Public Enterprise in Kenya: What Works, What Doesn't, and Why 9781685851620

Citation preview

Public Enterprise

in Kenya

Public Enterprise

in Kenya

What Works, What Doesn't, and Why

Barbara Grosh

Lynne Rienner Publishers

• Boulder & London

Parts of Chapter 3 were first published in "Performance of Agricultural Parastatals in Kenya: 1963-84," Eastern Africa Economic Review, June 1987. Parts of Chapter 4 were first published in "Parastatal-Led Development: The Financial Sector in Kenya, 1971-1987," African Development Review, vol. 2 no. 2, December 1990. Parts of Chapter 5 were first published in "Public, Quasi-Public, and Private Manufacturing Firms in Kenya: The Surprising Case of a Cliché Gone Astray," Development Policy Review 8, March 1990.

Published in the United States of America in 1991 by Lynne Rienner Publishers, Inc. 1800 30th Street, Boulder, Colorado 80301 and in the United Kingdom by Lynne Rienner Publishers, Inc. 3 Henrietta Street, Covent Garden, London WC2E 8LU © 1991 by Lynne Rienner Publishers, Inc. All rights reserved Libraiy of Congress Cataloging-in-Publication Data Grosh, Barbara. Public enterprise in Kenya: what works, what doesn't, and why / by Barbara Grosh Includes bibliographical references and index. ISBN 1-55587-209-3 1. Government business enterprises—Kenya. 2. Government ownership—Kenya. I. Title. HD4346.5.G76 1991 338.6'2'096762—dc20 91-20375 CIP British Cataloguing in Publication Data A Cataloguing in Publication record for this book is available from the British Library.

Printed and bound in the United States of America The paper used in this publication meets the requirements of the American National Standard for Permanence of Paper for Printed Library Materials Z39.48-1984.

Contents List of Tables and Figures Acknowledgments

vii xi

1

The Crisis in Africa and the Search for Solutions

2

The Public Enterprise Problem in Kenya

11

3

The Agricultural Sector

27

4

The Financial Sector

55

5

Development Finance Institutions

91

6

Infrastructural Parastatals:

7

1

Transport, Communications, and Electricity

119

The Public Enterprise Problem in Kenya Revisited

153

Appendix A: Tables Corresponding to Figures

173

Appendix B: Deviations of Consumer and Producer Prices from Competitive Levels

205

Appendix C: Auxiliary Tables

209

References Index About the Book and the Author

215 219 223

v

Tables and Figures TABLES 1 -1 1 -2 1-3 1 -4 1-5 1-6 1 -7 1-8 2-1 2-2 3-1 3-2 3-3 3-4 3-5 3-6 3-7 3-8 3-9 3-10 3-11 3-12 4-1 4-2 4-3 4-4

Growth Rates of GNP per Capita Average Annual Growth Rate of Merchandise Trade Volume Measures of Debt for Sub-Saharan Africa External Debt Situation of ID A-Only Countries Central Government Deficit Output and Investment Shares of Public Enterprises Contribution of Public Enterprises to Growth in Domestic Credit Savings and Investment Shares in GDP Sequence of Major Public Enterprises in Kenya Performance of Public Enterprises in Kenya Firms Included in the Study Sample Type of Market Served and Manner of Price Formation Nature of Monopsony Power Ratio of Net Worth to Net Assets Ratio of Current Assets to Current Liabilities Trends in the Financial Performance of Agricultural Parastatals Number of Firms Classified by Average Rate of Return on Assets Classification of Firms by Financial Performance Trends in Processing Margin and Classification of Performance Comparison of Commodity Price Levels Used by Public Enterprises with Competitive Prices Firms with Problems Cost Function Estimates Nominal and Real Interest Rates Distribution of Bank Branches Consolidated Commercial Bank plus NBFI Deposits Comparison of Commercial Bank Deposits vii

3 3 4 4 5 6 7 8 12 22 28 29 29 34 35 36 39 39 41 46 47 48 58 60 64 66

viii 4-5 4-6 4-7 4-8 4-9 4-10 4-11 4-12 4-13 4-14 4-15 4-16 5-1 5-2 5-3 5-4 5-5 5-6 5-7 5-8 6-1 6-2

Public Enterprise in Kenya Nonbank Financial Institution Deposits Regression Results K C B Fixed Assets Versus Deposits Rates of Return on Equity Effect of B a d Loans on N B K Profits Costs (excluding interest) as a Percentage of Loans Outstanding Financial Structure of A F C Average A F C Interest Rates and Rates of Return Transfusions of New Capital to A F C New A F C Loans Approved Characteristics of Loans and Borrowers Total Arrears on A F C Portfolio Development Finance Institutions I C D C Record of Investments and Dividends Administrative Expenses as a Percentage of Average Total Assets Measures of Performance in Manufacturing Firms Average Measures of Performance by Industry and Ownership Distribution of Capital Stock by Industry Grouping and Ownership Classification Quantity Index of Manufacturing Correlation Between Effective R a t e of Protection and Domestic Resource Cost Ratio Classification of Firms by Market Conditions Operational Indicators

67 69 72 73 74 76 78 79 80 83 83 84 92 100 104 108 110 112 113 114 120 150

FIGURES 2-1 3-1A 3-1B 3-1C 3-1D 3-1E 3-2 3-3A 3-3B 3-3C

Public Enterprise Shares Net Assets: Sugar Companies Net Assets: Cereals Marketing Firms Net Assets: K C C , K T D A Net Assets: C L S M B , Pyboard Net Assets: C B K , K M C , Uplands Average Rates of Return R e a l Unit Costs, Excluding Interest: Sugar Companies Real Unit Costs, Excluding Interest: Cereals Marketing Firms Real Unit Costs, Excluding Interest: K M C , Uplands, K C C , C B K

14 30 31 31 32 32 38 41 42 42

Tables and Figures 3-3D 3-3E 3-4 4-1 4-2 4-3 4-4 4-5 4-6 4-7 5-1 5-2 5-3 5-4 5-5 5-6 5-7 5-8 5-9 6-1 6-2 6-3 6-4 6-5 6-6 6-7 6-8 6-9 6-10 6-11 6-12 6-13 6-14 6-15 6-16 6-17

Real Unit Costs, Excluding Interest: KTDA, Pyboard Real Unit Costs, Excluding Interest: CLSMB U-Shaped Average Cost Functions Deposit Growth Rates Currency/M2 Ratios of Money to GDP Parastatal Market Shares Growth Rates of NBFI Deposits Measures of Returns to Employees Agricultural Credit Outstanding DFI Investment as Share of Key Macroeconomic Variables Rate of Return on Portfolio IDB Rates of Return on Portfolio ICDC Rates of Return ICDC Alternative Measures of Returns on Equity DFCK Rates of Return on Portfolio Administrative Costs as Percentage of Portfolio ICDC Administrative Costs as a Percentage of Portfolio Rates of Return: DFIs Price and Cost Performance: Kenya Pipeline Company Financial Performance: Kenya Pipeline Company Financial Condition: Kenya Pipeline Company Indexes of Output: Kenya Posts and Telecommunications Rates of Return: Kenya Posts and Telecommunications Financial Condition: Kenya Posts and Telecommunications Real Prices: Kenya Power and Lighting Company Rate of Return on Equity: Kenya Power and Lighting Company Financial Ratios: Kenya Power and Lighting Company Cargo Handled at Mombasa: Kenya Ports Authority Price Performance: Kenya Ports Authority Rates of Return: Kenya Ports Authority Financial Condition: Kenya Ports Authority Rate of Return on Total Assets: Kenya Airways Financial Condition: Kenya Airways Real Cost per Passenger-Kilometer: Kenya Airways Performance Indexes: Kenya Airways

ix

43 43 49 57 59 60 65 68 81 87 93 95 97 97 98 98 101 103 104 123 125 125 127 127 128 130 130 131 132 132 133 133 135 135 137 137

x

Public Enterprise in Kenya

6-18 6-19 6-20 6-21 6-22 6-23 6-24 6-25 6-26 6-27 6-28 6-29

Load Factors: Kenya Airways Output Indexes: Kenya Airways Financial Returns: Kenatco Financial Ratios: Kenatco Market Share: Kenatco Revenue Ton-Miles: Kenya Railways Index of Value of Output: Kenya Railways Railway Freight Traffic: Kenya Railways Traffic Units per Employee: Kenya Railways Cost and Revenue Performance: Kenya Railways Average Real Revenue per Ton-Mile: Kenya Railways Rate of Return on Net Assets: Kenya Railways

139 139 141 143 143 144 145 145 147 147 149 150

Acknowledgments In a project of the scope and duration of this one, so many debts are incurred that it is impossible to acknowledge them all. David Leonard was instrumental in every phase, especially in getting funding and research clearance. He showed infinite patience in reading endless drafts, and his comments and questions led to many improvements in the final product. Albert Fishlow's guidance in the final phases was always constructive and cheerful, and did much to improve the product as well. I am grateful to the Institute for Development Studies (IDS) at the University of Nairobi, where I was a visiting research associate during my fieldwork. Colleagues there made me welcome, and their joking threats to nationalize me were received as high compliments. I remember with special fondness the stimulation of exchanging views with George Ruigu, Shem Migot-Adholla, Charles Okidi, Patrick Alila, Kabiru Kinyanjui, and Njuguna Ngethe. I am grateful to the Government Investments Division in the Ministry of Finance, who gave me a desk and access to their files. I spent so much time there that many people thought I was a part of the division, and indeed I began to feel I was. I remember with special gratitude the late Dr. Maurice Dangana, and A. M. R. Odipo, Sam Ratnam, V. V. Ramanadham, George Mitine, Peter Karanja, Erastus Rweria, and Miss Mwakio. I received gracious cooperation from the Inspectorate for Statutory Boards, especially from Richard Ndubai, who read the entire manuscript carefully (but who should be absolved of blame for remaining errors and opinions). B. L. Makotsi helped collect data (and helped me understand it) on the sugar industry. Managers of the firms where I interviewed were almost without exception kind and helpful. The courtesy and generosity with which I was treated made the fieldwork a pleasure I was sorry had to end. In Kenya I had support from several groups of scholars. The Kenya Economics Association welcomed me back into "the tribe of the econs" after my sojourn in the wilds of the interdisciplinary IDS. Germano Mwabu, Aloys Ayako, Geoffrey Mwau, G. K. Ikiara, Peter Coughlin, and others were valuable supporters and critics. Other colleagues at the University of Nairobi working on public enterprise included Frank xi

xii Public Enterprises in Kenya

Okuthe-Oyugi, Walter Oyugi, and Wanyama Kulundu-Bitonye; I benefited from discussions and friendship with each of them. Alfred Saulniers shared with me the manuscript of his book on public enterprise in Peru, which helped me get oriented to begin my fieldwork. Fellow travelers working in Nairobi at the time provided a sounding board for ideas and frustrations. They included Jennifer Widner, David Throup, Rees Hughes, Amy Uyeki, Susan Hall, Corinne Kratz, Inez Sutton, Sara Scherr, Luis Malaret, Diane Rocheleau, John Nordin, R. J. P. Scott, Leslie Louie, David Bowen, Steve Orvis, John Cohen, Mike Roemer, and Paul Smoke. Judith Geist was an unfailing fount of good advice on how to get through the Kenyan bureaucracy. Officials at the World Bank and USAID shared documents on particular public enterprises and gave stimulating critiques of my work. I am especially grateful to John Nellis. Doris Jansen and Michael Selhorst were very cooperative in transmitting to me (with permission from the Kenyan government) the data on manufacturing firms in Chapter 5. Judith Tendler read much of the manuscript, as did Robert Bates, Laura Tyson, and Erick Otenyo. During the writing phase in Berkeley I had the good fortune to have a sympathetic dissertation group, which helped get me through the lonely days of writing. I am especially grateful to Rwekaza Mukandala, whose work on parastatals in Tanzania and Botswana helped me to get a little bit of comparative perspective. I am grateful for financial support from the Institute for International Studies at Berkeley and to the Fulbright-Hays Doctoral Dissertation Research Abroad program. The Metropolitan Studies program and Department of Public Administration at the Maxwell School of Syracuse University provided support during the final revisions. Fred Kwarteng, James Bower, and Brenda Bushouse all served as research assistants, helping me update and revise the final manuscript. Kathleen Staudt helped assemble the manuscript in its final phases and converted endless printouts into readable tables. Finally, I am grateful to my family, who have given me more patient encouragement than anyone has a right to expect.

Public Enterprise

in Kenya

The Crisis in Africa and the Search for Solutions The late 1970s and the 1980s were a time of economic crisis in Africa. External shocks in the form of the oil price increases of the 1970s coupled with declining demand for African exports due to world recession, as well as high real interest rates, afflicted most countries. The result was low or negative per capita income growth in most of the continent, and the excessive borrowing of the past promises to limit future growth. Public enterprises have been central in African states' development strategies. Controlling key sectors—including mineral and agricultural exports, transport and communications, manufacturing, and much domestic agricultural trade—public enterprises are alleged to be inefficient, to lose large amounts of money, and to cause allocative inefficiency by distorting prices and access to domestic and external credit. Naturally, as policymakers within Africa and in the international lending agencies look for ways out of the crisis, public enterprises are a focus of attention. Various parties have been quick to propose solutions to "the public enterprise problem." Some (especially in the donor agencies) have advocated that Africa follow the example of Thatcher's Britain and undertake wholesale privatization. Others (especially bureaucrats in African governments) have proposed measures to bring public enterprises "under control," shifting more and more decisionmaking authority from parastatals to government ministries. Yet, little concrete information has been available to show the dimensions of the problem, its causes, or how reforms might be expected to bring about improvement. This study provides that missing information for one important case—Kenya. Kenya is known for its relatively market-oriented approach to development and for its comparatively sophisticated administrative machinery. The country's experience demonstrates both the importance of the external shocks that precipitated the crisis and the potential for various proposed solutions to the public enterprise problem, be they privatization or increased controls on public enterprise managers. This study covers the operations of 32 public enterprises from independence in 1963 until 1988. The sample contains almost all those firms that are supported primarily from revenue from sales and that are at least 90 percent directly publicly owned. A few firms were dropped because 1

2

Public Enterprise in Kenya

their accounts did not provide enough information to give useful indicators of performance. The insurance companies were also excluded. The Kenya National Transport Company (Kenatco) was added, although it is a subsidiary, because of the high level of public interest in its activities. In addition, I provide data for 1984 on the operations of 30 manufacturing subsidiaries of some of the firms in the longitudinal sample, which I compare with 40 private manufacturing firms. For the first time, we get a fairly detailed view of "the public enterprise problem": when and how it developed, its scope, and its major causes. The results suggest that the policies being advanced will do little to improve the problem, and could make it much worse.

THE CRISIS IN AFRICA During the 1970s, GNP growth rates in sub-Saharan Africa fell to levels that were low for the region itself and also lower than other regions. The 1980s was a time of even greater distress, as per capita income shrank rapidly. Table 1-1 shows per capita growth rates for three different periods. These low growth rates are especially tragic since Africa is the poorest region of the world. Per capita income for the region was $411 in 1979. Of the 36 countries categorized in the World Bank's World Development Report 1986 as low-income, 21 are African. The grim economic statistics are reflected in equally grim social indicators. Death rates are the highest in the world and life expectancy is the lowest (47 years). Fifteen to 20 percent of children die by their first birthday, and only 25 percent of the population has access to safe drinking water. 1 The crisis in Africa was precipitated by external shocks and the adjustment policies (and failures to adjust) that followed in their wake. Table 1-2 shows that export volume for Africa stagnated during the 1970s and declined during the 1980s. The picture is not quite so bleak when Nigeria is excluded, but the failure to recover in the 1980s when the rest of the world did is striking. The low-income countries in Africa did even worse than the continent as a whole. Like the rest of the world, Africa had trouble contracting import growth in the face of declining export growth. Before the oil crisis the volume of exports for sub-Saharan Africa excluding Nigeria grew at 5.7 percent while import volume grew more slowly, at 3.4 percent. During the 1973-1980 period, import volume growth slowed, but export growth declined to a greater extent. During the 1980s, slow growth in export volume was compounded by declining terms of trade, so that the volume of imports was forced to contract throughout most of Africa. The negative effect of adverse world demand conditions was com-

Crisis in Africa & Search for Solutions

3

Table 1-1 Growth Rates of GNP per Capita

Sub-Saharan Africa SSA excluding Nigeria All low-income economies

1965-73

1973-80

1980-87

2.9 1.2 3.3

0.1 -0.7 2.6

-2.8 -1.2 4.0

Source: World Bank, 1989, page 221.

pounded by African policy mistakes. Real effective exchange rates appreciated by roughly 70 percent for sub-Saharan Africa from 1971 to 1984, during which period they declined in Latin America and Asia. 2 As a result, sub-Saharan Africa's share in developing country exports fell by half from 1970 to 1985. 3 African governments (as well as bodies like the IMF, World Bank, and O E C D ) expected the effects of the second oil price shock to be like the first—a sharp shock, a short slump, and a rapid recovery. Governments responded by using supplier credits and commercial bank loans to bridge what they expected to be a short slump. When the recession became prolonged and real interest rates rose, their debt service positions became unmanageable, despite sharp cuts in import volume.4 Table 1-3 shows the aggregate debt burden of sub-Saharan Africa. As both the African governments and their creditors began to realize that they had over borrowed, medium- and long-term finance became much more difficult to obtain. Countries scaled down new borrowing and devoted ever larger portions of their available foreign exchange to repaying previous borrowing. For some countries debt service ratios were

Table 1-2 Average Annual Growth Rate of Merchandise Trade Volume (percentages)

Exports

Terms of Trade 1980=100

Imports

1965 -73

1973 -80

1980 -87

1965 -73

1973 -80

1980 -87

1985

1987

Sub-Sarahan Africa 15.1 SS A excluding Nigeria 5.7 Low-income SSA excluding Nigeria 4.8 All low-income economies (world) 9.6

0.2 2.0

-1.3 1.7

3.7 3.4

7.6 2.9

-5.8 -1.5

91 92

84 84

0.6

-1.1

2.9

1.4

-2.5

91

84

2.3

3.5

1.1

8.0

2.3

92

84

Source: World Bank, 1989, page 241.

4

Public Enterprise in Kenya

Table 1-3 Measures of Debt for Sub-Sahaian Africa

Service on long-term debt as % of exports of goods and services Long-term debt as % of GNP Total debt as % of GNP

1970

1975

1980

1985

6 14

6 15

8 22 28

24 42 53

1987

18 85 100

Source: World Debt Tables, 1987-88, page 6.

intolerably high. Table 1-4 shows some examples. The result of declining export values and increasing debt burdens was a sharp compression in the volume of African imports—nearly 20 percent between 1980 and 1985. The results were grave. Shortages of spare parts and intermediate goods restricted economic activity, further depressing the volume of exports, worsening debt burdens, and reducing future growth prospects.5

Table 1-4 External Debt Situation of IDA-Only Countries Debt Service on Long-Term Debt (US$ mil) 1987 (actual) Benin Burundi C.A.R. Chad Comoros Gambia Kenya Madagascar Malawi Mali Mauritania Niger Rwanda Senegal Sudan Tanzania Togo Uganda Zaire Zambia

34 42 22 6 1 15 588 147 71 32 86 151 20 285 48 83 63 70 247 129

1988 (scheduled) 110 46 44 10 10 19 640 397 102 71 193 183 25 341 1,013 383 143 139 798 563

Source: World Debt Tables, 1987-88, page xxxvi.

1988 Scheduled Debt Service/1987 Exports (%) 52.1 41.8 22.3 6.1 33.3 16.7 36.8 95.7 33.5 21.8 40.8 57.0 14.1 26.4 142.3 85.3 31.6 37.5 41.3 59.0

Crisis in Africa & Search for Solu tions

5

Table 1-5 Central Government Deficit (as percentage of GDP)

ID A-eligible countries Other sub-Saharan Africa Total sub-Saharan Africa

1977-79

1980

1981

1982

1983

1984

-6.1 -4.4 -5.1

-7.1 -1.8 -3.7

-7.5 -7.6 -7.6

-7.9 -7.2 -7.5

-6.3 -9.0 -8.0

-5.6 -4.5 -5.0

Source: World Bank, 1986, page 58.

