99104 Marketing Aftica's High-Value Foods .Comparative Experiences of an Emergent Private Sector Edited by Steven Jaffee and John Morton [I KENDALL/HUNT 4050 Westmark Drive PUBLISHING P.O. Box 1840 COMPANY Dubuque, Iowa 52004·1840 The opinions and recommendations presented in this book are the result of research supported by the World Bank; however, they are solely those of the respective authors, and do not necessarily represent the position or policies of the World Bank, its affiliated organizations, members of its Board of Executive Directors, or the countries they represent. Cover photos courtesy of The World Bank Copyright© 1995 by The World Bank International Bank for Reconstruction and Development Washington, D.C. 20433 Library of Congress: 94-78572 ISBN 0-8403-9760-7 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted, in any form, or by any means, electronic, mechanical, photocopying, recording, or otherwise, without the prior written permission of the copyright owner. Printed in the United States of America 10 9 8 7 6 5 4 3 2 1 To our wives Gilat and Diana, and our children Danielle, Elan, and Helena. Table of Contents Acknowledgements _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ __ Vll I. Introduction _ _ _~--------------- 1 PART A: II. Transaction Costs, Risk and the Organization of Private Sector Food Commodity Systems _ _ _ _ _ _ _ __ 21 Steven Jaffee III. Private Sector High-Value Food Processing and Marketing: A Synthesis of African Experiences _ _ _ _ _ __ 63 Steven Jaffee and John Morton PART B: IV. Private Sector Response to Market Liberalization in Tanzania's Cashew Nut Industry _ _ _ _ _ _ _ _ _ __ 153 Steven Jaffee V. Perishable Profits: Private Sector Dairy Processing and Marketing in Kenya _ _ _ _ _ _ _ _ _ _ _ _ _ __ 199 Steven Jaffee VI. Market vs. Government Failures: A Case Study of the Vanilla Sector in Madagascar 255 Benoit Blarel and Diane Dolinsky VII. The Many Faces of Success: the Development of Kenyan Horticultural Exports 319 Steven Jaffee VIII. Fish Mammies and Tuna Conglomerates: Private Sector Fish Processing and Marketing in Ghana 375 Geoffry Ames and C.J. Bennett vi MARKETING AFRICA's HIGH-VALUE FOODS IX. Merchants and Middlemen in the Cattle Trade of Southern Somalia 417 Peter D. Little X. Reflections 455 Authors' Biodata 467 Bibliography 469 Index 497 Acknowledgements T he preparation of this book and most of the research upon which it is based was jointly funded by the World Bank and the Overseas Devel- opment Administration. The sponsoring departments within the World Bank were the Agriculture and Natural Resources Department and the Mrica Technical Department. The sponsoring department within ODA was the Africa Regional Desk. The opinions and recommendations pre- sented in the book are solely those of the editors and respective authors, and do not necessarily represent the position or policies of the World Bank, ODA, or their affiliated organizations. For their support in launching this project and providing continued guidance through its various stages, we would like to thank Gershon Feder, Graeme Donovan, Kevin Cleaver, Brian Grimwood, and Felicity Proctor. For their time, advice, and support in accessing relevant literature, we thank Alex Duncan (Food Studies Group), Mary Tiffen (ODI), Anthony Ellmann and Nick Hetherington (CDC), Helen Wedgewood (ITDG), Ingrid Vanore-Speirs and colleagues (ITC), Laurens van der Laan and colleagues (African Studies Centre), Nicolas Bricas and colleagues (CIRAD), Richard Roberts, Andrew Shepherd, Marc Bral and colleagues (FAO), John Holtzman (Abt Associates), Caroline Moser, Yvan Grouitch, Tyler Biggs, Jim Coates, Eduardo Loayza, and others (World Bank!IFC), and members of the Social Sciences Group and Food Science and Crop Utilization Divi- sion of the Natural Resources Institute. For superb editorial assistance and production support, thanks to Annette Bickel and Marianne Sabers. Special thanks are extended to the many African entrepreneurs and business execu- tives whose cooperation and insights have made this book possible. 2 MARKETING AFRICA'S HIGH-VALUE FOODS TABLE 1.1: Agricultural Performance Indicators for SSA (Average Growth Rates) Indicator 1975-1980 1980-,.1985 1986-1990 Value Added in Agriculture +1.9 +0.4 +1.5 Agricultural Production +0.5 +1.0 +1.6 Per Capita Food Production -2.3 -2.1 -1.5 Volume of Agricultural Exports -0.8 -2.9 -2.5 Value of Agricultural Exports +9.6 -2.9 -3.1 Value of Food Exports +9.9 -2.0 -B.7 Value of Food Imports +19.8 -5.7 +3.2 SOURCE: World Bank (1992). This gloomy aggregate picture obscures the progress which has been achieved by some countries in obtaining higher rates of agricultural growth, in expanding or adding value to traditional exports, and in diversifying their trade into higher value commodities with favorable world market trends. Between 1986 and 1991, Nigeria, Uganda, Tanzania, Chad and Botswana each achieved agricultural growth rates exceeding 4% per year. Moreover, there are now signs of recovery in industries which had previously faltered, such as Ghanaian cocoa, Ugandan coffee and Tanzanian cashew nuts. Among non-traditional exports, both Kenya and Cote d'Ivoire have become signifi- cant competitors in the world horticultural products trade. External or unavoidable factors, including declining world prices for major export commodities, frequent drought and civil strife, have undoubtedly con- tributed to the poor agricultural performance of some African countries. But they do not explain everything. Since the early 1980s there has been a grow- ing consensus that various forms of institutional failure, particularly within the public sector, have been a central factor in Africa's agricultural and more general economic stagnation.5 According to this view, African agriculture and agro-industry have been strongly and adversely affected by misguided macroeconomic and pricing policies and by excessive government interven- tion in factor and commodity markets. These interventions, together with the imposition and maintenance of inefficient, and frequently monopolistic, state- 5 A World Bank study published in 1981, Accelerated Development in Sub-Saharan Africa, linked Africa's economic problems to misguided macroeconomic and sectoral policies and to the overexten- sion and weak performance of public institutions. Although controversial at the time, within a few years analysts from across the political spectrum were extending or echoing its analysis and major findings. They include Bates (1981), Ellis (1982), Hart (1982), Hyden (1983), de Wilde (1984), Arhin, Hesp and van der Laan (1985), and Ghai and Smith (1987). INTRODUCTION 3 owned or state-sponsored processing and marketing organizations, are held to have undermined producer incentives and cro~ded out private investment in most of the region's important agro-industries/Despite sizeable amounts of state aid and international donor financial and technical assistance, parastatal agribusiness enterprises have largely been deficient in providing support ser- vices to farmers and in managing the logistics of trade. Their record of antici- pating opportunities and responding to changes in world markets has been even worse. The policy prescriptions following from this analysis have been straight- forward, even if the implementation by specific country or sector has proven to be complex and problematic. Over the past decade, the agenda pursued by African governments and international donors has centered around macro- economic reforms, trade and domestic market liberalization, and public sec- tor restructuring. More recently, measures have been geared toward expand- ing the role of the private sector.6 Attention has been directed at improving the enabling em•ironment for business (through legal and regulatory means), at reforms which strengthen domestic financial systems and at privatizing selected public enterprises. In many countries the agricultural sector has been the subject of far-reaching proposals and programs of structural reform. A curious feature of much of the recent literature discussing new policies (and even ofthe programmed reforms) is the abstract nature of recommenda- tions regarding 'private sector development' and greater reliance upon 'mar- TABLE 1.2: World Market Shares of sub-Saharan Africa for Major Agricultural Commodities(% of World Export Value) Commodity 1969-71 1989-91 Coffee 27.3 17.8 Cocoa 77.9 67.1 Cotton 16.0 13.9 Sugar 8.6 8.6 Tobacco 7.7 13.5 Tea 12.5 15.3 Combined 20.8 16.3 SOURCE: Data from FAO Trade Yearbook, various issues. 6 These have been the focal areas of the many structural and sectoral adjustment programs imple- mented within the region. For reviews of this experience, see Helleiner (1992), Jaeger (1992) and World Bank (1994). 4 MARKETING AFRICA'S HIGH-VALUE FOODS ket forces'. Despite the widespread advocacy of agricultural market liberal- ization and privatization in Africa, the actual forms which competitive mar- kets and private enterprises might take in the African context and the speed and direction of their development are rarely examined. Lele (1988) argues that the abstract nature of the recommended institutional reforms stems from the fact that so little is known about 'non-state' marketing institutions. Al- though it is true that there have been few detailed studies on private process- ing and trading activity, even the empirical evidence that is available has rarely been introduced into policy discussions.7 Significant policy reforms are now being implemented in many African countries. Yet there is growing recognition that such reforms alone are un- likely to bring about the desired supply response from farmers, private trad- ers and processors. 8 This is leading both African governments and interna- tional donors to move toward more direct measures to promote private sector agribusiness development. In a large number of countries-including but not limited to Uganda, Kenya, Guinea, Madagascar, Malawi, Burundi, Ghana, Zambia, Cameroon and Mali-donor-funded projects geared toward promot- ing private agribusiness are already being implemented or are in prepara- tion. Many additional projects have been proposed. To intervene successfully in this area, both African governments and donor agencies must acquire a greater understanding of the private sector in Africa, the constraints it faces, its modes and strategies of operation and its varied responses to trade and market liberalization. This improved understanding will facilitate the design of interventions which are consistent with private sector strategies while also promoting broader development objectives. 7 Social science research on African domestic food markets conducted during the 1960s and 1970s found that many of the prerequisites for efficient trading relations-including effective transport and communication systems, functioning capital markets, physical security, and consistently enforced laws-were either weak or absent in Mrican economies. This gave rise to high logistical and transac- tion costs in trade dominated by the private sector. Partly as a result of these conditions, long dis- tance trade occurred most often within oligopolistic market structures. It was dominated by particu- lar ethnic groups who had developed specialized information and fmancial arrangements (see Cohen 1969; Meillassoux 1971; Jones 1972 and van der Laan 1975). Contemporary research on export- oriented agribusiness in Africa has also emphasized significant efficiency and income distribution problems (Kaplinsky 1979; Mackintosh 1980; and Dinham and Hines 1983). It is not surprising that advocates of privatization tend to compare the inefficiencies or ineffectiveness of actual public enter- prises operating in real world political economies with the theoretical merits of a competitive and dynamic private sector. 8 See Platteau (1990) for an insightful analysis of technical and environmental barriers to agricul- tural growth in Africa. INTRODUCTION 5 Investigation should consider the following structural features of African private enterprise. Characteristics of Private Entrepreneurs/Enterprises Which private entrepreneurs and firms are able to take advantage of market opportunities and liberalization? How can they be characterized with respect to ethnicity, gender, ownership, management, sources of finance and opera- tional size? What is the relative importance of indigenous, ethnic minority, joint venture and foreign companies in different markets? Operational Strategies of Private Sector What are the operational strategies adopted by private firms and individuals in response to the market, regulatory, infrastructural, and social environ- ment in which they operate? What patterns of commodity, functional and market specialization or diversification can be distinguished? What prefer- ences are shown between different types of technology and between domestic and export market orientations? Crop/Raw Material Procurement Strategies How do private firms and individuals organize their supplies of raw materi- als or commodities? Do they simply buy on a spot market basis and have little or no linkage to the production system? Or do they maintain contractual ties with suppliers or engage in direct primary-level production? How do the techno-economic characteristics of commodities influence these patterns? Organization of Private Trading What is the structure of private marketing channels and privately-dominated markets? What institutional arrangements govern long distance and export trading? What is the competitive structure of these markets and what forms of competition occur? How do the characteristics of particular commodities and the demand for them influence such patterns? Performance of the Private Sector How have private processors and traders performed as measured by their cost efficiency, profitability, capacity utilization, market and product develop- ment, and innovativeness in applying processing and marketing technolo- 6 MARKETING AFRICA '5 HIGH-VALUE FOODS gies? Have these agents effectively supported primary producers (for instance farmers, herders, fishermen) through provision of technical, marketing and other services? Role of Government in Private Sector Development What has been the role of government in fostering private agribusiness devel- opment? What public goods (such as research and infrastructure) have been essential in the development of private markets and industries? What types of regulatory frameworks have best aided-or most hindered-private sector development? How has government reduced the risks and transaction costs faced by private firms-or has it increased them? OBJECTIVE AND FOCUS OF THE STUDY The objective of this study is to expand the knowledge base on private agro- processing and marketing activity in Africa. It seeks to enrich the on-going policy discussions and program initiatives directed toward private agribusiness development. The study addresses the empirical issues identified above through a systematic review of existing studies on food marketing and pro- cessing in Africa and six detailed case studies of commodity systems in par- ticular countries. COMMODITY Focus: Most prior research on agricultural marketing and pro- cessing in Africa has centered on either grain and other staple food market- ing, or the processing and trade of 'traditional' export commodities like cocoa, coffee, cotton, tobacco and tea. The reasons for this research focus are easy to understand. While cereals and starches feature prominently in African diets and in concerns about food security, beverage crops and other traditional industrial crops have for many years accounted for a large proportion of agri- cultural foreign exchange earnings in most African countries. Prior research focus is also related to and implicated in longstanding policy discussions. Throughout Africa, governments have intervened heavily in the markets for staple foods and in the pricing and marketing of traditional export ·Crops. Finally, pragmatic considerations have contributed to the concentration of researcher attention on staple food and traditional industrial crop marketing: in many countries, the only official data on the structure and performance of production, processing and marketing that has readily been available was that collected for these crops by public marketing boards and crop develop- ment authorities. INTRODUCTION 7 This book has a different focus and therefore seeks to complement the broader literature on African agricultural marketing. It examines private pro- cessing and marketing of high-value food products and raw materials which are defined here to include: • Fresh and Processed Fish Products • Live Animals • Fresh and Processed Meat Products • Milk and other Dairy Products • Fresh Fruit and Vegetables • Processed Fruit and Vegetable Products • Tree Nuts (e.g. cashews) • Oilseeds and Vegetable Oils • Spices and Flavorings Individually or in aggregate, these commodity groups have an important though under-recognized role in African economies in terms of household ex- penditures, agricultural and manufactured product value added and foreign trade. High-value foods account for approximately one-fourth of total house- hold cash expenditures and nearly two-thirds of food-related cash expendi- tures in sub-Saharan Africa.9 As Table 1.3 indicates, for African countries with above-average rates of urbanization and higher levels of per capita GNP, the share of high-value foods in total and food-related expenditures is even higher. Over the 1980-1991 period, Africa's urban population grew at an average annual rate of 5.8%, compared with a 3.1% per annum increase in total popu- lation. By 1991, 29% of Africa's population was urban.10 With continued rapid rates of u,rban population growth and future income growth, demand for high value foods is likely to increase significantly in Africa. 9 In most analyses of the food situation in Africa, it is generally assumed that cereals are the mainstay of the diet. However, for the region as a whole, cereals account for only 47% of total caloric consumption and a lower proportion of protein and fat consumption (FAO 1991). While roots and tubers account for a large share of caloric consumption in much of Central and Western Africa, items such as fish, meat, dairy products, fruits, vegetables and vegetable oils figure prominently in protein and fat consumption throughout much of SSA. Given their relatively high value, and given that only a small minority of producers are self-sufficient in their supply, cash outlays for the latter commodi- ties tend to be quite significant. 10 World Bank, 1993a, Table 31. 8 MARKETING AFRICA'S HIGH-VALUE FOODS TABLE 1.3: High-Value Foods in Total Expenditures1 Share of Share of Food in HVFin Share in Total Total HVFin Per %of HQ_usehold Household Food Capita Population Expenditure Expenditure Expenditures Country GNP($) Urban (%) (%) (%) Tanzania 100 34 64 32 50 Ethiopia 120 13 49 25 51 Madagascar 210 25 59 33 56 Kenya 340 24 38 22 58 Sierra Leone 210 33 56 34 61 Mali 280 20 57 35 61 Nigeria 340 36 48 30 63 Rwanda 270 8 29 19 66 Cote d'Ivoire 690 41 39 26 67 Benin 380 38 37 25 68 Senegal 720 39 49 34 69 Malawi 230 12 30 21 70 Cameroon 850 42 24 17 71 Mauritius 2410 41 24 17 71 Zimbabwe 650 28 40 31 78 Zambia 51 36 28 78 Simple Average 42 27 65 Weighted Average All SSA 350 29 1 Food expenditure data are from 1980 or 1985. Per capita GNP and urbanization rates are for 1991. Higher-value foods are defined here as foods other than cereals, roots and tubers. SOURCE: Data compiled from World Bank, World Development Report 1993. The available evidence suggests that expenditures on high-value foods are not confine4 to the relatively wealthy income groups in Africa and that even among the poorest groups, cash expenditures on high value foods as a group are as high or higher than those for staple foodstuffs. These patterns dem- onstrate that the efficiency of processing and marketing for high-value foods is important not only for middle or upper class consumers, as sometimes believed, but also for the relatively poor both as consumers and as producers. High value foods figure prominently within sub-Saharan Africa's agricul- tural trade. Although much attention has been given to Africa's growing ce- real imports and their adverse effect on national balance of payments posi- tions, cereals in fact account for less than one-third ofthe value of sub-Saharan Africa's food and beverage imports (Table 1.4). INTRODUCTION 9 In contrast, the high-value foods considered in this study account for more than one-half of such imports. Except in years of major drought over large parts of the continent, the value of imports of livestock products (that is, animals, meat and dairy products) alone approximate those for cereals. Although no single high-value food commodity is as important as either coffee or cocoa in Africa's food exports, as a group high-value foods register larger annual exports than either of the major beverage commodities. African exports of coffee and cocoa were valued at $2.4 and $1.9 billion respectively in 1988 whereas African exports of high value foods totaled $2.9 billion for that unexceptional year. Among high value foods, African exports are largest for fresh fruit and vegetables (especially citrus fruit, pineapples and bananas) and for various fish products (especially frozen shrimp and fish). Official ex- ports of livestock (and livestock products), oilseeds/oils, and nuts and spices are considerably lower, although unofficial exports of both livestock and oil~ seeds are probably quite substantial (see Table 1.5). The vast majority of African countries have markets or industries for these TABLE 1.4: Composition of sub-Saharan Africa's Food and Beverage Imports ($ Millions; 1988) Commodity (Group) Value Share of Total(%) Cereals 1754 32.1 High.Value Foods Fish + Products 700 12.8 Dairy+ Eggs 566 10.3 Meat + Products 533 9.7 Vegetable Oils 461 8.4 Live Animals 333 6.1 Vegetables/FruitJNuts1 222 4.1 Sub-total 2815 51.4 Other Sugar 272 5.0 Wine, Beer, Tea 262 4.8 Other ltems2 368 6.7 Sub-Total 902 16.5 Total 5471 100.0 1 Includes fruit and vegetable juices, but not other processed products. 2 Including roots/tubers, legumes, spices, canned fruit/vegetables, and other items. SOURCES: USDA World Agricultural Statistics; FAO Fisheries Commodity Yearbook. 10 MARKETING AFRICA'S HIGH-VALUE FOODS commodities. Insights from private sector experiences in particular countries are thus potentially relevant over large parts of the region. Contrast this situation with that of tea, cocoa or tobacco where only a handful of countries feature significant export-oriented development and minimal local market development.U In addition, marketing and processing activities for most of the focal commodities have traditionally been dominated by the private sec- tor, whether on a formal or an informal level. This differs greatly from the common pattern for beverage crops which has been one of single-channel marketing systems dominated by parastatalsP TABLE 1.5: African Exports of High-Value Food Products (1988) Product Group Export Value($ Million) % of HVF Exports FruitNegetables 1217.0 41.6 Fresh 930.6 Processed 286.6 Fish Products 957.0 32.7 Shell Fish 481.2 Other 475.8 Livestock 313.2 10.7 Live Animals 163.7 Meat 138.0 Dairy 11.5 Oilseed/Oils 270.3 9.2 Oilseeds 112.3 Oils 158.0 Nuts/Spices 165.0 5.6 Nuts 58.6 Spices 106.4 Totals 2922.5 100.0 Items: Total African Food Exports . 8160.6 Coffee Exports 2436.3 Cocoa Exports 1883.5 SOURCES: TARS Data Base (World Bank). 11 Only five African countries had exports of tea or cocoa exceeding $10 million in 1989. Only four countries had similarly large export-oriented industries for tobacco. By contrast, in that same year there were thirteen countries with sizable (e.g.>$10 million) exports of fish products and nine coun- tries with sizeable horticultural product exports, with significant domestic market development in these commodities as well (data from FAO Trade Yearbook and FAO Fisheries Statistics). 12 Varangis et al. (1990) highlight recent structural changes in some of the region's beverage crop industries. INTRODUCTION 11 Due to certain market characteristics and other properties that these high- value foods share, the prospects are favorable for expanded fUture trade and value-adding activities by the private sector in Africa. First of all, these com- modity groups feature relatively high income elasticities of demand in com- parison with staple food cropsP Second, these commpdities offer greater potential for the development of domestic markets and of intra-regional trade than do most of the region's traditional major export commodities (especially beverage crops). Third, many of these commodity groups. have exhibited very favorable international mar- ket trends which contrast with the patterns for the major traditional exports (See Tables 1.6 and 1. 7). Fourth, each of these commodities offers very wide scope for new product development and value-adding activities, frequently without requiring the use of large-scale, sophisticated technologies. Most of the commodities that Africa has traditionally exported do not offer the scope for horizontal diversification that is possible with tomatoes, fish, groundnuts and the other items included in our focal commodity set. ENTERPRISE Focus: The private sector that is engaged in marketing and processing high-value food products is highly varied, ranging from individual traders to large-scale corporations. In a few industries, most activities are conducted within an 'informal sector' by individuals who are not registered or licensed. Much of this activity is unrecorded and undocumented. Although there are many empirical studies on African micro-enterprises in general, such as those done through USAID's GEMINP4 project, few of them examine systematically the activities of food traders and processors. Primary research on informal activity was not possible within the present study, with the par- tial exception of the case studies in Somalia and Ghana, and the subject is covered here mainly through a literature review. This study concentrates on formal sector processing and trading activities because these play a predominant role in the high quality domestic markets and intraregional and international trade. Formal sector enterprises have also been the focus of much government and international donor attention through programs to privatize public enterprises and promote private sector 13 Cross-country analyses by Sarma and Young (1985) and Islam (1990) found that among develop- ing countries, the income elasticities of demand for cereals, vegetables, meat, fruit and eggs were 0.16, 0.61, 0.63, 0.68 and 1.00, respectively. · 14 Growth and Equity through Microenterprise Investments and Institutions. 12 MARKETING AFRICA'S HIGH-VALUE FOODS TABLE 1.6: Comparative Export Values and Growth Rates (Export Values for 1988/89; Average Annual Growth Rates, 1980-1989) High-Value Food Commodities Cereals and Traditional Exports Developing Developing World Country World Country World Export Export World Export Export Trade($ Growth Growth Trade($ Growth Growth Commodity Billions) (%) (%) Commodity Billions) (%) (%) Fresh/Frozen Meat 23.8 5.1 3.9 Wheat 15.6 0.4 2.9 Fresh Fruit + Nuts 15.8 4.4 4.1 Sugar 11.3 -3.0 -3.3 Fresh Vegetables 15.8 6.8 4.0 Coffee 11.0 -2.1 -3.1 Fresh/Frozen Fish 11.6 9.6 8.5 Maize 9.1 -1.9 -6.4 Oilseeds (Soft) 9.8 0.8 4.8 Cotton 8.1 0.5 -0.3 Live Animals 7.3 6.3 12.2 Cocoa 4.4 -0.5 -0.8 Processed Fruit 7.2 7.2 8.9 Rice 4.1 0.2 2.1 Processed Fish 5.0 6.9 11.4 Tea 2.3 4.2 1.7 SOURCE: UNCTAD (1991) Handbook of International Trade and Development Statistics. TABLE 1.7: Recent Commodity Price Trends (Indices of Unit Values oflmports for OECD Countries) (1976--80 =100; Current Prices) Commodity 1976-1980 1981-1984 1985-1988 Beverage Crops/Sugar Cocoa 100 66 75 Coffee 100 79 87 Tea 100 107 110 Sugar 100 107 93 Meat/Fish Canned Meat 100 106 114 Beef 100 111 121 Shell Fish 100 115 137 Fresh Fish 100 112 147 Frud~egetablesnsuts Tomato 100 106 111 Oranges 100 106 126 Bananas 100 130 157 Groundnuts 100 123 104 Cashew Kernels 100 128 153 SOURCES: Data from World Bank (1993) Commodity Trade and Price Trends; Gill and Duffus Edible Nut Reports, various years. development. Analysis of the experience of formal sector high-value food pro- cessing and trading thus has direct public policy relevance. This formal sector features a diverse array of private operators from small, family-owned units through to medium- and large-scale domestic and foreign owned firms. INTRODUCTION 13 In contrast to much of the earlier work, the present study places private traders and processors in the center of its analysis and examines the func- tional and temporal relations structuring their actions, their relationship to public regulatory· and support agencies, and their perceptions of the environ- ment in which they operate. Relatively little attention is given to the primary production of raw materials, except in the context of the procurement strate- gies which bring these traders and processors into contact with farming, herd- ing and fishing communities. Methodology r The study combines a series of detailed commodity case studies with a broader review of private agribusiness experience within the sub-Saharan region. The case studies focus on sub-sectors where the private sector has long played a dominant role or where it has already started to respond to recent market reforms involving the decontrol of prices and trade. Some of the focal com- modity sub-sectors have been export-oriented; others have been oriented partly or primarily to the domestic market. The case studies are based on recent field work where the researchers conducted structured interviews with samples of private traders and processors and met with government officials, research institutes, banks and other local institutions.15 They were selected in order to provide a cross section of countries, commodities, types of private sector de- velopment experiences, and degrees of success or failure. Each case high- lights the empirical issues listed earlier, setting patterns within the historical context of the particular commodity sub-sector and the wider context of Afri- can trade in the focal commodity. The case studies presented in this study relate to the following sub- sectors: • Cashew Nuts in Tanzania • Dairy Products in Kenya • Vanilla in Madagascar 15 Three of the case studies (Tanzania cashews, Kenya dairy and Ghana fisheries) are based on field work conducted in 1992 or 1993 specifically for this study. Field work for the Kenyan horticul- tural case was undertaken in 1985/86, with updated information coming from a brief 1992 field visit. Field work for the Madagascar case study was undertaken in 1991 as part of a sectoral analysis for the World Bank. Field work for the Somalia livestock case was undertaken in 1987/88 as part of work funded by USAID. Studies completed previously were revised or completely redrafted to follow the conceptual approach and to focus on the main themes of this study. 14 MARKETING. AFRICA'S HIGH-VALUE FOODS • Fresh and Processed Horticultural Products in Kenya • Fish and Fish Products in Ghana • Livestock (Cattle) in Somalia The analysis seeks to relate the organization and performance of pri- vate sector processing and marketing to the distinctive techno-economic char- acteristics of the individual commodities. These characteristics-which re- late to the commodities themselves as well as to their production, processing and handling-are expected to influence the nature and significance of oper- ating risks, entry and transaction costs, and the skill requirements for those active in processing and trade. 16 Systematic relationships between the techno- economic characteristics of commodities and forms of organization in private marketing channels will be hypothesizedP In general terms, it will be pos- ited that the range of feasible institutional arrangements for commodities which pose inherent problems for quality control aud vertical coordination and which are associated with economies of scale in production and/or pro- cessing will be limited to vertically-integrated systems or contract-based sys- tems. In contrast, for commodities with less demanding techno-economic char- acteristics and lower investment requirements, decentralized, small-scale trading and processing operations could well be the efficient institutional norm. Similarly, where fixed investment costs and operating risks are very high, one would expect that private sector investment in Africa would be limited, at least for certain types offirms.18 This type of analysis offers insights relevant to policy formulation. It can point to a range of commodities where private sector response to market liberalization is likely to be rapid and broadly-based. It can also indicate commodities for which technical and economic entry barriers are likely to be high and where the expected private response will be slow and limited to a 16 Examples of techno-economic characteristics include: the rate of perishability and degree of qual- ity variability of the commodity, its value/bulk ratio, the timing of harvest, storage and transport characteristics, the specificity of processing equipment, economies of scale in processing, the fixed costs in plant and equipment, and packaging/display requirements. 17 Such analysis parallels the work of Binswanger and Rosensweig (1986) which relates properties of productive assets (e.g. land, trees, animals) to patterns of organization in agricultural production. It also corresponds to work by Williamson (1979, 1985) and others operating under the rubric of 'transaction cost economics' which seeks to link organizational or contractual arrangements with technical, competitive and other variables in the market environment. 18 For some commodities, foreign investment or joint venture investment and management ar- rangements may be necessary, at least in the short- to medium-term. INTRODUCTION 15 few larger firms. The need for and types of appropriate public sector regula- tory support will differ considerably between these two scenarios. In these ways, a coriunodity orientation can provide specificity and nuance to policies and programs promoting private sector agribusiness development in Africa. Complementing the case studies is a broader review of the existing litera- ture dealing with marketing and agro-industrial development in the focal commodities in Africa. Cross-cutting themes related to private sector charac- teristics (gender, ethnicity, nationality, size), raw material procurement (for example, contract farming), food processing (both artisanal and larger-scale), and marketing (marketing channels, credit, product diversification) are ex- plored. A variety of vignettes from throughout the region provide specific applications of the broad themes identified. Structure of the Study Chapter II presents the conceptual framework for the study. This framework positions private processing and marketing activity within wider commodity systems which integrate producers and consumers. The study examines both generic and commodity-specific barriers to entry and organizational effective- ness in processing and marketing and outlines a range of alternative institu- tional responses to such barriers. Emphasis is given to problems of risk and transaction costs in commodity systems and to the ways in which private actors can reduce or spread their risks and costs. This microeconomic per- spective is necessary to fully understand the constraints facing private ven- tures in Africa as well as the nature and effectiveness of private organiza- tional and technical responses. Chapter III provides a synthesis of regional experience in private sector processing and marketing of high-value foods. The chapter reviews adverse features of the trading environment facing private agents, characterizes types of private high-value food operators and examines major patterns in private sector raw material procurement, food processing and marketing operations. It also touches on recent experience in the privatization of public food pro- cessing enterprises. The chapter draws on an extended array of academic, consultancy and project documents in addition to the case studies in order to provide a region-wide perspective on the issues. Chapters IV through IX are case studies examining the development of and private sector role and performance in particular commodity systems. Chapter IV examines the development of Tanzania's cashew nut industry, 16 MARKETING AFRICA'S HIGH-VALUE FOODS emphasizing the nature a!!d impact of recent policy changes which have lib- eralized the market and permitted private firms to participate. The case study delineates the role of policies in the long-term decline of the industry and the confusing nature of the recent reform process that was designed to reverse this decline. It provides insights into the nature of private sector responses to policy reforms in a still uncertain policy and market environment. Chapter V examines the development of the private sector role in Kenya's dairy sub-sector. While smallholder dairy production in Kenya has been a success, marketing has been dominated by a government-controlled coopera- tive which has used the powers of the state and donor financing to margin- alize truly private and cooperative dairies. The case study examines the pre- carious position of these private and cooperative dairies and their prospects under a liberalized market. Chapter VI deals with Madagascar's vanilla sub-sector. It provides a case in which processing and trade functions are conducted exclusively by private firms and individuals within a highly controlled framework set by govern- ment. In the 1960s and early 1970s, the government and private sector col- laborated effectively to organize what had been a chaotic market. This pro- duced a boom in exports. Starting in the mid-1970s, the government has increased its control over management in the sub-sector. Excessive market controls and a short-sighted commercial policy have dulled producer and trader incentives and channeled private entrepreneurship into rent-seeking modes. This in turn has undercut the country's international competitiveness and former dominant market share. The case of the vanilla industry is illustra- tive of circumstances in Africa where private firms have been forced to serve as agents for the state. Chapter VII deals with the case of Kenyan horticulture, one of Africa's major success stories in export diversification. In sharp contrast to Kenya's other agricultural export industries, private firms and individuals have played a leading role in the horticultural industry throughout its sixty year history. The case study highlights the diverse forms which private enterprises take and their varied links with producers and overseas markets. Kenya's experi- ence offers important lessons for other African countries aspiring to diversify into the high-value horticultural trade. Chapter VIII examines the case of Ghana fisheries, one of West Africa's largest and most influential fishery sub-sectors. The case highlights the im- portance of artisanal fishing and processing activities for domestic markets INTRODUCTION 17 and the adverse effects which structural adjustment has had on both artisanal and industrial components of the sub-sector. Chapter IX deals with the case of cattle marketing in Southern Somalia prior to the breakdown of civil order in the country. The case study illustrates the importance of informal, kinship-based trading networks in Africa's long distance and cross-border trade in live animals, and the responsiveness of traders to wider economic and political instability. Chapter X offers further reflections on private sector processing and mar- keting activity in Africa, including suggested initiatives to facilitate its future development, which may be useful to policy-makers. PART A II. Transaction Costs, Risk and the Organization of Private Sector Food Commodity Systems Steven Jaffee T his chapter provides a conceptual framework for examining private sector processing and marketing of high-value foods in Africa. It provides both systems and micro-level perspectives on marketing. After reviewing several generic barriers to private sector entry and effectiveness which are specific to high-value foods, the chapter presents a range of technical and institutional measures that can be adopted to overcome such barriers. MARKETING HIGH-VALUE FOOD COMMODITIES: A SYSTEMS PERSPECTIVES A high-value food product represents the outcome of a sequential series of investments, activities and decisions. This process begins with the articula- tion of consumer demand and leads to decisions by farmers and fishermen to produce, raise or catch particular crops, animals and fish. It continues through a series of activities which produce and subsequently transform the crop or animal product in form, time and place to match consumer demand (Breimyer 1976). Food marketing is the physical and economic bridge which links raw ma- terial production and consumer food purchases. It involves a set of in- terdependent decisions, investments, institutions, resource flows and phy- sical and business activities (Kohls and Uhl 1985). As the bridge between 21 22 MARKETING AFRICA's HIGH-VALUE FOODS producers and consumers, food marketing must fulfill several functions simultaneously: a) stimulate and support raw material production-Food marketing plays a critical role in stimulating, orienting and facilitating raw ma- terial production at the farm level. This requires the communica- tion of information to farmers regarding what and when to produce, the provision of financial and other incentives to produce food items for sale, the facilitation of farmer access to those production resources (for example, credit and material inputs) needed to respond to such incentives and the reduction of transaction costs between producers and consumers. b) balance commodity supply and demand-Food marketing institutions provide the organizational framework for coordinating production and consumption. This framework balances the supply and demand for food raw materials and commodities, not only in quantitative terms, but also in terms of quality, time and place. This coordination entails logis- tical and informational tasks, transactions for current or future sup- plies, quality control measures and physical transformation of the raw materials/commodities themselves. c) stimulate demand and enhance consumer welfare-Food marketing promotes increased demand, consumption and consumer welfare by in- troducing new products, improving product quality, reducing consumer costs, making foods available on a more consistent basis and educating consumers on the merits and alternative uses of products. These tasks will require the dissemination of information, the development and ap- plication of processing and logistics technologies, and efficient mecha- nisms for the exchange of goods. The process of marketing high-value food products transcends several dif- ferent industries and markets and may cross international borders (Marion et al. 1986). In this process, the physical commodity can be conceived of as 'flowing' from one value-adding stage to another. Each of these stages is asso- ciated with a particular industry as conventionally defined (transport, food processing, packaging and retailing). The product gains value as its form is changed and it is graded, stored, packaged and transported in order to more closely match consumer demand. Other important 'flows' precede, accompany and follow upon the flow of the physical product. These include (i) informa- FOOD COMMODITY SYSTEMS 23 tion flows pertaining to products, markets and technologies between consum- ers and traders, producers and traders, and among both traders and produc- ers; (ii) financial flows to finance production and remunerate participants which are channeled backward from the consumer through the production chain and occur between financiers, producers and marketing entities; and (iii) ownership flows whereby the right or title to the product accompany its physical movement and transformation. Food marketing involves several essential activities. Some are physical activities, such as grading, handling, transporting, storing and numerous foTil).s of processing. Others are activities which relate specifically to the trans- fer· of ownership or possession of the commodities. These are the exchange functions of buying and selling. Still other activities facilitate these physical and exchange functions. They include risk bearing, financing, standardiza- tion of quantities and qualities, and information collection, interpretation and dissemination. There is a high degree of technical and economic interdependence among many of these activities. Storage and transport functions are interconnected as what timing and modes of transport are feasible depend upon the scope and methods for storing a product or raw material. Sorting and grading are of little value if the subsequent storage and transport of a crop do not maintain its quality. Efficient processing operations require sufficient quantities and quality of raw materials, which in turn are at least partly a function of effec- tive input supply, production financing and raw material transport activities. In the marketing of high-value food products, many of the pertinent pro- duction, post-harvest and distribution activities require specialized technical or market knowledge, skills or assets and the presence of participants in particular locations. This suggests possible gains from a division of labor whereby many different individuals and organizations specialize in the per- formance of one or relatively few physical and business activities. Given the nature of food marketing, a strong degree of interdependence among these various individuals and organizations will remain, which must be reflected in the effective coordination of participant decisions and activities, whether through market, administrative or other means (Davis and Goldberg 1957). Once we recognize that production and food marketing activities are inter- dependent, that those individuals and organizations performing'these activi- ties are. themselves interdependent, and that all of these are linked through a network of exchange relations and additional coordinating mechanisms, it 24 MARKETING AFRICA'S HIGH-VALUE FOODS becomes useful to view them as elements of a comprehensive system (Arthur et al.l968). Faced with enormous problems in the conceptualization and em- pirical study of national food systems, analysts in agribusiness and agricul- tural marketing have focused their attention on individual commodity sys- tems\ defined by Marion et al. (t"986) as: "small economic systems ... incorporating an interdependent array of organizations, resources, laws and institutions involved in producing, processing and distributing an agricul- tural commodity." Commodity systems may involve the production, process- ing and marketing of a single commodity or a set of very closely related commodities (as in dairy product, poultry, oilseed or citrus fruit systems).2 Individual commodity systems exhibit widely different organizational char- acteristics, both within and among countries. Some commodity systems de- rive all of their raw materials from domestic sources, whereas others rely wholly or substantially on imported supplies. Some commodity systems are- exclusively export-oriented, while others are focused entirely on the domestic market or service multiple markets. Some systems generate only a single or very few types of products for consumers, while others yield dozens or hun- dreds of products, some of which have little observable connection with the initial farm level raw materials. Most commodity system studies by agricultural economists and agribusi- ness specialists describe both horizontal and vertical structural elements. The former relate to entry and competitive conditions prevailing at each industry stage (for instance processing or retailing); the latter to the location/timing/ clustering of marketing functions, inter-stage differences in size, seasonality, etc., the number of parallel marketing channels, and the incidence and forms of contractual or ownership integration. Other institutional elements of the commodity system landscape are government programs which affect the 1 Agricultural economists have tended to use the term commodity sub-sectors. This is an inappro- priate term since there is nothing 'sub' about them, except that they include particular sub-compo- nents of agriculture. A more appropriate term would be trans-sector since the focal system cuts across several sectors, markets or industries. The term filiere, widely used in French language literature (with a literal meaning of'channel'), also conveys this characteristic well (see Griffon 1989). 2 As with industries or markets, commodity systems are conceptual artifacts. The borders between one commodity system and another may be quite hazy. This is especially true where highly processed foods are concerned. Many individual foods are the product of several commodity systems. For ex- ample, the manufacturer of prepared soups and ready-to-eat meals frequently draws upon raw mate- rials from the vegetable, grain, and meat sectors. Technological change enables greater substitution among agricultural raw materials, and this blurs even further the distinctions between traditional commodity systems. FOOD COMMODITY SYSTEMS 25 commodity's production and marketing,3 and institutions such as banks, auc- tions, trade associations and insurance companies which perform specific trade facilitation functions. In commodity system analysis, the central focus is on the problems and mechanisms for coordination (Goldberg 1968; Marion 1976). Coordination is a general problem of arranging interdependent activities which requires link- ing the decisions and actions of different technical and ownership units when collective or overlapping tasks are performed. Vertical coordination poses a major challenge to food systems. Vertical coordination is the process of har- monizing the decisions and actions of input suppliers, farmers, processors arid traders so as to match the supply and demand for food raw materials and products (in terms of quantity, quality, timing, and location) at the various value-adding stages (Mighell and Jones 1963).4 This process entails signifi- cant flows of information and other resources which define and shift incen- tives. It also entails the definition and redefinition of required, permissible and impermissible patterns of behavior for system participants. In a food marketing context, the absence of effective vertical coordination is likely to result in resource misallocation, technical inefficiency and greater production and marketing risks. 5 Closely related to coordination is the issue of food system control, namely, the ability to exercise influence over key variables in the environment of others. Conventional industrial organization analysis focuses on horizontal control or market power-the ability to influence prices, incomes and other results in particular markets as a result of one's large market share and/or product differentiation. Vertical control is also important. It is the right or ability to make strategic decisions which will influence the activities and wel- fare of participants at different stages in the commodity system. Such deci- sions may be associated with what is produced and by whom, how it is mar- keted and how returns and risks are distributed. As will be discussed below, 3 Studies by Goldberg (1974), Morissey (1974) and Marion et al. (1986) typify this type of analysis. 4 AB will be discussed below, there are certain techno-economic features ofthe commodities and of the relevant production and processing functions which render vertical coordination especially diffi- cult in food systems. 5 For example, where the quantity of a commodity supplied exceeds the quantities purchased, extra storage costs will be incurred or part of the crop will be wasted. Where the quantity falls short of that demanded, the commodity buyer may not be able to fully utilize his processing or other marketing facilities (thus raising unit costs) and may not be able to meet demand further 'downstream'. On these and ot~er inefficiencies and resource misallocations in food markets, see Lang (1977). - 26 MARKETING AFRICA'S HIGH-VALUE FOODS coordination and control within food marketing systems can be achieved through a wide range ofinstitutional measures through market, administra- tive or regulative means. MARKETING HIGH-VALUE FOOD COMMODITIES: A MICRO PERSPECTIVE Viewing the marketing process in this way, it is not difficult to see why many past 'marketing' interventions involving the construction or rehabilitation of physical infrastructure (for example, marketplaces, processing plants, storage facilities) have had only limited beneficial impact. The marketing process, and market development more generally, depends upon effective flows of information and financial resources, and efficient exchange relation- ships. Market development entails the development of exchange networks since it is within the context of (potential) trading relationships that individu- als are stimulated to invest in specialized production and marketing activi- ties. Exchange networks enable individuals to transfer needed information and resources and capture the benefits of a division of labor and functional specialization. To properly understand markets and market development, one needs to examine patterns of exchange among existing and potential market partici- pants. Each transaction between food system participants can be viewed as a process, entailing a series of stages from the time when potential transactors make initial contact until the time when the terms of the ensuing agreement have been carried out and enforced.6 Important stages in the exchange pro- cess are those of: -searching for exchange opportunities and partners, -screeirlng information about the products/parties one wishes to deal with, -bargaining over the terms of trade, -transferring the goods, services, titles, cash, etc., -monitoring the exchange to assess whether the agreed terms are com- plied with, -enforcingthe stipulated terms. 6 Both Commons (1934) and Coase (1960) discussed exchange as a process. FOOD COMMODITY SYSTEMS 27 The centrality of information within the process of exchange is evident from this formulation. Exchange does not occur without the collection, pro- cessing and transfer of information regarding preferences, products and trans- actor behavior. It involves a two-way transfer of goods, services and money as well as the transfer of property rights associated with these assets. These are the rights to control the use, alteration or further transfer of assets. Finally, it is evident that exchange is always a rule-bound process, whether by pri- vate or social restraints on decisions and ·actions. Transacting parties offer one another certain benefits, but these are conditional upon certain patterns of behavior which the parties either explicitly agree to or they assume under the prevailing system of laws, norms and conventions.7 There are several generic problems that are associated with information, ' property rights and conflict of interest between self-interested parties that are well-known to social scientists, regulators and business people. The ac- quisition and processing of information can be a very costly activity, requiring investments in personal time and in the purchase of materials containing market intelligence or technology. The actual cost and effectiveness of'search- ing' and 'screening' will depend upon a variety of factors including (i) the nature (for example, complexity) of the good, service or asset being consid- ered, (ii) the similarity of new information to established knowledge, (iii) the physical and cultural proximity to the sources of information, and (iv) the overall skills, assets and experience of the searcher.8 The second generic problem concerns information which may be unevenly distributed with different transactors having different levels and qualities of exchange-relevant information. This situation creates opportunities for one or both parties to pursue individual gains by misrepresenting their actual intentions, capabilities and products and by manipulating information in a strategic manner. This behavior has been referred to as 'opportunistic behav- ior' (Williamson 1975). Third, the property rights accorded to individuals may be incomplete in terms of either imperfect excludability or non-transferability. Imperfect ex- cludability arises when ownership is conferred upon a group rather than an individual, or when an individual asset owner is unable to detect or punish unauthorized use or alteration of the asset. Even where individuals do have 7 Leblebici (1985) stresses the importance of working rules for trade. 8 Arrow (1974); Nelson and Winter (1981). 28 MARKETING AFRICA'S HIGH-VALUE FOODS exclusive property rights and can in fact exclude others from asset use, they may be restricted or lack altogether the right or ability to transfer assets to others. This may stem from trade regulations or from physical properties of the assets which render their transfer highly risky. Fourth, even though there is some mutuality of interests in any trading relationship, at least some degree of conflict over the distribution of benefits, costs and risks will be inevitable. Conflicting interests may provoke a lengthy and costly bargaining process or prevent agents from arriving at a mutually satisfying agreement. Conflicts of interest will also create problems in ex- change monitoring and enforcement, typically discussed under the heading of agency problems. Self-interested parties may behave in ways which violate the letter, if not the spirit of agreements, and in so doing, undermine the interests of their transacting partner. With information asymmetry, the prin- cipal cannot detect whether the outcomes derive from the agent's opportun- ism or whether some exogenous factors were more important. Agent oppor- tunism reduces the principal's gains from exchange and requires the latter to make expenditures in monitoring the agent's behavior.9 Many of the institu- tions which emerge in commodity systems are geared toward modifying or constraining self-interested acts so that some balance is achieved in the coop- erative and competitive motivations of the participating parties. These generic problems of information, property rights and conflicts of interest can create direct barriers to mutually beneficial exchange and greatly increase the costs associated with trade. These lost opportunities and expen- ditures fall under the general heading of transaction costs. In a food market- ing setting, transaction costs are the whole array of costs associated with buying, selling and transferring ownership of goods and services. They are the information costs incurred in identifying and screening different trading opportunities, outlets and partners, the costs of negotiating trading agree- ments, the costs of actually transferring goods, services, and ownership rights, the costs of monitoring trade conditions to determine whether the agreed terms are complied with, and the costs of enforcing stipulated terms through legal, social or other means. Table 2.1 below distinguishes between different types of transaction costs, identifies their origins and indicates various tan- gible forms which these costs may take. Despite their considerable impor- 9 See, for example, Jensen and Meckling (1976). FOOD COMMODITY SYSTEMS 29 tance, transaction costs other than for transport, storage and handling have rarely been examined in studies of food marketing in developing countries. The vast majority of these costs either are not considered at all or are lumped together under the category of overhead costs. What factors influence the level of transaction costs in any particular trad- ing context? According to Williamson (1979, 1985), the level of transaction costs is a function of three main dimensions in the trading environment: 1) asset specificity: the extent to which physical and other assets required for production and exchange are durable and specialized for a particu- lar product or trading relationship; 2) uncertainty: the overall degree of uncertainty surrounding the exchange (pertaining to the availability and quality of products, the assurance of market outlets, the operating 'rules of the game'); 3) competitive market structure: particularly the number of alternative buyers and sellers. Transaction cost economics places greatest emphasis on asset specificity problems. For any particular production and trading operation, individuals may undertake general or specialized investments. Certain types of plant, equipment, materials and knowledge have potentially generalized use across a broad range of products or trades. Other assets are highly specialized and therefore have little or no alternative use or value should a production scheme fail or a trade relationship break down. Making investments in specialized assets thus exposes the investor to potentially severe bargaining and contrac- tual enforcement problems since the investor will be locked into particular activities and encounter pressure from trading partners to improve the terms of trade. Examples of asset-specificity in agriculture include tree crops with extended gestation periods, large-scale specialized processing and post-har- vest facilities and use of highly specialized production inputs and technical knowledge. The relationship between uncertainty and transaction costs is apparent in Table 2.1. Much of the effort and expenditure to gather information, struc- ture terms of trade and monitor behavior is done to counter various forms of uncertainty. Fewer alternative buyers or sellers may lower screening costs while increasing bargaining and enforcement costs. When there are relatively few alternative trading partners, one might expect (i) less complete disclosure of interests to trade and less disclosure of product information, (ii) better 30 MARKETING AFRICA'S HIGH-VALUE FOODS TABLE 2.1: Transaction Costs in a Commodity Trading Setting Type of Tangible Forms of Transaction Cost Source/Origin of Costs Transaction Cost Search Costs Lack of knowledge about opportunities (e.g. Personal/Personnel Time products, prices, demand, supply, trading Travel Expenses rights, market outlets) Communication Costs Screening Costs Uncertainty about the reliability of potential Consulting Service Fees suppliers/buyers Advertising/Promotion Costs Uncertainty about the actual quality of goods/ Costs of Credit Rating Checks services offered Bargaining Costs Conflicting objectives and interests of transacting parties Licensing Fees Uncertainty about the willingness of others to Insurance Premiums trade on certain terms Uncertainty over transactor rights and obligations ------- Transfer Costs Legal, extra-legal or physical constraints on Handling/Storage Costs the movement/transfer of goods Transport Costs Bribery and Corruption Expenses ------- Monitoring Costs Uncertainty about transactor compliance Auditing Fees with specified terms Product Inspection Charges Uncertainty about possible changes in the Investments in Measurement quality of goods and services Devices ------- Enforcement Costs Uncertainty about the level of damages/ Arbitration, Legal, Court injury to a transacting party arising from Fees contractual non-compliance Costs to Bring Social Problems in exacting penalties through Pressure bilateral arrangements or through use of third parties. opportunities for strategic bargaining, and (iii) more transaction enforcement problems since threats to terminate trade and deal with competitors will be less credible. BARRIERS TO PRIVATE SECTOR ENTRY AND COORDINATION IN FOOD COMMODITY SYSTEMS Food products, raw materials, production, marketing infrastructure and mar- keting services have intrinsic technical and economic properties which fre- quently cause severe problems that are related to risk, inadequate informa- tion, transaction costs, logistics and overall marketing costs. These occur FOOD COMMODITY SYSTEMS 31 frequently in developing countries. Each can constitute a major barrier to coordination in commodity systems. In this section, we refine our analysis by discussing problems intrinsic to food marketing-problems which will arise even in favorable policy environments. The discussion reviews some of the generic technical and economic charac- teristics of food products, commodity production, production support services and downstream marketing functions. It points out their possible implica- tions for the risks and transaction costs faced by producers and marketing enterprises. In the subsequent section, this analysis is applied specifically to a sample of high-value foods commonly produced in Africa. Food Product Technical Characteristics Compared with most other products, food products and raw materials are bulkier and more perishable. Bulky commodities have relatively low values per unit weight or volume. They generate physical handling and transport problems related to the development and utilization of infrastructure capaci- ties and are associated with potentially high unit logistical costs. For very bulky goods, it may be necessary to establish processing facilities (and atten- dant power and water supply) in close proximity to farm production areas. Perishability limits the period of time during which a product can be mar- keted as a fresh commodity and used as raw material in processing.10 It greatly limits the marketing flexibility of producers and increases their market risks and has the potential to place them in an unfavorable bargaining position vis-a-vis buyers who have alternative supply sources. Commodity perishability enhances risk of product loss or decline in value during transport and storage. It may necessitate investment in highly spe- cialized and 'lumpy' transport and storage facilities and equipment. It also limits .the role of storage in balancing supply and demand over time, and raises the risk of contamination in food processing. In addition to these losses and special costs, rapid perishability raises transaction costs because it re- quires raw materials to be repeatedly screened and graded for quality at each point in the commodity system. 11 10 The perishability of a crop or commodity stems from its own physiological properties, its stage of maturity at harvest, the storage and handling procedures used, and the post-harvest environmental conditions. 11 Selling highly perishable commodities 'on description' poses risks for both sellers and buyers since their quality may deteriorate in the interim period before actual delivery. 32 MARKETING AFRICA'S HIGH-VALUE FOODS While agricultural commodities are often regarded as relatively homoge- neous, food commodities and raw materials do exhibit considerable variabil- ity in quality from unit to unit and from one supply period to another. This stems from several factors including: (i) the wide range of different biological varieties or species of crops, livestock and fish; (ii) the strong influence of environmental conditions and changes on production; and (iii) variations in the production and post-harvest practices used. Food commodities and raw. materials tend to have multiple quality attributes, some of which are difficult to measure (or observe). Most of these attributes are valued and weighted differently by specific groups of users and consumers. These features limit the scope for informative grading, create information asymmetries related to quality and reduce the likelihood that market prices will signal complete information about the quality of these goods. 12 Food Commodity Production Characteristics The physical production of many food commodities and raw materials has features which render it inherently risky, heighten transaction costs in a market setting and inhibit effective coordination of production with down- stream operations and consumption. When compared to manufactured prod- ucts, food products tend to be produced over a more dispersed geographical area and by individual producers who are smaller in scale and less special- ized. This is because of the immobility of land, the absence of significant economics of scale in primary production, and the risks of specialized agricul- tural production. This production pattern may result in high costs for crop intelligence and transmission of information to producers regarding consumer preferences. It also contributes to potentially high transportation costs for the collection of raw materials or animals, thus interrupting physical commodity flows. The output from each individual producer may be insufficient for him/ her to invest in a transport vehicle (a 'lumpy' investment) or for a buyer to conduct a procurement run specifically for one or a few farms. The output of a small producer may be insufficient to warrant investment in proper storage facilities or standardized containers, and this may lead to additional han- 12 Since the quality of food directly affects human health, informational asymmetries are poten- tially quite serious. In the absence of labels which indicate the true contents of food products, their nutritional value and other health implications, consumer choice may be poorly informed (Caswell and Padberg 1992). FOOD COMMODITY SYSTEMS 33 dling activities and require additional quality inspection. All of these imply higher transaction costs. 13 At the same time, small, dispersed producers can face a situation of monopsonistic competition with only one or very few active buyers in their area. Often, the most efficient scale of operation at the farm level does not match that of subsequent processing operations. In this instance, hierarchical market structures featuring one relatively large processor and multiple small suppliers may emerge with asymmetric information and considerable inequal- ity of bargaining power. The mismatch in efficient operating scales serves as a barrier to forward integration among rm-organized producers and requires the processor to develop multiple supply sources to enable it.to utilize its full capacity. A coordination problem arises because production schedules for dif- ferent suppliers must be spread out over time. Otherwise, processors would operate at well below their full capacities or incur higher costs by purchasing raw materials during a short period of time and storing them in a raw or semi-processed condition. A second common set of food production characteristics concerns the yield lag, yield uncertainty and seasonality of production. The production of most food crops and animal products is dependent upon the life cycle of plants and animals. In some cases (for example, tree crops, beef cattle), this life cycle involves an extended gestation period before commercial yields are attained. This in turn creates a need for medium-term financing and presents a poten- tially considerable commercial risk for the producer. Agricultural production is inherently highly risky due to the important influence of weather and the possible incidence of plant diseases or pests. Adverse natural events or social incidents can undermine total supply or the supply from one geographical area, resulting in farmer losses, un(der)-utilized marketing and processing facilities and unmet consumer demand.14 The seasonality of crop and animal production creates problems for cost-efficient utilization of transport and pro- cessing facilities. For perishable commodities, processing requirements may make it necessary to extend planting and harvest activities into riskier peri- ods of production. 13 From the point of view of marketing agents, such added transaction costs are at least partly counterbalanced by the reduced risk of total supply failure (due to weather, disease, or pests) when production is geographically dispersed. 14 These unplanned variations in annual production levels may result in similarly wide swings in producer and consumer prices. If either producer incomes or consumer purchases are narrowly based, such price variations can have major welfare implications. 34 MARKETING AFRICA'S HIGH-VALUE FOODS Production Facilitation Functions Food marketing enterprises can play an important role in stimulating and directly supporting raw material production. Food marketing enterprises can support production in various ways, including the supply of market and tech- nical information, financing and material inputs (seeds, chicks, fertilizers). The ability of these enterprises to. capture the benefits from such serVices will be influenced by the nature of the services themselves as well as the competi- tive environment. The incentives for marketing enterprises to provide ser- vices will depend upon their ability to appropriate the benefits deriving from them, benefits such as increased output, enhanced product quality and out- put better timed to marketing or processing requirements.15 The dissemination of technical and market information has the proper- ties of a public good: such information is non-rival in its consumption, and it is very difficult to exclude individuals who benefit from the information with- out contributing to its cost. The marketing enterprise is unlikely to capture the full benefit from its supply of information since in a competitive environ- ment, producers can utilize the information and then sell their product to a competing buyer. Where such 'free-riding' is widespread, there will be little incentive for private firms to provide more than minimal market or technical information. The provision of information may also be associated with what have been termed 'moral hazard' problems: the directed message may be biased toward the particular needs of the buyer rather than properly informing the pro- ducer about the wider range of technical and\ market options. The provision of technical information and the direct supply of production inputs can also give rise to negative externalities as when the recommended practice (for example, heavy chemical use) adversely affects neighboring farmers or residents. Firms might benefit in the short term from supplying low quality seeds, breeding stock, etc. but those companies active in the purchasing and sale of the farm product would have an incentive to supply high quality inputs. In these cir- cumstances where private and social benefits and costs diverge, there may be a justification for government intervention. With respect to production financing, barriers arise due to limited collat- eral and asymmetric information. The producer is generally better informed 15 See Umali et al. (1992) and Jaffee and Srivastava (1992) for more detailed analysis of the incen- tives for private sector supply of selected agricultural inputs and services. FOOD COMMODITY SYSTEMS 35 than the marketing enterprise about his/her creditworthiness. The firm's abil- ity to recover the loan may be better in a non-competitive situation than in a competitive market, since in the former case producers will have little or no alternative market outlet. This enables the lender to deduct the loan amount from payments due for the commodity. Processing and Distribution Functions Several types of infrastructure, information and other resources needed for efficient food processing and distribution functions have characteristics which may inhibit private investment in specialized activities and weaken the com- petitiveness of a commodity system. For example, certain types of infrastruc- ture necessary for marketing have either properties of a public good or are subject to such great economies of scale as to result in natural monopolies in all but very large countries. Roads are an example of the former and rail and port facilities of the latter. Private firms engaged in food marketing will gen- erally lack the capacity to invest in such facilities, yet their absence (or poor quality) will reduce producer incentives, raise marketing costs and restrain trade. Although other types of marketing infrastructure do not exhibit economies of scale as significant as for rail and port facilities, they do nonetheless entail 'lumpy' investments which can serve as a major barrier to entry. Investments in modern processing, storage, transport and trading facilities can provide the investor, at least initially, with an operational capacity far in excess of current supplies or supplies expected within a few years. The large unit oper- ating costs in these initial years, together with uncertainty about future raw material and commodity supplies, may inhibit many private firms from un- dertaking such investments.16 These markets may be perceived as 'non-con- testable' by those not currently active.17 In the absence of a well functioning stock or financial market, there may be few local means to pool the risk associated with such investments. Hence, in certain cases, private firms may 16 The problem may be circular: farmers may not expand production to levels which would make the marketing investment cost-effective without first seeing tangible evidence of facility development. 17 According to Baumol et al. (1982), a 'perfectly contestable market' is one which is subject to potential entry by firms that have no disadvantage relative to incumbents. This is generally associ- ated with circumstances where sunk costs are not incurred and where entry is easily reversible. Pressures felt by the incumbent from unimpeded potential entrants are expected to impose efficiency on the incumbent. 36 MARKETING AFRICA'S HIGH-VALUE FOODS not be able to perform the necessary risk-bearing function. For some types of infrastructure, there are also possible negative externalities such as the envi- ronmental effects of marketplaces and slaughterhouses. Risks will be especially high for a new or highly specialized product which requires specialized processing, facilities or equipment. In the short run, while supplies of the targeted product are being built up, the specialized equipment may not be applicable to other currently available raw materials and com- modities. Under such a condition of asset specificity, the processor/storer/ transporter is locked into a certain type of operation and becomes vulnerable to bargaining pressure from raw material suppliers and product buyers who possess alternative production, trading or consumption options. The processing, transport and storage of raw materials and commodities takes time. The ability to participate in these activities depends on access to finance, both to pay for raw materials/commodities purchased and to cover interest and costs of goods held in storage or transit. Thus, the availability and cost of credit is an important factor in the entry and viability of firms and individuals in the marketing system. Many private financial institutions have limited experience in lending for agricultural marketing and processing. Lim- ited recognized collateral and information asymmetries are again a potential problem. Several additional processing and development functions are associated with economies of scale that may inhibit entry and therefore result in con- centrated market structures. Crop intelligence and market research feature economies of scale and scope due to certain 'lumpy' investments in assets (for instance, computers and databases), and the advantages that accrue to those with multiple, diverse sources of information and those trading in several commodities simultaneously. Product promotion is also associated with economies of scale or scope, and a certain threshold level of trade and supply capability is necessary for such promotion to be worthwhile. There are both large sunk costs and large commercial risks associated with launching a new product. These represent potential barriers to entry and sustained com- petitiveness. Additionally, product promotion may give rise to externalities, though this may depend on the specificity of the promotion. Promotion of generic products (as opposed to individual brands) presents 'free rider' problems since some producers/traders will benefit from the promotion without contributing to its costs. Related to this is the promotion of an overall industry or country within international markets. It too features economies of scale and externalities. FOOD COMMODITY SYSTEMS 37 For some product groups, especially fresh produce, there are potential advan- tages in promoting a national image for quality. This image or reputation is a public good: all firms in the industry are associated with it and new entrants into the trade inherit it. Yet, the reputation needs to be protected. The supply of substandard produce by one producer/exporter can jeopardize the reputa- tion of a whole commodity system with adverse effects on market access and realized prices. Measures will need to be adopted to internalize these positive or negative externalities. Economies of scale are also associated with international transport of food products since minimum quantities are needed to fill individual pallets and containers and utilize the full freight capacity of a ship, train or plane. A certain minimum level of trade on a consistent basis may be necessary to attract those who provide international freight services in the first place. The supply of such services tends toward oligopoly (air carriers, ship conference lines) or localized monopoly (train serV:ices), which raises potential bargain- ing problems when the export trade is fragmented among many small firms. In an international context, transaction costs alone may be a major bar- rier to trade for many individuals and firms. In setting up a trade, exporters are likely to incur higher search and bargaining costs than domestic market traders. Physical distance to the target market constrains access to informa- tion about trading opportunities, and thinner information networks yield less complete information regarding the capabilities and financial solvency of buy- ers and agents. Physical distance and language barriers may prevent face-to- face negotiations and contribute to an extended bargaining process. Interna- tional traders may be unfamiliar with the standard trading practices in the target country and this may contribute to misperceptions about respective bargaining positions and tactics. Large geographical distances between trad- ers increase logistical costs and risks and limit the scope for direct monitoring of trade partner performance. Enforcement of international contracts may prove difficult and costly due to weak legal integration of trading countries and to the more limited scope for social group or trade association to pressure contract defaulters. Table 2.2 summarizes the economic properties of major marketing infra- structure and functions. The table indicates that while relatively few of the facilities and services have properties of public goods, many are associated with externalities, economies of scale/scope and 'moral hazard' problems. To- gether wit)l the important problems of risk, transaction costs and logistics management, these features inhibit private investment in specialized produc- 38 MARKETING AFRICA'S HIGH-VALUE FOODS TABLE 2.2: Economic Properties of Infrastructure and Functions Associated with Food Marketing Facilities/ Public Good Economies of Moral Function Properties Externality Scale/Scope Hazard Overhead Infrastructure Roads x x Rail and Port Facilities x Large Marketplaces x Some Power and Water Services x Large Production Support Services Inputs Supply x Some x Production Finance x Technical Info Supply x x x Market Info Supply x x Post-Harvest Assessment I Transformation Crop/Production Intelligence Some Initial Grading/Selection x x Product Assembly Storage Some Quality Control x x Processing Varied Marketing and Distribution Local/International Transport Some/Large Wholesaling/Retailing Some Market Research/Intelligence Some Product Promotion Some x Standardization X. Some Country/Industry Promotion x Some tion and marketing activities and reduce the degree to which such invest- ments are privately and socially efficient. Table 2.3 summarizes the discus- sion in an alternative way by identifying possible gaps or bottlenecks in the flow of physical commodities, information and financial resources within com- modity systems that might stem from the inherent technical and economic characteristics of food products, production, processing and marketing. Risk and Transaction Costs in the Processing and Marketing of High-Value Foods The above discussion has been very general in nature, pointing out character- istics of commodities, production, infrastructure, processing and marketing activities that pose barriers to entry, create risks and increase transaction costs for commodity system participants. These features differ significantly FOOD COMMODITY SYSTEMS 39 TABLE 2.3: Generic Barriers to Commodity System Flows Physical Product Informational Financial Problem/Barrier Flows Flows Flows Commodity Characteristics Bulkiness x Perishability x Heterogeneity x Production (Support) Characteristics Geographical Dispersion x x Unstable Production x x Extended Gestation Period x x Public Goods Nature of Market/Technical Information x Information Asymmetry in Credit x Processing I Distribution Characteristics Scale Economies/Public Goods Nature of Transport Infrastructure x Scale Economies/Public Goods Nature of Communication Infrastructure x Asset Specificity in Processing x x Information Asymmetry in Credit x x Scale Economies in Crop Intelligence and Market Research x by commodity and also by the prevailing regulatory and competitive environ- ment. Here we discuss commodity-specific factors that inhibit investment and trade, leaving points about the African operating environment to Chapter III. In order to operationalize the insights of transaction cost economics in the study of high-value food processing and marketing, it is necessary to identify quantitative or proxy qualitative indicators for uncertainty and asset-speci- ficity, and then to evaluate different commodities according to these indica- tors. This will enable us to develop hypotheses later on regarding the ex- pected organization of privately-based commodity systems for high-value foods. Three dimensions of uncertainty and four dimensions of asset-specificity are explored here. All of these pertain to the linkage between production and marketing or to post-harvest and processing activities themselves, rather than to primary production. These indicators are: UNCERTAINTY: 1) degree or rate of perishability of the raw material in its harvested, caught, or slaughtered state, 40 MARKETING AFRICA'S HIGH-VALUE FOODS 2) specificity of quality standards required for the raw material or com- modity, 3) seasonal variability of raw material supply; AsSET-SPECIFICITY: 1) technical sophistication and equipment specialization in post-harvest activities, 18 2) level of fixed costs and scope for economies of scale in post-harvest activities, 3) technical sophistication and equipment specialization in processing operations, 4) level of fixed costs and scope for economies of scale in processing operations. Table 2.4 rates these variables for selected high-value food commodities and products. Ii is assumed that the traded commodity or processed product is destined for export markets or to a segment of the domestic market that is quality-conscious and "up-scale". This assumption biases upward the quality of raw material required as well as the degree of sophistication and costs of equipment involved in post-harvest and processing operations. Quantitative measures have been used to provide ratings of 'high', 'me- dium' and 'low'. For perishability, a high rating signifies that the raw mate- rial loses its value or ability to be processed in less than two weeks after harvest, slaughter or catch. A medium rating allows a period of two to four weeks before deterioration. A low rating signifies that the product keeps its quality for longer than one month if storage conditions are favorable. For seasonality of raw material supply, a high rating signifies that the raw material is available only during a very brief harvesting period (less than four months). A medium rating signifies that raw materials are produced virtually year-round with well defined seasonal patterns. A low rating signi- fies limited seasonal change in raw material supplies even in the absence of sophisticated production technologies. For levels of fixed costs (in plant and equipment) for post-harvest and processing operations, a high rating has been assigned where costs exceed $50,000, a medium rating for costs of between $10,000 and $50,000, and a low rating for costs below $10,000. 18 Such as milk cooling, animal slaughter, fish icing, fruit and vegetable packing and nut drying. TABLE 2.4: Techno-Economic Characteristics of Selected High-Value Foods Commodity Elements of Uncertainty Elements of Asset-Specificity Degree of technical sophistication Fixed cost Technical Perish- Quality Seasonal and equipment or scale sophistication ability specificity in variability of specialization economies of and equipment Fixed cost or of raw raw material raw material in post-har- post-harvest specialization scale economies material or commodity supply vest activities activities in processing in processing Milk (Pasteurized) High Med Med Low-Med Med Med Med Cheese/Yogurt High High Med Low-Med Med Med Med Beef (Canned) High Low-Med Med Low Med Low High Poultry (Frozen) High Med Low Low Med Med-High High Fish (Dried) High Low Med Low Low Low Low Fish (Frozen) High Med-High Med Low Low Low High Fish (Canned) High Med Med Low Low Low-Med High Banana Med Med Low Low-Med Med Na. Na. Mango Low Med High Low Low-Med Na. Na. Pineapple (Fresh) Med High Low Low Low Na. Na. (Canned) Med Med Low Low Low Med High French Beans (Fresh) High Med-High Med Low Low-Med Na. Na. (Canned) High High Med Low Low-Med Low Med-High Cashew Nuts Low Med High Low Low Med Low-Med Vanilla Beans High Med-High High Low Low High Low-Med Oil Palm High Med-High Med Low-Med Low-Med High High Na. Not applicable. SOURCES: Brown (1991); Hyinan (1990); Keddie and Cleghorn (1980); Manchester (1983); Case Studies. 42 MARKETING AFRICA'S HIGH-VALUE FOODS For the other indicators, qualitative measures have been used. For quality specificity, a high,rating signifies very exacting standards below which trade and processing are not profitable. A medium rating signifies significant premiums or discounts for high (or low) grades. A low rating signifies that quality standards are not tightly defined and simple minimum standards must be met. For the technical sophistication and equipment specialization, a high rat- ing implies highly specialized skills are required and/or that certain pieces of equipment are designed only for the processing of the raw material in ques- tion. A medium rating signifies that skilled workers with close supervision are needed and that major pieces of equipment have limited alternative food processing uses. A low rating signifies that unskilled workers with minimal supervision can perform the function and that major pieces of equipment have widespread alternative uses. As Table 2.4 indicates, if we use these definitions, most of the listed high- value ', foods are highly perishable in their raw state and require rapid sale to retain their value or rapid processing to prevent deterioration or contamina- tion. Few of these raw materials can remain with the primary producer for more than a few weeks after harvest (or slaughter or catch). Cashew nuts are a major exception since they can be stored for a year or more once properly dried. Moderate to high levels of quality-specificity also apply in all cases except to fish which will be dried. Hence, significant price premiums or dis- counts will be associated with effective inducements of high-quality raw material supplies (and the subsequent maintenance of quality levels). Only a few of the commodities exhibit a very narrow seasonal yielding pattern. Yet most experience at least some seasonal variation in supplies, suggesting possible pressures on or sporadic underutilization of transport, processing and other infrastructure. In general, moderate to high levels of uncertain- ty will tend to surround the procurement of raw materials related to high- value foods. Turning to asset specificity, the table shows generally low levels of techni- cal sophistication or equipment specialization required for post-harvest ac- tivities and generally low to moderate fixed costs involved in undertaking these activities. This suggests that the technical or economic barriers to entry by firms or individuals in such activities may not be significant, and fairly broad patterns of participation would be expected where market licensing and other rules permit. Food processing is generally less technically sophisti- cated than other manufacturing fields and there are a variety of different FOOD COMMODITY SYSTEMS 43 processing technologies for most food product groups. Nevertheless, for some food product lines various pieces of specialized equipment have been designed to consistently meet international standards for quality. The same applies to the many technical innovations related to materials handling, product devel- opment, and packaging which food processors require to achieve and retain international competitiveness. While most of these innovations are developed outside of the food processing industry (indicating the importance of local institutions to conduct or transfer results from basic and applied research), large food manufacturers may benefit from in-house research, especially in the realm of new product development. Such large firms may have a strong competitive advantage in countries where overall research capabilities in food technology are weak. Among high-value foods, higher stage processing operations more com- monly feature significant fixed cost elements (and economies of scale) than they do technical sophistication or specialization. This applies to meat and fish canning and freezing and certain types of vegetable oil extraction and fruit and vegetable processing. This feature increases the costs of entry and intensifies the·challenge of obtaining a reliable source of raw materials. llius- trative examples of fixed costs incurred in the start-lip of high-value food processing and trading operations are indicated in Table 2.5. TABLE 2.5: Indicative Fixed Costs for Starting Food Processing Operations (Costs of Buildings, Civil Works, Equipment, and Installation Services) Fixed Unit Description Costs($) Stall for Street Foods (e.g. kebabs, fritters) urban Senegal, 1984 15 Open-air Restaurant urban Senegal, 1984 90 Food Wholesaling Business urban Senegal, 1984 2,500 Small Fermented Milk Unit 500 liters/day capacity, rural Kenya, 1992 13,670 Pig Slaughterhouse 7500 headiyear capacity, urban Kenya, 1991 87,000 Small Meat Processing Unit 25 pig + 2 beef carcasses/week capacity, urban Kenya, 1991 190,000 Small Cashew Nut Factory 500 tons/year capacity, rural Tanzania, 1993 464,000 Small Fish/Shrimp Processing Facility 235 tons/year capacity, urban Benin, 1991 552,000 Medium-Scale Vegetable Processing/ Freezing Plant 6000 tons/year capacity, urban Cameroon, 1992 2,000,000 Large-Scale Fish Processing/ Freezing Facility 13,000 tons/year capacity, rural Uganda, 1992 3,900,000 SOURCES: Vafious Feasibility Studies; Also Bricas (1984). 44 MARKETING AFRICA'S HIGH-VALUE FOODS TECHNOLOGIES, INSTITUTIONS, AND OTHER SOLUTIONS TO GENERIC FOOD MARKETING PROBLEMS A wide range of technical and institutional measures can be adopted to facili- tate the flow of information, goods, money and product ownership rights in commodity systems. These same measures can also facilitate economies of scale, reduce risks and transaction costs, internalize externalities, and pro- vide those public goods necessary for efficient market development. Many of these measures are essentially market or quasi-market responses on the part of private firms and individuals, while others entail government intervention to stimulate, re-direct or supplement private activity. This section examines a representative sample of such measures, noting their potential for promoting improved efficiency and coordination. Technological Measures19 Technologies can be introduced at the production level which have strong marketing implications. For example, crop varieties have been developed which can be harvested earlier or over a longer period of time and whose harvested product has improved storage, taste and aesthetic properties. Cultivation prac- tices and chemicals can induce or delay crop maturity. Production under con- trolled irrigation, temperature, lighting and other conditions can influence the timing of production as well as the quality of raw material output. These and many other production technologies reduce production risks, enable pro- ducers to diversify their crop/livestock mix and facilitate closer coordination between production and subsequent processing or distribution activities. There is also a wide range of well-established technologies and new ones are emerging that facilitate the post-harvest flow of food products and raw materials by countering their perishability and bulkiness, thereby lowering the risks and costs of commodity storage and transport. These include con- trolled atmosphere storage and transport, advanced mechanical handling tech- 19 Only a brief discussion is provided here on the many production, processing, storage, transport, and other technologies which can be adopted to facilitate more efficient food marketing and commod- ity system coordination. For more detailed coverage, readers are referred to the Economic Develop- ment Institute (1990-91) working paper series on agricultural processing technologies (edited by Brown) and the papers by Greeley (1991), the Economic Commission for Europe (1991), and Peters (1992) dealing with post-harvest technologies, food processing machinery, and logistics management, respectively. FOOD COMMODITY SYSTEMS 45 niques, vacuum and polyethylene packaging, wax coating, irradiation and containerization in internal and international transport. Computerized ware- housing and computer monitoring of goods in transit can improve physical commodity and informational flows, plus provide early warning regarding product quality deterioration. Technologies can be introduced to more effectively measure the quality of raw materials before processing. They provide improved information and in- centive to producers. Uniform measures enable processors to better sort raw materials for production of different types of products. Examples of such tech- nologies include those which enable more accurate measurement of the fat content of meat, the color, sizing and chemical make-up of fruit and veg- etables, the kernel out-turn of raw cashew nuts and the moisture content of grains and raw sugar. Processing and related food technologies can facilitate improved physical product flows, counter the uncertainties and costs of raw material procure- ment and enhance. consumer demand. Many processing functions directly reduce the bulk and perishability of raw materials. The use of ultra-high temperature processing, together with aseptic packaging, can extend the mar- ket life of commodities which are ordinarily highly perishable (for example, 'long-life' milk). Technologies enable the replacement of natural raw materi- als with tailor-made fat-, sugar- and beverage-substitutes. These obviate the need to coordinate processing with farm production. New processing tech- nologies can promote greater uniformity of output and better hygiene and lower unit costs of production. Information technologies can improve monitoring of demand and commu- nications between commodity system participants. Laser scanners at retail locations are an example of the former. Scanning not only speeds up retailer inventory management, but provides detailed information about the buying patterns of consumers, information which can be very valuable both to the retailer and to food manufacturers. Improved communications via facsimile, computer networks and other electronic devices can lower the costs of doing business and keep suppliers and buyers in up-to-the-minute contact regard- ing available suppliers, delivery times and prices. The investment costs and the skills required to utilize each of the above technologies vary enormously, thereby influencing who will be in a position to adopt them and to gain a competitive advantage from their use. 46 MARKETING AFRICA'S HIGH-VALUE FOODS Laws, Rules, and Standards All commodity systems operate within an institutional environment consist- ing of a set of fundamental political, social and legal ground rules. These ground rules establish the basis for production, exchange and distribution. According to Ruttan and Hayami (1984, p. 204), these rules: . . . facilitate coordination among people by helping them form expectations which each person can reasonably hold in dealing with others . . . [They] provide assurance respecting the actions of others and give order and stabil- ity to expectations in the complex and uncertain world of economic relations. The institutional environment governing a commodity system consists of several different types of rules. The most important are (1) rules defining, allocating and enforcing property rights (for example, property and bank- ruptcy laws), and (2) rules and conventions defining permissible and non- permissible forms of cooperation and competition (for example, standards, licensing rules, laws of contract and liability, company,and cooperative laws and 'fair trading' conventions). A well articulated and consistently enforced system of property rights is a fundamental precoridition for efficient exchange within commodity systems. Clarity over ownership and the rights to use, trade and alter assets is vital to market development since this assigns to individuals the right to benefits (and the burden of losses) from specialized production and marketing activi- ties (Bromley 1986). The general system of property laws places boundaries on participant behavior and expectations, and in doing so, increases the scope for coordination. As we will see later in Chapter III, in many African coun- tries there is a lack of clarity regarding 'the rules of the game' for private trading activity which deters private investment in marketing and processing facilities where significant fixed costs must be incurred. Rules and conventions specifying entry conditions and boundaries on co- operative and competitive practices may also facilitate exchange and coordi- nation. The establishment and enforcement of standards can reduce transac- tion costs by increasing the available information to buyers and consumers.20 Standards, which may include basic weights, measures and quality grades, provide farmers and marketing agents with a more exact language to com- 20 On standards and grades, see Kindleberger (1983), Bowbrick (1983), and Rottenberg and Yandle (1988). FOOD COMMODITY SYSTEMS 47 municate offers, a norm to compare actual with expected behavior, and a more ~etailed and objective view of the actual outcome. Quality standards may bEJ mandatory or voluntary; they may be minimum standards or include multiple grades. They are especially important when trade takes place over long df tances and among strangers because standardized goods can more easily be traded 'on description'. The development of appropriate standards and gr~des is not an easy task, especially when the commodity has numerous valued [characteristics. Standards can also be abused when designed to serve as a m;jor barrier to entry for smaller, less well-capitalized firms. The licensing of producers and marketing agents can also assist trade by reduci g transaction costs. This would occur where the criteria for. licensing centers around the asset holdings, past experience, financial solvency and other Jroxies for competence of the enterprise in question. In this case, a public br collective authority provides the initial screening of the capabilities and cr~dibility of alternative suppliers and buyers. Performance incentives I are built in since suppliers of substandard produce would risk revocation of license[ Of course, if licensing is based on criteria other than capability and compe~ence (such as ethnicity or payment of bribes), then arrangements will provid~I misleading information and inefficient restraints on trader behavior. In addition, licensing increases the cost of entry and may generate economic rents if only a limited number of parties are permitted to trade. I As food products become more complex and varied, consumers cannot be expect~d to have full knowledge about the choices available to them and the qualit~ of products which are offered. The food supplier has the incentive to supplyljust enough information to differentiate its product from its competi- tors, b~t no more. In order to enable consumers to make m:ore informed deci- sions dnd to protect themselves from potentially harmful products and prac- tices, r~gulations may be established requiring tests or inspections of products, certairl handling or processing procedures, nutritional or other labels and truth-+-advertising. For such regulations to maintain credibility in the eyes of consumers and other buyers, those implementing the regulations must be i:nimutle to financial inducements to 'look the other way' if proffered.21 • Thj definition~d ""'°"""'"'' """"" "°"' of the above of rule. geneOrtiM 'pine~Pples, :: ugan~a. onf1ifimi :comt)ines lhgustrial ~ooq distri~ution :operationt; (for pefa). :. : ··J.eurn. prociuct~, :vehictes, tractors) with extensive agrib~si~es$ ~cuv1t1~~~ \The/ ·::.