave this organization

I am extremely glad that I chose this organization to work for over others

0.736

I could just as well be working for a different organization as long as the type of work is similar

0.522

0.585

5

I find my values and the organization’s values to be very similar

0.754

4

0.786

0.690

0.763

3

I would accept almost any type of job assignment to keep working for this organization

I feel very little loyalty to this organization

0.654

0.689

I am confident of taking decisions about my job tasks

I talk about this organization with my friends as a great organization to work for

0.809 0.780

I take decisions about my job tasks

0.516

0.652

1

I care about the job tasks that I perform

I have always done the kind of work I do currently

Component

Table 9.5 Psychological factors’ component matrix 7

8

(continued)

0.465

9

9 Psychological Determinants of Employees’ Intentions to Retire … 193

8

Extraction method: Principal component analysis Rotation method: Varimax with Kaiser normalization Rotation converged in 9 iterations Source Survey Data (2013)

Cronchbach’s Alpha

Retirement means a new beginning

Retirement means end of work

0.791 0.621

7

Employees who are aged 50 years and more should retire 0.584

6

0.765

5

Employee skills have become outdated

0.565

4

0.657

3

My head of department would support me if I chose to continue working until the mandatory age of retirement

2 0.758

1

My immediate supervisor would like me to continue to work after attaining the mandatory age of retirement

Component

Table 9.5 (continued) 9

194 L. J. Rono and E. Agasha

9 Psychological Determinants of Employees’ Intentions to Retire …

195

The psychological factors in this study include established work relationships. The findings show that established work relationships are an important aspect that affects an employee’s intention to retire. This result is probably linked to the fact that when an employee establishes strong ties with his work colleagues they become socially bound and they possibly receive support from each other while taking various decisions including decisions about retirement (Henkens and Tazelaar 1997). This finding may also imply that an established work relationship as a predictor has an effect on the intentions to retire. This finding is similar to other studies’ findings. Existing research shows that retirement anxiety is a consequence of no structured work life and detachment from social relationships after retirement. A research carried out to review typologies showed that the older workforce had a wide range of workers whose work experience had an impact on their intentions to retire (Flynn 2010). Older worker stereotypes were found to be a vital component of psychological factors predicting an employee’s intention to retire. This finding is similar to existing research which shows that older worker stereotypes arouse a stronger desire to retire early (Desmette and Gillard 2008). This is probably explained by the older worker stereotype’s attitude towards retirement. In the event that a negative connotation is contained in the stereotype such as retirement is the end of work, an employee is unlikely to want to retire. On the other hand, if there is a positive stereotype like retirement is the beginning of a new life, there will be a positive attitude towards intentions to retire. Literature shows that the main challenge for an employee intending to retire is creating new relationships. Research further shows that social relationships at work may positively demotivate an employee to retire (Fletcher et al. 1992; Henkens and Tazelaar 1997).

9.6.2 Implications of the Study for Theory, Practice, and Policy This study contributes to knowledge on psychological factors determining an employee’s retirement intentions. This perspective gives future researchers a theoretical basis of undertaking further research in different contexts. Scholars can use the results of this research to study retirees as a follow up of this study. Confirmatory tests of the outcomes of this study will also emerge if a similar study is initiated in other economies. Going by the results of this study, employees who intend to retire will be better equipped to take responsible retirement decisions. Employees may appreciate the relationship between work and retirement. An employee’s level of attachment to work and its influence on his decision to retire may become clearer through the results of this study. Employees can consciously accept or reject retirement stereotypes that influence their decisions to retire. This may lead to the development or modification of an organization’s culture.

196

L. J. Rono and E. Agasha

Finally, the results can be used by policymakers and incorporated in human resource training for employees before retirement. The results of this study should also be of interest to organizations like the Retirement Benefits Authority of Kenya and the Government of Kenya in administering retirement education. This study’s results provide a way forward for changing policies so as to undertake reforms in retirement intentions and decisions in Kenya and other developing countries. This study recommends further research in work environment, organizational policies, retirement systems, organizational culture, the economy of a nation, legal structures, and the retirement process. This study can also be replicated in a different set-up such as for civil servants, various counties in Kenya, and small and medium enterprises.

References Adams, G. S. A. J. Prescher, T. A. Beehr, and L. Lepisto (2002). Applying work-role attachment theory to retirement, decision-making. International Journal of Aging and Human Development, 54: 125–137. Asch, B. S. Haider, and J. Zissimopoulos (2005). Financial incentives and retirement: Evidence from federal civil service workers. Journal of Public Economics, 89(2–3):427–440. Barlett, J. E. J. W. Kotrlik, C. Higgins, and C. Chadwick (2001a). Organizational Research; determining the appropriate sample size in survey research. Information Technology, Learning and Performance Journal, 19(1): 43–50. Barlett, J. E. Kotrlik, J. W. Higgins, C. (2001b). Organizational Research; determining the appropriate sample size in survey research. Information Technology, Learning and Performance Journal, 19(1); 43–50. Beehr, T. A. and M. M. Bennett (2007). ‘Examining Retirement from a multi-level perspective’, In K.S. Shultz and A. Adams (eds). Aging and Work in the 21st century. New York: Psychology Press, pp. 277–302. Beehr, T. A. Jex, S. M. Stacy, B. A. and Murray, M. A. (2000a). Work stressors and coworker support as predictors of individual strain and job performance. Journal of Organizational Behavior. Vol. 21, issue 4, 391 405. https://doi.org/10.1002/%28SICI%291099-1379%28200006%2921:4%3C391: :AID-JOB15%3E3.0.CO;2-9. Beehr, T. S. Glazer, N. Nelson, and S. Farmer (2000b). Work and non-work predictors of employees retirement ages. Journal of Vocational Behavior, 57(2): 206–225. Bringley, P. and P. Pedersen (2011). Pension reform, early retirement, health and Mortality; A30 year follow-up study. The Danish National Center for Social Research: 2–26. Brougham, R. and D. A. Walsh (2007). The International Journal of Aging and Human. Cochran, W. G. (ed.). (1977). Sampling Techniques, (3rd ed.). New York: Wiley. Conway, N. Kiefer, T. Hartley, J. and Briner, R.B. (2014). Doing more with less? Employee reactions to psychological contract breach via target similarity or spillover during public sector organizational change. British Journal of Management. Vol. 25, issue 4, 737–754. Cremer, H. J. Lozachmeur, and P. Pestieau (2008). Social security and retirement decision: positive and normative approach. Journal of Economic Surveys,22(2): 213–233. Creswell, J. W. (ed.). (2009). Research Design: Qualitative, quantitative and mixed methods approaches (3rd ed.). Thousand Oaks, California: Sage publications. Davies, E. and S. Cartwright (2010). Psychological and psychosocial predictors of attitudes to working past normal retirement age. Employee Relations, 32(3): 249–268.

9 Psychological Determinants of Employees’ Intentions to Retire …

197

Desmette, D. and M. Gaillard (2008). When a ‘worker’ becomes an older-worker: The effects of age related social identity on attitude towards retirement and work. Career Development International, 13 (2): 168–185. Development, 65 (3): 203–229. Title of article needed. Duflo, E. and E. Saez (2003). The role of information and social interactions in retirement plan decisions. Evidence from a randomized experiment. National Bureau of Economic Research paper number 8885, Massachusetts. Ekerdt, D. J. and S. De Viney (1993). Evidence for pre-retirement programs and firm performance. Academy of Management Journal. Volume 48, issue.2, S35–S43.https://doi.org/10.1093/geronj/ 48.2.S35. Feldman, D. C. (1994). The decision to retire early; A review and conceptualization. Academy of Management Review, Vol. 19, Issue 2: 285–311. https://doi.org/10.2307/258706. Fletcher, W. L., R. O. Hanson, and L. Bailey (1992). Assessing occupational self-efficacy among middle aged and older adults. Journal of Applied Gerontology, 2(4): 489–501. Flynn, M. (2010). Who would delay retirement? Typologies of older workers. Personnel Review. Vol, 39, issue 3, 308–324. https://doi.org/10.1108/00483481011030511. Henkens, K. and F. Tazelaar (1997). Explaining retirement decisions of civil servants in the Netherlands. Intentions, behavior and the discrepancy between the two. Research on Aging, 19(2): 139–173. Hershey, D. A. and Henkens, K. (2014). Impact of different types of retirement transitions on perceived satisifaction with life. The Gerontologist, Vol. 54, issue 2, 232–244. https://doi.org/10. 1093/geront/gnt006. Higgs, P. G. Mein, J. Ferrie, M. Hyde, and J. Nazroo (2003). Pathways to early retirement: structure and agency in decision-making among British civil servants. Aging and Society,23: 761–778. Kakwani, N. S. and K. Hinz (2006). Old age poverty and social pension in Kenya. International poverty Centre, working paper no. 24. Kinsella, K. (1992). Aging trends in Kenya. Journal of cross cultural gerontology, Vol. 7: Issue 3, 259–268. https://doi.org/10.1007/bf00122512. Loi, J. and K. Shultz (2007). Why older adults seek employment: Differing motivations among subgroups. Journal of Applied Gerontology, 26: 274–289. Mein, G., Martikainen, P., Stansfeld, S. A., Brunner, E. J., Fuhrer, R., & Marmot, M. G. (2000). Predictors of early retirement in British civil servants Age and Aging. Vol. 29, Issue 6, 529–539. https://doi.org/10.1039/ageing/29.6.529. Miller, D. C. and N. J. Salkind (eds). (2002). Research Design & Social Measurements (6th ed.). Thousand Oaks, California: Sage Publications. Mowday, R. R. Steers, and L. Porter (2004). The measurement of organizational commitment. Journal of Vocational Behavior, 14(2): 224–247. Shacklock, K. and Y. Brunetto (2005). Employees’ perceptions of the factors affecting their decisions to retire. International Journal of Organizational Behavior, 10(5): 740–756. Shimizutani, S. (2011). A new Anatomy of the retirement process in Japan. Japan and the world Economy, Elsevier,vol.23(3), 141–152. https://doi.org/10.1016/j.japwor.2011.07.002. Smith, P. C. L. M. Kendall, and C. L. Hulin (1969). The measurement of satisfaction in work and retirement: A strategy for the study of attitudes. Skokie IL: Rand McNally. Spector, P. E. (2006). Method variance in organizational research, Truth or urban legend? Organizational Research Methods, 9(2): 221–232. Szinovacz, M. E. and A. Davey (2005a). Predictors of Perceptions of Involuntary Retirement. The Gerontological Society of America. Volume 45, issue 1, 36–47. https://doi.org/10.1093/geront/ 45.1.36. Szinovacz, M. E. and Davey. A. (2005b). Retirement and Marital Decision Making; Effects on retirement satisfaction. Journal of Marriage and family. Vol. 67, issue 2, page 387–398. https:// doi.org/10.1111/.0022-2445.2005.00123x. Talaga, J. A. and T. A. Beehr (1995). Are there gender differences in predicting retirement decisions? Journal of Applied Psychology, 80: 16–28.

198

L. J. Rono and E. Agasha

Taylor, M. A. and L. Shore (1995). Predictors of planned retirement age: An application of Beehr’s model. Psychology and Aging, 10 (1): 45–57. Toossi, M. (2004). Labor force projections to 2012: The graying of the U.S. workforce. Monthly Labor Review, 127:3–22. Tversky, A. and Kahneman, D. (1974). Judgement under uncertainty; Heuristics and biases. Science. Vol. 185, issue 4157, page 124 131. https://doi.org/10.1126/science.185.4157.1124. Van Solinge, H. and Henkens, K (2014). Work related factors as predictors in the retirement decision making process of older workers in the Netherlands. Ageing and Society. Vol. 34, issue 9. https:// doi.org/10.1017/s0144686x13000330. Van Solinge, H. and K. Henkens (2007). Involuntary retirement; the role of restrictive circumstances, timing and social embeddedness. Journal of Gerontology: Series B, 62 (5): S295–S303. https:// doi.org/10.1093/geronb/62.5.s.295. Wang, M. and K. S. Shultz (2009). Employee Retirement: A Review and recommendations for Future Investigation, Journal of Management. 36(1), 172–206.https://doi.org/10.1177/ 0149206309347957. http://jom.sagepub.com/content/36/1/172. Wong, C. and M. Earl (2009). Towards an integrated model of individual psychosocial and organizational predictors of retirement adjustment. Journal of Vocational Behavior, 75: 1–13. Zaniboni, S. G. Sarchielli, and F. Fraccaroli (2010). How are psychosocial factors related to retirement intentions? International Journal of Manpower, 31(3), 271–285.https://doi.org/10.1108/ 01437721011050576.

Chapter 10

Implementation of Corporate Governance Strategy: An Overview of Africa Ambrose Kipruto Chepkwei

Abstract Corporate governance around the world differs according to the form capitalism in which it is embedded. In the United States, a corporation is governed by a board of directors, which has the power to choose an executive officer. The World Bank defines good corporate governance as a set of laws, rules, and policies which must be followed to motivate corporate sources to perform efficiently and produce a long-term economic value. The alarming rate of organizations’ failures owing to governance malpractices in developing countries coupled with the apparent failure of corporate governance within these organizations and their managements motivated this study. The objective of the study is establishing the implementation of corporate governance strategies in corporate organizations in Africa. Implementation of corporate governance strategies can be understood by using both the agency theory and the resource dependency theory. This study does an in-depth literature study by reviewing existing literature (conference reports, journal articles, and websites) on implementation of corporate governance strategies in Africa. It uses a random sample approach to study and survey existing research on the implementation of corporate governance strategies in 12 African countries. It concludes that corporate governance is the solution for global organizational problems and African countries in particular need to provide the right policies that enhance increased productivity that brings about economic growth. Corporate governance is needed to motivate the boards and managements of corporates to pursue their objectives in the interests of their companies and shareholders and for facilitating effective monitoring through the use the firms’ resources efficiently. Keywords Corporate · Implementation · Strategy · Board · Governance

A. K. Chepkwei (B) Jomo Kenyatta University of Agriculture and Technology, Nairobi, Kenya e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 J. Wood and O. Habimana (eds.), A Multidimensional Economic Assessment of Africa, Frontiers in African Business Research, https://doi.org/10.1007/978-981-15-4510-8_10

199

200

A. K. Chepkwei

10.1 Introduction The concept of corporate governance is fairly new as in the early 17th century this concept did not exist because in those days, ownership was divided among a small number of people (partnerships) who participated in organizations’ operations. Therefore, they controlled and safeguarded their own interests (Ali 2016). Corporate governance around the world differs according to the form of capitalism it is embedded in. A corporation in the United States is governed by a board of directors, which has the authority to choose an executive officer, usually known as the chief executive officer. There has been a recent revival of concern about large corporations closing down all over the world because of various reasons (Jayeola and Olufemi 2011). Corporate governance can be defined as an act of the management in a business environment in financial markets and it has recently been used as a key management strategy by corporates both developed and developing countries (Sonmez and Yildirim 2015). Corporate governance is the relationship between the corporate governance committee, the shareholders, and other parties who are interested in a company. The World Bank further defines good corporate governance as a set of laws, rules, and policies which must be followed to motivate corporate sources to perform efficiently and produce a long-term economic value which will be beneficial for shareholders and citizens (Siallagan and Januarti 2014).

10.1.1 History of Corporate Governance The concept of corporate governance dates back to the 19th century when state corporation laws were enhanced through the rights of corporate boards without consent of their shareholders. The idea at that time was that corporations should have good board structures to enhance their performance which was firmly rooted on the assumption that good corporate governance practices enhanced performance (Shungu et al. 2014). However, corporate governance and businesses became the focus of study in the 21st century. Corporate governance is about the manner in which authority is exercised over corporate entities. In today’s environment, stakeholders have high expectations that companies should be run following good corporate governance practices (Indermun and Bayat 2015). Corporate governance gained relevance and prominence in global society after big corporate scandals such as those involving Enron and WorldCom (Maune 2015). Corporate governance is concerned with the process, systems, practices, and procedures that govern institutions as well as the resolution of problems arising out of the collective actions of dispersed investors and the reconciliation of conflict of interest among the board of directors, the management, and shareholders. Sound corporate governance regimes are a positive development as they boost investor confidence and maximize shareholders’ returns (Ruparelia and Njuguna 2016). Corporate

10 Implementation of Corporate Governance Strategy …

201

governance brings in a new perspective through external independent directors; it enhances firms’ corporate entrepreneurship and competitiveness (Afande 2015). Good corporate governance has been highlighted as being vital for corporate organizations especially in transition and emergent economies. However, owing to corporates’ economic failures across the globe, there is no doubt that the implementation of the principles of corporate governance is significant for ensuring good governance in every economy (Agymang et al. 2013). Corporate governance is concerned with governing corporate entities. Corporate governance contributes towards economic stability by enhancing the performance of organizations and increasing their access to outside capital. Corporate governance also has to do with the relationships between the management, the board of directors, controlling shareholders, monitoring shareholders, and other shareholders (Latif et al. 2013). The global economy appears to be on the road to recovery after suffering its worst recession in the post-war era. Swift developments and transformations of any organization in today’s hypercompetitive environment depends on the quantum of strategic relationships to a large extent which further enhance service excellence making it an important strategic step that organizations can take for understanding and implementing the concepts and principles of corporate governance (Omolade and Tony 2014). The failure of corporates such as Enron, WorldCom, and Tyco shifted attention to governance in organizations. Organizations have given more emphasis to corporate governance studies mainly, after the passing of Sarbanes-Oxley Act in the United States in 2002 (Surej 2013). The first guidelines on public sector corporate governance were developed by the UK based Cadbury Report (Mulyadi et al. 2012). After major scandals of organization giants like Enron and WorldCom collapsing, further attention was directed towards corporate governance. The concept of corporate governance had been there even earlier, but it was not being implemented properly due to such things happened to large corporations. After such scandals, a need for corporate governance arose as shareholders and managements were seen as separate as were their interests, so a moderator was required to fill this gap and become a bridge between the two (Ali 2016). Corporate governance is growing in importance in today’s modern world. Although corporate governance is deemed as the final product of the 20th and 21st century economies, old economic growth theories too had been aware of its importance for growth and development. The roots of corporate governance can also be traced back to ancient economies of India and Greece (Skare and Hasic 2016).

10.1.2 Statement of the Problem The alarming rate of failure of organization sowing to governance malpractices and mismanagement in developing countries coupled with the apparent failure of corporate governance within organizations and their managements is the main motivation

202

A. K. Chepkwei

for this research (Opata and Awino 2017). One of the main reasons for business failures from the 1980s till now is improper ethical values and weak corporate governance in corporations (Al-Azzam et al. 2015). Lack of good corporate governance has led to poor performance of organizations as well as the suppressing of good and sustainable organizational decisions (Mbalwa et al. 2014). Corporate governance has recently become a central issue in corporations’ success. In particular, in the wake of a number of scandals, such as Enron, Parmalat, WorldCom, and Lehman Brothers, its importance has been understood by both developed and developing countries (Sonmez and Yildirim 2015). Failures related to corporate governance have taken various dimensions with different implications, especially in profit-oriented business organizations and has become an issue of global significance (Olannnye and Anuku 2014). Corporate governance as a subject has continued to draw a lot of interest in global academic circles. The extent to which corporate governance contributes to a firm’s value continues to attract a lot of debate from both academicians and corporate players (M’Ithiria and Musyoki 2014). Main Objective: Implementation of corporate governance strategies among corporate organizations in Africa.

10.2 Literature Review Implementation of corporate governance strategies among organizations can be studied using both the agency theory and the resource dependency theory. Therefore, this study examines the implementation of corporate governance strategies based on these two theories.

10.2.1 Agency Theory Corporate governance is developed based on the agency theory in which corporate governance must be monitored and controlled to ensure that governance is done according to rules and policies (Siallagan and Januarti 2014). The agency theory advocates that corporate governance can reduce agency costs which in turn leads to improved firm performance. The agency theory is important in corporate governance, because it forms the basis of policy regarding proper governance of corporations (Homayoun and Homayoun 2015). There is no doubt that the agency theory and its view of a firm as a complex nexus of contracts, constitutes one of the major pillars of theoretical values as it not only helps understand and explain the behavior of business actors, but also provides a rich fund of practical implications for designing governing structures (Balago 2014). Agency theory envisages that the chief executive officer (CEO) and the chairman’s positions should be assumed by different individuals to protect shareholders’ interests. Although in the last decades of the 20th century, the agency theory became

10 Implementation of Corporate Governance Strategy …

203

a dominant force in a theoretical understanding of corporate governance it does not cover all aspects of corporate governance. Finally, it can be posited that agency is a significant proposition in a firm’s discipline. The theory assumes that when ownership and control are separated in a firm, the CEO and manager act as agents on behalf of the principal (Mamun et al. 2013). Agency theory is a radical, powerful foundation which is predominantly used for explaining and predicting various aspects of corporate governance. However, the agency theory only gives a restricted view of corporate governance that is effective (Abid et al. 2014). According to agency theory, the relationship between the owners and the management is regarded with importance especially when introducing governance mechanisms to resolve potential conflicts (Afza and Nazir 2014).

10.2.2 Resource Dependence Theory Resource dependence theory explains that a company needs several resources for completing its operations successfully which is not possible without the assistance of directors or board members (Afza and Nazir 2014). This theory underscores the importance of the board of a company as a resource and envisages its role beyond its traditional responsibilities considered from the agency theory perspective (Yusoff and Alhaji 2012). The resource dependency theory concentrates on the role of the board of directors as giving advice, or providing strategies, and providing access to resources and it recommends higher representation of inside directors (Aduda et al. 2018). The resource dependence theory is realistic about the role of the management. The management’s ability to act is limited by the resources available. Resource dependency theory also assumes that bounded rationality applies to managers: the perception of the environment is directed and filtered by cognitive structures. This theory assumes that organizations create their own environments, change, and disapprove of resistance (Werner 2008).

10.2.3 Role of the Board of Directors The board of directors form the cornerstone of effective corporate governance. Efficient roles and responsibilities of board members with commitments to complying with rules and regulations can help in creating value and protecting the interests of the stakeholders. Designing a system of governance in which the board of directors monitor and ensure that the managers are fulfilling their responsibilities (Zerban et al. 2017) is therefore needed. A majority of the directors should have independent status and minds. Directors ought to be autonomous and not under the management and free of all business and other relationships which could materially interfere with their exercising independent

204

A. K. Chepkwei

judgement. Directors who are considered independent by the board should be so acknowledged in the statutory annual reports under the subject corporate governance (Mudashuru et al. 2014). The corporate board of directors contribute to corporate governance through supervision of the executive management and taking strategic decisions for the organization. The board is supposed to govern the corporation optimally on behalf of the shareholders and effectively act as their trustee. The board of directors can play an important role in making sure that a firm’s outward looking approach is good (Jan and Sangmi 2016). The board and the management ought to be trained on the demarcation of duties between them to avoid the board doing the duties of the management which may lead to sour relations between the two (Chisi and Gondwe 2017).

10.2.4 Corporate Governance Structures Corporate governance structures entail a board of directors, an executive, a management team, and departments that may be organized according to functions, divisions, or a combination of both. The board represents a higher level of power, control, and authority in an organization (Nmai and Delle 2014). An effective corporate governance structure improves investor confidence and it enhances the integrity and efficiency of the capital market (Alnaser et al. 2014). Effective ways of corporate governance provide structures through which the goals of the organization are set and also working out ways of accomplishing these goals in addition to the structures of product and factor markets. Effective governance requires a clear understanding of the roles of the board and senior management and their relationships with others in the corporate structure (Al-Azzam et al. 2015).

10.2.5 Role of Corporate Governance It has been argued that corporate governance plays a critical role in determining the strategic directions that corporations take. This can be achieved by formulating strategic choices that use a company’s internal resources to support it in the external environment for optimal performance (Kamau et al. 2018a). Good corporate governance is said to possess the following characteristics: it is participatory; consensus oriented; accountable; transparent; responsive; effective and efficient; equitable and inclusive; and follows the rule of law (Pal 2017).

10 Implementation of Corporate Governance Strategy …

205

10.2.6 Organizational Performance Corporations cannot operate automatically. Their operations require the involvement and coordination of staff members. If a corporate ignores the interests of its employees and managers, the process of corporate management will inevitably lead to moral hazards and adverse selection problems (Zhang 2015). For any organization, corporate governance is a key factor that determines the health of the system and its ability to survive economic shocks. Good governance, therefore, contributes to sustainable economic development by enhancing companies’ performance and increasing their access to external capital (Sarbah and Xiao 2015). It is argued that good corporate governance is a key determinant of organizational performance (Kamau et al. 2018b).

10.2.7 Information Technology Globalization and information technology have made corporate governance complex but at the same time relevant, because organizations have become powerful in that information can now be relayed faster than before across the world (Zuva and Zuva 2018). Corporate governance is a system in which we can give proper rights to various departments in the firm equally and the requirements of a particular department and the division of resources can be done according to their responsibilities through which the effectiveness and productivity of the firm also increases (Mujahid et al. 2014). Corporate governance as a new management system tool has recently provided solutions to different problems in the management of organizations. In fact, corporate governance is a new born concept and practice but is not a new born idea. The concept differs from one country to another as their economic environments and development levels are not equal. This is also true for corporate governance in small and medium sized enterprises which differ from each other (Brahim and Nourrendine 2017).

10.2.8 Composition of the Board of Directors The board of directors should be of a sufficient size relative to the scale and complexity of the organization’s operations and composed in such a way so as to ensure diversity of experience without compromising independence, compatibility, integrity, and availability of members to attend meetings. It should have a mix of executive and non-executive directors, headed by the chairman. A majority of the board members should be non-executive (Ejuvbekpokpo and Esuike 2013). The board’s composition is a critical element of corporate governance with the board mandated with supervisory and advisory roles in a company’s management. This has created the

206

A. K. Chepkwei

belief that the board of directors can influence a firm’s strategic decision making and subsequently its performance (Cherotich and Obwogi 2018).

10.3 Research Methodology This study engaged in an in-depth literature study, which is descriptive in nature by reviewing existing literature (conference reports, journal articles, and websites) on the implementation of corporate governance strategies in various organizations in African countries. The researcher randomly sampled and surveyed existing research on corporate governance strategies’ implementation in 12 countries in Africa with the principle objective of studying the corporate governance strategies and how they were being implemented by corporates in different African countries. Methodically, this study follows a qualitative approach by employing secondary data in which a document analysis of relevant empirical literature of published journal articles and a review of different corporate governance theories was used to formulate the discussion (Meressa 2017).

10.4 A Discussion of the Findings 10.4.1 Botswana Botswana has recently developed and adopted its own corporate governance code, giving it an opportunity to address the perceived gaps in the existing governance structures in its context (Josiah et al. 2016). Research on corporate governance, especially in developing countries such as Botswana is limited. Corporate governance in the consumer sector in Botswana is poorly understood and its specific governance issues remain unexplored despite the fact that the sector plays a very important role in the social and economic development of the country (Sathyamoorthy et al. 2017).

10.4.2 Egypt Corporate governance is believed to have major implications for the growth prediction of the Egyptian economy. Corporate governance strategies and practices are viewed as important in reducing risks for investors; catching the attention of investment capital; and improving the performance of firms in the country. However, the way in which the corporate governance strategies are pre-arranged differs between countries, based on their economic, political, and social circumstances (Emile et al. 2014). There is absence of effective application of appropriate corporate governance

10 Implementation of Corporate Governance Strategy …

207

strategies in listed firms in Egypt (El-Habashy 2018). As a global trend, corporate governance gained prominence in Egypt’s context in the 1990s when it started implementing its economic reforms program. One of these trends was the establishment of an Egyptian corporate governance code. The final version of this code was presented in 2005. This code plays an important role in making Egyptian companies more transparent and understandable for local and global investors (Desoky and Gehan 2012).

10.4.3 Ghana Ghana is an example of the failure of those in power in administering laws and regulations in relation to corporate governance (Agymang et al. 2013). The need to minimize corruption in the Ghanaian business and corporate sector is obvious and this calls for governmental interventions in the form of legislations for establishing strict governance structures at all levels. Corporate governance standards relate to both listed and non-listed firms and the private and public sectors in Ghana such as non-governmental organizations, charities, business and government boards, trusts, and agencies (Sarbah and Xiao 2015).

10.4.4 Rwanda Despite the fact that Rwanda has ratified most of the key international standards and codes in corporate governance and that a regulatory framework for promoting good governance is in place, there is still overall lack of awareness about this concept (Opata and Awino 2017). The Rwanda Private Sector Federation (PSF) has gone as far as publishing the guiding corporate governance code for member companies to use while designing their corporate governance codes since there exist no compulsory standards either for private or public companies. The new company law in Rwanda was promulgated in 2009 entrenching corporate governance policies for the first time in the companies’ codes (Kayihura 2013).

10.4.5 Morocco Even though corporate governance in Morocco has achieved some milestones during the past decade, in practice its corporate governance is still in an embryonic stage. Except for the banking sector, compliance with the best practices of corporate governance, is still done on a voluntary basis. Further, there is no existing practice that is suitable for the Moroccan market which measures the degree of compliance of Moroccan companies in terms of corporate governance (Samlal and Jahidi 2018).

208

A. K. Chepkwei

Since the 1990s, Morocco has implemented a review procedure for corporate governance legislations and regulations of its system. In 2000 and later it became necessary for Morocco to introduce self-regulatory rules by March 2008 and it introduced its first corporate governance code of best practices (Allali 2016).

10.4.6 Nigeria In Nigeria, some corporations’ failures have been seen as a fallout of non-observance of the code of corporate governance which is in place. Nigeria’s code of corporate governance (CCG) (2003) was formulated to address the decay and failure in corporate and public governance in the country. Therefore, it is expected that the right corporate governance should have a positive impact on the performance of Nigerian organizations (Ibadin and Dabor 2015). In Nigeria, observing the principles of governance has been secured through a combination of voluntary and mandatory mechanisms. In 2003, the Atedo Peterside Committee (APC) set up by the Securities and Exchange Commission (SEC) developed a Code of Best Practices for Public Companies in Nigeria (Kehinde et al. 2012). However, despite the recent introduction of the Code of Best Practices on Corporate Governance in Nigeria which gives evidence of Nigeria’s share of intelligent business practices, the country has had its share of failed corporate giants that once stood firm without any sign of trouble (Isaac 2014). Recently, Nigeria laid the foundation for corporate governance by sponsoring a series of legislative, economic, and financial reforms that are intended to promote transparency, accountability, and rule of law in the economic life of the country. For a developing country like Nigeria corporate governance is of critical importance. In its recent history, the lack of corporate governance has led to economic upheavals in the country (Isaac 2014).

10.4.7 Kenya Effective monitoring of corporate governance practices and long-term commitment to good practices are essential for listed companies in Kenya (Ruparelia and Njuguna 2016). In Kenya, the institutions that have been at the forefront in sensitizing the corporate sector on corporate governance are the Capital Markets Authority, the Nairobi Securities Exchange, Center for Corporate Governance, and the Central Bank of Kenya which regulates the banking industry. The Capital Markets Authority had a major impact on the development of corporate governance practices by public listed companies in 2002. These guidelines were published under gazette notice No. 369 of January 25, 2002 (62).

10 Implementation of Corporate Governance Strategy …

209

10.4.8 South Africa Since the 20th century, corporate governance mechanisms have been developed globally to curb the negative effects of the agency problem. South Africa published the first King Report on corporate governance in 1994. South Africa is widely considered as a pioneer in the field of corporate governance. The first King Report provided very specific corporate governance guidelines to the companies listed on the Johannesburg Stock Exchange. This report was revised in the early 2000s due to changes in legislation and developments in global corporate governance (Mans-Kemp et al. 2016). The likely direction of corporate governance reforms in South Africa is maintaining a general alignment with the United Kingdom approach. The commitment to shareholders’ interests combined with sustained attention on issues relating to South Africa’s development needs is the trajectory that will further develop a hybrid model of corporate governance in the country (Andreason 2011).

10.4.9 Libya Survival in a competitive market and good corporate governance are essential and important for the growth of modern corporations. Corporate governance affects corporations’ performance. In Libya, corporate governance practices were established in 1950 when oil and gas was discovered (Gabasi et al. 2014). Therefore, corporate governance issues are important especially in transition economies such as Libya, since many of the developing countries do not have established organizational infrastructure that can deal with issues that arise during the adoption of good corporate governance practices (Iswaissi and Falahati 2017). The concept of corporate governance is still new and is still misunderstood among business organizations in the country (Zagoub 2016).

10.4.10 Zimbabwe In Zimbabwe, corporate governance is gaining momentum as evidenced by the publishing of the corporate governance guidelines by the Reserve Bank of Zimbabwe in 2004. The financial crisis that affected the banking industry in 2003–04 led to failures which were attributed to poor corporate governance practices in Zimbabwe (Dube and Murahwe 2015). Several companies in Zimbabwe folded up especially at the turn of the new millennium, companies in both the private and public sectors were affected by corporate scandals (Zuva and Zuva 2018). The Government of Zimbabwe must set up rules and regulations on corporate governance and put in place measures against offenders to ensure that there is good corporate governance

210

A. K. Chepkwei

in the country. This will allow the law to take its course in case there are corporate governance breaches (Obert and Patience 2018).