The effect of declining imports, exports, and growth rates of domestic production wreaked havoc on African governments' budgets. In 1972,29.7 percent of total current revenue for sub-Saharan African countries came from taxes on foreign trade. With foreign trade declining, government revenues fell as well. Revenues from domestic excise taxes and income taxes were also affected by the recession. Spending commitments were cut only after a lag, and the result was increased central government deficits, as shown in Table 1-5. Public enterprises played a part in these increased fiscal strains, as many of them experienced financial deterioration during the period, as will be shown in the Kenyan case in later chapters. Against this background of contraction of economic activity levels, excessive foreign borrowing and internal fiscal balance problems, African governments and the major lending agencies have naturally searched for adjustment programs that will provide an exit from this vicious circle. It is not surprising that a great deal of attention has focused on the public enterprise sector.

Public Enterprise in Africa Public enterprises are central to the development strategies of most sub-Saharan economies. Table 1-6 shows that during 1965-1980 the share of public enterprises in the GDP of various African countries ranged from a low of 6.8 percent to a high near 40 percent and shares in capital formation range even higher. Table 1-7 gives the share of public enterprises in domestic credit during the late 1970s. The shares were high, up to 87.1 percent in Guinea, with shares in the range of a quarter to a third common. If public enterprises are strategically located, their actual importance can exceed that indicated by the overall figures just cited. In most African countries the transport and communications sectors are heavily dominated by public enterprise. 6 Primary exports, both agricultural and mining, are also usually dominated by public enterprises. The strategic placement of public enterprises means that their actual importance probably far exceeds that indicated by their share in GDP.

6

Public Enterprise in Kenya

Table 1-6 Output and Investment Shares of Public Enterprises (percentages)

Country

Year

Benin Botswana

1976 1974-77 1978-79 1976-77 197M0 1978-80 1979 1965-69 1970-73 1974-77 1978 1979 1964-65 1966-69 1970-73 1974-77 1978-79 1973 1974-76 1977 1969 1970-73 1974-77 1978 1975-77 1978 1973 1974-77 1978-79 1977-79 1970 1974 1979 1964-65 1966-69 1970-73 1974-77 1978-79 1980 1972 1979-80

Ethiopia Gambia Guinea Ivory Coast

Kenya

Liberia Malawi

Mali Mauritania Mauritius Senegal Sierra Leone Tanzania

Togo Zambia

Source: Short, 1983, pages 9-11.

Share in GDP at Factor Cost 7.6 7.7 7.3

25.0

10.5 7.5 8.1 8.7

6.8

11.2 9.4

8.4 19.9 7.6 9.3 12.7 12.3 11.8 37.8

Share in Gross Fixed Capital Formation

16.5 7.7 17.6 36.5 37.9 16.5 27.9 28.8 38.2 39.5 9.7 13.0 10.6 18.1 17.3 24.3 14.1 20.2 29.4 28.1 21.2 20.9 7.6 85.8 31.1 37.2 14.4 17.9 19.6 9.2 22.7 48.2 30.3 16.3 49.7 61.2

Crisis in Africa & Search for Solutions

1

Table 1-7 Contribution of Public Enterprises to Growth in Domestic Credit (percentages) Public Enterprise Total Growth in Growth in Credit to Share in Total Domestic Credit Public Enterprises Domestic Credit (per year) (per year) at End-Period Benin Gambia Ghana Guinea Ivory Coast Malawi Mali Mauritania Niger Senegal Somalia Sudan Togo Upper Volta

1978-80 1978-81 1978-80 1978-80 1978-79 1978-81 1978 1978-80 1978-80 1978 1978-81 1978-81 1978-79 1978

30.4 29.3 35.1 6.0 18.2 25.4 11.0 7.4 71.7 31.1 40.0 25.5 22.4 35.5

31.6 62.4 66.8 12.0 16.4 6.7 17.6 6.5 81.5 34.3 17.4 31.1 18.5 13.2

54.2 40.5 20.7 87.1 29.0 14.7 37.9 7.9 3X4 38.7 37.9 22.5 22.4 27.8

Source: Short, 1983, pages 57-59.

There is ample evidence that public enterprises have performed poorly in a variety of dimensions. Nellis summarizes it as follows: African public enterprises present a depressing picture of inefficiency, losses, budgetary burdens, poor products and services, and minimal accomplishment of the non-commercial objectives so frequently used to excuse their poor economic performance. 7

In financial terms specifically, numerous countries' public enterprise sectors have performed poorly: In a sample from twelve West African countries, 62% showed net losses while 36% had negative net worth. . . . In Tanzania, between 1976 and 1979, one third of all public enterprises ran losses. . . . Cumulative public enterprise losses in Mali reached 6% of G D P by the end of the 1970s; a 1980 study of eight Togolese public enterprises revealed that their losses alone equalled 4% of GDP.®

Public enterprises are thus deeply implicated in the fiscal problems of African governments. The large losses earned by public enterprises have contributed to negative public savings, decreases in gross domestic savings, and decreases in investment rates. The magnitude of the overall trend is alarming, as shown in Table 1-8. Profitability has many limitations as a measure of public enterprise

8

Public Enterprise in Kenya

Table 1-8 Savings and Investment Shares in GDP (percentages)

Low-income Africa excluding Nigeria Gross domestic investment Gross domestic saving Sub-Sabaran Africa excluding Nigeria GDI GDS Total sub-Saharan Africa GDI GDS

1965

1980

1987

14 14

19 9

16 7

15 15

20 15

16 11

14 14

20 22

16 13

Source: World Bank, 1989, pages 226-227.

performance, since a firm may be efficient but unprofitable or inefficient and profitable. Various studies have found the average public enterprise to be inefficient as well as unprofitable. Killick reports that average labor productivity in Ghanaian public enterprises was only about half that of private enterprises, despite the fact that public enterprises were found in the most capital-intensive sectors. Value added per worker declined in Tanzania throughout the seventies. Gross output per unit of operating capital in public firms was only 40 percent of that for private firms. Turnover per worker in public manufacturing firms in Zambia also fell in the seventies.9 Public enterprises are often involved in price distortions. These may occur in either direction; in some commodities, prices are kept down either as an anti-inflationary measure or because the commodity is considered a necessity that must be affordable to consumers. In others, prices are kept high via monopoly or trade barriers, in order to protect inefficient enterprises or provide resources for cross-subsidization. Public enterprises are often criticized for not exhibiting enough entrepreneurial flair. Their activities may be circumscribed or distorted by their attempt to fulfill noncommercial political objectives. A cursory survey, then, suggests that economic performance of public enterprises has been unsatisfactory. Public enterprises on the whole seem to have been unprofitable and inefficient and to have contributed to misallocation of resources. Given the important role of public enterprises, which derives both from their contribution to GDP, employment, and investment and from their placement at many strategic junctures in African economies, it is undoubtedly true that before the overall crisis can be reversed the problems in public enterprises must be brought under control.

Crisis in Africa & SearchforSolutions

9

HOW DO WE BRING PUBLIC ENTERPRISES UNDER CONTROL?

Two different approaches to bringing the public enterprise problem under control have gained great currency in the wake of the African crisis. Privatization of public enterprises is being pushed strongly by the World Bank and USAID, agencies on which the African countries depend heavily for external support. Centralization is favored by many African policymakers. The pressure from the donor agencies can be seen in several ways. Most IMF and World Bank structural adjustment loans contain agreements to restrain the public budget.10 Since many public enterprises depend on budget support, the curtailing of such support inhibits further public enterprise investment. Structural adjustment agreement limitations on public spending sometimes lead governments to run up arrears with major public enterprises, contributing to their collapse.11 Some structural adjustment loans have carried explicit agreements to privatize public enterprises.12 Among donors USAID has been the most blunt in its opposition to public enterprise. In 1985 Secretary of State George Schultz instructed USAID officials on major issues they should raise with LDC governments: Policy dialogue should be used to encourage LDCs to follow free market principles and to move away from government intervention in the economy. . . . To the maximum extent practical governments should rely on the market mechanism—on private enterprise and market forces. . . . Parastatals are generally an inefficient way of doing business In most cases, public sector firms should be privatised. 3

Many African policymakers resist the demand for privatization and focus instead on how to bring existing public enterprises under control. For them it is a short jump from the feeling that the enterprises' deficits are out of control to the view that the enterprises' managers are out of control. Thus, they have concentrated their efforts on tightening the control the government exercises over public enterprise managers. The present study of Kenya provides new data on the nature of the public enterprise problem and shows when and why major problems developed. Based on this analysis, which is far more detailed than that hitherto available for any African country, the prospects for the success of these two reform strategies are slim. It will be shown in the following chapters that privatization is not a panacea that can rescue Africa from either the public enterprise problem or the broader malaise in which it finds itself. To improve efficiency and profitability, it would be more effective to begin with detailed sectoral reforms of the small number of firms that constitute the bulk of "the public

10

Public Enterprise in Kenya

enterprise problem." The results of this study contain both good news and bad news for African policymakers. The good news is that there is hope: parastatals can work and have worked well for long periods and under difficult conditions. The bad news is that current African approaches to the problem, based on increased centralization of control, are likely to fail, and are in fact likely to make the problem worse. NOTES 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13.

World Bank, 1981, page 3. Christensen, page 80. World Bank, 1989, page 20. Faber and Green, page 16. Lancaster and Williamson, page 4. Short, page 22. Nellis, page ix. Nellis, pages 19-20. See Killick, 1983, pages 73-75. See Zulu and Nsouli. See Barad, page 39. See Killick and Commander, page 1467, and Barad, page 40. Quoted in Killick and Commander, page 1467.

The Public Enterprise Problem in Kenya The colonial government in Kenya established many public enterprises because they were believed to be the most efficient mechanism for providing certain services. These public services encompassed those that were not provided by the private sector or that were of a monopolistic nature. It was considered undesirable to leave monopsonies in private hands because of the exploitation that would have resulted. The Kenyan government created new enterprises following independence both for these same reasons and to help promote Africanization and regional income redistribution. In this chapter I examine briefly the philosophy of the Kenya government toward public enterprise, its perceptions of the nature and causes of the public enterprise problem, and its responses to it. In the second section I present a framework for how to analyze parastatal performance, considering that their performance is inherently multidimensional. The chapter concludes with a preview of the results from Chapters 3 through 6. PUBLIC ENTERPRISE IN KENYA—A SHORT HISTORY

As in most African countries, one of Kenya's colonial legacies was a large public enterprise sector. Table 2-1 shows the chronology of the country's major public enterprises. During the colonial period the infrastructural services, which are mostly natural monopolies, were organized as public enterprises. This included the ports and railways, airlines, and posts and telecommunications. Another major group of public enterprises set up during the colonial period was the crop marketing boards. These boards were organized primarily for the benefit of the white settler farmers, whose produce they marketed. Most of them had strong cooperative elements, including heavy grower representation on the boards of directors. They handled most agricultural exports, including coffee, pyrethrum, and meat and dairy products. Their role was to organize processing and overseas marketing as efficiently as possible, so that farmers could accrue the largest possible incomes. Hence, they complemented private European farming and did not compete with it or supplant it. It was no doubt considered desirable 11

12

Public Enterprise in Kenya

Table 2-1 Sequence of Major Public Enterprises in Kenya

Firm Uganda Railways (after several incarnations, currently Kenya Railways & Kenya Ports Authority) Coffee Board of Kenya East African Posts & Telecommunications (eventually Kenya Posts & Telecoms) Kenya Cooperative Creameries Uplands Bacon Factory Kenya Meat Commission Wheat Board of Kenya Pyrethrum Marketing Board East African Airways (eventually Kenya Airways) Industrial Development Corporation (later ICDC) Cotton Lint & Seed Marketing Board Maize Marketing Board Development Finance Company of Kenya Kenya Tea Development Authority Agricultural Finance Corporation Agricultural Development Corporation Housing Finance Company of Kenya Kenya Tourist Development Corporation National Construction Corporation East African Sugar Industries Maize & Produce Board Kenya National Transport Company Kenya Industrial Estates Horticultural Crops Development Authority National Bank of Kenya Chemelil Sugar Company Kenya Commercial Bank East African Power & Lighting (later Kenya Power & Lighting) Mumias Sugar Company Industrial Development Bank Kenya Pipeline Company Nzoia Sugar Company South Nyanza Sugar Company National Cereals and Produce Board

Year of First Incarnation or Nationalization 1903

1920s ? 1925 1946 1950 1952 1950s 1950s 1954 1955 1960 1963 1963 1963 1964 1965 1965 1966 1966 1966 1966 1967 1967 1968 1968 1970 1970 1973 1973 1975 1979 1980 1980

The Problem in Kenya

13

to keep the marketing boards out of the hands of private middlemen who could have extracted monopsony rents. Growers enjoyed disproportionate political power, so the state-owned marketing boards refrained from such extraction themselves. This distinguishes marketing boards in Kenya from those in other African countries where they were used primarily as tools to extract such surplus.1 The boards served purely an efficiency function—they did not engage in any important forms of cross-subsidization. As independence approached, the government set out under the Swynnerton plan to develop progressive middle-class African farmers. The existing marketing boards reoriented themselves to serve large numbers of smallholders, and some new boards were created (including the Cotton Lint and Seed Marketing Board and the Kenya Tea Development Authority). As Africans were incorporated, the role of the boards remained unchanged, costs were minimized, surpluses were paid to farmers in the form of the highest producer price consistent with marketing board financial health, and cross-subsidies were avoided. In addition to infrastructural public enterprises and agricultural marketing boards, there was a small participation in the manufacturing sector, under the auspices of the Industrial Development Corporation. Much of this activity dated from the war years, when manufactured imports were curtailed by wartime shortages in Europe. At independence the public enterprise sector contributed 11.2 percent of GDP. This share rose slowly until 1971, when it reached 14.4 percent. It then began a slow decline until 1977. Unfortunately, the Central Bureau of Statistics stopped publishing the series after that, so we cannot say with as much confidence whether the downward trend has continued. Other available measures suggest that the downward trend has probably continued and may have accelerated. Several measures of the role of public enterprise in the Kenyan economy are shown in Figure 2-1. The share of public enterprises in capital formation has been less stable than the share in GDP. In the first decade of independence there was no clear trend. The share of public enterprises in capital formation was sometimes higher and sometimes lower than their share in GDP. In 1974 the series was reclassified.2 The share of public enterprise in capital formation drifted slightly upwards from 1974 until the coffee boom. Since then it has fallen markedly, though the fall was interrupted in 1982. 3 Since 1982 the share of public enterprises has fallen precipitously. The most volatile measure of the importance of public enterprises has been their share in the development budget. This measure fluctuated widely from the mid-1960s until 1980. Since 1980 it has exhibited a steady decline. The data are not quite conclusive, because we are missing the share

14

Public Enterprise in Kenya

Figure 2-1 Public Enterprise Shares

Capital formation GDP Development budget

64

67

70

73 76

79

82

85

of GDP since 1978. But we can generalize the available information this way: the role of public enterprises was essentially stable from independence until around 1978 or 1979. Since that period the role of public enterprises has gradually declined. Several major events in Kenyan history occurred at that time. President Jomo Kenyatta died in late 1978 and was succeeded by Daniel arap Moi. The boom in coffee and tea prices ended in 1978 to be followed by the second oil price shock in 1979. Warning signs had appeared as early as the first oil crisis, but for Kenya the economic crisis began to cause serious discomfort only from 1979. Since then Kenya has been caught up in structural adjustment loans from the World Bank and balance of payments borrowing from the IMF, with all the conditionally such arrangements usually imply. Kenya has responded to economic crisis in part by curtailing investment in the public enterprise sector.

The Problem in Kenya

15

In 1965, shortly after independence, the new government expressed its philosophy about the role of the state in the economy in the famous Sessional Paper No. 10, "African Socialism and Its Application to Planning in Kenya." The government committed itself to promoting rapid growth and equitable distribution. The possible conflict between these two goals was acknowledged, and some fairly clear guidelines for how the tradeoffs would be made were spelled out: The most important of these policies is to provide a firm basis for rapid economic growth. Other immediate problems such as Africanization of the economy, education, unemployment, welfare services, and provincial policies must be handled in ways that will not jeopardize growth. The only permanent solution to all of these problems rests on rapid growth. If growth is given up in order to reduce unemployment, a growing population will quickly demonstrate how false that policy is; if Africanization is undertaken at the expense of growth, our reward will be a falling standard of living.4

In order to achieve high growth, high capital formation was needed, higher than could be financed through domestic savings alone. In such a scheme, nationalization made little sense, because it created no new assets, absorbed available state funds, and might cause flight of private capital. Nationalization, then, would be carried out only in very limited circumstances. In line with this commitment to growth, firms that were nationalized were expected to operate on commercial principles: It must also be clear that in most cases when an industry is nationalized it must be operated efficiently, cover its costs and earn a profit at least equivalent to the taxes paid when operated privately.5

While the new government was in no rush to nationalize existing firms, it had no qualms about creating new ones, either as sole proprietor or in joint ventures with private capital. The purpose of most of the newly created firms was to accelerate Africanization of the economy. In Kenya, in contrast to neighboring Tanzania, Africanization was perceived chiefly in terms of aiding the development of private African business. Especially popular were firms that provided credit and technical assistance of some sort. Thus, the Agricultural Finance Corporation provided finance to help Africans acquire and develop farms, the Kenya National Trading Corporation helped them enter commerce, the Industrial and Commercial Development Corporation helped them enter commerce and industry, the National Construction Corporation helped them enter construction, and the Housing Finance Corporation of Kenya helped them acquire and develop urban real estate. The new agricultural firms, including the Kenya Tea Development Authority and the five sugar

16

Public Enterprise in Kenya

companies, were all designed to foster and promote commercial farming among smallholder Africans. Even the Agricultural Development Corporation, which operated state farms, did so largely for the benefit of private farmers. Its research led to commercialization of new crop varieties and breeding stock for improved breeds. The general principles governing public enterprise laid down by the Kenyan government were these: they should be efficient at whatever line of business they were engaged in, and efficiency was not to be sacrificed for other goals; they should be financially solvent; and they should foster growth and development of the private sector, especially among African entrepreneurs—but this private sector activity was expected to be economically efficient and commercially viable. Since the major purpose of most of the public enterprises was to foster private sector activity rather than to grow themselves, high rates of return were not in general considered desirable. Such was the theory of how public enterprises should work, but clearly not everything went according to plan. Politicians were appointed to the boards of directors of many public enterprises, and allegations of nepotism and corruption were made periodically. There were demands to Africanize public enterprise management as fast as possible, and in some cases promotion of African managers proceeded at a faster pace than their training, with resulting declines in efficiency. There were demands for various types of cross-subsidy, and these were not always resisted. Import substitution was extended into areas where Kenya had no comparative advantage. In 1979 the new president appointed the Committee on Review of Statutory Boards to review and make recommendations with regard to urgent financial, administrative, and operational problems facing important boards. The committee found widespread evidence that all was not well among public enterprises: "There is clear evidence of prolonged inefficiency, financial mismanagement, waste and malpractices in many parastatals."6 The only specific evidence offered to illustrate the poor performance was a low overall rate of return on investments in public enterprises. Although the nature of the parastatal problem was not delineated very clearly, several causes of the problem were identified, especially poor selection of directors and senior staff and lack of proper financial control procedures. It is around these two problems that most of the recommendations of the committee revolve. Passing mention was made of undercapitalization and ineffective government supervision. N o recommendations were made regarding the former. Government supervision was discussed only in terms of a policing function, with little attention given to more positive aspects such as definition of mission,

The Problem in Kenya

17

setting of goals, or identification of necessary resources. No mention is made anywhere in the report of government regulations or directives that might affect parastatal performance. Market structure and competitive conditions are not mentioned, nor are sectoral policies that might affect public enterprises' ability to operate. The report, then, leaves the impression that the chief problem is corrupt and inefficient managers. The main recommendations follow directly. Parastatal managers should be brought under tighter control of central government. Terms and conditions of service should be harmonized with those of the civil service and managers made transferable between parastatals and between parastatals and the civil service. Salary harmonization was implemented via an annual series of circulars from the Office of the President, beginning in 1981. The circulars set maximum salaries of chief executives. The maxima set in 1981 were, for many firms, lower than salaries already in effect. Incumbents were allowed to continue to receive their former salaries, but when a position turned over, the new officer was to be paid in accord with the guidelines. Firms were left to harmonize their internal salary structures, subject to the approval of their parent ministries and the Parastatals Advisory Committee, later the State Corporations Advisory Committee. This regulation greatly magnified the role of government in setting terms of service and approving wage negotiations of all parastatals. Because the guidelines harmonize parastatal salaries with civil service salaries rather than with the private sector, some parastatals have found themselves in anomalous positions. For example, unionizable wage employees may earn more than the lower level managers who supervise them. The effect of this on management morale can be easily surmised. Uniformity and control was to be attempted in the sphere of financial management as well. Each enterprise would be required to prepare annual operating budgets and forward budgets, presented on the same fiscal year and using the same format. Procurement procedures were to be prescribed by the minister of finance7 and procurement outside the approved budget prohibited.8 Operating surpluses were to be surrendered to the government.9 The controls on budgeting and procurement were implemented via a series of circulars from the Government Investments Division of the Ministry of Finance. Until the State Corporations Act was passed in 1986, most firms ignored them. That highly punitive act, one quarter of whose sections dwell on the details of how managers and board members can be prosecuted personally for failure to comply with the many directives they receive, has elicited compliance.10 Recommendations were made for strengthening the government supervisory apparatus. These measures focused on such matters as account-

18

Public Enterprise in Kenya

ing, cost control, budgeting, and financial reporting. T h e work of the 1979 Ndegwa committee was followed up by the 1982 Working Party on Government Expenditure. T h e working party again found major problems among public enterprises: Examples of unsound and poorly controlled investments can readily be found in such areas of activity as fertilizer, sugar, textiles, and power alcohol. The amounts involved are of such a magnitude that if they had been directed toward the development of essential rural infrastructure, several districts could have been radically transformed in terms of both production and employment. 11

Once again, the only specific evidence of poor performance mentioned was financial.12 It is perhaps symptomatic of the financial crunch in which the government found itself by 1982 that even its financial focus had narrowed. Instead of reporting measures of financial rates of return, it discusses only flows of interest and dividends. Returns reinvested in productive assets were apparently of only secondary interest to a government strapped for cash. T h e report of the 1982 Working Party on Government Expenditure took more cognizance of the fact that much of the responsibility for poor financial performance lay with central government, but it offered f e w recommendations that might help alleviate these problems. T h e working party, highly critical of what it viewed as overextension of public enterprise into sectors that were strictly commercial, called for a program of divestiture. Had the working party's recommendations with regard to divestiture been followed, public enterprise would have been confined to enterprises with important social mandates, with the need for coherent government policy all the more pressing. T h e report did recommend that "the Government should not direct a parastatal to carry out policy-related activities which might not be financially sound without providing explicit subsidies for those activities." 13 However, there was no discussion of how this recommendation could be implemented or of how the problems originating in government policy could be resolved.

