MADIA DISCUSSION PAPER 4 FI 8325 WtPE&Q RU UX MEEg UMA LELE STEVEN W. STONE MANAGING ______ AGRicuLTURAL\_____ DEVELOPMENTI_____ IN AFRICA ____ 1 FOREWORD The MADIA study and the papers comprising this MADIA Discussion Paper Series are important both for their content and the process of diagnosis and analysis that was used in the conduct of the study. The MADIA research project has been consultative, nonideological, and based on the collection and analysis of a substantial amount of concrete information on specific topics to draw policy lessons; it represents a unique blend of country-oriented analysis with a cross-country perspective. The conclusions of the studies emphasize the fundamental importance of a sound macroeconomic environment for ensuring the broad-based development of agriculture, and at the same time stress the need for achieving several difficult balances: among macroeconomic, sectoral, and location-specific factors that determine the growth of agricultural output; between the development of food and export crops; and between the immediate impact and long-run development of human and institutional capital. The papers also highlight the complementarity of and the need to maintain a balance between the private and public sectors; and further the need to recognize that both price and nonprice incentives are critical to achieving sustainable growth in output. The findings of the MADIA study presented in the papers were discussed at a symposium of senior African and donor policymakers and analysts funded by USAID in June 1989 at Annapolis, Maryland. The participants recommended that donors and African governments should move expeditiously to implement many of the study's valuable lessons. The symposium also concluded that the process used in carrying out the MADIA study must continue if a stronger, more effective consensus among donors and governments is to be achieved on the ways to proceed in resuming broad-based growth in African agriculture. The World Bank is committed to assisting African countries in developing long-term strategies of agricultural development and in translating the MADIA findings into the Bank's operational programs. Stanley Fischer Edward V K. Jaycox Vice President Development Economics Vice President and Chief Economist Africa Regional Office MADIA DISCUSSION PAPER 4 POPULATION PRESSURE, THE ENVIRONMENT AND AGRICULTURAL INTENSIFICATION VARIATIONS ON THE BOSERUP HYPOTHESIS UMA LELE STEVEN W. STONE THE WORLD BANK ____ WASHINGTON, D.C. . I 1lU Copyright © 1989 All rights reserved The International Bank for Reconstruction Manufactured in the United States of America and Development/THE WORLD BANK First printing November 1989 1818 H Street, N.W Washington, D.C. 20433, U.S.A. MADIA Discussion Papers are circulated to encourage discussion and ment, at the address shown in the copyright notice above. The World Bank comment and to communicate the results of the Bank's work quickly to the encourages dissemination of its work and will normally give permission development community: citation and the use of these papers should take promptly and, when the reproduction is for noncommercial purposes, with- account of their provisionai character. Because of the informality and to out asking a fee. Permission to photocopy portions for classroom use is not present the results of research with the least possible delay the manuscript required, though notification of such use having been made will be has not been prepared in accordance with the procedures appropriate to appreciated. formal printed texts, and the World Bank accepts no responsibility for The complete backlist of publications from the World Bank is shown in errors. The findings, interpretations, and conclusions expressed in this the annual Index of Publications, which contains an alphabetical title list and paper are entirely those of the author(s) and should not be attributed in any indexes of subjects, authors, and countries and regions. The latest edition is manner to the World Bank, to its affiliated organizations, or to members of available free of charge from the Publications Sales Unit, Department F The its Board of Executive Directors or the countries they represent. World Bank, 1818 H Street. NW. Washington, D.C. 20433. U.S.A.. or from The material in this publication is copyrighted. Requests for permission Publications, The World Bank, 66. avenue dtlna. 7 51 16 Paris, Prance. to reproduce portions of it should be sent to Director Publications Depart- Uma Lele is the manager of Agricultural Policy in the Africa Technical Department at the World Bank. Steven W. Stone is a doctoral student at Cornell University Library of Congress Cataloging-in-Publication Data Lele, Uma J. Population pressure, the environment, and agricultural intensification: variations on the Boserup hypothesis. (MADIA discussion paper; 4) Includes bibliographical references. 1. Agriculture-Economic aspects-Africa, Eastern. 2. Agriculture and state-Africa, Eastern. 3. Managing Agricultural Development in Africa (Organization) I. Title. II. Series. HD2126.L44 1989 338.1'8676 89-22728 ISBN 0-8213-1320-7 Contents Summary and Policy Recommendations .............................. 5 Introduction ....... ,,...., 7 Agricultural Intensification: What Does It Mean? ....................... 8 Defining Intensification ............................................ 8 Limitations of the Hypothesis ...................................... 9 Present and Projected Land Availability .............................. 11 Introduction ...................... . .,. , 11 Aggregate Land Availability ........................................ 12 Aggregate Population Data ........................................ 14 Estimated Carrying Capacities ..................................... 14 Soil and Rainfall Constraints ....................................... 16 Deforestation .......................................... 19 Land Policy .......................................... 20 Interaction between Population Densities, Cultivable Area, and Land Productivity: Some Empirical Evidence ............................ 21 Distribution of Population on Land ................................. 21 Population Densities in Relation to Quality of Land ...... ............. 21 Population Densities and Incomes ................................. 26 Population Densities and Regional Crop Production ....... ............ 28 Food Crops ............................................ 28 Nonfood Crops ............................................ 31 Population Densities and Regional Public Expenditures ...... ......... 32 Population Densities and Input Use ................................ 34 Conclusion ................ ............................ 38 Annex I: Cameroon ............................................ 39 Annex 2: Tanzania ............................................ 44 Annex 3: Senegal ............................................ 50 Annex 4: Kenya ............................................ 54 Annex 5: Malawi ................. 63 Annex 6: Nigeria ................. 69 Annex 7: MADIA Tables ................. 72 Notes ........ 75 Bibliography ........ 77 Acknowledgements Peter Oram's earlier work has been a source of inspriration for this paper. The paper also benefited greatly from comments by G.M. Higgins, on whose work we have drawn extensively, and from Ridwan Ali, Stephen Carr, Jaya Sankar Shivakumar, T.N. Srinivasan, and Gert Stern. Special thanks are due to members of the MADIA Team-Manmohan Agarwal, Vishva Bindlish, Robert Christiansen, Juan Gaviria, Mathurin Gbetibouo, Kundhavi Kadiresan, Riall Nolan, and Manfred Schulz-who have contributed both intellectually and logistically to this paper. Ilustratioons 18. Official maize purchases in Kenya by region, 1970-83 ....................................... 29 Tables 19. Maize purchases for own consumption in Kenya ..... 29 I. Comparison of Total Land and Arable Land Per 20. Official maize purchases in Tanzania by region, Capita Availability, 1984 .......................... 12 1970-87 .29 2. Comparison of FAO and National Data on "Arable" 21. Official maize purchases in Malawi by region, Land . 13 1970-87 .29 3. Comparison of FAO, IBRD, and National Population 22. Production of estate tobacco in Malawi by region, Data ........................................ 15 1960-85 .31 4. Rates of Growth in Population, 1960-2000 ........... 15 23. Growth in tea production in Kenya by region, 5. Per Capita Land Requirements and Land Availability.. 16 1973-82 .31 6. Fertilizer Response Coefficients for Hybrid Maize in 24. Per capita regional expenditure in Kenya by region, Kenya, Malawi, Tanzania, and Nigeria .............. 16 1969, 1979, and 1983 .32 7. Kenya: Average Yields for Selected Crops by 25. Growth in primary school attendance in Kenya by Province .... 22 region, 1968-84 .33 8. Cameroon: Average Yields in the Traditional Sector, 26. Per capita government expenditures in Nigeria by by Province ...... 23 region, 1981-85 .33 9. Population Densities, Average Crop Yield, and Mean 27. Agricultural wage labor in Malawi by region, 1977-84. .34 Rainfall, by Region in Senegal .................... 24 28. Fertilizer purchases in Kenya by region, 1976-79 ...... 36 10. Average Yields for Selected Crops in Malawi, by 29. Fertilizer consumption in Nigeria by region, 1984 ..... 37 Region ....................................... 25 30. Fertilizer use in Malawi by region, 1981 ............. 37 11. Rural Incomes by Source and Region in Senegal, 1975 ....................................... 27 12. Percentage Distribution of Holdings by Household Acronyms Income Group and Mean Value of Assets per Holding, ADD Agricultural Development Division (Malawi) by Province (1974/75) .............. 28 ADMARC Agricultural Development and Marketing 13. Maize Deficit and Maize Surplus Areas by Province Corporation (Malawi) and District in Kenya .............. 30 ADP Agricultural Development Program 14. Regional Investment as Percent of Total in Senegal, (Nigeria) 1977-84 ... 31 DHS Deloitte, Haskins, and Sells 15. Family, Hired, and Total Labor Working on Farms by DPGA Direction Generale de la Production Province in Cameroon ........................... 35 Agricole (Senegal) 16. Smallholder Land Distribution in Malawi, 1980/81 .... 36 FAO Food and Agriculture Organization of the 17. Population Density, Proportion of Land Cultivated, United Nations and Ratios of Farms Using Purchased Inputs GTZ German Agency for Technical Cooperation in Cameroon ....... 36 IIASA International Institute for Applied Systems 18. Fertilizer Use, Purchased Seeds, and Irrigated Area Analysis In Tanzania by Region, 1980 ................... 37 IBRD International Bank for Reconstruction and Figures Development 1. Regional population growth in Kenya, 1969-79 ....... 10 IFDC International Fertilizer Development 2. "Arable' land per capita in the MADIA countries, rural Center and total populations, 1985 and 2000 .............. 1] IFPRI International Food Policy Research 3. Classification of arable land in Kenya .............. 12 Institute 4. Differences in arable land by country and source .....1 2 KTDA Kenya Tea Development Authority 5. Remaining area frontier in the MADIA countries, 1985 14 MADIA Managing Agricultural Development in 6. Mean annual rainfall in Senegal, 1960-84 ............ 17 Africa 7. Deforestation in the MADIA countries relative to per NCPB National Cereals and Produce Board capita cultivable land . ........................... 19 (Kenya) 8. Distribution of population on total land area ........ 21 NMC National Milling Corporation (Tanzania) 9. Kenya: Per capita high and medium potential land NRDP National Rural Development Program by province ...... 22 (Malawi) 10. Cameroon: Area planted and fallow by region, 1984 . 22 NSO National Statistical Office (Malawi) 11. Population densities in Nigeria by region, 1986 ......23 SAED Soci A nagement et dExploitation 12. Intensity of land use in Senegal by region .......... 24 des Terres du Delta du Fleuve Senegal et 13. Intensity of land use and population densities in de la Faleme (Senegal) Malawi by region ........... 25 SEMRY Soci&6 dExpansion et de Modernisation 14. Population density and per capita agricultural income de la Riziculture de Yagoua (Cameroon) in Cameroon ................................... 26 SODECOTON: Societe de Developpement du Coton du 15. Agricultural income by region in Cameroon ......... 26 Cameroun (Cameroon) 16. Population densities and rural incomes in Malawi .... 27 TCC Tobacco Control Commission (Malawi) 17. Sources of household income in Malawi by region, UNEP United Nations Environment Program 1981 .27 UNFPA United Nations Fund for Population Activities 4 Summary and Policy Recommendations In this paper we explore the relationship among population smallholders as land values increase. The limitations of the densities, agricultural production, land, labor, and rural hypothesis have not been easy to document because of incomes to expand the explanatory base of the Boserup contradictory and inadequate information about such hypothesis, which holds that with increasing population matters as the extent of arable land remaining, but the densities, a corresponding shift to greater agricultural scattered evidence presented in the paper suggests that production and more intensive use of the land takes place the environmental damage caused by deforestation, autonomously through the development of market forces. decline in soil fertility, and retrenchment into subsistence The movement away from traditional area-extensive farming and wage labor may well outweigh the effects of autono- methods is associated in the model with higher levels of mous intensification. The movement against autonomous technology, labor, and capital investment in land. In view of intensification is associated with rapidly declining farm the rapid rates of population growth in Africa and the sizes for the majority and marketed surpluses coming from decreasing frontier, the question arises: "how far can market fewer sources. forces alone induce a productivity-enhancing process of The second, less obvious type of intensification must agricultural intensification in Sub-Saharan Africa, and to therefore extend the Boserup hypothesis to include mea- what extent must it be complemented by an active public sures of output and productivity as well as the frequency of policy to support broad-based agricultural development?" cropping. The process of using an increased role of the The answer is critical to the increasing concern about food state to enhance productivity we call "policy-led intensifica- security and environmental degradation prompted by rapid tion." The paper shows that higher yields, better inputs, population growth on the one hand, and on the other, to and larger incomes for small farmers do not axiomatically the pressure on governments to privatize smallholder follow from higher population densities or more frequent services because of fiscal problems and questions about cropping of the land. Three measures of this latter type of the efficiency of the public sector. To address these issues, intensification are particularly salient. Research indicates the paper surveys existing literature and compiles data at that: the regional level for the six MADIA countries to isolate variables in the equation linking the intensity of land use, I. Shifts to Areas of High Potential (and subsequent the increasing opportunity costs of idle or fallow periods, expansions onto marginal areas) occur spontaneously, the effects of continuous cropping on the soil, and their but are in some cases restricted either explicitly by policy implications. public policy toward land use or by natural or social Two types of intensification are distinguished in the causes. In the MADIA sample, population naturally paper. The first type, identified by Boserup, occurs spon- gravitates toward the most productive land (where taneously as more land is cropped more frequently in returns per hectare are highest), except where disease response to higher population densities. The second and pests pose a significant health problem, or where depends more on policy and incentives for a shift to crops land policy proscribes this type of shift by giving a few of higher value or higher yields, or to more productive land. estates preferential access to land over small farmers (as The spontaneous movement toward better adapted tech- in Malawi) or constrains population movement (as did nology and higher levels of productivity was observed first the Ujamaa policy in Tanzania). In other cases (such as in the development of Europe and Asia, a process we have Kenya), smallholders have recourse to legal ownership, termed "autonomous intensification." This paradigm of but the process of titlement is fraught with unequal demand-led growth has served as the standard model, but access to capital and land, due to ethnic biases, conflict- worsening conditions in Africa are casting doubt on its value ing tenure customs, and registration fees. In situations of as a historical precedent. A combination of apparently more high population densities, the paper documents a fragile African soils, declining rainfall, and historically phenomenon of outward migration to marginal areas unprecedented population growth rates in circumstances of when land in high potential areas is no longer accessible. unequal political power between the mass of smaliholders This type of "regressive intensification," which simply and the privileged few makes the exclusive dependence on amounts to mining nutrients from the soil, is not the market for achieving rapid growth in productivity more sustainable but is becoming pervasive. questionable in Africa. The paper documents several inherent limitations in the original model, e.g., (i) the 2. Shifts to Higher-Yielding Crops by a large number of negative effects of extremely rapid population growth as small farmers are made urgent by population pressure compared to the slowly rising densities envisaged in the but remain dependent on policy. One way of improving hypothesis; [iii) the substantial concentration of population, crop yields is to promote high-yielding varieties of seed even in land-abundant countries; (iii) the conflict between and complementary modern inputs such as fertilizers. social and private gain of large family size at low levels of The extent to which research priorities are tailored to the labor productivity for poor households; (iv) the tendency to needs of small farmers will critically affect whether the "mine" the land for immediate survival versus the social "improved" planting material will have local appeal. If need to protect soils as a productive resource; and (v) new seeds require additional cash inputs, are vulnerable unequal access to land and even expropriation from to drought, do not store, process, taste good, or in any 5 way increase the element of risk in cropping, they will of data on rights to land, its use, potentials, and probably not be adopted even where population density availability. The paper documents that despite mas- is high. Adopting hybrids or using more inputs to boost sive amounts of external aid to Africa for nearly a yields will depend on the degree of farmer confidence in decade and a half, such most basic information is not the market to purchase crop surpluses. The case of widely available: it simply has not been a priority for hyb-id maize in Malawi is one such example. Similarly, in either governments or donors. Such data facilitate Senegal the paper documents a return to planting public debate within each country on the sensitive sorghum and millet, reflecting the farmers' desire for land issue and obviate the tendency for it to become greater food security over potential (but risky) gains from part of highly visible lending conditionality. Bilateral higher-yielding or higher value crops at international donors with lesser perceived power than multilateral market prices. agencies such as the World Bank need to take a lead 3. Shifts to Higher Value Crops depend as above on in this crucial but basic task of helping African farmer confidence in the market, but also on the legal countries to develop and analyze information on land right to grow such crops. Ironically, population density policy by encouraging African scholars to work on the appears to have little bearing on whether governments issue, and by helping to implement an equitable legal encourage or circumscribe smallholder production of framework. cash crops. (Nonfood crops mainly produced for export have in a traditional parlance been called cash crops and 0 Stabilizing production and consumption policies: data for a number of countries is reported as distinguish- Production policy must aim toward rapid, equitable, ing between cash and food crops although food crops and highly participatory growth. That process will are also frequently sold for cash.) Densities are extremely require stable buying and selling prices to increase high in Malawi and low in Tanzania, but each has farmer confidence to grow high value crops and rely pursued policies effectively curbing the supply response on the market to provide food staples. Predictable or of smallholders to export crops. Either they cannot grow reliable incentives and clearly stated national objec- high value crops, or they have until recently had no tives will help farmers to plan