10.4.11 Ethiopia Currently Ethiopia has several companies that have been formed by selling shares to the general public. The emergence of publicly held companies in Ethiopia has given rise to a multitude of complex corporate governance issues. A revision of the 1960 commercial code is underway by the Federal Democratic Republic of Ethiopia’s Ministry of Justice, which is vital for improving and upgrading corporate governance in the country (Tura 2014). The overall standard of corporate governance in Ethiopia is inadequate and needs overhauling with the adoption of good corporate governance principles (Ayele 2013). The issue of corporate governance has long been overlooked in developing economies like Ethiopia. Since the 1960s, financial institutions as well as other business entities have been using the commercial code of Ethiopia as their administrative tool (Yemane et al. 2015).

10.4.12 Algeria Although the concept of corporate governance is old in Algeria, it is new in terms of implementation. The implementation of corporate governance strategies in Algeria is determined through two aspects: the first relates to the ministry commandment and motivation and the second to corporate organizations themselves to help prevent administrative corruption (Nourredine and Brahim 2017). In 2003, the World Bank launched a Country Assistance strategy in Algeria for fiscal years 2004–06 aimed at increasing the state’s capacity to regulate the market and encourage the private sector to embrace sound corporate governance practices through technical assistance. However, significant deficiencies remain with regard to laws and regulations as Algeria lacks a modern corporate governance framework (Noureddine 2015).

10.5 Conclusion Corporate governance brings a new strategic outlook to countries in Africa including Rwanda through external independent directors and it enhances a firm’s corporate entrepreneurship and competitiveness. It is not a threat to value creation in entrepreneurial firms if the guidelines on corporate governance are properly applied (Sarbah and Xiao 2015). This study found that the adoption of good corporate governance practices enhanced transparency of a company’s operations, ensured accountability, and improved the firm’s profitability. Corporate governance also helps protect

10 Implementation of Corporate Governance Strategy …

211

the interests of the shareholders by aligning their interests with those of the managers. With a good corporate governance record, companies can generate more resources to create more employment opportunities, support businesses through prompt accident claims, claims payment of dividends, and taxes to the government (Mudashuru et al. 2014). Corporate governance is an internal system which includes policies, people, and processes that serve the needs of shareholders by controlling and directing the activities of the management with objectivity and integrity. It is important in ensuring that responsibility and accountability are incorporated in every part of an organization (Zaman and Quazi 2015). Hence, corporate governance contributes to sustainable economic development by enhancing companies’ performance and increasing their access to external capital (Sarbah and Xiao 2015). Legislation in relation to corporate governance and an analysis of corporate governance in Africa showed that the institutions and legal frameworks for effective corporate governance were poor and compliance and/or enforcement was weak or did not exist (Ayandele and Isichei 2013). Hence, the implementation and promotion of corporate governance strategies in Africa includes weak or non-existent law enforcement mechanisms, lack of commitment on the part of the boards to their responsibilities, lack of adherence to legal frameworks, and lack of transparency and disclosures (Okpara 2010). Good corporate remains a solution for global organizational problems especially in Africa. The right policy that enhances increased productivity and brings about economic growth and development globally, and in Africa as a continent (Oguntibeju et al. 2014) must be followed. Corporate governance is therefore about what the board of a company does and how the countries in Africa set regulations for implementing corporate governance, while organizational values are made distinct from the operational management of companies (Tihanyi et al. 2014).

10.5.1 Recommendations Corporate governance motivates the boards and managements of African companies to pursue objectives that are in the interests of the company and its shareholders and countries should facilitate their effective monitoring encouraging firms to use resources more efficiently for corporate governance (Yesser 2011). Corporate governance is recognized in Africa as one of the most important paths for building confidence in the market and attracting positive investors to organizations and the economy (Ahmed and Hamdan 2015).

212

A. K. Chepkwei

References Abid, A. B. Kahn, Z. Rafiq, Z. abd A. Ahmed (2014). Theoretical Perspectives of Corporate Governance. Bulletin of Business and Economics, 3(4): 166–175. Aduda, J. R. Chogii, and P. Magutu (2018). An Empirical Test of Competing Corporate Governance Theories on the Performance of Firms Listed at Nairobi Securities Exchange. European Scientific Journal, 9(13): 108–137. Afande, F. (2015). Adoption of Corporate Governance Practices and Financial Performance of Small and Medium Enterprises in Kenya. Research Journal of Finance and Accounting, 6(5): 36–73. Afza, T. and M. Nazir (2014). Theoretical Perspective of Corporate Governance: A review. European Journal of Scientific Research, 119(2): 255–264. Agymang, O. E. Aboagye, and A. Ahali (2013). Prospects and Challenges of Corporate Governance in Ghana. International Journal of Scientific and Research Publication, 3(5): 1–9. Ahmed, E. and A. Hamdan (2015). The Impact of Corporate Governance from Bahrain Stock Exchange. European Journal and Innovation Research, 3(5): 25–48. Al-Azzam, Z. A. Al-Qura’an, and H. Al-Mohameed (2015). How the Corporate Governance Affects Organization Strategy: Lessons from Jordan Environment. IOSR Journal of Business and Management, 17(4): 52–66. Ali, M. (2016). Impact of Corporate Governance on Firm’s Financial Performance (A Comparative Study of Developed and Non Developed Markets). Arab Journal of Business and Management Review, 6(6): 2–6. Allali, A. (2016). Corporate Governance and Corporate Social Responsibility in Islamic Banking: The Case of the Moroccan Banks in Italy. European Journal of Islamic Finance, 1: 2–10. Alnaser, N. O. Shaban, and Z. Al-Zubi (2014). The Effect of Effective Corporate Governance Structure in Improving Investors’ Confidence in the Public Financial Information. International Journal of Academic Research in Business and Social Science, 4(1): 556–569. Andreason, S. (2011). Understanding Corporate Governance Reform in South Africa: AngloAmerican Divergence, the King Reports, and Hybridization. Business and Society, 50(4): 647–673. Ayandele, I. and E. Isichei (2013). Corporate Governance Practices and Challenges in Africa. European Journal of Business and Management, 5(4): 21–59. Ayele, A. (2013). Revisiting the Ethiopian Bank Corporate Governance System: A Glimpse of the Operation of Private Banks. Law, Social Justice & Development (LGD), (1): 2–37. Balago, G. (2014). A Conceptual Review of Agency Models of Performance Evaluation. International Journal of Finance and Accounting, 3(4): 244–252. Brahim, Z. and F. Nourrendine (2017). Corporate Governance among Small and Medium size Enterprise in Algeria: Impediments to the Practice of Corporate Governance System. Asian Journal of Economic Modelling, 5(2): 154–166. Cherotich, D. and J. Obwogi (2018). Effect of Board Composition on Financial Performance of Listed Firms in Nairobi Securities Exchange. The Strategic Journal of Business and Management, 5(4): 1239–1250. Chisi, R. and L. Gondwe (2017). An Investigation on Corporate Governance on Savings and Credit Cooperatives (SACCOs) in Southern Malawi. IOSR Journal of Business Management, 19(4): 07–18. Desoky, A. and A. Gehan (2012). Corporate Governance Practices: Transparency and DisclosureEvidence from Egyptian Exchange. Journal of Accountancy, Finance and Economics, 2(1): 49–72. Dube, Z. and G. Murahwe (2015). An Analysis of Corporate Governance Practices in Government Controlled Versus Private Control Banking Institutions in Zimbabwe. International Journal of Development and Sustainability, 4(6): 759–771. Ejuvbekpokpo, S. and B. Esuike (2013). Corporate Governance Issues and Its Implementation: The Nigerian Experience. Journal of Research in International Business and Management, 3(2): 53–57.

10 Implementation of Corporate Governance Strategy …

213

El-Habashy, H. (2018). Determinants of Capital of Capital Structure Within the Context of Corporate Governance in Egypt. International Journal of Business and Management, 13(8): 26–39. Emile, R. A. Ragab, and S. Kyaw (2014). The Effect of Corporate Governance on Firm Performance: Evidence from Egypt. Asian Economic and Financial Review, 4(12): 1865–1877. Gabasi, A. K. Kertahadi, and N. Firdausi (2014). An Analysis of Corporate Governance and its Impact on the Firm’s Financial Performance (A Study in ENI Oil and Gas Company). IOSR Journal of Business and Management, 16(7): 61–70. Homayoun, S. and S. Homayoun (2015). Agency Theory and Corporate Governance. International Business and Management, 9(5): 805–815. Ibadin, P. and E. Dabor (2015). Corporate Governance and Accounting Quality: Empirical Investigations from Nigeria. Journal of Policy and Development, 9(2): 64–82. Indermun, V. and M. Bayat (2015). A focus on Corporate Governance and the Relationship with Business. Business and Economics Journal, 6(1): 2–8. Isaac, L. (2014). Corporate Governance and Organization Performance in the Nigerian Banking Industry. European Journal of Business and Management, 6(25): 110–118. Iswaissi, H. and K. Falahati (2017). Challenges to Corporate Governance Practices: Case Study of Libyan Commercial Banks. Corporate Governance and Sustainability Review, 1(1): 33–71. Jan, S. and M. Sangmi (2016). The Role of Board of Directors in Corporate Governance. Imperial Journal of Interdisciplinary Research, 2(5): 707–715. Jayeola, O. and O. Olufemi (2011). Corporate Governance and Performance of Nigerian Banking Sector. International Journal of Development and Management Review, 6: 1–12. Josiah, J. G. Themba, and T. Matenge (2016). Corporate Governance in Botswana: Exploring Developments and Compliance. Botswana Journal of Business, 9(1): 1–20. Kamau, G. E. Aosa, V. Machuki, and G. Pokhariyal (2018a). Corporate Governance, Strategic Choices and Performance of Financial Institutions in Kenya. International Journal of Business and Management, 13(7): 169–178. Kamau, G. V. Machuki, and E. Aosa (2018b). Corporate Governance and Performance of Financial Institutions in Kenya. Academy of Strategic Management Journal, 17(1): 1–13. Kayihura, M. (2013). The Corporate Governance Approaches: The Shareholder versus the stakeholder. The Case of Rwanda. Economic and Management, 1(1): 20–31. Kehinde, A. F. Adekunjo, K. Kayode, O. Musibau, and F. Adedamola (2012). The Impact of Corporate Governance and Leadership on Enterpreneurship Development in Nigeria. IOSR Journal of Humanities and Social Science, 6(2): 46–68. Latif, B. M. Shahid, M. Hag, H. Waqas, and A. Arshad (2013). Impact of Corporate Governance on Firm Performance: Evidence from Sugar Mills of Pakistan. European Journal of Business and Management, 51–59. M’Ithiria, E. and D. Musyoki (2014). Corporate Governance, Ownership Structure Perspective and Firm Value: Theory, and Survey Evidence. International Journal of Research in Management & Business Studies, 1(3): 57–61. Mamun, A. Q. Yesser, and A. Rahman (2013). A Discussion of the Suitability of Only One Vs More than One Theory for Depicting Corporate Governance. Modern Economy, 4: 37–48. Mans-Kemp, N. P. Erasmus, and S. Viviers (2016). Advances in the Corporate Governance Practices in Johannesburg Stock Exchange Companies. South African Business Review, 20: 93. Maune, A. (2015). Corporate Governance in Zimbabwe: An Overview of Its Current State. Asian Economic and Financial Review, 5(1): 167–178. Mbalwa, P. H. Kombo, L. Chepkoech, S. Koech, and P. Shavulimo (2014). Effects of Corporate Governance on Performing of Sugar Manufacturing Firms in Kenya. IOSR Journal of Business and Management, 16(11): 86–112. Meressa, H. (2017). Application of Corporate Governance Principles for the Sustainability and Competitiveness of Small Scale Enterprises: A Literature Review. European Journal of Business Management, 9(10): 31–37.

214

A. K. Chepkwei

Mudashuru, A. I. Bakare, Y. Babatunde, and O. Ishmael (2014). Good Corporate Governance and Organization Performance: An Empirical Analysis. International Journal of Humanities and Social Science, 4(7): 170–178. Mujahid, M. M. Zuberi, M. Rafiq, S. Sameen, and S. Manzoor (2014). Corporate Governance and its Effect on Decision Making of the Firm. Journal of Economics and Sustainable Development, 5(24): 30–34. Mulyadi, M. Y. Anwar, and M. Ikbal (2012). The Importance of Corporate Governance in Public Sector. Global Business and Economics Research Journal, 1(1): 25–31. Nmai, B. and E. Delle (2014). Good Corporate Governance and Employee Job Satisfaction: Empirical Evidence from Ghanian Telecommunication Sector. International Journal of Humanities and Social Science, 4(13): 209–217. Noureddine, A. (2015). Corporate Governance Application in Algerian Financial Sector. International Journal of Research in Business Studies and Management, 2(11): 142–149. Nourredine, F. and Z. Brahim (2017). Corporate Governance Perspective in Algerian Small and Medium Scale Enterprises. Global Journal of Management and Business Research, 17(2): 13–22. Obert, S. and D. Patience (2018). Corporate Governance and the Law Governing the Management of Companies in Zimbabwe. IOSR Journal of Business and Management, 20(2): 66–70. Oguntibeju, J. I. Olayinka, A. Kehinde, and F. Adedamola (2014). Leadership, Corporate Governance and Entrepreneurship Development in Nigeria: A Mutual Link. Journal of Economic and Sustainable Development, 5(25): 67–82. Okpara, J. (2010). Perspective of Corporate Governance Challenges in Sub-Saharan African Economy. Journal of Business & Policy Research, 5(1): 110–122. Olannnye, P. and D. Anuku (2014). Corporate Governance and Organization Performance in the Nigeria Banking Industry. Journal of Emerging Trends in Economic and Management Sciences, 5(6): 525–531. Omolade, O. and O. Tony (2014). Effect of Corporate Governance on the Survival and Sustainability of Banks in Nigeria. American Journal of Engineering Research: 73–83. Opata, E. and Z. Awino (2017). Corporate Governance Practices and Performances in the Mobile and Data Services Companies in Rwanda. Journal of Public Policy and Administration, 1(1): 21–42. Pal, R. (2017). Challenges and Prospects of Good Governance in Reduction of Poverty: A Case Study of Buee Town 01 Kebele, Ethiopia. Global Journal of Political Science and Administration, 5(3): 13–27. Ruparelia, R. and A. Njuguna (2016). The Evolution of Corporate Governance and Consequent Domestication in Kenya. International Journal of Business and Social Sciences, 7(5): 153–163. Samlal, Z. and R. Jahidi (2018). Empirical Evidence of Impact of Best Corporate Governance on Financial Performance: Case Study of Blue Chip Companies in Morocco. IAETSD Journal for Advanced Research in Applied Sciences, 5(3): 438–447. Sarbah, A. and W. Xiao (2015). Good Corporate Governance Structures: A must for Family Business. Open Journal of Business and Management, 3: 40–47. Sathyamoorthy, C. P. Baliyan, M. Dzimiri, and L. Wally-Dima (2017). The Impact of Corporate Governance on Financial Performance: The Case of Listed Companies in the Consumer Services Sector in Botswana. Advances in Social Sciences Research Journal, 4(22): 63–77. Shungu, P. G. Ngirande, and G. Ndlovu (2014). Impact of Corporate on Performance of Commercial Banks in Zimbabwe. Mediterranian Journal of Social Science, 5(15): 93–105. Siallagan, H. and I. Januarti (2014). The Effect of Good Corporate Governance Implementation and Proportions of the State Ownership on Banking Firms Market Value. International Journal of Business, Economic and Law, 5(1): 28–37. Skare, M. and T. Hasic (2016). Corporate Governance, Firm Performance and Economic Growth— Theoretical Analysis. Journal of Business Economics and Management, 17(1): 35–51. Sonmez, M. and S. Yildirim (2015). A theoretical Aspect on Corporate Governance and its Fundamental Problems: It is a Cure or another Problem in the Financial Markets. Journal of Business Law and Ethics, 3(1): 20–35.

10 Implementation of Corporate Governance Strategy …

215

Surej, J. (2013). An Introduction to IT Governance and Various IT Governance Framework. Asian Conference of Corporate Governance and Business Sustainability. Bangkok, Thailand: Asian Conference of Corporate Governance and Business Sustainability. Tihanyi, L. S. Gaffin, and G. George (2014). Rethinking Governance in Management Research. Academy of Management Journal, 57(6): 1535–1543. Tura, H. (2014). Reforming Corporate Governance in Ethiopia: Appraisal of Competing Approaches. Oromia Law Journal, 3(1): 169–211. Werner, M. (2008). Resource Dependence Theory: How well does it Explain Behaviour of Organizations? Mering, 19(1): 9–32. Yemane, A. L. Raju, and M. Raju (2015). Impact of Corporate Governance on Firm’s Performance: Evidence from Ethiopian Insurance Companies. Research Journal of Finance and Accounting, 6(9): 225–233. Yesser, Q. (2011). Corporate Governance and Performance (A Case Study of Pakistani Sector). International Journal of Trade, Economics and Finance, 2(3): 204–211. Yusoff, W. and I. Alhaji (2012). Insight of Corporate Governance Theories. Journal of Business and Management, 1(1): 52–63. Zagoub, A. (2016). Concepts and Practices of Corporate Governance in Libyan Commercial Banks: A Stakeholders Perspective. Journal of Economics and Business Studies, 3(2): 2–14. Zaman, F. and Z. Quazi (2015). Literature Review on Corporate Governance Structure and Performance in Non-Financial Firms in Bangladesh. Asian Journal of Finance & Accounting, 7(1): 96–104. Zerban, A. M. Abdullah, and M. Abdullateef (2017). Corporate Governance and Board of Directors Responsibilities: The Case of Saudi Arabia. International Journal of Accounting Research, 5(2): 1–5. Zhang, Q. (2015). Research of Corporate Performance Evaluation Indicators Based on Shareholder Theory. Open Journal of Social Science, 3: 111–116. Zuva, J. and T. Zuva (2018). Corporate Governance and Organizational Performance. International Journal of Business and Management Studies, 10(1): 16–29.

Chapter 11

Efficiency of Commercial Banks in Ghana Eric Baah Nyannor and Isaac Osei Mensah

Abstract This study investigates the efficiency of commercial banks in Ghana following the stochastic frontier analysis (SFA) approach. We analyzed the impact of ownership structure and the recent global financial crisis on the efficiency of the 26 commercial banks in the country. The data used cover the period 2001–10. Although the average efficiency scores declined after the financial crisis, the maximum likelihood estimation shows that the financial crisis reduced both cost and technical inefficiencies. All banks were efficient during the period. The maximum likelihood estimation (MLE) and the descriptive statistics also show that all bank ownership types reduced cost inefficiencies, while the ownership structure had no effect on technical inefficiencies. Keywords Capitalization · Bank efficiency · Ownership · Financial crisis · Ghana JEL Code G28 · G20 · G01

11.1 Introduction Firms’ efficiency and performance have attracted a lot of attention over the years. A considerable number of studies have examined the relationship between efficiency and a firm’s performance, or sources of firm efficiency (Alvarez and Crespi 2003; Keramidou and Mimis 2011; Keramidou et al. 2013; Wai et al. 2015), while others have looked at the effect of industrial ownership and performance (Wei et al. 2002; Gedajlovic et al. 2005; Nakano and Nguyen 2013; Yu 2013; Ahmed and Hadi 2017). An interest in investigating how firms’ ownership structures and control affect efficiency and performance is rooted in the seminal works of Berle and Means (1932) E. B. Nyannor (B) Zenith Bank (Ghana) Limited, Accra, Ghana e-mail: [email protected] I. O. Mensah Bank of Ghana, Accra, Ghana e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 J. Wood and O. Habimana (eds.), A Multidimensional Economic Assessment of Africa, Frontiers in African Business Research, https://doi.org/10.1007/978-981-15-4510-8_11

217

218

E. B. Nyannor and I. O. Mensah

and Coase (1937). Ownership structure can take various forms—family members, foreign investors, governments, and other groups—who are interested in a firm. Thus, firms may be owned by private individuals and corporations, family members, local or foreign investors, and many other groups, and the effects of the ownership structure on how well firms perform has been of interest. Studies have also examined banks or firms’ efficiencies and performance within the financial services industry. Their focus is on measuring and/or determining efficiency or performance; ownership structures and banks’ performance; diversification and banks’ efficiency; and [de]regulation and banks’ performance (Berger and Humphrey 1997; Doan et al. 2018; Hughes and Moon 1995; Reddy and Subramanyam 2011; Sarkar et al. 1998). The studies consider environmental, legal, and institutional factors while maintaining the traditional view that allows banks to optimize (either minimize their costs or maximize their profits). Different banks may have different efficiency levels and thus different levels of performance. As noted by Hughes and Mester (2008) and Das and Kumbhakar (2012), differences in the legal framework, environment, market conditions, population density, and institutional details across political jurisdictions can lead to differences in banks’ efficiency. Even in the same jurisdiction, increased competition for inputs and output, quality of services, shrinking profit margins, changing regulations, and technology may put pressure on banks’ efficiency and performance (Das and Kumbhakar 2012; Singh and Singh 2015). The ever-changing landscape of the banking industry across the globe may put banks’ performance under pressure, and may lead to mergers and acquisitions, or even failure. Banks contribute to economic development through their intermediation role. By mobilizing deposits and making these funds available in the form of loans to economic agents requiring capital their activities boost output (Allen and Santomero 1998). Thus, efficient banks are important for economic growth and development and their ownership structure may have several implications for their efficiency and performance (Doan et al. 2018). Since banks’ capital make-up is mostly debt structured and ownership is separated from the management, this ideally imposes a self-check and discipline mechanism on banking activities. Nonetheless, the extent to which this ownership structure improves efficiency and performance is still debatable. In the banking industry, ownership structure might constrain the effectiveness of bank regulations (Laeven and Levine 2009).1 There may be differences in the efficiencies of foreign and domestic banks, and though the advantages are usually in favor of foreign banks, empirical evidence on which of the two perform better is mixed (Lensink et al. 2008; Micco et al. 2007).

1 Ownership

does not matter in the Basel regulations.

11 Efficiency of Commercial Banks in Ghana

219

Given this background, this study finds out whether the ownership structure influences the efficiency of commercial banks in Ghana, and how such banks may be susceptible to external shocks. The analysis focuses on Ghanaian banks for various reasons. Existing studies on bank efficiency in Ghana have focused on rural and community banks without looking at commercial banking which accounts for a significant market share (Adusei 2016; Iddrisu 2014; Oteng et al. 2016). Also, a financial crisis or a financial boom that affects the structure and strength of large economies may influence banks’ efficiency in other jurisdictions, especially in small open economies like Ghana (Ackah and Asiamah 2014). The rest of the paper is structured as follows. Section 11.2 provides a brief overview of the banking sector’s performance in Ghana. A review of relevant literature and a discussion of the model, data, and the estimation strategy are discussed in Sects. 11.3 and 11.4. The results are reported and discussed in Sect. 11.5 and Section 11.6 gives the concluding remarks.

11.2 Overview of the Ghanaian Banking Industry The banking industry has undergone rapid growth in Ghana with banks and other deposit taking financial institutions growing in size and complexities (including increasing number of microfinance institutions). By regulation, financial institutions are categorized into commercial banks, rural or community banks, savings and loans companies, finance houses, and deposit-taking microfinance institutions. In the early years of independence, Ghanaian banks were mostly state owned. However, the structure of ownership changed first due to the financial deregulation policies in the 1980s and then under the Financial Sector Structural Adjustment Program (FINSAP) and the Financial Sector Strategic Plan (FINSSP) in the 2000s. Consequently, many private domestic and foreign banks have entered the industry (Bawumia 2010; IMF 2011). It is argued that foreign banks are beneficial for host economies as they bring in technologies that are more advanced, a highly skilled labor force, and better risk management practices which local banks can adopt (Lensink et al. 2008). Some crosscountry studies have suggested that the entry of foreign banks improves economies and boosts competition which leads to improved efficiency in the banking system (Bayraktar and Wang 2005), however empirical evidence supports this in countries with strong institutions (Lensink et al. 2008). The ownership structure, type of bank, and institutional factors in any economy promote and define efficiency and competition in the banking industry (Lensink et al. 2008). Although there are many commercial and rural banks in Ghana, a significant portion of the population is unbanked resulting in the growth of microfinance institutions of which many were weak. The industry has recorded some mergers and acquisitions in the last decade, especially after the 2007–08 global financial crisis. In 2012, Ecobank Transnational Incorporated (Ecobank), the parent company of the Ecobank Group acquired the Trust Bank (TTB) due to recapitalization requirements.

220

E. B. Nyannor and I. O. Mensah

Similarly, Access Bank Plc merged with the stressed Intercontinental Bank, while on the other front the Bank of Africa acquired Amalgamated Bank. Very recent events in the industry relate to the takeover of Capital Bank and UT Bank (private domestic banks) in 2017 by GCB Bank and the withdrawal of the operating licenses of five private domestic banks (Sovereign Bank, Beige Bank, Construction Bank, UniBank, and Royal Bank) in August 2018 to form the The Consolidated Bank by the central bank. These banks were found distressed and could not meet the capital and liquidity requirements of the regulator; there were also other regulatory breaches on their part. It is also suggested that many other banks and/or microfinance institutions facing liquidity and solvency challenges had their license revoked by the central bank and this affected the public’s trust in the banking system. While the number of banks was on the increase, their assets grew steadily with a high cost-income ratio. In the first half of 2017, gross non-performing loans had reached 21.2% from 17.3% in 2016 compared to an industry average of 12.8% in 2013 (Ackah and Asiamah 2014; Bank of Ghana 2016). Non-performing loans in the banking industry have been relatively high compared to countries like Kenya and South Africa (Ackah and Asiamah 2014). Commercial banks accounted for 85.1% and 85.6% of the total industry assets in 2016 and 2015 respectively. Thus, commercial banks control a large market share in the industry and hence their inefficiencies could drive the entire industry down.2 Hence, a continuous review of the capital requirements of the industry by the Bank of Ghana to ensure an adequate capital buffer to mitigate possible external shocks and trading losses is essential. Table 11.1 summarizes the indicators of commercial banks from July 2007 to July 2014.

11.3 Review of Relevant Literature A few studies have investigated banks’ performance in emerging and developing economies. In Asia, many studies have examined banking performance and their efficiency levels (see Sarkar et al. 1998; Kumbhakar and Sarkar 2003; Das and Kumbhakar 2010). Mokhtar et al. (2006), analyzed the efficiency of Islamic banks in Malaysia using the stochastic frontier technique. They showed that average technical and cost efficiencies of the conventional banks were higher than those of the Islamic banking system. Reddy and Subramanyam (2011) investigated the efficiency of 81 commercial banks in India cutting across domestic (public and private) and foreign banks. They concluded that foreign banks were far ahead of domestic banks in terms of achieving risk efficiency. This supports the argument that foreign banks have better risk management practices. Further, they observed that the exogenous and endogenous variables made no difference to the structural risk efficiencies of domestic banks.

2 The

data comes from the audited and published yearly financial statements for banks in Ghana (see Asiamah and Ackah 2014).

11 Efficiency of Commercial Banks in Ghana

221

Table 11.1 Ghanaian banking industry’s indicators Indicators (%)/year

July 07

July 08

July 09

July 10

July 11

July 12

July 13

July 14

Capital adequacy (CAR)

14.5

14.4

14.7

19.3

17.0

15.5

18.6

16.2

TIER 1 CAR

13.3

13.0

13.0

18.5

15.4

13.6

16.8

14.6

Gross yield

15.1

15.9

20.0

21.3

9.6

8.7

10.4

12.7

Int. payable

7.3

6.9

11.0

10.7

3.7

2.8

3.9

4.1

Spread

7.9

9.0

9.0

10.6

5.9

5.8

6.5

8.5

14.5

16.1

20.1

19.5

9.0

8.9

10.1

10.5

Int. margin to total asset

7.0

7.1

7.7

9.3

4.2

4.2

5.3

5.3

Int. margin to total asset

48.3

44.2

38.4

47.7

46.9

46.9

51.3

50.2

Profitability ratio

15.7

14.5

13.8

12.6

14.8

22.0

22.5

23.7

Return on asset

3.7

3.2

3.3

3.2

3.5

4.6

5.5

6.3

Return on equity

24.4

23.0

23.7

19.3

18.1

26.7

27.7

31.4

Asset utilization

Note According to the Bank of Ghana, commercial banks are required to hold a minimum CAR of 10% and a buffer of 13% to absorb unexpected losses Source Bank of Ghana

Altunbas et al. (2001), used the stochastic frontier approach to assess the efficiency of the banking system in Germany for the period 1989–96. They found little evidence to suggest that privately owned banks performed better than their mutual and public counterparts since all the bank ownership types benefited from widespread economies of scale during the period. The efficiency of Australian banks has been studied extensively. For instance, Walker (1998) examined the cost efficiency of 12 Australian banks from 1978 to 1990, and reported an average efficiency level of around 90% suggesting that Australian banks needed to reduce their actual costs by 10% to become cost efficient. Avkiran (1999), Neal (2004), Sathye (2001), and Sturm and Williams (2010) also suggest that banks need to be more efficient. Using data on about 6,000 US commercial banks that existed in the six-year period 1990–95 Berger and Mester (1997) and estimated scale economies, cost for banks in different size categories based on their preferred model that incorporated asset quality, financial capital, and off-balance sheet assets. Aljo and Oginnlyl (2010), analyzed technical and scale efficiencies of commercial banks in Nigeria using the data envelopment analysis (DEA). By sampling 13 banks, they found that 25% of these banks were productively inefficient as a result of

222

E. B. Nyannor and I. O. Mensah

disproportionate use of some inputs despite mergers and acquisitions. Iddrisu et al. (2014), investigated rural banks’ efficiency in Ghana in 2000–10 using DEA. They show that their efficiency improved with the establishment of the ARB Apex Bank ensuring best practices in the industry. Oteng-Abayie et al. (2016), confirmed this in a later study of the rural banking industry in Ghana. Both studies agree that the industry can further improve its efficiency by employing cost efficient techniques for maximizing profits. Despite a plethora of empirical studies on the banking sector’s efficiency, there are very few studies on Ghana. There is no consensus on the best method for measuring efficiency in financial institutions. Hence, this study contributes to the literature on the efficiency of commercial banks in Ghana.

11.4 Methodology 11.4.1 Empirical Estimation Strategy Literature on efficiency assumes that standard profit or cost efficiency estimates how firms that are close to the frontier make the most profits or operate at minimum costs, given a particular level of input prices and the cost of output (fixed inputs and environmental factors). Like the profit function, the standard cost equation shows variable costs instead of variable profits and considers variable output prices as a given, rather than maintaining all output quantities empirically fixed at their known, possibly inefficient, standards (Berger and Mester 1997; Mester 2005). The dependent variable in the cost function permits considering costs that are obtained by changing outputs as well as inputs. Mester (2005), specified the profit efficiency equation in log form as: ln(π + θ )i = ln g( pi , wi , z i , h i ) − u πi + vπi

(11.1)

The cost function can also be specified in a log form as: ln Ci = ln f (yi , wi , z i , h i ) + u i + vi

(11.2)

where π stands for variable profits of the bank and θ represents the constant computed on every firm’s profit.3 This helps the natural log to be taken as a positive number. w shows the vector of prices of variable inputs, h shows the environmental or market variables that affect performance (for example, regulatory restrictions), and ut represents the inefficiency variable that may lower profits above the bestpractice level. The random error that captures the measurement error and chance that may temporarily offer firms high or low profits is vi , while p is a vector of prices 3 The P symbol in Eq. (11.1) helps the equation to stay positive so that the natural log of the equation

can be found easily.

11 Efficiency of Commercial Banks in Ghana

223

of variable outputs. C measures variable costs, while y is the vector of quantities of the variable’s outputs. As in the case of banking, the profit technology features several outputs, where the profit frontier shows the maximum profit that is required to produce any scalar output given exogenous input prices (w). Like profit inefficiency, cost inefficiency is the cost that is added when contrasted with the amount of cost that can be reduced in an efficient bank (best practice bank). In the same vein, growth in productivity profits measures the differences in profits from period t to period t + k, maintaining the exogenous environmental variables (business circumstances) as constant at various periods t. Altunbas et al. (2001), explain that estimating the efficiency of different bank ownership forms poses two main problems. First, one has to decide whether to estimate an industry profit or cost frontier including all ownership types from which inefficiencies can be calculated, or alternatively to estimate frontiers for each ownership type. The former permits assessments of firm ownerships relative to the industry ‘best practice’ profit or cost frontier, whereas the latter solely allows assessments of banks in the same ownership category. In the case of cost functions, Mester (1993) makes the case for estimating a separate cost function for each type of ownership, arguing that it is inappropriate to compare inefficiency scores and ownership categories when firms in one category are using a different technology. While there are different approaches for measuring efficiency, for example, data envelopment analysis and total factor productivity indices (Worthington 1998), we opt for SFA because it appropriately specifies a bank’s Cobb-Douglas production function. SFA also includes the estimation of the statistical noise in the frontier while effortlessly allowing statistical tests of the estimates. In SFA, an assessment of profit or cost efficiency is tied to the relevant frontier. Since the ‘true frontier’ is not observable, an estimate is usually constructed which is known as a ‘best practice’ frontier. SFA is a typical technique for estimating efficiency relative to the estimated ‘best practice’ frontier. This technique assumes that a departure from the constructed frontier can be due to either random fluctuations or inefficiency (Aigner et al. 1977; Battese and Coelli 1995; Meeusen and van den Broeck 1977). Based on these studies, the stochastic frontier can be specified as: yit = αit + f (X it , β) + εit ; εit = vit − u it

(11.3)

where y represents output, X is a vector of inputs for the ith bank at time t, and β represents the vector of the parameters to be estimated, and α i accounts for the unobserved heterogeneity that exists throughout firms in a panel data system. εit , the error term, is a composite error term that is made up of a vit random variable which has an independent, identical, and normal distribution. The term uit is thought to behave as u it = u i e(−λ(t−T )) . Essentially ui are non-negative variables which assess technical inefficiency and λ is an observed efficiency variable to be estimated over time. To measure efficiency of Ghanaian banks, we adopted Battese and Coelli’s (1995), model using the SFA approach. This model uses the estimation of efficiency and the determinants of inefficiency at a chosen time by using parameters that estimate the

224

E. B. Nyannor and I. O. Mensah

mean of the inefficiency term as a function of exogenous variables. The profit and cost functions are given as: p

p

ln Pit − ln f (yit , wit , λ) + vit − u it

(11.4)

p

ln Cit − ln f (yit , wit , Θ) + vit + u itc

(11.5)

In Model (11.4), the subscript i shows the number of banks (i = 1, 2, …, N) and t is a particular year (t = 1, 2, …, T ); Pit measures profit before tax; C it is total costs; yit represents a vector of outputs; wit is a vector of inputs; and P and λ are vectors of the parameters to be assessed. In Eqs. (11.1)–(11.5), one can find a relationship between profit (P) and cost (C) depending on the vectors of the inputs and outputs p used in the bank production process. The inefficiency term uit assumes a non-negative distribution and is shown as a function of the variables that affect efficiency levels: p

p

Uit − f (git ; λ) + Ψit

(11.6)

where git represents a vector of explanatory variables, λ is a set of unknown paramp eters to be estimated, and the random variable, Ψit shows a truncated normal distribution with zero mean and constant variance.