ANALYZING PUBLIC ENTERPRISE PERFORMANCE—A FRAMEWORK In order to assess public enterprise performance, it is necessary to recognize that public enterprises serve a multiplicity of objectives. 14 Each firm is the focus of a set of pressure groups that push for policies regarding the firm that will be favorable to them. Firm behavior can be characterized as a vector of outcomes (output prices, input prices, efficiency, profitability, etc.), some of which are of interest to each of the pressure groups. T h e

The Problem in Kenya

19

firm's behavior will depend on the relative strength of the different pressure groups. The most important pressure groups that act on public firms are: • consumers, who press for low prices; • suppliers of inputs, who press for high input prices and procurement from themselves; • competitors, who press for the firm to charge high prices and restrict services offered; • employees, including management, who wish to ensure that the firm generates a continuing stream of surplus from which it can appropriate a portion, either directly or indirectly; and • shareholders, who press for the firm to earn profits. Each of these interest groups can seek to exert influence through a variety of means. They can act through the market, they can petition the firm, or they can petition government, either alone or in concert with one of the other interest groups. Organizational structures particular to the firm will provide other channels. Suppliers may be given places on the board of directors; employees may organize for collective bargaining. In addition to its role as shareholder, government has other interests in public enterprise. The other interest groups are constituents of government, and their demands must be mediated by government. In addition, government may have interests that override those of the immediate coalition members that make up any one firm. It follows from this model of the public firm that there is no one index of performance. The interests of the different interest groups conflict to some extent. For example, if consumers manage to get low prices, they will do so at the expense of some other party, either shareholders, suppliers, or employees. Likewise, high profits will please the shareholders but come at the expense of consumers, suppliers, or employees. I have relied mostly on data from the audited annual accounts of the parastatals themselves, supplemented by data from government publications such as the Statistical Abstract. I also interviewed managers of some of the firms, as well as officers of supervisory bodies such as the Government Investments Division of the Ministry of Finance, the Inspectorate for Statutory Boards (later the Inspectorate for State Corporations) in the Office of the President, the Kenya Sugar Authority, the Ministry of Transport, etc. Following the model, I calculate four different measures of performance. The first is financial rate of return on investment, defined as before-tax profits plus interest expense divided by total long-term investment. I believe that it is the best single measure of financial performance,

20

Public Enterprise in Kenya

and it provides a fair comparison for firms with different financial structures. However, differences in the rate of return on total investment (measured as described above) and the rate of return on equity (before tax profits divided by net worth) reveal important information about the adequacy of the firms' financial structures, so I report both measures where possible. Definition of adequate financial performance is debatable. A simple comparison with the opportunity cost of capital elsewhere in the economy is nonsensical as a measure of an individual firm's performance, because many Kenyan public enterprises are instructed in their enabling statutes to break even, after making provision for depreciation and finance costs. In other words, they are given target rates of return of zero return on equity. Their target rate of return on total investment will then depend on the type and source of their finance. If they have a high equity/debt ratio and a mature organization, their target rate of return may be near zero. This is the case with some of the agricultural firms. On the other hand, if they live on overdrafts at a 16 percent rate of interest, their target rate of return is much higher. Much of the parastatal sector has been financed from loans from international lending agencies such as the World Bank, some of which carry lower interest rates and hence give intermediate target rates of return. Some firms operate in sectors where there is need for rapid capital accumulation, so that higher returns to permit self-finance is desirable. No one rate of return is adequate for all firms. Broadly speaking, to be adequate the rate of return on equity must be positive, and the firm must be solvent and liquid. If this is achieved with low rates of return to equity, it makes little sense to criticize the firm for poor performance, since the enterprise is doing exactly what it was instructed to do. Somewhat different criteria of adequate performance are appropriate to each sector, and these will be discussed in each chapter. The next measure of performance is efficiency. Several different approaches to measuring efficiency are possible, depending on the different types of data available. In the agricultural sector a number of firms were mature at independence, providing what was considered to be good service, at a time when the employees did not constitute a group that had the ability to make claims on the firm over and above the opportunity cost of their labor. In these cases containment of unit cost margins is a valid criterion of adequate performance. In some cases there are several firms in the same sector, so efficiency levels and standards of performance can be compared. In some cases efficiency can be based on international data. A number of different measures, which will be explained and justified within the different chapters, are necessary. Efficiency is not only an indicator of rational resource allocation, but also an indicator of returns to employees of the public enterprise. Many

The Problem in Kenya

21

people, including much of the Kenyan civil service, believe that the major reason for poor performance is corrupt and inefficient managers, and hence that greater control over the managerial function is essential to improved performance. Managerial corruption or inefficiency will show up in the accounts of the firm as inflated unit costs. If the accounts show efficient operations, we can reject the hypothesis that corrupt and inefficient managers are the major sources of whatever problems exist. The converse is not true; there are many possible causes of inefficiency, so that a finding of inefficiency requires more information to establish its causes. The next measure of performance is returns to consumers. We generally measure this by comparing actual prices with some measure of the opportunity cost of the good. For tradable goods this opportunity cost is measured by the relevant international parity price. In other cases market share information provides clues as to whether the parastatal is offering prices to consumers that are competitive with prices offered by the private sector. The final measure of performance is returns to suppliers. Like returns to consumers, this indicator compares actual prices with opportunity costs. At times indirect measures, such as the ability of public firms to attract inputs in competition with private firms, are also useful. SUMMARY OF RESULTS

Table 2-2 summarizes the performance of public enterprise in Kenya on all four of our measures of performance. Financial and efficiency records that were judged to be adequate using the criteria discussed above are indicated as OK. Inadequate performance on these measures is indicated by the year in which the poor performance trend began. Prices are characterized along two dimensions. In columns 4 and 5 we show whether on average, over the years, there is evidence that prices have been consistently above or below the opportunity cost, as discussed in the last section. Column 6 shows whether prices have been fixed by administrative fiat or left flexible to fluctuate in response to supply and demand conditions. Table 2-2 summarizes a large volume of information, which is discussed in more detail in the sectoral chapters. Without attempting a complete argument, which must await that more detailed discussion, I will summarize the major conclusions. Some, but not all, of the conclusions can be derived from Table 2-2. Conclusion 1: There is a wide range of performance among Kenyan public enterprises, from excellent to abysmal. Use of average figures to characterize public enterprise performance obscures that range. The range of Kenyan experience with parastatals offers lessons relevant to

22

Public Enterprise in Kenya

Table 2-2 Performance of Public Enterprises in Kenya

Producer Financial" Efficiency" Priceb Agric. Dev. Corp. Agric. Finance Corp. Chemelil Sugar Co. Coffee Brd. of Kenya Cotton Mktg. Brd. Dev. Finance Corp. of Kenya East African Sugar Ind. Housing Fin. Co. of Kenya Ind. & Com. Dev. Corp. Industrial Dev. Bank Kenya Airways Kenya Commercial Bank Kenya Coop. Creameries Kenya Meat Commission Kenya Nat. Transport Kenya Pipeline Co. Kenya Ports Auth. Kenya Posts & Telecoms Kenya Power & Lighting Kenya Railways

1979 1965d OK OK 1978 OK OK OK OK OK 1978 OK OK 1971 1980 OK OK OK OK 1972

1967 1967" OK OK 1963 OK OK OK OK OK

OK 1981

Kenya Tea Dev. Auth. Kenya Tourist Dev. Corp. Maize Mktg. Brd. Maize & Prod. Brd. Mumias Sugar Co. National Bank of Kenya Nat. Cereals & Prod. Brd. Nzoia Sugar Co. Pyrethrum Mktg. Brd. South Nyanza Sugar Co. Uplands Bacon Factory Wheat Mktg. Brd.

OK 1968 1963d 1977 OK OK 1980 1979 OK 1980 1978 OK

OK 1967 1963d 1977 OK OK 1980 1979 OK 1980 1966 OK

OK 1963 1971 OK 1982

LOW HIGH COMP COMP COMP' HIGH COMP6 COMP COMP® COMP COMP COMP COMP COMP COMP COMP COMP COMP COMP COMP HIGH COMP COMP HIGH COMP HIGH COMP LOW

Consumer Price" LOW LOW HIGH COMP COMP HIGH COMP6 COMP COMP COMP COMP' COMP COMP COMP HIGH COMP COMP LOW ORCOMP COMP LOW COMP COMP HIGH COMP COMP HIGH COMP HIGH LOW LOW

Price Setting Mode c FIXED FIXED FIXED FLEX FIXED FIXED FIXED FIXED FIXED FIXED FLEX FIXED FIXED FIXED FLEX FIXED FIXED FIXED FIXED FIXED FLEX FIXED FIXED FIXED FIXED FIXED FIXED FIXED FLEX FIXED FIXED FIXED

Notes: a. OK indicates adequate performance. Date indicates year when adverse trend began. b. Indicates whether firm's price is HIGHer than competitive price or opportunity cost, LOWer, or COMPetitive with that price. Determination of competitive price is discussed on page 8. c. FIXED signifies price set by administrative fiat; FLEX signifies price left free to fluctuate with supply and demand. d. First year data available; problem may have existed earlier. e. Shown as competitive because equal to price paid by private banks, but regulated price below market clearing price.

The Problem in Kenya

23

improved performance in the future. Simplistic use of averages obscures the lessons of that history. When assessed on measures of profitability and efficiency, about half the firms have performed well throughout the period, some for a quarter century since independence. Most are fairly large and complex organizations, dealing with clienteles in the thousands, which makes the successes achieved all the more impressive. Some of the firms have weathered unfavorable market conditions, both domestic and international, but unfavorable external conditions have not resulted in institutional decay. The fact that so many firms have performed well for substantial periods of time contradicts the conventional wisdom on solutions to the parastatal problem. If the goal is to improve performance of public enterprise and the economy generally, there is no need to look to wholesale privatization as a solution. Public firms can perform well. Not all of them do, but the conditions for good performance are attainable within the context of the politics and administrative capabilities of an African state. Conclusion 2: There are important correlations between performance measures. There is no a priori reason for the four indicators to move together, so it might have been impossible to classify firms as successes or failures. However, performance in Kenyan public enterprises falls into a surprisingly small number of categories. Some firms perform well on efficiency measures, adequately on financial measures, and give competitive returns to their consumers and suppliers. Firms that have attempted to subsidize their consumers or suppliers have usually run into financial problems. Their poor financial condition causes illiquidity, which leads to inefficiency. The firm is ultimately not very successful in serving the consumers or suppliers it attempts to subsidize. Thus, problems tend to come in groups, so that firms can be characterized as successful or unsuccessful. Two reasons explain the high correlation of different measures of performance. First, budgetarily firms are expected to be fairly autonomous. Firms in financial problems do not receive prompt subsidies, so that operating problems become inevitable. Second, public firms in Kenya have generally not succeeded at lobbying for effective monopoly powers. Except for true natural monopolies, firms with problems have been vulnerable to competition from private firms less fettered by regulation. Conclusion 3: Public enterprises in Kenya have not been used extensively to subsidize suppliers or consumers via prices. The policy declared in Sessional Paper No. 10 stating the priority of growth over distribution has not always been observed, but it has been observed more often than not. There is no doubt that the clear-eyed policy of generally allowing

24

Public Enterprise in Kenya

public enterprise prices to cover costs contributed to the successes that have been achieved. Many of the problems that have existed are found where that policy has been ignored. The matter of flexibility in pricing is more subtle. There have been cases where prices were set at levels that were reasonable in the long run, but that were out of line with supply and demand conditions in the short run. Such inflexible prices have caused problems in at least two ways: (1) they have led to larger variation in volumes, requiring larger capacity and higher unit costs, and (2) they have hobbled public enterprises in competing with private firms. The latter factor has been especially important in the agricultural sector and will be discussed in more detail there. Conclusion 4: Many problems observed are of recent vintage, since 1978. While it is not observable from Table 2-2, many of the problems arising since 1978 are due to the tighter economic conditions in that period, as will be argued in the sectoral chapters. To blame poor public enterprise performance for causing the crisis may be to reverse cause and effect. The general economic problems have been passed on to public enterprises in several different ways. One is undercapitalization. Several firms started in this late period were begun with inadequate working capital and capital structures that were unrealistic for new firms. If performance of these firms is to be improved, they must be properly capitalized; treating them as though the problems were due mainly to corrupt and inefficient managers is a policy that cannot succeed in eliciting improved performance. Foreign exchange shortage has also contributed to poor performance. The devaluation of the shilling increased financial costs of firms with large overseas borrowing. Foreign exchange rationing caused shortages of spare parts, which meant idle capacity and higher unit costs. Again, it will be argued that treating firms in trouble because of these problems as though the problems were due mainly to corrupt and inefficient managers is a policy that cannot succeed in eliciting improved performance. Conclusion 5: Firms are often required to undertake investments or provide services that are motivated by distributional rather than commercial criteria, and firms are seldom promptly reimbursed for doing so. This will depress performance on profit and efficiency measures in any case, and where the firm competes with private firms it can cause irreparable damage to the firm. It is essential that the government learn to respect the very real limits that dependence on revenues from sales puts on firms' ability to finance social spending. Once again, treating firms in trouble because of these problems as though the problems were due mainly to corrupt and inefficient managers is a policy that can only fail. Conclusion 6: Use of overall financial returns gives a deceptively negative summary statistic of public enterprise performance. Firms have

The Problem in Kenya

25

been instructed to achieve target rates of return, not to maximize them. High rates of return have been subject to criticism, because such returns imply taxing consumers or suppliers. Hence, there will be few firms with high returns to counterbalance those few firms with severe problems. If the average is used to characterize the representative firm, it gives a misleading impression. It is possible for most firms to achieve their targets with the average still appearing unacceptable. These conclusions place the problem of Kenyan public enterprise in perspective. There are firms with serious problems, problems of a magnitude that is unacceptable. Since many such firms will remain in the public sector in the foreseeable future, it is essential to understand the problems that plague their performance. Solutions based on ideology or casual empiricism are inadequate. Policy toward parastatals in the 1980s seemed to ignore this evidence. It was predicated on the assumption that most problems are caused by corrupt or inefficient managers, and that appointment of such managers will continue to occur as a rule. Without denying that such managers are found in Kenyan parastatals, many other factors are the primary causes of the problems observed. These other problems are primary because until they are resolved, attempts to clamp down on managers cannot succeed in eliciting better performance. If resolved they could do much to improve performance, even if no greater pressure were applied to managers. Most of these other problems originate in decisions taken by civil servants who regulate public enterprises. Policies that centralize more and more decisions out of enterprises and into ministries will exacerbate these primary problems. They will also demoralize those public enterprise managers who are doing their best to operate firms under difficult circumstances. Current policies intended to improve parastatal performance could misfire badly, causing performance to further deteriorate. NOTES 1. See Arhin, Hesp, and van der Laan, and Bates. 2. In 1974 the series was reclassified. From 1964 to 1973 I have reported the totals labeled "East African Community" and "Parastatal Bodies" under the larger heading "Contribution of the Public Sector to Capital Formation." Since 1974 the subheading has been labeled "Enterprises and Non-Profit Institutions" under the same larger heading. While the jump in the level of the series in 1974 is not easily explained, it probably does not represent a genuine increase in the share of public enterprises in capital formation. Such a jump would eventually have resulted in an increased share of public enterprises in GDP, and no such increase occurred. 3. This temporary increase in the share of parastatals in capital formation in 1982 came because of a drop in private investment the same year. This drop is probably connected with the coup attempt in that year, which resulted in an emigration of residents of Asian origin and a drop in private investment.

26

Public Enterprise in Kenya

4. Republic of Kenya, 1965, page 18. 5. Ibid. 6. Republic of Kenya, 1979, page 3. 7. See Section 13 of the State Corporations Act of 1986. 8. See Section 12 of the State Corporations Act of 1986. 9. See Treasury Circular No. 12, January 25,1985. Profits may be retained with permission from government. Under the State Corporations Act, surpluses "shall be disposed of as the Minister, in consultation with the Board, may, in writing, direct." See Section 16. 10. At this writing, the act had never been enforced, and according to an interview with the inspector for state corporations in 1987, the inspectorate has no intention of ever enforcing it. Nonetheless, it stands as law and could be used at any time. 11. Republic of Kenya, 1982, page 42. 12. Ibid., page 41. 13. Ibid., page 49. 14. See Pfeffer and Salancik for a discussion of firms as coalitions of actors, each of whose needs must be met to remain in the coalition. See Aharoni for a discussion of the multiplicity of goals public enterprises serve.

The Agricultural Sector The agricultural sector contains a number of the biggest and most visible parastatals in Kenya, as well as some of its worst disasters. The sector has shown profoundly mixed results. About half the firms have performed adequately on efficiency and financial measurements. Numerous performance indexes will be presented as a basis for discussion of which policies are responsible for the differing results.

AGRICULTURAL PARASTATALS—BACKGROUND

The 16 agricultural firms included in this chapter have strong similarities in structure and mission, which makes comparison of their performance especially informative. Yet, important differences in market structure, regulatory environment, or historical circumstances have caused major variations in performance. Hence, the agricultural sector is an especially good arena in which to study the factors that lead to good and bad performance. The structure and mission of the agricultural firms are similar. These firms are trading and marketing firms only. None engages in agriculture directly; all purchase agricultural produce from farmers and groups of farmers. Each transports, stores, processes, and markets the produce of thousands of farmers. Table 3-1 gives a list of the firms, together with the years in which they operated and the years covered in the study. Within that similar structure, several important things that differentiate the firms cause their performance to differ. First, some firms handle crops are mostly or entirely exported. The prices the firms receive for their products are set in world markets, are not known in advance, and are not susceptible to political or bureaucratic manipulation. All of these firms are allowed to use a flexible pricing system in setting the prices they pay farmers (hereafter referred to as producer prices or supplier prices). Producer prices are determined as a residual; after revenues from sales are known, the firms' costs (including processing and transportation) are subtracted and the balance paid out to suppliers. 27

28

Public Enterprise in Kenya

Table 3-1 Firms Included in the Study Sample

Acronym Chemelil Sugar Company Coffee Board of Kenya3 Cotton Lint & Seed Marketing Board East African Sugar Industries Kenya Cooperative Creameries Kenya Meat Commission Kenya Tea Development Authority Maize Marketing Board Maize & Produce Board Mumias Sugar Company National Cereals & Produce Board Nzoia Sugar Company Pyrethrum Marketing Board South Nyanza Sugar Company Uplands Bacon Factory Wheat Board of Kenyab

Chemelil CBK CLSMB EASI KCC KMC KTDA MMB MPB Mumias NCPB Nzoia Pyboard SONY Uplands Wheat

Years of Operation 1968-present 1920s-present 1955-present 1966-present 1930s-present 1950-85 1963-present 1960-66 1967-80 1973-present 1980-present 1979-present 1950s-present 1980-present 1946-85 1952-80

Years Included 72-88 63-87 63-87 72-88 63-85 63-84 63-88 63-66 67-80 73-88 80-87 79-85 63-87 80-88 46-84 72-80

Notes: a. Before 1971 the coffee industry was handled by two firms, the Coffee Board of Kenya and the Coffee Marketing Board. In 1971 the latter was merged into the former. For comparability the accounts have been consolidated for previous years. b. Most of the functions of the Wheat Board were carried out by the Kenya Farmers' Association. Before 1972 the Wheat Board accounts presented only the accounts for its own administrative costs, excluding trading results. Therefore, most of the tables present data only after 1972.