·~!~tter'jncl~.Qe;dai·~.:· spices~ 'frt~it, and oil.seed. productiqn; ;grain mming; ·tii~>:· i 0 ·.: ;.at'.l<;l;cqffee: pr0ce~$irig/p"acking, ,.and tis~ J~f:i5cessing ·~nd hxpoJtf '.IQ ·Mo4aY}F.' biqµe,:m~iti-sectoral. firm$ :play· major roles'••in certain agro-h:idustrfes;•:For in~.·.·· ·• ,Starlc~1 • ·On~··· flrfu ·has . siteabl~ .Jtlterests; lfr ::aUtorn6!),ile'distri6titiq~r Ca~bew'. . • rutproce~ing; vegeta?le. dil:prod\JytiOfl • • fore9try prQduc~?' apetcotton .•.. :..•.. · . ·..-m~,l~rge;physical 'as~~ts' Of$UChJlrrns,· 'toeJr w.eli~evetoge9 disir;~ution; : ph~nn~hii ah~ t~elr.•favo'r~ble' ~cces~·. ~o flna~.ce. (iJ)clu~ing .Qffshore~ finance): • .•.· · PrOVid.e mem:\'Vitll conside~able. advanta~e§. Q~er their :Competitors lri Jf)divi~: • ::. •~al rn~rket~: . . w ·' i • • • • • .. • :1: •' ; •. < ....:.;. •· . . . ; ; Despite macroeconomic reforms and adjustments in investment codes in many African countries, potential foreign investors continue to have negative images of and imperfect information on the business environment in Africa.33 The majority of the foreign companies now operating food processing and marketing subsidiaries in Africa have been located there since the 1960s or earlier. The experience within our case study industries as well as wider anecdotal evidence suggests that much of the new agribusiness investment over the past decade by foreign companies has been made by firms which have already been established for a long time within the countries where they are operating. Investments have been geared toward expanding existing capacity and trade or toward diversifying product lines away from commodi- ties facing adverse market trends. For example, Kenya has witnessed the 33 Rhee et al. (1993). 80 MARKETING AFRICA'S HIGH-VALUE FOODS diversification of foreign-owned tea, coffee and sisal companies into horticul- tural production and trade. In Ghana, several mining companies have in- vested in pineapple production and exports. Much of the incremental involvement of foreign companies in the process- ing and marketing of high-value foods in Africa has not occurred through conventional investments in subsidiaries. It has taken other forms including: (1) management and marketing contracts with existing public and private companies, (2) equipment supply, consultancy services and technical contracts, frequently under the aegis of donor-funded projects, and (3) joint ventures with parastatal or local private enterprises (see Box 3.2). These arrangements have enabled foreign firms to profit from their own technical experience, mar- keting expertise and linkages, while incurring considerably lower risk than that associated with a 'green field' direct investment. In many recent export- oriented ventures, European import and distribution firms have entered into partnerships with local entrepreneurs, assuming a minority equity stake and making arms-length arrangements for marketing. Ethnicity Ethnicity is a very visible and much discussed characteristic of actors in food marketing and processing. Private traders may be ethnically distinct from the mass of the population because they are: (a) settlers identified with the former colonial power, such as 'Europeans' in Kenya and Zimbabwe; (b) members of other non-African minorities: the Asians in East Africa and the Lebanese in West Africa are the best known examples.34 These communities occupied an intermediate position in the colonial economic and social structure between the Africans and the colonial power. (c) members of ethnic groups from other African countries: this category includes traders from the Sahelian countries active in the livestock trade in coastal West Africa, Yoruba in Ghana; and Mauritanians in Senegal until their forcible removal in 1989.35 34 Greeks in the Nile Valley and Central Africa, and Yemenis on the East African coast (see Le Guennec-Coppens 1989) are two other examples. 35 Associated with this category but not traders or processors in a strict sense, are Somali trans- porters elsewhere in East Africa, and Ghanaian and Senegalese migrant fishing communities else- where in West Africa. PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 81 · Box 3.2: CAMEROON: .YES, \\fE NOW HAVE BANANAS! . Jeetinlcal and. marketing contr.act~ .between· para$tatal ·agricUltura1· enterprises . alld .· foreigr1 capital .can unde·r .·certain .conditions'.: amount.· to. •the •.·virtu~I privatization• 'of. the . enterprise.· 'Tl1is •. has .happened ·With .. the· pro.ductiori of··· · . bananas .bY .the C~meroon . ·Development. Corporation cpoC), • Productio.n :on . . . · Cofrs J::kona estate had fallen to 10,00Qtor,is per yean~ 1£)87 from apeaK· of16;000. eight year~ earlier.. This was sympto~a.tic of the decline of bal)ana . production in the. whole cou~try due .t<> .the failure·· pf th~ $fate .~atketlf1g.•. board, the Office·Camer6unais de la Banane, to prganize expcirt opportJnities · .·.and to provide .effective.•crop protection ·services. As· a result; camet6pn was · at risk of losing its share .of:ihe French banana· matketi . · · . a The solution:adopted~y.CDC was:to.e11ter into technical and:marketing . ·· · contract •V1lththe De! MOrite Gro.up to cover the operatipri of.a new plantation· · · , . ovei.tt)e period 1987-'1994, ·Under the contract; Del M~nte agreed to buy the entire production, subjecuo qualify, for CDC's ·costprice FOB;. plus a bonus • . of ·Ct=A. 15/kg. ··Del Mont~ .also agreed to provicte .• an interest~free.· loan of $12•••. million, td be:. repaid by withtlol(jing pFA 7.E) per kg .. of· bananas. Del Monte; : was. to provide te'.cihhical 'a$sistance entirely at• its· 0wn ·post. ··its ipte~ests in .···· ·this contract were·.·to diversify 'its .own. sources ·of. prottuc;:tion and.: to· us.e ;• . . existing• agreements· between C~meroon ·a.nd . . Frarwe.· to .·penetrat~ Eurc)pean.··: markets> The contract enabled it to do. thiswith.out ~irect ·investment ~nd ·.· •Without. becoming ,involveq in.the ·bureaucratic problem ()f r~gisteri~g as a. Cameroonian. company arid acquiring land; By contracting to buy .frof1'l· a .ne~ planiPtion, jt also avoidect.imyinvolvE)rr,ientin r~~ucingtnewdrkforceatEkon~; .• . ..•·.•·· . t.11e compariy's.'Interest in reducing costs . quickly tran.sfornied )ts 'techni~ · .•• • 'gc:tl as5l5tanc~''.into.a.c9mplete nia11ag~mentzsef\lipe,JnclUding·~()talcontrol . Of l9a~ed .flJn~s, .tttough this WaS notforeseen. in the :contract.· 1.n .• particular, it ........ intrbduced. computerized, •control systems . and usect Cameroonian' private • firms, rather than. CDC,· f~r construction wotk; .·transport.and. niatlitenarice~ . The·cpntrac{has.beeh successfuJ .• in.··improving qljalify ~nd •red~Cingcosts of·.· Cameroo~ic:in banan~s. Tl1e· Canieroongovemment h'as·adopteda procedure . that avoids ••some of the political and administratiye ·problems. of Plltright privclti~ation,· but still sees• its snare ot t.he. profits sPQSitliziQg other .QQ.C. losses. Jthasb.een suggested that .similar cofitracts :e1sev,rhere co.uld. i.nvolve. · • the sale ••()f. assets ·tO the foreign. partner after a: cert~in p~rlod pf :$u(:qessful. .... . 82 MARKETING AFRICA'S HIGH-VALUE FOODS d) members of ethnic groups originating within the country concerned, but conspicuously successful in trading outside their home areas: the Bamileke in Cameroon, Chagga in Tanzania, Kikuyu in Kenya and Gurage in Ethiopia are all good examples. Members of the first three categories are defined by some sense of 'out- sider' status to the countries in which they live and trade, even though some or all the members of these communities may be born in those countries and be citizens of them. This differs from case to case, according to the history of the community and the politics of the host country. For instance, many Leba- nese in Senegal now hold Senegalese nationality exclusively, whereas no Leba- nese in Sierra Leone (except for those of mixed parentage) have Sierra Leonean nationality. 36 Nor does a long history of settlement and national status neces- sarily relate to how these groups are perceived by the majority population or by governments. Malawian Asians, for example, are generally thought of as Malawian citizens, but are still subject to specific legal restrictions on their economic activity. There is ample evidence that ethnic minorities are particularly active in many sector markets. In Lagos' main livestock market in 1991, for example, 41% of the traders were Hausa, either from Northern Nigeria or neighboring countries, and a further 49% were from other identifiably Moslem ethnic groups. 37 In the early 1980s Bamileke made up 50% of food traders in Yaounde's Mokolo market and 36% of an aggregate sample from the city's five major markets.38 The seventh chapter discusses the prominent role of Kenyan Asians in that country's fruit and vegetable export trade over the past thirty years. The continuing ability of Tanzanian Asians to find new economic opportunities is documented in Chapter IV. There can be no single explanation for the success of such diverse groups, and many of the explanations advanced by concerned politicians and dis- gruntled competitors are little more than prejudices. The success in trade of certain ethnic minorities can be related to their development of mutual credit and information sharing institutions (see Box 3.3), their capacity for innova- tion, their association with specific commodities, the conditions in their re- gion of origin, and resistance to extra-economic pressures. 36 Boumedouha (1990), van der Laan (1975). 37 Kulibaba (1991). 38 Dongmo (1983). PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 83 One of the most important advantages of belonging to an ethnic minority is the tendency of members to give each other mutual support through pool- ing market information, providing credit and informally contracting to secure supplies and markets. Moreover, these arrangements are enforceable through community pressure.39 Traders outside their home areas have an additional advantage in being able to resist extra-economic demands that a trader from the area would accede to: demands for loans-perhaps interest-fre~gifts, lower prices, etc. Yemeni retailers in northeast Sudan are a good example of this. 40 Ethnic minorities who have left their home regions are also likely to depart from their traditional economic and social patterns. They may be will- ing to experiment with new commodities, techniques and ways of doing busi- ness. The Lebanese who settled in Sierra Leone are a good example. They fulfilled a vital function by developing a new unwritten law of business prac- tice that was neither traditional nor colonial.41 A simple explanation for the concentration of some ethnic minorities in certain sub-sectors is that the commodities are produced by the same group, or that the group has particular handling or processing skills. This is espe- cially marked with livestock where people from pastoralist societies are most likely to have the skills needed for herding and handling. The success of Kenyan Asians in the trade of 'Asian vegetables' (for example, okra, bitter gourd and chilies) is not surprising given their detailed knowledge of the quality, storage and consumption properties of these vegetables. A common but by no means universal factor is that successful trading minorities come from agriculturally poor and/or land-hungry areas. Members of these groups were forced at some point to seek careers away from home and then graduated from petty trade into larger-scale trade. The Bamileke, for whom land-hunger is exacerbated by primogeniture, fit this paradigm well as do the Chagga in Tanzania and the Kikuyu in Kenya.42 Beyond these generalizations, there are further specific reasons for each minority's success: links to political power (for example, white settlers, Arabs 39 These characteristics are described for a long-distance trading network in Cohen's classic study of Hausa livestock traders in Ibadan (1969). 40 Morton (1989). 41 van der Laan (1975). 42 Brenner et al. (1990). The Gurage of Ethiopia present a variation on this theme. Their particular cultivation system based on the 'false-banana' could support high population densities, but did not produce any marketable commodities. Thus, they were pushed towards urban wage-labor, and from there, moved on to trade (Baker 1992). 84 MARKETING AFRICA'S HIGH-VALUE FOODS · !lox 3~3;: ~UDAN: A• TRADltlONAL MERCHANT CLASS Sodan•J ·ppsition. on· the. ~i\e: Valley trade ·ro~1:~ between •sub-Saharan AfriCa and the Mediterrcioean world, and its· peculiar colonial history, have led to the . continued ?oniinatio~ of a traditional merchant·c1ass associated with the ... . dominant ethnic group, the Arabs, and the . -relative absence.of··the public sector in agric.~ltural marneting. . . . · . . ·. ·. . · . ···:While long!distance. tracle was practiced in a variety ofJorms by the pre~ · ?Olonialst~tes. of Central and Western Sudan .and by nineteenthceentury Egyp-; · tian. ~no Eur9pean adventurers; the presEmt trading system .experienced its ¢ritlcal development d4ring the Con(jominium period .(1898-1956). As. rural pr(>doctiOn systems became 111ore integrated into the.cash economy, .and as · transpb~ :ac·ross the· huge and varied .country became easier, family trading Ut in the. form Of croi;;mortgage contracts· . •· t:;o~cealing nigh ·intere~t r'ates); . Primary loyalty,. ho~ever, ·was to.·the family ;C1nd it~:tradin& concern, reinforc;ed by. continuous 'person~! links to the home .vm~ge ·in the· North'. < . . . . ······. . ·• . um • · ·•.•.. • . . · •. .. ·•. .· •. . • . • . . . . ·····•The Condo,minium government· intervened re111arKably little· in agricultural mar~eting, thqugh it ·w~~; heavily involved i!1 production. of cottoni the 01ajor ·•· expqrt.crop. ·rile gum arabic~. livestock; anq oilseed trades were,. significant •·.·•.{Sudan is the' worldfa )eading prqciUcer .·of both gum arabic and sesame) and . .•. ' ·.·. wnolfy in. the ha?ds. btthe P~ivate sector... Through these trades, . . so01e •Su· . .·d~O~seL either:the111~elves,frqm jelfaba backgrounds,'. or taking advantage of .. PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 85 je/laba .collection networks, were' able . to e~tablish ·.themselves .amcmg. the country's· wealthiest men. Carnal-drawn sesame presse~: producing 9prude · . oil for refinin~.in Britain gave.very highreturn:s forlovv irivestrl'lenfarlci'.w~te·~;.:· key ·pathway;for agricµlturat traders to diversify into industcy. . ..··. . •.. . . . .. ··••· . lndependence:andthe early policies :of th~ Nimeirl regin,iein···196.9~1~71. led to .the • depprture of.·• many foreigrt .businesses, ·and jellaba bµsin~ss~s: · .· . . expanded and diversified to fill the gap; Jel/aba had in)~ctbeen;ctiyersitying · : ··int? agricultural·. prod1.mtion for decades,· but .were qu!ck to seize :ne~ state-: .. sponsored opportunities ·in mechanized· cotton and. sorghum: prociuctic)ti~ If· was also business of jellaba qrigin that invested in· and. be.car:m~ lic~nsed. :. . collecting agents for the new semi-public monopoly companies that ;were· .· · i'ouhdecl to contrC>I the gum arabk: and oilseed tr.ades; • . · . . .. · · '·····. ·..... ·. •·. While an impo~ant min()rity.of ·Stid~nese businessmen· ar~ of 'Middle-East~'·. ern or, Greek ()rigin, and while foreign; pa:rtic~larly Gulf,A.rab,capital :is ipvest~> i[lg. in. tne agricultural and agr()-industrial .•sectors; .. the. jellab~ traditt<.>11 'tiirid: ' jellaba networks are··· stillyitally importi:mt. Tra~ltiQt) is ~e~r(.jr( continuing.'.: .patt~ms Qf expansion .• and· V~rticaf integration through family and ethnic ties, . •• and· a. widespread preference, even· befor;e they .were imposed by law; for.· :financial ·pn1ctices that avoid: the. form· of fixed:n:ite int~rest. The: Su(lan~$~ .· . : private sector historically has seen•tc:irfewerrestrictionsthan)nmany:other•. · counMes, and even Where the.·state· does intervene, ~he ethnic groups:from. · · .. which the jellaba have sprung are those that: domim~tti the g()vernrnent. There · ·· is therefo(e an jmportant poli,ical dimension; domination of trading networks· ' byjellaba has helped fuel supan's explosiye ethritc tensions;.. . ·s~dRcEs:,Ma~ger.(1984); Mahmoo~.(1984). --;:; -~ ., ' ~ ., ' ·" ' . ' . . , in Sudan, Kikuyu in Kenya43 ) and links to export markets (whites and Asians in East Africa). During the colonial period, Asians and Lebanese using family labor penetrated to remote areas and performed a labor-intensive collection function which Europeans were not prepared to do. These minorities also had the trading skills that rural Africans did not at that time possess.44 The dominance in certain sectors of ethnic minorities, particularly those from outside the country in question, raises concerns and fears among gov- 43 Some governments may actively favor a business class that is 'foreign' and therefore politically powerless: the Asians in Kenya (under Moi) and the Lebanese in Senegal (Boumedouha 1990) were so encouraged. 44 Bryceson (1993), van der Laan (1975). 86 MARKETING AFRICA'S HIGH-VALUE FOODS ernments and local populations. The mutual trust and information-pooling within minorities are interpreted as collusion and price-fixing, and the open- ing of markets in remote areas as exploitation, though there is little evidence that ethnic minorities per se are more (or less) guilty of these charges than other traders. 45 At a national level, non-African minorities are often accused of illegal export of capital, yet again there is little evidence that they indulge in this more than indigenous elites. Governments may feel that non-African minorities do not cooperate with official pricing policies, but this may be be- cause the policies are inconsistent with prevailing conditions of supply and demand. Yet, even if these fears are unfounded, they may still have real effects on government policy. The whole drift to state-controlled trade in British East Africa is partially explicable by fears that Asians would dominate an open market46, and these fears are still an influence, largely unacknowledged, on government views of liberalization in East Africa. Some attempts at positive discrimination towards indigenous (that is, non-Asian) businesses have been made by East African governments (see Chapter VII), but these tend to favor political elites rather than genuine entrepreneurs.. Gender Throughout Africa, women are responsible for a great deal of informal sector activity in food marketing and processing. Given their predominant role in food production and in domestic food preparation, this is hardly surprising. Women's involvement is most visible in West Africa. In Ghana, trading is considered to be part of everyday life for 80% of urban women. The central market in Kumasi, the country's largest and the key to national food distribu- tion, has over 15,000 traders, 70% of whom are women.47 This is a phenom- enon repeated on a smaller scale across Ghana and West Africa.48 One impor- tant feature of West African women traders is the degree to which they are organized among themselves by hierarchies which culminate in the 'market 45 Bryceson (1993) gives the example of a colonial administrator in Tanganyika who considered food prices charged by Asians in a remote area as excessive and as evidence of collusion, but was unable to match them using government transport on a no-profit basis. Haaland (1984) provides useful counter-arguments to charges brought against Sudanesejellaba. 46 Bryceson (1993). 47 Clark (1988, 1992). 48 Dongmo (1983) found that 67% of traders on the Yaounde food markets were women. PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 87 queens' (see Chapter VIII). Women also have a virtual monopoly on a range of artisanal processing activities. Women do not dominate marketplace trade in the same way in Eastern and Southern Africa; their involvement in food marketing and processing is less visible and is assumed to be more related to recent economic and social trends. 49 It is nonetheless diverse and very significant, both in simple trade and in artisanal processing. Women traders assure supplies of camel milk for Mogadishu and are well represented in food markets in Zambia. As sellers of cooked offal, they are an important source of food for urban workers in Cape Town.50 In Tanzanian towns a large proportion of low-income and middle- income women engage in some cash-earning activity, many of which involve high value food commodities. A similar pattern is found in rural areas. A review of recent studies of micro-enterprises shows the importance of female ownership (Table 3.3). Women's activities in food marketing and processing vary according to the commodities available and cultural patterns, but several sectors stand out. For example, artisanal processing of fish-salting, sun-drying and smoking- is associated with women throughout Africa. In many areas, women process fish caught by male family members, but they may also buy, process and sell in their own right, and loan money to their husbands and other men for boats and equipment. Some may graduate to become wholesalers or transporters. Women are also quick to become involved in new activities of clam and oyster collection.51 TABLE 3.3: Ownership of Micro-Enterprises in Eastern and Southern Africa by Gender Gender of Proprietors (%) Botswana Kenya Lesotho Malawi S. Africa Swaziland Zimbabwe Female 75 43 73 46 62 84 67 Male 19 53 25 52 32 12 32 Mixed 6 4 2 3 6 4 2 SOURCE: Liedholm and Mead (1993). 49 Tovo (1991). 50 Herren (1990), Due and White-Jones (1989), Karaan and Myburgh (1992). 51 Randall (n.d.), Cormier-Salem (1992). 88 MARKETING AFRICA'S HIGH-VALUE FOODS The palm oil sector in West Africa is sharply separated into industrial and artisanal sub-sectors. Artisanal processing is done by women throughout the region. 52 Although manioc lies outside the definition of high-value foods adopted in this study, artisanal processing of this tuber, principally into the roasted granules known as gari in Ghana and Nigeria, is a major female activity across West Africa. Where mechanized (but still small-scale) tech- niques have been introduced, there has been some move into the sub-sector by men, but women are still dominant in every stage of processing and sell- ing. 53 The small-scale sale of milk and dairy products is a female activity in many areas, especially since managing the milk supply for the family's own consumption is a female role in many pastoral and agro-pastoral societies.54 In some countries, men are equally involved, particularly in itinerant, urban milk sales on donkey or bicycle. Women's participation in informal processing and trade depends largely on cultural traditions which vary throughout Africa-social restrictions on movement, the ability of women to keep separate budgets from their hus- bands, and the extension of domestic tasks. These economic activities have gained the importance they have because they serve the needs of society at large. One advantage offemale marketing networks is their flexibility. Women traders in Ghana cultivate contacts among producers in different areas, deal in small quantities and in different crops as the year progresses, and will move into processing quickly if the supply of fresh produce is interrupted.55 Such flexibility derives primarily from the long hours women are prepared to work for relatively low returns. 56 This in turn probably reflects their lack of access to more lucrative employment or investment opportunities. Women make up a much smaller proportion of entrepreneurs in the for- mal sector than in the informal sector. To some extent this is definitional: women's enterprises, even if large in scale and with substantial fixed assets, risk being labeled as informal because they are women's. There is also a shortage of information, other than anecdotal, on women in the formal sector (see Boxes 3.4 and 3.5). 52 Ay (1990), Devautour (1990). 63 NRI (1992). 54 Kerven (1987), Waters-Bayer (1985). 55 Clark (1992). See also Ali-Gaye et al. (1989). 56 Due and White-Jones (1989) show that market women in Zambia spend longer hours at their trade than farming women do in the fields, although their overall working day is shorter. PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 89 .· 1'hrou~houtf~e West ,t\fri<1ansavann(l, aaddawci (also known •as·.sdumb~ra) is . . regularly used; as ·a'pro~in-rich seasonjng :1n.~oui)s, .stews .C1QdporridgE;J.· Traditlonally . ·rnade •from locust beans;::in Nigeria:.daddaw? .·is.now dedvect · ·• . Prim~ri,ly fro111 soybeans ••wn.!Je soybean;pµ1uvation is tiups, •••inot~ers, thes.e.(lctiyities are, clone····· by .intJMdual~ alone.·.· Sgni~· daC[t!Gl"'a production .is•·:· f()r .11ofl1·e use; 75% or· · .. ·niore •is.•• sold ·camm0rc,1any•.·. Perloclic· l'tlark~t days .tiave been set •pP. at the : ··vm~ge• lev~f,·pj?·s·~alesan~ rriacte inthe rnaintowns of centr~F~iieria; From· there; 'dadcJaw13· i§ trc:i11sP:orted tci ti)~ main'towns i,t) th~ south arid. r19!1h ()f' . : the ·country and•·• even .as· far. as :Niger,·· C~rneroon· and.· Chad~·.• Even though:· .· £laddawa :proce~s.in~· and trCJde .p:ovid~$ an. importan~ squrc~ :ol irlconie. fbr several .thousamd~omenJn parts of K~ara. and Kadun~rStates;:thiscott~ge ihpustry is· largely Ul)r~c()rqed:i~ ·offi()iatstatistics< ,[ · · · · · :· The flourishing informal sector in certain West Africa countries provides a basis for some women to move on to more formal trade and processing. In Ghana, several women fish traders have become lorry owners, and a number now own one or more ocean going vessels with foreign captains.57 At least one industrial fish processing plant in Ghana is owned by a woman. Cameroon also has a relatively high proportion of women entrepreneurs, but those in the formal sector are concentrated in small- and medium-scale enterprises. As entrepreneurs, they gain only a small proportion of government loans targeted at that sector. 58 Several of Kenya's small, peri-urban dairy process- ing firms are owned and/or managed by women, including one that is consid- ered to be the most dynamic single company. 57 Randall (n.d.); Marsden (1990). 58 UNECA (1987). 90 MARKETING AFRICA'S HIGH-VALUE FOODS ·Th~l.Jhi~n:of booperati\f~$(lJGC)Js·(;l privat~ :~nte(prise; ·started ib· 1~80wito > ·. ~h~$.UPPQrt'. of .{oreign NG9s, ·tl:le. tlGC bougtii <:ind· r¢habilitate~l f<:irms.·and · : ·',faciilitle$ t?rrrier!yio~ned by, stat~ enterprises, ;1"11e UGC ~ow· man9ge$ its 6~n .· · f,eeq rnm; n~tqheryi ~rqiler br~ding ·ho~se~. ~rid eAuipmeniwork~hop; while:•· . .·•.• 'providin~·· input:' supply,: ext~nsion;•. ?na:)tiarketing •services ·to: :=iB.1 co()pe.r&· ; · ·tiv~$ .and son;<:>eooo ~mallhoider'.tarn;er~ !oc~ted intnegr~en 4obe.s.arounci · :: :.rviaf!lut9(: : ;: • ·• · ·· · · · · · ··· • · •,; ..•.·• }Ytii!e tne ·Vnibn ·~ •initiaLactiviti¢s· per,iter~(i around: pig• a~d ····l101ticurtu~a1 •... ·. · . •:Pro(fu¢tion: and:marketing, ;since :t~8$; ·poultry: pro~~qtioni .pr()ce~sing, ·~n(f ::: •::Fsal~~ :h(;lV~jJ.~~l~e·. Qaci{boOe. of, its. operatio~s: ·Union sai.es,of'Pf,ult~.irb.·•<' :"Crease1ea~ 1 •• 1~~~i,Q$7i·tne \9~ci£ffl.r~t!ye:.> ·: ;'. mo\(ert,lent ·pl~yeda major role ih t~e :soppl~ Of f()O~ 10 f\lla~utQ, more recetitl}'· ~'.. · •it tt.as·takC:Jn'the lead.in r~sponding to more, H~e,~al 111'.?r:lset p~l!C,ies 9n ··t~~· ~art' •~•r Of~ove'rf)meht and :tO. the i,rnproved·secutitysituation. ·.... . <, .< '. • > ',;' •. :•.'Tfle: µ§C is. le~ .bY. an····lt~lian ~tiest,. ~orki~g'fuU-tim~. on:~ yPl~~ta,Y ~~(~ls;. <·'.T •·.·. Its jiqara ·.qf dir~tor$ :an~ most of.i~a 960 workers: arewo01en.~sar~rn<)st: of: j >:~~~ . $m~1111~1cierswhom it . ·serve$. ine :OGQ's;recent:t1nan9i.al resu1ts·hc;h1~·····' . ": ·~€>en vefy .favijrable; Pr9fits h~ye,< been. ·used '·i() invest Jo. $Oci~F{irogr~tQS:~Jrn~c. : .· · clu~ing il>rn~$£>Jnranf nurseri~s att~p'1ed to. the ·co?per~tiVe.~ and. a·s~(;op9-01.:: '· .: • ;:~1.$c.n~()\~!Pr:·~.~B§. ~:~49~~ts~;,for~~r;9~~rn.~.o:t;f9.~J!i~~ .i~1 m~ .?toP~~~l~~,~~:;.; 1 · ?QORce~.~FA0(1~~1);1'w~~E:in·shooperatiVe.oeve1cipmentin MaJ)utoGre4nzooeP::.·. : · ··<;, ,. ·:. :. '~<:'. :;~': ; :=: ~; '., : :; \ Women entrepreneurs face many problems in surviving in the market and in expanding and formalizing their businesses. A problem frequently cited is the availability of credit. Women may have problems negotiating with credit agencies entirely staffed by men. They may be specifically prohibited from applyn'.ig for credit in their own right. Or they may not have the necessary collateral for loans, especially in societies where they cannot own land. There are donor-funded credit programs targeted at women both for the formal and informal sectors, but much work needs to be done before women in sectors such as food processing will have generalized access to sustainable credit institutions. PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 91 Another problem is the possibility, as mechanical technology is introduced, that women will lose control to men of their roles in processing. Men may also gain control of niche markets targeted at expatriates and the middle classes.59 A broader problem is overtly political, and it is a combination of the physi- cal vulnerability of women traders and a lack of understanding among deci- sion-makers of their role. There is a widespread assumption that simple trade is a parasitic activity. Yet, this ignores the risk-bearing, collection and de- bulking functions of petty traders. This assumption becomes more potent when held by men about women, particularly in societies where petty trade is primarily associated with women. Thus, there have been repeated calls by Cameroonian academics for the suppression of the female petty traders known as buy'em sell'em. 60 Such calls have been translated into action in Tanzania and Ghana. Militants of the first Rawlings regime in Ghana attacked women traders in 1979 and subsequently proved unable to negotiate respectfully with women's leaders, causing serious disruption to food supply all over the country. Such occurrences are only particularly extreme examples of the wide- spread incomprehension of and hostility towards women's trading roles. TRADER AND PROCESSOR RAW MATERIAL PROCUREMENT ARRANGEMENTS One of the major challenges facing private traders and processors lies in the procurement of raw materials in the quantity and quality required at the necessary time and at a cost that enables the firm to profit from its own value-adding services. Failure to properly manage this interaction with the primary productive sector will limit the ability of traders and processors to operate efficiently and develop markets.61 Failure to stimulate increased raw material production and to coordinate it with processing and trading requirements is a severe and widespread prob- lem in sub-Saharan Africa. It is one of the major reasons why most of the region's medium- to large-scale food processing facilities have operated at 59 As with Congolese vegetable traders (Ali-Gaye et al. 1989). 60 Dongmo (1983) makes such a proposal. 61 An evaluation of the experience of the International Finance Corporation in agricultural process- ing and storage emphasized the significance of raw material supply problems in undermining the performance of numerous ventures. Raw material availability, price, quality and unpredictability were all problems, especially in ventures relying on 'non-captive' sources, that is, where suppliers could sell to alternative buyers (IFC 1987). 92 MARKETING AFRICA 's HIGH-VALUE FOODS well below their rated capacity, and as a result, incurred relatively high unit processing costs. Many factors have contributed to this pattern. Some are not enterprise- or industry-specific, but simply part of the overall environment in which African agribusinesses commonly operate. These include the frequent incidence of drought, the impassability of many rural roads, the limited availability of bank finance for crop purchases, and restrictions on imports and land owner- ship. Others may be specific to a firm or sub-sector, as are price controls, crop movement restrictions, regulations requiring raw material purchases solely from parastatals, the non-availability of key inputs and the poor manage- ment of integrated production operations. Private trader and processor strategies to procure raw materials have taken several, non-mutually exclusive forms. One strategy has been to import raw materials rather than organize supplies locally. This approach predominates in the dairy industries in much of West Africa. A combination of ecology, geography and policy has rendered it easier and less costly to import pow- dered milk and butter oil for local reconstitution, rather than organize local raw milk collection systems. Overvalued currencies, the availability of milk powder through aid programs or at subsidized rates, donor financing of re- constitution plants, and the failure of programs to promote local milk delivery have all contributed to this pattern. Imports of fish, beef and oilseeds are also important for the food processing industries in many African countries, al- though to a lesser extent than for dairy industries. In Nigeria and Cote d'Ivoire, imports accounted for 35% and 25%, respectively, of the raw materials used in their food and beverage processing sectors during the mid-1980s.62 When firms are highly dependent upon imported raw materials, they face risks associated with sudden and prolonged bans or restrictions on imports, controls on access to foreign exchange and sharp currency movements. This happened in Nigeria during the 1980s. When it does, in order to maintain access to imported raw materials, frequently firms resort to smuggling and remunerating licensing and customs officials. Port congestion and other diffi- culties related to physical logistics are also common problems encountered by firms reliant upon imported raw materials. 62 Data for Nigeria are from a UNIDO study cited by Stevens (1990). Data for Cote d'Ivoire are from Ministrie de l'lndustrie studies cited by Riddell (1990). PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 93 The second common approach is to procure raw materials and commodi- ties locally through open market transactions direct with suppliers or at es- tablished market places. This might involve buying runs through production areas, purchases at village market places, or purchases from intermediaries who themselves are in contact with primary producers. While there may be some regularity in buying patterns or in the suppliers they deal with, traders and processors pay current market prices and have no involvement in the primary production process. This approach is commonly used by firms which trade on a part-time or highly seasonal basis, firms for whom raw material quality need not meet very high standards, and firms which operate on too small a scale to warrant making 'investments' in longer-term trading rela- tionships. It is also found among fruit and vegetable processing firms geared toward African domestic markets, which tend to buy during periods of sea- sonal glut (at low market prices) and then store raw materials in a semi- processed state for use over a longer period. A variant of market-based procurement is the trader's or processor's use of agents who are either based in the major production areas or have extensive experience and social ties within these areas. The agents act to recruit suppli- ers (for instance, farmers, fishermen, cooperatives), screen quality, bulk up supplies and manage physical logistics all the way through to delivery to the buyer. The latter is never in direct contact with primary producers. This type of arrangement operates in the export-oriented cattle marketing channels in Somalia and among private exporters in Tanzhnia's recently lib- eralized cashew nut market. In the first case, exporters rely upon a limited number of agents to procure animals from herders and small local traders and to arrange the trekking of those animals to a collection point or to the port for export. The agents partially pay the herders and then incur costs for labor, water, veterinary supplies, etc. during the trek, which they only re- cover later from the exporters. Without utilizing agents who have in-depth knowledge of producing areas and herders, exporters would be hard pressed to organize their shipments of large numbers of animals on a timely basis. Family and clan-based ties between exporters and agents are common. In the Tanzanian case, cashew nut exporters have relied upon local trad- ing firms to manage crop procurement in the main production areas and the logistics of preparing the crop for shipment. The exporter indicates the quan- tities needed and the price which the agents should pay to farmers or coop- eratives. Each agent deals exclusively with one exporter, although low grade 94 MARKETING AFRICA'S HIGH-VALUE FOODS nuts are sold by the agents to other buyers. In contrast with the Somalia pattern, these agents work on a commission basis. The exporters provide them with large sums of cash in order to pay farmers at the time of purchase. Again, it is not surprising to find that several of the agents are linked to exporters through family ties, or at least have had long-standing trading relationships for other commodities. Cash (in)security is likely to be an im- portant barrier to potential exporters who lack strong ties with traders within the region who are well equipped with transport and other facilities. Recurrent trading linkages, based on kinship, common origin, habit or simply shared interests are common between primary producers and traders. For example, in the Brazzaville vegetable trade, village-level wholesalers will regularly buy from two or three producers, usually their relatives, and then sell to an urban woman originating from the village. Such a choice of part- ners is natural within a village setting. The enhanced trust involved also counteracts the effects of distance. A special case of trade between kin occurs widely in artisanal fisheries where fishermen habitually sell to their wives who then process the fish. It should not be assumed that this is anything but strictly commercial-in these communities husbands and wives tend to keep their own budgets. In Ghana the system has the additional feature that wives (or other unrelated women traders) may claim a share of the catch against equity they hold in the fishing boat or unpaid loans for equipment. Recurrent farmer-trader market ties whereby particular traders are given the right of first access to or refusal of supplies are also commonly found in the fresh produce export trades of Kenya and Senegal.63 Another common strategy is procurement through formal contractual ties with producers. Contract farming is widespread in African. agro-industry. A mid-1980s review of the literature on African agriculture and agribusiness identified eighty-eight contract farming schemes in twenty-one countries. Most likely this underestimates its incidence since little information on small-scale, privately organized contracting schemes has been collected. In Kenya and Cameroon, major contract farming schemes have been developed for a half- dozen or more commodities that occupy a central role in these countries' agro- industrial sectors. In at least five African countries (Kenya, Cameroon, Burkina Faso, Senegal and Mali), contract farming schemes involve the participation of over 75,000 smallholder farmers. Table 3.4 shows the incidence and selected characteristics of documented 63 Jaffee (1990); Horton (1987). PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 95 contract farming schemes in Africa as of the mid-l980s. Nearly two-thirds of the schemes have involved the focal commodities of this study, with schemes for fruits and vegetables being the most numerous. Contract farming has also been an important institutional framework in traditional agro-industries for sugar, cotton, tobacco and tea, and indeed, some of the largest individual schemes have been for such crops. The commodities which have commonly been procured through contract farming share certain technical and economic TABLE 3.4: Documented Contract Farming Schemes in Africa (Mid-1980s) #of #of Ave.# of Market Ownerships Patterns Commodity Schemes Countries ·Farmers Orientation (% of Schemes) Joint Private Public Venture Fresh Fruit/ Vegetables 25+ 6 <100 Mostly Export 90 5 5 Sugar 12 6 5000 Domestic/ Export 33 33 33 Processed Fruit/ Export/ Vegetables 9 7 3000" Domestic 33 22 45 Palm Oil 8 5 2000 Domestic/ Export 0 75 25 Cotton 7 7 64000b Domestic/ Export 0 57 43 Tea 7 5 26000 Export/ Domestic 28 72 0 Tobacco 6 6 5500 Domestic 17 33 50 Dairy 3 3 4000 Domestic 0 100 0 Rubber 2 2 1250 Export 0 50 50 Poultry 2 2 Na. Domestic 50 0 50 Rice 2 2 2000 Domestic 0 100 0 Spices 2 2 Na. Export 100 0 0 Oilseeds 2 1 Na. Domestic 50 0 50 Gari 1 1 141 Domestic 100 0 0 Totals 88 21 41 33 26 •Based on information on seven schemes only. •Based on information on four schemes only. Na. Not available. SOURCES: Watts et al. (1988); Jaffee (1987). 96 MARKETING AFRICA'S HIGH-VALUE FOODS characteristics that ordinarily contribute to high levels of risk for producers and processor/traders. These commodities are generally (a) highly perish- able in their raw form, (b) valuable per unit weight or volume, and (c) het- erogenous in quality. They also require specialized skills and inputs to pro- duce, have extended production cycles or gestation periods, and feature economies of scale in processing, though not in farm-level production. Under these circumstances, farmers are likely to require specialized services and they especially value market assurances. Processors and traders themselves seek greater security in raw material supplies and can appropriate signifi- cant benefits (for example, improved quality) from providing production sup- port services. In contrast, there have been relatively few contract farming schemes in Africa for livestock and dairy products. This differs from patterns in North America, Western Europe and Latin America where contracting of such com- modities is common. In the case of livestock products, the reason may stem from the still limited effective demand for high-quality meat and the limited availability of refrigeration and other technology to maintain the quality of high-priced meat. 64 Where contracting does occur, it commonly involves ho- tels, restaurants and upscale retailers or else processing companies which serve this type of establishment (Box 3.6). For dairy products, contracting has not occurred because of the widespread reliance on imported raw materials and the common presence of price and milk movement controls in industries based on local raw materials. Chapter V examines the experience of private dairies in Kenya. Until 1992, these firms were legally prevented from dealing directly with farmers. They had to rely upon supplies from the dominant parastatal firm, ad hoc extra-legal purchases from farmers or cooperatives, or the production from their own cows. With market liberalization, there are likely to be efforts by the dairies to contract supplies and to provide inputs and technical services to farmers. 65 Cases of formal contracting between fishermen (or fisherman groups) and 64 This is discussed by Kane (1987). 65 There are relatively few cases of contract farming for staple grains in Africa. Food grains lack most of the technical and market characteristics that would lead to a demand for and supply of market and production contracts. They have low value per unit weight or volume, their markets and prices are widely controlled, they do not require specialized inputs, have relatively short production cycles, and readily identifiable quality variations. Carney (1987) documents an interesting case in the Gambia. PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 97 : Box ·s.e:·.·. RA'l:IONING CONtRACTS IN AGLUTTEI> MARKET . .. :: PERI-URBAN.POULTRY.PRODUCTION .AND:TRADE IN SENEGAL •Folfowi6g t:l19~()· ban on imported poultry·•'br()ducts, a.· 1oc~1 ·poJ1try ·in~ostry / .· d?velop~d ~apidly in s~n.egat. T:h~ugh initially servi11g ~he smau/esld~rt ,Eur~.· .. . pean C1nd t.eb8,nese con;ununitles!· then'Jarketexpandetl tO:ll)ctude fh<:iiaig?r·• ·. . ~· ·middle,.class urbar .po·p~lation. ·By 19f3~. there:· .were so?1e 300:enter!}ri.se~ .••. ·using• imported ••chick9~ .• prpducing 11early 1QOO tons of meat ~nd •36o;o~::: •· •· dozen. eggs· per year. With the.. e11~ouragemenf Qf agricultlJral'offtciais,; ·l~rge .· .·.~· . numbers . . of; i;ldditional grower$ ·•·invested 1ln.·· .broiler proouction; · . ~~ing local. ": '. ····/unimprov~d•••stock,:.Theres!Jitwas a·Sev~re markeiglut)doil)j~ated t)y'.lqW' :. 9uality pr()duc~F .•. . •. .•. . . . · . .· ·· • . . · · .·. ..·.· · •. · · · • • · :" · . · :2: . · . . Produc~rs. sought mark?t secu,r~W through informell' :agre~IJ)e~ts :with' .buy1 ; . · er$:Which gave. th~Jfltter fin~t ;option ~ta price riegotiate~ at the·. ~irlt~ of saie>• .·· ·.In .thls··t)uyers; market;· any ~fo~ard.commitfnents were.a~price~ ·9eiow· $pet : · ··IJ)a'rket levels and· left buyers withdiscretionacyjudgyment'o~er qu~Uty.i . •. · ·•· . · .· .·.·• • ·.· : More ena~~lng w11tractual: ties have J>e€ln (tevelope·.•·•<;'.:m···:,~. "·i ,,, ., ·:.,,., . . ·· <• ;, ,'.Qyer'ttie past: uiit'uP' ~ ·~ucf th~ir cS(tctl '.:f/P~ Rff-$Hdre.!~pqtj~ ijfjpr~ :.: ; •· . t.·•· . . .• . .•·· ' agr~ed. 'prices;• •1q . . . . . Pi~m~n~;ihi$:supply :~nd\t(l ~xten~ its ·protj.~ct. rfl1x: t~. · · · .•• .\>···. ~~~~~~i~~tt~.~~~'~°:M~t:~J:. .••• :.·~aS,'.·s~bseqveritJy:~pr~~ci.1is.~et.VfidefbY.~nt~rir~;into.:a]qint.•~~ntµre'.shr1111~····:.·. ·.· <'i:>,roieci:J~.:.~.e~in,: io~xliutpiJ,t.,pf: ~hiqh; jwn( b~ roarl(ete~t;thfpl!gb;.;~he'; firm'..~., .> • • e,~1~tlhg eUropeari ,· . . .';i dfiahnels~ · .. '· ,· ·: ? ··. . ··. r ...· ,. ""..;.' ~ ,, . .. . '~= : F ..... : ., ,,; ·.. · 67 Jaffee (1987) discusses a case in western Kenya where more than 20,000 resource-poor farmers, 70% of which are women, have been contracted to grow french beans for subsequent processing and export to Europe. PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 99 As Table 3.4 indicates, the public sector has been heavily involved in con- tract farming in Africa, especially for traditional industrial crops. Many gov- ernment-initiated schemes have involved financial, managerial and technical ·assistance from international financial institutions, in particular the Com- monwealth Development Corporation and the World Bank. A large propor- tion of these schemes, including most of those involving sugar or tree crops and the construction of large processing facilities, have also featured the de- velopment of nucleus estates. This has been done in order to guarantee mini- mum throughput into the factories, undertake technical research and develop high-quality planting materials. Another common feature of government-ini- tiated contract farming schemes has been the participation of multinational corporations in technical and managerial capacities, and who have frequently taken on a minority shareholding. There are a great number of these hybrid institutions combining government ownership, donor financing and private management, and they belie the stark analytical dichotomy of the public and private sectors found in most of the literature on African agribusiness. Private sector mvolvement is widespread via joint ventures and manage- ment contracts. Only for horticultural crops and some minor, specialty com- modities (for example, spices, high-quality poultry) have there been many cases where private traders and processors have initiated and managed con- tracting schemes on their own. In examining the history of horticultural de- velopment of countries such as Kenya, Cote d'Ivoire, Senegal and Zimbabwe, one finds that most of the formal contracting schemes have been undertaken by the larger firms in the industry. More often than not, these latter are owned and managed by foreign companies or by non-African local minorities since, for the most part, smaller indigenous horticultural exporters have lacked the technical capacity and financial resources to organize significant contract farming schemes and to extend inputs and credit to farmers. The effectiveness of contract farming schemes has varied by country, sub- sector and venture, calling for caution in drawing generalized conclusions. Advocates o( contract farming have viewed it as a salvation for African agri- culture, which involves a "dynamic partnership [between] agribusiness and the small farmer."68 While contract farming has proven to be an effective institutional framework in a number of African agro-industries, it is not a viable model for many crops and types of farmers and firms. Where it has worked, contract farming has served to reduce the technical and market risks 68 Williams and Karen (1985). 100 MARKETING AFRICA'S HIGH-VALUE FOODS faced by farmers, processors and traders and to better coordinate their activi- ties in operating environments generally characterized by poorly functioning input and commodity markets and limited agricultural support services. Yet, many individual schemes, including a majority of those initiated and run by private firms, have either failed entirely or not been sustainable over time. In contrast with the (generally larger) public sector schemes, few pri- vate schemes have benefited from preferential access to land, finance and improved roads, protection against competitive imported goods, and, perhaps most importantly, protection against competing buyers of the contracted crop. As a result, contractual enforcement problems have been extremely common in private schemes, in turn resulting in the non-sustainability of most of them (see Box 3.8). Unlike the case of many state-run schemes-where the contractor has de jure or de facto monopsony powers and where alternative, artisanal processing is either discouraged or banned outright (the case with both Kenyan tea and palm oil from Cote d'Ivoire)-private contractors often face competition and experience the 'leakage' of the crop into alternative mar- keting channels. Competing firms can afford to make higher cash payments to farmers because they haven't invested in inputs, technical services, etc. The 'leakage' of the crop not only undercuts raw material supplies to the contractor, but also accentuates problems in the recovery of input loans. Le- gal remedies to contract default have been costly or near impossible to imple- ment since government officials have typically sided with farmers in dis- putes. Other frequent problems of private contractors include: (a) opportunism and inefficiency on the part of intermediaries (for instance, cooperatives), (b) poor quality in the technical assistance provided, (c) difficulty in gaining ac- cess to inputs, and (d) interference from government officials. · . :aoi·a.s;. BROKEN PROMISES~ CONTflACT :ENm::'ORCEM!NT \ PROBL~MS IN U~ANDA.ANI) KENYJf ·. .· . .. .. . . .. ..· ' ~ : ' . . .• . ~~gi~niJ1g.i,ri·:1~90,.$tie1i. ·ug~hda ·.4td..•b~gari. a·c~~p~i~n toi;enciour~e·.farmer ·. •;, ·.~ro,duqtioJ1'0f:drif!)d chili~s Which ~hell WOUid' thf!)n· ~xpprt to rograrns · · .: .te~hn.1ciu~$; Jtttien coiitracted: a local. qppperative to .supply· ~r agre~d quan~ ·. .:} i!ty ()f '.:ch]U~s, providi@g tt)e cooperative •wlih $ee~s. · tractors,: ploug11s, and ·• ·~. cr?~;pre-flnah~in~ to p~$s\ cm·. to its m;efl)b~rs~ V\fhile ;droµght wiped: oilt'the :·: Jir$t cf:op,., •Shell persisted and ,supplied: additiQnEiL inputs fOr renewed ·pl> bers·with seeds, technical .advice and .a·. commitment to purchase.·vario.us types.· pf vegetable~ •. . in· particular, quantities ·•at particular• prices•. o~eri aQ· e~.· ;tended period. For two years this schemefu~ctionecfrelatively welt with KHI::: .. expanding.·. its .trade and· injecting h;irge. arylounts .of n1oney. ·into; tt:l.e . Iopa! .··. ·economy.. $tiil,·farmers were·$pparEmtly;usingthe K~E qontratt.asa'safetY .... net, planting slleculatively; selling··some suppli~s·.to othertrader$, and PfO-·:· vidingthe .•reniaind~r tt>. ~HE. On ah individual comrlltldltyb(isis, :ttle Matut! s~ppHes did notmatc.h .KHE's requ.irements•.·.. · · .· · . . •. · · . ·..·. . . · · . · •· . ·ourtng the.third· year, the . •. scheme . tota11y.·c0Uaps~d;. A drought .·in other parts of the country had result.ed in·. a substantial· increas.e ·in matket priers .. · . In order. to• meet the.ir downstream .commitments, many exporters came to·•. the Matuu area and offered cash prices several· times those of ·KJiE'.g:. :c~n::.; · . tract pr.ic~,· even.though theY would sustain a. to~s of their ¢>fingnegafive lecSspns from its. experience:, KHE mo~ed tcr secure .• much c)fits com111()d.ity . requirement$ byleasing and managing it$ ow~ farmi;;: ·...•..· . .. so~Rc~: .. ~erl~hey (199~); faff~~ (1Q~;;, Kenya's experience with contract farming is the most extensive in Africa and illustrates the mixed record of private companies in organizing such schemes. Contract farming in Kenya dates to World War II when the colonial government organized a smallholder vegetable scheme to produce dehydrated vegetables for the Allied forces based in East Africa. Since then, formal con- 102 MARKETING AFRICA'S HIGH-VALUE FOODS tracting schemes have been developed by public, joint venture or fully private enterprises for a wide range of crops, including tea, tobacco, sugar, oilseeds, barley, seeds and a broad range of horticultural commodities. In 1991, nearly 350,000 smallholder farmers were registered in contract farming schemes. Assuming that about 10% of such smallholders produce more than one crop under contract, then some 315,000 smallholders, representing about 16% of Kenya's two million smallholder families, were growing crops under contract.69 Contract farming in Kenya has generally been most successful in generat- ing raw supplies of materials and commodities for processors and traders, in raising smallholder incomes and in raising farmer productivity. An illustra- tion of the latter is provided in Table 3.5 which compares long-term trends in yield patterns for three types of operations: (1) smallholders growing crops under contract, (2) smallholders growing other crops within the framework of general support services and marketing board procurement, and (3) large- scale, plantation-based agribusiness operations. The table makes visible the TABLE 3.5: Comparative Long-Term Yield Patterns in Kenyan Agriculture Average Annual Rate System Category Commodity Time Period of Change(%) Contract Farming" Sugar6 1966--85 6.7 Tea' 1971-86 2.7 Tobacco6 1966--85 13.8 French beansd 1982-86 37.9 General Support I Marketing Board• Cashew Nuts6 1961-84 -1.6 Coffee' 1971-86 2.1 Cotton' 1965-86 -1.3 Maize6 1966-87 1.3 Vertical-Integration Agribusiness PineappleC 1973-86 2.5 Sisal6 1961-84 2.7 Estate Teah 1971-82 5.6 • Smallholders only. • FAO Production Yearbook, various years. ' KTDA Annual Reports, various years. a Njoro Canners Ltd. Production Farmer Records. ' Coffee Board of Kenya, Annual Reports r Lele et al. (1989), Table 3. • Kenya Canners Ltd., personal communication. h Kenya Tea Board as reported in Lamb and Mueller (1982). 69 On the origins and development of contract farming in Kenya, see Jaffee (1990). PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 103 fact that over an extended period, smallholder productivity under contractual schemes has been far greater than when support services were limited to the general official system of extension and credit. Smallholders under contract have also matched or exceeded the productivity gains achieved by relatively high-input, large-scale agribusiness. The expansion and improved productiv- ity in smallholder tea, tobacco, sugar and french bean production has been directly linked to the dedicated support and supervisory system for these crops70 , while inefficient support systems and payment delays have been ma- jor factors in the weaker performance of smallholder coffee, cotton, maize and cashew nut production.71 Yet, success within the private sector is not universal and many contract- ing schemes have failed or proved to be unsustainable. Several programs run by private processors to encourage production of sunflower and sesame seed were unsustainable due to the 'leakage' of the contracted crop to competing firms or to artisanal processing activities. Jaffee (1994) has documented the rocky road of contract farming in Kenya's horticultural sub-sector, where pri- vate firms have experienced general problems related to competition, farmer and intermediary opportunism and maintenance of high quality standards. Even so, contract farming has proven to be an important basis for the initial development and subsequent expansion of trade and processing operations for many of the leading firms in that sub-sector. A final organizational mode for raw material procurement is own produc- tion, that is, vertical integration ofprimary production and processing and I or trade. In Africa, both complete and partial vertical integration have occurred. Partially integrated firms continue to rely upon 'non-captive' sources for part of their supplies or else sell part of their own crop, herd or catch to other traders or processors. Many agro-industrial ventures have combined nucleus estates and outgrower programs. Vertical integration in African agribusiness has taken three main forms. One form involves small farmers who _process or .directly retail their own production. The woman farmer who brings to market and sells her chickens, eggs and tomatoes undertakes vertical integration on a tiny scale. A second form of vertical integration involves medium-to-large-scale farmers (or farmer 70 See Lamb and Mueller (1982) for tea, Allen (1983) for sugar, Buch-Hansen and Kieler (1983) for tobacco, and Jaffee (1987, 1994) for french beans. The latter two are privately-initiated and -run schemes, while the major sugar schemes are privately-managed. 71 Schluter (1984); Ommeh (1984); Lele et al. (1989). 104 MARKETING AFRICA'S HIGH-VALUE FOODS cooperatives) who integrate their activities with processing and trade. By making investments in marketing and processing infrastructure, and by un- dertaking sales direct to retailers, consumers and overseas buyers, these farm- ers seek to add greater value to their produce and bypass intermediaries. During the colonial period, groups of European settler farmers in Kenya, Zimbabwe and Malawi took this route, frequently obtaining protection by government from competition from other firms. The third common form of vertical integration involves relatively large companies, which make simulta- neous investments in estates and processing infrastructure, or which inte- grate backward due to dissatisfaction with alternative raw material procure- ment arrangements (see Box 3.9). , . . kyny~·~ pork prodllct industiY is relative.ly sman; :accounting' for• 1¢ss than;1% ';: ·of agricultural poP.. Domestic• consumption has historically been limitecl by . . . high: prices relative to.those of beef, religious constraints (forthe local Mu~~ · lirn·pppulation}. and the genen:~I absence of a.traditional consumption pat- . tern> Howe:ver·,. tne 1n(josfry 11as strong links to the domestic touri.srn industry . •. i.• 'and)t\vas historic~Uy a significant earner pf foreign exchange througn exports. · · .. ·from .190G ·until the early 1980s, the pprk products in~ustry ;Was •. qomi~ . . ·.11at~p by. a single firm~Upland~ J3acon: Factory. Orlginally a p~ivate comp · ...•.• :·~ory brea~downs •.. . lJplands .Bacon Factory pig . slaughters.•fell. to4Q,OOO·· in · ·.... 19?9 an~t all·· the way. to .6000 ·by 1f)85. FacU1g a huge debt,. the com,pany .·•: wa~ li~uidated i'nt~$6 :and someiofits asset~ Were :auctioned off. ·. · .. ·.• . .·• · L. ·•••.: .TllEr l,Jpla,nds Bacon. Factory'was. never •tt)e sOle professor·.· of pork prod: .·. uct5;·although it lollg dpmin~ted the. 111arket. Sever~Lprivate firms had pro~ ····duced·pig meat; serying···foca1·markets or'pa.rticular market niches.·(such . a.s ·.• PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 105 tri ..··supplies the airlin~s}. It Was not un~il. tne l~te• 1~80s th~t a maj?r private·· .· ·company stepped·· in to···· fill the. vacµum created by the··. demise.·• of :Upli:m<;ls . ·• Babon~ the company Fatfoer's ·. c601ce.ltd ..was initlaily established. in 1975 .. inorderto supply fresh •meats and producetothe t)otels and restaurants.of its .parent company·. Block Hotels~ tater on it developed ·.a marketing an~ 'distribution system to. service the country's ·retail outlets catering to miadfo~ and upper~inc~m~ •. residents. ··1n•.1989 Farmer's ·cti6ice was.·:acquired by • the ·.local Sl!bsidiary of· Lon rho, providh1g . it, wi.th 'additional· resources.·.: ahd. with.··. closer links to Lonrho's local farming9perations. ·. . .· ...... · . · ··.. · ..... •The· pig .•. rneat·.·.processing . •. and marketing operations •of Farmer's Choice, . 'ha~e increa~e(j SllbstantiaUy sinpe ~he mid:198()s .The. number of pigs slaugh- :fi~red. by the. company increaseid from about .•·22,000 ·.iri •. f984 tp ·35;000. in a . ·1988 '.(when large,imodeirn processing facility was constrllcte.q) to an aver~ . . .age. of· more than 46,000 dµring the .1991:-1993 .·period •.. Much: of: the re9ent . · expansion . Jn proqessing .has bee.n fa9Hitated. by·t11e development of tlighlY; •· :efficient; larg&scale .pig-raising units on. farms. owned by the company and by·· .·.another Lomho affiliate,· In 1~92, .nearly tialf qf the .company's supply of.pigs · · \~as from affiliated farming operations: . · ·. · · · · Fanner!$ Ctioi9e currently aqcol1nt$ .·for about. 85% qf th? domestic pig meat mar!(et: .Approximately ~0% Of its s·ales ~re directed to hotels,. resta~- .· · .. rants, and ihstitutfons'. The corripany has also .made a considerable effort tp ; ·.promote con~umption.·.among.the .local foiddle-elast;.·Population, conducting: in-s.t()re promotions~. advertising through various mediai; an~. taking "mpQile ' . .·.; kitchens" into shopping and residential: areas to offer sami:;iles;, promote . . . . :pr~du~tS/ ~nd conduci;m~rket. research. A. successful export trade hCIS. alsq:• . .tieeni redeve1pp~d ••and is expected to,grow in. tl)e fut(.Jre-. The incidence of the last form of vertical integration is strongly influenced by government policies pertaining to access to land, land taxes, etc. Some- times, company access to land is restricted or is otherwise frowned upon politically. To circumvent this, firms have entered into joint ventures or tech- nical and management contracts with prominent landowners, at times guar- anteeing the latter a certain level of revenue or profit per hectare. There have also been cases where governments have leased or sold large tracts of land to companies as part of their encouragement of agro-industrial investment. Nigeria provides an example of this. During the mid-1980s, the govern- ment encouraged vertical integration by agro-industrial firms through the sale or lease of large land areas to them. It adopted an import licensing 106 MARKETING AFRICA'S HIGH-VALUE FOODS system which gave preference to processors who could demonstrate that they sought to increase local raw material supplies through the development of their own farms. Many of the country's largest conglomerates and food and beverage processors took this route, developing large farms for grains, veg- etables, oilseeds, soybean and dairy production. The Kenyan government has also supported vertical integration in certain industries (Box 3.10). More generally, vertical integration of primary production and processing and trade has been relatively common in Africa in oil palm processing and fish freezing operations and in export~oriented horticulture. Less common have been cases of vertical integration for tree nuts, oilseeds and meat pro- cessing. Among our case study industries, vertical integration has featured prominently only in certain segments of Kenya's horticultural industry, but there are indications that larger firms may take this route in the future for Tanzanian cashew nuts and Kenyan dairy products. Despite much-touted statistics giving a favorable impression of the pro- ductivity (yet not the resource use efficiency) of large-scale estates by com- parison with smallholdings, the actual record of vertical integration in Africa is quite mixed. For the most part, farmers have not been particularly adept at food processing or export marketing because they lack the skills, contacts and time to be effective (see Box 3.11 for an example). There are exceptions of course. 72 At the same time, many processors and traders who have tried to integrate production into their activities have experienced major problems with regard to land acquisition, farm management, labor, and input and spare part availability. Traditional plantation companies have generally fared bet- ter. Financial Strategies Lack of access to credit (and foreign exchange) is perceived as a constraint on private enterprise, both informal and formal. Without minimizing the need for reform and expansion of credit institutions, our discussion now turns to the strategies enterprises have developed for obtaining capital which may or may not be defined as credit. In the informal sector, the most common source of start-up capital is the entrepreneur's own resources saved from previous empioyment together with loans or gifts from relatives and inheritances. For many micro-enterprises in 72 For example, since the mid-1980s, several dozen (locally-owned) integrated farming/trading op- erations have built up a successful export trade in Ghana for fresh pineapples. PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 107 B,oxa.10: . VERTICAL IN.TE~RATION wltH' GOVERNM~NTSUPP.ORT: .... THE PINEAPPLE CANNING INDUSTRY.OF KENYA •.. · .~:nya's can~~d pine~~~le indust~· d~~e.s . .to t,he··lat~ 1940s v~hen 'two •firms •. 'built processing facilities and orgfinized raw material supplies from ·Eur()pea~ .. settler· farmers. As. such .raw· material . s'upplies were·• iqadequat~.: the ·firms · · · b~g~oJo buy the;surpfus Pineapple crop ofsfl'la.llh()ldertanners, then growi.ng fbr tile tocai fresh market.· By the mid~1950s, gove111ment •launched a pine~·• .· apple ·support prpgram. in•• prder to: expand. the industry and enHan9e the incomes .•. of smallholder ffirmers.·. Inputs •. technical adviOe. cr~dit, and .·m.ini~ . ·mum. pri~e guarantees•. INere ···prpvided. rhe. scheme. and ·the pmcessors en- .dol.)ntereo .• majortechnicaJ;.· 1ogistical;. and .competitive problems and' the.• irt7. dustry survived only because of government subsidies; .: . ·. ·. . ' . · · Afte; independence, ()ne, ()fthe prope~sors-~enya Canners~entered Jnto. · · a management an.d marketing 9ontract:with 'Del Monte'. With •.i:m invest~ent·•··.· .·!TlaQe to expand .. Kenya Cann.er's•• processing .• cClpacizy, .• the Kenya°- .• govern·· · ment offered to expand th.Er smallholder; productiOn prograrl1 an,d .to'purchase • and ieas~ wttie firm.an :e$tate of 5000. acre~ tor itto.·asses~·tne feasibilftf qf devel()ping .a nucleus .estate. . . . . . .• •. •. . . . . •· • ·. ·.... ·. . . .•. . . .·•. . . . .•. ....• · 'Ral,\I material $UPplY. problems persisted ill the suBsequent~o.years as .. a. · . i result of drought and the rapid expansion a~d better return~ to coffee in' tile main· pinea.Pple gro\Yin~ .area.s~· .·Faced ···with·.this sitOation; and haying. Del· ·M?nte1s pfiqr experie11qe in plant~tipn. agr'jc~lture, Kenya Canner~ prop.os~d .• 'that the furtner expansion :ohhe industrybe 6ased on ~state productio,n~ The ; government .:a~reed to a:proposeci pro~ram; pµr9hasing and teasiog tp Ke~ya: i :~apn~rs ~ 01.1tstandirig loan~. By: +974, Ke,my~ :Ca~n~rs' qpe:r~tion .Wfi~ •..• · ··ru11y·ver:tica11y·1ntegrat~Q.• ·.. ·.·. ·. . . • .. · · .•..... ··'. .•.·. ... . . . ..· . ·.• ·.· ..· Tlie tjevelopmeot of a large. . estate. by Kenyci C.anners and .the linkage·.· of: .. ··the Kerwan operatio11 into the ·international .marketing ahd d.istributiotl ~ystem ofOef' Monte were·key factors in the. huge expansion of theJhdustry,, Between. · .+S>74-<~hcj :t,977, canned pine~pple exports. increaseci·fivefqld, from?,6!S to···· .45.~2(: ton.s. ~ince then, Kenya's exports.li13ve generaUyremained )Nithihthe •••. . ·rarige. -0f 4?,ooqt?.• 50,o()~)tons/Ye;ar; placing. the •. country among the. fiv~.· .,. .· .: •. [argesi:.e~pprters W()rldwide al)~ .keeping Canped: pineapple ra~ked firstam(mg . .. • Kenya's ,·' . ' ,.. , ~ .manufact1ired .. ,. . . ;· ' '·' product . .. ., . .. exports: . . . ..._, . . . . 108 MARKETING AFRICA'S HIGH-VALUE FOODS the food sector, no fixed investment in the true sense is necessary, merely expenditure on the first stock. For itinerant micro-retailers in Dakar in 1984, this could be equivalent to as little as twelve hours work at the formal sector minimum wage. There are many informal credit institutions, most notably the rotating credit associations known as susus in Ghana, tontines in Francophone Africa, and sanaadig in Sudan. While these share a common form, their scale and content are very diverse. Some are purely social, others have welfare or in- surance rather than credit functions. Many are organizations for disciplined saving and at a certain scale can provide useful capital for informal economic activity. In some countries, rotating credit associations are large in. scale and may include formal-sector entrepreneurs as participants. In Cameroon there is a system of bidding for use of the funds that introduces an element of real Bo){ 3.~1: LEA~NIN~ BY DOING: ~ARQ UJCK IN UGA~DA ·Jn the: ffiid~1980~, a group of c;oIT)mercialJarrners •.and businessmen formed .· the N~angauzi and \fegetable Growers Association(NVGA). By.1989, th~ g~~up · ··sentsarnples offresh···_and<.iried·gingerto Europe and the_.Middle East:Ex- .···.•·ports ijuring .1~8~/90 \fvere sm~H••. yef the product qualify. was sufficient to . . .: ; ~~c.ure ·_ 1arger.ordt:irs. to. _the ·. pext year... The .NVGA . quickly ?rganized .an •.•. ·. outgroWer .·scl:lem~ ·(wit~ 500. farmers)··· and.·. borrowed ·a . t;iz~ble sum troma •· · commercial bank in order to modify aboffe~ryingfactory.forthe .Purpose qf dr}lin~ gihg~r. rhe factory modifiqations and additional investments ·were•-made withoutcon~>Ulting with gingerspecialists. . ·. · . . · . .. . . · ·... The fir,rnwas not in' close cp,ntact w!th ·its ou~rowws; most ?f ,whom h~d .· no'. prlor ~inger~rowfrig:exf)erience; The q~ality ofproductioO varied• and was . ·exacerbated btdelays in crop _collection and grading.. Factory. operations were .'·· also,.proble~atic .as the diesel powdered drier~;cpuld not lower the moisture . :conten~ .of the·• gingerto international •standar,d levels. As.·a·_·result iofthese qu~lity i:>robtems,, the NYGAls main foreign ~uyer deducted 60% from the · originally quoted price.· Now starved of cash, NVGA ceaseq its buying,ope.ra~ tic)ns •anct •·sought :to rene~otiate .• i.t~ · loans •. ·.This... being unsuccessful, ... NYGA · . was closed pown: . . .. . PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 109 interest. 73 A development from these in both Ghana and Cameroon are susu collectors or tontiniers, small operators who collect daily savings and return them at the end of the month. Again these are savings rather than credit systems, and clients accept a small negative interest rate (the collector's com- mission) for the convenience involved. In Ghana some of these operators have attempted to formalize their businesses as 'savings and loans companies' with mixed results. 74 In many countries there are money-lenders, either legal or illegal, who usual charge high interest rates.75 Less visible and more amorphous habits of savings and patterns of infor- mal credit may be just as significant. Susu collectors in Ghana were found to be important for market women and traders, but not for urban small or me- dium enterprises. Of 103 small and medium enterprises recently surveyed, none had used susus and only two had used susu collectors. 76 One other common form of credit consists of traders deferring payments to producers until they themselves have realized a profit, a widely accepted practice, especially in the livestock trade. In the formal sector this is referred to as 'trade credit' and is increasingly seen as an important part of the busi- ness strategy of the enterprise. Even near the formal end of the spectrum, credit from banks may be inaccessible, expensive or simply less convenient than trade credit. In a study of small and medium formal Ghanaian enter- prises, 71 % of investment in capital stock by enterprises in the food sector was financed from internal sources and only 17% from banks. Of the total debt portfolio of food enterprises, 31 % was with banks, 37% with suppliers, 15% with clients and 17% with family and friends. 77 Of the ten small to medium scale commercial cheese and yogurt producers in Kenya interviewed for this study, only three have utilized loans from commercial banks for in- vestments or working capital. Similarly, only a small minority of firms which have entered Tanzania's liberalized cashew nut market at the primary level 73 Henry et al. (1991). 74 Duggleby, Aryeetey and Steel (1992). 75 Indicative figures of 30% per month for short term loans, and 50-100% per year for long term loans are given by Duggelby, Aryeetey and Steel (1992). 76 Duggleby, Aryeetey and Steel (1992). 77 Cuevas et al. (1993). The sub-sample for the food sector is extremely small, but figures are comparable with a larger sample including other sectors (43 for source of capital, 33 for composition of debt portfolio). 110 MARKETING AFRICA'S HIGH-VALUE FOODS have relied upon (or been able to obtain) credit from commercial banks for crop purchasing. While access to formal credit will obviously be greater for larger-scale traders, the overall picture even for the formal sector is for a preponderance of informal credit in agricultural trade which is only now beginning to dimin- ish. Terpend (1992) gives thumbnail sketches of the credit situation in agri- cultural trade in six francophone coUn.tries. In Congo, despite an apparently active government-sponsored Credit Union in rural areas, the flow of credit is from producers to traders through differed payments.78 In Mali, village asso- ciations which engage in marketing can obtain credit from the Agricultural Bank, while large and small wholesalers organized into groups borrow from commercial banks .. The. bank loans are secured against stocks, which has necessitated a state-sponsored storage capacity. Retailers and village-level purchasers are not reached by either channel. In Burkina Faso up to 1990, government views on private agricultural trade were so negative that no formal credit was available. In Guinea-Bissau there was no private banking system until 1989, and the entire rice-cashew trade (cashews are grown in grain-deficit areas so cashew traders also deal in rice for consumption) was financed informally. Until 1986 debts could not even be repaid in cash. Traders are happy with the current modified informai system. Credit for domestic food marketing in Guinea has been almost entirely informal, apart from pilot projects of govern- ment credit institutions. Food importers could receive short-term loans from private banks. In Niger credit is advanced down the hierarchy of traders from the half-dozen large scale merchants to wholesalers, small wholesalers and village-level purchasers. A variety of donor-supported initiatives to remedy the lack of access to credit have been tried. These can be targeted at enterprises of particular scales or degrees of formality, at particular sub-sectors, or special target groups such as women. They can be highly concessional or approximate to varying degrees market rates and conditions on collateral. They can operate through special structures, government institutions or commercial banks.79 In Ghana, '-- 78 Terpend sees this as a source of abuse. Micro studies of informal domestic agricultural market- ing in Congo see it rather as part of the prevailing trust relations (Ma-Mfuka 1989). 79 The SME Credit Program in Ghana is a good example of the latter, see Duggleby, Aryeetey and Steel (1992). PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 111 there are presently at least ten projects or institutions involved or potentially involved in credit to small-scale agribusiness (see Box 3.12).80 In Malawi, there are at least two government institutions and seven donor programs as well as a multitude of NGO initiatives geared toward the supply of credit to small-scale, off-farm, rural enterprise. Despite these initiatives, a rece:nt sur- vey found that only 1.2% of Malawi's small enterprises have taken loans from formal credit institutions and projects.81 Informal means of obtaining foreign exchange are also prevalent. These include the practices of over- and tinder-invoicing, delaying the repatriation of foreign exchange, mislabeling of goods for export, and offering inducements to bank officials to disregard missing currency forms and other trade docu- mentation. Such measures are a response to overvalued exchange rates, strict controls on access to foreign exchange, expectations about future devalua- tions and overall uncertainty about one's future economic (and perhaps civil) rights. In countries where macroeconomic stability has been restored, where exporters have been permitted to retain at least part of their foreign ex- change earnings, and where auctions or other mechanisms for the legal pur- chase or sale of foreign exchange have been created, private firms have had significantly less need to employ informal mechanisms. FORMAL AND INFORMAL FOOD PROCESSING Food processing is the largest single component of the manufacturing sector of most African countries. The discussion here first focuses on formal me- dium-to-large-scale activities and then on artisanal and small-scale activities. These components feature quite different actors and results. It must again be noted that the statistical and wider information base for examining private processing activities in Africa is extremely poor. Most industrial censuses are now quite dated and few surveys make a distinction between public and pri- vate enterprises. 82 80 RIO (1993). 81 Daniels and Ngwira (1993). 82 Under the research program Africa Regional Program on Enterprise Development, which is coordinated by the World Bank, an effort is being made to systematize the available data on manufac- turing industries in selected countries and to gather survey data on private firms in a number of industries. 112 MARKETING AFRICA'S HIGH-VALUE FOODS . . . ·. .:at>x'3.-12: CRiDIT ~HA.NA•AN . ~INE_. ENtREs:tRENEOR~::cHoossyouR .. .. . . .... ·· $mall~s9~rerufat ~ntrepreneurs·.·can be offered: a great variety ·of Credif~ppor· ._ . .... tunitles_:and related schemes._··reflectlng _the·dlv~rse· and'qverlapping triterests,. .. ofdo.nbr~ ~~d .t~eltcllqs~nc6unt~rpat1:swithiri ·the government • • .• . . . :. • . . . : ... · _ .. · Thi$ is Jllustrated by (lhana's ri:lcentiexperjence~·Thefargest inltiat_ive is: : -· the ;-~Cin.k'of Gtiarya!~ ·fund ·for ·Small and: Medi.um •. -Enterpr_ise ·:. Developm~nt : · . _wnictr•has· .a $25• .mmion~redit ·line.·:f@nthe WOrld ~~nH/$OJ5 _milHon·J~:·•. ea.rriiark~d for enterprise$ with assets t>elo'JV· $t(),0()() ;. :J'hese fund.S are to be. used_indirectly i~ -mutual savin~s s.c~~rries_. that can tnen: serve. as collateral< < ;fQr.·.cornril~i:cial bank 1oa11s; Jhe"Worid •BankJs also involved in·tlie;~eveloP:;· ... :inent of sman~scale palm oit prtjc~ssing thro~gh the Agriculturaf ~eveloprf)ent •_ •: B~nk"and iecnnpserve .(an American NGO),. the strengthening ptrural. ban~s;' •. · ·.a ~r~dltline f()r small.·scale en:erprise.sthroµgt,i the·. National "Bo$rd. f()r_ $mall:• •.. · .• _sca.reJnaustries.(which ~ake~roan$ of,uptcr ~4t)OO}, and.training and credit·•.: · · throug(1Woments World Banking;-· ·· · · ·· ·. . · . _-._-_TheE.urope~n,_9ori1fl1unity'$ Micro pr0~ctsPr()gr~nilnc1u~~s sbm~. fur'lding> of market ihfrastnfoture :and coQperative agro!irldustrial enterprise,, It. is about ·_._._. . ··to ~egin loans to small _ and_mediu~ e~terprises underJtieGRATIS_sc.be,me.·· ·. ·. :J.iNDP funds _th~ :·_.Cehtr9l ·Regioh oevefbpme~t Prpgrarn witb ·c~edit; "training;·.-_····· . women's ~ups,thro~gti. tt{e :· : .''.'National Board-.forS(Tlalt SC?t_le ··_Industries·· and tWO hatiori~Jwp~en*s o_rgani~t .. .zat1C>ris:·.1FAD'.s sm~11hp1aer '.• cr~dit an~. Marketi11g ·Prolecti .irnprernent~ci; :tw' ._·; · the· Ministcy of:AgrictJfit!re,;_ the Ghana Foo~ [)istribµtion. Corpqration: ahd 'the •· >Associqtionote~ople:f(:ic Practical lif¢ Educatiof1~, Jnctuaes·-.a. credit-• cornpo.:.• . . nent fQrmar~eting acti~itles. ' < ' > . . ·. ·. . . · >.. . . ·. · . .·. . .· ....... ·.Th~ initiatives li$te<;t :abpve· inol_ude. only_ those active in '.rural areas Of two : .• 0 .· ~t·GQana·s· .• iaine· regions. Many•• pf th~se:d~tojects target" :.coope(ativ~styler .·. \ · • ~rPii~sratn~r.t~~n_.individu.al,eritrepren~ur~ .•-_whatis:less • c1_ear ·;s h.ow' entr~. :preneµr~· and. would-be ·.·entrepreneurs·•.• a6tt1aily :.make use: of. these ·diverse. · :ppp9yt:unities: ·· · · · · · ·· · · · · · · ·· · ·· PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 113 Medium-to-Large-Scale Processing After independence and up through the rnid-1980s, the policies and direct investments of African governments favored relatively large-scale industrial units in the food industries and elsewhere. The objectives of industrial policy were to generate employment in urban areas, save or earn foreign currency, and meet short-to-medium-term goals ·of financial viability. Most manufac- turing activities, including food and beverage manufacturing ventures, were geared toward import-substitution which was encouraged through high lev- els of nominal protection. In many countries, overvalued exchange rates en- hanced an anti-export bias. Government attention to food processing centered around the establishment and operation of parastatal enterprises. African governments received extensive support from multilateral and bi- lateral donor activities in this import-substitution, parastatal-led industrial- ization strategy. A World Bank review of its own experience in supporting agricultural processing ventures found that over the 1974-1985 period, parastatals were the focal institutions in 84% of its projects in East Africa and 66% of its projects in West Africa in which there were agricultural mar- keting/processing components.83 Parastatals were also the focus of most sup- port efforts in African food processing by the FAO and by bilateral agencies. Where new investments were supported, this typically involved turnkey factories based on imported plant and equipment. These factories were capi- tal-intensive and required well developed infrastructure, strong maintenance and repair capabilities, and skilled, labor to operate them on an efficient ba- sis. Precisely these resources and capabilities were lacking in many African countries. Nonetheless, initial investments were high. Other common fea- tures included the following: OVERSTAFFING AND INAPPROPRIATE STAFFING: Hiring and wage policies were determined by government and were subject to outside interference. The manual workforce was likely to be paid more than its counterparts in the private sector and traditional agriculture. The size of the workforce could be increased for political reasons, but reduced only with difficulty, and there was little flexibility to hire seasonal or part-time workers. Management was drawn from the ranks of generalist civil servants without commercial or industrial backgrounds and was subject to sudden transfer elsewhere. Appointments became a form of patronage. 83 World Bank (1990), p. 29. 114 MARKETING AFRICA'S HIGH-VALUE FOODS POOR OVERALL MANAGEMENT: Management was of low quality not only be- cause of the individuals recruited as managers, but also because of unneces- sary procedural rules (for example, cumbersome requirements to tender and restrictions on the purchase of second-hand equipment, requirements to buy from or sell to other parastatals, fixed prices, lack of flexibility in product development and confused objectives). Government protection meant that there was little pressure to rectify poor decisions. COSTLY PuRSUIT OF NON-COMMERCIAL OBJECTIVES: Some agro-industrial parastatals also had social or development functions with mandates to pro- vide a variety of services to rural populations. Several performed both these and their primary functions well,84 and the extra services if worth performing should not be regarded in the same light as commercial losses. There is, however, reasonable doubt as to whether agro-industrial enterprises were the most suitable organizations to fulfill these functions. As a result of these structural weaknesses, many of the investments were failures. They were sustained only through government subsidies, additional donor resources (for enterprise rehabilitation), and restrictions or outright bans on competitive activities by other firms. To some extent, private agro- industries of the same scale experienced similar problems, but there is direct evidence, particularly vivid for Cameroon, that private enterprises operating in the same sub-sector and in geographical proximity to parastatals outper- formed them by a considerable degree. 85 There is more general evidence that private firms have achieved higher levels of capacity utilization than have public enterprises. Table 3.6 provides a list of some of the many government and donor- supported high-value food processing enterprises which have experienced se- vere operational problems over the past decade. It can be argued that the failures of these firms have been far more costly than simply the low (if not negative) return on initial investment and additional funds subsequently in- jected for working capital and coverage of losses. The dominance of these firms in their respective industries crowded out private investment. At the same time, parastatal illiquidity contributed to a decline in produc- tion support and raw material procurement services and prevented the para- 84 For example SODECOTON, the Cameroon Cotton Development Organization, see Schiavo-Campo et al. (1983). 85 Grouitch (1992). PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 115 TABLE 3.6: lliustrative Cases of Problematic (Donor-Supported) Food Processing Parastatals Country Enterprise Activity Experience Kenya Kenya Cooperative Dairy Processing Expanded production, yet Creameries capacity under-utilization Large financial losses Restricted private competition Kenya Meat Meat Processing Capacity under-utilization Commission Loss of domestic + export markets Large financial losses Zambia Zamhort Fruit/Veg. Processing 'White Elephant' Inadequate raw materials Poor location · Lack of operating capital Dairy Produce Dairy Processing Weak procurement system Board Capacity under-utilization Tanzania Tanzania Cashew Cashew Nut Processing Extremely low capacity utilization Nut Board Negative processing margins Poor processing efficiency Poor product quality Decline of procurement system Gambia Gambia Produce Groundnut Processing Capacity under-utilization Marketing Board Negative processing margins Large financial losses Guinea Salguidia Pineapple Processing Capacity under-utilization Loss of markets Burkina Faso O.N.E.R.A. Meat Slaughter High unit costs Cold Storage Large subsidies req¢red SOURCES: FAO (1992); Holtzman (1990); World Bank (1990); Woodward et al. (1989); Case Studies. statals from increasing producer prices for raw materials. Parastatal ineffi- ciencies thus undercut the incentives for producers to develop the focal high- value commodities and may have presented an additional burden on any new private investor. Such an investor would have had to overcome the disillu- sioned attitude among producers which the parastatal nurtured during its waning years of operation. Among our case studies, these ·problems were borne out in the develop- ment of Tanzania's cashew nut industry and Kenya's dairy industry. In the former case, a multilateral donor supported a major government effort during the 1970s to build up local cashew processing facilities in order to add value to nuts formerly being exported in raw form. State-of-the-art, capital-inten~ 118 MARKETING AFRICA'S HIGH-VALUE FOODS prises sold have gone out of business, and that the objectives of increased efficiency and stimulus to private sectors have been attained in only a few cases. The liquidation of state owned enterprises has been more vigorously pursued, but even here benefits have been small, partly because many of the liquidated enterprises had closed their doors years ago, partly because claims by workers and creditors absorb much of the receipt from asset sales. (p. 19) The actual record of privatization in Africa is still unclear. Most studies on the subject merely provide a 'body count' indicating the number of liquidated or privatized enterprises, but offering little insight into the size and signifi- cance of such enterprises, the terms of their liquidation/privatization and the subsequent experience of the firm and its sub-sector. Some of the parastatals counted as 'dead' are merely firms included on some government list of privatization/liquidation candidates. Conversely, some 'dead' parastatals find their way back into public or semi-public ownership after failed privatizations. Too few detailed country and enterprise-level case studies exist to make any conclusive remarks. According to Shaikh et al. (1993), between 1986 and 1992 there were 287 privatizations (including partial sales) in Africa, two-thirds of which took place in only four countries-Mozambique, Nigeria, Guinea and Ghana. Nearly one-third of all privatizations in Africa appear to have been of small retail outlets in Mozambique and Guinea. While Nigeria's privatization effort has been relatively well organized (with sales of several insurance companies, cement plants and other companies), total receipts from these sales were less than 1% of the book value of state-owned assets. By comparison, Guinea's privatization program was much more poorly executed. Many of the enter- prises were sold on the basis of single offers after non-transparent valuation exercises, and a large majority of the industrial enterprises never resumed business or failed within a few years of privatization. Similarly in Niger, privatized enterprises for groundnut oil processing, cattle marketing and groundnut and cowpea marketing were liquidated shortly after sale.87 Several factors have slowed the progress of public enterprise divestiture in Africa, as described by Berg (1993), Grouitch (1992) and others. The first is the overall lack of public support for privatization. In Africa, privatization is widely viewed by intellectuals and others as an externally imposed process which is often directed by foreigners for the benefit of foreigners. This feeling 87 This experience is discussed by Berg (1993). PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 119 is re-enforced by the inclusion of privatization conditionality in donor support programs and by the frequency with which foreign investors do in fact ac- quire divested assets. The effects of divestiture on unemployment are also widely feared and can stimulate industrial unrest. Parastatal staff them- selves form a powerful vested interest against privatization. Both govern- ments and donors have done an inadequate job of convincing people that the benefits of privatization exceed the clear and immediate costs. Lack of trans- parency during negotiations and the undoubted tendency of some privatiz- ations to benefit political elites have made privatization's public image even more negative. The second factor that has slowed divestiture are problems of defining and valuing enterprises for transfer. Privatization can involve disaggregation of large enterprises and thus remedy diseconomies of scale or inappropriate mixes of activities. It may be difficult to compromise between the interests of potential buyers in acquiring well-defined profitable units and the interests of governments in not being left with the residual loss-making components. This is related to the question of valuation and whether governments remain liable for debts incurred by the enterprise during its public phase. If too many liabilities are left for governments, they will have little interest in pursuing privatization. The last major factor is a problem of finding willing and acceptable buyers. Given the poor performance to date of the industries and difficulties with their valuation, in many cases it is difficult to find willing purchasers. The problem is more acute if there is opposition to the acquisition of divested enterprises by foreign investors, by local non-African groups or even by par- ticularly active indigenous ethnic groups.BB Weaknesses in capital markets have exacerbated this problem, by limiting the scope for 'acceptable' buyers to raise investment funds and by limiting the scope for diverse parties to pool investments. The available information does not enable us to provide an in-depth analy- sis of privatization of high-value food processing and marketing enterprises. Limited data and information from a few countries provide the basis for only a preliminary generalization of patterns. ·A first, tentative generalization is that privatization has often benefited 88 A very particular case is when potential buyers are existing importers of a product, and thus suspected of a conflict of interest. This did not prevent the Malian tomato paste industry being sold to the leading importers (World Bank· 1993). 120 MARKETING AFRICA'S HIGH-VALUE FOODS foreign interests. In Mali, out of seven agro-industrial parastatals slated for the first and second phases of the privatization, it seems as if three will end up under the control of French or American interests.89 Numerous other cases are cited in the literature. Sixty percent of the Office Camerounais de la Banane was sold to French private interests in 1990. The assets of Stalpeche, a joint Togolese Libyan parastatal venture in the fish sector, were sold in 1989 to a private company with 74% foreign, mainly French, shareholding. Government-owned pineapple plantations in Benin have typically become joint ventures between local businessmen and large French fruit distributors. A U.K.-based company took over management and 40% ownership of the Zam- bian Poultry Processing Company. A South African-owned company has taken over management of Mozambique's poultry processing plant and also sells the products through its own retail outlets.90 In Uganda and Zaire some privatization has taken the form of restitution of confiscated assets to former nationals of Asian origin and former European settlers, respectively. While there may be serious political arguments against divestiture to for- eign buyers, as a trend it is in accordance with the economic argument pre- sented throughout this report. International companies, or local interests closely bound to them, are much more likely than indigenous business in- terests in most countries to be able to undertake the financial commitments necessary to rehabilitate medium-to-large-scale agro-industries and manage the risks and transaction costs associated with export-oriented agricultural trade. Some of the perceived problems of foreign ownership can be limited through joint ventures, with either the government or local entrepreneurs, or both, as local partners. Joint ventures in which the government retains some stake. seem particularly common in Francophone Africa. In Mali in particular, en- terprises have been privatized under complex equity arrangements involving foreign and local entrepreneurs, the Malian government and the French gov- ernment. In one case, the fruit and vegetable company FRUITEMA, a local private company, has been able to increase its share gradually to 65%. An interesting variation has been the involvement as equity holders of multilateral and bilateral finance organizations such as the International Finance Corporation, Britain's Commonwealth Development Corporation and 89 World Bank (1993). 90 In both of these cases, enterprise performance has improved markedly since the assumption of (foreign) private management. See FAQ (1991, 1992). PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 121 France's PROPARC0. 91 Their involvement is useful in simultaneously as- suaging government fears of irresponsible or exploitative private manage- ment and private sector fears of abrupt turnarounds in government policy leading to re-nationalization. Our second preliminary generalization is that so-called non-traditional privatization has been more successful than the outright sale of assets to private parties. To test this, we can look at different forms of privatization, and problems that arise:92 1) Classic sale of assets: Outright sale of the entire concern is the most common and most straightforward form of privatization, but .the one in which the problems outlined above can be felt most acutely; problems of valuation, lack of transparency, fears of foreign domination and lack of willing buyers. A specific problem is one of land tenure because land cannot be sold freehold, especially to foreign enterprises, in many Afri- can countries. 2) Management contracts: Many of the advantages of privatization can be achieved through management, technical or marketing contracts, gen- erally with foreign companies, that do not involve the sale of assets.93 Besides Cameroon, such contracts are widespread in Mozambique and Ghana. A management contract was the initial basis for Del Monte's entry into Kenya's pineapple canning industry. These contracts limit the investments, particularly fixed investments, required of private par- ties, avoid problems of land acquisition and are less vulnerable to po- litical criticism of 'selling the country's heritage'. Problems arise when the rights and duties of the management company are left vague or experience slippage. The private partner can find itself either power- less to effect real reform or compelled to adopt a greater role than that specified in the contract. 3) Leasing: An alternative solution is the leasing of operations to private parties; a state holding company may retain ownership of some or all of the assets. Responsibilities for maintenance and replacement of fac- 91 A semi-autonomous unit of the Caisse Fran~aise pour le Developpement set up to assist the private sector in developing countries. 92 Using a classification developed by Grouitch (1992). 93 Or only involve a partial sale of assets, as in the case of the Zambian Poultry Processing Com- pany referred to above. 122 MARKETING AFRICA'S HIGH-VALUE FOODS tories have to be clearly specified. The advantages are largely the same as for management contracts, though a lessee would normally have more powers to effect reform. The problems lie in fixing a fair and attractive rent or rental formula, given fluctuating world markets for the product and in the fact that leasing requires less of a screening process than purchase. A lessee may turn out, once agreements have been finalized, to lack capital for reinvestment and operation.94 Leas- ing appears to be the most viable solution to Tanzania's current white elephant problem regarding its large-scale cashew nut processing factories. 