11.4.2 Model Specification To specify the production frontier function, we adopted the general preference in previous studies (Berger and Mester 1997), with respect to appropriate flexible functional forms by using a generalized Cobb–Douglas production function. The study uses the maximum likelihood estimation (MLE) technique for the analysis. Compared to the least square dummy variable (LSDV) and the generalized least squares (GLS) estimators, MLE gives more efficient estimates. The reduced trans log stochastic cost and profit frontiers are given as: ln Cit (wit , yit , z it ; b) = α0 +

K 

βk ln wik + 1/2

k=1

+

M 

M 

.

M 

k=1

δmr ln yim ln yir

m=1 r =1

K J   . K k j ln wik ln wi j k=1

+

l=1

δm ln yim + 1/2

m=1

+

K K   . βkl ln wik ln wil

M 

j=1

·

N 

m=1 n=1

K ξmn ln yim ln yin + vi + u i

(11.7)

11 Efficiency of Commercial Banks in Ghana

ln Pit (wit , yit , z it ; b) = α0 +

K 

225

βk ln wik + 1/2

k=1

+

M 

M 

.

M 

k=1

δmr ln yim ln yir

m=1 r =1

K J   . K k j ln wik ln wi j k=1

+

l=1

δm ln yim + 1/2

m=1

+

K K   . βkl ln wik ln wil

M 

j=1

·

N 

K ξmn ln yim ln yin + vi + u i

(11.8)

m=1 n=1

The outputs are loans and advances (y1 ), number of branches (y2 ), and investments (y3 ). The three inputs that the banks use for producing output are labor, deposits, and capital. The study defines the prices of these inputs as: the price of labor or wages (w1 ), the percentage share of each bank’s deposits (w2 ), and the real cost of the stated capital (w3 ). One factor that plays a crucial role in measuring efficiency is the role of financial capital (Berger and Mester 1997; Maudos et al. 2002). A bank’s insolvency risk lies in its asset portfolio and the financial capital available for absorbing any losses and defaults. Banks characterized by a higher aversion to risk will demand a level of financial capital that is higher than the optimum (that is, that which reduces costs or maximizes profits). If capital is not included, banks that are most prudent or most risk-averse will be penalized, even if they behave optimally in terms of their preferences regarding risks (Maudos et al. 2002). Financial capital or equity, w3 , is incorporated in the cost and profit frontier functions to capture the effects of risk propensity on efficiency in the model. Other inputs that can feature in an efficiency model are the amount of deposits per branch. The idea behind the number of branches per bank as an output is to proxy “the quality and convenience that a bank offers its customers” as explained by Kumbhakar and Sarkar (2003, p. 410). The same idea used for the construction of other inputs shows the extent of branch penetration in the collection of deposits and advances per branch. It proxies the extent of branch penetration in Ghana and in advancing loans and credit. There are some restrictions on Eqs. 11.1 and 11.2 (that is, Eqs. 11.7 and 11.8): (1) For constraints on symmetry, βkl = βlk and δmr = δr m . (2) For homogeneity in prices, kk = βk = 1. k k k (3) k βkk = l βkl = j K k j = 0. To assess the differences in efficiencies for different ownership structures over the study period, we use time and ownership dummies (state-owned or public, private foreign, and private domestic).

226

E. B. Nyannor and I. O. Mensah

11.4.3 Data and Variables Data for the study comes from annual financial statements of 26 commercial banks and gazettes of the Bank of Ghana. Of these 26 banks, 15 are foreign owned, 8 domestic private, and 3 state-owned banks. The data covers the period 2001–10. All monetary values are deflated to 2002 prices using the consumer price index (CPI). The independent variables in the model are time trend, advances (loans), investments, labor, deposits, capital, dummy for ownership type, and the financial crisis, while the dependent variable is operating costs. Some of the variables are also interacted. The selection of the variables was based on literature.

11.5 Results and Discussion Tables 11.2 and 11.3 present a summary of the efficiency scores of the studied banks for the period 2001–10 and by ownership structure. Basically, a distinction is made between pre-financial crisis years and the years of the financial crisis. Table 11.2 Table 11.2 Mean efficiency of banks during 2001–10 Year

Mean

Std. dev.

Minimum

Maximum

2001

0.9805

0.0085

0.9794

0.98175

2002

0.9803

0.0094

0.9796

0.98991

2003

0.9813

0.0023

0.9808

0.98193

2004

0.9886

0.0004

0.9805

0.98206

2005

0.9709

0.0004

0.9802

0.98195

2006

0.9708

0.0003

0.9801

0.98133

2007

0.9807

0.0004

0.9800

0.98124

2008

0.8953

0.1757

0.4421

0.99798

2009

0.9678

0.9005

0.9601

0.97557

2010

0.8732

0.8432

0.8111

0.85645

Note Values closer to unity are considered more efficient Source Authors estimation

Table 11.3 Mean efficiency of bank ownership (with capital) Ownership type

Mean

Std. dev.

Minimum

Maximum

Foreign

0.9693

0.0007

0.9681

0.9717

Domestic private

0.9708

0.0009

0.9687

0.9729

State-owned

0.9736

0.0009

0.9714

0.9752

Note Values closer to unity are considered more efficient Source Authors estimation

11 Efficiency of Commercial Banks in Ghana

227

shows that in the pre-crisis period (2001–07), the average efficiency stood at 98.0%. At the end of 2010, the average efficiency had declined to about 87%. In effect, bank efficiency was affected by the financial crisis. One reason for this could possibly be exposures to counter parties in overseas banks in the form of placement and correspondent balances which increased from 48.12% and 55.46% of the net-worth of all banks in 2007 and 2008 respectively (Gockel 2010). Similarly, the efficiency scores by ownership structure show that all banks were efficient (Table 11.3). The study analyzed two models: first with capital and the second without capital in the regressions. The purpose was capturing the effect of banks’ risk aversion in the model. The results are given in Tables 11.4 and 11.5. In these tables, the study identifies cost inefficiency variables as P2u and the idiosyncratic error term as P2v (technical inefficiency). In Table 11.4, labor, capital, and deposits inputs have positive coefficients. The coefficients are 0.8146, 0.1863, and 0.8832 respectively and are statistically significant at the 1% level. The positive coefficient of labor means that as more payments, in the form of salaries and employee benefits, were made to staff members and the management, the banks’ operating costs increased. The sign of the coefficient capital is mixed in literature as it is an applied matter of either using too much or too little capital. Capital is regarded as a buffer for covering losses, thus possessing too little capital may mean a risk of default. At the same time, possessing too much capital may impose higher costs for banks and lead to poor returns on equity. Hence, the exact amount of capital (optimal) that a bank needs is hard to define. In Table 11.4, the coefficient of capital suggests that banks’ operating costs increased with an increase in capital. In Table 11.5, banks’ capital was omitted from the regression; only advances and deposits were significant at the 1% and 10% error levels respectively in this regression. However, they had a negative impact on the banks’ operating costs which is contrary to the results in Table 11.4. This may be due to exclusion of capital. The interactions of the variables in the first model were all significant with the exception of the interaction of advances and deposits coefficients of −0.1107. The square of labor was found to significantly reduce banks’ costs if labor and other inputs performed efficiently in the production process. The interaction of advances with investments produced a significant coefficient of −0.0219 at the 5% level. Advances squared remained highly significant at 1% as well as the square of capital emphasizing the importance of capital in the banks’ production processes. Similarly, the interaction of labor and capital was also statistically significant at 1%. The coefficient of the 2007–08 financial crisis is −1.8 which is significant at the 5% level (Table 11.4). In Table 11.5, the estimated coefficient of labor is similar to that in Table 11.4, however, the coefficients of deposits and advances are negative and significant implying that they added to banks’ operating costs. The dummy for the financial crisis period (2007–08) is positive and significant at the 1% error level. Therefore, the impact of the financial crisis on the efficiency of banks in Ghana was inconclusive when we considered the signs in the two models with and without capital. A possible reason

228 Table 11.4 Model with unrestricted effects of capital

E. B. Nyannor and I. O. Mensah Variable

Coefficient

Std. error

P-value

Constant

−0.0059

0.0609

0.923

LnAdvance

1.3160

0.7413

0.002

LnInvestment

0.7767

0.8712

0.001

LnLabour

0.8146

0.7037

0.002

LnDeposits

0.8832

1.0789

0.003

LnCapital

0.1863

0.2101

0.001

LnAdvance * LnAdvance

0.2803

0.1105

0.007

LnAdvance * LnInvestment

0.0219

0.0119

0.043

LnCapital * LnCapital

0.0033

0.0129

0.002

LnInvestment * LnInvestment

0.0172

0.0500

0.000

LnAdvance * LnDeposits

−0.11073

0.0999

0.268

LnDeposits * LnInvestment

0.2060

0.1072

0.000

LnLabour * LnLabour

−0.1052

0.0886

0.003

LnLabour * LnCapital

0.0399

0.0451

0.001

P2v (Technical inefficiency/idiosyncratic errors) State owned Private domestic

0.6816

237.4299

0.998

0.4492

237.4047

0.998

Foreign

−0.0760

237.508

1.000

Year

−0.0002

0.1185

0.998

Fin. crisis (08–10)

−1.8071

0.8463

0.033

State owned

−5.1272

1446.414

0.004

Private domestic

−5.1299

1438.446

0.000

Foreign

−5.1171

1441.172

0.004

P2U (Cost inefficiency)

Year

−0.0237

0.7186

0.003

Fin. crisis (08–10)

−0.0032

0.523134

0.001

Log likelihood

−141.9928

Observations

154

Note Dependent variable: log of operating cost Source Authors estimation

11 Efficiency of Commercial Banks in Ghana Table 11.5 Model without capital

229

Variable

Coefficient

Constant

−0.0472

0.0515

0.359

LnAdvance

−1.0236

0.2648

0.000

LnInvestment

0.6985

0.8040

0.385

LnLabour

0.6730

0.4262

0.114

−1.6109

0.9547

0.092

LnAdvance * LnAdvance

0.2651

0.0469

0.000

LnAdvance * LnInvestment

−0.0272

0.0078

0.001

LnInvestment * LnInvestment

−0.0025

0.0431

0.952

LnAdvance * LnDeposits

−0.0213

0.0471

0.650

0.0631

0.002

0.0465

0.144

LnDeposits

LnDeposits * LnInvestment LnLabour * LnLabour

0.19362 −0.0679

Std. error

P-value

P2v (Technical inefficiency/idiosyncratic errors) State owned

0.3952

210.9262

0.999

Private domestic

0.2066

210.9233

0.999

Foreign

0.5172

211.0052

0.998

Year

0.0004

0.1053

0.997

4.1816267

0.002

Fin. crisis (08–10)

13.2285

P2U (Cost inefficiency) State owned

−5.4743

2054.9429

0.998

Private domestic

−4.9817

2052.9198

0.998

Foreign

−4.9684

2052.9597

0.998

Year

−0.0020

1.0223

0.998

0.3453

0.001

Fin. crisis

7.0828***

Log likelihood

−132.19909

Observations

155

Note Dependent variable: log of operating cost Source Authors estimation *** Indicates significance at 1% level

for this is that Ghana’s financial system has not been fully integrated with the global financial system. In Table 11.4, state-owned and private domestic banks have positive coefficients on technical inefficiency, while the coefficients of foreign banks and time trend are

230

E. B. Nyannor and I. O. Mensah

negative. However, these results are insignificant. The coefficient of the financial crisis dummy is negative and significant at 5% suggesting that the crisis led to declining technical inefficiency. In relation to cost inefficiency, the time variable also indicates falling cost inefficiencies. The time variable’s falling cost inefficiencies are an indication of the improved efficiency in the banking system. Even though cost inefficiencies in the banking industry have declined over time (Table 11.4), there is room for improvement on the efficiency scale (Oteng-Abayie et al. 2016). The results also show that banks with all ownership types performed better in cost efficiency, as they show statistically significant negative coefficients. This improvement could be due to proper regulations and higher competition.

11.6 Concluding Remarks This chapter examined the efficiency of commercial banks in Ghana with particular emphasis on ownership structure and the financial crisis. Using data from 26 commercial banks for the period 2001–10, we estimated a production frontier function using the MLE technique. Our results do not support the hypothesis that banks’ efficiency differed by ownership structure. Thus, we did not find support for the idea that foreign banks are more cost efficient than other bank ownership types. We also found that banks’ average efficiency scores declined after the 2007–08 global financial crisis. The MLE estimation (preferred model, Table 11.4) also showed that cost inefficiency was negatively related to the time trend and ownership structure (all banks). Contrary to the descriptive statistics, the financial crisis was negatively related to inefficiency. Except the financial crisis variable, the effects of all other independent variables on technical inefficiency were insignificant. The study concludes that the effect of the financial crisis on efficiency is inconclusive. Capital as an input is an important factor which signifies the risk component of the bank production process. Hence, policies should proactively assess the optimal (economic) capital framework for commercial banks in accordance with the Basel capital framework to ensure an efficient banking sector in Ghana. Acknowledgments The authors would like to thank Mustapha Immurana, Micheal Kofi Boachie, and Ernest Ngeh Tingum for their valuable comments on earlier versions of this paper. The views expressed are the authors and do not reflect the views of and should not be attributed to affiliated institutions or third parties.

References Ackah, C. and J. P. Asiamah (2014). Financial regulation in low-income countries: Balancing inclusive growth with financial stability. ODI Working paper 410. Adusei, M. (2016). Determinants of bank technical efficiency: Evidence from rural and community banks in Ghana, Cogent Business & Management, Taylor & Francis Journals, 3(1): 1199519-119

11 Efficiency of Commercial Banks in Ghana

231

Ahmed, N. and O. A. Hadi (2017). Impact of Ownership Structure on Firm Performance in the MENA Region: An Empirical Study. Accounting and Finance Research, 6(3): 105. Aigner, D. C. A. L. Lovell, and P. Schmidt (1977). Formulation and estimation of stochastic frontier production function models. Journal of Econometrics, 6(1): 21–37. Ajao, A. and L. Ogunniyi (2010). Technical and scale efficiency of commercial banks in Nigeria. IUP Journal of Applied Economics, 9(4): 59–68. Allen, F. and A. M. Santomero (1998). The theory of financial intermediation. Journal of Banking & Finance, 21: 1461–1485. Altunbas, Y. E. Lynne, and P. Molyneux (2001). Bank ownership and efficiency. Journal of Money, Credit and Banking: 926–954. Alvarez, R. and G. Crespi (2003). Determinants of technical efficiency in small firms. Small Business Economics, 20(3): 233–244. Avkiran, N. K. (1999). An application reference for data envelopment analysis in branch banking: helping the novice researcher. International Journal of Bank Marketing, 17(5): 206–220. Bank of Ghana (2016). Annual Report 2016. Retrieved from: https://www.bog.gov.gh/ privatecontent/Publications/Annual_Reports/Annual%20Report%202016%20final%2024th% 20June.pdf (accessed: 15 November 2018). Battese, G. E. and J. T. Coelli (1995). A model for technical inefficiency effects in a stochastic frontier production function for panel data. Empirical economics, 20(2): 325–332. Bawumia, M. (2010). Monetary Policy and Financial Sector Reform in Africa, Ghana’s Experience. Combert Impressions Ghana Ltd. Bayraktar, N. and Y. Wang (2005). Foreign bank entry and domestic banks’ performance evidence using bank level data. Pennsylvania State University and World Bank. Berger, A. N. and D. B. Humphrey (1997). Efficiency of financial institutions: International survey and directions for future research. European journal of operational research, 98(2): 175–212. Berger, A. N. and L. J. Mester (1997). Inside the black box: What explains differences in the efficiencies of financial institutions? Journal of banking & finance, 21(7): 895–947. Berle, A. A. and G. C. Means (1932). The Modern Corporation and Private Property (revised edition, 1967). New York: Harcourt Brace and World, Inc. Coase, R. H. (1937). The nature of the firm. Economica 4, (16): 386–405. Das, A. and S. C. Kumbhakar (2010). Productivity and efficiency dynamics in indian banking: an input distance function approach incorporating quality of inputs and outputs. Journal of Applied Econometrics. Doan, A. T., K. L. Lin, and S. C. Doong (2018). What drives bank efficiency? The interaction of bank income diversification and ownership. International Review of Economics & Finance, 55: 203–219. Gedajlovic, E., T. Yoshikawa, and M. Hashimoto (2005). Ownership structure, investment behaviour and firm performance in Japanese manufacturing industries. Organization Studies, 26(1): 7–35. Gockel, F. A. (2010). The world financial crisis and its implications for Ghana’ Friedrich Ebert Stiftung. Parliamentary briefing paper. Retrieved from: http://library.fes.de/pdf-files/bueros/ ghana/10494.pdf. Hughes, J. J. and C. G. Moon (1995). Measuring bank efficiency when managers trade return for reduced risk. Rutgers University, Department of Economics. Hughes, J. S. and L. J. Mester (2008). Efficiency in Banking: Theory, Practice, and Evidence. Working Paper ID No. 08-01. Retrieved from: http://apps.olin.wustl.edu/FIRS/pdf/MemberPapers/61/ 19.pdf. Iddrisu, A., A. Q. Q. Aboagye, and K. Osei (2014). Measuring rural bank efficiency in Ghana: An application of the data envelopment analysis (DEA) approach. Journal of Human and Social Science Research, 3: 41–59. International Monetary Fund (IMF) (2011). Ghana: Financial System Stability Assessment Update. IMF Country Report, No. 11/131. Retrieved from: https://www.imf.org/en/News/Articles/2015/ 09/28/04/53/socar052208a.

232

E. B. Nyannor and I. O. Mensah

Keramidou, I., A. Mimis, A. Fotinopoulou and C. D. Tassis, 2013. Exploring the relationship between efficiency and profitability. Benchmarking: An International Journal, 20(5): 647–660. Keramidou, I. and A. Mimis (2011). An application of the double-bootstrap data envelopment analysis to investigate sources of efficiency in the Greek poultry sector. World’s Poultry Science Journal, 67(4): 675–686. Kumbhakar, S. C. and S. Sarkar (2003). Deregulation, ownership, and productivity growth in the banking industry: evidence from India. Journal of Money, Credit and Banking: 403–424. Laeven, L. and R. Levine (2009). Bank governance, regulation and risk taking. Journal of Financial Economics, 93: 259–275. Lensink, R., A. Meesters, and I. Naaborg (2008). Bank efficiency and foreign ownership: Do good institutions matter? Journal of Banking and Finance, 32: 834–844. Maudos, J., F. Perez, J. Quesada, and J. M. P. Monsálvez (2002). Cost and profit efficiency in European banks. Journal of International Financial Markets, Institutions and Money, 12(1): 33–58. Meeusen, W. and J. V. Broeck (1977). Efficiency estimation from Cobb-Douglas production functions with composed error. International economic review, 18(2): 435–444. Mester, L. J. (1993). Efficiency in the savings and loan industry. Journal of Banking & Finance, 17(2): 267–286. Mester, L. J. (2005). Efficient production of financial services: scale and scope economies. Business Review: 15–25. Micco, A., U. Panizza, and M. Yaez (2007). Bank ownership and performance: Does politicsmatter? Journal of Banking and Finance, 31: 219–241. Mokhtar, H. S., N. Abdullah, and M. S. Al-Habshi (2006). Efficiency of Islamic banking in Malaysia: A stochastic frontier approach. Journal of Economic Cooperation: 37–70. Nakano, M. and P. Nguyen (2013). Foreign ownership and firm performance: evidence from Japan’s electronics industry. Applied Financial Economics, 23(1): 41–50. Neal, P. (2004). X-Efficiency and Productivity Change in Australian Banking. Australian Economic Papers, 43(2): 174–191. Oteng-Abayie, E. F., B. Owusu-Ansah, and K. Amanor (2016). Technical efficiency of credit unions in Ghana. Journal of Finance and Economics, 4(3): 88–96. Reddy, C. S. and T. Subramanyam (2011). Data Envelopment Analysis Models to Measure Risk Efficiency: Indian Commercial Banks. IUP Journal of Applied Economics, 10(4): 40–69 Sarkar, J., S. Sarkar, and S. K. Bhaumik (1998). Does ownership always matter? Evidence from the Indian banking industry. Journal of Comparative Economics, 26(2): 262–281. Sathye, M. (2001). X-Efficiency in Australian Banking: An Empirical Investigation. Journal of Banking and Finance, 25(3): 613–630. Singh, P. and K. Singh (2015). Efficiency assessment parameters of public sector banks in India. Global Business Review, 16(6): 1112–1126. Sturm, J. B. W (2010). What determines differences in foreign bank efficiency? Australian evidence. Journal of International Financial Markets, Institutions and Money, 20(3): 284–309. Walker, G. (1998). Economies of Scale in Australian Banks, 1978–1990. Australian Economic Papers, 37(1): 71–87 Wai, P. W., K. L. Soh, C. L. Chong, and N. Karia (2015). Logistics firms performance: efficiency and effectiveness perspectives. International Journal of Productivity and Performance Management, 64(5): 686–701. Retrieved from: https://doi.org/10.1108/IJPPM-12-2013-0205. Wei, Z., O. Varela, and M. K. Hassan (2002). Ownership and performance in Chinese manufacturing industry. Journal of Multinational Financial Management, 12(1): 61–78. Worthington, A. (1998). The Determinants of Non-Bank Financial Institution Efficiency: A Stochastic Cost Frontier Approach. Applied Financial Economics, 8: 279–289. Yu, M. (2013). State ownership and firm performance: Empirical evidence from Chinese listed companies. China Journal of Accounting Research, 6(2): 75–87.

Chapter 12

Cooperative Societies as a Distribution Channel for Insurance Services in Kenya: A Situational Analysis Robert Kuloba, Esther Gicheru, and Silas Maiyo

Abstract Cooperative enterprises the world over have had tremendous social, economic, and environmental impact. In Kenya, cooperatives operate in different sectors like agriculture, finance, housig, insurance, fisheries, transport, and arts and culture. Their contribution to GDP is estimated at 45% (Chambo 2008) while they contribute 31% to national savings and deposits. The cooperative movement is therefore a key aggregator and accelerator for growth of the financial services sector in which insurance is a key player. Considering the role that the cooperatives play in national development, this study analyzes how cooperative societies can be used as distribution channels for insurance services in Kenya. The study uses a mixed method approach for collecting data through observations, personal interviews, focus group discussions, and an analysis of documents and case studies for collecting relevant qualitative and quantitative data as per its objectives. The study was carried out in five purposively and conveniently selected counties. The findings show that awareness and knowledge about insurance in the cooperatives remains low. Cooperatives have a strong capital base, effective governance structures, competent managerial and technical staff, are up to date with ICT, and are close to their members. Cooperatives in Kenya have strong regulatory frameworks that can conveniently accommodate the insurance distribution business. The study recommends that the insurance sector should actively engage with the cooperatives in the distribution of insurance services. The SACCO assurance model is an aggregator that is key in the distribution value chain for insurance services in Kenya. There is also a need to develop SACCO assurance regulations to ensure that the cooperatives play their role as aggregators in distribution of insurance services as intermediaries to ensure compliance with provisions of the insurance act. R. Kuloba (B) Insurance Regulatory Authority, Nairobi, Kenya e-mail: [email protected] E. Gicheru · S. Maiyo Cooperative University of Kenya, Nairobi, Kenya e-mail: [email protected] S. Maiyo e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 J. Wood and O. Habimana (eds.), A Multidimensional Economic Assessment of Africa, Frontiers in African Business Research, https://doi.org/10.1007/978-981-15-4510-8_12

233

234

R. Kuloba et al.

Keywords Distribution · Aggregator · Cooperatives · SACCO assurance

12.1 Introduction Cooperative enterprises have had a tremendous social, economic, and environmental impact across the world. In Kenya, cooperatives operate in different sectors including agriculture, finance, housing, insurance, fisheries, transport, and arts and culture. They contribute 45% to the national GDP (Chambo 2008) and 31% to national savings and deposits. The growth of the Kenyan cooperative movement provides an opportunity for further business integration and diversification, and therefore an opening for investing in the insurance sector. An Agency Task Force report recommended that a new agency framework be adopted, and non-traditional players be allowed to sell insurance to help reduce transaction costs and improve service delivery to consumers (IRA working paper, 2011. Among the institutions that the task force said should be considered for licensing as insurance agents in the new agency framework were microfinance institutions, travel agents, estate agents, SACCOs, supermarkets/retail chains, investment clubs/chamas, shops, hire purchase firms, trade unions, welfare societies, trade associations, opinion leaders, professional societies, professionals, and other viable entities. Enhancing access to insurance for most of the Kenyan public remains one of the key policies and strategy issues, especially because of the growing uptake of insurance services among the low-income groups, women, youth, and persons living with disabilities (PLWD). Insurance penetration in Kenya remains low at 2.8% compared to an average of 3.8% for Africa (IRA, 2018). Various studies have shown that insurance and the SACCO sub-sectors have high potential for growth given the available market opportunities that can be explored in both the sub-sectors (Business Sweden 2016). Insurance is basically an intermediary driven business with over 90% of the business being done through brokers and agents. Over time, these conventional channels of distributing insurance have not only proved insufficient in enhancing access to insurance for the low-income groups and at the grassroots level but are also reaching saturation. Considering the level of the development impact, national penetration, and wide network of the cooperatives in Kenya, this study tries to understand cooperatives regulatory frameworks and structures as an alternative distribution channel for insurance services. The purpose of this study is assessing how cooperative societies can be used as a distribution channel for insurance services, in particular: Assessing the level of insurance awareness among cooperatives’ members and managements; Reviewing the operating structures of cooperative organizations in line with the purpose of the study; Examining the characteristics of the cooperatives and their appropriateness for insurance distribution;

12 Cooperative Societies as a Distribution Channel …

235

Reviewing existing regulatory frameworks for cooperatives as enablers for the distribution of insurance; and, Recommending an appropriate regulatory model for use by cooperatives in the distribution of insurance. The study targeted both rural and urban based cooperative enterprises and more specifically financial cooperatives (SACCOs), agricultural marketing cooperatives, consumer cooperatives, dairy cooperatives, and coffee marketing cooperatives.

12.2 Review of Related Literature Economic theory recognizes two kinds of common risks: economic risks arising from direct participation in economic activities and extra-economic risks arising from the participation, involvement, or occurrence of extraordinary events outside the economic activity’s arena, but which may also adversely affect an economic activity and its outcomes. This explanation is helpful in identifying, classifying, and ascertaining the nature of risks and their occurrence that can ultimately be secured through insurance products and services. From the perspective of behavioral economics and insurance, Unfeather and Pauly, argued that considerable empirical evidence shows that many consumers fail to take advantage of insurance protection against losses to property and health, and do not invest in efficient loss reduction measures in low probability-high consequence (LPHC) events. They have a hard time collecting and processing information to determine the likelihood and consequences of risks for which they have little or no experience. Consequently, they fail to behave in ways that would benefit them personally and also enhance social welfare. Behavioral economics explains these aspects and suggests some solutions. However, designing and implementing solutions requires interventions by both public and private institutions. This reasoning justifies the need for active partnership initiatives between IRA, the insurance industry, cooperatives, and the public in promoting awareness about potential risks and insurance solutions available for the common good of society. Globally, cooperatives have remained significant in the role that they play as drivers of the economy. Their global reputation has seen the establishment of the International Cooperative and Mutual Insurance Federation (ICMIF) as a global representative body for the cooperative/mutual insurance sector. ICMIF has an important role to play in measuring, building, and protecting the sector’s reputation across the world (Global Reputation Report, 2016). It also aims to support, strengthen, and grow the cooperative and mutual insurance sector, and to ensure that insurance protection is accessible through its unique business models to all socioeconomic groups across the world. According to the Global Reputation Report (2016), there was an increase of 11% in global visibility for the cooperative/mutual insurance sector, 4.8% increase in the sector’s association with global influencers, and a 184% increase in references to cooperative and mutual values.

236

R. Kuloba et al.

An empirical study’s findings showed that the African insurance market was valued at US dollars 69 billion in 2016 (African Insurance Organization 2016). Accordingly, Africa’s economic boom and subsequent growth of its insurance market, which is well above the global average, were the region’s most significant strengths according to most of the executives surveyed. The market’s robustness has improved markedly, partly not only due to implementation of a more responsive regulatory regime – risk-based supervision regulations, but it has also benefitted from enhanced distribution of insurance products driven by the growing success of bancassurance and mobile phone distribution. According to the African Insurance Organization (2016) study, Africa’s largest opportunities lie in low insurance penetration, which in some countries is less than 1% of gross domestic product (GDP). Gatuguta et al. (2014) note that there has been tremendous growth in cooperative financial organization and it has become a giant financial power house which has surpassed the normal commercial banks and other financial institutions; the Cooperative Bank of Kenya is the 3rd largest commercial bank in Kenya, while the Cooperative Insurance Company of Kenya (CIC) is the 2nd largest insurance entity in Kenya and the only one of its kind in Africa. There are two major aspects of insurance distribution mainly channel distribution and physical distribution. According to Kortler and Armstrong, physical distribution involves planning, implementing, and controlling the physical flow of materials, services, final goods or related information from the point of origin till the point of consumption to meet customers’ requirements at a profit. Distribution of insurance can be categories in three main strategies as follows (a) selective; (b). exclusive; and, (c). intensive: A selective distribution strategy involves fewer channels but focusses on the economies of scale by leveraging on large numbers. Manufacturers position their brands as high quality and so have reason to pursue a highly selective distribution policy; An exclusive distribution strategy approach is the opposite of the intensive distribution strategy as the focus is on niche products using limited channels. This strategy works for brands for fashion related products for purposes of uniqueness. The main aim of this distribution strategy is maximizing margins as opposed to volumes; In an exclusive distribution strategy, organizations limit the number of retailers or intermediaries to one in a given market area. This helps them in maintaining control and the partnership. Please rephrase. One of the ways in which insurance penetration and its uptake is to rethink the distribution model of insurance through intermediaries, especially agents.

12.3 Conceptual Framework The conceptual framework (Fig. 12.1) shows the factors that need to be critically assessed to determine if cooperatives can be effective channels of insurance distribution in Kenya. The conceptual framework shows the relationships–that exist

12 Cooperative Societies as a Distribution Channel … Independent Variables

Intervening Variable

Knowledge of Insurance Business Structures Cooperatives

237

Government Policy Cooperative

Dependent Variable

on

of

Regulatory Frameworks of Cooperatives

Cooperatives as effective Channels of Insurance distribution in Kenya

Characteristics of SACCOs / other types of cooperatives Appropriate Regulatory Model for Coop

Fig. 12.1 Conceptual framework (Colour figure online)

between the independent and dependent variables of the study. Between them is the intervening variable that is likely to influence the relationship positively or negatively. The depicted scenario is such that a cooperative society’s knowledge of the insurance business determines the extent to which it will function as an effective channel for insurance distribution. Further, established structures of cooperative organizations influence the flexibility of cooperatives as effective channels of insurance distribution. Cooperatives’ regulatory frameworks might hinder or support cooperatives as effective channels of insurance distribution. Also, since SACCOs and insurance institutions share a common characteristic of being financial institutions, unlike other forms of cooperatives they have a greater potential to function as effective channels of insurance distribution. Finally, the development of an appropriate business model for cooperatives to participate in the insurance distribution business enhances their effectiveness as channels of insurance distribution. However, government policy on cooperatives’ development with regard to the cooperative business’ innovations and diversification is likely to encourage or hinder the strategic positioning of cooperative organizations to take advantage of emerging opportunities in the insurance industry.

12.3.1 Research Methodology This study adopted a mixed-methods approach that was both quantitative and qualitative using a descriptive survey and case study design. Both primary and secondary

238

R. Kuloba et al.

Gender of respondents Missing 1% Female 39%

Male Male 60%

Female Missing

Fig. 12.2 Distribution of members of cooperatives by gender (Colour figure online)

data was collected and analyzed. Other methods used included observations, interviews, and analyzing documents and case studies to collect relevant data in line with the study’s objectives. The study purposively targeted 25 cooperatives in selected counties in Kenya. The target respondents constituted: board/management committee members, managers/chief executive officers, and general members of cooperatives. The data was coded, entered, and analyzed using SPSS and presented as graphs, tables, and photographs.