ahead and finance incentive to do so. At the other extreme are Kenya and investments in the land and sustain broad-based Cameroon. Although densities run much higher in Kenya, productivity increases. The following means are avail- both have adopted policies enabling the small farmer to able to ensure that end: reap the fruits of higher value crops. These policies include ensuring rural transport, passing along close to * Targeting crop research: introducing seed varieties world prices, and providing a variety of support services that reduce risk and complement traditional farming that enable small farmers to grow these crops. strategies. Integrating soil management techniques, The paper demonstrates how over time the changing such as nitrogen-fixing fodder crops and leguminous demography of a country will alter relative opportunity trees that retain soil and moisture. In land-scarce costs and factor endowments; these changes will be most countries, developing higher-yielding varieties, visibly manifest in the first type of intensification. High on- which may require complementary inputs, that meet land densities, however, do not lead directly to progress in consumer and producer preferences. intensification as defined in this paper. The shift to higher- * Improving rural physical and social infrastructure: yielding and higher value crops and more productive land, especially in high potential areas, investments in as opposed to merely cropping the land more intensively roads, input and output marketing channels, and "mining" soils, requires that changes in factor costs be schools, and Ihealth services will show high returns. reflected in agricultural pricing and marketing, land tenure, They will also be vital in bringing new information in and crop research policy. Three countries in particular- primary life expectancy and encouraging migration Kenya, Malawi, and Cameroon-have provided a stable into lower density but potentially more productive policy environment and performed well, but broad-based areas. growth was achieved only in Kenya, and even there gains in the smallholder sector came mostly through shifts to higher * Accelerating fertilizer use: introducing and main- value crops such as tea rather than improvements in yields taining affordable prices and physical access for per hectare, as was the case in the large farm sector. In smallholders to increase the productivity of scarce circumstances where price distortions are not compensated land in the short and medium term including the for by public initiatives or policies do not facilitate the judicious use of subsidies when necessary. move to intensification, environmental degradation will Although a more holistic and appropriate strategy increase as a very rational response to the conditions of will rely more on locally produced inputs, hybrid rural households. seed may have to be accompanied by other chem- The paper finds that the most direct means of addressing ical inputs, such as herbicides, for adequate returns. the problems of rapid population growth and environmen- * Extending credit: increasing the availability of rural tal stress include among others the following: capital to sma,lholders will facilitate the adoption of * Redefining land policy: The land base and the better tools, seeds, and other inputs. Institutional degree of population depending on it for their credit will be required until rural financial markets livelihood need to be assessed. When left to market develop and rural savings can be mobilized. mechanisms, access to land must be ensured by policy measures to overcome the various constraints * Granting access to export markets: giving small (social, cultural, economic) to equal access. Land farmers the right to grow high value crops and the policy must be complemented by a detailed inventory means to market them. 6 * Rethinking population policy: In absence of the use of modern inputs, and without investment in above, population policy by itself may be incapable of human capital (education, health, water) that increases reigning in problems of food security. In addition to labor productivity and life expectancy, this conflict the above, governments must think about reevaluating between private and social gain will not be reduced a laissez-faire approach to population growth given the and population growth will continue unabated. conflict between private household and social gains. Although traditionally considered land-surplus, weak agricultural performance and accelerating rates of Failure to address these crucial policy areas will lead to population growth in Africa are making international increasing stress on the environment. Neglect in one policy donors and a growing number of African policymakers area will not remain isolated, but will because of interde- question the benefits of high population growth. But pendence between the environment, agriculture, and without pursuing policies that increase household economic performance impact with negative repercussions labor productivity, which among other things includes in other sectors of the economy. Introduction The interaction between population growth, the environ- and policy attention on areas most responsive to chemical ment, and agricultural intensification raises the most fertilizers and improved seed (see also Lele, Christiansen, compelling and most controversial issues currently facing and Kadiresan 19891. Raising agricultural productivity in such developing countries. Given low initial population densi- areas offers the prospect of achieving quicker relief to the environmen- ties, the benefits of increasing population on agricultural tal problems such as soil depletion and deforestation. The faster the development have been widely documented (Boserup 1961, improvement in factor productivity, the smaller the propor- 1981; Ruthenberg 1982); these authors have argued that tion of land and population needed for employment in slowly increasing population densities have desirable agriculture to feed the total population and the greater the effects on technical change, land and labor productivity, and possibility that increased area can be left fallow or refor- rural per capita incomes through changes in relative factor ested. Given the higher rates of population growth and the prices. Others have pointed out that while high population absence of options to migrate, the movement to enhanced densities may be desirable in stimulating rural markets and productivity will hinge on policy-led initiatives. technological adaptation, rapid population growth is very In the past, political pressures within countries to spread costly to countries at early stages of development (World resources and government services to as much of the Bank 1984). This paper shows that the environmental population as possible after independence led to the damage from the reduction of bush fallow, the more expansion of development projects in virtually all parts of intensive use of land without supplementary biological and the countries. The equity and food security concerns of the chemical inputs, and the depletion of forestry resources donor community in the 1970s also led it to support complicates the transition from low to more densely development projects in areas of marginal physical poten- populated areas as originally envisaged in the Boserup tial and indeed even in the areas of medium potential in hypothesis. support of subsistence food crops (Lele 1988a). The result Many of the benefits associated with high population for many countries has been a regional redistribution of densities are seen by Boserup, Ruthenberg, and more production but without substantial growth. Thus to effectively lately Binswanger et al. (1986) as being derived mainly address both growth and equity concerns in Africa-while through market forces, with relatively little emphasis on the simultaneously conserving the environment-will require role of public policy. They have described the effect of both active production policies to stimulate growth in areas population densities on agricultural intensification assum- of high potential and consumption and welfare oriented ing a benign or at least policy-neutral environment. This efforts in areas of lower productive potential. Focusing on paper departs from the conventional view and demon- high potential areas alone risks increasing regional inequal- strates that a policy-led approach to intensification is ities, as weak transport networks can prevent markets from critical to maintaining and preserving resources otherwise functioning, integrating effectively, and allowing marginal degraded through more intensive use. it argues that areas to share in the gains. Alternatively, attempts to autonomous intensification, the result of population growth develop areas of lower productive potential, while justifi- on factor scarcity and the freeplay of market forces, is by able on grounds of nation-building and encouraging partic- itself unlikely to achieve the expected gains in per capita ipation of all groups in economic growth, carry implicit agricultural production and rural income. economic costs that must be recognized. Finally, in areas of In the study, the environmental consequences of growing lower potential but high densities (such as in northern population pressure without gains in agricultural productiv- Nigeria) or remote but productive areas (such as the ity in six Sub-Saharan African countries are documented., Southern Highlands of Tanzania) development policies The paper demonstrates that the most pragmatic means of must inevitably be accompanied by a willingness to tolerate achieving rapid growth in agricultural production, employ- slower growth while an appropriately targeted long-term ment, and incomes in circumstances of rapid population strategy is given a chance to work. growth and declining extensive margin is to focus resources 7 Agricultural Intensificatdon: What Does it Mean? Defining Intensification growth on the environment. This occurs when the positive Agricultural intensification is traditionally associated with effects of population growth (as seen in the more intensive changes in land use and fallow periods. Following loosten use of land) are superseded by the detrimental effects of (1962), Ruthenberg (1980) defines the intensity of cultivation, continuous cropping (soil degradation and fertility loss) and among other ways, by measuring the length of fallow deforestation. This is an especially serious problem given periods between plantings.2 Ester Boserup (1965, 1981), the fragile nature of African soils, their dependence on whose work forms a theoretical foundation for the hypothe- vegetative cover for moisture and stability, and the effects sis, also argues that as the population density increases, of continuous cultivation. Recent data show, for example, changes occur in cropping techniques such as first expand- that for each 4,000 kilogram crop of maize produced on a ing the area under cultivation, or when that is no longer hectare, 200 kilograms of nitrogen, 80 kilograms of phos- possible, shortening fallow periods and increasing the labor phate, and 160 kilograms of potassium are removed from input to satisfy the higher demand for food. The theory the soil (Higgins, personal communication). Other agrono- rests on the assumption that the "problem" of population mists, while conceding these general effects, question the pressure gives rise to its own solution; the very scarcity of magnitude of losses being claimed, but few systematic land, by altering factor prices, results in its more intensive studies exist that analyze these long-term effects. It seems use. clear that the role of policy in channeling "autonomous" Two basic concepts integral to the Boserup hypothesis forces and their long-term effects on the environment may are factor substitution and technological change. Rising be understated in Boserup's work. Developing countries opportunity costs of holding land fallow are compensated facing heavy population pressure must adopt a strategy for for by working the land harder, often with decreasing policy-led intensification. This is a particularly serious issue returns from each additional unit of labor. Instead of a in Africa. Not only is the environment more fragile, but the "peak season" for agricultural labor, the shift to intensive capacity of the governments to put together complex and agriculture implies year-round activities such as water finely-tuned packages to meet the diverse needs of a large collection, soil management practices, and staggered crop number of small farmers and achieve marginal improve- production. The surplus generated from more intensely ments in productivity is limited, especially in view of the cultivated land contributes to growth in other sectors lack of a clear consensus on appropriate policy. There is an through linkage effects in infrastructure, markets, credit, and acute need for policies that promote the interests of small services. This view of intensification is further elaborated by farmers to ensure broad participation in economic growth. Binswanger, Pingali, and Bigot (1986) and is consistent with Intensification of agriculture in this paper is therefore the "induced innovation" argument presented by Hayami considered somewhat differently than in the Boserup- and Ruttan (1985), who contend that changes in factor related literature, in that it considers output as well as proportions will lead to conservation of the more scarce changes in the length of fallow period. It can be measured resource-in the case of several MADIA countries, land- in three interrelated ways: a shift from low to high value and to increased use of the abundant resource in produc- crops on any given land; increases in yield per hectare of tion-in this case, labor. any given crop; and a geographical shift in crop production A critical dimension to Boserup's model of intensification from areas of poor potential to those of higher potential. is that the higher population densities increase agricultural Over the period 1960-1987, the three countries experiencing production and off-set the diminishing returns to inputs on the fastest growth in per capita GNP-Cameroon, Kenya, a fixed land base. Thus, even though the regenerative fallow and Malawi-achieved their growth not through increases in cycle that restores organic matter to the soil may initially be productivity, but through shifts to higher value crops (coffee abandoned, savings from higher output can later be and maize in Kenya being exceptions to this rule) (Lele reinvested in land, labor, and tools to keep soil productivity Forthcoming). With less land available for expanding area high and prompt growth in other sectors of the economy. under crops, especially high value crops, more attention The assumption of induced innovation in situations of and hard empirical study will have to be applied to the task extremely high population growth rates, however, may not of raising productivity in the agricultural sector. Research be valid. In the face of high population growth from carried out in the MADIA study indicates that with increas- preexisting high levels of population density, the external- ing population pressure and the movement of people into ities of agricultural research to bring about technical change marginal areas (reducing average yields), an increasing will require an active public role (Lele, Kinsey, and Obeya proportion of land in many countries is being allocated to 1989). food crop production. The number of people dependent on Two important dynamics are simultaneously at work. The the market for food, even in rural areas, is increasing first concerns changes in cropping patterns occurring more rapidly. There are very clear signs of reduced soil fertility or less autonomously in response to population pressure, or and declining rainfall (Lele, Christiansen, and Kadiresan persons per square kilometer. (Cultivable land per capita is 1989). While some question the extent of decline in soil also considered as an indication of population pressure. fertility, the relationship of reduced rainfall to environmen- Available agricultural land per person is often less than tal stress, or the decline in rainfall, they concede that more densities might reveal due to semiarid conditions.) These often than not public policy stands in the way of the shift "pressures" are normally reflected in the frequency with to higher value crops, to increased input use or improved which land is cropped. The second, perhaps unforeseen, resource management that would otherwise occur. In dynamic concerns the damaging effects of rapid population Malawi, for instance, the practice of issuing licenses 8 prohibits smaliholders from producing burley tobacco, a precondition of technological change, but it alone does not lucrative crop on the world market that is reserved exclu- insure that new techniques will be adopted. sively for estate production.3 It is not known, however, how If it is true, as suggested here, that certain types of unique the case of Malawi is. Another constraint is land technical change will occur only when a certain policy. In Kenya and Malawi, having either no access to density of population has been reached, it of course capital or constrained by ethnic and cultural barriers to does not follow, conversely, that this technical change land acquisition, many people are forced onto marginal will occur whenever the demographic prerequisite is land. Finally, initiatives to develop national research present. It has no doubt happened in many cases that capacity, such as the programs in Senegal and Malawi, are a population, faced with a critically increasing density focusing first on investment in large physical capital and was without knowledge of any types of fertilization technical assistance; their emphasis on the substance of techniques. They might then shorten the period of agricultural research issues and on building human capital fallow without any other changes in methods. This resources or developing seeds, fertilizers, and farm man- constellation would typically lead to a decline of agement practices appropriate to specific physical circum- crop yields and sometimes to an exhaustion of land stances and requirements of low income households is resources. The population would then have to face relatively weak. the choice between starvation and migration In defining intensification, the crucial issue is not the (Boserup 1965, p. 41). (emphasis added) frequency of cropping, which with population pressure Whereas in the process of autonomous intensification that appears inevitable. That frequency is instead only one of occurred in Europe and Asia the option to migrate was many determinant variables, which include the choice of more widely available, especially to overseas colonies, in crops actually grown, the quality of land designated for the case of many countries in Africa it seems no longer a cultivation, the permission to grow high value crops, and viable solution. It is for this reason that agriculture and the size of output land the market where it can be sold). research policy must concern itself with the environment. limitations of the Hypothesis While acknowledging their intellectual debt to Boserup, In this section we briefly outline the first dynamic inherent Pingali and Binswanger (1984) also express skepticism about in autonomous changes in cropping patterns outlined in the the situation in Sub-Saharan Africa. They observe that Boserup hypothesis and point out its limitations before farmer-based innovations appear to be "incapable of taking up the second dynamic of environmental damage supporting rapidly rising agricultural populations and/or and its implication for policy. rapidly rising non-agricultural demand for food." Further- The direct bearing of population density on frequency of more, they suggest that "large-scale irrigation systems and Thndse dirt baring ofvpopulation dhensitymont frequncyf, science- and industry-based technical changes must land use is mnore obvious than the movement to higher bcm ao ore ftert fgot narclua levels of technology and more efficient resource use. The become major sources of the rate of growth in agricultural latter phenornenon is induced by what Pingali and Bin- output" (1984, p.2). Science- and industry-based innovations swanger 119841 call "farmer-based innovation." It depicts the include technological and mechanical inputs that can be ,low evolutionar proces of adapting the means of padministered or overseen by the state to speed up the tion to changes in factor costs. As labor and credit, tools natural" process of farmer-based innovations. Even though and other inputs become less costly relative to land, the large-scale irrigation for the most part has turned out to be andothr npus ecoe esscoslyreltie t lnd,th costly and difficult to maintain in Kenya, Nigeria, and farmer will naturally select the cheapest combination of cStly and diffcuhato mani in K en Nigia, an inputs that maximize output and lower his or her opportu- eng, tpintat the sermust testhei in nity costs. However, the process of change is a slow one; exploring, maintaining, and conserving resources is well Europe and Asia had centuries to perfect locally suited taken.4 Elsewhere the MADIA study documents the benefits techniques of intensification to their high-density condi- of promoting small farmer organizations for the develop- tions. The relevant question in the context of Africa is ment of low-cost irrigation as a viable alternative to the more costly large-scale irrigation projects previously whether the catalyzing factor of population is ahead of or desired by governments and donors. However, potential for behind the pace of farmer-based innovation. The question even by irrnments not Hown, potenting reveals a major limitation of the Boserup model. High rates even small-scale Irrigation is not fully known, and existing of population growth in an initially high density area information suggests it to be much more limited in Africa (5 jeopardize the perceived benefits of autonomous percent of total cultivated area in Nigeria is irrigated) than intensification. in Asia (where irrigated area represents 22 percent in India .. the intensification of agriculturemaycompeland 28 percent in Indonesia) iLele and Meyers 1986; Lele, ...the Intensification of agriculture may compel culti- Oyejide, et al....1989;.Lele..1988.). vators and agricultural laborers to work harder and Oyejide, et al. 1989; Lele 1988al. more regularly ... landl facilitates the division of labor "Speeding up" the natural evolution of intensification is and the spread of communications and education.... a complex task, especially for Africa's relatively young state This condition may not be fulfilled in densely bureaucracies. A crucial distinctiorn separating Binswanger peopled communities if rates of population growth and Pingali from Boserup is their gteater attention to policy are high. (13oserup 1965, p. 118i. (emphasis added) and the role the state must play to encourage intensifica- tion. In addition to identifying salient technological priori- Rapid population growth eats away at capital savings and ties for Sub-Saharan Africa, they also warn that: investment and the physical resource base. That this is arguably the case in many parts of Africa constitutes a .. .the transition to these new technologies depends primary reason to try to extend Boserup's original on many factors-the relative cost of labor, capital, hypothesis. and fertilizers; the cost and availability of credit; the Another obvious limitation of the model is revealed when reliability of markets for inputs and output; the access countries are confronted with a diminishing land frontier to spare parts and repair facilities; and the adequacy and none of the expected gains from population growth. of information and training systems (Binswanger and Boserup explains that the high population density is a Pingali 1988, p.84). 