4) Divestiture in favor of employees: Management buy-outs and similar ideas fashionable in developed countries seem to have been little tried in Africa, or at least seldom admitted to. Although parastatal manag- ers are becoming involved in private food-processing in Mali, this is by founding new companies, not by buying parastatal assets.95 To the ex- tent that privatization is aimed at reducing overstaffing or replacing unsuitably qualified managers (most often, generalist civil servants), the reluctance to try these approaches is understandable. Management may be more likely to become involved when relatively small units serving domestic markets are sold off individually, as Box 3.13 illus- trates. It also illustrates some of the pitfalls. These do not exhaust the types of divestiture possible. There are also, for example, hybrid forms involving a leasing or management contract for a fixed period followed by transfer of assets. The discussion does show that if divesti- ture is to be successful, the form it takes must be tailored to circumstances of the case and the country; and that transfer of ownership to the private sector is neither necessary nor sufficient for the successful running of an enterprise. While much past government and donor attention was directed at the development of parastatal food processing ventures, the private sector plays an important role in the processing of high-value foods in many of the rela- tively large and commercially well-developed countries within the region. This can be seen in Table 3. 7 which indicates the relative importance of private firms in various sub-sectors in a selection of ten countries. Private sector activity predominates in the formal processing of poultry, fish, oilseeds, fruits 94 As happened in the case of a pineapple plantation in Guinea (see Grouitch 1992). 95 CIRAD (1992). PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING ].;t;:s TABLE 3.7: Private Sector Involvement in 'Formal' (Medium/Large-Scale) Food Processing Tan- Zim- South Mada- Cote Cam- Industry Kenya zania babwe Africa gas car Ghana Nigeria d'Ivoire eroon Beef 2 4 2 1 1 - 1 3 Poultry 1 1 1 1 1 1 1 1 1 Dairy 3 3 3 1 1 1 1 1 1 Fish 1 2 1 1 1 2 1 1 1 Oil Palm - - 1 - 4 1 2 2 3 Oilseed 1 2 1 1 3 - 2 2 1 Fruit 1 1 1 1 1 1 1 1 1 Vegetable 1 1 1 1 1 - 1 1 1 Tree Nut 2 4 1 - 4 2 1 1 1 CODES: 1 Exclusive or Dominant (e.g. >60%) share of output by privately owned companies. 2 Important (e.g. 25-60%) share of privately owned companies. 3 Minor (<25%) share of privately owned companies. 4 No private sector participation -No formal sector processing exists. SOURCES: Literature Review; World Bank Agricultural Operations Staff. In 19sa ··rvi~1awi'S,~gricultura1.: Development ~nci •·Mar~~tin~. Ag~n~Y. pega9Jo of tfies~,i tlW . .· djve~t ,.lt9~lf $Jf unpfOfit~qle; ~Sta.te~ .ahd '.?Ctiyiti.~s•• ,~t· l~ast one · · Phir~trioe Gpi~c .C1$set: •~~ck ()f crE:)ciif I~~. to cielC1Y$J~ impq~~ 1r~t7ed•.: so ~9rl'l~ ct)i9ken7. died an(i •ot~E!rl3J(:lile~ tQ rE!ach ·the.Ir expected·· l.?pk ·township~. has . . co~tinH~d. fof .,,. ~.any y~ar$.JyUddlerfien: PPt~l~ shee.r,rfmrri. auctions·()rtm.rough fiXf:d. arra~g~· ,. ·melit? }vith isO)~llh61df;lfs.: l'fi~y self t() but¢h,er~ for high: p~ices; seeki~g; porfi: • • · ·. ·. ~.• · pensati6J1 for .the .~isks··· now .:·inherent ip to\ViJst)ip tr~cte, ·a~d "the risk (?f djs~ ~ :.· .·quali~cation··from fQrn'l€JI •. marketing ~han~els~ :iownst\ip :ctemanQ is for fatty ·•: .• :meat qonsidereci :lof'erior J:>y'.the ·i~flexiblk o~icial grading sYstem •• so butch~rs :•.. •··· for ; ,. :p~y·f11ore f?t.this'than.·tne ,t:iuctlon'Pr!~e~ the highest 'gra~e nwµt· · ·• ••.....•..•· · .•.· , •....••.•..:. rvleat.:moVes• quipklY .fltlhe' butctiery st~l~s, clustered .~round p.~blip trans"·· . ~: ~··•:pory points .thrqugh Whiqh ·pJa(}Rworkei,s returh .tb ••th7ir; qut~r-clt}I fesipe~tial ( .areas. The. bUtch~rstogether .with: r()qdsidf3 •·steaktipu.ses ~nd iem~ie. d~arers • <: ·in .r~ad)!-()Ook~d ·gffatcater effici~l'l,t.ly~o: urpan 9~~s~ --·.~ ''····,;~·;,~~-~ (. ,.. ,'···~~~:'· ·._. , and tfl~t~s. • · .,,.''~,·.~:;·. ···;'.-· ...., ·'·~"··· .. . ··.. .' ". The advantages of the informal sector stem from prices, flexibility and taste preferences (Box 3.14). Prices will generally be lower because overhead costs such as rent are lower, labor is generally remunerated at lower rates and less (expensive) packaging is used. However, informal retailers of items produced by formal sector enterprises may still have to pass on high prices to consumers. The informal sector is flexible in its location; artisanal processes generally do not need specific sorts of premises and are less likely to depend on specific infrastructure provisions. Retail and the sale of street food are often carried out by mobile vendors who circulate in the markets and busy, central districts. Actors in the informal sector can also meet changing de- mand by dealing in different commodities and switching between retail and different sorts of processing as appropriate. Taste preferences are one of the biggest advantages of employing artisanal processes. People have strong attachments to the foodstuffs that form part of their culture. Even when local foodstuffs can be industrially processed at a comparable cost to the consumer, the market for them is usually very small. This may be an ideological preference by consumers for a hand-processed 134 MARKETING AFRICA'S HIGH-VALUE FOODS ' ' ., , . ,,.,,.. ,,, ,. :·,' ..,,.,, ,,,,:~, , ··.··.:Slncethe 1~te :t,980~.·· u:s)based, NGQ$JechnoserVe'andAppropriate'rec~:: .. · ·· ho1ogx lhternation~I have work~d ~ith farmer.groupg;: v1nag~ased.s6cietie$,: • ·ana family cornpanies to develop smaU~scale milk proc~ssing;enterpri~e~ri~: :.· . xl.Jral ~e~ya: The ,focus• nas been. on '.the! prodpciian .ota cultur~d mnl<:. b~ver~ :. . • 'age ("~al,le f~cilities and·~ ;sl~ple. technol?gy; 1 · • ·· ,qevetopecJJnitially by . an expatriate visiti'ngfood'technplpgistto the u!liver~lty ; • : df ·NairPbi .. The Cultureq mill< :j~ Yery. po~ulat· anti. is far less' Peri~hable. than .• .• · .: :tre~h .mukas.it.~a~be '.~tdredJor·uRtote.n·?~ysyvithput:refriger~tidl):.:: :.• ····•·· .•. :.: · · . •·.· ..· '> ::rne. N~()$, ~~V~ ~rQvide~ .assistance'. in. fn~!'ke( re$ear~h, '(~Ot~rpri,~e q0yel·, · . ·. :·}?Pment, and •.tra!hlng, an.d;'rn t~~. pases of farmer groups, haye· tnad~ Jnitial; • .· . ••.. ~gUlty i~V,~stmen,ts:throu~··•. a r1ol1·Pr()fjt local trust.• While, ttlere: jia,v? 'peen ar : : ; . .'te'9, failed '.en~~rprises;, mo'st: ()f.tr~ ·uP.$~art ·.~rocessorS ~·a,ve beeh; highly; ' •.. :·• · .sµ~qess™! •· ~hd ,aern~nd for. .the tech!Ji~a1.•. a,nd·. ma~ageria1 · s~fyice$1 '9r:tb~ • . · .• :. ..:N~·o~ oy aaditiona1 far&ier·groups Jn Jse~Ya~as far exceed~<:l'their capacitY t~ ~. · ~ ·.'> h ~· ••• '¢ ·.. . . "· ~- ~ ;. ~ .. ·,. ·, . ' . • ·~,· ~ ~ • '.< "c ., apply, and there is a more natural progression from farming into small-scale processing to somewhat larger-scale processing on a custom basis or using outgrowers. It has been argued, for example, that village-level, camel-drawn sesame milling was a key rung up the economic ladder for many of Sudan's largest trading families. 112 Several of the 'success stories' featured in the FAO's (1982) The Private Marketing Entrepreneur and Rural Development come from among the class of rural artisanal processors. This area deserves further investigation. STRUCTURAL FEATURES IN HIGH-VALUE FOOD MARKETING AND TRADE This section highlights several prominent structural features of African pri- vate sector high-value food trading. It discusses a range of measures fre- quently adopted by private firms in order to overcome prevailing policy, infor- 112 Mahmoud (1984). PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 135 mational and other barriers to trade. This material brings out the consider-. able importance of extra-market relationships in trade development and the often intertwined roles of formal and informal market institutions. Given the underlying weaknesses of infrastructure and market support services within the region, the economies of scale associated with certain trading functions and the relatively high risks and transaction costs faced in HVF trading, it is not surprising that the marketing channels for these commodities exhibit high levels of concentration, especially in export marketing. The evolution of HVF markets and marketing channels will be further examined in the indi- vidual case study chapters which follow. The Private Sector Under State Domination Private trade in agricultural commodities has shown a remarkable capacity to survive decades of government policies that have attempted to limit its role, exclude it from certain commodities or suppress it altogether. Paradoxi- cally, the very adaptations that the private sector adopted in order to survive now limit its effectiveness in a new, liberalized context. Many sub-sectors in numerous countries experienced little state interven- tion, particularly with domestically marketed foodstuffs not considered stra- tegic. Thus, even in Congo, the first African 'People's Republic', the important domestic trade in vegetables and bananas remained private.113 The most im- portant national commodity sub-sectors, however, were subject to attempts at state control. These took two forms: bans on private trade in the commod- ity or limitation of the private sector by restricting the issue of trading li- censes and maintaining official prices, which transformed private traders into 'licensed buying agents' (or some similar phrase). The first strategy was more often applied to staple foodstuffs or export beverage crops. It was also adopted by Tanzania for commodities such as cashews, and by Guinea for virtually all trade at points in its history. The second strategy, sometimes resulting from a partial liberalization of the first, has been used for a variety of export com- modities within and outside the high value foods category and included ground- nuts in Senegal, oilseeds in Sudan, and meat and dairy products in much of Southern Africa. The effects of these two forms of state intervention on the private sector overlapped to a great extent. Ironically, the skills that the private sector was 113 Ali-Gaye et al. (1989), Ma-Mfuka (1989). 136 MARKETING AFRICA 's HIGH-VALUE FOODS forced to acquire and maintain were not those that would serve it in a liberal- ized market. Specifically,114 there was a growth in corruption, as private en- trepreneurs suborned officials to be allotted licenses, quotas or regional mo- nopolies or for a blind eye to be given altogether to their operations. The incidence of this and other sorts of corruption was significant in most African states. In trades that had been prohibited, more private efforts were expended on the survival of enterprises, including bribes of officials, than on their growth and development. In licensed enterprises, efforts were directed to legal, bor- derline and illegal means of gaining patronage. In either case, the private sector collectively lost many of the skills necessary to 'normal' commercial operations, those of business administration and entrepreneurship. In sub-sectors where private traders were enlisted as licensed buying agents, the competitive spirit of enterprise, of finding and defending new markets, was repfaced by rent-seeking behavior as traders obtained quotas and regional monopolies and lived off of them. No value was placed on taking risk, and in systems where a state body provided advances to its private agents, there was no need to seek alternative sources of capital. Government interventions also contributed to a climate of uncertainty; even during periods of liberalization, private operators have been uncertain whether policies favorable to them would be maintained or rescinded. There- fore, they have chosen low risk strategies, diversifying their source of income and avoiding large or long-term investment. Uncertainty has strengthened existing preferences for forms of savings that do not involve reinvestment in formal productive enterprise-gold, real estate, livestock, and social invest- ment in consumption. Some of these effects are illustrated in Chapter VI on the vanilla trade in Madagascar. Licensing has been the most important entry requirement into all levels of production, processing and trade, and most crucially into export- ing. Only a dozen or so firms have been given export licenses. A small num- ber of well-connected firms were awarded most of the export quota. In ad- dition to bureaucratization and rent-seeking, the structure of incentives established by the state has led to the declaration of fictitious product stocks (whose costs are paid for by government) and the covert sale of such stocks at prices below the official rate. Pockets of private enterprise can sometimes exist legally within state- 114 To adapt a framework used by Terpend (1992). PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 137 controlled marketing chains. For example, in Zimbabwe between 1980 and 1989 a small and dynamic family firm was able to manufacture and develop new export markets for corned beef by a combination of buying from the parastatal Cold Storage Commission and importing meat. 115 Cases like this one and those of commodities that were never subject to state intervention are important for the lessons they can provide. Nonetheless, in many coun- tries there has been a major distortion of the private sector such that many private enterprises are wholly unequipped to search for capital, raw materi- als and markets in a liberalized economy. They are even less capable of enter- ing into competitive export trading. Marketing Channels in Informal and Formal Export Trading Systems Export marketing entails potentially quite significant transaction costs, given the considerable physical distance separating the exporter from the target market(s), differing cultural and business norms, potential language differ- ences, and legal barriers to trade and effective contract enforcement. These transaction costs-related to communications, market assessments, legal and accounting fees, transport negotiations and quality claims-have proven to be among the most important barriers to export marketing by indigenous African firms who lack previous experience in international trade. A common feature of successful and sustained export-oriented operations in Africa has been the development of extra-market trading networks which have economized on transaction costs, provided exporters with improved ac- cess to information, finance and other resources, and provided exporters with assurances of market outlets, even in unstable or declining markets. Extra- market networks have taken various forms and have involved both informal and formal systems. The best examples of informal trading networks come from the trade in live animals, both the Sahelian and Coastal West African trade and that taking place in East Africa among producers and traders in Somalia, Kenya and Ethiopia. 116 The livestock trade is largely informal as it involves little institutional or judicial intermediation, except when a dispute cannot be settled within the social and trading community. Livestock trading channels are Riddell (1990). 115 116 · This discussion is based on Kulibaba and Holtzman (1990) and on the case study dealing with Somalia cattle marketing. 138 MARKETING AFRICA'S HIGH-VALUE FOODS highly personalized and based on agreements and sanctions internal to the community. While certain ethnic or familial ties are found, ethnic solidarity tends to be less important than established trading networks involving pro- ducers, intermediaries, traders and final buyers (for example, butchers and processors). Various intermediaries including brokers play important roles in transmitting production and market information and in aggregating supplies. They also negotiate deals and provide assurance to buyers that animals are not stolen. These intermediaries shorten the transacting distance, if not the physical distance, between producers and final processors. Within these trading networks, informal lines of credit tend to be far more important than commercial bank credit. Trust-based credit systems prevail in which traders or processors are required to make only partial or delayed payment for animals until the time that sales receipts are realized.117 Social pressures and threats of future loss of reputation, business or access to ani- mals contribute to relatively low default rates.118 Kenya's export trade in horticultural products has exhibited a blend of informal and formal trading arrangements which have been critical to that country's success (Chapter VII). Informal trading networks have been espe- cially important in the fresh produce trade, where each of the leading firms has based much of its overseas trade on family ties or long-term personalized arrangements with selected firms (see Box 3.16). Relationships such as these have provided for superior access to reliable and up-to-date market informa- tion, reduced trading and price risks, and better flexibility in financial man- agement. In contrast, a large number of firms which have been dependent upon arms-length, short-term marketing arrangements have been highly vul- nerable to changes in market demand and to opportunistic behavior on the part of overseas market importers. The life-expectancy of firms who have conducted trade on a spot market basis has been extremely short, frequently less than one season. For trade in processed fruit and vegetable products, the successful export- ers have been those whose trade has been governed by intra-firm relations or long-term contractual ties. As noted earlier, export markets have only been 117 This is discussed by Samatar et al. (1988) for Somalia and in Chapter IX. 118 Although Kulibaba and Holtzman (1990) indicate growing problems of payment delays or de- fault by West African butchers, partly as a result of structural adjustment programs and weakened consumer spending patterns. PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 139 ... Fp~the.pastquarter century, one of the; core .cbmpon~nt$ of Keny~;s e)(panti; .· . ing horticultural tra(f~: has been the export of· nearly·. two. dozen. 'Asian veg-: . etle marketing sYstem. · · · ·· · · ·· · · · ·· · ·· · .. ·· Beginhing.inthe 196()s and continuing :through the 1970s, there .was. a • . • rapidexwi:Jnslon in the u~K's population of individuals 9f.$o(!thAs1an ethnic : ·•·.:origin,: Fuele(f by ·immigration as.well as'.relatiyelY liighb.irth . rates •. this popu~ ·lation·•·reached' .5~(,),00() . in.1971 and over o~e. million ~y .•19SO'. ·EXhibiting a. ; . preference for cert<)irt tradition~I. fpods: and lacking acceptable: local SUbSti- . tutes :tqrmariy. ot.~11ese. fooas,·Jnere .de,ve1opect a lc;trge Hn:roet demaMJor·~t .. ..• Vlt!tfe range of fresh •yegetabl~swhic!'J.• for. agro·9litnatic reasons. could not h~ i ·· grQwri :1ocally. An import trade' in 'Asian vegetables' first. emerged in the early · '· .·196()$ ~h~n. sev~rcll srn.aU-scale ln~iad: merchants began receiving, produce .. capsignments 'from lhdia,.and•·• seHing···this produce.·from.the :Pari{ing:. lot of . Hea~tirow Airport; Within ~rfew ·years; .Kenya would replace:• India as the· pri- ,. ,maty '}\sian vegetable; .supplier•.·Kenya' s . entry into this market was ea~edby . . }he f~¢tJha~ rn()s~ b( that countYy's 'teadi ng produc~ exporters· .were Asian~ · . ·gWned·•tarnilycompaniesi,vi~h experience in the pro.duction and trade··ot'As!an vegetal?le;$'.,;: . . . ... .. . . . . . . . . ' ' .. . . . . . . .. .. .•. .•.··•·····~~the' e~rly ~970s; ,some. dfthe: ~·ar:king l?t im~qf-ters acquired van.~ and. · .· · • ,; ·:~egan• maKin~ :deHverie~J-0i~tjl)d.o~~as.ed: [~t~iler~;' bther~. ~eveioped :wnol~··· .: , · .: 5aJ~ distribOtio.r cen1:~r$· rye,arby t~. •the .Pfimar)I ITl~i~strea~ ~hol~sale mar~ 1 ·..·, 'Jution C()rnpany. ·This .urm .·would re~shape the. ~Asi~n ~e~etable~ ··market py·developing a ·neti,vork of s,ecoMary wholes~l.ers .· · : c;tnd·retait~rs.within Lond~~an~I distrib:utioo· ne.N.!orks withi~:the majorc.ities :;(n:•: tb~Mid.18,nds: The urm'.sfamHy-~ifiii~tec{supplier Jin .• ~enya··· responded ~o ••.. ·' the .riar~et o~~~ing .byrapitjly expan~ipg supplies ,Jh!~ integriltedexp?rt-1,J:K. :. •. · .·i ··~i$\rjbu~ion: operation: so~n became the. leading. act?r iri this market, .•. setting. · ··: ·.·.·tne· standarps: f~r qu~litx and·•· strqngty infll1encing··the. levels .of . import and• .•WhOl~sale.prJties~ , · . ·· . . ., · 140 MARKETING AFRICA'S HIGH-VALUE FOODS . . ...... :. By th~ late 19?0s,· ~eny~:s other; lead\ng.prOdUce· exporters.al~<> •. deve.1~· · . oPe.d: Hi~hly· ~ersonelized,. lon~~term trading arrengements ·with . sele.cted... irn~ .··.po~Eii;aistribt1tor~ i~; the .LJ.K. )11 :s~tne··· cases, .• the tradl~g fi.rtns were linked by farnlly ties; !ltherWls~/ connections were:rnade through mutual friends and· · .· ;busi~ess .associates~ Both on 'the Kenyan and U~K. sides,the primary actors .if} ihi.s'. 'Asian ·veget~bler trade. remained smalHo-mediurn~scale Asian~owhed . : : .:fal11ilY '.9peraticms.. Whether or not there were family ties, the"typical fn:irne- ·•·. :. wprk f~r treid~ was season.a! contraqts specifying the range of prod~cts to···b~ . ·.traded, required·. q4ahtittes and .delivery days: .n.egotiateq ~xed ···prices,. ~nd · •· . P~Ytile11t· ~rrangemen~s ..•. Most ·Kenyan··· exporter~··. geveloped .and· IYlaintained . :traae .links fo ·~uyers based··· In •··severol. different cities so . . to achieve·· wiqer .•. · · . ·· . · ci1stribut10~ aham1111T1~e airect competition among their buyers; .·.....·· · ... · .·.· · .· .······ · ·.. Jhese::~rsona.lized. an~l . . 1,QnaH~rm .· tradin~ arrangements·· ;fecilitated• im- . prove~ ;inform:ation fiovvs' ·.lowered .exporter: .market risks and .• buyer ·procure- ••··•· rne,nt·risks',r~dl,tced therlsks.faqejjby e){potters of iO)PO~er payment failUre . '. i' :Qf .fa!s~ :9U~l!tY Claims'.. enabl~d· the firm$ ;Jo: bypass. exchange. CbY}troJS and ·'. oth~rWl~e take. ~dvi:mtage <>lcurre~cy fluctuations; ·anti. en?bled the· firms to •. •.uooertalUtlonchannels, inc1uciing t11ose se~ingmult!·i)1e-Ghair;i retailer$ ..• the focus of foreign-owned and joint venture firms or firms which already have had contracts with foreign manufacturing and distribution companies. The Kenyan firms or subsidiaries have not been engaged in international marketing in the usual sense. Instead, Europe-based affiliates have been re- sponsible for most decisions regarding product line, sales volume, delivery time, market destination and prices. Market research and development ac- tivities have been left to the European affiliates. So have the determination of distribution channels for the product and the negotiation of sales agreements with individual distributors and manufacturers. While these integrated sales arrangements have led to concerns and alle- gations about transfer pricing and neglect for the financial profitability of the Kenyan enterprises, such ties have been critical to the success and sustain- ability of trade. Long-term and exclusive trading ties with major Europe- PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 141 based firms have provided Kenyan and joint venture firms with immediate and continued access to markets, even those that are highly concentrated at import and wholesale levels or that have experienced stagnant or declining demand. These connections have improved flows of market and especially technical information and have reduced logistical problems. While these non- market trading ties have generated certain social costs (for instance, the fail- ure to provide Kenyans with true international marketing experience), those fruit and vegetable processors who lacked strongties with international firms have generally been unable to compete in European and Middle Eastern mar- kets on the basis of cost and quality, and as a consequence, have not been able to develop sustained and profitable export trades. A similar contrast is found in the recent market liberalization of cashew nuts in Tanzania (Chapter IV). Since 1991, more than ten private firms and cooperatives have sought to export raw nuts to India alongside the trade being conducted by the Tanzania Cashew Nut Marketing Board. A few of the private firms with prior international trading experience and experience serv- ing the Indian market have developed successful trades over the past year that are based on links with affiliated companies overseas. The other work- able alternative has been to engage trade brokers who have provided access to international sources of finance and obtained guaranteed purchase orders from individual Indian processors. These half-dozen successful exporters have not had to deal directly with Indian buyers or to handle the significant paper- work associated with trade between Tanzania and India. In contrast, other Tanzanian exporters and cooperatives have entered the trade by negotiating deals with individual Indian buyers with whom they have had not prior business dealings. The results have been largely disas- trous with regular cancellations of deals, large quality claims, 'missing' quan- tities of nuts in Indian ports, and other unexpected events which have re- sulted in losses for the Tanzanian exporters. The Pyramidal Structure of Export Marketing Channels Export marketing systems under the exclusive control of parastatal market- ing boards have typically featured a pyramidal structure with significant num- bers of licensed buying agents at the primary level, often a secondary level of cooperatives or other processors and transporters of the crop, and then the final level involving a monopoly exporter. With official producer prices, fixed 142 MARKETING AFRICA'S HIGH-VALUE FOODS trader margins and frequently exclusive trading zones, there was virtually no competition within such systems which would stimulate improved services to farmers and economies in marketing costs. In parallel fashion, export marketing systems within the private sector have also tended toward pyramidal structures with numerous buyers or agents at the primary level and far fewer firms concentrated in export trade. How- ever, except where strong government intervention remains (for example in the case of Madagascar vanilla trade), these exporting systems differ signifi- cantly from the parastatal-led systems because the oligopsonistic exporters and their primary level agents/buyers must compete for the business of farm- ers through price and non-price means. Although market liberalization, exchange rate devaluation and foreign cur- rency retention schemes have led large numbers of individuals and firms to enter into the export trade in unprocessed high-value agricultural commodi- ties, most of these entrants operate on a very small, periodic or seasonal basis. Failure rates are high. In most of these industries, only a handful of firms tend to account for two-thirds or more of total trade (see Table 3.10). These are the core firms which not only remain in business but actually develop the trade through their investments in marketing facilities, invest- ments in trading relationships and devotion of time and effort to lobbying government for improvements in services and the trade regime. While there are cases where trade oligopoly is a matter of government protection and rent-seeking, in the more typical scenario, trade concentration results from economies of scale, differences in management and marketing skills, and dif- ferences in access to (off-shore) financing and overseas marketing channels.119 Unfortunately, whether it is due to the ethnic characteristics of the lead- ing firms, to objectives related to 'spreading the benefits' from particular trades or to other factors, these trade leaders are typically omitted in government and donor programs to expand and increase the competitiveness of non-tra- ditional agricultural trade. Yet, at least in the short-to-medium-term, it is likely to be these firms which will underpin non-traditional export growth and provide the models in terms of products, markets, technologies, organiza- tion which other firms will seek to imitate. 119A recent review of major 'success stories' of HVF exports in Latin America and Asia similarly found high levels of concentration at the export stage, although in some industries there were large numbers of participants at the primary production and post-harvest stages (Jaffee and Gordon 1993). PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 143 TABLE 3.10: ffigh Rates of Export Concentration Among Private Firms Number of Industry/Country Share of Trade(%) Leading Firms Horticulture Kenya FruiWegetable 67 6 Senegal FruiWegetable 81 5 Ghana Pineapple 63 6 Fish/Animals Nigeria Shrimp 74 3 Cote d'Ivoire Fish >75 3 Somalia Cattle 70" 3 Spices/Nuts Tanzania Cashews 3 Madagascar Vanilla 3 • For overseas trade from southern Somalia. ' Market share among private firms. Also public exports. Cross-border trade No study of private trade in Africa can ignore the existence of enormous unofficial flows of goods including high-value food commodities between Afri- can countries. 120 While accurate statistics are hard to collect, those who have studied the phenomenon have come up with astonishing estimates. In 1983 the total value of contraband between ECOWAS states was estimated at US$ 1.7 billion. Togo's unofficial trade with Ghana, valued at CFA 5.7 billion in 1983, was equivalent to 25% of its total official foreign trade. Gambia did not feature in Senegal's official trade statistics in 1985, but 45% of its actual foreign trade that year was with Senegal. Benin's official imports of sugar, which it does not produce, were less than the published consumption of its brewery. 121 The large flows of livestock between African countries are de- scribed in Chapter IX. This trade appears in different forms from organized smuggling opera- tions to local people trading with neighbors as they did until a national bor- der was put between them. The extent to which such trade is explicitly ille- gal-or merely unrecorded-varies. In some cases, goods pass into an official export market through an intermediary country:· coffee from Tanzania and Zaire has been known to enter Rwandan official marketing channels, and 120 And in the case of Northeast Africa, between those countries and Saudi Arabia. 121 Various authors cited in Coussy and Hugon (1985). 144 MARKETING AFRICA'S HIGH-VALUE FOODS Gambian groundnuts to have been bought by Senegal's marketing organi- zation. Even when the trade appears to be a local and traditional phenom- enon, it is often very sophisticated: in markets in the area where Zaire, Uganda and Sudan meet traders deal in U.S. dollars and gold, as well as the three national currencies. 122 Unofficial cross-border trade is generally explained as the consequence of government policies or as a straightforward manifestation of comparative advantage. The first argument is based on differences between parallel and official exchange rates (such as the former overvaluation of the Nigerian N aira and the Ghanaian Cedi against the CFA Franc) and on differences in prices due to official pricing and taxation policies. The latter argument is based on differences between a country with an interventionist marketing board and one with free trade in a commodity, or between two marketing boards, as in the case of Gambian and Senegalese groundnuts. While it would be unwise to deny the relevance of these factors, argu- ments about them are often couched in extreme terms and become untenable. Barad (1990) cites 'conventional wisdom' as holding that: African states do not trade with each other because they all produce the same tradable goods. Unofficial transborder trade flows, if they are recognized, are usually explained as a response to price differentials induced by government policies. Many obvious examples contradict this 'wisdom': for example, where unof- ficial trade flows in agricultural commodities are the result of the most straightforward form of comparative advantage: ecological differences. The huge flows of livestock from semi-arid zones to the coastal countries of West Africa, Egypt and the Middle East are a ready example of this, but there are others. There are also intermediate cases where the differentials between countries are in general levels of prosperity and purchasing power; this ex- plains the export of fish over short distances from Tanzania and Uganda to Kenya. 123 One significant feature of unofficial transborder trade is that entry barri- ers are usually lower than those to the official export trade. Potential buyers 122 Meagher (1990). 123 But not the export of dried fish (by bicycle) from Tanzania to Burundi, which seems much more specifically dependent on exchange rates. It is estimated that 20% of Tanzania's exports ofnile perch are undeclared exports by boat or road (van der Hoeven and Budeba 1993). PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 145 are likely to be known through existing networks. When, as is so often the case, these are ethnic- or kin-based networks which span national bound- aries, they provide pre-existing relations of trust and possibilities for informal enforcement of contracts. There is also more scope for small and irregular shipments, and fewer bureaucratic skills are required than with sea or air freight. These points are explored further in Chapter IX. Commodity Diversification as a Trading Strategy Multi-sectoral holding companies have played a prominent role in high-value foods processing and trading in certain African countries. These companies or groups have been able to spread their risks and better utilize their access to in-house or external financing. Their access to considerable infrastructural and managerial resources helps them gain a competitive advantage in the individual sub-sectors where they are active. Commodity diversification is a prominent feature of medium-to-large scale private agricultural trade. This pattern contrasts sharply with the commod- ity specialization of most (export) marketing boards. Commodity diversifica- tion is common among private firms oriented to both domestic and foreign markets. Among the latter, one commonly finds exporters who use part of their foreign exchange earnings to legally (or illegally) import food, spare parts, appliances and other goods. This is especially common in intra-regional trade where the exporters import other items on the return trip and therefore do trade coming and going. In many cases, this trade in imported products has been more profitable than the export trade. Oftentimes, it has been necessary simply in order to conduct export trade. Examples in this last category come from Guinea Bissau and Mozambique where the lack of basic goods and financial liquidity in the local econoillies led cashew nut exporters to pay farmers in the form of rice and other foods which the exporters imported. Also, in Somalia during the mid-1980s, trade liberalization provided profitable opportunities for the im- portation and distribution of vegetable oil, rice and other foodstuffs. The only firms which had the resources (and foreign exchange) to conduct this trade were the principal livestock exporters whose overseas trade was at the time only marginally profitable. 124 Policy reforms on exports ~d imports may thus have greater complementarity than sometimes realized. 124 See Jaffee and Well (1985). 146 MARKETING AFRICA'S HIGH-VALUE FOODS Aside from this blend of export and import trading, trade diversification is widely practiced in order to counter the seasonality of individual crops, and thus to utilize more completely transport and other infrastructure. An additional factor may be a move to minimize policy-related risks related to possible government interventions in the marketing of any particular commodity. Among our case studies, trade and investment diversification are quite common and usually provide for a complementarity of activities, products and market outlets. For example, many of the emerging private exporters of cashew nuts in Tanzania have been actively involved in trade in a range of non-traditional agricultural exports for many years. Each of these commodi- ties, including various types of oilseeds and legumes, is storable and does not require specialized marketing facilities. These traders did not have to under- take any new investments to enter this trade. Since each of the crops has a different seasonal yielding pattern, the firms and their procurement agents have been able to engage in trade throughout the entire year and spread the investment costs for trucks and other infrastructure over a broad range of commodities. Revenues from other crop sales have provided working capital for these firms to make their initial cashew nut purchases. Finally, commodity diversification has been an important ingredient in successful fruit and vegetable export marketing operations within the region. While a high-quality lead item is critical for initial penetration of particular markets, successful exporters have quickly diversified their product range so as to differentiate themselves from competitors and to spread their own com- modity supply and market risks and to maintain steady clients of air freight carriers through both the peak and slow export seasons. By contrast, most of the enterprises who have failed in Kenya, Senegal, Mali and elsewhere have been totally dependent on only one or two commodities for their horticultural trade. Product Promotion as a Marketing Strategy As in much of the developing world, product promotion has something of a bad name in Africa. It has been argued that sophisticated product promotion has largely benefited affiliates of multinational firms and has served to marginalize local firms. 125 Me:rµories of the breast milk vs. baby bottle for- 125 See, for example, Langdon (1975), Mufson (1985), and Jouet (1984). PRIVATE SECTOR HIGH-VALUE FOOD PROCESSING AND MARKETING 147 mula controversy, efforts to increase local tobacco use and similar campaigns by multinational firms to engineer a 'transfer of tastes' by introducing 'inap- propriate products' from industrialized to developing countries have tarnished the image of advertising. Advertised products suffer by (negative) association. While recognizing the social costs and frivolous nature of some promotion campaigns, it would be wrong to regard all food product promotion in Africa as anti-social and a mistake to ignore its potential advantages both in domes- tic market development and in exports of high-value foods. Yet, other than the horror stories,- the literature on agricultural processing and marketing in Africa is virtually devoid of actual experiences in marketing or merchandis- ing, including market research and product and brand name promotion. This applies to both public and private sector enterprises. Advertising and product and brand name promotion are fairly significant in a number of African col.Intries, especially Nigeria, South Africa, Zimbabwe, Cote d'Ivoire, Senegal and Kenya. 126 In Kenya, research in the early 1980s found that advertising for food and drink products exceeded Ksh. 10 million per year, and ranked fourth behind transport, personal care and household products in advertising expenditures. In both the food and drinks sectors, as well as more generally, many of the most heavily advertised products be- longed to affiliates of multinational companies. Some local companies were also active. 127 More recent analysis of the marketing strategies employed by Kenya's small-to-medium-scale local private dairies reveals that few engage in media advertising. Yet, several do promote their products through in-store demonstrations and the distribution of samples at public events. Box 3.9 pre- viously discussed the experience of one firm in Kenya in successfully promot- ing pork products among middle- and higher-income consumers. One of that firm's promotional devices are visits of "Bwana Sausage" to stores where he offers cash prizes to individuals who are seen ordering the company's prod- ucts. Food product promotion hasn't only been directed at upper-income groups. In Nigeria, successful efforts have been made in recent years by research institutes and by small-to-medium-scale private firms to promote various types of soybean-based food products among all strata of income groups. The available evidence does suggest that in many African countries, the most aggressive product promotion efforts are undertaken by affiliates of 126 For Nigeria, see Newswatch (May 13, 1991). For Zimbabwe, see Katsande (1987). For Fran- cophone West Africa, see Jeune Afrique Economie (December 1987). 127 Jouet (1984), op cit. 148 MARKETING AFRICA'S HIGH-VALUE FOODS multinational companies. This can be attributed to their experience, manage- ment and to the up-market clientele whom they serve. Local firms appear more inclined to sell non-branded products to institutional buyers or to rely upon longstanding links with retail outlets and chains to distribute their goods. In markets where competition is weak, there has been little perceived need to communicate directly to consumers or to listen to them. Such passive marketing strategies are likely to result in missed opportunities for local firms and relatively weak market development. Product promotion is even more important (and less controversial) with regard to the export of high-value foods. Brand name recognition and promo- tion are prominent features of international trade in canned food products, and increasingly, fresh horticultural produce. For the most part, African ex- porters sell their products under somebody else's brand. With the exception of South Africa, nearly all of Africa's exports of canned fish, vegetables and fruit are marketed under the brand of a European, American or Japanese com- pany. In many cases, these companies operate or are joint venture partners in the African firms. Many African exporters of fresh horticultural produce have sought to es- tablish a recognized brand name in European markets. These efforts have not been successful since few firms have been able to effectively promote their brand and to back it up with consistently reliable service and product quality. Thus, there are dozens of different West African brands for fresh pineapple, yet there has been little effective promotion of any of these brands for the past decade. 128 All of Kenya's more than one hundred fresh produce exporters have their own brand name, yet few of these are recognized by European distributors and consumers. Only South Africa's Outspan and Cape brand names (for fruit) are widely recognized in Europe. 128 Afrique Agriculture (July 1993). PARTB: CASE STUDIES MARKET LIBERALIZATION IN TANZANIA'S CASHEW NUT INDUSTRY 151 INTRODUCTION Chapters IV through IX are case studies examining the development of and private sector role and performance in six commodity systems in five sub- Saharan African countries. Each study begins with an analysis of the techno- economic characteristics of the focal commodity and its production, process- ing and marketing. This yields insights into the potential risks, transaction costs and logistical problems which private processors and traders face as well as the potential barriers to their entry. In each case study, this analysis is followed by a brief overview of the importance of the focal commodity in the production, domestic consumption and international trade of sub-Saharan Africa. The background and historical development of the particular country commodity system is then outlined. The subsequent and core section of each chapter concerns the characteristics and operations of the private sector. The order of presentation of the studies relates to the degree of government inter- vention in the focal commodity systems, whether directly in trade and pro- cessing or indirectly through regulation and control of private activity. While in some cases governments have facilitated private sector investment and trade, at other times governmental controls and direct participation in mar- kets have deterred private sector entry or undermined its performance. Chapter IV examines the development of Tanzania's cashew nut indus- try, emphasizing the nature and apparent impact of recent policy changes which have liberalized the market and permitted private firms to participate in trade. For more than two decades this industry featured monopoly trad- ing and processing operations by a parastatal enterprise. With the indus- try approaching almost complete collapse, the government moved in 1991 to liberalize cashew trading. The case provides insights into the nature of private sector responses to policy reforms in a still uncertain policy and mar- ket environment. Chapter V examines the development of and private sector role in Kenya's dairy sector. While smallholder dairy development in that country has been a considerable success, formal marketing of milk and other dairy products has been dominated by a government-controlled cooperative which has used the powers of the state and donor financing to marginalize private and indepen- dent cooperative dairies. The case study examines the entry and competitive difficulties faced by the private and cooperative dairies, their strategies for coping in a constrained market environment, and their prospects under a liberalized market. PART 8: CASE STUDIES 152 Chapter VI deals with Madagascar's vanilla sub-sector. It provides a case in which processing and trading functions have been conducted exclusively by private firms and individuals, yet within a highly controlled framework set by government. In recent years, excessive market controls and a short- sighted commercial policy have dulled producer and trader incentives, chan- neled private entrepreneurship into rent-seeking modes, and undercut the country's international market position. This case is illustrative of circum- stances in Africa where private firms have been forced to serve as mere agents of the state. Chapter VII deals with the case of Kenya's horticultural sub-sector, one of Africa's major success stories in non-traditional export development. This sub- sector has featured a dominant role for the private sector in processing and trade, although during its post-independence take-off stage the government facilitated a series of foreign investments which proved to be of strategic importance to the overall development of the sub-sector. The case study high- lights the diverse forms which private processors and traders take and their varied linkages with producers and with overseas marketing agents. The importance of developing long-term crop procurement and marketing arrange- ments in this highly risky line of trade is emphasized. Chapter VIII examines the case of Ghana-fisheries, one of West Africa's largest and most influential fishery sub-sectors. The case highlights the cen- trality of artisanal fishing and processing activities for domestic markets and the short-term adverse effects which that country's structural adjustment pro- gram has had on the sub-sector, both its artisanal and industrial components. Chapter IX deals with the case of cattle marketing in Southern Somalia prior to the breakdown of civil order in that country in 1992. The case illus- trates the importance of informal and often kinship-based trading networks in Africa's long-distance and cross-border trade in live animals, as well as the responsiveness of private traders in the face of general economic and political instability. Iv. Private Sector Response To Market Liberalization in Tanzania's Cashew Nut Industry · Steven Jaffee INTRODUCTION I n the period between World War II and the early 1970s, Tanzania devel- oped one of the world's largest cashew nut industries, accounting for some 30% of total world production and ranking among the leading exporters. Cash- ews became one of Tanzania's largest sources of foreign exchange and pro- vided an important source of income for some 250,000 smallholder farmers. Although Tanzania's cashew nut trade was intitially developed by private traders, during the 1960s a multi-tiered marketing system-featuring local cooperative societies, regional cooperative unions and a national marketing board-was imposed, and the role of private traders marginalized and even- tually removed entirely. As analyzed by Ellis (1979, 1980, 1982), this multi-tiered marketing sys- tem failed to operate efficiently. Rising marketing costs together with declin- ing real producer prices and the forced 'villagization' (a program of collectiv- ization) of smallholder farmers greatly undermined the incentives to produce cashew nuts. From the mid-1970s the industry underwent a steady and dra- matic decline. By the late 1980s, marketed cashew nut production was only 15% of the level it had reached in the early 1970s. Most government process- ing factories lay idle and Tanzania's reputation for a high quality product had been badly damaged. In 1991 with the industry on the brink of complete collapse, the Tanza- nian Government announced that the cashew nut market would be liberal- ized, and private firms permitted once again to undertake trade. It viewed 153 154 MARKETING AFRICA'S HIGH-VALUE FOODS liberalization of the cashew nut market as somewhat of a test case whose results would influence the speed and nature of institutional reforms in the processing and marketing of Tanzania's other major industrial crops, includ- ing coffee, cotton and tobacco. This chapter examines the liberalization process within Tanzania's cashew nut industry over the past two years, placing particular emphasis on the nature of the private sector response to market reforms. What types of firms have. been able to take advantage of the market reforms? How has the re- entry of private firms into the cashew trade affected the trade of cooperatives and the Tanzania Cashew Nut Marketing Board, and the welfare of farmers? How competitive is the liberalized market and what factors have constrained the emergence of a greater degree of competition? What residual market con- trols have remained, either officially or unofficially? To what extent has mar- ket liberalization placed the industry on a road to recovery? The liberaliza- tion experience within Tanzania's cashew nut industry offers useful insights for other sub-Saharan African countries where uncertainty remains about the appropriate role for marketing boards in liberalized markets, the ability of cooperatives to compete in such markets, and the ability of indigenous firms to take advantage of new trading opportunities. This chapter is comprised of four sections. The first section examines im- portant techno-economic characteristics of cashew nuts and their production and processing, noting possible implications for the organization of cashew nut processing and marketing. The second section reviews the world market sittiatiOn for cashew nuts and the changing position of Tanzania and other sub-Saharan countries therein. The third section traces the rise and decline of Tanzania's cashew nut industry from the early 1930s until 1990. Finally, the last section examines the process of market liberalization which has taken place since 1991, emphasizing the nature and effectiveness of the private sector response. Cashew Nuts: 'Poor Man's' Crop and 'Rich Man's' Food The cashew (/\nacardium occidentale L.) belongs to the Anacardiaceae family of plants, which also includes the mango, the pistachio nut, and the noxious poison ivy, poison oak, and poison sumac.1 The cashew tree is an evergreen perennial found in many tropical areas. Native to northeast Brazil, in the 1 The English word cashew is derived from the Portuguese caju, which in turn came from the Tupi- Indian acuji. This section of the case study is based on FAO (1969), Northwood and Kayumbo (1970), Rosengarten (1975), and Ohler (1979). MARKET LIBERALIZATION IN TANZANIA'S CASHEW NUT INDUSTRY 155 16th Century the cashew was taken by Portuguese colonists and missionar- ies to East Africa and Goa (India) where it thrived. The cashew has since spread to several dozen countries, primarily between latitudes 15 degrees north and south. Under favorable conditions, the tree can grow to a height of 40-50 feet. While used for shading and anti-erosion purposes, the cashew tree is most valued for its fruit. The fruit is unusual in that it has two connected parts. One part is a fleshy, pear-shaped stalk, known as the 'cashew apple'. This apple is juicy and thin-skin:µed, from two to four inches in length. The second part of the fruit is a kidney-shaped nut, one to one and one-half inches long. The nut shell is smooth and oily and contains a toxic, resinous material. Contained within the nut is a white kernel-the cashew nut which is eaten- approximately seven-eighths of an inch in length. While the cashew nut kernel is frequently the focus of commercial ac- tivities, various components of the cashew fruit as well as the cashew tree itself can be regarded as industrial raw materials. In this regard, the cashew is similar to the coconut tree. Having an astringent taste when eaten directly, the cashew apple has multiple potential uses in juices, jams, syrups and alcoholic beverages.2 The vitamin C content of the apple is several times that of citrus fruit. The toxic liquid found in the nut shell, known as cashew nut shell liquid (CNSL), has been used in tropical medicines and has had various industrial applications (as a friction-modifying material in brake linings and as an ingredient in paints and varnishes). The cashew nut shell can be ground into a powder for use as an ingredient in building materials and pest-repel- lents. The wood itself provides excellent raw material for charcoal production and for the construction of boxes. Our discussion focuses attention on the cashew nut and kernel since these are the dominant commercial products within Tanzania's industry. Cashew nuts (and kernels) have certain properties which may influence the organization of their trade. They are characterized by: 1) Low perishability: AB long as the nuts are properly dried immediately after harvest (to a moisture content ofless than 8%), they can be stored 2 Ill most countries the cashew apple is wasted or discarded. In Africa, despite the fruit's very high ·nutritional properties, the 'apple' is used primarily to make alcoholic beverages. In contrast, both in India and Brazil the apple is widely. used, contributing to increased farmer incomes and increased employment and value-added in industry. In India, a carbonated beverage and several types of alco- holic drinks are made from the apple; in Brazil, a wider industry has developed which uses the apple to produce juice, jam, chutney, etc. 156 MARKETING AFRICA'S HIGH-VALUE FOODS without significant loss of quality for up to one year.3 This implies that production and subsequent processing operations need not be tightly coordinated with one another as processors can conceivably maintain stocks of nuts to guarantee even throughput into their factories. Post- harvest drying involves a simple technique of placing the nuts on bam- boo or metal sheets under the sun. Hence, unless the harvest period coincides with precipitation or very high humidity, there is no technical requirement for a rapid collection of the crop from farmers. Cashews can be stored in heaps or in bags and do not have special storage re- quirements. The kernels also have a long shelf-life as long as they are stored in air-tight containers or similar forms of packaging.4 2) Bulkiness of the raw material: Although cashew nut kernels are a very high value commodity (see below), the kernel makes up only 20- 30% of the weight of the harvested raw nut. Given the presence of foreign matter, empty shells, and diseased or deformed nuts, the actual yield of kernel material from raw nuts is generally only 18-23%. Un- less other by-products are to be used, the transport of raw nuts will involve movement of much wasted material. This indicates potentially large cost savings by locating processing facilities in close vicinity to major production areas. 3) Wide quality variability in the raw nuts and kernels: As a result of genetic, agro-cl~atic, and other factors, there are significant differ- ences in the size and shape of raw nuts, in the thickness of the shell, and in the proportion of kernel material contained. These variations occur between locations, trees and even on the same tree. The quality (and size) of the kernel cannot be judged by the outward appearance of the nut, except when there is clear evidence of disease or deformities. Even though kernel size/quality is not necessarily linked to raw nut size and color, these crude indicators are frequently used to determine grades. Only by doing a cutting test on a sample of nuts can a trader detect the true quality of the nut. There is thus a likelihood for conflict and uncertainty over grading. 3 Nuts which are not properly dried will develop mould or bacteria or experience irreversible dam- age to the kernel as a result of enzyme action. 4 In contrast, the cashew apple is highly perishable and must be processed within 24 hours after being picked. MARKET LIBERALIZATION IN TANZANIA'S CASHEW NUT INDUSTRY 157 Wide quality variability also occurs for kernels. Major distinctions are be- tween 'wholes' and 'brokens', between 'white' and 'scorched', and between different sizes of kernels. There exist international quality standards which provide the vocabulary for international trade. Very large price differentials exist between the different grades of whole and broken cashew kernels. It is the production and proper grading of high standard kernels which separates profitable from unprofitable processing operations. The production and processing of cashew nuts have several characteristics which also influence organizational patterns: 4) Cashews are tolerant of varied, marginal soil and climatic condi- tions: Due to their extensive root system, cashew trees can grow in areas with relatively poor soils and with relatively low rainfall. Cashew can grow on poor or stony soils, although best results are obtained on well-drained sandy loam soils. It can set fruit where rainfall is as lim- ited as 40 inches per year as long as there is a distinct dry period of two to four months. Its resilient properties have frequently led to its culti- vation in areas where few other crops provide an economic return. Yet, although the tree will grow under adverse conditions, it will not reach its full productive potential and consequently yields in such regions are well below their genetic possibilities. 5) Cashew production features relatively low entry costs and little in- tensity of cultivation or maintenance: Cashews can develop in the wild through natural cross-pollination but are otherwise generally cul- tivated from selected or unselected seeds, rather than from purchased seedlings. Entry costs are minimal unless planted on a large scale. The fact that cashews can be intercropped with other crops (at least until the development of the tree canopy), does not lock the farmer into long- term monocrop cultiva~ion of cashews. Few, if any, specialized produc- tion inputs are required for cashew cultivation. Tree maintenance, in- volving weeding and pruning, is less demanding than for most other tree crops. Depending on genetics and local conditions, the cashew will normally yield a commercially worthwhile crop by the third or fourth year and reach premium yields by the eighth to tenth year. The cashew has a very long productiv~ life of thirty years or more. 6) Cashews have a seasonal and uneven yielding pattern: Cashew trees yield mature fruit (and nuts) during a distinct season, which lasts three 158 MARKETING AFRICA'S HIGH-VALUE FOODS to four months. Effective storage is thus critical for continuous utiliza- tion of processing facilities. If the 'apple' is not being used, the grower can wait for the fruit to drop to the ground signifying its maturity. One problem is that the yielding pattern is very uneven with small quanti- ties maturing initially, building up to larger quantities, and then drop- ping down to smaller amounts again. Under humid conditions or where harvesting coincides with precipitation (as in Tanzania), the harvesting must be very regular. The uneven yielding pattern creates barriers to efficient use of hired labor for this task. This, together with features (4) and (5) above partially explain the dominance of smallholder cultiva- tion with family labor for cashews (and its nickname, the 'poor man's crop'). 7) Cashew nut processing is a complicated process not amenable to on-farm activity: Cashew nut processing is more. costly and difficult compared with that for other dessert nuts. The cashew nut shell is more difficult to break and requires conditioning, via humidification and roasting, before the kernel can be removed through decortica- tion. This problem is exacerbated by the presence of the toxic, sulphur- containing CNSL. Physical ·contact with CNSL can result in severe bums or dermatitis. Care must also be taken to prevent the CNSL from coming into contact with the kernel and therefore contaminating it. Cashew nut processing thus requires some technical skill and care- ful management. 8) There are a range of technologies available for cashew nut process- ing:5 Traditionally, the various functions of cashew nut processing were performed manually, using semi-skilled workers with experience. This remains the situation in India, the world's largest producer of cashew nut kernels. Since the 1960's, however, various mechanized processes and equipment have been introduced in several countries for roasting, CSNL extraction and decortication (deshelling). For the most part, raw material cleaning and sizing and kernel grading have remained very labor-intensive, manual operations. There are significant differences in investment requirements, labor skills (and health) requirements, and levels of efficiency between the Indian, largely manual technology and the various types of medium-to-large-scale mechanical and semi-me- 5 The technologies of cashew nut processing are examined in FAO (1969) and in Ohler (1979). MARKET LIBERALIZATION IN TANZANIA'S CASHEW NUT INDUSTRY 159 chanical operations. In general, the Indian processing system involves lower investment and variable costs and achieves far greater efficiency in terms of kernel material yield· and the proportion of whole kernels. Nonetheless, that system requires large numbers of experienced work- ers who work at unhealthy levels of exposure to the CNSL. The mecha- . nized systems are more vulnerable to breakdown (due to shortages of spare parts, lubricating oils, etc.), require large quantities of nuts for efficient operations,6 and operate well below manufacturer specifica- tions wherever strict grading and sizing activities are not in place prior to the decortication process. THE WORLD MARKET SITUATION FOR CASHEW NUTS International trade in cashew nuts began from India after World War I. Fol- lowing the introduction in the 1920s of improved packaging for long-distance transit, the trade expanded rapidly, especially between India and the United States. By World War II, cashew nuts had become the second most important traded dessert nut (after almonds). They have remained in this position for the past half century. Although consumed in the countries of production, the cashew nut is widely regarded as a luxury snack food, chiefly destined for markets in countries with above average per capita incomes. Cashew nuts are the most expensive of the major dessert nuts with a unit value typically 40% higher than that of almonds and three times that of all nuts as a group. 7 Some 90% of cashews are eaten out-of-hand, in contrast to the pattern for other nuts which are more commonly ground down into paste or used for confectionery purposes. This explains the very large price differentials between whole and broken cashews and among whole nuts of different sizes. World Production Trends World production of cashew nuts grew rapidly during the 1950s and 1960s, reaching a peak of 510,000 tons (of raw nuts) in 1974.8 Three countries- India, Mozambique, and Tanzania-accounted for the majority of this pro- 6 Most of the major manufacturers have produced systems with a rated capacity of more than 3000 tons/year. There are a few manufacturers whose systems have rated capacities of 500 tons/year or less. It is the larger systems which have been put in place in Tanzania and Mozambique. 7 Only macadamia nuts are more expensive than cashew nuts. 8 Gill and Duffus Edible Nut Market Reports, various years. 160 MARKETING AFRICA'S HIGH-VALUE FOODS duction, while smaller industries had developed in Brazil, Kenya and several other African countries. Beginning in 1975, however, and continuing through the mid-1980s, there was a sharp fall in world production, which contributed to a tight international market. Indian production remained stagnant, and the global fall can be attributed to the huge decline of production in Mozambique and Tanzania. Since the mid-1980s, there has been some recov- ery in world production due primarily to a significant expansion in Brazilian production. Table 4.1 summarizes major patterns in world production over the past two decades. The table indicates that the share of sub-Saharan Africa in world produc- tion decreased from nearly 60% in 1969-71 to only 28% in 1989-91. A small recent expansion in West African production has not compensated for the extreme decline in Mozambique's and Tanzania's production levels. While in 1969-1971 Brazil's production was only 8% that of sub-Saharan Africa:, by 1989-1991 Brazil's output exceeded than of the African continent. World Trade Trends Cashews are one of the many varjeties of edible nuts traded worldwide. World trade in edible nuts has experienced relatively rapid growth, averaging about 2. 7% per year since the early 1970s and increasing in value from $1.94 in TABLE 4.1: World and Country Cashew Nut Production, 1969-71to1989-91 Average Annual Production Share of World Production Country/Region (000 Tons) (%) 1969-71 1979-81 1989-91 1969-71 1979-81 1989-91 India 175.7 159.0 140.0 33.8 36.2 28.1 Brazil 24.1 71.3 145.5 4.6 16.2 29.2 Africa 311.l 174.9 140.4 59.9 39.8 23.2 Mozambique 180.0 69.4 44.7 Tanzania 108.4 54.2 27.0 Kenya 16.2 15.8 11.5 Nigeria Na. 25.0 25.0 Guinea Bissau 2.4 3.8 19.0 Others 1 5.0 6.7 13.2 World 519.5 439.7 497.9 100.0 100.0 100.0 1 Including Cote d'Ivoire, Madagascar, Angola, Benin, Togo, Mali and Zambia. Na. Not applicable. SOURCE: FAO Production Yearbooks, various years. MARKET LIBERALIZATION IN TANZANIA 's CASHEW NUT INDUSTRY 161 1980 to $2.84 billion in 1990.9 Only seven varieties of nuts-groundnuts, al- monds, cashews, brazil nuts, desiccated coconut, hazelnuts, and walnuts- account for 90% of international trade in edible nuts. While international trade in cashew nut shell liquid is not insignificant, the discussion here fo- cuses on trade in raw cashew nuts and kernels. Iri.ternational trade in raw cashew nuts has traditionally involved ship- ments from East Africa to India. India was the first country to build up a processing industry. Domestic production has been unable to meet the re- quirements of that country's hundreds of small-to-medium-scale processing factories. Government restrictions on internal movements of cashew nuts fur- ther constrained raw material supplies ·and necessitated imports from the emerging East African industries during the 1950s and 1960s. Peak levels of raw nut imports into India were reached in the early 1970s and exceeded 175,000 tons per year. The imports, generally concentrated over the Decem- ber to May period, complemented the local harvest which begins in May and continues until July. Indian processors could thus operate over an extended period without having to maintain large stocks of raw nuts. This trade declined when production in East Africa declined and as a result of periodic constraints on access to foreign exchange by Indian import- ers. Raw nut imports fell to as low as 20,000 to 30,000 tons/year during the early 1980s. Since then, import levels have increased due to greater supply availability, particularly from Southeast Asia. In the first six months of 1992, imports totaled over 80,000 tons with the largest supplies coming from Viet- nam, Indonesia and Tanzania,.:10 While Tanzania once obtained premium prices over those of other suppliers, this has not generally been the case in recent years due to growing uncertainty about the quality of Tanzanian shipments.11 World trade in cashew kernels is far better documented than is the trade in raw nuts. According to Gill and Duffus, world imports in 1990 totaled over $500 million, ranking cashews second (behind almonds) among dessert nuts. India has long been the world's largest supplier of kernels with its prices and quality setting the benchmarks or standards for the industry. At least in Europe, India has been the preferred supplier, with long-standing trading relationships based on confidence in product quality and on fast and regular deliveries. India has more than 150 cashew kernel shippers, many of which have offices in the United States and Europe. 9 Such figures are for world exports as reported in United Nations, Yearbook of International Trade Statistics. 10 The Cashew, October-December 1992 (Indian trade magazine). 11 As reported by international cashew brokers and as indicated in the Indian import data. 162 MARKETING AFRICA'S HIGH-VALUE FOODS Between .\'Vor!d W,ar 11 anci: tH.e early 19105, MozalJ1bique:~ emerged as the .·.... •wor!c;i'§i:.·largest .pro~uperof. Pa~bew :n4ts, ··a9countil1g .• for··sQme.·.40'.-50% of· . . t:()tal world·.·. prod.uqtion.. Initially•. priYate :trijd~r$. ~f lndi~~ • origin . initially · •.• ~rcr: ... · . :qu.red t,he cmp anq·shil)ped raw nuts t:oJndia trirpr~cessi 0g. Ho\Never,with a· · ••: :ste~qy buflclQp of JocaLprocesstng cap~city after: 1960, Mozambique began •.. . ·•··... to exP:~rt.· .m()st ·of; its ·J)roductiop:·. in·•. Kern.el ,foff"I'!, •·undercutting :India's ;f()tmer . ,;:·rylQf\opoly orthis triJde· . . . . : : . . .. .. . .....· .. ·. .... . • . . . < .· ·. •. : •......·. . ·• ,. ;: . J\ft~r fl/l·azanipique's irldependence 1~7~ •. t~ing~ began to :fail aparf ih ·· Jn . . the 9,ashew •. ~lJt. industry. T~e t;l~cline·wasbrOljgt}t on:.bythe exQdus Qf experi-: ... ·.,e,npetltrad~rs . •and.facfory . ·01~nage£5,::comb.i6ed.with~ciV!f:war····andsevere· .·· :w Shortages of Joreigh.exchange, and PasiC.COQsl!mer.goo#s:$eQt;Jrify .Pf6:Plems . .·.·t~d :farm~r~ to :tle~·fr~mimportant 'productiqnareas and; re$tricted :the ovetaU.. · · :: .; movement of goods' '.Shortages :C)f $pare PClrts further h.C101P~red ·the tra.bs: ·. •· · · ':Port ()f c~s~ew,nµ!S 9ndJorced ~iiny tre1defsoui of busine,f$s; ·• · . ,•. ·. .·..•.... ·.· ...... • ···· . · .·. . ]>roducer. incentives ·a1sq ·decUn~d· sharply.·. The ~xp<>rt· tr~de ~as :br0qgh(. · uncte,r ,~h~· ponirot qf a·.parastataLagency {although. private tracler~·continoe(f·: ?.···· .•.· . to·. act ~s .bl!yiog ~ge11ts)i and ·official ;J:lro.ducer· pric~s ;and Jr~ding. margins : •. :3·:wet~~et:Fr()111them,ir11~19strroug~:to.tre,•.roid-:t_~s9s,Je,C1I prodl!cer.prices··~· · : :·. declined sh,arply.•This ·; a~d .the' ove·ra,U .sh~rtag~. G)f oonsq111er: gooq~: .unde.r~:; ·.·. . ;\c m~ne. '. :·.··'.: .'· ·.'.. ·, ., :' ~ .:·. .: .· · ..· o.,~· •overauf),roductiyify:dec;Jj~'3d ·~·< :":. · . ,,.:: ,··... : ·.: :·.. a~ l'.>ti:>dtic;ti6rcsupp()rfser~.; ;: ·;'· '. .;- ·.':...; : . ; '. ·., ··.' : ;·.' ·. ,· .'. :, :. ··:: ~ . ·:. ·, ~. '. .· :, : ·. ·.'. ·:· < . ;·: . :·. :~; ~·, ·'. . ·. · · · ., .;.··:. ~~··, ·., ·: ;·; '. .; ;., Vices>deterioreit¢1d and age·ofmo$ttree$•extended' bey~nd· tn~ir prOdt;J~~· ·· th~ · · •· tive yef,lrs~: ~r~duction .•.• wmich 'had .·aJe~~~tJ:over: 200 ;oOt:r ton's/yr~ ·durmg·. :.· · i~72.-1974Jelr to'<>n!y 25,00~30,ooo·dunngthe .m1~19s~ •• '.Near1y:n~rr~~f · .;: th~· ~puntry!~ cashew •. n,J~;pr~ce~~irigJC1CJl!ti~$. w~;~ clC!$eQ; .anq. the .• others' :. ·. •:'.oper~t~d·atwe11.beJowcapa~i1y. . · ..... · j .. · .. .·•·.··.·.·.· . . • •. . • •· ..· · · · . . .•.."$t9tting in th~ mid~1980s; the in:Qustcy•began ~orecover..Thi$ process has : .. t• t>een • . ~ccelerated .bY•. the rec~nfending of :tJb$tf1iti~$ ·111 tile ma1n·.~roduction, :.·.. :areas, Private firms· have once again· ~een PermitteI'.faytorl~s ·and . cona?ct.exports; and ·Jn.i992·tf1ey.~aceoun~~crfor a5% 6tj~e· ; • ~~10~ ottr~de.:TtJe •P~rastataLCsju (:Je·~q.cambiqwe: "is p~e:$ently t>eiog.pre-:·. .•. . ···· l. ·• pareq for Privatiz~ii.on. AWorl~ ·aank:f1,19ded project i~~uppo~ingth{:5}7hcll)(n~ : •· or tatlprl cash¢.w·gtov~~ ·ei'nd marketing .inrrastruqt~re~ :rli~ 'ittiprovea: s~curity. •. · . situation . i~ the countryside:has' aided :.cfl'sheW•. 9°11ebti0n, fl/l~~l<~ted produc"'• : tlbrt' reco\i~~ed to54:,ooo 't~~s i~ .19~1/9~·~nd. ex~ort 1~~e1s are increasing··•· 0 ,···~ ·,.···. ·, .. ·:.'·..~·.· .·:.·=·· , .: . .,:· ... ~.· ,, ..· ·····,, ..; ;, .. ·' .., :.>·~,· ·,,. '''····-~ ·.~.·· ... MARKET LIBERALIZATION IN TANZANIA'S CASHEW NUT INDUSTRY 163 v. N....,.,h.. ;\"' •• w• ••, w v·~·., "h ,.w,,·.w,ww·~:->":""<'"~"'<-""·'" •r••••,•••w•·;·"~ »"<"~""<-" "<-'""~ •• , •••• , •• ~t:·:•"YN"N•W" '.~"\ •;••··;·'.~ ;:····:"' ..··.•.fr<>~: :ye~rto ye~d ·c~snew~ .r~~~rri ~Lv~~· 'iro.~oha~t: }~rob:1'<>r. ·~ina1ir~ic1~i ·· agriculture ·in· M~zanioique: and the fur,ther ·recovVer~ll e,bdn~rniC ·(~C~VeyY .•.. ·•· ·. · 'soµijct;s: Edlb;~:NµJ.NJatket.'~eWrt:'. AUstin (i9$1);; Jaffee (1993)>: .world 'sank 'dqcumen~s;<.·. • • ·' • : • • •• ' •••• ~ • • , ..,., < •• • • •• ·;.~ • •• ••• ••••• , ' ,· • ' ••• • •• •• •• •• ''~ •; •• ' • ,. ·' •, ' • • • • • , '..:."' Through the early 1970s, Mozambique was also a major exporter. Its de- cline has been paralleled by a major expansion in Brazilian exports. While Brazil doesn't obtain the technical results which prevail in India's manual processing system, it has been competitive in the U.S. market due to lower transportation costs and the unique larger-sized variety of the nut produced there. Other suppliers, including Tanzania, Kenya and China have a repu- tation for irregular quality, contributing to substantial price discounts (of 25-30%) in international markets. Table 4A (see Annex at end of this chap- ter) traces the levels of kernel exports by the leading suppliers, while Table 4B (see annex) illustrates the substantial price premium obtained by Indian suppliers. On the import side, the United States is by far the largest single market, accounting for more than 50% of world imports. The bulk of such imports are eventually marketed as a snack food under three main brand names-Plant- ers, Eagle Snacks, and Fisher. Cashews are the only major dessert nut for which U.S. domestic production is negliglble and the U.S. is in fact a major exporter of almonds, walnuts, hazelnuts and other varieties. Table 4C (see annex) indicates import levels of the. other major import markets (Japan, Australia, Canada, the former USSR and several EEC countries). With the chief exception of the former USSR (which had previously been a large im- porter of lower quality nuts for confectionery use), import levels increased in these major markets during the 1980s. Long-term trends in international prices in kernels are shown in Figure 4.1. The data refer to U.S. CIF prices per ton for benchmark size 320 whole kernels. Current prices have been deflated by the G-5 MUV index to obtain real prices. The graph shows that since the early 1970s, sharp increases in (nominal and real) prices were recorded during three periods-1976 to 1977, 1979 to 1982, and 1984 to 1987. Through most of the 1980s, real prices were higher than those which prevailed during the early 1970s. 164 MARKETING AFRICA'S HIGH-VALUE FOODS FIGURE 4.1: International Prices for Cashew Nut Kernals 8000 7000 6000 ~ 5000 c ---+-- Current Price ~ 4000 ~ ---0--- Real Price 3000 2000 1000 Real Price is Current Price Deflated by G-5 Manufactures Unit Value Index THE RISE AND DECLINE OF TANZANIA'S CASHEW NUT INDUSTRY Cashew nut production and trade has long played an important role in the Tanzania economy. Over much of the period since the early 1960s, cashews have been Tanzania's fourth largest source of foreign exchange, following coffee, cotton and sisal. Cashews are grown by more than 250,000 smallholder farmers. Although cultivated throughout the coastal area, cashew nuts have long been the most important source of farm incomes in the southern regions of Mtwara, Lindi, and Ruvuma. The decline of the industry has had a strong adverse effect on living standards in these regions. The cashew was probably introduced into Tanzania from Mozambique in the 16th Century and was initially used to check soil erosion in the coastal areas. Cashew groves developed naturally and there is no evidence of any specialized production or commercial trade of the crop prior to this century. This section traces the early development and subsequent decline of the Tan- zanian industry through four periods, namely: 1) 1930s to 1962-Private merchants and the development of trade 2) 1962 to 1974-Expansion and multi-tiered marketing system 3) 1974 to 1985-Industrialization and production decline 4) 1985 to 1991-Reversing the downward slide MARKET LIBERALIZATION IN TANZANIA'S CASHEW NUT INDUSTRY 165 Private Merchants and the Development of Trade (1930s to 1962) A commercial cashew nut trade did not develop until the 1930s when plantings- on Mafia Island and on sisal estates in Tanga yielded an export crop of 210 tons in 1938.12 This trade was interrupted during World War II, but immedi- ately afterwards there was an expansion in plantings, especially in the Mtwara and Ruvuma Regions in the southern zone. Both production and trade grew steadily and the rapid progression of exports is illustrated in Figure 4.2. This trade was entirely in raw nuts sent to India for processing. An attempt in the early 1950s to develop a small-scale processing operation at Mtwara was unsuccessful due to the instability in the available labor force. While there were some cashew plantings on estates, including on lands leased by the Tanganyika Agricultural Corporation (famous for the failed groundnut scheme), most of the expanding production took place on small- holdings. Cashews fit into the traditional cropping system and often cashew was intercropped with cassava until the tree canopy developed. In parts of Newala District, the high population density and previous intensive cultiva- tion had impoverished the soils so greatly that only cashews and cassava provided a reasonable return. A drought-resistant tree, the cashew has the FIGURE 4.2: Tanzania Raw Cashew Nut Exports, 1945-62 60000 50000 40000 Sulphur > Sesame--> Yellow Gram - - - > Beeswax - > Cassava--> Pigeon Pea Timber~------------------------~> At the first handler level, competition was still not strong during this second year of a liberalized market. Several factors _contributed to this. One was the overall shortage of finance available. The number of traders in pos- session of cash. present at any one time in the villages was limited to one or very few. A second reason is a common practice among traders to bypass villages where another private trader is known to be active. The reasons for this are not very clear. However, it is the case that private traders have sought to (and have apparently succeeded) 'buy' primary cooperative society secretaries through the payment of a small commission (e.g. TSh. 1/kg.) or through the provision of a gift such as a radio or bicycle. These cooperative officials seek to guarantee village supplies for the favored buyer by telling other would-be buyers that the local crop is already committed to another party.41 Third, price competition was weakened by the announcement of guide- line or indicative prices. Given the many years of government-directed prices, the notion of an indicative price is not well understood, and most traders and primary societies tended to settle on prices at or very close to the indicative prices. The fact that cooperative unions were paying the indicative price rein- forced this notion of an official price. While the majority of the local first handlers are indigenous traders, some of the larger traders and all of the exporter buying agents are long-estab- lished Tanzanian Asian trading companies. Table 4.10 provides a breakdown 41 AB one trader noted, the objective is to have the cooperative secretary tell would-be competitors that there are 'no vacancies' in the village for cashews. · 186 MARKETING AFRICA'S HIGH-VALUE FOODS TABLE 4.10: Cashew Nut Suppliers to TCMB Lindi (1992/93 Season) First Handler/Supplier Quantity Supplied Proportion Coop. Society/Union 293.6 6.5% R.T.C. (Public Enterprise) 202.5 4.5% Private Traders 3988.2 89.0% Of which: Tanzanian Asian (14) 2048.6 45.7% Indigenous (17) 1939.6 43.3% SOURCES: Based on TCMB Lindi Buying Records; Lindi Regional Agricultural Staff. of the suppliers to the Lindi branch of the TCMB during the 1992/93 season. It indicates that, at least for the trade directed through the TCMB, indig- enous firms have been able to hold their own. The pattern for private exporters is quite different. During the 1992/93 season, twelve private companies undertook (or were preparing) exports, ac- counting for about 40% of the country's raw nut trade. Nine of these export- ers were interviewed, and their profiles are indicated in Table 4.11. By Tan- zanian standards, each of these firms is quite large or is part of a larger group of companies. In contrast with the mixed pattern found for local buyers and traders, each of the exporters is owned and/or managed by Tanzanian Asians, typically family-based companies or partnerships. Only one firm con- centrates on cashew nut trading, but even this firm is part of a conglomerate. Each of the other firms is active in trading of a wide range of commodities. Most are also active in the transport business and in agro-processing ven- tures. During the 1992/93 season, cashew nut trading accounted for 20% or less of the turnover of most of these firms. These firms are also among the country's largest traders of non-traditional agricultural commodities, includ- ing sesame and cassava roots. Table 4.12 outlines the trade mix of one of the largest emergent cashew nut exporters. Notice both the extensive range of products and the diversity of markets served. These new cashew nut exporters are generally quite expe- rienced in international trade or else have hired experienced managers from abroad to develop this business. Several of the new exporters had experience in cashew nut trading during the 1960s and 1970s. TABLE 4.11: Characteristics of Surveyed Private Cashew Nut Exporters s ): ::>: ;:.. Cashew !:'. Cashew Turnover Prior r- Year Fixed Export as% Cashew o: rr Estab- Assets Volume of Total Trading Source of i r- Company Ii shed Ownership ($ 000) (Tons)' Earnings Experience Financing Other Activities!'l'rading i': 2: TWD 1989 Tanzanian Asian 3700 No External Commercial Import and Export c < Individual "' ): ETC 1992 Tan. Asian + U.K 120 2800 20% No Own Commodity Exports ~ Partnership Relief Agency Sales ): < :;; FDHN 1947 Tan. Asian Family 2500 2045 13% Yes Own+Bank Commodity Exports (J Oilseed Milling c: l: (J :J n BDS Na. Tan. Asian Individual 500 2150 (1150) 20% No Own + External Commodity Exports ~ ~ Seed Beans "' c Transport RPR. 1989 Tan. Asian Family 170 2000 (1000) 20% Yes Own+Bank Commodity Trade c c (J Textiles ~ AMKT Tan. Asian Partnership 100 1300 50% No Own+Bank Commodity Exports Na. Gems Pharmaceuticals KRSH 1991 Tan. Asian+ U.S. 15 1200 (1200) 100% Yes Own+Bank Part of Large Holding Company Partnership ABTR 1973 Tan. Asian Family 2500 270 6% No Own Commodity Exports Transport/Clearing Spare Parts MENT 1983 Tan. Asian Family 4050 0 (600) Na. No Own Commodity Exports 1 Figures in parentheses are tonnage of 1991192 crop exported in 1992/93. Na. Not available. SOURCE: Author's Survey (March 1993). " - Cl 188 MARKETING AFRICA'S HIGH-VALUE FOODS TABLE 4.12: Commodity Export Mix of Major Private Cashew Trader (1992/93 Season) Commodity Market Destination Quantity (Tons) Value($ 000) Cashew Nuts India 2045 1496 Cotton Seed Cake Western Europe 16,000 1436 Cocoa Beans U.K; Netherlands 670 589 Sesame Seed Turkey; Japan 1026 540 Beeswax Japan 152 413 Cassava Roots Westtirn Europe 3000 345 Yellow Gram India, Ethiopia 1714 257 Pigeon Peas India; U.K. 755 217 Castor Seeds Europe; Japan 736 146 Miscellaneous Minor ltems 1 Pakistan; Zaire 219 27,300 5658 1 Including betel nuts, cardamom, copra cake, cooking oil, and gum arabic. SOURCE: Based on Private Exporter Trade Records provided to author. While several of the firms financed their cashew nut purchases out of their own funds, most obtained additional funding, either from local commer- cial banks or from external buyers or affiliated companies. Such external finance provided a strong competitive edge, since money was available on a timely basis and at interest rates below 10%, compared with the 27-31% interest rates on local overdraft facilities. As noted above, the private exporters are generally not directly involved in the local-level procurement of the crop. They have instead relied upon private traders (or less commonly cooper11tives) based in the main production areas to arrange crop collection, grading, farmer payments and the prepara- tion of consignments for shipment. The buying agents are not strangers; sev- eral are linked through family, while others have previously dealt with the exporter with other commodities (e.g. sesame, cassava). Trust is essential in this type of trade, given the relatively large sums of money which the ex- porter provides the agent to make purchases, and given the fact that the exporter is simultaneously negotiating forward export contracts which the agent's supplies will help fulfill. Similarly, export transactions are not conducted with strangers. Several firms direct their trade toward affiliated companies or through buyers/bro- kers with whom they have had past trading experience. Few of the firms deal directly with Indian processors. Instead, they rely upon offshore brokerage firms (based in places such as Singapore and the Virgin Islands) which pro- MARKET LIBERALIZATION IN TANZANIA'S CASHEW NUT INDUSTRY 189 vide reliable financial and shipping services and which guarantee product quality to the eventual buyer. The interviewed exporters were asked to rate various potential problems on a scale of 1 to 5, with 5 signifying a 'major problem or bottleneck' and 1 signifying that the factor is not a problem faced by them in their procurement or sales. Table 4.13 aggregates the results of this exercise. According to the exporters, the most significant problem is social infra- structure (telephones, roads) linking them or their buying agents with the main production areas. This raises overall transaction costs in crop procure- ment. Taxation is the next concern of significance, especially the (increasing) local taxes imposed by district authorities. In some locations, the sum of edu- cational, building, and other funds totals up to TSh. 18/kg. While traders are normally required to pay part of the fees, these fees are in fact simply taxes on local producers. 42 Exporters indicated that the storage and transport facilities which they own or lease are more than adequate for their present trading activities. In fact, not a single firm indicated that it needed to make any additional invest- TABLE 4.13: Private Exporter Assessment of Problems (1 to 5 Scale; 1 =No Problem, 3 =Moderate Problem, 5 =Severe Problem) Factor/Problem Rank Average Rating Poor Communications to Producing Areas 1 4.2 Poor Roads 2 3.7 Taxation/Cesses 3 3.1 Availability of Cashew Nuts 4 2.8 Access to Short-Term Finance 5 2.7 Lack of Time/Conflicting Activities 6 2.7 Inflation/Price Instability 7 2.4 Lack of Market Information 8 2.3 Availability/Cost of Packaging Materials 9 2.3 Official Restrictions on Cashew Procurement 10 2.1 Inadequat.e Transport/Storage Facilities 11 2.0 Lack of Market Outlets 12 2.0 High Cost of Cashew Nuts 13 1.9 Unreliability of Buyers 14 1.8 SOURCE: Author's Field Survey (March 1993). 42 Data on producer prices for cashews provided by the Marketing Development Bureau are thus misleading as they do not take into account these taxes. When taken into account, the share of farmers in the export value this past season was only 40-45%, depending upon location. · 190 MARKETING AFRICA'S HIGH-VALUE FOODS ment in order to enter into the cashew trade. In the southern zone, there is significant unused storage facilities, especially at the depots of the National Milling Corporation and at the idle cashew nut factories. Varied responses were given to the question about finance. This was not a problem for those firms who receive financing from overseas affiliates or buyers. Still, it was reported to be a costly and difficult barrier by those firms having to rely upon local commercial banks. The exporters do not see marketing (including mar- ket information, market outlets, and buyer reliability) as a serious constraint to the current trading activities. Private exporters were asked about their potential interest (or emerging plans) to enter into cashew nut processing to replace or supplement raw nut exports to India. Several firms, each with external partners, have been en- gaged in discussions with the TCMB to lease one or more of the existing factories. At least one of these deals is expected to go through in time for operations ·during the 1993/94 season. A few other exporters expressed skep- ticism regarding the financial viability of the complete mechanical processing systems. They have been exploring alternative, semi-mechanized technolo- gies whose operating scales are significantly lower than the existing factories. In March of 1993, no investments of this type were imminent. A third option, suggested by one of the firms, was to cannibalize a few of the existing facto- ries, utilizing the best features from the Oltremere and Cashco systems, while building in less mechanized features for several processes. Dimensions of Performance During the 1992/93 season, the overall performance of the industry was mixed. Total marketed production declined to about 35,000 tons because of an unseasonal cold spell which damaged flowers and reduced fruit setting. There was some evidence of increased effort to combat the powdery mildew problem (through sulphur spraying) and rehabilitate cashew shambas.43 One of the major objectives beyond the market liberalization was to im- prove the reliability of the crop procurement system and to raise farmer in- comes. To some extent, progress has been made. In contrast with past sea- sons when crop buying was not completed until April or later, this year virtually the entire crop was purchased by the end of February, and in some 43 The evidence coming from surveys by the ODA Cashew Research team and from the loan appli- cations of the commercial banks does indicate that it is primarily farmers with above average land- holdings and other assets which are obtaining finance and making the investments to restore cashew productivity. The vast majority of smallholders have not made such investments. MARKET LIBERALIZATION IN TANZANIA'S CASHEW NUT INDUSTRY 191 areas, a month earlier. This occurred despite the delays in bank overdraft approvals. Furthermore, in contrast with past experience, this season farm- ers were paid on a timely basis, especially when private traders paid cash directly to farmers or to the cooperative societies. The activities of private traders undoubtedly put pressure on the cooperative unions and societies to improve their services to farmers. 44 Relatively few farmers benefited from the higher prices (TSh. 140-160/kg. for standard grade) which private traders were offering at the very beginning of the season. While trade figures for the entire season were not finalized by the time that research field work was completed, the share of farmers in the export value is likely to have fallen back below 50% (see Table 4.14). Given the profit margins earned on exports (see below) and the level of producer prices prior to the announcement of guideline prices, it would appear that official interventions resulted in a transfer of income from farmers to private traders/cooperatives/banks amounting to some TSh. 525 million.45 Direct taxes on producers, in the form of local cesses, amounted to about TSh. 350 million (that is, an average of TSh. 10/kg). During the season, both farmers and traders did move cashew nuts across district and regional borders in order to counter or take advantage of differ- TABLE 4.14: Recent Trends in Producer Prices (TSh./kg.)1 Season Standard Grade Undergrade Weighted' Share of Export Value 1989/90 84 56 82.6 57% 1990/91 110 73 108.2 56% 1991/92 137 89 . 132.2 66% 1992/93 120-140 80-90 110.0-127.5 47-54%3 1 Does not take into account local cesses paid by farmers which vary significantly by district (e.g. between TSh. 2 and 18 per kg.) 2 For 1989/90 and 1990/91 weighted 95% standard+ 5% undergrade. For 1991/92 and 1992/93 the respective ratios are 90/ 10 and 75/25. 3 Only raw nuts exported. Estimated average FOB value at $700/ton. SOURCE: Marketing Development Bureau and Author's Field Work. 44 One future difficulty is foreseen here. It is still the cooperative unions and societies which pro- vide most of the input loans and inputs to farmers, with repayment coming through deductions on the delivered crop. With private traders taking an increasing proportion of the cashew crop (and other crops), input loan recovery may become a severe problem for the cooperatives. 45 Given export prices and the shilling exchange rate, still very· comfortable margins would have been earned by the private trade with a producer price of TSh. 140/kg. At this producer price, the cooperative unions who sold to the TCMB could have covered their finance charges and still had a modest profit to begin repayment of arrears. 192 MARKETING AFRICA'S HIGH-VALUE FOODS entials in guideline/actual producer prices and in local cesses. This 'migra- tion' of the crop also took the form of chasing the available cash because in some villages, districts and regions there were periods in which cooperatives, the TCMB, and locally operating private traders did not have cash to make purchases. As a result of these 'migrations' district-level figures for marketed production showed considerable anomalies with huge recorded increases in some areas and major declines in others. Due to differential access to transport facilities, it was traders rather than farmers who were able to best take advantage of these inter-district price and tax variations. The available evidence suggests that local (that is, first han- dler) trader operations were quite profitable this season as a result of the set TCMB buying prices and the guideline producer prices. An indication of this is provided in Table 4.15. The first column is for a Masasi-based trader who procured nuts within a 25 km radius of the town and then sold to TCMB in Masasi. For this low risk operation, the trader obtained a margin of 12.8%. The second column concerns a trader in N achingwea who transported nuts to Lindi to sell to TCMB at much higher inter-store price. The trading margin was considerably higher. Overall, private traders accounted for about 43% of the crop procured di- rectly from either farmers or cooperative societies and, by the completion of sales, probably accounted for a similar proportion of the export volume. How- ever, with the TCMB and several of the cooperative unions not obtaining finance until December or January, the export-oriented private traders (through their buying agents) were able to procure most of the high quality nuts from the 1992/93 harvest. That is, all or a large proportion of the nuts from Tinduru and Nachingwea Districts (known for high-quality nuts) were eventually procured by private traders. In contrast, most of the nuts in Newala District (the quality of which is uneven or poor) were left for the TCMB and MARCU. While procurement was more or less equally divided between MARCU and eight private traders in Masasi District, the majority of MARCUs stock was classified as undergrade. Traders supplying both private exporters and TCMB tended to provide the former with standard grade nuts and the latter with undergrade (selling them as standard grade if possible). While the TCMB was still negotiating with foreign buyers/brokers in March, most of the private exports were shipped by the end of February. Financially hand-cuffed. by the banks, TCMB obtained a lower quality stock of supplies and it is likely that the average export prices it earned will turn out to have been considerably lower than those obtained by the private traders. MARKET LIBERALIZATION 1i/J TANZANIA'S CASHEW NUT INDUSTRY 193 TABLE 4.15: Private Cashew Trader Margins on Sales to TC:MB (1992/93 Season) (All Values are TSh./kg.) Masasi Trader Supplying Nachingwea Trader Cost or Price TCMB in Masasi Supplying TCMB in Lindi Producer Price 127.00 120.00 Levies/Cesses/Duties 14.40 9.40 Village Secretary Commission Na. 1.00 Transport Costs 2.00 12.00 Packaging Costs 4.35 4.35 Finance Charges Na. 3.60 Depreciation (On Vehicles) Na. 4.10 Overhead, Labor, Insurance 1.70 5.50 Total Costs 149.45 159.95 TCMB Inter-Store Price 166.45 190.00 Trader Margin 19.20 (12.8%) 30.05 (18.8%) SOURCE: Author's Trader Interviews. As noted earlier, private trader efforts during the 1991/92 season to export were largely unsuccessful, with both product quality and buyer unreliability problems. Firms entering the trade during the 1992/93 season faired much better. FOB prices were mostly within a range of$675 to 715/ton. While these were lower than for the previous season (when average exports were $813/ ton), the exporters benefited from a devaluation of the Tanzania Shilling from 290 per $1 in March 1992 to 345 per $1 a year later. Together with the lower producer prices, this provided comfortable trading margins (12-20%) for most exporters this season. The trading costs and earnings for two compa- nies (two transactions) are illustrated in Table 4.16. During the 1992/93 season, all processing factories were completely idle. Prior to the season, the large permanent work staff at several of the factories was let go. Both the TCMB and the cooperative unions were under pressure from the banks to repay their loans. Had they chosen to process the nuts themselves, this would have added several months' delay before the crop would be sold and thus before payments were received. Although discussions were held between several private firms and the TCMB about the possibility of TCMB performing custom processing services, no arrangements were ever implemented.46 46 Neither private firms nor cooperative unions have much confidence in TCMB's ability to process nuts on an efficient basis. For its part, the TCMB was requiring firms to deliver a minimum of 1000 tons before it would consider custom processing activities. 194 MARKETING AFRICA'S HIGH-VALUE FOODS TABLE 4.16: lliustrative Exporter Trading Costs and Profits (1992/93 Season) (Figures in TSh./kg. or $/ton) Cashew Nuts Procured in Cashew Nuts Procured Masasi in January and in Tinduru in December Exported via Mtwara Port and Exported via Mtwara Port Cost or Price Item in Late February in Late January Price to Producer 127.00 130.00 Levies/Cesses 18.00 15.00 Commissions to Buying· Agents and/or Cooperative Officials 6.00 6.50 Transport Costs to Mtwara 16.00 19.00 Packaging Materials 4.35 4.35 Weight Loss 1.40 1.50 Finance Charges 13.25 13.56 Overhead Costs 4.00 4.50 Port/Inspection/Handling 6.60 6.60 Total Costs 196.60 (at TSh. 345 = $1) 201.01 (at TSh. 330.3 = $1) $570/ton $608.8/ton FOB Earnings 1 $689.5/ton $691.5/ton Exporter Margin $119.5/ton (20.9%) $ 82.7/ton (13.6%) 1 Assuming 1.5% weight loss in transit and $68/ton for freight and insurance. SOURCE: Author's Field Interviews. CONCLUSION In the period between World War II and the early 1970s, Tanzania developed one of the world's largest cashew nut industries. By 1973/74 marketed pro- duction reached 145,000 tons (about 30% of total world production) and ex- ports were among the highest in the world. Cashews were the country's fourth largest source of foreign exchange and provided an important source of in- come to some 250,000 smallholder farmers. During the 1960s a multi-tiered marketing system-involving local cooperative societies, regional cooperative unions and a marketing board-was imposed on what had been a trade car- ried on by private firms and private traders were removed from the market- ing system. The steady rise in production and trade during the 1950s and 1960s has been mirrored by a dramatic and steady decline of the industry since then. Despite a buoyant international market, Tanzania's cashew nut production fell to less than 17 ,000 tons by the late 1980s and its early reputation for a high quality product was badly damaged. Several factors contributed to this downward spin, including the country's 'villagization' program, a sharp and MARKET LIBERALIZATION IN TANZANIA'S CASHEW NUT INDUSTRY 195 virtually continuous decline in real producer prices, and major inefficiencies in cooperative and marketing board crop collection and downstream activi- ties. Government 'education' campaigns (and threats) were insufficient to counter the loss of farmer incentives due to declining services and incomes. The massive decline in production has not only had a sharp adverse affect on living standards in the southern regions, but has made most of the large- scale, donor-funded, government-owned, state-of-the-art processing factories into 'white elephants'. With the industry on the brink of complete collapse, the Tanzanian gov- ernment announced in 1991 that the cashew nut market would be liberalized and private firms permitted to undertake trade once again. The reform pro- cess has been confusing and full of irregularities, although there are signs that the industry may be set for a recovery. Whether by design or default, the policies of liberalization were poorly communicated by the central govern- ment, leaving local authorities to use their own discretion and ad hoc proce- dures in implementing the new set of rules regarding trade. The first year of the reform (1991/92) was largely a failure: few private traders were licensed, there was no competition, controls over prices and crop movements contin- ued, and results were poor when private and cooperative enterprises sought to market cashew nuts outside of the TCMB. The second season of the liberalized market featured some improvements, though considerable confusion still existed. Private traders were more widely licensed, yet there was little commercial bank finance available either to them, to cooperatives or to the marketing board. In order to insure that official loans extended to the cooperatives and marketing board were recovered, an agreement was reached between these parties and the regional governments to set 'indicative' producer prices at levels 10-15% lower than the previous season. Private firms, which had been paying farmers considerably higher prices at the beginning of the season, complied with the 'indicative' prices by reducing their own prices and pocketing the windfall. Government and com- mercial bank interventions thus resulted in a huge transfer of income from poor farmers to a combination of banks, cooperatives and private firms. Despite these limitations, some positive trends are emerging. With the entry of private traders, crops are collected soon after harvest and farmers are being paid promptly (not only by private traders but also by cooperatives which are feeling the competitive pressure). This is leading farmers to har- vest their cashews, rather than neglect them, and has led some of the better endowed farmers to rehabilitate their farms and plant new cashew seedlings. 