DRD DISCUSSION PAPER Report Noo DRD273 r GROl.JTH AND STRUCTURAL CHANGE IN EAST AFRICA: DOMESTIC POLICIES~ AGRICULTURAL PERFORMANCE AND WORLD BANK ASSISTANCE, 1963-1986 Part I by Uma Lele I and l L~ Richard Meyers I ·=------·-----------·- May 1987 Development Research Department Economics and Research Staff World Bank The World Bank does not accept responsibility for the views expressed herein which are those of the author(s) and should not be attributed to the World Bank or to its affiliated organizations. The findings, interpretations, and conclusions are the results of research supported by the Bank; they do not necessarily represent official policy of the Bank. The designations employed, the presentation of material, and any maps used in this document are solely for the convenience of the reader and do not imply the expression of any opiriion whatsoever on the part of the \~orld Bank or its affiliates concerning the legal status of any country, territory, city, area~ or of its authorities~ or concerning the delimitations of its boundaries, or national affiliationa Abstract A comparative overview will be presented of domestic policies (macroeconomic and agricultural) in Kenya, Malawi and Tanzania and their influence on the extent and character of agricultural growth. The contribution of the World Bank to agricultural growth in the three countries will then be reviewed, from the perpectives of both policy advice and lending prov.ided by the Bank. The presentation will be based upon the findings of case studies of the Bank's involvement in the three East African countries, carried out under DRD's Managing Agricultural Development in Africa (MADIA) Research Project. TABLE OF CONTENTS OVERVIEW •••••••••••• " ••••••••••••••••••••••••••••••••••••••••••• i INTRODUCTION •••••••••••••••••••••••••••••··~··••••••••••••••••••••••1 PART I-- A COMPARATIVE OVERVIEW OF THE THREE COUNTRIES •••••••••••••• 6 I. STRUCTURE AND PERFORMANCE OF THE AGRICULTURAL SECTORS OF KENYA, MALAWI AND TANZANIA ......................... 6 A. Initial Conditions at Independenee ••••••••••••••••••••••• 6 B. Agriculture's Role in Overall Economic Development and the Process of Structural Transformation •••••••••••• 6 C. Agricultural Performance •••••••••••••••••••••••••••••••• 14 1. Export Crops ••••••••••••••••••••••••••••••••••••••• 14 2. Food Production••••••••••••••••••••••••w•••••••••••l9 D. Agriculture and GDP Growth ••••••••••••••••••••••••••••• ;22 E. Macroeconomic Environment ••••••••••••••••••••••••••••••• 24 1. Levels and Sources of Government Expenditures •••••• 26 2. Inflation Rates •••••••••••••••••••••••••••••••••••• 31 3. The Effects of External Shocks ••••••••••••••••••••• 31 4. Patterns of Government Expenditures •••••••••••••••• 35 5. Indicators of Investments in Hu~an Capital ••••••••• 39 II. LEVELS OF FOREIGN ASSISTANCE •••••••••••••••••••••••••••••••42 III. AGRICULTURAL POLICIES•••••••••••••••••••••••••••••••••••••••48 A. Natural Resource Endowments and Policy Choices ••••••••• 49 1. Smallholder Intensification •••••••••••••••••••••••• 52 B. Small Versus Large Scale (or Estate) Production ••••••••• 52 1. Defining "Large" and "Sma11"•••••••••••••••••••••••53 2. Differential Marketing Institutions •••••••••••••••• ss 3. Taxation of Smallholder Agriculture •••••••••••••••• 60 c. Land Policies ••••••••••••••••••••••••••••••••••••••••••• 67 D. Policies Affecting Labor Markets •••••••••••••••••••••••• 69 E. Stability of Agricultural Service Institutions •••••••••• 71 F. Agricultural Research ••••••••••••••••••••••••••••••••••• 72 G. Fertilizer Policies••••••••••••••••••••••••••••s••••••••74 PART II-- ACTIVITIES OF THE WORLD BANK•••••••••••••••••••••••••••••82 IV. BANK ECONOMIC AND SECTORAL ANALYSIS AND POLICY ADVICE ••••••••• A. Introductory Comments ••••••••••••••••••••••••••••••••••••• B. Policy Advice to Kenya •••••••••••••••••••••••••••••••••••• 1. Changes Over Time in the Overall Policy Context •••••• 2. The 1974 Basic Economic Report ••••••••••••••••••••••• a. Evaluation of the BER•••••••••••••••••••••••••••• 3. Economic Policy Assessment in Subsequent Reports ••••• 4. Strategy Recommended by the 1982 BER ••••• ~······••ooo 5. Interaction of Poverty with Growth •••• oa•••••••••e••• 6. Recommendations of the 1986 Agricultural Sector Reportoooee•oooooooooooeooooooooooooooeooooo Co Policy Advice tO Tanzaniaoooooaoooooooecooooooooooocoooooo 1. The 1974 Agricultural Sector Reportoooooooooooooooooo 2. The 1977 Basic Economic Reportooceoeeoooooooooooeocoo 3. Critical Issues of the 1970s.ooccoooooaoooooo•~o•o•oo a. Equity Versus Growth•••••••••••••••ooooo•o••••••• b. Public Versus Private Sector ••••••••••••••••••••• c. Large Scale Agriculture •••••••••••••••••••••••••• d. Weather •••••• ., •••• ., o • • o. o ••• ., o" •• o . . . . . . . . . . . . . . . . . e. World Commodity Markets ••• ., .. o•o••.,•••••••• ... ., ..... f. Industrial PolicyoooomeoeoooCJoooocooeecooooooocoo 4. New Policy Emph£ses of the 1980s • .,oo•••••o•s•••••••o• De Policy Advice tO Malawioooooo~<>eooooooeoeooooeooeooeooeooo 1. The 19 60s ...... e e •• 0 • ., • • 0 0 • 0 0 0 II • • 0 ........ 8 • 0 0 " 0 0 •• 0 0 ., 0 • 2. The 19 70s ...... c • o• o • o • o • • ., o ., o • oo o o., • ., • o • ., • o.... oo... (' ., o • Q 3 o The 1980 s ., ..... o o o o . . . . . . . . o . . . ., • o o o o • " • o o o o o o o ... o o . . . . o • o V. Bank Lending to Kenya, Malawi and Tanzania .. o~•ooeooo•o•o•ooooo A. O?erall Lending PatternSooooooooooooe>coooooooooaooooeooooo B. Types of Projects Financed ••• o • .,••o•••••••••••••••o•co~o•• Co Lending tO Kenyaeoooooo<>eoooeeoeoooooeoooeoooooooooooooooo lo Themes in Lending in Kenyaocooocooooooooooooeeooooooo Do Lending tO Tanzaniaeoosooooocoooooeoooeooooooo<>oocoooooooo lo Support for Crop ParastatalSooeooooooooocoooooocooooo 2c Support for Regional Development Projects •• ., •• o•••••• 3o Structural Adjustment Lending -- Recent DevelopmentSoooosooooccooeoooo~••••o•oojoooooeeoo&o E. Lending tO Malawieeoeoooeooooeoooceoooeoooooeoeoooooooesoo 1. Overall Project Performanceoocoooooo••••••oooocco<>ooo 2.. Project Impacts on Constraints to Agricultural Growth 0 0 0 0 0 0 • 0 0 c Q (I 0 0 •• 0 c c 0 0 •• 0 Q G 0 •••••• 0 s 0 0 e 0 c ••• Cl 0 a. Slow Growth of Smallholder Exports ............. e•o• bo Diversification of the Export Base ......... ~······ c~ Budgetary Implications.o••••••••••••••o••o••••••• do Parastatal Financing ••••••••••••••• $••••••a•••• .. o 3. Structural Adjustment Lendingooooeoo••••••••••••••••• CONCLUSIONS oocoeu&aegoceoeooaec•eoogo~eooc•oo~•••••••~••••••o•o•a••G ! OVERVIEW Comparisons between Kenya, Tanzania and Malawi are of interest because all three started with somewhat similar initial conditions at Independence but have followed quite different policy paths with very different economic outcomes. Agriculture is nonetheless the most important source of employment, income and exports in all three countries. Not surprisingly, the performance of the agricultural sector and the agricultural policies pursued in each have been closely related to the country's overall economic performance and policies$ Kenya and Malawi have both done quite well in terms of growth of export crop production, but Kenya's performance has been far superior in reconciling growth with equity. Tanzania has done least well on growth of export crops, including those produced by smallholders. Tanzania's efforts to sustain policies to achieve equity have been hampered by the lack of growth of the economy. Malawi's strong export growth has until recently diverted the attention of many observers, including the Bank, from the sources of that growth, including examination of the basic structural policies the government has pursued as well as the technological constraints that have adversely affected Malawi's smallholder sector performance. The relative performance of each country in the food sector is more difficult to compare due to weak data. Again, however, Kenya appears to be more advanced in promoting the process of technical change in the smallholder sector, especially in maize production. The breadth of participation in growth has had a profound impact on the process of economic development in each country. Achieving equitable - ii - growth requires the development of a sophisticated network of institutions to service the needs of a large number of small, geographically dispersed producers with diverse resource endowmentso Kenya, which admittedly started out with the most favorable institutional base at Independence derived from its large (European) farm structure, cashed in on this base and greatly broadened small farmer access to institutional serviceso Malawi's historical base of institutions serving a modern European agriculture was narrower than Kenya 1 so Its subsequent growth has maintained this narrow base of a European estate sector along with an evolving but equally narrow indigenous estate sector in which growth appears to have occurred at the cost of incentives and investment opportunities for the smallholder sectoro Tanzania pursued policies aimed at dismantling its historical institutional base, and experimented with many new institutional arrangements 9 which greatly destabilized the environment for smallholder productions The structure of agricultural production and its growth is, however, not simply determined by institutional and microeconomic factors but by the quality and the stability of the macropolicy environment within which agricultural production is carried outo Kenya's macroeconomic and sectoral policies were far more conducive to growth than Tanzania's throughout the 1970so Depending on the particular policy under examination~ Kenya and Malawi exchange places in terms of demonstrating superior macroeconomic management however, if the interaction of structural (estate-oriented) policies with macroeconomic policies is conside~ed, Kenya was certainly superior to Malawio Both Kenya and Malawi have in addition provided a more stable institutional environment for development than has Tanzaniao Also, external shocks were more adverse in the case of Kenya and Malawi than of Tanzaniae - iii - Tanzania's resource base is far more diversified and favourable for growth than that of Malawi and perhaps even Kenya. Land availability, as reflected in land person ratios, is much greater in Tanzania compared to Kenya or Malawi although a small proportion of Kenya's land (4%) is of very high quality. Per capita ODA levels have, however, been substantially higher in Tanzania than in Kenya and Malawi. While they began to decline from tr~ir 1981 peak due to Tanzania's tardiness in adjusting its macroeconomic and sectoral policies, they were still higher than in Malawi and Kenya in 1984 as donors were slow to recognize the adverse effects created by Tanzania's domestic policy environment. The above arguments lead us to conclud·~ ·,hat policy variables explain much of the growth or stagnation that has occurred in the three countries. Similarly, they help to explain how the benefits of growth have, or have not, been distributed. The Role of the World Bank With the exception of smallholder tea, coffee and dairying in Kenya, there appears to be relatively little connection between where growth has occurred in the agricultural sectors of the three countries and where the Bank has provided about $994.1 million worth of agricultural project assistance as of 1986. In addition the Bank provided $440.9 million of assistance in the form of sectoral or structural adjustment lending in the three countries during the 1980 to 1986 period. The fungibility of resources diverted to the estate sector explains this phenomenon in Malawi, where the Bank concentrated its resources in the smallholder sector but in which there has been little - iv - growth. Growth in smallholder tea and coffee in Kenya -- the main source of its agricultural growth-- occurred contrary to the Bank's worldwide advice on tea and coffee expansion to countries producing these commodities (although, paradoxically, the Bank's lending for agro-processing was crucial for expansion of smallholder production in Kenya). In Tanzania the Bank's 1973 Agricultural Sector Report correctly identified the constraints to growth and stressed the need for a sequential approach to the development of smallholder agriculture that could capture the most obvious sources of growthe However~ this approach conflicted with Tanzania's policies. The Bank's policy analysis after that was very constrained by the Bank's reluctance to directly question Tanzanian policieso Its project portfolio was~ until about 1981, very conditioned by Tanzanian policies that were not growth-orientedo By the early 1980s macroeconomic difficulties were reinforced by external shocks in all three countrieso These were combined with severe project implementation difficulties being encountered, especially in Kenya and Tanzania~ but also in Malawia This was partly a result of the rapid expansion of Bank lending, as well as that of other donors, to the agricultural sectors of these countries, often for quite marginal activities under conditions of weak planning and institutional capacityo The World Bank financed a total of 68 agricultur~l project operations in Kenya~ Malawi and Tanzania between 1965 and 1986 -- 26 operations in Kenya with commitments of $500o50 million; 17 in Malawi with commitments of $172.69 million and 24 in Tanzania with commitments of $320c95 millionD Of 24 World Bank agricultural project operations completed in the three countries between 1965 and 1986f involving investments of $266 million, only 14 had positive - v - rates of return; ten had EERs equal to or exceeding ten percent. Not all these poor realized returns were the result of unanticip~ted problems. Many marginal investments were approved in support of political objectives of the governments, especially ones concerning interregional income distribution. While taking these concerns into account, it can nonetheless be argued that the projects financed were often not necessarily the most cost effective way of addressing such concerns. This and other evidence suggests that the countries would have been better off if they had not borrowed from the tiank for many of the activities funded. This is more true for Kenya and Tanzania than for Malawi where ERRs for a larger number of projects suggest a more positive impact. However, economic evaluations are done immediately upon the completion of projects. More recent data on Malawian smallholder a~riculture raise questions about the long-term effects of projects initially regarded as favourable. Another noteworthy feature is that until quite recently the Bank's assistance (as well as aid levels) were not positively related to the conduciveness of policies or the level of performance of the three countries. Pressure to lend in the 1970s resulted in indiscriminate growth in lending and weak project portfolios that did not clearly reflect the positive features of the Bank's macroeconomic and sector analysis. The early 1980s ushered in an era of greater appreciation within the Bank concerning the nature of the interactions between macroeconomic, sector and micro constraints to growth and the need to more directly relate the level and the composition of lending to the macroeconomic and sectoral policy environment. This realization had three consequences: (1) attempts by the Bank to seek macroeconomic and sectoral policy and institutional reforms in - vi - each of the three countries;ll (2) cancellation of poorly performing projects; and (3) development of new projects that were geared to improving the capacity of the governments to more eff&ctively deliver basic agricultural services~ e.g.