9 A successful transition to more intensive (i.e., sustainable) are not.6 In Kenya, for example, the fastest population use of the land thus depends largely on the specifics of growth between 1969 and 1979 occurred within those sectoral policies toward agriculture. Besides population provinces (aside from Nairobi) that were least populated densities, Boserup (1981) introduces other variables into her and least fertile, including districts in the North-Eastern, original formulation of the hypothesis to explain the weak Rift Valley, Eastern, and Coast provinces (see figure 1). The showing of autonomous intensification in Africa; these provinces that grew least had the highest initial densities, include lack of infrastructure, inefficient extension and suggesting a spillover effect into Kenya's less densely marketing, and rural-urban migration. A formal reading of inhabited but more marginal regions. If the intensive margin her previous work suggests that these constraints would be in the high density area yields a lower return than the lifted as population grew and new technologies were extensive margin in the low density area, a resource shift adopted. Our research indicates that they persist and (including population) to the latter area is appropriate, but indeed become compounded with population growth and this has not been empirically established. Given the rising high population densities, especially in circumstances demographic pressure developing in low income areas, the where there is no correspondence between population environment may prove to be the weak link in the chain densities and land quality. binding population densities to autonomous agricultural The second, related dynamic that does not receive intensification. This paper argues that it is the government's enough attention concerns, among other things, changes in role to reinforce that link with appropriate institutional and soil quality as land is cropped more intensively. Boserup policy support. notes that when "the analysis is based on the concept of frequency of cropping, there can be no temptation to regard soil fertility exclusively as a gift of nature bestowed upon certain lands once and for all" (Boserup 1965, p. 15). Figure 1 She argues quite rightly that the soils' structure and Regional population growth in Kenya, 1969-70 nutrient levels will depend not only on initial status but also on the farming techniques selected. However, it is Annual rate almost implicitly assumed that in the transition from 4_ 5; extensive to intensive cultivation the farmer will invest more (Population densities, 1985) in labor (mulching, terracing) to minimize the negative 4. ON effects of continuous cropping.5 23 24 A more realistic assessment may be that in the short-run 3.X - not only does it makes economic sense to "mine" the land 3,0_ but also it may be inevitable. Ruthenberg, for instance, remarks that an agricultural surplus in the industrial 2.5X 2 countries came from the "exploitation of natural resources L in terms of nutrients and humus which were used to feed laborers cheaply to facilitate industrial capital formation" 1.5 _ (1980, p. 12). He argues that whereas the process of soil . mining in the industrialized countries was accompanied by _ the accumulation of a surplus, in developing countries the 0 5L practice is employed merely to maintain current levels of consumption. The natural process of intensification is far 0 ox too slow in relation to the rate of mining, given the rapid North Rift Valley Eastern Coast Central Western Nyanza growth of population. As pointed out earlier, according to Eastern the FAO, such soil mining is occurring in Africa on a large scale, causing much more irreparable damage than would be the case with soils in temperate climates, which tend to Source: Government of Kenya 1981 b. be structurally more sound (Higgins, personal communication). Similarly, shifts onto marginal areas and stagnating overall crop yields may signal that intensification in terms of frequency of cropping is occurring, but that the envi- sioned reinvestments into productive assets (e.g., the land) 10 Present and Projected Land Availability Although the primary focus of this paper concerns policy Figure 2 measures available to tap regional advantage in cropping, an "Arable" land per capita in the MADIA countries, rural and overview of the land constraints facing Kenya, Malawi, total populations, 1985 and 2000 Tanzania, Cameroon, Nigeria, and Senegal will provide an idea of the urgency with which these issues must be Rural Total addressed. In the discussion that follows, for more detailed information on disaggregated data, the reader is referred to Hectares per person of cultivable land the statistical annexes. Introduction 6 In most MADIA countries, population has doubled since the independence period, and will double again shortly after the turn of the century. Yet even though population 5 meroon densities have reached quite high levels in some parts of a Africa-up to 300 persons per square kilometer in the highlands of Kenya, in southern Malawi, and in coastal Nigeria-it is unclear whether reinvestments in land and labor are occurring in compensation. A very important 4 consequence of higher density is that as the share of population depending on the market for food increases, including those moving to marginal land, the internal terms of trade move in favor of food crops (Lele 1988al. In theory 3 growth in food imports should keep up with internal demand growth. However, in practice import capacity tends Tanzania to be limited at early stages of development by slow growth of exports. Policies toward exchange rates and taxation of crops can make a significant difference in the speed at 2 which relative prices between food and export crops move, Senegal but they may not be able to avert this shift altogether. This is especially true when the price of imported food Nigeria increases with devaluation and population growth reduces 1 land productivity while simultaneously increasing the demand for food. As relative prices shift, agricultural Malawi production moves away from traditionally high value export crops, posing a potential problem in the move to intensi- 0 fication as we have defined it. 1985 2000 1985 2000 At present, Kenya, Malawi, Nigeria and, to a lesser extent, Senegal are experiencing substantial population pressure. Note: A similar graph was used by Binswanger and Pingali 11988) using The first three countries constitute 75 percent of the MADIA agroclimatic densities, based on FAO/UNFPA/IIASA figures (Higgins population and 30 percent of the total population in Sub- 1982). The figures presented here are per capita amounts of arable land Saharan Africa. By government definitions, not one of them based strictly on government definitions. For treatment of FAO/UNFPA/ currently has more than three-quarters of a hectare of IIASA findings, see Section 3. cultivable land per person (using total population). By FAO Source: Government data (see Tables 2 and 3). definitions, the per capita amounts are even smaller. Projecting to the year 2000, this figure will fall to less than half a hectare (see Figure 2) and to a miniscule 0.1 hectare and North provinces have already begun to threaten fragile per person in parts of Kenya, Malawi, and Nigeria (see ecologies. The causes manifest in decreasing land availabil- Annexes 4, 5, and 6). Using only rural population to ity are very much more complex than simple growth in calculate the figures improves the ratio of people to land population: they include such politically sensitive issues as (especially in the more urbanized West African countries) the original expropriation of land by colonials and its but does not relieve the demand for food or lessen the subsequent transfer to elites, policy bias toward estate degree of population pressure (see Table 3). agriculture, health factors such as river blindness, tsetse The per capita land figures are deceptive, however, in infestation in more watered areas, and ethnic discrimina- that they mask very important differences in land quality tion.7 These and other factors can prevent the shift to more and in regional concentrations of population. It is note- productive land. Such is the case in Senegal, where the worthy that even in land-surplus countries, population is predominance of the Mandinque tribe in Casamance concentrated on small amounts of land. In Cameroon, for presents an obstacle to Wolof migration from the crowded instance, millions of hectares of well-watered land in the and less productive Groundnut Basin and in the Middle eastern tropical rainforests go unused while population Belt of Nigeria, which has more assured rainfall and greater pressure and declining rainfall in the semiarid Far North fertilizer responses but low densities. This section can only 11 indicate the broad statistical parameters of the problem. Figure 3 Because of these complexities, the regional dimension of Classification of arable land in Kenya policy-led intensification, in terms of where governments have been and should be investing their resources, should 1982 564,162 sq.km. form the substance of policy debate. It makes a big difference whether populations are concentrated in areas of High potential-4% high or low potential, and whether the emphasis is on long- potential or short-term gains. If reaching the most people and l increasing production with the quickest return on invest- ment is the priority, it obviously makes sense to focus Low potential resources on high-density, 'high-potential" areas. By the 14% same token, it may then be essential to have education, employment, income, and consumption policies that pro- tect those in the low potential areas. It is to the exploration of these issues that we now turn. Aggregate Land Availability There are vast differences in the amount of land classified Arid-74% as "arable" in Sub-Saharan Africa, ranging from 26 percent in Kenya to 75 percent in Cameroon and Nigeria. For Kenya, this means discounting the 400,000 square kilometers of land that receives less than 300 mm of rainfall per year and is considered barren (see Figure 3). Apart from variations in climate, these differences also result from methodology in Source: Jaetzold and Schmidt 1982. land classification. Kenya's figure reflects a detailed analysis of agroclimatic growing zones and land potential for the entire country, carried out by the Ministry of Agriculture and the German Agency for Technical Cooperation (GTZ/ Figure 4 Jaetzold and Schmidt 1982). As a result, their estimate of 26 Differences in arable land by country and source percent arable is thought to be quite reliable. In the cases Arable land as percentage of total land in '000 hectares of Nigeria and Cameroon, by contrast, government esti- -__ mates of arable land are based on less extensive analysis * Government 3 C and tend to be more optimistic.8 The more accurate 0 FAO Atlas information on land availability makes it easier to assess 5.1 - el the land constraint on a regional basis, and therefore to s* ; determine where to focus resources for intensification. For Xu many countries in the MADIA sample and elsewhere in I 0 Africa, however, there is very little authoritative information on either land quality or land availability. Effective policy will depend directly on the quality of information about i - land as it becomes more scarce. The extreme diversitv in land quality between countries 0 can be seen by comparing estimates of overall land Kenya Malawi Tanzania Cameroon Nigeria Senegal availability. For example, Table I (and Figure 4) shows that, at any given level of population, there are dramatic Source: National Estimates (Table 2 and 3) and FAO Atlas of African differences between total land and arable land area. Land Agriculture 1986b. unsuitable for cultivation is considered to be quite high at 74 percent of the total in Kenya, 47 percent in Senegal, and 44 percent in both Malawi and Tanzania. cultivated rather than cultivable land (FAO 1986a). Most Second, Table I indicates that using "arable" as a generic governments, however, choose to include forests and term to mean "cultivable" can be misleading. The FAO permanent crops in their definitions as it yields a more (Production Yearbook) definition of arable land excludes generous estimate of arable land. Using the government areas under permanent pastures or permanent crops, figures sheds a more optimistic light on the room left for forests/woodiands, and "other" land, and therefore reflects "extensification" than seems desirable from an environmen- Table 1 Comparison of total land and arable land per capita availability, 1984 (government and FAO Production Yearbook definitions in hectares per person) Item Kenya Malawi Tanzania Cameroon Nigeria Senegal Per Capita Total land 2.79 1.31 4.13 4.59 0.94 3.04 Arable land (government) 0.73 0.73 2.30 3.45 0.71 1.62 Arable land (FAO) 0.09 0.32 0.19 0.58 0.30 0.81 Source: See tables 2 & 3. 12 tal point of view. Clearing forests has direct ecological hectares-or 12 percent of total area-shy of government consequences on the long-term effectiveness of intensifica- figures. Despite the lack of consensus over how much land tion; whether rainforests, tropical cover, or bush trees are is actually available for farming-in Malawi, government removed, the effect is to destabilize soils and render them estimates have ranged from 19 to 56 percent of total more vulnerable to wind and water erosion. The distinction area'0-it is clear that using FAO figures merely increases between forests and potentially cultivable land also under- the urgency for a policy-led intensification, including among scores the crucial omission of environmental sustainability other things the fundamental importance of improving the that needs to be added to the list of Boserup's original land statistics. concerns.9 A final important point must be made before leaving this More recently the FAO published an Atlas of African section. Elsewhere it has been documented ILele and Agriculture (1986), which lists "potentially cultivable" land Meyers 1986; Oram 1987) that gains in agricultural output in figures that appear to share the broader definition used by the past few decades have come from increasing the area government sources (see Table 2). In most cases, the Atlas under cultivation.' I According to the FAO Atlas, up to two- figures are still more cautious about the absolute size of the thirds of total cultivable land for Kenya, Malawi, and Nigeria available land base than are national estimates (see Figure is already in use (see Figure 5). When available, government 4). In Nigeria, the difference between the two estimates is data appear to confirm this trend: in the Groundnut Basin on the order of 20 million hectares, or one-quarter of total in Senegal, for instance, area under crops reached 70 to 80 land area. In Kenya, the new FAO figure is less than half of percent of total cultivable area in 1976, the last year for the national estimate. In Tanzania, the FAO Atlas figure is which such data are available (see Annex 3). Likewise, also 12 million hectares less than the government estimate, government surveys in Malawi indicate that 60 to 70 while in Malawi, the FAO estimate is about I million percent of the declining amount of customary cultivable Table 2 Comparison of FAO and national data on "arable" land (in thousand hectares) East Africa West Africa Land Year Kenya Malawi Tanzania Cameroon Nigeria Senegal Total land area National 1985 56,416b 9,428c 88,366d 46,540d 90,241 f 19,6729 FAO Yearbooka 1984 56,925 9,408 88,604 46,944 91,077 1 9,200 Area under cultivation National 2,577h 3,639i 4,465 6,830k 12,5421 2,612m (as % of total) 5% 39% 5% 15% 14% 13% FAO Yearbook' 1984 2,335 2,345 5,190 6,965 31.035 5,225 (as % of total) 4% 25% 6% 15% 34% 27% FAO Atlas0 1980 4,400 2,500 9,200 7,700 32,300 5,200 (as% of total) 8% 27% 10% 16% 35% 27% "Arable" land FAOP 1985 1,850 2,320 4,130 5,91 0 28,500 5,220 (as % of total) 3% 25% 5% 13% 31% 27% FAO Atlas (potentially cultivable)p 1980 6,700 4,100 36,600 31,500 47,900 9,700 (as % of total) 12% 44% 41% 67% 53% 51% National Arable Estimate 1985 14,703b 5,280r 49,100s 34,905t 67,951u 10,481v (as % of total) 26% 56% 56% 75% 75% 53% Source: FAO 1985, FAO 1986, and National Data. and Land Under Permanent Crops ("land cultivated with crops that a FAO 1986. occupy the land for long periods and need not be replanted after each b By Jaetzold and Schmidt 1982. Arable land estimate includes low harvest.. . but excludes land under wood and timber"). potential land area. °Atlas of African Agriculture 1986. Land Under Cultivation given as c Malawi Population Census 1984. "Annual and Permanent Cropland," for 1980. d Bureau of Statistics 1983. P FAO "Arable Land" (unadjusted) defined as "land under temporary e Ministry of Agriculture, Cameroon 1980. crops, temporary meadows for mowing or pasture, land under market f Federal Ministry of Science and Technology 1985. and kitchen gardens, and land temporarily fallow or lying idle." 9 Direction Statistique 1982. qAtlas of African Agriculture 1986. h Smailholder Land: Central Bureau of Statistics 1981. Large-farm land: 'Compendium of Agricultural Statistics 1977. We use the government Central Bureau of Statistics 1980. estimate of 53 percent cultivable, based on the 1965 land survey. I Mkandawire and Phiri 1987. This is a 1983 estimate. However, the figure is considered optimistic. A more conservative I Bureau of Statistics, Tanzania 1970. estimate of 37 percent is cited in Mkandawire and Phiri 1987. k Cameroon Ministry of Agriculture, 1980. Defines area as "surfaces STanzania Bureau of Statistics 1970. Calculated by subtracting from total mobilizes," under cultivation or temporarily lying fallow. area lands designated as swampland, desert, and urban areas. If "Other I Federal Office of Statistics 1983. Compiled from area under production Woods, Forests" are included, the area for Tanzania rises to 86,760 figures for crops (mostly food crops) for the year 1983. hectares. m Direction Statistique 1982. Land under cultivation defined as "terres tCameroon Ministry of Agriculture 1980. Arable land defined as "surface agricoles: superficies cultives." agricole utile." n FAO 1986. Land Under Cultivation defined as Arable Land ("land under u Lele, Oyejide, et al. 1989. temporary crops, temporary meadows for mowing or pasture, land under vSenegal Direction de la Statistique 1982; and S6nbgal Direction d'Eaux, market and kitchen gardens, and land temporarily fallow or lying idle") Forets et Chasses 1978. Figure includes woodlands. 13 Figure 5 Remaining area frontier in the MADIA countries, 1985 Kenya Cameroon Malawi Nigeria (34.3%) (24.4%) (39.0%) 32.6%) (65.7%) (~~75.6%) 6.% 6.