196 MARKETING AFRICA'S HIGH-VALUE FOODS During 1991/92 and 1992/93, marketed production averaged 37,000 tons and export revenues recovered to their level of the early 1980s. More and more private firms are entering the trade. Collectively, they accounted for 43% of the purchased crop during the 1992/93 season. Most individual firms are very small and have undertaken the low risk activity of buying from farmers and selling to the TCMB at a guaranteed price. The larger players are firms which are active in domestic and export marketing of a wide range of non-traditional commodities. Some have excellent interna- tional trading contacts which they are using in developing their own cashew nut exports. Due to uncertainty about government policies, private firms have not yet made investments in production or marketing infrastructure for cashews. They continue to utilize trucks, warehouses and infrastructure used for other commodity trade. With the confidence gained and profits earned during the 1992/93 season, this is likely to change. Some firms are likely to begin to establish their own cashew farms and develop closer links with producers. Several firms have expressed an interest in processing the nuts locally. Under the current policy environment, and with considerable uncertainty as to whether Tanzania's highly-mechanized processing factories can be fi- nancially viable even if operated on a commercial basis by private firms, the most appropriate strategy is to enable interested private firms to lease se- lected TCMB factories. At the same time it is important to encourage these and other firms to investigate alternative, semi-mechanized technologies. Leas- ing of the large-scale factories for periods of one to five years would expand the resources available to the private sector without significantly increasing its exposure to policy reversal risk. This approach would signify the govern- ment's commitment to the continued liberalization of trade, plus provide the government with some return on its factory investments. The long-term com- petitiveness of Tanzania's industry will require changes in the technolo- gies used toward a lower degree of mechanization and reduced dependence on fuel oil as an input. While some adaptions can be made within the existing factories or by combining certain features of the Oltremare and Cashco sys- tems, it would be expedient for would-be processors to explore and test alter- native technologies. MARKET LIBERALIZATION IN TANZANIA'S CASHEW NUT INDUSTRY 197 ANNEX TO CHAPTER IV TABLE 4A: Exports of Cashew Kernels by Major Producers ('000 Tons) Country 1975 1980 1985 1988-90 (Ave.) India 58.0 37.4 31.6 34.9 Mozambique 24.4 17.1 2.5 3.2 Brazil 7.6 11.9 25.0 21.7 Tanzania 4.0 5.5 0.5 1.4 Kenya 0.0 2.7 1.9 0.5 SOURCE: Gill and Duffus, Edible Nut Market Reports, various years. TABLE 4B: Price Comparisons of Major Cashew Kernel Suppliers (United States Import Values (FOB) All Grades; 1989-90 Averages) Quantity Unit Value % ofindian Country (Tons) ($/ton) Price Comments India 19,833 4835 100 Manual Processing; High Quality Brazil 19,429 4205 87 Mechanical Processing; Medium Quality Mozambique 2,845 3577 74 Mechanical Processing; Low Quality Tanzania 1,332 3673 76 Mechanical Processing; Low Quality China 744 3568 74 Manual Processing; Low Quality Kenya 403 3262 68 Mechanical Processing; Low Quality SOURCE: AgriconsultJAMEC (1992). TABLE 4C: Cashew Nut Kernel Imports into Major Markets (Tons) Country 1980-82 Average 1984-86 Average 1989-91 Average United States 30,917 43,073 48,372 (Former) USSR 20,816 3302 3328 Netherlands 3080 2301 3669 Germany 2796 2736 3661 Canada 2666 3235 4309 United Kingdom 2638 2934 4919 Japan 2371 2717 4520 Australia 1535 3011 2930 Total 66,819 63,309 75,708 SOURCE: Gill and Duffus Edible Nut Market Report, various issues. ... C.I Q TABLE 4D: Operations of Cashew Nut Processing Factories Est. Total # of Days Date of Source of Maximum Operated Through Factory Test Run Finance Capacity Annual Raw Nut Throughput (Tons) Sept. 1991 1985/86 1986/87 1987/88 1988/89 1989/90 1990/91 Tanita 1965 GOT 10,700 0 0 218 2455 3000 1345 >1000 Mtwara 1968 GOT+ES 6800 0 0 0 0 0 0 ? Lindi 1978 WB 8600 0 0 0 0 0 0 301 Mbagala 1978 ES 10,700 0 0 0 0 0 0 ? Mtama 1979 WB 4300 0 0 0 0 0 0 358 Likombe 1980 WB 8600 0 0 2402 3144 6000 1473 >800 Kibaha 1980 WB 8600 0 0 1383 2536 3500 1345 720 Nachingwea 1981 WB 4300 0 0 0 0 0 0 95 ~ Masasi 1981 WB 8600 0 0 0 0 0 0 91 :i: :> Newala II 1981 WB 8600 0 0 0 0 0 0 116 " < c: Tinduru Never ES 8600 0 0 0 0 0 0 0 :i: ..,. ~ Newalal Never WB 8600 0 0 0 0 0 0 0 §; (J SOURCES: Agriconsult/Amec (1992); MDB Annual Cashewnut Report, various years. ::t c: :i: .:: )> r- c: " ..,. c c c: (J v. Perishable Profits: Private Sector Dairy Processing and Marketing in Kenya Steven Jaffee D airy products account for approximately 4% of the agricultural GDP of sub-Saharan Africa and 11% of the total value of food production in the region. 1 Dairy products provide an estimated 2% of calories and 4% of protein in the average diet in SSA, and these shares are considerably higher in the region's main producing countries.2 Primarily for agro-climatic reasons, SSA is a large net importer of dairy products. In fact, for much of the past decade, the value of SSA's dairy imports exceeded the region's exports of all indi- vidual agricultural products other than coffee, cocoa and sugar.3 The histori- cal experience from many countries points to the importance of the dairy industry in agro-industrial development and in the structural transformation of agriculture. This has certainly been true in Kenya since Independence. This chapter focuses on dairy processing and marketing in Kenya. Kenya is one of SSA's largest milk producers and has developed one of the region's most commercialized dairy industries. The industry has long been dominated by the private sector, unlike the situation in Zimbabwe and South Africa which have developed Africa's two other commercially advanced dairy indus- tries. The Kenyan experience is illuminating for two reasons. First, much of the history of the industry has centered around the development of a large private cooperative which per~istently sought to utilize the powers of the 1 Mbogoh (1984); Walshe et al. (1991). 2 Walshe et al. (1991). See FAO Food Consumption Patterns for dietary profiles of individual countries. 3 FAQ Trade Yearbooks, various issues. 199 200 MARKETING AFRICA'S HIGH-VALUE FOODS state to protect its interests and minimize competition. In what amounts to an ironic historical twist, over the past decade this same cooperative has essentially been 'publicized' and now serves as a de facto parastatal. A second interesting feature of the sub-sector has been the attempts by other private firms and cooperatives to develop dairy processing operations on the geo- graphical and competitive margins permitted by government and the domi- nant cooperative. With recent changes in governmental competition and pric- ing policies within the industry, these private and cooperative activities are expected to expand and be replicated in many parts of the country. Thus, the Kenyan experience can provide insights into the costs of market controls and the likely degree and forms of private sector response to market liberalization. The chapter·has three sections. The first section examines major techno- economic characteristics of dairy products and production, noting possible implications for entry into dairy processing and marketing and the forms this may take. The second part provides a concise overview of the scope and fea- tures of dairy production and markets in sub-Saharan Africa. The last part examines the Kenyan experience. It discusses the historical development of the industry and summarizes current patterns of operation by private enter- prise engaged in dairy processing. MAJOR TECHNO-ECONOMIC CHARACTERISTICS OF MILK AND DAIRY PRODUCTS There are several techno-economic features of milk, milk production and dairy processing which generate risks as well as opportunities for dairy producers and marketing enterprises and which might be expected to influence the industrial organization of dairy commodity systems in Africa.4 Among the inherent properties of raw milk itself are the following: 1) High perishability: Since milk provides a good medium for micro-bio- logical development, quality deterioration is rapid unless the raw milk is cooled or processed within a few hours of 'harvesting'. Raw milk's · perishability minimizes the scope for storage and places a premium on the development of a reliable milk collection and cooling system and on investments in equipment and processes which convert the raw prod- uct into less perishable forms. 4 This discussion is based in part on Hopcraft (1978), Manchester (1983), Brown (1991), and Walshe et al. (1991), and on helpful comments by Cees de Haan. ·PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 201 2) High bulkiness: Since milk is 87% water, its valuable components (fat, proteins and casein) constitute only a small share. Together with the perishability factor, this pr~sents major logistical problems. Raw milk carries relatively high transport costs per unit value, reducing the scope for economical long-distance trade and leading toward localized supply and distribution arrangements. When producing areas are distant from major markets, it may be profitable to convert the raw product into less bulky forms by separating the water from the valuable components. 3) Wide quality variability in terms of fat content, micro-biological qual- ity and the presence of diseases and of antibiotic residues: Quality varia- tions stem from differences or changes in animal breeds, climate, feed availability, animal health and the hygienic conditions during milking. Milk can also be adulterated by producers and traders. The quality of raw milk delivered by producers and traders can have major public health implications and influence the economics and product quality of dairy processing. This implies a need for quality standards and moni- toring measures, as well as potentially significant benefits from mea- sures to support animal husbandry and health at the farm level.5 In addition, the production of milk at the farm level has several generic properties which might be expected to influence the market structure for dairy products. These include: 4) Relatively high entry and exit costs: Lumpy investments must be made in specialized assets (grade cows, milk cans and perhaps milking ma- chines and buildings) whose alternative uses yield far lower returns (for example, culling grade cows for meat). Even at the smallest scale, such as one or two milking cows, dairying entails a relatively signifi- cant fixed cost element and must be regarded as a long-term invest- ment, Thus, smallholder entry will require credit facilities or signifi- cant savings from other activities. Prospective producers must consider · the overall reliability of livestock support and dairy marketing services and the long-term remuneration of dairy market outlets. 5) Relatively low short-term production elasticity: Due to multi-year biological lags, relatively small adjustments in output can be made by 6 While quality standards have public good properties, measures to improve milk quality (for in- stance, through production support) may be associated with externalities and the service supplier being unable to appropriate the benefits from the activity. See Umali, Feder and de Haan (1992). · 202 MARKETING AFRICA'S HIGH-VALUE FOODS changing the feeding regime. More significant output changes require either adding or culling cows. Significant production lags might be ex- pected following changes in prices or other incentives. However, where only a small proportion of milk output is marketed, changes in the incentive system could result in a more rapid and significant supply response as shifts are made between household consumption and mar- keted supply. 6) Continuous product harvest: Cows produce milk throughout the year rather than during a limited season of a few weeks or months.6 While this yielding pattern has the attraction of providing a continuous source of income for the producer, it requires that the milk collection and mar- keting system also be continuous and reliable. Otherwise, much milk will be wasted or the producer will be forced to utilize greater quanti- ties than preferred. 7) Seasonal variability in output: This can lead to surpluses and short- ages at the local and/or national levels with heavy pressure placed on transport, processing and other infrastructure during the 'flush season' and underutilization of such infrastructure during the 'dry season'. Be- cause the demand for liquid milk has seasonal patterns which may not match those of production, the result can be wide swings in market prices from one season to the next. Production seasonality implies a need to offer incentives to producers to increase dry season production · (through use of manufactured feeds, the timing of calving, etc.) and a need for processors to make special arrangements to procure sufficient raw materials during the season. Inherent trade-offs are faced between using milk for calf-rearing versus human use. Finally, the processing, commercialization and demand for dairy products have distinctive features which might be expected to influence dairy system organizational patterns. 8) Liquid milk can be consumed in its raw harvested form or pre- served through pasteurization or fermentation. Raw milk provides the cheapest source of milk-based nutrients and its direct supply to consumers avoids most of the coordination problems associated with other dairy processing and distribution modes. Raw milk consumption may entail health hazards although boiling the milk can destroy all 6 In Kenya, harvesting is typically done two times per day-in the morning and evening. PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 203 pathogens. Raw milk can be treated and preserved for a few days through pasteurization and preserved for a longer period in liquid form either through exposure to ultra-high temperatures (UHT) and aseptic packaging or through fermentation. The costs of these pro- cesses and associated packaging vary significantly. This multiplicity of potential milk products offers considerable flexibility for dairy systems to adapt to local demand, infrastructure and technological conditions. 9) Raw milk can be converted into numerous processed products, in- cluding soft and hard cheeses, yogurt-based products, butter and but- ter oil (ghee), cream and skimmed milk, and powdered milk. The pri- mary ingredients of raw milk can be disassembled and then recombined in different ways or mixed with other raw materials to yield different products. This again offers considerable flexibility in processing and marketing operations. Because of this technical as well as financial complementarity among certain products (that is, cream and skimmed milk, butter and butter oil, cheeses of different rates of maturity), one would not expect to find processors with single-product specialization. 10) There are many technological options for the preservation or processing of most dairy products. Both batch and continuous processes are available that utilize different degrees of mechanization and different (efficient) operating scales. Some products can be effi- ciently treated in rural areas on a small-scale basis using relatively simple non or semi-mechanized technology. This applies to fermented milk, butter, cream and most types of cheeses and yogurt products. Relatively little asset-specificity is involved-most of the equipment and inputs used (boilers, fuelwood, vats, refrigerators) can be used for other purposes. For these products, technical and financial barriers to entry are not expected to be significant, and subject to demand, many small producers are likely to be active. Other products require more complex processes and higher invest- ments in plant and equipment. Milk pasteurization can be done on a relatively small scale, although some economies of scale are realized in (semi-)mechanized continuous heating and cooling systems, and 'lumpy' investments in c~ld stores are required. Still, both pasteuriz- ers and cold stores can be put to alternative uses (for example juice- making), limiting the degree of asset specificity. Where there is suffi- 204 MARKETING AFRICA'S HIGH-VALUE FOODS cient market demand (along with or instead of raw milk), one would expect to find several small-to-medium-scale milk pasteurizers and distributors serving each market center or urban area. The production (and packaging) ofUHT milk and evaporated pow- dered milk require more sophisticated, capital-intensive techniques and more specialized skills. The production of these products is associ- ated with greater economies of scale and higher degrees of asset speci- ficity. Due to its high-cost packaging, UHT milk might be expected to encounter relatively limited local demand in Africa in spite of the absence of household refrigerators. For these commodities one would expect relatively few, larger-scale producers. 11) Sources and patterns of demand will vary significantly among prod- ucts with market segmentation being common. The highest valued ingredient in milk is fat. In Africa, products with a very high milk fat content are likely to face a relatively limited market, restricted prima- rily to middle-to-upper-income groups and tourists. In contrast, raw milk is likely to be the predominant dairy product sold in milk-pro- ducing rural areas for reasons of both logistics and purchasing power. DAIRY PRODUCTION AND MARKETS IN AFRICA The livestock sub-sector plays an important role in the economies of many sub-Saharan African countries. For the region as a whole, the sub-sector ac- counts for about 25% of agricultural GDP. This share exceeds 40% in several countries. 7 For the region as a whole, dairy products account for an estimated 20-25% of the value of livestock sector output.8 Dairy production and processing in Africa is based upon cattle, camel and small ruminant production. Dairy production data for the region should be regarded as 'guess-timates' since they are based on estimated herd sizes and very rough indicators of average yields. Data are especially tentative for camel and small ruminant milk production because this production occurs largely in poorly monitored arid and semi-arid areas, and in many cases, is primarily for subsistence purposes. Based on FAQ data for cow and small ruminant milk production and on specialists' estimates of camel milk production, sub- 7 Winrock International (1992). 8 Walshe et al. (1991). PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 205 Saharan Africa's 1988 milk output of 15.2 million metric tons consisted of 9.62 million metric tons of cow milk (63%), 2.94 million metric tons of small ruminant milk (19%), and 2.68 million metric tons of camel milk (18%).9 Cow's milk constitutes the bulk of local production in the vast majority of countries. Only in five countries (Somalia, Chad, Mali, Mauritania, and Niger) does estimated small ruminant and camel milk production exceed local cow milk production. 10 Sub-Saharan Africa's production of cow milk in 1991 was 11.9 million metric tons, representing 2.6% of world production.11 Due to agroclimatic and other reasons, this production is heavily concentrated in relatively few South- ern and East African countries (Table 5.1). The three largest producers- South Africa, Sudan and Kenya-accounted for nearly 59% of the region's total reported cow milk production in 1991. Annual cow milk production on a per capita basis ranges from less than one kilogram for many West and Cen- tral African countries to nearly 100 kgs. for Kenya. While reported cow milk production in sub-Saharan Africa grew rapidly during the 1960s, since the mid-1970s production has barely kept pace with average annual population growth of 3.1%. 12 Most of this expansion in pro- duction can be attributed to increased herd size, rather than improved pro- TABLE 5.1: Concentration of SSA's Reported Cow Milk Production Production Share of SSA Total Cumulative Share Country ('000 Metric Tons) (%) of SSA Total (%) South Africa 2515 21.l 21.1 Sudan 2299 19.3 40.4 Kenya 2189 18.3 58.7 Ethiopia 752 6.7 65.4 Somalia 497 4.2 69.6 Madagascar 472 4.0 73.6 Tanzania 463 3.9 77.5 Uganda 437 3.7 81.2 SOURCE: FAO Production Yearbook (1991). 9 This calculation is based on Walshe et al. (1991, Annex 2) plus data on South African cow milk production from the FAO Production Yearbook (1990). 10 Ibid. Sudan and Somalia account for about three-fourths of SSA small ruminant and camel milk production. 11 This total SSA production is similar to the milk output of the Netherlands. 12 This covers the period from 1975 to 1987. 206 MARKETING AFRICA'S HIGH-VALUE FOODS ductivity. 13 With the exception of a few areas and production systems, milk production is based predominantly on low-yielding, yet disease-resistant, in- digenous zebu stock. Excluding South Africa, the average annual cow milk yield in SSA is estimated at 353 kgs. per cow, compared to an average of 820 kgs. per cow for all developing countries as a group.14 The available data suggest that regional per capita dairy product con- sumption has remained steady since the early 1970s at slightly below 30 kgs. per annum. Most of this consumption-estimated at 90o/o-is of liquid milk, either fresh or in sour or fermented forms. Sour or fermented milk consump- tion is especially prominent in rural areas. Evidence from some countries indicates that most fresh milk is consumed in te~ or in cooked meals rather than drank separately. 15 The remainder of dairy consumption is in the form of cheese, butter oil, butter and yogurt-based products. While FAO data for cheese and butter suggest stagnant production within the region over the past two decades, such data is highly questionable given that most produc- tion and consumption occur in rural areas and do not enter formal marketing channels. Such production is unlikely to be captured in national statistics. Even though much of the available dairy production and consumption data may be underestimated, it is clear is that sub-Saharan Africa has be- come a large net importer of dairy products over the past two decades. Com- mercial imports of dairy products rose in value from $113 million in 1970 to $705 million in 1980. Although imports declined during the 1980s, they re- mained at $613 million in 1989.16 Over this period, expanding quantities of skim milk powder and butter oil (for recombination to form liquid milk) were imported into Africa as part of food aid programs of the United States, the EEC and the World Food Program. In recent years, commercial and donated imports have accounted for 20-25% of total estimated dairy product consump- tion in the region. 17 13 Over the 1974-1989 period, the number of cows in SSA increased by 56% from 14.88 million to 23.24 million. Average milk yields increased only by 15. 7% from 305 kg./head to 353 kg./head. (Shapiro et al. (1990, p. 50). 14 FAO Production Yearbook (1989). Average yields in South Africa have exceeded 2500 kg./head in recent years. 15 Walshe et al. (1991, p. 9) refer to 70% in Zimbabwe. 16 Of which $499 million was of condensed/dry/fresh milk, $69.6 million was of butter, and $44.6 million was of cheese and curd (FAQ Trade Yearbook 1989). 17 Imports have been especially important in West Africa, where they accounted for 40-50% of total regional consumption during the mid-to-late-1980s. See Shapouri and Rosen (1990) for a country-by- country analysis of the importance of and growth in dairy imports. PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 207 Several factors contributed to the rapid growth in SSA dairy imports dur- ing the 1970s and their maintenance at significant levels in more recent years. On the supply side, price support and other incentive programs con- tributed to the development oflarge dairy product surpluses in North America and within the EEC. In the case of the United States, large quantities of surplus production were made available for bilateral and multilateral food aid programs. For the EEC, subsidies were provided which enabled produc- ers to export at prices well below those of domestic prices. On the demand side, imports were stimulated by (1) overvalued exchange rates which made commercial imports even less expensive, (2) donor assistance in building large- scale milk reconstituting factories, and (3) rapid urbanization rates in some countries. Dairy imports into SSA have come predominantly from outside of the region. Only Kenya, Zimbabwe and South Africa have undertaken lim- ited exports to other African countries. Dairy systems in sub-Saharan Africa· are characterized by dualistic pro- duction and marketing structures. Most production, trade and consumption occurs within traditional or informal ·systems. Here, production is derived from indigenous breeds of animals and either consumed by the producing household or sold locally. Production, processing and trade are generally con- ducted on a very small scale. Although constituting the bulk of dairy activity in the region, these traditional or informal systems are poorly documented and not well understood by outside analysts.· Modern or formal dairy sub-systems account for less than 10% of the milk production and flow in most SSA countries. This rises. to slightly more than 25% in Kenya and over 50% in Zimbabwe and South Africa. These systems are based on grade or pure bred cattle, have a significant or predominant large farm component, and historically have featured dominant or mono- psonistic national cooperatives and parastatals serving the urban markets. In many parts of East and Southern Africa, during the colonial period these formal systems were controlled by European settlers. In the period since in- dependence; large state farms have played major roles in the formal sub- systems in both these countries and several others. Past· efforts to improve dairy product supplies and processing have in- volved the construction oflarge-scale processing and milk reconstituting plants to meet urban demand for liquid milk.18 These efforts were influenced by the 18 This discussion is based on Walshe et al. (1991), pp. 58-68. 208 MARKETING AFRICA'S HIGH-VALUE FOODS availability of cheap or free imports of skim milk powder and butter oil and by an initial emphasis (particularly by UNICEF) in supplying pasteurized milk to children. Many of these investments were financed by donors whose technical advisors were most familiar with capital-intensive dairy systems. These ventures have encountered substantial problems in developing effi- cient milk collection and dairy marketing systems. Some of these problems have stemmed from the general features of African production, infrastructure and demand including high ambient temperatures, highly seasonal rainfall (and hence production) patterns, low milk production densities, poorly devel- oped and poorly maintained transportation systems, low per capita incomes, and relatively low rates of urbanization. Milk production densities in Africa are illustrated in Table 5.2. These patterns call into question the advisability of centralized, 'large-scale processing facilities, at least those based on a local production base. Many of these dairy investments have also suffered from major institu- tional problems. Parastatal enterprises have operated most of the large-scale dairy processing facilities in Africa. They have typically experienced major managerial problems and have had little incentive to perform efficiently.19 In many cases, they have been granted exclusive milk collection and marketing rights, thus limiting·competitive pressures. Weak livestock support services, together with government pricing policies that have provided only minimal incentive for farmers to supply formal marketing channels and left processors and distributors with inadequate operating margins, have further inhibited dairy development. Dairy investment schemes have commonly resulted in low rates of factory capacity utilization, large financial losses and low levels of development of local milk supplies. Africa's dairy potential remains far from being effectively realized. At the same time, a long-term strategy of meeting increased demand through im- ports will become increasingly more costly. Adjustments in dairy policy in North America and the EEC have already contributed to significant reduc- tion in production surpluses and in the availability of dairy products on concessional or food aid terms. 20 World prices for non-fat dry milk powder have increased from $600-700/ton in the early-to-mid-1980s to $1750-2000/ ton in 1989/90. 19 See Nellis (1986). 20 Donated dairy products declined from one million tons (liquid milk equivalent) in 1985 to 315,000 tons in 1988. (Shapiro et al. 1990, p. 52). PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 209 TABLE 5.2: Comparative Milk Production Densities (kgs./Km2 per day) African Patterns Comparative Patterns East Africa 4.2 Netherlands 930 West Africa 0.7 Switzerland 256 Southern Africa 0.5 Ireland 247 Central Africa 0.2 New Zealand 82 Total SSA 1.5 India 28 SOURCE: Nell (1990), p. 37. Among African governments and donor agencies, there has been a grow- ing recognition of the costs and retarding effects of monopsonistic/monopolis- tic milk procurement and marketing systems and reliance upon centralized, large-scale processing facilities to meet diverse and expanding demand. The need to focus more attention on building up local production and supporting small-to-medium scale processing activities in rural areas and market towns has also been emphasized. In several countries, including Zimbabwe and Ni- geria, there have been moves to privatize dairy enterprises and liberalize dairy marketing systems. The Kenya dairy industry, the primary focus of this chapter, is in the midst of a transition away from a centralized, near monopolistic formal mar- keting system toward one featuring multiple participants and different mod- els of milk procurement and dairy product distribution. The Kenyan experi- ence is interesting because the firm which has dominated historically has remained, at least in its formal legal status, a private commercial enterprise while operating as a de facto parastatal since the late 1970s. Competing pri- vate and cooperative processors have been forced to operate on the geographi- cal and consumer margins of the formal dairy marketing system due to gov- ernment restrictions and anti-competitive behavior by the state-sponsored firm. The ongoing market liberalization will lead to significant structural changes as well as a likely reversal in the sub-sector's recent decline. THE KENYA DAIRY SUB-SECTOR: BACKGROUND AND PRIVATE SECTOR DEVELOPMENT Livestock production plays an important role in the Kenyan economy. In 1990, sales of livestock products, including meat, milk and hides accounted for 26% of agricultural GDP and nearly 7% of total GDP. Livestock produc- 210 MARKETING AFRICA'S HIGH-VALUE FOODS tion is especially important for Kenya's smallholder and pastoral populations as a source of food, income and employment, and as a form of savings.21 The production and sale of dairy products have long comprised one of the most important segments of Kenya's livestock sector. Since independence, the dairy sub-sector has continued to account for some 30% of livestock product sales through formal marketing channels. The very large informal trade in dairy products adds to the sub-sector's importance. In this section we provide an overview of the origins and evolution of commercial dairy production and marketing in Kenya. After summarizing current patterns of dairy production, consumption and distribution, we move to an examination of the background, operating characteristics .and perfor- mance of private and cooperative dairy processors. A Short History of the Development of the Dairy Sub-sector The raising of cattle has traditionally been an important production and in- vestment activity for many of Kenya's indigenous communities.22 In these communities, milk has occupied an important place in the daily diet. While milk has been exchanged at the village level for many years, the indigenous zebu cattle generally provided little surplus milk over and above household requirements. Production was for subsistence rather than market purposes. The development of commercial dairy production and trade in Kenya can be divided into five historical stages, namely: 1) the origins of export-oriented butter production (1900-1930s), 2) discovering the domestic milk market (WWII-mid-1950s), 3) dualistic development (late 1950s-1970), 4) market consolidation and the 'publicization' of the private sector (1971- early 1980s), 5) high cost expansion and creeping liberalization (mid-1980s-present). The main developments of each of these stages are summarized here.23 21 According to Kenya's Integrated Rural Survey (1), during the mid-1970s sales of livestock prod- ucts accounted for 47% of smallholder agricultural sales earnings nationwide, Only in two Prov- inces-Nyanza and Eastern-were sales of crops more important than livestock product sales. 22 Heyer (1976) discusses the traditional role oflivestock production in Kenyan rural communities. 23 This summary draws upon the work of Hill (1956), Klemm (1966), Hopcraft (1976), Raikes (1981), Minae (1981), Stoltz (1979), and DANIDA (1991). Readers interested in a more in-depth review of the historical development of the sub-sector should consult these sources. PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 211 The Origins of Export-oriented Butter Production Commercial dairy production in Kenya originated with the import of pedigree dairy stock from Europe at the beginning of this century. These animals were cross-bred with indigenous stock on European settler farms and research and breeding stations.24 Commercial butter production started in 1901. By 1908, a private cooperative called the Lumbwa Creamery was exporting the product to Uganda, and other small dairy processing units followed. During World War I the emergent industry was called into service and the entire Veteri- nary Department was called up. The government's breeding program was interrupted and large numbers of cattle were slaughtered to provide meat to the army. · Following the war, commercial dairy stock was rebuilt and butter produc- tion and trade grew steadily. The Lumbwa Creamery expanded production and many smaller makers of cheese, cream and butter emerged. Construction of a large cold store at Kilindi port by the government made possible exports of cheese and butter to London. In 1925, a group of large European farmers in the N aivasha and N akuru areas formed the Kenya Cooperative Creamery to produce cheese, share technical information and pool sales revenues. An- other group of European settler farmers formed the N anyucki Cooperative in 1928. Periodic efforts by these three main cooperatives to set common prices and quality standards and to undertake joint export sales campaigns were not successful. The Depression significantly reduced international butter and cheese prices and led to losses for Kenya's cooperatives and European dairy farmers. In order to protect their investments and markets, the three main dairy coop- eratives merged in 1:931 under the banner of the Kenya Cooperative Cream- ery (KCC). A year later, the KCC was the first company to register itself as a cooperative under the Co-operative Societies Registration Ordinance. In or- der to counter the effect of declining world prices, the KCC successfully lob- bied the government to apply a levy on domestic sales of butter that would be used as a subsidy on exported butter.25 Since two-thirds of KCC's sales were sales abroad, it was the smaller private producers which bore most of the cost of the subsidy. 24 Much of this discussion on the early development of commercial dairying is drawn from Hill (1956). 25 This was instituted under the Butter Levy Ordinance (1931). 212 MARKETING AFRICA'S HIGH-VALUE FOODS Although KCC's membership, processing capacity, production and sales all grew steadily during the 1930s, the organization continued to face compe- tition from smaller dairies and private milk dealers which sought supplies from members of the cooperative. These conditions, on top of widespread eva- sion of the Butter Ordinance, led the KCC to advocate the creation of a Dairy Control Board which would pool the marketing of all dairy products under a single direction. Underlying the proposed scheme was the KCC's assumption that it would be appointed as the board's sole marketing agent and that other private dairies would either be closed or merged into the KCC operation. Although the government did not create such a board, it did provide the KCC protection against competitive imports and restricted potential competition from African farmers. 26 KCC's advocacy of a Dairy Control Board marked the beginning of more than a half-century of efforts by the firm to use the legal powers of the state to create and maintain exclusive marketing rights within the domestic dairy prod1:1cts market. Discovering the Domestic Market During World War II, the Kenyan government and the British Ministry of Food contracted with the KCC to provide large supplies of butter and cheese. It was in serving the military forces and agencies responsible for feeding locally-held prisoners of war that the bulk ofKCC's sales were directed at the local market for the first time. At the same time, rapid wartime population growth in Nairobi and Mombasa resulted in greater local demand for whole milk which the KCC and other firms sought to supply. During the late-1940s and early 1950s, the commercial dairy sub-sector prospered in spite of periodic droughts, termination of the government bulk commodity purchase scheme, and political and security problems in parts of the highland areas. In 1950, the KCC began to market fresh milk on a pilot basis in Nairobi and Sotik. This proved profitable and milk marketing activi- ties were extended to other locations. As in the 1930s, the KCC made another attempt to monopolize this expanding market. 27 The general manager of 26 African farmers were discouraged by the Veterinary Department from acquiring grade cattle. Technical support to African livestock owners was limited to disease control programs. The only support for milk marketing was a scheme at the coast, distant from any KCC procurement or sales activity. 27 After a feasibility study determined that it would not be profitable for the KCC to construct a large milk distribution depot in Nairobi, the organization proceeded to lobby the government to pre- vent any other firm or cooperative from making such an investment. This episode is discussed by Hill (1956). PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 213 Britain's Milk Marketing Board was invited by the KCC to advise it and the Kenyan Government on measures to improve dairy marketing. Not surpris- ingly, he recommended compulsory control over milk sales with a single sales agent-the KCC. Although the government again rejected the scheme for monopolization of milk marketing, it did organize a voluntary milk marketing program involv- ing the KCC and an association of other European private and cooperative dairy producers and traders. 28 A Milk Committee was created which would issue contracts to producers and firms to supply the Nairobi market. The Committee guaranteed minimum producer prices and channeled unsold milk to factories for dairy product manufacturing. The Committee would later ex- tend its activities to other markets. While the KCC continued to face competition from several local and re- gional suppliers, it retained its dominant market position. In 1954, it ac- counted for 77% of milk production by settlers from Europe and 90% of the European milk suppliers were members. The membership of the KCC contin- ued to be exclusively Europeans. Expanding milk and ghee production by African farmers still did not enter formal marketing channels other than the government's milk program on the eastern coast and a few other isolated schemes for ghee marketing. Dualistic Development The period from the mid-1950s to 1970 witnessed rapid and steady develop- ment of smallholder dairy production and its initial incorporation into the formal dairy marketing system. During this period, the KCC continued its efforts. to expand its operations and consolidate its dominant market position. Even though small farmers would come to account for about two-thirds of total national milk production, large farmers continued to form the backbone of this formal marketing system due to the discriminatory milk procurement and pricing policies employed by the KCC. The expansion of smallholder dairy production stemmed from several sources. One set of stimuli was a series of government- and donor-financed programs following the development of the Swynnerton Plan and in the years just prior to and after Kenya's independence in 1963. Some of these pro- grams, including the UNICEF-supported Rural Dairy Development Project 28 This was the Kenya Dairy Cooperative Association. 214 MARKETING AFRICA'S HIGH-VALUE FOODS and the USAID-supported Dairy Training Center, were geared specifically toward assisting smallholder dairy producers. Other initiatives, including a land titling program, the Smallholder Credit Project, a massive land transfer program from European to African owners and various settlement schemes, contributed to smallholder dairy development. Smallholder dairying further benefited from the rapid growth in small- holder tea, coffee and pyrethrum production which took place in Central and Rift Valley Provinces during the 1950s and particularly the 1960s. This ex- panded cash crop production not only generated increased demand for milk (and other dairy products) in rural areas, it also provided the capital for smallholder purchases of grade cows.29 Because most smallholder producers had only one or very few cows and the expansion of smallholder production occurred simultaneously across many districts, the issue arose of how to develop a commercially viable milk collec- tion system for such farmers. Parallelling concurrent developments in the coffee and pyrethrum sub-sectors, local dairy cooperative societies were formed. The first of these cooperatives was established in the mid-1950s. A decade later there were 125 of them with tens of thousands of members. Cooperative development benefited from infrastructure investments (especially in rural roads) and from a UNICEF-funded program which supplied milk cooling and other equipment. In the late 1950s, the KCC began to procure, process and market milk provided by smallholder cooperatives. The portion of its milk supplies from cooperatives would remain very small, accounting for only 5% of the total volume procured in 1960.30 By 1965, when smallholder and settlement farm- ers accounted for some two-thirds of total production, only 22% of KCC's supplies came from these farmers. 31 Perhaps as much as 75% of smallholder production was consumed by the farming households who produced it and much of the marketed milk was sold locally through informal channels.32 Several factors contributed to the limited incorporation of smallholders into the formal dairy marketing system. First, KCC's procurement arrange- 29 This point is made by Raikes (1981) and Bates (1990). Some 80-90% of grade cow purchases by smallholders during this period were made with savings from off-farm employment and cash crop production. See Stolz (1979). 30 World Bank (1963). 3 1 Klemm (1966), p. 10. 32 Hopcraft (1976), p. 2; Brown (1963). PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 215 ments for smallholder milk were largely passive rather than active.33 The firm played no part in ongoing government technical support programs, pro- vided no market information to smallholders, and left milk collection and financial matters to the cooperative societies. The KCC did not in fact have a presence in rural areas. The geographical scope of KCC's procurement and processing operations was limited to the central highland areas, leading farm~ ers from other areas to face enormous transport bottlenecks and costs when supplying KCC plants. Second, the system of procurement quotas and prices applied by KCC during this period offered little financial incentive for smallholders to supply the organization. In 1954, the KCC replaced the former contract system with a three-tier pricing structure which offered premium prices year round to those suppliers which could meet agreed upon dry season quotas ('quota milk'), next highest prices for 'contract milk' which would be used to produce manu- factured products, and relatively low prices for 'separation milk' which would be used to produce cream and skimmed milk. In a typical year, prices paid for 'quota' and 'contract' milk e4ceeded those for 'separation' milk by 100% and 66%, respectively. Quota contracts were initially offered exclusively to large European farmers, based on their past deliveries and their access to irriga- tion facilities to meet dry season requirements. By the mid-1960s, these farm- ers still accounted for 70% of quota milk deliveries, whereas smallholders , were largely paid at the lower price scales.34 This is not surprising given that KCC's membership and management continued to be comprised mostly of European farmers. AB the KCC continued to consolidate its market dominance, smallholder farmers faced a dwindling range of alternative channels to market milk· in urban areas. During the mid-to-late-1950s, the KCC built several new facto- ries, expanded its product line (to include pasteurized cheddar cheese, ice cream, skim and whole· milk powder, condensed milk, casein, ghee and but- ter), and was the first firm outside of Europe to adopt the Tetra Pak technol- ogy to extend product shelf life. The organization's in-take of milk and butter fat, stagnant at a level between 125,000 and 150,000 liters/year between 1955 and 1963, rose to 200,000 liters/year by the end of the 1960s. At that time the firm had an 88% share of the formal market for milk and butter. Throughout this period, the KCC operated at a profit. 33 Minae (1981), op cit. 34 Klemm (1966), p. 19. 216 MARKETING AFRICA'S HIGH-VALUE FOODS In 1958, in the midst of a downturn in world dairy prices, and at the instigation of the KCC, a Kenya Dairy Board was established. The Board had a broad mandate to stabilize prices, improve milk quality, conduct market research, and implement regulations and support services which would aid the development of dairy production and marketing. From the beginning, the KCC dominated the Board, which relied on the cooperative organization's tax payments for most of its operating funds. The KCC used the Board's regula- tory powers to restrict the activities of real and potential competitors. By the mid-1960s, the Dairy Board had withdrawn the licenses to sell milk in major urban areas independently of the KCC of all producers, cooperatives and retailers. Henceforth, such organizations would be permitted to sell milk only in smaller townships and rural areas. Market Consolidation and the 'Publicization' of the KCC In 1970, an official_;Dairy Working Party was set up to examine the state of the dairy industry, including issues related to competition and pricing. In the midst of spirited discussions among the Working Party members, the KCC announced the abolition of independent dairies and its monopolization of the urban dairy market. The KCC argued that it could provide the services of the independent dairies at lower cost and that its control of the market would lead to full compliance with product quality standards. The Dairy Board ac- quiesced to this intervention, which was one of its last actions before the Board's management was dissolved and its functions absorbed by the Minis- try of Agriculture. A year later, the three-tiered milk quota and pricing system was abol- ished. It was replaced by a system of uniform producer prices set by the government, with no differentials by season or location. The government also set consumer prices for pasteurized milk. Under the new rules, the KCC would be obligated to accept all milk delivered to its factories and depots, regardless of its ability to market the liquid milk. In effect, the KCC would become the 'buyer of last resort' for farmers who first sought potentially more remunerative local milk sales. As compensation, the KCC was given virtual veto power over the entry of competing dairy manufacturers and control over competitors' raw material procurement. Competing firms would be required to source milk in collaboration with the KCC and pay it an administrative fee for handling producer payments. Competing firms were restricted in the prod- ucts which they could handle and the locations where they could sell. PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 217 The creation of a guaranteed market outlet, the significant short-term price increase (for formerly non-quota producers), and favorable weather pat- terns led to a large increase in KCC's milk intake in 1972 and 1973.35 Com- pared to prior years, deliveries also showed' much greater seasonality. This surge in supplies surpassed KCC's processing capacity during the rainy sea- son and led to considerable waste, large inter-factory transfers of supplies and the need to convert \arge quantities of milk into storable products. For some of these storable products, including cheese and powdered milk, the KCC incurred large storage costs and sales losses. For the first time since the 1930s the firm recorded considerable financial losses. Thus, the requirement that the KCC purchase all (delivered) supplies at the higher official price brought about a financial crisis that was eased only by reduced production in 1974 and 1975 due to drought. During the mid-to-late-1970s, (estimated) national milk production grew at a slower pace. KCC's milk intake varied from year to year in line with rainfall patterns and remained below the peak 1972-1973 levels. Most of the incremental production during this period was used for smallholder house- hold consumption and informal rural market sales. By 1979, only one-half of all milk production was ever marketed. The KCC, holding a near monopoly on the formal dairy market, handled only 20% of national production. As Table 5.3 illustrates, by this time smallholder farmers had come to account for two-thirds of total production and 50% of marketed milk. However, be- cause only a small proportion (11%) of smallholder production was channeled to the KCC, larger- (and medium-)scale farmers remained the backbone of its operation. Large farms alone accounted for 68% of KCC's milk intake. By the late 1970s, the operational characteristics and financial structure of the KCC had changed so greatly that the organization could ~o longer be regarded as a private, commercially-oriented firm. This also occurred with several other national cooperatives during this same period. KCC's opera- tions were increasingly affected by political intrusions which influenced its appointment and retention of managers, its patterns of investment, the ser- vices it provided and the parties with whom it could contract for supplies and services.36 Despite incurring large financial losses, the KCC was required to 35For a detailed analysis of developments during the early to mid 1970s, see Hopcraft (1976). 36Other cooperatives and agricultural associations which were heavily politicized included the Kenya Farmers Association, the Horticultural Cooperative Union, and the Kenya Coffee Planters Union. I thank Shem Migot-Adholla for this point. 218 MARKETING AFRICA 's HIGH-VALUE FOODS purchase milk supplies in excess of its marketing needs for conversion into powdered milk which would be maintained as a strategic reserve. When the KCC became the sole supplier of a new school milk program, the government (that is, the Ministry of Education) became one of the largest buyers of KCC products. Huge financial losses totally eroded the KCC's equity base and ren- dered it completely dependent upon government-backed loans and subsidies.37 High Cost Expansion and Creeping Liberalization KCC's operations have undergone considerable expansion over the past de- cade. The firm has accumulated large financial losses as the returns to milk producers have declined and several private and cooperative enterprises have begun to chip away at its dominant market position. The KCC's expansion involved the extension and modernization of several factories, the construc- tion of about a dozen milk collection and cooling centers in or near small TABLE 5.3: Kenya Milk Production and Usage (1979) (Millions of Liters) Medium-Scale Farmers Large-Scale Smallholders (20-50 Acres) Farmers Total Total Production 769.0 192.8 195.6 1157.4 Of which consumed or fed to calves 436.4 62.7 14.8 513.9 (56.7%) (32.5%) (7.6%) (44.4%) Of which sold 295.4 114.0 177.8 587.2 (38.4%) (59.1%) (90.9%) (50.8%) Sold to KCC 87.7" 47.6b 160.1' 234.4 (11.4%) (24.7%) (81.2%) (20.3%) Other disposal 37.3 16.2 2.9 56.4 (4.9%) (8.4%) (1.5%) (4.9%) • Residual after calculating medium- and large-scale farm supplies. b As reported in Integrated Rural Survey (1976-79). ' Estimated at 90% of total sales. SOURCE: Integrated Rural Surveys, 1976-1979. 37 Over the 1977-79 period, cumulative losses totalled Ksh. 82.7 million. In 1978, the firm's ratio of net worth to net assets was -1.22 and its ratio of current assets to c~ent liabilities was 0. 76. Grosh (1987) provides data showing the KCC's dramatic financial deterioration during the 1970s. PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 219 towns in central Kenya, and the further extension of its product line. It was made possible by large, government-backed soft loans from bilateral donors. With the collection stations providing improved smallholder (and coopera- tive) access to the KCC, milk deliveries to the organization increased sub- stantially. From an average of 228 million liters/year over the 1980-1985 period, the KCC's milk intake surged to over 350 million liters/year in the late 1980s. Similar to what had happened in the early 1970s, this sudden expansion in milk purchases contributed to a financial crisis as the KCC accumulated large stocks of powdered milk which it sold at considerable loss. KCC's losses reached a cumulative level of nearly Ksh. 250 million by 1989. Mismanagement, declining .product quality, delayed or no payment for milk supplied to the school milk program and increases in unit processing and marketing costs above the margins provided by official prices also further eroded its profitability.38 While most government and KCC attention was directed at building up the milk marketing infrastructure, Kenya's dairy production base suffered. Since the mid-1980s, official producer prices have not kept pace with produc- tion costs and overall inflation rates (Figure 5.1). This has reduced the profit- ability. of dairy production and thus the available resources for animal main- tenance and feeding. Many large-scale farmers whose sales have been almost entirely directed through the KCC have responded to declining profitability by slaughtering or selling parts or all of their dairy herds. Between 1981 and 1989, the number of dairy cows on large-scale farms declined from 167,800 to 110,200, thus reducing the available supply of high-quality breeding stock.39 Over the same period, the quality of government livestock and dairy support services has declined significantly because most public research, extension and veterinary service institutions have increasingly suffered from shortages of funds and inadequate staff training and mobility. Although the government supported the expansion of the KCC, at the same time it has announced a policy of promoting rural dairy processing activities. This was signaled in the National Livestock Development Policy (1980), which stated that, "Cooperatives will be helped to assume the major milk collection, cooling and transportation functions in those areas where the 38 See the Dairy Master Plan, Technical Report #4 for an analysis of recent KCC operations and performance. 39 Statistical Abstract, 1990. 220 MARKETING AFRICA'S HIGH-VALUE FOODS FIGURE 5.1: Nominal and Real Producer Prices for Milk 5 .................. 4 . .. ..···\ ..·• • • • • • • • • • • • • • • • Nominal Producer Price 3 .... ---- ··~I 2 ~- ···········-----··· .... ...... / - ----- ~ Real Producer Price ~ ~~ ----- - ------- --- ~ ~ Purchasing Power Value** 1 0 I I I I I I I I I I I I 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 *Deflated by index of prices of material and service agricultural inputs *Deflated by index of purchased consumer goods in rural areas souRcE: Economic Survey, Various Issues KCC does not operate. Rural dairies will be developed and equipped with the necessary facilities." The Sessional Paper No. 4of1987 went further: The dairy cooperatives will be encouraged and assisted to establish rural small-scale dairy processing plants, particularly in surplus areas where KCC does not have processing factories. The paper recognizes the national role of KCC, but at the same time sees it necessary that KCC's development plans and those of rural dairies should be coordinated in order to harmonize milk collection, processing, and marketing for the benefit of farmers and the country. In practice, government support for rural dairies has been very limited and selective. Only two cooperative investments, each involving substantial donor grant or soft loan components, have received financial or other support from the government. Both involved the construction and operation of UHT processing plants in locations formerly not covered by KCC operations. In spite of the relatively small-scale of these two ventures, KCC attempted to subvert them by establishing milk collection and sales depots in nearby loca- tions to compete for the local supplies and markets, and by refusing to pur- PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 221 chase the surplus milk collected by these cooperatives. The governm:ent has provided little or no support to several other cooperatives and private firms which have sought to enter the market for processed fluid milk and other dairy products. License applications have been rejected or considerably de- layed while firms which have engaged in processing have had their range of products and their market areas restricted by the Dairy Board. Despite the growing financial burden of KCC's losses, the fact that many parts of the country are not served by the KCC infrastructure, and the appar- ent viability of small dairy processing operations even in the face of severe restrictions, the government has moved very slowly to liberalize the dairy market. Even after a high-level work group produced the Dairy Master Plan (1991)-which emphasized the inefficiencies of the present system and its unsustainability-reform measures were not introduced. Among the likely contributing factors for this inaction were: (1) conc~rns about the large sunk investments in KCC infrastructure and the need to repay government-guar- anteed external loans; (2) strong vested interests in the prevailing system involving KCC staff and firms providing inputs and services to the KCC, and (3) concern that liberalization would result in a reduction in government con- trol over the sub-sector while foreign or other non-indigenous firms emerged as new leaders. Initial market liberalization was forced upon the government in early 1992. Drought in major production areas, a KCC liquidity crisis which led to delays in payments for three to four months, and the inability of KCC to reconstitute large quantities of imported powdered milk due to a lack of butter oil resulted in severe shortages of milk in urban areas over the February-May period. KCC milk purchases declined by some 35% during the first half of the year, compared with the previous year. With the KCC unable to meet urban de- mand for milk, the government was forced to relax controls on private milk distribution. This relaxation allowed raw milk to be brought into Nairobi and other major cities. The formal milk collection and marketing system had completely broken down and most producers, processors, and distributors scrambled to obtain non-KCC sources and outlets for their milk. To·counter the disarray, the government removed its controls on producer and consumer prices (on June 2; 1992). The KCC promptly increased its pro- ducer price from Ksh. 5.20 to 6.50 per liter and its price to retailers for pas- teurized milk from Ksh. 9.50 to 13.10 per liter. Although the process of decon- trolling prices was taking place in a leg&l and regulatory vacuum, it provided 222 MARKETING AFRICA'S HIGH-VALUE FOODS additional stimulus to private traders, cooperative societies and large farms to buy, process and sell milk throughout the country. In the two months following the price decontrol, the Dairy Board received over one hundred applications from prospective processors and traders of dairy products. Nearly a year later (April 1993), very few of these applications had been acted upon. In the interim, multi-party political elections attracted most offi- cial attention and government decisions regarding the future direction of the dairy industry were placed in abeyance. Nevertheless, the price decontrol brought a significant response from informal sector traders of raw milk, who captured a major share of the urban market. Partially as a result of such raw milk sales, the KCC's sales of pasteurized and UHT milk in the Nairobi area fell from about 400,000 liters/day in early 1982 to less than 230,000 liters/day in early 1993.40 RECENT STATUS OF DAIRY PRODUCTION, CONSUMPTION, AND MARKETING Dairy Production and Farm Economics In 1990, Kenya had an estimated cattle population of 11.9 million. This herd consists of nearly 2.4 million high milk-yielding grade cattle (mostly cross- breeds of European breeds and indigenous zebu cattle) and some 9.5 million zebu cattle. While the indigenous cattle are heat tolerant and disease resis- tant, their milk yields are very low. Kenya has basically three dairy herds: 41 1) The pastoral herd is estimated at 4. 7 million head and composed en- tirely of indigenous stock. Milk is central to the pastoral diet. However, milk yields are less than 200 liters/year per lactating cow and output rarely enters the formal market. The pastoral herd accounts for an estimated 8% of total milk production. 2) The smallholder herd which numbers some 7.0 million head, of which 2.1 million head are grade cattle. Production systems vary from region to region and consist of open-, semi-zero, and zero-grazing feeding sys- tems involving different levels of technology and use of purchased 40 Staal and Shapiro (n.d.). 41 This summary is based on Price Waterhouse (1991) and information provided by Tetra Pak. PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 223 inputs. The smallholder herd accounts for an estimated 69% of total production. 3) The large-scale herd presently numbers some 220,000 head of pure bred and grade animals. This herd is maintained on private, corporate and state-owned farms with relatively high levels of management and ani- mal nutrition. Open-grazing supplemented by feed concentrates is the common feeding program. This herd now accounts for some 23% of total production. While accounting for only 20% of the total cattle population, grade cattle account for some 69% of total estimated milk production and virtually all of the marketed milk production. They are kept in the country's high-altitude, high-rainfall areas, principally in Rift Valley (47% of the total herd), Central (31%) and Eastern (11%) Provinces. According to available estimates, total Kenyan (cow) milk production has experienced a slow but steady increase since the early 1970s, with the excep- tion of the drought years of 1975, 1976, 1978 and 1984. This is illustrated in Figure 5.2. FIGURE 5.2: Estimated Kenyan Milk Production (Million Liters) ~ Q) 51300-1-~~~~~~~~~~~~~~~~~'--~~---~~~~~~ "' c ~1200-1-~~~~~~~~~~~~~~--=::;_~~~~~~~~~~~ ~ souRcE: Agricultural Growth Prospects and Strategy Options p. 70. Vol. 3. 224 MARKETING AFRICA'S HIGH-VALUE FOODS TABLE 5.4: Profitability of Milk Production (1989) (Ksh. per kg. of Milk) Official Local Payment Payment KCC (Informal) Production Production from from Producer Market Location System Cost Coop#l Coop #2 Price Price Nakuru L-S Open 2.86 3.63 4.00 S-S Open 2.48 2.90 3.04 3.63 4.00 S-S Semi-Zero 2.64 2.90 3.04 3.63 4.00 S-S Zero 3.50 2.90 3.04 3.63 4.00 Nyeri S-S Open 3.09 3.00 3.35 3.63 5.50 S-S Semi-Zero 3.58 3.00 3.35 3.63 5.50 S-S Zero 3.80 3.00 3.35 3.63 5.50 Kakamega S-S Open 3.00 3.20 Na. 3.63 7.00 S-S Zero 4.51 3.20 Na. 3.63 7.00 Costs per kg. for 1988/89 season (Sellen et al. 1990). Average producer payments from area cooperatives (1989) MOCD. Local prices paid by private buyers (1989) MOCD Survey. L-S Large-scale. S-S Smale-scale. Historically, dairy production in Kenya has been quite profitable, espe- cially for smallholders. However, the recent deterioration in livestock support services and decline in real producer prices paid by the KCC since the mid- 1980s have weakened production incentives. In several important production areas, payments made by the KCC and local cooperatives did not cover pro- duction costs (Table 5.4). Only farmer milk sales in local markets provide significant profits in these locations. In Nyeri and Kakamega, smallholder milk production generated little or no profit when milk was supplied to local cooperatives or sold directly to the KCC. Only local market sales through informal channels provided large profit margins. In Nakuru, large-scale open- grazing systems were minimally profitable for sales to the KCC, considering the added costs for transport and various taxes. Smallholder sales through local cooperatives similarly yielded relatively small profit margins. Formal market prices were inadequate to cover the additional labor and other costs incurred in production under zero-grazing. Table 5.5 illustrates that the returns to labor in smallholder milk produc- tion are generally below those for other crops and below the prevailing rural daily wage. These figures make clear that the sustainability of smallholder dairy production has required incentives over and above those provided by formal marketing channels. These added incentives have included: (1) house- hold consumption needs, (2) remunerative, informal (rural) market outlets PERISHABLE PRORTS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 225 TABLE 5.5: Returns to Labor from Milk Production and Competing Activities (1990 Prices) Milk Production Competing Uses Open-grazing in Rift Valley with herd of 5 improved Cows 20.8 Daily rural wage 20-25 Semi-zero grazing in Central and Eastern Provinces with herd of 2 cows 17.8 Tea 47.4 Open-grazing in Western Province with herd of 5 improved cows 17.3 Maize 22.6 Zero-grazing in Central and Eastern Provinces with herd of 2 cows 8.4 Coffee 20.0 SOURCE: Kenya Dairy Master Plan. for some part of production, (3) use. of by-products from dairy animals (for example, manure and hides), and (4) the use of cattle as a form of investment. Consumption, Demand and Prices Only rough estimates can be made about current levels of milk consumption and demand for marketed dairy products in Kenya. The most recent surveys recording information on milk consumption and expenditures date back to 1983 (Rural Household Budget Survey-1981/82 and Urban Budget House- hold Survey-1982/83). The authors of the Dairy Master Plan drew upon this information, updated population estimates and other data to project forward dairy demand to 1990. ~ According to the 1981 survey, out of total national milk production of 1.28 billion liters, about 52% (668 million liters) was consumed by the farmers who had produced it. Some 191 million liters were sold to the KCC and 290 million liters sold through other channels, with the remainder used to feed labor and livestock. In certain areas, especially in parts of Rift Valley, Cen- tral and Western Provinces, part of the household consumption was probably 'forced', since it is higher than what it would have been normally if adequate marketing infrastructure and market outlets had been available. The 1982 and 1983 surveys indicate per capita milk consumption of 64 kgs./yr. and 125 kgs./yr. in rural and urban areas, respectively. Table 5.6 provides the Dairy Master Plan's estimated distribution pattern of milk consumption in 1990. The table shows that while urban areas ac- counted for only 13% of the national population, they accounted for 24% of 226 MARKETING AFRICA'S HIGH-VALUE FOODS total milk consumption and 50% of the marketed milk. The higher urban consumption rates can be attributed to several factors, including higher dis- posal incomes, better developed food distribution infrastructure, greater ac- cess to refrigeration, and demand from the hotel and catering trades. Al- though 76% of milk is consumed in rural areas, less than one-third of this milk is purchased, either from formal or informal distribution channels. · The 1983 survey found that overall milk expenditure patterns were highly skewed. While the 58% of the urban population which comprises the lowest income category accounted for only 8% of milk expenditures, the highest in- come group, representing only 2% of the population, accounted for 45% of expenditures. High-income groups accounting for 19% of the population made 75% of milk expenditures. These patterns suggest that it has been the rela- tively wealthy who have benefited most from the government's controls on pasteurized milk prices. Milk expenditures in rural areas are much less skewed as a result of lower income disparities· and the fact that many of the more wealthy rural households own cows and thus do not need to purchase milk. Using survey data, one recent study calculated the income elasticity of milk expenditure to be 1.01 in rural areas and 0.58 in urban areas.42 Some 10-15% of milk (and other dairy product) consumption in Kenya occurs outside of the household. This distribution comes through various in- stitutions such as hospitals, the army, schools, hotels and restaurants/can- teens. The largest of these outlets is the School Milk Program. Some 38 mil- TABLE 5.6: Milk Consumption Patterns (1990) Per Capita Proportion Proportion Popu- Consump- Consumption of Total Con- Marketed Area lation tion (Liters) (Million Liters) sumption (%) Milk(%) Nairobi+ Mombasa 2.4 125.0 300.5 19.7 41.3 Other Urban 0.5 125.0 62.5 4.1 8.6 Rural 18.9 61.6 1165 76.2 of which: 365 (purchased) 23.9 50.1 800 (own production) 52.3 Total 21.8 70.1 1528 100.0 100.0 SOURCE: Based on Dairy Master Plan, Technical Report #1, Tables 1 and 3. 42 World Bank (1990). PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 227 lion liters of milk were distributed through this channel in 1990, although in the past two years delayed government payments (to the KCC) have under- mined the program. The hotel and catering_ trades have also absorbed in- creasing volumes of milk and other dairy products as Kenya's tourist indus- try has undergone a major expansion. Holiday and business visitors doubled during the 1980s from 336,000 to 696,000.43 It is tourists, together with expa- triate residents and Kenya's middle- and upper-income population, who are the primary consumers of cheese, yogurt and other high-value products. The bulk of milk consumed in Kenya is raw milk which is neither pro- cessed nor packaged. Raw milk is normally boiled before use and in many areas, the largest proportion of consumption is as a whitener in tea or an ingredient in cooking. More than 90% of the milk consumed in rural areas is raw milk, including milk retained on the farm and milk purchased locally. Little is known about the quantities or proportion of marketed milk in urban areas which is sold in raw form, as this is illegal. A recent, limited household survey in Nairobi did indicate that raw milk may account for 10-20% of total purchases.44 In 1990, consumer and institutional sales of pasteurized milk totaled about l~O million liters. Sales of long-life UHT milk totalled about 105 million li- ters. Most sales of pasteurized niilk were in urban areas, with the Nairobi and Mombasa metropolitan areas together accounting for some 75% of the total. In contrast, 82% of packaged UHT milk is sold in rural areas.45 The school milk program took another 38 million liters of pasteurized and UHT milk. The milk-equivalent consumption of fermented milk and high-value manufactured dairy products did not exceed 80 million liters.46 Therefore, the remaining 1.1 _billion liters of consumption was of raw milk. Prices Since 1971, consumer prices for pasteurized milk have been controlled by the government. During the 1970s and through to the mid-1980s, price adjust- ments were made in line with overall inflation levels. However, between 1987 and early 1992, official consumer prices did not keep pace with overall rates 43 Visitor numbers are drawn from the Statistical Abstract, 1990. 44 Unpublished survey sponsored by Tetra Pak (Kenya). 45 Price Waterhouse (1991), Annexes 18 and 19. 46 The total Kenyan market for cheese is only about 800 tons per year, while sales of other high- value dairy products are considerably lower. 228 MARKETING AFRICA'S HIGH-VALUE FOODS FIGURE 5.3: Retail Prices for Pasteurized Milk 10 / Price Adjusted For Inflation _,," Q) 8 " ~ _J ........ 6 ..c: Cl) ::s:: 4 2 0 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 of inflation (Figure 5.3). By the early 1990s, consumer prices for pasteurized whole and semi-skimmed milk in Kenya were considerably below those found in several other African milk-producing countries, including Zimbabwe, South Africa and Tanzania.47 In 1992, Kenyan retail prices for pasteurized milk were considerably lower than import parity prices.48 Consumer price controls have been one of the contributing factors to KCC's poor financial performance. For example, over the 1988-90 period, KCC's cumulative loss for pasteurized market milk sales was Ksh. 42.6 million, accounting for about 20% of the firm's total losses.49 Consumer prices for other types of fluid milk and for manufactured dairy products are not determined by government, except when the government is the buyer (the School Milk Program, supplies for the army). Time series data on prices for such products are not available. In recent years, consumer prices for UHT milk have been 25o/o-33 higher than for pasteurized milk as a result 47 This is one of the findings of a survey conducted by Tetra Pak (Kenya). 48 Staal and Shapiro (n.d.). 49 See Dairy Master Plan, Technical Report #4, p. 32. PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 229 of higher packaging costs and the absence of government price controls. This has led to a recent expansion of KCC's UHT production, despite its higher cost to consumers. The price for raw milk sold in rural areas is generally well below that for processed milk due to lower transport costs and the absence of packaging costs. This is illustrated in Table 5.7 which compares KCC milk prices with the average local prices reported by cooperatives in several districts in Cen- tral, Rift Valley, and Western Provinces. Milk Collection, Processing, and Marketing Channels A substantial proportion of smallholder milk production is retained on the farm, whether by choice or because of transport or other marketing bottle- necks. According to the Dairy Master Plan, 52% of smallholder production was retained nationwide, with this rate ranging from 32% to 68% among different provinces. Ip. contrast, less than 15% of milk produced on large-scale farms is retained for domestic consumption. Some milk is sold by producers directly to consumers, institutions and private processors. Part of these sales have been illegal, especially those tak- ing place in or near urban areas. Several thousand smallholders and the vast majority oflarge farmers deliver their milk directly to KCC processing plants and cooling centers, using their own vehicles, tractors and bicycles. Most milk sales however occur through intermediaries. For smallholders, the most important outlet has been primary cooperative societies. In 1990, there were 183 registered cooperatives with nearly 120,000 members who had dairy marketing as their main activity. However, 105 of these coopera- TABLE 5.7: Average Consumer Prices for Processed and Raw Milk (Ksh. per Liter) Source or Locationffype of Milk 1988 1989 1990 KCC Pasteurized 6.90 7.20 7.80 KCCUHT Na. Na. 11.00 Nyeri (Raw; Local) 5.00 5.50 6.00 Kirinyaga (Raw; Local) 4.00 4.60 6.07 Eldava Ravine (Raw; Local) 4.00 4.00 4.80 Busia (Raw; Local) 5.50 6.50 7.10 Kakamega (Raw; Local) 5.80 7.00 7.00 Na. Not available. SOURCES: Kenya Cooperative Creameries; Ministry of Livestock Development Cooperative Survey. 230 MARKETING AFRICA'S HIGH-VALUE FOODS tives were either dormant or under liquidation and there were tens of thou- sands of non-active members. A few dozen other cooperatives, registered as 'multi-purpose' cooperatives, do undertake a significant level of dairy market- ing. There has been considerable volatility among Kenya's cooperatives with frequent start-ups and liquidations and widespread management and finan- cial problems. Dairy cooperatives have organized milk collection systems, generally us- ing their own vehicles. In most cases, milk collection is done only in the morning and members need to consume, sell or otherwise dispose of the pre- vious evening's milk. Although the Rural Dairy Development Project pro- vided milk coolers for evening milk to close to sixty cooperatives, more than 50% of the coolers were either misplaced, never used or used only occasion- ally. 50 Few cooperatives perform milk quality checks, providing little incen- tive for farmers to improve quality. The cooperatives sell milk directly to local consumers and institutions, to the KCC, and in recent years, to private processors. In milk deficit areas and in areas with difficult or long distance transport to a KCC factory or cooling center, most cooperative sales occur locally with only seasonal or periodic surpluses being directed to the KCC. On a national level, probably no more than 25% of the milk collected by cooperative societies is sold to the KCC. Transportation is a major bottleneck and the leading cost item for coop- eratives. Smallholder milk supplies are greatest during the rainy season when rural roads are in their worst condition. Some roads are simply impassable, even with tractors. One multi-district survey in the early 1980s found trans- port costs to account for 70% of total cooperative costs and more than 22% of the cooperative milk sales prices.51 KCC is the predominant dairy processor in Kenya. With ten active pro- cessing plants, the KCC has the capacity to handle up to 700 million liters per year for fluid milk processing. During the 1980s, the KCC accounted for some 98% of the milk processed by formal organizations. The company pro- duces seventeen products, the most important being pasteurized, UHT and dehydrated milk in whole or semi-skimmed forms, fermented milk, cheese and butter. 50 Dairy Master Plan, Main Report, p. 2.15. 61 Ministry of Livestock Development data as reported in Price Waterhouse (1991). PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 231 Until recent years, the volume of KCC purchases has been closely associ- ated with levels of annual rainfall in major producing areas-and hence the production surplus available for household and rural consumption needs (see Figure 5.4). It reinforces the notion that the KCC has long been a residual buyer of smallholder (surplus) production. The 1986-1987 break in this asso- ciation between rainfall and KCC purchases can be attributed to the large increase in the official producer price in 1986 and to the start-up of opera- tions for several KCC milk buying centers. Two cooperatives, the Meru Cooperative Union and the Kitinda Dairy Farmers' Society, have UHT processing plants with capacities of 15.8 million liters/year and 7.9 million liters/year, respectively. The Kitinda plant was inoperative in mid-1992 due to management and other problems. Both coop- eratives have utilized an expensive technology supplied under a bilateral do- nor program. There are approximately fifty private enterprises involved in dairy processing. Most of these are backyard, cottage industry-type opera- tions, with only a dozen or so firms being formally registered. Most of these FIGURE 5.4: KCC Milk Purchases and Rainfall, 1971-1989 400 350 / ,..--- 25 I ~ 300 2 I :.:J c: I 20 E' ~E ~ 250 I en~ "C- \ I c:- ~ co ~ en rmal training or prior experience in dairy processing or food science. The backgrounds of the entrepreneurs are varied. Many are professionals with a business or academic background who view dairying and dairy pro- cessing as either a second career, a source of supplementary income or as a worthwhile hobby for themselves and their spouse. Among the proprietors of these firms include several former university teachers and administrators, a former World Bank and government economist, and a Minister of Agricul- ture. In several cases, these entrepreneurs have had academic training and prior experience in marketing or one of the social sciences which has proven useful in developing their dairy business. Only in the case of the largest firm has the owner-manager had many years of dairy processing experience. Nearly all of the cheese and yogurt makers have investments in additional agricultural ventures, most co~only in pig farming, beverage crops and PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 235 horticulture. Pig farming can be well integrated with cheese-making because the whey makes excellent pig feed. One firm spreads the costs from its pas- teurizing and cold storage equipment by also producing a range of fruit juices. Only six of the firms rely on sales of dairy products for 50% or more of their income, and only two family companies are totally specialized in this field. Most of these firms have built up their operations slowly, ploughing back earnings to acquire additional pieces of equipment or to upgrade technolo- gies. Batch, rather than flow, process technologies are used and most opera- tions are manual or semi-mechanized. The majority of firms have relied upon locally produced or reconditioned equipment. The limited introduction of so- phisticated technologies in this industry is a function of several factors: the small size and uncertain growth of the product market; the high cost of im- ported machinery; limited capital resources; and a generally conservative busi- ness development attitude in the face of potential reactions or intrusions by the KCC. Most of these firms produce a range of dairy products in order to make maximum possible use of the fat content of milk and to spread their market risk and production over time. The seasonality of both the available milk supplies and the demand for high-value products (linked to peak tourist and slow seasons) is one important reason why firms produce many different types of cheeses with varied maturity and storage periods. KCC actions and pres- sures have at times hindered the private processors from fully developing their product lines. For example, for many years registered processors were restricted from marketing their own cheddar cheese in competition with the KCC's product. Uncertain KCC supplies of dehydrated milk inhibited private investments in ice cream production. If there were no restrictions on milk marketing, many of the firms would also have sought to integrate production and sales of whole and skimmed milk into their operations.55 Most of these cheese and yogurt makers are based in rural or peri-urbah areas, with the largest clusters located on the outskirts of Nairobi and Nakuru. 56 This locational pattern contrasts sharply with that of the KCC's processing factories which are all urban-based. Given the bulkiness and rapid 55 A few firms have by-passed market restrictions, meeting special orders for skimmed and whole pasteurized milk from hotels and others. 56 The largest private processor is, however, urban-based. This firm has had long-term linkages with the KCC and has been dependent upon the latter's nearby milk collection/processing plant for its raw materials. · TABLE 5.8: Private Producers of High-value Dairy Products (Cheeses, Cream, Yogurt Products) in Kenya N Col Ol 1991 Dairy Dairy Share Product of Turn- Milk #of Total over In-take Year Source of Em- Other Turn- (Ksh. (Lts./ Company Started Ownership Finance ployees Ventures over' Dairy Products Mill.) day)2 DL 1964 K.European 80%, Savings 45 Pig Farming, A Many Cheeses, Tinned 20.0 11,000 Employees 11%, Meat Canning Milk, Ice Cream Foreign 9% KMP 1952 K. Asian Other Ventures 17 Hardware, Paints, D Many Cheeses, 5.5 6000 Partnership Paper Products Yogurt ABD 1984 K. Afr. Family Savings 15 None A Many Cheeses 6.5 4200 EF 1976 K. Afr. Family Savings 15 Fresh/Frozen Vege- B Yogurt, Cream, Soft 8.5 2500 tables; Pig Farm- and Cottage Cheeses, ing; Meat Products Skimmed Milk SNP 1982 K. European Savings 16 None A Soft Cheeses, Cream 7.0 1600 Family Cheese BFD 1990 K.Asian/Foreign Savings+ 20 Grain Milling; B Flavored Yogurt 9.0 700 ~ :u Partnership Commercial Agro-industry ~ Loans consulting :::! <: Gl EDV 1984 K. Afr. Family Savings 7 Poultry, Eggs, B Cream, Yogurt, 3.0 560 l> Fruit Juices Milk ~ ~ er) MLS 1972 K. Afr. Family Savings/AFC 3 Tea, Flowers, D Cream, Yogurt, 1.5 250 Beekeeping Dip bases :!; Gl =t: 1 A= 75-100%, B = 50-74%, C = 25-49%, D = Less than 25%. :;;:: r- 2 Peak season supplies (1991 or 1992). ,.,, c:: Na. Not available. 2i 0 Cl en ~ TABLE 5.8: Private Producers of High-value Dairy Products (Cheeses, Cream, Yogurt Products) in Kenya (continued) ~ (/) 1991 ~ Dairy Dairy OJ Share Product ,.,, r- of Turn- Milk ~ Cl #of Total over In-take :n Year Source of Em- Other Turn- (Ksh. (Lts./ ~ Company Started Ownership Finance ployees Ventures over1 Dairy Products Mill.) day)2 ~ :!:! KWG 1986 K. Afr. Family Savings 4Pr Consulting D Cheddar Cheese 0.5 150 § ;tj (J) LDW 1982 K. Afr. Family Savings/AFC 3Pr Tea, Flowers D Yogurt, Cream 0.75 100 f!J Cl ::0 Firms Not Interviewed ~ :0 CT 1980s K. Afr./Foreign Na. Na. Tea D Soft Cheeses, 7.0 1500 -< Partnership Mala Milk (Est.) ;a Cl C) GA Na. Kenya African Na. Na. Coffee, Trade, D Yogurt, Cream, 3.0 500 fJl (/) Industrial Mala Milk (Est.) ~ Investments l> <': Cl K. Afr. Family TK 1980s Na. Na. Politics, Cereals D Soft Cheeses 1.5 250 ~ (Est.) :>:i ::s; ,.,. MP 1980s K. Afr. Family Na. Na. Na. Na. Yogurt 1.5 250 ~ (;': (Est.) :2 1 :i: ,.,. A= 75-100%, B = 50-74%, C = 25-49%, D =Less than 25%. 2 Peak season supplies (1991 or 1992). < ):; Na. Not available. .. t 238 MARKETING AFRICA'S HIGH-VALUE FOODS perishability of milk, rural and decentralized placement of dairy manufactur- ers is more appropriate. Yet, the KCC's main cheese-making factory, at Dandora just outside of Nairobi, has generally processed milk transported over distances of 100-300 kilometers. Background and Characteristics of Fluid Milk Processors Table 5.9 provides basic information on nine existing or prospective private/ cooperative fluid milk processors. Eight of these firms were interviewed. At the time of field work, the Kitinda Farmers Dairy Society was inoperative and problems of security prevented access. The interviewed firms are repre- sentative of the wider set of firms already active or preparing investments in this area. The table shows the varied profiles of these firms. Some are large mixed farms with significant dairy and beef herds. These firms have sought to add value to their output and to by-pass the KCC and deal directly with quality- conscious consumers or institutions. Others are cooperatives which have ob- tained the financial and technical backing of international donors and equip- ment supply companies. The 'BD Company' is representative of about a half-dozen small firms which have undertaken the production of fermented milk with the technical and financial support of an American PVO (Tech- noserve).57 Several multinational companies, including two interviewed for this study, are considering major investments in a liberalized Kenyan milk market. Most of the interviewed firms have diverse agricultural and other interests. Dairy-related sales are not their dominant source of income. Only small firms producing fermented milk had specialized in this area. Unlike the pattern for cheese and yogurt makers, none of these firms are small family companies. Until now, entry into the fluid milk market has required either political muscle or external sponsors, neither of which apply to small family companies. The pattern that is slowly emerging is one where large farms, larger com- panies and cooperatives undertake the production and sale of pasteurized, UHT or powdered milk, while smaller companies and cooperatives focus on fermented milk products. This follows directly from the differential invest- ment requirements for the different products and the, differential capacities 57 Some of these firms are owned by individuals while other are owned jointly by groups of small farmers and run on a cooperative basis. The latter firms are registered as private companies rather than cooperatives in order to avoid intrusions by the Ministry of Cooperative Development. ~ TABLE 5.9: Characteristics of a Sample of Actual and Prospective Private and Cooperative Milk Processing/Market Enterprises 5l OJ : Year Dairy Prod- Dairy Milk .,., r- Started (Intended) uct Sales Share of In-take in Pro- Source of Dairy (Ksh. Other Total Liters/day' ~ a Company cessing Ownership Finance Products Million) Ventures Turnover' (Planned) :n ii1 KLF Early K. Euro. Family Savings Pasteurized Milk, 20.0 Sisal, Beef B 8500 \J 1960s Cream, Butter (1991) 32 :;;; ;tj NGK 1978 K. African Other Ventures Pasteurized Milk Na. Coffee D 1000 .,., Vi C) Meru Co- 1982 Cooperative Concessional Foreign UHTMilk 26.4 Coffee, Maize D 20,000 Cj ::0 operative Loan, Gov't Grant, (1990/91) Milling, Animal (50,000) Union Member Contrib. Feeds ~ :0 -<: Kitinda 1987 Cooperative Donor Grant; Member UHTMilk 14.7 None A ;a a Farmers Contributions (1989) C) Dairy Coop. Ul en ~ Buit 1990 PVO/K. African PVO Equity Finance, Mala and Raw 0.8 None A 1000 Gl ;,. Dairy Co. Member Contributions Milk (1990/91) < CJ MNE 1992 K. European Company Savings Pasteurized Milk, 12.0 Cereals, Beef, D 8000 ~ :t (Yogurt) (1991) Stockbreeding, (24,000) :>: rr Tourism ~ c;; Limuru Dairy Planned Cooperative To be Determined (UHT and Fermented 38.5 Agricultural A 34,000 ~ Farmers Co- operative Donor Concessional Loan likely Milk, Yogurt) (1989/90) Inputs, Post Office (50,000) " rr <: :;:; EAT Planned Foreign To be Determined (Pasteurized Milk, Na. Pig Farming, Ani- D 4000 Cream, Butter) mal Feeds, Part of (10,000) Conglomerate NST Planned Foreign To be Determined (Powdered Milk) Na. Weaning/Baby Na. (80,000) Foods, Con- fectionary 1 A= 75-100%, B = 50-74%, C = 25-49%, D =Less than 25%. " " u 2 Current peak season in-take or farm production. Planned in-take for new or expaned facilities is in parentheses. 240 MARKETING AFRICA'S HIGH-VALUE FOODS to service regional or national markets. Fermented milk is popular in rural areas and small townships. Relatively simple, batch-process technologies can be used to produce fermented milk and the product can be kept for more than one week without refrigeration. In contrast, in the production of pasteurized, UHT and powdered milk, both cost economies and improved hygiene can be gained by adopting mechanized or semi-mechanized flow-process technolo- gies. Marketing pasteurized milk requires refrigerated storage, while mar- keting UHT milk requires use of sophisticated, high-cost packaging materials. Given the relatively large market potential, actual and planned producers of pasteurized, UHT and powdered milk are considering milk intake levels which are much higher than even the largest processor of high-value dairy products. Simply for the firms listed here, planned processing capacity would exceed 200,000 liters/day, a figure which is several times larger than the total processing capacity of all high-value product manufacturers. Enterprise Development: Registration, Technical Information, and Finance To register a dairy processing enterprise and comply with other official oper- ating regulations, a firm must obtain a wide range of licenses and certificates. In addition to the initial company registration, a firm has to obtain a man- ufacturer's license, a Kenya Dairy Board license, a food and drug license, a health certificate, a Kenya Bureau of Standards certification, a VAT license, and for marketing, a wholesaler's license. Acquiring these licenses and certifi- cates is costly. Depending upon whether the licenses are for single or mul- tiple years and the level of side payments made to licensing officials, the annual costs are reported to be Ksh. 8000-15,000.58 Even more important than the actual costs are the delays from rejected applications and the restrictions placed on company operations. Delays of between one and three years have been the norm for obtaining the Kenya Dairy Board license, caused by KCC efforts to deter or limit potential com- petitors. The Kenya Dairy Board license restricts what the holder can pro- duce and where the product can be sold. In general, licenses also restrict the quantity of production and sales. The usual delay for receiving health, food and drug licenses is between six months and a year. Bureaucratic inefficiency artd[bribe-seeking behavior account for this delay. 58 The firms report that proprietors or senior managers must spend two to three weeks at the beginning of the year procuring the necessary licenses/certificates. PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 241 Manufacturers of dairy products are not permitted to obtain wholesalers' licenses. Instead, they are required to use the services of a commissioned trading agent. This regulation, applied throughout the manufacturing sector, was instituted during the 1960s in order to encourage the development of indigenous trading companies at a time when the manufacturing sector was dominated by foreign and Kenyan Asian companies. The KCC has been ex- empt from this law and has developed its own sales depots. By contrast, the largest private cheese-maker has been prevented from wholesaling its prod- ucts, placing it at a competitive disadvantage vis-a-vis the KCC because of the additional 20% commission fees charged by distributors. Other private dairy firms have been able to bypass this regulation, either by setting up . sister companies (usually run by relatives) to perform distribution tasks or by dealing directly with hotels and caterers for a large proportion of their sales. The established and emergent dairy processors have had several sources of technical information available to develop product ideas and recipes and to make decisions about processing technology. For the recently developed high- value product producers, the most important sources of technical information have been: local and visiting dairy specialists at the University of Nairobi and Egerton College, representatives of equipment supply companies, foreign coun- try trade representatives, personal visits and short training courses abroad, and technicians and food specialists from hotels and from a private company which provides food services to international airlines (for example, the Na- tional Airline Service). The technology for fermented milk production, devel- oped and initially tested at the University of Nairobi, was later extended to several small companies by a private voluntary organization. None of the firms who were interviewed mentioned any public, non-uni- versity source of technical information, though several firms do have staff who attended courses at the Dairy Training Center. The weakness of public services applies also to the area of quality control. Most firms complain that the Kenya Bureau of Standards lacks qualified staff. Results from Bureau tests take months to arrive and are said to be incomprehensible at times. In contrast, the quality inspection and advisory services of both hotels and the private airline service are universally praised. Among the producers of high-value products, the dominant source of fi- nance for investment and working capital has been savings from employment and agricultural activities. Of the ten firms interviewed, only one had ob- tained a commercial loan to develop its processing factory. Two others had 242 MARKETING AFRICA'S HIGH-VALUE FOODS received loans from the Agricultural Finance Corporation to purchase their farms. Several firms report being rejected for long-term loans for dairy pro- cessing investments. 59 The cooperatives which have invested in milk processing plants have fi- nanced their projects through a combination of concessional loans and grants from external donors, small Kenyan government grants and contributions received from members. The external concessional loans or grants were linked to the purchase of equipment and technical service contracts with companies from the donor country. In the case of the Kitinda Cooperative, 85% of the initial investment was a grant from a foreign government.60 The Meru Coop- erative will complete the construction of a large new UHT processing facility in 1993 with tied bilateral donor funds. Several schemes to produce fermented milk, including that of the Buit Dairy Company, have involved a joint ven- ture between a PVO (registered locally as a foundation) and a group of small farmers, the plan being to sell the PVO's equity shares to the farmers over time. Raw Material Procurement Arrangements Restricted product sales and KCC controls on all milk flows within the formal sector have prevented private processors from developing sustained raw ma- terial procurement arrangements which best match their needs in terms of both quantity and quality. At the time of field work, just after the decontrol on producer prices had been announced, raw material procurement arrange- ments were in a state of flux. Both farmers and processing firms were just beginning to develop new types of linkages. Impending changes seem most apparent in the milk procurement arrange- ments of processors of high-value product processors. Until early 1992, the two largest producers were required to obtain their milk supplies from the KCC. The system involved direct deliveries by (primarily) large-scale farmers to the plants, invoices were made on a KCC account and payments passed from the firms to the KCC and then to the farmers. Farmers were paid the prevailing official KCC prices and the factory paid the KCC an 'administra- tive fee' for its bookkeeping. In recent years, this administrative fee was Ksh. 59 At least for cheese-makers this has created cash flow problems given the time needed for the product to reach maturity and thus yield income. 