In contrast, the rest of the firms handle crops that are primarily or entirely consumed domestically. The price controller sets prices for domestic food traded by the parastatals in this study. The prices the firms pay suppliers are generally set in advance of the crop year by the Ministry of Agriculture. Table 3-2 shows which kind of market each firm serves. A second factor that differentiates the agricultural firms is their monopsony power in their input market. Table 3-3 shows the degree of monopsony power each firm holds, with comments on its source. While farmers' obligation to sell to parastatals varies by crop, each parastatal is obligated, in theory, to buy all produce offered to it. Finally, the firms in the sample vary in age, as shown in Table 3-1. Some were old at independence, others were started as late as 1980. While maturity is no guarantee of health, as will be seen below, all of those firms started after 1978 have suffered severe financial and operating problems. These problems are due largely to severe undercapitalization at the firms' inception.

The Agricultural Sector Table 3-2 Type of Market Served and Manner of Price Formation Firm

Market

Chemelil CBK CLSMB EASI KCC KMC KTDA MMB MPB Mumias NCPB Nzoia Pyboard SONY Uplands Wheat

domestic export export turned domestic domestic domestic domestic turned export export domestic domestic domestic domestic domestic export domestic domestic domestic

Price Formation fixed flexible fixed fixed fixed fixed flexible fixed fixed fixed fixed fixed flexible fixed fixed fixed

Table 3-3 Nature of Monopsony Power Firm

Source & Degree of Monopsony Power

Chemelil EASI Mumias Nzoia SONY

Near total monopsony on regional basis due to monopoly on processing facilities and high transport costs. Small diversion of cane to private jaggeries.

KTDA Pyboard CLSMB

Near total monopsony due to monopoly on processing facilities.

CBK

Near total monopsony due to legal restrictions.

MMB MPB Wheat NCPB KCC Uplands KMC

Legal monopsony, but existence of large parallel markets.

No legal monopsony. Vigorous competition from private sector.

30

Public Enterprise in Kenya

FINANCIAL PERFORMANCE Trends in Financial Condition Figure 3 - l ( A - E ) shows the net assets of agricultural parastatals. These figures reflect access to capital, whether from retained earnings, from new equity investment by government, or from borrowing. Net assets (total assets less current liabilities) are shown instead of total assets, because it is more robust to coincidences in timing of financial transactions.1 Several conclusions can be drawn from the figure. First, from independence to 1976, only one parastatal, the Cotton Lint and Seed Marketing Board, experienced difficulty maintaining its net assets. The rest, though they suffered occasional bad years, or even several years of stagnation, were able to maintain or increase their capital bases. From 1976 four firms, including the Cotton Lint and Seed Marketing Board, the KCC, Uplands, and the Wheat Board, began suffering serious erosion of their capital bases. In 1977 they were followed by the KMC. In 1978 the Maize and Produce Board started the slide. In 1980 Nzoia and South Nyanza Sugar companies began serious and prolonged erosion of their capital bases. In 1981 the newly formed National Cereals and Produce Board began a steep decline in its already negative net assets. For those firms that suffered significant declines in capital base, recovery has been difficult and delayed. The decline has been so severe Figure 3-1A Net Assets: Sugar Companies

Chemelil EASI Mumias Nzoia SONY

-200

.400 —I

I

I

I

I

I

I

I

1 —

72

74

76

78

80

82

84

86

88

The Agricultural Sector Figure 3-1B Net Assets: Cereals Marketing Firms 3000

MMB M PB NCPB

2000

Wheat

e>

sz

CO

v> c o

1000

-1000

Figure 3-1C

63 65 67 69 71 73 75 77 79 81 83 85 87

Net Assets: KCC, K T D A

31

32

Public Enterprise in Kenya

Figure 3-1D Net Assets: CLSMB, Pyboard 300

200

-100

-

63 65 67 69 71 73 75 77 79 81 83 85 87

-20

.40 I I

l

l

l

l

l

l

l

l

l

l

l

l

63 65 67 69 71 73 75 77 79 81 83 85 87

The Agricultural Sector

33

that by 1980 one firm, the Wheat Board, was reporting negative net assets (that is, its current liabilities exceeded its total assets). By 1983 four firms had negative net assets, including the CLSMB, the NCPB, the KMC, and Uplands. In 1987 the South Nyanza Sugar Company reached the same state. The K C C was able to reverse the trend, a recovery attributable to the large increase in demand for its product when the school milk program was introduced. Only in the late eighties did the CLSMB, the NCPB, and S O N Y receive sufficient infusions of capital as to show positive net assets. The K M C and Uplands closed during the mid-eighties. In 1989 the K M C reopened, though its financial health remains problematic. Many parastatals are highly leveraged, as can be seen from Table 3-4, which gives the ratio of net worth to net assets. In 1983, seven out of 13 firms had virtually zero or negative net worth. Four had net worth so negative as to constitute negative net assets. Since the seventies, liquidity has been a chronic problem for many firms. Liquidity is measured using the current ratio (the ratio of current assets to current liabilities), as reported in Table 3-5. The convention on current ratios used to be that a ratio between two and three would give sufficient liquidity without excessive capital costs. Recently, as interest rates have risen and short-term capital markets have grown more sophisticated, the generally accepted guideline has fallen as low as 1.4. A current ratio below 1.0 reflects a firm with liquidity problems, one bound to make late payments to its suppliers.2 Firms that are illiquid are likely to become inefficient and exhibit high costs, because they are unable to buy in a timely manner and on the best terms. Plants may stand idle because of missing items. Table 3-5 shows that during the first decade of independence, only two firms suffered chronic liquidity problems: Chemelil and East African Sugar Industries. These two firms later recovered their liquidity. The K C C hovered at the edge of liquidity problems during the first decade of independence. During the second decade of independence, illiquidity was a more general problem. The KCC, KMC, Mumias, NCPB, Nzoia, SONY, Uplands, and the Wheat Board all suffered severe and long-lasting problems. The fact that six new firms (the five sugar firms and the N C P B ) suffered illiquidity suggests that government didn't understand the importance of liquidity. The sugar firms were starved of liquidity at the same time they were paying heavy excise taxes. The low levels of liquidity shown in this table must be a direct cause of the late payments to farmers that are so often reported. Such late payments clearly sap farmers' incentives to produce. This credit from farmers to the parastatals is probably large relative to the level of seasonal credit from parastatals to farmers.

34

Public Enterprise in Kenya

Table 3-1 Ratio of Net Worth to Net Assets 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 -1.07 -1.01 -.96 -.37 Chemelil CBK 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 CLSMB 1.00 1.00 1.00 1.00 1.00 .72 .78 .75 .80 .86 .87 .88 .70 .92 .62 .62 .56 EASI KCC .77 .80 .82 .77 .80 .86 .77 .82 .85 .87 .71 .69 .58 KMC .31 .32 .34 .34 .23 .28 .33 .40 .46 .55 .61 .51 .64 KTDA -14.09 -4.10 -8.90 -3.72 -3.43 -2.21 -2.31 -1.69 -1.47 -1.04 -.56 -.27 -.08 MMB .90 .85 .89 .91 MPB 1.00 1.00 .90 .90 .62 .25 .12 .53 .30 .12 .48 .52 .63 Mumias NCPB Nzoia Pyboard .52 .47 .44 .46 .48 .50 .46 .49 .55 .64 .65 .69 .73 SONY .67 .68 .69 .70 .73 .75 .76 .77 .79 .80 .83 .90 .75 Uplands Wheat 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 Chemelil CBK CLSMB EASI KCC KMC KTDA MMB MPB Mumias NCPB Nzoia Pyboard SONY Uplands Wheat

.51 .58 .63 .65 .73 .72 .70 .72 .84 .80 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 1.00 .83 .78 .40 -1.57 t t t t t t .49 .46 .49 .55 .78 .76 .66 .58 .56 .65 .59 .07 -1.22 -.78 -.05 .14 .20 .25 .26 .29 .32 .70 .54 .16 -.72 -1.95-10.06 t .06 .27 .31 .29 .22 .16 .04 .01 .17 .37 .28 .13 .49

-.04 .56

-.05 .36

.14 .76 1.00 .75 .90 1.00 .94

.20 .25 .52 .92 .84

.76

-.46 -1.16 .34 .38 .40 .42 t t t .09 .08 -.10 -.41 .75 .78 .11 .16 .45 .30 .15 .00 .86 -1.36 -5.06 t -1.45 t

.81

.82

.57 t .48 .22 .33 .31

.33

.48 .65 .86 .91 -.68 -.39 -.30 t t 1.07-20.39 -1.12 .84 .84 .84 .84 .89 -.78 -1.72 -3.35-18.60 t t Î t

.95 .90 .89

Source: Annual reports, various years. Notes: t Indicates negative equity and negative net assets.

Rates of Return to Capital The poor popular image of parastatals stems largely from the large losses they incur. However, profit is not a suitable measure of the return to capital. In a public enterprise, government chooses how the firm is capitalized, i.e., debt versus equity, and how its returns are realized, i.e., through profits or interest payments. These differences are arbitrary, and

The Agricultural Sector

35

Table 3-5 Ratio of Current Assets to Current Liabilities 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 Chemelil CBK CLSMB EASI KCC KMC KTDA MMB MPB Muraias NCPB Nzoia Pyboard SONY Uplands Wheat

1.07 1.00 1.02 1.05 30.95 18.92 5.36 1.38

.46 .74 1.02 1.09 1.07 1.06 1.07 2.00 1.51 4.63 5.70 3.94 .24 .34 .93 .94 1.12 1.05 .96 3.00 2.93 2.33 1.94 1.39 1.40 1.56 1.18 1.35 1.49

.75 1.17 1.40 .32 1.32 .75 1.11

1.25 1.08 4.96 .38 1.45 .95 1.46

1.02 .80

1.01 .71

1.30 1.28 1.26 1.93 2.24 2.16 1.45 .64 .72 1.75 2.28 1.19

1.22 1.30 .75

1.14 2.69 1.61

1.28

.95

1.34 1.67 2.58 3.79 13.76 7.42 13.17 2.26 1.86 1.65 1.10

.99 1.09 1.59 1.50

1.15

1.37

1.52 1.76 1.92 2.22 1.85 1.15 1.29

1.31 2.08 1.01 1.34

2.27 2.32 2.51 2.52 2.26 2.41 2.29 2.33 2.21 1.49 1.57 1.22 1.55 ... 74.26 144.94 221.79 108.74 1.26 6.29 2.59 1.41 .66 .93 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988

Chemelil CBK CLSMB EASI KCC KMC KTDA MMB MPB Mumias NCPB Nzoia Pyboard SONY Uplands Wheat

3.11 3.79 3.02 1.03 1.10 1.04 1.15 1.02 1.26 1.25 1.04 .81 .76 .75 3.34 1.50 .60 1.19 1.39 1.43

4.19 1.09 .83 1.19 .76 .60 1.38

3.69 1.04 .62 .94 .93 .63 1.47

2.88 1.02 .50 .76 .98 .52 1.47

3.22 2.80 1.69 1.30 1.22 1.00 1.00 1.00 1.00 1.00 1.00 .32 .32 .32 .61 1.63 .80 .75 .44 .45 .26 1.28 .170 1.32 1.28 1.32 .59 .66 1.27 1.06 1.19 1.23 1.29 1.32

6.67 8.52 7.54 3.88 .97 .89 .58 .50

2.92 .60 .39 2.18 1.70 .94 .63 .57

.60 .51 1.47 1.46 .83 .61

.53 .68 .85 1.31 .63 .29

.67 .82 .93 .65 .79 .67 .63 1.31 1.54 1.71 .52 .32 .52 1.33 1.53 2.34 2.39 3.85 4.11 .54 .49 .47 .43 .28 .19

2.51 1.12 2.53 1.31 1.55 1.14 1.35

4.28 .79 .93 1.43 1.47 1.90 1.81 .02 19.83 2.28 1.24 1.06 1.00 .69 1.46 .59 .47 .51

Source: Annual reports, various years.

to consider only the return to equity capital gives a distorted picture. Therefore, returns to capital is defined to include pretax profits plus interest payments, and it is compared with total assets, rather than equity.3 The use of profit alone gives a different impression of firm performance than one gets from examining total financial returns. The top half of Table 3-6 gives pretax profit. It is from this set of figures that the popular impression of parastatals as sinkholes for public money is derived. Caution must be used in interpreting the totals on this table because of missing

36

Public Enterprise in Kenya

Table 3-6 Trends in the Financial Performance of Agricultural Parastatals (thousands of Shs) Profits Only

Period 1964-66 1967-69 1970-72 1973-75 1976-78 1979-81 1982-84 1985-87

Profits of Profitable Enterprises

Losses of Unprofitable Enterprises

16,909.5 28,832.3 82,8621.1 178,069.0 391,128.9 182,726.9 480,981.8 674,104.3

46,653.4 22,163.3 24,065.6 110,360.8 374,090.8 1,300,354.3 2,178,656.6 4,152,291.2

Net -29,743.9 6,669.0 58,796.7 67,708.4 17,038.1 -1,147,627.1 -1,697,674.9 -3,478,186.9

Total Returns to Capital

Period 1964-66 1967-69 1970-72 1973-75 1976-78 1979-81 1982-84 1985-87

Returns to Firms with Positive Total Returns

Returns to Firms with Negative Total Returns

19,872.5 42,901.1 94,477.2 216,891.5 493,884.2 367,983.8 727,388.6 1,060,397.3

39,299.4 16,576.3 14,261.2 69,043.0 241,416.3 929,416.3 1,009,049.3 2,462,678.0

Net -19,426.9 26324.8 80,216.1 147,848.4 252,468.2 -561,432.5 -281,660.7 -1,402,280.7

data. The totals for the top and bottom halves of the table can be compared, but the time trends are imperfect measures for the entire sector, since some firms' data were not available in years before 1972 and after 1985. The bottom half of the table shows profits before finance charges, i.e., the total return to all capital, both equity and debt.4 Several facts emerge from Table 3-6. First, the popular perception that parastatals cannot make money is clearly inaccurate. The positive total returns of the healthy firms grew fairly consistently, with only one downturn in the 1979-1981 period, followed by a healthy recovery. The overall negative picture in recent years comes from the explosive growth of the size of the negative returns earned by those firms in trouble. These negative returns were dominated by a few firms, including the maize trading firms—the Maize and Produce Board up to 1980 and the National Cereals and Produce Board since then. Other major loss earners were the Kenya Meat Commission, South Nyanza Sugar Company and Nzoia Sugar Company. Second, Table 3-6 reveals the importance of high leveraging and lack

The Agricultural Sector

37

of adequate capital base in the negative financial performance. The lower half of the table shows that from the period 1976-1978 to 1979-1981, when the price of Kenya's exports collapsed, oil prices jumped and the world slid into economic recession, the total negative returns to capital increased by about 285 percent. That increase was followed by a smaller increase in 1982-1984, of 9 percent. Meanwhile, the total returns of the healthy enterprises more than doubled in the later period. On net, from 1979-1981 to 1982-1984, there was considerable recovery, with net losses for the sector falling by 50 percent. The shocking deterioration in 1985-1987 is largely due to the NCPB, whose losses accounted for 47 percent of the total losses. The top half of Table 3-6, which shows only profits rather than total returns to capital, reveals a rather different picture. The profit-making firms show an even stronger recovery on this measure. However, the unprofitable enterprises show an accelerating slide toward disaster. Whereas using total returns showed losses growing only 9 percent during the 1982-1984 period, using profit only, these firms' losses grew 64 percent. The net shows further deterioration, rather than recovery. Again, the debacle of 1985-1987 is largely a matter of the NCPB, which again accounted for 47 percent of the losses (the raw data on profits and returns are included in Appendix C). Two conclusions emerge from these data. First, use of profits only, rather than total returns to capital, provides a picture that is distorted and alarmist. Substantial recovery occurred by the mid-1980s. Second, the financial structures of the loss-making enterprises make it virtually impossible for them to recover financial health. During the very difficult years 1979-1981 they financed their losses by borrowing, and even their partial recovery has not been adequate to permit retirement of that debt, so that their debts and interest payments are continuing to mushroom in absolute terms and as a portion of their profits. Of the negative net profit for the whole sector in 1982-1984,83 percent went to interest payments, most of which were necessary to finance past losses. Had these firms not been so highly leveraged, they would have been more able to recover. Figure 3-2 shows average rates of return for each firm by decade. Detailed data on rates of return by firm and by year are found in Appendix C. All firms that lost money on average in the 1960s improved their performance in the 1970s, including the CLSMB, the KTDA, and the Maize Marketing Board, which was reorganized into the Maize and Produce Board. Three firms started after 1978 (the NCPB, SONY, and Nzoia) incurred large losses. The other large losers were the livestock trading firms, the KMC, and Uplands. Several firms showed modest profitability throughout the period, including Chemelil, the CBK, EASI, the KCC, Mumias, the Wheat Board, and the Pyrethrum Marketing Board.

Public Enterprise in Kenya

1 1 1 1 III

y

• •il 11

M

M

II I

£ to o

o

§ o

3 lu

o O O 2 * SÉ

_ » S S

-2 f E iL 3 S 2

6%) Marginal (0 to 6%) Poor (< 0%)

1963-69

1970-79

1980-88

Overall

2 5 3

4 4 5

3 3 7

3 4 9

Table 3-7 categorizes financial performance as healthy if the average rate of return exceeds 6.0 percent, marginal between zero and 6.0 percent, and poor if negative. The choice of 6.0 percent to define healthy performance was arbitrary, but the firms so categorized were indeed healthy. They were liquid and did not suffer erosion of equity or net assets. Performance improved a bit from the 1960s to the 1970s but worsened in the 1980s. Furthermore, it became more polarized in the 1980s, with fewer firms in the marginal category. Even in the 1980s, when performance was

Table 3-8 Classification of Firms by Financial Performance Rate of retuma(percentage) Adequate Performers Chemelil CBK KCC KTDA Mumias Pyboard Wheat Inadequate Performers CLSMB EASI KMC MMB MPB NCPB Nzoia SONY Uplands

5.2 0.4 2.7 4.0 14.2 10.3 6.2 -6.4 -2.2 -9.3 -19.4 -11.0 -13.5 -11.4 -11.0 -7.6

Note: a. Figures show average rate of return on assets (earnings before interest and taxes over total assets) for all years from 1963 to 1988 for which data were available, weighted by assets.