12.4 Analysis, Discussion, and Findings 12.4.1 Demographic Data Out of the 1,006 members interviewed, 60% (602 members) were male and 39% (390 members) were female. The remaining 1% (14 members) did not respond (Fig. 12.2). This suggests that the cooperative movement in Kenya is dominated by men.

12.4.2 Age of Cooperatives’ Members The cooperatives’ members were categorized as per age brackets. The highest number, 46.3% (466 members), were in the age bracket of 36–50 years, followed by 34.1% (343 members) in the age bracket of 18–35 years.

12.4.3 Awareness About Rules and by-Laws Governing Cooperative Societies Figure 12.3 shows whether cooperatives’ members were aware of the rules and by-laws governing their cooperative societies.

12 Cooperative Societies as a Distribution Channel …

239

Aware of rules and by -laws governing your cooperaƟve society No 16%

Missing 1% Yes No Missing Yes 83%

Fig. 12.3 Are you aware of rules and by-laws governing the cooperatives? (Colour figure online)

Most of the members of the cooperatives (83%) were aware of the rules and bylaws governing their cooperative societies, 16% of the members were not aware of the rules and by-laws governing their cooperatives while the remaining 1% did not respond.

12.4.4 Knowledge About the Listed Policies of Cooperative Societies Figure 12.4 shows members’ knowledge about cooperatives’ policies. This study established that there were various policies for cooperative societies to govern and operationalize their cooperative business processes including (i) credit (ii) education and training; (iii) insurance rules; (iv) human resources; (v) rules about dividends; (vi) rules about bonuses; and, (vii) election rules.

Frequency

Aware of the listed Policies of the Society? 800 700 600 500 400 300 200 100 0

704

684

676

605 390

318

289

12 Credit Rules

11 Educ. & Training

665

603

532 443

372

329

280 33

22

31

Insurance rules

Human Resources rules

Dividends rules

31 Bonuses rules

12 ElecƟon rules

Response Yes

No

Missing

Fig. 12.4 Know about the listed rules of cooperative societies (Colour figure online)

240

R. Kuloba et al.

Most of the members (67% or 676 members), 66% (665 members), 60% (605 members), and 53% (532 members) said that they were aware of credit policies, election policies, education and training policies, and dividend policies of their cooperatives respectively. The study also found that most of the members (70% or 704 members), 68% (684 members), and 60% (603 members) were not aware of human resource policies, insurance policies, and policies about bonuses respectively of their cooperatives. As per our analysis, members of the cooperatives knew about the policies that directly affected their business interests such as credit, election, education and training, and dividend policies. Insurance may not have a direct effect on members’ interests so they were not knowledgeable about their cooperatives’ insurance policies.

12.4.5 Services or Products Received by Members from Cooperative Societies Figure 12.5 gives the different products and services that members of the cooperatives received from their cooperative societies. The study established the following products and services offered to cooperatives’ members by their cooperatives: (i) savings; (ii) loans; (iii) insurance; (iv) salary processing; (v) education and training; (vi) Marketing; (vii) Advances; and, (viii) payments Most of the members (73% or 734 members), 68% (682 members), and 53.3% (536 members) of the cooperatives indicated that they received loans, savings and education and training services/products respectively from their cooperatives. While 56% (567 members), 52% (524 members), 51% (517 members), 50% (507 members), and 48% (481 members) said that produce marketing, salary processing, produce payment, produce advances, and insurance services/products were not

Frequency

Services or Products that the members recieved from their CooperaƟve SocieƟes 800 700 600 500 400 300 200 100 0

734

682

481 295

536 458

524 451

567 423

480507

517 475

248 144 135

29 Savings

31

24 Loans

Insurance

Salary processing

12 Educ. & Training

16 Produce markeƟng

19 Produce advance

13 Produce payments

Services or Products Yes

No

Missing

Fig. 12.5 Services or Products that the members received from their cooperative societies (Colour figure online)

12 Cooperative Societies as a Distribution Channel …

241

received from the cooperatives; 13.4% (135 members) did not respond. This implies that many members did not know about insurance services. The members did not list any other services/products that they received from their societies.

12.4.6 Insurance Services Figure 12.6 shows awareness about insurance and the level of awareness about insurance services/products in the cooperative movement. Most of the members (82%) of the cooperatives had heard about insurance other than NHIF, 17% had not heard about insurance other than NHIF, and 1% did not respond. It was found that 70.2% of the members of the cooperatives had heard about insurance and knew what it meant, 16.4% had heard about insurance but they did not know what it meant, 8.8% had never heard about insurance before, and 4.6% of the members did not respond. This means that most of the members of the cooperatives were knowledgeable about insurance in general. In general, most of the members (72%) were aware of the insurance products, 24% were not aware of the insurance products, and 4% did not respond (Fig. 12.7). This means that most of the cooperatives’ members were aware of insurance products. Most of the members (60% or 600 members), expressed a positive interest in buying an insurance product, while 36% (368 members) were not interested in buying any insurance products (Fig. 12.8). The remaining 4% did not respond. This means that many members of the cooperatives were interested in buying insurance products if an opportunity existed. Ever heard about Insurance other than NHIF? No 17%

Missing 1% Yes No Missing Yes 82%

Fig. 12.6 Ever heard about Insurance other than NHIF? (Colour figure online)

242

R. Kuloba et al.

Are you aware of any Insurance Products? Missing 4% No 24%

Yes No Missing

Yes 72% Fig. 12.7 Are you aware of any Insurance products? (Colour figure online) Would you be interested in buying any Insurance products? Missing 4% Not Interested 36%

Interested Not Interested Interested 60%

Missing

Fig. 12.8 Interested in buying any insurance products (Colour figure online)

12.4.7 Reasons for Being Interested in Buying/not Buying Insurance Products by Cooperatives’ Members Members of the cooperatives provided different reasons for their interest in buying or not buying insurance products (see Box 1). Box 1: Reasons for being interested or not interested in buying insurance products

12 Cooperative Societies as a Distribution Channel …

243

Reasons for being Interested in buying

Reasons for not being Interested in buying

It will be of great benefit to me, my family, and children I am already insured through my SACCO I am satisfied with the current insurance policy Because I receive information that is relevant for the policy that I have, I can buy more from my SACCO Because insurance is good and payments of claims is done effectively, for example, NHIF Because getting back to the original status on recovery of the losses with support from insurance claims is enormous There is a settled mind and business because of assured protection and compensation I have other insurance products Insurance is good since it helps one to plan for the future especially education and health

I am about to retire from my job and active service I have planned to use all my money for paying my loans, after I am done with that, I may consider buying insurance At times, the insurers are unreliable about their claims Challenges in accessing insurance offices Premiums are paid well and on time but claims in the event of a loss are not I don’t have adequate knowledge and information about insurance I don’t trust insurance companies I don’t want to incur extra expenses If I can get a good affordable health policy, I can buy one other than NHIF due to its high premiums

In summary, the cooperatives’ members were interested in buying insurance products mainly because they had relevant information about the importance of insurance for individual members, businesses, and their families; the insurance services that they had experienced before were good and motivated them to buy more, while others already enjoyed the benefits of insurance.

12.4.8 Marketing of Insurance Products by Cooperatives Most of the respondents (64.3%) said that they would buy insurance products if marketed by their cooperatives, 17% members said that they would not buy insurance products if marketed by their cooperatives, 16.3% of the members did not know their position, and 2.4% did not respond. This means that most of the members of the cooperative societies were willing to buy insurance products if marketed by their cooperative societies (Fig. 12.9).

12.4.9 How Can the Existing Legal Framework Support Cooperatives for Engaging in the Distribution of Insurance? A clause should be considered that targets cooperatives directly as distribution channels, specifically outlining the guidelines clearly.

244

R. Kuloba et al.

Would you buy Insurance product if marketed by your CooperaƟve?

Percentage

80.0%

64.3%

60.0% 40.0%

17.0%

16.3%

No

Don’t Know

20.0% 0.0%

Yes

2.4%

Missing

Response Number of Respondents Fig. 12.9 Would you buy Insurance products marketed by the Cooperatives? (Colour figure online)

Amendments to the law to allow distribution of insurance products by the cooperatives. Review the law that SACCOs should engage purely in savings and credit thereby restricting their non-core businesses that could be an avenue for distributing insurance products. Cooperatives is coming up with the policies and regulations to guide distribution of insurance products, awareness creation among its membership and training of relevant Committees and management. Consider making it mandatory for cooperatives to distribute insurance products. Reviewing the policies an operating procedure. The law should be made flexible and friendly for the cooperatives to engage in insurance distribution. The policy should be clear during registration and should restrict the services you can offer. Unlike other insurance agencies, cooperatives should be allowed to create insurance agencies without strict requirements such as the need to have Certificate of Proficiency qualifications (COP), IRA to harmonize their regulations with SACCO regulations to facilitate insurance distribution. The legal framework should be expanded to ensure that cooperatives are covered in the distribution of insurance products.

12.4.10 Do by-Laws Allow Cooperatives to Engage in Other Businesses? Table 12.1 shows whether the by-laws allow cooperatives to engage in any business other than that they were established for. Fifty percent of the cooperatives’ by-laws allowed cooperatives to engage in businesses other than those which they were established for, 45.5% of the cooperatives’

12 Cooperative Societies as a Distribution Channel … Table 12.1 Do your By-laws allow you to engage in any business other than that of SACCOs/Cooperatives?

Response Valid Missing Total

245 Frequency

Percent

Yes

11

50.0

No

10

45.5

Missing

1

4.5

22

100.0

by-laws did not allow this while 4.5% of the respondents did not respond. This means that most of the cooperatives’ by-laws allow cooperatives to engage in businesses other than those for which they were established.

12.4.11 What Are the by-Laws for Engaging in Other Businesses? The respondents provided different views on cooperatives’ by-laws which allowed them to engage in other businesses in addition to the one they were established for. The most common view about the by-laws was that cooperatives are registered and licensed to undertake only those businesses for which they were initially registered. There is only provision for diversification into products and services that are relevant for supporting the core business of the cooperatives as opposed to totally venturing into different businesses. Some suggestions and observations gathered from different cooperatives regarding how by-laws allow cooperatives to engage in other businesses are: Common services or those that are similar to the core business of the society. Allowed to engage in housing activities only. The by-laws do not deny new investments. The by-laws only allow marketing coffee. The cooperative is multipurpose. There is room for provision of more services. The law does not allow cooperatives to enter into any business other than what is authorized by law. The regulator requires the cooperatives to adhere to the core business only. The rules do not allow any other business. The SACCO already has a shop among other services like transport and issuance of materials on loan like maize to members. The SACCO is licensed to take deposits only. There is room for other services like cleaning and courier services. The regulator has very stringent rules. Those activities are aimed at supporting the core activities. A majority (81.8%) of the cooperatives agreed that the by-laws allowed the cooperatives to engage in insurance services, while 13.6% did not agree (Table 12.2).

246

R. Kuloba et al.

Table 12.2 Can the same laws allow the cooperative societies to engage in insurance services?

Response Valid Missing

Frequency

Percent

Yes

18

81.8

No

3

13.6

Missing

Total

1

4.5

22

100.0

Do exisƟng regulaƟons enacted by the county or naƟonal governments support the CooperaƟves in engaging in the distribuƟon of insurance services? Missing 14%

No 13%

Yes No Missing Yes 73%

Fig. 12.10 Whether existing regulations support distribution of insurance (Colour figure online)

The findings of this study with regard to enabling cooperatives to. engage in the distribution of insurance services include: Amendments to the clauses of the Insurance Act. Review of SASRA regulations to accommodate insurance services in SACCOs’ operations. A majority (73%) of the respondents agreed that existing regulations enacted by the county or national governments supported cooperatives in engaging in the distribution of insurance services (Fig. 12.10).

12.4.12 Cooperatives’ Staff Capacity for Insurance Matters Table 12.3 shows whether the cooperatives’ staff members were trained in insurance matters. Table 12.3 Are any of your staff members trained in insurance matters?

Response Valid Missing Total

Frequency

Percent

Yes

10

No

9

45.5 40.9

Missing

3

13.6

22

100.0

12 Cooperative Societies as a Distribution Channel …

247

The study found that 45.5% of the cooperatives’ staff members were trained in insurance matters while 40.9% were not trained for this; 13.6% did not respond. The Cooperatives can improve their value as distribution channels of insurance services through the following interventions: Assuring sustainability of the cooperatives’ businesses and their members. By directly engaging in the insurance business. Coming up with customized insurance products and creating awareness among the members. Education and training for members, developing relevant products customized to suit individual members’ needs. Enhancing income earning through commissions earned. Facilitating the uptake of insurance products. Financial support for distributing insurance products. Generating additional incomes, creating employment, and serving members’ needs effectively. Increased incomes and business growth through an increased revenue base. Increased incomes and an enhanced business image. Members’ education and training. Insurance will help the cooperatives in mitigating risks. Cooperatives will increase their revenue streams. The Appropriate Strategies to be used for introducing cooperative members to the insurance business and products. The leadership of the cooperatives suggested different appropriate strategies to be adopted by IRA while introducing the insurance business and insurance products to the cooperatives’ members. Some of the suggested recommendations are: Adopting a bottom-up approach where a cooperative’s members are sensitized first on insurance matters and then given education and training is to the leadership of the cooperatives; this is supposed to be done before the actual roll out of insurance products. Undertaking a market needs assessment (product research and development) for finding the kind of insurance products that are relevant for the needs of the cooperatives’ members in Kenya. Intensive marketing of insurance products relevant to the cooperative sector using brochures, newsletters, internet, TV ads, mobile phone SMS’s, and also through conferences in members’ education forums. Use of appropriate communication channels which are enhanced using technology. Promoting competitive interests in the market, training and education, using underwriters instead of brokers. Publicity for the insurance services through posters, fliers, AGM, and members’ meetings. Are Cooperatives best placed to be distribution channels for insurance products? All departments of cooperative development (100%) at the county level agreed that cooperative societies are best placed to be distribution channels for insurance products in Kenya because of various reasons including:

248

R. Kuloba et al.

The cooperative movement provides a huge market for insurance use/consumption—insurable resources and membership. Cooperatives have operational strategies already in place like premises/offices, management teams, employees, members, and visibility. Low level of business risks (are sustainable businesses) due to existing commitments. Have networks and linkages with the communities and society at the grassroots level thus having a presence even in the rural set-up and in various classes of the community (rich and poor). Members of cooperatives are already aware through education, training, and information forums hence it will be easy for them to accept insurance services.

12.5 Summary of the Key Findings/Observations The cooperative societies should be considered as distribution channels for insurance products. This research will help improve the financial performance of the insurance sector with the entry of cooperatives to this sector as this will be important and beneficial for the insurance sector. There is still a need for providing more necessary information, knowledge, and skills about insurance to the cooperatives’ members and other players. More marketing and sensitization should be rolled out on a large scale in the rural parts and countrywide. All the cooperatives and their members should be insured. SACCOs, especially deposit-taking SACCOs, should be strengthened and allowed to venture into the distribution of insurance products. There is need for introducing relevant insurance products to the people in villages such as farmers and families with low incomes. IRA should give the necessary advisory services such as on claims and compensation processes to the cooperatives directly rather than through agents for their effective acceptance. IRA should take up all the issues raised by this study and address them through effective implementation by involving the cooperatives and other stakeholders. Insurance is good for the people but regulators are advised to be effective enough by ensuring that the consumers are protected from exploitation especially during compensation for claims. This means that IRA should ensure that the insurance companies and agents are trustworthy, transparent, and effective in their service delivery. Insurance is currently not affordable for all the people. Insurance premiums are very high, so they should be lowered so that all Kenyans can afford them. Insurance products relevant to the cooperative sector should be developed and promoted across the country. IRA should ensure promotion and registration of relevant insurance products for the cooperatives and their members.

12 Cooperative Societies as a Distribution Channel …

249

12.6 Conclusion and Recommendations From the study cooperative enterprises and its membership are aware of insurance. The existing structures of the cooperatives are robust, strong, and sustainable since they are able to handle the management members and their deposits. Their governance structures are reliable because they are participatory and inclusive with clear mechanisms for appointment. The study further concludes that as aggregators, cooperative enterprises are appropriate for distributing insurance services in Kenya given their wide reach and captive membership. They have a strong capital base, effective governance structures, competent managerial and technical staff, have made investments in ICT, are membership driven, and have the essential business networks.

12.6.1 Recommendations Need to create awareness about insurance and educating SACOO members. Encourage cooperatives to participate in insurance agencies’ business by sensitizing them about insurance in partnership with other stakeholders. They are strong at the grassroots level, enjoy massive support, and are trusted by their members. The SACCO assurance model is an aggregator that is key in the insurance distribution value chain for insurance services in Kenya. Although the regulatory framework is enabling, there is need for its further review and harmonization by defining SACCO assurance in the legal framework and developing SACCO assurance regulations to ensure that the cooperatives play their role. The insurance regulator needs to actively engage with cooperatives and help structure appropriate capacity building programs on insurance so as to enhance awareness and understanding.

References African Insurance Organization (2016). African insurance barometer: market survey. Retrieved from: http://www.african-insurance.org/documents/Both%20Books.pdf. African Insurance Organization. Business Sweden (2016). The Financial Services Sector in Kenya. Nairobi. Chambo, S.A., M. M. Mwangi, and O. O. Oloo (2008). An Analysis of the Socio-Economic Impact of Cooperatives in Africa and their Institutional Context. ICA and the Canadian Cooperative Association. Gatuguta, E. M., P. Kimotho, and S. Kiptoo (2014). History and organization of cooperative development and marketing sub-sector in Kenya. Retrieved from: http://www.industrialization. go.ke/images/downloads/history-and-organization-of-cooperative-development-and-marketingsub-sector-in-kenya.pdf. IRA Website (2017). www.ira.go.ke, Annual Insurance Industry Report 2017.

Part IV

Economic Integration, International Trade and FDI

Chapter 13

The FDI-Domestic Investment Nexus in SSA Yemane Michael

Abstract This study empirically investigates the nexus between foreign direct investments (FDI) and domestic investments (DI) in sub-Saharan African (SSA) countries for the period 1986–2015. The study uses panel data for 40 SSA countries. Using the flexible accelerator model of investments, the dynamic common correlated effects estimator, and other types of dynamic and static estimation methods, the study finds that FDI crowds-out domestic investments in SSA countries. Specifically, by applying the Chudik and Pesaran (2015) dynamic common correlated effects estimator it finds that a 1% increase in FDI inflows results in a reduction of domestic investments by 0.037 to 0.126%. Keywords FDI · Domestic investments · SSA · Dynamic common correlated effects estimator JEL Codes C51 · C52 · F21 · O11 · O55

13.1 Introduction This chapter investigates the various dimensions of the relationship between foreign direct investments (FDI), domestic investments (DI), and economic growth in subSaharan Africa (SSA). The main focus of this paper is analyzing whether FDI crowdsout or crowds-in domestic investments in SSA. The study presumes that FDI influences economic growth by promoting domestic investments and exports and by developing human capital, infrastructure, and institutions. Of all these, domestic investments are probably the most important channel through which economic growth in the host country is influenced by FDI because FDI influences employment and incomes more directly through this mechanism than through other channels. Y. Michael (B) Department of Economics, College of Business and Economics, Mekelle University, Tigrai, Ethiopia e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 J. Wood and O. Habimana (eds.), A Multidimensional Economic Assessment of Africa, Frontiers in African Business Research, https://doi.org/10.1007/978-981-15-4510-8_13

253

254

Y. Michael

13.1.1 A Review of Direct and Indirect Links Between Financial Sources and Economic Growth The amount of FDI flows to developing countries grew steadily in the 1990s and reached $583 billion in 2009 in current US dollar terms (The World Bank Group 2010). However, UNCTAD’s (2015), World Investment Report puts the figure at 681 billion US dollars with a 2% increase. It also stresses that developing countries had extended their lead in global inflows. China had become the world’s leading recipient of FDI. Five of the world’s top-10 recipients of FDI in 2015 developing countries. The tremendous increase in FDI inflows happened mainly due to the policies that these countries adopted such as reducing FDI barriers and offering tax incentives and subsidies for attracting FDI. The increasing importance of FDI flows as a source of external funding for recipient countries encouraged research on the channels through which FDI might promote economic growth. The key point in evaluating the FDIgrowth nexus is captured by the link between foreign and domestic investments. As a result, a number of studies have investigated whether FDI and domestic investments complement each other or if they are substitutes in recipient countries. Colen et al. (2008), state that FDI has a more advanced level of technology, managerial capacities, skills, and know-how which result in higher levels of efficiency and productivity. Hence, FDI contributes directly and more strongly than domestic investments to accelerating growth in the host economy. However, some scholars refute the argument that foreign firms are more productive than domestic ones. For example, Mutenyo (2008), _ENREF_33investigated FDI’s impact on economic growth in 32 sub-Saharan African countries by applying crosssectional and dynamic panel data for the period 1990–2003. He found consistent results that FDI had a positive impact on economic growth but it was less efficient than domestic investments. This is because FDI inflows might not be accompanied by improved technological and managerial capacities or organizational structures all the time especially when FDI takes the form of mergers and acquisitions (M&A). Believing that FDI could create positive externalities and spillover effects in the form of technology transfers to their economies, various developing countries have offered special treatment to foreign enterprises and multinational corporations (MNCs) (Aitken and Harrison 1999). These countries have invited MNCs to invest in their economies thinking that this will enable them to access technologies that cannot be produced by domestic firms (Blomström et al. 1999). FDI is regarded as the primary channel through which technological transfers take place. The effect of FDI on domestic economic growth depends on the diffusion of best practices through the local economy (Ajayi 2006). MNCs produce different spillover effects and there are different channels through which these effects occur.1 Host countries are tempted to attract MNCs and FDI in the hope that they will boost the productivity of domestic firms. Developing countries presume that foreign enterprises own intangible assets and cutting-edge technology that can be passed on to 1 For

more information on FDI’s various spillover effects see Blomström et al. (1999).

13 The FDI-Domestic Investment Nexus in SSA

255

domestic firms which will eventually improve their productivity levels. Productivity distribution is related to externalities which are often referred to as productivity spillovers (Crespo et al. 2008). The crowding-in and crowding-out effects of FDI on DI work as follows: FDI’s crowding-in effects happen when FDI by foreign firms builds new investments in downstream or upstream production in the host country which would not have happened if they were not there, particularly when investments are made in undeveloped sectors of the economy. Conversely, FDI’s crowding-out effects happen when FDI firms distort domestic firms and other foreign affiliates from undertaking investments by driving them out of business (Bende-Nabende and Slater 2003). De Mello (1997) and Apergis et al. (2006) argue that FDI can affect DI through its effect on the profitability of domestic investors which leads to the crowding-out effect of DI. But on the other hand, FDI can have an impact on adjustments in ownership structures of the total investments in the host country; this offers additional financial support to DI. This effect leads to crowding-in of domestic investments in the receiving countries. To put it differently, if FDI has no effect on domestic investments, any increase in FDI ought to be reflected in a dollar-for-dollar increase in total investments. If FDI crowds-out or supplants investments by domestic firms, the increase in total investments should be smaller than the increase in FDI. Finally, if there is crowding-in or a complementarity effect, total investments should increase more than the increase in FDI. The entry of MNCs may create competition that crowds-out investments by domestic firms. But it is also possible that FDI might stimulate DI and lead to the crowding-in effect of investments by domestic firms (Colen et al. 2008). However, Borensztein et al. (1998) argue that the effects of FDI on domestic investments can be different: MNCs that compete in product and financial markets may crowd-out investments by domestic firms. However, there is an avenue whereby FDI could support the expansion of domestic firms—complementarity in production or by increasing productivity through the spillover of advanced technologies. Policies that offer special tax treatment and other incentives such as export free zones and tax exemptions to attract FDI inflows may introduce distortions that affect domestic investments. These distortions, in turn, could have a greater deleterious impact on domestic investments which will limit the growth spillover effects that are meant to be generated through FDI’s crowding-in effects (Borensztein et al. 1998; Colen et al. 2008). FDI has a positive influence on DI when it contributes to the introduction of new industries to the host country. This means, if FDI has a positive externality in the creation of new industries at home, then it is presumed to enhance DI. Lipsey (2002), offers new investment prospects for local firms through the provision of machinery and technology (Sun 1998); and generates new demand for local inputs (Cardoso and Dornbusch 1989). Foreign and domestic investments are also likely to be substitutes if foreign firms compete with domestic firms for the use of domestic resources as this will inhibit investment opportunities for domestic investors (Agosin and Mayer 2000; Fry 1992;

256

Y. Michael

Jansen and Stokman 2004). In this case, FDI’s effect on economic growth can be dampened and its role in the economies of recipient countries can be ambiguous. One thing that has to be known for analyzing the relationship between FDI and domestic investments is the linkages between FDI, public investments, and private investments which help identify the policy implications that can be drawn to maximize FDI’s gains. A strong private investment climate which acts as a signal of high returns to capital as well as improved public infrastructure via public investments that cut down the cost of doing business are crucial for attracting foreign capital. It is also possible that FDI may supplant or complement different types of domestic investments. Ndikumana and Verick (2008) studied sub-Saharan African countries and found a two-way relation between FDI and private investments. But they also noted that public investments were not a driver of FDI. Ang (2009), found that both public investments and FDI were complementary to private investments in Malaysia. MNCs could affect domestic investments in host economies in two ways: directly through their own investment activities and indirectly by affecting investments in the host country’s firms (UNCTAD 1999). Herzer et al. (2008), argue that as endogenous growth theory depicts, positive knowledge spillovers cannot run from FDI to DI especially in developing countries. Görg and Greenaway (2004), assert that there is a positive spillover from FDI to DI only in developed countries and not in developing countries. Borensztein et al. (1998), found a crowding-in effect of FDI in a sample of 69 developing countries. A one-dollar increase in net FDI inflows led to an increase in DI in the host country by more than one dollar. Moreover, their findings also suggest that the complementarity between FDI and DI was not sensitive to FDI’s productivity. Some studies have also found evidence of negative spillover effects from FDI to domestic firms in developing and transition economies by using firm level panel data from manufacturing industries. For example, Haddad and Harrison (1993), used data from Moroccan manufacturing industries and found that horizontal spillovers did not take place in all the industrial sectors. Aitken and Harrison (1999), probed the impact of MNCs on domestic firms in Venezuela and Djankov and Hoekman (2000), in the Czech Republic. They found that the MNCs shifted the demand for intermediate inputs from domestic to foreign producers thereby reducing the scale of output and productivity in local production and causing negative spillover effects rather than positive ones. Generally speaking, FDI’s positive contribution to economic growth occurs when FDI crowds-in DI. Contrarily, FDI can decrease DI when it takes away investment opportunities through licenses, skilled labor, and credit facilities. This shows the dominance of FDI over DI (Herzer et al. 2008). However, some studies on this relation have concluded that there is a strong relationship between FDI inflows and DI over time (Lipsey 2004). However, the relationship between FDI and DI is not necessarily unidirectional only running from FDI to DI. It is highly plausible that DI can also affect FDI in several ways. For example, increased investments in physical and human infrastructure can lead to increased FDI profitability thus further enhancing FDI’s efficiency (Apergis et al. 2006). In addition, DI can act as a signal of the state of the investment

13 The FDI-Domestic Investment Nexus in SSA

257

climate if information regarding investment opportunities and threats is unavailable or incomplete in the host country. Agosin and Mayer (2000), developed a theoretical model of investments based on the neo-classical investment model to test whether FDI crowds-in/out DI in three groups of developing countries (Africa, Asia, and Latin America) from 1970 to 1996. They found strong crowding-in effects for DI in Asia and lower effects in Africa, whilst there were strong crowding-out effects in Latin America. Their study mainly focused on FDI’s impact on DI and did not consider the dynamic interaction between FDI, DI, and economic growth. Growth-driven FDI occurs when the growth of the host economy attracts FDI. Economic theory gives us different fundamental reasons about MNCs’ decisions to invest in advanced or developing countries. Growth-driven FDI has been strongly supported by the empirical findings of Baliamoune-Lutz (2004), which are based on data from developed economies and Asian countries. The strong links between FDI and growth are a result of either growth-driven FDI or FDI-led growth; this can make it possible for the two variables to move together through feedback or bidirectional causality (Zhang 2005).

13.2 A Model of the Nexus Between Domestic Investments and FDI This section develops a model of domestic investments placing FDI at the center of the analysis. The model is empirically estimated to see whether FDI crowds-in or crowds-out domestic investments in SSA. Empirical literature focuses on four approaches for modelling investments in developing countries—the accelerator model, the expected profits model, the neoclassical model, and Tobin’s Q model. Of these, the flexible accelerator model is the most widely used due to data unavailability for some of the variables, especially for capital stock and the returns on investments. The model assumes that the speed of the adjustment parameter is influenced by a number of variables for which data is available. The empirical model used in our study is based on the flexible accelerator model, which assumes that the desired capital stock is proportional to the level of expected output (Abdullah 2017; Blejer and Khan 1984; Mody and Murshid 2005; Ramirez 1994) expressed as: K t∗ = αYte

(13.1)

where K t∗ denotes the desired capital stock in period and Yte is the expected level of output in period . The expected level of output can also be considered as future aggregate demand.

258

Y. Michael

This model assumes that the actual stock of capital adjusts to the difference between the desired stock in period t and the actual stock in the preceding period (t–1). The relationship is given by:   K t − K t−1 = K t = β K t∗ − K t−1

(13.2)

where is the coefficient of adjustment. Equation (13.2) implies that the extent of adjustment per period in the current actual capital stock is a fraction of the difference between the desired stock in the current period and the actual stock in the previous period. If = 1, the actual capital stock adjusts fully to the desired level immediately within one period. On the other hand, a value of = 0 signifies that there is no adjustment. However, due to technical and capital cost constraints and the time required to plan, build, and install new capital and other adjustment costs, instantaneous adjustment to the desired level of capital ( = 1) is less plausible and it is commonly assumed that lies between 0 and 1. In most developing countries, data on investments (capital flows) is more readily available than that on capital stock. Hence, we use gross investments which are defined as: It = (K t − K t−1 ) + δ K t−1

(13.3)

where is the depreciation rate of capital stock and It is gross investments. Using the lag-operator, we can rewrite Eq. (13.3) as: It = [1 − (1 − δ)L]K t

(13.3a)

where L denotes the one-period lag operator defined as L K = K t−1 . For all practical purposes, the partial adjustment mechanism of investments can be expressed as: It − It−1 = β(I ∗ −It−1 )

(13.4)

Following Blejer and Khan (1984), Ramirez (1994), Erden and Holcombe (2006) and Abdullah (2017), we assume that the speed of adjustment coefficient which defines the gap between the actual and desired investments is influenced by FDI and other relevant macroeconomic variables. This helps us incorporate more dynamism into our empirical model. Hence, β can be defined as a linear function as:  β = α0 +

 1 (φ1 F D It + φ2 X t ) It∗ − It−1

(13.5)

where 0 is the intercept, is a vector of the other determinants of investment, and FDI is usual FDI. The coefficient of adjustment captures the impact of FDI’s current and lagged values and other determinants on domestic investments. If FDI complements or supplements domestic investments, an increase in FDI speeds up the

13 The FDI-Domestic Investment Nexus in SSA

259

adjustment coefficient and closes the gap between the actual and desired investments. On the other hand, if FDI supplants domestic investments, an increase in FDI slows down the adjustment coefficient and widens the gap between the actual and desired investments. Subsuming the value of β from Eq. (13.5) into Eq. (13.4) and after some small simplifications, yields:   It − It−1 = α0 It∗ − It−1 + (φ1 F D It + φ2 X t )

(13.6)

The steady-state value of the investments can be expressed from Eq. (13.3a) as: It∗ = [1 − (1 − δ)L]K t∗

(13.7)

We now surrogate the expression for K t∗ from Eq. (13.1) into Eq. (13.7) to get = [1 − (1 − δ)L]αYte . Finally, we plug this expression for It∗ in Eq. (13.6) which gives us: It∗

  It − It−1 = α0 [1 − (1 − δ)L]αYte − It−1 + (ϕ1 F D It + ϕ2 X t ). Rearranging this expression and adding the subscript i to give it a panel dimension and the composite error term, yields: Iit = α0 α[1 − (1 − δ)L]Yite + (1 − α)Iit−1 + (φ1 F D Iit + φ2 X it ) + μit

(13.8)

where = 1, 2, …, represents cross-sectional units, = 1, 2, …. denotes the time dimension of the panel data, and μ is the composite error term. By letting (1 − α0 ) = ϕ1 , α0 α[1 − (1 − δ)L] = ϕ2 , φ1 = ϕ3 , and φ2 = ϕ4 , the flexible accelerator the model of DI from Eq. (13.8) can be written as: Iit = ϕ1 Iit−1 + ϕ2 Yite + ϕ3 F D Iit + ϕ4 X it + μit

(13.9)

Here, domestic investments are considered a partial adjustment process between the current actually existing and the desired capital stock due to the presence of liquidity constraints, adjustment costs, and the considerable time that elapses before full adjustments take place. Besides, since investments are a structural component of the economy, they are expected to show strong autoregressive behavior. Thus, we incorporate the lagged value of domestic investments (Iit-1 ) in our empirical equation. This allows us to consider the persistence of the domestic investment rate. It also highlights the dynamic nature of the domestic investments framework by manipulating long-term coefficients. The coefficient of Ye which captures the accelerator effect is expected to be positive. Since, the expected output Ye is not observable and hence cannot be measured directly, we need to use a proxy for it. To this end, we estimate a first-order autoregressive model, AR(1) of real GDP for every country in the sample and calculate

260

Y. Michael

the predicted values. We then use these predicted values as a proxy for the expected output (see Blejer and Khan 1984; Erden and Holcombe 2006; Ramirez 1994). The variables which are supposed to influence the adjustment coefficient β and are included in the baseline regression model include inflation rate, official exchange rate, openness, lending interest rate, financial development whose proxy is broad money, and foreign aid. The other variables included in the baseline regression are the ones for which sufficiently long panel data series exist. The rationale for adding each of these regressors in the baseline model is given as: A hike in the lending interest rate increases the cost of borrowing and results in a reduction in investment demand. Therefore, the lending interest rate is expected to have a negative influence on domestic investments. The lending interest rate is used as a proxy for the cost of capital. The inflation rate reduces the real returns on investments and hence dissuades investors from undertaking more investments, ceteris paribus. Moreover, the inflation rate can be taken as an indicator of macroeconomic instability. Higher inflation, especially above a certain threshold, causes jitters in the economy and reduces the returns on investments. A combination of lower expected returns and higher uncertainty caused by a panic due to excessive inflation discourages investments. But, from a different perspective the Tobin-Mundell model predicts that an increase in the inflation rate reduces the real interest rate and encourages investments. Thus, the final impact of inflation on investments is uncertain. An increase in the official exchange rate (measured by units of local currency per US dollar) could raise the price of tradables relative to non-tradables. This, in turn, may lead to an increase in investments in tradable-goods producing sectors and a decline in investments in the sectors that produce non-tradable goods. Thus, the depreciation of the domestic currency may result in an increase in overall investments. On the flip-side, depreciation of the local currency may lower the real value of income and assets which could lower the demand for investments. In addition, depreciation may also increase the real cost of imported capital goods and machines that are destined for investment purposes thus discouraging investments. It is also possible that depreciation increases the burden of foreign debt which may influence investments negatively. Therefore, whether depreciation stimulates or retards domestic investments is not clear on a priori grounds and is an empirical question. Trade openness: This bolsters greater utilization of capacity, realization of economies of scale, and technological improvements due to competition in foreign markets (Helpman and Krugman 1985). Moreover, trade openness mitigates foreign exchange constraints and facilitates import of capital goods and machinery which promotes domestic capital formation. This, in turn, enhances countries’ capacity to export more which likely promotes economic growth in the long-run by speeding-up the learning process from abroad and improving technological innovations. A high degree of trade openness is expected to boost investments especially in export-oriented sectors. Thus, we use openness measured by the sum of exports and imports of goods and services as a percentage of GDP in the baseline regression model. However, later we use exports as a percentage of GDP to check for robustness.