60 This is discussed by Tribe (1989). PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 243 0.50 per liter, representing a tax on processors of more than 10% of the pre- vailing official procurement price. While this procurement arrangement was . generally satisfactory during the 'flush' season, whenever there were short- ages of milk, the KCC would seek to ration supplies to the processors in order to increase the utilization of the KCC's own processing facilities. Mid-1992, these firms were in the process of breaking their link with the KCC and forming direct procurement channels. Kenya Milk Products Limited reported developing a collection system for a network of forty large-scale suppliers and some 150 smallholders. The firm was also considering supplying private vet- erinary services to its farmers at cost. The moderate size processors (those handling 1000-5000 liters/day) have relied upon local cooperative societies for the bulk of their raw material needs, supplemented by clandestine deliveries from large-scale and other neighbor- ing farmers. The common practice has been to make bi-weekly or monthly pre-payments to the cooperatives and then to draw down the accounts through deliveries. These processors have been favorable outlets for the cooperative societies, because they offer a steady and fairly sizeable market outlet, pay prices slightly above KCC rates and feature lower transaction costs than those incurred when dealing with the KCC. Not only is payment more prompt, the cooperatives also do not encounter the milk intake congestion which is common at KCC collection stations during the 'flush' season. For their part, the processing firms have found these arrangements to be acceptable, al- though all complain about periodic quality problems. Due to the presence of antibiotics in some deliveries, the firms have generally relied upon their own small dairy herds for starter culture. The procurement of milk directly from farmers had been fraught with risk until the recent decontrol with fines, bribes and threats of or actual confiscation of milk containers occurring regu- larly. The smaller cottage industry firms have relied mostly on their own dairy herds for milk, supplemented by supplies from neighboring farmers. Among the existing and prospective fluid and powdered milk processors, there are varied procurement arrangements and plans. One model integrates large-scale dairying activities with milk processing. These farms have herds numbering 500 or more milking cows. Animal feeding and calving is under- taken or timed either to maintain steady year-round supplies or to target supplies for the dry season when market prices are higher. These farms indi- cate that they will buy supplies from other (large) farms in order to supple- ment their own production and reach desired levels of capacity utilization. As 244 MARKETING AFRICA'S HIGH-VALUE FOODS these farms are seeking to tap the premium quality end of the milk market, preparations are being made for extensive quality checks of both own-farm and out-supplier milk. Successful, relatively large-scale dairy cooperative societies and unions have developed extensive milk collection and service systems for their mem- bers. The Meru Cooperative Union and its affiliated primary ~ooperative soci- eties have an extensive milk collection network. The Union provides credit to members and sells them feeds, agro-chemicals and many other production inputs at cost. The Limuru Dairy Cooperative Society has gone further and provides its members with private artificial insemination and clinical ser- vices, charged at cost. A similar, multi-dimensional procurement system is being considered by Nestle, should it invest in a dehydration factory. The firm will likely concentrate its milk procurement activities in one high-poten- tial region, offering a full range of services within the framework of a contract farining scheme. This would be similar to those operated by Nestle in other countries. Marketing Arrangements The Kenyan market for cheese, yogurt, cream and other high-value dairy products remains small due to the limited purchasing power and lack of an acquired taste for such products by most of the population. As noted above, consumers of these products are mostly tourists, expatriate residents and Kenya's relatively small middle- and upper-income population. Therefore, dis- tribution of these products is primarily through hotels, restaurants and up- market grocers and chain stores. There has been relatively little supplier promotion of high-value dairy products, whether by the KCC or by private companies. For many years, the KCC has produced cheese largely as a means of using surplus milk and serv- ing large institutional needs (military, hospitals). Private firms have also been very cautious, producing only types and quantities of products for which they have firm orders or know that they can sell. Only a few firms have advertised their products in magazines or on television or conducted promotions in stores, through clubs or at public events. While all products sold through stores have company or brand labels, only in the past few years have firms recognized the importance of product presentation and invested in higher quality packaging and labeling. For !fiOSt dairy processors, market research has consisted only of discus- PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 245 sions with hotel and restaurant chefs and supermarket managers about cur- rent demand. A few firms, with the cooperation of hotel chefs, have test mar- keted their products in hotel buffets. Two firms have benefited from market- ing advice provided by a specialized distribution company which carries their products. The sales strategies and channels generally differ between the cottage industry type firms and the larger firms. The former companies direct most of their sales to hotels and up-market shops, without using distribution agents and without paying the required value added tax on sales. These firms tend to channel their products to buyers in only one or two major towns. In con- trast, the larger firms combine sales to hotels and restaurants, sales at the farm and factory gate or through affiliated companies, and sales to shops. Some firms, particularly those operating under proper hygienic conditions, also supply the National Airline Service. The larger firms tend to use com- mission agents for some sales, especially for distant markets. Most serve Nairobi and Mombasa and selected other sites. Their preferred outlets are hotels and the airline service since they get immediate feedback on quality and advice from them, there is no risk of returned unsold items, and the processors can escape paying VAT on such sales. Sales are on a cash basis to stores and a credit basis to hotels. Although prices are negotiated, it is gener- ally the buyers which have the final say, especially in the cases of five-star hotels, the airline service and large retail chain stores. For cheddar cheese, the high-value product with the largest market volume, private company prices (and overall profitability) have been kept down by KCC sales of the product at or below its variable production costs. The KCC has consistently shown losses on this product. Although the market for fluid milk is far larger than that for high-value products, Kenya Dairy Board restrictions on sales quantities and locations and the predominance of KCC is certain areas have handicapped marketing efforts by private and cooperative firms. While Kenya Dairy Board restric- tions have been less aggressively enforced in recent years, their presence has still prevented firms from establishing long-term distribution arrangements outside of their licensed market areas. Whenever their own target market areas have been temporarily saturated, these firms have had to develop ad hoc sales arrangements in other areas. Some lucrative market outlets, such as the School Milk Program, the Army and other public institutions have been monopolized by the KCC. 246 MARKETING AFRICA'S HIGH-VALUE FOODS The cooperatives engaged in milk processing have generally sold both raw and processed milk. Their sales, made at cooperative society depots and fac- tory gates and through local shops and distribution agents, have been concen- trated in close vicinity to the factory. Surplus quantities have been sent to distant major cities, such as Nairobi or Kisumu. In contrast with the pan- territorial sales prices for KCC milk, these cooperatives employ different price schedules by location and type of buyer. Previously prevented from doing so, several of the large cooperative processors are now assessing the feasibility of setting up large wholesale depots in Nairobi to compete directly with the KCC in that market. Indicators of Performance Both private and cooperative dairy processors have faced official restrictions in their milk procurement arrangements, choice of products and areas and channels to or through which they could market these products. At the same time, these firms have faced aggressive efforts ( in milk collection, marketing, and pricing) by the KCC designed to weaken or limit the growth of its com- petitors. Not surprisingly, these constraints and aggressive intrusions have undermined the sales and financial performance of private and cooperative processors and greatly limited the extent to which private firms could sup- port farm-level milk production. Table 5.10 examines rates of capacity utilization for the KCC and for pri- vate and cooperative processors for which data (or estimates) are available. (Private firms are indicated by initials.) For fluid and fermented milk produc- ers, the indicated utilization levels are for peak production years and/or for 1992, and based on one-, two- or three-shift operations. For cheese and yogurt makers, the estimates are for the peak production season during 1992, with all calculations based on an assumption of two shifts. In practice, most of these firms operate for only one long shift of 9-10 hours, but conversions are made to render utilization rates comparable to those of fluid milk processors. The table does not include some of the-- small, cottage-industry firms, whose processing capacity is difficult to estimate and which lack records on monthly or quarterly milk intake. The table shows capacity utilization rates varying between 26% and 75%. The KCC's capacity utilization rates in its peak production year (1989/90) were higher than comparative rates among private and cooperative milk pro- cessors. However, with reduced milk purchases, KCC's average capacity utili- zation fell below 36% in 1992. The average figures for the KCC do not reveal . PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 247 TABLE 5.10: Rates of Capacity Utilization in the Dairy Sector Fluid Milk Processors Cheese/Yogurt Makers #of % Capacity % Capacity Company Year Product Shifts Utilization Company Utilization KCC 1989/90 Pasteurized 2 61 DL 56 (Peak) Milk 3 55 UHTMilk 3 69 1992 Powdered Milk 3 36 All Products Meru Coop. 1988 UHTMilk 2 26 KMP 40 Kitinda Coop. 1989 UHTMilk 2 30 ABD 57 1992 1 5 KLF 1992 Pasteurized Milk 2 51 EF 75 MNR 1992 Pasteurized Milk 1 75 SP 57 (Aug-Dec.) (est.) Buit 1992 Fermented Milk 1 40 BFD 35 Average Average 1992 38 (Weighted) 52 SOURCES: DANIDA (1991); company interviews and financial data. the very low rates of capacity utilization at some of its factories. These fac- tories, in a state of disrepair and experiencing major congestion at intake points during the peak season, have been used in recent years primarily as collection points for milk to be transferred to operational factories several hundred kilometers away. Only three of KCC's factories-at Kitale, Kiganjo, and Nyahururu-process large quantities of milk produced from within a 75 kilometer radius. Neither Meru nor Kitinda has ever achieved annual utilization rates above 30%, due to the seasonality of milk supplies, periodic factory and infrastruc- ture-related breakdowns, profitable cooperative society sales of raw milk, and aggressive KCC milk collection and sales programs in nearby locations. Nev- ertheless, with the support of an external concessional loan, the Meru Coop- erative is presently completing construction of a new factory with more than double the capacity of the existing factory. 61 The relatively low capacity utili- 61 A recent consultant report claims that neither Meru nor Kitinda have been able to produce UHT milk with the equipment that they have. The shelf life of their product has been about five days, and both cooperatives receive back up to 10% of their output returned to them (spoiled) from distributors (Avezard 1992). 248 MARKETING AFRICA'S HIGH-VALUE FOODS zation for the Buit company stems from shortages of water (for cooling the heated milk) and the profitable diversion of some raw milk supplies to fresh market sale. While the Kilifi Plantation could nearly double its milk intake, its market area has been restricted by the Dairy Board. Among the larger cheese and yogurt-makers, average capacity utilization rates during the 'flush' production season are about 50%, given the assump- tion that there should be two operating shifts. The limited market size and its seasonality (because of tourist arrivals), past restrictions on milk procure- ment arrangements, and the narrow management and labor supervision base of all of these firms have prevented more complete capacity utilization even during the 'flush' production season. For similar reasons, most of the cottage- industry processors operate only a few hours per day. Table 5.11 indicates recent patterns in the sales and financial performance of the KCC, the two leading cooperative processors and the two largest pri- vate high-value product processors. The table highlights the large and contin- ued financial losses incurred by the KCC. Operational inefficiencies, official price controls and huge losses incurred in converting seasonal surplus milk supplies into storable products have been the major contributing factors to its negative balance sheet. KCC losses have been covered by increased borrow- ing, especially from the government. Still, liquidity problems have resulted in delayed payments of several months to farmers, cooperatives and other input and service suppliers. TABLE 5.11: Sales and Financial Performance in the Dairy Sector (Ksh. Millions) Company Indicator 1987/88 1988/89 1989/90 1990/91 KCC Sales 2,203.9 2,644.4 2954.3 3031.6 Profit (Losses) (159.1) (57.1) (103.2) (66.7) Accumulated Profit (Losses) (191.1) (248.2) (351.4) (418.1) MeruCoop. Srues 22.6 23.5 22.6 26.4 Profit (Losses) (2.5) 0.8 (0.1) (0.02) Accumulated Profit (Losses) (11. 7) (10.9) (11.0) (11.0) Kitinda Coop. Sales 13.3 14.7 16.9 12.0 Profit (Losses) (0.4) (0.6) (0.9) (1.1) Accumulated Profit (Losses) (0.8) (1.4) (2.3) (3.4) DL (Private) Sales Na. 17.6 19.0 20.0 Profit (Losses) Na. 0.3 0.4 0.4 KMP (Private) Sales 4.3 4.2 5.1 5.5 Profit (Losses) 0.4 0.5 0.5 0.6 SOURCES: DANIDA (1991); Company financial statements. PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 249 TABLE 5.12: Payments to Producers as a Share of Sales Turnover Company 1988 1989 1990 1991 KCC 55.1 54.3 53.2 47.7 MeruCoop. 65.6 55.7 54.4 49.1 Kitinda Coop. 54.3 62.3 63.1 Na. Buit Mala Co. Na. Na. Na. 52.4 Engineer Mala Co. Na. Na. Na. 51.4 SOURCES: DANIDA (1991); Company financial statements. In addition, both Meru and Kitinda Cooperatives have exhibited unsatis- factory sales and financial performance in recent years. As with the KCC, their accumulated losses have been covered by government and external do- nor grants. The private firms listed in the table have generated steady, if modest profits, yet sales turnover has increased only moderately in recent years. Table 5.12 provides a comparison of the share of company sales paid out to producers by the KCC, the two main cooperative milk processors, and two small private fermented milk producers which obtain milk directly from com- pany shareholders/members. Similar calculations are not possible for firms converting milk into high-value products. Since the cooperatives and small companies have sought to increase their payout to members in order to mini- mize tax liabilities, we would expect to find higher producer shares of final realized sales for these firms. The table indicates that the small-to-medium.- scale processing cooperatives have indeed had a larger share of their sales earnings paid out to milk producers. The payout to producers from the KCC has fallen below 50% in recent years.62 Table 5.13 compares the cost/price structure for UHT milk produced by the KCC and the Kitinda cooperative in 1990. KCC milk is packed in multi- layered Tetra Pak containers, and Kitinda milk packed in plastic sachets. The table shows lower net costs for Kitinda, largely due to the high cost of Tetra Pak packaging and KCC's relatively high overhead costs. These extra costs more than compensated for Kitinda's higher unit distribution and labor costs. Because of the KCC's higher transport cost and the high import content 62 The share was 48% in 1992. For comparison, Eriksen (1992) found that farmers received 60% of the value of dairy processor sales to retailers in Morocco. He also reported that among urban north- east dairy plants in the United States, the relevant farmers' share was nearly 62%. 250 MARKETING AFRICA'S HIGH-VALUE FOODS TABLE 5.13: Cost/Price Structure of UHT Milk in Kenya (1990, Ksh/Litre) Kitinda Coop. KCC (plastic sachets) (Tetra Pak) Net Farm-gate Price 3.31 2.98 Transport to Collection Station and Factory 0.75 0.84 Cooperative Society Expenses 0.30 0.30 Taxes/Dairy Development Cess 0.34 0.32 Cost Ex-factory 4.70 4.44 Processing Costs Salaries/Wages 0.69 0.26 Power and Fuel 0.27 0.19 Machinery Repair 0.14 0.22 Depreciation 0.67 0.12 Packaging 1.20 2.75 Other Costs 0.24 0.22 Sub-total 3.21 3.76 Distribution Costs Transport 0.27 0.33 Selling Costs 1.00 0.47 Returns and Leakages 0.30 0.11 Sub-total 1.57 0.91 Overhead Costs 0.44 1.71 Total Costs 9.92 10.s21 Retail Sales Price 11.00 11.00 1 KCC receives milk with 3.5% butter fat and sells milk containing 2.3% butter fat. Taking this into account the unit costs for production are reduced by Ksh. 0.30/litre. SOURCE: FAO (1991). of Tetra Pak packaging (for the materials and machinery rental), the foreign exchange costs for KCC milk are considerably higher than those for Kitinda, Meru and private milk processors. CONCLUSIONS Although milk exhibits several technical properties which constrain produc- tion and distribution, it can be consumed in many different forms and be preserved or processed using a range of different technologies with varied efficient operating scales. The organizational patterns and technologies adopted for daicy processing and marketing should reflect the structure and PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 251 location of farm-level milk production, as well as the nature of demand for dairy products. For many years, the development of Kenya's dairy industry has not followed this dictum. While Kenya's dairy industry is one of the larg- est and most commercialized in sub-Saharan Africa, its recent pattern of development has been poorly suited to the structure of local production and demand. AB a result, the majority of Kenya's producers and consumers have benefited little from major government-backed dairy industry investments and have in fact opted out of the formal dairy marketing system. Throughout much of the colonial period, dairy industry development was largely geared toward serving the interests of its producers who were settlers from Europe. Local consumers were taxed to subsidize loss-making exports and most African farmers were discouraged from undertaking dairy produc- tion. The Kenya Cooperative Creamery, a commercial cooperative with largely European membership, used the powers of the state to protect its members' interests and to thwart the activities of competitors. Following Kenyan independence, smallholder dairy production was suc- cessfully promoted. Yet, smallholders were incorporated into the KCC's milk procurement system on highly disadvantageous terms. This, together-"with the limited geographical spread of KCC operations, the absence of direct KCC linkages with smallholder farmers and of services to assist them, led most of the latter to consign the bulk of their production to household use and infor- mal rural area sales. After formal market competition for the KCC was sup- pressed (in the 1960s) and the KCC itself transformed into a "public" entity (in the 1970s/80s), a single model of dairy processing began to emerge. With the backing of external donors, the government and the KCC undertook a large investment program featuring urban-based, medium-to-large-scale pro- cessing facilities. These facilities operated at well below their capacity and generated a stream of products affordable to only a small proportion of the local population. Most producers and consumers have continued to rely upon the informal market for much of their sales outlets and consumption needs. The KCC has experienced worsening problems of management, product qual- ity and illiquidity. Despite policy pronouncements to the contrary, many past efforts to de- velop competitive small- and medium-scale private or cooperative processing facilities have been discouraged, if not directly prevented. Facing delays or rejections for license applications, subsequent restrictions on raw material sources, products, and market areas, and periodic or sustained anti-competi- tive measures by the KCC, private and cooperative processing and marketing 252 MARKETING AFRICA'S HIGH-VALUE FOODS activities have been pushed to the margins of the formal market in terms of both product line and geogTaphy. Enmeshed in a web of controls, these firms have encountered difficulties developing effective long-term arrangements for milk procurement and downstream product sales. With the partial exception of a few cooperatives, these firms have also been unable to fill the vacuum created by a deteriorating system of official livestock support services. A climate of change appears to be emerging. The Kenya Dairy Master Plan, produced in 1990/91 by a high-level team of local and foreign consult- ants, argue~ strongly that the centralized, KCC-dominated system was both inefficient and unsustainable and that measures needed to be taken to liber- alize the dairy market and stimulate additional private and cooperative pro- cessing and marketing activities. There is considerable interest within the country in undertaking such ventures. The potential participants in a liberal- ized market are very varied, as are their intended operational scales, prod- ucts and linkages with dairy producers. Such new ventures could complement as well as compete with a re-organized, decentralized and commercially- oriented KCC. ln'"-future progTams to support Kenya's dairy ingustry it should be recog- nized that no single model or operating system can meet the diverse produc- tion support, processing and consumption needs of the country. A healthy, re- structured dairy industry would combine: a) rural small-scale processing units (perhaps linked to NGOs), b) regional cooperative processors offering a range of production support and marketing services to their members, c) large farms doing their own milk pasteurizing and processing, perhaps providing production and marketing services to nearby farmers (who supply them), d) small-to-medium-scale private firms producing and selling a range of processed dairy products to middle- and upper-income population gToups and tourists, e) large processing firms with integrated or contracted milk supplies and serving both domestic and international markets for processed dairy products, and f) a reconstituted Kenya Cooperative Creamery competing in the major urban markets and redeveloping export sales. PERISHABLE PROFITS: PRIVATE SECTOR DAIRY PROCESSING AND MARKETING IN KENYA 253 Each of these enterprises will have to compete with the raw milk market, both in the procurement of supplies and for consumer custom. In the absence of government price controls, competition from this source will likely drive up producer prices and put significant pressure on processors to increase their operating efficiency and alter their product mix (for example, away from costly UHT products) to attract consumers from the lower income strata. VI. Market Imperfections and Government Failures: the Vanilla Sector in Madagascar Benoit Blarel and Diane Dolinsky INTRODUCTION V anilla is the second most expensive spice traded on the world market.1 After dominating the world market with more than 95% of total supplies, sub-Saharan Africa is in the process of rapidly, and perharps permanently, losing its dominant position to more competitive and commercially aggressive countries in the Pacific Basin. By 1991, SSA's share of the expanding and promising world market for vanilla had already fallen to about 50%. If the last ten years' trends were to continue unchanged, SSA would cease export- ing vanilla by the year 2000,.foregoing potentially iarge export revenues from a US$ 100 million market. This dramatic loss of competitiveness cannot be explained by an unfavor- able cost structure. Madagascar, which alone accounts for about three quar- ters of all SSA vanilla exports, produces vanilla for a third of the cost of its most direct competitor, Indonesia. It cannot be explained either by a produc- tion of sub-standard quality. According to the world's largest user of vanilla, a U.S.-based flavor manufacturing house, the vanilla produced in Madagascar continues to constitute, quality-wise, the reference for the flavor industry worldwide. The explanation for this loss of competitiveness lies elsewhere. 1 This case study is based on the results of a background paper on the international vanilla market prepared for the World Bank by Thomas J. Payne Market Development (1990), and on field work conducted by the World Bank, in February/March and July 1991, for the preparation of a vanilla sector report (World Bank, 1991). Besides the main author of the present paper, Messrs Thierry Aube, Francis Masson, Andriantsoamiraho Rasolojaona and Velonjara participated in the World Bank mission. 255 256 MARKETING AFRICA'S HIGH-VALUE FOODS The cause for the decline of Madagascar on the world market lies in a sectoral organization and policy which are proving inflexible and unable to react to a fast changing world market. Yet the same sectoral organization and policy were also the main factors behind Madagascar's past successful achievements on the world vanilla market. Until 1960, the market was small, stagnant, highly unstable and prone to speculation. This can be traced back to the numerous, structural market imperfections and failures which charac- terize both the supply and demand for vanilla. After independence, the Gov- ernment of Madagascar quickly intervened to establish an orderly domestic vanilla market which would promote consistency and reliability in production and quality among the tens of thous ands of small-scale vanilla growers, and would take the necessary collective action to develop external markets. Public sector interventions to overcome market failures were necessary and initially very successful: in about 15 years, Madagascar export volumes in- creased sixfold and export earnings fivefold. Madagascar was also able to maintain a dominant share of an expanding world market, as well as to sustain the highest quality standards. However, Government intervention in the vanilla market eventually proved counter-productive. The organization and policies governing the Malagasy vanilla sector since 1960 were too rigid and heavily biased towards direct Government intervention rather than relying on more market-based instru- ments, and on greater private sector involvement in the management of col- lective market institutions. The resulting incentive structure has generated myopic and collusive behavior on the part of a handful of exporters and an over-powerful public administration. After the mid-1970s, Madagascar made increasing use of its monopoly power on the world market by exacting an ever increasing price from its world buyers. Vanilla growers, however, did not reap any gains from the dramatic increase in monopoly rents which Mada- gascar extracted from the world market. Despite important technical barriers to entry into vanilla production, In- donesia rapidly expanded its production capacity and became increasingly su,ccessful at overcoming the same market failures which Madagascar had faced in the past. Indonesia has now supplanted Madagascar to become the world's largest vanilla exporter and is starting to challenge Madagascar in the quality of its production. The myopic and collusive behavior resulting from the Malagasy sectoral organization has so far proved unable to adapt to the competitive situation it helped establish on the world market. MARKET VS. GOVERNMENT FAILURES: A CASE STUDY OF THE VANILLA SECTOR IN MADAGASCAR 257 The Madagascar vanilla example offers an almost perfect illustration of a number of instances where African countries have lost their world market shares and competitiveness as a result of policy failures. The private sector continues to perform all commercial functions, yet under the pervasive con- trol of Government policies. This case study is also illustrative of numerous instances where the private sector's initiatives, operating under the heavy control of Government policies, engages in rent-seeking activities rather than developing markets and product quality. Finally, it is a good example of an apparent need for collective action to overcome multiple market failures and imperfections. Such collective action, when it is dominated by Government and too distanced from market signals, can lead to even greater inefficiencies and inequities, as this case study will show. In the second section of this case study we review the technical character- istics of the product and show how they contribute to market imperfections and failures. In the third section we analyze the structure of world vanilla demand and supply and identify the corresponding market f~ures and im- perfections. In the fourth section we describe the technical and economic char- acteristics of vanilla production in Madagascar, at each of its three stages (growing, curing and packing), and derive for each level the nature of the possible market failures and imperfections which may emerge. In the fifth section we review the successive policy regimes (free market, orderly market, and government managed) which have governed the organization of the va- nilla sector in Madagascar until today, and how these relate to the failures identified previously. In the sixth section we assess the impact of these differ- ent policy regimes on the performance of the Malagasy vanilla sector using both economic efficiency and equity criteria. We then conclude by draWing out the main policy implications of this case study. TECHNO-ECONOMIC CHARACTERISTICS OF VANILLA Vanilla has low perishability, at least once it is cured.2 Processors can hold cured vanilla beans in stock for years if the beans have been harvested at the 2 An explanation of terminology is necessary. Production of the vanilla beans actually consists of three stages-vanilla growing, curing and packing. At the growing stage, vanilla beans are referred to as "green", and need to undergo the curing process to acquire their flavor and aroma characteristics, hence providing them with their commercial value. They are then referred to as cured vanilla. The packing stage consists of the sorting, grading, conditioning and storage which are necessary before exports. 258 MARKETING AFRICA'S HIGH-VALUE FOODS appropriate time, and the curing has been carried out properly. If either of these two conditions is not met, there is heightened risk of mold. The spread of mold is checked by examining beans periodically during conditioning and storage to permit removal and treatment of infected beans. As long as there are regular inspections, low perishability permits packers to maintain stocks for sale at optimal times. It is low in bulkiness, and high in unit value. Vanilla beans are com- pact in size and easy to pack together. Vanilla beans have high unit value, and after saffron, are the second highest priced spice on the world mar- ket, with an average world FOB price of US$ 53/kg. in 1991. While vanilla beans accounted for only 0.5% of U.S. imports of spices and herbs in volume terms, their share of value was as high as 16%. Vanilla beans' low bulki- ness and high unit value make distribution over long distances feasible and highly profitable. Because of its high value, the vanilla trade is also prone to speculation. There are many dimensions to the quality of a vanilla bean. The major quality components are: flavor profile, natural vanillin content, bean length, moisture content, appearance, color, and presentation. Flavor profile is the most important quality characteristic of a vanilla bean, and differs greatly from one variety of vanilla plant to another .3 Differences in quality make the substitution between vanilla beans from different origins difficult. At least, four major types of vanilla beans can be distinguished: 4 the Bourbon vanilla (grown in Madagascar, Comoros and Reunion), the Java vanilla (grown on the island of the same name in Indone- sia), the "Bourbon-like" vanilla (mainly grown on the island of Bali in Indone- sia), and the Mexican vanilla. They differ in flavor profile, organoleptic and analytical properties as a result of variations in growing conditions (e.g., qual- ity and age of the vine, number of fertilized flowers per inflorescence), har- vesting and curing practices.5 Malagasy vanilla ranks top in terms of quality, a result of improved and well-established production, harvesting and curing 3 Three species of vanilla plants are commercially grown throughout the world. Vanilla fragrans cultivated in the Indian Ocean (Madagascar, Comoros, Reunion), Mexico and Indonesia dominates the world supply with over 95%. Two other varieties are grown in French Polynesia (Tahiti) and in Guadeloupe, respectively. They do not directly compete with the V. fragrans since they are used mostly in the perfume industry. 4 McCormick & Company (1992). 5 McCormick & Company (1992). MARKET VS. GOVERNMENT FAILURES: A CASE STUDY OF THE VANILLA SECTOR IN MADAGASCAR 259 practices. As the world leader, it sets the standard for the industry. Vanilla grown in Indonesia ranks fourth (after the three Bourbon-producing coun- tries) because of lower natural vanillin conten~ and less attractive flavor pro- file (the result of premature harvesting practices and poor curing techniques). Some of the quality characteristics are not directly observable, and grad- ing standards help reduce transaction costs and the level of uncertainty for buyers. The flavor profile and natural vanillin content are not directly mea- surable without special testing equipment. There is a small degree of correla- tion between the natural vanillin content and other more easily observable characteristics such as the length of the vanilla bean, whether the bean is split or not, its appearance, etc. Experienced vanilla specialists can detect appreciable differences in the quality of vanilla beans. To manage these wide quality differences, grading standards have long been elaborated in Mada- gascar and are schematically represented in Table 6.1. All producing coun- tries follow some sort of grading system, although these may differ greatly. Because important quality characteristics are not directly observable and are largely determined by growing conditions, time of harvest, and the curing process, there is scope for marketing externalities at the production (farm, curing, packing) level. Together, vanilla growers financially benefit from fol- lowing similar growing, harvesting and curing practices which safeguard con- sistency and high quality (flavor profile and vanillin content). This reduces uncertainty and risks for the buyer, potentially raising the demand for and market value of vanilla beans. Because direct observability of intrinsic qual- TABLE 6.1: Madagascar Vanilla Export Grades Standards of Vanilla Beans 1st Quality 2nd Quality 3rd Quality 4th Quality Whole Split Whole Split Whole Split Whole Split Good Flavor Broken or Cut Length> 14cm Length< 14 cm Tan Brown Uniform Color Supple Supple or Full Supple Dry No Spot Some Spots & Spots & No Scratch Scratches Scratches Moisture Moisture Moisture Moisture Content 38% Content 38% Content 30% Content 25% SOURCE: Ministere du Commerce, Direction de la Qualite et de la Metrologie Legale. 260 MARKETING AFRICA 's HIGH-VALUE FOODS TABLE 6.2: Technico-Economic Characteristics of the Vanilla Product and Relationships with Market Imperfections and Failures . Ll.rhited. Ob~etvabilit}' Market Imperfections and Failures: 1) Market Instability (speculative market) x x 2) Marketing externality: -Transaction costs x x -At production level x ity characteristics is difficult, individual growers and/or curers can be tempted to free-ride. 6 THE WORLD MARKET FOR VANILLA Structure and Composition of the World Demand 7 The world market for vanilla beans is small. Aggregate global demand for vanilla is estimated at about 2,000 tons a year. Over the period 1965-1989, world consumption of vanilla beans has grown at the average annual rate of about 2% (see Annex Table 6A). Between 1980 and 1989, demand expanded rapidly, in particular in the U.S. where it grew at 7% a year (in volume terms). In Europe, the rate of progression was more modest at 2-3%. The growth in aggregate demand has not matched the rates of income and popu- lation growth· in developed countries, despite the fact that the demand for vanilla is associated with highly income-elastic demand for end-products9 (e.g., ice creams, colas). The advent of synthetic vanilla (vanillin) whose production has been booming9 helps to explain this phenomenon. On the other hand, a number of factors have strengthened the demand for vanilla beans over the 6 To follow stringent growing, harvesting and curing practices increases quality but also tends to reduce production levels; e.g., the number of vanilla flowers pollinated on an individual vine raises production, but can also lower quality. 7 This section draws heavily on Thomas J. Payne Market Development (1990). 8 Purseglove, Brown, Green and Robbins (1981.) 9 Synthetic vanilla production in the U.S. increased by 135% between 1984 and 1988. MARKET VS. GOVERNMENT FAILURES: A CASE STUDY OF THE VANILLA SECTOR IN MADAGASCAR 261 past decade: increased health awareness for natural products, escalating con- sumer demand for processed foods (which use new flavors and spices), and an explosion in popularity of gourmet ice creams (which tend to use exclusively pure natural flavors). The world market for vanilla beans is highly concentrated in a few, rich countries. Three countries only (U.S., France and Germany) account for about 80% of world imports, the U.S. absorbing 50--60%, France and Germany be- tween 10 and 15% each (see Figure 6.1). These three countries are also major re-exporters of both vanilla beans and processed vanilla products.10 Other industrialised countries are significant minor importers.11 There are a very small number of importers of vanilla beans,12 which makes for limited competition, although they cater to the need of a much larger, diverse and competitive food industry. Vanilla beans are imported by brokers specializing in the spice business who sell them to flavor manufactur- ers and extractors for further processing. Vanilla bean brokers provide a vari- ety of services to their clients: they ensure the necessary consistency in qual- ity and in supplies by keeping inventories, they offer quality control services, they select the beans according to their quality requirements, and they main- tain close commercial contacts with suppliers in producing countries, often providing technical assistance and expertise to the packers. Reputation, ex- clusivity and technical expertise on the part of the brokers lower transaction costs which, more than financial constraints, act as important barriers to entry in the vanilla importing business. The marketing policies of Madagas- car, Comoros and Reunion have voluntarily restricted the number of buyers. This policy is not being pursued by other exporting countries, such as Indone- sia, which may sell to a larger number of brokers or directly to the food industry. However, it is believed that most imports from these sources are 10 Germany only consumes 30-40% of its imported beans, and its re-exports are by far the most significant, reaching approximately 135-210 metric tons per year in the late 1980's and early 1990's. France's annual re-exports were equalled 70-120 tons, while the United States' re-exports were 40- 60 metric tons. 11 Canada, Japan, and Switzerland each import about 35-60 tons each, the Netherlands, United Kingdom, Denmark, Italy, and Saudi Arabia 18-35 metric tons each, and Brazil, other EC countries and Australia between 10 and 20 metric tons each. Highest per capita consumption is found in Denmark (4.5 grams), the United States (3.85 grams) and France (2.5 grams). 12 In the U.S., there are only two major importers of vanilla beans produced in the Indian Ocean, with a third representing a small and declining share of U.S. imports. In France, there were seven importers of beans from the Indian Ocean islands in 1990-1991, although the two larger ones account for about half of all imports. In Germany, only two dealers import vanilla beans from the Indian Ocean islands. 262 MARKETING AFRICA 's HIGH-VALUE FOODS FIGURE 6.1: Share of World Vanilla Beans Imports (1989-91) Rest of the World 23% Germany 13% France 11% USA 53% still being imported (in particular in the large consuming countries such as the U.S. and France), by the same dealers/brokers who also enjoy exclusivity contracts from Madagascar (and Comoros and Reunion). Vanilla beans are primarily used as an ingredient in the food industry. The corresponding demand assumes special characteristics peculiar to the food ingredients market. Besides price, quality standards and consistency, reliability in supply, technical characteristics of the product, changes in food technology, regulations pertaining to food labelling and technology are all major determinants of the demand for vanilla beans. In the United States, over 95% of the vanilla beans are processed into extracts sold to flavor manu- facturers or to the retail trade. Some extract processors are vertically inte- grated with flavor manufacturers. Three types of vanilla extracts and flavors are manufactured: pure vanilla extracts, vanilla with other natural flavors, and vanilla flavors which may contain artificial substances. Only a small proportion of vanilla beans are consumed directly by consumers, except in France where the proportion is about 20%. The dairy industry (e.g., ice cream, frozen desserts) is the largest consumer of vanilla extracts and flavors,13 the confectionery, baking, and beverage industries (e.g. cola) the other major in- dustrial users. 13 44% of all vanilla extracts in the U.S. are used in the production of ice creams. MARKET VS. GOVERNMENT FAILURES: A CASE STUDY OF THE VANILLA SEC.TOR IN MADAGASCAR 263 Asymmetric information between buyers (the food industry and its bro- kers) and sellers of vanilla beans may arise. Confidentiality about quality and technical characteristics of the vanilla bean and its processed products, because of formulation secrets and proprietary processes in the food industry, is of paramount importance in the commercial relationship of the food indus- trialist and the vanilla broker. This provides the vanilla buyer with superior knowledge and information about the true demand for vanilla, resulting in the possible presence of asymmetric information. Vanilla is the spice most subject to competition from imperfect substi- tutes (low-cost artificial flavorings). Aggregate demand for vanilla beans therefore remains concentrated on the upper quality end of the vanilla flavor- ing market, where it is fairly price-inelastic, hence creating the potential for an unstable market. Four types of substitutes exist to date: synthetic vanil- lin, ethyl vanillin, other natural flavors, and tissue culture products. Syn- thetic vanillin accounts for more than 90% of the U.S. vanilla flavoring mar- ket and about 50% of the French market, the lowest national share. Synthetic vanillin costs one-hundredth the price of the natural product and serves not only to substitute for vanilla, but also to supplement or adulterate vanilla extracts. 14 Technological changes in the food, extract and flavor industry are major determinants of the extent to which these substitutes can compete with vanilla beans. 15 Competition from substitutes, and the significance of technological change in the food and ingredient industries on the demand for vanilla beans, strongly suggest the presence of marketing externalities at the demand level in three areas: generic promotion and advertising, food product and production research, and food regulations. Generic advertising and promotion directed to the food and ingredient industries, as well as to consumers, could increase the demand for natural vanilla, at the expense of substitutes.16 To be effec- tive, programs of generic advertising and promotion require substantial and sustained investments which are too large for most individual producers and 14 When vanilla extract is made from inferior quality vanilla beans, the resultant may contain less than the required percentage ofvanillin content. In order to raise the latter, synthetic vanillin is often mixed with the extract, thereby keeping a lid on the cost of the mixture. 15 For a more detailed discussion of these technological considerations, see Thomas J. Payne Mar- ket Development (1990). 16 In addition, the scope for generic advertising and promotion-Le. for the same country of origin- is underscored by the technical characteristics of the commodity: flavor profile tends to be similar within a same country of origin, and quality characteristics are not always directly observable. 264 MARKETING AFRICA'S HIGH-VALUE FOODS exporters of vanilla beans. An individual exporter's financial incentives to invest in such generic advertising and promotion programs are also reduced because other exporters from the same country of origin may benefit from the investment. These problems associated with economies of scale and free-riding give rise to marketing externalities that result in inefficiencies in a free mar- ket (i.e. too little generic advertising and promotion programs). A similar marketing externality prevails in the case of funding of research to increase potential uses of vanilla beans in the food and ingredients industry, or of extract certification to detect adulterations. As a result, in a free market, potential uses of vanilla beans in the food and ingredients industry are not being fully exploited, and increased competition from substitutes is not being addressed to the extent possible. Another example of marketing externality arises in the case of food regu- lations. In major consuming countries (U.S., EC), vanilla is the only spice which benefits from a Standard of Identity which helps shield vanilla beans from competition by substitutes. These regulations govern both food labelling and vanilla extract production. Food labelling regulations, which differ mark- edly in the U.S. and EC, determine the ease of substitution between natural vanilla beans and artificial substitutes in the food industry. In the U.S. and EC, food labelling regulations that impose specific labelling rules for vanilla flavored food products, allow for a clear distinction between food products that contain natural vanilla and those that contain synthetics. Current U.S. food labelling regulations offer a positive regulatory environment for vanilla beans. By contrast, the supra-national European Community food labelling regulations are less protective than some of the national regulations now being superseded (e.g. France), but are clearly more favorable than others (e.g. Italy). Extract regulations influence the scope for substitutability between va- nilla beans of different origins and quality, as well as the potential for market segmentation. French regulations on extracts stipulate the minimum percentage of natural vanillin content rather than the method of extraction. This favors high quality vanilla beans (those beans with a high vanillin con- tent). By contrast, U.S. standards, by making possible the use of lower qual- ity vanilla beans (i.e. with low vanillin content), encourage price competition and market segmentation in the largest world market for vanilla beans. Madagascar increasingly faces a price-elastic demand for its vanilla beans. Improved farming and curing practices recently adopted in Indonesia MARKET VS. GOVERNMENT FAILURES: A CASE STUDY OF THE VANILLA SECTOR IN MADAGASCAR 265 have enhanced the flavor profile and overall quality of its vanilla beans, in particular those grown in Bali, Celebes and Sumatra islands. As a result, there has been a marked convergence of quality, and concurrently, of importer's preferences between Indonesian and Bourbon vanilla.17 The Malagasy and top-of-the-line Balinese beans are increasingly seen as substitutes by the ma- jor world users of vanilla beans. Balinese output has been characterized by the world's largest producer of vanilla extracts and flavors as "Bourbon-like" in reference to the world standard.18 With increased substitution possibilities between Madagascar and Indonesia, the individual demand for Malagasy vanilla is becoming increasingly price-elastic.19 To summarize, world aggregate demand for vanilla beans is small, highly concentrated, and relatively price inelastic. Conditions for an unstable mar- ket prevail as a result of vanilla beans' high value and inelastic demand. vanilla beans ire primarily used in the food industry as a food ingredient. The nature and characteristics of this demand for vanilla beans are such that competition among importers is limited, and there are potentially severe prob- lems of asymmetric information between vanilla producers and users. These demand characteristics also lead to marketing externalities at the demand level. Finally, the scope for price competition among vanilla beans of different quality and origins varies from one consuniing country to another, and ap- pears to be much more important in the largest world market for vanilla beans (U.S.) than in France. These distinctive features of the demand for vanilla give rise to potential market imperfections and failures, which are being summarized in Table 6.3. 17 Technical and organizational efforts at the production stage (quality of vines, agronomic prac- tices, time of harvest), and at the curing stage are being successfully undertaken in Indonesia, espe- cially on the islands of Bali, Celebes and Sumatra. 18 McCormick & Company (1992). 19 In addition, the ability of Madagascar to charge a premium for its aromatic and other sensory traits tends to diminish at least proportionately with its decline in world trade. The share of high quality vanilla beans produced in Indonesia is extremely difficult to know, but estimates provided by vanilla dealers indicate that, by 1991, at least a third of total Indonesian exports were of good to excellent quality. The continuous improvement in quality of the Indonesian vanilla is reflected in the observed increase of the average price of the vanilla imports from Indonesia, in particular between 1988 and 1992 when the average import price increased from US$ 19/kg. to US$ 31/kg. (see Annex Table 6C). It is also reflected by the much later introduction of the Indonesian vanilla bean on the French market (1986) than on the U.S. (at least since 1980), where a high vanillin content is required on the former market but not on the latter as a result of differences in extract regulations (see above). It is also reflected by the much smaller average import price differential between the Malagasy and the Indonesian vanilla beans in the French than in the U.S. market (35% as opposed to 82% in 1988). 266 MARKETING AFRICA'S HIGH-VALUE FOODS Structure and Composition of World Supply Competition on the supply side is limited, and two countries-Madagascar and Indonesia-dominate the world supply of vanilla beans (see Figure 6.2 and Table 6.3). This leads to a situation of market power by the producing countries on the world vanilla market. Market power, when combined with a price-inelastic demand, potentially gives rise to significant instability on the world market. It also provides countries possessing market power with the ability to extract monopoly rents from the world market. This concentration of vanilla production in a handful of countries partly results from the exact- ing agro-climatic conditions its cultivation requires, limiting entry by new countries.20 Madagascar is by far the largest vanilla bean producer, with an average production of 1,200 tons per year over the period 1980-1990. Mada- gascar has until recently dominated the world market for vanilla, supplying at least 70% of world demand until 1980, and as much as 90% in the early 1970s. After a spectacular boom during the 1980s, Indonesia has now become the world's largest vanilla supplier, its exports jumping from 80 tons in 1980 to 666 tons in 1991,21 surpassing total exports from Madagascar (644 tons in 1991). Comoros ranks third in production and exports, its exports averaging 140 tons over the period 1980-1990, but with large fluctuations from one year to another. The French island of Reunion was the fourth largest pro- ducer and exporter, but its exports have been steadily declining from 58 tons in 1965 to only 13 tons in 1991, and it is now being overtaken by Tonga. Other minor producing countries include Mexico (13 tons exported in 1991), French Polynesia (Tahiti), Costa Rica, and more recently (since 1993) Uganda. Vanilla production is highly unstable.22 Instability in production partly results from agronomic and climatic factors that make vanilla highly suscep- tible to disease, rainfall and wind patterns. Unstable production of a high- 20 The vanilla plant, a member of the orchid family, originates from Central America and only thrives in areas located within 25 degrees on either side of the equator, and also requires rich soils. It needs certain elevations and rainfall patterns to survive. It grows best in areas sheltered from strong winds, on gently sloping land, and supported by trees that offer shade. It is highly susceptible to pests and diseases under conditions of excess moisture and overcrowding. In addition, because vanilla production is highly labor-intensive with limited scope for mechanization, it is rarely produced in countries where wage levels are high. This may help explain the production decline in the French island of Reunion. 21 No production data are available, but Indonesia keeps virtually no stocks. 22 For instance, the coefficient of variation of commercialized vanilla production in Madagascar stood at 15% between 1961 and 1979, and 41 % between 1979 and 1990. MARKET VS. GOVERNMENT FAILURES: A CASE STUDY OF THE VANILLA SECTOR IN MADAGASCAR 267 FIGURE 6.2: Share.of World Vanilla Bean Exports in Volume (Re-exports excluded) 1989-1991 1970-1972 Indonesia Other 41% Reunion 43 Indonesia 1% Comoros 11% Madagascar 43% TABLE 6.3: Characteristics of the International Demand for Vanilla and Relationships with Market Imperfections and Failures Food In- gredient. : ml;irl·;: 1 :.~~·,,'·.~x~~.;::x~:;t} ,·,;, 'Price 1 QU'bta .. ·s4yer:s ; ;·· • • •••• ' •• , •• < - ·.<. ·~ Market Imperfections and Failures: 1) Limited Competition x 2) Duopoly Market Structure x x x 3) Market Power x x x x 4) Market Instability x x x x x 5) Assymmetric Information x x x 6) Marketing Externality: ~ -Transaction Costs x x x x 3 rT -At Production Level x x x x ~ -Promotion & Advertising x c: -Food Product Research x ,. ): !!; -Food Regulations x ~ Equity Considerations x " :i: G: =f :;; ,.. c: rr, .., a a a (/ MARKET VS. GOVERNMENT FAILURES: A CASE STUDY OF THE VANILLA SECTOR IN MADAGASCAR 289 physical handling of all vanilla-related operations from the farm to the point of exports, but also in the commercial relationships with external buyers, as well as in the collegial management of the marketing institutions (GNIV, VSF, and even the Vanilla Alliance) and instruments (differentiel, storage fee, commercial and trade policy, quality standards) governing the vanilla sector. In addition, despite an already dominant position on the world mar- ket, Madagascar through the Vanilla Alliance did not exploit its monopoly powers unduly by exacting a high price on the world vanilla market. Instead, it used its monopoly power to build-up- confidence and trust with its few selected buyers; to reduce uncertainty by elaborating with them (and Comoros and Reunion) yearly marketing plans with the view to satisfying their needs at a pre-announced and stable export price; to build-up, maintain and finance important stocks (about a year worth of production, or 1,000 tons) with the view to discouraging speculation and preventing supply shortages in the event of production shortfalls. As we will see in the next section, this framework for collective action and a cooperative stance between the public and the private sector was extremely successful in promoting quality in production and growth in external demand, in triggering a formidable expansion of vanilla exports from Madagascar, as well as in distributing fairly the gains from trade be- tween the growers, the packers and the State. Before turning to the analysis of the performance of the vanilla sector, we need to briefly describe the impor- tant policy changes that took place in the management of the vanilla sector, and the underlying reasons for these changes. From Orderly Market Organization to Perverse Government Control: 1975-1992 Important policy changes in the vanilla sector started to take place in 1975 which perversely affected its performance, and were foremost in causing Madagascar's rapid loss of competitiveness on the world vanilla market and dismal performance since the mid-1980s. To a large extent, these changes have been triggered by events, political and macroeconomic, that are largely exogenous to the vanilla sector. However, the internal weaknesses and incon- sistencies of the (internal and external) trade policies and marketing institu- tions have facilitated, if not precipitated, these changes. It is important to underscore from the outset that between 1975 and 1992, minimal changes have been made to the legal and regulatory framework and to the (internal and external) trade policies governing the vanilla sector. What 290 MARKETING AFRICA'S HIGH-VALUE FOODS changed dramatically, however, was the selective interpretation and use which was made of the policy instruments and marketing institutions by the GOM over time. This points at the structural weaknesses of the vanilla policy frame- work, in particular the lack of flexibility and market-based mechanisms in the trade policy instruments, and the limited involvement of the GNN in the management of collective actions by the marketing institutions, which could have provided a safeguard against perverse intervention either by Govern- ment or individual private sector agents. Three major changes were imple- mented which transformed a sectoral organization predicated on the long term goals of achieving economic efficiency and equity in the sector, to one that was predicated on the short term goals of surplus extraction from both producers and consumers. Progressive Crowding-out of the Private Sector The private sector was progressively marginalized, and its collective role in the management of the sector virtually disappeared. Starting in 1975, the GNN became progressively defunct in its policy advisory role to the GOM; its recommendations with regard to the official prices and storage fee schedules were progressively ignored by GOM; the export levy financing the GNN re- mained unchanged in nominal terms, and payments came increasingly late if ever, starving the GNN of resources. Instead, the GNN became a simple extension of the GOM apparatus for distributing growers' registration cards and other similar administrative functions. The management and strict qual- ity and price controls of the official domestic markets by the GNN were abandoned, and replaced by administrative and nominal controls performed by local administration officials. Between 1975 and 1982, the Vanilla Stabilization Fund lost entirely its financial and management independence from the GOM, becoming a mere extension of the Ministry of Commerce which was overseeing all external trade operations. By contrast with the pre-1973 period, export quotas are now rigidly allocated by the GOM in an increasingly non-transparent man- ner, and this is confirmed by the abolition of the Sales Committee in 1982. It was during the 1980s that the concentration of exports among fewer export- ers and, eventually, the dominance by a single packer-exporter took place. Similarly, the private sector as a group played a lesser role in the annual preparation and discussions of the Vanilla Alliance marketing plans, which become increasingly dominated by GOM representatives. The above noted MARKET VS. GOVERNMENT FAILURES: A CASE STUDY OF THE VANILLA SECTOR IN MADAGASCAR 291 weaknesses of the trade regime and marketing institutions, together with its resulting incentives, played a large part in this concentration of exports among a few packer-exporters, with the collusion of the GOM. To some extent, the socialist revolution explains the increasing reluctance of the new GOM to share its prerogatives with a poorly trusted private sector, in particular the packer-exporters. Surprisingly, even the vanilla grower co- operatives which had been actively promoted by the earlier administration, were also marginalized and quickly disappeared. Macroeconomic events also explain these important policy changes, and their acceleration in the early 1980s. In 1975, the GOM import-substitution, public sector-led industrialization development strategy required important public financial resources. The vanilla sector, as well as other traditional exports (coffee, cloves), became obvious targets for nationalization. In the case of vanilla, however, the GOM did not proceed with the nationalization of the packer-exporters. It did not need to, since the already established marketing institutions (the VSF, the GNIV, the Vanilla Alliance) and instruments sim- ply needed to be brought under its tighter and exclusive control to achieve the same goals. In the early 1980s, as a result of the financial, foreign exchange and macroeconomic crisis unfolding in Madagascar, further financial and management control of the VSF by the GOM was implemented. In particular, in 1982, the GOM instituted a new system whereby the nationalized commer- cial banks distributed the proceeds from vanilla exports between the VSF (by crediting it with the stabilization levy), the packer-exporters (by paying them the guaranteed export prices), and the exchequer (corresponding to the ex- port taxes); this completed the total loss of financial and management au- tonomy by the VSF. This loss is best exemplified by its inability to fulfill its financial obligations, in particular the payment of the monthly storage fees to the packers since January 1990 despite (nominal) accumulated financial sur- pluses amounting to at least FMG 125 billion in 1989/90, and a (nominal) budget amply covering the estimated storage expenditures. Since the mid-1980s, following a stabilization program, Madagascar has embarked on a program of structural adjustments, liberalizing its domestic and external trade, lowering its import tariffs, and gradually lowering and eventually abolishing all export taxes, including coffee and cloves export taxes in 1992. The vanilla sector, however, has remained untouched by these struc- tural reforms: GOM continues to set all domestic and export prices, export taxation is persistently high, and export quotas are still set and allocated by the GOM. 292 MARKETING AFRICA'S HIGH-VALUE FOODS Increasing Domestic Taxation Since 1975, the GOM has increased dramatically its taxation of the vanilla sector. This was done in two complementary ways: by explicitly levying new and additional export taxes (e.g., in 1977 and again in 1985), and in 1982 by implicitly absorbing the annual revenues generated by the stabilization levy as well as the operating surpluses accumulated by the VSF into the GOM budgetary revenues. In other words, the stabilization mechanism increas- ingly became a pure instrument of taxation rather than stabilization. As soon as the accumulated surpluses were controlled and absorbed by the GOM, the VSF effectively ceased to provide any stabilization function. The share of overall taxation in the actual export price (i.e. combining both explicit export taxes and implicit export taxes accruing from the stabili- zation levy) increased from 45% in 1975 to 82% in 1989 and 1990 (see Annex Table 6D and Figure 6.4). Reflecting the increasing weight of taxation, vanilla growers and packers have seen their share of the actual export price drop sharply. While vanilla growers still received a third of the actual export price in 1975, they only received (officially) about 8% of the actual export price in 1990. In practice, FIGURE 6.4: Evolution of the Vanilla FOB Price and its Distribution 140000 120000 D Taxes & Stab. Levy D Packer & Curer 100000 Ill Producer 'QI) ~ 80000 ........ C!J ~ LI.. 60000 40000 20000 0 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 YEAR MARKET VS. GOVERNMENT FAILURES: A CASE STUDY OF THE VANILLA SECTOR IN MADAGASCAR 293 FIGURE 6.5: Evolution of the Official Producer Price and Packing-curing Margins in Real Terms 170 150 0 130 ..., 0 II 110 LO I'- 0) ~ 90 x Q) -0 E 70 50 30 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 YEAR vanilla growers in 1990 were receiving a still lower price for their vanilla beans. Similarly, while packers-exporters received 27% of the actual export price in 1975, their share plummeted to less than 10% in 1990. In practice, most packers did not even receive this price since the overwhelming majority of them were not allowed to export; they only received the monthly storage fees, at least until January 1990 when the GOM ceased to make the monthly payments. It is interesting to note that vanilla growers were even more dis- criminated against than packers. Producer prices in real terms remained constant between 1975 and 1990,34 while packers' margins increased by 40% over the same period (see Figure 6.5 and Annex Table 6G). Clearly, the objec- tive of the stabilization fund had evolved from providing stable producer prices to providing stable and low producer prices. The stabilization fund had be- come a powerful instrument of agricultural taxation. 34 The official producer price improved significantly between 1979 and 1981, a period when vanilla was in short supply on the world market (vanilla was then traded on the world market at about US$ 100/kg.). 294 MARKETING AFRICA'S HIGH-VALUE FOODS Monopoly Rent Extraction from the External Market As illustrated by Figure 6.4, GOM made increasing use of its monopoly power on the world market and extracted large monopoly rents. The official export price, established for each marketing plan by the Vanilla Alliance, was first kept basically constant at about US$ 10/kg. between 1963/64 (the first such marketing plan) and the 1974/75 marketing plan, Madagascar making little use of its already existing monopoly power on the world vanilla market. Be- tween the 1974/75 and the 1978/79 marketing plans, the official export price jumped from US$ 14.50/kg. to US$ 30/kg., a 100% increase in only four years. By the 1984/85 marketing plan, the official export price was further increased to US$ 70/kg., a 133% jump in six years. By the 1989/90 marketing plan, the official export price had reached US$ 74/kg. level, where it has stabilized since. As illustrated by Figure 6.4 and all of the successive increases in taxation (multiple export taxes and stabilization levy) by Madagascar were passed on into higher export prices, made possible by its monopoly position. The rate of overall taxation jumped from 45% of the actual export price in 1975 to 82% in 1990. This policy was highly successful: Madagascar vanilla exports in value terms more than quadrupled in a decade, jumping from about US$ 20 million in the mid-1970s, to close to US$ 90 million in 1987, its highest level ever reached (see Figure 6.6 and Annex Table 6H). Most if not all of these rents accrued to the Government in the form of (gross) overall revenues, either as export taxes or stabilization levy, increasing by 700% in real terms over the same period, from about FMG 1 billion in 1975 to FMG 70 billion in 1990. PERFORMANCE OF THE VANILLA SECTOR: IMPACTS OF SECTORAL POLICIES The Vanilla Export Boom: 1960-1975 After 1960, the vanilla sector entered a "take-off' period, with Madagascar vanilla export volumes increasing fivefold between 1960 and 1974. The pro- gression was steady, growing from 270 tons in 1960 to close to 1400 tons in 1973, even reaching a historical maximum of 1700 tons in 1977. Total value of vanilla exports tripled between 1960 and 1974, increasing from US$ 6. 70 million in 1960 to US$ 20 million in 1974, and reaching US$ 35 million in 1977. MARKET VS. GOVERNMENT FAILURES: A CASE STUDY OF THE VANILLA SECTOR IN MADAGASCAR 295 FIGURE 6.6: Madagascar Vanilla Exports in Value (US$ Million) go-,..--,~~~~~~~~~~~~~~~~~~~~~----,=-~-. ao-(-l~~~~~~~~~~~~~~~~~~~~~- 70-(-1~~~~~~~~~~~~~~~~~~~~~ 60---1-4~~~~~~~~~~~~~~~~~-----, so-(-!~~~~~~~~~~~~~~~~~~ 40-(-1~~~~~~~~~~~~~~-----=~----, 30-(-1~~~~~~~~~~~~~ 20-(-1~~~~~~~~~~---,"'""""~ 10 -H-----==----: 0 M ~ M ~ ~ ro n ~ ~ n ~ ~ M oo ~ oo YEAR The sectoral organization instituted by GOM, despite its weaknesses, was therefore successful in overcoming some of the major market imperfections and failures which hindered the development of the vanilla market. Reliabil- ity and consistency in volume and quality provided the needed incentive for demand in the consuming countries to expand. Complementacy programs of generic advertising and market promotion (e.g., a vanilla label was created and developed on both the U.S. and European markets during that period), as well as lobbying efforts to elaborate and adjust food and extract regula- tions were simultaneously launched in the major consuming countries (U.S., France, Germany) with obvious and positive results on the consumption side. Total world demand increased steadily from 1500 tons in 1965 to 2700 tons in 1977; Madagascar was also successful in establishing itself as the quality standard on the world market, building a positive brand image and reputa- tion for providing high and consistent quality to its clients, which enabled it to maintain a dominant share of the expanding world market (75% on aver- age between 1965 and 1975, even reaching 87% in 1971 and 1972). On the domestic market, the sectoral policies were also successful since Madagascar was clearly able to respond to an increasing demand without jeopardizing its quality standards. It was also successful in extending the gains from trade to the vanilla growers. AB suggested by the almost equal share of the export price between growers, packers and GOM/VSF in 1975 296 MARKETING AFRICA'S HIGH-VALUE FOODS FIGURE 6.7: Madagascar Vanilla Exports in Volume (Tons) 1800 1600 1400 1200 en 1000 c: t2 800 600 400 200 0 60 62 64 66 68 70 72 74 76 78 80 82 84 86 88 90 YEAR (33.3%, 27.4% and 39.4%, respectively), an equitable distribution of income seemed to have been achieved. In addition to being equitable, results also suggest that the distribution was also economically efficient since vanilla grow- ers responded positively: in terms of increased levels of production, but also in terms of maintained high quality production. The latter suggests that the more equitable distribution of income may have helped lessen the inefficien- cie~ resulting from asymmetric information between the growers and pack- ers; other interventions may have also contributed to lessening these ineffi- ciencies, such as the extensive market regulations designed to promote and enforce quality standards, and the extension efforts undertaken at the grow- ers' level; the extent to which the promotion and training aspects were more effective than the regulations and enforcement remains difficult to ascertain. Efficiency and Equity Losses: 1975-1992 The policy changes initiated since 1975 triggered dramatic reactions on the world market, both on the demand and the supply side, which eventually percolated into the domestic sector only to result in a full-blown crisis of the vanilla sector. The costs, both on efficiency and equity grounds, were ex- tremely large, which Madagascar, the tenth poorest country in the world, could ill-afford. MARKET VS. GOVERNMENT FAILURES: A CASE STUDY OF THE VANILLA SECTOR IN MADAGASCAR 297 Efficiency Losses on the External Market On the external market, the policy changes induced an extreme loss of com- petitiveness for the Malagasy vanilla. As illustrated by Figure 6.8, reactions on the world market to the monopoly and inflexible pricing strategy (FOB prices are indicated by the solid line in Figure 6.8) pursued by Madagascar since the mid-1970s, throughout the 80s and until 1993, were swift. Although Madagascar supplied 87% of the world vanilla market in 1971, its market share was reduced by more than half to only 31% in 1991. Export volumes fell dramatically from its historical maximum of 1700 tons in 1977 to only 640 tons in 1991. This spectacular fall in market share and in export volume was for a long time masked by the equally dramatic progression in export earnings caused by the steadily escalating export prices through the mid- 1980s. Notwithstanding the spike in export receipts in 1987, vanilla export earnings (in nominal terms) reached their maximum as early as 1983 and declined ever since. Export earnings in 1991and1992 are no higher (in nomi- nal terms) than they were ten years earlier. Reactions from the Supply Side Competitors were fairly quick in their reactions given the relatively low short- term elasticity characterizing vanilla production, and the natural, agro-cli- FIGURE 6.8: Vanilla Export Performance 90 80 80 70 70 60 ~ Ill .r:: 50 (/) Qi ~ ta 40 ~ 30 30 20 20 10 10 0 0 ~~~~~~~n~~~ronnNM~~~M~oo~~~oo~ Year 298 MARKETING AFRICA'S HIGH-VALUE FOODS matic constraints on it. The major gainer was Indonesia whose total exports jumped from 80 tons in 1980 to 666 tons in 1991, becoming the world largest exporter in 1989 and again since 1991. It is interesting to note that Indonesia's exports had already progressed in the second-half of the 1970s but collapsed in 1979 (17 tons) and 1980 (80 tons), most probably as a result of a production crisis. This suggests that the international price and market stability intro- duced by the Vanilla Alliance since the first marketing plan of 1963/64 ben- efited other non-Vanilla Alliance producing countries as well, prompting them to increase their own production. The marketing policy pursued by the Va- nilla Alliance (pre-announced quotas and prices, stability) permitted Indone- sia to plan production levels, develop a marketing strategy, and estimate revenues. Indonesia seized the opportunity to implement an aggressive com- mercial strategy based on under-cutting the price of Bourbon vanilla. The high pricing policy pursued by Madagascar since the mid-70s increased even further the incentives for Indonesia to expand its production, accelerating Madagascar's decline. The marketing strategy pursued by Indonesia on the world market is im- portant to detail because it carries important implications for the future of Madagascar's vanilla exports. Indonesia's marketing strategy consisted of the following elements: (1) growth in production, (2) progressive and selective extension of geographic coverage by first entering the U.S. market followed by the European market; (3) progressive improvement in quality; (4) market segmentation by price, according to quality and country; and (5) distribution through a large number of buyers without quota. Following this marketing strategy, Indonesia first targeted the U.S. market where, as a result of favor- able extract regulations, low quality vanilla beans could easily enter the mar- ket and where price competition was promoted. As indicated earlier, Indone- sia progressively took the necessary technical and organizational measures on its domestic market to start improving the quality of its vanilla beans, and gradually moved up-scale within the U.S. market segments, as well as even- tually into the European market where high quality (high vanillin content) imposed by the extract regulations constituted an important barrier to entry and to competition. As illustrated by Annex Table 6C and Figure 6.9, Indone- sian vanilla exports concentrated first on the U.S. market, where it became the largest exporter since 1989; Indonesia started to export to the German market in 1984, and to the French market in 1986, its market share rapidly expanding over time in these two countries. MARKET VS. GOVERNMENT FAILURES: A CASE STUDY OF THE VANILLA SECTOR IN MADAGASCAR 299 FIGURE 6.9: Evolution of U.S. Market Shares 80 70 60 50 Market 40 Share 30 20 10 0 80 81 82 83 84 85 86 87 88 89 90 91 92 YEAR AB a result of Indonesia's ability to gradually satisfy increasing quantities of the many quality segments of the vanilla market in the main consuming countries, Madagascar is not only losing its overall market share. It is also becoming a residual exporter which no longer enjoys a monopoly position on the world market. Its exports are now only confined to the highest quality segments of the world market: Madagascar can now only export its highest quality beans (longer than 14 cm in length) at the request of its buyers, whereas it used to be able to export all its grades, including the vanilla beans below 14 cm and even its cuts. Even within these highest quality segments, Madagascar faces increasing competition from Indonesia which will continue to expand as long as Madagascar does not modify its own export trade policy. Madagascar has until now been able to maintain a hand in world trade due to constraints on Indonesia's production capacities and in improving its pro- duction and curing capabilities (the same market failures Madagascar had been facing). Thus, unless it changes its commercial strategy, Madagascar's export future in vanilla increasingly depends on the climatic uncertainties and other production shortfalls in Indonesia, and Indonesia's difficulties in overcoming its supply problems. 300 MARKETING AFRICA'S HIGH-VALUE FOODS REACTIONS FROM THE DEMAND SIDE. In addition to the above reactions on the supply side, Madagascan policy also triggered significant reactions on the demand side. The first and obvious reaction on the demand side was to in- duce a change in the consumption patterns away from Bourbon vanilla and towards Indonesian vanilla. This is extremely damaging for Madagascar's future since it implies that Madagascar has lost its quality edge, no longer being the standard of reference in a market where quality considerations are paramount. The demand for vanilla faced by Madagascar has therefore be- come increasingly price-elastic over time as well as subject to price competi- tion by Indonesia, making its monopolistic policy even more incongruous. In addition to triggering lasting changes to the shape of its individual vanilla demand curve, Madagascar's monopolistic and inflexible policy has also accelerated, or at least facilitated, the development of vanilla substitutes (synthetic vanillin, vanilla with other natiral flavors, and bio-technology prod- ucts). The development of these substitutes in the course of the 1980s may have resulted in reducing the aggregate demand for natural vanilla, as well as in making the aggregate demand for vanilla more price-elastic. Recent econometric estimation35 of the vanilla demand curves faced by Madagascar and Indonesia would appear to confirm the increasing price- elasticity of the demand faced by Madagascar, as well as the greater substi- tutability and competition from Indonesia vanilla. The econometric results suggest that, for the individual demand curve faced by Madagascar, the own- price elasticity is becoming larger (in absolute value) over time, jumping from -0.0477 in 1974 to -0.954 in 1986, and becoming larger than unity by 1988; the cross-price elasticity of Indonesia vanilla also increases over time, from 0.096 in 1966 to 2.88 in 1996. The econometric estimation of the individual demand curve faced by Indonesia indicates an own-price elasticity of -0.071 in 1966 and of -2.13 in 1996, becoming larger than unity by 1981; and a cross-price elasticity of Madagascar vanilla equal to 0.09 in 1966 and to 2. 7 in 1996. Efficiency Losses on the Domestic Market Madagascar's loss of competitiveness on the world market is also having im- portant efficiency repercussions on the domestic market: export earnings are falling, the domestic stabilization mechanism has collapsed, yields from va- 35 See de Melo, Olarreaga & Takacs (forthcoming). MARKET VS. GOVERNMENT FAILURES: A CASE STUDY OF THE VANILLA SECTOR IN MADAGASCAR 301 nilla taxation are plummeting, the framework for safeguarding quality has disappeared, and the domestic sector is now being dominated by a single packer-exporter. FALLING EXPORT EARNINGS. The fall in export earnings is even more damag- ing to Madagascar in that, following the successive collapse of the clove and coffee markets during the 1980s (its former two main exports), vanilla has become its largest export earner. Except for shrimp, non-traditional exports do not yet represent a viable alternative until major efforts are made by Madagascar to rehabilitate its basic infrastructure and to attract foreign investments. COLLAPSE OF THE STABiLIZATION SCHEME. While exports dwindled, domestic production remained high as a result of the stabilization scheme which, by maintaining a low but stable producer price, sheltered domestic producers and packers from the lack of performance on the world market. This excess- supply naturally resulted in an impressive stock build-up throughout the 1980s, reaching close to 4,000 tons in 1992 (see Figure 6.10) at enormous financial costs to the Vanilla Stabilization Fund and the exchequer (see An- nex Table 61). Escalating stocks imply huge financial outlays for the VSF, whose man- date is to pay storage fees to the packers. Storage costs exploded from FMG 10 million in 1979/80 to more than FMG 9 billion in 1989/90, a 17,860% increase in real terms in only ten years, accounting for an ever increasing share of the VSF operating budget. Eventually, the system collapsed in Janu- ary 1990, when the GOM simply ceased to make any further monthly storage fee payments to the packers. From that point on, free market forces fully resumed their course on the domestic market, and actual prices to the grow- ers dropped dramatically (to, at best, less than 50% of the official price). The FMG 125 billion accumulated surpluses by the VSF until 1989/90 had long been spent, and the GOM was financially unable to face its contractual obligations with the packers as well as the vanilla growers who had been finandng the VSF since its inception. By 1990, the vanilla stabilization scheme had therefore ceased to perform the functions for which it was initially man- dated (guaranteed price and market/storage outlet), although it continued to restrict exports and to levy a 50% stabilization tax on vanilla growers. Even before its financial collapse in 1990, the VSF was unable to provide for stability on the domestic market throughout the 1980s. The increasing 302 MARKETING AFRICA'S HIGH-VALUE FOODS gap between export and production levels implied that fewer packers were permitted by GOM to export. Packers became increasingly dissatisfied with their inability to export and therefore to receive the guaranteed export price for their vanilla. For the large majority of packers who could no longer export, their only source of income became restricted to the monthly storage fees only. This increased the pressure for illegal exports,36 leading in turn to more coercive measures by GOM to maintain its monopolistic strategy on the world market, as well as to increasing instability and disruptions on the world market. Packers who could no longer export faced increasing finan- cial constraints from the commercial banks, and had no choice but to reduce and adjust their level of purchases to the level of credit they were able to obtain. Increasingly, a situation of excess-supply developed on the domestic market and domestic prices adjusted downward accordingly, the packers no longer complying with official prices. Vanilla purchases by packers became increasingly erratic as Figure 6.10 clearly illustrates (the coefficient of varia- tion, an indicator of instability, in the volume of officially commercialized production jumped from 15% over the period 1961-1979, to 41% over the period 1979-89). The VSF attempted to remedy the situation by buying back about 800 tons from packers in 1986/87, but the solution was short-lived since it did not address the root cause of the problem, namely the loss of external competitiveness. FALLING YIELDS FROM VANILLA TAXATION. In addition to the above efficiency costs, vanilla taxation is also becoming less and less attractive for the GOM. The explosion in the costs of the stabilization scheme associated with storage costs means that the net returns and yields from the taxation imposed on the vanilla sector are falling rapidly. While taxation yields were as high as 91 % in 1983, they had fallen to 54% in 1990. In real terms, the net returns to the GOM from the vanilla taxation system were no larger in 1989 and 1990 than they were in 1983. The financial implications for the GOM are significant since vanilla taxation represents its single largest source of revenues: 14% of total budgetary revenues in 1989, but only 8.8% in net terms once the costs of the stabilization scheme were deducted. This helps explaining why the GOM ceased all storage fee payments since 1990. 36 See in Annex Table 6H the gap between theoretical stocks and officially reported stocks. MARKET VS. GOVERNMENT FAILURES: A CASE STUDY OF THE VANILLA SECTOR IN MADAGASCAR 303 FIGURE 6.10: Escalating Stock Levels 5000 Production 4500 Exports 4000 Theoretical Stocks 3500 en Q.l 3000 c: c: 2500 i2 2000 1500 1000 500 0 80 86 87 88 89 90 YEAR DISINTEGRATION OF THE SECTORAL ORGANIZATION. The disappearance of both a guaranteed price and a guaranteed outlet, combined with that of any qual- ity control and enforcement (for domestic markets as well as for curing and storage practices and even at export) led to a total collapse of the sectoral organization designed to maintain high quality standards at all levels in the sector. Independent curers and even sometimes packers can no longer afford the neccesary material to ensure quality curing; agricultural extension ser- vices are no longer being provided to growers, no new planting material is being introduced to rejuvenate the vanilla vines. Quality is therefore deterio- rating and consistency in quality can no longer be maintained, making it increasingly difficult for Madagascar to satisfy the needs of the high quality segment of the world market it has restricted itself to. The development of illicit exports, more often than not of poor quality, also undermines the confi- dence of foreign buyers by damaging the reputation of Madagascar. DOMINANCE BY A SINGLE PACKER-EXPORTER. An additional loss of efficiency results from the fall in exports. Rather than sharing more equally the shrink- ing export volumes among packers in order to maintain a semblance of com- 304 MARKETING AFRICA'S HIGH-VALUE FOODS petition within the private sector, GOM instead has promoted the concentra- tion of the domestic vanilla sector into the hands of a single packer-exporter, effectively supporting the creation of a monopsonist37 who is at the same time a monopolist.38 This concentration does not result solely from a vastly supe- rior technical efficiency on his part. Instead it is consistent with the lack of transparency and perverse incentives inherent in the trade policies which result in a collusive behavior between the large exporter(s) and GOM. As a result of such a privileged position, efficiency losses for the sector are poten- tially large since the monopsonist can extract huge rents from the vanilla growers. Additional efficiency losses may also result for Madagascar if the monopolist-monopsonist uses its dominant position to capture rents from the external market which would otherwise accrue to the StateNSF as taxation revenues. Equity Losses on the Domestic Market In addition to the above efficiency losses, the disintegration of the vanilla sector organization in Madagascar also carries important equity implications. Clearly, the vanilla growers have suffered the most from a system which was originally designed to protect them from market imperfections and failures. They paid a high price to finance the VSF by accepting a small and declining share of the value of their production, and received no return from their investment. The VSF has virtually ceased to exist, no longer able to guaran- tee a minimum price and a market. outlet for their production. For example, in 1992, in a desperate political effort to defuse the mounting social and economic crisis in the vanilla growing region, GOM almost doubled the offi- cial producer price without taking the other necessary steps to make that price effective (e.g., paying the overdue storage fees, radically changing its commercial policy on the world market in order to boost exports and reduce the excess supply on the domestic market). At the same time, GOM announced unilaterally that, from now on, only vanilla beans of at least 14 cm in length will be "supported" by the VSF, introducing new restrictions which never existed before. Despite the high taxation levels extracted by the GOM from the vanilla region, little was ever plowed back into the region. The vanilla region, despite 37 A single buyer of green and cured vanilla on the domestic market. 38 A single exporter of vanilla on the world market. MARKET VS. GOVERNMENT FAILURES: A CASE STUDY OF THE VANILLA SECTOR IN MADAGASCAR 305 its wealth and agricultural potential, is virtually isolated from the rest of the country; there are no operating phones, little in terms of public services, and local roads are in extremely poor conditions; the only possible means of com- munication are either by boat or plane.39 The current dominance of the vanilla sector. by a single packer-exporter also carries important equity implications, besides the welfare costs imposed on vanilla growers by the virtual monopsonistic structure of the domestic vanilla sector. The interests of the dominant single packer-exporter are now at odds with the interests of the sector as a whole. It is in the economic interest of the dominant packer-exporter to pursue an inflexible monopolist commercial strategy on the world market, to maximize and capture rents (from both the external and domestic markets) in the short-to-medium-term, investing the profits in other activities. Instead, it is in the economic interests of the sector as a whole to radically change its. commercial policy on the world market, in order to re-establish the conditions for a sustainable source of revenues and income. CONCLUSION More often than not, market failures and imperfections pervade agricultur- al markets. The vanilla case study provides such an illustration. There are multiple solutions for dealing with such market failures and imperfections, which were presented in a more general context in Chapter II. The respective outcome of each of these solutions will differ in their efficiency and equity impacts. The solution to these supply problems, adopted by Madagascar at the time of Independence, consisted of creating an orderly market that would rely on collective actions between the growers, packers, Government and importers. Initially, these marketing institutions and policy instruments were successful in achieving a more efficient and equitable market outcome. Exports multi- plied several fold, both in value and volume, high quality standards were attained, stability and consistency in supplies provided the needed stimulus 39 A General Equilibrium Model developed for Madagascar (Dorosh 1993) also points to the fact that increased Government revenues and spending worsen rather than improve income distribution, favoring urban areas at the expense of rural areas, and within urban areas the higher income groups, further questioning the economic and social justification of the high level of taxation imposed on the vanilla growers. 306 MARKETING AFRICA'S HIGH-VALUE FOODS for demand to expand, Madagascar's dominant position on the world market was managed in a sustainable fashion, and the gains from trade were equita- bly distributed domestically between growers, packers and Government. Very quickly, however, the marketing institutions and policy instruments previously so successful were progressively diverted by the Government from their original objectives. The instruments of collective action were no longer used to serve long-term efficiency and equity objectives, but became increas- ingly a powerful instrument in the hands of the Government and a handful of packer-exporters for rent extraction from both growers and the world market. This diversion can be traced back to Government failures in setting these marketing institutions and policy instruments. AB the Madagascar vanilla case study illustrates, these Government failures have resulted in large effi- ciency and equity costs to its vanilla sector as well as its economy. Five main Government failures in its intervention on the vanilla market can be identified. Combined together, these Government failures have con- tributed to the myopic over-exploitation of Madagascar monopoly power on the world market, and to the concentration of exports among a handful of exporters. First, because the external policy regime is over-determined (since it sets both the export price and the export quota), it provides for little flex- ibility in the face of changing external conditions on the world vanilla mar- ket. The latter could result from structural (and/or random) changes either on the demand side (e.g., changes in price-elasticity, shifts in demand), or on the world supply side (e.g., development of competition). Second, the lack of market-based mechanisms (such as an auction system) for setting and allo- cating both the individual export and import quotas promotes rent-seeking activities among the packer-exporters and Government officials. Third, the vanilla price stabilization scheme through its collective storage policy, by delinking the domestic market from external market conditions, magnifies the development of structural imbalances between domestic supply and ex- ternal demand. Fourth, the price regime (i.e., unique prices irrespective of quality) does not provide any financial incentives for improving quality, un- dermining the effectiveness of other complementary measures (e.g., market regulations; quality grades; production, curing and packing standards). Fifth, because Government of Madagascar increased, rather than decreased, its ex- MARKET VS. GOVERNMENT FAILURES: A CASE STUDY OF THE VANILLA SECTOR IN MADAGASCAR 307 elusive role in the conduct of the marketing institutions (GNIV, VSF, Vanilla Alliance) designed to play a central role in the collective management of the vanilla sector, the voicing of protest by different interest groups (growers, packers, importers) became virtually impossible.· Consequently, there was no countervailing mechanisms for correcting the growing inefficiencies and ineq- uities of the vanilla sector policies.40 Market failures or imperfections often impede the efficient and equitable development of agricultural markets. Failures in Government interventions may also be as, if not more, damaging. What the Madagascar case study indicates is that Governments should be cautious, selective and flexible in their market interventions, and to the maximum extent rely on market-based intruments as well as the articulation of interests in organizations (such as trade associations, professional or inter-professional organizations). Con- versely, when contemplating the liberalization of agricultural markets (or the promotion of agricultural diversification), policy analysts should be aware of the often genuine concerns of Governments for the efficiency and equity im- pacts of market failures and imperfections, and the design of efficient solu- tions to these supply problems should be addressed. 4 ° For a full treatment of market and non-market mechanisms, and their complementarity, for addressing conduct and performance issues in economic systems, see Hirschman (1970). ANNEX TABLE 6A: World Vanilla Imports (Tons) North & Central America Europe South Asia Sub- Ger- Sub- Amer- Sub- Oce- Year USA Canada Other Total France many Other Total ica Japan Other Total Africa ania World 1965 1008 29 0 1037 160 76 99 335 11 10 79 89 1 14 1487 1966 864 39 0 903 156 95 125 376 11 13 4 17 1 21 1329 1967 591 38 0 629 171 103 129 403 15 14 6 20 4 12 1083 1968 1017. 41 2 1060 175 126 134 435 14 17 9 26 7 18 1560 1969 840 43 2 885 233 118 132 483 17 22 8 30 13 19 1447 1970 1016 53 1 1070 234 118 124 476 10 40 11 51 5 20 1632 1971 923 45 2 970 245 113 111 469 8 30 6 36 6 20 1509 1972 991 47 2 1040 263 111 104 478 8 27 6 33 6 20 1585 1973 1056 54 3 1113 296 114 153 563 19 38 16 54 10 20 1779 1974 977 47 4 1028 277 160 176 613 19 40 32 72 12 30 1774 1975 963 50 35 1048 199 125 141 465 15 23 18 41 24 25 1618 1976 1014 65 1 1080 325 192 161 678 14 62 19 81 13 17 1883 1977 1554 183 0 1737 417 266 213 896 14 57 7 64 13 18 2742 1978 1185 43 33 1261 459 291 182 932 30 75 59 134 20 24 2401 1979 497 40 38 575 302 164 132 598 32 52 48 100 9 28 1342 1980 343 26 2 371 133 124 130 387 17 34 69 103 9 8 895 1981 640 19 29 688 248 183 169 600 30 21 79 100 20 9 1447 1982 884 30 18 932 301 182 134 617 21 19 8 27 1 5 1603 1983 977 29 10 1016 261 200 144 605 19 42 9 51 6 13 1710 1984 841 31 2 874 202 175 165 542 5 36 73 109 12 7 1549 1985 743 31 2 776 154 137 135 426 1 40 31 71 2 7 1283 1986 1001 28 1 1030 199 190 166 555 3 60 13 73 3 7 1671 1987 1387 58 4 1449 220 256 163 639 12 38 66 .104 5 6 2215 1988 1224 71 5 1300 267 203 195 665 5 79 59 138 8 10 2126 1989 1107 36 13 1156 253 232 189 674 23 48 65 113 37 15 2018 1990 975 110 15 1100 220 294 218 732 8 59 116 175 16 16 2047 1991 1311 56 18 1385 245 279 216 740 17 60 49 109 13 29 2293 SOURCE: FAO Trade Yearbook. ANNEX TABLE 68: World Vanilla Exports (Tons) North & Central America Europe Asia Re- Sub- Re- South Re- Sub- Year Mexico Other exports Total exports America Indonesia exports Total 1965 39 2 0 41 46 0 59 5 64 1966 22 1 0 23 40 0 56 4 60 1967 70 0 0 70 62 0 108 0 108 1968 54 0 0 54 67 0 86 4 90 1969 18 0 0 18 102 0 15 1 16 1970 22 0 0 22 80 0 15 1 16 1971 23 0 0 23 83 0 15 4 19 1972 18 0 0 18 77 0 15 4 19 1973 19 0 0 19 109 0 317 7 324 1974 15 0 0 15 118 0 53 3 56 1975 0 0 0 0 122 0 237 7 244 1976 0 0 0 0 161 0 334 0 334 1977 0 0 0 0 239 0 420 0 420 1978 2 0 0 2 334 0 389 11 400 1979 1 0 0 1 285 0 17 25 42 1980 0 0 0 0 150 0 80 24 104 1981 0 0 0 0 192 0 138 81 219 1982 0 0 0 0 205 0 116 0 116 1983 0 0 0 0 222 0 234 3 237 1984 24 0 0 24 245 ' 0 154 1 155 1985 1 0 0 1 199 0 175 7 182 1986 0 0 0 0 222 0 298 10 308 1987 30 1 8 39 252 0 410 77 487 1988 27 0 67 94 296 0 506 16 522 1989 36 20 31 87 290 0 677 32 709 1990 14 0 78 92 338 11 607 23 630 1991 13 2 92 107 333 0 666 6 672 SOURCE: FAO Trade Yearbook. ANNEX TABLE 68: World Vanilla Exports (Tons) (continued) Africa Oceania FR. World Mada- Sub- Poly- Re- Sub- Exel. Re- gascar Comoros Reunion Total nesia Tonga Exports Total World export 984 167 58 1213 132 0 0 132 1496 1445 885 134 16 1046 115 0 0 115 1284 1240 666 145 18 837 59 0 0 59 1136 1074 961 138 25 1132 67 0 0 67 1410 1339 1097 207 32 1344 55 0 0 55 1535 1432 1218 144 26 1392 33 0 0 33 1543 1462 1160 73 31 1269 27 0 0 27 1421 1334 1215 100 21 1341 25 0 0 25 1480 1399 720 34 21 775 19 0 0 19 1246 1130 1353 160 4 1517 12 0 0 12 1718 1597 858 211 8 1077 8 0 0 8 1451 1322 1101 124 19 1244 7 0 0 7 1746 1585 1713 100 13 1826 7 0 0 7 2492 2253 1459 117 9 1585 3 0 0 3 2324 1979 436 170 23 629 3 0 0 3 960 650 410 13 8 431 1 0 0 1 686 512 635 160 15 810 1 0 0 1 1222 949 887 259 11 1157 1 0 0 1 1479 1274 1099 177 10 1286 5 0 0 5 1750 1525 827 26 19 872 2 13 0 15 1311 1065 628 181 28 837 3 4 8 15 1234 1020 689 150 21 860 4 8 0 12 1402 1170 1260 63 14 1337 4 14 0 18 2133 1796 625 247 20 892 3 17 1 21 1825 1445 594 164 16 774 4 26 1 31 1891 1537 829 127 7 963 7 36 2 45 2079 1638 644 240 13 897 6 37 13 56 2065 1621 SOURCE: FAO Trade Yearbook. ANNEX TABLE 6C: Respective Performance of Madagascar and Indonesia Vanilla Exports on the ... Main Consuming Markets "' ~ 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 U.S. Total Volume (T) 343 640 884 977 841 743 1048 1387 1217 1107 975 1308 1262 Average CIF Price ($/kg.) 54 49 51 52 59 64 58 56 51 42 49 53 52 Madagascar Volume (T) 205 438 580 651 602 462 672 820 576 421 398 493 471 %Total 60% 68% 66% 67% 72% 62% 64% 59% 47% 38% 41% 38% 37% Average CIF Price ($/kg.) 50 52 55 60 65 69 67 70 69 59 72 74 74 Indonesia Volume (T) 90 123 102 240 192 184 306 443 432 526 482 616 602 % Total 26% 19% 12% 25% 23% 25% 29% 32% 36% 48% 49% 47% 48% Average CIF Price ($/kg.) 42 35 22 28 43 50 37 28 19 21 26 32 31 France Total Volume (T) 133 249 301 261 201 154 198 219 247 219 Average CIF Price (FF/kg.) 350 317 380 479 581 646 517 410 391 Na. Madagascar Volume (T) 92 147 167 202 149 79 92 148 165 84 %Total 69% 59% 55% 77% 74% 51% 46% 68% 67% 38% Average CIF Price (FF/kg.) 588 661 512 409 390 Na. Indonesia ~ :0 Volume (T) 0 0 0 0 0 0 1 11 9 55 ~ :j % Total 0% 0% 0% 0% 0% 0% 1% 5% 4% 25% < Gl Average CIF Price 0 0 0 0 0 0 451 298 256 Na. )> ~ Germany ~ Total Volume (T) 125 183 182 200 185 165 199 265 181 234 ui ~ Madagascar Gl Volume(T) 100 136 143 175 151 137 177 213 130 141 :p. %Total 80% 74% 79% 88% 82% 89% 80% 60% :;:; 83% 72% r- ,.,, c:: Indonesia ..,. 0 Volume (T) 0 0 0 0 2 0 1 11 11 43 0 % Total 0% 0% 0% 0% 1% 0% 1% 5% 8% 30% ~ SOURCE: U.S. Spice Trade, TJP Market Development. MARKET VS. GOVERNMENT FAILURES: A CASE STUDY OF THE VANILLA SECTOR IN MADAGASCAR 313 ANNEX TABLE 60: Madagascar Vanilla Official Domestic Prices, Actual Export Prices and Government Proceeds from Export Taxes and Stabilization Levy Official Guaranteed Export Price to Packer-Exporter Packing & Official Producer Price Curing Margins Proceeds from %of: as %of: Export Taxes Guar- Guar- FOB & Stabilization anteed anteed Price Levy (FMG/ Export (FMG/ Export (FMG/ Billion asa% Year kg.)' FOB Price kg.)2 FOB 3 Price3 kg.)• FMG5 ofFOB 6 1975 250 33.3 54.9 2096 27.4 45.1 3457 1.17 39.4 1976 250 26.8 51.7 2226 25.0 48.3 4299 2.28 48.2 1977 280 25.2 53.6 2402 21.8 46.4 5109 4.64 53.0 1978 330 25.6 55.4 2739 20.6 44.6 5925 4.65 53.8 1979 500 32.2 59.8 3849 21.7 40.2 7140 1.43 46.1 1980 600 28.7 64.8 4260 15.6 35.2 9622 2.20 55.7 1981 700 27.3 58.4 5518 19.5 41.6 11,797 3.99 53.2 1982 700 17.1 56.4 5711 13.3 43.6 18,795 11.61 69.6 1983 1000 18.6 59.0 7801 12.9 41.0 24,760 18.64 68.5 1984 1000 12.5 55.6 8275 10.0 44.4 36,759 23.56 77.5 1985 1000 10.0 53.2 8640 8.8 46.8 45,868 23.38 81.2 1986 1100 11.7 53.7 9431 10.1 46.3 43,280 23.32 78.2 1987 1200 7.8 48.6 11,365 8.3 51.4 70,821 74.92 84.0 1988 1700 8.4 46.3 16,893 9.8 53.7 92,850 47.47 81.8 1989 2000 8.1 47.1 19,525 9.1 52.9 113,704 55.94 82.8 1990 2000 8.5 47.1 19,525 9.6 52.9 107,749 69.70 81.9 1 Official Producer Price for green vanilla; 4.6 kg. of green vanilla yields 1 kg. of cured vanilla. 2 Official Guaranteed Price for Cured, Packed & Conditioned Vanilla to the Packer-Exporter. 3 Packing & Curing Margins= [Official Guaranteed Price to the Packer-Exporter] - [Official Producer Price, Cured vanilla equivalent]. ' Unit Export Value. 5 Proceeds = {[FOB price] - [Official Guaranteed Price to Packer-Exporter])*[export volume]. 6 = {[FOB price] - [Official Guaranteed Price to Packer-Exporter])![FOB price]. SOURCES: Annual Differentiels VSF, BDE. ANNEX TABLE 6E: Madagascar Vanilla Financial Results of the Vanilla Stabilization Fund w ~ 1979/80 1980/81 1981/82 1982/83 1983/84 1984/85 1985/86 1986/87 1987/88 1988/89 1989/90 I. Operating Budget D.400 0.550 0.210 0.660 0.820 0.990 0.138 0.132 0.348 0.203 0.206 IL Market Promotion Budget 0.640 0.190 0.130 0.356 0.450 0.427 0.436 1.811 0.882 1.493 0.000 III. Stabilization Budget: 2.141 2.039 3.555 3.886 1.010 1.158 2.759 4.401 4.275 7.421 9.409 Storage 0.100 0.160 0.320 0.651 1.010 1.156 2.759 4.401 4.275 7.421 9.409 Transfer to GOM 2.101 2.023 3.235 3.235 0.000 0.000 0.000 0.000 0.000 0.000 0.000 IV. Other 0.534 0.590 0.100 --0.350 -0.398 -0.416 -0.395 0.880 0.143 0.395 5.071 V. Total Expenditures 2.779 2.172 3.590 4.273 1.144 1.268 2.938 6.432 5.648 9.512 14.686 VI. Total Receipts from Stabilization Levy 3.158 1.530 1.592 11.886 15.563 17.543 11.287 19.850 35.848 26.495 31.066 VIL Annual Surplus -0.821 -1.997 9.570 14.992 15.337 7.491 13.749 30.193 15.894 15.040 VIII. Surplus from Previous FY 5.800 4.979 2.982 12.552 27.544 42.881 50.372 64.121 94.314 110.208 IX. Cumulated Surplus 4.979 2.982 12.552 27.544 42.881 50.372 64.121 94.314 110.208 125.248 Madagascar Vanilla: Proceeds from Export Taxes ~ :0 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 :s:; ni :::! DROITS DE SORTIE 0.9 2 3.27 3.65 4.0. 5.0 12.0 7.6 8.6 9.9 <: t;') ~ TAXE CONJONCTURELLE 0.66 1.47 2.39 2.68 3.0 3.0 9.0 5.8 8.5 7.2 ~ §;? PRELEVEMENT SPECIAL 0 0 0 0 1.3 1.3 7.0 4.7 2.3 5.7 cii ::!.= t;') TOTAL 1.6 3.5 5.7 6.3 8.3 9.3 28.0 18.1 19.4 22.8 =!