40

Public Enterprise in Kenya

at its worst, about one-quarter of the firms earned healthy rates of return. To summarize, over the entire period, seven firms performed adequately and nine performed inadequately, as shown in Table 3-8. The criterion of zero average profit is used to define adequate performance. This seemingly lenient definition was used because the charters of most of the agricultural enterprises state that their financial target is to break even, on average, over the years. In firms with flexible pricing, this target has been achieved by adjusting upward the price paid to producers whenever the firm had accumulated too much profit. EFFICIENCY—PROCESSING MARGINS

One important measure of performance for a firm is the efficiency with which it transforms inputs into outputs, the cost of processing a unit of output. Real processing margins were calculated.5 The GDP deflator was used. The price paid to farmers is excluded. So are interest payments, since these were included in returns to capital. The processing margin includes the costs of transport, processing, marketing, storage, depreciation, administration, etc. Time trends were fit on real processing margins. The firms were classified as good performers if their margins didn't rise significantly, and as poor performers if they did. The trends, together with the classification of performance, are found in Table 3-9. The processing margins for each firm, for each year, are shown in Figure 3-3 (A-E) and in Appendix A. The processing margin at the beginning of the period was used as a standard for measuring efficiency because the pressures against efficient public enterprise operation were weak before independence. Public firms in the colonial period were set up to serve the interests of a well-educated, politically organized group. Public enterprise employees were not an interest group with a claim on protection, as they became later. The boards were mostly appointed from among those who stood to benefit from efficient operations, and there was no managerial class that received jobs in state firms as political sinecures. Several firms began operations after independence, in some cases well after independence, so the use of data from their first years of operation would be problematic. Fortunately, in each case there is a better standard of comparison. The Maize and Produce Board can be compared with its predecessor, the Maize Marketing Board. The National Cereals and Produce Board can be compared with its predecessor, the Maize and Produce Board. The five sugar firms can be compared with each other. An alternate standard would be to compare public firm processing margins with those of private firms in the same line of business. Unfortunately, there are no data publicly available from any private sector

The Agricultural Sector

41

Table 3-9 Trends in Processing Margin and Classification of Performance

Firm

Years

Annual % Change

T-Stat

Chemelil CBK CLSMB EASI KCC KMC

72-88 63-87 63-87 75-88 63-84 63-71 71-83 63-88 63-66 67-77 77-80 73-88 80-87 79-85 63-87 80-88 63-66 66-84 73-80

0.1 0.4 3.1 1.2 6.2 -0.5 7.0 -4.6 9.3 -11.4 53.6 -0.0 4.5 -0.5 0.8 1.8 -1.9 2.7 -0.9

0.3 1.4 6.0 2.4 10.3 -1.9 6.1 -8.9 1.2 -2.0 11.5 -0.2 4.3 -0.3 1.7 1.0 -1.3 6.6 -0.1

KTDA MMB MPB Mumias NCPB Nzoia Pyboard SONY Uplands Wheat

Performance Good Good Poor Poor Poor Good Poor Good Poor Good Poor Good Poor Poor 3 Good Poor b Good Poor Good

Notes: a. Classified as poor despite downward trend, due to high absolute level of margins relative to other firms in same industry. b. Classified as poor despite low t-statistic, due to high absolute level of margins relative to other firms in same industry. Figure 3-3A Real Unit Costs, Excluding Interest: Sugar Companies

42

Public Enterprise in Kenya

Figure 3-3B Real Unit Costs, Excluding Interest: Cereals Marketing Firms 350

MMB MPB

-

300

NCPB Wheat

250 c3 CO Q

+ A3>

Q2/ASSETS

This specification of total costs yields an average cost equation of the form AC = Bo+Bi • l/Q + B2 - ASSETS/Q

+ B3 •

QlASSETS

This average cost function gives the classic envelope of U-shaped shortrun cost functions, as shown in Figure 3-4. It was estimated separately for each commodity. The dependent variable was real processing margins, reported above. Real assets were calculated using a deflator for capital goods and the declining balance method of calculating depreciation. The results are shown in Table 3-12. In this equation the regression coefficients should all be positive. Good results were achieved for all firms whose operating margins either

Table 3-12 Cost Function Estimates: Dependent Variable = Real Processing Margins (1976 Sh^ton or per head) Firm CLSMB

Constant l/Q

3,369.9 19,957.9 (3,402.3) (9,005.3) -213.4 .174x10® KCC (489.9) (.393xl08) .222xl08 KMC (,081x10s) Pyboard 1,135.3 ,148x10s (430.8) (.036x10®) Sugarb -.286xl07 (,371xl07) Uplands 647,681.2 (2,313,721)

Assets/Qa

Q/Assets"

-,320xl0"3 -,250xl010 (,255xl0"3) (.167xl010) .998 18,495.4 (.332) (180,061.2) .225 61,501.8 (.080) (13,016.5) .048 (.025) 1,328,199.6 .135 (.012) (300,200.9) .350 135,914.9 (.105) (37,254.1)

D.W.

1.67

RHO .934 (.095) .381 (.277)

2.55

R2 .77 .89 .90 .60

.286 (.101) .758 (.175)

.81 .74

Notes: a.Assets are measured in 1976 Shs. Standard errors are shown in parentheses. Ordinary least squares are used for the Kenya Meat Commission and the Pyrethrum Board. Cochrane-Orcutt iterative technique was used for the Cotton Lint and Seed Marketing Board, the Kenya Cooperative Creameries, and Upland Bacon Factory, in order to correct for the presence offirst-orderserial correlation. b. Pooled data from all five sugar companies. The pooled sample was heteroskedastic and had mutual correlation of errors across firms, in addition to first-order serial autocorrelation. The regression was estimated according to the procedure reported in Kmenta, pages 512-514. The estimates of the coefficients and standard errors are consistent and asymptotically efficient. In addition to the coefficients reported above, a dummy was included that allowed fixed costs for firms started after 1978 to differ from those of older firms (Bi was allowed to differ). This coefficient was estimated to be 3.473xl07, with a standard error of ,248xl07.

The Agricultural Sector

49

Figure 3-4 U-Shaped Average Cost Functions

grew or fluctuated significantly. All but four coefficients had the right sign, and those four were not significantly different from zero. Capacity utilization explained between 60 percent and 90 percent of the observed fluctuation in processing margins. Thus, capacity utilization is a major cause of public enterprise operating problems. Two primary factors contribute to capacity utilization problems, which lead in turn to high processing margins and poor financial performance. The reasons differ by firm and may reinforce each other. First, a firm may not be allowed to vary prices to help smooth natural cycles in supply. Firms are often not allowed to offer premiums to attract produce during production troughs, or to drop their prices during supply peaks, so they are completely at the mercy of weather-induced agricultural cycles. This has been a problem for the KCC. After the abolition of wet/dry season price differentials for milk, seasonal fluctuations in supply increased markedly. The ratio of maximum monthly production to minimum monthly production was 1.29 in 1969/70, but rose to 2.02 in 1973/74.8 When a public enterprise competes with private firms that have the advantage of more flexible pricing, the parastatal will have throughput cycles greater in magnitude than the underlying production cycles. Suppliers will take their produce elsewhere during times of scarcity when premiums are being offered by private firms. During times of abundant supply, when the price offered by the public firm is supracompetitive, they will switch to the public firm. This has been a problem for the KMC.

50

Public Enterprise in Kenya

Second, a firm may for some reason have liquidity problems. The firm must respond by failing to purchase produce offered to it, or by lowering real prices to farmers. This can be done either explicitly or through delayed payments or other much-publicized "malpractices." Farmers respond by offering less produce in the future. If the problem is not quickly reversed, the firm will be chronically undersupplied. This happened with Nzoia, SONY, and Uplands. Many factors could cause the original liquidity problem: a bad year, a delay in payments from a large customer, a delay in receipt of government investment funds during a time when the investment expenditure has already been incurred, a failure of the government to capitalize the firm properly at the outset, or an unexpectedly large crop. A temporary liquidity problem can turn into a permanent problem that will cause high processing margins. In conclusion, price stabilization policies have contributed to public firm inefficiency. In several firms they have caused supply problems, which have then been translated into liquidity problems, which in turn became permanent operating problems. In contrast, those firms that operate with prices left free to fluctuate have all shown consistent adequate financial performance and have been consistently efficient. Implicit in the use of a flexible pricing regime is the political decision that producers will bear the full risks of the market and will not be subsidized either by consumers or by shareholders. Under a flexible pricing regime, public enterprise cannot be used to redistribute wealth on a regional or other basis. Thus, there is a political cost in switching to a flexible pricing regime, in that it means the end of cross-subsidization.

POLICIES THAT CONTRIBUTE TO PARASTATAL PROBLEMS

The government has pursued four policies toward certain agricultural parastatals that have contributed to their failure. Each of these policies could be construed as an effort to keep pressure on the firms' managers to perform, i.e., to reduce managerial discretion. The examples of the successful agricultural parastatals show that there are more productive ways of applying such pressure. The first policy that has led to agricultural parastatal failure is price control. A firm that can set supplier prices as a residual after subtracting costs from revenues is a less risky firm by definition, one that will not run large deficits or become illiquid. It might seem that such a policy invites abuse by managers. However, in all of the firms in Kenya that use this flexible pricing procedure, none has shown upward trends in real unit processing margins. Short of a residual price determination policy, a firm that can buy at going market rates is less risky than one that must buy at

The Agricultural Sector

51

prices set before the crop year. This strongly suggests that a reexamination of domestic price stabilization is in order. These policies seem to stem from several different motives. For example, the domestic price of sugar has been stabilized well above world levels for a number of years in order to support the incomeearning ability of those sugar farmers who were drawn into production when world prices were high. The success of Mumias and Chemelil demonstrates that the policy can succeed. Price stabilization in cereals is supposed to protect the interests of urban consumers. The main source of price variation is fluctuation in supply due to changes in weather. If price were allowed to vary along the demand curve, then revenue, the product of price times quantity, would be more stable than if price stayed constant while quantity varied. Apparently government is willing to destabilize farmers' incomes in order to stabilize urban consumer prices. This policy has proven expensive, and there are several alternative policies that might achieve stabilization at lower cost to the economy. 9 Price stabilization in livestock is designed to support producer incomes. Yet, the policy whereby this was attempted was nonviable. From 1971, K M C competed with a vigorous private sector, so during nondrought years it could not amass excess profits with which to support livestock prices during droughts. The result was a predictable cycle of collapse. If government truly desires to support producer incomes during droughts, it must do so explicitly through budget appropriations. There is no magic by which a firm in a competitive industry can do so. The second policy that has increased the riskiness of agricultural parastatals is the use of highly leveraged financial structures. Leveraging increases risk because where capital comes from loans rather than equity, the firm must pay a return on capital whether it has earned one or not. If a large loss is incurred, this can cripple a firm beyond the point of recovery. Where a firm has flexible pricing it can survive high leveraging, but where the two policies are combined the result may be disaster. The third policy that has helped cripple several agricultural parastatals is that of allowing them inadequate working capital. In some cases this has been done fairly consciously. For example, the working capital designated in the investment plans for Nzoia and South Nyanza Sugar companies was never paid in; meanwhile, liquidity was withdrawn in the form of excise taxes. The result was a failure to provide short-term finance to sufficient outgrowers, resulting in low capacity utilization and high unit costs. In other cases the policy was an unintended by-product, as when government failed to finance KMC's price support activities during drought years. Even a well-managed firm like the Pyrethrum Board can suffer liquidity problems if it has to buy an unusually large crop unexpec-

52

Public Enterprise in Kenya

tedly, as it did in the early eighties. Finally, some agricultural parastatals have been hurt by competition from the private sector. This is most notably true for KMC and Uplands. The policy has undoubtedly benefited beef and pork producers and perhaps consumers as well. But where, as with KMC, the firm is supposed to support a price well above the going market price at some times, it is quixotic to hope it can do so without direct subsidy. CONCLUSIONS

The current fashion in some circles of speaking in blanket terms of how poorly public enterprises perform is ill founded. In Kenya's agricultural sector about half the firms have performed well, some for two decades since independence. All have served large numbers of smallholder farmers, a role that requires a fairly large and complex organization, which makes the success achieved all the more impressive. Several of the firms have weathered unfavorable market conditions, both domestic and international, but unfavorable external conditions have not resulted in institutional decay. It has been shown that where problems have occurred with inefficient management, they have resulted in financial problems. A major cause of increases in processing margins is low capacity utilization. This occurred in firms that function with inflexible pricing policies and firms that are highly leveraged and/or illiquid, in firms that must carry out a distributive policy of government for which they receive no funding, and in firms that face competition from the private sector. Firms can survive one of these problems, but where two or three coincide, increasing losses and increasing inefficiency are bound to result. NOTES 1. For example, a firm may have large current assets and large current liabilities in one month, because it has received either the crop or the revenue from sale of its crop but hasn't yet paid farmers. In the next month it does so, reducing both the current assets and the current liabilities. This type of transaction doesn't reflect a change in the capital base of the firm, hence the use of net assets. The definition may seem unsatisfactory when the firms misuse the accounting categories. For example, in the 1980s some firms ran up large losses, which they are financing through "short-term" borrowing. In several cases these current liabilities amount to more than the total assets of the firms; thus, the definition used shows negative capital invested in the firm, an anomalous concept. However, it was judged that the problem is not in the definition of capital invested, but rather in the classification of liabilities. In the cases cited, there is little doubt that the accumulated losses will never be repaid, certainly not within one year, as their classification as current liabilities implies. In any case, the definition shows that disinvestment has occurred in these firms, which is an accurate perception. 2. These guidelines, taken from industrialized countries, assume that inven-

The Agricultural Sector

53

tories are accurately valued. To the extent that inventories are obsolete or overvalued the current ratio needs to be higher to give adequate liquidity. Obsolete inventories were probably common among the poorest-performing parastatals. 3. We revert to total assets as a measure of capital invested, rather than net assets, because of the following anomaly: some firms that incurred losses had negative net assets. This would yield a positive (in many cases large positive) rate of return. 4. Data for each firm for each year, from which Table 3-6 was compiled, are found in Appendix C. 5. There were problems defining units for some firms. For example, for the NCPB, it was necessary to aggregate tons of all types of grains and produce. To the extent that handling costs vary between the different products handled and the composition of the aggregate product changed, the results may be misleading. Similarly, the KMC and Uplands measure units handled in terms of head. It would probably be more meaningful had they given kilos of meat produced. While these measurement problems are real, it was judged that they were not so severe as to negate the usefulness of examining the degree to which unit costs have been controlled over the years since independence. 6. This margin includes only the costs of the central KTD A administrative apparatus and excludes factory processing costs. Factory costs were fairly constant from year to year, and were comparable to those of private tea factories. See Leonard, pages 139-140. 7. See Ruigu. 8. See Ruigu. 9. See Pinckney.

The Financial Sector This chapter evaluates the performance of parastatals in the financial sector. It is organized in two parts. The first half covers depository institutions and the last half covers a nondepository financial firm, the Agricultural Finance Corporation. Depository institutions are firms that accept deposits from the public, which they then reloan to borrowers. They include both commercial banks (CBs) and nonbank financial institutions (NBFIs). The distinction will be important in assessing the performance of the parastatal firms. The period under study is 1971-1987. Performance is shown to differ between the early part of the period and the later part, with the break occurring around 1979. The first section will provide (1) background information on the growth patterns of the financial sector, (2) an account of the entry of the Kenyan government into direct participation in the sector, and (3) an analysis of the performance of the parastatal firms relative to the performance of the financial sector as a whole. I will show that the depository financial institutions have functioned effectively, competing successfully with private firms and earning consistent high profits. During the 1970s the depository institutions pioneered in the development of general purpose NBFIs, which have contributed importantly to the development of financial intermediation in Kenya. During recent years they have been used for more distributional goals, spreading financial intermediation to rural areas, with no apparent drop in efficiency. The second section will provide stark contrast. It will analyze the activities of the Agricultural Finance Corporation (AFC) from 1963 to 1987. It will show that this firm has been highly politicized, is commercially nonviable, and is unprofitable and inefficient, functioning mainly to transfer funds to its elite group of clients. The A F C has largely failed in its role of mobilizing credit to the agricultural sector, a role that has been increasingly filled by the depository firms. Performance is analyzed in terms of the four groups who are potential beneficiaries of the parastatals: depositors, borrowers, stockholders, and employees. Data relating to each will be presented. 55

56

Public Enterprise in Kenya DEPOSITORY FINANCIAL INSTITUTIONS

Growth and Development of the Financial Sector Depository institutions are regulated under the Central Bank of Kenya Act, 1966 and the Banking Acts of 1968,1986, and 1989. Regulations differ between CBs and NBFIs. However, within those categories the regulations apply uniformly to institutions that are owned by government, by foreign banks, or by private citizens. CBs accept deposits in the form of checking accounts, savings accounts, and time deposits. They are not permitted to pay interest on checking deposits. Minimum interest rates are set by the Central Bank for the latter types of deposits. CBs make loans, subject to maximum interest rates set by the Central Bank. Other charges, such as application fees and commitment fees, are also limited by the Central Bank. NBFIs also accept deposits and make loans. They are not permitted to offer checking accounts. Minimum interest rates paid on savings accounts and time deposits are also set by the Central Bank, at rates higher than those for CBs. Maximum interest rates charged on loans are also set by the Central Bank, also at rates higher than those allowed for CBs. Other charges, such as application fees and commitment fees, are not regulated and can be quite significant. Paulson reported that, while the maximum interest rate on NBFI loans in mid-1983 was 16 percent, the effective rate reported on some NBFI loans exceeded 30 percent.1 CBs have been limited in the amount of lending for consumer durables and real estate they are allowed to do.2 The differential interest rate regulations have favored the growth of NBFIs over CBs, due to the fact that NBFIs can charge higher interest rates. Since the interest rate ceiling has been binding throughout the period, this makes them more profitable. Because they can charge more for loans, they can afford to pay more to attract deposits. Thus, depositors are attracted to put their money in NBFIs rather than CBs. CBs and NBFIs tend to operate as pairs. This can and has happened from either direction: some banks have found it profitable to start NBFI subsidiaries, while some NBFIs have found it profitable to start CB subsidiaries. The latter was ruled out by the 1986 revision of the Banking Act. The incentive for banks to create NBFIs is clear: they can steer loan applications to the NBFI and make more money on the loan. The incentive for NBFIs to start banks derives from the structure of the reserve requirement. NBFIs do not have accounts with the Central Bank; instead they may count as liquid reserves current account deposits held with commercial banks, cash and Treasury Bills. The reserve requirement for NBFIs ranged from 15 to 24 percent from 1974 to 1983. Since NBFIs receive no interest on the current account deposits they are forced to keep with

The Financial Sector

57

banks, they profit by placing them in banks they own, which can make loans against them. The growth in deposits in both CBs and NBFIs is shown in Figure 4-1. The growth was much faster prior to the coffee boom than after it. The fast-growth period will be distinguished as 1971-1979 and the slow-growth period as 1979-1987. In 1971 deposits in NBFIs were only 14 percent of deposits in CBs. CB deposits grew rapidly, 20 percent p.a., compared with inflation of 10.5 percent. Deposits in NBFIs grew even more rapidly, 24 percent p.a. As shown in Figure 4-1, those growth rates accelerated during the seventies. After the coffee boom, deposit growth slowed. Overall deposit growth slowed from 21.7 percent p.a. in 1971-1979 to 14.4 percent p.a. in 19791987. But this fall in deposit growth was very unequally distributed. CB deposit growth dropped most dramatically, from 20.0 percent p.a. in 1971-1979 to 11.9 percent p.a. in 1979-1987. CB deposit growth only slightly exceeded inflation, which averaged 9.1 percent. Meanwhile, NBFI deposit growth dropped less, from 30.6 percent p.a. in 1971-1979 to 21.5 percent p.a. in 1979-1987. Since NBFI deposits grew faster than CB deposits throughout the period, they naturally increased their market share. Whereas in 1971 NBFI deposits were only 14 percent of CB deposits, by 1979 they were 27.2 percent, and by 1987 they were 52.6 percent as large as CB deposits. While NBFIs have been considered a significant phenomenon and have attracted a great deal of attention since the late 1970s, this is slightly Figure 4-1 Deposit Growth Rates (two-year moving average)

Source: Central Bank of Kenya, Quarterly Economic Review, various years.

58

Public Enterprise in Kenya

misleading. NBFIs were growing even faster during the seventies than they have since then. Their dramatic gain over CBs is due to a sharp decrease in growth of CB deposits since that time rather than to increasing growth of NBFI deposits. Interest rates have been set administratively at levels below market clearing rates and also, throughout the seventies, below inflation. In the face of excess demand for loans, the Central Bank issued guidelines to banks that were intended to influence the allocation of available funds. These guidelines encouraged lending to agriculture and discouraged lending for luxury consumption.3 Loans to foreigners are also restricted.4 Beginning with the budget of 1980, interest rates have been adjusted upward several times.5 Table 4-1 shows that positive real rates were achieved throughout the eighties. As interest rates have risen, this greater differential between real rates of return on interest-bearing and non-interest-bearing deposits has helped fuel the large shift of deposits from CBs to NBFIs. Unfortunately, no time series on deposit rates for NBFIs are available, but their rates have exceeded those paid by CBs.6 It is widely Table 4-1 Nominal and Real Interest Rates Interest Rate3 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988

4.0 4.0 4.0 4.0 4.0 4.0 5.6 5.6 5.6 5.6 5.6 5.6 6.4 12.3 13.8 13.0 12.0 12.0 12.0 12.0 13.0

Inflation Rateb 1.8 1.0 2.7 4.9 7.7 9.3 18.0 10.4 16.3 18.0 3.3 5.1 8.7 9.2 10.7 8.6 10.9 8.4 10.4 6.1 10.3

Real Interest Rate 2.2 3.0 1.3 -0.9 -3.7 -5.3 -12.4 -4.8 -10.6 -12.4 2.3 0.5 -2.3 3.1 3.1 4.4 1.1 3.6 1.6 5.9 2.7

Source: Author's calculations. Notes: a. Commercial bank deposit rate, 9-12-month time deposit, b. Inflation is measured by changes in GDP deflator.