13 The FDI-Domestic Investment Nexus in SSA

261

When it comes to FDI, if its coefficient is positive, it will be deemed that complementarity exists between FDI and domestic investments, that is, a dollar of inward FDI adds more than a dollar to gross fixed capital formation (GFCF) and induces more domestic investments. This scenario is referred to as the crowding-in of domestic investments. On the other hand, when crowding-out occurs, a dollar of FDI inflows adds less than a dollar to GFCF and domestic investments decrease. Therefore, there is substitutability between FDI and domestic investments. Here, the coefficient of FDI is expected to be negative. The third and final scenario is a situation in which there is a dollar-for-dollar increase in GFCF from FDI whose impact on domestic investments is neutral. FDI has neither a positive nor a negative spillover effect on domestic investments. Other variables used in the robustness check, some of which could crowd-out domestic investment include: Another variable that could influence investment decisions is uncertainty. If investors are uncertain about the future course of events and the macroeconomic climate, they may be reluctant to invest. Hence, an increase in uncertainty is presumed to have a deleterious effect on investments and investment decisions. A number of methods can be used for measuring uncertainty such as volatility in inflation, exchange rate, output growth, terms of trade, and institutional quality. We measure uncertainty by a three-year rolling window standard deviation of inflation following Mody and Murshid (2005), Wang (2010) and Lee et al. (2012). This variable is addressed as an extension of the study. Institutions: There is ample empirical evidence that countries with better institutions attract more FDI than those with poor institutions (Busse and Hefeker 2007; Daude and Stein 2007; Globerman and Shapiro 2002). There are many reasons to justify this argument. Wei (2000), asserts that poor institutions increase the cost of operations. Moreover, poor institutions increase uncertainty (Daude and Stein 2007). And finally, productivity is lower in countries with poor institutions as evidenced by Daniele and Marani (2011). All these factors motivate MNCs to prefer countries with better institutions as their investment destinations. A country that attracts more FDI is likely to have a thriving DI due to backward and forward linkages. But when viewed from a different perspective, countries with better institutions are more likely to have a stronger and stricter property rights system which limits the scope of spillovers of knowledge and technology. When this happens, more FDI inflows triggered by better institutions could intensify competition in domestic markets and dampen or crowd-out domestic investments. Therefore, FDI’s impact on domestic investments triggered by institutional factors is ambiguous. This variable is added to the model in the section that deals with robustness checks and a sensitivity analysis. Human capital: MNCs’ have a tendency to choose host countries and investment destinations on the basis of the availability of a more skilled, educated labor force since the MNCs mostly produce knowledge and technology-intensive goods and services. Gao (2005) and Du et al. (2008) found a positive association between human capital and FDI inflows. Sometimes, MNCs train local workers to enhance their skills. The moment these workers move to local firms by leaving foreign firms,

262

Y. Michael

knowledge flows take place to these local firms. Hence, labor turnover from MNCs can lead to technological spillovers which promote domestic investments. Infrastructure: This is another important variable that could have a bearing on MNCs’ investment decisions. MNCs prefer to invest in countries with an adequate and reliable supply of physical infrastructure such as electricity, networks, highways, ports, and roads. In an attempt to attract more FDI, host countries engage in more investments in this type of infrastructure. Kose et al. (2006), call these the indirect impacts or ‘collateral benefits.’ In their endeavor to attract FDI, governments in developing countries are tempted to implement sound macroeconomic policies and creating a conducive environment for doing business including developing institutions and improving governance. The availability of this infrastructure boosts productivity and increases returns on investment, which also leads to the promotion of more domestic investments. Financial development: A well-developed financial system helps promote domestic investments through MNCs’ backward linkages. On the other hand, if the financial sector is poor and the local entrepreneurs face credit constraints, demand for inputs created by the presence of MNCs alone does not achieve the desired and intended targets. A financial system helps the smooth functioning of an economy in many ways. First, a strong and well-developed financial system disseminates more information in a more efficient way than individuals do. This results in reducing the costs of investing in firms and facilitates more efficient allocation of capital (King and Levine 1993). Second, a financial system improves corporate governance which promotes investments. Financial intermediaries serve as a balancing act by monitoring the way firms’ agents and managers use the funds of the corporate businesses that they run on behalf of the shareholders. This helps shareholders and creditors and reduces the cost of monitoring which in turn reduces credit rationing and promotes investments (Bencivenga and Smith 1993). Third, a financial system creates a conducive environment for investors to diversify their investment portfolios which reduces risks. This in turn is likely to promote investments in riskier projects whose returns are higher than those of safer projects. Besides these benefits, a financial system helps reduce liquidity risks and promotes investments in projects with long gestation periods which usually yield higher returns (Levine 1997). Financial systems also mobilize a larger pool of savings and mobilize it more cheaply than individuals due to economies of scale which lead to higher investments and a faster rate of capital accumulation (Levine 2005).

13.3 Data Sources and Methodology The data used for this study was retrieved from the World Bank’s (2016) online database for 40 SSA countries for the period 1986–2015. There are very few missing values. The main variables for which data was retrieved include gross fixed capital formation, official exchange rate, lending interest rate, personal remittances, GDP,

13 The FDI-Domestic Investment Nexus in SSA

263

broad money (which proxies financial development), mobile cellular and fixed line subscribers out of 100 people (the sum of which is used as a proxy for infrastructure), inflation rate, openness (the sum of exports and imports of goods and services as a percentage of GDP), foreign aid, FDIG, and total population. We use two dynamic common correlated effects estimators (DCCE): dynamic common correlated effects mean group estimator (DCCEMG) and dynamic common correlated effects pooled estimator (DCCEP) and two static common correlated effects estimators (CCE): common correlated effects of the mean group estimator (CCEMG) and the common correlated effects pooled estimator (CCEP) to find the results of our baseline regression for analyzing the relationship between domestic investments and FDI. The common correlated effects estimator models’ unobserved common factors are controlled for by incorporating cross-section averages of the explanatory variables and the dependent variables in the regression models. The method for common correlated effects’ estimators: We use Chudik and Pesaran’s (2015), dynamic common correlated effects estimator (DCCE) and the static version of this estimator, namely the common correlated effects mean group (CCEMG) estimator to study the relationship between DI and FDI. These estimators allow parameter heterogeneity and control for unobserved non-stationary common factors and endogeneity that arises due to the presence of such common factors and is robust to cross-sectional dependence, the absence of cointegration, and the presence of structural breaks. It can cautiously be claimed that this study is the first of its kind that applies these methods to study the relationship between domestic investments and FDI. The common correlated effects estimator is based on: yit = βi xit + u it 

u it = νi + γi f t + eit

(13.10) (13.11)

The first term νi in Eq. (13.11) represents time-invariant, country-fixed effects; f t is a vector of unobserved common factors; and γi is country specific heterogeneous factor loading. The heterogeneous coefficients are randomly distributed around a common mean, such that βi = β + νi , νi ∼ I I D(0, Ων )(Pesaran and Smith 1995). Pesaran (2006) shows that Eq. (13.10) can be consistently estimated by approximating the unobserved common factors with cross-sectional means x¯t and y¯t under strict exogeneity. The estimated equation becomes: yit = βi xit + δi x¯t + ηi y¯t + eit where x¯t = 1 N

N  i=1

1 N

N 

(13.12)

xit is the cross-section average of the regressor(s) and y¯t =

i=1

yit is the cross-sectional average of the dependent variable.

264

Y. Michael

Chudik and Pesaran (2015) and Everaert and De Groote (2016) show that the common correlated effects (CCE) estimator is consistent only in static panels. The dynamic version of the model should incorporate the lagged dependent variable as a regressor. As this regressor is not strictly exogenous, the CCE estimator becomes inconsistent. Chudik √and Pesaran (2015), show that the efficiency of the CCE estimator improves when 3 T lags of the cross-section means of the dependent and explanatory variables are included in the model. Once these are included in Eq. (13.12), the new model becomes: yit = αi yit−1 + βi xit +

PT p=0

δi, p x¯t− p +

PT

ηi, p y¯t− p + eit

(13.13)

p=0

where PT is the number of lags. The model given in Eq. (13.13) is known as the dynamic common correlated effects (DCCE) estimator. To estimate the mean group, a separate regression is run for each cross-sectional unit and the αi and βi estimators are derived by averaging them while in the estimation of the pooled mean group the estimated coefficients are constrained to be equal across all cross-sectional units. We used the dynamic fixed-effects (DFE) estimation approach by pooling each group’s time-series data and allowing the intercepts to differ across groups. The DFE estimator, however, produces inconsistent and potentially misleading results if the slope coefficients are not identical. On the other hand, the model can be fitted separately to each group and a simple arithmetic average of the coefficients can be calculated. This approach is that of the mean group (MG) estimator proposed by Pesaran and Smith (1995). With the MG estimator, the intercepts, the slope coefficients, and the error variances are allowed to differ across groups (Blackburne and Frank 2007). Pesaran et al. (1997, 1999), also propose a pooled mean group (PMG) estimator which combines both pooling and averaging. This intermediate estimator allows the intercept, short-run coefficients, and error variances to differ across groups (as is the case with the MG estimator) but it imposes a constraint that the long-run coefficients be equal across groups (as is the case with the DFE estimator). For this kind of a nonlinear model in parameters, Pesaran et al. (1999), constructed a maximum likelihood estimation method to estimate the parameters. We used PMG, DFE, and the system GMM methods to check the robustness of our findings.

13.4 A Discussion of the Main Findings 13.4.1 Descriptive Statistics This sub-section presents descriptive statistics of the variables used in the GMM estimation. Table 13.1 gives the summary statistics of the variables while Table 13.A1

13 The FDI-Domestic Investment Nexus in SSA

265

Table 13.1 Summary statistics Variables

N

Sum

Domestic investments

1,200

21,197

Mean 17.66

13.38

28.08

2.423

2.423

Domestic investment lag

1,160

20,435

17.62

13.49

28.11

2.465

2.465

Uncertainty

1,200

61,261

51.05

FDI

1,200

4,821

4.018

sd

656.5 10.23

363.0 74.65

18.66 5.973

18.66 5.973

1,146

87,139

76.04

1,200

69,510

57.92

737.9

597.7

24.42

24.42

Inflation

1,200

62,368

51.97

727.1

948.3

29.50

29.50

Lending interest rate

1,200

22,733

18.94

Official exchange rate

1,193

470,524

Infrastructure

1,200

27,821

23.18

36.90

Institutional quality index

600

18,387

30.64

19.42

685.1

3.484

J. Bera

Broad money

16.64

23.42

Skewness

Openness

394.4

48.12

Kurtosis

3.484

66.59

6.548

6.548

47.34

5.644

5.644

6.758

2.004

2.004

2.421

0.612

0.612

Note sd refers to standard deviation and J. Bera is Jarque-Bera Source Own calculation

in the Appendix gives the correlation matrix. The summary statistics form the basis of the quantitative analysis of the data in the empirical investigation. From Table 13.1 we can observe that the values of the Jarque-Bera test statistics for some of the variables fall around 3, signifying that the null hypothesis of the normal distribution of the residuals of the variables cannot be rejected. However, the kurtosis for each variable is very far from 3 which indicates that the variables are not normally distributed. The skewness values of all the variables are positive implying that the variables are more tilted to the right. The standard deviation is high when compared to the mean which indicates a high coefficient of variation. But the ratio of the mean over the median is closely equal to one, representing the normality of the distribution. A normal distribution is supposed to be symmetric with a skewness value close to zero. However, the extremely high skewness and kurtosis values for some of the variables can be attributed to their heavy tails. Extreme values in the tails can distort the mean and standard deviations. This is also true for skewness and kurtosis. Statistical literature recommends taking the log of a dataset of a variable that exhibits moderate skewness to the right. We apply this transformation. As indicated in Table 13.A1 in the Appendix, there is no high correlation among the variables. This is backed by the insignificance of most of the variables in the econometric model’s estimation. The correlation results show that the relationship between most of the variables is low. Some correlation coefficients are negative while others are positive. For example, the correlation coefficient 0.441 in Table 13.A1

266

Y. Michael

indicates that domestic investments, the dependent variable, and openness, one of the explanatory variables, are positively related. Besides, the coefficient implies that when openness increases by 1%, domestic investments increase by 0.441% ceteris paribus. The other results can be interpreted analogously.

13.4.2 Main Findings of the Domestic Investment-FDI Nexus Table 13.2 presents the coefficients of our regression models obtained from the dynamic common correlated effects mean group (DCCEMG) and the dynamic common correlated effects pooled mean group (DCCEPMG) estimation methods. FDI’s coefficients are negative in three out of the four specifications of which two are significant. The other specification yields a positive coefficient but it is not statistically significant. As stated earlier, two of the four models produced statistically significant evidence that FDI crowds-out DI between 0.037 and 0.126%, that is, a 1% increase in FDI inflows is associated with a reduction in domestic investments by 0.037 to 0.126%. The other model shows that FDI inflows actually increase domestic investments, but this does not have any statistically significant effect. Thus we can say that the estimated coefficients of the regression models in Table 13.2 support the argument that FDI crowds-out domestic investments. Morrissey and Udomkerdmongkol (2012); Mutenyo et al. (2010); and Titarenko (2006) also found that FDI crowded-out domestic investments. The lagged value of domestic investments, trade openness, expected output (a proxy for future aggregate demand), and uncertainty have a positive and statistically significant effect on domestic investments. Uncertainty is expected to have a negative effect on domestic investments and the sign of the coefficient of uncertainty in two of the four models is as expected on a priori grounds. But the significance of the positive effect of uncertainty in one of the four models is an unexpected one and it is counterintuitive. Inflation has a significant negative effect on domestic investments. The signs of the other variables are more or less in line with what is expected a priori but they are insignificant. Table 13.3 gives a combination of short-run and long-run impacts of various variables on domestic investments. The table shows that trade openness has a statistically significant positive impact on domestic investments in one of the four specifications. Moreover, it has a positive but insignificant effect in two specifications in the longrun. The coefficient of the model with a significant effect indicates that for a 1% increase in trade openness, domestic investments increased by 0.44% which seems a high figure. In each of the specific models given in Table 13.3 which are combinations of short-run and long-run effects, the lagged value of domestic investments captures the error correction term. The value and sign of the error correction coefficient is significant and as expected lies between 0 and −1 in each of the two DCCEPMG estimators. However, in the other two models related to DCCEMG estimators, the value or the sign deviates from what is expected a priori. For example, the first model

13 The FDI-Domestic Investment Nexus in SSA

267

Table 13.2 FDI’s impact on domestic investments using the dynamic common correlated effects estimator models DCCEMG

DCCEMG0

DCCEPMG

DCCEPMG0

Explanatory variables

1

2

3

4

Lag of domestic investments

0.144***

−0.088

0.801***

0.103

(0.05)

(0.18)

(0.02)

(0.12)

0.541***

0.054

0.114***

−4.405

(0.14)

(0.39)

(0.03)

(6.53)

FDI

−0.126***

−0.047

−0.037***

0.580

(0.04)

(0.07)

(0.01)

(0.80)

Uncertainty

−0.001

0.093

0.003***

−0.026

(0.01)

(0.09)

(0.00)

(0.03)

0.607

1.341*

0.002

−0.833

(0.46)

(0.78)

(0.00)

(1.90)

Official exchange rate

−0.003

1.014

−0.005

0.018

(0.17)

(1.74)

(0.00)

(0.28)

Infrastructure

−0.007

−0.028

0.000

0.029

(0.01)

(0.03)

(0.00)

(0.04)

−0.002

−0.032

−0.001***

0.015

(0.00)

(0.02)

(0.00)

(0.02)

Lending interest rate

0.079

0.059

0.001

0.058

(0.08)

(0.05)

(0.00)

(0.09)

Observations

1031

1031

1031

1031

Trade openness

Expected output

Inflation

R-squared

0.72

0.76

0.98

0.11

Number of groups

40

40

40

40

Residual stationarity

I(0)/I(1)

I(0)

I(0)

I(0)

Residual CD test

0.47

−1.17

0.36

−0.04

CD test p-value

0.636

0.241

0.719

0.968

Note CD is cross-sectional dependence. I(0) means integrated of order 0, that is, stationary. DCCEMG is the dynamic common correlated effects mean group. DCCEPMG is the dynamic common correlated effects pooled mean group. The suffixes to DCCEPMG show the cross-sectional lags used in estimating the models. * p < 0.10, ** p < 0.05, *** p < 0.01 show that the coefficients are significant at 10%, 5%, and 1% respectively. Standard errors are given in parentheses Source Own calculation

has an expected negative sign of the error adjustment coefficient but its value is less than −1 whereas the second model has a positive fractional value that lies between 0 and 1 instead of −1 implying that the model diverges and long-run convergence is unlikely to occur. The official exchange rate, FDI, and uncertainty all have a negative and significant impact on domestic investments in the long-run. Regarding the official exchange rate, depreciation/devaluation seems to affect domestic investments by increasing

268

Y. Michael

Table 13.3 FDI’s impact on domestic investments using the dynamic common correlated effects estimator model (Dependent variable: domestic investments) DCCEMG1

DCCEMG2

DCCEPMG1

DCCEPMG2

Explanatory variables

1

2

3

4

D. domestic investments

0.278*

−0.912

−0.095

0.313

(0.17)

(0.66)

(0.26)

(0.22)

0.203

−0.317

0.173

−0.198

(0.31)

(1.01)

(0.66)

(0.52)

D. FDI

0.085

−0.152

−0.023

−0.677

(0.12)

(0.17)

(0.04)

(0.85)

D. uncertainty

0.005

0.090*

0.045

0.011

(0.04)

(0.04)

(0.05)

(0.02)

D. official exchange rate

−0.621

−0.981

−0.814

−0.926*

(0.59)

(1.41)

(0.69)

(0.51)

D. inflation

−0.027

0.099*

0.047

0.003

(0.02)

(0.05)

(0.04)

(0.01)

D. lending interest rate

−0.193

0.274*

0.002

−0.066

(0.15)

(0.15)

(0.02)

(0.11)

Lag of domestic investments

−1.327***

0.201

−0.217***

−0.641***

(0.29)

(0.83)

(0.03)

(0.23)

Trade openness

−1.644

2.531

0.442*

0.078

(3.10)

(1.84)

(0.25)

(0.34)

Official exchange rate

−0.247

0.127

−0.127***

0.076

(0.91)

(0.43)

(0.05)

(0.08)

FDI

−0.512

0.243

−0.194***

−0.145**

(0.42)

(0.20)

(0.07)

(0.07)

0.053

−0.398

0.051

0.080

(0.62)

(0.33)

(0.05)

(0.05)

Inflation

−0.115

0.060

0.014***

0.017

(0.15)

(0.09)

(0.00)

(0.02)

Uncertainty

0.294

−0.035

−0.005*

−0.008

(0.27)

(0.11)

(0.00)

(0.03)

0.461*

−0.080

0.027***

0.009

(0.26)

(0.13)

(0.01)

(0.01)

Infrastructure

–

–

0.005**

0.005**

–

–

(0.00)

(0.00)

Observations

933

933

933

933

R-squared

0.86

0.96

0.48

0.71

D. trade openness

Expected output

lending interest rate

(continued)

13 The FDI-Domestic Investment Nexus in SSA

269

Table 13.3 (continued) DCCEMG1

DCCEMG2

DCCEPMG1

DCCEPMG2

Number of groups

40

40

40

40

Residual stationarity

I(0)

I(0)

I(0)

I(0)

Residual CD test

2.23

0.22

1.05

2.35

CD test p-value

0.025

0.827

0.294

0.018

Note CD is cross-sectional dependence. I(0) is integrated of order 0, that is, stationary. DCCEMG is the dynamic common correlated effects mean group. DCCEPMG is the dynamic common correlated effects pooled mean group. The suffixes to DCCEPMG show the cross-sectional lags used in estimating the models. * p < 0.10, ** p < 0.05, *** and p < 0.01 show that the coefficients are significant at 10%, 5%, and 1% respectively. Standard errors are given in parentheses. The prefix ‘D’ for the variables preceding ‘lag of domestic investments’ stands for differenced. These variables are meant to capture short-run impacts on domestic investments Source Own calculation

the real cost of imported capital machines which discourages domestic investments and increases the burden of foreign debt which in turn could influence investments negatively. The coefficients of the FDI variable indicate that for each dollar increase in FDI, domestic investments decline between 14 and 19 cents confirming FDI’s crowding-out effect on domestic investments. This result is in line with Morrissey and Udomkerdmongkol (2012), who in their study of 46 low-and middle-income countries documented a negative relationship between FDI and domestic investments. Abdullah (2017), also claims that there is unambiguous support for the hypothesis that FDI crowds-out DI. His regression results suggest that countries that have weak institutions, less developed financial systems, less human capital, and less developed infrastructure, and economies that are more open, are more likely to experience crowding-out effects of FDI. The negative and significant impact of uncertainty shows that investors become reluctant and are hesitant to invest in the face of a political, social, and economic environment in the future. Inflation, the lending interest rate, and infrastructure are the other variables that have a positive and significant effect on domestic investments. The positive significant effect of inflation looks counterintuitive but it can be argued that inflation becomes harmful and destabilizing only when it surpasses a certain threshold level. The lending interest rate can positively affect domestic investments. Stiglitz and Weiss (1981) and Stiglitz (1994) argue that a moderate increase in lending interest rates leads to a higher volume of lending. However, an additional increase in rates beyond a certain level will prompt a lower level of lending by adversely changing the pool of quality borrowers in favor of the errant and unsafe ones. Infrastructure which is proxied by the sum of the number of fixed line and mobile cellular users per 100 people has a positive and significant effect on domestic investments. The availability of communication networks as a part of the broader infrastructure increases productivity which increases the returns on investment and this in turn eventually boosts domestic investments.

270

Y. Michael

We also observe that the differenced value of domestic investments has a statistically significant positive impact on domestic investments in the short-run. Uncertainty, inflation, and lending interest rates also have a positive and significant impact on domestic investments in the short-run while the official exchange rate has a significantly negative effect on domestic investments. However, the other variables included in the short-run regressions are found to have a neutral impact on domestic investments. The diagnostic tests for the evaluation of the empirical models are reported in the bottom half of Table 13.3. We applied the Maddala and Wu (1999) panel unit root test, the Pesaran (2007) CIPS of panel unit root test in the presence of crosssectional dependence, and the Pesaran (2003) CADF test for identifying the stationarity of the test residual series. The results show that the residual series obtained from the dynamic common correlated effects pooled mean group (DCCEPMG) and the dynamic common correlated effects mean group (DCCEMG) estimation methods are stationary. The diagnostic tests’ results show that the models are well specified and the empirical specifications capture the long-run equilibrium relationship. The results of the residual cross-sectional dependence tests for all the models suggest that there is weak cross-sectional dependence in each of them as indicated by the high p-value of the cross-sectional dependence (CD) test. The null hypothesis here is that the residuals of the models exhibit weak cross-sectional dependence. From the results of the various regression models of static common correlated effects estimations presented in Table 13.4, we can observe that the lending interest rate is the only variable that has a significantly negative impact on domestic investments in the short run. All the other variables have a neutral impact on domestic investments in the short run. In the long run, uncertainty has a significantly positive impact on domestic investments at the 10% significance level, a result which is difficult to justify. Expected output on the other hand, is positively and statistically significant in explaining domestic investments at the 1% significance level. The other variables incorporated in the model do not have a statistically significant effect on domestic investments. When we compare the outcomes of the regressions of the various models of the dynamic and static common correlated effects estimators, the dynamic models seem to fare better than the static ones in terms of explaining domestic investments in SSA countries. Stationary/I(0) estimated residual series indicate that a cointegrating relationship exists among the variables meaning that the empirical specification captures the longrun equilibrium relationship and the model is well specified. However, if the residual series is non-stationary/I(1), there is a possibility of the existence of a spurious relationship among the variables. We observe that the residual series obtained from all models estimated by the CCEMG and CCEPMG methods are stationary. We apply the Maddala and Wu (1999) panel unit root test (MW), the Pesaran (2007) CIPS test, and the Pesaran (2003) CADF tests to test the stationarity of the residual series obtained from different models. One of the main differences between the static forms of common correlated effects (CCE) estimators reported in Tables 13.4 and 13.5 is that the results reported in

13 The FDI-Domestic Investment Nexus in SSA

271

Table 13.4 FDI’s impact on domestic investments using static common correlated effects CCEMG

CCEMG0

CCEPMG

CCEPMG0

Explanatory variables

1

2

3

4

D. trade openness

−0.455

−1.423

0.146

0.098

(0.70)

(1.53)

(0.29)

(0.72)

D. FDI

−0.051

0.180

−1.364

−0.407

(0.11)

(0.26)

(1.31)

(0.34)

D. uncertainty

0.061

0.047

−0.024

0.034

(0.05)

(0.07)

(0.03)

(0.04)

0.644

2.680

−0.108

−0.251

(0.64)

(2.73)

(0.76)

(0.96)

D. expected output

0.992

−0.486

3.114

0.267

(1.47)

(5.03)

(2.39)

(1.59)

D. inflation

−0.019

0.063

0.008

0.047

(0.02)

(0.06)

D. official exchange rate

(0.06)

(0.01)

−0.137*

0.792

−0.217*

0.041

(0.08)

(0.70)

(0.11)

(0.09)

D. infrastructure

0.081

–

−0.623

−0.499

(0.06)

–

(0.81)

(0.57)

Trade openness

0.878

1.690

−0.095

0.023

(1.10)

(1.82)

(0.09)

(0.23)

−0.719

−0.038

−0.666

1.844

(1.39)

(0.07)

(0.63)

(17.99)

Uncertainty

0.510*

0.012

0.059

0.262

(0.31)

(0.03)

(0.05)

(2.53)

Official exchange rate

5.525

0.014

−0.705

−1.074

(5.48)

(0.13)

(0.65)

(11.30)

0.037

0.132***

0.252***

0.248

(0.23)

(0.03)

(0.08)

(1.12)

Inflation

0.175

−0.016

0.027

−0.470

(0.34)

(0.01)

(0.03)

(4.55)

Lending interest rate

−0.398

0.019

0.071

−0.063

(0.80)

(0.03)

(0.07)

(0.76)

1.093

–

−0.002

−0.067

(1.25)

–

(0.01)

(0.61)

939

939

939

939

D. lending interest rate

FDI

Expected output

Infrastructure Observations R-squared

0.74

0.89

0.43

0.70

Number of groups/countries

40

40

40

40 (continued)

272

Y. Michael

Table 13.4 (continued) CCEMG

CCEMG0

CCEPMG

CCEPMG0

Residual stationarity

I(0)

I(0)

I(0)

I(0)

Residual CD test

−0.11

−1.17

−0.26

1.44

CD test p-value

0.9149

0.2435

0.719

0.1505

Note Residual non-stationarity: The order of integration of the residuals is determined by applying the Maddala and Wu (1999) panel unit root test (MW), the Pesaran (2007) CIPS test, and the Pesaran (2003) CADF test. Null for MW, CIPS, and CADF tests in series is I(1). Stata routines multipurt for MW; and CIPS and pescadf for CADF tests are used. I(0) indicates that the residual series is stationary; I(1) implies non-stationary; an ambiguous result is denoted by I(1)/I(0). RMSE reports the root mean squared error. CD test: The Pesaran (2004) test is applied to test cross-sectional dependence. The null hypothesis of this test is that the residuals are cross-sectionally independent. Cross-sectional p-values are reported Source Own calculation

Table 13.4 control for cross-sectional dependence while those in Table 13.5 do not. Table 13.5 presents the results of the regression models of dynamic fixed effects (DFE), mean group (MG), and pooled mean group (PMG) estimators. The error correction terms in all the three models are significant with the expected signs and within the range of values that the error term is supposed to assume. Error correction mechanisms (ECMs) are useful for estimating both short-term and longterm effects of a given time series on another time series. The idea of error-correction is related to how the last period’s deviation from a long-run equilibrium, the error, affects its short-run dynamics. Hence, ECMs directly estimate the speed at which a dependent variable returns to equilibrium after a change in other variables. The DFE model shows that trade openness and the lending interest rate positively and significantly affect domestic investments while the official exchange rate negatively and significantly impacts it in the long-run at the 1% significance level. In the short-run, trade openness has a positively significant effect on domestic investments at the 5% level of significance. On the other hand, FDI and the official exchange rate have a negatively significant impact on domestic investments at the 1% significance level. The remaining variables have a neutral impact on domestic investments. From the findings reported in Table 13.5, we can conclude that the static version of CCE estimators provides evidence that FDI’s impact on domestic investments in the host countries is either negative or neutral. The diagnostic test statistics for each of the models indicate that the residuals of the regression are stationary implying the existence of long-run cointegration. Table 13.6 presents the regression results obtained using various forms of the system GMM model using different lags of the dependent and explanatory variables as instruments. The lagged value of domestic investments positively and significantly affects the current value of domestic investments in six out of the seven models. In four of the models, the lagged value of domestic investments affects its current value at the 1% significance level while on two occasions, it positively and significantly affects

13 The FDI-Domestic Investment Nexus in SSA

273

Table 13.5 FDI’s impact on domestic investments using the dynamic common correlated effects estimator models (Dependent variable: domestic investments) DFE

MG

PMG

Explanatory variables

1

2

3

Ec

–

–

–

Trade openness

0.479***

–

–

(0.15)

–

–

FDI

−0.038

−0.119

0.002

(0.03)

(0.10)

(0.00)

−0.082***

0.085

−0.007

(0.02)

(0.12)

(0.01)

Infrastructure

0.001

0.027

–

(0.00)

(0.02)

–

Inflation

−0.000

0.001

0.000

(0.00)

(0.01)

(0.00)

Official exchange rate

0.012***

−5.036

−0.001

(0.00)

(5.05)

(0.00)

SR

–

–

–

Ec

−0.411***

−0.897***

−0.435***

(0.03)

(0.06)

(0.05)

D. trade openness

0.208**

–

–

(0.09)

–

–

D. FDI

−0.037***

0.033

−0.073***

(0.01)

(0.05)

(0.02)

D. official exchange rate

0.158**

0.173

0.447

(0.08)

(0.36)

(0.29)

0.001

−0.008

–

(0.00)

(0.02)

–

D. inflation

−0.000

0.001

−0.001

(0.00)

(0.00)

(0.00)

D. lending interest rate

−0.002

−0.053

−0.001

(0.00)

(0.05)

(0.01)

Constant

0.322

88.955

1.177***

(0.26)

(87.05)

(0.17)

Observations

40

989

989

Number of groups

I(0)

40

40

I(0)

I(0)

Lending interest rate

D. infrastructure

Residual stationarity

Note The prefix ‘D’ stands for differenced. Those variables are meant to capture short-run impacts on domestic investments. ‘ec’ stands for error correction while ‘SR’ is short-run Source Own calculation

274

Y. Michael

Table 13.6 FDI’s impact on domestic investments using the system GMM Explanatory variables Domestic-investment-lag Trade openness FDI Expected output Infrastructure

Model22

Model23

Model33

Model34

Model44

Model45

Model55

1

2

3

4

5

6

7 0.648*

0.851**

0.881***

0.633

0.701***

0.532***

0.437***

(0.33)

(0.29)

(0.49)

(0.19)

(0.12)

(0.09)

(0.34)

0.511**

0.437

1.148

0.586*

−0.037

0.435

0.499

(0.22)

(0.40)

(0.90)

(0.33)

(0.27)

(0.28)

(0.37)

−0.019

−0.017

−0.014

−0.005

0.019

−0.001

−0.020

(0.02)

(0.04)

(0.02)

(0.01)

(0.02)

(0.01)

(0.01)

0.098

0.026

−0.177

−0.224

0.129

0.242

0.287**

(0.36)

(0.28)

(0.36)

(0.33)

(0.29)

(0.26)

(0.12)

−0.002

−0.001

−0.001

0.001

−0.002

−0.003

−0.003**

(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

(0.00) 0.000

−0.000

0.000**

−0.000

−0.000

−0.000

0.000

(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

0.000

−0.000*

0.000

0.000

−0.000

−0.000

−0.000

(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

−0.001

−0.002

−0.002

−0.005

−0.040**

−0.021*

−0.017*

(0.00)

(0.00)

(0.01)

(0.01)

(0.02)

(0.01)

(0.01)

0.020

0.019

0.013

−0.014

−0.075*

−0.001

0.009

(0.02)

(0.02)

(0.03)

(0.03)

(0.04)

(0.03)

(0.02)

−3.875

−2.032

0.266

3.620

−0.351

−5.228

−7.140**

(7.80)

(5.27)

(8.72)

(8.15)

(6.81)

(5.95)

(3.00)

Observations

1090

1090

1090

1090

1090

1090

1090

Number of instruments Number of countries F-test (P-value) Hansen P-value AR(1) P-value AR(2) P-value

16 40 0.000 0.266 0.018 0.546

22 40 0.000 0.106 0.020 0.184

16 40 0.000 0.447 0.073 0.687

22 40 0.000 0.315 0.010 0.724

16 40 0.000 0.312 0.002 0.689

22 40 0.000 0.190 0.006 0.606

16 40 0.000 0.800 0.030 0.371

Uncertainty Inflation Lending interest rate Official-exchange-rate Constant

Note * p < 0.10, ** p < 0.05, *** p < 0.01 Source Own calculation

domestic investments at the 5% and 10% significance levels. Trade openness also has a significantly positive effect on domestic investments at the 5% and 10% significance levels. Uncertainty too has a positive and significant effect on domestic investments at the 5% significance level which is counterintuitive. Inflation and the lending interest rate have a negatively significant effect on domestic investments at the 10% level of significance while the lending interest rate has a statistically significant negative impact on domestic investments at the 5% and 10% significance levels. The regression coefficients from the system GMM estimation method given in Table 13.6 indicate that six coefficients of the FDI variable are negative and only one is positive. Nevertheless, none of these coefficients are statistically significant. All the models pass the diagnostic tests. Failure to reject the null hypothesis of ‘no over-identification problem’ leads us to conclude that the instruments as a

13 The FDI-Domestic Investment Nexus in SSA

275

group are exogenous. As far as the serial correlation test is concerned, our empirical findings lead us to reject the null of the absence of the first-order serial correlation, AR(1), and not to reject the null of the absence of the second-order serial correlation, AR(2), which is expected for consistency of the GMM estimators. The number of instruments is less than the number of cross-sectional units. These findings confirm that the instruments are valid and the models are correctly specified. The unobserved panel level effects are correlated with the lagged dependent variable in linear dynamic panel data models which makes the use of standard estimators inconsistent. Though it is not ideal to use the standard static models for linear dynamic panel data models we apply them to see how they compare with the other models that take into account the dynamism of the system. All the five models in Table 13.7 indicate that the lag of domestic investments is positively and significantly associated with the current value of the variables at the 1% significance level. FDI has a negatively significant effect on domestic investments in three of the five models at the 1% significance level and at 5% in one of the five models. In the only other remaining model, the sign is negative but it is statistically insignificant. Uncertainty has coefficients with a negative sign in four of the five models, as expected a priori, but in the only remaining model, that is, the between effects (BE) model it has a positively significant effect at the 5% significance level which is highly unexpected. Infrastructure and human capital also have statistically significant effects on domestic investments at the 10% level of significance. Inflation has a statistically significant negative effect on domestic investments at the 1 percent significance level. All the other variables have a neutral effect on domestic investments.