= ~ ,... NOTE: The fiscal year for the VSF runs from May 1 to April 30. c: ni SOURCE: VSF, Ministry of Finance. (51 0 0 (/) ANNEX TABLE 6F: Madagascar Vanilla Marketing Export Plans Export Quotas Volumes Actually Exported FOB Plan U.S. Europe Other Total Price U.S. Europe Other Total Number Year (T) (T) (T) (T) (US$/kg.) (T) (T) (T) (T) 2 1963/64 10.20 261 28 289 3 1964/65 10.20 579 49 628 4 1965/66 10.20 821 163 984 5 1966/67 10.20 533 175 4 712 6 1967/68 10.20 428 217 15 660 7 1968/69 10.20 685 270 20 975 8 1969/70 10.70 745 331 20 1096 9 1970/71 10.70 780 413 25 1218 10 1971/72 11.30 810 317 33 1160 11 1972/73 12.40 771 407 36 1214 12 1973/74 13.40 680 390 22 1092 13 1974/75 14.50 740 450 31 1221 14 1975/76 17.50 400 400 40 840 15 1976/77 19.30 776 437 61 1274 16 1977/78 23.00 950 600 50 1600 17 1978/79 30.00 720 384 5 1109 18 1979/81 45.00 495 266 28 789 19 1981/82 55.00 420 238 8 666 20 1982/83 57.00 566 300 43 909 21 1983/84 62.50 575 300 44 919 22 1984/85 404 211 35 650 70.00 461 211 46 718 23 1985/86 659 367 87 1113 70.00 556 243 50 849 24 1986/87 456 209 88 753 70.00 735 315 50 1100 25 1987/88 770 331 83 1184 72.00 698 336 50 1084 26 1988/89 421 166 56 643 72.00 460 268 50 778 27 1989/90 330 165 136 631 74.00 375 237 50 662 28 1990/91 420 244 80 744 74.00 450 256 50 756 29 1991/92 500 305 100 905 74.00 400 210 40 650 Marketing Plan runs from April 1 to March 31. SOURCE: Vanilla Stabilization Fund. 316 MARKETING AFRICA'S HIGH-VALUE FOODS ANNEX TABLE 6G: Madagascar Vanilla Official Producer Prices, Marketing Margins and Government Proceeds from Export Taxes and Stabilization Levy in Real Terms (Index 1975 = 100) Official Official Marketing Proceeds Producer Margin From Export Taxes Year Price (Curing & Packing) & Stabilization Levy 1975 100 100 100 1976 95 108 145 1977 103 108 183 1978 114 111 202 1979 152 124 184 1980 154 102 253 1981 138 120 227 1982 105 99 360 1983 125 106 390 1984 114 111 597 1985 104 111 712 1986 99 104 560 1987 94 121 852 1988 105 148 862 1989 113 155 980 1990 103 140 831 SOURCE: Differentiel VSF, BDE. MARKET VS. GOVERNMENT FAILURES: A CASE STUDY OF THE VANILLA SECTOR IN MADAGASCAR 317 ANNEX TABLE 6H: Madagascar Vanilla Production, Exports in Volume and Value Terms Theoretical and Officially Declared Stocks Commer- Export Export Export Export Theo- cialized Exports Value Unit Value Unit retical Declared Production Volume (US$ Value (FMG Value Stocks Stocks Year (tons) (tons) Million) (US$/kg.) Million) (FMG/kg.) (tons) (tons) 1960 270 6.70 24.80 1642 6081 1961 585 7.69 13.14 1885 3222 1962 650 640 8.82 13.78 2161 3377 10 "'.), 1963 1050 292 4.27 14.64 1048 3589 768 1964 856 628 6.46 10.28 1583 2521 996 1965 850 984 9.94 10.10 2437 2477 862 1966 950 885 8.95 10.11 2216 2504 927 1967 983 666 6.81 10.22 1672 2511 1244 1968 654 961 10.23 10.64 2530 2633 937 1969 1310 1097 10.84 9.88 3013 2747 1150 1970 1330 1218 13.07 10.73 3610 2964 1262 1971 1508 1160 13.64 11.76 3565 3073 1610 1972 1690 1215 14.91 12.27 3819 3143 2085 1973 660 720 9.41 13.07 2217 3079 2025 1974 800 1353 20.13 14.88 4475 3307 1472 1975 764 858 13.84 16.14 2966 3457 1378 1976 859 1101 19.82 18.00 4733 4299 1136 1977 847 1713 35.62 20.79 8751 5109 270 1978 726 1459 38.31 26.25 8644 5925 ---463 1979 772 436 14.74 33.81 3113 7140 -127 1980 900 410 18.67 45.53 3945 9622 363 1981 1034 635 27.79 43.77 7491 11,797 762 1982 1195 887 46.97 52.95 16,671 18,795 1070 1983 526 1099 62.45 56.83 27,211 24,760 497 1984 1508 827 52.82 63.86 30,400 36,759 1178 1985 1761 628 43.80 69.74 28,805 45,868 2311 1986 695 689 48.02 69.69 29,820 43,280 2317 1987 2164 1260 88.75 70.43 89,235 70,821 3221 2194 1988 804 625 43.61 69.77 58,031 92,850 3400 2552 1989 1681 594 41.94 70.61 67,540 113,704 4487 2019 1990 972 790 56.97 72.11 85,122 107,749 4669 3256 NOTE: Commericalized production refers to cured vanilla production purchased by packers. Theoretical stocks in year tare assumed to equal to: stocks in year (t-1) plus, the difference between production and exports in year t; it is also assumed to start from zero stock level in 1962. SOURCE: Vanilla Stabilization Fund, BDE. 318 MARKETING AFRICA'S HIGH-VALUE FOODS ANNEX TABLE 61: Madagascar Vanilla Gross, Net Proceeds and Yields from Vanilla Taxation System 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 Gross Proceeds (FMG billion): [FOB price- Official Guar- anteed Export Price to Packer] 3.99 11.61 18.64 23.56 23.38 23.32 74.92 47.47 55.94 69.70 Net Proceeds: CurrentFMG (billion) 0.3 5.3 17.0 21.4 21.0 18.9 47.2 43.5 35.0 32.8 Constant 1980 FMG (billion) 0.1 2.9 9.2 10.4 9.3 7.1 14.9 11.9 8.2 6.9 Yields: Net Proceeds/ Gross Proceeds 9 46 91 91 90 81 63 92 63 54 NOTE: Net proceeds = [proceeds from export taxes] + [annual surplus from VSF]. SOURCE: VSF, Ministry of Finance. VII. The Many Faces of Success: The Development of Kenyan Horticultural Exports Steven Jaffee 0 ver the past two decades, international trade in horticultural products, such as fresh and processed fruit and vegetables, cut flowers and other . ·ornamental foliage, has been one of the most dynamic components of interna- tional agricultural trade, featuring a 10% average annual increase in value.1 With this growth in horticultural trade and the stagnation or decline in world trade for many other agricultural commodities during the 1980s, the value of fruit and vegetable products imported by OECD and developing countries now exceeds those for any other category of agricultural products, including cereals, oilseeds, fish products, meats and tropical beverages. Many developing countries have participated in this expanding trade in horticultural products. Some countries have based their trade on high volume commodities such as bananas, pineapples and citrus fruits; other countries have targeted niche markets for high value, low volume specialty items and OECD market 'off-season' produce. Developing countries have competed well in many horticultural product markets. While the share of these countries in total world agricultural trade declined between the early 1960s and mid- 1980s, their share of world horticultural trade increased, reaching nearly 37% by the end of this period.2 Many sub-Saharan African countries have a potential comparative advan- tage in the production of fresh and processed horticultural products as a 1 GATT (1990); Islam (1990). 2 Islam (1990), p.17. 319 320 MARKETING AFRICA'S HIGH-VALUE FOODS result of their agro-ecological conditions, location and relatively low labor costs. Very few African countries and entrepreneurs, however, have effec- tively translated these resource advantages into competitive and profitable horticultural trades with Europe, the Middle East or other regions. The rea- son is that most African countries lack the physical infrastructure and the technical, managerial and marketing skills necessary for competitive horti- cultural export development. For the vast majority of SSA countries, nascent horticultural export industries have floundered due to weak production sup- port systems, international transport and other infrastructural bottlenecks, ineffective marketing institutions, and poor trade linkages with overseas dis- tribution systems. The potentially high risks and transaction costs associated with trade in highly perishable and quality-variable horticultural products have frequently been exacerbated by the underdeveloped transport and com- munications infrastructure and weak contract enforcement mechanisms found in many countries. Despite the considerable interest among African entrepre- neurs, governments and external development agencies in horticultural ex- ports as an engine of growth or source of trade diversification, there are few cases of successful and sustained development. Kenya is one of the very few SSA countries that have emerged as major participants in international horticultural markets. Compared with the other successful exporters (which include South Africa, Cote d'Ivoire and increas- ingly Zimbabwe), Kenya's horticultural trade has been far more diverse in terms of the range of products traded, the participants and the institutional arrangements which have emerged to govern the trade. From a narrowly based trade in the 1960s with a few dominant products and small numbers of participating farmers and firms, Kenya's horticultural export sub-sector now delivers some seventy-five products to more than a dozen markets. The pro- duction base ranges from very small farmers to large plantations. The trade passes through a dense network of traders and processors of varied· origins and diverse operational characteristics. This chapter examines the development of Kenya's horticultural export sub-sector. Kenya's case is of interest for several reasons. First, this sub- sector has always featured a dominant role for the private sector, which con- trasts to the general historical pattern of agricultural development in Kenya whereby the government has either monopolized trade through marketing board operations or strictly controlled the functioning of the market through THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORTICULTURAL ExPORTS 321 controls on prices, crop movements, market eri.try, etc.3 Kenya's horticultural export experience illuminates the complex relationship between the state and private sector development and the changes in that relationship as the indus- try matures. Second, the case clearly illustrates the diverse forms of organi- zation which may emerge in African agro-industries, and the influence of technical, economic and political factors on the emergence and sustainability of these different organizational forms. Finally, Kenya's long experience in horticultural exports offers both positive and negative lessons for entrepre- neurs, governments and donors interested in promoting horticultural exports in other SSA countries. This chapter is composed of five sections. The first section reviews the major techno-economic characteristics of horticultural products, their produc- tion and trade. On the basis of this review, we develop hypotheses regarding organizational patterns to be found within Kenya's private sector-dominated horticultural sub-sector. The second section summarizes the patterns of par- ticipation of SSA countries in global horticultural products trade. Section three traces the historical development of Kenyan horticultural exports and identi- fies the major roles played by the private and public sectors, the importance of foreign investment in the sub-sector, and the various factors contributing to growth and stagnation in the exports of particular commodities. Section four examines the ownership characteristics of Kenya's horticultural traders and processors and reviews their institutional linkages, both backward to- ward production and forward to overseas markets. The last section offers lessons that can be drawn from Kenya's experience.4 3 See Mosley (1983) and Bates (1990). Bates' critical stand against the alleged "miracle of the market" in the Kenyan context is greatly compromised by his failure to examine the only major agro- industry in that country where private firms and individuals have been largely permitted to develop their own institutional arrangements and respond to international market signals-the horticultural industry. 4 This chapter is based on field and desk research conducted by the author during the mid-1980s, supplemented by a brief field visit to Kenya in 1992. The initial field work included interviews with forty-four exporters and processors (who at the time accounted for 97% of trade), interviews with farmers, extension workers, transport agents and others in Kenya, as well as interviews with a sample of importers, wholesalers, and retailers of horticultural commodities within the United King- dom and the Netherlands. This chapter draws heavily from Jaffee (1990, 1991). 322 MARKETING AFRICA'S HIGH-VALUE FOODS MAJOR TECHNO-ECONOMIC CHARACTERISTICS OF HORTICULTURAL COMMODITIES Horticulture is defined as "the science and art of growing fruits, vegetables, flowers and ornamental plants". There are hundreds of different horticultural crops, rendering dubious any attempt to define generalized characteristics for these crops and their production. The considerable variability among horti- cultural crops with regard to certain techno-economic characteristics actually provides a basis for an interesting set of hypotheses regarding the emergence of different organization forms to govern horticultural production and trade. In this section, an effort is made to identify several 'general tendencies' among the broad range of horticultural crops and to compare Kenya's most impor- tant horticultural export crops according to the identified criteria. It applies elements of transaction cost economics to predict the organizational forms which one would expect to emerge in Kenya's horticultural sub-sector. In later sections, the actual institutional patterns within this sub-sector are dis- cussed and analyzed. In comparison with most other agricultural crops and commodities, the broad range of horticultural crops can be characterized by: 1) High rate of perishability: They experience rapid quality deteriora- tion, severely limiting their marketable life as a fresh commodity and the period of time during which they can be used as a raw material for processing. Even under optimal post-harvest conditions, the market- able life of many horticultural crops is only several weeks or even sev- eral days. 5 Horticultural crops continue to ripen after harvest and this process can be slowed or delayed only slightly. Rapid perishability gives rise to high risks of product and revenue losses and puts a premium on effective post-harvest treatment and handling, transaction speed, and close coordination within the marketing and logistical chain. While per- ishability helps to clear the market of surpluses, it also affects the bargaining or power relationships within the marketing chain as grow- ers and other sellers cannot afford to hold the product waiting for higher prices or other more favorable terms. Rapid perishability implies espe- cially high transactional risks in long distance and international trade, 5 The rate of perishability of horticultural crops stems not only from their own physiological prop- erties, but also from their stage of maturity at harvest, the handling and storage procedures used, and the prevailing post-harvest environmental conditions (e.g. temperature, humidity). THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORTICULTURAL EXPORTS 323 as suppliers are vulnerable to erroneous quality claims and protection- ist inspection or clearance delays at ports of entry or other destination points. 2) High variability of quality from unit to unit and from one supply pe- riod to another: This heterogeneity derives from many factors, includ- ing the numerous different biological varieties which exist for most horticultural crops, variations in production and post-harvest practices, and the impact of climatic or other environmental factors. The hetero- geneity of horticultural crops is compounded by the fact that the qual- ity of such produce is very complex, meaning that it is typically associ- ated with many individual attributes. For example, commercially important attributes of fruit and vegetables may include such physical characteristics as color, shape, size, weight and texture, and such chemi- cal characteristics as moisture, sugar or acidic content. With cut flow- ers, fragrance is also a valued quality attribute. Some of these attributes may be hidden or otherwise difficult to measU.re (for example, freshness), weakening the informational value of grades and complicating the language of trade. 6 Quality variability and complexity create uncertainty on the part of suppliers and buyers and can be a major source of conflict between them. While much qual- ity-related information must change hands, this normally needs to be supplemented by other measures, including direct observation of pre- shipped produce and/or the intervention of buyers in the production process through the provision of inputs, technical advice and so on. It is not surprising that reputations are a central element in the fresh pro- duce trade, serving as a close proxy for (expected) quality and overall reliability. Product heterogeneity implies that those pursuing a strat- egy of market segmentation can earn higher returns. 3) High seasonality of production and demand: Harvests and sales fre- quently need to be carefully scheduled in order to maintain high, stable input levels for processing facilities or to take advantage of 'market window' opportunities in the fresh produce trade. Seasonality in horti- cultural production is partly a result of seasonal differences in sun- 6 AB Produce Business columnist Max Brunk observes (September 1989), ''Everyone talks about quality, [but] when asked, few people can come up with a consistent answer. Two traders speaking on the telephone can be talking about quality with little or no understanding because each trader has a different point ofreference in mind. This is a cause of much conflict in marketing." 324 MARKETING AFRICA'S HIGH-VALUE FOODS light, temperature and other environmental conditions which exhibit greater variability in higher latitudes. By breeding new varieties or employing certain techniques, it is possible to alter or extend natural production seasons to some degree, although this may be costly (for example, use of heated greenhouses) or lead one to sacrifice on certain quality or other objectives. One consequence of natural seasonality is wide seasonal variations in prices, with potentially lucrative 'market windows' for suppliers with less or counter seasonal patterns of pr9duc- tion. AB some horticultural crops provide aesthetic in addition to con- sumptive value, demand frequently shows strong peaks near particular holidays. 7 Seasonality of production combined with rapid perishability raises problems for processors and traders in obtaining steady suppli- ers of particular commodities/raw materials, generally requiring them to diversify their product mix and their sources of supply. 4) Long gestation periods or extended production cycles of many horti- cultural crops (including most fruit crops and many flowers): These expose producers to long-term market entry (and price) risks, lower the elasticity of supply faced by processors and traders, and frequently lead both sides to seek supply and purchase assurances. 5) Labor intensiveness and need for careful crop husbandry. While some horticultural crops do feature production activities amenable to mecha- nization, many do not and labor requirements, especially for planting, weeding and harvesting, are as high or higher than for most other agricultural crops. Most horticultural crops also require particularly careful husbandry with close attention given to the timing of specific cultivation and post-harvest tasks. These characteristics generate cer- tain diseconomies of scale at the production level. Centralized, large- scale production systems frequently encounter high costs associated with labor recruitment, training, supervision and maintenance. 6) High level of technical and management skills required and need for specialized production and post-harvest inputs: The cultivation and post-harvest care of many horticultural crops require a level of 7 This is especially the case for cut flowers, for which demand in the United States and parts of Western Europe is especially high at the times of Christmas, Easter, Mother's Day, and Valentine's Day. Varietal and color preferences also differ considerably at different times of the year. THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORT/CULTURAL ExPORTS 325 technical and managerial skill which is greater than that for most other agricultural ventures due to the high environmental sensitivity of these crops, the strong influence of varietal choice and cultural (and post- harvest) practices over eventual market value, the potentially high ben- efits (as well as costs and risks) associated with particular techniques, and the characteristics of perishability, heterogeneity and labor inten- siveness noted above. The quality of some horticultural crops and the timing of their harvest and availability is enhanced by the use of spe- cialized chemicals, irrigation facilities and other purchased inputs and equipment This implies potentially high barriers to entry for less skilled and less well-endowed farmers in the absence of effective channels for credit, inputs and technical advice. 7) Economies of scale in some post-harvest and processing operations: There are many types of marketing infrastructure-including produce collection stations, pack houses, cold stores and transport vehicles- which require 'lumpy' investments. Their appropriate scale will vary by circumstances. 'Lumpiness' and economies of scale also apply to inter- national freight arrangements where a certain minimum level of trade is necessary to attract sufficient, competitive sea or air freight services and relatively high and predictable trade volumes are needed to enter into special freight chartering arrangements. While there are many technologies to process fruits and vegetables, most of the standard tech- nologies for producing canned, frozen and dehydrated for international sale are fairly standardized and exhibit economies of scale associated with the full utilization of major pieces of equipment. There is scope for joint marketing, common use investments and coordinated negotiation of and allocation of transport services. Potential economies of scale in processing operations as well as in certain quality control and trans- port operations can be realized only through close coordination of pro- duction and downstream activities, whether through private or collec- tive action. Descriptions of techno-economic characteristics of agricultural commodi- ties (and their production) are a useful starting point for surmising expected organizational forms within private agribusiness. To do so, one can draw upon and apply certain concepts from the expanding literature of transaction 326 MARKETING AFRICA'S HIGH-VALUE FOODS cost economics. 8 This literature focuses on the organization of transactions or, more broadly, exchange relationships. It demonstrates that there does not exist any single institutional structure which is superior to all others on effi- ciency grounds. Rather, different institutional arrangements are shown to have particular advantages and disadvantages. Their relative suitability de- pends on the actual operating conditions surrounding the trading relation- ship. In defining the operating conditions, emphasis is given in the literature to two main elements: 1) Asset-specificity-For any particular production and trading opera- tion, individuals may undertake either generalized or specialized in- vestments. Certain types of plant, equipment, materials and knowl- edge have potentially generalized use across a broad range of products or trades. Other assets are highly specialized for a particular product or trade outlet and have little or no alternative use or value outside of this product or trading area. Examples of asset specificity in horticul- ture include crops with extended gestation periods or production cycles (such as fruit trees), large-scale specialized processing and post-harvest facilities, and use of highly specialized production inputs and technical knowledge. 2) Uncertainty-In any particular trading context, the degree of uncer- tainty may vary-uncertainty regarding the availability of supplies or market outlets, the quality of the products on offer, the timing of sup- ply and demand, the trading terms being offered, possible political in- terventions, and so on. Such uncertainties tend to be more pronounced in agriculture than in industry because of the important influence of changing weather conditions and the wider geographical dispersion of primary producers and intermediate users. In the literature of transaction cost economics, it is postulated that spot market exchange, long-term contracts and vertical integration will each be 8 In contrast with traditional microeconomic analysis where institutions are exogenously deter- mined, transaction cost economics, together other branches of the so-called ''New Institutional Eco- nomics", seeks to explain the origins or current existence of institutions, ·to account for the particular forms which they have taken, and to examine the efficiency and distributional properties of alterna- tive institutional forms. The major elements and propositions of transaction cost economics are pro- vided in Williamson (1975, 1985), Langlois (1986), and Bromley (1989). John (1980) and Reve (1980) apply transaction cost concepts to the study of industrial marketing channels. See Bardhan (1989) of applications of elements of transaction cost economics to the study of agrarian institutions in develop- ing countries. THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORTICULTURAL EXPORTS 327 efficient modes of organization, defined in terms of economizing on a combi- nation of production and transaction costs under different degrees of asset- specificity and uncertainty. The table below summarizes the hypothesized relationships. The theory contends that under conditions of high asset-specificity, the most efficient mode of organization is the vertical integration of the two adja- cent stages of production or marketing. The firm investing in specialized as- sets, particularly those which are durable and involve large sunk costs, will be highly vulnerable to opportunistic bargaining on the part of suppliers or buyers since they will know that this investor has little or no alternative use for such assets and thus must come to terms unless it is in a monopolistic or monopsonistic position. Vertical integration is also viewed as an effective means of countering high levels of operational uncertainty, since a central management gains control over the different stages and direct supervision can be introduced. On the other hand, when asset specificity is moderate to low, other insti- tutional arrangements are viewed as more suitable than vertical integration since they tend to provide greater· flexibility of action, have lower 'start-up' costs, and do not incur the heavy overhead costs associated with integrating separate operations. Long-term contracts enable buyers and sellers to counter market uncertainties by offering mutual assurances and supplementing price signals with other informational devices related to the quantity, quality and timing of expected deliveries and purchasing requirements. Where both asset specificity and uncertainty are low, spot market arrangements may be most TABLE 7.1: Operating Conditions and Appropriate Institutional Arrangements Asset-Specificity High Medium Low High Vertical Vertical Long-term ~ Integration Integration Contract ·~ t: Q) Medium 1:l Vertical Long-term L-T. Contract or p Integration Contract Spot Market Low Vertical Long-term Spot Integration Contract Market 328 MARKETING AFRICA'S HIGH-VALUE FOODS efficient since they give the participants greatest flexibility of action and im- mediate signals about performance. It is generally easier (and less costly) to negotiate an adjustment in price levels than to agree upon and implement changes in trading rules or lines of command. In order to operationalize the insights of transaction cost economics for the purpose of developing and testing hypotheses regarding the organization of crop procurement systems, we have developed proxy indicators for both asset specificity and uncertainty for which quantitative or qualitative measures can be ·obtained in Kenyan horticulture. These proxy indicators and the rat- ing system employed for our empirical analysis are as follows: AsSET·SPECIFICITY: 1) the length of the crop production cycle or gestation period between the initial planting of the crop and the first commercial harvest: A produc- tion cycle of six months or less will be rated as short, one of six to twelve months will be rated medium, while one of over twelve months will be rated long. 2) the scope for scale economies in processing and post harvest opera- tions. In this case, qualitative assessments of low, medium, and high are based upon the mimmum efficient scale for the most restrictive processing activity, the needs for post-harvest treatments (for example, cleaning and waxing), and the advantages of cold storage. 3) the degree of specialization of material production inputs and technical knowledge: A high rating signifies that important inputs or cultivating techniques are used exclusively for the particular crop in Kenya. A medium rating indicates that important inputs (and techniques) have few alternative applications in Kenyan agriculture. A low rating indi- \ cates general applicability of material inputs and technical knowledge used. UNCERTAINTY: ·1) the degree or rate of commodity I raw material perishability: A high rating signifies that the crop maintains its quality before deterioration (under appropriate storage conditions) for less than one week. A me- dium rating is assigned for crops maintaining their quality for one to three weeks, while a low rating is assigned for crops maintaining their quality for more than three weeks after harvest. THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORTICULTURAL EXPORTS 329 2) the degree of specificity in the commodity I raw material quality required: A high rating indicates that quality must meet exacting (high) stan- dards. A medium rating indicates that quality standards are set within a specified range, while a low rating indicates that quality standards are not tightly defined and simply minimum permissible standards are set. 3) the degree of specificity in the timing of harvests and crop deliveries: A high rating indicates that harvests and deliveries must be timed to meet daily or two-day processing/marketing requirements. A medium rating indicates that harvests and deliveries must be timed for bi- weekly to weekly procurement requirements, while a low rating indi- cates that more flexible harvesting and delivery patterns are accept- able to the processor/exporter. Expected Organizational Forms in Kenyan Horticulture Table 7.2 rates Kenya's most important horticultural export crops/commodi- ties according to the above criteria and indicates the expected mode for orga- nizing the link between exporter/processors and farm producers of the crop. As there is no a priori way of weighting the different proxy variables, we take a crude average of the ratings for the three variables under each heading when making predictions about institutional arrangements. The table indicates that on the basis of technical crop characteristics, pro- duction characteristics and demand-based pressures for product quality and the timing of crop deliveries, one would expect the linkages between Kenyan producers and processors/traders to be governed predominantly by long-term (seasonal or other) contracts and vertical integration of production and down- stream activities. Spot market sales and purchases are expected to occur primarily as supplementary, market-clearing devices and not to be a primary mode of sale/procurement for any of these major crops, with the possible exceptions of french beans and chilies for the fresh market. SUB-SAHARAN AFRICA IN INTERNATIONAL HORTICULTURAL TRADE Many countries and entrepreneurs in sub-Saharan Africa have attempted to develop horticultural export trades. Some have sought to supply the expand- ing West European and Middle Eastern import demand for a range of high- " " c TABLE 7.2: Asset-Specificity, Uncertainty and Kenya's Major Horticultural Crops Asset-Specificity Uncertainty Scale Inputs/ Economies in Production Technical Post/Harvest' Quality Timing Expected Mode of Commodity Cycle Specificity Processing Perishability Specificity Specificity Coordination Pineapple (for processing) Long Med High Low High Med Vertical Integration Mango Long Low Med Med Med Low Long-Tenn Contract Passion Fruit (for processing) Med Med Med Low Low Low Long-Tenn Contract Strawberry Med High Low High High High Vertical Integration French Beans: Fresh market Short Low Low Med Med Med Spot Market/L.T. Contract For processing Short Low Med Med High Med Long-Tenn Contract Chilies Med Low Low Med Med Med Spot Market or Long-Tenn Contract Carrots (for processing) Med Low High Low Med Med Long-Tenn Contract Carnation Short Med Med Med High Med Long-Tenn Contract Chrysanthemum Short High Med High High High Vertical Integration Statice Med Med Med Med High Med Long-Tenn Contract THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORT/CULTURAL EXPORTS 331 value temperate, sub-tropical and tropical fruits, vegetables and cut flowers; others have developed trade within the African region for lower value and less perishable commodities. On an aggregate basis, the performance of sub- Saharan Africa's horticultural export trade can be considered satisfactory, especially in comparison with the stagnation and decline in Africa's exports of more traditional agricultural commodities over the past decade or two. Table 7.3 indicates that between 1976 and 1989, sub-Saharan Africa's horticultural exports increased two and one-halftimes from $636 million to more than $1.5 billion. To put this in perspective, the listed cluster of horticultural products would rank third among agricultural exports within sub-Saharan Africa, trail- ing only coffee and cocoa. In recent years, the region's horticultural exports have been two and a half times (in value) its exports of either cotton or tobacco, and nearly four times its exports of tea. These are all major tradi- tional exports and have been the focus of considerable attention in analyses dealing with SSA's agricultural trade. While the rate of growth in Africa's horticultural trade has been slower than that experienced by developing countries in either Latin America or Asia, the region has not experienced the same precipitous decline in world market share as it has for some of its major traditional exports (including coffee, cocoa, and cotton). As Table 7.4 indicates, SSA's world market share increased between 1973 and 1989 for fresh fruit/nuts and for cut flowers. It was maintained (at a very low level) for preserved vegetables and declined for fresh vegetables and processed fruit products. Over this period, a small ag- gregate decline in market share was recorded.9 While SSA's horticultural export performance has been satisfactory on an aggregate basis, country-level patterns have been highly divergent. Only three countries-South Africa, Cote d'Ivoire and Kenya-have developed sizable trades, enabling them to become major participants in international horticul- tural markets. These three countries accounted for 79% of the region's horti- cultural export value in 1988 and for nearly 82% of the value of EEC horticul- tural product imports from sub-Saharan Africa in 1992. Several other countries 9 See Islam (1990, Chapter 3) for a comparison between rates of change in African versus other regional exports of horticultural products. Islam's very negative assessment of Africa's performance stems partly from his inclusion of North African countries whose horticultural trade has declined (i.e. Morocco; Algeria), his non-inclusion of cut flowers in the basket of horticultural products, and his reliance upon FAO data which appear to underestimate the horticultural exports of many minor exporting countries. On the significant decline in SSA's world market shares for major traditional commodities, see Svedberg (1991). 332 MARKETING AFRICA 's HIGH-VALUE FOODS TABLE 7.3: Expansion of sub-Saharan Africa's Exports of Horticultural Products (US$ Millions; F.0.B. Value) Product Category 1976 1980 1989 Fresh Fruit +Nuts 307.5 659.1 1025.0 Processed Fruit 207.7 277.4 284.0 Fresh Vegetables 97.3 122.8 117.1 Preserved Vegetables 11.2 14.9 18.1 Cut Flowers 12.6 18.0 67.6 Totals 636.3 1092.2 1511.8 SOURCE: Data from United Nations, Yearbook of International Trade Statistics, various years. including Cameroon, Zimbabwe, Swaziland and Ghana have experienced an expansion of horticultural exports in recent years, although such trade re- mains relatively small. Most other African countries have either experienced a stagnation in their traditional fruit and nut export industries (for instance, banana exports in Somalia; cashew nut exports in Tanzania and Mozambique), or have encoun- tered significant production, transport and marketing problems which have severely limited the development of trade in non-traditional products (Mali, Burkina Faso, Senegal, Uganda, Zambia and Ethiopia). Surveys of European fresh produce importers consistently indicate the perception of high transac- tion costs in dealing with African suppliers and high degrees of uncertainty regarding product quality, continuity and reliability of supply, provision of information, and the exporters' overall familiarity with the requirements of the trade. 10 TABLE 7.4: Sub-Saharan Africa's World Market Share for Horticultural Product Exports Product Category 1973 1989 Fresh Fruit +Nuts 5.4% 6.2% Processed Fruit 12.6 4.2 Fresh Vegetables 3.1 0.9 Preserved Vegetables 0.6 0.5 Cut Flowers 1.0 3.0 Total (Weighted) 4.3 3.6 SOURCE: Based on data from United Nations, Yearbook of International Trade Statistics. 10 The survey results of Hormann and Will (1987), covering fifty importers in six West European countries, is indicative of broad perceptions of the hazards of doing business with African suppliers. THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORTICULTURAL EXPORTS 333 Only a few of the successful SSA horticultural exporting countries have developed a broad-based trade, which includes a range of different products and market outlets. This pattern appears to apply only to South Africa, Kenya and Zimbabwe. The horticultural trades of these countries consist of a range of fresh and processed fruits and vegetables and cut flowers. In contrast, the horticultural exports of Cote d'Ivoire, Cameroon and Swaziland are domi- nated by only one or two commodities/products. Their trade is either linked to preferential treatment in particular European markets (for example, Ivoirian and Cameroonian banana exports to France) or an outgrowth of the export- ing industry of a neighboring country (for example, the linkages between the industry of Swaziland and that of South Africa). Hence, for different categories of horticultural products, a limited few coun- tries account for the bulk of SSA's trade. As indicated in Table 7.5, only two countries account for 77% of the region's fresh fruit and vegetable exports to the EEC, nearly 87% of processed fruit and vegetable exports, and 81 % of cut flower exports to the EEC. THE HISTORICAL DEVELOPMENT OF KENYA'S HORTICULTURAL EXPORTS This section examines the development of Kenya's horticultural exports over the past half century. It traces changes in the organization of production and trade, the forms and intensity of government interventions, the international TABLE 7.5: Concentration of SSA Horticultural Exports to the EEC (1992) (Value and% Figures Pertain to EEC Imports; CIF $Millions) Fresh Fruit Processed Fruit and Vegetables and Vegetables Cut Flowers Totals %of %of %of %of African African African African Country Value Total Value Total Value Total Value Total South Africa 627.7 56.7 133.4 56.5 8.3· 8.0 769.4 53.2 Cote d'Ivoire 226.6 20.5 1.9 0.8 1.8 1.7 230.3 15.9 Kenya 55.6 5.0 71.2 30.1 56.1 53.9 182.9 12.6 Cameroon 88.2 8.0 1.6 0.7 - - 89.8 6.2 Zimbabwe 13.6 1.2 1.5 0.7 28.4 27.3 43.5 3.0 Swaziland 15.1 1.4 14.5 6.1 - - 29.6 2.0 Other 79.4 7.1 11.8 5.0 9.5 9.2 100.7 7.0 Total 1106.2 100.0 235.9 100.0 104.1 100.0 1446.2 100.0 SOURCE: Eurostat Data Base. 334 MARKETING AFRICA'S HIGH-VALUE FOODS market environment faced by Kenya's exporters, and the levels and composi- tion of trade. Early Origins and Development Through World War II The historical origins of Kenya's horticultural export trade date at least as far back as 1500 when traders in the coastal ports of Kilwa and Malindi provided fruit and vegetables to Portuguese ships sailing to and from India. In a more recent era, Indian and Arab traders based in coastal towns in- cluded fruits and vegetables among the various agricultural commodities sent to Zanzibar during the late 1800s. Small quantities of tropical fruits contin- ued to be exported from the coastal ports on into the early decades of the twentieth century.11 During British rule in Kenya, first as a protectorate (in 1895) and later as a colony (from 1920), relatively little attention and investment was directed to the production and trade of horticultural products. Prior to World War II, the horticultural sub-sector was characterized by an absence of modern mar- keting facilities, nation-wide marketing organizations, well-recognized grades and standards for fresh produce, and research and advisory services. Although small export trades in potatoes and passion fruit juice were' developed by private firms during the inter-war years, the bulk of local fruit and vegetable production was of very low quality and Kenya remained a net importer of horticultural products. Other than a few backyard jam- and juice-making units, the only commer- cial processing plants were four small factories constructed during the 1930s to extract juice from passion fruit for export to South Africa and Australia. Raw materials for the factories came from about one hundred European farm- ers based in Western Kenya who had planted passion fruit to supplement depressed incomes from cereals production. A Passion Fruit Board provided technical support to growers and regulated the competition between the factories. Although this trade was interrupted during World War II, the war-time restrictions on non-essential imports stimulated investment in im- port-substituting fruit and vegetable processing operations. These companies would subsequently benefit from high post-war tariffs on competing imported products. 11 This coastal trade is discussed by Martin (1973). THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORTICULTURAL EXPORTS 335 ·· aox: 7.i: LOSING AND GAINING GROUND~ ATALE OF '1wo · · CO..,NTRIE;$ . .. . .... The·. recent (:lxperiences in .Cote· d'l~oire :and Zimbat:>we illustr~te the wi.dely divergent trends taking. place in .horticultural exports within SSA. Duringthe. i97.0's; Cote d'Noire emerged as a major, and th.en dominant,. source of..· · · · fresti. pineapples• to .theJ/Vest European market. It benefitted from both indig- enous and fqreign·:· investme~ts:~nd. high·•.•levels of. integration into European distribt,Jtion channels.• · . particularly in France. The c9untry's close proximity to Europe and the· development ()f strollg infrast.ructure for pro1:Woe grading: and sea shipment enabled it to unde.rcut prices of more distant suppliers {eig: .· Kenya, South ·Africa and Southeast Aslan. counUies);····Pineapple exports in: . . creased frorn 17,440tons in 197Q to 157, 700 tons in :J.980. These exports. Pefiked in 1985 al.196,00Q tons,· atwhich time Cote d'Ivoire accoµnted for: .. :neatly :93% of ·EEC . imports'. Since· then, . rapidly·· rising costs, an· overvalued·· · cµrre.ncy; the restructuring ofthe industry .and the col laps~ qf the country's .. pineapple processing cQmpanies :reduced production and.cau.sed aff overall. declin~ in '.competitiveness ofthe. industl'}'. During the early 1990s, exp.arts .had fallen :to . below 120;000 .tons p¢r year: G1ndthe:country's •t:EC.market·· .: share had fallen by nearly one~third fo e3%~ . . .. · ·.... In sharp contrast; fimbabwe has experiericecl a rapid eiort facil itie~. Th.e. country's export • trade in fresh Vegetables is also 8){panding, and there haVe b~en several re.c~nt inv~stments in· factories ..to. proc~ss fruits anct v~geJablesfor: export·• . 336 MARKETING AFRICA'S HIGH-VALUE FOODS Probably the niost significant development during the war years was the implementation of a Dried Vegetable Project by the Agriculture Department. The project was dedicated to providing the Kenyan army and Allied troops in East Africa with supplies of dehydrated vegetables for use in soups and other foods. In 1942, dehydration factories were constructed at Kerogoya and Karatina in the center of the country. African smallholder farmers in sur- rounding areas were contracted to grow potatoes, carrots, cabbages and other vegetables to supply the factories. Although many of the farmers had prior experience growing vegetables for the local fresh market, the project offered participating farmers access to high-quality seeds and technical advice, a guaranteed outlet for their entire production at remunerative prices, and exemption from recruitment into mili- tary labor service. A special unit of agricultural officers was assigned to the project. Supplementing smallholder outgrower production was production from factory nucleus estates on land leased from local fariners and production on collectively farmed swamp areas, where small irrigation systems were con- structed. While initially encountering organizational and weather-related prob- lems, by 1945 the project was a major success, involving over 13,500 farmers, an agricultural and factory staff of more than 3500 people, and deliveries of more than 22,000 tons of produce to the factories. The project was by far the largest food processing operation involving African farmers in Kenya to that date. 12 When the war ended so did the military's need for dehydrated vegetables. Still, the closure and subsequent dismantlement of the factories created con- siderable controversy. The institutional and technical features of the project would influence later developments in smallholder agriculture in Kenya.13 For example, the project represented the first widespread use of irrigation in African farming in Kenya. The experience gained in irrigated cultivation and irrigation management would prove important in later irrigation schemes. The scheme also demonstrated that an expansion in smallholder cash crop production could be achieved over a relatively short period. In this regard, the project anticipated subsequent programs to expand smallholder coffee, tea, pineapple and pyrethrum production during the 1950s. In addition, the 12 In the food processing industry, only the Mumias sugar scheme developed in the 1970~ and the Njoro Canners french bean scheme developed in the 1980s have featured larger numbers of smallholder outgrowers. 13 Some of these points are noted by Moris (1973) in his analysis of the background to the Mwea irrigation scheme. THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORT/CULTURAL EXPORTS 337 project's use of a specialized extension service and input distribution system for a targeted set of crops and locations anticipated innovations built into later smallholder contract farming schemes in Kenya. The project would also serve as a model for post-independence program to increase the incomes of farmers settled under the Million Acres Settlement Scheme. Post-World War II to Kenyan Independence Following the war, the British Government maintained restrictions which it had imposed on food imports from the United States and other dollar coun- tries. Since the United States was the leading supplier of many canned fruit and vegetable products to the U.K., the import restrictions resulted in major shortages of these products.. The colonial government of Kenya viewed this trade regime as an opportunity to expand the colony's exports. The Agricul- ture Department soon set up a horticulture section, whose initial mandate focused on experimental research into fruit and vegetable varieties suitable for canning and advisory services to farmers on cultivation practices and post-harvest procedures. Fruit and vegetable canning was also an area in which the government sought to promote foreign investment through tax relief and other incentives. These research, advisory and investment promotion efforts were valuable contributions to the establishment of a pineapple canning industry. In 1948, two British companies in collaboration with European settler farmers set up canning factories to supply the U.K. with processed pineapple products. They got off to a very slow start and over the next fifteen years, these firms would experience a continuous stream of problems related to raw material procure- ment, transport, and intermediate input (e.g. cans, sugar) which weakened their competitiveness in international markets and generally resulted in fi- nancial losses. Although exports reached over 9000 tons by the early 1960s, the Kenyan product was largely a 'filler' when supplies from more preferred sources were inadequate. This was reflected in the far lower unit value of the Kenyan product compared withthat of its competitors.14 The pineapple canning industry survived during this period due in large measure to interventions by government, including price guarantees for (Eu- ropean) farm producers, technical and financial support for African smallholder outgrowers, taxes imposed on pineapples sold in the fresh market, the en- forcement of exclusive crop buying zones for firms, duty exemptions on im- 14 The early history of this industry is discussed by Kay (1965) and Swainson (1980). 338 MARKETING AFRICA 's HIGH-VALUE FOODS ported inputs, negotiations for improved international freight rates, and other measures. While the industry had initially been based on raw material sup- plies from a limited number of large-scale European settler farmers, by the early 1960s, some 75% of factory supplies came from smallholder farmers. A program undertaken by the Agriculture Department to provide technical ad- vice and inputs (on credit) to smallholders was a major factor in the availabil- ity of raw materials (albeit of uneven quality) to the factories. During the period between the war and· Kenyan independence, export trades were also developed or redeveloped for other processed fruit and veg- etable products. This was initially on a very small scale. In the case of pas- sion fruit juice, Kenya's trade remained at tiny levels through much of the 1950s due to a decline in European farmer plantings of the crop, frequent outbreaks of plant diseases in the main producing areas and increased com- petition in world markets. Beginning in the late 1950s the Agriculture De- partment launched a program to encourage smallholder production in parts of Nyanza District. The result was a recovery of production and trade to pre- war levels. On a similarly small scale, other Kenyan companies entered into contracts to supply European manufacturers with canned vegetable products (e.g. green beans and tomato paste). Government support was also important in these ventures. The varieties used had been selected from among those tested by the Agriculture Department, and the farmers in the government- backed Perkerra Settlement Scheme were an important source of raw mate- rials for the factories. An export trade in fresh fruit and vegetables was also initiated during this period. In 1955, a local (Asian-owned) farming and trading company began to send consignments of lettuce, tomatoes and other temperate zone vegetables by sea to Aden in order to supply ships calling at that colony and oil compa- nies operating in the Persian Gulf. Over the next two years, small trades in fresh vegetables were conducted with buyers in Djibouti and Somalia. In 1957, the Horticultural Cooperative Union (HCU) sent small quantities of high-quality produce by air to the U.K 15 While the East African-Near East 15 The Horticultural Cooperative Union was created in 1952 to provide marketing services to Euro- pean growers. During its first decade of operations, the HCU became the largest wholesaler and exporter of fresh produce in East Africa and the most important distributor of imported horticultural seeds and temperate fruits in Kenya. It entered the export market only after several years of local market experience and investment in marketing infrastructure. While some African primary coopera- tive societies were brought into the Union, the HCU's senior management was entirely European and 80% of its sales was of produce from European farmers. THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORTICULTURAL EXPORTS 339 trade represented an extension of the earlier regional trade in potatoes, the air-freighted trade was an entirely new development, made possible by the initiation of passenger air carrier links between East Africa and Western Europe. The fresh produce trade with the U.K. initially consisted of small volumes of strawberries, mangoes, pineapples, green beans and chilies, produced by European farmers and directed to a single wholesaler in London's Covent Garden Market. The air-freighted supplies were very expensive and sold only through 'up-market' restaurants, hotels and department stores during the winter months. Although this 'off-season' fresh produce market was as yet undeveloped, the U.K. demand for Kenyan produce far exceeded the latter's supply capability throughout the late 1950s and early 1960s. Kenya's annual exports were limited to a few hundred tons, primarily as a result of the very limited availability of air cargo facilities. The colonial government played a very minor role in the development of this air-freighted trade, regarding it as having prestige value for Kenya, yet little growth potential. While the gov- ernment did provide the initial financing for the establishment of the Horti- cultural Cooperative Uri.ion, its support for the export trade was limited to research on packaging materials and quality assessment trials on fruit shipped by sea to Europe. A very small commercial trade in cut flowers was also developed during this period, with a limited number of European settler farmers participating. One nursery company, started by a retired British Army officer, developed an export trade in carnations and sent consignments to neighboring countries as well as to the U.K Another producer/exporter specialized in more exotic flow- ers such as orchids and strelitzia. Cut flower production among African smallholders also began, although this was limited to a few locations and only involved sales to Nairobi hotels and florists. At Independence, Kenya's horticultural exports were r_elatively insignifi- cant, amounting to about 1.2 million Kenyan Pounds or 2.8% of Kenya's total exports in 1963. The country's horticultural exports were dominated by an unstable pineapple canning industry which required extensive government support and whose future commercial viability was in doubt. While the Euro- pean market for fresh horticultural produce appeared to provide Kenya with significant trading opportunities, the country lacked the level of production and infrastructure required for economical sea shipment of produce and its trade in highly perishable commodities was restricted by limited air freight facilities. The minor significance of horticultural exports at independence and 340 MARKETING AFRICA'S HIGH-VALUE FOODS the limited regard given to its future development were reflected in the World Bank's Economic Survey Mission Report (1963), which devoted only two para- graphs to the horticultural sub-sector and did not mention it in its extended series of conclusions and recommendations. 16 First Decade of Independence The first decade of Kenyan independence witnessed a moderate expansion and considerable diversification of Kenya's horticultural exports. Major struc- tural changes occurred in the sub-sector as a result of new foreign invest- ment, ·major financial and managerial problems encountered by a 'Kenyanized' Horticultural Union, and the entry of additional firms into the fresh produce trade. By the early 1970s the sub-sector would have the infrastructure and international market linkages to begin a rapid expansion in production and trade. During this period, the government directed very few resources to the further development of broad-based production support services, marketing infrastructure or regulatory mechanisms within the sub-sector. Instead, the it intervened primarily by facilitating and sometimes. by directly participat- ing in specific projects involving foreign investment. Horticultural production support services were badly neglected during this period. With respect to crop research, most of the activities previously under- taken at the National Horticultural Research Station and its various sub- stations were terminated after independence as a result of severe shortages of qualified staff and financing. A joint Kenyan Government/FAQ research project was initiated in 1971, but the work undertaken focused on varietal testing, largely duplicating work done during the 1950s. No field tests or post-harvest research were undertaken. While the Development Plan 1966-- 1970 indicated government intentions to establish a specialized horticultural extension service, no unit was ever formed and the number of extension agents with specialized horticultural training remained very limited. Throughout this period, no official research and advisory capability was developed for cut flower production. Similarly, very limited public resources were directed toward developing the infrastructure for horticultural marketing. Studying the domestic mar- keting system for hortic1,1.ltural crops in the late 1960s, Wilson (1971) found 16 The mission's main comment regarding horticulture was to warn that the exports of fresh pine- apples to Europe was 'uneconomic'. Actual Kenyan experience proved this to be incorrect. In fact, fresh pineapples served as the 'growth leader' in Kenya's fresh produce trade during the 1970s, enabling firms to penetrate European continental markets. THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORT/CULTURAL EXPORTS 341 that only two of Kenya's five largest urban areas had separate wholesale market facilities and that the transport facilities in major fruit and vegetable producing areas was very inadequate. By the early 1970s, the Nairobi airport still lacked a cold storage facility and had inadequate numbers of loading platforms and trolleys to handle the country's fresh produce exports. Despite previously successful trials in sea shipment of fresh produce, when the gov- ernment expanded and re-equipped the port at Mombasa, it did not invest in cold storage, handling or loading facilities for horticultural produce. During this period, little official attention was given to developing a proper regulatory framework for horticultural production and trade. In 1967, a Hor- ticultural Crops Development Authority was created and endowed with an extremely broad yet vague set of powers. The Authority was in fact given little authority. It would operate under the direction of the Ministry of Agri- culture, which provided it With few operating funds and little policy guidance. As a result, the Horticultural Development agency was for many years un- able to build up a competent staff and experienced frequent changes in senior management. Instead of performing regulatory functions, the agency was initially used by the ministry as a conduit for channeling foreign donor re- sources into specific projects. The agency's support to the export trade was limited to operating a small shed at the airport, convening meetings between exporters and air freight suppliers, and instituting common standards for produce packaging. Foreign Investment in Fruit and Vegetable Processing The recovery of West European economies during the 1950s and the growing incomes and demand for convenience food that resulted from it produced a major increase in international trade in a wide range of processed fruit and vegetable products. This lasted throughout the 1960s and until the mid-1970s recession. Accompanying this expanding demand and trade was a partial re- location of processing operations for temperate fruits and vegetables away from Northern Europe and toward sites in Southern and Eastern Europe and a relocation of tropical fruit industries away from high-income countries (e.g. United States and Australia) and toward sites in developing countries. This relocation trend, stimulated partly by labor cost-saving objectives, was a preva- lent strategy of European and U.S. multinational corporations.17 Kenya would participate in this industrial relocation, becoming the site for foreign invest- 17 This trend was discussed by Mackintosh (1977) and Groosman (1982). 342 MARKETING AFRICA'S HIGH-VALUE FOODS ments in several product lines. In each case, the Kenyan government would make substantial resource commitments either in the form of equity holdings in a joint venture operation or through the provision of land, technical staff and financial support to expand raw material supplies. One of the major foreign investments concerned the dehydration of veg- etables for export. The initial investment took place in 1964 and was a joint venture between a British agricultural trading company and the parastatal Development Finance Company of Kenya. For the British company, the es- tablishment of a small dehydration factory at Naivasha represented a pilot project to examine the technical, economic and market prospects for a larger venture. The firm had no prior experience with vegetable dehydration, but wished to retain a presence in Kenya following its ouster from the country's pyrethrum trade at independence. For the Kenyan Government, social and political objectives. were paramount. The project would be used as a vehicle for enhancing the agricultural skills and incomes of smallholder farmers who were being resettled under the so-called Million Acres Settlement Scheme. By targeting support services to these farmers and by controlling the result- ant crop (via a contracting arrangement), the government would be in a bet- ter position to recover its land purchase loans and strengthen its administra- tive control over the settlement area. Many of the project's components were modeled on the war-time dried vegetable project.. A targeted farmer advisory unit was set up, seeds and other inputs were provided on credit, and guaranteed prices were offered. The dehydration factory was also to receive raw material supplies from larger European farmers who had irrigation facilities and were otherwise producing for the fresh export market. While the smallholder outgrower component largely achieved company and government objectives, no other dimension of the venture had any mea- sure of success. Large farm supplies were less than expected because the fresh produce market offered more remunerative prices. Overall, raw materi- als supplies to the factory were highly seasonal, leading it to operate at less than half of its capacity. The factory itself was poorly equipped and managed and its final product was a low quality carrot powder, exported to the U.K. at a loss. In 1968, the Kenyan Government rescued the bankrupt company, paying off its past debts and buying out the former owners. Over the next few years, the company continued to struggle and required cash infusions from the Treasury. Despite very favorable international market conditions for de- THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORTICULTURAL EXPORTS 343 hydrated vegetables, the Kenyan company achieved only modest increases in production and trade. Because of the uneven quality and generally low value of its product mix, it accumulated financial losses annually. · In 1970, a Working Party on the Horticultural Industry criticized the veg- etable dehydration company for its failure to develop a commercially sound, long-term plan to develop its trade and to utilize experts in its product field. The Working Party argued that in the longer run, the profitability of the industry was dependent upon a linkage being formed with a major European or U.S. company which would provide finance, technical know-how, and a secure distribution outlet. After several aborted contacts, the government did enter in a joint venture partnership in 1972. The venture would involve the construction of a large modem factory at N aivasha and a shift toward higher value crops and more reliance upon large-scale production, perhaps on nucleus estates. While the government would have a majority equity share holding, a key element of the project would be a management and marketing contract with (and minority shareholding for) Germany's largest manufacturer/dis- tributor of dehydrated vegetables. The pineapple canning industry was also transformed during this period via a government-supported foreign investment. In 1964, the firm Kenya Can- ners approached a parastatal financing company for a loan to expand the capacity of its factory. This was agreed to on the condition that Kenya Can- ners develop a strategic link with an international canning company. A year later, a management and marketing contract was entered into with the Cali- fornia Packing Company (Del Monte), giving the latter a low cost and low risk opportunity to assess the potential for diversifying its international pro- duction base. In addition to being awarded sole marketing rights to Kenya Canners' product and an option to purchase a majority of the company's eq- uity, the American company received a commitment from the Kenyan gov- ernment to expand its smallholder production support program. It also leased 5000 acres of land to the company so that they could examine the potential for developing a nucleus estate. In the first few years under Calpak management, Kenya Canners contin- ued to face major problems with raw material supply, in both quantity and quality. An extended drought in 1965-1966 and the sale of some former Eu- ropean estates were the proximate causes. Moreover, in the main pineapple- growing areas, smallholders were finding coffee production more remunera- tive and were scrambling to plant coffee before the International Coffee 344 MARKETING AFRICA'S HIGH-VALUE FOODS Organization's ban on additional plantings was effectively applied. Faced with this situation, Kenya Canners proposed to the government that further ex- pansion of the industry be based on estate production. This proposal encoun- tered opposition from smallholder producers and their parliamentary advo- cates. Yet, the government agreed to it and purchased and then lea~ed an additional 18,000 acres ofland to Kenya Canners. The government also agreed to a Calpak demand giving Kenya Canners a ten-year monopoly on local pineapple processing and providing the firm with reductions on rail, wharf- age and handling charges. 18 Soon afterwards, Kenya Canners undertook a major investment program to develop its nucleus estate and construct a modern factory with an annual processing capacity of 170,000 tons, five times that of the existing factory. Still, through 1972 the company continued to operate at a loss with its vol- ume of trade remaining at or below that at Independence. The official pro- gram of smallholder pineapple support was phased out and the. government had to write off a large part of the loans provided to farmers. By 1974, outgrower supplies were totally phased out and the Kenya Canners operation was based entirely on estate production. A joint venture involving foreign participation was also developed in the passion fruit juice industry soon after Independence. This joint venture, formed in 1965, involved an Australian firm (Cottees Limited) which was experi- enced in the exotic fruit juice industry and the Kenyan parastatal Agricul- tural Development Corporation. The new company, Kenya Fruit Processors Limited, modernized an existing factory, developed an extended fruit collec- tion system and raised producer prices considerably. This stimulated increased plantings, but a plant disease wiped out much of the crop. When Cottees quickly abandoned the project, its place was taken by a Swiss company, Passi AG, one of Europe's leading distributors of exotic fruit juices. Passi would blend the Kenyan product with passion fruit juice sup- plies from other sources. The firm's entry into Kenya necessitated little finan- cial investment, yet it obtained sole control over marketing and received a commitment from the Kenyan government to provide financial and technical support to increase smallholder production of passion fruit. During the late 1960s and early 1970s, Kenya Fruit Processors' production and trade ex- 18 These developments were reviewed by Kaplinsky (1979) and Swainson (1980). Both writers ar- gued that the Kenyan Government was ill-advised to provide significant investment incentives to Del Monte and enable the firm to develop a nucleus estate for pineapple production. THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORT/CULTURAL EXPORTS 345 panded rapidly, leading Kenya to emerge as one of the leading international suppliers of passion fruit juice. Expecting a major growth in market demand, the company undertook the construction of a large modern factory in 1972 and launched an expanded production support scheme. Foreign Investment in Flower Production During the 1960s, commercial cut flower production in Kenya expanded. Most of the increased production which came primarily from small European-owned companies, was directed to the domestic market. Growing numbers of African smallholders, particularly in the Limuru and Tigoni areas west of Nairobi, began to experiment with flower production. Kenya's earlier flower trade was diversified to include chrysanthemums, alstromeria, molecella and other flow- ers. However, the limited local investment in post-harvest facilities and lack oflocal marketing expertise limited Kenya's cut flower exports to only 75 tons in 1969. This cottage industry was transformed by a multi-million dollar invest- ment in 1969 by a Danish firm, Dansk Chrysanthemum Kultur (DCK), then the world's largest producer of chrysanthemum cuttings and a major player in the European market for other types of flowers. While DCK had initially concentrated its operations in northern Europe, in the 1960s the firm began to develop production sites in the Mediterranean and other areas in order to extend its product range and supply season and to take advantage of lower labor costs. As in the case of the foreign investors in Kenya's fruit and vegetable pro- cessing industry, DCK was offered highly favorable terms to invest in Kenya. For example, the Kenyan Government provided the firm with a long-term, low-cost lease of a 15,000 acre estate (at Masongaleni), exclusive growing and trading rights for several types of flowers for eight years, extensive work permits for expatriate workers and a twenty-five year status quo passus with regard to changes in the laws about foreign investor taxation and profit repa- triation. The Danish Government reduced the start-up costs for DCK by pro- viding a cash grant equivalent to about one-third of the initial investment costs. 19 19 While the investment did offer significant employment and foreign exchange earning potential, the very favorable investment terms accorded to DCK may also have stemmed from the fact that the Kenyan Government's primary negotiator and signatory for the investment was a high-ranking of- ficial who had initially invited DCK to Kenya and who later became a Director in its Kenyan sub- sidiary. 346 MARKETING AFRICA'S HIGH-VALUE FOODS Within a few years, the Dansk Chrysanthemum Kultur Production Com- pany (Kenya) had some 100 hectares under chrysanthemums and other flow- ers/foliage at the Masongaleni estate and had purchased two additional farms. The Masongaleni operation featured relatively advanced production and post- harvest technologies (including an advanced irrigation system, shaded net- ting and a large grading and cold storage unit) and employed some 1600 people. One of the firm's newly acquired farms, a 3000 hectare estate on the shores of Lake Naivasha, was developed to grow carnations, a flower whose European market was then expanding at a very rapid pace. The third farm, at Updown near Nairobi, was used by the firm primarily as a nursery and experimental station. Although DCK (K) initially experienced major techni- cal production problems, by 1973 large increases in both the quantity and quality of production were recorded. Kenya's cut flower exports that year were nearly 1500 tons, with DCK accounting for some 90% of the total. Expansion and Diversification in the Fresh Fruit and Vegetable Trade During Kenya's first decade of independence, the market conditions within the European fresh produce market were highly favorable for new as well as established suppliers. West European import demand for 'out-of-season' and 'exotic' produce, though still small compared with imports of more traditional commodities (for example, citrus fruit, bananas), expanded rapidly as a result of increased incomes, greater consumer awareness of tropical products, grow- ing health awareness, and expanding European populations of 'guest work- ers' and immigrant communities from the Mediterranean, Asia and other origins. Kenya faced relatively few competitors in the product markets dur- ing the seasons in which its supplies were concentrated. Its competitive posi- tion was enhanced by duty free access to the U.K market, tariff reductions which accompanied the Agreement of Association between the East African Community and the EEC, and relatively low, Kenyan government-directed air freight rates for exported produce. Kenya's ability to service the growing European market was further en- hanced by investments in production infrastruCture and the expanded avail- ability of air-freight facilities. During the 1960s, several investments in irri- gation systems were undertaken on large farms around Lake N aivasha. The development of the nucleus estate at Kenya Canners also made available to the fresh market trade small quantities of very high quality pineapples, which came to serve as the 'product leader' of this trade. The increased availability THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORT/CULTURAL EXPORTS 347 of air-freight chartering facilities in the late 1960s and the subsequent intro- duction of wide-bodied carriers for passenger air traffic between East Africa and Western Europe facilitated the flow of greatly increased volumes oftrade. Kenya's fresh fruit and vegetable exports increased from only 1400 tons in 1968 to 8600 tons in 1973. During this period Kenya's fresh produce trade not only expanded in size, it also became increasingly diversified. More than two dozen different com- modities were exported on a seasonal or more extended basis. These included tropical and sub-tropical fruits (pineapples, mangoes, avocado and passion fruit), temperate zone vegetables (french beans, sweet peppers and zucchini), and a broad range of 'Asian vegetables' (chilies, okra, karela and dudhi). Kenya's trade was concentrated in the U.K market, although some com- modities found a profitable market on the European continent. Trade was oriented to specialty markets or the high-quality end of more mainstream markets. The rapidly expanding trade in 'Asian vegetables' was directed to the South Asian (immigrant) communities in London, whose population was \ growing due to immigration and whose dietary habits included a preference for traditional fruits and vegetables. Kenya's growing trade in fresh pine- apples consisted of large fruit with large crowns which were used as well for ornamental purposes. While three companies accounted for the bulk of this growing trade, a large number of companies began to involve themselves in the fresh produce trade on a seasonal or full-year basis. By 1973, there were thirty-six regis- tered exporters. Entry into this trade came from two major sources. One set of ne~ entrants were domestic wholesalers and/or retailers of fresh produce. These firms acquired experience in grading and packing, established trading relations with producers and built up their storage and transport facilities prior to entering the export trade. The second set of new entrants consisted of medium-scale farmers integrating trading into their operations in order to circumvent existing exporters. During this period, the Horticultural Coopera- tive Union's position in the trade was eclipsed due to internal financial and managerial problems and the Union's politicization. Despite the growth and product diversification of the fresh produce export trade, this trade still had a very limited base with only about 150 to 200 (medium-to-large-scale, non- African) farmers accounting for the bulk of supplies. 348 MARKETING AFRICA'S HIGH-VALUE FOODS Horticultural Export Development Since 1974 Over the past two decades, the horticultural export sub-sector has developed into an important component of the Kenyan economy, providing a major source of foreign exchange earnings as well as substantial employment and farm income opportunities. During this period when Kenya's exports of many 'tra- ditional' commodities either declined (sisal, meat, pyrethrum) or fluctuated greatly from year-to-year (coffee, tea), the aggregate volume and value of horticultural exports increased substantially and virtually continuously, mount:lng into double-digit rates most years. As Table 7.6 indicates, horticul- tural exports have increased nearly five-fold in volume and thirteen-fold in dollar value terms since 1973. By 1991, the value of horticultural exports-at $133 million-accounted for 12% of Kenya's total merchandise exports. While the sub-sector has achieved substantial and virtually continuous export growth, the trade patterns for particular (categories of) commodities have varied considerably. Such variations and the factors contributing to them are examined below. Fresh Fruit and Vegetables Table 7.7 traces the progression of Kenya's export trade in fresh fruit and vegetables. The most rapid growth in this trade occurred between 1974 and the early 1980s, with aggregate exports doubling from about 10,000 tons to more than 20,000 tons. Since then, trade has expanded at a slower pace, and except for one year in which there were unusual circumstances (1988), it has remained more or less stagnant in volume and value terms since 1987. While maintaining a diverse product mix (spanning some fifty different commodi- TABLE 7.6: The Growth of Kenya's Horticultural Exports", 1973 to 1991 Value (Millions Value Average Volume of Kenyan ($Millions Unit Value Year (OOOs Tons) Pounds) Equivalent) ($Per Ton) 1973 27.4 3.6 10.4 380 1976 47.3 13.2 31.8 672 1980 66.3 23.l 61.0 920 1984 90.0 51.6 65.4 727 1988 111.l 99.9 107.4 967 1991 125.1 187.3 133.4 1066 • Includes fresh and processed fruits and vegetables; cut flowers. Does not include nuts for dried leguminous vegetables. SOURCES: Kenya Customs and Excise Reports; Horticultural Crops Development Authority. THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORT/CULTURAL EXPORTS 349 ties), Kenya's fresh produce trade is dependent upon only a few items for the bulk of its turnover. For example, in 1990 a single commodity-frerich beans- accounted for more than half of the value of fresh fruit and vegetable exports, while four commodities-french beans, mango, avocado, and chilies-ac- counted for three-fourths of the total. Kenya's trade performance for many fresh fruits has lagged far below the country's potential. Kenya's fresh pineapple trade, having served as the lead product for the penetration of continental European markets in the early 1970s, succumbed to competition from less expensive, sea-freighted supplies from Cote d'Ivoire during the 1980s. A lack of public and private investments prevented Kenya from developing its own sea-freight capacity for pineapples and o_ther less perishable fruits until the late 1980s, by which time major markets in Western Europe and the Middle East were well supplied from competing sources. Kenya's fresh pineapple trade was also inhibited by the Kenyan Government's restrictions on Kenya Canners (the dominant pine- apple producer) in making sales to the fresh market. Only in the late 1980s was Kenya Canners given a license to export fresh pineapples. Although sales boomed in 1988, the subsequent restructuring of the parent Del Monte com- pany (and the sale of its fresh produce marketing division) reduced the incen- tive for the company to begin exporting and it refocused solely on canning. Kenya's trade performance for mangoes, avocadoes and passion fruit has also been disappointing. The country has failed to take advantage of the ma- jor expansion in European (import) demand for the former two during the 1980s, and Kenyan exporters and their downstream distributors have failed to more fully develop the European market for passion fruit. Lacking the major varieties preferred by European traders and consumers, much of Kenya's mango trade has been directed to Middle Eastern countries where the Kenyan product is used primarily for juicing rather than for direct consumption. Al- though Kenya's avocado exports have increased steadily, trade volumes re- main trivial compared with booming European import levels which have ex- ceeded 100,000 tons in recent years. The Kenyan product has served merely as a supplement or gap filler for shipments from Europe's more preferred suppliers (Israel and South Africa). Trade in other fruits, including 'off-sea- son' supplies of strawberries and melons, and limited supplies of apple, ba- nanas and various exotic fruits have performed somewhat better. Kenya's trade performance for fresh vegetables has been more favorable. Trade expansion through the mid-1980s was broad-based. Since that time, 350 MARKETING AFRICA'S HIGH-VALUE FOODS TABLE 7.7: Fresh Fruit and Vegetable Exports (Tons) Commodity(s) 1974 1976 1978 1980 1982 1984 1986 1988 1991 Fresh Fruits Pineapple 2445 1878 3717 1882 517 807 864 16,745 580 Avocado 281 407 717 654 963 1400 2151 3753 4193 Mango 674 718 878 1308 1735 2472 2941 3485 1745 Passion Fruit 80 101 138 264 324 425 646 733 619 Other Fruita 123 202 772 486 432 816 665 1728 833 Total Fruits 3603 3306 6222 4594 3971 5920 7267 26,444 7970 Fresh Vegetables French Beans 921 2324 3187 4964 6306 7524 9674 12,269 14,855 Asian Vegetables6 3741 4811 5953 7079 8051 9050 10,102 8582 9230 Sweet Pepper 854 573 497 111 50 81 16 37 14 Courgette' 758 881 787 780 891 814 231 54 20 Other Vegetables' 159 294 1096 950 1003 949 761 787 956 Total Vegetables 6433 8881 11,520 13,884 16,307 18,418 20,679 21,729 25,075 Total Fruit and Vegetables 10,036 12,187 17,742 18,478 20,278 24,338 27,946 48,173 33,035 • Includes strawberry, melon, papaya, lime, banana, and several other minor items. ' Includes about twenty-five different vegetables, the most important being chilies, okra, karela and eggplant. ' Includes bobby beans, tomatoes, carrots, and other temperate vegetables no specified elsewhere. SOURCE: HCDA Trade Statistics. virtually all growth has come from expanded shipments of french bean, supple- mented by small incremental gains for chilies and okra. Survival in the in- creasingly competitive markets for these items has been dependent upon ex- porter moves to upgrade their product through customized pre-packing for individual retail chains. Bulk consignments (in cartons) of these and many other vegetables are now only marginally profitable for Kenyan exporters. Kenya's trade in the broader range of 'Asian vegetables' has been stagnant since the mid-1980s as a result of increased competition from other countries and the adverse effects of increased air-freight costs on profitability. Kenya's once significant trade in sweet peppers and courgettes has fallen victim to competitive products from within Europe and the Mediterranean region. These divergent patterns in trade development can be linked to several sets of factors: • Rising air-freight costs and persistent air-freight bottlenecks have con- tributed to a long-term shift in Kenya's product mix away from rela- tively low-value commodities to those with higher unit values. Export- THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORTICULTURAL EXPORTS 351 ers facing limited freight allocations have sought to maximize sales turn- over by adjusting their product mix. • Belated and limited investments in horticultural research and advisory services, belated and limited arrangements for planting material propa- gation, and official barriers on imports of planting material have greatly constrained the long-term development of competitive exports for sev- eral commodities, including most fruit. Weakness in official horticul- tural support services proved to be less important in the case of french beans and many 'Asian vegetables' since the older varieties grown in Kenya continued to be accepted by European consumers. • Most of Kenya's fresh produce exporters are small-to-medium-scale trad- ing companies whose overall size and limited, sometimes intermittent, involvement in fresh produce exports have prevented them from invest- ing in proper post-harvest packing and storage facilities. Together with the failure to develop common use packing and cold storage facilities, these factors have constrained the development of trade in highly per- ishable or easily damaged commodities (such as strawberries and pa- paya). Another institutional factor is that much of Kenya's fresh pro- duce trade is undertaken by ethnic minority-owned firms whose political vulnerability has undoubtedly weakened their incentive to make long- term investments in marketing facilities or produce crops with extended gestation periods (for example, many fruits). With only a few excep- tions, most of the leading exporters have concentrated their trades on crops which mature rapidly, require less intensive post-harvest treat- ment and are relatively less costly to produce. • The differential trade results also reflect the intensity and sources of competition which Kenya's exporters have faced. For fresh fruits and vegetables, Kenya has not been successful when it has tried direct com- petition with suppliers from within Europe and the Mediterranean re- gion. The declining fate of Kenya's trade in courgettes, sweet peppers and other 'off-season' temperate vegetables is a testament to this. For mainstream commodities, Kenya's trade has largely been based on supplementing supplies of leading competitors (e.g. for avocado) or in supplying narrow 'market windows' (e.g. for strawberries in December/ January). In contrast, for high-quality french beans and for many 'Asian vegetables', Kenya's primary competitors have been African and Asian suppliers whose transport costs are comparable to those of Kenya and 352 MARKETING AFRICA'S HIGH-VALUE FOODS whose production and marketing systems are frequently less well devel- oped than those of Kenya. The expansion of Kenya's fresh produce trade has been accompanied by, and is in many ways dependent on, the entry of many new participants, especially at the production level. While in the early 1970s only a few hun- dred smallholders produced for the fresh export market (accounting for some 10-20% of supplies), by the mid-1980s some 15,000 smallholders were active in this trade, accounting for nearly half of its volume. More recent informa- tion is not available. Table 7.8 provides a breakdown of the estimated num- bers of smallholder producers of selected horticultural crops and their esti- mated share of export volumes. Expanded smallholder production and sales have been a major component of Kenya's increased exports of french beans, 'Asian vegetables' and several fruits. In the case of vegetables, this expanded production came in response to a rapid increase in foreign demand during the 1970s and was stimulated by competitive pressures within Kenya which led exporters to extend and diversify their sources of supply. Smallholder adoption of these crops was facilitated by informal technical advice provided by larger farmers in several locations and the production support services offered by several exporters. The fact that several crops have technical production characteristics which are quite similar to more traditional smallholder crops further eased their introduction. In the case of the avocado, the main stimulus for smallholder production came not from the encouragement or support of exporters, but from a seed- ling propagation program undertaken at the National Horticultural Research Station. During the late 1970s and early 1980s, NHRS employees distributed avocado seedlings free to farmers in nearby areas and provided them with informal technical advice. This informal assistance contradicted a stated gov- ernment policy that official support for avocado production go only to farmers planting a minimum of five hectares. Once the smallholder plantings ma- tured, the government, through the Horticultural Crops Development Au- thority, was obliged to collect and market the crop since there was inad- equate demand on the part of private exporters. Several other changes have taken place in the structure of production underpinning the fresh produce trade. Over the past decade, large-scale cor- porate farms owned by foreign companies or private local firms have become increasingly important for this trade. In most cases, the firms involved have THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORTICULTURAL EXPORTS 353 TABLE 7.8: Smallholder Participation in the Fresh Produce Export Industry (Estimates for the Mid-1980s) Estimated Estimated Share Typical Number of of Export Plantings Crop(s) Smallholders Volumes(%) Main Locations of the Crop French Beans• 4000--4500 40 Embu, Kikuyu, 1/8 to 112 acre Athi River Asian Vegetables• 2750-3000 50 Matuu, Mtito 1/2 acre Andei, Kibwezi Mangob 1500-2000 65 Malindi, Lower 1/2 to 1 acre Tana River area Avocado' 2000-2250 50 Kandara, Kangema 1/2 acre Passion Fruit'1 3000--4000 40 Embu, Kirinyaga 3/4 acre Sub-Total' 13,250-15, 750 50 •Information based on Hormann (1981), District Annual Reports, interviews with MOA officials, and field visits to major production areas. ' Participating in export marketing channels only. Based on Retief and Tucker (1985) and interviews with exporters. 'Central Province only and producing only Hass and Fuerte varieties. Based on Muranga District Avocado Census (1984), interviews with exporters and MOA officials, and field visits to major production areas. a Based on information provided by Kenya Fruit Processors and fresh produce exporters. ' Limited numbers of smallholders are also producing other fruits and vegetables for export. also had extensive interests in other lines of agriculture, with horticultural production serving as a diversification away from less or non-profitable cof- fee, cattle and/or sisal operations. Another important change relates to the ownership and management of medium-scale horticultural farms. Such farms, especially those at N aivasha and Thika, were the backbone of the fresh pro- duce trade until the late 1970s. Until that time, the vast majority of them were owned and managed by Kenyan Asians and Europeans. However, in the wake of the iate 1970s coffee price boom and more recently, a large number of African entrepreneurs and civil servants have leased or purchased farms near Nairobi or N aivasha in order to grow vegetables for export. The structure of export trade has also changed at the entry level. After the government relaxed its licensing policy, enabling all interested locally-owned firms to register as exporters, the number of licensed exporters increased from thirty-six in 1973 to eighty by 1978 to over one hundred by the mid- 1980s. Fresh fruit and vegetables were among the very few major agricul- tural commodities whose marketing had not been controlled by parastatal marketing· boards or well-established multinational corporations. The fresh 354 MARKETING AFRICA'S HIGH-VALUE FOODS produce trade thus offered what was perceived to be the easiest access by local firms into international markets and the foreign exchange which ac- crues through this trade. Processed Fruit and Vegetables Like the fresh fruit trade, Kenya's trade performance for processed fruit and vegetable products has beeri uneven. Following the major joint venture in- vestments in the early 1970s, production and export levels expanded rapidly and substantially, going from just over 10,000 tons in 1974 to more than 49,000 tons in 1977. Over the next five years, exports remained in the 45,000 to 50,000 ton/year range before increasing again to 55,000-60,000 tons/year during the mid-1980s. Over much of this period, a single firm-Kenya Canners-whose expan- sion of canned pineapple exports and subsequently pineapple juice exports has more or less dominated the broader trends within the industry, has ac- counted for 80-90% of Kenya's processed fruit and vegetable exports. The boom in processed fruit and vegetable product exports during the mid-to-late- 1970s can primarily be attributed to Kenya Canners' success in building up its integrated nucleus estate/factory operations and its effective linkage with Del Monte's European marketing operations. This success made Kenya the leading source of canned pineapple to Western Europe in some years. As the West European market for canned pineapple became increasingly saturated during the 1980s, and the fresh fruit juice market showed greater vibrancy, Kenya Canners undertook to expand its own production and exports of pine- apple juice. Kenya's broader experience in processed fruit and vegetable product ex- ports has been more varied and generally less successful. The joint venture operation at Naivasha for vegetable dehydration experienced a build-up of trade levels in the mid-to-late-1970s, but subsequently faltered due to raw material procurement problems, frequent breakdowns in factory operations and the high cost of intermediate inputs (especially fuel). Financial losses were persistent and the company eventually passed into receivership in 1982. Similarly, the performance of Kenya's longstanding passion fruit juice in- dustry has been poor. Raw material supply problems as well as management problems prevented the major joint venture company from fully utilizing its modern processing plant and taking advantage of the large expansion in world trade which took place between the mid-1970s and mid-1980s. For a broad THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORT/CULTURAL EXPORTS 355 range of other processed fruit and vegetable products, poor quality and high production and intermediate input costs have prevented Kenyan firms from being competitive in international markets.· The only non-pineapple product for which the industry has achieved some sustained success in export development has been canned french beans. Dur- ing the 1980s, the basis for this expansion was a joint venture between a local firm. and a major French manufacturing/distributing company. While the Eu- ropean market for canned green beans has generally been stagnant or in decline, the Kenyan venture has generated a superior quality product which has filled a particular niche in the market, especially in France. Kenya's labor cost advantages (over European producers) have compensated for the higher packaging and other intermediate input costs faced by the firm. In the past few years, there have been several additional investments in french bean canning and freezing operations by local and joint venture companies. Flower Production and Trade The production and trade of cut flowers has been one of the most dynamic components of Kenyan agriculture over the past two decades. This industry has experienced several new medium-to-large-scale investments, a secondary yet substantial growth in smallholder production, and a considerable diversi- fication in the mix of crops produced and marketed. Two major stimuli are responsible for this expansion and diversification of production and trade. One has been the large and steady expansion in West European use and trade of cut flowers since the early 1970s and the sustained interest on the part of European flower trading companies in broadening their product range and their supply sources. The second stimulus was the successful example provided by DCK(K) during the early to mid-1970s. When employees left the firm. to strike out on their own, the technical and marketing experience they had gained there served them in good stead. Rapidly expanding its production and exports during the early to mid- 1970s, the DCK venture was plunged into a financial crisis in 1976 when the company's major shareholder suddenly pulled out of the project. The DCK farms were passed into Kenyan ownership and the Updown Farm branched off as a separate company. Following two seasons of continuing financial and managerial problems, ownership of the estates changed hands again. Brooke Bond acqurred the Naivasha and Masengoleni Estates, while the Agricul- tural Development Corporation acquired the Updown Farm. ·' 356 MARKETING AFRICA 's HIGH-VALUE FOODS While Brooke Bond had no prior experience in flower production or trade, it was able to retain the services of several of DCK's expatriate production specialists. The company soon closed down the flower operation at Mason- galeni, absorbing many of its 2500 workers in the company's tea estates. At Naivasha, the company (under the name Sulmac) soon undertook to expand carnation production. By the early 1980s, Sulmac had approximately 120 hectares planted under fifty different varieties of carnations, making this the largest operation of its kind in the world. Since then, Sulmac has diversified its flower mix to include roses and gypsophilia and the company has remained Kenya's dominant flower company. When Updown Ltd. branched off in 1976, a cabinet-level government deci- sion was made to develop a smallholder outgrower scheme for flowers. Within two years, Updown had some 425 outgrowers and a list of 1200 others wait- ing to join the project. Although the outgrower project collapsed soon after Updown was purchased by the Agricultural Corporation who dismissed its expatriate staff, many of the farmers who had learned floricultural skills under the project continued to produce cut flowers for commercial purposes. Some have formed marketing cooperatives, while others have sold flowers to one or more of the private firms which were established by former expatriate staff of DCK or Sulmac. The experience of DCK and Sulmac and the efforts of particular individu- als have contributed to the development of additional foreign investments in Kenya's flower industry. The largest such investment has been that of the Oserian Development Corporation,. which is owned by a Dutch family now based in Kenya. The Oserian Estate, located adjacent to the Sulmac Farm at Naivasha, diversified into cut flower production (from vegetable production) during the early 1980s with the technical and marketing support of former DCK employees. Within a few years, the scale (although not the value) of Oserian's cut flower operations rivaled that of Sulmac, with production in- cluding many different types of flowers (such as delphinium, roses, dille, statice). The Oserian flower project has subsequently given rise to several other spin-off investments in flower production and trade. A former DCK employee was also instrumental in a mid-1980s investment by Yoder Broth- ers (U.S.) for the production and export of chrysanthemum cuttings. In the past five years, there have been several additional investments in flower production and trade. Most involve either joint ventures or efforts by fruit and vegetable export companies. to diversify into flower production and THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORT/CULTURAL EXPORTS 357 TABLE 7.9: Kenya's Exports of Cut Flowers 1975 1978 1980 1982 1984 1986 1988 1990 1991 Volume (Tons) 2746 3214 3499 4194 6941 8264 9809 14,423 16,405 Value($ Millions) 8.9 8.3 8.1 6.4 8.9 15.3 35.8 37.8 Na. Unit Value $/Ton 3241 2582 1743 1526 1282 1851 3650 2621 Na. SOURCE: Central Bank of Kenya. trade. Much of the new investment has been for rose production and other relatively high value crops, the result being dramatic recent growth in the value of Kenya's cut flower exports. Table 7.9 traces the expansion of Kenya's cut flower exports since the mid-1970s. PRIVATE SECTOR CHARACTERISTICS AND THE ORGANIZATION OF THE SUB-SECTOR Over the past two decades, participation in Kenya's horticultural export sub- sector has broadened considerably, with growing numbers of smallholder and other producers providing commodities and raw materials and with increased entry into trade, especially for fresh produce exports. Nevertheless, a close analysis of published and unpublished trade statistics indicates that only a few individual firms continue to account for the bulk of Kenya's horticul- tural exports and that firms owned by Kenyan Africans account for only a tiny proportion of this trade. These patterns (for 1991) are summarized in Table 7.10. The last column of the table indicates that while some seventy-six compa- nies did export horticultural product~ in 1991, three companies accounted for more than half of the trade. The siX largest exporters accounted for nearly three-fourths of the $133 million trade. 20 The table also shows that foreign-owned companies (including joint ven- ture companies in which multinational firms have controlling shares) pres- 20 Such a concentration of trade is not atypical for horticultural-exporting developing countries. For example, between three and six firms account for the majority of exports within the Colombian cut flower, Brazilian frozen concentrated orange juice, Chilean temperate fruit and Mexican fresh veg- etable export industries. 358 MARKETING AFRICA 's HIGH-VALUE FOODS TABLE 7.10: Competitive and Ownership Structure of Kenya's Horticultural Exports, 1991 Fresh Fruit Processed Fruit and Vegetables Cut Flowers and Vegetables Combined Number of firms 47 17 13 76 Proportion of trade: 3 Leading firms 57% 81% 92% 55% 6 Leading firms 75% 93% 98% 71% Export shares for Foreign vs. Local Firms: Foreign-owned 7% 69% 89% 58% Private-Locally owned 93% 31% 11% 42% of which: Kenyan Asian 64% 2% 9% 24% Kenyan European 17% 25% 0% 13% Kenyan African 11% 4% 1% 5% Parastatal 0% 0% 1% 0% Total 100% 100% 100% 100% SOURCES: Based on unpublished HCDA data and personal communications (for cut flowers and fresh fruit and veg- etables) and data provided in UNIDO (1992) (for processed fruit and vegetables). ently account for about 58% of Kenya's horticultural export earnings, with most of the remainder accounted for by firms owned by Kenyan Asians (24%) or Kenyan Europeans (13%). African-owned enterprises accounted for only 5% of the trade in 1991. This represented only a small improvement over their 3% share of trade during the mid-1980s. There has been widespread entry into the fresh produce trade over the past two decades. This trend has given rise to some degree of trade fragmen- tation and firms have competed with each other for the available air-freight space. Kenyan produce continues to be sold under a number of different brands. Overseas buyers have been able to play off one Kenyan exporter against another and negotiate export prices downward. The same dominant core of firms have accounted for the majority of the trade over the past twenty years. Since 1970, there has not been a single year in which the ten leading firms accounted for less than 80% of the fresh produce trade, and in most years this share exceeded 90%. Most of the other players in this trade have conducted small and highly seasonal trading activities and have generally not remained in business for more than a year or two. THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORTICULTURAL EXPORTS 359 . Over half of the leading firms are Asian-owned family companies or part- nerships. Most had considerable experience in fruit and vegetable production and trade in the domestic market before entering the export trade. Most have their own wholesale and/or retail establishments and farms; some also have complementary interests in freight forwarding and transport. Several of the leading exporters benefit from having relatives in Europe with whom they conduct a significant proportion of their trade. These operating features have significantly reduced the risks and transaction costs which such firms have faced when trading in highly perishable and heterogenous horticultural commodities. Much of the new entry into the trade in the past decade has been by African-owned firms. Despite their having preference in government-spon- sored trade fairs and training programs and periodic preferential access to the air-freight facilities of Kenya Airways, there has been a very high failure rate among African owned firms and only a few ofthe survivors have achieved sustained trade growth. Of twenty-two licensed exporters with 50% or more African share ownership in 1979, only four remained in business by 1985. Several factors have contributed to the high failure rates of new. African- owned trading companies: 1) Most of the African entrepreneurs entering this field had no prior experience in horticultural trade or in international marketing per se. They have instead relied upon 'learning by doing', a strategy which is very risky in the field of horticultural marketing where information, contacts, and trader reputations are key components of competitive ad- vantage and disadvantage. 