~ research, extension and input supply. Policy distortions in the three countries have been the least in Kenya although ~orne difficult institutional problems remain with regard to land tenure and the role of the private sector in agro-process1ng and marketing. The Bank, however, was slow to appreciate the complexity of these issues. This led to an untimely effort at grain marketing liberalization attempted through the second SAL in a period culminating in a severe drought. By 1985, the Bank's policy and project dialogue in Kenya had returned to a more balanced effort to address the problem of priorities in the sector as well as a number of institutional issues of a long standing nature that had repeatedly been confronted in the course of project lend~ng. The past and future sources of growth in Kenya center on the issue of intensification in tea, coffee, maize, dairying, etc. The Bank would appear to now be on the right track in Kenya in concentrating on improving agricultural research and extension, credit, marketing, etc. in order to to achieve intensification. Nevertheless, the relatively limited diagnosis, through pri.mary data collection and analysis, of the precise constraints to achieving growth and hence the speed of reform may continue to be problems unless the balance of resources devoted to lending versus analysis changeso Secondly, the Bank needs to seriously reconsider its policy advice to Kenya ll $440o90 million were provided in the three countries ($220.9 in Kenya, $170.0 in Malawi and $50.0 in Tanzania) in support of macroeconomic and sectoral reform. - vii - concerning the development of coffee and tea. The policy has been prompted by concerns about limited world market prospects for tea and coffee and the I collective good of beverage producing countries whose interests are served by limiting production. However, this advice has not served Kenya well and has been inconsistent with the realization of a dynamic comparative advantage. Equally important, the treatment of risks has been quite weak -- including I those related to the non-realization of the Bank's price forecasts in the estimation of economic benefits. At a more general level the issue of the prospects for primary commodities produced in Africa and its implications for country and project specific advice needs serious review by the Bank. The effects of macro and sectoral distortions on agricultural performance and on the Bank's portfolio have been the greatest in Tanzania. The Bank was tardy in taking into consideration the importance of the policy environment for the size and the content of its lending program and in several ways reinforced the government's worst tendencies through project assistance i.e., support for the government's import substitution industrialization I strategy and its excessive focus on equity. These problems were identified in the Bank's 1983 Agricultural Sector Report, which repeated many of the themes of the 1973 report. Once recognized, the difficulties of the large project portfolio combined with the government's slowness in responding to these problems, brought the Bank's agricultural lending activity to a virtual standstill from about 1983 to 1986 when the government began to reconsider structural reform. In Malawi, on the other hand, the Bank, through the SAL process and several new projects in agricultural research and fertilizer distribution, has since the early 1980s helped the government to correct some of the more - viii - important policy distortions those that favoured estates at the cost of smallholders in the 1970's. On other sectoral policy issues which will have profound long term effects on development~ eGg. 9 the land issue, the speed of removal of fertilizer subsidies and the restructuring of ADMARC (the agricultural marketing parastatal) 9 the Bank in our view needs to go further in analysing the basic sources of Malawi's structural problems and needs to help develop a long term strategy of development that will address the question of how to better reconcile growth with equity. In this analysis the political economy aspects of policy reform need far greater emphasis than is true of the more narrowly defined economic analysis usually conducted by the Banke The most important conclusions of our research concern the recognition of the Bank's ob~~ous comparative advantage in policy analysis and in the articulation of long-term country specific development strategies in support of which donor assistance and domestic resource mobilization can be organized through aid coordination. However, we observe a pattern in the Bank's operations of insufficient analysis of specific constraints to long term development 9 including consideration of the implications for sequencing and phasing of policy reforms and investments 9 before reform packages are put in placeo This has been accompanied by the lack of a long term view of development 9 one that in particular places greater emphasis on human capital/institutional development in the recipient countries relative to the emphasis phased on financial resource transferso There is also inadequate effort at aid coordination in which the comparative advantages of other donors to undertake specific activities in support of a long term strategy are explicitly recognized. - ix - The issues of donor comparative advantage and lack of analysis of specific constraints are closely related. Lack of attention to critical constraints is due in part to the insufficient attention paid to micro level factors that potentially might inhibit the success of investments. This in turn stems from limited analytical capacity in recipient countries with which to undertake the necessary microanalysis. Thus, donors need to devote greater attention to building up such analytical capacity~ While this is an area where the Bank does not have a particular comparative advantage, it needs to both recognize and encourage the efforts of those donors who do possess such comparative strengths. GROWTH AND STRUCTURAL CHANGE IN EAST AFRICA: DOMESTIC POLICIES, AGRICULTURAL PERFORMANCE AND WORLD BANK ASSISTANCE, 1963-1986 Uma Lele . L. R1chard Meyers * INTRODUCTION This paper on three East African countries, Kenya, Malawi and Tanzania, draws upon the results of a wider study of the role of foreign assistance in African agricultural development, which is in turn a component of a major World Bank research project, "Managing Agricultural Development in Africa" (MADIA).!/ The focus of the MADIA study on agriculture is the result of the wide recognition among African governments and ·donors of agriculture's central importance in overall economic development and their acknowledgment that past failures in Africa have been largely a result of the failure of the agricultural sector. The foreign assistance component of MADIA consists of case studies that examine the experience of eight donors, including the World Bank, in providing assistance to agriculture in six African countries since their Independence.~/ This synthesis paper on three East African countries * We are grateful to Henry Russell, Lien Tran, Natasha Mukherjee, Linda Nunes-Schrag, Jan Sundgren, Pierre Seka, Narendra Rustagi, Robert Geraci and Maria Cancian for research assistance and computer work and to Kim Tran and Estela Zamora for tireless typing of drafts. !/ The other two major components of MADIA consist of: 1) a study of the politics of agricultural policy and 2) a study of the relationship between agricultural policies and performance. 2/ Other donors are USAID, SIDA, DANIDA, ODA, EEC, W. Germany (BMZ) and France. - 2 - summarizes the World Bank's experience in Kenya, Malawi and Tanzania, drawing on detailed reviews of the Bank's activities in eachal/ It draws on the World Bank's Tanzania Agricultural Sector Report prepared under Uma Lelets direction in 1981-83, and subsequent comparative work on Tanzania~ as well as reports prepared for the MADIA study on the World Bank's role in Kenya and Malawivs agricultural developmenta~/ The data on macro economic structure and performance were developed by Yaw Ansua Those on agricultural performance and policies for Kenya were developed by Michael Westlake and Kevin Cleaver, Chandra Ranade for Malawi, and Uma Lele and Ellen Hanak for Tanzaniao The analysis of the Bank's role has involved a detailed review of the Bankes formal and informal economic and sector work as reflected in Basic Economic Reports, Annual Economic Memoranda, Agricultural Sector Reports and various project related documents, including Staff Appraisal Reports (SARs), Supervision Reports, Project Completion Reports (PCRs) and Project Performance and Audit :&.eports (PPARs)o Consultation with the concerned Bank's operational staff has also been an equally important independent source of data and analysis. Field investigations were carried out by Uma Lele in April 1985, January 1986 and July 1986, involving discussions with co-financers of Bank 1/ The other three countries included in the MADIA study are Nigeria, Cameroon and Senegale 3_/ Uma Lele and Lc Richard Meyers, nAgricultural Development and Foreign Assistance~ A Review of the World Bank's Experience in Kenya''; and Jo G. Kydd and N.Jo Spooner, uThe World Bank's Analysis of Malawian Agriculture: Changing Perspectives, 1966 to 198Sn. (Drafts). - 3 - projects (in particular CDC, ODA, and USAID) and interviews with Kenyan and Malawian government officials and farmers. Lele's investigations in Tanzania extended from 1972 to 1974 and from 1977 to 1982. During 1979-82 she was responsible for the Bank's agricultural sector analysis and lending operations in Tanzania. The study also draws on other research on each of the countries by Bank staff and outside researchers.!/ The study would not have been possible without the cooperation and encouraging support of the Bank's operational staff and the governments concerned. While the analysis presented here is necessarily historical in approach, it is intended to be more than just a retrospective treatment of the Bank's involvement in the agricultural sectors of the three countries. Its objective in tracing past developments is to try to understand the relative roles of domestic policies and the Bank in the agricultural development of the three countries over the past two decades and then to explore priorities for future government policies and donor (especially World Bank) assistance. Thus, the paper examines in detail the nature and patterns of agricultural growth in Kenya, Malawi and Tanzania and then reviews the contribution of the World Bank to the process of agricultural growth in each of the three countries. The process of agricultural growth is examined in the paper from a specific point of view, derived from the work of a number of agricultural 1/ For instance, more than 500 Ph.D. theses on MADIA countries have been written in the U~S: a number are on agriculture -- 43 deal with Kenya, 20 are on Tanzania and 4 are on Malawi. These have been drawn upon extensively as they often contain valuable detailed information on specific problems. We have also drawn on the publications of FAO, ILO, IFPRI and the Institutes for Development Studies in Sussex. and Nairobi as well as those of scholars in MADIA countries. - 4 - economists who have been concerned with the process of structural trans- formationo!/ These economists have explored patterns of agricultural growth which simultaneously created increased employment and incomes while expanding output. They recognized that at the early stages of development there are diminishing returns in agriculture under traditional technologies. However they point out how efficiency increasing reallocation of resources in agriculture which favours the increased use of labor in agricultural output growth shifts effective demand outwards~ while at the same time achieving rapid growth in employment and incomes. Further that this increased effective demand has important growth linkage effects which make the process of development self- sustainingo This paper therefore examines the policies and performance of the three East African countries in their po~t- Independence period from the viewpoint of the extent to which they have achieved equitable growth and have created growth linkages. The process of agricultural growth has also been considered from the viewpoint of intensification, which is defined in three different and inter- related ways: (i) a shift from low to high value crops on any given land; (ii) increase in yields per hao of any given crop; and (iii) a geographical shift in production of crops from areas of poor land quality to those of higher land quality. The policies affecting agriculture in each country are evaluated from this perspective. Lastly, it is hard to review relationship~ between donors and the 1/ See J. Wo Mellor, The New Economics of Growth: A Strategy for India and the Developing World. Ithaca~ Cornell University Press, 1976; B. F. Johnston and Pa Kilby, Agriculture and Structural Transformation: Economic Strategies in Late Developing Countries. New York: Oxford University Press; U. Lele and John Wo Mellor, hTechnical Change, Distributive Bias and Labor Transfer in a Two Sector Economy,u Oxford Economic Papers, 33, 3 (November~ 198l)g 426-441. - 5 - recipients of aid without seeming to be second guessing the protagonists with the benefit of hindsight. To some extent this is unavoidable if lessons are to be learned for the future. Nevertheless the study devoted considerable effort to identifying the information that was available at the time decisions were made. It discusses how this information was (or was not) applied, and uses decisions and outcomes to suggest policy, analytical, procedural, staffing and management implications for Bank/government interactions in the future. Organization of the Paper The paper is divided into two parts. Part I provides a comparative overview of key macroeconomic indicators as well as the post-Independence structure and performance of agriculture in the three countries. It also briefly summarizes the effects of external factors, including terms of trade effects, on economic growth. The cumulative effects of overall foreign aid levels are briefly reviewed, with particular attention to the relative contribution of the World Bank. Lastly, the domestic agricultural policies of the three countrieJ that have contributed to the agricultural performance of each are discussed. Part II describes the evolution of the Bank's assessment of the agricultural potential~ performance and policies of each country. It then reviews the nature of policy advice provided by the Bank to each. A description of the Bank's agricultural lending by subsector follows in which lessons learned by the Bank and the governments are identified as well as some that might still be learned.l/ 1/ Development assistance considered includes both World Bank loans and more concessional funding provided by IDA credits. - 6 - PART I A COMPARATIVE OVERVIEW OF THE THREE COUNTRIES I. STRUCTURE AND PERFORMANCE OF THE AGRICULTURAL SECTORS OF KENYA, MALAWI AND TANZANIA A. Initial Conditions at Independence All three countries are former British colonies (or protectorates)~ have relatively similar ecological conditions and grow many of the same crops. At Independence they inherited similar agricultural structures, consisting of a large number of small African farms and a modern agricultural sector operated by colonial settlers. Of the three countries, Kenya had the most favorable conditions in terms of the size and development of its modern, largely European agricultural sector, and of its economy, physical infrastructure and institutions. Both Kenya and Tanzania enjoy good port~, while Malawi is landlocked and has faced serious transportation difficulties since the late 1970s. In 1965 Kenya•s per capita income of $103 in 1965 was the highest followed by Tanzania's $77 and Malawi's $63. Kenya and Tanzania had similar size populations (9.5 and 11.7 million in 1965, respectively) compared to Malawi's 3.9 milliono B. Agriculture's Role in Overall Economic Development and the Process of Structural Transformation Agriculture accounted for quite similar employment shares (84 to 91 percent) in the three economies at Independence in the mid-1960s (Table 1) and by 1980 the percentages were still quite high (78% in Kenya, 86% in Malawi and Table 1 EMPLOYMENT SHARES BY SECTOR (PERCENT) YEAR KENYA MALAWI TANZANIA Agriculture 1960 86.0 92.0 89.0 1965 84.1 90f)6 87~6 1970 82.0 89o0 86o0 1975 80"1 87o6 84e6 1980 78.0 86.o0 83o0 Industry 1960 SoO 3o0 4.0 1975 So9 3o5 4o5 1970 7o0 4o0 SeO 1975 8o4 4o5 SG)S 1980 lOoO SeO 6o0 Other Sectors 1960 9o0 stio 7,0 1965 10.0 5f)9 7o9 1970 lloO 700 9.0 1975 11.5 1e9 9.9 1980 12.0 9o0 11.0 Note~ Industry is defined here as Manufacturing, Mining and Quarrying 9 Construction, and Public Utilities (electricity, water~ gas, sanitary services) o Source: World Bank, World Tables, Volo II~ 1983o - 8 - 83% in Tanzania). In a Kuznetzian "normal" pattern of growth, agriculture's share in a country's GDP tends to be high early in the development process and to fall during later stages. Surprisingly, during the 1967-73 period agriculture already contributed a relatively low 34.3% to Kenya's GDP, compared to 44.4% in Malawi and 40.8% in Tanzania (Table 2), once again reflecting Kenya's more advanced stage of structural transformatione By the 1982-1984 period, however, agriculture's share had fallen slightly to 32.6% in Kenya and 39.8% in Malawi, but had risen to 51.6% in Tanzania. Meanwhile, industry's share in GDP, which was around 12% during 1967-1973 in each country, had risen by 1982-84 to 15.6% in Kenya and 11.8% in Malawi, while falling to 9.7% in Tanzania. Tanzania's Basic Industrial Strategy of the 1970s (which strongly discriminated against agriculture -- see Section IV.C.3.F below) actually had the reverse of its intended effect: it raised agriculture's share in GDP and reduced industry's. The share of trade in GDP was initially also similar in the three countries (Table 3), in which agricultural exports dominated. Trad~ shares amounted to 58.5% in Kenya, 51.2% in Malawi and 53.8% in Tanzania for 1967- 1973 with agriculture's share in trade being over 90% in Malawi ar1d about 80% in Tanzania (Table 4). In Kenya, however, it had fallen over time, from 74.7% during 1961-73 to 56.6% in 1979-81. By 1974-1978, trade shares had risen to 67.5% and 56.9% in Kenya and Malawi respectively, reflecting increases in agricultural export volumes as well as prices. In Tanzania, however, trade shares had declined to 48.5% despite the relatively more favorable movements in its barter terms of trade compared to Kenya and Malawi (Figure 1). Tanzania's stagnancy or decline in exports is reflected in its much poorer income terms of trade {Figure 2). All three countries registered falling _, 9'- Table .2 nn~ nF.CO~O~t~!ON RY ~Er.TORS ( 'D'F.ltCF.NT OF Gnt») YEAR KENYA TAN~ ANI A l 19o7 lq7tl.... Al AQ;ricul ture 31o0 18o0 51oi Industry 11()3 lloA 11 .. 