The Financial

Sector

59

Figure 4-2 Currency/M2

Source: Central Bank of Kenya, Quarterly Economic Review, various years.

reported that NBFIs, in addition to paying higher interest rates, engage in non-interest rate competition for deposits by offering cash payments and preferential treatment to large depositors.7 High interest rates contribute positively to economic development by permitting greater intermediation. When firms and households are not constrained to self-finance, funds should flow to higher return investments, increasing the average quality of investment and raising the rate of economic growth.8 During the years since the establishment of the Central Bank and since government began direct participation in the financial sector, substantial development of the financial structure has been achieved. Figure 4-2 shows that the public has grown increasingly willing to hold money in the form of deposits, instead of as currency. Figure 4-3 shows that the ratio of money (broadly defined) to GDP grew substantially in the period 1968-1979. Figure 4-3 also shows that most of this growth was in time and savings deposits, included in M2, while the ratio of Ml (currency plus demand deposits in CBs) to GDP did not change substantially. Although the number of banks operating in Kenya has grown, the sector is still quite concentrated, especially outside of Nairobi and Mombasa. Table 4-2 shows the distribution of bank branches as of June 30,1983. Of 161 bank branches in Kenya, 123 belonged to one of the four large retail banks: Kenya Commercial Bank, Barclays, Standard, or National Bank of Kenya. Of the 87 bank branches located outside Nairobi and

60

Public Enterprise in Kenya

Figure 4-3 Ratios of Money to GDP

Source: Central Bank of Kenya, Quarterly Economic Review, various years.

Mombasa, 78 belonged to one of these four large banks. Of these four banks, the National Bank of Kenya is in a distant fourth place. It had only eight full branches, compared with Kenya Commercial Bank's 49, Barclays' 30, and Standard's 36. Thus, the retail banking sector in Kenya is highly concentrated, with only four significant actors.

Table 4-2 Distribution of Bank Branches (as of June 30,1983)

Bank

Nairobi Full Sub

Mombasa Full Sub

Full

Sub

Full

Sub

KCB Barclays Standard NBK Cooperative

10 11 10 4 1

5 4

4 3 2 1 1

2 1 1 2

35 16 24 3 4

29 24

49 30 36 8 6

36 29 1 3 0

Subtotal Other banks

36 17

10 3

11 10

6 9

82 5

53

129 32

69 12

Total

53

13

21

15

87

53

161

81

1

Source: Paulson, 1984, page 54.

Other

Total

The Financial Sector

61

Government Participation in the Financial Sector Participation by the Kenyan government in the financial sector dates from the late 1960s. The primary goal of the government in entering the sector was to make the sector more responsive to the borrowing needs of the Kenyan public. Foreign banks had lent very little to Kenyans of African origin, choosing instead to invest surplus funds abroad.9 The Kenyan government used a twofold approach to entering the commercial banking field: creation de novo and acquisition of existing financial institutions. In 1965 the government, together with the Commonwealth Development Corporation, created the Housing Finance Company of Kenya (HFCK). The HFCK accepts deposits from the public and reloans them in the form of mortgages. The HFCK was set up with a very small initial capital base. Its size was boosted considerably in 1970 by the acquisition of another NBFI, First Permanent (EA) Ltd. In 1968 the government created the National Bank of Kenya (NBK). The NBK operates an NBFI subsidiary, the Kenya National Capital Corporation (known as Kenyac), which was set up in 1976. In 1970 the government acquired 60 percent of the shares of the National and Grindlays Bank, which it renamed the Kenya Commercial Bank (KCB). National and Grindlays began operating in Kenya in 1896 and had an extensive branch network. Takeover of control of commercial banking was not intended "to bring about new thinking or practices in commercial banking," and government-owned banks were "to operate on commercial principles.10 Operating on such commercial principles, the newly nationalized KCB found it possible to expand lending to Kenyans of African origin by 225 percent between January and September 1971.11 The takeover, which was friendly, was completed in 1976 when the remaining shares were acquired from Grindlays. Grindlays signed a management contract and provided most of the high-level managers, with the proportions declining as Kenyans were trained to fill more and more positions. The management and training contracts were renewed several times but finally expired in 1986, at which time the last expatriates left. The Kenya Commercial Bank owns an NBFI, the Kenya Commercial Financial Corporation Ltd, which was set up in 1971. The financial parastatals have in many ways enjoyed a privileged status among parastatals. They have been headed by a group of managers generally seen as astute businesspeople and/or extremely well connected politically. This group has included John Michuki, Stanley Githunguri, Nicholas Nganga, Philip Ndegwa, George Saitoti, and Benjamin Kipkorir. Like the successful Kenyan public managers David Leonard studied, these men were active in shaping the policy framework in which they operated.12 These firms have enjoyed considerable operating autonomy, at least autonomy from the normal civil service bureaucratic controls, if not

62

Public Enterprise in Kenya

political autonomy. For example, although it is required that all parastatals submit development plans for approval and inclusion in the national forward budget, none of these firms has done so. Likewise, parastatals are required to submit investment projects for Treasury approval. Projects that cannot project financial profitability are generally unlikely to receive approval. The banks do not submit their plans for expansion of facilities for approval. Given that the Treasury is the parent ministry of the depository financial institutions, such autonomy is unusual. The KCB produces one of the least informative annual reports of any parastatal. KCB justifies its secrecy by arguing that fuller disclosure would handicap it in competing with the private Barclays and Standard banks. The banks are also exempted from the State Corporations Act of 1986. The depository financial parastatals are subject to at least three types of political pressure, which they find difficult to resist. First, they are pressured by politicians to provide loans to politicians and their friends. It is not clear to what extent such loans are made against the better judgment of the bankers. It is impressive that once the loans are made, the bankers seem to be able to pursue their recovery with full vigor. Occasionally even a minister is foreclosed on due to failure to pay. The bank staff have learned that so long as the government is advised before such stories hit the headlines, they will usually be allowed to proceed, or quiet pressure will be put on the politician to pay before the matter hits the headlines.13 Paulson found that only 16 percent of the accounts had arrears, and all those accounts with arrears over three months had been turned over to lawyers for seizure of security.14 The second type of political pressure under which the parastatal banks find themselves is potentially more serious. The parastatal banks are naturally under pressure to act as bankers to the rest of the parastatal sector. They have loans to many other parastatals, including some of the insolvent ones that cannot qualify for loans on commercial criteria. Such loans always carry government guarantees, but these guarantees are often not made good, or only after long delays. The banks have been given some leeway in protecting their interests in these cases. For example, Kenatco owed money to other (private) sources that seemed on the verge of moving against Kenatco. NBK was allowed to protect its interests by putting Kenatco under receivership. A few details of such bad loans came to light when Parliament discussed the guarantees. While the complete numbers for all the bad loans were not released, bad loans to other parastatals constitute a minimum of 10-15 percent of the total loan portfolios of KCB and NBK. 15 Clearly, if the practice of forcing the parastatals to carry bad debts of other parastatals became too widespread, it could ruin their effectiveness as banks.

The Financial Sector

63

A third type of political pressure has been concentrated on the KCB, which has been chosen to help decentralize resources out of the cities by bringing banking services to wananchi}6 Its declared goal is to open a full-service branch in every district, regardless of demand conditions. Such a program is fraught with dangers, but the major pitfalls have been avoided to date. For example, such a program is expensive and if pushed too far could cause insolvency. This has not occurred. Furthermore, it could be interpreted that since KCB must take banking to wananchi, then it must extend them loans on concessionary terms. This also has not happened. It appears that KCB is operating in a very businesslike way with proper controls and procedures. There is a cross-subsidy implied in running facilities in areas that don't have enough demand to support them, using profits from more densely populated areas.17 KCB has never published any figures on the size of the cross-subsidy, and indeed asserts that no such cross-subsidy is occurring. Instead it conceives of the program of branch expansion as a long-term investment that will pay off eventually. Managers implicitly argue that as a public institution the KCB should have a lower discount rate than private banks interested only in profits for shareholders. Such an attitude undoubtedly contributes to the KCB's ability to resist pressures to grant loans on concessionary terms. The KCB's explanations of its policy are logical, and its actions seem consistent with its rhetoric. Nonetheless, the same policy could be used to hide large-scale misappropriation of resources. It is puzzling why the KCB is so reluctant to discuss the magnitudes involved, or the time horizon when they expect to see the investments pay off. It seems that the autonomous parastatal nature of the KCB is being used to circumvent the budget process where such interregional transfers receive more scrutiny and debate. This policy contrasts with that used with the Kenya Power and Lighting Corporation (KPL), which also invests in building facilities in rural areas with insufficient demand to justify investment on commercial terms. That corporation clearly states the amounts involved and makes such investments with financing provided only through the district focus budget of the Ministry of Energy. In 1988 KCB floated 20 percent of its shares to private shareholders. It will be interesting to see if the new shareholders will demand that KCB switch to a policy more similar to KPL. Comparison of the performance of the public and private sector banks is hindered by several limitations on data availability. Standard Bank's operations in Kenya are carried out as part of their international operations and not incorporated separately. Thus, no figures are available on deposits, assets, profitability, etc., on operations in Kenya. The same was

64

Public Enterprise in Kenya

true for Barclays until 1979, when it incorporated a local subsidiary. However, the Central Bank's Quarterly Economic Review gives data for the commercial banking sector as a whole, so we can track market shares for the public firms. Performance of Financial Parastatals How well financial parastatals treat their depositors can be assessed by comparing the growth in deposits in parastatals with growth in total deposits in all financial institutions. Those institutions that increase their market share can be assumed to be offering depositors better returns. These returns may be monetary, i.e., interest paid, or nonmonetary, e.g., convenient branch locations, prompt and courteous service, etc. A great deal of information on deposits is available and is discussed at some length here. Another indirect indicator of returns to depositors is the number and Table 4-3 Consolidated Commercial Bank plus NBFI Deposits (millions of Shs)

Year 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988

KCB 1,108.2 1,189.5 1,568.8 1,768.3 1,849.8 2,251.9 3,297.0 3,796.7 4,033.3 4,217.7 4,391.4 4,889.6 5,333.6 6,138.1 6,775.2 8,947.2 8,516.4 10,019.7

NBKa 211.8 337.5 528.0 747.1 950.3 1,173.6 1,461.3 1,901.2 2,254.7 2,323.2 2,316.4 2,304.0 2,446.6 2,634.5 2,928.4 3,406.2 3,748.2

Summary of Growth Episodes (growth in % p.a.) 71-89 18.8 34.4 9.8 79-87 6.6

HFCK

Total

Public Share (%)

83.3 109.7 128.2 144.3 187.2 180.9 213.6 246.8 375.7 472.6 688.0 886.5 1,020.7 1,366.3 1,632.2 1,828.9 1,892.2 1,934.4

3,628.9 4,128.2 5,236.1 6,657.4 8,411.2 11,936.4 14,237.4 15,425.4 17,444.7 18,854.7 21,377.6 25,568.7 28,121.7 34,313.3 38,627.4 47,070.6 51,276.8 58,027.2

38.7 39.6 42.5 40.0 35.5 30.2 34.9 38.5 38.2 37.2 34.6 31.6 31.3 30.9 29.4 30.1 27.6

21.6 22.4

21.7 14.4

21.5 9.9

Sources: Annual reports and Central Bank of Kenya, Economic and Financial Review, various years. Note: a. The NBK uses a balance sheet date of June 30, while all other firms use December 31. For purposes of comparison, the deposits reported here are the linear interpolation of the deposits they reported six months before and six months after the end of each year.

The Financial Sector

65

Figure 4-4 Parastatal Market Shares 60 CBs NBFIs CBs plus NBFIs

«Q.

71

73

75

77

79

81

83

85

87

location of branches, which will also be discussed. Comparative Deposit Growth. The first aspect of performance of the financial parastatals to be assessed is their ability to compete for deposits. First, the growth in the consolidated (CB plus NBFI) parastatals will be compared with growth in total (CB plus NBFI) deposits. Then, growth in commercial bank parastatals relative to the rest of the commercial bank sector will be examined. Next, their success in attracting nonbank deposits will be compared with the rest of the NBFI sector. Table 4-3 compares the growth of the public sector firms (consolidating the CBs with their NBFI subsidiaries) with total growth in deposits. During the early, fast-growth period, the public firms' deposits grew on a par with total deposits. Of the three institutions, the HFCK grew at about the same rate as total deposits, the NBK grew much faster, and the KCB grew somewhat more slowly. Figure 4-4 shows that there was no change in the share of parastatals in total deposits from 1971 to 1979. However, during the later, slow-growth period the public firms experienced a greater slowdown than the industry as a whole. Total deposits were growing at 14.4 percent during 1979-1987, while deposits in public financial institutions were growing at only 9.9 percent p.a. Only the HFCK

66

Public Enterprise in Kenya

Table 4-4 Comparison of Commercial Bank Deposits (deposits in millions of current Shs, market shares in percentage)

1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988

KCB

Market Share

1,021.3 1,083.3 1,438.6 1,580.5 1,529.5 1,866.2 2,653.1 2,886.7 2,900.3 2,933.4 3,252.1 3,682.1 3,982.7 4,611.2 5,228.2 6,782.4 6,644.1 7,637.8

32.0 30.6 34.4 27.1 21.4 17.8 21.9 22.9 21.2 20.9 20.7 19.5 20.5 22.4 21.3 21.9 19.8 20.5

NBKa 211.8 337.5 528.0 747.1 950.3 1,173.6 1,374.0 1,669.7 1,948.0 1,874.8 1,825.3 1,835.5 1,926.4 2,109.7 2,282.7 2,645.7 3,029.1

Market Share 6.6 9.5 11.6 12.6 13.3 11.2 11.3 13.3 14.2 13.4 11.6 10.0 9.9 10.2 9.3 8.5 9.0

Barclays 3,186.7 3,545.3 4,559.1 5,940.2 7,157.1 10,506.8 12,130.5 12,597.5 2,556.9 3,023.7 3,208.7 3,969.9 3,963.8 4,539.9 4,665.4 6,420.3 7,490.4 8,384.3

Summary of Growth Episodes (growth in deposits, % p.a.) 71-79 13.9 34.8 79-87 10.9 5.7 14.4

Market Share

Total

18.6 21.6 20.5 21.6 20.4 22.0 19.0 20.7 22.3 22.5

13,711.5 14,002.2 15,686.0 18,392.2 19,449.8 20,625.3 24,593.2 30,954.9 33,602.3 37,270.7 20.0 11.9

Sources: Annual reports and Central Bank of Kenya, Economic and Financial Review, various years. Note: a. The NBK uses a balance sheet date of June 30, while all other firms use December 31. For purposes of comparison, the deposits reported here are the linear interpolation of the deposits they reported six months before and six months after the end of each year.

managed deposit growth faster than the sector as a whole, while KCB's growth rate was about two-thirds that of total deposits and NBK's was less than half. Figure 4-4 shows that the market share of parastatals in total deposits declined since 1979. Table 4-4 shows that during the rapid growth of the pre-coffee boom period the NBK grew very fast, 34.8 percent p.a., much faster than commercial bank deposits as a whole. On the other hand, deposits in the Kenya Commercial Bank grew significantly slower than total CB deposits, at a rate of only 13.9 percent p.a. During the slow-growth period following the coffee boom, the KCB improved its relative performance. Its deposits grew almost as fast as total CB deposits. Barclays grew even faster. The NBK languished after the coffee boom, with some shrinkage of nominal

The Financial Sector

67

Table 4-5 Nonbank Financial Institution Deposits (deposits in millions of current Shs, market shares in percentages)

KCB 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988

nil 106.2 130.2 187.8 320.2 385.7 643.9 910.0 1,133.0 1,284.3 1,139.3 2,207.5 1,367.7 1,526.9 1,550.1 2,164.8 1,872.3 2,381.9

Market Share 0.0 18.2 19.2 23.0 25.5 27.0 30.6 32.2 30.3 26.5 20.0 30.8 15.8 12.6 11.0 13.4 10.6 11.5

NBKa nil nil nil nil nil nil 87.6 231.5 306.7 448.4 491.1 468.5 708.1 524.8 654.6 760.5 719.1

Market Share 0.0

HFCK

Market Share

Total

83.3 109.7 128.2 144.3 187.2 180.9 213.6 246.8 375.7 472.6 688.0 886.5 1,020.7 1,366.3 1,632.2 1,828.9 1,892.2 1,934.4

18.8 18.8 18.9 17.7 14.9 12.7 10.1 8.7 10.1 9.7 12.1 12.4 11.8 11.2 11.6 11.3 10.7 9.3

442.1 582.9 677.0 817.3 1,254.1 1,429.6 2,106.9 2,828.0 3,733.2 4,852.5 5,691.6 7,176.4 8,671.9 12,153.9 14,034.2 16,115.7 17,674.5 20,757.0

Summary of Growth Episodes (growth in deposits, % p.a.) 71-79 30.4 ... ... ... 20.7 79-87 6.5 ... 11.2 ... 22.4

... ...

30.6 21.5

4.2 8.2 8.2 9.2 8.6 6.5 8.2 4.3 4.7 4.7 4.1

Sources: Annual reports and Central Bank of Kenya, Economic and Financial Review, various years. Note: a. The NBK uses a balance sheet date of June 30, while all other firms use December 31. For purposes of comparison, the deposits reported here are the linear interpolation of the deposits they reported six months before and six months after the end of each year.

deposits after 1979. The net effect was a shrinking of parastatal commercial banks' market share since 1979, as shown in Figure 4-4. Table 4-5 shows that the public sector firms were important in the early growth of the nonbank financial sector. KCB set the pace for this growth, with its market share growing from zero in 1971 to 32 percent in 1978. HFCK grew somewhat more slowly but still contributed a large share of the absolute growth of the sector. Figure 4-5 shows that KCB NBFI deposits grew faster than private NBFI deposits every year during the period 1972-1978. The market share of parastatals grew from 18.8 percent in 1971 to 49.1 percent in 1978. In later years KCB's nonbank deposits grew only 6.5 percent p.a., much more slowly than the market growth rate of 21.5 percent. During this period, however, HFCK held its

68

Public Enterprise in Kenya

Figure 4-5 Growth Rates of NBFI Deposits 200 HFCK KCB NBK Private NBFI

72

74

76

78

80

82

84

86

own, turning in a growth rate from 1979 to 1987 of 22.4 percent p.a. The battle for market share has been fierce, but Figure 4-5 shows that at least one parastatal has achieved growth rates exceeding those of private NBFIs during every year but one. Deposit Growth—Summary and Conclusions. In the first period under study, the high growth years of 1971-1979, the public sector firms' deposits grew at about the same rate as the market as a whole. The public sector growth was led by the newly started National Bank of Kenya, the NBFI subsidiary of the KCB (also newly started), and, to a lesser extent, by the HFCK. Under most models of oligopoly, the presence of more firms leads to more competitive behavior, and this seems consistent with the evidence on the Kenyan financial industry. Up until 1970 the banking sector was lacking in dynamism and was unresponsive to the needs of the public. Private sector firms were free to enter throughout the period under study (and many did so). The public sector firms competed with private sector firms for deposits. Their success at attracting funds demonstrates that they offered depositors better terms than did the private sector. One factor might negate this interpretation of the deposit growth rates of the para-

The Financial Sector

69

statal banks. The state controls a significant portion of deposits in Kenya directly via the national government accounts and indirectly via local government and other parastatals. The proportion of deposits controlled by the government sector grew during the fast-growth period. If parastatal deposits grew because government preferred to deposit government funds in government-owned banks, parastatal banks could not be said to have competed for funds. The direct evidence that would permit us to test this explanation is not available: the data on what percentage of deposits in parastatal banks originate in the government sector. However, the importance of government deposits in explaining deposit growth of parastatal banks can be tested indirectly, econometrically. The model estimated was the following: Pardep = A i Deppub + A2 Deppriv where Pardep = deposits in parastatal financial institutions Deppub = total deposits controlled by the national and local governments and other parastatals Deppriv = total deposits originating from the private sector In this model, A i is the portion of deposits the government places in parastatal banks. Likewise, A2 is the portion of deposits the private sector puts in parastatal banks. If A i were 1 and A2 were 0 it would indicate that parastatal banks depend entirely on government deposits. The estimates for the equation are shown in Table 4-6. Both coefficient estimates have the expected positive sign. The Cochrane-Orcutt iterative estimation technique was used to correct for the presence of serial autocorrelation. The estimates reported in Table 4-6 allow us to reject at the 90 percent