% Tanzania Senegal (4.4%) 46.4%) 53A6%) Land available for use Land under cu'livation Source: FAO Atlas oflAfrican Agriculture 1987. land under the control of the Malawian smallholder popu- all cases since the 1960s, and are projected to continue at lation in the more crowded agricultural development high rates through the year 2000. The question of how to districts (ADDs) of Blantyre and Lilongwe were already effectively channel new demands for land, food, income, under crops in 1985 (see Annex 5). As more area comes and fuel into a productive force for development becomes under crops and is cultivated more frequently, soil degra- all the more pertinent in view of the limitations of the dation and ultimately complete loss of fertility become Boserup model and the inadequacy of autonomous inten- more likely. This is the most compelling evidence for policy- sification to accommodate high rates of growth. led intensification; the area frontier acts more or less like an Urbanization is more advanced in West Africa, where hourglass by which to gauge the time remaining for about one-third of the population lives in cities and towns autonomous intensification. of at least 5,000 persons, but rates of growth in urbanization Aggregate Population Data are much higher in East Africa. 3 For Cameroon, Nigeria, and Senegal, the urban population is projected to increase by Population data (using government estimates) were held 5.7, 5.4, and 4.5 percent a year, respectively, from 1985 to constant in the preceding calculation of per capita arable 2000, whereas in East Africa, the urban populations in land to highlight differences between FAO and government Kenya, Malawi, and Tanzania are expected to grow by 7.0, estimates of land availability at the national level. Nonethe- 8.3 and 7.6 percent a year, respectively (see Annex 7). Very less, wide differences also exist in the population data. little is known about the important subject of rural Ethnicity and population growth are much more explosive migration or the nature of the rural/urban/rural migration issues in some countries than in others and affect statistics in most MADIA countries, with the exception of Kenya and differently. Whereas Kenya and Malawi have routinely Malawi, which have much better data than the other published data on changes in the ethnic composition of the countries. Migration away from agriculture can relieve on- population, a census cannot be conducted in Nigeria land pressure in per capita terms, but if there is no because the publication of such information might spark a technological change in agriculture that increases labor political controversy. Cameroon and Senegal did not productivity, increased urbanization only changes the terms conduct national censuses in the 1960s, and Nigeria has not of trade in favor of the food crop sector (Lele, Oyejide, et conducted a census since 1963.12 The 1963 census figures al. 1989). were themselves considered overinflated, with the result that government and World Bank projections to the year Estimated Carrying Capacities'4 2000 differ by as much as 16 percent, or 23 million persons. The FAO, in coordination with the United Nations Fund The lack of agreement between population estimates is For Population Activities (UNFPA) and the International reflected in Table 3. Institute for Applied Systems Analysis (IIASA), has calcu- Despite the inconsistencies in data, overall high rates of lated the maximum amount of calories that could be population growth-generally about 3 percent a year and produced in each country to determine its carrying capac- around 4 percent in Kenya-leave little room for doubt ity, based on agroclimatic conditions and varying levels of about growing demographic pressure on the resource base. input use.'I The results are necessarily rough, because they Table 4 shows that rates of population growth have risen in rely on a technical rather than a social estimation of ideal 14 Table 3 Comparison of FAO, IBRD and national population data: Initial, present, and projected (in thousands) East Africa West Africa Popullation Year Kenya Malawi Tanzania Cameroon Nigeria Senegal Initial (Census of 1960s) 10,942a 4,040k 12,313m NA 55,670X NA Present Total National (Census of 1970s) 15,327a 5,547k 17,036h 7,761s NA 5,069bb National (current estimate) 1985 20,200b 7,200k 21,3830 10,1 30t 96,1259 6,478c FAOC 1985 20,600 6,944 22,499 9,873 95,198 6,444 IBRDd 1985 20,000 7,000 22,000 10,000 100,000 7,000 Present Rural National 1 6,596e 6,2761 18,389P 6,571 a 67,288z 4,340dd (as % of total) 82% 87% 86% 65% 70% 67% FAO9 1985 16,242 5,440 18,574 6,036 63,484 5,121 (as % of tctal) 79% 78% 83% 61% 67% 79% IBRD9 1985 16,000 6,160 18,920 5,800 70,000 4,480 (as % of total) 80% 88% 86% 58% 70% 64% Projected Total 2000 National 37,505n 11,783k 34,066q 16,682v 140,220Y 10,093ee IBRDI 36,000 11,000 37,000 17,000 163,000 10,000 Projected Rural 2000 National 26,103J 8,837k 25,073' 8,341w 77,121aa 5,955f (as % of total) 70% 75% 74% 50% 55% 59% Source: World Bank 1987; FAO 1986; and National Data. Rural population derived from WDR estimates of 6 percent urbanized in a Kenya Central Bureau of Statistics 1981. 1965 and 14 percent in 1985 and then projecting to 2000. b Kenya Central Bureau of Statistics 1987. s Cameroon Bureau Central du Recensement 1978. c FAO 1986a. t Projected from 1981 at 3.1 percent per year to 1985. Rate of growth d World Bank 1987. cited in Sixth Plan. e Kenya Central Bureau of Statistics 1987b. Calculated from 15.1 percent u Level of urbanization calculated by government to be 35.13 percent in level of urbanization in 1979 and projected level of 30.4 percent for 1985. From Sixth Plan. 2000, to obtain 1985 figure. v 1985 base projected at 3.23 percent growth per year. From Sixth Plan. f FAO 1 986a. Referred to as "Agricultural Population." w World Bank Cameroon Country Economic Memorandum 1987. 9 World Bank 1987, except for Malawi, World Bank 1986. x Nigeria Federal Census Office 1963. h Kenya Central Bureau of Statistics 1985. Y Lele, Oyejide, et al. 1989. Population projected from 1963 base at 2.5 World Bank 1987. percent per year growth, except for Lagos, which was projected at I Kenya Central Bureau of Statistics 1985. Level of urbanization for an estimated rate of 4 percent. Kenya in year 2000 given as 30.4 percent. z World Bank 1987. No government estimates available. k Malawi National Statistical Office 1984a. aa World Bank (Nigeria) 1981. 'Malawi National Statistical Office 1 984a. Rural population derived by bb Senegal Bureau National du Recensement 1982. projecting from urbanization level of 8.5 percent of total population in cc Ministbre du Plan et de la Cooperation 1985. 1977 to 25 percent in 2000. dd Senegal Ministere du Plan et de la Cooperation 1984. Latest available m Tanzania Central Statistical Bureau 1969. Government of Senegal estimate for rate of urbanization is for 1982 * Tanzania Bureau of Statistics 1981. (at 32 percent). Projected to 1985 at 1.45 percent. 0 Calculated by projecting 1978 base to year 1985 at 3.2 percent. The ee National 1985 figure projected at 3 percent (World Bank 1987) to rate of growth came from Vol. IV of the Demography of Tanzania. 2000. Ministry of Finance and Planning and the Demographic UniV ff Rural population derived from urbanization estimates for 1965 (27 University of Dar es Salaam, p. 231. Table 14.3. percent) and 1985 (36 percent) (World Bank 1987) to get rate of 1.45 P World Bank 1987. Government estimates unavailable. percent, projected to 2000 from 1982 level of urbanization. q Projected from 1978 base at 3.2 percent per year to 2000. crop allocation. Despite inaccuracies, the study is important Table 4 as the first and only systematic attempt to quantify land Rates of growth in population, 1960-2000 (in percent per potential in Africa. The results have been applied in many annum) forms (Binswanger and Pingali 1988; Oram 1987) and are Country 1960-70 1970-82 1980-86 1986-2000 highly relevant to our current study. (Projected) The evidence presented in the FAO/UNFPA/IIASA study Kna32041e on carrying capacities suggests that of the six MADIA Kenya 3.2 4.0 4.1 3.9 countries, Kenya is least able to produce enough food at Tanzania 2.7 3.4 3.5 3.4 low input levels to sustain its present and projected population. Looking strictly at arable land availability using Cameroon 2.0 3.0 3.2 3.3 government definitions, we find that Malawi, Nigeria, and to Nigeria 2.5 2.6 3.3 3.3 a lesser extent Senegal all face similar land constraints. The Senegal 2.3 2.7 2.9 3.0 study results are complicated by the various assumptions Note: The Nigerian government uses the rate of 4.0 percent growth for used in the assessments, which include production from Lagos and 2.5 percent for the rest of the country. As a result its rangelands and fallow lands. estimates are 16 percent lower than the Bank's for the year 2000. The most meaningful way of interpreting FAO's assess- ment is to use the data for carrying capacities from rainfed lands alone and to translate them into terms of minimum 15 amounts of rainfed arable land required to support one Table 6 person. A minimum land requirement indicates the relative Fertilizer response coefficients for hybrid maize in Kenya, average productivity of the land, based on FAO/UNFPA/ Malawi, Tanzania, and Nigeria IIASA assumptions.'6 We compare these figures with those Source Kenya Malawi Tanzania Nigeria we have already calculated, the government-estimated amounts of arable land available per person. The results Government 15-26 29 - 5-14 are presented below. One observes that for Kenya, Senegal, FAO 12-25 27-37 11-14 4-18 and Nigeria, the minimum "low-input" requirement is World Bank - 30 - 5-8 greater than the 1985 per capita available land and that this Source: Lele, Christiansern, and Kadiresan 1989. situation will extend to Malawi by 2000. Only with increased input levels and/or major land improvements (such as irrigation) will these countries be able to meet food needs calculated by FAO/UNFPA/IIASA are higher on average in on a sustainable basis. Another possible way of interpreting Tanzania, Nigeria, Malawi, and Cameroon on an aggregate the results is that growing conditions, including land quality, basis. Obviously, some parts of Kenya and other countries are poorest in Kenya and Senegal because minimum will be much more productive in certain crops than will rainfed land requirements are highest there. others. We turn briefly to an analysis of regional cropping patterns, rainfall, and population densities before closing Table 5 the section on aggregate data with a look at deforestation. Per capita land requirements and land availability, rainfed Soil and Rainfall Constraints arable land (in hectares per person) This section points out some sources in the literature for Rainfed land requirement Available rainfed land analyses on climate and soils in Africa. It cannot be Low Intermediate (government estimate) authoritative, but will try to indicate prominent research Country Inputs Inputs 1985 2000 and its relevance to intensification. In addition, it attempts to correlate population densities and rainfall, and rainfall Kenya 2.8 0.6 0.7 0.4 level and production possibilities. Production possibilities Senegal 2.7 0.5 1.6 1.0 afforded by the resource endowments of a given country Nigeria 0.9 0.2 0.7 0.5 determine the income opportunities available to different Malawi 0.6 0.2 0.7 0.5 regions. A region's comparative advantage in growing high Cameroon 0.4 0.2 3.5 2.1 value crops such as tea, coffee, or cocoa can increase foreign exchange earnings, employment, and income to the Source: FAO 1978; National Data (see Tables 2 and 3). benefit of different groups. It can speed the process of intensification, depending on price incentives and invest- At first sight such a conclusion is counterintuitive. Who ments in government services. On the other hand, equity would imagine the more fertile parts of Kenya to just be concerns may overshadow the investment and price incen- reaching par with areas in Nigeria, Tanzania, or Malawi? A tives governments are willing to allow particular regions, look at two further sets of data, however, confirm this view. especially if regional income inequalities threaten political First, a comparison of the proportion of cultivable land stability. occurring in the subhumid or humid tropics shows how It has been observed by Matlon (1987) that the soils of "moisture" advantaged countries such as Cameroon and West African semiaricd tropics and parts of the humid and Malawi are in comparison with Kenya and Senegal. In such subhumid tropics farther south are far more susceptible to moist environments, double cropping (e.g., of rainfed rice) rapid degradation with continuous use than was previously is a possibility, and hence the FAO/UNFPA/IIASA study thought. Low and variable rainfall makes intensifying accords them a higher value than areas where only a single fertilizer use a risky and sometimes marginally productive rainfed crop can be grown. Data showing the percentage of proposition, especially' in Sahelian countries such as total cultivable area formed by subhumid and humid Senegal, where fertilizer application can go unused in a dry cultivable tropical areas are shown below: season or can be washed away in a sudden downfall (see Figure 6). Even in eastern and southern Africa, considered Kenya Senegal Tanzania Nigeria Malawi Cameroon to have slightly more stable agroclimatic conditions, 10% 30% 51% 60% 64% 91% increasing frequency of cropping and shorter fallow periods are reducing the soil's fertility and undermining its nutrient Second, a look at response coefficients for food crops content. The process of degradation has accelerated as provides further support. Although coefficients of variation, more people are moving onto marginal land with long length of growth cycles, and rainfall dependability vary fallow requirements.17 These conditions complicate the between Kenya and Nigeria, the high potential lands of evolutionary movement toward higher levels of technology Kenya appear roughly twice as responsive to fertilizer than and weaken the causal linkage between increasing popula- land in Nigeria (see Table 6). The MADIA study on fertilizer tion densities and agricultural output implicit in the documents in detail the range of response coefficients by Boserup hypothesis. ecological zones and population densities (Lele, Christian- Broadhl speaking, in the semiarid tropics of West Africa sen, and Kadiresan 1989). between the 200 and 800 millimeter isohyets (8 to 20 A great deal of documentation accompanies these inches-see map), crop production is generally limited to coefficients and so one should be cautious about general- lower value commodities through systems of mixed crop- izing them. The responses for Kenya, for instance, refer to ping: sorghum/millet, groundnuts, and cotton. According to the so-called high potential areas that receive high levels of some, research priorities in these areas (central and moisture and enjoy deep, fertile soils. Since 74 percent of northern Senegal, northern Nigeria, and Far North Came- the land in Kenya is arid, however, carrying capacities as roon) should focus on faster maturing varieties that can 16 Figure 6 Soils with few exceptions are vulnerable to acidification and Mean annual rainfall in Senegal, 1960-84 other factors, have poor structural stability, and when cultivated intensively will be more susceptible to erosion. MM perannum Likewise the removal of tree cover has grave implications for the structural stability of these soils. The problems of soil degradation and erosion are especially acute in this 8Do_ / t zone owing to high population densities, e.g., in Nigeria. Some of the more interesting material still in experimental 70O - / \i/ \stages coming out of the International Institute for Tropical Agriculture (IITA) in Nigeria to cope with these conditions includes alley cropping with leguminous trees and shrubs, 600_ - \ new cassava varieties, and no-till cropping that increases / t \;/\ \ \ 1\ soil fertility and retains vegetative cover, thus minimizing S00 \ moisture loss and reducing erosion. j \ t \ t \/ R The higher level of rainfall in this area, between the 1400 and 3200 millimeter isohyets, is well suited to the produc- 400- tion of tropical tree crops such as cocoa, oil palm, and . \j rubber, and to the root crops yams and cassava. The higher 2_ __ t returns per hectare from the higher value crops, assuming 1960 /6 1965566 197D/71 1975/76 1980/Bt adequate yields in the humid and subhumid zones, give Source: DPGA 1961-85. the government greater latitude in shaping its intensifica- tion strategy.18 High value crops such as cocoa enable the government to extract a margin and still pass along profit to deliver stable yields in the face of declining or erratic farmers; Cameroon is a case in point. Likewise, in Nigeria growing seasons (Oram 1987). A counter argument is that in returns from planting improved cocoa were fully competi- the case of sorghum, early-maturing varieties conflict with tive with wages outside the agricultural sector even at the traditional mixed cropping with millet and may even impair peak of the oil boom, but policy and institutional con- yields if they flower before the rainy season ends (Lele, straints inhibited expansion of new cocoa plantings (Lele, Oyeiide, et al. 19891. Oyejide, et al. 1989). The MADIA paper on fertilizer (Lele, In higher rainfall regions, between the 800 and 1000 Christiansen, and Kadiresan 1989) explains how exploiting millimeter isohyets (see map), soils are typically ferrugi- regional comparative advantage is constrained by high costs nous, crusty, and prone to leaching. Clay content is of internal transportation and political and institutional generally below 20 percent (Matlon 1987). As a result, these barriers. soils tend to be shallow and have low natural fertility and In East Africa, below 400 millimeters of rainfall, few crops poor moisture retaining capacity, as opposed to soils other than sorghum and millet can survive; the diet is containing more clay or organic matter. Crop production in supplemented by livestock products such as meat, milk, this climate extending into southern Senegal, the Middle and blood. Between the 400 and 800 millimeter isohyets, Belt states of Nigeria, and northern Cameroon, include including large parts of Kenya and Tanzania, crop produc- more cereals such as wheat and maize and a variety of tion is again limited to hardy and quickly maturing cereals tubers such as yams and cassava. Soils are by comparison like sorghum and millet, and to a smaller extent cotton, much more fertile in the Asian semiarid tropics (Matlon groundnuts, and tobacco. In regions with higher rainfall, 1987). As a result, response coefficients to fertilizer tend to between 800 and 1,200 millimeter isohyets, higher value be low in many parts of tropical Africa and crop research grain crops like wheat and maize are possible, as is the must begin to consider new ways of maintaining soil fertility production of tea, coffee, and pyrethrum in the higher and increasing output. Even so, fertilizer response is higher altitude areas of East Africa (see map). The returns to labor than in the drier northern regions, indicating an untapped per hectare are especially high for tea, coffee, and tobacco; potential. The threat of trypanosomiasis, as well as other but in Tanzania and Malawi, for instance, poor prices and pests and diseases, prevents the extensive use of draft other institutional constraints to export crop production animals in the humid and subhumid tropical regions, and have shifted incentives in favor of food crops. Other MADIA keeps population densities low, despite apparently higher papers that address issues related to the development of potential for a wider range of crops than is possible in the cotton in anglophone and francophone Africa or structural North. adjustment in Malawi point out why, without intensification Eastward and to the South, in the lower parts of Nigeria efforts on cotton in anglophone Africa or with improved and Cameroon, one finds similar problems with soils in the maize in Malawi, the elasticity of acreage with respect to humid and subhumid tropics. Greater moisture and rainfall relative prices tends to explain much of the production do not translate into better growing conditions. One response. These papers document how, with increasing popular study notes: population pressure and stagnant or declining yields, Rainfall in tropical areas generally is highly erosive. overall production increases are unlikely to occur simply Rain causes erosion when it falls at more than 25mm through price corrections (Lele 1989a; Lele, van de Walle, an hour. Only five percent of rainfall in temperate and Gbetibouo 1989). areas is erosive. The proportion in tropical areas is Soils in East Africa are thought to be structurally more around forty percent-much of that at even higher sound than those in West Africa, but with the exception of and more destructive velocities. Downbursts of 100- subhumid highlands still thin and low in nutrients. They will 150mm an hour are not uncommon-as much rain as initially give higher yields using higher inputs, such as New York gets in an average month (Harrison 1987, p. chemical fertilizer, but will lose that capacity with repeated 36): cultivation unless supplemented by organic matter, such as 17 40 5 ATZAN71C