13.5 Conclusion and Policy Implications 13.5.1 Conclusion This chapter investigated various dimensions of the relationship between FDI, domestic investments, foreign aid, and economic growth in SSA. It focused on analyzing whether FDI crowds-out or crowds-in domestic investments in SSA. FDI is presumed to influence economic growth by promoting domestic investments and exports and developing human capital, infrastructure, and institutions. Of all these channels, domestic investments are probably the most important through which the host country’s economic growth is influenced by FDI. This happens to be the case because FDI influences employment and incomes more directly through this mechanism than through other channels. Using the flexible accelerator model of investments and the dynamic common correlated effects estimators and other types of dynamic and static estimation methods, the study found that FDI crowds-out domestic investments in SSA countries.

276

Y. Michael

Table 13.7 FDI’s impact on domestic investments using various types of static panel data models (Dependent variable: domestic investments) POLS

RE

FE

BE

PA

Explanatory variables

1

2

3

4

5

Lag of domestic investments

0.690***

0.690***

0.359***

0.979***

0.518***

(0.06)

(0.07)

(0.11)

(0.02)

(0.09)

0.052

0.052

0.447

−0.056

0.168

(0.06)

(0.07)

(0.32)

(0.04)

(0.14)

FDI

−0.012***

−0.012**

−0.017***

−0.004

−0.015***

(0.00)

(0.00)

(0.01)

(0.00)

(0.01)

Uncertainty

−0.005

−0.005

−0.014

0.010**

−0.011

(0.01)

(0.01)

(0.01)

(0.00)

(0.01)

0.006

0.006

−0.128

0.001

0.005

(0.01)

(0.01)

(0.14)

(0.01)

(0.02)

Infrastructure

0.001*

0.001

0.001

0.000

0.001*

(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

Inflation

0.001

0.001

0.006

−0.007**

0.004

(0.01)

(0.00)

(0.00)

(0.00)

(0.01)

0.001

0.001

0.006

−0.000

0.003

(0.00)

(0.00)

(0.00)

(0.00)

(0.00)

Broad money

0.006

0.006

−0.071

0.012

−0.022

(0.03)

(0.03)

(0.06)

(0.03)

(0.04)

Institutional quality index

0.007

0.007

0.108

−0.005

0.026

(0.02)

(0.03)

(0.24)

(0.01)

(0.05)

Human capital

0.145*

0.145*

1.125

0.053

0.152

(0.08)

(0.08)

(0.96)

(0.04)

(0.14)

Trade openness

Official exchange rate

Lending interest rate

Constant Observations F-statistic(p-value) R-squared Overall R-squared Wald-chi-squared Wald-chi-square test (p-value)

0.504**

0.504

−0.252

0.266

0.469

(0.25)

(0.32)

(1.61)

(0.19)

(0.60)

589 0.000 0.503

589 – – 0.503 324.01 0.000

589 – – 0.191 22.65 0.000

589 – – 0.485 310.96 0.000

589 – – – 339.89 0.000

Note * p < 0.10, ** p < 0.05, *** p < 0.01 Source Own calculation

13 The FDI-Domestic Investment Nexus in SSA

277

Specifically, on the basis of Chudik and Pesaran’s (2015) dynamic common correlated effects estimator, the study found that a 1% increase in FDI inflows resulted in a reduction in domestic investments by 0.037 to 0.126% which is significant at the 1% significance level. The finding that FDI crowds-out domestic investments should not be misconstrued to mean that FDI is not important. The argument here is confined to mean that profitable investment opportunities are limited to foreign investors only. MNCs have greater advantages over local investors in that they have better access to investment finance, technology, global markets, and management skills.

13.5.2 Policy Implications To mitigate the adverse effects associated with FDI crowding-out domestic investments, it is essential for policymakers to come up with the right policies that suit their economic realities and the investment climate in host countries. One possible avenue of doing this is by sifting and screening FDI projects to make sure that they do not displace domestic investments made by domestic firms. Instead, the authorities should opt for MNCs that promote linkages with local producers. These linkages could take several forms. They can be related to technology transfers, supplying contracts, training workers, and skill upgrading through apprenticeship and on-the-job training. The preferential treatment afforded to foreign firms is another area that needs to be looked at and revised. Believing that FDI positively affects domestic investments and economic growth, many SSA countries have gone in for fiscal reforms such as tax holidays, low corporate tax rates, and abolishing import duties on intermediate inputs to attract FDI. Here, it might be reasonable to argue that now is the time to abandon this type of discriminatory treatment that favors foreign firms to the detriment of domestic firms. Ideally, domestic investors should enjoy the same privileges and incentives as their foreign counterparts, yet incentives are necessary for competing with alternative destinations of capital and for attracting foreign investments. Acknowldgements I would like to thank my mentor and advisor Professor Almas Heshmati, Jonkoping University, Sweden, for his contributions in improving the quality of this paper. I am also grateful for the help that I got from my former co-supervisor Dr Adane Tuffa, Addis Ababa University, Ethiopia. All errors that remain are mine.

Appendix See Table 13.A1.

0.0332 0.0049 −0.069

−0.0161

0.0530

−0.0048

−0.031

−0.0350

−0.105

0.0109

0.205

Inflation

Interest rate

BM

IQI

−0.00096

−0.058

−0.0039

0.163

0.738

−0.090

−0.027

−0.010

−0.034

1 0.0179

−0.0461

OER

−0.0410

1

Uncertainty

0.106

−0.010

−0.158

−0.0396

1

Infrastructure

−0.0372

−0.0027

0.194

1

Inflation

0.0556

0.0621

1

Intrate

−0.057

1

BM

1

IQI

Note Dominv is domestic investments, OER is the official exchange rate, intrate denotes the lending interest rate, BM is broad money, and IQI is the institutional quality index Source Own calculation

−0.0139

0.117

0.203

−0.0126

0.0528

−0.104

1 −0.0198

Infrastructure

Uncertainty

FDI

OER

0.552

−0.0382

−0.0074

−0.0462

FDI

1

0.441

Openness

Openness

1

Dominv

Dominv

Table 13.A1 Correlation matrix

278 Y. Michael

13 The FDI-Domestic Investment Nexus in SSA

279

References Abdullah, M. (2017). Three Essays on the Macroeconomic Impact of Foreign Direct Investment In Low and Middle Income Countries. PhD thesis, University of Manitoba, Manitoba, Canada. Agosin, M. R. and R. Mayer (2000). Foreign investment in developing countries: Does it crowd in domestic investment? UNCTAD Discussion Paper 146. Geneva. Aitken, B. J. and A.E. Harrison (1999). Do Domestic Firms Benefit from Direct Foreign Investment? Evidence from Venezuela. The American Economic Review, 89: 605–618. Ajayi, S. I. (2006). FDI and Economic Development in Africa. Paper presented at the ADB/AERC International Conference on Accelerating Africa’s Development Tunisia, November 22–24. Ang, J. B. (2009). Financial development and the FDI-growth nexus: The Malaysian experience. Applied Economics, 41(13): 1595–1601. Apergis, N. C.P. Katrakilidis, and N. M. Tabakis (20006). Dynamic linkages between FDI inflows and domestic investment: a panel cointegration approach. Atlantic Economic Journal, 34: 386– 394. Baliamoune-Lutz, M. N. (2004). Does FDI contribute to economic growth? Knowledge about the effects of FDI improves negotiating positions and reduces risk for firms investing in developing countries. Business Economics, 39(2): 49–63. Bencivenga, V. R. and B. D. Smith (1993). Some Consequences of Credit Rationing in an Endogenous Growth Model. Journal of Economic Dynamics and Control, 17(1): 97–122. Bende-Nabende, A. and J. Slater (2003). Private capital formation: Short- and long-run crowding-in (out) effects in ASEAN, 1971–99. Economics Bulletin, 3(27): 1-16. Blackburne, E. F. and M. W. Frank (2007). Estimation of Nonstationary Heterogeneous Panels. The Stata Journal, 7(2): 197–208. Blejer, M. and M.S. Khan (1984). Private investment in developing countries. Finance and Development, 21(2): 26. Blomström, M. S. Globerman, and A. Kokko (1999). The Determinants of Host Country Spillovers from Foreign Direct Investment: Review and Synthesis of the Literature. Working Paper No. 76. Borensztein, E. J. de Gregorio and J.W. Lee (1998). How does foreign direct investment affect economic growth? Journal of international Economics, 45: 115–135. Busse, M. and C. Hefeker (2007). Political risk, institutions and foreign direct investment. European journal of political economy, 23(2): 397–415. Cardoso, E. A. and R.B. Dornbusch (1989). ‘Foreign Private Capital Flows’, In H. B. Chenery and T. N. Srinivasan (eds). Handbook of Development Economics, Vol. 2,. Amsterdam: Elsevier. 1387–1439. Chudik, A. and M. H. Pesaran (2015). Common correlated effects estimation of heterogeneous dynamic panel data models with weakly exogenous regressors. Journal of Econometrics, 188(2): 393–420. Colen, L. M. Maertens, and J. Swinnen (2008). Foreign Direct Investment as an Engine for Economic Growth and Human Development: A Review of the Arguments and Empirical Evidence. IAP P6/06. Working Paper No.16. Crespo, N. I. Proença, and M.P. Fontoura (2008). FDI Spillovers at Regional Level: Evidence from Portugal. Working Paper - 05/08. Economics Research Center, Lisbon University Institute. Daniele, V. and U. Marani (2011). Organized crime, the quality of local institutions and FDI in Italy: A panel data analysis. European Journal of Political Economy, 27(1): 132–142. Daude, C. and E. Stein (2007). The quality of institutions and foreign direct investment. Economics & Politics, 19(3): 317–344. De Mello, L. R. (1997). Foreign direct investment in developing countries and growth: A selective survey. Journal of Development Studies, 34(1): 1-34. Djankov, S. and B. Hoekman (2000). Foreign Investment and Productivity Growth in Czech Enterprises. The World Bank Economic Review, 14(1): 49–64. Du, J. Y. Lu, and Z. Tao (2008). Economic institutions and FDI location choice: Evidence from US multinationals in China. Journal of comparative Economics, 36(3): 412–429.

280

Y. Michael

Erden, L. and R.G. Holcombe (2006). The Linkage between Public and Private Investment: A Co-Integration Analysis of a Panel of Developing Countries. Eastern Economic Journal, 32(3). 479–492. Everaert, G. and T. De Groote (2016). Common correlated effects estimation of dynamic panels with cross-sectional dependence. Econometric Reviews, 35(3): 428–463. Fry, M. (1992). Foreign Direct Investment in a Macroeconomic Framework: Finance, Efficiency, Incentives, and Distortions. Working Paper Series 1141. International Economics Department, The World Bank. Gao, T. (2005). Labor quality and the location of foreign direct investment: Evidence from China. China Economic Review, 16(3): 274–292. Globerman, S. and D. Shapiro (2002). Global foreign direct investment flows: The role of governance infrastructure. World development, 30(11): 1899–1919. Görg, H. and D. Greenaway (2004). Much ado about nothing. Do domestic firms really benefit from foreign direct investment? World Bank Research Observer, 19: 171–197. Haddad, M. and A. Harrison (1993). Are there positive spillovers from direct foreign investment? Evidence form panel data for Morocco. Journal of Development Economics, 42: 51–74. Helpman, E. and P. R. Krugman (1985). Market structure and foreign trade: Increasing returns, imperfect competition, and the international economy. MIT press. Cambridge, MA, Herzer, D. S. Klasen, and D.F. Nowak-Lehmann (2008). In Search of FDI-led Growth in Developing Countries: The Way Forward. Economic Modelling, 25: 793–810. Jansen, W. J. and A. C. J. Stokman (2004). Foreign Direct Investment and International Business Cycle Co-movement Working Paper Series No.401. European Central Bank,. King, R. and R. Levine (1993). Finance and Growth: Schumpeter Might Be Right. Quarterly Journal of Economics, 108(3): 717–738. Kose, M. A. E. Prasad, K. Rogoff, and S.J. Wei (2006). Financial globalization: a reappraisal. IMF Working Paper, 06/189. Lee, M. I. H. M. M. H. Syed, and M. L. Xueyan (2012). Is China Over-Investing and Does it Matter? (No. 12–277). International Monetary Fund. Levine, R. (1997). Financial Development and Economic Growth: Views and Agenda. Journal of Economic Literature, 35(2): 688–726. Levine, R. (2005). ‘Finance and Growth: Theory and Evidence’, In P. Aghion and S. N. Durlauf (eds). Handbook of Economic Growth. Amsterdam: Elsevier, pp. 865–934. Lipsey, R. E. (2002). Home and Host Country Effects of FDI. NBER Working Paper 9293. Lipsey, R. E. (2004). Home- and Host-Country Effects of Foreign Direct Investment. A chapter in Challenges to Globalization: Analyzing the Economics from National Bureau of Economic Research, Inc 333–382. Maddala, G. S. and S. Wu (1999). A comparative study of unit root tests with panel data and a new simple test. Oxford Bulletin of Economics and statistics, 61(S1): 631–652. Mody, A. and A. P. Murshid (2005). Growing up with capital flows. Journal of International Economics, 65: 249–299. Morrissey, O. and M. Udomkerdmongkol (2012). Governance, private investment and foreign direct investment in developing countries. World Development, 40(3): 437–445. Mutenyo, J. (2008). Does Foreign Direct Investment Stimulate Economic Growth in Sub-Sahara Africa? Paper presented at the Paper Presented at the ESRC Development Economics Conference September 17–18, University of Sussex, London, UK. Mutenyo, J. E. Asmah and A. Kalio (2010). Does Foreign Direct Investment Crowd-Out Domestic Private Investment in Sub-Saharan Africa? The African Finance Journal, 12(1): 27–52. Ndikumana, L. and S. Verick (2008). The Linkages Between FDI and Domestic Investment: Unravelling the Developmental Impact of Foreign Investment in Sub-Saharan Africa. Development Policy Review, 26(6: 713–726. Pesaran, H. (2003). A Simple Panel Unit Root Test in the Presence of Cross Section Dependence. University of Cambridge: Cambridge Working Papers in Economics 0346, Faculty of Economics (DAE).

13 The FDI-Domestic Investment Nexus in SSA

281

Pesaran, M. H. (2004). General diagnostic tests for cross section dependence in panels. Pesaran, M. H. (2006). Estimation and inference in large heterogeneous panels with a multifactor error structure. Econometrica, 74(4): 967–1012. Pesaran, M. H. (2007). A simple panel unit root test in the presence of cross-section dependence. Journal of Applied Econometrics, 22(2): 265–312. Pesaran, M. H. Y. Shin, and, R. P. Smith (1997). Estimating Long-run Relationships in Dynamic Heterogeneous Panels. DAE Working Papers Amalgamated Series 9721. Pesaran, M. H. Y. Shin, and R.P. Smith (1999). Pooled Mean Group Estimation of Dynamic Heterogeneous Panels. Journal of the American Statistical Association, 94(446): 621–634. Pesaran, M. H. and R. Smith (1995). Estimating long-run relationships from dynamic heterogeneous panels. Journal of econometrics, 68(1): 79–113. Ramirez, M. D. (1994). Public and private investment in Mexico, 1950–90: An empirical analysis. Southern Economic Journal: 61(1): 1-17. Stiglitz, J. E. (1994). The role of state in financial markets. Proceedings of the World Bank annual conference on development economics, Washington. 1(1): 1–44. Stiglitz, J. E. and A. Weiss (1981). Credit rationing in markets with imperfect information. American Economic Review, 71(3): 393–410. Sun, H. (1998). Macroeconomic Impact of Direct Foreign Investment in China: 1979–96 The World Economy, 21(5): 675-694. Titarenko, D. (2006). The influence of foreign direct investment on domestic investment processes in Latvia. Transport and Telecommunication, 7(1): 76–83. The World Bank Group (2010). Investing Across Borders: Indicators of foreign direct investment regulation in 87 economies. Washington, DC: The World Bank Group. UNCTAD (1999). Foreign Direct Investment and the Challenge of Development. World Investment Report 1999. New York and Geneva: United Nations. UNCTAD (2015). World Investment Report: Reforming International Investment Governance. New York and Geneva: United Nations. Wang, M. (2010). Foreign direct investment and domestic investment in the host country: evidence from panel study. Applied Economics, 42(29): 3711–3721. Wei, S. J. (2000). How taxing is corruption on international investors? Review of economics and statistics, 82(1): 1–11. World Bank (2016). World Development Indicators Database. Zhang, K. H. (2005). Foreign Direct Investment and Economic Growth in China. Illinois State University, USA: Department of Economics.

Chapter 14

China’s Unprecedent Move and Its Repercussion on African Economies: Empirical Evidence from Ethiopia Zerayehu Sime Eshete

Abstract It has been that China has been increasingly moving into the African economies over the last two decades. This has been viewed either positively or negatively from different perspectives. This study’s aim is examining the effects of China’s FDI and trade on the Ethiopian economy. The study uses the autoregressive distributed lag model (ARDL), using quarterly data for the period 2001–17. It finds that Ethiopia’s trade with China and the rest of world is not equally important to the Ethiopian economy. Trade with China is statistically insignificant and does not influence GDP growth in Ethiopia, but trade with the rest of the world (excluding China) is statistically significant. On the contrary, FDI from China is statistically significant and has positive effects on the growth of the Ethiopian economy. However, FDI from the rest of the world has negative implications for growth. This implies that the Chinese have a comparative advantage for the Ethiopian economy as China has been engaged in a fundamental area of developmental activities. Though Chinese FDI is important, its indirect effects are not recommended. From the interaction of moderating variables in the ARDL model, it is seen that Chinese FDI has an indirect negative effect on the Ethiopian economy through a transmission mechanism of facilitating foreign trade, local investments, and labor productivity in Ethiopia. However, it promotes growth via infrastructure and financial development measured by the size of private credit to the private sector as a proportion of GDP. This suggests that the Ethiopian government should pursue a policy considering both the direct and indirect effects of China’s FDI, promote joint ventures with local firms, and provide priority to local employment opportunities. Keywords FDI · Foreign trade · Spillover effects · ARDL model · China and ROW JEL Codes F41

Z. S. Eshete (B) Addis Ababa University, Addis Ababa, Ethiopia e-mail: [email protected] © Springer Nature Singapore Pte Ltd. 2020 J. Wood and O. Habimana (eds.), A Multidimensional Economic Assessment of Africa, Frontiers in African Business Research, https://doi.org/10.1007/978-981-15-4510-8_14

283

284

Z. S. Eshete

14.1 Introduction Chinese engagement in Africa can be divided into three parts: 1850–1950 when there was demand for colonial labor which served the Chinese labor market with small exports; 1960–80 when China’s relationship with Africa became political in the framework of south-south approach. In this period, China was on a series economic reform program and allowed private investment through a gradual liberalization process at special economic zones. The third period is from 1990 to the present in which Chinese companies have moved to all African countries in areas of construction, mining, and oil extraction (Ishengoma and Bo 2014a, b). After the end of the Cold War in the 1990s, the western interest in Africa started declining which created an opportunity for China to influence Africa through trade and FDI ties. This relationship became visible when President Jiang Zemin visited six African countries (Kenya, Egypt, Ethiopia, Mali, Namibia and Zimbabwe) in 1996 and put forward a five-point proposal for long-term development and a cooperative relationship between China and the African countries which led to the creation of the Forum for China-African Cooperation in Beijing in 2000. Sino-African exports increased rapidly from US dollar 5 billion in 2000 and US dollar 104.75 billion in 2018, generating a positive trade balance on average. The import value also increased from US dollar 4.85 billion to US dollars 99.28 billion in the same period. However, trade with USA accounted for US dollars 21 billion in exports and US dollars 34 billion in imports in 2017 leading to a trade deficit.1 China’s move to Africa was a part of the Chinese one belt-one road policy of repossessing the lines of international trade. This policy enables China to connect around 68 countries including Ethiopia through maritime and land roads. China also budgeted around US dollar 1 trillion for building the silk road and silk maritime projects. There are around 900 asphalt roads, railways, airports, ports and electricity generation, oil generation and dissemination projects which are worth mentioning. These are financed by the China Development Bank, the China Ex-IM Bank, the Asia Infrastructure Investment Bank, Silk Road Projects, the State Owned Investment Fund, and private banks in China. The money is partly allocated to projects via support from the Chinese government and through FDI. Some portion is also allocated via loans at concessionary and market rates. As a result, around 23 countries are also highly indebted to China as stated in the report of Center for Global Development, questioning their sovereignty (CGE 2018). Nowadays, it is common to see Chinese companies everywhere in Africa; these are also seen as examples of hard work and commitment. Many of these have helped address the acute infrastructural problems in Africa. African leaders also consider China’s role as an alternative way out from being squeezed by the preconditions of the west for getting foreign grants and loans. China’s extensive moves have been questioned recently and have also received a great deal of attention from the perspective of their long run effects on the performance of the African economies. 1 http://comtrade.un.org/data/;

2018 data from Chinese Customs.

14 China’s Unprecedent Move and Its Repercussion …

285

China contributes to African countries’ GDP through an upward swing in prices by importing natural resources to satisfy the demands of industrialization. However, China’s exports are low cost consumer goods like textiles at the cost of local producers (World Bank 2007). According to (COFACE 2017), China-Africa relations are not balanced. Although bilateral trade has been increasing for more than ten years, the region experienced a trade deficit with China. Ninety percent of the total exports to China concentrated on natural resources while imports were more diversified including capital or manufactured goods, equipment, and machinery. The index established by COFACE indicates that sub-Saharan Africa had a significantly higher export dependency ratio as compared to the other emerging countries and relied more on China than the European Union and United States (COFACE 2017). China’s engagement in Africa has faced strong criticism from African scholars due to unfair advantages taken by Chinese businesses in terms of utilizing scarce local resources where domestic business owners and workers have received less attentions. China’s complex relationship and its policy of non-interference in the affairs of African governments has also been strongly challenged (CFR 2017). China is not only providing finance but almost all the projects are being implemented by Chinese constructors and both the management and workers come from China. Not only this, the major imported materials also come from China. Besides, most of these projects are systematically designed to link and facilitate Chinese exports and international trade in general. This means that projects are intended to facilitate raw material demand in China and promote Chinese exports to these countries. On top of this, most African countries have been suffering with accumulation of debt and they are now in the fear of sovereignty debt. This gives us enough reasons to investigate whether this relationship is benefiting both the sides or only China. China’s focus on natural resources and disregard for good governance where it has a poor record and its focus on poor governance in countries raise many questions about its role in Africa. Chinese ODI (outward direct investments) are indifferent to the rule of law in the host country but positively correlated to political stability. Chinese investments are more visible in countries with poor rule of law from where western investments stay away. Chinese involvement in African infrastructure development is another issue because it is not clear whether Chinese participation in infrastructure is as FDI or not. This involvement can be FDI if the Chinese partners have equity stakes in the infrastructure (Chen et al. 2015). In general, there are two arguments about Chinese investments: Theory of political welfare gives more emphasis on examination of Chinese involvement whether oppressive and distressing effects on the host countries. Contrary to this, the proponents of South-South cooperation believe that China’s increased aid, trade, and investments in Africa are a means of fostering Africa’s self-sufficiency and sustainable development. Therefore, the objective of this research is examining the effect of Chinese foreign direct investment inflows and bilateral trade on economic growth in Africa with a special focus on Ethiopia using time series econometrics in a particular vector error correction method using quarterly data over the period 2001–17.

286

Z. S. Eshete

14.2 An Overview of Chinese Investments and Trade with Ethiopia Ethiopia’s trade exchange relationship with China can be traced back to about 100 BC in the period of Axumite Empire and the Han dynasty respectively. It got strengthened in 1972 due to the agreement on the Ethio-China economic and technical cooperation framework (SHINN 2014). This relationship grew and reached a climax over the last couple of decades. The comparative advantages of working with China include there being no western type preconditions (including liberalization and the Structural Change Program), low rate of foreign grants assistance over a long period, and professional and technical training. To take advantage of these comparative benefits, the Ethiopian government has turned East and offers a range of FDI incentives to China (Geiger and Goh 2014). As a result, China has been Ethiopia’s largest trading partner since 2006 and it has invested US dollars 25 billion over the period 2014–18 in various projects in the country, giving China a chance to lead in the capital and manufacturing industry, followed by India and Turkey (Tibebu 2017). However, the share of Ethiopian exports to USA and China is small at around 4% on average over the same period (Fig. 14.1). In addition, the share of imports from China stood at 7% in 2001 which increased to 33% in 2017. However, the share of exports to China were 1 and 10% respectively in the same period. This implies that Ethiopia is highly dependent on Chinese trade and this may lead to concentration risks in particular on the imports side. To examine this aspect, we need to verify the composition and quality of the imports that Ethiopia gets from China. Imports from China were composed of machinery, electronic equipment, vehicles, iron or steel products, iron and steel, clothing, plastics, rubber, knit or crocheted clothing, and furniture in 2017. Ethiopia’s most valuable exported products were coffee followed by miscellaneous oil seeds and oleaginous fruits. In 2017, the composition of exports was: coffee, tea, spices (33.6%); vegetables (18.8%); oil seeds (15.6%); live trees, plants, cut flower (7.8%); gems, precious metals (4.4%); meat (3.4%); raw hides, skins not China_export

China_Import

USA_Export

USA_Import

16.00

MILLIONS (USD)

14.00 12.00 10.00 8.00 6.00 4.00 2.00 2000

2002

2004

2006

2008

2010

2012

2014

2016

2018

Fig. 14.1 China and USA’s exports and imports to Ethiopia. Source UN COMTRADE statistics

14 China’s Unprecedent Move and Its Repercussion …

287

furskins, leather (2.6%); live animals (2.2%); electrical machinery, equipment (2%); and footwear (1.6%) (Fig. 14.2). Report from the Ethiopian Investment Commission (EIC 2015) also indicate that Chinese investments in Ethiopia accounted for 20% of the total FDI in the period 1998–2015 (Table 14.1). This is a reflection of the political commitment of Ethiopian government to replicate the Chinese development model to advance the Ethiopian economy. Of the total licensed FDI projects, 54% were personalized and providing goods and services for end users. Chinese companies had invested 70% of their money in the manufacturing sector as a priority and strategic sector, followed by the real estate and construction sectors. In line with this, the Chinese government also launched an encouraging policy that promotes Chinese private firms to move to Ethiopia (EIC 2015). A business strategy that allows firms to work through joint ventures with local firms is attributed for the remarkable performance of FDI in Ethiopia (Glans 2014). As per the Ethiopian Share of China Export in Ethiopia

Share of China import in Ethiopia

0.40 0.35

PERCENTAGE

0.30 0.25 0.20 0.15 0.10 0.05 2000

2002

2004

2006

2008

2010

2012

2014

2016

Fig. 14.2 Chierna’s trade share in Ethiopia. Source UN COMTRADE statistics

Table 14.1 Licensed Chinese investment projects in Ethiopia Sector Agriculture Manufacturing

Total

Pre-implemented

Implemented

Operation

17

11

4

2

746

271

92

383

Mining

5

1

1

3

Education

1

0

0

1

Health

12

2

1

9

Hotels

45

16

5

24

Tours

7

0

1

6

Real estate

127

28

12

87

Construction

123

26

26

71

4

1

1

2

1087

356

143

588

Others Total Source EIC

2018

288

Z. S. Eshete Chinese FDI Flow

Chinese FDI Stock

2500 2000 1500 1000 500 0 2002

2004

2006

2008

2010

2012

2014

2016

2018

Fig. 14.3 Chinese FDI flows and stock in Ethiopia (US dollars 000, unadjusted). Source UNCTAD’s Chinese FDI data

Investment Cooperation, over 90,000 temporary and permanent job opportunities have been created for Ethiopians, of which, 45% are in the manufacturing sector, followed by construction (28%), and real estate (23%) over the period 1998–2015. In terms of capital inflows, Chinese FDI in Ethiopia grew remarkably at an annual rate of 24% on average over the period 2014–17. The FDI flows reached at peak in 2016 at US dollars 282 million (Fig. 14.3). Following the influx of Chinese resources into the Ethiopian economy, some studies focused especially on different segments of the economy including output, exports, technology transfers, and efficiency local firms (Geda and Meskel 2008; Tegegne 2006). However, they did not do a time series dynamic analysis due to lack of data for a long period. They put out important findings that show the significant role of Chinese investments and trade though some facts have also emerged based on longer period experiences. For instance, Chinese involvement becomes questionable in light of crowding-out local investments, limited skill transfers, and poor quality of imported goods, exposing the country to inefficient utilization of the scarce foreign currency. Investments and trade are also done in a questionable manner and raise concerns about whether Chinese involvement is for mutual benefits or simply as a means of building its own Silk Road program that will create a market for Chinese inputs and products.