Table 4-6 Regression Results

Label Ai Az

Estimate 0.18 0.15

Degrees of Freedom=14 Sample period 1971 to 1986

Standard Error 0.20 0.10

T-Statistic 4.10 1.58 R 2 =.98

70

Public Enterprise in Kenya

confidence level the hypothesis that parastatal deposit growth was due solely to growth in government deposits. The point estimate for the coefficient on private deposits (A2) is nonzero, whereas that for public deposits is significantly different from one. The econometric results confirm that the depository parastatals did compete for both public and private funds. Hence, their presence made the financial sector more competitive than it would have been in their absence. The public sector firms were especially important in the development of the NBFI sector. When KCB entered the market for NBFI funds the market was very small. It consisted of seven privately owned institutions plus the HFCK. Of the seven, at least two were owned by and operated for the Asian community (Diamond Trust of Kenya Ltd and Ismailia Corp Ltd).18 Several were specialized in mortgage lending. KCB immediately captured a large market share (18 percent in the first year of operations), which it could only have done by offering more competitive terms to depositors. They continued to perform better than the private sector, with an explosive 30.4 percent annual growth rate, compared with private sector NBFI growth of 23.5 percent. At the end of the fast-growth period, in 1979, the public firms held 48.1 percent of the market, compared with 18.8 percent at the beginning of the period. Thus, the public sector led the way in the creation and development of the dynamic NBFI sector. It is perhaps paradoxical that it was the parastatals that pioneered in the development of the general-purpose NBFIs. The main attraction of NBFTs was the high interest rate paid on deposits, which, of course, is made possible by higher lending rates. Thus, the development of NBFIs can be seen as a means of evading the interest rate restrictions imposed by government regulations. Their evasion permitted the growth of credit to local entrepreneurs who would not have qualified for loans from the commercial banks under the tighter interest rate restrictions applied to CBs, with its resultant credit rationing. The success of KCB and NBK in operating these aggressive and dynamic NBFIs demonstrated the potential of this virtually untapped market, in addition to providing the training ground for the personnel who would tap it. A s the attractiveness of the market became apparent, many private competitors followed, including firms staffed by former employees of the parastatals. Those parts of the public financial sector that were most dynamic during the high-growth years of the seventies were very urban oriented. The commercial bank part of the KCB, with its extensive network of rural branches, grew only about two-thirds as fast as total CB deposits. The dynamism came from the KCB's NBFI subsidiary, from the NBK, and from the HFCK. All of these firms had all or most of their branches located in major urban areas, primarily Nairobi. Their growth did much to improve services in the urban areas and to increase the competitiveness of

The Financial Sector

71

the financial sector there. But little effort was devoted to extending financial services to the rural areas. The public sector financial institutions apparently changed their strategy during the slow-growth post-coffee boom period. Those parts of the public sector that primarily cater to the urban markets exhibited the worst performance. The NBK (consolidated CB plus NBFI deposits), for example, grew at a dismal 6.6 percent p.a., a growth rate about two-thirds that of the total market and well below inflation. Likewise, the NBFI subsidiary of the KCB had deposit growth of only 6.5 percent, less than one-third the rate of the NBFI sector as a whole, and also below inflation. Of public sector firms serving the urban markets, only the HFCK continued to perform on a par with the rest of the sector with which it competed. The commercial bank portion of the KCB, with its extensive rural branch network, has performed better than the urban-oriented financial parastatals during the slow-growth period. It managed deposit growth almost equal to growth in total CB deposits. Although data that differentiates KCB bank deposits by branch of origin are not available, this success must have stemmed in part from an aggressive expansion into rural areas. This expansion into the rural areas will be discussed more below. One source of the decline in relative performance of parastatals following 1979 stems from the change in interest rate policies that occurred around then. As the policy switched from low nominal rates to higher ones, CBs suffered a dramatic fall in growth of deposits relative to NBFIs. Furthermore, as noted above, rumors abound of cutthroat competition among NBFIs, who offer noninterest cash payments to attract large depositors. To the extent that the parastatals have refrained from these practices, which are of questionable legality, they have lost out relative to the private sector. This switch of deposits from CBs to NBFIs is an unintended effect of the high interest rate policy. It calls into question the existence of interest rate ceilings on the CBs and especially differential policies between CBs and NBFIs. NBFIs operate almost entirely within the largest cities.19 The draining of deposits from CBs toward NBFIs means that less credit is available to rural borrowers.20 District Focus and the Banking Sector. Part of the success of the parastatal banks in attracting a larger share of deposits derives from Kenya Commercial Bank's aggressive branching strategy. In 1986 KCB had 59 full branches, compared to 41 for Barclays. In addition, they had 43 subbranches and 126 mobile centers, compared to 7 subbranches and 34 mobile centers for Barclays.21 Thus, part of their growth came from serving markets not otherwise served, and not merely by competing for deposits within the cities. Paulson lists 143 locations that are served by only one

72

Public Enterprise in Kenya

Table 4-7 KCB Fixed Assets Versus Deposits Deposits Shs mil. 1978 1979 1980 1981 1982 1983 1984 1985 1986

2,886.7 2,900.3 2,933.4 3,682.1 3,682.1 3,982.1 4,611.2 5,228.1 6,782.4

Fixed Assets3 Shs mil. 87.0 87.8 144.4 147.7 163.4 185.9 229.0 276.3 333.3

Ratio of 1 to 2 33.2 33.0 20.3 22.2 22.5 21.4 20.1 18.9 20.3

Source: Author's calculations from annual reports, various years. Note: a. Replacement cost, current Shs. Calculated using category of asset and appropriate inflation indexes.

bank in 1983. Of those 143 locations, 130 were served by KCB. 22 This aggressive branching strategy has continued and will continue. Since 1984 it has been a declared KCB goal to have at least one full branch in every district.23 Considering that in September 1984 only 27 of the 42 districts had full branches, this represents a considerable planned expansion. These extensive rural facilities represent a large investment in fixed assets. Since 1979 the value of these assets has expanded faster than the deposits they have attracted. Table 4-7 compares the value of fixed assets to deposits. Fixed assets were revalued so that the figure given for each year represents the replacement cost in that year. Whereas in 1978 and 1979 each shilling of fixed assets supported Shs 33 of deposits, that dropped to 20 in 1980, where it has remained. Profitability. Table 4-8 shows rates of return on equity in the financial sector. A brief glance at the table reveals that the sector has been highly profitable for all the firms for which data are available, in almost every year. From a simple financial point of view, financial parastatals have been extremely good investments. A more detailed examination of Table 4-8 reveals interesting details. First, throughout the period 1971-1979 the NBK was less profitable than the KCB or HFCK. Out of 17 comparisons of NBK's profits with those of another firm, the NBK surpasses one of the others only four times. This is consistent with our view of the NBK as an aggressive entrant into a highly concentrated oligopolistic market. It probably spent more than KCB to attract depositors, since it had to attract a clientele to become established.

The Financial Sector

73

Table 4-8 Rates of Return on Equity3 (percentages) Year 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988

Barclays

50.8 38.9 29.8 67.0 55.4 55.5 59.5 54.2 49.6

HFCK

KCB

NBK

69.1 50.5 60.3 61.2 106.2 46.2 27.2 31.1 52.7 38.3 45.3 87.1 52.6 30.2 31.5 9.3 25.0 54.8

32.9 9.7 42.8 37.4 47.1 62.6 62.0 37.9 25.8 25.1 22.3 32.4 28.0 35.2 38.7 40.3 32.9

17.3 20.5 27.1 29.9 54.4 30.6 35.8 38.4 -128.9 28.8 25.0 15.8 21.6 -3.5 5.0 11.7 13.9 19.0

Source: Author's calculations from annual reports, various years. Note: a. Using average of equity at beginning and end of year.

Second, the HFCK has been more profitable than the commercial banks. Out of 44 comparisons (17 with KCB, 18 with NBK, and 9 with Barclays), the HFCK's profits were higher in 29. This reflects the greater profitability of the NBFI sector. The profit rates reported for the KCB and NBK are based on the consolidated performance of the banks with their NBFI subsidiaries. Through 1985 all the instances in which the KCB and NBK surpassed the HFCK in profitability occurred in the earlier era during which their nonbank deposits were growing rapidly. In the later period, when KCB and NBK didn't compete effectively for nonbank deposits, they were always less profitable than the nonbank HFCK. In 1986 and 1987 the HFCK suffered due to the banking crisis. Its deposits grew slowly as the public panicked and withdrew from NBFIs generally. The HFCK also had substantial deposits in the institutions that failed, on which it lost substantial interest income. Third, the parastatal banks were less profitable in the later era. Before 1979 KCB's rate of return only once dipped below 30 percent, but after 1979 it achieved that rate only half the time. Since 1979 the NBK has experienced two years of losses, the only losses reported by any firms in the sample. During the rest of the later period, in the years in which the NBK reported profits, its performance was equivalent to its worst years during the early period. On the other hand, the HFCK's profits improved

74

Public Enterprise in Kenya

Table 4-9 Effect of Bad Loans on NBK Profits (millions of Shs)

Year

Reported Profit

Bad Loans Written Off

1979 1980 1981 1982 1983 1984 1985 1986 1987 1988

-89.4 21.7 29.7 26.6 49.9 -8.5 12.3 29.4 38.3 52.5

152.8 22.8 5.4 9.4 19.7 42.9 23.7 12.8 57.2 40.2

Source: Annual reports, various years.

a bit during the later period. The worse profit performance of the NBK and KCB during the later period does not stem from worse market conditions generally. For the KCB at least, the drop in profitability is tied to its reorientation away from the urban, NBFI market and towards the rural, commercial bank sector. Through its policy of opening facilities in unprofitable remote areas, it has sacrificed about half its potential profits.24 The worsening of NBK's profitability in the later period has another cause which merits discussion. That is the matter of alleged embezzlement on the part of the bank's executive chairman. In 1981 allegations were made in Parliament that the chairman had stolen Shs 293 million from the bank during 1978 by taking out loans that would not be repaid. Documents were introduced into evidence showing the details of Shs 50 million of the alleged stolen money. While no charges were ever filed related to the matter, annual reports in later years confirm that large bad loans were made. These loans have been written off from 1979 until at least 1988. By 1984 the total amount written off came to some Shs 250 million.25 Table 4-9 shows that the bad loans written off were large relative to NBK's profits and explain most of NBK's decline in profitability. There is no way to know what fraction of the bad loans were due to the alleged malfeasance; probably some were normal losses. It is not known how high loan losses would normally be. The NBK didn't report writing off any loans prior to 1979. KCB and Barclays don't report their loan write-offs. Finally, the entirely private Barclays Bank was much more profitable than KCB during the period 1979-1988, with profit rates averaging 64 percent higher. A large part of this difference must stem from KCB's

The Financial Sector

75

policy of serving unprofitable rural markets, which will be discussed more below. It could also reflect differences in efficiency, but since neither KCB nor Barclays report their costs, this is impossible to assess. Cost Margins and Returns to Employees. It is frequently alleged that parastatals are run for the benefit of their employees at the expense of the other potential beneficiaries. This might happen through a firm hiring more employees than it needed, or by rewarding them more than needed to retain them. Such rewards could come in the form of high salaries, fringe benefits, on-the-job consumption, or through opportunities to divert company assets to personal use. All these practices can exist at any level, be it manager or sweeper. All these practices will show up in a firm's accounts as an elevation of the firm's costs. Unfortunately it is difficult to evaluate whether costs are elevated by such practices, since there is no good standard by which to judge what costs should be. Table 4-10 shows available data on costs relative to loans outstanding. If no upward trend in this ratio exists, then at least there is no evidence that employees are increasingly exploiting their positions at the expense of other interest groups. The table shows the effect of the bad loans written off by the NBK from 1979 to 1985, as mentioned above. The limited data available, for NBK and HFCK, suggest that aside from the problem with NBK management mentioned above, there was no upward trend on cost margins before 1982. There may be cause for concern since then, though the HFCK seems to have arrested and reversed its problem of the mid-1980s. There is indirect evidence that, at least at managerial levels, employees are not being overpaid. Most of the mushrooming locally owned private sector financial institutions are staffed by former managers from KCB and NBK. The fact that these managers chose to leave proves that the parastatal sector is not overcompensating the value of their skills. Conclusions—Performance of Depository Parastatals The parastatal financial institutions have been used as effective tools of national development. The manner of their use differs before and after 1979, when the general government development strategy changed. Before 1979 the parastatal firms were used to foster economic growth pure and simple. Private foreign capital was hesitant about economic prospects in Kenya following independence. Private local capital was not yet in a position to provide adequate investment and services. The parastatals filled a vacuum, providing financial services on terms that drew business away from private foreign firms, and were highly profitable. By 1979 local financial markets were more developed and had more participants, including local entrepreneurs.

76

Public Enterprise in Kenya Table 4-10 Costs (excluding interest) as a Percentage of Loans Outstanding

HFCK 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988

3.0 2.5 2.9 2.2 1.9 1.9 2.4 2.5 2.9 2.5 2.3 15.8a 7.0a 5.5a 5.6a 3.3a 3.8a 4.1a

NBK

4.1 3.9 3.8 15.4 5.1 4.3 5.4 5.4 7.7 7.1b

NBK (excluding loan loss write-offs)

4.1 3.9 3.0 3.5 3.6 4.0 4.8 4.4 5.1 5.6

Notes: a. Costs were arrived at indirectly in these years. They were defined as total income less profit and interest expense. It is possible that some change in accounting categories could explain the surprising jump, b. Costs were not reported after 1985.

Since 1979 the financial parastatals have competed less vigorously in urban markets, which are well served by private firms. Major emphasis has instead been put on expanding services to rural markets not previously served. This reorientation of the financial parastatals is consistent with the entire government policy since 1979, which has put less emphasis on simple numerical growth and more on questions of regional distribution. This policy has been less profitable financially. The impression obtained from observing branches and interviewing staff is that it seems to have been carried out in an efficient and businesslike way.

THE AGRICULTURAL FINANCE CORPORATION The Agricultural Finance Corporation ( A F C ) is a specialized nondepository financial institution. It was established in 1963 for the purpose of providing loans to Africans who purchased farms from departing white settlers around independence. It was later (1966 unofficially, 1969 officially) amalgamated with the Land and Agriculture Bank, which had been

The Financial Sector

77

the main source of agricultural loans to settlers before independence. This section first analyzes the performance of the A F C as reflected in its audited accounts. However, there is serious reason to doubt the economic relevance of these accounts. Probable sources of error and resulting bias in the results will be discussed.

Financial Performance The A F C was started as a small organization, with only 25 employees. By 1966 the combined A F C and Land Bank had 147 employees, including 17 expatriates. It had nine branch offices and seven subbranches. In 1987 the A F C had 50 branches and subbranches and 1,100 employees. Unlike the financial institutions discussed in the previous section, the A F C does not accept deposits. The capital the A F C lends comes from grants, loans, and the accumulated general reserve. Much of the capital is obtained from soft loans from international agencies such as the World Bank and USAID, which is on-lent by the government to AFC. The financial structure of the A F C is shown in Table 4-11. The A F C was started in 1964 with a small equity base of Shs 7.8 million, which was then eroded away due to losses, leaving only Shs 1.4 million of equity by 1969. During this early period the A F C received infusions of new capital in the form of loans. The combination of eroding equity and growing net assets means that the ratio of equity to net assets fell from 15.7 percent in 1964 to 1.7 percent by 1969. During 1970 large capital transfusions were received. Equity jumped fiftyfold, while net total assets more than doubled, giving the A F C a much more favorable equity to net assets ratio of 25.7 percent. Following this transfusion, performance continued to be lackluster. Equity stagnated in nominal terms from 1970 to 1976, during which time price levels rose by 87 percent. Even this overstates the health of the AFC. Equity levels were maintained only by an inflow of interest-free grants, which balanced the declining general reserves as the A F C continued to earn losses. From 1976 to 1987 the A F C recorded a healthier performance. Equity grew at 15.3 percent p.a. while inflation averaged 9.0 percent p.a. While total equity nearly quadrupled, general reserves grew twentyfold, the first self-financed growth the AFC ever recorded. Net total assets also increased by three and one-half times, due to continuing infusions of loan capital. Since 1980 there has been an alarming increase in current liabilities. Current liabilities as a share of total assets hovered in the range 2 to 5 percent from 1973 to 1979. Then they began a rapid climb to 34.6 percent in 1986, an astonishing ratio for an organization whose business is longterm lending. Table 4-12 provides data on the interest rate structure of AFC. Column 1 shows interest payable by A F C as a percentage of its interest-

78

Public Enterprise in Kenya

Table 4-11 Financial Structure of AFC

Equity (mil. Shs) 1964 1965 1966 1967 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987

7.8 7.8 6.0 4.9 1.4 51.0 50.4 49.9 50.4 52.8 49.8 51.1 59.0 70.4 83.5 100.5 109.4 112.8 120.2 143.7 175.4 192.1 245.3

Net Total Assets (mil. Shs) 49.7 55.7 65.3 67.1 82.8 198.7 211.3 227.7 255.8 283.7 329.2 414.1 492.0 524.7 625.9 705.2 767.8 821.7 828.8 812.7 987.4 1,424.0 1,855.8

Equity/Net Assets

(%)

15.7 14.0 9.3 7.4 1.7 25.7 23.9 21.9 19.7 18.6 15.1 12.3 12.0 13.4 13.3 14.3 14.2 13.7 14.5 17.7 17.8 13.5 13.2

Current Liab./ Total Assets

(%)

1.5 0.6 7.9 7.4 3.5 8.0 9.0 6.0 4.4 3.5 3.5 2.8 2.6 2.5 2.7 7.8 20.2 19.1 23.4 34.4 29.9 34.6 26.0

Source: AFC annual reports, various years.

bearing debt. These interest rates were low until 1985. Especially in the late seventies and early eighties, when interest rates elsewhere were so high, the rates the AFC was charged on its debt were highly concessionary. Column 2 shows the interest receivable by AFC as a percentage of total loans outstanding. These figures can be contrasted with those in column 3, the maximum loan rate charged by commercial banks. AFC loan rates were consistently lower than rates charged by commercial banks, which were in turn lower than those charged by nonbank financial institutions. A comparison of columns 1 and 2 shows that the degree to which AFC passed on its own interest rate subsidy to its borrowers decreased markedly. During the periods 1965-1967 and 1981-1983, AFC's average interest rate on debt was the same, 3.4 percent. During the former period the average interest rate AFC charged its clients was 6.8 percent, while during the latter period its interest rate charges averaged 10.4 percent. Column 4 shows the rate of return on investment in AFC. This is

The Financial Sector

79

Table 4-12 Average AFC Interest Rates and Rates of Return Interest Payable Interest Receivable CB Loan as % of Debt as % of Loans Rate (max.) 1965 1966 1967 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987

2.9 3.4 3.8 5.5 4.0 4.4 4.8 4.6 4.3 3.9 3.8 3.6 3.4 3.4 3.2 3.4 3.7 4.0 3.9 6.0 8.2

8.4 6.3 5.8 9.3 6.3 7.1 7.4 7.3 7.5 7.7 8.9 8.1 8.9 9.2 9.1

11.0 11.0 12.3 12.2 12.6 13.2

9.0 9.0 9.0 10.0 10.0 10.0 10.0 10.0 10.0

11.0 14.0 16.0 15.0 14.0 14.0 14.0 14.0

ROI

ROE

2.5 0.1 1.8 4.1 3.3 3.5 3.6 4.1 2.5 3.5 5.1 5.4 5.0 5.5 5.5 6.1 5.9 7.3 6.7 7.2 9.6

-0.2 -25.5 -20.1 -40.7 -17.2 -2.4 -1.0 2.2 -6.4 0.8 14.5 17.3 15.2 18.9 19.3 22.7 19.3 24.7 19.9 14.0 19.0

Source: AFC annual reports, various years.