0 ('f.anufaeturin~) 13e0 llo8 10o4 (Minin~) 01)2 OoO 0()5 Infrastructure 141)2 l3c4 lOoS Public Admin1stre & Defense l4oa 9o8 9o9 Others 24.1 27 oO li ol 1982-84 Agrieul tu re 32o6 39e8 5lo6 Industry 13o6 llo8 <)o7 ('Manufacturing) 15Q4 lle8 9o3 (Minin~' Oc2 OoO 0.4 !nf!'astructure <)el 12.0 10.2 Public Administr. Iii nefense ~OC) lleO 11.1 Othel.*'s 12.A 25.4 l7c2 * Industrv is rlefined as ~1ning (fuel and other metals) and ~nufacturin~() •~ !fifrastruct:ure is rtet'ined as r.onstruction and 't'ransoot:t and r.ol'lmYUnication. *** others includes Trades 11 'Rank/!ntJuranc:e/'R.eal Estate ~ervi.i~&tl .and Unst)eeifierl. Sources: World ~nk, EPn for data u~ to 19A3 for ~enya and ~lawi, to 1982 for tanzania. lqA4 data are obtained from CEMS for ~enya and Malawi. TJn-puhlished data obtained from the '1\snk of Tanzania and other "t'anzania Authorities for 19A1-19R4 data. Table 3 -fo- TRADE SHARES IN GDP* (PERCENT OF CURRENT VALUES) KENYA MALAWI TANZANIA Share of ExEorts** 1967-73 28o5 194P7 25.6 1974-78 32.0 22o8 19e5 1979-81 26.8 25a2 14.9 1982-84 26.5 20.2 10.9 ImEorts** 1967-73 30.0 31o5 28.2 1974-78 35.5 34ol 29.0 1979-81 35.6 39.1 26.2 1982-84 29.3 27.3 22.7 Share of Trade 1967-73 58.5 Sle2 53.8 1974-78 67.5 56.9 48G5 1979-81 62.4 64.3 41.1 1982-84 55.8 47.5 33.6 Sh~re ~· of Net Ex:eorts 1967-73 -1.5 -11.8 -2.6 1974-78 -3.5 -11.3 -9.5 1979-81 -8 .. 8 -13.9 -1le3 1982-84 -2.8 -7.1 -11.8 * GDP is at market priceso ** Both Exports and Imports include go~ds and non-factor services. Source: World Bank, EPD for data up to 1983 for Kenya and Malawi, to 1982 for Tanzania, 1984 data are obtained from CEMs for Kenya and Malawi. Unpublished data obtained from the Bank of Tanzania and other Tanzania authorities for 1983-1984 data. Table 4 -!)- COMPOSITION OF TRADE * (PERCENT) YEAR KENYA MALAWI TANZANIA A. EXPORTS 1967-73 Agriculture 74e7 97.0 (Food) 78c2 ( )oO 91.7 Manufacture 48o2 12.5 2o7 ll.,2 Fuels 12.0 O.l 1o8 ~etals and Minerals 0.7 Oo2 Oe8 1974-78 Agriculture 66.1 95o5 (Food) 84oJ 54e5 93o5 58o0 Manufacture lJol 4.4 Fuels llcl 19o9 401 Metals and Minerals Oe9 0.1 OoS 1979-81 Agriculture 56.6 93.8 (Food) 19o5 481)8 92.2 60o7 Manufacture 12$5 6o2 Fuels 14ol 28o5 2o7 Metals and Minerals 2o4 3.7 1982-34** Agriculture 61 .. 9 (Food) 55o4 Manufacture llo8 Fuels 24o0 Metals and Minerals 2e3 B. IMPORTS 1967-73 Agriculture 9oS 15.3 (Food) 9.3 7e4 14.2 8.4 Manufacture 78.4 76.3 Fuels 79(18 10o8 7.1 9.4 Metals and Minerals ls3 lo3 1.6 ·-1:;1- Table 4 (continued) YEAR KENYA MALAWI TANZANIA 1974-78 .. Agriculture 8.8 9.9 15.2 (Food) 6.7 9.1 14.0 Manufacture 66.3 76.& 68.4 Fuels 23.5 11.8 14.4 Metals and Minerals 1.4 1.5 2.0 1979-81 Agriculture 8.2 9.1 9.3 (Food) 2.0 8.3 8.3 Manufacture 58.8 73.9 67.3 Fuels 31.7 15.6 21.9 Metals and Minerals 1.3 1.4 1.5 1982-84** Agriculture 10.5 (Food) 8.6 Manufacture 51.5 Fuels 6.7 Metals and Minerals 1.3 Food Imports Per Capita (in constant 1967 US$) 1967-73 2.3 2.5 1.5 1974-83 2.6 1.7 (74-81) 2.3 (74-81) * Legend: The Standard International Trade Classification (site) code has been used: Agriculture SITC (0+1+2+4-27-28) (Food) SITC (0+1+22+4) MANUFACTURE S!TC (5+6+7+8+9-68) Fuels SITC 3 Metals and Minerals SITC (27+28+68) Total Merchandise Exports; Imports 3 Agriculture+Manufacture+Fuels+Metals and Minerals. ** 1982-84 figures are for 1982-83 for Kenya and are not available for Malawi and Tanzania. Sources: The World Bank Trade System (EPI) for trade data. IMF - International Financing Statistics (1985) for population. Figure 1 -/3 .-. terms of trade index (1972 bon) 140 130 120 , 10 100 90 70 1969 1970 1971 1972 1973 1974 1975 1976 1977 197! 1979 19!0 1981, 1982 1983 1984 o ken + mal Figure 2 income terms of trade (base 1967) 160 ~---------------------------------------------------------, 150 - 140- 130 ""i I ! 120 -i ' 100 I 90 -t' 80 -: 70 _; 50 ...., 19671968196919701971 1972197.319741975197619771978197919801981 198219831984 0 kenya + malawi <> tanzania Source: IFS (IMF), 1985. - 14 - trade shares in the 1982-84 ~eriod. For Kenya and Malawi the terms of trade declines were far greater relative to the 1972 base, than for Tanzania, and indeed during the 1983 and 1984 period Tanzania's terms of trade had recovered from a low of 1982, whereas in Kenya and Malawi they continued to remain depressed. Tanzania's trade share had fallen to 33.6% by 1982-1984, over 20 points below its 1967-1973 level, as volumes for a number of its agricultural exports (sisal, cashews, tobacco) stagnated or declined. Table 3 also shows that Kenya has had the best record for net exports, which were -1.5% of GDP in 1967-73~ dropped to -8.8% in 1979-81 and rose to -2.8% in 1982-84. Tanzania's net exports were -2.6% of GDP in the 1967-73 period and worsened to more than -11% after 1979. Malawi has consistently had much larger negative net exports (slightly over -11% of GDP) than Kenya or Tanzania until 1982-84, when its negative balance improved to -7.0%. Current account deficits in the three countries showed roughly similar patterns to those of net exports. The reasons for the differential behaviour of the trade shares and net exports are better understood through the more decomposed picture of export and food crop performance presented below. C. Agricultural Performance 1. Export Crops Coffee and tea are the two major export earners in Kenya. The share of coffee ranged between a quarter to a half of agricultural exports, depending on international prices, and averaged ~0-28% of total exports in Kenya during the 1970 to 1985 period. The share of tea in agricultural exports ranged between 13% and 36%. On average, tea constituted 12-19% of total exports. Coffee and tea export volumes increased at 3.8% and 7.5% - 15 - respectively in the 1970 to 19.85 period (Table 5) 11 Kenyan exports of horticultural crops also grew rapidly (12.7%) between 1970 and 1985 --albeit from a small base. Smallholders have played an active role in Kenyan export crop production through a shift in ctopping patterns to higher value crops. The growth rate of smallholder coffee production was 6% compared to 1% for estates leading to an increase in the share of smallholder coffee production from 35% (60% of total area) in 1964 to around 60% in the 1980s (75% of area)~~ Smallholder tea production increased at an impressive 13.5% compared to the 5.5% growth in volumes from estates and the share of smallholder tea production (and area) increased from around 5% in the mid-1960s to 48% in 1985 (65% of total area)o Smallholder sugar production grew at 16~~9% and estate production at 5.3% with the smallholder share of sugar growing from 11% of the total in 1973 to 48% in 1984. Much of the impressive growth in the volume of horticultural crop exports in 1970-85 came from smallholders. The smallholder share of marketed maize sales, which was negligible at Independence, is currently 45%. Smallholder dairying production increased at 8.5% with the smallholder share increasing to 50% of the country's milk (to which smallholders contributed little at Independence)~~ Almost all the marketed production of rice, pulses, cotton and pyrethrum comes from smallholders. In sharp contrast to Kenya~ tobacco~ tea and sugar, the three important exports of Malawi, have all been estate crops, with tobacco earnings ranging from 43% to 51% of the total earnings during 1970 and 1985, those of tea between 15% and 22% and sugar rising from a low base of 1.7% in 1970 to 17% du~ing 1979-81, but then declining to 9.3% during the 1982 to 1985 period. -------------------------------------------------------~------------------------------------------~------------------------------------------------------~ Table 5 STRUCTURAL CHPJWE AND GROWTH: AGRICULTURAL PERFORMANCE IN KENYA, HALAWI AND TANZANIA 1970-1985 a/ (Growth Rates in-volumes) Exports Production Food Production Kenya Kenya Malawi Maize# Coffee 3.8 Coffee ---rea Production Purchases Sales (Net Sales) Tea 7.5 Smallholder 6.0 - Estate 4.5 Kenya 3.9 2.4* 9.2 (6.8) Horticultural crops 12.7 Estate 1.0* Tobacco Malawi 1.5* 19.1 23.7 (4.6) Tea - Smallholder 0.3* Tanzania 2.1 1.1 * 1.9 (0.8) Malawi Smallholder 13.5 - Estate - Burley 15.4 ~ceo Estate 5.5 Estate - Flue- Food Imports Burley b/ 14.1 Sugar cured 10.4 Kenya 6.4 - Flue-cured b/ 9.2 Smallholder 16.9 Sugar Malawi 3.1 Tea b/ - 5.2 Estate 5.3 - Estate Rice 14.7 Tanzania 3.0* sugar b/ 28.1 Dairying Groundnuts c/ -13.2 Smallholder 8.5 - Smallholder -2.7* Food Aid (Total Cereals) Cotton .EJ - -12.5 - Large Farm 0.0* Ground nuts - Kenya 43.1 e/ Rice - Smallholder -7.2 Malawi 28.6- Smallholder 2.8 Cotton Tanzania 23.5 Cotton - Smallholder 1.1* Smallholder 4.9 Tanzania Tanzania Coffee 0.8* Coffee Cloves -2.7* Smallholder 2.3 Tobacco -4.7* Estate -4.1 Cotton c/ -2.3 Tea I '1 Sisal - -5.9 Smallholder 13.7 Cashewnuts s..l -6.8 Estate 1.0 ~ Tea 1.9 Tobacco Smallholder -4.8* Estate -7.5 Sugar 0.8* Cotton Smallholder 1.6~ * Statistically insignificant (all other figures significant at the .05 level). !H See accompanying .. Years and Sources for Table 5." In some cases data are not available for the complete 1970-85 period~ ~ Estate crop. ~ Smallholder crop. ~ Purchases and sales refer to "official" purchases and sales. "Net sales" are sales minus purchases. ~ Started from a very low base duriug 1970 to 1978 and then dramatically increased in 1979. YEARS AND SOURCES FOR TABLE 5 Export Vol uaes Haize Production KENYA 1970-1985 Er:onoaic Surveys KENYA Produdion 1970-1.1184 NCPB tiALAIH 1970-1985 ADI1ARC Purchases « Sales 1970-19ij5 NCPB TANZANIA HAJ:.:n Coffee 1970-1981 1970-1978: Hin. of Agriculture; 1979-1981: lBRD Production 1970-1984 FAO Cloves 1970-1991 IBRD Purchases 1970-1'185 ADHARC Tobacco 1970-1985 1970-1978: tlin. of Agriculture; 1979-1985: HDB Sales 1972-1985 AD MARC Ci>tlon 1970-1981 1970: Hin. of Agriculture; 1971-1975~ HDB; 1976-1981: iBRD Sisal 1970-1981 HOB TANZANIA Cashewnuts 1970-1985 1970-1971: liio, of Agriculture; 1972-1985: HDB Production 1970-1984 FAO Tea 1970-1984 1970-1977: Internat'l Tea Cottittee; 1978-1984: USDA Purchases 1970-1985 HDB Sales J970-l983 KDB Producbon Volu•es food Iaports 1970-1985 World Bank Trade Systea tEPD> and Country Econoaic tleuorandi KENYA Coffee 1970-1985 Kenya Coffee Board Food Aid Tea 1971)-1'185 Kenya Tea Developaent Authority Sugar ......... --Snall holder 1973-1985 Econoai c Surveys KENYA 1970/71-1984/85 FAD, •food Aid in Figures", Dece1ber 1983 v --Estate 1970-1985 Econoai c Surveys Dairying 1970··\9B5 Data coapiled by I. R. McDonald 11ALAN1 1970171-1983/84 FAD, •food Aid in Figures•, Dece11ber 1983 l Rice 1970-1982 Kenya Statistical Abstract Cotton 1970/71-1984/85 Cotton Seed and lint Harketing Board TANZANIA 1970/71-1984/85 FAO, •rood Aid in Figures•, Deceaber 1983 11ALAWI 1970-1985 ADtiARC TANZANIA Coffee 1970-1985 1970-1972: EEC; 1973-1985: l'tDB Tea 1970-1985 11DB Tobacco 1979-1985 HDB Sugar 1970-1985 11DB Cotton 1971-1984 tiDB - 18 - Production of burley tobacco increased by 15.4% annually, flue cured tobacco by 10.4%, tea by 4.5% and sugar by 14.7%. Groundnuts (the only crop grown by smallholders), which was previously a major export generating about 11% of total export earnings in the 1970-73 period, declined to 1.2% in the 1982-85 period. Groundnuts export volumes declined by 13.2% annually and the production of cotton (another smallholder crop) also declined by 12.5% annually. Tanzania has had a much more diversified export base compared to either Kenya or Malawi with coffee contributing 23% to 35% of total export earnings during the 1970 to 1985 period, but with other important exports either stagnating or declining in shares, e.g., cloves and tobacco remained steady at around 10% and 4% of earnings, respectively, but cotton declined from 20% to 14%, sisal from 11% to 6%, cashewnuts from 9% to 4%, etc. Only coffee and tea export volumes grew in Tanzania, but by only 0.8% and 1.9% respectively. Export volumes of all other major crops showed a decline e.g., cloves annually by 2.7%, and tobacco by 4.7% (both of these albeit at statistically insignificant levels), cotton by 2.3%, sisal by 5.9% and cashew nuts by 6.8%. Unlike in Malawi where estate crop production sho\·l~·~d a sharp rise, in Tanzania it declined even more sharply than agricultural exports leading to an increased share of smallholders in export crop production, if only by default. For example, estate coffee production declined by 4.1% annually whereas small- holder production increased by 2.3% annually. Smallholder tea production increased by 13.7 annually albeit from a small base, whereas estate production increased only by 1%. Bot.h smallholder and estate tobacco production declined by 4.8% and 7.5% annually from 1979 on. Sugar production increased only by 0.8%. - 19 - 2. Food Production The comparative performance in food production in the three countries is of interest from the viewpoint of its effect on welfare as well as balance of paymentso However, the relevant data from FAO and the respective ministries of agriculture are relatively less consistent than in the case of export crops. With the exception of Malawi we have relied on the data from the ministries of agriculture, which appear to be internally more consistent, although they are based on subjective reporting systems whose validity is not very certaino We focus on maize as the most important foodcrop in the three countrieso Kenya shows a growth rate of 3.9% annually between 1970 to 1985, compared to 2.1% for Tanzania and only 1.5% for Malawi (Table 5)o While Kenya appears to have done better, other related indicators of maize performance need to be considered to make a firm judgement given the paucity of reliable production data. ~rom this viewpoint, Kenya's food imports show a faster rate of growth (6.4% annually) compared to Tanzania 9 s (3.0%) and Malawi (3.1%)o There are three possible explanations for thiso First, food i~ports volumes are greatly influenced by the incidence of droughtso Kenya's largest imports took place in 1985, i.e~, at the end of the period under consideration whereas Tanzaniaes imports were great in 1974 and 1975. The timing of these increases affects the growth rates.!/ Second 9 food aid data show higher receipts for Tanzania (1,647 million of cereals in the 1970/71 - 1984/85 period), compared 1/ Also, the greater year to year fluctuations in Tanzania are reflected in a higher year to year grow rate average than for the other two countrieso - 20 - to Kenya (1,058 m.ton) and Malawi (only 37,000 tons). Although rates of cereal food aid once again show higher growth rates for Kenya (43.1%) compared to Malawi (28.6) and Tanzania (23.5%), albeit from a small base, Tanzania also received more skim milk powder than Kenya which had a much more thriving domestic dairying industry.!/ Thirdly, higher food imports may be a result of more rapid growth tn domestic demand, which would be a function of urbanization and overall population as well as income growth.~/ Kenya's population growth rate has been higher than Tanzania's or Malawi's. However, its income growth was also greater and more broadbased than the other two countries, supporting Mellor's observation that the 29 most rapidly growing developing countries have experienced the most rapid rate of growth in food imports.l/ This point also applies to the supply of maize offered to official channels by producers. Because much of the growth in Kenya resulted from the settlement of small farmers on formerly European lands, the productivity per ha of maize increased. However, marketed surpluses did not increase commensurately due to increased domestic demand.~/ 1/ Needless to say, Kenya's food import capacity was greater than that of Tanzania or Malawi. Its food imports ranged from 6.5% to 22% of exports, Malawi's between 9.8% and 23.8% and Tanzania's fluctuated from a low of 6.8% to a high of 42% of export earnings in 1175. 2/ Reliable estimates of urbanization have not been available. However, there appears to be no significant difference in the rate of urbanization in the three countries. 3/ J. W. Mellor and B. F. Johnston, "The World Food Equation: Interrelations Among Development, Employment and Food Consumption," Journal of Economic Literature, 22 (June 1984): 531-574. 4/ See Chapter VI, Lele and Meyers, op cit. - 21 - Data on "officially" marketed surpluses and officie1ll maize sales allow determination of the extent of net official sales, which provides additional insights concerning the growth of domestic food production, performance and consumption. In Kenya growth of volumes sold during 1970 to 1985 to the official monopsonist National Cereals and Produce Board (NCPB) were 2o4% (a statistically insignificant growth rate) and maize sales by NCPB were 9.2% annually, or a growth in net sales of 6.8% (see Table 5). In contrast, in Tanzania official maize purchases by the National Milling Corporation (NMC) increased by only 1.1% annually and sales by lo9%~ reflecting a growth rate of net sales of 0.8% annually. In Malawi, on the other hand, ADMARC purchases increased by an unprecedented annual rate of 19.1% annually whereas ADMARC sales increased by 23.7% annually suggesting a growth of net sales of 4.6% annually. It is thus evident that of the three countries Kenya has had the highest rate of growth of net sales, even in the face of the highest growth rate of maize production, suggesting.