14.3 Review of Related Theoretical Literature Both FDI and trade play an important role in driving growth by enhancing capitalization and competition. FDI narrows down financial gaps, helps transfer knowledge to local firms, fills managerial skill gaps, and is expected to create a competitive environment for local firms. However, the net effects depend on the level of local firms and the policy and legal frameworks in the host country. In the same manner, trade also facilitates the transaction of goods and services from one country to another due to gaps in local demand and supply and also helps in earning adequate foreign

14 China’s Unprecedent Move and Its Repercussion …

289

currency. Trade is also a means of sharing goods and services, as also experiences and knowledge across trading partners on a mutual basis.

14.3.1 FDI Theories There is no universally accepted general theory on the effects of FDI on economic performance in developing countries leading economists taking two fundamental approaches on the effects of FDI on growth depending on the level of development in the host country. For instance, the macroeconomic FDI theories that focus on the flow of capital across national borders include theories of market size, GDP, infrastructure, natural resources, and institutional factors (Lipsey 2004). The market size theory, as developed by Robert Aliber in 1970 (Alibert 1970) explains that FDI emerges due to capital market imperfections and currency differences between the host and source countries. A weaker currency in the host country has a higher capacity to attract FDI and take advantage of the differences in the market capitalization rate. However, the location based FDI theory emphasizes that FDI inflows are high between countries which are geographically, economically, and culturally close reducing the cost of transportation. The institutional FDI fitness theory, as developed by (Wilhelms 1998) states that FDI inflows depend on four fundamental pillars: government, education, market, and socio-cultural fitness assuming that these pillars are used for measuring the host country’s ability to attract and absorb FDI inflows. (Hymer 1976) microeconomic theory of FDI underlines the motives of foreign firms as access to labor, raw material, intangible assets, and economies of scale. The eclectic paradigm/OLI model takes ownership, location, and internalization (this implies the capacity of producing firms to utilize their core competencies) as a precondition for attracting firms towards FDI since the source country’s firms should possess a certain level of advantages over the host country’s firms. This implies that these three conditions should be fulfilled otherwise FDI will not happen. However, this theory is criticized as it lacks an explanation for any subsequent increase in FDI and it incorporates many variables which are difficult to apply practically as it fails to recognize the financial factors in taking decisions about FDI. The oligopolistic theory sets out two motives for choosing a particular country as a host country to set up new facilities for FDI: firms seeking increased access to the host country’s market and wanting to use the relatively abundant factors available in the host country. The third motivation for firms investing in a country is matching a rival’s move. Hymer also suggests that oligopolistic reactions increased with concentration levels and decreased with diversity of products. In general, this theory is valid if there is uncertainty in costs in the host country though Hymer fails to explain the motivation for the first firm to opt for FDI. The theory of exchange rate also underlines the importance of exchange rate that affects the relative advantage of international firms as compared to local firms and depreciation of the host country’s currency is likely to attract FDI inflows due to two

290

Z. S. Eshete

reasons. Firstly, foreign firms have an advantage over domestic firms because of their ability to obtain financing in international capital markets on strong currency terms at lower costs. Therefore, they can pick up projects with higher profitability because they can acquire higher value from the same projects as compared to the local firms due to lower capital costs. So, countries with weak currencies tend to be recipients of FDI while countries with strong currencies tend to be sources of FDI. Secondly, currency depreciation reduces production costs in the host country, thereby making it attractive for FDI that is seeking production efficiency and revenues. In other words, FDI can be a tool for foreign exchange risk hedging with the assumption that foreign firms may be more efficient in hedging risks. Diversification is another motive for FDI but political instability has an adverse effect particularly on FDI and generally on development.

14.3.2 Theory of International Trade A number of attempts have been made to integrate the FDI theory with the theory of international trade. The theories related to international trade like those put forward by (Smith 1776) of absolute advantage and (Ricardo 1817) of comparative advantage depend on labor costs of production in setting up trade among countries. Such classical theories indicate that a country produces and exports commodities in which it has a cost advantage and imports those commodities in which it has a cost disadvantage. Smith argued that international trade was important if a given country had absolute advantage in the cost of production so that country specialized in a commodity in which it had an absolute advantage. Ricardo’s theory suggests that a country engages in the production and export of commodities in which it has comparative advantage and imports products in which it has less comparative advantage in the framework of perfect competition. Modern international trade theories, in particular the (Heckscher and Ohlin 1991) factor endowment theory explains that a country’s exports depend on its resource endowment whether it is a capital abundant or a resource abundant country. If the country is capital abundant it produces and exports capital intensive goods that are relatively cheaper than in other countries. If it is labor abundant it will produce and export labor intensive goods relatively cheaply than other countries. This model has faced strong criticism because it assumes that there is no difference in aggregate preferences between different countries and the only difference is resource endowments. The factor price equalization theorem, the Stolper-Samuelson theorem, the Rybczynski theorem, and the Heckscher-Ohlin theorem are important in the Heckscher-Ohlin model. The factor price equalization theorem states that if factors of production are freely mobile within free trade countries, factor prices will be the same leading to equalized prices of outputs in the countries. This means that the rents and wages will be almost the same across countries. Thus, trade in goods is a perfect substitute for trade in factors. But this theorem fails to consider differences in factor quality, production

14 China’s Unprecedent Move and Its Repercussion …

291

technology, and output prices across countries for the same product. The StolperSamuelson theorem on the other hand indicates that an increase in the price of a good leads to an increase in the price of factors used intensively in its production or industry and a decrease in the prices of other factors. The Rybczynski theorem also emphasizes that an increase in factor endowments leads to an increase in the output of the industries that use such resources intensively. So, a change in resource endowments has a direct effect on the industry using these resources. Finally, this theorem considers small open free trade countries. In general, the Heckscher-Ohlin theorem stresses that capital abundant countries export capital intensive goods whereas labor abundant countries export labor intensive goods. The Helpman-Krugman model combines the monopolistic competition theory with the factor endowment theory because it takes the comparative advantages as a result of trade and economies of scale into account. This model considers intraindustry trade and inter-industry trade flows which means that intra-industry trade depends on specialization due to economies of scale whereas inter-industry trade flows reflect specialization due to countries’ factor endowments. Generally, in the Ricardian model and in its extension the determinants of comparative advantage lie outside the model. In the new trade theory, Paul Krugman brings determinants of comparative advantages into the model.

14.3.3 Review of Related Empirical Literature Most of the existing research emphasizes on the positive contributions of FDI to economic growth depending on the intensity and extent of transferring and catching up with technology. Studies also show that FDI has a comparative advantage over domestic investments and sustainably thus promoting growth (Neto and Veiga 2013). One of the most important factors that makes FDI more effective is the existence of strong macroeconomic and institutional frameworks to boost the absorptive capacities of local firms to internalize the merits of technology transfers. Education, infrastructure, and sound financial development are worth mentioning in this regard as positively influencing FDI and hence helping in achieving a high growth rate through productivity and the accumulation of factors of production (Algiacil et al. 2011; Alfaro et al. 2009; King and Levine 1993). This can be seen in the experiences of countries with strong financial sector performances that earn more from FDI via better allocation of resources. On the macro side, higher inflation in the monetary policy and debt accumulation in the fiscal policy also increase uncertainty and slow down economic growth (Lensink and Sterken 1999). The experience of Bangladesh has also been a witness for the existence of long run and positive relationship between GDP and FDI over the period 1972 to 2013 and the Granger causality test also confirm that there is unidirectional causal relationship that runs through GDP to FDI (Ibrahim and Mohammad 2015).

292

Z. S. Eshete

Aga (2014), examined the effects of FDI on economic growth in Turkey using time series data by doing the cointegration and Granger causality tests. The results showed that there was no long-run relationship and there was no Granger Causality relationship between the variables. Another study which examined the impact of FDI on economic growth in China and India using multiple regression models found that India was less affected by FDI than China because China used more FDI than India (Gaurav and Aamir 2011). Another study investigated analyzes the link between foreign direct investment and economic growth in 65 countries using macro panel data model and found that there is variation in their long run relationship taking into account country specific characters and also indicated that a unidirectional causality from FDI to GDP (Sahraoui et al. 2015). An empirical study by (Athukorala and Sarath 2003) examined the relationship between FDI and GDP using the Granger causality test for the Sri Lankan economy showed that FDI inflows did not exert any independent influence on economic growth; rather the causality was from GDP growth to FDI. Shimul (2009), re-examined the relationship between FDI and GDP using the time series data analyzed through ARDL and Granger causality tests. The results revealed that all variables used in the study were cointegrated in the long run and there existed unidirectional Granger causality from FDI to GDP. Masipa (2014) examined the impacts of FDI on economic growth and employment in South Africa by using the unit root test for stationary, the Johansen cointegration test for the existence of long run relationships, and the Granger causality test to establish the causality relationship. The findings show that causality ran from FDI to GDP and from FDI to employment and there was a positive long run relationship between FDI, GDP, and employment in South Africa. Foreign direct investment inflows are expected to affect economic growth positively through technology transfers, know-how, and capital formation. They also negatively affect the host economy through crowding-out of domestic industries that leads to an exploitation of the host country’s economy. So, FDI inflows have both positive and negative spillover effects. These different effects of FDI inflows are a function of the host country’s absorptive capacity. Chinese MNCs use imported inputs from China making it the number one exporter of capital and intermediate goods into Ethiopia. So, this bilateral trade enables Chinese foreign direct investors to have low costs of production as they use imported inputs which makes them strong enough to compete even with other MNCs in Ethiopia. So, China’s bilateral trade with Ethiopia plays a big role in Chinese direct investments in Ethiopia. On the other hand, the World Bank (WB 2012) conducted a survey of Chinese FDI in Ethiopia in 2012 and indicated that FDI inflows increased at splendid rate and such inflows were growing at a rate of 10% over the last half decades. This is just followed in response to the incentives provided by the Ethiopian government. However, Chinese FDI into Ethiopia is also constrained because of the local conditions in the country. However, the World Bank report failed to contextualize this and make it compatible with the reality of Chinese engagement in Ethiopia. Further, studies undertaken in Ethiopia are case studies and follow a qualitative method and are not supported by a statistical method of data analysis (Geiger

14 China’s Unprecedent Move and Its Repercussion …

293

and Goh 2012). For instance, Desta (2009) did four China-Ethiopia investment case studies, emphasizing the impact of Chinese investment on human resources, management, exports, technology transfers, and environment. The focus on the specific issue of human resources concluded that “given the Chinese investors in Ethiopia are unfamiliar with cultural makeup of the local situation and the Ethiopian labor laws, Ethiopian employees seemed to be in charge of the human resources management in these companies. Therefore, Geiger and Goh concluded that it difficult to draw appropriate, more general policy conclusions.

14.4 Method 14.4.1 Model Specification To understand the effect of Chinese FDI and bilateral trade on economic performance of Ethiopia, we specify the model on the basis of the Solow growth model which this study extends to reflect the Ethiopian economy: Y = f(L, K)

(14.1)

where Y is output or GDP produced by K units of capital and L units of labor at time t, L is the economically active labor force, and K is physical capital. According to macroeconomic foreign direct investment theories, FDI is capital flows across borders which affect the host country’s economy through increasing capital stock. In this model, foreign direct investments are divided into foreign direct investment inflows from China and from the rest of the world. Y = F(Labour, Investment, FDIchi , FDIROW )

(14.2)

where FDIchi : FDI inflows from China and FDIw : FDI inflows from the rest of the world. International trade theories incorporate trade flows between two countries. Through this interaction, a country gains or loses. In this chapter, bilateral trade between Ethiopia and China is a major issue to be investigated. Thus, exports to China and imports from China are variables included in the model. Y = f(labour, Investment, FDIchi , FDIROW , TradeChina , TradeROW )

(14.3)

where tradechi denotes Ethiopia’s trade with China and tradeROW denotes Ethiopia’s trade with the rest of the world, excluding China. Another crucial issue is examining whether FDI inflows promote growth with good institutional and macroeconomic frameworks. The open trade policy facilitates greater capital flow openness. Hence, FDI inflows combined with higher trade

294

Z. S. Eshete

flows spur economic growth. Finally, the countries with higher educational attainments can better exploit the positive externalities of FDI. FDI inflows combined with educational attainments promote growth. Y = f(labour, investmetn, FDIchi , FDIROW , tradechi , tradeROW , Interactive terms) (14.4) The interactive terms consist of moderating variables of Chinese FDI with local investments, infrastructure, financial development, and labor productivity. The study used secondary data collected from ADB, WDI, ITC, and UNCTAD for the period 2001–17. It generated quarterly data by applying some decomposition and smoothing techniques. Even though there are many decomposition methods, this study used the linear and quadratic methods.

14.5 Results and Discussion 14.5.1 Unit Root Test Results Testing for the stationarity of economic time series is crucial since standard econometric methodologies assume stationarity in the time series while they are, in fact, non-stationary thus leading to inappropriate statistical tests and erroneous and misleading inferences. The augmented Dickey-Fuller (ADF) test was used for determining whether the variables followed a non-stationary trend and were in fact integrated of the order one or zero. The null hypothesis for the test is that the series contained a unit root (that is, non-stationary) and the alternative is that the series did not contain a unit root (that is, the series was generated by a stationary process). In this case, if the unit root exists on the level of the series in the model, the variables are first-differenced to correct for the evidence or existence of unit roots. The unit root diagnostic test is given in Table 14.2. The ARDL method can be applied to time series data irrespective of whether the variables are I(0) or I(1) (Pesaran and Pesaran 1997)as it provides unbiased estimates of the long-run model and validates the t-statistics even when some of the regressors are endogenous. For these reasons this study used the ARDL technique to examine the macroeconomic dynamic relationship GDP and explanatory variables. It also re-parametrized the general model into an error correction model in this sequence of testing and estimating for two reasons: firstly, the error correction model is easier to interpret in economic terms as the distinction between short term adjustment responses and long-term relationships between the variables are easier to see in this form. Secondly, there are some statistical advantages in working with ECM as it satisfies all misspecification test criteria, is parsimonious, and parameterized in line with economic interpretations.

14 China’s Unprecedent Move and Its Repercussion …

295

Table 14.2 Results of Augmented Dickey-Fuller (ADF) test Level t-Statistic LNGDPQ

Difference Prob.

t-Statistic

Prob.

0.598677

0.9887

−3.11849

0.0300* 0.0153*

0.107600

0.9639

−3.383657

LNLABQ

−1.761393

0.3963

−8.928237

0.0000*

LNTRADECHINA

−1.891130

0.3345

−8.298797

0.0000*

LNTRADE

−1.357241

0.5980

−7.759374

0.0000*

LNFDIQ

−0.396000

0.9029

−3.621406

0.0080*

LNFDICHQ

−1.566048

0.4942

−8.611849

0.0000*

LNTARIFFQ

−1.181158

0.6779

−4.147675

0.0016*

LNEREQ

−0.614483

0.8589

−2.192072

0.2113

LNINVEST

Note *significant at the 5% level Source EVIEWS

14.5.2 ARDL Bounds Tests for Cointegration Even though the ARDL framework does not require pre-testing the variables, the unit root test can tell us whether or not the ARDL model should be used. The results of the unit root test showed that there was a mix of I(0) and I(1) of the underlying regressors. So, the appropriate technique is the ARDL approach to cointegration. As the first step in the ARDL model, this study looked at the long run relationships between the variables by carrying out a partial F-test. The calculated F-statistics for the cointegration test are given in Table 14.3. The calculated F-statistics (F-statistics = 13.42) are higher than the upper bound critical value at the 1, 5, 10% levels of significance. This implies that the null hypothesis of no cointegration can be rejected at 1%, 5%, and 10% levels and therefore, there is evidence to support a long run relationship between the variables. Table 14.3 The Cointegration bounds tests Test statistic

Value

K

F-statistic

13.42985

13

Significance (%)

I0 bound

I1 bound

10

4.14

3.79

5

4.85

4.41

2.5

5.52

5.15

1

6.36

4.19

Critical value bounds

Source EVIEWS

296

Z. S. Eshete

The bounds test shows the existence of a long run relationship among variables in each equation in a situation when the sample size T is small and it is applied to a mix of stationary and non-stationary time series. The estimation process sometimes incorporates deterministic terms such as intercept and time trends to improve the interpretability of the model’s coefficients and for addressing the issue of outliers by ensuring that the variables are normally distributed.

14.5.3 The Long Run Dynamics of the ARDL Model The regression model is not spurious as R-squared is less than the DW statistic (0.99 < 2.38). The long-run ARDL and elasticities model selected on the basis of AIC, are given in Table 14.4. From the long run part of the error correction model, the study found that investments, labor, trade with ROW, and FDI from China were statistically significant and had a positive effect on the growth of the Ethiopian economy. However, FDI from the rest of the world (excluding Chinese FDI) had negative implications for growth. The factors tariff rate, effective exchange rate, and trade with China were statistically insignificant and did not influence GDP growth in Ethiopia (Table 14.5). Taking the interaction effect and moderating variables, Chinese FDI had a negative effect on growth through trade, local investments, and labor productivity in Ethiopia. However, it promoted growth via infrastructure development and financial development measured by the size of private credit to the private sector as a proportion of GDP. Table 14.4 Model criteria

R-squared

0.999999

Mean dependent var

22.14087

Adjusted R-squared

0.999998

S.D. dependent var

0.456715

S.E. of regression

0.000565

Akaike info criterion

−11.81559

Sum squared resid

1.15E-05

Schwarz criterion

−10.82029

Log likelihood

419.9145

Hannan-Quinn criter

−11.42230

F-statistic

1463175.

Durbin-Watson stat

2.384356

Prob (F-statistic)

0.000000

Source EVIEWS

14 China’s Unprecedent Move and Its Repercussion …

297

Table 14.5 The long run dynamics of the ARDL model Long run coefficients Variable

Coefficient

Std. error

LNINVEST

0.090802

0.014303

6.348298

LNLABQ

1.120346

0.029607

37.840630

0.0000

LNTRADECHINA

0.000013

0.001584

0.008423

0.9933

LNTRADE

t-Statistic

Prob. 0.0000

0.079407

0.014427

5.503930

0.0000

−0.004421

0.001344

−3.288913

0.0023

1.607004

0.016521

97.267876

0.0000

LNTARIFFQ

−0.004811

0.006491

−0.741123

0.4634

LNEREQ

−0.003609

0.004241

−0.850894

0.4005

LNFDICHINAINVESTMENT

−0.006040

0.000893

−6.763634

0.0000

LNFDICHINATRADE

LNFDIQ LNFDICHQ

−0.004572

0.000875

−5.226117

0.0000

LNFDICHINAFD

0.003009

0.000369

8.148424

0.0000

LNFDICHINAINFR

0.001171

0.000415

2.819488

0.0078

−1.885937

0.017140

−110.028777

0.0000

0.540028

0.484256

1.115171

0.2722

LNCHINALPQ C Source EVIEWS Output

14.5.4 The Short Run Dynamics of the ARDL Model As can be seen in Table 14.6, the short run estimated coefficients give results that are almost consistent with the long run results at different lag lengths. However, the higher tariff rate puts significant and negative pressure on GDP growth in the short run, but it is statistically insignificant in the long run. The error correction term (EC) that captures the speed of adjustment of a given dependent variable shows that a variable moves towards its long run equilibrium position if there is a single shock. It also shows how quickly the variable converges to restore equilibrium in the dynamic model and it should have a statistically significant coefficient with a negative sign. The highly significant EC term further confirms the existence of a stable long-run relationship. It also implies that the deviation from the long run equilibrium level of the current period is corrected by ECM percentages in the next period to bring back equilibrium (Banerjee et al. 1998). It is well known that the ECM coefficient is theoretically expected to be between −1 and 0. If there is positive ECM, the process does not converge in the long run, leading to problems in the model’s specification, data issues including structural breaks, and the presence of autocorrelation. Therefore, the result shows that the ECM term is statistically significant with a negative sign of 0.45, implying that the speed of adjustment is 45% per quarter so that 45% of shock in the system of equation can be corrected in a quarter.

298

Z. S. Eshete

Table 14.6 The short run dynamics of the ARDL model Cointegrating form Variable D(LNINVEST)

Coefficient

Std. error

t-Statistic

Prob.

0.175281

0.011768

14.894194

0.0000

−0.003653

0.002164

−1.687958

0.1001

D(LNLABQ)

1.105686

0.078699

14.049633

0.0000

D(LNTRADECHINA)

0.000006

0.000716

0.008424

0.9933

D(LNTRADE)

0.035876

0.007469

4.803524

0.0000

−0.003317

0.000798

−4.159601

0.0002

D(LNFDIQ(−1))

0.001053

0.000564

1.868045

0.0699

D(LNFDICHQ)

1.529880

0.019484

78.519250

0.0000

D(LNTARIFFQ)

−0.019002

0.004259

−4.462129

0.0001

D(LNTARIFFQ(−1))

−0.006177

0.003375

−1.830310

0.0755

0.001700

0.003890

0.436929

0.6648

D(LNEREQ(−1))

−0.005758

0.003423

−1.682372

0.1012

D(LNFDICHINAINVESTMENT)

−0.010875

0.000764

−14.225003

0.0000

0.000402

0.000131

3.062410

0.0041

−0.002066

0.000452

−4.572909

0.0001

D(LNINVEST(−1))

D(LNFDIQ)

D(LNEREQ)

D(LNFDICHINAINVESTMENT(−1)) D(LNFDICHINATRADE) D(LNFDICHINAFD)

0.000421

0.000361

1.166663

0.2510

−0.001268

0.000404

−3.140526

0.0034

0.000529

0.000221

2.391119

0.0221

D(LNCHINALPQ)

−1.766930

0.023966

−73.726191

0.0000

CointEq(−1)

−0.451802

0.056778

−7.957323

0.0000

D(LNFDICHINAFD(−1)) D(LNFDICHINAINFR)

Note ***, **, and * significant at the 1, 5, and 10% level Source EVIEWS

On top of the speed of adjustment, the ECM term allows the long run behavior of the endogenous variables to converge to their long run equilibrium relationship while allowing a wide range of short run dynamics. The highly significant error correction term further confirms the existence of a stable long run relationship. These estimates provide additional evidence of the complicated and often inconsistent dynamics that exist between trade balance and its main determinants. Some coefficients of ECM in the model are statistically significant and negative as expected and support the validity of the equilibrium relationship between the variables in the long run. This also indicates a high rate of convergence to the equilibrium, which implies that a deviation from the long-term equilibrium is corrected by 45% over each quarter.

14 China’s Unprecedent Move and Its Repercussion …

299

Table 14.7 Model diagnostic tests No.

Test

1

Breusch-Godfrey serial correlation LM test

Test statistic

2

Heteroskedasticity Test: Breusch-Pagan-Godfrey

3

Ramsey’s RESET test/functional form

3.249703

0.0801

4

Normality test

0.6246

0.7317

3.818024 26.48474

Prob. value 0.0507 0.5995

Source EVIEWS

14.5.5 Diagnostic Test’s Results Post estimation and diagnostics tests, the autocorrelation test, heteroskedasticity test, stability test, specification test, normality test, and omission of relevant variable tests are satisfied. Note that the normality test does not contribute to bias or inefficiency in the regression models. Stability test using CUSUM and CUSUMSQ is confirmed as these are satisfied (Table 14.7).

14.5.6 Stability Test The stability of the long-run coefficients is used to form the error-correction term in conjunction with short run dynamics. Some of the problems of instability could stem from inadequate modelling of the short-run dynamics characterizing departures from the long run relationship. Hence, it is important to incorporate the short run dynamics for the consistency of the long run parameters. A structural stability test for parameter consistency was carried out to ascertain the best fit ARDL model using recursive estimates of the cumulative sum of recursive residuals (CUSUM) and the cumulative sum of squares of the recursive residuals (CUMSUMQ). Note that the straight lines graph in Fig. 14.4 represent critical bounds at the 5% significance level. The stability of the long run and short run parameters are examined for the robustness of the results. The CUSUM test is based on the cumulative sum of recursive residuals based on the first set of n observations. It is updated recursively and is plotted against the break points. If the plot of CUSUM statistics stays within the 5% significance level, then the estimated coefficients are said to be stable. A similar procedure was used for CUSUMSQ that is based on the squared recursive residuals. A graphical representation of the CUMSUM and CUSUMSQ statistics are given in Fig. 14.4. Based on the plots of both the CUMSUM and CUSUMSQ statistics it can be observed that the variables remain within the 5% critical bound which supports the null hypothesis that all coefficients are stable. Therefore, these statistics confirm the stability of the long run coefficients of the estimated equations.

300

Z. S. Eshete

20

1.4

15

1.2 1.0

10

0.8

5

0.6 0

0.4

-5

0.2

-10

0.0

-15

-0.2 -0.4

-20 2009 2010 2011 2012 2013 2014 2015 2016 2017 CUSUM

5% Significance

2009 2010 2011 2012 2013 2014 2015 2016 2017 CUSUM of Squares

5% Significance

Fig. 14.4 Plot of cumulative sum of recursive residuals and sum of square of recursive residuals

14.6 Policy Implications Both FDI and trade are needed for driving long run economic growth. This is confirmed by the experiences of different countries. Their particular effects depend on the countries’ situation. The direct and indirect effects also score differently, leading us to examine the different transmission mechanisms. In particular, the spillover effects of FDI should be examined in depth as FDI affects domestic economies through several channels. It also depends on the host country’s absorptive capacity measured by infrastructure development, an efficient financial system, education, and institutions. And, the effects of trade openness rely on the competitiveness capacities of local firms and terms of trade position of Ethiopia. Among international trading and investment partners, China has been receiving great attention from African leaders, who turn to China more often in response to less preconditions set by China. However, contentions are emerging in the frameworks of politics and economics. This study examined the effect of China’s current moves in the Ethiopian economy since it became its principal trading partner and had investments worth US dollars 25 billion over the period 2014–18, making China indispensable for the Ethiopian economy. To do so, the study used the ARDL model and satisfied the existence of long run cointegration among the variables and fulfilled all relevant diagnostic tests. In the long run, the study found that local investments (both private and public investments), and an active labor force were important for the Ethiopian economy for improving its GDP. Trade with China and the rest of world are not equally important: trade with China is statistically insignificant and does not influence GDP growth in Ethiopia, but trade with the rest of the world (excluding China) is statistically significant. This might be due to the quality of tradable goods, not the quantity of tradable goods associated with ROW. FDI from China is statistically significant and has a positive effect on the growth of Ethiopian economy. However, FDI from the rest of the world (excluding Chinese FDI)

14 China’s Unprecedent Move and Its Repercussion …

301

has negative implications for growth. This implies that Chinese FDI has a comparative advantage for the Ethiopian economy as it has been engaged in fundamental areas of development activities. Though Chinese FDI is important, its indirect effects are not recommended. From the interaction and moderating variables in the ARDL model, we found that Chinese FDI had an indirect negative effect on the Ethiopian economy through a transmission mechanism of facilitating foreign trade, local investments, and labor productivity in Ethiopia. This means that it lowers the role of trade, local investments, and labor productivity in the economy. Chinese companies maintain that they are unable to find skilled labor in domestic markets and are forced to import their own professionals which affects the host economy through negative spillover effects. It is obvious that workers with more educational qualifications earn more and are able to absorb technological changes which increase economic growth. Chinese industries create job opportunities for the local population but the types of jobs are very low quality with low wages for the unskilled labor force and uses its own imported inputs. Chinese companies are more favored in policy as a result of which local firms are unable to face stiff competition from China’s huge companies. Thus, cheap unskilled labor together with cheap imported inputs from China are able to control and exploit the domestic economy which results in the crowding-out of local industries, lowering the role of local investments. However, FDI promotes growth via infrastructure and financial development measured by the size of private credit to the private sector as a proportion of GDP. This is the good part contributed by Chinese FDI to boost infrastructural development which attracts foreign direct investments and helps in fueling financial development and intermediation roles in the Ethiopian economy. Therefore, the Ethiopian government should develop policies that exploit the relationship with China considering both its direct and indirect effects. It is also recommended that there must a policy that does not allow a crowding-out of local firms by restricting foreign investments through joint ventures with local firms so that they can share all the benefits of the policy, technology transfers, knowledge diffusion, and lead to risk reduction. The job market created through Chinese FDI should provide priority to local labor to enhance their productivity and fair competition to sustain the China-Ethiopia friendship on a mutual basis. Further, the Ethiopian government should be aware of the costs of poor quality tradable products as a factor of scarce foreign currency reserves and should develop a standard that eliminates unproductive foreign currency consuming tradable items. Moreover, there must be diversification of exports to reduce dependency on primary agricultural commodities and stabilizing price fluctuations in the international market. In general, the Ethiopia government should reduce unnecessary foreign dependence on China by designing a way-out that helps it to play the game well. Acknowledgements My Special thanks to Abel Teshome who helped with data management.

302

Z. S. Eshete

References Aga, A. AK. (2014). The impact of foreign direct investment on economic growth: A case of Turkey 1980–2012. International Journal of Economics and Finance, 6, 71–84. Alfaro, L., Kalemli-Ozcan, S., & Sayek, S. (2009). FDI, productivity and financial development. The World Economy, 111–135. Algiacil, M., Cuadros, A., & Orts, V. (2011). Inward FDI and growth: The role of macroeconomic and institutional environment. Journal of Policy Modeling, 481–496. Athukorala, & Sarath, R. (2003). Foreign direct investment in crisis and recovery: lessons from the 1997–1998 Asian crisis. Australian Economic History Review, Economic History Society of Australia and New Zealand, 197-213. Banerjee, A., J. J, D., & R, M. (1998). Error-Correction Mechanism Tests for Cointegration in a Single-Equation Framework. Journal of Time Series Analysis, 19(3), 267–283. CFR. (2017, July 12). China in Africa. Retrieved 2018, from Council on Foreign Relation: https:// www.cfr.org/backgrounder/china-africa. CGE. (2018). Examining the Debt Implications, Center for Global Development. Washington DC 20036: Center for Global Development. Chen, W., Dollar, D., & Tang, H. (2015). China’s direct investment in Africa: Reality versus myth. Brookings Education. COFACE. (2017). CHINA’S AMBITIONS IN SUB-SAHARAN AFRICA: EFFORTS TO REBALANCE BI-LATERAL RELATIONS STILL NEEDED. Coface SA. is listed on Euronext Paris – Compartment B. Desta, A. (2009). https://www.ezega.com/News/NewsDetails/1009/Chinese-Investments-inEthiopia-Political-Warfare-Operations-or-south-south-cooperation. Retrieved from Ezega: https://www.ezega.com/News/NewsDetails/1009/Chinese-Investments-in-Ethiopia-PoliticalWarfare-Operations-or-south-south-cooperation. EIC. (2015). Report and Database on Chinese Investment in Ethiopia. Addis Ababa: Ethiopian Investment Commission. Gaurav, A., & Aamir, K. (2011). Impact of FDI on GDP: A Comparative Study of China and India. International Journal of Business and Management. Geda, A., & G. Meskel, A. (2008). China and India’s Growth Surge: Is it a curse or blessing for Africa? The Case of Manufactured Exports. AFRICAN DEVELOPMENT REVIEW, 163–199. Geiger and Goh. (2012). Chinese FDI in Ethiopia. DC: World Bank. Geiger, M., & Goh, C. (2014). Chinese FDI in Ethiopia: A World Bank survey. World Bank. Glans, N. (2014). Chinese OFDI and Private Companies in Ethiopia—Industrialization and Employment Opportunities. LUND UNIVERSITY LIBRARIES. Heckscher, E., & Ohlin, B. (1991). Theory of International Trade. Hymer. (1976). Determinants of Foreign Direct Investment Inflows to ECOWAS Member Countries: Panel Data Modelling and Estimation. he International Operations of National Firms: A Study of Foreign Direct Investment., MIT Press, Cambridge, MA. Ibrahim, K., & Mohammad, R. (2015). Foreign Direct Investment (FDI) and Gross Domestic Product (GDP) in Bangladesh: A Cointegration Analysis. Journal of Economics and Sustainable Developmen. Ishengoma, & Bo, S. (2014a). Different reports about Chinese investment in Africa. Retrieved from Journal of mass communication and Journalism: https://www.omicsonline.org/open-access/todiscuss-different-reports-about-chinese-investment-in-africa-and-cif-2165-7912.1000214.php? aid=31075. Ishengoma, J., & Bo, S. (2014b). Different reports about Chinese investment in AfricDifferent Reports about Chinese Investment in Africa. Journal of mass communication and Journalism, https://www.omicsonline.org/open-access/. King, R., & Levine, R. (1993). Finance and Growth: Schumpeter Might Be Right. The Quarterly Journal of Economics, 717–737.