MrELWEA NifA 5tA 10 OCEAN -3 20 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~20, 4TOO 400 EjaL 1400 DJ~:20 CUINEA43 ~ ~ QINMIL AJINEA SUB SAH-ARAN AFRICA: ~AGROCLIMATICZONES ZA MADIA: COUNTRIES. Average Annual Rtainfaflland Lengt iof Growing Seasion impor.y) ICLIMATIC:ZONES 20~~~~~~~~~~~ 427 * I~~~~~~~~~~~~~~~ 7 MYSU 71 18~~~~~~~~~~~~~~~~~~~~~~~~~~~~......... animal manure or humus.'9 The need for constant biological Figure 7 input underlies the growing importance of agroforestry for Deforestation in the MADIA countries relative to per capita farm management (Boehnert 1988). Tree cover also helps cultivable land reduce high rates of water evaporation that shorten the Deforested as % of total forested area effective growing season; solar radiation in East Africa (150- 180 kcal per square centimeter annually) is the highest in the world (Collinson 1987). Low levels of rainfall and high 22 7 rates of transpiration limit the utility of high solar intake, ao government estimates) which more often than not just bakes the earth. te 7 Deforestation 16 Deforestation relates back to the second, unforeseen ii dynamic of autonomous intensification. It represents an 12 acute form of overexploitation of, rather than investment in, natural resources, contrary to the Boserup hypothesis. The central importance of forests for energy (fuelwood), for food (browse and fodder), and for environmental stability (soil a l and water retention), and their threatened position as the 4 last easily accessible frontier for development, reinforce the 2 2 3 argument for introducing a more comprehensive land policy L to protect forests and encourage the use of trees in farming systems. This is why the FAO definition of arable land Malawi Nigeria Senegal Kenya Cameroon Tanzania available for cultivation is of extreme relevance for policy governing the use of the forestry sector. Considering that Note: Includes broad-leaved, coniferous, and bamboo forests. fuelwood constitutes an estimated 90 to 95 percent of Source: FAO/UNEP 1981. energy needs for rural populations in Sub-Saharan Africa and that it is also gathered for sale in urban areas, one would expect this resource to be in high demand in land- fuel and potential cropland; trees are an indispensable scarce countries and to observe a close correlation component of soil fertility management in tropical agricul- between population densities and a decline in forest ture. In the drier Sahelian and Sudanian zones, for instance, area.20 it has been shown that trees not only protect soils against The aggregate data presented in this section support this wind and water erosion and restore subsoil nutrients by premise. The four countries under greatest population shedding leaves, but they also provide fruits and leaves, pressure correspond exactly to those suffering most from firewood and building poles, bark for cord and medicine, deforestation.2' Nigeria, for instance, is often cited as a case and thorn branches for fencing, as well as serving as a where the sheer magnitude of deforestation is causing critical source of browse for livestock in the dry season serious environmental damage. With only .71 hectares of (Gorse and Steeds 1985). Forestry research in these climatic arable land per person (by government definition), the zones is said to be promising, including the use of plant country faces depletion of its tropical forest resources, as tissue cultures for propagating well-suited clones and over one-quarter million hectares are cleared for agricul- symbiotic root microorganisms to enhance the nitrogen- tural and other uses each year (FAO 1981). Equally alarming fixing capacity of certain species.23 The importance of in relative proportion is the case of Malawi, where popula- maintaining soil fertility and stability in the humid rainforest tion pressure is intense at just .73 hectares per capita of regions and the potential use of trees as part of integrated arable land. It is estimated that 120,000 hectares of farming systems in the tropics have been pointed out woodland are cleared annually, almost half as much as is previously. cleared in Nigeria on a yearly basis. If one extrapolates over What are the long-term effects of deforestation? The 10 years, one finds that, because of its small size (91,000 reduction of tropical, high altitude, and other forests has square kilometers), Malawi faces losing up to 24 percent of spawned a great deal of controversy. The Tropical Forest its forest area in a decade (see Figure 7). Resources Project undertaken by the FAO observes: New recording methods should among other things One of the most serious consequences related to consider removing this category from the calculations of forest clearing is the loss of genetic plasma and of the arable land (as FAO Production Yearbook does); most seed bearers which leads to the complete disappear- governments-gauging by their definitions of arable land ance of many species. On the contrary the impact of and vague or unarticulated policies-assume that forests deforestation on the neighboring zones is much more can be brought under cultivation with relative ease and with complex to assess: changes in water regimes, erosion, few damaging consequences. A controversial issue is climatic modifications, spreading of diseases, diffusion whether the Kenyan government's clearing of high altitude of polluting agents, change of carbon dioxide content rainforest to make way for state tea plantations is causing of the air (FAO 1981). permanent damage. Forest proponents argue that tree crops serve the double function of retaining soil cover and Evidence turned up in the MADIA study points to marked generating export revenues, but there is no consensus on changes in rainfall patterns over the past 20 years. For the issue, nor is there likely to be until more research is instance, annual rainfall in Senegal has decreased by 2.2 completed.22 Other high priority policy areas include percent a year over the past two decades (lammeh and promoting tree-planting campaigns at the national level and Lele 1988). The MADIA studies of Nigeria and Cameroon moving more land into state parks. also note a sharp decrease in rainfall in the northern parts The role of forests extends well beyond being a source of of both countries (Lele, Oyejide, et al. 1989; Lele, van de 19 Walle, and Gbetibouo 1989). These trends are alarming in in total registered land was 43 percent overall, but it was West Africa because of the more intense pressure on the well over 80 percent in Western, Nyanza, Central, and land in the lower rainfall Sudano-Sahelian zones. Although Eastern provinces, the heart of smaliholder production these trends may be temporary, there is little evidence to areas in Kenya (Lele and Meyers 1987). Institutional rights suggest that they do not reflect a permanent change to the land for smallholders have played a critical role in resulting from tree loss. Most will agree that consuming encouraging intensification, but differential access to insti- forest resources faster than they grow back is causing a slow tutional credit and a combination of social and ethnic but steadily growing environmental crisis.74 factors have rendered the land market in Kenya imperfect. Slowly rising population densities may have once been In Malawi, customary rights to cultivate and transfer enough in themselves to bring about positive changes smaliholder land are conferred by traditional tribal chiefs, associated with technological adaptation in production, while the expansion of estate agriculture has been deter- resource conservation, and consumption behavior, but mined by explicit government policies. Burley and flue- arguably this is no longer the case in Africa; the transition cured tobacco production has been reserved for estates to high density populations has been too rapid. There has through a licensing policy that accompanies the establish- been little technological change in agriculture. The tradi- ment of leaseholds on "unused" customary land. The tional farming systems of bush fallow were meant for low transfer of land from smaliholders to estates has contrib- levels of population, not rapidly rising densities. They make uted to economic growth through estate production but has the need for "intensification" and changes in farming worsened land distribution over time and led to a decline systems more urgent. Limited resources, fragile ecosystems, in average farm size in both sectors (Lele 1988a). Although and skewed incentives make it more difficult for the small- the process of technical change may be slower for small- holder to plan beyond the subsistence horizon. They make holders than for estates, land policy will be for Malawi one the short-term overuse of resources such as trees and land of the most important factors determining future growth in rational, if only for immediate survival. smallholder productivity (Lele and Agarwal 1989). Without a Land Policy clear policy, a three-tier land ownership of estates, small- holders, and marginal or landless will emerge. In this section a brief presentation is given of the various Similarly in Tanzania, smallholder control over land has approaches taken toward land policy in the MADIA sample suffered as a result of state policy. Tanzania formally and the impact they have had on the intensification abolished traditional tribal village authority, replacing it process. The analysis focuses on the East African countries with public ownership of land whereby an individual has no as they have experienced the more rapid and abrupt right of ownership or sale. In fairness to Tanzania, it should changes in land tenure patterns; despite growing popula- be added that the World Bank's 1963 report provided a tion pressure in at least two of the three countries, land in major intellectual justification for the so-called "transforma- West Africa has been a surprisingly unimportant issue in tion approach." The policy of forced "villagization" resulted public discussion and policy formulation. in the resettlement of more than 9 million people-about In Kenya, land titles and licenses to grow export crops 60 percent of the population-into 6,000 villages by mid- have been far more freely available than in Malawi, as 1975. Given the weak soils (the reason for traditionally shown by the fact that smaliholder tea hectarage has sparser settlements), the Ujamaa policy toward land increased almost tenfold and coffee hectarage doubled increased environmental stress and led to greater problems between 1970 and 1985 (Lele 1989). The World Bank has of erosion and deforestation (Lele, van de Walle, and consistently supported land registration in Kenya, since the Gbetibouo. 1989). Attempts at collective woodlots failed early 1960s. The amount of land registered in Kenya (according to one source because when one sites and increased from 1.8 to 6.5 million hectares between 1970 and plants a tree, it is tantamount to claiming ownership (Leach 1983, constituting 97 percent of all high and medium and Mearns 19881), and production of wood-related crops potential land, or, including semiarid and transitional areas, like tobacco and pyrethrum has declined (Lele 1988a). 44 percent of the cultivable land. The share of smallholders 20 Interaction between Population Densities, Cultivable Area, and Land Productivity: Some Empirical Evidence Distribution of Population on Land Figure 8 A relevant question for designing a policy-led intensifica- Distribution of population on total land area tion strategy involves the location and degree of population concentration in relation to land quality. Are people more Cameroon, 1985 Kenya, 1979 densely settled in the fertile "high potential" areas ,0% (defined by agricultural production and income possibili- 80/ ties), or are settlements-because of such factors as health 70X. hazards-located in drier areas of more limited crop 60X production potential? To the extent that population densi- 40% / ties are highest in the areas of high land potential, the 30% / answer will determine where future investment priorities in 20%X physical infrastructure such as roads, schools, clean water, 10%/ and health facilities will have the greatest impact. Regional 0% 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100. concentrations, and the land base beneath them, will also Senegal, 1985 Tanzania, 1978 figure in policy discussions of where it makes most sense 100% , to promote the use of chemical fertilizers and to direct the 90% efforts of agricultural research for the quickest returns. 80% / This section therefore tries to sketch the proportions of 70X population density on a regional basis and to assess the 50% implications for development planning. Surprisingly, there 40% is a high degree of population concentration in both land- 30% scarce and land-abundant countries. Even in large countries 20% considered to have ample land, the population is very 10% 0% much more concentrated than usually believed: In Camer- 0o 20% 40% 60% 80% 100% 0% 20% 40% 60% 80% 100% oon, for example, between 70 and 80 percent of the rural Nigeria, 1985 Malawi, 1987 population is concentrated on only 20 percent of the land 100% 1 (see Figure 8). According to government estimates, over 80 90% 1 percent of the land remains to be brought under cultiva- 80%/ tion. In land-scarce Kenya, the same proportion of the 70% 60% population is concentrated on even less land, just 10 to 15 50/ percent of total area-but for very different reasons (see 40X Figure 8). In the case of Cameroon, most people have 30% / / tended to avoid the humid tropical rainforest areas (despite 2 / the higher agricultural incomes reported there from the 10% production of cocoa, coffee, and oil palm) and farm in the 0%0% 20% 40% 60% 80% 1005 0% 20% 40% 60% 80X 100% milder climates; whereas in Kenya arable land forms such a small fraction of the total that the population is highly (X-axis: Area, Y-axis: Population) concentrated by necessity. In fact, only in Kenya was there more or less a complete Sources: Cameroon: Sixth Plan 1986; Kenya: ISNAR 1986; Senegal: congruence between high population densities and high Seventh Plan 1985; Tanzania: FAO/lBRD 1987; Nigeria: Lele, Oyejide, et land potential, although as pointed out earlier people are al. 1989; Malawi, National Statistical Office 1988. now moving to more marginal areas. This congruence has profoundly affected the regional development of crop higher rainfall Sudano-Guinean zone to the South (e.g., the production. It has intensified regional specialization in food Casamance and Tambacounda regions of Senegal, the and cash crops, rather than promoting shifts into areas of Middle Belt states of Nigeria, and the Adamaoua region of lower density but good cropping possibilities. In Cameroon, Cameroon). Not all the movement has been spontaneous, Nigeria, Tanzania, and Senegal, the population has settled as the Boserup model would suggest: Governments have in the areas of highest "potential" or best cropping used a range of policy inputs to affect the movement onto possibilities or lowest risk of disease, but large amounts of these lower density lands, including producer prices, apparently fertile land remain with low population densi- regional public investment, and the development of small- ties. In these latter countries and in the geographically holder institutions. Before analyzing shifts in production, we smaller and climatically less diverse Malawi, the issue of first consider the regional distribution of population, land population pressure on land has been framed largely as a use, and land productivity by region. "North-South" phenomenon; length of growing season and Population Densities in Relation amount of rainfall are critical in determining the range of possible population movement. Especially in the West to Quality of Land African countries, there is extreme population pressure in "High potential" can be considered in terms of yield and the drier northern reaches (between 500 and 800 millimeter response to fertilizers or in terms of income-producing isohyets) but an apparent gap of low density areas in the capacity, such as the capacity to grow high value crops. The 21 two are not always synonymous. An analysis of the price Figure 9 effect on shifts in production is given in Lele (1988a) and Kenya: Per capita high and medium potential land by will not be repeated here; we will focus exclusively on yield province data insofar as they are available. The substantial regional variation in population densities Hectares per person is not apparent in either the aggregate figures or the distribution curves. The degree of regional variation can be 0. High potent5i seen more clearly by looking at the annexes, which give the Hium potential regional breakdown in population for each country by province or other geographical subunit. As mentioned earlier, the data on land quality and land use cannot be treated as authoritative in most cases. They are used here D to give a rough idea of how population is distributed over what kind of land. 0 According to the FAO/UNFPA/IIASA study, Kenya faces the worst land constraint and has the greatest need for intensification. This observation is supported by the fact ilft Nyanza Western Central Eastern Coast North that population is heavily concentrated in the high produc- Valley Eastern tivity areas and, as we saw earlier (Figure 1), is migrating into more marginal areas. Roughly 65 percent of Kenya's Source: Jaetzold and Schmidt 1982. population is concentrated on just 9 percent of the land, which constitutes three-quarters of all high potential (i.e., humid and subhumid zones) land. As a result, the amount of per capita arable land is lowest in the three provinces with the greatest proportions of high potential land: Western, Nyanza, and Central provinces (see Figure 9),25 In fact, while Table 7 constituting only 6 percent of the total area in Kenya, these Kenya: Average yields for selected crops by province, three provinces support almost 50 percent of the total selected years (in kilograms per hectare) population. Province Coffee Tea Maize Two points bear mentioning with respect to crops yields: Smallholder Estate (Smaliholder) first, yields in the high potential areas of Central province for tea, coffee, and maize are on the order of two to three Central 723 1,286 711 1,700 times higher than in the drier parts of the country such as Nyanza 465 - 536 1,760 the Coast province and sections of the Eastern province. Eastern 420 818 524 850 Second, yields have not improved significantly in the small- RWf v'alley 250 219 526 ,1960 holder sector due to increased production on marginal Coast 250 - - 920 areas (Lele and Meyers 1986). Table 7 indicates that the Central province has a clear advantage in the production of Average 538 1,024 688 1,650 coffee and tea. Striking in the data is the difference in yield Source: Coffee: de Graaff 1986; for 1981/82 only. between smaliholders and estates; estate yields for coffee Tea: Kenya Ministry of Agriculture; for 1973-1981. approached I metric ton per hectare in 1981, whereas small- Maize: Kenya Ministry of Agriculture; for 1970-1981. holder production lagged behind at an average of .53 metric ton per hectare (see also Lele and Agarwal 1989). Small- holder coffee yields were highest in the Central province, as were smallholder tea yields-generally 25 percent higher than in its closest competitor, the Nyanza province, for the Figure 10 period 1973 to 1981. Nonetheless, the tea subsector in Cameroon: Area planted and fallow by region, 1984 Kenya is also remarkable for its consistently equitable high Percent of total area rates of growth. For all provinces, growth in production fluctuated less than 1.1 percent, between 11.8 percent in 71= Central and 10.7 percent in Nyanza. Thus, while output (Population density, 1985) 6 shares and yields may differ significantly, growth in produc- tion was largely balanced over the 1973-82 period (see Annex 3). The Rift Valley and Western provinces have a ZV 9 7 distinct advantage in the production of staple foods, reflected in their superior yields in maize production. In Cameroon, almost 50 percent of the population is do =X concentrated in the fertile Western Highlands and the high4 rainfall western lowlands, which cover less than 20 percent of total area. Data from the 1984 Agricultural Census and the C Bilan Diagnostic indicate that on the whole the intensity of F., North Adak.es. Ewot Center South Littoral South. North. Vest land use (as measured in percentage of area planted and Nor1h test st fallow) is below 35 percent, but as expected is most intense Planted Fallow in the more densely populated areas such as the Western Highlands region (see Figure 10). Similarly, the proportion Sources: Bilan Diagnostic 1986; 1984:Agricultural Census. 22 lying fallow appears to be lower in the higher density Figure 11 regions such as the Far North. The data should be taken as Population densities in Nigeria by region, 1986 a rough approximation as they are derived from two Persons per square kilometer. different sources, but they indicate a correspondence between cultivated area and population densities. Further- more, Table 8 indicates that yields of high value crops in these regions were generally higher than elsewhere in the country. Although the yield data are for a single year (1984, ,U a dry year) and do not represent an average, they are still indicative of relative land productivity. One of the principal 0- characteristics of Cameroon, in addition to its overall land abundance and relatively high concentration of population, - is the use of parastatals to promote regional smaliholder development (see Lele, van de Walle, and Gbetibouo 1989). 