2) Most of these entrepreneurs have maintained diverse business and other professional interests which have little or no relation to horti- cultural marketing, yet which have tended to consume a great deal of their attention and resources. As most such horticultural trading com- panies are .single proprietorships or family enterprises, the diversion of management attention to other, unrelated activities has undermined trading effectiveness. 3) Most of these African-owned firms have sought to operate only on a small-scale or part-time basis in the form of commercial clearing houses. They have sought to operate without investing in post-harvest treatment, packing and storage facilities. Their scale and seasonality of 360 MARKETING AFRICA'S HIGH-VALUE FOODS operations has given them virtually no bargaining power vis-a-vis me- dium-to-large-scale growers and air freight carriers, while their lack of marketing infrastructure has weakened their capacity to supply pro- duce of high and uniform quality. 4) Most of these firms have focused their trade on only one or rela- tively few commodities. This has exposed such firms to considerable crop procurement and market risks and has given their trade a high degree of seasonality. Their concentration has rendered them far less attractive to potential overseas buyers and distributors than is the case for Kenya's exporters carrying a broad product mix. Over many years, selected politicians and others have contended that the dominant position in the fresh produce trade by a few firms and the high failure rates of new entrants can be attributed to anti-competitive practices on the part of the former, including predatory pricing, bribery of airfreight officials, and various forms of 'exploitation' of farmers (such as non-payment, under-weighing, and false grading). Accompanying such arguments have been accusations of under-invoicing and failure to repatriate foreign exchange earn- ings. Such practices have at times occurred although their use has certainly not been restricted to the leading export companies. Over the years, many of the new entrants into this business had as their primary objective gaining access to and retaining foreign currency, in circumvention of existing foreign exchange control laws. Some of the small, unstable companies have been particularly delinquent in their relationships with small-scale farmers. The above allegations, together with the vested interests of particular gov- ernment officials, have led to various proposals and interventions geared to- ward restructuring the fresh produce trade in order to provide Kenyan Afri- cans with a greater share. Promotional measures have been taken, such as the provision of technical support to African-owned firms and the prominent featuring of their products at trade fairs. Other measures have given prefer- ential treatment to Kenyan Africans, such as the reservation of Kenya Air- ways freight space exclusively for African-owned firms. At times, Asian-owned firms have been harassed and 'encouraged' to take on African partners who in most cases turned out to be prominent government officials. While the market share of African-owned firms is higher today (11 %) than it was dur- ing the mid-1980s, this expansion can be attributed to the continued develop- ment and growth of a few individual firms, rather than to any positive impact THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORTICULTURAL EXPORTS 361 of racially preferential policies. Many of the people who have benefitted from preferential policies have been government officials active in the fresh pro- duce trade. 21 In the case of cut flowers, the government has long recognized the need for foreign investment and technical support but never developed an official sys- tem of floricultural research and extension other than on an ad hoc, project- level basis. Many more firms are active today than a decade ago. Yet Kenya's · cut flower exports continue to be dominated by Sulmac and Oserian, whose combined export share was 74% in 1991. Several African-owned companies have developed successful floricultural operations, and many have strong tech- nical and marketing linkages with an overseas company. However, most of these companies are small and African-owned firms accounted for only 4% of cut flower exports in 1991. Kenya Government efforts to promote the 'Kenyanization' of this industry have historically been limited to some direct floricultural investments by the parastatal Agricultural Development Corpo- ration and to pressures on foreign companies to limit the size of their expatri- ate staffs and provide training to Kenyans. Kenya's exports of processed fruit and vegetable products are even more concentrated and dominated by foreign-owned companies. Kenya Canners, a subsidiary of Del Monte, has accounted for 80% or more of such exports over much of the past two decades. The only other processing firms which have been oriented primarily to export markets have also featured the involve- ment of a foreign company, either in a joint venture with a Kenyan parastatal or private company or within the framework of a management and market- ing contract with a local firm. All operations have been linked into the global production and marketing operations of major multinational corporations. The Kenyan government has sought to increase 'Kenyan control' in this industry by entering into joint ventures with foreign partners. However, such local control has generally been ~usionary as the local partners have typi- cally been 'sleeping partners' (that is, equity holders without influence), while most fundamental production, marketing and financial decisions· have been made by the foreign partners and their overseas parent companies. 21 For example, during the mid-1980s, among those active in the fresh produce trade were high- ranking officials in the ministries· of Agriculture and of Lands and Settlement, in the Horticultur- al Crops Development Authority, in the Agricultural Finance Corporation, and several Members of Parliament. 362 MARKETING AFRICA'S HIGH-VALUE FOODS In contrast, the vast majority of locally-owned fruit and vegetable process- ing companies have focused on serving the Kenyan domestic market, a strat- egy made profitable due to strong tariff and non-tariff barriers on imports of competing products into the country. These firms did service the neighboring markets of Tanzania and Uganda under the tariff umbrella of the East Afri- can Community. However, after the dissolution of the Community in 1977 and the subsequent political and economic crises in these neighboring coun- tries, these exports declined substantially. Most such locally-owned firms have not been cost or quality competitive in international markets. The structure of ownership in Kenya's trade has both positive and nega- tive elements. On the positive side, without the technical skills and market- ing expertise and links of foreign companies, Kenya's horticultural trade, especially that for cut flowers and processed fruit and. vegetables, would have been considerably lower. A few foreign investments have had a demon- stration effect (especially for cut flowers) and encouraged local entrepre- neurs to make similar investments in production and post-harvest facilities. On the negative side, such an ownership pattern poses potential political problems. This sector is now one of the fastest growing components of the Kenyan economy and concerns about the distribution of benefits from this trade continue to be expressed. The government has experienced problems in monitoring and regulating the pricing and financial practices of firms which are trading with subsidiaries or associated firms overseas. From a long-term development perspective, this structure of trade has also served to provide relatively little experience for Kenyan Africans in the international market- ing of Kenya's horticultural crops, experience which could be applied to other agro-industrial areas. It is ironic that the government's most persistent at- tention to the 'Kenyanization' issue has been directed toward a component of the horticultural trade-the export of fresh fruit and vegetables-in which Kenyan-owned firms have long accounted for more than 90% of sales. Institutional Linkages Between Producers and Processors/Exporters In the discussion in the first section dealing with the techno-economic charac- teristics of horticultural crops, it was hypothesized that due to problems of risk, transaction costs and logistics, the coordination of horticultural produc- tion with downstream activities in Kenya would be governed more by con- tractual ties and vertical (ownership) integration than by spot market pur- chasing arrangements. Here the actual vertical linkages in the Kenyan THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORT/CULTURAL EXPORTS 363 sub-sector are summarized. These institutional arrangements are classified into five categories along the following continuum from open market arrange- ments to internal organization:22 ----> ------- > ------ > ------ >------ Spot Market Forward Interlinked Vertical Market Reciprocity Market Factor and Integration Purchase Agreement Contract Market Contract ----> ------- > ------ > ------ >------ On one end of the spectrum are spot market purchases. Here the processor or exporter purchases requirements from the market at a particular time at market prices. The firm may buy from farmers, local merchants or truckers. In this institutional framework, the processor or exporter will play no role in the production process itself. On the other end of the spectrum is the vertical integration of production and subsequent processing and marketing activi- ties. Here the marketing firm develops its own farm and flows of information and resources for production are internalized within the firm. In between these two extremes are various intermediate arrangements. For example, there may be market reciprocity agreements. These are infor- mal, yet highly personalized repeat trading ties in which some degree of loy- alty is built up between the exporter and a certain sub-set of growers. Still, produce is exchanged at the current market price and the exporter is not involved in the production process. Two types of intermediate contractual arrangements are indicated. Crop procurement can be made via forward mar- ket contracts which feature formal commitments to buy and sell specified quantities and qualities of produce at particular times. Prices may be agreed to in advance or at the time of actual exchange. A more intensive arrange- ment combines these forward purchase and sale commitments with buyer promises to provide specified production inputs and technical advice on credit and farmer agreement to follow the buyer's instructions regarding produc- tion. Such interlinked factor and mark.et contracts are frequently referred to as contract farming. 23 22 This section draws heavily on Jaffee (1994). 23 Along the continuum from spot market purchases toward vertical integration, the relationship between the exporter/processor and the crop producer changes in several fundamental ways. For example, i) decisions regarding production and sales become more centralized in the hands of the exporter or processor, ii) the relationship between producer and buyer becomes more formalized, iii) the terms of exchange become more extended into the future, and iv) the role of market prices in coordinating production and marketing is reduced and progressively supplemented by operating rules and direct supervision. 364 MARKETING AFRICA'S HIGH-VALUE FOODS TABLE 7.11: Institutional Arrangements Governing Private Processing and Trader Crop Procurement (Percentage Share of Raw Material/Commodity Volume, 1985/86) Alternative Institutional Arrangements Inter- Consistent linked With Expecta- Factor tions Based Forward and Vertical on Techno- Spot Market · Market Market Inte- Economic Crop/Industry Market Reciprocity Contract Contract gration Characteristics Fresh Produce and Cut Flowers Trade French Beans 20 20 30 20 10 Yes Asian Vegetables 30 20 20 20 10 Yes Avocado 40 40 0 0 20 No Mango 35 40 20 0 5 No Strawberry 10 0 90 0 0 No Carnations 0 5 0 0 95 No Chrysanthemum 0 0 0 0 100 Yes Other Cut Flowers ' 0 15 0 5 80 Na. FruiWegetable Processing Pineapple 0 0 0 5 95 Yes French Beans 5 0 0 95 0 Yes Passion Fruit 60 0 20 20 0 No Carrots• 0 0 0 95 5 No Orange 80 0 20 0 0 Na. •For 1980. SOURCE: Author's Field Survey. Table 7.11 depicts the actual institutional arrangements by which private processors and traders procured horticultural commodities and raw materi- als during the mid-1980s. The data, obtained from interviews with proces- sors, exporters and other sources, indicate the proportion of the exported or processed volume which was procured via the various institutional linkages with growers. The table indicates that there are great variations in the procurement arrangements used for different horticultural crops, although some of the patterns are not consistent with our expectations based solely on the crops' techno-economic characteristics. Other factors, including the capacities and experience of companies, the (non-)availability of official technical support services, geographical, logistical, political and historical factors have also in- fluenced the emergence and evolution of institutional arrangements linking growers, processors and exporters. THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORTICULTURAL EXPORTS 365 The table represents the institutional patterns at a single point in time. The earlier discussion of the historical development of the sub-sector indi- cated that while institutional patterns have gradually evolved in the case of the fresh produce trade, there occurred more sudden and significant shifts in institutional arrangements within the cut flower and fruit and vegetable pro- cessing industries. These typically resulted from foreign or joint venture in- vestments which brought with them not only new capital but also new forms of organization and new strategies for crop procurement. . Several interesting patterns can be observed from the table. One is that contrary to expectations, spot market linkages between growers and down- stream firms have been quite prominent, even for mango and avocado in the fresh trade and in some components of the processing industry. These ar- rangements were thought to entail too high a level of risk for the producer and too much uncertainty with regard to product quality, supply continuity and timeliness for. the processor/exporter. What is interesting is thB;t export- ers (and some processors) have tended to deal with smallholder farmers on an open market basis, while gaining most of their supplies from medium-to- large-scale producers governed by seasonal or longer term contractual and ownership ties. Instead of having direct ties with smallholder farmers, most exporters have relied upon selected local farmers and merchants who serve as interme- diaries. These appointed agents have fulfilled a number of functions includ- ing: (1) identifying and 'recruiting' farmers, (2) communicating short-term information to farmers regarding exporter quantity and timing requirements and regarding expected prices, (3) informing the exporters about local supply and competitive conditions, (4) distributing packaging materials to farmers, and (5) issuing payments to farmers. These agents typically have a shed or store to which farmers deliver their crop. The exporter's vans and trucks pass on these pick-up points on their collection rounds. For their services, the local agents are paid a commission based on the volume of sales. Based on instruc- tions from exporters (who must contend with fluctuating availability of air freight space and fluctuations in the deliveries of medium-to-large-scale farm- ers), and the extent of competition from other exporters and agents, the local agents may adjust upward or downward the quality standards sought and the prices offered to farmers. Reliance upon local agents has enabled exporters to economize on the infrastructural and transaction costs which would ordinarily be associated 366 MARKETING AFRICA'S HIGH-VALUE FOODS with procuring produce from large numbers of individual smallholders. By using local agents, exporters need not establish collection stations of their own within the smallholder areas. Rather than attempt to communicate and negotiate directly with dozens or hundreds of smallholders, exporters need only deal with a limited number of agents who have superior knowledge of local conditions and people. The residence of agents in the producing areas facilitates a continuous, if somewhat superficial monitoring of the crop. The buying procedures used also provide great flexibility and enable ex- porters to shift at least part of the transport and market risks which they face back to the growers. In the face of unexpected cutbacks in air freight allocations, downturns in market prices or surplus supplies from larger farm- ers, exporters and their buying agents can adjust their purchases from smallholders and farm-gate prices to them so as to minimize stocks and pre- vent losses on sales. The typically poor communication links between Nairobi and major smallholder producing areas creates an information asymmetry which is used to their own advantage by exporters and their agents.24 For many farmers and processors/exporters, spot market linkages have proven to be a sub-optimal arrangement. Yet, the adoption and maintenance of alternative institutional arrangements have been constrained by difficult logistics, weak bargaining power, competition and other factors. For small- holder farmers, spot market linkages have not provided a context for enhanc- ing or even maintaining farm-level productivity as little or no technical ad- vice is available. Such arrangements have also provided little basis for improving product quality because the quality standards for several crops are either poorly defined or highly variable. Sometimes, spot market prices are completely divorced from quality considerations. High levels of waste have also occurred as a result of poor coordination between growers and down- stream buyers. In the early-to-mid-1980s, some 25% of the smallholder crop of avocadoes, french beans, and 'Asian vegetables' remained unharvested or 24 There are of course limits to the opportunistic manipulation of information and adjustment of quality standards by exporters and their agents. Competitive pressures provide one restraint on behavior with different firms and agents acquiring reputations for reliability and fairness. Opportu- nistic behavior is also restrained by the need to obtain the loyalty of at least a core set of growers so that the agent can economize on his own search costs and consistently fulfill the exporter's product requirements. Pressures from local politicians and from HCDA officials also restrain opportunistic behavior. THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORTICULTURAL EXPORTS 367 unsold as a result of seasonal production gluts and short-term exporter re- ductions in crop purchases due to transport or other bottlenecks. Processing companies which have relied primarily on spot market pur- chases of raw materials have tended to operate at very low rates of capacity utilization. For example, the country's largest producer and exporter of pas- sion fruit juice had an average capacity utilization rate of 29% during the late 1970s and early 1980s. In the previous decade, that firm had developed a very successful contract farming scheme involving some 4500 smallholder farmers. However, with increased overseas demand for fresh passion fruit and growing demand from the local hotel and catering trade, the firm was unable to effectively compete for the crop, and most of its contracted produc- ers abandoned it except for the part of the year when there was a market glut. Unable to maintain control over product flows, the processing company cut back on its technical assistance and input supply program and became increasingly dependent upon highly seasonal spot market purchases from wholesalers/transporters. Another interesting pattern is the wide variability of institutional arrange- ments used to procure individual commodities. This is most evident for french beans (for the fresh trade and for processing) where the prevailing arrange- ments span the full institutional spectrum. A more detailed analysis of the procurement arrangements for this crop indicates close positive associations between the quality standards and total volume required by the buyer and the intensity of the buyer's involvement in the production process. Firms requiring mostly or solely beans of the highest quality grade (that is, 'extra- fine') have tended to involve themselves in the production process via sea- sonal contracts and the provision of inputs and technical assistance to farm- ers. For example, Njoro Canners, whose target market is the French market for premium quality beans, has developed a highly intensive and hierarchi- cally~organized contract farming scheme with smallholder farmers which features: 25 1) close controls on farmer participation in the project, strict limits on the planted area for any contracted farmer, and close controls on the tim- ing of bean plantings; 25 This case is examined in greater depth in Jaffee (1990, 1994). 368 MARKETING AFRICA'S HIGH-VALUE FOODS 2) the organization of farmers into distinct groups to which specific com- pany staff are assigned. The company's field staff has numbered more than 500 people; 3) the provision of a mandatory and standardized package of production inputs; 4) the close specification and supervision of each stage of the french bean production process; 5) the specification of fixed procurement prices and input costs; and 6) strict quality standards, below which beans are subject to rejection. Other companies with lower quality standards need far lower volumes of beans, have fewer technical and financial resources and have developed far less intensive linkages with french bean producers. Institutional Linkages Between Kenyan Exporters and Overseas Buyers/Distributors26 The organization of the Kenyan horticultural export sub-sector has also fea- tured diverse patterns in the institutional arrangements linking Kenyan ex- porters with overseas buyers, agents and other distributors. Table 7.12 sum- marizes the main patterns from the mid-1980s. Trading relations are classified as being governed by consignment transactions, long-term contracts and ver- tical integration. At least for the main products, Kenya's exports of processed fruit and vegetables were governed almost entirely by intra-firm trade or long-term contractual arrangements. While exports of canned french beans and dehy- drated vegetables were technically governed by annual or longer-term mar- keting contracts, all or nearly all trade was conducted through European firms which also played important management and technical roles in the Kenyan operation. Ninety percent of passion fruit juice sales were made through the majority shareholder of Kenya's leading producer. Kenya's ex- ports of pineapple products have been handled exclusively by a U.K.-based subsidiary of Del Monte. While actual product shipments have been made from Kenya to more than one country or final buyer, in each of the above cases practically all 26 A more detailed analysis of this such institutional patterns and their implications for the devel- opment and growth of trade is provided in Jaffee (1991). THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORT/CULTURAL EXPORTS 369 marketing activities have been orchestrated by the Europe-based affiliate and nearly all exports are invoiced to them. In each case, it has been the Europe-based affiliate and not the Kenyan firm which has been responsible for short- and long-term decisions regarding product lines, sales vol~es, delivery times, market destinations and prices. Hence, rather than speak of 'export marketing arrangements', it is more appropriate to regard Kenya's major processed fruit and vegetable exports as elements of the interna- tional product supply and trading operations of major multinational food corporations. Not all of Kenya's processed fruit and vegetable exports are carried out via exclusive trading links with ownership- or management-affiliated firms. Kenya's smaller processors, including those which are oriented primarily to servicing the domestic market, sell their products to several African and Middle Eastern buyers, usually on the basis of periodic purchase orders at agreed prices. Kenya's trade in cut flowers has similarly been dominated by intra-firm trade or long-term contractual arrangements with particular companies. For many years, Kenya's exports in carnations were made almost exclusively un- der a long-term contractual arrangement with a subsidiary of Europe's larg- est flower distributor. The marketing contract, which was renewed several times between the mid-1970s and mid-1980s, provided the importer/distribu- tor with full control over the marketing of Sulmac's carnations and potential veto power over additional elements of Sulmac's production and trade opera- tions.27 While Kenya's exports of chrysanthemum cuttings have been under- taken through a single vertically-integrated operation, the bulk of Kenya's sales of other cut flowers have occurred within the Dutch flower auction sys- tem. In the mid-1980s only one Kenyan exporter was consigning its flowers directly to particular auctions. Other companies had either subsidiaries or partners in Holland to whom flowers would be invoiced or sold before the . Dutch partner sold the flowers within the auctions or direct to other buyers. Kenya's fresh produce trade has featured greater variability of exporting arrangements. Various forms have been used for the sale of individual crops and individual firms have diverse sales strategies depending upon their prod- ucts, markets, contacts and experience. Still, even here, a majority of trade is governed either by long-term contracts or vertical integration. In this case, 27 Sulmac has since diversified its marketing channels. 370 MARKETING AFRICA'S HIGH-VALUE FOODS TABLE 7.12: Institutional Arrangements Governing Export Sales (Percentage Share of Trade) Institutional Arrangements Seasonal or Consignment other Long- Vertical Crop/Industry Transactions Term Contracts Integration Processed Fruit and Vegetables Canned Pineapple and Pineapple Juice 0 0 100 Canned French Beans 0 100 0 Dehydrated Vegetables 10 90 0 Passion Fruit Juice 0 10 90 Fresh Produce French Beans 40 40 20 Asian Vegetables 30 30 40 Avocado 85 0 15 Weighted Average for all FruiWegetables 43 31 26 Cut Flowers Carnations 0 100 0 Chrysanthemums 0 0 100 Other Cut Flowers 15 30 55 vertical integration involves trade between members of the same family. Many exporters, both Asian and African, have family members resident in the U.K. or elsewhere with whom trade is conducted. The development of long-term contractual arrangements with overseas buyers and the prominence of intra-family and intra~firm trade has played a critical role underpinning the sustained development of Kenya's horticultural exports. Such institutional arrangements have: 1) provided 'Kenyan' firms with immediate and continued access to mar- kets, even those which are highly concentrated at the import and whole- sale levels; 2) facilitated the deeper penetration of Kenyan products within the tar- geted import markets, enabling the Kenyan suppliers to continue trade or even gain market shares during periods when market demand was stagnant or declining; 3) provided some Kenyan firms with access to a continuous stream of technical and market information, enabling them to better match their product mix, quality, packaging, etc. to the tastes and requirements of consumers and end users; THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORTICULTURAL EXPORTS 371 4) reduced the transaction costs faced by Kenyan suppliers, by reducing company exposure to opportunistic quality claims, sales misreport- ing, and payment delays, and as the overseas partner organizes the downstream distribution and sale of product to potentially quite diverse buyers; 5) improved logistics and quality control, with the overseas partner main- taining some product stocks, providing quality control information, do- ing re-processing or re-packing were necessary, and helping to organize or negotiate international freight arrangements; and 6) provided Kenyan and foreign companies with considerable scope to manipulate export and production input prices and to by-pass currency, tax, and other provisions in order to best meet their own financial objectives. CONCLUSIONS While many African countries have a potential comparative advantage in the production of fresh and processed horticultural products, few such countries have been able to develop competitive and profitable horticultural trades to the markets of Europe and the Middle East. For the vast majority of African countries, nascent horticultural export subsectors have exhibited little or no growth over the past decade, despite favorable international market condi- tions for many commodities. Weak technical support systems, international transport and other infrastructural bottlenecks, inappropriate marketing in- stitutions, and poor international trade linkages have been among the com- mon constraints. With the possible exception of South Africa, Kenya has developed the most diversified and successful horticultural export trade within sub-Saharan Af- rica. However, even for Kenya, trade development patterns have been very mixed, both among commodities and individual firms. In the processing in- dustry, export success over an extended period has only been achieved by a limited number of firms which developed strong ties with major multina- tional corporations either through joint ownership or through long-term tech- nical, management and marketing contracts. Strong long-term technical, managerial and marketing linkages between Kenyan and foreign firms have also been essential to the development of Kenya's cut flower export industry. In both industries, critical 'infant-industry' support was provided by the colo- 372 MARKETING AFRICA'S HIGH-VALUE FOODS nial, and then the independent, government of Kenya during the 1950s and 1960s. In the fresh fruit and vegetable trade, profitable and sustained export growth has been achieved by less than a dozen of the several hundred firms which have entered this trade over the past quarter century. For the most part, the successful firms had long experience in domestic horticultural trade and production before entering the export market. They thus had an under- standing of the quality control, storage, packaging and more general risk elements associated with horticultural production and marketing before searching out export opportunities. Most of these firms initiated exports dur- ing the 1960s or 1970s, a period in which the European market for exotic and 'off-season' fruits and vegetables was just developing. The successful Kenyan firms built up long-term trading links, sometimes with family relations, which enabled them to more completely understand the requirements of the trade, reduce their risks and transaction costs, and achieve greater penetration of their products in foreign markets. In stark contrast, the large majority of entrepreneurs who have entered the horticultural trade have done so on a part-time, highly seasonal basis. Large numbers of businessmen, civil servants and politicians have entered this trade as a leap in the dark. Lacking both technical knowledge of horticul- ture and relevant trading experience, the tenure of such entrepreneurs in the horticultural trade has typically been quite short. For African countries and entrepreneurs seeking to replicate aspects of Kenya's horticultural development experience, several lessons can be drawn from that experience. First, while many African countries share with Kenya some ecological, locational and labor cost advantages in the production of certain horticultural crops, few such countries are well endowed with the additional technical, financial, infrastructural and managerial assets which are critical for developing a competitive horticultural trade. In the absence of foreign investment or a considerable initial investment by government and/or the local private sector in horticultural research, training and marketing in- frastructure, the development of horticultural exports will be a very slow process in Africa with a high incidence of failure among individual firms. Second, and relatedly, the development of African trade in many horticul- tural commodities will require the attraction of foreign investment or at least the development of long-term technical and marketing contracts between do- mestic and foreign firms. This may be especially important in commodity THE MANY FACES OF SUCCESS: THE DEVELOPMENT OF KENYAN HORTICULTURAL EXPORTS 373 lines requiring highly specialized knowledge and inputs, featuring rapid tech- nical and market changes, and featuring highly concentrated international markets. Governments need to improve the incentives for foreign investment and the legal and informational basis for the development of complex contrac- tual relations between local and foreign companies. Third, 'learning by doing' may no longer be a viable means of developing a horticultural export trade in light of the intensification of competition in Eu- ropean and other markets and the increased requirements for quality, supply continuity, and product variety in these markets. Prospective exporters would be advised to gain experience first in the domestic market, perhaps by devel- oping crop procurement and distribution arrangements to supply high-qual- ity produce to restaurants, hotels and 'up-market' retail outlets. In encourag- ing new entry into horticultural exports, attention should focus on private firms and cooperatives currently active in domestic fresh produce wholesal- ing, large individual growers, and firms or individuals which currently carry out exports of other agricultural or manufactured products and thus have market contacts and an understanding of the risks and mechanics of interna- tional trade. VIII. Fish Mammies and Tuna Conglomerates: Private Sector Fish Processing and Marketing in Ghana GeoffAmes and C. J. Bennett O ver 4.01 million tons of fish were landed in sub-Saharan Africa in 1989.1 Fish provides up to 25% of the protein in the diet in some countries,2 and fishing and fisheries related industries such as boat-building, net-making, fish processing and marketing provide employment and income for large num- bers of people. While certain East African and Indian Ocean countries have fishing industries, and while fish is an important item of domestic consump- tion in the inland countries adjoining the great lakes of Africa, it is West Africa that accounts for most of African fish production and employment in fisheries. This case study focuses on fish processing and marketing in Ghana, which presents several interesting features. Despite its short coastline (536 km), Ghana has a very prolific fishery, ranking about 35th in the world's fish catches. It also has one of the highest per capita fish consumption figures in Africa. Ghana has a wide range of fishing operations, with large-scale indus- trial and artisanal marine fisheries, and also inland fishing. Most fish pro- cessing and marketing is done by the private sector, with minimal inputs by official organizations. Women play a major role in both processing and mar- keting, and in fishing itself, as they own many of the fishing boats, even at the industrial level. Ghana's fishing industry has considerable regional im- portance, as much of the fish landed in other West African countries is caught 1 FAO (1990). 2 Essuman (1992). 375 376 MARKETING AFRICA'S HIGH-VALUE FOODS by Ghanaian fishermen. They operate for short and long periods from coun- tries as far away as Senegal and The Gambia. Finally, the Ghanaian case· vividly illustrates the impact of an ongoing Structural Adjustment Program on the fisheries sector. The case study has three sections. The first section examines major techno- economic characteristics of fish production, noting possible implications for entry into different forms of fish processing and marketing. The second sec- tion provides a short overview of the scope and features of fish production and markets in sub-Saharan Africa. The third and core section examines the Ghanaian experience, with particular reference to the negative effects of Struc- tural Adjustment policies on both parastatal and private industrial fisheries, and the relative resilience of the artisanal sector. The latter is distinguished by the persistence of traditional methods and traditional marketing relation- ships, and by the continued dominance of female fish buyers, the "fish mammies". GENERAL CHARACTERISTICS OF FISH, FISH PRODUCTION AND FISH MARKETING Fish as a commodity, together with fish production and marketing, have cer- tain techno-economic characteristics which influence the incentives and risks of undertaking activities in this sector, as well as the organizational patterns that are adopted. Techno-Economic Characteristics of Fish as a Commodity Perishability: Fish is very perishable: at ambient temperatures in the trop- ics, and without some form of preservation, serious spoilage of small fish occurs within 6 to 8 hours. Small fish become putrid within about 12 hours, and even large fish are of poor quality within a day. Packing fish in ice is a very effective means of preserving fish, which retains its natural appearance, flavor and texture, all qualities favored by consumers. However, in many areas ice is expensive. Also, many artisanal fishing boats are not equipped to take ice to sea, which limits the distance over which they can operate if the fish is to be landed in marketable condition. Fish can often be sold un-iced at landing sites, and trade in un-iced fish can be significant in some major coastal cities. However, most fish has to be iced or preserved in some other way. When fish is intended for export to PRIVATE SECTOR FISH PROCESSING AND MARKETING IN GHANA 377 · developed countries it may be frozen, but the scarcity of frozen storage, distri- bution and marketing facilities, and the cost of operating them, mean that there is little sale of frozen fish in most developing countries. (There are exceptions, such as Ghana, where much fish is frozen, either for smoking out- of-season, or for sale while thawing.) The other alternative is curing, that is reducing the moisture content so that the spoilage processes in the fish are greatly reduced. Mixed species catches: Whatever kind of net is used will retain all the fish inside it, other than those small enough to escape through the mesh. This is a minor problem with pelagic (near-surface) fish, which tend to move in single species shoals. Trawling, on the other hand, tends to retain all the species living near the bottom, resulting in a very mixed catch. This is a major factor in shrimp trawling, where the non-target species, referred to as by-catch, can make up as much as 95% of the catch. The need to use available space for the much more valuable shrimp catch generally leads to this by- catch being thrown back, representing an economic loss to the country as a whole. Characteristics of Fish Production Glut catches are the most important ~haracteristic of the production process, occuring in many marine fisheries. It is very unusual for catches to be evenly distributed throughout the year. Weather conditions such as high seas or monsoons may make fishing difficult or impossible for a substantial part of the year. Often the fish can only be caught at a particular season for climatic or oceanographic reasons. This applies particularly to pelagic fish such as sardines, anchovies and mackerel. In the period July-September each year, off the coast of Ghana, there is a huge upswelling of cold water rich in miner- als. This leads to a sudden proliferation of plankton and of the fish which feed on it. Large quantities of fish are caught over a short period, far in excess of the current consumer demand, and putting pressure on processing and mar- keting facilities. Such a seasonal production peak is in many ways analogous to the harvest of annual crops, but is given extra importance by the very high perishability of fish. Common access aggravates the problems of glut catches. A characteristic feature of marine fisheries (and some inland lake fisheries) is that the re- source is not owned by any individual or group. In principle the fish are available to anyone who can catch them. To some extent common access is 378 MARKETING AFRICA 's HIGH-VALUE FOODS limited by Exclusive Economic Zones where nations limit fishing by licens- ing, but this is not usually relevant to artisanal fishermen. In some cases access is limited by agreement between groups, for example where canoes fish in a coastal zone up to 10 km. out, small boats operate in the next zone, and bigger boats further offshore. However, even within these groups there will inevitably be competition between individuals. Every fisherman will want to land as much fish as possible while he can do so. It will rarely be possible for any fishermen to defer catching fish, to avoid a glut in the hope of a price rise; someone else is likely to catch those fish. Risks are inevitable in anything related to fisheries. First, fishermen have to find the fish. In an industrial fishery, this is less difficult, with modern technology. At the artisanal level, much depends on the acquired skills of the fishermen. The resource itself is uncertain. Pelagic fish especially are greatly affected by climatic and other changes. With a change of only a few degrees in the temperature of the water near the surface, pelagic fish may disappear, perhaps to deeper water or to a completely different area. Allied to this, risk within the marketing system due to the high degree of perishability, means that fishing has often remained a marginalized activity, financed by the in- formal sector and under-valued by government. Fishing communities are of- ten transient, lacking in secure land tenure and alternative income sources, and considered low status. The unpredictable supply of fish creates risk for fish processors. They have to decide when to buy, knowing that at any time large quantities of fish may be landed, causing a sudden drop in price. Alternatively, if little fish is landed, prices may rise equally quickly. Processors and traders will have to judge the price consumers will be prepared to pay for cured fish many months ahead. Formal information systems are very rare in fisheries. Hence, informal sources and forms of information are critical to the functioning of many fish marketing systems and in linking producers, processors mid traders. High initial fixed capital costs; these include large single payments for nets, boats and increasingly, motors. Due to the unpredictable nature of fish- ing income streams, repayment ofloans for such investments can be problem- atical. This explains why the financing of the sector, especially the artisanal fishery, is commonly characterized by informal financial arrangements be- tween boat operators/owners and shore-based money lenders and middlemen/ women who will lend where other institutions are reluctant to do so (formal credit systems among fishing communities have a poor record). This financ- PRIVATE SECTOR FISH PROCESSING AND MARKETING JN GHANA 379 ing role is likely in many areas to be taken by those involved in the process- ing and marketing of fish. Considerable economies of scale are available in the fisheries sector. Larger vessels can have greater catch per unit effort and longer times at the fishing grounds, reducing the unit cost of travelling to and from these grounds. Fish Processing Value added products from fish are not very important in Africa. In South East Asia, for example, there is a big market for products su~h as fish sauces, fish soups, fish crackers, etc. In Africa the consumer requirement is for fish in its wet state or cured. Cured fish cannot properly be described as a value- added product. Its price per kg. or per ton may be higher than that of wet fish, but that overlooks the fact that lkg. of cured fish is obtained by remov- ing water from up to 3 or 4kg. of wet fish. In terms of value per unit of dry matter or food value, cured fish is a lower value product than wet fish. Al- most always, consumers want wet fish and many will pay more for it when they can get it. Therefore, fish is cured not to increase its value but to pre- serve it during marketing and distribution. Cured :fish is usually produced by sun-drying or smoking, often after pre- liminary treatment with salt. Salting reduces the moisture content of the fish, retarding the processes involved in spoilage. More moisture can then be removed by sun-drying or smoking, depending mainly on the size of the fish, and whether it is oily. Small fish such as triggerfish (Balistes) can be sun- dried readily. Small oily fish such as anchovies can be sun-dried directly, but usually they are salted first to remove some of the water and to reduce the rancidity of the final product. Large fish can be difficult to sun-dry, as the surface hardens as it dries, preventing moisture from within the fish evapo- rating; the outside may be dry while the inside continues to spoil. Large fish can instead be hot-smoked, heated in a kiln or oven so that it is cooked and dried at the same time. Salting before smoking can accelerate the drying process. Depending on the extent to which it is dried, cured fish can be kept · for periods of many months. Glut processing is a convenient term to describe the problem .wh,ich arises from the combination of glut catches with the need to process most of the catch. Demand for labor and utilization of processing capacity are both highly seasonal. In artisanal processing, family labor including women and children is utilized to spread, turn, collect, and pack the product. Fish smokers are 380 MARKETING AFRICA'S HIGH-VALUE FOODS limited by the capacity of their kilns, and have to consider that the capacity needed at glut seasons will not be used for most of the year. Options for speeding up processing, or making it less-labor intensive, are limited and tend to require more fuel, or larger-scale operations. Attempts to create comple- mentary forms of non-fisheries employment in fishing communities have not generally been successful, partly because such communities often lack secure access to land. In the industrial sector, besides a similar tendency to seasonal employ- ment patterns, there are serious problems of low overall utilization of pro- cessing capacity, and of the difficulty in predicting returns to capital, espe- cially returns to scale. Fish Marketing Low entry costs are typical of artisanal level fish marketing and processing. However, this statement should be balanced by noting that income in this sector is also low, inflating potential capital investment as a proportion of overall financial resources. Processing (salting, drying, smoking or curing) tends to require low levels of technology and inputs. Fixed capital costs in the industrial sector are high, and demand high utilization of available catch. Additionally, canneries are commonly dependent upon imported packaging materials and must compete with imported processed products. Vertical ties are the norm among artisanal fish marketing systems and are common in industrial fisheries. This mitigates risk and develops patron- client linkages between buyer and seller. In the face of variable and some- times unpredictable supplies, these ties provide some degree of assurance to • both sellers (assured markets, ad hoc loans, etc.) and buyers (first rights to available supplies and a means to guard against short-term price fluctua- tions). Fishermen need to know where to land their catch, which will depend on market requirements and processing facilities in each area. Fishermen usually manage to avoid landing fish for which there is no outlet, resulting in a complete loss. At the industrial level, integration is common. Fleet owners often devel- op processing and marketing facilities to maximize returns on investment in ships. At the artisanal level, integration within family and cultural groups is common with family labor often divided between fishing, process- ing and marketing. Cooperative marketing is not typical of fisheries and many attempts to implement such schemes have failed. In the authors' expe- PRIVATE SECTOR FISH PROCESSING AND MARKETING IN GHANA 381 rience, mutual trust exists among fishing families and between buyers and sellers offish products with strong patron-client links; less so between fishing households. ROLE AND IMPORTANCE OF FISH IN SUB-SAHARAN AFRICA Fish and meat are the two main sources of animal protein in Africa, and are nutritionally superior to vegetable proteins. Their relative importance varies throughout SSA. Key factors are the length of coastline and/or the amount of inland water which is available and the quantity of fish available in those waters. While many African countries are major cattle producers and meat is much more important in the diet than fish, fish may be a much cheaper source of animal protein than meat. For example in Zambia and Zimbabwe fish pro- duction is much smaller than that of meat, but the fish which is produced sells very readily as demand exceeds local supply at prevailing prices. Fish, especially cured fish, is much cheaper than meat in many African countries. In many countries fish provides much of the animal protein of low income groups. In a substantial number of SSA countries, fish is a very important food for most or all levels of society. This applies particularly to the coastal states with rich off-shore fisheries such as Cote d'Ivoire, The Gambia, Ghana and Senegal. It also applies to some inland countries such as Uganda, where fish from lakes provides much of the nation's protein food. Fish consumption data for selected countries are given in Table 8.1. Fish catches and landings for SSA countries are shown in Figure 8.1. There is a huge variation in production levels, ranging from only 40 tons in Lesotho to more than 500,000 tons in South Africa and Nigeria. The available data indicate no clear pattern of change over recent years. As noted above, catches fluctuate substantially. This is particularly true for the pelagic fish species which represent the greater part of all the principal countries' catches. Employment in fish production is substantial, but difficult to define and measure. Some countries produce data for the number of fishermen. For ex- ample, in 1990 Ghana claimed to have 91,400 fishermen whilst Nigeria had 164,870.3 UNIDO (1991) estimated that Ghana employed 83,000 full-time 3 From Infopeche (1990). 382 MARKETING AFRICA'S HIGH-VALUE FOODS TABLE 8.1: Fish Consumption in Selected African Countries and Regions Per capita Fish products consumption Fresh Cured as share of Country (kg/year) (%) (%) protein supply % Chad 8.5 10 90 9.9 Cote d'Ivoire 17.7 35 65 8.7 Gambia 16.4 30 70 Na. Ghana 26.4 20 80 18.7 Mali 7.1 10 90 3.5 Senegal 20.7 70 30 9.8 Sudan 2.0 70 30 0.5 Uganda 13.0 45 55 7.2 West Africa 10.0 Na. Na. Na. Central Africa 13.2 Na. Na. Na. East Africa 6.0 Na. Na. Na. SOURCES: Essuman (1992), (protein figures (1990) from World Bank 1993). and 58,000 part-time fishermen in the capture fisheries sector, and 30,000 in post-harvest support industries. Seasonality and the wide geographic spread of employment in the artisanal sector makes estimation difficult and may lead to underestimation in the post-harvest sector. Employment in pre-har- vest support industries such as boat-building and net-making plus fish pro- cessing and marketing mean that the total number of people employed as a result of fish production is many times the number of actual fishermen. It is difficult to generalize about African external trade in fisheries. Some countries import and export similar quantities, but these can be high-value frozen shrimp for export and low-value dried fish as imports. Increasingly, there are exports of high quality iced fish, such as the export of Nile perch to Europe from Kenya, Uganda, and Tanzania. Table 8.2 gives some idea of the complexity of import and export patterns. Most of the trade within sub-Saharan Africa is in wet or cured fish to neighboring countries. Examples are Chad, which exports nearly all its fish to Nigeria, and Malawi, which exports dried fish to Zambia. Uganda and Tanzania export substantial quantities of cured fish to Rwanda and Burundi, although a growing proportion of their trade is of fresh and frozen fish to European countries. Intra-regional trade is widespread but unmeasured. Much is unofficial and related to traditional migration and ethnic linkages. Migration of peoples along the African coastline over a period of many centuries has produced vast socio-economic spaces which were only later constrained by territorial limits PRIVATE SECTOR FISH PROCESSING AND MARKETING IN GHANA 383 FIGURE 8.1: Fish Production in sub-Saharan Africa, 1989 (OOO's Tons) SOURCE: FAO (1990). 384 MARKETING AFRICA'S HIGH-VALUE FOODS TABLE 8.2: International Trade in Fisheries Products from Selected African Countries, 1991 Country Imports Exports Angola v 49,502 5896 Q 45,289 1224 Cote d'Ivoire v 114,947 173,067 Q 186,444 109,405 Ghana v 31,320 15,916 Q 32,630 16,490 Nigeria v 191,460 15,590 Q 492,580 5815 Senegal v 38,635 233,400 Q 50,316 123,143 V =value in US$ '000, Q = quantity in tons. SOURCE: FAO (1991). superimposed by colonial powers.4 People migrate to catch, process and mar- ket fish. 5 Short term migration to trade in fish seems widespread.6 This trade is commonly conducted by women and is primarily export oriented. Recent stud- ies7 suggest a number of constraints to such trade. These include: • high costs of intra-regional transport; • high import duties; • poor infrastructure (i.e. lack of cold-chains); • non-tariff barriers (i.e. complicated import-export procedures); • devaluation of currencies; this has, to some extent worked in Ghana's favor since the Cedi has been devalued considerably against the CFA. The Ghanaian fisheries sector has considerable regional advantages in terms of skills. Currency ·devaluation and reduced trade restrictions have 4 See Jean Pierre Chauveau's paper in Haakonsen (ed.) 1991 for an attempt at periodization of the regional fishery migration. 5 Migration is defined here as the movement of peoples to seek profit or employment outside their traditional area of abode. 6 See for example the map of the international cured fish trade in the West African sub-region in lnfopeche, 1990, Appendix 1:88. 7 INFOPECHE (1992, 1993). PRIVATE SECTOR FISH PROCESSING AND MARKETING IN GHANA 385 allowed them to exploit these advantages more fully. Further benefit would accrue from improved intra-regional communications and reduced bureau- . cracy at borders. The Policy Environment Policy within the fisheries sector has reflected both the worsening economic predicaments of some African countries and the changing agendas of the international donors. In the post-independence period there was a general development focus on industrialization. As the private sector was weak in this area, governments played a significant role in enterprise development through the formation of state-run fishing and fish processing enterprises.8 Four successive approaches to intervention in the fisheries sector can be seen. In the 1950s the focus was on large-scale, vertically integrated projects aiming to increase food supply and employment. These projects were mainly in the industrial fisheries sub-sector and designed to exploit unused resources. Lack of expertise in the private sector meant that much of this activity was channelled through parastatals. Added impetus was given to these national programs by the extension of the 200 mile national exclusive fishing zone in the 1970s. This policy continued into the 1960s and 1970s with an additional empha- sis being on resource assessment and institution building. By the 1970s it had been realised that artisanal fisheries played a much greater role in food supply and employment than previously thought. By the 1980s evidence that fishery resources were nearing full exploita- tion switched the emphasis from the pre-harvest aspects of fisheries to the post-harvest aspects in an endeavor to increase the value ofexisting resources and reduce losses. Most recently, support has reflected a more general switch to holistic approaches to rural development with national and regional inte- grated community development projects. THE FISHERIES SUB-SECTOR IN GHANA General Characteristics Ghana is one of Africa's major fish producers. In view of its small coastline it must be the world's largest fish producer ill terms of catch per unit area of its 8 This section is based mainly on Campbell et al. (1993, 30-35). 386 MARKETING AFRICA'S HIGH-VALUE FOODS Economic Zone. 9 Fish is the major source of animal protein in Ghana, and a relatively sophisticated fish processing and marketing system operates al- most entirely within the private sector, despite early governmental policies of promoting public sector enterprises. Artisanal fishing in Ghana has been a traditional activity since before recorded history. 10 The traditional fishing vessel is a large dugout canoe as wood is plentiful in the coastal areas. The most significant change in the artisanal sector for the purposes of the present discussion occurred in the 1970s, when outboard motors became available. This enabled the canoes to travel further, and take advantage of better fishing grounds. It also led to a considerable increase in the number of canoes. Meanwhile, industrial and semi-industrial fishing industries were developing, both in the public and in the private sector. For the purposes of describing the operational framework of Ghanaian fisheries, industrial and artisanal fisheries can be considered as two separate but interacting sub-sectors. Industrial fishery refers to all fishery propelled by inboard engines. This implies larger, more sophisticated vessels requiring relatively high initial in- vestment and operational costs but needing low per unit fishing effort. In Ghana's case, this relates mainly to inshore wooden hulled vessels, distant- water steel .hulled purse-seiners and trawlers, and tuna vessels. Artisanal fishing boats use out-board engines or sails for propulsion. This implies small vessels, often built of traditional materials, and fishing short distances from the shore with relatively low initial investment and opera- tional costs but high per unit fishing effort. Canoes are the main vessel in this category in Ghana. In addition, land-based fishing systems (i.e. beach netting) are included in this category as entirely artisanal in character be- cause of their low use of capital goods. The entire inland fishery, in the rivers and the Volta Lake, is artisanal in nature, and accounts for about 20% of total national catch by weight. The relative economic importance of the two sub-sectors in the Ghanaian economy is shown in Table 8.3. For the purposes of this study of the private sector, data for marine and inland artisanal fisheries have been combined. All inland capture fishery is artisanal in character. Especially in the artisanal sector, the number of fishermen cited here does not exhaust the number of 9 Excepting countries such as North Korea which catch most of their fish in distant waters. 10 For a review of the early history of coastal fishing in West Africa see Chauveau in Haakonsen (ed.) (1991). PRIVATE SECTOR FISH PROCESSING AND MARKETING IN GHANA 387 TABLE 8.3: Comparison of Artisanal and Industrial Fishing Sub-Sectors Total catch 1991 Vessels Fishermen Sector Tons % 1988 (No.) 1988 (No.) Artis anal 272,850 80 14,372 138,000 Industrial 68,400 20 346 4460 Total 341,250 100 14,718 142,460 SOURCES: Department of Fisheries Statistics for catch, UNIDO (1991) for vessels and fishermen. people employed in fisheries. There is considerable, but declining, employ- ment in boat building, boat repairing and net-making, as well as in the down- stream activities of processing, marketing and distribution. Figure 8.2 attempts to indicate some of the interactions and linkages be- tween the two sectors in their production, processing and marketing systems. Once fish enters the domestic marketing chain, distinctions between indus- trial and artisanal products become less well defined. Recent Trends The recent history of private sector. fisheries and fish-processing will be dis- cussed below. At this point it is worth discussing the broader context of fish- eries development in Ghana, in particular the relevant policies of the Ghana- ian Government, and the effects on fisheries of the Structural Adjustment Program initiated in 1982. Government intervention in fisheries became significant in the 1960s and 1970s. A State Fisheries Corporation (SFC) was set up, originally in 1961 using a Norwegian company to manage a fleet of deep sea vessels. At that time a separate company, Ghana Cold Store, bought the fish and marketed it. The price of the products was dete~ed by the Prices and Incomes Board and was based purely on input costs, in the manner of centrally planned economies. During this period prices were commonly set below operating costs. The fishing and marketing companies merged in 1974 to form the SFC itself. At this time it was operating twenty-five large trawlers in Angolan waters. The SFC set up forty cold stores in various places in Ghana. A second government intervention was the establishment, in 1973 of the Terna Food Complex Corporation (TFCC). This consists of a single large-scale integrated food processing plant with fish canning, smoking and freezing op- erations sharing the site with a flour mill (the core activity), an animal feed mill, an oil mill and a fish meal plant. FIGURE 8.2: Interactions Between Artisanal and Industrial Fisheries in Ghana PROCESSING/ RESOURCES EXTRACTION HANDLING DISTRIBUTION CONSUMPTION CAPITAL GOODS ~ ~- Local: Canning Greater nearer the coast and canoes, nets, INDUSTRIAL SECTOR ...._ (not operating) major inland water bodies. wooden hulls, Exported traditional smoking Inshore direct Likelihood of fresh or frozen kilns. Imported: Steel vessels, ..... Tuna Distant water -I Filleting & freezing ~ supplies diminishes with distance from landing site. Wealthy prefer large, fresh, engines, refrigera- -~ frozen or quality smoked. tion equipment, ice Fresh plant, packaging Poor eat small, frozen, materials, Improved smoked, cured or fermented. ~ smoking kilns. SERVICES ARTISANAL SECTOR ~~ Fresh on ice I Fish mammies: wholesalers, Harbours, repairs, fuelwood, salt, fuel, extension and research. f - .... Canoes offshore Canoes inland H Cold Storage I agents & retailers. NATURAL RESOURCES Land-based systems H Smoking (wet & dry) I Deep sea fishery Coastal fishery Inland fishery: ...._ H Drying I '"" ~ -~ "' Q) <>-"' o~ _J "' E - Consumptiom 1988 Canned Frozen Smoked (%) - 16 73 Capture Aquaculture H Curing/salting I Fresh 11 POLICIES .... IMPORTS Fermenting Macro-economic: Foreign exchange, domestic tax', f-- Imports/exports. Natural resource management I I ""-T I NOTE: Based on UNIDO (1991, p.8). PRIVATE SECTOR FISH PROCESSING AND MARKETING IN GHANA 389 The main reason for setting up the TFCC facilities seems to have been to take advantage of the hydro-electricity available from the Volta Lake devel- opment. It is not clear whether the facilities were ever fully utilized. The TFCC started large scale fish smoking, mostly of sardines but also of tuna, croakers and mackerel. It processed. canned sardine and mackerel, using fish imported from Senegal and South Africa. The fish meal plant processed some cannery waste but mainly sun-dried anchovies, purchased locally. The SFC and TFCC are still in existence, but with a greatly reduced role. When other African countries set up Exclusive Economic Zones, the SFC trawlers' catches declined, from a peak of 33,000 tons per year to none now. Despite the end of fishing activities, staff were kept on. In 1986 the Corpora- tion had 11 boats and 3,000 employees. Now it has 530 employees. In 1987 the SFC was placed on the Government's divestiture list, where it remains to date. This effectively prevented the SFC from raising further loans from the formal sector. To continue operation the current Chairman has organized a consortium of fish mammies11 who provide working capital in return for free cold storage. Total SFC cold storage capacity is 12,000 tons in 40 cold stores spread around Ghana. Some, like those in Terna harbor, are potentially highly prof- itable. Others, many in remote inland areas, are not. About 8,000 tons (66%) is operational at present. Profitability is low due to lack of a fast freezing facility, rigid operating arrangements, poor operating ~tandards, a heavy debt burden and aging plant. Some . inland cold stores have been rehabilitated using capital borrowed from fish mammies in return for rent free use. Lack of income from these stores is leading to problems with covering key recurrent costs such as electricity and staff salaries. Opportunities exist for the impor- tation of frozen fish, but lack of liquidity and an inability to borrow prevent the company benefiting from this. The SFC also owns a flake ice plant, recently rehabilitated with donor assistance. This lies idle because demand for ice from the inshore fishing fleet has collapsed. As a result of trade liberalization, the TFCC has found it very difficult to compete with cheap imports of canned fish. This problem has been compounded by the widespread provision of canned fish as food for work as part of donor 11 The term 'fish mammie' is widely used in West Africa to describe women involved in fisheries. It includes all women traders offish: wholesalers, retailers, agents, processors and financiers. 390 MARKETING AFRICA'S HIGH-VALUE FOODS assisted structural adjustment amelioration programs. Cold storage capacity at the TFCC is surplus to needs and poorly situated for use by private fish traders. Seasonal demand for cold storage means that the cold stores may not be economically viable on their own. The TFCC fish smoking plant has a capacity to produce twenty tons per day. This is far in excess of the demand likely from any individual traders. Furthermore, the high cost of firing the kilns means that they can only be used "to order". At present the canning plant is reaching 9% capacity utilization. A review of the Corporation in 1992 showed that trading losses have been incurred due to lack of raw materials, an inability to find markets for its products, a lack of investment in machinery and maintenance and the high cost of inputs, such as aluminium cans, due to currency devaluation. TFCC staff themselves point to a number of bureaucratic and non-commercial man- agement practices that have hindered profitable operation and reinvestment. The decline of the two parastatals is a result both of the over-centralized, statist policies that set them up in the first place, and the Structural Adjust- ment policies. that have replaced them. An Economic Recovery Program was adopted in Ghana in 1982, with the support of the IMF, to reverse negative growth rates and reduce massive budget.deficits, by measures including an increase in agricultural producer prices, withdrawal of subsidies on consumer prices, and devaluation. It has also entailed the withdrawal of support from, and government divestiture of parastatals. In the short- and medium-term it has undoubtedly had some negative effects on many sub-sectors, including both artisanal and private-sector industrial fishing. Some of the effects of Structural Adjustment policies on the economy are indicated by the figures in Table 8.4. Whilst inflation has been largely con- trolled, there has also occurred a credit squeeze and a reduction in the pur- chasing power on the part of many consumers. This has had very damaging effects on an industry heavily dependent on imported fuel, and, especially for the industrial sub-sector, on imported spare parts. Devaluation of the Cedi has been the primary cause of this, though constraints on the money supply have also had their effects in terms of poor access to credit and diminished demand for fish. Interest rates during this period have often been negative, falling behind the current rate of inflation between 1988 and 1990. However, access to the formal banking system is so limited by other barriers, that credit cannot be said to be effectively rationed by the interest rate. Rates for borrowing in the informal sector are much higher than those in the formal sector. PRIVATE SECTOR ASH PROCESSING AND MARKETING IN GHANA 391 . TABLE 8.4: Inflation Rate, Exchange Rate and Indicative Interest Rates Year Inflation (%) Cedis per US$ Lending Rates 1 1980 50 1.02 1981 117 1.15 1982 22 2.75 1983 123 20.33 1984 40 35.99 1985 10 54.37 1986 25 89.21 1987 40 162.37 1988 31 202.35 23.0-30.0 1989 25 270.00 22.5-30.0 1990 37 326.33 22.5-29.5 1991 18 367.78 19.5-31.5 1 Lending rates are for agriculture, forestry and logging. SOURCE: Statistical Services (1992). While the Government of Ghana's policies for fisheries12 now implicitly include assistance to the private sector, both artisanal and industrial, this has tended to concentrate on provision of equipment and infrastructlire for the industrial fisheries. 13 Recent reviews14 have noted the absence of inte- grated support to the artisanal fisheries sector, especially with regard to credit provision, post-harvest losses and resource management. Consumption and Prices Fish is Ghana's most important source of animal protein. Official statistics suggest consumption per capita ranges between 14kg. and 21kg. per annum, well above the average for West Africa. Consumption is much higher in the south of the country than in the north where poor communications make marketing fish more problematic. Neither quantitative nor qualitative market research into fish consump- tion patterns in Ghana exists. Anecdotal evidence, and discussion with key informants, suggest the following: 1) Meat (including poultry) is preferred to fish. The widespread availabil- ity of cold chains in urban and peri-urban Ghana has encouraged the development of markets for frozen meat and poultry. The authors' ex- 12 As described in UNIDO (1991, pp.10-11). 13 With the notable exception of work on the Chorkor kiln. 14 FAO (1989), UNIDO (1991). 392 MARKETING AFRICA'S HIGH-VALUE FOODS perience in other countries suggests that higher income groups will switch from fish to meat and poultry unless fish products of better quality (usually this means fresh fish) are available. While it is often believed that fish are effectively the last resort for poor income fami- lies, such families often in fact choose meat as a low-cost protein source. It can be purchased in smaller quantities and is perceived to result in less waste, to be more filling and very versatile. 2) Fresh and frozen fish are (jointly) preferred to cured fish. Generally, wealthier consumers prefer larger, fresher fish for which they are pre- pared to pay a premium. Less wealthy consumers tend to buy less fresh fish, and of a smaller size. The cheapest fish available is that which has deteriorated, or "stink fish". Poorer consumers buy smoked fish that has been re-smoked, is broken into smaller pieces by over-handling, or subjected to attack by mould and/or insects. 3) Consumer preferences between freshwater and marine fish are hard to define. Migration of people from areas of traditional freshwater fish consumption to cities in the south of Ghana has led to a significant trade in fish between these areas. Similarly, since large freshwater fish can be sold profitably in the south, the gap in the market is often filled by marine fish, either smoked or recently frozen. However, in areas distant from sources of supply, for example northern Ghana, choice is limited and demand for some types of fish products, such as fresh and frozen fish, goes unfulfilled. 4) In Ghana, unlike most countries, there is no consumer resistance to frozen fish as compared to fresh fish. Fresh fish statistics quoted in much of the literature actually refer to recently de-frosted fish. No pre- mium is available for fresh or iced fish (with the possible exception of the catering trade which makes its own arrangements for fresh sup- ply). Frozen fish is also thawed for subsequent smoking, and is also the basis of the increasingly popular dish of fish fried in vegetable oil, cooked in roadside restaurants and valued as a present from travellers. 5) Niche markets exist for crustaceans, snails, eels, bivalves, shark fins, crabs and numerous other aquatic products. These items are generally more readily available in the markets of more populous areas where a strong demand exists. PRIVATE SECTOR FISH PROCESSING AND MARKETING IN GHANA 393 FA015 estimated the demand for fish in Ghana in 1990 to be between 30.6 and 36.2 kg. per capita, implying a shortfall in supply of 50%. This might have been met by imports but a shortage of foreign currency caused fish imports to decline to a low of 22,000 tons in 1987 from a peak of 50,000 tons in 1977. Subsequent devaluation has led to a recovery in this business with about 34,000 tons being imported in both 1992 and 1993. It is suggested that imports may now have peaked as consumer purchasing power has not kept pace with the increase in prices of imports. One of the consequences of the Structural Adjustment policies16 was that inflation has been higher than the growth of nominal incomes of a large part of the population. By July 1993, the average per capita consumption of about 20kg. fish per annum, if supplied by frozen mackerel, could cost 11,000 Cedis or 4% of a typical laborer's annual income. Loss of consumer purchasing power may have been to the benefit of fisherfolk because there seems to be some degree of substitution between fish protein and other more expensive sources of protein such as beef or pork. Various studies by researchers affiliated to Cornell University17 show that changes in income result in a high degree of substitution between the differ- ent calorie sources of the varied Ghanaian diet. There is, however, other evidence that, despite diminished real incomes consumers are spending more on meat and poultry. While data on animal offtake (and thus meat availabil- ity) were not available for the present study, the national populations of sh~ep, goats and pigs have increased markedly since 1987 (the poultry population has also increased, while cattle numbers have remained static). The advan- tages. of meat over fish perceived by lower-income groups, as detailed above, may be a factor here. Information on the division of consumption between income groups and in different geographic areas might reveal more about the impact of Structural Adjustment on fish consumption patterns. Seasonal price fluctuations are determined by market forces. Fish supply peaks in August/September, and the glut of pelagic species depresses the whole market. In. addition, the price of fish is determined by the supply of alternative protein sources such as chicken and beef. Fish are sold either whole or by the piece making comparisons of unit prices and assessment of 15 FAO (1990). 16 As noted by FAO (l989). 17 Alderman (1990) and (1992), Alderman and Shively (1991) and Alderman and Higgins (1992). 394 MARKETING AFRICA'S HIGH-VALUE FOODS relative marketing margins difficult. Retail and wholesale price information is therefore limited. Fish prices exhibit all the characteristics of trade uncertainty and fluctu- ating supply. Prices can change fourfold in a day depending on the size of landings. Marketing margins appear considerable, but transaction costs and risks (due to perishability and irregular supply) subsumed by marketing agents are relatively high compared with other commodities. Prices reflect seasonal supply variability and physical distance from the source of supply. The State of the Resource There are three areas of resource constraint in the Ghanaian fisheries sector, that of the fish itself, the timber required to construct the fishing boats, and the biomass needed to process the fish once landed. Evidence for the pressure on existing marine fish stocks is limited, though anecdotal evidence suggests that catch per unit of fishing effort is decreasing and that the use of illegal fishing methods (i.e., reducing the mesh size of nets) is also increasing.18 Fishermen and middlewomen alike talk of increasing proportions of juvenile fish in the catch. In recognition of this problem, and to prevent exploitation of the resource by industrial fishermen from other countries, the Government of Ghana has taken several measures. Ghana declared an Exclusive Economic Zone of 200 nautical miles through the Territorial Waters and Continental Shelf Act (amended) of 1977. Other important regulations effecting fishing are as follows: 1. licences are required to fish in Ghanaian waters; 2. tuna vessels registered in Ghana are required to land at least 10% of their catch in Ghana; 3. importation or construction of new fishing vessels (this applies espe- cially to trawlers) is only allowed as replacements for existing ones; 4. minimum mesh size is specified for different types of gear; 5. at least half the crew of each vessel must be Ghanaian. These and other regulations effecting the production and marketing of fish are not effectively enforced because insufficient provision has been made. for monitoring, surveillance and control. 18 For example, many canoe operators now find it economic to use ice on board their vessels to reduce spoilage and to allow them longer over the fishing grounds. This is a sign that they are travelling further out to sea and spending longer meeting catch targets. PRIVATE SECTOR FISH PROCESSING AND MARKETING IN GHANA 395 A similar situation occurs in the lake fishery, though this is protected from over-exploitation for the time being by large trees which were covered when the lake was constructed and which effectively prevent trawling. Anec- dotal evidence suggests that fishing for fresh-water fish is highly profitable and that large numbers of fishermen have migrated to the lake from the coast to exploit this resource. There, however, is no research available to support this contention. The use of large forest trees for canoe construction is traditional in Ghana. The great expansion of the canoe fleet during the 1970s and 1980s with the development of motorized fishing, coupled with a tendency to develop larger canoes which can go further to reach more productive fishing grounds, has meant that canoe carvers now have to travel considerable distances to find suitable trees. This is reflected in the increased costs of commissioning ca- noes (around Cedi 2 million each} 19 Use of wood to smoke fish, hitherto not considered a great problem in Ghana, is potentially a severe resource constraint for the future. In areas of intensive smoking, forest wood has become scarce and must be purchased from the hinterland, representing a significant recurrent cost for fish smok- ers who must pay at the beginning of the peak season in cash. Discussions with traders from the Volta Lake suggest that wood for smoking is becoming scarce on the lake islands where this activity is carried out and wood must now be purchased and transported from the mainland. Monitoring and control of open access resources such as fisheries and for- ests is not easily managed in the private sector. However, despite commit- ments by the Government of Ghana to police resource use, no action has been taken. The issue of resource management requires further attention. Current Production, Imports and Exports Recent catch data for Ghana are given in Table 8.5. These clearly show the fluctuation in anchovy catches, with catches varying from 15,200 tons in 1986 to nearly 88,000 tons the following year. Cured fish production is described in Table 8.6 and imports and exports offish and fish products are given in Table 8.7. What these figures represent, partial as they are, is the Ghanaian fisher- ies sector in a state of stagnation relative to the rapid growth of the 1960s 19 See the innovative paper by Chevas in Haakonsen (ed.) (1991, pp. 233-242). 396 MARKETING AFRICA'S HIGH-VALUE FOODS TABLE 8.5: Fish Production in Ghana, 1985-1992 (tons) Fishery 1985 1986 1987 1988 1989 1990 1991 Marine Canoe 159,230 190,200 262,380 244,560 220,880 242,020 215,850 Round Sardine 54,070 45,490 45,670 75,850 Na. Na. Na. Flat Sardine 22,230 16,630 25,480 10,450 Na. Na. Na: Anchovy 27,590 15,210 87,890 75,900 Na. Na. Na. Inshore 17,980 21,890 14,930 7410 12,660 9250 7360 Round Sardine 9430 5510 1740 70 Na. Na. Na. Flat Sardine 1830 2090 1700 100 Na. Na. Na. Distant waters 56,340 57,060 53,640 51,470 55,740 68,120 66,460 Shrimps 380 730 780 Tuna 34,410 34,270 33,470 35,430 32,290 40,800 37,790 Total Sea Fishery 233,550 269,150 330,950 303,440 289,280 319,390 289,670 Total Inland Fishery 55,000 57,630 57,660 58,000 57,000 Na. Na. SOURCE: Department of Fisheries, personal communication. and early 1970s. This stagnation is particularly marked in the industrial sub- sector, where some indices of production are actually declining. What ac- counts for the stagnation in fish production: is it a supply side constraint or is it demand led? The evidence seems to suggest that both forces are at play. On the supply side, the down-tum in the industrial fisheries as a result of re- duced resources (i.e. the imposition of EEZs) and higher costs (i.e. a result of currency devaluation) has not, it seems, been reflected in radically higher prices for landed fish. Furthermore, though imports have risen (see Table TABLE 8.6: Ghana: Cured Fish Production ('000 tons)a Year Dried6 Smoked 1981 6.8 44.0 1982 6.3 43.2 1983 6.5 44.5 1984 6.2 48.8 1985 6.5 49.6 1986 8.3 58.1 1987 8.3 44.0 1988 8.6 46.1 1989 8.6 46.1 1990 8.7 46.4 NOTES: a) 1 ton of cured fish is derived from 3 to 4 tons of wet fish. b) Freshwater dried fish only. SOURCE: FAO (1984, 1991). PRIVATE SECTOR FISH PROCESSING AND MARKETING IN GHANA 397 TABLE 8.7: Ghana: Fish Imports and Exports Year 1987 1988 1989 1990 Imports Fresh, chilled, frozen Q 21.7 33.1 30.4 31.0 v 14.0 26.5 24.6 30.0 Dried, salted, smoked Q 0.7 0.4 0.1 0.1 v 1.5 1.0 0.4 0.3 Exports Fresh, chilled, frozen Q 20.7 24.9 12.7 12.4 v 14.2 20.1 11.4 19.7 Dried, salted, smoked Q 0.3 0.4 0.2 1.0 v 0.1 0.2 0.2 0.5 Crustaceans, molluscs Q 0.4 0.3 0.5 0.7 v 0.9 0.7 1.6 2.7 Q = quantity in '000 tons, V = value in million US$. SOURCE: FAO (1991) . . 8.7), production growth still lags behind the expansion in population. This suggests that in addition to increased input costs, Ghana fisheries have suf- fered from the deflationary fiscal and monetary policies of the Ghanaian Gov- ernment during the period of Structural Adjustment. Consumers' willingness or ability to buy fish has been limited by declining real income growth. Most of these imports are oflow value small frozen pelagic fish from North- ern Europe, though some tinned fish (mainly mackerel and tuna) is also im- ported. Traditionally, frozen fish is imported to meet the fish supply shortfall outside the main local fish catching season. It is believed that little or no fish is imported into Ghana from other West African countries at present, though historically, some Ghanaian migratory fishermen have returned their catch to home. Also it is said that fresh water fish used to.be imported from Mali to Kumasi. Marketing Fish production, processing and marketing in Ghana20 can be considered as four distinct systems, shown schematically in Figures 8.3--8.6. These are: 1) Artisanal: canoes and beach seines, often processed and marketed by immediate family members. 20 The most comprehensive review of fish marketing in Ghana is contained in Infopeche (1990). Descriptions in DANIDA (1988), FAO (1989) and UNIDO (1991) are also useful. 398 MARKETING AFRICA 's HIGH-VALUE FOODS 2) Inshore: larger wooden hulled vessels with inboard motors, product usu- ally marketed by an immediate family member. 3) Industrial (deep sea and tuna fishery): Ocean going vessels, product usually marketed by a. middlewoman with strong financtal ties to the boat owner. 4) Freshwater fishery: canoes using nets and traps on the Volta Lake system selling through family members to middlewomen from the same ethnic group. FIGURE 8.3: Market Chain for the Marine Artisanal Fishery Canoe operators Commission agents Wholesalers Short-term cold storage Export agent Market retailers Export wholesaler Export NOTE: A box denotes transfer of ownership. A circle denotes other physical movements not involving transfer of ownership. Unboxed denotes a process or activity. PRIVATE SECTOR FISH PROCESSING AND MARKETING IN GHANA 399 FIGURE 8.4: Market Chain for the Marine Inshore Fishery Boat owners Freeze Freeze Wholesalers Cure/smoke/dry/fry Cure/smoke/dry/fry Consumers NOTE: A box denotes transfer of ownership. A circle denotes other physical movements not involving transfer of ownership. Unboxed denotes a process or activity. In addition, there are fisheries imports and domestic exports which link into these systems. Fish mammies may conduct a variety of marketing activi- ties. For convenience these have been disaggregated in the diagrams into specific marketing functions such as wholesaling, retailing, commission agents etc. Artisanal Sector Marketing and Processing The great majority of domestically produced and imported frozen fish (as much as 80%, estimated by the Department of Fisheries) is processed into traditional products. Fr'esh fish at landing sites is sold to fish mammies, who 400 MARKETING AFRICA'S HIGH-VALUE FOODS may be wives of the fishermen, and who may be owners of the boats, or creditors of the fishermen. Fish mammies may act as buyers on their own account, or as agents sell- ing to wholesalers, in which case they do not take title to the fish, but take a commission on its sale. In either case, fish mammies often have a financial interest in the fishing operation. Transactions between wholesalers and re- tailers are often done on credit with wholesalers deferring payment until after on-sale has been achieved. On the whole, the marketing chain is relatively short, with fish changing hands about twice before purchase by the consumer. In the case of dried fish travelling to distant rural markets, two or three wholesalers may be involved before retail. Fish is bulked up and wholesaled in large baskets (of dried or smoked fish) or packaged blocks (of frozen fish). A considerable ancillary trade in the sup- ply of baskets, packaging material and packaging exists, centered around FIGURE 8.5: Market Chain for the Tuna Fishery Boat owner/tuna fishing companies Cold store Small fish Large fish As per Figure 2 Exporters/processors NOTE: A box denotes transfer of ownership. A circle denotes other physical movements not involving transfer of ownership. Unboxed denotes a process or activity. PRIVATE SECTOR RSH PROCESSING AND MARKETING IN GHANA 401 FIGURE 8.6: Market Chain for Freshwater Fishery Fishermen Wives of fishermen •· Smoke, dry or cure Locally-based City-based middle women middle women Retail on Retail on local market urban market NOTE: A box denotes transfer of ownership. Unboxed denotes a process or activity. major producing areas and markets. Fish landed at Terna fishing harbor is auctioned in wooden crates. Transportation involves hiring a truck or pick-up. Groups of wholesalers often combine to rent a truck. Transport costs reflect the increasingly high cost of fuel. For example, a trader in Kumasi market pays Cl,000 ($1.61) per 20kg. carton of frozen fish for transport from Terna, a distance of about 110 kilometres. This represents around 12% of the purchase price of C8,500 per carton. Processing is conducted at the major landing centres like Terna, Elmina, Chorkor etc., and also in the main inland consumption/marketing centres such as Kumasi. Cured fish in its various forms is the main fish product in Ghana. Only low-cost technology is used, or needed. Salting can conveniently be carried out in tanks built of concrete or baked brick. Sun-drying can be done on the sand or earth, though it is better to use racks made from wire netting, reed slats, or old fishing nets supported by a wooden frame and 402 MARKETING AFRICA'S HIGH-VALUE FOODS resting on wooden poles, to reduce contamination with dirt and animal excre- ment. The fish is then raised above ground where there is more wind, and moisture loss can take place on both sides of the fish. Drying is often done on a tarmac road, where the hard surface serves as a barbecue. Smoking fish is the most common method of processing fish in Ghana.21 Two methods gre practiced, hot-smoking and smoke-drying. The former has a higher moisture content and therefore reduced (2-3 days) shelf life in com- parison with smoke-dried fish. Sardinella, horse mackerel, mackerel, snap- per, threadfin bream, croaker, grouper, seabream, barracuda, tuna, catfish, sharks and rays are the most common fish which are hot-smoked. Smoking is carried out by fish mammies both at the landing sites and inland. For inland smoking, frozen fish is brought from cold stores at the coastal landings and transported to the smoking area. Once smoked, fish is either sold directly to retailers or taken to weekly markets and sold in bulk to wholesalers both from Ghana and neighboring countries. The Tuesday mar- ket in Chorkor village and the Thursday market in Djemeni are good ex- amples of wholesale markets for smoke-dried and hot-smoked fish. Retailers of the relatively perishable hot~smoked fish must re-smoke the product regularly to prevent deterioration. In Kumasi, an ancillary trade in transporting fish from the Central Market back to the smoking areas for re- smoking has developed. Quality and therefore price declines with each re- smoking. Smoke-dried fish can last up to nine months. Anchovies, shrimp and most freshwater fish are smoke-dried. Mackerel and sardines can be processed in either way. Fish were formerly smoked in primitive kilns, mud-brick or dug-out, as still used in other parts of Africa. Soon after international trade in oil devel- oped it was found that forty-four gallon oil drums could conveniently be used as fish smoking kilns. With both ends cut off, and holes pierced in the sides to take rods to support the fish, and set on blocks over a wood fire, an oil drum serves as an effective kiln. In 1970 the FAQ introduced an improved, fuel- efficient kiln made mainly of mud-bricks, known as the Chorkor kiln after the fishing village where the project was based. This has become very popu- lar in some areas, but requires a larger investment than oil-drum kilns, and is only efficient with large quantities of fish. Many oil-drum kilns, more effi- 21 Koranteng (1992). PRIVATE SECTOR FISH PROCESSING AND MARKETING IN GHANA 403 cient with smaller quantities, are still iii use, alongside or instead of the improved kiln. The establishment by SFC and TFCC of cold stores throughout the coun- try had a marked effect on fish marketing and processing. In most tropical countries fish is frozen only when it is intended for export. The lack of retail outlets with freezers prevents the local sale of any significant quantity of frozen fish. However in Ghana glut catches of fish are frozen and stored until the slack season, when it is thawed, either for direct sale to consumers (in a part-frozen state, for immediate use) or for smoking. As far as the authors are aware, there are no similar operations in any other country. Fish-drying occurs in two forms: sun-dried and salt-dried. These processes take up to three days, involve :qrinimal capital outlay and are therefore com- monly practised during gluts. Salt-drying is a preferred method of processing Tilapia. Fermented products are used mainly as flavor enhancers in the Ghanaian diet. Fermented fish, often referred to as "stink fish", is usually made with fish that has deteriorated beyond a marketable stage. Ethnicity, Gender and Traditional Marketing Methods While ethnicity and gender are not irrelevant to industrial sector fisheries, it is in the artisanal sector that they are most clearly seen as structuring the division of labor, and they are therefore best discussed at this point. A number of different ethnic groups are involved in fishing activities. There are a number of cases in which identifiable ethnic groups work in fisheries outside their home areas, or where markets are divided along ethnic lines. For example: (a) Ga women in Winneba migrate seasonally to Elmina, a predominantly Fante area, to smoke fish; (b) Ewe women from Volta Region smoke fish on the coast and market it in their home towns, whilst fresh water fish smoked in Volta Region is marketed in Accra and on the coast, also by Volta people; (c) Women traders in Kumasi use family members resident in Terna to purchase and ship frozen fish for processing and sale. Using family members or fellow-members of an ethnic group is a way of reducing the high transaction costs in both the artisanal and the industrial sectors. Such costs are increased by poor communications, undeveloped fi- 404 MARKETING AFRICA'S HIGH-VALUE FOODS nancial systems, uncertainty and insecurity. Middlewomen complain that huge amounts of cash are required to conduct business as a result of inflation and devaluation. This is both inconvenient and insecure. Banks have begun to issue Cedi travellers cheques, but banking hours do not generally fit well with trading activities and few small-scale fish mammies have bank accounts. An example of traditional marketing methods is the communication of price information between family members in the fish marketing chain, from landings to inland marketing areas. Beans or stones tied in a piece of cloth are sent with the fish denoting the purchase price and expected sale price. Individual families have their own codes which are passed on through gen- erations. Women, many of whom are illiterate, often group together to share transport costs, and an elaborate system of markings has evolved using spe- cific knots, cloth strips, leaves or painted symbols to distinguish baskets and parcels on arrival. Women are involved at all levels of Ghana's fishery, both artisanal and industrial, pre-harvest and post-harvest. Given the unique position of women in organising and financing private fish trading in Ghana, this would seem to be an under-researched area. 22 It has been estimated23 that, including fuel and food supplied to crew, pre- financing of fishing activities, processing and markets, women manage 40- 60% of the cash flow of artisanal fisheries in West Africa. In Ghana this may be an underestimate. Increasingly, women pre-finance fishing operations in order to ensure continuity of fish supply. Another estimate is that up to 90% of marine produce is distributed by fish mammies in Ghana.24 In a typical artisanal fishing community, men harvest fish and sell the product to their wives or other local women. 25 Individual fishermen sell their whole catch to a single fish mammy. Fish mammies are often also entitled to shares of the catch because they are owners or part-owners of boats, or are entitled to buy at low prices because they have loaned money for nets, out- board motors and fuel. Husbands and wives pursue their own separate objectives and maintain their own budgets in these transactions, even when they deal with each other. 22 But studies by the FAO (d'Assise 1992 and Randall n.d.) and IDAF (Haakonsen (ed.) 1991 and IDAF 1991) provide the main background. 23 D'Assise (1992). 24 Rapdall (n.d.). 