14 China’s Unprecedent Move and Its Repercussion …

303

Lensink, R., & Sterken, E. (1999). Does Uncertainty affect Economic Growth? Weltwitschaftiches Archive, 379–396. Lipsey, R. (2004). Challenges to Globalization: Analyzing the Economics. EconPaper, 333–382. Masipa, T. (2014). The Impact of Foreign Direct Investment on Economic Growth and Employment in South Africa: A Time Series Analysis. Mediterranean Journal of Social, Vol5, No. 5. Neto, D., & Veiga, F. (2013). Financial globalization, convergence and growth: The role of foreign direct investment. Journal of International Money and Finance. Pesaran, M., & Pesaran, B. (1997). Interactive Econometric Analysis. Oxford: Oxford University Press. Ricardo, D. (1817). The Ricardian Law of Comparative Advantage. Sahraoui, A., Belmokaddem, M., & Guellil, M. (2015). Causal Interactions between FDI, and Economic Growth: Evidence from Dynamic Panel Co-integration. Procedia Economics and Finance, 276–290. Shimul, S. (2009). An Examination of FDI and Growth Nexus in Bangladesh: Engle Granger and Bound Testing Cointegration Approach. BRAC University Journal, 69–76. SHINN, D. (2014, Jun 11). Ethiopia and China: When Two Former Empires Connected. Retrieved May 08, 2018, from International Policy Digest: https://intpolicydigest.org/2014/06/11/ethiopiaand-china-when-two-former-empires-connected/. Smith, A. (1776). An Inquiry into the Nature and Causes of the Wealth of Nations. Tegegne, G. (2006). Asian Imports and Coping Strategies of Medium, SMALL and Micro Firms: The case of Footwear Sector in Ethiopia. Journal of Modern African Studies, 647–679. Tibebu, H. (2017, February 18). allAfrica. Retrieved September 20, 2017, from https://www. allafrican.com/stories/201702180271.html. WB. (2012). Chinese FDI in Ethiopia: A World Bank Survey. World Bank. Wilhelms, S. (1998). Foreign Direct Investment and Its Determinants in Emerging Economies. United States Agency for International Development, Bureau for Africa, Office of Sustainable Development. United States Agency for International Development, Bureau for Africa, Office of Sustainable Development. World Bank. (2007, April 20). The growing relationship between China and Sub-Saharan Africa: macroeconomic, trade, investment, and aid links. Retrieved 2018, from World Bank: http://documents.worldbank.org/curated/en/617301468003942566/The-growing-relationshipbetween-China-and-Sub-Saharan-Africa-macroeconomic-trade-investment-and-aid-links. Z.Alibert, R. (1970). A theory of direct foreign investment. pp. 17–34.

Chapter 15

Does Free Trade and Institutional Quality Affect the Economic Community of the West African Trading Bloc? Luqman Olanrewaju Afolabi

Abstract Regional integration is a very significant concern for the Economic Community of West African States (ECOWAS). Even though a lot of empirical work has been done on regional trade including its determinants, very little work has been done to examine the effects of institutional quality and free trade predominantly concentrating on ECOWAS, which is one of the most corrupt, politically unstable, and less governed regions in the world. This research empirically examines the effects of institutional quality and free trade on ECOWAS during 1996–2017 using the gravity model. It also uses the Poisson Pseudo-Maximum Likelihood (PPML) method to establish the relationship between the effects of institutional quality and free trade on ECOWAS. As suggested by Baldwin and Taglioni (2006) and Baier and Bergstrand (2007) it incorporates a multilateral resistance trade term in the model The results confirm that ECOWAS is a regional force where feeble governance and high political instability have hindered trade performance over the years. Historical/colonial ties spurred ECOWAS trade flows and our results show the fundamental significance of no-log (zero trade) for export trade flows and the need to properly account for heterogeneity and endogeneity bias when the trade policy effect is empirically examined. The study recommends that ECOWAS should address the poor institutional quality within the region and proceed to the next stage of integration to enhance economic growth and development. Keywords ECOWAS · Free trade · Gravity model · Multilateralism · Institution JEL Codes F150 · F1 · F120 · F130 · F5

L. O. Afolabi (B) Faculty of Economics and management Science, Department of Economics and Statistics, Kabale University Uganda, Plot 364 Block 3 Kikungiri Hill, Kabale Municipality, Uganda e-mail: [email protected]; [email protected] © Springer Nature Singapore Pte Ltd. 2020 J. Wood and O. Habimana (eds.), A Multidimensional Economic Assessment of Africa, Frontiers in African Business Research, https://doi.org/10.1007/978-981-15-4510-8_15

305

306

L. O. Afolabi

15.1 Introduction One of the most significant phenomenon in the global economy in the last 20 years has been the tremendous growth in the number of regional trade agreements (RTAs). Global trade is the most significant beneficial economic factor, pushing economies into trade integration. Worldwide trade flows are seen as a signal linking economic centers of the world and global trade is seen as a solution for achieving economic growth and improving societies, including the mitigation of poverty and hunger (Winters 2004). Africa’s contribution to world trade has remained comparatively low (Bouet et al. 2008). According to UNCTADstat (2018), the total annual trade growth rate for ECOWAS in 2000–17 stood at 26.37–25.27% while in 2014–16, it recorded a negative annual trade growth rate of −4.99, −36.46, and −17.62 respectively. Table 15.1 shows the contributions of select regional trading blocs across the globe; it also shows that ECOWAS’s trade performance in terms of export and import contributions has not improved over the years. A comparison of ECOWAS with other regional trading blocs shows that the other areas performed better in their export and import levels from 1980 to 2015 (UNCTADstat 2016). International researchers have shown that institutional quality is the main factor in determining global trade and patterns of production. According to Anderson and Marcouiler (2002), corruption remains the key obstacle when undertaking business in global markets. Fosu (2011), found that a major factor affecting trade performance was institutions in the respective countries. Hence, the essential forces driving trade development such as institutional quality, financial development, infrastructure, and the social environment for investments must be encouraged to enhance trade and development. In this regard, poor institutions, bureaucratic hassles, and other factors act as inhibitors to trade (Ndomo 2009). Bannon and Collier (2003), proved that a clear link exists between the high dependence on primary commodities and conflicts. This is because the struggle to take charge of natural resources and illicit smuggling leads to fights. The ECOWAS region is typically characterized by the smuggling of resources and conflicts, most of which are internal and small rather than largescale wars (McGowan 2006). Since 1998, more than 35-armed cliques have been operational in more than two-third of the 15 ECOWAS states (Florquin and Berman 2005). Easterly and Levine (1997), found that political unrest encouraged unfavorable developments which affected public choice. Fosu (2003), discovered that coup d’états also had a negative effect on African export advances, which was more than their effects on GDP. Another critical problem facing ECOWAS is corruption and bad governance. According to Heinrich (2006), corruption at all levels was very high in all ECOWAS countries when compared to other areas, with the region ranking as the most corrupt region in the world. Quartey (2012), reported that a recent survey of the West African trade hub showed that in every 100 km, there were 17 controls from which, on average, US dollar 54 was collected as a bribe. He further identified bribery as a major barrier in the movement of goods, people, and services across the area, with an average delay of 55 min per control point existing across the border of each

11.61

2.39

3.18

2.37

1.62

1.67

0.43

0.11

ASEAN

EFTA

CENSAD

MERCOUSUR

ECOWAS

COMESA

ECCAS

EAC

0.07

0.59

0.76

0.75

2.28

1.62

2.16

25.25

33.87

0.07

o.61

0.63

0.77

2.24

1.45

2.73

25.1

32.05

2014

0.07

0.55

0.49

0.72

2.04

1.28

2.42

25.79

32.39

47.97

2015

0.08

0.38

0.42

0.51

1.83

0.99

2.43

27.84

32.68

50.5

0.19

0.28

1.12

1.23

2.37

2.39

2.59

11.92

45.4

35.33

2010

0.17

0.27

0.86

0.54

1.98

1.56

1.67

22.5

34.5

48

2013

0.19

0.31

0.92

0.6

2.1

1.62

2.2

24.01

31.24

49.47

2014

0.21

0.34

0.96

0.61

1.96

1.64

1.94

23.9

31.7

49.4

2015

0.21

0.32

0.99

0.58

1.75

1.58

2

23.1

31.3

49.9

Note All figures are in billion USD. EU European Union; APEC Asian-Pacific Economic Cooperation; ASEAN Association of Southeast Asian Nations; EFTA European Free Trade Association; CENSAD Community of Sahel-Saharan States; MERCOUSUR Common Market of the South; COMESA Common Market for Eastern and Southern Africa; ECOWAS Economic Community of West African States; EAC East African Community; and ECCAS Economic Community of Central African States Source UNCTADstat (2016)

41.46

EU

2013 47.17

1980

2010

47.34

1980

34.3

Imports

Exports

APEC

Regional groups

Table 15.1 Share of world exports and imports by various regional trading groups across the globe (1980–2016)

15 Does Free Trade and Institutional Quality Affect … 307

308

L. O. Afolabi

country within the ECOWAS region. Such indications of the poor quality of existing governance speaks volumes about the poor economic development and growth in the bloc. With there being no operational checks and balances, corruption has remained unrestricted over the past four or more decades in Africa. Thus, after 38 years of existence as a regional group, ECOWAS’ performance remains stagnant with little or no progress to show despite the continuous reformation and implementation of new policies. Hence, there is an urgent need to remove the main barriers in ECOWAS’ growth including improving its economic performance. If this is not done, attaining the last stage of integration might not be an easy task. This research is among the first to comprehensively investigate the effects of free trade arrangements and institutional quality on ECOWAS’ trade by considering the time–varying multidimensional resistance as proposed by Silva and Tenreyro (2006), including zero trade in most sectoral data focused on trade flows. The rest of this study is organized as follows. Section 15.2 presents a summary of the empirical and theoretical background. Section 15.3 develops the PPML estimation technique. Section 15.4 deals with the data and empirical analysis and presents the result using PPML. Section 15.5 gives a conclusion.

15.2 Methodology This section discusses the gravity model that can be used to answer our objective. This study uses Thede and Gustafson (2012), Sharma and Chua (2000), Hassan (2000, 2001), Afolabi et al. (2016a, b, 2017) and Abidin et al. (2015) models. It makes minor adjustments to these models by bringing in dummy and other economic, political, and institutional variables to address its objectives. Our model examines the impact of institutional quality on ECOWAS’ trade. Looking beyond the major determinants for examining the likely impact of institutions on ECOWAS’ trade is necessary. The estimation technique that the study uses is the Poisson Pseudo-Maximum-Likelihood Estimation (PPML) method. This technique was introduced to gravity modelling to capture the zero-trade matrix that usually occurs in trade (exports and imports). Santos et al. (2010), designed a method for capturing zero trade matrices, which are part of the non-linear method of estimation. By default, the PPML estimation technique is semi-robust against any likely bias. Our main focus here is on three variables—the corruption index, political instability, and regulatory quality. However, other issues affecting the gravity equation include the question of the log or log stand-off. Log-linearization tends to change the error term’s properties. Consequently, this results in an estimation which is not efficient, largely owing to heteroscedasticity. Assuming that the data is homoscedastic, the variance in the error term and the anticipated value must be continuous. Trade data is always heteroscedastic and the projected value based on the error term is a function of the regressors. When the conditional distribution of the dependent variable is altered, OLS cannot be efficient. This argument has been emphasized numerous times by Silva and Tenreyro (2006, 2007, 2008). The significant fact is that “the log linearization of the empirical

15 Does Free Trade and Institutional Quality Affect …

309

model in the presence of heteroscedasticity leads to inconsistent estimates due to the fact the expected value of the logarithm of a random variable largely depends on higher-order moments of its distribution” (Silva and Tenreyro 2006, pp. 642–658). Heteroscedasticity in the data is not unique. When it comes to omitted or dependent variables, the error term variance differs according to the regressors. In line with Silva and Tenreyro (2006), PPML performs better than OLS in the existence of heteroscedasticity. Other issues emanating from OLS include endogeneity and the omission of zero values. Helpman et al. (2008) introduced the firm heterogeneity model to address the issue of zero values. Numerous researchers have employed PPML to estimate trade flows. Recently, Burger et al. (2009), Martin and Pham (2008), Martínez-Zarzoso et al. (2007), Westerlund and Wilhelmsso (2007) obtained results that are different when compared to the result of the PPML estimators and others that deal with heteroscedasticity and zero value issues. Zeroes in trade data and heteroscedasticity were found by Silva and Tenreyro (2006), who identified two issues that can generate a bias that is substantial. First, the trade dataset shows that it is unlikely that the variance of nij (the multiplicative stochastic term) will include several measures of distance independent of the country’s size. Due to the fact that the projected values of a logarithm of a random variable rests on both the mean and higher-order moments of its distribution, the variance of error term nij relies on the regressors, violating the condition of OLS consistency. Santos and Tenreyo (2006), suggest that this violation is a serious source of bias in the application of the gravity equation. Secondly, based on the logarithmic transformation, the pairs of nations for which bilateral exports recorded zero were dropped automatically from the sample. Typically this led to about 30% loss of data points. A substantial sample selection is both critical and problematic when considering poor or small nations. These two complications can be addressed when estimating the gravity equation in its multiplicative form. The Poisson estimator has extra properties for policy-applied investigations within gravity models. It is reliable in the existence of fixed effects which can be incorporated into our model as a dummy. This is an uncommon property for maximum likelihood which is regarded as a non-linear estimator. Numerous studies have failed to understand the properties including fixed effects. This key point is particularly significant for gravity modelling due to the fact that it is consistent with theory which requires the inclusion of fixed effects by both the importer and exporter. In conclusion, using PPML to estimate the gravity model adds another dimension to the gravity model in trade literature. We used PPML to capture the zero-trade matrix’s presence in the export trade flow data. The PPML regression’s results are compared to other estimations to see whether there is a variation in the final regression results. Relying on Santos and Tenreyro’s (2006) submissions, the importer and exporter dummy was included as a fixed effect. This is included in all ECOWAS estimation models.

310

L. O. Afolabi

15.3 Model Specification We considered the static gravity model within the context of a one-way model which includes time fixed effects, time effects, and a country effect. We examined a linear system that considers firm heterogeneity and exports. Referring to the comprehensive gravity model of trade, the export size among pairs of nations, X ij , is a function of their respective market size (GDP), their geographical distance, their population and a set of dummies: β1

β2

β3

β4

β5

β6

X i j = β0 Ci C j P O PI P O PJ disti J Fi j ∈i j

(15.1)

C i (C j ) represents the GDP of the exporter (importer), POPi (POPj ) is the exporter (importer) population, DIST ij represents the existing distance between the two countries’ main capitals (economic centers), and F ij indicates any other factors preventing or aiding trade among the pairs of nations. ∈i j represents the error term. For the purpose of our estimation, we can postulate an augmented type of model in a log-linear procedure which includes the time dimension: ln X i jt = ∝i j1 +∅t1 + X i1 + ϕ j1 + β1 ln Cit + β2 ln C jt + β3 ln P O Pit  + β4 ln P O P jt + β5 ln D I STi j + β6langi j + β7 ad ji j + γk1 Dki Dk j +

 k

δk1 Dki +



k

δk1 Dk j + ∈i jt

(15.2)

k

where ln represents the variables in the logarithms and X ijt represents export from nation i to nation j in period i (time). C it and C jt indicate the GDP of nations i and j, POPit and POPjt denote the populations of nations i and j respectively in period t. DIST ij is the existing distance between the countries capitals of i to j. The model included the dummy variables of the trading partners sharing common borders (adjij ) and a common language (langij ) as well as the trading blocs which were also represented with a dummy variable that was incorporated in the model to measure the effects of free trade agreements (FTAs). ∝i j measures the specific effects associated with each bilateral trade flow. They also effectively control for all omitted variables that are essential for each trade flow, which are also time invariant in nature. ∅t1 represents the time effects that are specific and that control for the omitted/ignored variables. This is usual for trade flows but they can also vary over the time period. X i1 + ϕ j1 are the exporter and importer effects that are proxies for the multilateral resistance factors. ∈i jt signifies the error term that is presumed to be well-behaved. However, there is another amendment to the existing conditions, which include country and time effects and account for time–variant multilateral value terms as suggested by (Baldwin and Taglioni 2006). According to Baldwin and Taglioni (2006), time-varying nation dummies can totally remove the bias emanating from the ‘gold medal error,’ that is, the omission or incorrect specification of the terms which Anderson and Wincoop (2003), call multilateral trade resistance. The resulting equation is:

15 Does Free Trade and Institutional Quality Affect …

311

ln X i jt = ∝i j1 +∅t1 + X i1 + ϕ j1 + β1 ln Cit + β2 ln C jt + β3 ln P O Pit + β4 ln P O P jt + β5 ln D I STi j + β6langi j + β7 ad ji j + β8 F T Ai j + β9 P T Si j + β10 R E E Ri j + β11 Corri j + β12 R E Q i j  + β13 ln DG D P PCi jt + β14 ln SG D Pi jt γk1 Dki Dk j +

 k

δk1 Dki +



k

δk1 Dk j + ∈i jt

(15.3)

k

The new adjustment includes the institutional quality variable in the model which includes ij which is the political unrest variable index for all the ECOWAS countries. Corr ij measures the level of systematic, systemic, and incidental cases on the corruption index for the ECOWAS countries. REQij denotes and captures the regulatory quality index for ECOWAS countries. REERij represents one of the macroeconomic indicators called the real effective exchange rate. We estimated the aggregated bilateral exports flows of 15 ECOWAS countries over the period 1996–2017. Our data was unbalanced with a maximum of 4,480 observations (16 × 14 × 20). The data was extracted from the Directory of Trade (DOT) from IMF, GDP, CPI, and GDP per capita. The good governance index was obtained from the IMF and World Bank databases. The corruption perception index was obtained from http://www.transparency.org. The political instability index was obtained from http://www.politicalterrorscale.org.

15.4 Interpretation of the Results This section examines the impact of free trade and institutional quality on ECOWAS’ trade. Looking beyond the major determinants to examine the likely impact of institutions on ECOWAS’ trade is important. We used the PPML estimation technique which was introduced to the gravity modelling to capture the zero-trade matrix that usually occurs in trade (exports and imports). Santos et al. (2010) designed the method to capture zero trade matrices, which are a part of the non-linear method of estimation. The default PPML estimation techniques are semi-robust against likely biases. Figure 15.1 shows that using the combined economic mass of importing and exporting nations (the GDP of ECOWAS countries) represents the explanatory variables. The scatter plot indicates a positive relationship between the two variables. Further, the best line is upward sloping. The graphical representation offers further proof that larger country pairs tend to trade more. The scatter plot in Fig. 15.2 is suggestive of a negative relationship. This impression can be demonstrated further using the line of best fit, which has a strong downward slope. Thus, the nation pairs that are far apart tend to trade less. In addition, the estimated results (PPML) reported in parentheses are semi-robust as indicated

L. O. Afolabi

5

10

15

20

25

312

35

40

45

50

ln_gdp_both logexport

Fitted values

5

10

15

20

25

Fig. 15.1 Line of best fit and the scatter plot for export versus GDP combined (1996–2017). Source Authors’ computations

5

6

7

8

9

logdist logexport

Fitted values

Fig. 15.2 Line of best fit and scatter plot for export versus distance (1996–2017). Source Authors’ computations

in Table 15.2 correspond with Santos et al.’s (2010) study. PPML includes timevarying importers and exporters with a fixed effect as recommended by Anderson and Wincoop (2003). Discussing the elasticity of GDP’s exports and imports in the ECOWAS countries is important. The gravity model using export bilateral flows as the dependent variable shows that the coefficient of GDPi (exporting countries) was significant and positive at 1%. A 1% upsurge in GDPi of exporting countries to other ECOWAS nations led to an increase of 1.23% in exports. The GDP of the importing nations within ECOWAS was also negative and significant at the 1% significance level. This implies that with

15 Does Free Trade and Institutional Quality Affect … Table 15.2 Poisson pseudo-maximum likelihood estimation results (1997–2017)

DV: export

313 PPML Coef

Contig

T-test

0.136***

9.80

ln populationi

−1.638***

6.34

ln populationj

1.119***

4.21

ln GDPi

1.234***

6.71

ln GDPj

−1.284***

5.16

ln GDP similarity

0.017

1.39

ln GDPPC differences

0.014

1.35

ln Exchangeij

0.005***

4.70

FTA

−0.472***

4.14

ln PTSj

−0.702***

2.52

ln distance

−0.035***

4.10

0.055***

6.43

Common lang ln Corruptj

−0.219

0.88

ln Regulatoryj

−0.314*

1.75

Adj R2

0.86

Note *, **, and *** denote significance at the 10%, 5%, and 1% levels respectively Source Authors estimation

a 1% decrease in the GDP of the importing nations within ECOWAS, their level of importing will decrease by −1.28%. In our study, GDP represents economic size and represents what Pöyhönen (1963) and Pulliainen (1964) have called the trading volume of a nation. The exporting countries’ GDP represents productive capacity and is also a metric used for measuring a range of product varieties available for exports. GDP under imports is an indicator of the size of the imported goods. It is expected that the incomes of both countries within the region should positively impact bilateral trade (Linnemann 1966; Pöyhönen 1963; Pulliainen 1964; Tinbergen 1962). Populationi representing the exporting countries was both positive and significant at the 1% level. With a 1% increase in population, the exporting ECOWAS countries tended to produce more. This means that ECOWAS’ exports will increase by 4.39%. Logically, countries with larger populations tend to have a diversified production base, which means that they have the capacity to capture economies of scale in production. This gives them an edge when it comes to trading more as compared to smaller economies. This finding is in line with Brada and Mendez’s (1983) findings. Populationj represents the population of the importing countries within ECOWAS. The coefficient of populationj was both negative and significant. The negative sign is consistent with the works of Blomqvist (1994), Linnemann (1966), Mátyás et al. (1997), and Oguledo and MacPhee (1994) who concluded that trading populations of countries affected trade flows negatively and significantly. With a 1% decrease

314

L. O. Afolabi

in populationj , imports will reduce by −1.63%. Generally, the expected sign of the population parameters is always unclear (Brada and Mendez 1983) (Table 15.2). Logically, the importer income of a populationi is a measure of the potential import demand while exporter income and populationj can be an indicator of exporter supply (Aitken 1973; Linnemann 1966). The distance variable was included in the model to capture transaction costs. Distance had a strong harmful effect on the overall size of the trade between countries from i to j within ECOWAS by 1%, which led to a reduction in exports by 0.035%. This result is in line with Breuss and Egger (1997, 1999), Frankel et al. (1995), Soloaga and Wintersb (2001), Thoumi (1989a, b). Overall, these results suggest that transportation expenses have a great impact in most small ECOWAS economies. The variable common border (Contig) was found to be positive and significant in all the models. Sharing a common border positively affected trade, which is in line with the results of gravity models. A shared border is projected to have a positive coefficient because the countries selected for this study were from the same region. The West African nations sharing the same boundaries within ECOWAS trade 0.136% more than those without a shared border. The coefficient of the common language variable for our study was positive and significant in the two estimated results, signifying that two or more countries within ECOWAS sharing the same official language tended to spur trade between them. An increase of 1% in the coefficient of common language of the countries within ECOWAS tends to increase export trade within ECOWAS by 0.05% at a 1% significance level. This finding is in line with Stack and Pentecost (2011). We now shift focus to an examination of the impact of institutional quality variables incorporated in the model because observing the behavior of the series is important. For purposes of interpretation, three indices of institutional quality were incorporated in the models—the level of corruption, political instability, and regulatory quality. The political instability variable was found to be negative and statistically significant at the 1% level. ECOWAS’ exports will increase by 0.70% if there is one unit of improvement in the political instability rating for all ECOWAS countries. This result is in line with Bannon and Collier (2003), Easterly and Levine (1997), Florquin and Berman (2005), Fosu (2003), and McGowan (2006). These authors discovered that political instability encouraged poor growth through its effects on public choice. They also showed that coups d’états also had adverse effects on African export growth, which was greater than their effects on GDP. The regulatory quality variable was negatively significant and indicated that a one standard deviation decrease in regulatory quality would lead to a 0.31% increase in bilateral trade within the ECOWAS region. This result is in line with Mehlum et al. (2006) and Fosu’s (2011) findings who showed that institutional considerations remain essential if African countries want to enhance their trade and development. The exchange rate variable coefficient was positive and significant under CCR and PPML at the 1% significance level. This positive coefficient suggests that an appreciation of the real exchange rate will encourage ECOWAS exports from i to j. The estimated results, which indicate depreciation suggest that the depreciation of the real exchange rate by 1% would increase ECOWAS’ exports from i to j by

15 Does Free Trade and Institutional Quality Affect …

315

0.005%. Under the estimation technique (PPML), the variable was negative and significant. The results showed that ECOWAS’ exports would increase by 0.47% if there was a one unit improvement under FTA. The results also show that a trade diversion exists. The proliferation of FTAs has been widely criticized. One criticism is the fear of trade diversion through bypassing an effective non-member nation in favor of a member nation that is less efficient in production. FTA is a tool used for reducing trade barriers among the signatories and this leads to an expansion of trade at the bilateral level among its affiliates, which could also be at the expense of non-members.

15.5 Conclusion Regional integration agreements have been a very significant issue in economic literature that studies the reasons for their existence. The theoretical background and its impact on these arrangements in both the overall trading system and that of individual countries has increased. This study emphasizes the effect of free trade and institutional quality on the ECOWAS trading bloc. It also provides further evidence of a systematic multifaceted crisis effect on trade. Our study recommends that to solve the issue of weak institutional quality, ECOWAS countries need to advocate clean accountancy, transparency, and good governance. Another crucial finding of the gravity model is the issue of political instability. It is time that the ECOWAS countries address the issue of political unrest as it is a serious hindrance in these countries’ penetrating the available markets within the ECOWAS region. In addition, the issue of good governance should be addressed with strict mechanisms and policies, allowing ECOWAS to come up with a policy that will serve as a check and balance for all ECOWAS countries to have sound institutions. The study recommends that the ECOWAS regional group should integrate further. This recommendation is further reinforced because the FTA stated that as a force it was still not yielding the desired results, thus showing the need for strengthening the regional group further. ECOWAS needs to move to the next stage of integration, which is, having a customs union, an economic union, and a monetary union. ECOWAS needs to address the matter of the partial implementation of the regional integration arrangement process as it hinders the performance of ECOWAS’ trade. For instance, ECOWAS has not yet fully implemented free trade agreements and it has started implementing the customs union. ECOWAS needs to fully achieve a particular stage before moving to the next stage to ensure that all ECOWAS countries fully comply with the rules. Otherwise, jumping from one integration process to another without full success in the first stage will hinder ECOWAS’ performance.

316

L. O. Afolabi

References Abidin, I. S. Z. N. A. Bakar, and M. Haseeb (2015). Exploring trade relationship between Malaysia and the OIC member countries: A panel cointegration approach (1995–2012). Asian Journal of Scientific Research, 8(1): 107. Afolabi, L. O. A. B. Aznin, and A. A. Izraf (2017). ECOWAS Regional Integration and Trade Effect: A Dynamic Panel Cointegration Method Using the Gravity Model. International Journal of Economic Perspectives, 11(1): 165–172. Afolabi, L. O. N. A. A. Bakar, and A. Aziz (2016a). The Gravity Model Approach: An Application on The Ecowas Trading Bloc. The South East European Journal of Economics and Business, 11(1). Afolabi, L. O. N. A. A. Bakar, and M. I. A. Aziz (2016b). Openness and Ecowas Regional Trade: A Panel Cointegration Analysis Using A Gravity Model. Aktual’ni Problemy Ekonomiky = Actual Problems in Economics, 177: 319. Aitken, N. D. (1973). The effect of the EEC and EFTA on European trade: A temporal cross section analysis. The American Economic Review, 63(5): 881–892. Anderson, J. E. and D. Marcouiller (2002). Insecurity and the pattern of trade: An empirical investigation. Review of Economics and statistics, 84(2): 342–352. Anderson, J. E. and E. Van Wincoop (2003). Gravity with gravitas: a solution to the border puzzle. American economic review, 93(1): 170–192. Baier, S. L. and J. H. Bergstrand. (2007). Do free trade agreements actually increase members’ international trade? Journal of international Economics, 71(1): 72–95. Baldwin, R. and Taglioni, D. (2006). Gravity for dummies and dummies for gravity equations (No. w12516). National bureau of economic research. Bannon, I. and P. Collier (2003). Natural resources and violent conflict: Options and actions. World Bank Publications. Blomqvist, H. C. (1994). Trade and Industrial Policy in Developing Countries. A Manual of Policy Analysis. Bouet, A. Mishra, S. and Roy, D. (2008). Does Africa trade less than it should and why? The role of market access and domestic factors. In IFPRI, December, IFPRI Discussion Paper. Burger, M. Van Oort, F. and Linders, G. J. (2009). On the specification of the gravity model of trade: zeros, excess zeros and zero-inflated estimation. Spatial Economic Analysis, 4(2), 167–190. Brada, J. C. and J. A. Mendez (1983). Regional economic integration and the volume of intraregional trade: a comparison of developed and developing country experience. Kyklos, 36(4): 589–603. Breuss, F. and Egger, P. (1997). Use and misuse of gravity equations in European integration research (No. 93). WiFo. Breuss, F. and P. Egger (1999). How reliable are estimations of East-West trade potentials based on cross-section gravity analyses? Empirica, 26(2): 81–94. Easterly, W. and Levine, R. (1997). Africa’s growth tragedy: policies and ethnic divisions. The quarterly journal of economics, 112(4), 1203–1250. Florquin, N. and Berman, E. G. (Eds.). (2005). Armed and aimless: Armed groups, guns and human security in the ECOWAS region. Small arms survey. Fosu, A. K. (2003). Political instability and export performance in sub-Saharan Africa. Journal of Development Studies, 39(4), 68–83. Fosu, A. K. (2011). Terms of trade and growth of resource economies: A tale of two countries: Working paper//World Institute for Development Economics Research. Frankel, J. Stein, E. and Wei, S. J. (1995). Trading blocs and the Americas: The natural, the unnatural, and the super-natural. Journal of development economics, 47(1), 61–95. Hassan, M. K. (2000). Trade relations with SAARC countries and trade policies of Bangladesh. Journal of Economic Cooperation, 21(3): 99–151. Hassan, M. K. (2001). Is SAARC a viable economic block? Evidence from gravity model. Journal of Asian Economics, 12(2): 263–290.

15 Does Free Trade and Institutional Quality Affect …

317

Heinrich, F. (2006). Transparency International Corruption Perceptions Index: Transparency International, Bertelsmann Foundation. Helpman, E. M. Melitz, and Y. Rubinstein (2008). Estimating trade flows: Trading partners and trading volumes. The Quarterly Journal of Economics, 123(2): 441–487. Linnemann, H. (1966). An econometric study of world trade flows. Amsterdam: North Holland. Martin, W. and C. S. Pham (2008). Estimating the gravity equation when zero trade flows are frequent. MPRA Paper. Martínez-Zarzoso, I. A. Bengochea-Morancho, and R. Morales-Lage (2007). The impact of population on CO2 emissions: evidence from European countries. Environmental and Resource Economics, 38(4): 497–512. Mátyás, L. (1997). Proper econometric specification of the gravity model. The World Economy, 20(3): 363–368. McGowan, P. J. (2006). Coups and conflict in West Africa, 1955–2004: Part I, theoretical perspectives. Armed Forces and Society, 32(1): 5–23. Mehlum, H. K. Moene, and R. Torvik (2006). Institutions and the resource curse. The Economic Journal, 116 (508): 1–20. Ndomo, A. (2009). Regional economic communities in Africa: A progress overview. Study Commissioned by GTZ, Nairobi. Oguledo, V. and C. R. MacPhee (1994). Gravity models: a reformulation and an application to discriminatory trade arrangements. Applied Economics, 26(2): 107–120. Pöyhönen, P. (1963). A tentative model for the volume of trade between countries. Weltwirtschaftliches Archiv: 93–100. Pulliainen, K. (1964). A World Trade Study: an Economic Model of the Pattern of the Commodity Flows in International Trade in 1948–1960. Institute of economics, University of Helsinki. Quartey, L. (2012). ECOWAS The African Report. Retrieved from http://www.theafricareport.com/ WestAfrica/ecowas-lagging-behind-sadc-eac-comesa-in-integration.html. Sharma, S. C. and Chua, S. Y. (2000). ASEAN: economic integration and intra-regional trade. Applied Economics Letters, 7(3), 165–169. Silva, J. S. and S. Tenreyro (2006). The log of gravity. The review of economics and statistics, 88(4): 641–658. Santos Silva, J. M. C. and Tenreyro, S. (2008). Comments on the log of gravity revised. retrieved on September, 20, 2012. Silva, J. S. and S. Tenreyro (2010). On the existence of the maximum likelihood estimates in Poisson regression. Economics Letters, 107(2): 310–312 Silva, J. S. and Tenreyro, S. (2011). Further simulation evidence on the performance of the Poisson pseudo-maximum likelihood estimator. Economics Letters, 112(2), 220–222. Soloaga, I. and Wintersb, L. A. (2001). Regionalism in the nineties: What effect on trade?. The North American Journal of Economics and Finance, 12(1), 1–29. Stack, M. M. and E. J. Pentecost (2011). Regional integration and trade: A panel cointegration approach to estimating the gravity model. The Journal of International Trade and Economic Development, 20(1): 53–65. Tenreyro, S. (2007). On the trade impact of nominal exchange rate volatility. Journal of Development Economics, 82(2), 485–508. Thede, S. and N. Å. Gustafson (2012). The multifaceted impact of corruption on international trade. The World Economy, 35(5): 651–666. Thoumi, F. E. (1989a). Bilateral trade flows and economic integration in Latin America and the Caribbean. World development, 17(3): 421–429. Thoumi, F. E. (1989b). Trade flows and economic integration among the LDCs of the Caribbean Basin. Social and Economic Studies, 215–233. Tinbergen, J. (1962). An analysis of world trade flows. Shaping the world economy: 1–117. UNCTADstat (2016). United Nation Conference on Trade and Development Retrieved 24/06/2018. http://unctadstat.unctad.org/wds/TableViewer/tableView.aspx?ReportId=102.

318

L. O. Afolabi

UNCTADstat (2018). United Nation Conference on Trade and Development. Retrieved from: http:// unctadstat.unctad.org/wds/TableViewer/tableView.aspx?ReportId=102. Westerlund, J. (2007). Testing for error correction in panel data. Oxford Bulletin of Economics and statistics, 69(6): 709–748. Winters, L. A. (2004). Trade liberalisation and economic performance: an overview. The Economic Journal, 114(493), F4–F21.