40 I Soil surveys completed in 1986 also support the premise that the West and Northwest provinces show better poten- tial than either the rainforest areas to the East or the savannah zones to the North (IFDC 1986). The large concentration of people in these high potential areas ,,, i makes the provision of services and the creation and maintenance of roads and physical infrastructure relatively a efficient, and consequently a smaliholder-led strategy of Northern Middle Belt Southern intensification a realistic and cost-effective way to raise states states states rural incomes. Investments in these areas, especially for transport capacity and human capital, are likely to have Source: Lele, Oyejide, et al. 1989. strong multiplier effects throughout the economy, not unlike those envisioned by Boserup as occurring spontaneously. The lack of accurate data in Nigeria on either land, Nigeria has lost its former position as a supplier on the population or crop yields makes an accurate assessment world market. Given intense population pressure in the difficult, but it appears that almost 50 percent of the South, which has from 200 to 500 persons per square population is concentrated in the southern rain forest area. kilometer, the government is moving to develop its less Population densities in the southern states are as high as densely settled areas. Since the oil boom and Sahelian those found in the East African highlands (see Figure 11). drought of 1973, two important policy instruments used to Before the oil boom, this area-which covers just under 20 promote its objective of increasing food production have percent of the total area-earned a high agricultural income been fertilizer subsidies and the construction of large-scale from the cocoa and oil palm tree crops. Since then, an irrigation schemes in the Northern region (Lele, Oyeiide, et overvalued exchange rate, a shift in terms of trade toward al. 1989). Those familiar with Nigeria expect that the food crops, and unstable marketing institutions have greatest room left for area expansion is in the lower density undermined the returns from growing these crops, and Middle Belt states with an estimated 53 persons per square Table 8 Cameroon: Average yields in the traditional sector, by province, 1984 Province Average yield per hectare (in kilograms) Robusta Arabica Oil Palm Cocoa Coffee Coffee Yams Maize Cassava (liters) The North (Savannah) Far North - - - - 665 - North _ - - - Adamaoua - 1,445 - - 1,811 2,768 South-Center (Tropical Rainforest) East 202 1,119 - - 2,012 6,906 2,107 Center 377 699 - 6,535 1,327 20,925 4,438 South 273 341 - - 1,455 15,097 709 Western Lowlands (Tropical Rainforest) Littoral 531 1,321 - 4,295 983 19,154 2,891 Southwest 597 387 - 4,953 1,581 19,550 1,413 Western Highlands (Guinea Savannah) Northwest 200 726 440 4,213 2,820 17,466 2,627 West 580 706 358 4,406 1,894 29,716 1,323 Cameroon 381 885 392 4,900 1,987 12,011 1,646 Note: Yield totals may include more than one harvest for certain crops. Source: 1984 Agricultural Census of Cameroon. 23 kilometer. As pointed out earlier, densities are lower here Figure 12 than in the North for reasons of health, a factor contributing Intensity of land use in Senegal by region to labor shortages in the Middle Belt. The government's Cultivated as percent of cultivable emphasis on promoting regional food crop specialization 116 and mechanization in the Middle Belt is underscored by .30l the recent import ban on grains and extension efforts in the North using the World Bank-sponsored agricultural devel- (Population density, 1985) opment projects (ADPs). eq In Tanzania, a land-abundant country, as much as 60 percent of the population lives on 20 percent of the land. D Population in this case is concentrated around the Lake Victoria Basin (26 percent of the total) and the coffee Z.L3 producing Northeastern Highlands (11 percent), areas of traditionally higher value and higher-yielding crops. Both these regions have a history of intensive land use, including Dakar Thies Diourbel Kaolack Louga Ziguinchor St. Louis Tramba irrigation, but the farming techniques that have evolved et Fatick et Kolda there have to date not been complemented by public policy to intensify production. Smailholders, for instance, Source: Seventh Plan 1985. receive only one-third to one-half of the world price for dark-fired and sun/air-cured tobacco (see Lele 1988a). Concerned about population pressure, the government has the majority of the Senegalese-72 percent, including the tried to open up new areas of high potential in the population of Dakar--live in the drier Sahelo-Sudanian Southern Highlands. This strategy makes sense in the long zone (350-600 millimeters per annum). Rainfall is likely to term, but in the short run it has high opportunity costs in become more of art issue insofar as it has declined terms of returns foregone that would occur more imme- significantly over the past two decades. The relatively better diately in the more accessible Northeastern Highlands. The performance for Casarnance and its more favorable place- fiscal problems encountered by Tanzania illustrate the ment in the Sudano-Guinean zone suggest greater produc- dilemma of giving regional equity a higher national priority tion possibilities for groundnuts as well as other crops; than growth in overall production. according to government estimates, over two-thirds of the In Senegal, there appears to be even less congruence arable land in Casamance remains to be brought under between population and land potential. It may be that cultivation. historical and health-related factors have militated against Finally, the population in Malawi is largely concentrated the movement into high response areas. The purposeful in the Southern region-a full 49 percent of the total. Of the concentration of infrastructure-roads, schools, railways-in 5.3 million hectares classified as cultivable in the most the Groundnut Basin of Senegal, and subsequent settle- recent land survey, 42 percent is already under cultivation .26 ment by Wolof visionaries' may, for instance, help to The extent of land use, expressed as the proportion of explain why its densities are higher than in the regions to cultivable land that is already cultivated, is highest in the the South. If the data are reliable, Figure 12 shows that in crowded Blantyre agricultural development district at over the most densely populated areas (those in the Groundnut 60 percent, followed by Lilongwe and Kasungu ADDs, with Basin) farmers are reaching the limit of the area frontier. just below 60 and 50 percent, respectively (see Figure 13). Data presented in Table 9 suggest that crop yields for These figures do not include land held fallow; they are groundnuts and sorghum/millet are on average as much as merely the crop estirnates for total area. Were they to two times higher in Casamance than in the rest of Senegal include land held fallow it is likely they would approach 100 but that average population densities there are substan- percent of cultivable land. In fact, if one uses the more tially below those found in the Groundnut Basin: 14 as conservative estimate of only 22 percent arable (without compared to 45 persons per square kilometer. Almost half forests), over 100 percent of available land would already be of the total population lives in the Groundnut Basin. In fact, under cultivation. Table 9 Population densities, average crop yield, and mean rainfall, by region in Senegal (densities in persons/square kilometer; yields in metric tons per hectare; rainfall in millimeter per annum) Province Population Average,yields (1960-1987) Average rainfall density Groundnuts Millet/Sorghum in mm./annum Dakar 2,673 560 470 438 Thies 130 790 460 520 Diourbel 116 730 48C0 500 Kaolack et Fatick 54 840 61 Cl 585 Louga 17 690 32CI 347 Ziguinchor et Kolda 31 1,020 84CI 1,118 Fleuve 14 490 39CI 284 Tambacounda 6 840 67CI 825 Source: Jammeh and Lele 1988. 24 Figure 13 Intensity of land use and population densities in Malawi by potassium are removed from the soil (Higgins). Others region argue that soils either have or do not have the major plant Percent of cultivable land under crops food elements, which if they are there are not easily 70 i 9S exhausted by cropping. If not there they must be added. 1 3 3 l | Nitrogen is an exception as it is generally very quickly' exhausted. The drain on soil nutrients caused by continu- sc 64 1 | ous cultivation and reduction in fallow periods underscores 8 l 5 _ the need for more resourceful cropping patterns, such as 3 2 ~~~~~~~~~~~including leguminous, nitrogen-fixing shrubs in the plot. Also, changing the structure of output to higher-yielding and higher value crops-both a function of policy-will by 3 4 ~~~~~~~~~~~~~producing higher incomes alleviate the pressure brought to bear by increasing population. 0 ~~~~~~~~~~~~~~~~The main thrust of this section has been to point out the Karonga Mzuzu Kasngu LIlongwe Salima Liwonde Blantyre Ngabu Total production possibilities of the various regions where populations are concentrated, and what the implication is region region for a policy-led approach to intensification. In countries where population is highly concentrated on the most productive lands, investment choices are easier from an Source: National Statistical Office 1977; 1988. economic standpoint: The marginal cost per head of extending smallholder services, such as credit, marketing channels, and inputs are small given the potential returns. A probable cause for the high concentrations in southern Elsewhere, investments in infrastructure and social services Malawi is the location of the former capital of Blantyre; are more costly but may be required to attract population although the capital has moved since independence to to underutilized land. The Casamance region in Senegal or Lilongwe in the Central region, the area around the former the Southern Highlands in Tanzania are cases in point. capital-Blantyre ADD-still contains over one-quarter of A final note before closing this section: Investment Malawi's population. The problem is complicated by decisions must be extremely sensitive to the social con- refugee movement onto the land from neighboring Mozam- straints to migration, such as ethnicity. Latent antagonisms bique. Contrary to what one might expect in the Boserup may rise to the surface with migration. The long-standing model, the yield data presented in Table 10 suggests higher antipathy between the Wolof, for example, who dominate yields in the Central region for maize, groundnuts, and the Groundnut Basin, and the Diola, a non-Muslim group tobacco, apparently unrelated to population densities. inhabiting the lower Casamance, is likely to complicate Two other important features in Malawi bear mentioning; migration in Senegal. One observer notes that: first, Table 10 indicates much higher yields for estate-grown if the relatively well-watered Casamance is to become tobacco, generally twice as high as those found in the small- an agricultural growth area for Senegal, the Diola will holder sector. Lele and Agarwal (1989) document that the have to be given a greater share of national resources lower yields for smallholders reflect their lack of access to and be represented in the elite ...if "development" inputs and better performing burley and flue-cured varie- comes in the attache cases of northern technocrats, ties. It is estimated that more than 80 percent of estate the unhappy story of the Southern Sudan or even of tobacco area is underutilized (Deloitte, Haskins, and Sells East Pakistan may well be repeated (Waterbury 1989). 1986). The second salient feature of land use in Malawi is Similarly, ethnic tensions between the Hausa of the North the apparent decline in yields over time: In the most and Yoruba and Ibo of the South may interfere with intensely cropped areas like Lilongwe, a decline in soil planned development to induce migration into the lower fertility has reduced response coefficients for fertilizer on density Middle Belt states. Interregional migration has hybrid maize from 23 to 13 between 1957-62 and 1982-84 reportedly more or less stopped since the civil war, but (TWyford 1988). This observation squares well with recent even before that, migration to the Middle Belt from data from FAO showing that, in general, for each 4,000 adjacent areas in the North and South appears to have kilograms per hectare crop of maize, 200 kilograms of been largely confined to homogeneous ethnic groups (Lele, nitrogen, 80 kilograms of phosphate, and 160 kilograms of Oyejide, et al. 1989). Table 10 Average yields for selected crops in Malawi, by region (in kilograms/hectare) Region Maize Groundnut Tobacco Smaliholders Only Estate Smaliholder (1984-87) (1984-87) (1970-85) (1984-86) Northern 1,190 410 900 400 Central 1,280 480 1,160 430 Southern 880 360 1,200 400 Average 1,110 450 1,160 420 Source: Ministry of Agriculture Spreadsheets 1987. Estate Tobacco: Tobacco Control Commission Circulars. 25 Population Densities and Incomes Figure 14 Integrally related to the "potential" of the land is the Population density and per capita agricultural income in income that derives from agricultural production. In terms Cameroon of income, land potential is determined by such variables Persons per square kilometer as the length of growing season and quality of soil (i.e., the - standard agronomic definition), as well as by access to land and secure land tenure, legal right to grow high value crops, extension, input and output marketing services, and the degree of implicit or explicit taxation of crops. As the level and quality of these services in developing countries are largely tied to government initiative to provide them, income potential by region is as much a function of policy as it is of regional resource endowments. The interaction of smallholders and government changes the simple dynamic outlined in Boserup's model, especially when land Xl becomes scarce throughout a country. The evidence uncov- ered from the MADIA sample suggests that income levels do not always follow population densities, in either land- w A abundant or land-scarce countries. 0 The point can be illustrated by taking two extremes. In ' ' o | Bo s Cameroon, for instance, it is estimated that 80 percent of its arable land remains to be brought under cultivation. Given Rural per capita income in '000 CFA appropriate cultivation techniques (retaining vegetative cover), it has a wide margin for area expansion. The highest source: Government of Cameroon 1985. agricultural income-earning areas in Cameroon were those areas least densely populated (see Figure 14); the high density areas received lower incomes, based more on the sale of Figure 15 food crops than of what are traditionally termed cash crops Agricultural income by region in Cameroon (see Figure 15). These findings relate back to the definition of high potential land that looks more at income and Thousand CFA therefore uses value of crops grown to measure land potential, as opposed to FAO definitions that classify (Population density by province, 1985) potential simply in terms of soil quality and rainfall patterns. The data suggest that people chose to forgo the better income opportunities of the tropical rainforest areas and instead are concentrated in regions of more moderate climate. In Senegal, by way of contrast, populations are concen- trated in the high income areas. The production of ground- nuts, Senegal's principal export crop, is concentrated in the high density Groundnut Basin. Four-fifths of total groundnut production accrues to the regions in the Groundnut Basin, Far North Ad..aota East Cnter Sooth North- West So.th- Littoral and close to 50 percent is produced in the Sine-Saloum North west west IKaolack and Fatick) region alone (see Annex 3). The latest MCashcrop _ Foodcrop available data from 1975 suggest that almost one-third of total crop income, or 22 billion CFA, accrued to the Sine- Saloum (see Table II). The higher density Groundnut Basin, Source Government of Cameroon 1985. with 49 percent of total population, received 58 percent of total rural income in 1975, with well over two-thirds of its total income derived from crops. Waterbury (1987) argues that the Groundnut Basin also had preferential treatment in land in Senegal would force people into the Casamance institutional arrangements in the colonial era for marketing area to exploit its apparent higher yields in crop produc- groundnuts and in some years received substantially more tion. However, variables other than population density than the world price. The lack of more recent data on appear to be affecting the natural processes of autonomous income makes it hard to assess what has happened in more intensification (here, area expansion) observed elsewhere recent years, especially in light of erratic and declining by Boserup. These variables include social and ethnic rainfall and soil erosion. But it is evident that development factors, the choice of crops grown, the prices received for of the Basin is no more the priority it once was. Lele, those crops, and public expenditure. Even though one Christiansen, and Kadiresan (1989) have documented that region may be densely settled-for reasons of better fertilizer consumption has virtually collapsed in the infrastructure, social services, or climate-it does not Groundnut Basin, and that investment has shifted to axiomatically lead to higher incomes. Incomes also depend irrigated rice production in the Fleuve region. on the congruence between land potential and adequate These two cases contradict a commonsense interpreta- labor to produce high value crops. Incomes are higher in tion of the Boserup hypothesis: One would expect, all the low density areas of Cameroon, for example, because things being equal, that the acute demand for productive the crops grown there, cocoa and oil palm, fetch a premium 26 Table 1 1 Rural incomes by source and region in Senegal, 1975 (in billions of 1975 CFA; per capita income in 1975 CFA) Crops as Total rural Total per capita Region Crops Livestock Fishing Forestry % of total Income Income 1975 Dakar 1.3 0.7 5.7 1.3 144% 9.0 10,088 Groundnut Basin 45.3 7.5 4.4 4.7 73% 61.9 31,313 Thies 9.9 1.4 3.1 1.9 61% 16.3 24,291 Diourbel 13.7 3.0 0.6 79% 17.3 41,051 Sine-Saloum 21.7 3.1 1.3 2.2 77% 28.3 28,597 Louga Outlying Regions 20.3 9.8 3.9 2.4 56% 36.4 24,493 Casamance 12.8 4.4 1.5 1.0 65% 19.7 28,143 Saint Louis 3.4 3.4 2.4 0.8 34% 10.0 19,563 Senegal Oriental 4.1 2.0 0.6 61% 6.7 24,367 Total Senegal 66.9 18 14 8.4 62% 107.3 21,992 Sources: Jammeh 1987. Population densities from Seventh Plan tor 1985,19, table 4. Per capita incomes calculated by dividing total rural income by 1976 population figures. on the world market, and even though the government of Figure 16 Cameroon taxes cocoa production heavily, farmers still Population densities and rural incomes in Malawi receive a healthy margin. This would be impossible under Kwacha per household, 1980-81 the production of low value crops. lust as in land-abundant Cameroon, incomes in land- °°t 64 l scarce Malawi are highest per household in regions of lowest (Population density, 1987) population density (see Figure 16). Kasungu, Ngabu, and 32 9 7e Karonga are the lowest density ADDs (57, 54, and 34 ,a} persons per square kilometer, compared to the national average of 76) but they have the highest average household incomes, at 213, 143, and 142 kwacha a year, respectively (about US$70-100). In the high density areas, land is so short that small farmers have difficulty earning an adequate income from crop production. Despite the greater role of K agriculture in the Malawian economy, crops contribute only K.,ang. Muu K.sungu Selim. Llbng Ui BISn'yO. Ng.bu 34 percent to total rural household income. The dominant source of income for smallholders is business or trading, at Labor Transportation Cash crop 27 percent of the total; the second largest source of cash s Livestock Food crop _ Business income is food crops, at 23 percent of the total. Cash crops- generally higher value crops typically grown for export- Source: National Statistical Office 1984b; 1988. contribute only II percent on a national average to small- holder incomes (see Figure 17). The data indicate that where crops have contributed significantly to total house- Figure 17 hold incomes, the absolute level of income is higher. The Sources of household income in Malawi by region, 1981 observation is consistent with the literature on Asia that Total - 157 million kwacha emphasizes the importance of agriculturally-led growth (Mellor and Johnston 1984). Were the government to allow or encourage the production of higher value crops, it could i potentially alleviate the land constraint by raising the K incomes of smallholders. l In Kenya and Tanzania, high incomes were found in areas s ness-28 of high population density, but lack of recent data makes it difficult in the case of Tanzania to assess the income effects of shifts in production to the Southern Highlands. For transfers Kenya, a high correspondence between population