pretax profit plus interest payable as a percentage of total net assets. The A F C earned a small positive R O I every year but one. That such returns are inadequate can be seen by comparing column 4 with columns 1 and 5. The returns on total investment reported in column 4 are less than the average interest rate on A F C debt in column 1 every year until 1977, which explains the string of negative returns on equity (pretax profit over equity) reported for the first decade of A F C ' s existence. Even if government never intended to turn a profit from loaning money for agriculture, the financial performance of the A F C in its first decade was so poor that the capital base was being eroded away. The A F C was not even self-sustaining, let alone profit-making. After 1977 the profit performance of A F C improved. Over the period 1977-1983, A F C reported an average R O I of 5.5 percent, more than double the average of 2.4 percent reported during the 1965-1976 period. While A F C ' s R O I in the post-1977 period cannot be considered high, it at least surpassed the cost of borrowed funds, so that returns on equity were positive, mostly in the range of 15-20 percent for the period 19771987. A F C ' s improved profitability derives from the reduction of interest

80

Public Enterprise in Kenya Table 4-13 Transfusions of New Capital to AFC ('000 Shs) Retained Earnings 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986

New Grants

-18.1 -1,763.2 -1,101.1 -3,508.1 -1,683.6 -1,087.4 -673.6 -481.6 1,127.9 -3,261.7 392.4 7,961.4 11,181.7 11,668.1 17,397.9 7,810.5 2,465.5 6,939.6 14,755.2 18,603.0 1,846.2

51,213.7 549.6 166.0 996.0 1,257.5 300.0 874.9 -44.9 140.0 1,480.4 -402.6 1,051.4 996.2 477.8 8,828.6 73,157.0 220,797.2

New Debt

Total

6,000.0 11,390.6 2,874.5

5,981.9 9,627.4 1,773.3

19,246.0 66,328.8 13,200.4 16,838.2 27,587.9 26,764.8 48,535.0 83,637.3 69,906.2 21,464.2 87,962.1 62,356.3 53,738.2 50,417.2 669.0 -40,660.6 82,993.3 213,908.6

15,246.0 115,859.0 12,662.6 16,330.6 28,102.3 27,892.7 45,573.4 84,904.8 77,822.8 32,785.8 101,110.6 79351.6 62,600.2 53,878.8 8,086.4 -17,076.8 174,753.3 436,552.0

Source: AFC annual reports, various years.

rate subsidy to AFC borrowers, as seen from comparing columns 1 and 2. The increase in interest rates AFC charges on its loans and the resulting improvement in financial performance do not reflect a decision that AFC must rely on internally generated funds. Table 4-13 shows sources of capital for the AFC. AFC continued to receive large infusions in the form of loans as well as small infusions in the form of grants. Paulson noted that since 1976/77 appropriations from the budgets of the ministries of Agriculture and Livestock Development to AFC fell as a percentage of the total development budgets for those ministries.26 Table 4-13 shows that the transfers to AFC after 1977 were large relative to previous transfers. The reported improvement in AFC's financial performance reflects one aspect of a commitment to maintaining AFC operations, not a cutting off of AFC from its access to soft funds. Efficiency and Returns to AFC Employees Figure 4-6 shows AFC costs as a percentage of AFC's portfolio of loans to farmers. These costs include staff costs, general administration, depreciation, and board members' fees and allowances. They do not

The Financial Sector

81

include provisions for bad debts, nor the opportunity cost of low cost loans made to staff. Figure 4-6 shows no particular trend in cost levels in the sixties and seventies. However, the level of costs grew steadily during the eighties, indicating growing inefficiency or increased exploitation by the employees of the resources of the AFC. 27 The A F C could have hidden excessive returns to A F C staff by giving excessive staff loans rather than high salaries. Data on loans to employees are also presented in Figure 4-6. There was a steady decline in loans to staff as a percentage of loans to farmers from 1969 to 1976. The level skyrocketed in 1986 and 1987. Most of the provisions for bad debts are applied against loans to farmers. Hence, there is no evidence that the A F C could be making loans to its staff that it would then write off. To summarize, the data reported by the A F C show that A F C costs have risen steadily in relation to its loan portfolio since 1980. This may mean that A F C staff are increasingly exploiting their position and capturing increasing returns. This could be done via increasing salaries, employment, or expenditure on fixed assets, including staff housing, or any combination thereof. Comparison of data from Figure 4-6 with that from Table 4-10 shows that before the escalation of costs began in 1980, the Figure 4-6 Measures of Returns to Employees: Agricultural Finance Corporation

10

Staff loans/loans to farmers Administrative costs/loans to farmer

8

6

c o 2

x: CO

74

75

76

77

78

79

80

81

82

83

84

85

86

Sources: Paulson, page 23 for 1974-1983. Later years from AFC annual reports and Quarterly Economic Review. AFC credit excludes seasonal credit schemes.

88

Public Enterprise in Kenya

sionary terms, at low interest rates and for long terms, and the choice of who receives those loans is politicized. Its greatest departure from commercial practice is its failure to collect on its loans. The repayment performance has been so poor as to endanger its ability to continue operations.

CONCLUSIONS This chapter has examined the performance of four financial parastatals. The first three are depository institutions. The performance of these institutions during the 1970s was excellent. They were highly profitable and competed successfully with the private sector. They helped to. expand the financial sector and cause it to behave more like a competitive market and less like a tight oligopoly. During the period 1979-1986 the behavior of the parastatal commercial banks changed. Emphasis shifted from urban markets to rural ones. Kenya Commercial Bank embarked on a policy of aggressive branching in rural areas. This policy is costly in terms of foregone profits, but appears to have been accomplished without sacrificing efficiency or changing from commercial criteria for loans to political ones. Several dangers to the sector were identified. Key among these was political pressure to make bad loans. The depository institutions appear to have been successful at resisting pressure to make unsound loans to politically well connected individuals. They have been less successful at resisting pressure to loan to bankrupt parastatals. To date the situation is manageable, but if the situation were to worsen, it could ruin the effectiveness of the banks. The potential for trouble due to politicized loans was illustrated by the case of the Agricultural Finance Corporation. This firm was shown to be unprofitable and inefficient. Its loan procedures are politicized and its collection procedures even more so. These problems are so severe that the corporation is not self-sufficient. Though it has continued to receive large transfusions of new capital on soft terms, the A F C is so ineffective at debt collection that it has repeatedly run out of funds and been forced to suspend new lending. The A F C can only be seen as a mechanism to transfer resources from the exchequer to the AFC's clientele. This clientele is, on the average, far wealthier than that of the commercial banks. In contrast to the positive performance of the depository institutions, the A F C has contributed little to national development. Its chronic squandering of funds and failure to mobilize finance for agriculture calls into question the use of soft funds and noncommercial criteria to allocate resources that are supposed to be invested in private sector, commercially viable agricultural enterprises.

The Financial Sector

89

NOTES 1. Paulson, page 74. 2. Ibid., page 62. 3. Central Bank of Kenya, 1972, page 20. 4. Swainson, page 189. 5. Paulson, page 74. 6. Ibid., page 75. 7. Ibid., page 65. 8. See McKinnon. 9. Central Bank of Kenya, 1972, page 19. 10. J. N. Michuki, chairman of KCB, as quoted in Leys, page 134. 11. Quoted in Leys, page 157. 12. Leonard, page 253. 13. Interview, KCB staff. 14. Paulson, page 187. 15. "Kenya's White Elephants: The Government Continues to Put Good Money into Failed Projects," Financial Review, November 24,1986,3-5. 16. Swahili for citizens, or "the people." 17. The subsidy is about as big as reported profits of the KCB. In other words, half the potential profits of the KCB are consumed supporting branches that have low demand. Confidential personal interview, KCB staff. 18. Central Bank of Kenya, 1972, page 22. 19. In 1986, 68 out of 92 NBFI offices were in Nairobi or Mombasa. See Mullei and Ng'elu, page 270. 20. Paulson, page 67. 21. See Mullei and Ng'elu, page 269. 22. Paulson, Appendix C. 23. Report of a speech by the executive chairman of Kenya Commercial Bank, Dr. Benjamin Kipkorir, reported in the Sunday Nation, September 9,1984. 24. Interview with KCB staff. 25. In fact, the total written off exceeds that amount. The accounts of the NBK are consolidated with the Kenya National Capital Corporation, its NBFI subsidiary. However, NBK owns only 60 percent of the Kenyac, the Kenya National Assurance Corporation owning the other 40 percent. Hence, of loans written off by Kenyac, only the NBK's share of the losses is reflected in the NBK accounts. Most of the 1984 write-offs were Kenyac write-offs. The breakdown is not available for previous years. 26. Paulson, page 118. 27. This trend is probably exaggerated because of inflation. See the discussion on page 103 of this book, as it pertains to the Development Finance Institutions. 28. Paulson, pages 130-133. 29. Ibid., page 155. 30. Ibid., page 123. 31. Ibid., page 136. 32. Ibid., page 157. 33. Ibid., page 160. 34. Ibid., pages 108-109.

90

Public Enterprise in Kenya 35. 36. 37. 38. 39.

Ibid., page 133. Leys, page 102. Paulson, page 114. Ibid. Ibid., page 115.

Development Finance Institutions This chapter describes and evaluates the performance of five public enterprises that share the goal of fostering development of one or more sectors of the economy. These five parastatals, termed Development Finance Institutions (DFIs), primarily foster development in the private sector. This goal makes evaluation of their performance difficult. A complete evaluation would require examining the firms they aided, and evaluating the performance of those firms. I will take a threefold approach, first describing the DFIs and their role in the capital markets of Kenya. Second, I will examine three aspects of the performance of the DFIs per se: (1) whether DFIs subsidize their subsidiaries1 by offering long-term finance on concessionary terms, (2) efficiency at the level of the DFI, and (3) the record of DFI returns to shareholders. Finally, I will evaluate the manufacturing subsidiaries of the DFTs, using a rich data source for examining parastatal manufacturing subsidiaries. No data on the performance record in other sectors are available. DESCRIPTION

The DFIs have used diverse tools in their common goal of fostering development, which will complicate comparison of performance indicators. Table 5-1 lists the firms, with their year of incorporation and their major tools. All the firms offer credit to private firms and make equity investments in subsidiaries, as ways of encouraging development in their sector. They also offer management services. The Agricultural Development Corporation (ADC) operates state farms and manages private ones on contract. With the exception of ADC's farming activities, all of the major activities are geared to encouraging private sector activity. All of the firms have both aided African entrepreneurs and encouraged foreign investment. There is a marked tendency toward organizational proliferation and duplication among DFIs. The first DFI was the Industrial Development Corporation (IDC), started in 1954 to assist and encourage medium- and large-scale investment in the industrial sector. Around independence the IDC was changed to the Industrial and Commercial Development Cor91

92

Public Enterprise in Kenya

Table 5-1 Development Finance Institutions

Firm

Year of Incorporation

Agricultural Development Corporation (ADC)

1964

Development Finance Company of Kenya (DFCK)

1963

Industrial and Commercial Development Corporation (ICDC)

1954

Industrial Development Bank (IDB) Kenya Tourist Development Corporation (KTDC)

1973 1965

Major Activities Management of state and private farms. Equity and loans in agro-industrial enterprises. Equity and loans in medium- and small-scale projects, mostly manufacturing. Loans to smallscale projects. Equity and loans in medium- and large-scale industrial and commercial projects. Loans for small-scale projects. Equity and loans in medium and large-scale industrial projects. Equity and loans in tourism projects. Hotel management.

poration (ICDC). Programs were started to lend to small- and mediumscale African entrepreneurs. Around the same time the Development Finance Company of Kenya (DFCK) was started to lend to medium- and large-scale industrial projects. ICDC held equity in DFCK, along with European foreign aid agencies. From the beginning, ICDC and D F C K overlapped in the area of medium- and large-scale projects. Later D F C K added small-scale loans to its lending program, completing the duplication. Again in 1973 the ICDC established a new subsidiary whose role would be equity and loan participation in medium- and large-scale industrial enterprises—this time the Industrial Development Bank (IDB). Again ICDC did not withdraw from direct participation in the sector. ICDC has functioned parallel to its subsidiary, the IDB, participating in many of the same projects. There has also been redundancy in promotion of small-scale enterprises. I C D C also loans to small enterprises for industrial, machinery, and property loans. In addition to these loans ICDC started a subsidiary in 1967, the Kenya Industrial Estates (KIE), which provided the same kind of loans to the same clientele. In 1978 the ownership of K I E was transferred to the Treasury and the Ministry of Commerce and Industry. Reportedly, at»the time of separation, K I E was saddled with many bad loans made earlier by ICDC. ICDC has continued to lend in the sector. The K I E would have been included in this chapter had their annual reports been more informative.

Development Finance Institutions

93

Figure 5-1 gives two indicators of the importance of the DFIs in capital markets in Kenya. The role of the institutions increased fairly steadily until 1977. By then portfolios of the DFIs were equal to almost 15 percent of total credit from the banking sector (commercial banks and nonbank financial institutions combined) and new net investment by the DFIs amounted to 5 percent of capital formation in the enterprises and nonprofit institutions sector of the economy. After the coffee boom the relative importance of the DFIs declined. The combined portfolio of the DFIs gradually declined to 5 percent of banking sector credit, and net investment by DFIs was less than 5 percent of capital formation. The role of the DFIs has been crucial to the development of the sectors in which they participate. They are major sources of long-term finance. Commercial banks provide practically no term loans suitable for financing fixed-capital investment.2 Although Kenya's capital markets are far more developed than those of most sub-Saharan African countries, they are still highly underdeveloped. Only six corporate debt instruments are traded on the Nairobi Stock Exchange. The market for equities is slightly better developed but has actually declined in the past decade. It is not an attractive source of finance for new

94

Public Enterprise in Kenya

or fast-growing companies that want to retain most of their earnings for reinvestment. Because the market is so thin, stockholders cannot be sure of realizing the value of retained earnings when they sell their stocks, so equity prices are set with respect to dividends only. A n example of this came to light in 1986 in the case of the profitable Kulia Investments Ltd. The majority owners of this firm wanted to retain substantial portions of its earnings (80 percent in 1985) for reinvestment, while the minority shareholders have pressed for higher dividends. The conflict of interest became severe enough that the majority shareholders were willing to buy out the minority shareholders at prices two to four times recent market prices for the stock. 3 Eight other formerly publicly traded companies have exited the stock market since 1980, cutting the number of firms traded from a high of 59 in 1968. Only three firms have gone public in recent years: Jubilee Insurance Ltd, in 1984; Barclays PLC, in 1986; and the Kenya Commercial Bank, in 1988.4 All three firms are mature, so the conflict over whether to retain earnings to finance rapid growth is less of a problem. For new businesses or ones that are growing fast, the stock market has not provided an attractive source of investment funds. Given the absence of other sources of long-term credit, the DFIs' role is crucial to the firms in which they invest, especially those local firms with no access to overseas borrowing. The accounting conventions among parastatals make the interpretation of the accounts of the DFIs problematic. Equity investments are generally listed at cost. Subsequent profits or losses by DFI subsidiaries are not reflected in the DFIs' accounts, either in the income statement or in the balance sheet. It is not clear what bias this introduces. Some of the subsidiaries have been highly profitable, and the value of the DFIs' shares far exceeds their cost. Other firms have suffered chronic losses, so that the value of the DFI shares is less than their cost—in some cases zero. Without tracking the accounts of the dozens of subsidiaries, it is impossible to state the effect on the DFIs. ICDC carries its investments at cost on the balance sheet, but in the notes to the accounts, it gives the market value for investments in firms quoted on the stock exchange and the directors' valuations of total unquoted investments, based on the retained earnings history. The reported revaluations have exceeded cost, at least since the mid-1970s. For example, in 1981 the directors' valuations exceeded cost by 61 percent. This suggests that the balance sheet has consistently understated the value of investments. General inflation, especially since the early 1970s, has contributed to this understatement. This chapter relies on the book values given for the portfolios, but reference will be made to other possible reinterpretations if the directors' valuations are accurate.

Development Finance Institutions

95

MEASURES OF PERFORMANCE OF THE DFIs PER SE Terms on Which DFIs Invest The role of the DFIs is to supply long-term capital in the form of loans or equity investments. The recipients of this finance constitute the clients of D F I services. In this section I will examine rates of interest or dividends clients paid in return for the capital received from DFIs. The available data on returns to the investments of the DFIs are presented in Figure 5-2 through Figure 5-6. Figure 5-2 compares the average rate of return on the total investment portfolio for the five holding companies with the interest rate charged by commercial banks. The rate of return included interest and dividends receivable, divided by loans plus equity investments. Since Figure 5-2 excludes unrealized capital gains, it probably understates the true returns on D F I investments. Figure 5-2 shows that the rates of return on the portfolios of the five holding companies have varied considerably. The K T D C and A D C consistently had returns below those of the manufacturing DFIs. The average rate received by A D C never surpassed 5 percent. With the exception of the ADC, all of the firms showed a marked upward trend in returns on their portfolios beginning in 1977 or 1978, though the trend was interrupted in 1982 and/or 1983. Figure 5-2 Rate of Return on Portfolio 30

64

67

70

73

76

79

82

85

88

96

Public Enterprise in Kenya

The ICDC, IDB, and DFCK all had average returns only slightly below the commercial bank lending rate in most years. Figures 5-3 through 5-6 show this in more detail. Figure 5-3 demonstrates that since 1977 the yield on IDB loans has nearly always equaled or exceeded commercial bank lending rates, but the yield on equity has been very low. The yield on equity improved dramatically in the eighties, as the IDB's investments "came of age" and began paying dividends. Figure 5-4 shows a different pattern for ICDC. From 1966 to 1980 yields on both loans and equity approximated commercial lending rates. During the early eighties ICDC's dividend yield suffered, but it had recovered to rates approaching the yield on commercial bank loans by 1984. The high yield on ICDC's loan portfolio probably reflects their policy of generous provisions for bad loans in their small loans scheme. In addition to writing off known bad loans, they make a general provision of 20 percent of commercial loans, 25 percent of industrial loans, and 50 percent of machinery loans. If the default rates are actually lower than those figures, the effect on the reported yields could be dramatic. In addition to reporting the value at cost (or book value) of its equity portfolio, ICDC has, since 1977, reported a directors' valuation based on the value of retained earnings in its subsidiaries. The annual change in the value at cost represents new investment at cost. Any difference between this and the change in the directors' valuation must then represent unrealized capital gains in the portfolio. Adding these unrealized capital gains to the realized dividends yields a more economically meaningful measure of the total return on the equity portfolio (measured at the directors' valuation). This measure is shown in Figure 5-5, along with the measure based on dividends and book values alone. The more comprehensive measure is far more volatile but moves in the same direction. The lower volatility of the dividends-only series reflects the common practice in Kenya of smoothing dividend streams relative to profit streams. The average rate of return during the period is 16.4 percent using the more comprehensive measure, compared with 11.5 percent using only dividends relative to book value. The historical cost accounting convention has distorted ICDC's performance by making it appear more concessionary than it was. ICDC's subsidiaries have not paid high dividends to ICDC, but as equity holder, ICDC owns a proportional share of the retained earnings. ICDC could realize these returns if it chose, either by selling shares or by insisting on higher dividend rates. Instead, it has been satisfied to allow its successful investments to retain the capital for internal reinvestment. Whether the figures for the other firms are similarly distorted depends on the extent to which their subsidiaries were financially successful. Figure 5-6 shows an intermediate pattern for DFCK. Its yield on loans

Development Finance Institutions

Figure 5-4 ICDC Rates of Return 30 Equity Loans CB lending rate

65

68

71

74

77

80

83

86

98

Public Enterprise in Kenya

Figure 5-5 ICDC Alternative Measures of Returns on Equity 60 Realized returns, book value Realized + accrued, revalued

c

8

o «o

i

r2

äS "

§

ce en

a

00

§

a

172.98

r-

1

S

oo

1

o a

oo ci

>o r-

S

?rl

M

n

§

rj

a

©

2

r-

-

1

79.75

i

î

U

Ò


o ri £

S

3

u

"E

£ §

eo et

S Fi

n ofj

r-

a

00 ©

fO o a o

£

f»> a a s a ci E o o a o o

Jh

5

ro ft

3

s

9s 00 (N

O 0\ ri

R

1

•n r*>

i

n

a

s

8

«o

8

5

R

--

a E o

«

©

„E

m a o o a o PI o

H

1

s

Os f>

5

9« M

n

E

9

00 «n

N

§

S

o

Os

00

n

s

a

O H O O E o R

Pi

§

9s

E

r00

S

s

•«t M

t->

m §

»

«

a

r»

ft

0 00 >c SrO « 0 o E s o ©

©

¡2

£ g s 1 s

ft a.

p s 2 *

B

£

a

"a

S e

g A

3

"5.

e g e

V) ri

ri to S ci

r» cn ri

se

«n

© ©

^

© o

E

I

ri ri fi

ri

V O S

E

E

§ a

«E

00

ri

n