~ncreased dependen~0 on the market by the Kenyan populationo The growth of ADMARC's maize purchases has baffled many observers in view of the low growth rate of maize production, although they seem to be less puzzling when considered in relation to saleso Some observers have argued that growing official maize purchases reflect an element of distress in the post harvest season resulting from the lack of growth of agricultural incomes and purchasing power among the vast majority of small producers.!£ 1/ See, for instance, "Food Pricing Policies and their Implications on Nutrition," Ministry of Agriculture, undated. - 22 - Critics of Malawi, including government officials, point to the high level of malnutrition and infant mortality -- one of the worst in Africa -- in support of this conclusion.!/ We will examine the role of price and technology policies 1n explaining the likely growth of food production and consumption in these three countries in Section III. D. Agriculture and GDP Growth The three countries have had strikingly different GDP growth records since the mid-1960s, which have been closely related to the growth of their agricultural sectors. Over the 1967-1973 period, GDP at factor cost grew at 7.8% in Kenya, compared to Malawi's 4.6% and Tanzania's 4.4% (see Table 6). Agricultural GDP grew at a high 5.4% in Kenya during the same period mainly due to the growth of smallholder coffee, tea, maize and dairying compared to 2.8% and 2.3% in Malawi and Tanzania, respectively. During 1974-78, GDP growth was similar (at 4.5% and 5.1%) in Kenya and Teinzania as was agricultural GDP growth (4.1% and 4.7% respectively). In Malawi however GDP growth accelerated to 6.4% and, due to the policy of estate expansion detailed below, agriculture GDP was also much higher at: 5.8%. Kenya, however, experienced a robust 4.3% GDP growth rate during 1979-81 after the second oil shock, whereas Malawi's growth rat~ declined to -0.8% and Tanzania's to 1/ "Food Production and the Nutrition Status in Malawi." A paper prep~:red by a study team comprised of members from the Department of Economic Planning and Development, Ministries of Agriculture, Finance, Education and Culture, Health and Community Services, for the Interministerial Symposium on Nutrition and Development held in Mangochi from 31st of July to 2nd of August 1986. '-;z)- Table 6 ~EAL Gn? mtOW'm ftA"ff'!S (a) ( 'j)~'R.C'F.NT) Y'F.AR JCF.NYA MALAWI TAN7.ANtA lqn7-7'3 M(riculture 5.4 2o1J 2o'3 !nrlust:ry 14o0 io3 4oA (Manufacturin~) 14o2 5.'5 1oR (Minin~) 12.8 -~.. 2 Infrastructure 1o7 8.2 9 .. 5 ~ublic Administr. & nefense 10.1 n. 1 1oB Others Ro7 5 .. 0 2o7 GD'P FoCo 7"8 4.6 4o4 lq74-78 A~riculture 4ol 5.8 4.,7 Industry F;.t; 6o7 4ol (Manufacturing) 6o6 6.7 4o7 (lii\'iin~) 6a5 -2o/ Infrastructure 1o3 2.6 3o9 Public Administr. & Defense 6o4 5.1 1lofi Others '3oR 7.6 4o0 l';Y)'P FQ~CI 4o5 6.4 5.1 1q7('j-FJ1 As;rieu1ture lo'> ""'1e9 -loO Industry io1 2.R -~o4 (~nufacturinQ;) 5a5 2.R -10.2 (~ininJJ;) -~L,:J 2o7 !nfl:a.structure 5e9 -2.2 5.,6 ~ublic Administr. & nefl~nse 6.1 8.2 lOoO OthEt'9 6o3 -2.n -0.3 rmp FoCo 4o1 -o.A 1.7 lqAz·. . R4 A~riculture 4o4 5.8 1.8 Industrv 3o0 1.4 -10.0 ('Manufacturin~) 3o0 3.4 -9.9 (Mining) OoO -2o7 Infrastructure lo6 0.8 -5.6 Public Administr~ & Tlefense 4.0 7.7 8.2 Others 4.0 5.1 o.q GDP FoCo 1.7 4.6 1.1 • tndustey is defined as Minin~ (fuel and other metals) and manufacturing a •• Infrastructure is definerl as r~nstruction and Transport and Communication • ..... r.n~ is at factor cost to bE consistent w{th in~ividua1 secrors whose output are at factor cost (a) Avera~e annual ~rowth rates. qource: WOrl~ ~nk, Y.OP for data up to 1q~~ for Kenya and Malawi, and up to 19R2 for Tanzania. lqR4 rlata a~ ohtained from ~ for Kenva an~ Malawi. TTnnuhlished ciata obtained from the 'Rank of Tanzania and other Tanzania Autho~ities for lqR1-R4 data. - 24 - 1.7%. During this period agricultural growth once again mirrored overall growth performance. Kenya's agricultural growth rate was also the best (1.5%) of the three compared to a sharp drop of -3.9% in Malawi (due to a severe drought during the 1979/80 growing season) and -1.0% in Tanzania. By 1982-84, GDP growth had been restored to 4.6% in Malawi, rose to 3.7% in Kenya, but remained only 1.1% in Tanzania. Once again, the agricultural sector growth rates contributed to the GDP growth rates, being 4.4% in Kenya, 5.8% in Malawi but only 1.8% in Tanzania. Several other sectors of the Tanzanian economy also experienced negative growth (-10.0% in industry~ -5.6% in infrastructure, etc.) resulting from the foreign exchange crisis created by stagnant exports. Thus, the gap between the performance of Kenya (and to a lesser extent of Malawi) with that of Tanzania had widened considerably by the beginning of the 1980s with Kenya's performance being the best of the three. E. Macroeconomic Environment Kenya's superior agricultural performance is a result in part of the extent to which Kenya adopted generally moderate macro economic policies. Figure 3 shows the trade weighted real effective exchange rates for Kenya, Malawi and Tanzania using the 1972 base. Malawi and Kenya each avoided overvaluation of their currency through regular adjustment of the nominal rates whereas Tanzania's exchange rate overvaluation increased sharply over time with the index having reached 33.9 in 1985 compared to 98.9 in Kenya and 112.1 !/ in Malawi. Kenya's gross domestic saving rate was also the 1/ The results do not change much if a current 1982 base is used. Figure 3 Figure 4 real effective exchange rates index budgetary deficit conatont -'9tlb percent of gdp ·------- ---------- --· ~ r=--------·--------------· ~- 1~,----------------------------------------------, 110 -t -2 100-~ -3 -4 -5 vo -6 -7 -8 eo -9 -10 70 -11 -12 -1J &0 -14 -15 00 -16 -17 -18 ~~--~--~--~--~--,---,---~--~--~--~--~--~--~~ -19; t • 1 1 r '1 r 1 1 -- r·· r··--1-·-r--r- 1970 1971 1~72 1973 1974 1~75 1976 1g77 1978 t97S 1U50 1981 1982 tg!J 1984 1971 1972 t973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 D ken + mol <> tQI'I D kenya + malawi <> tonzonla Figure 5 Figure 6 over- . .JII budget deficit inflation rates 1 percent of gdp -J (cpl index, percent) 40 - . - - - - - - - - - - - - - - - • ----·-- - - - - - - - .. ~ (/1 -4 ·a---a-. :S5 \ r -5 30 -6 -7 25 -a 20 -9 -10 15 -11 10 -12 5 -13 -14 0 -d.L----------··---- -15 1 ·· 1 r- ·r 1 T · r· ·-,·- --,-· · 1 · -r---.- ---,-·- -,---,-- -5-t I J J I I 1 1 I f I y · - - 1 - - r - - r .. -T-- 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 69 10 11 12 73 74 75 76 11 76 79 so at e2 a.1 84 65 D kenya + malawi <> tanzania D kenya + malawi <> tanzania SOURCES f1gure 3: lfs flllf>, 1985, and DirecHon of Tnde Shtistics (JI'lFJv 1985., figure 4 1- 5; Keny• Shtisticil Abstr11ct U0-741, Central Bureau of Statistics (75-BOJ 1 and ltinistry of finance and Plinning 181-Btl. Fiqure 6: IFS IIIIFJ, 1985. - 26 - highest (20.8% in 1967-73), falling slightly to 18o3% in 1982-84 (Table 7). In Malawi it was 8.2% in 1967-73, rose to 18.3% in 1974-78 and fell to 14.9% in 1982-84. In Tanzania the domestic saving rate fell sharply from 18.2% in 1967-73 to 8.3% in 1982-84. Gross capital £,ormation as a share o£ GDP was 22.3% in Kenya in 1967- 73, rose to 27% in 1979-81 following the coffee boom and declined to 21% in 1982-84. Concommitment with Malawi's higher growth rate in this ?eriod, gross capital formation reached a high of 29.6% in Malawi during 1974-78 (through a combination of increased borrowing and savings), but by the early 1980s it had fallen to a level similar to Kenya's. In Tanzania, the rate remained at approximately 21% from 1967-73 to 1982-84. Kenya's savings and investment behaviour involved broader participation than in Malawi as in the latter case it was closely associated with the growth of estates and other investments financed by implicitly taxing the smallholder sector through a producer price policy (see Section III.B.3). 1. Levels and Sources of Government Expenditures In this section we examine budget deficits (i.e., total revenues exclusive of grants minus total budgetary expenditures exclusive of lending minus repayments) and overall deficits (to include foreign grants and net lending). Kenya's expenditure (less net lending) as a percent of GDP ranged between 21% in 1974 and 1984 to 32% in 1979. However, the avera~e share was very stable and accounted for about 24% except between 1979-81, when it reached 30% (see Table 8). -·_.*::?~,­ Table 7 INDICATORS OF AVERAGE ECONOMIC PERFOE~CE (PERCENT) KENYA TANZANIA MALAWI --------------------·----~--------------~---------------------~~--~~~ 1967-73 Real GDP Growth Rates (gdp m~p.) 8'o 5 5o2 5o2 Growth Rate of ~opula~:ion 3s5 2.7 2o5 Per Capita GDP Growth 5.0 2o5 2.7 Investment to GDP Ratlo 22.3 20.8 20o0 Total Saving Ratio 20o8 l8o2 8o2 Net Exports Ratio =1.• 5 -2.6 -11()8 Current Account Deficit to GDP Ratio -3.,0 -2 .. 9 -llo7 Total Debt to Exports 61 .. 4* 120.6* 148.7* Debt Service to Exports 4.7-11 5.6• 7ell* Fiscal Deficit to GDP Ratio Inflation Rate (cpi) -3.4 -s.o ""'2c4 4e2 a.. s 12()5 Inflatiot.~ Rate (gdp deflator) 3.3 692 5o8 Real Growth Rate of Agri. Output 5.4 2.3 2e8 Real Growth Rate of Manufo Output 14..,2 7o8 5oS Real Growth Rate of Mining 12.8 -6.2 Real Growth Rate of Exports 3.1 3.6 Real Growth Rate of Imports 4.0 6el Borrow1.ng from the Central Bank ~u1 % GDP LaO 3e6 Rural Population as % of Total 89.7 92.8 1974-79 Real GDP Growth Rates (gdp a.p.) 4.7 4e7 6o6 ~rowth Rate of Population 3.6 4·.0 3o4 Per Capita GOP Growth 1.1 0.7 3.2 Investment of GDP Ratio 23.5 20.6 29e6 Total Saving Ratio ZOeO 11.0 l8o3 Net Exports Ratio -3.5 ... 9o6 -llc3 Current Account Deficit to GDP Ratio -6.2 -8.1 ""~8c9 Total Debt to Exports 74o6 187.1 181.7 Debt Service to Exports 6.4 6.6 l2o5 Fiscal Deficit to GDP Ratio -3e6 -7.6 -Oe9 Inflation Rate ((!pi) 16~~0 15.1 8.5 Inflation Rate (gdp defl.-ator) 14,.,5 14~7 10.2 Real Growth Rate of Agriculture Output 4.1 4.7 5$8 Real Growth Rate of Manufacture Output 6.,6 4.7 6.7 Real Gro~rth Rate of Mining· 6e5 -2.7 Real Growth Rate of Expurts 2.0 -6a8 -o.7 Real Growth Rate of Imports 7 .. 8 2.8 6ol Borrowing from the Central Bank a.s % gdp 3.6 8.2 2.9 Rural Populationas % of Tota~ 87.4 90ol 91.6 . . . . ~g­ Table 7 (continued) YEAR KENYA TANZANIA MALAWI 1979-81 Real GDP Growth Rates (gdp m.p.) 4.2 1.8 2.0 Growth Rate of Populat!r~n 5.3 3.2 2.6 Per Capita GDP Growth -1.1 -1.4 -0.6 Investment to GDP Ratio 27c.O 22.2 27 .o Total Saving Ratio 13.2 10.8 13.2 Net Exports Ratio -8.8 -11.4 -13.8 Current Account Deficit to GDP Ratio -10.5 -10o2 -12.7 Total Debt to Exports 1'20.? 261.1 211.4 Debt Service to Exports 14.3 9.4 24.8 Fiscal Deficit to GDP Ratio -4.6 -10.1 -1.0 Inflation Rate (cpi) 11.2 23.2 4.0 Inflation Rate (gdp Deflator) 9.0 21.9 9.1 Re.al Growth Rate of Agriculture Output 1.5 -1.0 -3.9 Real Growth Rate of Manufacture Output 5.5 -10.2 2.8 Real Growth Rate of Mining -8.3 2.7 Real Growth Rate of Export!iJ -1.3 8.5 11~9 Real Growth Rate of Imports -9.9 -8.3 -4.6 Borrowing from the Central Bank as % gdp 6.0 18.6 9.3 Rural Population as % of Total 85.8 88.2 90.5 1982-84 Real GDP Growth Rates (gdp m.p.) 1.8 1.1 3.5 ·Growth Rate of Population 4.1 3.2 3.8 Per Capita GDP Growth -2a3 -2.1 -0.3 Investment to GDP Ratio 21.2 20.1 21.9 Total Saving Ratio 18e3 8.3 14.9 Net Exports Ratio -2.9 ..,11.8 -7.0 Current Account Deficit to GDP Ratio -4.7 -11.9 -3.0 Total Debt to Exports 1)8.8 490.6 257.0 Debt Service to Exports 20.-9 12.7 11.7 Fiscal Deficit to GOP Ratio -3.9 -3.9 -0.7 Inflation Rate (cpi) 14.0 30.6 5.4 Inflation Rate (gdp deflator) 9.6 12.9 9.4 Real Growth Rate of Agriculture Output 4.4 1.8 5.8 Real Growth Rate of Manufacture Output 3.0 -9.9 3.4 Beat Gt'owth Rate ot Mining o.o -2.7 Real Growth Rate of Exports 2.9 -15.8 -0.9 Real Growth Rate of Imports -6.4 -4.1 -0.4 Borrow~ng f~vm the Central Bank as % GOP 10o4 21.9 19.3 Rural ll opulation as % of Total 1 * The A.verage is over 1970 to 1973. Source: World Bank (EDP, GEMs, World Tables), IMF (IFS, 1985). -;2Cf-,- Table 8 Caapositian of Sovern1ent Financial Operitions (Percent of GOP at "arket Pricesi ----------------------~--~-----------------------------------~------------------------------------------------------- 1971-73 1974-79 1979-81 1992-86 --------------••-••-•••-----------------------------•-•o-Q•-------·~----•a-----------------------••••••·--~---•a••--- ~enyi l 6overn1ent Expenditure 24.8 24.1 30.4 24.5 2 6overntent Revenue 1'1.3 18.8 23.9 22.6 3 Budget•ry Deficit (2-1) -5.5 -5.3 -6.5 -1.9 4 6overntent Expenditure &Net Lending 25 24.3 31.8 28.9 5 Sovern1ent Revenue ~ Foreign Grants 19.5 19.3 24.6 24.2 b Over-all Deficit (5-4) -5.5 -5.1 -7.2 -4.7 7 Financing The Deficit (8+9-101 5.5 5.1 7.2 4a8 8 External Borrowing 2.1 !o 9 3.5 lo2 9 Doaestic BorroMing 3.2 2e9 4.2 3.4 10 Change in Cash Balances ..0.2 -0.3 o.s -0.2 ----Q·-----------------------------------------------------~--~-----~-0~-------------------------------------------- Malawi 1 6overn1ent Expenditure 23.5 22.1 33 30.2 2 Sovern1ent Revenue 14.2 14el 18.9 18.1 3 Budgetary C!ficit (2-1) .. 9.2 -B -14.1 -12.1 4 Sovern1ent Expenditure &Net Lending 25 23.3 35.1 30.3 5, 6overn1ent Revenue &Foreign Grants 15 15.3 22.3 20.8 a Over-all Deficit (5-4> ~to -B -12.8 -9.5 7 Financing The Deficit (8+9-10) 10 a 12.9 9.5 8 External Borrowing 6.9 5 7.4 5.9 9 Do•estic.Borrowing leo 10 7 3.,9 4.3 10 Change in C&sh Balances -1.5 -1.3 ·laO 0.7 G_C_G ______________ a-~------------------------------~----~--~---------~--------------~---~---------------q-~-------~~ Tinzania 1 Soverneent Expenditure 21.1 28 34.,5 38.3 2 Bovernaent Revenue 16.,7 18.7 21.,1 22ol 3 Budgetary Deficit (2-1) .:4.5 -9.3 -13.5 c16a3 4 Boverntent Expenditure ~ Net Lendtng 22.3 28a2 34oS 38.3 5 GovernMent Revenue &Foreign Grints 17 20aO 24oO 26.2 6 Oter-all Deficit (5-4> -5.2 -7.6 -9.9 -12.2 7 Financing The De~icit <8+9-10) 5.3 7.7 9.8 12.1 8 External Borrowing 1. 9 2o7 3o2 1.4 9 Doaestic Borrowing 2 2.6 8.3 9 10 Change in Cash Balances -1.4 •2e4 lo 1 .. 1.7 ---~~----o~-~----~---·-------------------------~~----------~~--------------•--e----------~-~-----~----0-~-------o---- Nates Most af the Data are reprinted in Corresponding CEHs Tanzania Series Start in 1972 and ends in 1983 MalaM1 , series ends in 1985 Kenya,s 1985 and 198b data are fro1 Budget EstiMates Overall Deficit and total Financing aay not aatch exactly,because of rounding Sources For Kenya ,Kenya Stast.Abstract, Central Bureau of Statistics,"inistr.of Finances For Halawi,Public Sector Financial Statistics,Econoaic Reports 1985. For Tanzania, Bureau of Statistics, Econoaic Surveys, Statistical Abstract - 30 - In Malawi, the GDP share of total expenditure (less net lending) ranged from a low of 20% in 1974 to a high of 36% in 1981. In Tanzania, the GOP share of total expendi~ure (less net lending) showed a continuous increase ranging between a low of 20% in 1972 to a peak of 40% in 1982. Expenditure as a share of GDP was the highest in Tanzania followed by Malawi, then Kenya. Also the share grew faster in Tanzania than in Malawi and Kenya. Shares showed a decline after 1982 in all three countries. The share of revenue (exclusive of foreign grants) in GDP does not differ greatly among the three countries. It ranged between 19% in 1974-79 and 24% in 1979-81 in Kenya; between 14% in 1974-79 and 18.9% in 1979-81 in Malawi; and between 16.7% in 1971-73 and 22% in 1982-83 in Tanzania. Both Malawi and Kenya show a very similar pattern in terms of domestic fiscal revenues. Revenues fell between 1974-79, then increased between 1979-81, then fell in subsequent periods. On the other hand, Tanzania showed a steadily increasing trend throughout all these periods. In terms of magnitude, Kenya had the highest revenue/GDP share, followed by Tanzania and then Malawi. The lowest expenditure share combined with the highest revenue share meant that Kenya's "budgetary deficit" was the lowest in terms of share of GDP. Malawi comes second with a relatively higher expenditures share but with a relatively lower rev~nue share. Tanzania showed the highest budgetary deficit in terms of GDP share with the highest expenditure share and relatively higher revenue share. The pattern for the budgetary deficit is the same as the pattern for the revenue share in all the three countries. However, overall deficits, showed different ranking, altering slightly the position of Malawi and Tanzania. In Kenya, the budgetary deficit and overall deficit were very similar up to 1980, after which the overall deficit worsened significantly due to the combination - 31 - of steady foreign grants and increasing net lending (see Figures 4 and S, pe 25). Similar trends were noticeable in Malawi up to 1980. From 1980 on, while the overall deficit worsened in Kenya relative to the budgetary deficit 9 in Malawi the overall deficit improved over the budgetary deficit, due to an increasing GDP share of foreign grants to Malawi~ especially since 1978o In.Tanzania, the overall deficit was smaller than the budgetary deficit throughout the period, due to an increasing share of foreign grants in GDP, with the share of foreign grants in GDP being largest since 1979o Malawi, with its largest overall deficit relied more on foreign borrowing as compared to domestic borrowing to finance its overall deficit while Kenya and Tanzania used relatively more domestic borrowingso 2c Inflation Rates Malawi had the lowest overall inflation rates of 12o5% in 1967-73 9 declining to 5.4% in 1982-84o Tanzania had the highest 9 rising from 8oS% in 1967-73 to 30.6% in 1982-84. Kenya's inflation rates have been intermediate 9 ranging from 4.2% in 1967-73 to 14e0% in 1982-84. While these general patterns obtain overall, there is considerable year to year variation in inflation rates (see Figure 6, po 25). To the extent that the stability of inflation rates is important as well as their level 9 Kenyais rates showed less. fluctuation than did Malawi or Tanzaniao Tanzania's discount rates were the least adjusted to its inflation rates compared to Kenya's or Malawi's 9 although all had negative discount rates. 