densi- ot'_ 1 ties and good land quality means that incomes have tended to be higher and remain localized in the areas growing high value crops. Table 12 indicates that the 1974- 75 survey, the last to include income data, shows that over half of the households surveyed in the Central and Rift Valley provinces of Kenya earned more than 3,000 Kenya shillings. Likewise, the mean value of farm assets for the two provinces was substantially above those found in other provinces. The Nyanza, Eastern, and Coast provinces had Source: National Statistical Office 1984b. 27 Table 12 Percentage distribution of holdings by household income group and mean value of assets per holding by province (1974/75) Income Group Central Coast Eastern Nyanza Rift Valley Westem Total Less than 0 10% 4% 5% 4% 16% 5% 7% 0- 999 8% 10% 9% 12% 10% 21% 12% 1,000-1,999 14% 21% 26% 26% 9% 29% 22% 2,000-2,999 14% 17% 13% 13% 15% 15% 14% 3,000-3,999 10% 15% 15% 11% 10% 10% 12% 4,000-5,999 15% 20% 13% 14% 15% 9% 14% 6,000-7,999 11% 4% 11% 4% 8% 7% 8% 8,000 and over 1/7% 8% 8% 17% 17% 3% 12% Mean Value of Farm Assets 11,233 7,397 6,438 4,357 10,327 4,471 6,905 Note: Mean Value of Farm Assets includes land, buildings, farm equipment, transportation equipment, livestock, crops in store, planted crops, and inputs in store. Source: Integrated Rural Surveys 1974-1975. lower proportions of households above the 3,000 Kenya the low density Southern Highlands. In Senegal, invest- shilling mark, and the Western province had the lowest-70 ments in irrigation in the Fleuve region have caused rice percent below the 3,000 Kenya shilling line. Remarkably, production to shift to the North and away from Casamance. only the Rift Valley had a higher proportion of households Reliable time-series data for Nigeria, Cameroon, and earning no cash income than did the Central pro,vince, at 16 Malawi are not available, but there too it appears that versus 10 percent, indicating a concentration of subsistence production has shifted into lower density regions. The farmers in the two most well-to-do provinces. The apparent spontaneous movement into new areas sits well with the distribution problems in these two provinces point to the Boserup model, but as the following sections try to need for more accurate and up-to-date information to demonstrate, the picture is somewhat more complicated. assess the effects of rapid growth in high potential areas. A]though data on regional income are even more limited Food crps in Tanzania, it appears that the traditionally most densely We begin by looking at maize in East Africa, because it is populated districts (Kilimanjaro, Mwanza) also received the marketed and records of official purchases are readily highest incomes. However, owing to shifts in production available. While there is a good deal of informal marketing, from North to South, the picture may have changed. In official sales nevertheless provide some important insights. former times the coffee-producing Kilimanjaro region had In Kenya, for example, maize is produced throughout the the second-highest regional GDP (1970), after Dar es Salaam country but in largest quantities in the Rift Valley. Between (see Annex 2). More recent data on regional incomes are 1970 and 1985, 38 percent of maize production on average not available, making it difficult to distinguish whether came from Rift Valley, from 27 percent of the area under incomes still follow population densities as they do in cultivation (see Annex 4). The Eastern province, which grew Kenya. Other piecemeal data on fertilizer consumption, by I percent in production and 4 percent in area, registered investment in roads, and marketed surpluses of tobacco, a 26 percent share of total area on the average, but tea, coffee, and maize suggest a clear shift away from the produced only 13 percent of total output. The lower returns northeastern and Lake Victoria areas toward the South. on the increased area in Eastern province may indicate an The specialization in high-value crops by certain regions expansion onto marginal lands. such as the Central province or Northeast Highlands raises Furthermore, the Rift Valley sold the highest percentage interesting questions about regional comparative advan- of maize to the National Cereals and Produce Board (NCPB) tage. In the next section we examine the shifts in produc- with 63 percent on average, followed by Western province tion in the most important crops, treating the shifts as with a 24 percent share (see Figure 18). These figures are outcomes to autonomous changes arising from localized substantially higher than those given above for total output, population pressure (autonomous intensification); supply where the Rift Valley had a 38 percent share of production. response to price changes; supply response to regional This suggests that a large part of the Rift Valley and investment patterns, and supply response to other non- Western maize output is channeled through the NCPB, price factors such as institution-building at the regional whereas for other provinces, such as Eastern and Central, level. output bypassed the parastatal and was consumed locally. Central and Eastern provinces, for example, produced 13 Population Densities and Regional Crop percent and 12 percent of total output for maize, but Production accounted for only 3.4 percent and 2.5 percent, respectively, Data on regional crop production over time-insofar as they of maize sold to the NCPB for the 1970-84 period. Percent- are available-indicate a shift in production among regions, age amounts of maize purchased for consumption are shown in generally away from high density areas, and apparently Figure 19, and confirm this observation; they indicate that owing more to policy initiatives than to spontaneous households in Eastern and Central provinces purchase over migration. Only in Kenya was there no perceivable shift in 40 percent of their grain (for their own consumption) on the marketed production, a fact attributable to the apparent market. What Figure 19 cannot show is the extreme congruence between population concentrations and crop- fluctuation in regional market dependence, especially in ping potential of the land. In Tanzania, as mentioned above, drought-prone or marginal areas. government investment policy encouraged production in The problem of market dependence is complicated by 28 Figure 18 Figure 19 Official maize purchases in Kenya by region, 1970-83 Maize purchases for own consumption in Kenya Thousand 90 kg. bags Percent 50 40 30 10 -Western Eastern Rift Valley of Rift Western Nyanza Eastern Central Coast 0 Nyanza ,. Central Valley Source: NCPB 1985. Source: Government of Kenya 1983. projected decreases in per capita arable land. Little Figure 20 agricultural land is available per person in these areas Official maize purchases in Tanzania by region, 1970-87 already, and Table 13 shows that in most cases those amounts will fall by up to 40 percent by the year 2000. As Thousand metric tons population grows, more of the land will be allocated to 12° maize production. Table 14 indicates that projected maize . deficits will grow in many districts and that many districts formerly showing a surplus will record a deficit. Especially in the more marginal districts, the difference between a 70i good year and a bad one can have serious implications; based on the projections in Table 14 four out of the six districts in Eastern province would slip from maize self- sufficiency to a deficit in maize without a "good" harvest. a00 The rapid population growth and shrinking per capita land io supply emphasize the need for policy-led intensification, , especially in countries such as Kenya where little remains of the area frontier. The MADIA paper on fertilizer explores the implications for input use in high and low potential + South A Plateau * Northeast O Basin C coast areas for both growth and equity; another explores the implications for food policy (Lele, Christiansen, et al. 1989). Source: Government of Tanzania 1980; 1987. In the case of Tanzania, as a result of policies such as pan-territorial pricing that discriminated against the North- east Highlands, the production of marketed maize shifted Given the high population densities in the northeast, there over time from the Northeast Highlands and Dodoma is an urgent need for intensification of high value crops. province to the Southern Highlands. In 1970, the Kiliman- In Malawi, yield differences across regions are not so jaro, Arusha, and Dodoma provinces accounted for over 64 large (see Table 4) as to confer a regional advantage in percent of National Milling Corporation (NMC) purchases; maize production. Nevertheless, because of extreme pop- by 1987 that figure had dropped to less than one-third. ulation densities in the South, regional surpluses have Regions in the Southern Highlands, by contrast, rose from shifted over time, and two trends stand out. The Central about 22 percent in 1970 to over 55 percent in 1987 (see region emerged in 1974 to become the leading supplier of Figure 20). The shift in marketed production is away from maize (see Figure 21). Concurrently, the limited data for the relatively high density regions to the North to lower sales show an increasing dependency on the market in the density highlands in the South. For the 1978-87 period, for Southern region, where population pressure is most which data are available, between 40 and 60 percent of the intense. Between 1983 and 1986, the Southern region officially marketed surplus was sold in the Coast region, accounted for one-half to three-quarters of total maize sales including Dar es Salaam (see Annex 2). Even though the from the Agricultural Development and Marketing Corpora- high potential Northeast Highlands have stopped selling tion (ADMARC) to smallholders. As referred to earlier, surplus maize to the NMC, it appears they are roughly self- Twyford documents the decline in response to fertilizers in sufficient and-with the notable exception of Dodoma- this region as it has been most intensively cropped, which have not increased purchases of officially marketed maize. could signal mining of the soils and perpetuate the circle of 29 Table 13 Maize deficit and maize surplus areas by province and distrAct in Kenya, and distribution of population on high and medium potential land, 1985 and 2000 Province Maize balancea High and medium potential land ('000 MT) Total Percent Hectares per person District Moderate year Good year square km. of total 1980 2000 1980 2000 1985 2000 Nairobi -79.82 -224.65 -79.82 -224.65 Kiambu -46.73 -169.25 -34.34 -151.79 1,248 51% 0.14 0.08 Kirinyaga 5.84 -7.17 19.10 18.36 950 66% 0.29 0.15 Muranga -35.49 -125.27 -21.29 -97.69 1,808 73% 0.21 0.11 Nyandarua -8.38 -32.24 0.18 -16.94 1,988 56% 0.67 0.39 Nyeri -33.93 -97.93 -24.22 -80.39 1,380 42% 0.22 0.12 Central -118.69 -431.26 -60.57 -328.45 7,374 56% 0.25 0.13 Kilifi -21.44 -71.57 -5.91 -41.47 2,541 20% 0.45 0.25 Kwale -29.43 -60.10 -26.50 -69.80 2,085 25% 0.58 0.30 Lamu -3.02 -10.22 -2.02 -7.40 3,887 60% 6.54 3.02 Mombasa -34.32 -79.85 -33.99 -79.20 0 0.00 0.00 Taita/Taveta -6.67 -24.13 -1.11 -14.73 703 4% 0.37 0.21 Tana River -8.18 -29.74 -6.59 -26.64 418 1% 0.32 0.15 Coast -103.06 -275.61 -76.22 -239.24 9,634 122% 0.55 0.30 Embu -11.37 -11.37 -3.99 14.84 800 29% 0.23 0.12 Isiolo -2.18 -3.03 -0.86 1.49 0 0.00 0.00 Kitui -36.98 -38.89 -29.04 0.65 2,902 10% 0.48 0.27 Machakos -22.91 -3.73 53.51 234.70 3,657 26% 0.27 0.14 Marsabit -9.01 -28.33 -7.80 -24.20 0 0.00 0.00 Meru -34.17 -20.63 -16.38 40.19 2,870 29% 0.27 0.14 Eastern -116.62 -105.98 -4.57 267.47 10,229 7% 0.29 0.15 Garissa -2.37 -8.77 -2.37 -8.76 0 0.00 0.00 Mandera -1.96 -3.74 -1.97 -3.75 0 0.00 0.00 Wajir -2.62 -7.88 -2.62 -7.88 0 0.00 0.00 North Eastem -6.95 -20.39 -6.95 -20.40 0 0.00 0.00 Kisii -13.35 -65.20 -0.12 -38.52 1,925 88% 0.16 0.09 Kisumu -35.77 -87.48 -33.12 -82.61 1,597 76% 0.24 0.13 Siaya 3.81 -26.68 23.15 6.64 2,039 81% 0.31 0.19 South Nyanza -1.43 -35.30 18.00 2.78 4,124 72% 0.37 0.22 Nyanza -46.74 -214.66 7.92 -111.71 9,685 77% 0.27 0.15 Baringo -18.76 -43.21 -16.42 -40.29 1,976 20% 0.77 0.46 Elgeyo Marakwet 21.32 35.83 33.91 51.55 1,104 48% 0.67 0.63 Kajiado -8.31 -40.49 -5.50 -34.33 311 2% 0.15 0.07 Kericho 44.77 81.85 72.94 144.90 3,354 85% 0.41 0.23 Laikipia -5.10 -28.98 -0.63 -19.19 1,330 14% 0.69 0.30 Nakuru -8.49 -24.38 0.57 4.39 2,678 46% 0.36 0.17 Nandi 99.13 177.40 127.27 229.89 1,926 70% 0.49 0.30 Narok -10.71 -53.23 -6.49 -44.00 5,435 34% 1.87 0.89 Samburu -9.71 -17.09 -9.37 -16.39 0 0.00 0.00 Trans Nzoia 98.21 183.09 121.01 236.88 1,550 75% 0.41 0.18 Turkana -20.80 -20.62 -20.77 -20.68 0 0.00 0.00 Uasin Gishu 43.72 74.70 52.10 93.83 2,781 82% ; 0.68 0.33 West Pokot -2.54 -50.14 1.78 -43.54 1,368 15% 0.60 0.27 Rift Valley 222.73 274.73 350.38 545.09 23,840 15% 0.55 0.29 Bungoma 28.63 53.53 43.80 88.72 1,992 65% 0.30 0.16 Busia 0.08 -26.05 8.69 -9.63 1,349 83% 0.35 0.18 Kakamega 43.58 101.39 85.37 198.23 2,548 73% 0.20 0.11 Western 72.29 128.87 137.86 279.33 5,889 72% 0.25 0.14 Total -176.86 -868.95 268.03 167.43 66,652 12% 0.33 0.18 Note: For maize balance 15% deducted for fodder and losses. Assumes 2.5% overall yield growth distributed in accordance with districts' growth potential. Area growth 1% in Central, Nyanza, and Western provinces, otherwise 2%. Some have expressed doubts about the district maize balance results in this table. For instance, G. Stern observes, ". . . Machakos production fluctuates between feast and famine depending on the weather, but it is hard to believe that in a favorable year, by 2000 its surplus would be second in the country and very close to first.... Kakamega data [are also] surprising. At one time, the district (called North Nyanza) included Busia and Bungoma, and it was Bungoma that generated major surpluses.... [it is] hard to believe that Kakamega with some of the most densely populated areas could generate sizeable surpluses. One can divide the district into the heavily populated South that will be as or more food deficient than Kiambru district; a reasonably self-sufficient, fairiy heavily populated center and a potential surplus, less densely populated North. The surpluses in the North could not do more than meet the deficit of the South" (Personal communication with the authors). Source: Maize Balance and Population Data: Githongo & Associates 1983. Agricultural Land Statistics: Farm Management Handbook of Kenya Vol. il, as 30 reported in ISNAR 1986. Table 14 Figure 21 Regional investment as percent of total in Senegal, 1977-84 Official maize purchases in Malawi by region, 1970-87 Region Population Fifth Plan Sixth Plan Thousands o metric tons density 1985 Investment Investment -__ per/sq.km Dakar 2,673 31.2% 21.7% Groundnut Basin 49 28.2% 13.5%0/ Thies 130 10.7% 5.0% on Diourbel 116 3.4% 0.2%0/ Kaolack et Fatick 54 10.7% 5.0% Louga 17 3.4% 3.3% Outlying Regions 14 30.0% 23.7% Ziguinchor et Kolda 31 11.5% 9.7% Saint Louis 14 11.5% 10.0% Tambacounda 6 7.0% 4.0% i 70 1940 29.4 1996 199 1900 1003 1990 Nonlocal - 6.0%/ 40.0% * Northern K Southern + Central Total Senegal 26 95.4% 98.9% Source: John Waterbury 1986; Population Densities from Seventh Plan Source: ADMARC/DHS 1987. 1985. lower crop yields and greater market dependence. Figure 22 Data for West Africa are more scarce, making it hard to Production of estate tobacco in Malawi by region, 1960-85 point to areas of food surplus or deficit. In Senegal, for instance, it appears that the country as a whole is shifting Thousand metrc tons into sorghum and millet. Its share of total cultivated area grew from 42 percent in 1963 to 53 percent in 1987. jammeh / - and Lele (1988) argue that the shift into millet and sorghum reflects an attempt to manage climatic uncertainties and 30 - reduce risk. The most dramatic increase in area and production occurred in the densely populated Groundnut Basin, particularly in the Sine-Saloum (Kaolack and Fatick) region, where between 1961 and 1976 area and production doubled, from 157,000 metric tons to 322,000 metric tons, dropping slightly in 1987 to 290,000 metric tons (see Annex 3). The problem of area expansion in this high density region is compounded as we saw earlier by the fact that, according to government estimates, little arable land remains to be brought under cultivation in these regions (refer to Figure 22). Area and production of sorghum and millet rose much less in the lower density Casamance ±96o ti7 oo7 o97 0490 0900 0902 region, which instead showed a steady increase in maize production and variable performance in rice production. K Southern + Central w Northern Rice production increased in the irrigated northern Fleuve region. The lack of data on officially marketed production Source: Tobacco Control Commission Circulars 1972-86. makes it difficult to pinpoint food surplus areas, but from production data it appears that the shift in food crops has consisted mainly of a diversification in the better watered flows are from the West province, while the primary regions to the South and more rice production in the North. destination is the Littoral province. As we saw earlier, the While this is a desirable move in principle, the remoteness proportion of income deriving from food crops was highest of these regions and their very small populations make in the Northwest and West provinces, at 79 and 57 percent improvements in employment and income generation less of the total, respectively. effective than would be the case if the Groundnut Basin Nonfod crops were the focus of development. In Cameroon, information on marketed production is Shifts in the production of high value export crops among available from survey data only for 1984, which was a regions were most significant in Nigeria, Tanzania, Malawi, drought year. The Northwestern Highlands accounted for and to a lesser extent Cameroon. In all four cases, the shift over half the marketed maize (100,000 metric tons), just away from the traditional centers of export production under a third (122,000 metric tons) of the plantain, and resulted from explicit policy objectives, not from spontane- about one-quarter of marketed cassava (85,000 metric tons), ous or autonomous migration as might be thought under making it a food-surplus region despite its high densities the Boserup paradigm. Although it is common sense that (see Annex I). Gaviria (1988) points out that the major food policy will figure largely in the structure and location of 31 agricultural production, it is important to underline this crops. Especially in Nigeria, traditional export crops in the point to dispel the belief that a laissez-faire approach to South declined as oil revenues supplemented them. The population growth, by allowing market forces to operate, effects of this shift away from the South and on the will correct for factor scarcities. economy as a whole are documented in Lele, Oyejide, et al. In East Africa, two points emerge: Production shifted into (1989). In Cameroon, no time series data are available, but low density areas in Tanzania and Malawi, and production important gains in cotton and rice production in the North concentrated in the high potential regions of Kenya. In the are documented by Lele, van de Walle, and Gbetibouo case of Tanzania, as pointed out above, the government (1989). These authors point out how parastatals played a encouraged a shift in production away from the Northeast- vital role (SODECOTON, SEMRY) in encouraging this ern Highlands to the Southern Highlands. Although total regional shift. The allocation of resources to develop the production of coffee grew at only 2.3 percent and tobacco dry northern area raises questions about optimal efficiency at -4.8 percent, the relatively low density Southern High- that must be reconciled with the government's agenda of lands doubled its share of total coffee production to 25 equitable development as a nation. Similarly in Senegal, percent in 1981-85, and increased its share of tobacco large investments in the North do not provide the govern- production from 18 percent in 1970-74 to 60 percent in 1982- ment with the highest economic return but may meet other 86 (see Annex 2). The redistribution in production was not politically important criteria. It is to a brief analysis of associated with substantial growth in overall output, due to expenditures that we now turn. a decline in traditional areas. . In Kenya, the data indicate little change in relative shares Populatlon Denslies and of cash crop production. For the period 1973 to 1981, for Regional Public Expenditures instance, the Central province dominated, accounting for Data on regional public expenditures must be treated with half of all tea production. A striking feature of tea produc- caution, as there is no preexisting methodology to calculate tion in Kenya is that it grew evenly among the provinces, rates of return, nor are there enough adequate or reliable generally above 10 percent a year (see Figure 231. In view of data on which to base such an analysis. However, it is the country's very tight land constraints, the story of tea possible to make some tentative observations based on development there is a model of policy-led intensification. the limited data available. The most important point to Data on coffee production, while more limited, again point emerge is that, beyond the simple mechanics of increasing to a concentration in the Central province, where growing population densities, regional and sectoral allocations by conditions are the best, and to a lesser extent the Eastern governments will shape the pace, direction, and location of province (see Annex 5). intensification. The point can be simply illustrated by considering expenditure patterns in Kenya and Tanzania. Both countries inherited fertile highlands endowed with an indigenous Figure 23 labor supply. Yet their responses were almost exactly Growth in tea production in Kenya by region, 1973-82 opposite. Kenya chose to develop its high potential areas explicitly (some would say was compelled out of political Percent expediencyl whereas Tanzania shifted expenditures in favor t1TZ [- | . of its high potential but less developed, less populated regions. lO- l _ F __ In Kenya, for instance, expenditures on main services between 1970 and 1983 grew fastest in the high income, 8 ~~~~~~~~~~~~~~~~high potential Central province, at 6.2 percent in real terms. In the second half of this period, subsequent to the death of President lomo Keniyatta in 1978, the Central province 6- E l g __ _!1 received consistently up to one-third of regional expendi- tures; similarly, per capita expenditures were substantially 4 | i | I above those in other provinces (see Figure 24). It was followed by the Western province, where expenditures grew by 4.9 percent in real terms, compared to the national 2 average of 2.4 percent real growth. The provinces exhibiting the fastest growth in expenditure also showed the greatest 0 t I degree of ethnic homogeneity: The Kikuyu dominated the Central Western Eastern Nyanza Rift Valley Central province, composing 95 percent of its population in 1979, as did the Luhya, with 86 percent, in the Western Source: KTDA Annual Reports. province, with both groups exceeding 1.5 million persons. The data suggest that rather than trying to reduce regional income disparities, as was the case in Tanzania, the In Malawi, it is striking that the government policy government used its expenditures to reward its most vocal, favoring estate agriculture led to the dramatic expansion of active, and vital constituents. In the process, the govern- such production throughout the country, even in the high ment spent more to develop high potential areas than it density Central and Southern regions. One consequence of did on other provinces, a policy that paid off in high rates estate agriculture in areas of tight land supplies was to of growth. Significantly, growth rates for primary school increase environmental stress on land under smallholders enrollment for the 1968-84 period show that, despite higher (see Lele and Agarwal 1989). spending in the Central province, other provinces benefited In West Africa, a series of price, investment, and institu- from more rapid growth in jobs and education (see Figure tional policies affected the regional production of export 25; see also Annex 4). 