25 Note that fishing itself is not an exclusively male domain in Ghana. Particularly in the inland fishery and coastal land-based fishery, women also participate. PRIVATE SECTOR FISH PROCESSING AND MARKETING IN GHANA 405 But it is difficult to discern how and by whom prices are fixed, and how this varies with the degree of kinship between fisherman and fish mammy. Fish- ermen see fish mammies, through their power as creditors, as fixing pur- chase prices below going market rates. Discussion with a number of fish mammies suggests that prices are usually fixed by the husband/male family member in an arbitrary manner at what is considered a "fair" price for that day. The woman is then expected to sell at a price over that figure. This is not always possible and many women complained that they were forced to loan money to their husbands whether they had made a profit or not. The likelihood of immediate kinship links between fish mammies as buy- ers and fishermen as sellers seems to diminish with the scale of operation. In canoe fishing, processing and marketing is typically conducted within an ex- tended family group. Sales of the catch to unrelated fish mammies occur but are not considered as mutually beneficial as those within family groups. Fish mammies operate on several levels and considerable vertical mobility exists. Women often start out as retailers offish supplied by related wholesal- ers, or assistants to processors in the same kinship group. With experience it is possible to accumulate capital quite rapidly and enter the long distance smoked and frozen fish wholesaling trade. Discussion with fish mamrpies travelling long distances to buy fish gives the impression that the effort and hardships involved in this type of trade fil"e considerable. Women traders from Togo, for example, must spend three or more days travelling day and night to markets in coastal Ghana. Since many fish markets commence at dawn, particularly at landing sites, it is common for women traders and their children to sleep in the market. In some mar- kets, such as Chorkor village, businesses have developed to cater for this, by providing secure places for visiting traders with large amounts of cash to sleep. Farther away from main trading centres where roads are poor, fish mammies must travel for several days to buy fish and return in time for regular village markets. For example, traders from Tamale in the north of Ghana travel to Kumasi, Terna and Elmina on the coast to get supplies of smoked fish. It should be remembered that the work of fish mammies is combined with child rearing, small-scale farming and household management. Women also play a major role in the industrial fisheries of West Africa. Ghanaian fish mammies own or manage a considerable proportion of the inshore fleet. Many of these individuals have invested profits from earlier pre-financing activities in larger vessels, or have taken over management due 406 MARKETING AFRICA'S HIGH-VALUE FOODS to accumulated debts. Many, even at this level of operations, are illiterate and do not keep accounts. Nevertheless, they have continued operations de- spite rising input costs and increased competition from deep-sea trawlers in Ghanaian waters. AE far as the authors are aware, this group has not been directly involved in donor funded activities to date and very little is known about their scale, scope or operating patterns. Industrial and Semi-industrial Fisheries The story of industrial fisheries in Ghana is one of unabated decline (with the notable exception of the tuna fishery) since the introduction of Structural Adjustment measures in 1983. To some extent this decline is disguised by the statistics, especially those for number of vessels, because they do not describe the sea-worthiness of the ships, which has declined drastically. Several of the larger companies now have fleets of trawlers which are unable to put to sea because they need repairs or because they are the subject of legal action with regard to unpaid debt. There are three main types of industrial sea fisheries in Ghana: the in- shore fleet (coastal purse-seiners and trawlers), the distant water fleet (deep- sea refrigerated trawlers and purse-seiners) and the tuna fleet. The inshore fleet is characterized by locally constructed, wooden, multi- purpose vessels. Small vessels (9-12m in length) operate from Mumford, Elmina and Sekondi and were designed to harvest trigger fish. Larger, more sophisticated vessels (12-23m in length) operate from Terna. The FAQ (1989) observes that the relatively high level of technology adopted by this latter group (i.e, echo-sounders, sonar) does not seem to have been justified and is not currently in use. In 1987 only 257 of the 388 vessels registered at Terna were operating, and fishing effort was low except during the sardine season. This reflects the poor economics of this fishery. Although up-to-date figures are unavailable, discussions held with boat owners and boat builders in Elmina reiterate the pattern of decline experienced in Terna. Respondents mentioned longer peri- ods at sea necessary to reach target catches, high input costs, heavy debt burdens and the disappearance of the trigger fish as the causes of decline. The inshore fleet contributed 8% (21,890 tons) of the total marine catch in 1986, its peak year. Subsequently it has declined to 3% (7,360 tons) in 1991. This is a result of intensified competition from vessels that used to fish in distant waters, increased input costs and declining catch per unit effort. PRIVATE SECTOR FISH PROCESSING AND MARKETING IN GHANA 407 Characteristics of the inshore fleet which may constrain further devel- opment include: aging boats, a decline in fleet size, high indebtedness and lack of bargaining power. There is increased hegemony of fish mammies who hold debt, though this is not necessarily a problem as many are related to boat owners and therefore have vested interests in maintaining production. They are also often on-lending down the marketing chain to wholesalers, retailers and cold-store operators. The dramatic example of the State Fisher- ies Corporation raising working capital from fish mammies has already been mentioned. Distant water fishing (outside Ghanaian waters) was effectively curtailed by the imposition of the EEZ in 1979. In the 1960s and 1970s the Ghanaian fleet ranged along the coast of west and south-west Africa from Mauritania to Namibia. In 1989 there were twenty-eight vessels registered of 30-70m in length, of which ten were out of service and eleven were only used irregu- larly. A number of even larger vessels are owned by the State Fisheries Cor- poration and a few companies in Terna: all are unserviceable and few if any will ever put to sea again. The experience of one company involved in distant water fishing is illumi- nating. Kaleawo Fishing Company purchased eight trawlers from Poland in 1969 at a time when fish prices were fixed by the Government of Ghan:a and all duties on inputs had been waived, making industrial fishing financially very attractive. Under a joint venture agreement, the vessels were skippered by expatriates and crewed by Ghanaians. All fish were landed and sold in Ghana. The bank loan for these vessels was paid off in full by 1974. Between 1974 and 1983 these vessels earned high rates of profit for the company. Structural Adjustment measures introduced in 1983 had two immediate effects: firstly, the company suddenly found itself unable to extend its credit at the bank to pay for day-to-day operational costs due to the anti-inflation- ary credit squeeze. Secondly, rationing of foreign exchange meant that the supply of spare parts for their foreign built ships dried up. By 1983 only five of the eight original vessels were operating. Subsequently, these vessels too had to be abandoned. To maintain activities, the company diversified into cold storage whilst at the same time undertaking a joint venture to fish with a Soviet fishing company. This fizzled out with the collapse of the former Soviet Union. As of July 1993 the company had only one serviceable boat wholly under its own name. This boat does not operate, but is incurring harbor dues (a 408 MARKETING AFRICA'S HIGH-VALUE FOODS common problem). The company is buying a shrimper under a hire-purchase agreement with a Greek company; the Greeks provide officers and the vessel in return for 70% of the catch. Kaleawo dispose of the remainder of the catch locally to pay for inputs and the salaries of the expatriate officers. Under the joint venture agreement, 30% of the foreign exchange earned by the opera- tion can be repatriated to Greece and 70% must stay in a foreign exchange account in Ghana to pay for inputs. Kaleawo claim that of the original loan of US$176,000 over five years, none of the principal has been repaid (after two years) due to the inflated cost of inputs as a result of devaluation of the Cedi. The tuna fishery was developed in the early 1960s using Japanese, Ko- rean and US vessels selling to Starkist Foods for export to the United States. Domestic companies started fishing in 1973, and by 1987 the fleet consisted of thirty-six vessels. The history of Mankoadze Fishing Company illustrates well how compa- nies became involved in this sector and the problems they have encountered maintaining viability in the wake of both Structural Adjustment and the greatly reduced fishing area available since the imposition of the EEZ. Mankoadze started shortly after the end of World War II in a small fish- ing village using canoes. The original capital was provided by the family of the owner, a retired army officer who was well connected in the village and within Government circles. This proved helpful in expanding the fleet during the early 1950s when various Government schemes were available for boat building. In the late 1950s when distant water fishery was expanding, the Company got loans from U.K. financial institutions to build distant water vessels in the U.K. for use in Ghanaian waters. In the 1960s, the Company began a series of joint ventures with Ameri- can, Japanese and Korean companies to catch tuna for canning in the company's recently opened canning factory for both re-export and domestic sales. This business continued successfully throughout the 1970s, but in the 1980s the American company (Starkist) pulled out due to the poor security situation in Terna harbor, and Structural Adjustment measures changed the economics of the activity: fuel and spares became expensive and foreign debt grew alarmingly. Eventually the canning operation had to be closed down because the cost of importing cans was prohibitive. Efforts to diversify the business included setting up a farming enterprise in the mid 1980s. In a sense this was vertical integration, because the main objective was to provide food for fishing crews, an important variable cost. PRIVATE SECTOR FISH PROCESSING AND MARKETING IN GHANA 409 They also built repair shops for their boats to reduce reliance on government yards which were perceived to be expensive and inefficient, cold storage facili- ties to reduce reliance on the State Fisheries Corporation facilities, and ware- housing facilities to store inputs which had to be purchased in large quanti- ties when available due to supply uncertainties. Thus, in order to maintain efficient supplies of inputs and continue regu- lar operation, the company had to tie up capital in operations which contrib- uted little to overall profitability. The motivating force behind this diversifi- cation and integration was to ensure maximum time at sea to enable its fleet to cover its fixed operating costs. The net result was a large and cumbersome company with a wide variety of interests and top-heavy administration. In the late 1960s Mankoadze had thirty-two boats. In 1973 it built a can- ning plant, and in 1981 it purchased eight large vessels from Norway. By July 1983 only seven of these vessel were serviceable. Of the 2100 workers employed in 1989, 500 remain. The existing operation is based upon three vessels fishing for herring in Senegalese waters. The catch is landed in Dakar, frozen and trans-shipped to Ghana, where it is held in store and sold to fish mammies for distribution. The company used to freeze in Terna, but finds it cheaper and easier to do it in Dakar. The company's remaining boats are engaged in tuna fishing, selling about 20% of the catch locally through fish mammies and the remainder to Starkist for export. The contribution of tuna fishing to the total Ghana catch has remained fairly stable, peaking at 15% (34,410 tons) of the total marine catch in 1985. Tuna production has averaged roughly 13% of marine fish production during the period 1985-1991. Demand for tuna for export has been relatively stable. The ability of the product to earn hard currency has partially hedged the sector as a whole against currency devaluation which has hit the rest of the industrial fishing industry so severely, but high input costs have still left the fleet in a very poor state. As illustrated by the case study, individual busi- nesses have struggled to survive under the conditions of the Structural Ad- justment program, partly because of problems inherited from the previous period when the economy was centralized, and partly due to the changes in cost structure resulting from currency devaluation. A poor herring/sardine catch ~uring the 1992 season has led many to believe that this resource is in decline, although this could be the result of a single-year fluctuation in sea temperatures-the 'El Nino effect'--or other 410 MARKETING AFRICA 's HIGH-VALUE FOODS factors. For a number of reasons the industrial fishery would have problems responding to an upturn in its fortunes such as improved catches and en- hanced prices. High levels of debt to indigenous banks and to foreign donors and investors will make getting further loans or re-capitalizing the industry problematic. Boats are lying idle. Plant is out of date and needs replacing. Overmanning is widespread, especially in what remains of the public sector. Banks lack confidence in existing management capability. The single largest element of operating costs for all industrial fisheries is fuel, possibly accounting for up to 70%.26 Fuel costs, therefore, are the largest single determinant of profitability. Industrial fisheries have been exempted from a recent 60% increase in taxation on fuel. Anecdotal evidence suggests anyway that many in the semi-industrial sector are still buying fuel at sea from tenders bringing fuel from other West African countries, notably Nige- ria. In July 1993 marine diesel cost US$220/ton in Terna harbor but was available for between US$185-190/ton off-shore. Fuel purchased off-shore must be bought in US Dollars. Despite the apparent free convertibility of the Cedis, this is still considered to be a problem since the banks periodically face short- ages of US dollars. Industrial Processing and Marketing The tuna export market is dominated by one ns. company (Starkist) which buys the great majority of large landed tuna for re-export in frozen form. Some is converted to frozen loins. The price variation with fish size is consid- erable. For export, a premium is paid for large fish (see Table 8.8), but for domestic fresh consumption and smoking/curing, heavy competition has bid up the price for smaller fish (less than 3 lb) to C476/kg. Starkist will shortly commence construction of a 100 ton per day canning plant for tuna in Terna. They have negotiated highly beneficial conditions for this plant with the Government of Ghana, including tax holidays and tariff exemption. They plan to conduct a bonded operation27 using locally landed fish, imported cans and imported soya oil. The product will be exported to the U.S. and particularly the European Community, where it will not incur im- 26 This figure is disputed by a World Bank Report of 1989 as an exaggeration, putting the figure nearer 50%, which is still significant. However scrutiny of one company's annual balance sheet during the field work for this report adds credence to the 70% figure. 27 The term bonded means that goods imported into the factory and re-exported will not be counted as imports and exports and will not, therefore, be liable for duty. PRIVATE SECTOR FISH PROCESSING AND MARKETING IN GHANA 411 TABLE 8.8: Indicative Landed Tuna Prices, July 1993 Weight (lbs) Landed Price (Cecli/kg.) >20 600 7.5-20 600 4--7.5 600 3-4 278 <3 118 NOTE: C630 = US$1, June 1993. SOURCE: Starkist International. port duty due to Ghana's status as a signatory of the Lome Convention. Ex- isting total landings will not meet this demand for some years to come. The project is a joint venture with a large local fishing company which has experience in managing a canning operation. The operation will be vertically integrated from the canning plant onward, with Starkist's international mar- keting operation being used to distribute the product. Within Ghana it will not be integrated, but will rely on either buying tuna under contract from boat owners or through fish mammies. Potential benefits to the Ghanaian economy of the development of this large canning operation will be: firstly, the creation of 370 jobs, mainly for women manual workers; secondly, the increased use of currently over-sup- plied electricity utilities; thirdly, increased demand for locally landed tuna using locally owned vessels and crew. Finally, the additional supply of fish waste from the factory will be sold locally providing a valuable input of pro- tein meal into local animal feed ingredients which have decreased in quality and increased in price in recent years. Potential draw-backs of this approach to inward investment include: firstly, that profit will be held in US dollars and will probably be repatriated rather than re-invested; secondly, the quality of employment created will be low and transient, i.e. low skilled, low waged, on a day-to-day basis; thirdly, that the investment may not be as long-term as it appears. Canning operations are easily moved. Starkist has moved its operation in Ghana to Cote d'Ivoire in the past when it was not happy with conditions in Terna. Finally, the addi- tional demand for tuna may result in a reduction in fish available for domes- tic consumption, though tuna: canneries demand large fish and the domestic market is for small ones. Starkist have a long track record in the Ghanaian tuna fishing industry, 412 MARKETING AFRICA'S HIGH-VALUE FOODS know their local partners well, dominate demand for tuna, have assured mar- kets for their finished products and are being offered highly advantageous terms for setting up this business. In the present, stable, open business envi- ronment in Ghana, they might be expected to do very well from this enter- prise. The only potential cloud on the horizon is the future supply of tuna, the state of the resource being the subject of considerable debate, though Starkist believes that supply is "nearly in-exhaustable". Credit Access to and the cost of credit are key issues in any discussion of fish produc- tion and marketing. Fishing using motors requires considerable working capi- tal for fuel which must be financed before fishing commences. Demands for fixed capital at irregular intervals are considerable. Fishermen's income is, as stressed earlier, highly unpredictable. In the marketing system, selling pro- cessed fish involves bulk transactions with relatively small margins requir- ing a high degree of liquidity on the part of the middle-person. Credit provision can be divided clearly between formal and informal sys- tems. The former consist of Development Banks and Cooperative Societies and usually entail compliance with certain criteria on the part of the bor- rower such as provision of surety, collateral, evidence of savings and testimo- nials. Informal credit is made up of a wide range of credit arrangements between relatives, ethnic groups, buyers and sellers characterized by their lack of formal fixed terms and conditions and often repaid in kind. The experience of the formal sector in providing credit to fisheries over the decades since Ghana's independence has not been favorable. 28 Recovery levels on loans to artisanal fisherfolk for canoes, nets, motors and working capital for one bank are still only running at 50% of total disbursements. For the inshore fishing sector, recovery rates are even lower and the banks have pulled out, because fishing in this sector is no longer economically viable. To improve recovery rates, banks in Ghana have tended to increase the stringency of requirements for loans. For example, the Agricultural Develop- ment Bank requires applicants for loans to undertake the following: • Provide evidence of their ''background" i.e., be known to the bank; • Have their existing assets valued by bank valuation officers to offset against potential default; 28 Neither has its experience with agriculture-see Obben (1991). PRIVATE SECTOR FISH PROCESSING AND MARKETING IN GHANA 413 • Have savings deposited with the bank which are greater than the value of the loan; • Have documentary evidence of ownership of property for collateral; In lieu of this last requirement, the bank will, under exceptional circum- stances, accept a personal guarantor as surety for a loan. Effectivelr however, these requirements exclude the majority of artisanal and industrial fisherfolk from bank loans at discounted rates. In the past banks have lent money to middlewomen for marketing and processing purposes. Usually this involves lending money at the beginning of the herring season to be paid back over the following eight months. However, very few of these loans are still undertaken. In one case all the borrowers in this sector defaulted, it is believed because they on-lent the loan to their husbands for fishing activities instead of investing in trading. The issue of fungibility of loans has yet to be resolved. Even when formal sector companies need credit, the smooth functioning of the credit market can be obstructed by traditional values and expectations. Large capital assets, be they companies, land or fishing vessels are often felt to be owned "in trust" for family members of the next generations.29 While the kinship system allows companies access to loans, often interest free, from within the family group, it can limit a company's ability to raise outside equity to fund capital expenditure. For example, banks now request borrow- ers seeking loans for family concerns to provide collateral in the form of real estate. In interviews boat owners revealed that they either had no proof of ownership of their family land or were not prepared for the social stigma associated with "mortgaging'' their children's birthright. It also means that companies with excessive liabilities and debt may be unwilling to file for bankruptcy and recommence more profitable working free from debt. The deep-sea fishing fleet in Ghana is a case in point. One company invested heavily in trawlers from Europe and then found that the economics of trawling had changed, making it uneconomic to put to sea. De- valuation of the Cedi during the past decade has massively increased the Cedi cost of financing the debt of the trawlers which are now unseaworthy. The net effect is ~hat the company is unable to raise capital from formal 29 Companies are registered either under Sole Proprietorship laws, or as 'exempt companies' regis- tered by a restricted number of extended family members. In this way, traditional family businesses can enjoy limited liability protection. There are few public companies in Ghana, and none in fisheries. 414 MARKETING AFRICA'S HIGH-VALUE FOODS institutions, or to spread its equity base to finance potentially profitable new areas of businesses because creditors are fearful of their liability position. Owners are reluctant to allow businesses to go bankrupt and the absence of an enforceable insolvency law compounds this situation. All the Ghanaian banks have been involved in the donor-funded program of concessional loans for outboard motors. Default rates on these loans were high and collateral requirements minimal. Borrowers experienced further problems with spare parts especially when the Cedi was not freely exchange- able for other currencies. A second tranche of donor money made available in 1987/88 for engines was handled more circumspectly but the collapse of the trigger fish catch effectively made repayment of these loans impossible. One new area in which most banks have become involved is the financing of construction of cold stores. These are mostly used for the storage of im- ported frozen herring and sardines for sale during the low season. Informal Credit Provision The loaning of money (or deferment of payment of accounts) is an important part of informal sector fisheries and fish marketing. Of particular interest are the credit relationships between individual members of fishing families. These have been discussed above, in terms of the relations between artisanal fisher- men and fish mammies, but are also important in industrial fisheries. Anecdotal evidence suggests that, with the increasing costs of both artisanal and industrial fishing, the levels of indebtedness within families and between fisherfolk and middlewomen is increasing. Traditionally, there has always been a degree of debt, but of late, with the devaluation of the Cedi and re- duced resource base, this system has been placed under further stress. When fish prices fall sharply, this can occasionally result in a loss on the part of the fisherwomen causing considerable family tension. Little is known about the impact of these changes on fishing societies and particularly on income distri- bution within family groups. There exist a number of different informal credit and savings arrange- ments between groups of individuals. An example is the susu or revolving credit and savings union which accepts regular savings from members who each in turn receive a lump sum equivalent to their total expected savings over the period. 30 There exist a number of informal fish mammies associa- 3 ° For a comprehensive study of informal credit systems with many references to West Africa see Adams and Fitchett (1992). PRIVATE SECTOR FISH PROCESSING AND MARKETING IN GHANA 415 tions which collect savings from members, provide health insurance and give loans. One such association in Kumasi (the Kumasi-Mopti Cooperative Fish Marketing Society) has nearly 2,000 members who are shareholders. Credit is provided by some societies. Others will stand as guarantors for loans from the formal sector. CONCLUSION Because fisheries are an extreme example of an open-access resource, there will always be an important role for government in regulating exploitation. Ghana has seen heavy public-sector involvement in fisheries, through the establishment of two large parastatals, but it is debatable whether this core function has been performed effectively in the past, still more so after the withdrawal of government from fisheries production and processing. There is therefore a threat to the resource base behind the rest of the story presented here, the differential effects of Structural Adjustment policies on artisanal and industrial fisheries. In Ghanaian fisheries family and kinship units are involved in various activities from production through processing to marketing. The ability of these units to benefit from intra-family vertical linkages in both the artisanal and industrial sectors is demonstrated by their capacity to survive the severe economic cleavages of the 1980s and early 1990s. However, within these units there has been a continued but unmeasured transfer of debt, mainly from 'male household members to female household members. Direct results of Structural· Adjustment policies, such as high input costs, constrained con- sumer demand and competition from imports, combined with increasing com- petition for a finite resource, seem to have changed the economics of fishing in Ghana. Generally this has been in the favour of the artisanal sector. Traditional, informal linkages in processing and marketing fish products have proved more robust than large-scale industrial integration (in both pri- vate and public enterprises), though the objectives of the two sectors differ. The ability of the artisanal sector to respond to the inherent uncertainties of fish production (i.e., risk, product variability, supply fluctuation, perishability etc.), contrasts markedly with the managerial and institutional problems that the large public sector enterprises, and even larger private sector enterprises, have had. .. ·. ~ 416 MARKETING AFRICA'S HIGH-VALUE FOODS METHODOLOGICAL NOTE AND ACKNOWLEDGEMENTS The research for this chapter consisted of two elements: a study of existing literature and data sets on Ghana fisheries, and a field trip to investigate the fishery first hand. Major sources of secondary data have been indicated throughout, but of particular use for the section on industrial fisheries were a series of sector reports completed during the last five years for various international donors considering rehabilitation of the industrial sector (UNIDO 1991; World Bank 1989; FAO 1989; DANIDA 1988 and CTA 1987). The field work was divided roughly equally between key informants in the production, marketing and processing of artisanal and industrial fisheries. For individual fisherfolk, middlewomen and processors, a purposive sampling system was necessitated by the brevity of the period of research. Interviews were, in the main, semi-structured, to cover the key issues while allowing additional insights to be pursued. Where possible, key informants were sought, such as chief fishermen, "Queen" fish mammies and heads of cooperative trading and processing bodies. Key informants in the industrial private sec- tor and managers of fisheries parastatals were also interviewed about their businesses. The bias inherent in this approach is recognised by the authors. Other potential sources of error/bias are the lack of a seasonal perspective on the fishery and the necessity to seek informants in accessible areas of Ghana. A number of key informants in Government institutions and agencies as well as donor agencies and universities were also interviewed in order to. gain a wider perspective on the framework within which the fisheries sector operates. The authors wish to express their gratitude to the following individuals and institutions without whom this report could not have been completed. John Manful and Lawrence Abbey of the Food Research Institute, Accra. Florence Lamptey, Department of Fisheries, Accra. Andrew Wilson, British High Commission, Accra. Ernest Tettey, Infopeche, Abidjan. We would like to thank all the informants, both in the private and public sectors. In July 1993 US$1 = 640 Cedis IX. Merchants and Middlemen in the Cattle Trade of Southern Somalia1 INTRODUCTION T his case study of cattle ma,rketing in Southern Somalia focuses on the unofficial cross-border trade with Kenya and demonstrates the inter-re- latedness of domestic trading and official and unofficial exports. The analysis reveals that the orientation of export trade is highly contingent on economic and political factors that have led at times to reversals in trade flows. The relationships between different types oftraders and their diversification strat- egies are examined in detail. Conditions of market uncertainty, macroeconomic decline and political in- stability have reached extreme proportions in Somalia during the last few years. Those traders who have become agents ·of large, export-oriented mer- chants focusing on a single market have suffered the most, whereas traders who are based in small villages and involved in both domestic and export markets have sometimes prospered. The unofficial trade in livestock to neigh- boring countries has allowed certain groups of Somali traders to withstand an environment of extreme economic and political volatility that is excep- tional even in the African context. The chapter begins by discussing the volume and general characteristics of trade in livestock from the arid and semi-arid countries of Northeast Africa and the Sahel. After discussing the historic precedents of market crisis and instability in southern Somalia, it goes on to suggest that: (1) dyadic and mul- tiple relationships among livestock traders, often based on some form of ma- terial exchange, are the norm during periods of both economic decline and 1 This is an edited version of "Traders, Brokers and Market 'Crisis' in Southern Somalia" previ- ously published in Africa 62 (1), 1989. John Morton edited the paper for re-publication and contrib- uted additional material on the dimensions and general characteristics of the livestock trade in Sahelian and Northeast Africa. 417 418 MARKETING AFRICA'S HIGH-VALUE FOODS relative prosperity; and (2) economic differentiation among livestock merchants makes it difficult to generalize about trader behavior without paying atten- tion to differences in scale, access to markets and position in the market 1 chain. Market channels are distinguished spatially and sociologically in order to demonstrate that certain groups of traders are associated with particular patterns of trade and geographic location. The chapter concludes with a dis- cussion of the importance of social relations in marketing and the responses of traders to macroeconomic changes. The countries of the Sahel and the Horn of Africa all export large numbers of livestock on the hoof either to neighboring African countries or to the Middle East. In each country there are large arid and semi-arid areas where live- stock production by traditional pastoralist methods or agropastoralism is the most rational land-use strategy. Contrary to some stereotypes, most pastoral societies have customarily and systematically sold live animals and animal products to agricultural societies near or distant.2 They have engaged in ani- mal trade because there existed an external market for animals as meat, as a means of transport or for social reasons,3 because the terms of trade fre- quently allowed pastoralists to obtain more calories than they could through the direct consumption of animal products, and because male animals could be sold without endangering herd reproduction. Prior to the advent of modern freezing and chilling technologies, there were few alternatives to taking live animals to market.4 Before motor trans- port, they were necessarily trekked overland. Because the international mar- kets for Sahelian and Northeast African livestock are relatively close,5 live animals continue to be exported. Now, they can be taken by truck or train or trekked in the traditional manner according to cost considerations that vary by time and place. The continued use of trekking through sparsely inhabited areas allows many livestock exports to bypass frontier controls and avoid associated taxes and duties. 6 In many cases, the trade existed long before the frontier. This element must be factored into official export figures for live- stock from the countries. 2 Kerven (1992). 3 This refers principally to the use of animals as bridewealth. 4 A traditional meat-drying technique appears to have allowed a longstanding trade in meat from Northern to Southern Nigeria (Silverside 1992). 5 By comparison with Southern Africa, where end markets for exports are in Europe. 6 This is not to say that unofficial exports of livestock are not in some cases effected by truck or by boat; livestock are shipped unofficially by dhow to Saudi Arabia from both Sudan and Somalia (Morton 1989; Jaffee and Weli 1985). MERCHANTS AND MIDDLEMEN IN THE CATTLE TRADE OF SOUTHERN SOMALIA 419 While governments are involved in export of livestock, particularly by sea, the trade has remained in private hands throughout the Sahel and Northeast Africa. Several studies7 suggest that trader profits are not excessive, and that the apparent involvement of large numbers of intermediaries in many mar- keting systems is in fact an efficient response to the difficulties inherent in the nature of livestock as a commodity. CHARACTERISTICS OF LIVESTOCK AS A COMMODITY Animals on the hoof, particularly those raised by pastoralists or agropas- toralists, have distinctive characteristics that affect marketing patterns. 1) Livestock are both a commodity and a form of capital. There is no clear division in pastoralist economies between 'capital', 'inventory' and 'surplus'. 8 While under optimal conditions, pastoralists will sell only surplus males, in times of stress even reproductive females can be sold, eventually leading to the pastoralists' destitution. Traders can continue to be active, even or especially when the productive sys- tem is in crisis. 2) High 'perishability': Livestock on the hoof are much less perishable than meat, but need considerably greater care than many harvested crops. Livestock must continue to be fed and watered, and in some cases, protected against disease. Care involves trade-offs between transport costs, labor costs, speed of marketing and animal condition. For instance, before the advent of large-scale trucking of livestock in Nigeria, the necessity of trekking stock quickly from the semi-arid zone through the trypanosomiasis belt to the end markets resulted in significant weight loss (and therefore lower prices) but was judged preferable to mortality en route. 3) Scattered supply: Livestock are necessarily produced on natural range- lands at low population densities. AB a result, time and transport costs in collecting together marketable herds of livestock are high. These may fall on the producers or the first traders in the marketing chain. 7 Usefully reviewed in Sandford (1983). 8 It is partly because of this that pastoralists are sometimes alleged to display a 'perverse supply response', a debate which need not be entered into here. See Sandford (1983). 420 MARKETING AFRICA 's HIGH-VALUE FOODS 4) Individual variability: Livestock vary individually by a number of criteria, all of them relevant to market value. These include weight, age, sex, proportion of fat and presence of parasites or diseases.9 As a result, a great deal of time must be invested in examining animals and negotiating prices. In many markets there are brokers who per- form this function on behalf of larger merchants without necessarily taking ownership of the stock. 5) Multiple uses: Animals have multiple uses to pastoralists: milk, meat, manure, transport, social uses. How they are ranked will vary in dif- ferent situations. Once animals enter the marketing chain, there is also at least a theoretical choice (depending on the infrastructure of the countries in question) between keeping them on the hoof, or hav- ing them slaughtered and marketed as red or frozen meat.10 Market- ing livestock on the hoof involves continuous decision-making, more so than marketing an agricultural surplus (assuming the choice of crop has been made). 6) Livestock production entails relatively low requirements for pur- chased inputs: Even though there is increasing use of purchased fod- ders (such as oilseed cake), medicines and veterinary services, pastor- alists in Africa still purchase very little of what could be considered a production input. Traders are therefore less likely to become providers of credit than are crop traders, although they may provide credit to pastoralists for grain and other necessities for household maintenance. 7) Skill requirements: Considerable knowledge is necessary to keep stock healthy, as well as docile and tractable, during transit to market. Such knowledge is acquired through experience, which largely accounts for the high degree of ethnic specialization in long-distance livestock marketing. 8) Varied investment requirements: Market systems are typically made up of many levels with widely differing entry requirements in terms of capital. Because of the demand for skilled animal handlers and brokers, there is some scope for people to work in the trade with no 9 Where stock-raiding or theft is a problem, the legitimacy of the seller's title may also have to be verified. 10 There is some specialized export of animals that are not to be slaughtered at all, for instance the trade in riding camels from Sudan to the Gulf. There is little information on this, and it is unlikely to be significant compared with the trade in animals destined for slaughter. MERCHANTS AND MIDDLEMEN IN THE CATTLE TRADE OF SOUTHERN SOMALIA 421 capital at all. A few may later graduate to become small traders in their own right. Especially if motor transport is involved, however, there are likely to be levels of trade with heavy investment require- ments, namely, the capital (or at least the credit) to purchase one truckload of animals and prepay for truck hire. Such an investment is less than that required for industrial processing equipment in many other sectors, but more than that required for artisanal processing or simple trade in those sectors. It is also likely to be permanently be- yond the reach of ordinary pastoralists, even if they enter the trade at lower levels. 9) N~n-financial entry requirements: Traders who export officially re- quire ability to deal with bureaucratic procedures. Even where live- stock leaves one country unofficially, there may be bureaucratic pro- cedures necessary at the end market. For example, in Nigeria a requirement for veterinary clearance certificates is successfully en- forced, even on livestock unofficially exported from Niger.11 Bureau- cratic procedures are likely to be more cumbersome where livestock ·are shipped rather than trucked or trekked. 10) Seasonality in trade: Marketing oflivestock by pastoralists may show some seasonality due to both supply and demand factors. On the sup- ply side, animals are in their best condition at the end of the rainy season. Pastoral migrations and road conditions may make it is easier to take them to market at another time of year. Demand may exhibit seasonality with maximum consumption occurring at religious festi- vals and other holidays. 12 Yet, seasonal peaks in sales are unlikely to be as sharp as they are for many annual crops, and under certain conditions sales can take place at any time of year. 11) Possibility of ''back-haul trade": Pastoralist producers living in zones poor in other productive possibilities are likely to present permanent markets for a variety of commodities: grain and other basic foods, sugar, beverages and stimulants, cloth. In many, but by no means all, situations these commodities are imported from the countries to which livestock are exported by the same merchants using the same means of transport. 11Kulibaba (1991). 12 This is complicated by the fact that Muslim festivals move forward against the Western calendar every year. 422 MARKETING AFRICA'S HIGH-VALUE FOODS EXPORT MARKETING, COMPETITIVE STRUCTURE AND GOVERNMENT INTERVENTION The export systems under consideration divide generally into overland trade, whether official or unofficial, and trade by sea. In overland trade, govern- ments are not heavily involved, except by taxing traders and ensuring health standards. 13 Overland trade is likely to be relatively open to competition, even at the highest levels, and one is likely to find successful enterprises in both exporting and importing countries. Even where the trade is completely official, enterprises will be characterized by a degree of informality with inte- gration of the marketing chain achieved by kinship, ethnic or client ties- 'agreements and sanctions internal to the community'14-rather than by for- mal contract or partnership. In the case of trade by sea, governments have become more involved, and there is greater tendency for a small number of firms to dominate export marketing. In Ethiopia the parastatal ELIMCOR has accounted for over 70% of all sheep exports in recent years. 15 In Somalia, official exports of cat- tle to Yemen and Egypt during the mid-to-late 1980s were dominated by three companies that were set up under government supervision. Trade with Egypt took place under a specially negotiated agreement between the two governments. The reason for the differences is fairly obvious. Shipping involves more bureaucratic procedures-letters of credit, bills of lading-than even official overland export is likely to. These are unavoidable costs of using interna- tional shipping and represent potentially large economies of scale. In addi- tion, when pre-existing links to end markets are not established, there will be high costs involved in searching for a purchaser and enforcing contracts. Fi- nally, overland trade usually involves countries sharing the same currency (for example, the CFA Franc), or countries for which there is an easily acces- sible parallel currency market. This is less likely to be the case where coun- tries do not have a land border. 13 There were attempts to set up parastatal livestock companies with both production and market- ing functions in the Sahel in the 1970s, but these failed or have been dissolved in the course of Structural Adjustment programs. See Kulibaba and Holztmann (1990). 14 Kulibaba and Holtzmann (1990). 15 FAQ/World Bank (1993). ELIMCOR dominates official exports for all species, but only for sheep is official export more important than unofficial. MERCHANTS AND MIDDLEMEN IN THE CATTLE TRADE OF SOUTHERN SOMALIA 423 Export Volumes and Trends The FAO now attempts to include unofficial livestock exports in its annual trade statistics (Table 9.1). These figures, together with data from other sources on Ethiopia and Somalia, give an indication of the volume and importance of livestock exports from the Sahelian and Northeast African countries. The table makes no attempt to quantify marketing of livestock within the coun- tries concerned; yet, there are major flows of stock to urban centers for con- sumption in all the countries listed. Both Cote d'Ivoire and Nigeria, men- TABLE 9.1: Live Animal Exports from Sahelian and Northeast African Countries Exports as% of National 1 Country Exports ('000 head) Destination Export Earnings Mauritania 70 cattle Senegal, Middle East Na. 450 small stock Mali 190 cattle Cote d'Ivoire and other 26 4 70 small stock coastal countries Burkina Faso 81 cattle Cote d'Ivoire and other 6 92 small stock coastal countries Niger 100 cattle Nigeria Na. 170 small stock Chad 40 cattle Cameroon and elsewhere 19 160 small stock Sudan 600 small stock Saudi Arabia. No official 18 14 cattle export of cattle or goats since 1986. Substantial unofficial exports of sheep to Saudi Arabia and camels to Egypt and Libya Ethiopia1 176 cattle Official cattle exports only to ,3 359 sheep Yemen, small stock to rest of 266 goats Middle East. Unofficial exports to Kenya and Somalia Somalia2 68 cattle Official cattle exports to Egypt 41 734 small stock and Yemen, small-stock to Saudi Arabia. Unofficial cattle exports to Kenya 1 Ethiopia: 1987/88 figures including estimated unofficial exports; 85% of cattle, 28% of sheep, 75% of goats. Export earnings figure based on official exports only. 2 Somalia: 1987 official figure for cattle, plus estimated 40,000 unofficial exports. 1984 official figures for small stock. SOURCES: Export figures from FAO Trade Yearbook 1991, except Ethiopia (FAO/World Bank 1993), Somalia (Little 1989; Jaffee and Weli 1985). Export earnings figures all from FAO Trade Yearbook 1991. 424 MARKETING AFRICA'S HIGH-VALUE FOODS tioned above as importers, both also produce large quantities of livestock in their northern regions which are trucked south to the cities. Domestic and import/export marketing systems may be organized by the same merchants or may compete for livestock supply and markets. The table excludes meat exports, which are generally small yet growing in some countries. Livestock exports from the countries under consideration have proved vul- nerable both to demand factors-non-economic barriers and international com- petition in the importing markets-and supply factors. There is a legitimate fear that imports of livestock on the hoof may infect the national herd, but policies based on such fears have sometimes been taken to extremes by Saudi Arabia and the other Gulf states. A Saudi ban on Somali cattle to prevent the spread of rinderpest has been in force since 1984, despite the lack of evidence of rinderpest in Somalia. Livestock from the countries under consideration now has to compete in export markets with stock or meat from developed countries. Despite high transport costs, ranched livestock from Australia is taking over markets in the Gulf states because a reliable supply which meets health requirements is more ~ssured. The European Community exports subsidized meat to West Africa and Egypt, which has serious effects on inter-African trade. Supply of livestock has been affected by the drought and environmental degradation that has taken place in the arid and semi-arid zones in the last two decades. It is hard to find reliable data or to distinguish short-term from long-term trends within it. Livestock numbers in the Sahelian and North- east African countries all declined in aggregate over the 1980s, but much of this was due to a sharp fall in 1984--85. At that point, exports from Mali, Burkina Faso and Sudan (three countries for which data is at hand) peaked sharply, probably because distress sales by pastoralists were converted into exports. While one might expect a long-term decline in exports, the relation- ship between desertification (itself a controversial concept), herd numbers and livestock marketing remains extremely complex and does not simply spi- ral downward. Another supply problem is that of political instability in exporting coun- tries. Somalia is the worst case here. The present case study will show the resilience of trading networks up to around 1990, but there has probably already come a point where trading networks have been broken up. On a smaller scale, political instability may affect livestock production and export in other countries of the region. MERCHANTS AND MIDDLEMEN IN THE CATTLE TRADE OF SOUTHERN SOMALIA 425 THE LOWER JUBBA REGION OF SOMALIA The Lower Jubba Region16 in the far south of Somalia consists of a semi-arid plain, in which most cultivation is carried out in the immediate vicinity of the Jubba river and in scattered settlements along the coastal plain to the south of Kismayo town. Approximately 25% of this population is classified as pasto- ral nomadic, 30% as urban, and 45% as settled agropastoralist or agricultur- alist.17 Average rainfall is 560 mm, much lower in the west, and falls mainly between April and June, with uncertain 'short rains' between October and December. Compared to other regions of Somalia, the lower Jubba is a rela- tively favorable rainfall area. 18 The seasonal floodplain of the Jubba River and the Lag Dera basin are important dry-season grazing resources. The former support the highest concentration of cattle in any region of Somalia, which is among the highest in eastern Africa.19 The Lag Dera basin was developed as a grazing resource by the World Bank which financed the Trans- Jubba Livestock Project from the mid-1970s. Larger export-oriented herders and traders were among the chief beneficiaries, and the project was discon- tinued in 1986 amid local conflicts and environmental controversy. The other important geographical feature of the region is its 150 kilometer coastline and Kismayo port, in particular, which is the third largest port in Somalia and serves as a major livestock export facility. The port provides access to lucrative international livestock markers for nomadic herders and local traders. Their counterparts elsewhere in the pastoral regions of sub- Saharan Africa are not similarly endowed. The characteristics discussed above help explain the importance of the livestock sector (particularly cattle) to the regional economy. In most years Somalia derives more than 80% of its export earnings from the livestock sector. Largely because of its excellent water and range resources, the Lower Jubba Region has the largest population of cattle in the country (appro:Xi- mately 860,000) and also a sizable camel population (estimated at 222,000).20 These represent about 22% and 4%, respectively, of Somalia's cattle and camel 16 The Region is subdivided into four administrative districts. Fieldwork concentrated on two: Kismayo and Afmadow. 17 Evans et al. (1988). 18 Hubl (1986), Conze and Labahn (1986). 19 Murray Watson, personal communication. 20 Hubl (1986). 426 MARKETING AFRICA'S HIGH-VALUE FOODS populations. In the aggregate, livestock numbers seemed to stabilize in the late 1980s after a rapid expansion of herds, especially of cattle. It is estimated that from 1952 to 1983/84, regional cattle and camel herds grew annually at rates of 7.9 and 5.2%, respectively. At these rates, herds of cattle would have doubled approximately every nine years, and camel herds every thirteen years. 21 Growth in regional herds during this period was related to: (1) growth in human populations due in part to in-migration of herders (and herds); (2) development of export markets for cattle; (3) establishment of water points in underutilized parts of the region, which allowed cattle to use these areas regularly; and (4) improvement in veterinary health services and the eradica- tion of such animal diseases as rinderpest. Families of ~he region now have cattle herds that are well above the country's average and among the highest in East Africa.22 Ownership of both camels and cattle in the region is highly skewed. Among a sample of forty- two households owning camels, the richest 12.5% control an estimated 52% of these animals. While ownership of cattle in the lower Jubba is not as skewed as this, considerable differences do exist. The richest 12.5% of households own 39% of total cattle, while the bottom 50% of cattle herders control only 15% of the herd. Such inequalities in livestock ownership are not unusual for pastoral economies. 23 In most years the Lower Jubba Region, because of Kismayo port, accounts for 35 to 40% of the country's total overseas exports of cattle and about 20% of its camel exports. The unofficial export of cattle to neighboring countries, especially to Kenya, has been even more significant. During the 1980s, the Lower Jubba Region was exporting about 30,000 cattle a year to Kenya on the unofficial market, or approximately 60% of Somalia's total exports to that country.24 During 1987 and 1988, unofficial sales to Kenya exceeded official exports from Kismayo port by a factor of six. Because of the recent civil war 21 Hendy (1985). · 22 See Schneider (1979) for general East African comparisons, and Ensminger (1984) and Grandin (1988) for particularly cattle-rich groups. Lower Jubba herds are also well above averages recorded in West Africa; see Sutter (1987) and Swift (1986). 23 Sutter (1987); Little (1985). 24 It is obviously very difficult to estimate the annual volume of unofficial trade to Kenya. Cassam (1987) conducted research in both Kenya and Somalia to derive his estimate of about 50,000. Using this figure, with the estimated herd sizes of the Lower, Middle and Upper Jubba Regions and the author's market data, it can be estimated that some 30,000 cattle, or 60% of the total, originate from the Lower Jubba Region. MERCHANTS AND MIDDLEMEN IN THE CATTLE TRADE OF SOUTHERN SOMALIA 427 TABLE 9.2: Average Herd Size and Composition Among Herder Households Livestock Kismayo District · Afmadow District All Cattle 42.98 74.74 58.14 Camels 22.37 1.73 12.52 Sheep/goats 6.54 8.26 7.36 SOURCE: Author's survey of forty-six households in Kismayo District and forty-two households in Afmadow District, 1987-88. in the country, the differential has probably increased since official exports of livestock have virtually halted since 1990. Livestock production systems in the region exhibit substantial variability that relates to both environmental and social variables. Major differences in livestock production systems occur between Kismayo and Afmadow Dis- tricts, for example, the most important being the significance of camels in Kismayo as opposed to cattle in Afmadow (see Table 9.2). The average num- ber of camels controlled by herders in Kismayo District is more than ten times that of Afmadow herders who, in turn, own considerably larger cattle herds than do Kismayo households. Neither district holds large numbers of goats and sheep, a finding that further distinguishes the Lower Jubba Region from most other areas of Somalia where small stock-especially sheep-are so important. 25 Environmental parameters partially explain interdistrict variation in the ownership of cattle and camels: Afmadow, particularly its southern half, con- tains larger expanses of perennial pastures suitable for bovine production than does Kismayo, which has a browse/shrub environment suitable for camel pastoralism. These production differences also correlate loosely with the clan structure of the region. The area's two most important subclans, the Maxamed Zubeer of the Ogadeen clan and the Herti, are found in Afmadow and Kismayo Districts, respectively. There are significant differences in their production systems. The Maxamed Zubeer, who inhabit most of Afmadow District and a territory extending more than 100 kilometers inside the Kenya border, are closely associated with cattle pastoralism. The Herti, residing mainly in Kismayo District, focus considerably more attention on camel pastoralism. In addition, many Herti are actively engaged in commerce (including livestock trade) and control many of the businesses in Kismayo town. The Maxamed 25 See Samatar (1989). 428 MARKETING AFRICA'$ HIGH-VALUE FOODS Zubeer are more strictly pastoral, and their commercial activities are limited mainly to small retail trade and minor roles in the overseas export trade. While clan is not the only social factor determining regional production relations and systems, it is, along with such variables as class and territory,26 critical to understanding relationships among traders and between traders and herders. The current chaos in Somalia has solidified clan relations, and most of the regional political and military movements are based on particular clans. Both the Maxamed Zubeer and the Herti migrated to the Lower Jubba Region last century. The major influx of Maxamed Zubeer was in the 1870s and 1880s after they and other subclans of the Ogadeen clan were pushed by the expanding Ethiopian Empire and a series of inter-clan conflicts.27 During approximately the same period, Kismayo began to grow as a commercial cen- ter and attracted Herti families from northeastern Somalia who moved there to trade and settle in the hinterland as pastoralists.28 The Herti were experi- enced traders who had been involved with northern Somalia's trade of live- stock and other products to the Arabian peninsula. This commerce predates the export trade from Kismayo. The Herti were able to adapt to a settled lifestyle, and by the latter part of the nineteenth century they had become "the dominant group of petty traders along the coast between the Jubba and Tana Rivers". 29 The influx of the Herti clan brought them into direct confrontation with the Ogadeen over control of the lower Jubba. In the 1880s the Herti "were engaged in a struggle for political supremacy and control of local commerce with elements of the neighboring Maxamed Zubeer clan." The Maxamed Zubeer clan along with other Ogadeen subclans controlled caravan move- ments in and out of Kismayo town. 30 While livestock and livestock products assumed some significance in the caravan commerce, the most important 26 See Samatar (1989). 27 For more detail on the history of different groups in the Jubba area, see Dalleo (1975), Menkhaus (1989), and Turton (1970, 1975). 28 Menkhaus notes that the Herti traders "were accompanied by clansmen who relocated in the region with their herds, in part to offset competition for control of local land and trade from the Muhammad Subeyr sub-clan of the Ogadeen clan" (1989). 29 Cassanelli (1982). The Herti also were to become strong allies of the British colonial state, serv- ing as policemen, clerks, and even "spies" for the administration (see Turton 1970, 1972). This close association further strained their relationship with Ogadeen groups in the region. 3° Cassanelli (1982), see also Dalleo (1975). MERCHANTS AND MIDDLEMEN IN THE CATTLE TRADE OF SOUTHERN SOMALIA 429 commodity was ivory. The export of cattle and other products from Kismayo became more important in the 1880s and 1890s with Herti merchants play- ing a major role. Today, the region's most prosperous export trader traces his roots to this period when an earlier member of his family moved to Kismayo to establish an export business.31 A few of these northern traders eventually moved out from Kismayo into smaller centers like Afmadow in order to establish market alliances with the Ogadeen nomads and to be closer to sources of supply. A small number of Herti merchants married Maxamed Zubeer women. The town-based mer- chants and large middlemen32 often extended credit to herders and small middlemen, developing a network of market alliances in the area that ex- tended into what is today northeastern Kenya.33 When the cattle trade was focused on Kenyan rather than Somali markets, the Ogadeen nomads and middlemen who occupied much of northeastern Kenya, as well as the lower Jubba of southern Somalia, played a greater role in the commerce. Market networks of this type continue today and are crucial to understanding how traders have responded to recent economic and political crises. Historical Context of Market Crisis and Instability The decade of the 1980s was a period of devastating political and economic decline in Somalia that took its toll on the livestock sector. Since the mid- 1980s, the national economy has gone from bad to worse. As of 1989 the value of annual imports was about twice the value of Somalia's annual exports, while more than 70% of the annual operating budget of the government was directly paid out of foreign aid. The market value of the Somali shilling de- clined from 37 SSh = $1.00 in 198334 to about 1500 SSh = $1.00 in 1989. Annual inflation in the late 1980s exceeded 300%. Even before the most recent crisis, Somalia's economy exhibited "most of the major symptoms of African underdevelopment, for example, growing food imports, balance of payment problems, declining agricultural production, malnutrition, and starvation."35 31 Cassanelli (1982). 32 While women traders dominate the milk and other agricultural markets (Herren 1990), virtually none are involved in the cattle or camel trade. 33 See Hjort (1979). 34 Samatar and Samatar (1987). 35 Samatar (1987, p. 356). 430 MARKETING AFRICA'S HIGH-VAq.JE FOODS The current national political situation in Somalia, in which clan-based political factions control different regions, hardly needs further description here. 36 Even prior to the overthrow of the Siad Barre regime in January 1991, the southern-based Somali Patriotic Movement controlled most of the Lower Jubba region with the exception of Kismayo town. Major armed clashes be- tween Barre's forces and the Somali Patriotic Movement, who are mainly drawn from the Ogadeen clan, took place as early as August 1989. The state, weak even during periods of relative stability, had effectively lost sovereignty over most of the lower Jubba by the end of 1989. During the decade of the 1980s the livestock sector-the country's most important economic specialty-declined dramatically after more than fifteen years of rapid growth in livestock exports. During the 1970s the dominance of Saudi Arabia as a market for Somali cattle had grown rapidly, and by the early 1980s it accounted for more than 95% of the external market. In volume alone, the expansion of the Saudi market was a dramatic departure from the past. Cattle exports from Kismayo increased more than threefold from the late 1960s to the late 1970s. The amount of revenue also increased dramati- cally since Saudi importers were willing to pay relatively high prices to guar- antee meat for their burgeoning domestic market. This growth in Saudi de- mand correlated with the general oil boom of the 1970s and early 1980s that drove up Arabian incomes. The period 1983 to 1989 saw a reduction in total animal exports to Saudi Arabia. Fierce competition from Australia and other countries for the lu- crative Saudi Arabian market reduced the volume of sheep exports to that country. The export of Somali cattle was also hurt by other actions that were only marginally market-related. In 1983 Saudi Arabia imposed a ban on cattle imported from Somalia in response to fears of rinderpest in south- ern Somalia. The immediate effects of the ban were catastrophic: annual cattle exports declined from 157,000 in 1982 to less than 8,000 in 1984 (Fig- ure 9.1). While an international team from the FAQ verified early on that the area's cattle are not infected by rinderpest-nor were they in 1983-the ban on cattle exports to Saudi Arabia remains in force. The loss of the Saudi market was particularly disastrous to the economy of the lower Jubba region because of its dependence on cattle exports. In two 36 It is not surprising that current political factions draw much of their support from the clan structure, since the Siad Barre regime reinforced "clanism" by strategically manipulating one group against the other. MERCHANTS AND MIDDLEMEN IN THE CATTLE TRADE OF SOUTHERN SOMALIA 431 years (1982 to 1984), annual cattle exports from Kismayo declined from approx- imately 51,000 to less than 1,000. This hurt certain merchants more than others and resulted in a redirection of the cattle trade to Kenyan markets. New markets in Egypt and Yemen partially compensated for the loss of the Saudi market, but annual exports from Kismayo iri 1985 and 1986 re- mained at only 50% of the 1982 volume. The next few years brought evidence that these new export markets might also prove volatile. From 1985 to 1987 annual cattle exports from Kismayo declined from 26,213 to 4,168 and pros- pects for 1988 looked only marginally better. Although trade to Yemen may continue, Egypt has stopped importing cattle from Somalia, including Kismayo, and is unlikely to resume imports in the near future. The civil unrest in , Somalia has also reduced exports, the major port of Berbera being closed since mid-1988. AB a result, the total value oflivestock exports from Somalia fell from US $51 million in 1987 to US $22.4 million the following year. Dur- ing much of 1989, the last year of reliable data, no official exports of cattle from Somalia took place. The trade situation could only have become worse in 1990 and 1991.37 For the Lower Jubba Region, the Kenya market proved to be the most reliable during this period. While Kenyan cattle had been exported overseas via the Kismayo port during the Saudi boom period, the trade was now re- versed. AB of 1989, Somali cattle were being sold at Kenyan markets and Kenyan cattle were no longer being sold at Kismayo. How unusual is this sort of political and economic instability? Although the current situation is an extreme case, turbulent conditions have character- ized the region for much of the past century. The cattle trade, for example, was disrupted by colonial policies of both the British and Italian govern- ments, policies that included regulation of livestock movements, imposition of quarantines and restrictions on foreign currencies. It was also affected by warfare: in the 1890s and early 1900s between the Maxamed Zubeer and the . British, in the 1930s between the Italians and Ethiopians, in the 1940s be- tween British and Italian troops, and in the 1960s between Somalis and the Kenyan government. Severe droughts have occurred during almost every de- cade of this century. Fighting among local clan factions flares up periodically. Ambiguity and risk have surrounded livestock trade in the region during much of this century. Dalleo examines the structural transformation of the 37 See Woodward and Stockton (1989) for export figures and prospects, and Greenfield (1991) for an update after the end of the Barre regime. 432 MARKETING AFRICA'S HIGH-VALUE FOODS livestock trade in the 1920s due to of the imposition of a 20-year long ban on exports of Somali cattle to zones of Kenya settled by Europeans. The policies of the fascist Italian state also added to the uncertainties surrounding re- gional livestock trade: In July of 1936, Italian officials at Bardera [in the Jubba valley] gave a five day quit notice to the Somali who had been waiting there to sell their live~ stock. The Italian action put the Somali at a disadvantage because the water pools necessary for the safe return of their livestock to Wajir [in Kenya] had already dried out. The Italians purchased the Somali cattle at 'cutthroat prices' .38 Several times during the past century reversals in the flow of cattle be- , tween Kenya and Somalia occurred, depending on political and market condi- tions in the two countries. The shift toward Kenyan markets since the mid- 1980s is only the latest in a series of transitions. During the periods of approximately 1890-1910, 1921-1940, 1945-1950 and 1968-1983, the cattle trade went mainly from northeastern Kenya to southern Somalia, especially to Kismayo. By contrast, during the periods 1910-1920, 1941-45, 1951-1967 and 1984 to the present, movements of cattle went mainly from Somali mar- kets to Kenyan markets in the Northern Frontier Province, and then eventu- ally to.Kenya's commercial ranching areas and towns. These market trends were not absolute, and during any given year cattle flows could be reversed or go in both directions, depending on seasonal and local market conditions. For example, the lack of water points in northeast- ern Kenya restricts movements of cattle to its markets during the dry season, even when prices are more favorable there than in Somalia. At any rate, the risks associated with cattle trade, regardless of their origins, have always required astute merchants to maintain multiple relationships with other trad- ers on both the Kenyan and Somali sides of the border. Overview of Different Markets The rapid expansion in cattle exports from 1970 to 1983 resulted in two important changes. First, it inserted into the region a class of very large export traders from Mogadishu and other areas to the north. These traders 38 Dalleo (1975, pp. 171-172). MERCHANTS AND MIDDLEMEN IN THE CATTLE TRADE OF SOUTHERN SOMALIA 433 FIGURE 9.1: Cattle Exports-for Kismayo Region and National Total Exports from 1980 to 1988 170 160 150 140 130 - A - National Total 120 - o - Kismayo Region 110 VJ "C en 100 +-' c: 90 ~"' 0 VJ c. :::> 80 ~o B 70 60 50 40 30 20 10 0 1980 1981 1982 1983 1984 1985 1986 1987 1988 YEAR Based on Cassam (1987); Woodward (1989). were not associated with the pastoral sector of the Lower Jubba,39 and de- pended on large brokers and middlemen to procure their animals. The growth in cattle exports and the insertion of large outside traders in turn altered local and regional market relations, as local middlemen and brokers began to serve as agents for the large traders. In lower Jubba this resulted in a scale of trading enterprise and specialization not characteristic of earlier periods. The contemporary cattle trade is essentially focused on four types of mar- kets which account for about 90% of the region's sales. First is the regional domestic market, concentrated in regional towns like Kismayo, Jamaame and Jilib. This trade is for local consumption and rebuilding of local herds, and it involves both low-quality, low-priced cattle and young heifers or bulls (less than 3-4 years). On the consumption side, it is centered on local butcheries and accounts for an estimated 40% of cattle sales in the area. This figure includes purchases by local herders to replenish their own herds, but not animals bought for eventual resale in other markets .. 39 Of the more than 100 registered companies in Somalia involved with livestock exports, none are from the Kismayo area. 434 MARKETING AFRICA'S HIGH-VALUE FOODS The second market is the national domestic market located in Mogadishu, Somalia's largest city; it requires the trekking of cattle over a distance of more than 300 kilometers. The Mogadishu market is growing rapidly along with the city's population, and cattle prices there tend to be about 20-25% higher than in the lower Jubba. Herti merchants are strongly involved in this trade and usually have partners in Mogadishu on whom they rely. The Mogadishu market is the final destination for approximately 16% of cattle sold in the region. The third and fourth market channels in the lower Jubba involve interna- tional export trade. In the case of the Kenya market, animals are sold and moved unofficially across the border to Kenyan markets, particularly to Garissa. This trade accounts for about 25% of cattle sales. Because the trade is controversial and not officially sanctioned, its importance has probably been understated by traders. Ogadeen middlemen and traders are heavily involved, and those Herti who participate usually have forged some type of alliance with an Ogadeen middleman. This trade is seasonal: during the dry season (January-March), virtually no cattle move from the lower Jubba to northeastern Kenya. This is the trade that has captured the bulk of cattle exports since the ban on Saudi Arabian imports. The Kenyan trade involves medium- to high-quality male and fe- male animals, which are used for fattening and eventual slaughter in urban centers and for restocking and breeding purposes on commercial ranches in the Rift Valley. The sale of Somali cattle to European-owned ranches in central Kenya began early in this century and frequently took place ille- gally even during periods of market quarantine. At least part of the reason for the recent Kenyan market boom was the 1984 drought, which devastated approximately 50% of Kenya's national cattle herd. The result was a surge in demand for cattle in Kenya, and the country's herders have been rebuild- ing their herds through unofficial imports from Tanzania, Ethiopia and Somalia. 40 The fourth market, the overseas export trade, is very different from the other three channels, and it is the only livestock business in which the largest merchants and companies are involved. In 1988, it accounted for fewer than 8% of the cattle sold, but because it involved high-quality animals, this trade was responsible for about 15% of the aggregate value of marketed cattle. The 4 ° Cassaro (1987). MERCHANTS AND MIDDLEMEN IN THE CATTLE TRADE OF SOUTHERN SOMALIA 435 overseas export trade entails the grafting of a modern contractual system of marketing onto a customary pattern of trade. The export traders who are typically registered members of a company (see below) have contracts with a private importer, usually from a Middle Eastern country, to supply a speci- fied number of high-quality animals at an agreed price. The contract indi- cates the party responsible for shipping costs and the vaccination and quar- antine requirements. When the Somali trader has written proof of an order, he can request a letter of credit from the state-owned Somali Commercial and Savings Bank and receive half of its value in local currency. Until the mid-1980s the state's role in this trade was minimal, except for providing financial, shipping and veterinary services. Since then, however, the government has assumed a more active posture. It has implemented leg- islation requiring individual exporters to cooperate and form companies and has negotiated government-to-government contracts on behalf of the com- mercial sector. Sales of cattle to Egypt, which were very important for Kismayo during 1985 to 1988, resulted from a negotiated contract between the govern- ments of Somalia and Egypt which the private sector has implemented. The Somali government took these actions because Somalia was losing market share in its traditional export markets (especially in the Middle East), and competition by export traders for cattle in Somalia was boosting local prices too high, thus allegedly reducing traders' profits and competitiveness at the international level.41 The formation of compallies was a way to reduce competition by limiting the local arena to a smaller number of actors-in this case, companies. How- ever, competition has not been reduced since most of the companies are domi- nated by two or three large traders who operate essentially as individual merchants rather than as members of a collective. They procure their own contracts and register them in the name of the company, but keep the pro- ceeds from sales for themselves. After the export crisis of the 1980s, coopera- tion among export traders declined even further, as larger merchants sought to attain a bigger slice of a diminishing trade by forcing out smaller competi- tors and consolidating their.own market advantage. Some of the individual contracts in the 1970s were for as many as 10,000 cattle over a period of 18-24 months. One of the largest traders in Kismayo town exported 47,000 cattle and 18,000 camels to Saudi Arabia between 1971 41 Samatar (1987). 436 MARKETING AFRICA'S HIGH-VALUE FOODS and 1983. Other traders had contracts in excess of 2,000 cattle that had to be met within five or six months. Because formal credit is usually only extended for a period of two to three months, the exporter must procure the animals within that brief period. When the Saudi trade was in place, many Somali exporters received private financing from their Saudi Arabian contacts. With- out it, they must depend on local informal credit or government sources and they count on herders accepting payment only after the final sale is made which amounts to obtaining a form of credit from them.42 The need to pur- chase large numbers of cattle in a relatively short period of time further dis- tinguishes the overseas export trade from other markets. The local system of procurement for overseas trade is essentially the same as that used for other markets with a few important distinctions. First, over- seas traders usually do not purchase animals at local market centers but rely instead on their own chain of middlemen and agents for supplies. Second, export traders often employ some full- or nearly full-time people whose func- tion it is to procure animals. Third, in contrast to other markets, the overseas commerce is restricted to young male cattle aged four to seven years. Female cattle may not be exported, and only the largest males, those that exceed 275-300 kg. live weight, are purchased. Finally, because of the sheer size of their orders, exporters are forced to procure animals throughout the year and to employ hired herders to manage the animals. Exporters usually keep only their head herders employed throughout the year. Other herders are hired as needed. To manage a herd of 100 cattle, one head herder and three tempo- rary employees are required. The benefits and costs of these different market channels vary consider- ably. While there are increased risks with the overseas trade, the largest net returns for traders take place in the overseas export trade, followed closely by the Kenya trade (see Table 9.3). Time-series data are not available on trader profitability in the overseas trade, but profits are likely to have declined since the imposition of the Saudi ban. While the FOB price of US$950 per live weight ton for cattle changed very little between 1984 and 1988, costs for veterinary and other inputs for traders increased considerably during this period. That cattle exporters still received a relatively favorable return in 1988 strongly implies that profits were likely to have been very high before 1983. 42 See Samatar et al. (1988). MERCHANTS AND MIDDLEMEN IN THE CATTLE TRADE OF SOUTHERN SOMALIA 437 TABLE 9.3: Cattle Trader Margins and Net Returns (Somali Shillings per Head of Cattle) Kismayo Overseas Kenya Mogadishu Town Trade Trade Trade Trade Traders Final Price 38000 30000 17500 10592 Purchase Price (23000) (21000) (13029) (8820) Trader Margin 15000 9000 4471 1772 Trader Costs (8135) (3758) (2298) (905) Net Return 6865 5242 2173 867 As Percent of Final Price 18.07 17.47 12.42 8.19 Middlemen• Selling Price 23000 21000 13029 data Purchase Price (19000) (16000) (10620) not Middleman Margin 4000 5000 2409 available Middleman Costs (2225) (1985) (1324) Net Return 1775 3015 1085 As Percent of Selling Price 7.72 14.36 8.33 • This reflects a case where the middleman buys directly from a herder and then sells to a larger trader. In many cases, however, the middleman may be buying from smaller middlemen ("bush traders") who in turn purchased the animal from the producer. The table has-been greatly simplified and does not take into account the full range of market transactions that occur. Average exchange rate during 1987 to 1988 (until February 1988) is calculated at 130 SSh=US $1.00. SOURCES: Author's data, and Stockton (1987) for overseas trade. The lowest profits among traders are found in the regional/Kismayo town trade, which is based predominantly on sales of low-value animals. Since the costs of engaging in this trade are considerably below those of other markets, it is open to most small-scale traders. To a lesser extent, the same can be said for the Kenya and Mogadishu markets (see Table 9.4). Not surprisingly, the highest trader costs are associated with the overseas export commerce, where in 1988 the operation costs per animal were 31,135 Somali shillings (includ- ing the purchase and other costs). The scale of investment per head of cattle, even though the exporter often has credit, precludes most traders other than the wealthiest from participating at the upper levels of this market. In com- parison to other trade channels, the overseas trade in cattle is particularly distinctive in its volume and types of transaction costs. The net returns of middlemen, on the other hand, vary according to the different markets, but they tend to be highest for the Kenya trade. Middle- 438 MARKETING AFRICA'S HIGH-VALUE FOODS men acting between a Kenya-based trader and the producer can earn net returns equivalent to 10% of the sale price in Kenya, or about 3,015 Somali shillings per animal. By 1988, the net returns per head of cattle for agents involved in the overseas trade were about 40% below those of middlemen in the Kenya trade, and were equivalent to about 4.5% of the FOB price at Kismayo. Because the Kenya trade is so lucrative for local middlemen and traders, they are directing most of their efforts to it rather than to other markets. This pattern, coupled with the general decline in overseas markets, has made the official export trade from Kismayo increasingly problematic. Trade in other animal species-such as camels, goats, and sheep-reveals neither the complexity nor the volume that cattle marketing does. The mar- keted volume of cattle exceeds regional sales of small stock and camels by facfors of three and fifteen, respectively.43 An examination of the figures in Table 9.1 shows why this might be so: average cattle herds in the region are considerably higher, especially in Afinadow District, than those of other live- stock types. Despite this discrepancy, some camels and small stock are ex- ported from Kismayo to the Middle East, though it is likely that many of the camels for export originate from the Middle and Upper Jubba, not the Lower Jubba. Neither animal type is exported from the region to Kenya. Camel markets are poorly developed in Kenya and relative demand for the animal there is low. Exports of sheep and goats to Kenya are inhibited by the ad- equate supplies of the animals in the country. Regional trade in small stock is almost strictly oriented to local markets, especially those in the larger towns. With the collapse of the overseas cattle trade, merchants who were involved at least partially in the trade of small stock or camels were better prepared to confront the market crisis of the 1980s. Traders Several types of actors are involved in livestock trade, distinguishable by market, scale and location in the market chain. Local terms are used to de- fine some of the roles in the different markets, but many of these terms have been redefined as market conditions have changed. Cattle merchants, in par- ticular, can be classified into five general types although distinctions among certain types may be blurred at times. Recognizing social and economic differences among traders is important 43 Little (1989). MERCHANTS AND MIDDLEMEN IN THE CATTLE TRADE OF SOUTHERN SOMALIA 439 TABLE 9.4: Estimated Costs Incurred by Cattle Traders (Somali Shillings per Head of Cattle) Overseas Kenya Mogadishu Kismayo Expenditure Item Trade Trade Trade Town Trade Traders• Water 180 90 45 90 Local Transport 200 300 663 Vet Costs 100 34 30 60 Hired Labor 450 271 35 200 Fodder 150 200 100 25 Feed-Related Costs 1000 Risk/Loss 690 1050 261 265 Broker Fee 140 525 326 265 Tax/Fees 591 438 Credit 445 600 300 Insurance 534 Ship Agency 630 Port Charge 10 Kenya Tax 688 Market Broker Fee 575 Association Fee 100 Quarantine/Holding 840 Fee 500 Communications 1000 Trucking Costs 23000 21000 13029 8820 Purchase Price 31135 24758 15327 9725 Total Trader Costs Middlemen• Water 90 90 90 Fodder 50 25 25 data Tax 575 525 266 not Risk/Loss 570 480 212 available Broker Fee 475 400 266 Vet Costs 60 60 60 Hired Labor 405 405 405 Purchase Price 19000 16000 10620 Total Middlemen Costs 21225 17985 11944 •These categories do not reflect the full range of traders and middlemen involved in cattle trade (see discussion of traders later in the article). The costs included here are based on estimates from. a sample of 27 traders and are only illustrative. Average exchange rate during 1987 to 1988 (until February 1988) is calculated at 130 SSh=US$1.00. SOURCES: Author's data, and Stockton (1987) for overseas trade. for understanding the Somali livestock marketing system. At one end of the scale are the largest traders, those involved in overseas commerce who usu- ally have offices in Mogadishu or another large town. Their operations in- volve substantial expenditures of capital on veterinary services, office and communications equipment, and in some cases, trucks. At the other end of 440 MARKETING AFRICA'S HIGH-VALUE FOODS the spectrum are bush traders, who may account for no more than twenty cattle sales annually. These are usually indistinguishable from the herders. Certain traders, especially in the major market centers, serve only as brokers (called dilaal), while some act both as dilaal and trader. Brokers match buy- ers with sellers and charge a percentage of the sale price or a fee per animal for their service. As with most commodity market systems, the level of com- petition among traders diminishes as one moves further up in the market chain. Market concentration at the highest levels of the system seems to have increased with the steep decline in the volume of overseas exports. In the case of Kismayo, three large traders accounted for more than 70% of overseas cattle exports in 1987. Approximately half of the traders interviewed specialized in buying and selling only one animal type, usually cattle. From their d1scussions, it ap- pears that this sort of specialization has increased in the past 20 years. Among the twenty-seven traders in the sample, 30% traded both cattle and camels, 39% traded only cattle, 13% traded only camels, and 17% traded in all animal species. More than 30% of livestock traders owned water points (usually sur- face ponds) in the pastoral areas, and sometimes they provided water to herd- ers in exchange for the right to purchase cattle at a price below market level. Part-time Bush Traders The smallest traders or 'bush traders' are the most numerous and usually are involved in the regional domestic or Kenyan trade rather than in the over- seas trade. They are distinguished from other local traders and middlemen by their small scale-frequently they buy only twenty to twenty-five cattle a year-and their lack of full or nearly full-time commitment to trade. Many of these traders started in the livestock business by first accumulating some capital through trading in goats and then moving into the cattle trade. This category of trader operates in the more remote areas of the region, buying animals to sell to larger traders and middlemen such as those represented in Tables 9.3 and 9.4. They frequently buy from small- and medium-scale herd- ers rather than larger ones and they pursue a pastoral lifestyle that is indis- tinguishable from local herders. Many of these traders are young (under thirty- five years) and they use profits earned from livestock trade to build up their own herds and marry. Their position in regional livestock trade has improved as the Kenyan market has grown. They reside in the interior of the Lower Jubba Region at some distance from the large towns but often close to the Kenya border. MERCHANTS AND MIDDLEMEN IN THE CATTLE TRADE OF SOUTHERN SOMALIA 441 Jeeble Traders The term jeeble can only roughly be equated with the concept of a middle- man. 44 The scale ofjeeble enterprises varies considerably. Approximately 70% of middlemen sold animals to exporters during 1986 to 1988, but fewer than 20% did so on a regular basis. Middlemen frequently cooperate with two or three other jeeble, pooling resources and assisting each other with purchases. Seventy-one percent of middlemen have trading ties with other middlemen, usually of the same clan. Some of these relationships endure for years. The larger middlemen buy from smaller middlemen and then sell directly to ex- port traders. These larger middlemen are likely to be involved in at least one of the export trade channels. Among the larger middlemen, who handle more than 250 cattle per year, 83% are associated with the overseas trade, the unofficial trade with Kenya or both. It has been noted in the literature that the role of jeeble has changed in response to the internationalization of the Somali livestock trade. In discuss- ing the livestock trade in northern Somalia, where it is considerably more developed than in the south, Samatar and his co-authors suggest that middle- men became more important as the overseas export trade grew and exporters increasingly relied on them to procure animals in a timely fashion, especially during the pilgrimage season when demand in the Middle East is at its maxi- mum. In northern Somalia by the 1970s, the system of depending on brokers (dilaal) to procure animals ''was incapable of collecting livestock from thou- sands ofpastoralists in time for delivery during the Haj [pilgrimage] season." Jeeble were needed to augment the supplies that export traders could obtain from their own agents and brokers.45 In southern Somalia the situation seems to have had a similar development, althoughjeeble clearly were prominent in the Kenyan trade before the growth in overseas exports. In the Kismayo region the roles of both jeeble and broker (dilaal) assumed more importance 44 By definition, a middleman in the cattle trade is not involved in final transactions, but rather is a person who buys and then sells to agents of large traders or to the large traders themselves. Jeeble however refers to any type of petty trader, whether a middleman or not, and thus, it represents the most ambiguous category of traders. In the cattle trade of the lower Jubba, the term is used for most traders except for brokers and export traders. In the export trade, a jeeble is clearly a middleman, operating between the producer and the export trader, but in other market channels ajeeble may be involved in final transactions. In the overseas trade, some jeeble have begun to provide brokerage services to exporters, rather than actually buying and selling export-quality animals. For the pur- poses of this analysis, the termsjeeble and middleman will be used interchangeably. 45 Samatar, Salisbury and Bascom (1988). 442 MARKETING AFRICA'S HIGH-VALUE FOODS with the expansion of the overseas export trade. Yet, the supply of livestock to Saudi Arabia during the period of pilgrimage never assumed the impor- tance in the south that it did in the north. Market Brokers (dilaal) Cattle brokers are found in each of the major market towns, and in the large towns they are often required to be registered with the local government. They provide a market service to buyers by directing them to potential sellers and negotiating a price. At the smaller markets, it is the broker that guaran- tees to the buyer that the animal is not stolen and thus will not be reclaimed at a later date. 46 In the past, the broker was usually from the local area and knew the herders and townsmen well, but this has not always been the case since the growth in the overseas trade. What is the difference between a jeeble and a dilaal? Some jeeble also serve as dilaal, but this was recorded in only 15% of cases. The major differ- ence between the two is that unlike the jeeble, the broker does not actually buy or sell the animal; rather, brokers match buyers with potential sellers. Usual fees are 2.5% of the value of the animal, or a per animal fee of around $2.00 to $3.00. Although the broker merely performs a minimal service-- pointing out a potential seller to a buyer, for instance-traders in towns al- most always use a broker for procuring cattle since that is a way of sanction- ing the sale and provides a measure of insurance that the animal is not stolen. The number of brokers in a market varies according to its size. The largest numbers are active in Kismayo and Afmadow towns (in excess of fifteen in each). Over the years, individual traders have built up strong associations with certain brokers, using particular dilaal for procuring animals. Export traders especially rely on brokers, and some brokers work for only one large trader. By the late 1970s certain brokers and middlemen had begun to specialize in procuring cattle for only one or two export traders. The growth in the export trade resulted in some middlemen turning to brokerage activities full-time where they served as agents in the export trade. In providing brokerage ser- vices, middlemen do not have to spend large amounts of money to procure cattle themselves. Since purchase prices of export-quality cattle (US $150 to 46 Brokers are frequently found in livestock markets in Africa: see Dupire (1962); Cohen (1969); Manger (1984). MERCHANTS AND MIDDLEMEN IN THE CATTLE TRADE OF SOUTHERN SOMALIA 443 $200 each) are often 200% or more above prices of other bovines, outlays for cattle would be substantial. Agents of Large Traders Agents of large traders are middlemen and brokers who have attached them- selves to a large export trader (or traders). The agent position is a direct result of the growth in the overseas export trade, and reflects the movement of certainjeeble into procurement for export traders on a full-time basis. Dur- ing the Saudi boom period, agents profited more than other local traders. Since it is less costly for export traders to deal withjeeble than with bro- kers, jeeble more often serve as agents. They remain middlemen despite the fact that many would prefer to be paid a brokerage fee rather than to buy and sell animals. It is in the interest of the export trader, however, to force the jeeble to be responsible for purchasing the cattle. Under the brokerage sys- tem, the export trader has to pay a fee and account for the actual purchase. In dealing with a middleman, he avoids responsibility for lower-level transac- tions. It is the jeeble, then, who has to make arrangements for buying from smaller traders or directly from herders. In most cases the seller is paid at least half the price at the time of sale, and the remainder when the animal is exported. It is only after the exporter receives his money that the jeeble is paid in full and he in turn can pay off the herder. Even in cases where the full price of the animal has been paid to the herder, the middleman may be paid only 30-40% of the price until the animal is exported. Because these outlays are costly, only the larger jeeble can afford to deal with overseas exporters, and even these try to pass on the costs to herders by delaying their payments. After transaction costs are calculated, in 1988 middlemen earned a return of about 8% in the overseas export trade as opposed to approximately 14% in the Kenyan trade (see Table 9.3). Another reason why export traders prefer to deal withjeeble rather than with brokers is because they can shift responsibility for the management of the animals to thejeeble. Because profits have been increasingly squeezed in the overseas export trade, export traders have delayed purchases from middle- men so as to force them to incur more of the production costs. Once the animal is purchased, labor, water, veterinary and other production costs must be met. If the date for export is delayed or postponed, as was common in the late 1980s, the costs of maintaining large, high-quality cattle can bite deeply into a trader's profits. Not surprisingly, many jeeble have resisted these tac- tics by entering into the less restrictive trade to Kenya where possible. 444 MARKETING AFRICA'S HIGH-VALUE FOODS Most of the middlemen who moved toward the status of full-time agents for the export trade were from the major towns of Kismayo and Badhaade Districts, rather than from the smaller towns of Afmadow District. While there are important exceptions, many of those who are tied into the overseas export trade are Herti. Agents are frequently from the same clan as the ex- porters, or from a related clan. Yet, clan relationships do not eliminate the possibility of domination and exploitation by export traders. The manipula- tion of clan ideologies by larger traders disguises to some extent what are markedly class-based relations. In an activity as risky as livestock trade, it should not be surprising that family- and clan-based ties permeate many aspects of the business. This is a pattern in trading enterprises that is not unique to Somali, Africa or even to the developing countries.47 Another reason for Herti involvement in the overseas trade is that they live in or near the port of Kismayo and historically have been closely associ- ated with the town's commerce. On the other hand, the Maxamed Zubeer and other Ogadeen sub-clans do not usually reside in Kismayo town or its imme- diate environs, and thus, they are less likely to be involved in the overseas export trade. Export Traders · In the 1970s, several large-scale traders and trading companies from outside the area opened branches in Kismayo to take advantage of increased cattle exports. Most of them operated a diversified trade, using all three of Somalia's major ports (Berbera, Mogadishu, and Kismayo) for animal exports. Capital requirements for the trade were such that very few locally based traders became exporters (see Table 9.4). The overseas export traders operate procurement systems based on their own networks of agents. On average, export traders employ three to four agents full-time or nearly full-time throughout the year, as well as several middlemen on a part-time basis. Those exporters who do reside in the area live in Kismayo town, the only regional center that provides some of the banking and communications facilities required for the export trade. They do not maintain strong links to the nomadic sector or to smaller traders, but work instead through a small number of large brokers and middlemen. When interviewed in Mogadishu, many export traders had little familiarity with 47 See Clark (1988). MERCHANTS AND MIDDLEMEN IN THE CATTLE TRADE OF SOUTHERN SOMALIA 445 the actual day-to-day operations of their enterprises. Specific information about numbers and names of local traders with whom they worked had to be ob- tained from their employees in Kismayo. The overseas export trader buys cattle throughout the year, but restricts his purchases of large numbers until he has a contract from an importer specifying price, volume and date of export. He relies on middlemen to incur at least a portion of the costs of managing the herd. Since letters of credit from the Somali Commercial Bank are granted for only two to three months, the export trader cannot pursue much speculative buying prior to export. However, orice the merchant receives information on a definite export date and number of cattle, large-scale buying begins. Despite the recent slack in external markets, traders often cannot meet their contract terms because time constraints are too restrictive and cattle may have migrated away from the main purchase areas or, as in 1987, may be in very poor condition. Even reliance on local brokers and middlemen has not always allowed exporters to meet the conditions of contracts. For ex- ample, when representatives of an Egyptian import company came to Kismayo in May 1987 to check on the progress of their agreement with a Somali ex- porter, they found that their allotment of cattle had not been procured. The date of export had to be postponed for several months. At least part of the 1987 decline in cattle exports can be traced to difficulties in procuring cattle, especially since some jeeble have reoriented their trade to Kenyan markets. The time limitations on letters of credit make it difficult, costly and risky to procure large numbers of export-quality animals several months prior to ex- port.48 In recent years traders have depended on buying cattle from outside the region (for example, in the Lower Shebelle) to meet quota levels. Unlike the overseas exporters, traders who deal with the Kenya market are from the smaller towns such as Libooye and Afmadow. The logistical requirements for this trade are minimal, differing very little from those needed to sell animals to Mogadishu or any other domestic market. Thus, a mer- chant ofthis type is not likely to differ from local traders in the region and usually will orient a portion of his business toward regional and national domestic markets as well. The overseas export trader, on the other hand, 48When the Saudi Arabian ban was imposed, one exporter had more than 7500 cattle in the lower Jubba awaiting export. The trader claims to have lost almost 5000 of these during the severe dry season of 1984. 446 MARKETING AFRICA'S HIGH-VALUE FOODS differs considerably from these traders in terms of both scale of enterprise and degree of market specialization. Social Relations of the Cattle Trade Credit in a variety of forms permeates all levels of cattle markets in the region. Ironically, it is producers and middlemen of more modest means who provide credit to wealthier traders in the overseas export trade, a phenom- enon that also occurs in northern Somalia.49 Because they do not receive full payment until after their animals are exported, local middlemen and herders are providing credit to the export trader. Whereas in other livestock markets, traders often provide some form of credit to herders. Moreover, in these latter markets herders are usually paid in full at the time of sale, which is why many pastoralists prefer to trade with Kenya rather than overseas. Jeeble who own retail stores may advance maize flour, sugar and other necessities to herders. The same is true for traders who own water points and provide this resource to herders before sales are consummated. Credit is also pro- vided by Kenyan traders who regularly buy cattle in the lower Jubba. These loans to local middlemen may not involve a direct interest charge, but they usually obligate the borrower to sell cattle to the trader at a price equal to or below market levels. A large jeeble may have dozens of credit relationships outstanding at any given time, supported by few or no formal credit instruments. The practice has been referred to as a 'trust-based credit system'.50 In the lower Jubba region, it may equal the amount of credit provided through the formal credit system (that is, the government bank). In the absence of formal contracts in the 'trust-based system', the credit relationship is often reinforced by clan or kinship ties. The informal system works because relationships are not based on financial exchange alone. Even where clan is not a binding mechanism, local traders invest in activities that are not strictly commercial yet that reinforce market alliances. These can be as small a gesture as when a trader donates a few cattle to help a herder's son to marry or pays the Koranic school fees of a local community member. The overseas export trader taps into these relations by working through a small number of local brokers and middlemen who have strong ties to the community. Because most of the export traders are non-Ogadeen and their 49 Samatar, Salisbury and Bascom (1988). 50 Ibid. MERCHANTS AND MIDDLEMEN IN THE CATTLE TRADE OF SOUTHERN SOMALIA 447 business is focused on the Kismayo port, they recruit agents from among the local population. It is then up to local agents to invest in the relationships they need to ensure a regular supply of cattle. Dealing with a Market Crisis . The disintegration of the overseas trade and the subsequent collapse of the Somali economy left exposed the export traders and those local merchants and middlemen who had become agents in the overseas trade. Although ex- porters lost a substantial source of revenue, they were able to shift their emphasis from Kismayo to begin exporting from Berbera and Mogadishu and to focus on small stock and camels, rather than cattle. A number of export traders had already diversified into urban real estate and other business, using revenues derived from animal exports; these were additional buffers against the effects of the decline in cattle exports. This is not to imply that export merchants did not incur considerable losses because of the closure of the Saudi market. Nonetheless, it was the local traders and brokers who had invested most of their efforts and social capital into maintaining market channels for the export trade who suffered most. Because their alliances, including those with herders,51 were geared to the overseas trade (and to a lesser extent the Mogadishu trade), many of these local merchants and middlemen have not benefited from the unofficial trade to Kenya. In fact, they have been hurt by it because it has made it more difficult for them to procure animals. Nor do they have supplementary invest- ments on which to rely. Two examples oflocal traders demonstrate the injuri- ous effects of recent changes (Box 9.1). · Neither trader A nor B, both of whom are Herti, has been able to capital- ize on the booming unofficial trade· to Kenya. They do not have the ties- either with middlemen in Kenya and Somalia, or with Ogadeen herders -that would enable them to take advantage of opportunities in the Kenyan trade. Moreover, they are older traders, unlikely to want to endure the rigors of the Kenyan trade, which frequently requires moving animals long dis- tances over poorly watered areas of the lower Jubba and northern Kenya. In contrast to these traders, the changes ·in the cattle trade have placed many small-scale bush traders and Afmadow jeeble in favorable positions. 51 Large herd owners in particular benefit from the overseas trade because they are the ones most likely to own export-quality steers. 448 MARKETING AFRICA'S HIGH-VALUE FOODS Trat:{erA, . :~pproJ,s;.t1 in cattle exports •.at·tt1e end 1983, he cont1n,ued to.export a sma·n . • · . nUf))ber of camels 'but this \\I.as a limited. business. ·ouring •1986 to 198S. he · : .• ·l)either.·bo~gh(nor.soltl':ca~le. and has.•traded •less:..than 10.camets.peryear' on a\/erage.: 1:hus; his. revenue. from the .animal trad7dr()pped c~nsid~rably . . sihce. th.e •. rn.id·1980s; '(;lncL he: oow spends·· tl1ost ·ot hi·s time .• managing tlis. .small retaii store ·rather than trading.livestock/ · . •. · . · • ·.. · ... · · · .· .•.•....· ... · ·. · • · a: .,trade.r whors:9!t.y~ars 01q;·has.ppn.centrated his efforts qntne· over, · . . ; seas trade···auilng :the pf;!St 1~ .Year5. while maintaining SOrnetrade 1n·.·the'...: ·r$1ativelysmall.dome.stic ·market in, ~isn1ayotown~.··He s~rvesa.s an agent fof . . ·· ' . fiye eXP())'.t. tradt1rs (three of them. ~ased: in' the Mogadishu area) rat11er th~ri ... A: •only ~ne, :a~ was the c.aseJort.~der When ·theexi:iort tra~~ ~as g0pd, he· ·• .·.•·.· · .· . workec:J withfoUrsmall' rnidcllEfrne11inKi~mayo Ci,!StrictWt1o .supp!ied.t!im.wittt·· · •. ·y~ttle.~i.$ reveryjJehas :been,, hurt by the l()Ss:'of oyersea~ .mark~ts; •b!Jt riot•t