3. The Effects of External Shocks A comparative examination of the effects of external shocks was undertaken in this study as African~ especially Tanzanian, policymakers have - 32 - emphasized the adverse effects of shocks on their economies. Ansu decomposed the sources of variations in current accounts in terms of shocks (i.e., variation in terms of trade, interest payments and the growth of foreign demand less interest rates) and policy-based changes (changes in market shares, in real GDP, changes in import demand due to expenditure switching as import substitution policies and interest payments due to increased debt). Tanzania's terms of trade index fell the least over the period from 1967 to 1984, while Kenya's fell the most (see Table 9).!/ This may be because Tanzania and to a lesser extent Malawi have more diversified trading partners including African, Asian and OECD countries while Kenya relies only on the latter" It is further noteworthy that Tanzania's pure terms of trade effects are negligible while these are substantial in both Kenya and Malawi. On the other hand policy based changes in market shares were the most unfavourable in Tanzania, followed by Malawi and Kenya, as Tanzania lost market shares to a greater extent than did Malawi and Kenya. All three countries compressed imports, but to a lesser degree in Kenya than in Malawi and Tanzania. This is because the income elasticity of imports is greater in Malawi and Tanzania than in Kenya. The income growth effect contributed the most to the worsening of the current account through increased imports. Import compression, on the other hand, had a positive effect on the current account. 1/ Yaw Ansu, "Macroeconomic Shocks, Policies and Performance: A Comparative ·Study of Kenya, Malawi and Tanzania-- 1967 to 1984". Draft Paper prepared for MADIA Study, July 1986, p. 64. Table 9 DtcOipasition of £xhrnil Shocks and Policies fptrcent of current gdp) ---------·------------------------·-----------------------------------------------------------------------------------------------------------------------------------------------------------; Edernil Shoch ind PolitltS Kfllya I I i'iilitU I Tuzuh . I ·-----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------: Aveuges Averages I I Avengfl 1974-84 1967-73 1974-78 1979-81 1982-84: 1974-84 1967-73 1974-78 1979-81 l982-841 1974-84 1967-73 1974-79 1979-81 1'182-84' V.Jrution an The Current Account -6.1 -1.2 -5.3 -9.8 -3.6 : -9.9 -4.1 -6.3 ··13. 7 -12.0 : -6.9 -4.1 -6.7 -8.2 -5.7 I I ~ Sboch -6.5 -0.6 -5.7 -JO.l -4.4 : -3.7 3.9 0.0 -JI.O -2.5 ' -2.6 0.6 -1.8 -4.4 -2.4 Teru of Tride -9.0 -1.8 -7.~ -12.! -8.5 : -6.9 1.5 -6.0 -ll.t -4.1 -1.1 -0.6 -6.3 -8.6 -1.0 Fortt gn Dl'iind 4.8 2.8 4.1 3.9 5.11 I 4.1 0.9 4.2 4.1 3.8 3.7 0.5 3.9 3.7 3.5 lahrest Rite Pnce EffKt -li.l -0.1 -0.4 -o.t : -0.4 6.0 -1.3 -O.l 0.3 0.3 0.2 0.6 Ml!t hctor lncote Hess Capital lncoae) Policies -2.0 -3.8 -1.6 -3.8 -2.9 -3.9 -1.5 -l.l -1.2 I -4.1 . l : -0.5 -7.2 1.5 -6.2 1.8 -7.9 -2.8 -6.7 -2.0 -6.3 0.4 -10.0 0.7 -6.0 0.3 -12.5 0.4 -7.9 0.5 -7.1 Exports ttarket Share -1.9 -2.0 -1.6 -2.0 I -2.2 -2.4 -2.l -1.8 -5.9 -7.3 -5.0 -4.6 Real lidp Grotfth 3.7 3.8 3.9 J.l 3.9 : 9.4 6.2 9.7 9.1 9.0 8.6 6.0 8.7 8.4 8.6 Chaage 1n laport runctaoo -2.6 0.0 -2.2 -2.4 -3.2 : -5.6 o.o -4.9 -6.1 -6.4 : -5.2 0.0 ...3.9 -6.1 1 Inbrest Payaent Due to Increased Debt -o.a -0.3 -0.9 -1.4 . : -J.2 -7.2 -6.2 -0.1 -1.9 -1.5 -6.8 -1.8 -6.2 : : -0.7 -0.5 -0.6 -6.6 : -1.1 : . I LV Others Fluctuations Around Avg ttarket Share 5.2 0.4 3.2 0.2 5.6 2.0 4.1 -0.2 5.7 -1.1 : l 2.4 -2.1 3.4 5.9 -2.5 I 5.4 1.3 7.5 4.1 3.0 : w 2.4 -0.2 ' 2.6 2.7 1.8 : 0.4 0.1 2.0 0.3 -2.2 : Fluctu.ations Around laport EJastJClty -6.3 0.2 0.2 0.3 -l.b : -0.1 Trsnshr Piyaent 1.5 l. 7 1.2 1.6 1.8 : 0.4 0.6 0.5 -i.B : -0.4 0.5 -0.4 o.o -0.9 I Net Export of Non hdor Services 1.7 -2.1 0.4 4.0 1.3 : 2.5 o.o 2.8 3.0 1.1 : 3.1 1.5 2.6 3.0 4.n : -1.1 0.5 0.9 -o.J Residual o.o I I -7.4 : : 2.0 t.l 2.4 0.9 2.6 : . I lteao -l.li -1.3 -0.6 -0.8 l . I -1.4 -0.3 -1.8 -1.9 -0.7 : ! O.l 0.0 o.o 0.0 1.4 : labdanced Teras of lndtli -1.2 -1).~ -1.9 -l.l 2.0 : -2.7 -0.6 -5.3 -2.0 1.1 : -6.5 -0.2 -6.7 -6.9 -5.6 : Pure Tetl.li of Trade -1.8 -1.2 -5.5 -9.0 -10.5 : -4.2 2.1 -0.1 -9.l -5.2 : -0.6 -0.4 0.5 -1.1 -1.4 : th!t Polley Effectee -Ct.l v.o 0.0 -0.1 -0.2 I 2.2 0.1} 1.8 2.4 2.7 : -1.4 o.o -3.8 0.5 0.9 : ---------------------------------------------------.----------------------------------------------------------------------------------------------------------------------------------------------: Motu Wlttn laports excefd Exports The Current Account could Dehnoratf t Even as the tens of Tradt icproves n Policy effects; l1ss real gdp grouth effect Source : IFS,Uif,l985,191Ua - 34 - Both Kenya and Malawi experienced greater shocks due to changes in interest rates and increased payments due to an increased debt burden, but each of these was greater for Malawi than for Kenya due to Malawi's higher and increased borrowings. Kenya on the other hand was paying the highest interest rates on foreign loans, followed by. Malawi. Tanzania was a distant third with even declining rates due to greater reliance on public loans. Table 9 shows that the contributions of policies have been negative in each of the countries. However, this picture is misleading for the following reasons. In all three countries, the income growth effect contributed the most to the worsening of the current account through increased imports. As Ansu has put it "Certainly no one would argue that it was bad policy to allow real GDP to grow".l/ One can also argue that growth itself is not a policy; it is the result of policies. Hence, to better understand the effects of policy-controlled factors, one has to substract the induced import effect of real growth. Table 9 indicates that in Kenya the net effect of policies (policies less GDP growth) was virtually nil; positive in Malawi, and negative in Tanzania, thus, reinforcing the effects of shocks in the latter. The effects of other shocks such as Tanzania's involvement in the Ugandan war or the breakup of the East African Community were not measured as reliable data on these subjects are difficult to obtain. In any case each country had some unique shocks (e.g., the return of the migrant workers in Malawi in the early 1970s, and the effect on Malawi's transport routes of the Mozambian war -- costing it $50 million dollars or 20% of its export earnings in 1985). l/ Ibid. - 35 - 4c Patterns of Government Expenditures Figures 7a and b show the pattern of capital and recurrent budgetary expenditures of the three governments by sectors over the 1970 to 1986 periodo The share of agriculture in total expenditures was highest in Malawi (ranging between 10% to 16%) and lowest in Tanzaniao Also in Tanzania agriculture's share declined sharply from 11% in the early 1970s to 7% by the end of the 1970s whereas that of the industrial sector increased from 2% in 1972 to 11% in 1980, a result of Tanzania's adoption of the Basic Industrial Strategy. Reflecting a steadier policy environment~ agriculture 9 s share fluctuated between 8% to 12% in Kenyao In Kenya 9 s case the share of the developmental budget going to agriculture increased sharply reaching a peak of 23o3% in 1974 from a base of 13% and declined to a low of 12% in 1985 while showing some fluctuations from year to yearo The share of the recurrent budget going to agriculture had declined from a high of 8% in 1970 to 4o4% in 1979-80 resulting in a shortage of recurrent fiRance in agricultural projects.!/ Recurrent funding problems were also prevalent in Malawi and Tanzania due to a similar switch in shares of recurrent and capital expendituresG Total governmental expenditures (developmental and recurrent) showed major differences in Kenya 9 Tanzania and Malawio Kenya showed the highest share of central government expenditures going to social services (ranging between a low of 25% and a high of 33%)o 1/ The decline in agriculture's share may be explained partly by the trippling share of public debt in rerurrent expenditures by the late 1970s from the 1970 base, going from 6% to 17.8%c Figure 7a -3 b- . " 10 " 10 o4-~--~~~~--r-~-r~--r-~~~~~~~ o+--r~--~~~--~~-r~--r-~~~--r-~-+ t~tmtmt~3t~•m•~•~•m•m•~•~•••~•~•••• t~tm•mm.a•~••m•m•~·~·~·~·~~-~~•••-•• 0 agriculture + transpoJ:t & ¢electricity b, other CJ education +health ()other & livestock communications & wateJ:' kenyo:recurrent expenditure percent _......, kenyo:recurrent expenditure percent •a~------------------------------------~-, ~~--------------~~~-~--------------------~ ~ 10 10 • • • ~· ,. 22 20 ,. • t.O 12 10 •t-.-~~~----~--~~J • 0~-r~--~~-T--~~-r~--~~-r~--~~~ '*'••• .tSJOt~t tm tm tn.s tmtm tm tm ,.,. ,.., 111111 • • • • C agriculture +transport & ()electricity Aother C education +health .()other & livestock communications & water kenya:totol expenditure percent ··~------------~---==~'=·~---==~----------------, ........... kenyo:totol expenditure percent t4 u~--------------------------------------~ 26 1.1 ,., u ft 10 ,.. • , • ,. ., II • tG • • 2 0~~~~~~~--~~-r~--r-~-r~--~~~ •~t~tmtmm•tmt~t~tmtm••••••t•t••••~ tl agriculture + tranaport & <> electrit;.ity A other D education + health ¢'other & livestock communication• & watar Sources: Kenya Statistical Abstract (70-74>, Central Bureau of Statistics (75-80), and Ministry of Finance and Planning (81-86). Figure 7b - ,S ~ ·- _.......... malawi: development e)(pendityre percent .. malawi: development e)(penditure percent _,......, . •• Ill 12 " to • 10 0~~~~~~~~~~~~~==~~~~ o4--;~~~-T--~~~--~~-r~--~~~~ IC'Jtl 1m ltJl'l tm tSf• IWJII II'N Ifni ,.,. !~We fGIIII , . . , . . UIU tiM , . . l~t~t~tmt~IWJIII~I~tmt~l-l_t_t~t~l- 0 agricultun +transport & & livestock communications ¢electricity 4other &. water c education +health '¢othe~ malawi: recurrent expenditure percent 10~------------~~~---~~--------------~ -- ma•:ifil: recurrent expenditure percent s,--r~--------------------------------~ M 24 22 •Ul Ul 10, Ill IG 8 &t-~----~--~----~---.-4--~~._ ·~-.--~~~--~~~--~~~--P-~-.--~~ •~•m•~•~·~~.,.,~,~~.,.,~,-••'-'~'~'• ·~·~·~•m•~•~•~·~·~•m•-,••••~••q~ tl agriculture + transport ¢electricity 4 other & C education +health. <)other & livestock communications & water malawi: total expenditure percent malawi: total expenditure percent ·1-------------~~~~---~----------------~ ............ ~~--------------~----------------------, '" IC " IG ~e 14 t! Ill 10 ICI ·~~--r-~~--r-~-T--r-~-T--r-~~~~ ·~-?--r-~~--~~-T--r-~-T~~~~--~ ,.,. ten 18111 lin ISM 1111'111 1118 te1 tiN 1 - IM 11111 1 - IIIIU II!M tflilll! '~'~•m•~·~·~·.,.,~,••~•-•••••~·~·~ C agriculture +transport & '¢electricity A. other C education +health (>other & livestock communications & water tanzania:totot expenditure percent tonzonia:total expenditure percent .. .-------------~==~1~---==~----------------, 1$,---------------~~~---~------------------, ... ,. If n• •• 10 90 " I! tO 13 tt ,. H Ill It tU ! •4-~~-r--~~--~--~~--~~~-P~ ten! tWS tlf4 1m liM ,_, ,... ,.,. ,.. ,.. IIIII ,. . C agriculture -t trans"Dort & ¢' electricity A othet" C education +health (>other & livestock comaunicationt & water Sources Malawi: Public Sector Financial Statistics 1973, 1975, 1976, 1979, ~ 1982, and Econo1ic Report, 1985o Tanzania: Econo1ic Surveys, Bureau of Statistics - 38 - Commensurate with its poor social indicators (see next section), Malawi had both declining and the lowest shares of expenditures on social services with the levels being well below 20% -- 15% in 1980 and 1981 compared to the high of 25% in 1970. Malawi's expenditures on education were the lowest of the three and Kenya•s the highest. Tanzania's share of social services declined in the latter half of the 1970s from a high of 28% to a low of 21%. In contrast, the share of its defense expenditures rose from a low of 9.7% in 1973 to a high of 24.4% in 1979. It is paradoxical that Tanzania, a large country with low population densities, poor initial infrastructure and population concentrations mostly in the areas bordering on other countries, devoted a smaller share (averaging about 7%) of its resources to transportation and communications compared to Kenya and Malawi.!/ Kenya and Malawi's expenditures on transport and communications have been similar {12% to 14%) although Malawi's rose to a high of 26% between 1976 and 1979. Additional expenditures in Malawi, however, included the construction of the capital city and government buildings, including the State House, which did ~ot benefit the rural population. Tanzania's expenditures on t~ansport and communications declined to !1 Tanzania has a total road network of about 50,000 km. This means it has about 2.9 km of road per 1,000 inhabitants, one of the lowest ratios in Africa. Thirty-five percent of the agricultural traffic goes by rail. This is high by other countries' standards, again reflecting lack of investment in roads. - 39 - 4% to 6% in the late 1970s from a high of 18% in 1972.!/ However even this limited budget reflects the fact that like Malawi, Tanzania devoted resources to non-productive construction of a new capital city in Dodoma. A factor contributing to the growth of expenditures and the shortage of recurrent funding for operating expenses was rapidly increasing public sector employment in all three countriese The growth of public sector employment was, however, the sharpest in Tanzania, i.e. 15.6% annually between 1976-77 and 1977-78. Reflecting the lower priority attached to agriculture relative to other sectors, public sector employment growth in the agricultural sector was nonetheless slower than in other (especially education and industry) sectors. A small fraction of these employees, (10%) were involved in the provision of basic services i1:1 the Minis tr.y of Agriculture, over which the Ministry of Agriculture had lost control subsequent to the decentralization of the Tanzanian administration. Seventy seven percent were working for paratatalso So Indicators of Investments in Human Capital The e~~~nt to which each government allocated resources to the building of human capital has affected rural welfare as well as growth possibilities by increasing capacity to plan and manage the economyo Rapid Tanzania also ranks among the lowest in Africa with respect to the share of public capital resources (13%} allocated to transport (compared to 22% to 25% for Kenya and 31% for Malawi). The private trucking fleet in Tanzania has been shrinking in size. Before 1974 private truckers provided all the commercial freight haulage services~ By 1977 this share had declined to about half of the total market. Truckers left the agricultural industry and concentrated their operations in urban areas, leaving rural areas to parastatal trucking (40% of the total vehicle fleet was in Dar-es-Salaam in late 1970s). - 40 - population growth can erode many welfare gains. It is therefore important to note at the outset that Kenya started out with higher population growth rates (3.8%) between 1967 and 1973 compared to Tanzania (3.2%) and Malawi (2.8%). While population growth rate accelerated in all three countries, in Kenya it reached the highest level -- 4% during the 1973-84 period. Tanzania and Malawi started out with poorer initial conditions than Kenya in 1965 as regards social welfare (see Table 10). Malawi had the lowest life expectancy for males and females compared to Tanzania and Kenya. Kenya continued to hold a lead in female life expectancy over Tanzania and Malawi. However the percentage gain in Tanzania's female life expectancy was greater (20%) compared to Malawi's (15%) and Kenya's (22%). The same applied in the case of male life expectancies, with Kenya holding the lead but with Tanzania showing a slight edge (22%) compared to Kenya (21%), and Malawi showing the least percentage change (16%). With regard to the number of persons relative to physicians, Malawi started out over three times as badly off (46,900 persons/physician in 1965) compared to Kenya (13,450). Tanzania was about one and a half times as badly off (21,840) as Kenya. Malawi's situation, ho~wer, appears to have deteriorated over time (with the persons/physician ratio rising to 52,960 in 1981) while Kenya reduced the persons/physician ratio by half from 13,450 persons to 7,540 during the same period. Tanzania's ratio was 19,810 persons per physician in 1978 (the latest year for which figures are available). Kenya also continued to hold the lead in the spread of primary education, increasing the percent of those of school going age attending school from 40% to an impressive 97% for females and from 69% to 104% for males. Its lead in secondary education, with 4% to 19% of the age group Table 10 SOCIAL INDICATORS Kenya, Halawi &Tanzania :===~===================================================================a======•=====~===================================================~===============================:c:::::::::::::a: Population life E21pettimcy Physicians Nuaber Enrolled in School Safe Mat~ .f I at birth -----------Pr-!~ary-------------- ---Seconduy--- Access Total SroMth Rates Non-urban Feaale Hale -----Feaale------ -------"ale------- 1984 1965-73 1973-84 Pop. as • 1965 1984 1965 1984 1965 1981 1965 1983 1965 1983 1965 1983 1973 1980 (llil.) fl) (I) of Total lyears> fyeusJ {years) (years) (population per) (as 1 of age group) (as t of age group) (1 of populatio~) ~ --------- :-----~-------------------------------------------------------------------------------------------------------------------------------------------------------------------------· :Kenya 20 l.B 4.0 82 ~ 46 56 43 52 I 13450 7540 I 40 97 69 104n 4 19 : 15 26 : :ttahwi 7 2.8 3.1 oo: 40 46 38 44 : 46900 52960 : 32 52 55 73 2 5 : 33 41 : :Tanzania I 21 l.2 3.4 86: 44 53 41 50 : 21840 198101 : 2S 84 qo 9! 2 3 : 13 34 : l=~==============================================·====================================================================================================s========~===c=====================l tl978 (1981 figure not ~v~ilable) I uGross enrollaent ntios !lay exceed 100 percent because so1e pupils are above M belOM shndard pri•ary-school age. ~ 1 Sourcl!& 198b Morld Developaent Reportv World Develi1p~Sent indicators Anne~, Tables 25-ll; ' Social lndicator5 of Development 1~86, Cotparative Analysis and Data Division Econoaic Analysis and Projections Department 0 World Bank Tanzania ientrai Statistical Bureau, Statistical Abstract li978 pop. per physician) - 42 - r.eceiving secondary educatiqn, is even more impressive. Tanzania's gain in primary education (84% of the age group) was the greatest given the low base (40% in primary education) but ,there was no significant movement in the numbers attending secondary education in Tanzania (an increase from 2% to 3%). Malawi made the least progress on the growth of primary education although its relative position on secondary education improved from 2% to 5% bOth over time and relative to Tanzania's. Malawi had the lead in access to safe water (33% of the population) in 1973 over Tanzania (13%) and Kenya (15%). However~ Tanzania's relative gain (an increase to 34%) was the greatest compared to the total coverage " achieved in 1980 in Malawi (41%) or Kenya (26%)a Encouraged partly by the donors, Tanzania adopted a policy of p·r;oviding piped water to the entire rural popl;llation by 1991.. However, in / 1984 half of the rural water supplies provided were either inoperative or operating at reduced capacity.l/ It is thus not possible to indicate the actual extent of rural water access. II. LEVELS OF FOREIGN ASSISTANCE Judging the appropriateness of Official Development Assistance (ODA) levels is not easy. Nor is it easy to sort out the degree to which overall ODA levels have contributed to agricultural growth. ODA levels may be considered on a "need" basis, in which case it could be argued that Kenya {with its higher per capita income than Tanzania's and Malawi's) may "need" 1/ Marian Radetzki, ''Swedish Aid to Kenya and Tanzania: Its Impact on Rural Development". Paper prepared for MADIA Study, August 1986e - 43 - less ODA than the other two countrieso On the other hand, if demonstrated overall economic performance is the criterion, then Kenya would clearly seem to have been more qualified for higher ODA levels than Tanzania and perhaps Malawio In actuality, Tanzania received substantially higher ODA, in both absolute ($669.0 million in 1981) and per capita terms (a peak of $35 in 1981) than either Kenya (a peak of $428.3 million in 1981 and $26 per capita) or Malawi (a peak of $140o3 million in 1979 Ad $24 per capita) (see Figures 8a and 8b)o Although Tanzania's ODA peaked in 1981 once a poor commitment to policy reform was noted by donors it nevertheless remained higher in absolute and per capita terms ($25) in 1984 than Kenya's ($21) and Malawi's ($23)o Kenya's showed a significant rise from 1977 to 1982, declining thereafter. ODA is also given by donors so as to induce policy changes, an approach initiated in the early 19&0so On this basis, the country with the least distorted policies, Kenya, should have received less assistance than Tanzania or Malawi, because Kenya's performance had the least need to improveo In fact, however, since Kenya and Malawi appeared to be more willing to undertake policy reforms than Tanzania, they received greater structural adjustment support than Tanzaniao Thus, ''non-project lending'' for the period 1980-86 was 24% and 37% of the Bank's total portfolio for Kenya and Malawi, respectively, compared to 10% for Tanzania. As of June 1986 Malawi had received three structural adjustment loans with amounts of $170 million by the Bank, and Kenya three structural and/or sector loans amounting to $245.9 million whereas Tanzania received no funding from the Bank for projects in the agricultural sector from 1982 nor for SALs until 1986. Overall, ODA constituted a smaller share of GDP and government expenditures in Kenya than in Malawi or Tanzaniaa Between 1970 and 1984, ODA Figure Sa -- 1/~-- Total ODA: 1970-84 tn Conatant 1183 SUe& 1870 1871 1172 1173 1974 1175 1176 1SJ77 1978 1171 1180 1H1 1182 15283 1H4 c l(enya + Malawt ~ Tanzania Figure 8b Per Capita ODA: 1970-84 fn 1183 tue. 1870-M 1170 1971 1972 1173 1974 1176 1171 1117 1971 1979 1180 1911 1182 1183 1184 Source: OECD, 'Geographical Distribution of Financial Flows•, various years. - 45 - averaged 5.3% of Kenya's GDP (21.9% of government expenditures). In Malawi and Tanzania, ODA was 9.6% and 9.7%~ respectively, of GDP~ (43.8% and 31.9% of their respective government expenditures). The Bank's share in ODA could be a reflection of its influence in two quite different ways. A sntaller share might mean lower financial influence but great need for aid coordination. It might, however, also mark the fact that through co-financing of projects or programs ~ith other donors the Bank could influence project level investments or sector level policies and institutions. ODA was a small 9.5% in Kenya during 1970-84, compared to nearly 20% in Malawi and Tanzania (see Figure 9). The share of Bank lending (as distinct from IDA credits) to Kenya was larger than in Tanzania and Malawi. The Bank's 19.4% share in net resource transfers (TRN) to Kenya was similar to the 20.3% in Malawi; TRN to Tanzania was only 11.6%. The Bank's influence on Kenyan policies and investment allocations has been distinctly more significant than that suggested by ODA levels, owing to three factors: (i) other donors' perceptions of the Bank's professionalism in the field and status as an international institution -- which have tended to give its presence and advice to the government more weight than its direct contribution to ODA might suggest; (ii) the Bank's extensive practice of co- financing individual projects with other donors;!/ and (iii) its recent active role as an aid coordinator, especially given the growing importance of macro and sector policy reforms. The Bank's influence has, however, been less strong in Kenya than in Malawi where the sources of assistance are less diversified. World Bank and 1/ See Lele and Meyers, op cit. i ._..y,-! I ~ ::. 1 ~ I I i I i I ., '* C'll I ~ I C'll i 0 i 0 ;J ~~. ~· ~. f' Sf' 5 t ~ ! a I I I I I I I I ~ ..., ... ,._ 11'1 ::1 .. Cl ·;: . > 11'1 lll Cl i:i: -;;; ~ 10 -+ .... Cll liD Cll ~ .:::: 0\ cu ... = i ~I i ~I - u: Cl c Cl ~· 00 zf _. ..... l ~ t ::1 ~ 0 ..CI i :E ~ U1 i ~ = -;; .:: .t::. .::.. ... "' r:r Cl QJ Sf' i I ~ ! =" Q ~ ....., '-" I I ~ Cl ~ I I ... ::1 1:1 en ! I ~ ~ I I -+ 10 Cll -+ 10 0'1 I I I I ;I ~I i i• .Zi 0 ! I= t < ;c. ffi :s: ffi :s: - 47. - UK assistance together constitute close to half of ODA and over half of TRNo Malawi's debt has also been increasing more rapidly than Kenya's. Also, the UK has tended to align itself with the Bank on major policy issues.!/ The Bank's influence has, however, been stronger in Kenya relative to Tanzania because Tanzania receives over a third of its ODA from the so-called "friendly donors" (Scandinavian countries and the Netherlands) who have been far more tardy in recognizing the adverse consequences of Tanzania's industrialization and other public sector dominated policies on the efficiency of the Tanzanian economy and have been reluctant to tie their assistance to macropolicy adjustments, as have the Bank and USAIDo~/ The Tanzanian government has also been reluctant to call an aid coordination meeting on grounds that the donors would gang up and push for macropolicy reforms that it was not ready to undertakeo The first aid coordination meeting for Tanzania after 9.years was held in Paris in June 1986e Despite greater and more effective attempts at aid coordination by the Bank in recent years, we will argue that failure on this important front continues for a variety of reasons: these include (1) the lack of an overall agreed upon long tP-rm strategy of development for each country; (2) the lack of donors' willingness to focus on those aspects of assistance which they have the greatest comparative advantage to address, within the confines of the limitations posed by the recipient countries' absorptive capacities; (3) the !1 John Howell, "UK Agricultural Aid to Kenya and Malawi". Paper prepared for MADIA Study. 2/ See M. Radetzki op cit; Ec Hanak and Me Loft,"Dsnish Development Assistance to Tanzania and Kenya, 1962-1984: Its Importance to Agricultural Development"; B., F. Johnston and A. Hoben, et al, nAn Assessment of A.I.D. Activities to Promote Agricultural and Rural Development in Sub-Saharan Africa". Papers prepared for MADIA Study. (Drafts). - 48 - limitations posed by the recipient countries' absorptive capacities; (3) the lack of flexibility among donors in providing assistance to countries in the areas of the countries' greatest need due to aid tying and assorted pressures from domestic constituencies; and (4) inadequate emphasis in the recipient countries on maximizing aid effectiveness by directing aid toward alleviating their most important long-term developmental constraints, rather than for meeting immediate short-term political objectives.~/ II!o AGRICULTURAL POLICIES This section examines why Malawi has had a high rate of growth of exports, achieved through an estate oriented strategy, and why i t:s smallholders have done poorly. It also examines why Kenya has performed the best in reconciling the objectivep of growth with equity. Tanzania's performance has been poor both in terms of income generation for the poor as well as growth. We will also show why Tanzania's welfare-oriented policies had become unsustainable by the end of the 1970s. Our argument will be that the relative performance of the three countries is not explained by a single policy, nor indeed even by a few key policies that can be "fixed" relati v\~ly quickly. Rather a combination of macroeconomic and sectoral policies, structural factors influencing the mobilization of land, labor and capital and institutional and human capital factors brought to bear on the development and 1/ Support for these arguments is provided in Part II of the paper which focuses on the Bank's policy assessment, advice and lending in the agricult~ral sectors of these countries. .... 49 - application of technology to smallholder agriculture all explain agricultural performance. A. Natural Resource Endowments and Policy Choices Agricultural performance is to a large extent determined by the quality of natural resource endowmentso Land quality differences and regional differences 1n population densities~ however, make intercountry comparisons difficulto In addition, each country uses a different classification system for categorizing land by agricultural potentiate Table 11 presents rough estimates of the ratio of rural population to agricultural land for all three countries. Tanzania clearly has a much larger land resource base than Kenya and Malawi where land pressures are far greaterc Only 26 percent of Kenyavs land (sqc krn.) is classified as agricultural compared to 56% for Tanzania and 38% for Malawi .. Aggregate figures can, however, mask considerable internal variation.. Only 16% of the agricultural land in Kenya consists of Zone I and II ("humid and sub-humid" land), the primary area for production of high value cash crops such as coffee and tea. Another 29 .. 3% consists of Zone III and IV ("semi-humid and transitional") land, suitable for basic cereals production.!/ Finally, 54c7% of all agricultural land consists of Zone V and VI ("semi arid and arid") areas in which farming is mainly limited to 1/ The limited amount of good quality land is reflected in the size of holdings. Thus, the average size of Kenya's lc7 million smallholdings is 2.3 ha.; over three quarters of these holdings are under 2 ha~ By the year 2000, Kenya is projected to have only half an acre of high potential land per person. Table 11 Population and Agricultural Land Kenya Malawi Tanzania Total Land 56,416 9,300 88,360 Agricultural Land ('000 ha) 14,703 3,550 49,100* Ag. Land as % of Total Land 26.1% 38.2% 55.6% 1984 Rural Population (millions) 16.0 6.0 18.4 Rural Pop./Hectare of Agric. Land 1.09 1.69 0.37 Source: Land Areas: Kenya -- Farm Management Handbook of Kenya, Vol. II, as reported in Kenya: Agricultural Research Strategy and Plan: Priorities and Programs, Vol. II, Draft Report, ISNAR, March 1985. Malawi -- Malawi Land Policy Study, 1986, Table 3.2. Tanzania -- Agricultural and Rural Development Sector Study, 1974) Table 23 ( 11 Agricultural Land" and "High Altitude Forest"). Population: International Financial Statistics (IMF), 1985. Rural Population: World Development Report, 1986, and World Bank Annual Report, 1986. * If "Other Woods, Forests" is included, the area for Tanzania rises to 86,760 hectares. - 51 - subsistence production and (non dairy) livestock -- the latter especially in Zone VI. While breakdowns for Tanzania using exactly the same categories are not available, it appears that while high quality land comprises only a small proportion of total land in Kenya, it is more abundant relative to Tanzania's endowment of this type of land.!/ However, Tanzania clearly has proportionately vastly more of what is usually referred to as medium quality land suitable for rainfed annual cropping of various food crops, cotton, tobacco and certain perennials like cashews and sisal (in areas where in the 1950s and 1960s there was impressive growth through expansion of area under cultivation). Kenya has relatively little medium potential land of the quality enjoyed by Tanzania. Malawi has proportionately more medium potential land than Kenya but proportionately considerably less than Tanzania. The important point about land quality in Malawi is that almost all of the best quality land suitable for tea and coffee is already under use and therefore there is virtually no room for smallholder expansion into these areas. There is land available that could be brought under smallholder tobacco production, however. The evidence suggests th~t much of this land, currently under estate control, is greatly underutilized. Thus, the issue with respect to increased smallholder acreage centers on reallocation of underutilized land rather than expansion into new unallocated areasc 1/ The situation is less clear in absolute terms. Tanzania may have almost as many total hectares of high quality land as Kenya with only a slightly larger population to support. - 52 - 1. Smallholder Intensification As indicated in the Introduction, intensification of agriculture is considered in this paper in three different and interrelated ways: (i) a shift from low to high value crops on any given land; (ii) increase in yields per ha. of any given crop; and (iii) a geographical shift in production of crops from areas of poor land quality to those of higher land quality. In view of the overall shortages of high quality agricultural land in Kenya, both external analysts of Kenyan agriculture and Kenyan policies have emphasized all three types of intensification of agricultural production as principal means of increasing employment and income generating opportunities. In contrast, the Malawi government's concern about smallholder intensification in the sense of shifting from low to high value crops and a geographical shift to high potential areas has been a rela~ively minor until recently. Tanzania has made major strides in geographically diversifying the production of maize, coffee and tea to the previously uninhabited high potential southern highlands from regions of Northern Arusha, Kilimanjaro and Bukoba. But its poor agricultural policies have generally arrested intensification. (In Part II of this paper we will show how donor policy analysis and investments reinforced domestic policy tendencies in Malawi and Tanzania in the 1970s.) B. Small Versus Large Scale (or Estate) Production How production units (small and large) are defined, how factors of production are mobilized, the prices at which those factors are mobilized, the markets in which produce is sold and the prices at which it is sold are all - 53 - policy issues that have major implications for the p~~cess of intensification. These topics are discussed in the sections below. 1. Defining "Large" and "small" Differentiating between small and large holders is important~ but not easy. Apart from the problem caused by differences in land quality~ conventions vary across countries with respect to definitions of small and large farmersQ In Kenya, for instance, the definition of a smallholder holding (with significant implications for access to institutional credit) is one with less than 20 hao Yet three quarters of all smallholder holdings are under 2 ha. Different policy and institutional rules also apply to "small" versus "large"o Thus, there are differences among countries in terms of the way in which the right to cultivate, own or transfer land, produce specific crops, end have access to the markets in which specific crops are sold are conferred by the government and/or traditional authoritieso In Malawi, for instance, estates (regardless of size) are defined in terms of whether cultivation takes place on leasehold or (in a small number of cases) on freehold land or titled land. Customary right to cultivate and transfer the land through lineage, on the other hand, is conferred by the traditional tribal chiefs and this area is designated for smallholders. The term estate infers a large scale farm, but this is not the case in Malawi where many estates are farming hectages similar to those of larger scale smallholders. Further, the size of estates has declined sharply overtime although, initially, the expansion of burley tobacco estates involved the very large farms.. Since then the demand for "estates" has increased dut.: - 54 - to the rent element they confer. It is for this reason that the distinction in nomenclature is significant. Rights to grow export crops such as burley and flue cured tobacco are conferred by the government through the granting of licenses to estates. Estates are also allowed to sell their crops in auctions at close to world market prices. But produce grown on customary land is sold to ADMARC at lower prices determined by the government. Access to input supply, credit and extension is similarly determined by the distinction in nomenclature. In Kenya the right to grow certain crops is also restricted by licenses. Thus due to world market demand concerns, smallholders in Western Kenya were not permitted to grow coffee in the 1970s. However, on the whole titles to land and licenses to grow crops such as tea and coffee have been made far more freely available in Kenya in areas that formerly were European settled or were used as grazing land. For instance, the area under tea estates in Kenya increased from 19.6 thousand to 25.9 thousand ha. or a 1.8% annual growth rate between 1970 and 1985, but that under smallholder tea increased from 4.9 thousand ha. to 48.9 thousand ha. or at 15.3%. The area under coffee estates increased from 29.9 thousand ha~ to 35.7 thousand ha. or 1.3% in the same period compared to that under smallholders from 54.1 to 116.3 thousand ha. at 5~5%. Data for total area under large scale farming for recent years are not available but there appears to be little reason to believe that much new area has been brought under large scale farming. On the contrary considerable numbers of large farms have been broken up over time and there is currently political resistance to expanding tea and coffee production through estates, although a general policy of expanding production of these crops has been adopted. - 55 - In Malawi on the other hand, the area under flue cured tobacco estates increased from 5.8 thousand ha. in 1970-71 to 16.3 thousand ha. in 1984-85, or at 12.1% annually, under burley tobacco from 7o2 thousand ha. to 22.6 thousand ha. or l4o3% annually, and under sugar from 2.6 thousand hectares to 14.9 thousand ha. by 1981-82 or 47.0% annually. It is noteworthy, however, that the area reportedly cultivated under burley is only 6 percent of the area licensed as estate land, reflecting substantial underutilization of land use •.!/ This alienatic,n has been taking place in Malawi despite the fact that overall land pressure is greater than in Kenya. Table 12 shows both the increase in the amount of land under estate control in relatiorr to total available arable land and the negative balance in arable land suggesting that smallholders may be cultivating marginal land. 2. Differential Marketing Institutions We explore in this section the implications of the more diversified and more decentralized nature of the markr.~ting institutions in Kenya and (to a lesser extent) Tanzania compared to Malawi. Both cooperatives and the private sector had played a more active role in the two coun~ries than was true for MalawL, These differences are first outlinedo Then we show that Kenya broadened the scope of its marketing arrangements to increase participation of small farmers, whereas Tanzania's policies created very great institutional instability in marketing institutions contributing to the growth of the parallel market in foodcrops. Malawi continued to have a relatively 1/ See CG Go Ranade, "Agricultural Marketing and Pricing in Malawi." Paper prepared for MADIA Study, January 1986 (Draft). -~6- Table 12 ·~< AVAILABILITY OF LA1ID IN MAU.WI ( ---· T"'T ~IT"' i• -.U L r'- • ) --A. NORTEE!UT CE:::'?.AL SO'CJ'!'HER1T ~.tA.IA~~r~ 1. Total Hectares 2.66 3. 51 3.13 9.30 2. Area Possible for cultiva.'tion un&:er present tee"-hnological 0.89 1. 51 1.15 3.5~ limitations (1) 3e Customar.y land under Agric. (1) 0.49 1.20 '1. 15 2.B4 Area. U..'tder cultivation in Oustomai"J La.nd (2) 0.14 o. 68 o. ;1 1.33 Area under all estates 1979 - 80 (3) 0.05 0.24 0.15 1984 - 85 (4) 0.11 0.44 0.23 6.i Balance of Arable I.a.nd (5) 0.20 -0.28 -0.43 1) Source: Ma.la.~<~i Ae;ricul tura.l Est a. t e Sec.o;or, Devres Inc. 1 1979, :9• 10 2) Sottrce; 1~tional Sample Survey of A~iculture 1980-81, Vol. II, Gover~ent of !~Ta.la.wi, April 1984. p. 3 3) Source: Compiled from (a) Tobacco Control Commlssion and (b) Malawi Statistical year Book, 1981 4) Sov.rce: Compiled from Records a.t !J!inis-::rJ s:"' Ag:-icul ture, Gcverr.ment of r~la.'tri 5) Row~ 2 - 0.10 x Ro~t 2 (from i~rac-:ura:".l.!"-e) - ?..o~·i 3 - R~~i 5 I *This table is from c. G. Ranade, op. cito - 57 - monolithic centralized marketing structure. This allowed cross-subsidization of the smallholder agricultural sector and provided a mechanism for resource transfers from the smallholder to the estate sector. There are some differences and some similarities in the structure and diversity of marketing institutions responsible for various crops in each country. For instance, in all three countries the governments have a de facto monopsony on the purchase of cereals -- all have discouraged the role of the private trade in grain markets (on grounds of curtailing the activities of Asians or other African ethnic groups dominant in trade)o Malawi went the farthest by formally mandating that Asians not reside in smaller towns and rural areaso Donors have however traditionally somewhat mistakenly viewed Malawi as a strictly private sector oriented country. ADMARC 9 the only marketing board in Malawi (until 1973 the Farmers Marketing Board)~ buys all smallholder crops, whereas in Tanzania and Kenya there are separate marketing organizations for each major export crop grown by smallholders. The National Cereals and Produce Board (NCPB) in Kenya and the National Milling Corporation (NMC) in Tanzania have had responsibility for the purchase and sale of cereal crops in the 1970s either directly or through private or cooperative agents. ADMARC's responsibility for buying all smallholder crops enabled it t:o cross subsidize maize producers from the proceeds of implicit taxation on tobacco, a possibility that Tanzania's and Kenya's maize parastatals have not had available (see Table 13)o This partly explains the fact that the NCPB in Kenya and thE~ NMC in Tanzania experienced substantial financial difficulties resulting from year to year fluctuations in official maize purchases and sales -.::>~- Table 13 * ADMARC'S PROFIT.: (LOSSES) ON CROP TRA.:i)I:IG ~B 1~7iL7' ~o 198~,811 (1~ (MILLIONS OF 1280 K'tlACHA) - Y1"'~R 71/72 TOBACCO 11.3 - RICE (0.4) MAIZE (0.1) GROU~IDNUTS 4-3 ·~- COTTON ... GN PD 1. 5 ~ ( 1. 2) --~LOSS) 1;.6 72/73 '7 q I •"" (0.3) (0.,1) 4.4 1.9 (2~7) 10.5 73/74 4·9 0.3 3e0 3o4 2o8 1. 9 12.4 74/75 7·9 o.g 0.4 1. 7 5.1 (2.5) 13.4 75/76 1~.3 (o.o) (4.3) 1..9 0.. 7 0.8 14 .. 3 76/77 21.3 ( 1. 5) (2.2) 3.1 ?,. 6 2.3 30.6 77/78 30.3 ( 1. 0) (2.8) 3.0 1. 7 1.7 33.4 78/79 5.0 (o.s) (4.0) 2.6 1. 4 0.6 s.o 79/80 3.1 (1.4) (4.3) 4.3 0.5 (1. 2) 0.1 80/81 3.2 ( 1. 4) (4. 5) 4.2 (0.3) (o. 4) 0.3 ,., ,., 81/82 7.9 (0.6) (4.4) 2.3 2.4 (0.5) I• I 82/83 13.4 (0.8) (4.1) (0.6) 1.6 (0.3) 8.8 33/811 11. 1 (0 . 4) (.1.6) 0.3 (0.3) (O.Ll.) ...... - 0 I 1) Da.:a. :""rom 1971-72 -o;;hrougr.. 1981-32 is :'!"'J:':l ~Tid. a.::i Christiansen en. ci.!; p 42 2) Data. for 1982-83 and 1983-34 are :rom AJ~~RC'S a::r.ua.l rspor~s *This table is from C. G. Ranade, op. cit. - 59 - in the 1970s. The governments' failure to provide adequate working capital to these agencies contributed to their high interest payments and debts and the subsequent financial difficulties of these boards received considerable donor attentiono In contrast, only when ADMARC's financial difficulties were accentuated by the rise in the maize producer price and the less than expected profits on the tobacco account due to lower world market prices in 1981 did ADMARC's pricing policies receive donor attentiona ADMARC was considered efficient despite the substantial growth in its purchasing centers in the 1970s, which increased its overhead in much the same way as occurred in Tanzania and Kenyao As in Malawi, export crop produce from small and large holders goes through different channels in Kenya (e.go, tea through KTDA and coffee through the coffee marketing cooperatives). Large private estates process and sell their own produce at local auctions and export directly to international markets~ However, in Kenya there is no differential taxation of smallholders and estates for coffee and tea as reflected in the prices received by the two except for those resulting from differences in marketing costs where scale economies are enjoyed by estateso Because the Kenyan marketing. agencies (the coffee cooperatives and KTDA) are generally quite efficient, the marketing margins are low in relation to actual costs. In Malawi on the other hand, the prices received by smallholders are considerably lower than those by estates in the case of tobacco mainly due to the price policy which has involved implicit taxation of tobacco as well as that of cotton and groundnuts. This is shown by the profits and losses made by ADMARC on the various crops as presented in Table 13. The extent to which export marketing arrangements are stable and allow reflection of the world market conditions in the terms received by - 60 - producers also affects incentives for maintaining the quality of the export produce. At Independence, some exports such as coffee and sisal, produced by estates in Kenya and Tanzania, were handled on private account through local auctions (in the case of coffee) and direct sales in Europe. Asian traders handled food crops £or small African farmers and sold rural consumer goods and agricultural implements to them. Marketing and credit co-operatives operated by smallholders were far more active in Kenya and Tanzania than in Malawi. Dairying, pyrethrum and coffee were handled by cooperatives in Kenya, as were cotton, tobacco and coffee in Tanzania. In the case of export crops, especially tea and coffee, Kenya has retained and nurtured its earlier export marketing strategies consisting of local auctions in Mombassa and Nairobi and sales in EuroRean markets and has as well brought a large number of small producers into the process of export marketing. After the breakup of the East African Community and the closure of the Tanzanian border, Tanzania sq£fered from lack of access to the Mombassa tea auctions. In the case of coffee, Tanzania's bilateral sales outside the quota market increased, in part reflecting a poorer quality product. The quality of Tanzanian tobacco and cotton has also deteriorated due to poor handling and processing of the products, and poor export arrangements.!/ 3. Taxation of the Smallholder Sector We have seen that Malawi's exchange rate policies have been superior to those of Tanzania. In Tanzania the exchange rate appreciation explains the ll See 1983 World Bank Tanzania Agricultural· Sector Report. - 61 - implicit taxation of export agriculture resulting in adverse effects on export crop production, as hated earlier. However direct taxation of smallholder agriculture has been prevalent in Malawi through ADMARC producer prices. This issue was overlooked until recently by analysts of Malawi's performance, who explained Malawi's good performance mainly in terms of the outwardness of the economy as reflected in Malawi's exchange rate and trade policies.!/ The prices paid by ADMARC to smallholders for tobacco were as low as 13% to 27% of the international prices throughout the 1970s, whereas the estate producers selling on the auction floor enjoyed prices which were between 250 to 300 percent higher than those received by smallholders (see Table 14).~/ Profits made by ADMARC from paying low prices to smallholders were invested in estate agriculture and in Press Holdings with equity interests by President Banda. Table 15 shows the sharp increase in ADMARC's equity shares in estates from K.072 million in 1972 to K14.9 million in 1984~ and in unsecured loans from K0.5 million to K3.0 million. The sharpest increases was in income notes in Pr.ess holdings during 1983 to 1984 from 0 to K29.2 million. In Tanzania's case, at official exchange rates the nominal protection rates as reflected in the ratios of domestic to international prices were considerably better than those for smallholder tobacco producers in Malawi (see Table 16) but worsened in the 1970s. They recovered their 1970 levels in the early 1980s. However, given that the overvaluation of the Tanzanian 1/ B. Balassa, ''Policy Responses to External Shocks in Sub-Saharan African Countries, 1973-1976," World Bank Reprint Series No. 270. 2/ See C. G. Ranade, op cit. Quality differences in the types of tobacco require that these comparisons of the extent of discrimination be tempered to some extent. However, the general point holds. Table 14 1< Malawi: Average Tobacco Prices Paid to Estate and Smallholder Growers and Ratios of Average Producer Prices to Export Unit Values by Nomenclatures (1970-81) Estate Tobacco Smallholder Tobacco Burley Flue-cured Dark-fired E:q:ort Average Patio of Export Average Ratio of ExpJrt Average .~tig of Unit ~r Q!"'jMSr Unit Gr:wer Gn:v.er Unit Grcwar Gn:w=r Value Price Price to Value Price Price to Value Price Price +:c t/kg t/kg Exp:>rt Unit tlkg tlkg E:q:ort Unit :/kg t/i, "DB (i8-85) - 65 - shilling increased sharply by the end of the 1970s and the early 1980s (see Figure 3)~ the implicit taxation of agriculture had increased greatly in Tanzania. It is also important to consider taxation of export crops relative to that of foodcrops competing in production. In this respect~ while in Tanzania the ratios of cotton and tobacco prices to maize moved sharply in favour of maize by the end of the 1970s and early 1980s (see Table 17), they remained relatively stable in Malawi over the 1970s. In Tanzania maize showed even more favourable ratios if parallel market prices are considered as these prices were between 4 to 6 times as high in the early 1980s as the official priceso In Malawi, on the other hand, the higher incentives for export crops relative to those in Tanzania were eroded partly by the increased cost of purchased inputs (especially fertilizers) as a result of devaluation and removal of fertilizer subsidies (see Section III.G below). The situation has been different in Kenya. Not only has the exchange rate not been greatly overvalued (see Figure 3)~ but in the case of tea and coffee -- Kenya's major exports -- smallholders have received more than 85% of the international prices, net of the costs of handling and processing (see Table 16)o Indeed, since the government is bearing the increased cost resulting from devaluations of the previous international loans of KTDA, rather than recovering them through an increased cess on tea~ smallholder producers of tea are being subsidized, albeit inadvertently.!/ In all three countries partial price elasticities of supply appear to have been quit~ high {perhaps between 1 and 3) in the case of major export 1/ See Chapter II, Lele and Meyers, op cit. Table 17 TREND~ 1/ '~/.IDS OF EXPORT TO FOOD CROP PRICES IN KENYA, TANZANIA, AND MALAWI KENYA TANZANIA Coffee/ Tea/ : Tobacco/ Coffee/ Sr'ndnuts/ Cotton/ : Cotton/ Tobacco/ Cshwnuts1 Coffee/ Maize Maize I 11aize t1aize 11aize Maize I Mai'~ Haize Hiize Maize 1967 6.09 9.79 3.30 2.73 I 1968 ~.30 10.07 3.07 3.23 : 1969 6.83 14.69 3.31 3.38 I 1970 27.2 7.84 11.66 3.31 3.28 I 1971 19.1 19.5 I 7a71 8.03 le03 3,37 I ~.23 22.31 3.46 1972 20.0 15.5 l 7.32 9.90 3.61 2.87 : 4.58 24.17 3.75 18.75 1973 23.7 15.2 : 5.97 9.49 3.51 3.43 I 4.35 21.88 3.46 15.96 1974 21.7 15.5 : 4.86 10.73 3.59 4.34 I 3.42 18.91 2.73 13.33 1975 15.3 11.6 I 6.05 11.19 3.70 3a 77 I 2.73 14.29 1.87 7.00 1976 32.9 13.8 I 5.40 8.75 3.11 2.25 I 2.50 9.66 1.29 10.00 1977 44.7 24.2 : 6.24 9.70 3.39 3.52 I 2.50 10.90 1.33 18.75 1978 31.7 17.8 : 7.80 11.28 3.70 3.94 I 2.71 10.67 1. 31 12.81 1979 36.8 17.6 : 7.88 12.54 5.81 4.19 I 2.82 10.~1 1. 92 10.67 1980 27.6 16.7 I 6.31 0.40 4.60 3.25 : 3.00 8.95 1. 73 11.42 1981 22.6 17.7 I 6.53 7.58 4.65 3.24 I 3.20 9.64 2.75 12.36 1982 25.8 18.0 I 4.03 4.50 2.87 2.45 I 2.47 7.41 3.09 9.93 1983 22.7 14.2 I 7.56 9.35 4.64 3.39 I 2a09 9.96 2.65 8.67 1984 22.0 29.6 I 6.61 8.33 4.89 3.31 l 2.73 7.61 2.95 10.40 1985 21.2 18.0 : 8.11 ERR 5.57 3.56 l 2.10 6.30 2.42 6.75 Sources Kenya: Econo1ic Surveys Tanzania Cotton, Tobacco, and Cashewnuts: HDB Tanzania Coffee: IBRD (72-77} 1 "DB 178-85> - 67 - crops. Kenya's growth of tea and coffee hectarage is at least partly explained by the increase in the international prices of tea and coffee which the farmers received.!/ c. Land Policies In addition to those for land distribution other land policies have also been more supportive of productive smallholder farming in Kenya than in Malawi and Tanzania. The amount of land registered in Kenya increased from 1.75 million ha. in 1970 to 6.5 million ha. in 1983 constituting a quarter of the total cultivable land (i.e., in zones I to IV). The share of smallholders in total registered land was 43% overall but it was well over 80 percent in Western, Nyanza, Central and Eastern provinces, the heart of the smallholder production areas in Kenya. In addition to progress in land registration there is also an active land market in Kenya, although due to differential access to institutional credit and ethnic factors, land access is far from equal (The Bank's credit projects may have facilitated further inequality of land ownership through providing unequal access to credit).~/ In Malawi, on the other hand, there has been very little registration of customary land. Land registration has also not been encouraged in Tanzania. Tanzania formally abolished the traditional tribal village authority, replacing it witb public ownership of land whereby an individual has no right of ownership or sale. Communal land rights nevertheless obtain !/ These were accompanied by investments in the crucial agricultural processing sector for which the government borrowed from the World Bank. See Lele and Meyers, op cit. 2/ See Lele and Meyers, op cit. - 68 - (at least informally) in most parts of Tanzania except in parts of Arusha, Kilimanjaro, Moshi and Irringa and Mbeya where coffee, tea, tobacco and estate agriculture prevailed. The government nationalized many private estates in the 1970s and prevented the development of further private property ownership in land as well as a land market. Tanzania has, moreover, pursued another land policy which has been detrimental to production. Villagization, enforced without the consent of villager~, led by the middle of 1975 to the settlement of over nine million people or about 60% of the population into 6000 villages making "Operation Vijiji" the largest settlement effort in African history. Poor citing of villages and their large size increased the distances villagers had to walk to farms. The intrnduction of communal cultivation following villagization came at a time of severe drought in 1973-74 and was achieved through minimum acreage laws. By the end of the 1970s agroeconomic evidence had begun to accumulate that increasing doses of inorganic fertilizer and the introduction of block-farming would be unlikely to counteract the damage to the environment or reverse the decline in soil fertility being caused by continuous production on fragile soils without a substantial improvement in land resource management. Increased walking distances to production units also increased the cost of fuelwood and caused deforestation and reduction in soil fertility. Deforestation had a major adverse effect on smallholder tobacco and pyrethrum production as obtaining fuelwood for curing these crops had become a greatly more 1abor intensive activity. Government also attempted to promote village wood lots with little response from producers. Attempts at collectivization - 69 - ~ were followed by the local party bosses of TANU and later CCM dictating the amount of acreages that should be allocated to different crops as well as the type of husbandry practices that producers should follow. D. Policies Affecting Labor Markets Labor policies have profound effects on incentives to intensify crop· production especially as export crops tend to be highly labor intensivp. There are major differences in the way labor markets have evolved in each countryo Labor markets (including intra-rural ones) are quite active in Kenya. A minimum wage provides guidelines for rural earnings but does not hinder the growth of labor markets, especi~lly in the smallholder sector. Hired labor accounts for as much as SO% - 60% of smallholde::- tea and coffee employment explained partly by an equivalent extent of urban male migration from smallholder families with females heading up rural households. While cut-migration is high from the semi-arid parts of Kenya, where there are limited production and income generating possibilities, labor markets are surprisingly tight in areas of high value crops~ despite the high growth rates of population and evidence of increases in real rural wages in tea and coffee areas since the early 1970s.!/ The wage increase is, of course, partly explained by structural obstacles to the migration and settlement of populations from low potential areas into the highlands, constraints imposed by ethnic and political barriers. 1/ See Annex to the 1983 Kenya Basi-c Economic Report by P. Collier. - 70 - Malawi's minimum wages similarly do not hinder the growth of labor markets. On the contrary, prevailing discriminatory producer pricing, and the land policies mentioned above that have favored the estate sector, have limited income earning opportunities in the smallholder sector. They appear to have created a highly elastic supply of sm~llholder labor for wage employment in the estate sector, thus facilitating estate growth.l/ Wage employment in the estates is estimated to have increased from 30,000 in 1969 to 148,000 in 1978. The total estimated wage employment in Malawi in 1983 vtas 387,000. This impressive growth still leaves about 60% of the total rural labor force that lives near subsistence existence. Not surprisingly, unlike in Kenya, the real rural wag~ rate in Mala~ri has not increased due to the fact that employment generating possibilities in the smallholder s~ctor, which contain the bulk of the labor force, have been so limited. In Tanz.~nia, active implementation of a minimum wage, restrictions on interregional movements of labor, encouragement of trade unions in the case of estate agriculture, and the ideologically prompted discouragement of the use of hired labor by small ep ~ medium holder export crop producers (to avoid creation of a laboring class) have tended to create artificial labor shortages. This has provided a disincentive for the production of labor intensive crops such as coffee, tea, sisal, tobacco, etc. Indeed~ crops such 1/ See R. E. Christiansen and J. G. Kydd, "The Political Economy of Agricultural Policy Formulation in Malawi, 1960-1985." Paper prepared for MADIA Study, January 1986. (Draft). - 71 - as tea and sisal have suffered from acute labor shortages (see 1983 Tanzanian Agricultural Sector Report).l/ Ec Stability of Agricultural Service Institutions We mentioned earlier that Kenya and Malawi have generally provided a relatively more ~table institutional environment whereas in Tanzania there has been great instability of institutions. Disruption of marketing and processing arrangements has occurred in Tanzania due to many changes in institutional arrangements. These have involved first the discouragement of private trader~ in the early 1970s, then the rapid promotion and the subsequent abolition of cooperative unions (in 1976), then the establishment of crops parastatals followed by their abolition in 1983 and replacement with the cooperatives in the early 1980s, and then the creation of marketing boards with the introduction of some liberalization in agricultural marketing in 1985. In ad<<: .' Jn, Tanzania also decentralized its administration, which greatly reduced the role of the parent technical ministries such as agriculture. This had an especially adverse effect on agriculture as a result of the transfer of responsibility for planning and implementation, including control of the field staff, from the Ministry of Agriculture to the Prime Minister's office. Tanzania's institutional problems have arisen as well from increased public sector control through a multiplicity of institutions. The number of !/ The Amboni Sisal estate and the Tanzania Sisal Authority reportedly had to make do with an aging labor force in 1981 of perLons ranging from 40 to 60 years old. The Bank funded smallholder tea development project similarly suffered greatly in the Bukoboa area where the hiring of migrant labor was discouraged by Governmentc - 72 - parastatals increased from 64 in 1967 to 373 in 1979/80 and over 1000 prices were controlled in 1979. In addition to the rapid growth of employment mentioned earlier, the management of parastatals suffered from ad hoc political interference and commandeering of public resources for party and political objectives, inadequate financial control, shortage of working capital and erosion of assets leading to a virtual lack of agricultural services. There are indications that in recent years Kenya's institutional responses are also becoming politically prompted and thus more unstable and centrally directed. Like Tanzania, Kenya split the ministries of agriculture and livestock in 1980 and then reunited them, and recently shifted the responsibility for grain marketing from KFA (a cooperative of large European and African producers) to the Kenya Grain Grower's Association,!/ a step Tanzania took in the mid-1970s by reducing the role of the Tanzanian Farmer's Associatian (TFA) consisting of large farmers in input marketing by declaring it a uprivate" institution. F. Agricultural Research Kenya and Malawi both have excellent agricultural research systems for their major export crops financed through a cess on the crops, i.eG 9 tea, sugar and tobacco in the case of Malawi and coffee and tea in the case of Kenya. One indicator of the productivity of research systems is the specific innovations they generate. Clonal teas were developed and issued by the Tea Research Foundation in the late 1960s in Kenya and the Coffee Research 1/ This now includes all producers. While this approach is more participatory, it is also more inefficient. .... 73 - Foundation has recently issued a new CBD resistant variety of coffee, Ruiri II. The research systems for foodcrops and those export crops in which the three countries are marginal exporters or importers, e.g., sugar, cotton etc. have, on the other hand, suffered from uncertain general budgetary support, too rapid a pace of indigenization of research management, frequent staff turnover, lack of clear research priorities, too many research stations and fragmented donor support for various low priority activities.l/ Although very weak on adaptive on-farm research, Kenya's hybrid maize program has nevertheless been quite successful in developing an effective improved seed distribution program and facilitating rapid adoption of hybrid or improved maize. Sixty percent of area under maize in Kenya was under improved maize at the end of the 1960s. This cannot be said for the hybrid maize research program in Malawio Malawi's national research system was reorganized in 1986 under the auspices of the National Agricultural Research Project funded by the Bank. A similar reorganization is under active consideration in Kenya under the umbrella of a World Bank funded project involving several donors. Tanzania's research system has been the weakest even for export crops such as tea and coffee. One reason is that Tanzania sufffered from the break-up of the East African community upon which it r.~d depended for research input on export crops prior to the community's break-up when sub-research stations only were 1/ See D. Jha, "Diffusion and Generation of New Agricultural Technology in Africa," Paper Prepared for MADIA Study. June 1986. (Draft); Ko Anthony, "UK Agricultural Research AID to Kenya, Tanzania and Malawi." Paper prepared for MADIA Study. January 1986. (Draft). - 74 - located in Tanzania. Cotton research suffered from the sudden withdrawal of the British Cotton Research Corporation (CRC) in 1975.!/ Tea and tobacco research has similarly been weak in Tanzania due to the shortage of qualified personnel, severe recurrent budgetary shortages, lack of foreign ~xchange for importation of critical supplies and breakdown of the transport system, which ha$ greatly inhibited supervision of field trials. The fragmentation caused by donors is equally prevalent. In 1981, a regional research station in Mbeya supported by the Nordics had a budget larger than the entir~ national agricultural budget. Unfortunately, although the World Bank and the USAID took an early lead in the reorganization of the research system in 1979, due to internal political difficulties Tanzania never made the basic political decisions necessary to act on this effort. The current stagnant or declining crop production and deteriorating quality of marketed produce is at least partly a result of the poor quality of Tanzanian research. G. Fertilizer Policies Increased use of fertilizer along with improved planting material is frequently an important source of growth in factor productivityo Fertilizer supply and pricing policies are therefore of considerable interest in understanding the sources of production and productivity increases. The profitability of fertilizer use is determined by the relative prices of fertilizers and ~rops and by the nature of the production function as !/ See J. Howell, op cit. - 75 - reflected in input/outpu~ coefficients. Fertilizer subsidies often have been considered an acceptable way of promoting fertilizer use in the crucial learning period of early adoption when adoption risks are high. The historical trends in the nutrient price/maize price ratios faced by smallholders in Malawis Kenya and Tanzania are compared in Table 18; ratios for a few selected Indian states are also presented in that table for illustrative purposes. The most striking aspect of Table 18 relates to the contrast between the ratios for Malawi on the one hand, and those for Kenya and Tanzania on the other. Not only are the ratios for Malawi (in the 9-10 range) generally much larger than those for Kenya (4-5 range), but they are also characterized by a substantial increase over time.!/ These differences in ratios are all the more striking in light of the fact that fertilizer response coefficients in the areas most favourable for maize production in Kenya are about 30% higher than the best responses in Malawio Parenthetically, it is also interesting to note that the ratios for the Indian states are similar to those for Kenya and Tanzania. Table 19 compares the maize and nutrient prices used to compute the above price ratios for the three countries.~/ In most years covered by the datas the maize prices prevailing in Kenya (more closely approximating the international prices) were substantially higher than those in Malaw1 at the 1/ Thus, maize smallholders in Malawi have needed to sell 9-10 bags of maize in order to buy one bag of nutrienta Their counterparts in Kenya and Tanzania (or for that matter in the Indian states) have needed to sell only 4-5 bags of maize. 2/ Note that since the fertilizers used on maize vary by country-- ioe., CAN in Malawi, DAP in Kenya and urea plus TSP in Tanzania -- the prices are expressed in nutrient terms to facilitate comparison. Table 18 Nutrient Price/Maize Price Ratios for Malawian, Kenyan and Tanzanian Smallholders, and for Selected Indian States* Selected Indian States (Urea) d/ Malawi a/ Kenya b/ Tanzania c/ Andhra Madhya Uttar (SA/CANT (DAP) -