32 Figure 24 Figure 25 Per capita regional expenditure in Kenya by region, 1969, Growth in primary school attendance in Kenya by region, 1979, and 1983 1968-84 Kenya pounds Percent Increase over population growth 2.4a 2 7 ~~j I 54I 0.B 0. ~~~~~~~~~~~~~~~~~~~~~ 1969/70 ~~~~~~~~~~~~~Rift North Nyanza Western Coast Eastern Central Nairobi Central Rift Valley Eastern Valley Eastern Nyanza =Coast Western Source: Kenya Statistical Abstracts. Source: Kenya Statistical Abstracts. In Tanzania the government adopted a totally different in both absolute and relative terms, more money was approach: Rather than try to develop the high density, high directed to the outlying regions. Significantly, the drier income areas, as was the case in Kenya, it used regional Fleuve region in the North of Senegal received as much expenditures to try and narrow regional income disparities, investment as the Casamance region in the Fifth Plan, at This was politically feasible because no one particular 11.5 percent each, and slightly more in the Sixth Plan, at 10.0 ethnic group dominates in Tanzania. Total expenditures compared with 9.7 percent, despite the fact that Casamance were lowest in the high potential Northeast Highlands, at has a greater share of the total population (14 compared to roughly 12 percent of total for the period, while a greater 9 percent), higher population densities, and according to share (in both absolute and per capita terms went to the the latest land statistics four times more "unused but lower potential coastal belt which received 25 percent, and potentially cultivable" area. in fact, investments in the the central and western plateau, which got 20 percent) (see Fleuve area (mostly in irrigation) fell less than in any other Annex 2). Tanzania's regional redistribution problem was region in the Sixth Plan, indicating the government's complicated by changes in intersectoral patterns. Govern- commitment to (or inability to withdraw from) costly ment expenditures on the directly productive agricultural investments already made. one might be led to conjecture sector declined, while increasing on social services, espe- that investments in the Fleuve region have a good deal to cially education. For Tanzania, the emphasis on equity and do with local and ethnic allegiances: The largest proportion provision of social services to the exclusion of growth of "fonctionnaires" in the government, roughly one-fifth, caused many problems. Chief among them was the inability were born in the Fleuve region (Le Senegal en Chiffres 1982/ to finance recurrent expenditures (Tanzania Agricultural 83). Our judgment is that investments in Casamance, a low Sector Report 1983). Total expenditures rose rapidly until density/higher rainfall region, will pay off more quickly and they peaked at 3.4 billion Tanzania shillings in 1983, before do more to ease population pressure in the Groundnut falling to one-third of that level in 1984. Further, expendi- Basin. tures on transportation declined, aggravating the already Finally, capital expenditures in the agricultural sector in poor mobility of labor and goods in Tartzania. Whether Nigeria have shifted since the early 1970s from the highest redistributing national income on equity grounds is a density Southern regions to the relatively less dense North. prudent approach toward intensification is debatable. This In 1981-1985, for instance, less than 10 percent of the paper argues that when resources are scarce, the most regional budget, or 1.3 billion naira, was allocated to productive investments are in areas with the highest investment in agriculture in the Southern states, whereas returns. the figure for the Northern states (thought to be more We close this section on regional expenditures by citing economically depressed yet politically quite important) is the cases of Senegal and Nigeria where neither the Kenyan higher, at 1.5 billion naira, and accounts for a larger share nor the Tanzanian pattern is repeated. Senegal chose to of its regional budget, at 18 percent. The expenditures in invest a slightly higher proportional share in the outlying the North increased in 1981-85 because of the statewide (i.e., non Groundnut Basin) areas. In fact, although almost agricultural development projects in Sokoto, Kano, Bauchi, half of the population is concentrated in the Basin, only 28 and Kaduna. On a per capita basis, however, the Middle and 14 percent, respectively, of total investment went to Belt states came out favorably, given its lower population this area in the Fifth and Sixth Plans (1977-84, see Table 14). (see Figure 261. Another tack pursued by the federal 33 Figure 26 the other MADIA countries, set a striking contrast in Per capita govemment expenditures in Nigeria by region, patterns of labor use. In Cameroon, the fact that land is still 1981-85 abundant is reflected in the low proportion of hired labor Naira per person in the agricultural labor force, just under 2 percent in 1984. Significantly, the highest proportion of hired labor in agriculture (roughly 6 percent) obtained in the high income Southwest province (see Table 15). This province alone produced one-third of the total cocoa (35,000 metric tons) 300 - _ | _ _ | lll 111 and one-fifth of the oil palm production (17,000 liters), earning over one-fifth of the country's total cash crop income in 1985. A strong correlation between high income and high hired labor input would seem to be borne out, regardless of population densities: The Southwest province had one of the lowest densities in lower Cameroon, at 33 _U _ _ _persons per square kilometer. The absolute amount of labor per farm is highest, by contrast, in the higher density Northwestern Highlands, at roughly 4.5 workers per farm, 5D0 | 111 _ _ compared with the national average of 3.7. Hired labor is higher where cash crops are grown, but total labor input Northern Middle Belt Southern corresponds more to population densities. states states states The case of Malawi presents an extreme contrast. Accord- ing to the 1980-81 rural survey, 55 percent of all households _ Federal expenditures _ Total expenditures cultivate less than one hectare of land. Even more striking, those 55 percent account for a meager 25 percent of the Source: Lele, Oyejide, et al. 1989. total area cultivated (see Table 16). Lele and Agarwal (19891 document the implications of land distribution and shrink- ing plot size, including the effects on intensification. In the government was to subsidize fertilizer sales, two-thirds of Southern region, population densities reach 200 to 300 which were consumed in the North. The salient point is that persons per square kilometer. There is a growing number of public policy plays a crucial role in the intensification individuals selling their labor to earn an income; the process, and that regional expenditures are an effective way Southern region accounts for over half of the number of of guiding the autonomous forces that arise out of popula- people earning wages through agricultural work (see Figure tion growth. 27). Plot size has become so small that the "normal path" of intensification is bypassing Malawi. The negative effects Population Densities and Input Use of Malawi's emphais on growth is a sobering counterpart to One of the main tenets of the Boserup hypothesis holds the extreme emphasis on equity in Tanzania. that the incentive to use more inputs (land, labor, and That the traditional path of moving to higher levels of capital) grows in proportion to population densities. The production has not been achieved is also shown by the most common and readily available input is labor; it is means of cultivation used in Malawi. In the most densely estimated that on average, up to 80 percent of value added populated regions, over 90 percent of the land is cultivated in Africa's agriculture comes from labor. In this section we by hand (see Annex 5). Oxen are used more extensively, in therefore survey available evidence on labor use by region the lower density Northern regions, where almost one-third before turning to examine the use of other inputs such as of the total area is cultivated using draft animals. This farm implements, seeds, and fertilizer that can increase the option is precluded in the Southern region as no land is productivity of land and labor. Three findings are signifi- available for growing fodder. The prevalence of hand tools cant: First, on-farm labor use increased commensurately in Cameroon, used by 85 percent of the farming population, with higher densities, especially in areas that tended to is less of a handicap to land productivity given the specialize in export crops or food crops for the market; abundance of land that can be brought under crops and second, the use of hired labor is correspondingly higher in consequently the initially much higher returns to labor (see high income areas; and third, data on consumption of Annex 1). fertilizer and improved seed indicate that the model of We now turn to examine other inputs that increase the increasing input use with higher densities is at best only productivity of labor, such as fertilizer and seed. partially true, even for the most land-scarce countries; In countries that have pursued a deliberate policy of government priority for promoting fertilizer use has been smallholder intensification, such as Kenya, the use of determined by other priorities (Lele, Christiansen, and purchased inputs like fertilizer and seed is much more Kadiresan 1989). In Kenya improved seed adoption has common and corresponds to areas of high potential and increased to 60 percent, but fertilizer use on small farms is high density (see Figure 28). According to the 1978 survey, apparently growing less impressively. The reverse is true for farmers in the Central province of Kenya applied four times Malawi, suggesting the absence of a well coordinated more fertilizer per hectare than did those in its closest strategy emphasizing the complementarity of inputs. The competitor, the Eastern province-116 as compared to 27 evidence supports the contention that at early stages of kilograms per hectare. The Central province also accounted development, national and regional policy initiatives will be for over half of all sprays, seeds, feeds, and hired labor of critical importance in adopting inputs to improve factor used in the smallholder sector for that year (see Annex 4). productivity. Because world prices for coffee and tea were reflected in The cases of Cameroon and Malawi, both of which have producer prices, the production of higher value crops and excellent and up-to-date rural survey data compared with the more intensive use of land naturally gravitated to the 34 Table 15 Family, hired, and total labor working on farms by province in Cameroon Family Labor Hired Labor Total Labor Percent hired Province Average Average Average Labor in Number farm Number farm Number farm Total The North Far North 978,000 3.4 9,000 987,000 3.4 0.91% North 286,000 2.9 4,000 290,000 2.9 1.38% Adamaoua 171,000 3.1 15,000 0.3 186,000 3.4 8.06% Subtotal 1,435,000 28,000 1,463,000 1.02% Tropical Rainforest East 209,000 3.1 1,000 * 210,000 3.1 0.48% Central 542,000 3.3 8,000 550,000 3.3 1.45% South 172,000 3.1 2,000 174,000 3.1 1.15% Subtotal 923,000 11,000 934,000 1.18% Western Lowlands Littoral 201,000 3.1 9,000 0.1 210,000 3.2 4.29% Southwest 276,000 3.7 17,000 0.2 293,000 3.9 5.80% Subtotal 477,000 26,000 503,000 5.17% Western Highlands Northwest 546,000 4.1 9,000 0.1 555,000 4.2 1.62%h West 763,000 4.8 4,000 767,000 4.8 0.52% Subtotal 1,309,000 13,000 1,322,000 0.98% Total 4,144,000 3.6 78,000 0.1 4,222,000 3.7 1.85% Notes: Total number who worked on farm 30 days or more during 1984 crop year. Hired labor includes permanent labor only. *Less than 0.1 worker average. Source: 1984 Agricultural Census. Table 16 Smaliholder land distribution in Malawi, 1980/81 Size of Households Area Cultivated holding Cumulative Total area Cumulative Average per (hectares) Total % % ('000 Ha) % % Household Total 1135.6 100.0 - 1332.0 100.0 - 1.2 Under 0.5 267.4 23.5 23.5 80.6 6.1 6.1 0.3 0.5-0.99 356.0 31.4 54.9 258.5 19.4 25.5 0.7 1.0-1.49 215.9 19.0 73.9 265.2 19.9 45.4 1.2 1.5-1.99 121.5 10.7 84.6 209.9 15.8 61.1 1.7 2.0-2.99 118.2 10.4 95.0 283.8 21.3 82.4 2.4 3 and Over 56.6 5.0 100.0 234.1 17.6 100.0 4.1 Source: Govemment of Malawi 1 984b. Central province. As a result, incomes there were the COTON projects in the region, reinforcing the argument for highest in Kenya outside Nairobi, but distribution was the policy-led intensification. worst, confirming the Kuznetzian view that income inequal- Similarly, many attribute high rates of input use in the ities may initially worsen with growth before they improve. Southern region of Tanzania to explicit public policy In Cameroon, another case presents itself: Input use is objectives. Less than 10 percent of all fertilizer was applied concentrated both in the higher density Western Highlands in the high potential Northeast Highlands, but the Iringa and in targeted cotton-producing regions in the North. Table region of the Southern Highlands (with a relatively low 17 indicates that the ratio of farms using fertilizer and density of 20 persons per square kilometer) accounted for purchased seed in the highlands reached 74 and 64 22 percent of all fertilizer and 13 percent of the seed in percent, respectively-about 20 percentage points above 1980; by no small coincidence it also had five of the twelve the national average. Surprisingly, in the lower density state-financed national retail outlets serving farmers in 1980 Northern region (with 17 persons/per square kilometer), the (see Table 18). This suggests room for increasing yields and ratio of farms using fertilizer was not much less: 61 percent. adds weight to the idea that input use follows regional It would be useful to have data on levels of fertilizer planning more closely than it does population density. application by regions and family size to carry out more Fertilizer use in Nigeria is directly related to state policy. detailed work, but such farm surveys are limited in Africa. Since 1977, the subsidy on fertilizer has been on average Those familiar with Cameroon attribute greater fertilizer use about 25 percent of the total agricultural budget. Nearly in the North to the success of state-sponsored SODE- two-thirds of the total 580,000 metric tons of product 35 Figure 27 Figure 28 Agricultural wage labor in Malawi by region, 1977-84 Fertilizer purchases in Kenya by region, 1976-79 Thousands Thousand Kenya pounds 130 32 120 - 30 " FODRESTRY A KD FISHEI NG, B Y P' vI ON a r 7 - / 4 \ ~~ ~ ~~~~~~~~~~~~~~~~~~~~~~~20 l Ito - 2 1 26 100 24 7 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~22 70 -0 16 I0 02 -990 190 1 I907 1931 1302 103321I2I 2 a Northern + Central K Southern Central Nyanza Eastern Rift Western Coast Vallev Source: National Statistical Office 1985 Source: Rural Integrated Surveys 1976-79. Table 17 Population density, proportion of land cultivated, and ratios of farms using purchased inputs in Cameroon (persons per square kilometer) Region Population Proportion of Farms Ratio of farms Farms Ratio of farms Province density land cultivated purchasing purchasing seeds using using fertilizer to 1986 (%) seeds to total farms fertilizer total Far North 50.4 12.0% 103,400 39% 182,900 68% North 9.0 2.2% 40,500 42% 61,100 63% Adamaoua 6.8 1.3% 12,700 24% 13,400 25% The North 16.8 3.9% 156,600 37% 257,400 61% East 4.4 1.3% 27,500 41% 17,700 27% Center 25.4 3.8% 90,700 56% 4,700 3% South 8.6 2.4% 24,100 44% 400 1 % Tropical Rainforest 11.7 2.3% 142,300 50% 22,800 8% Littoral 83.0 4.0% 43,900 69% 29,300 46% Southwest 33.1 8.0% 48,600 66% 17,400 24% Western Lowlands 55.4 6.2% 92,500 67% 46,700 34% Northwest 70.6 13.2% 92,800 71% 58,400 45% West 95.8 21.1% 121,600 77% 126,700 80% Western Highlands 81.8 16.7% 214,400 74% 185,100 64% Total 22.4 4.2% 605,800 54% 512,000 45% Source: Land data from Bilan Diagnostic, Ministry of Agriculture 1986. Agricultural Census 1984 table 38. consumed in 1984 went to the Northern states (see Figure average of 33 percent for fertilizer and 17 percent for 29). Food crops account for 80 percent of all fertilizer use purchased seed, at 23 and 8 percent, respectively (see (Lele, Christiansen, Kadiresan 1989). The strong regional Figure 30). The percentage of households purchasing seeds emphasis to fertilizer policy apparently does not comple- from ADMARC is also highest in the Central and to a lesser ment regional potential; responses are reportedly higher in extent the Northern region (see Annex 5). Both the North- the low density Middle Belt states. Data on soils from FAO ern and Southern regions of Malawi have a relatively lower (see Annex 6) suggest that the majority of low productivity population density than the Central region. In the North, soils in Nigeria are located in the South. the resultant greater land availability has contributed to the In Malawi, land has become so scarce in the Southern low level of intensification through increased use of inputs, region that small farmers can no longer produce enough whereas in the South the small farmers have lacked the food to feed their own families, let alone purchase inputs financial means and tlhe ability to undertake the risks on the market. In the southern parts of Malawi, the ratio of associated with the purchase and utilization of fertilizer and households using inputs is significantly below the national hybrid seed. The degree of population pressure in the 36 Table 18 Fertilizer use, purchased seeds, and irrigated area in Tanzania by region, 1980 Population Fertilizer use Purchased Retail Estimated As % of Area density 1980 (MT) grain seed outlets irrigated cultivated 1986 1980 area 1973 area Region Export Food Share of 1980 Share of (ha.) per/sq.km crops crops total (MT) total Northeast Highlands 25 4,071 4,639 8.8% 1,213 22.1% 2 63,854 18.8% Arusha 15 1,800 846 2.7% 973 17.7% 1 19,394 11.8% Kilimanjaro 85 2,271 3,793 6.1% 240 4.4% 1 44,460 25.4% Coastal Belt 21 4,211 5,973 10.3% 1,327 24.2% 4 11,692 0.9% Coast 18 550 678 1.2% 331 6.0% 2 660 0.3% Lindi 9 Mtwara 54 251 177 0.4% 11 0.2% 238 0.1% Tanga 48 3,410 465 3.9% 436 7.9% 1 4,535 1.3% Morogoro 17 - 4,653 4.7% 549 10.0% 1 6,259 1.6% Central and Western 19 5,800 7,946 13.9% 374 6.8% 3,687 0.4% Dodoma 29 - 319 0.3% 243 4.4% 1,857 0.7% Singida 15 287 0.2% Tabora 15 5,200 6,743 12.1% 81 1.5% 1,213 0.5% Kigoma 22 600 884 1.5% 50 0.9% 1 330 0.1% Southern Highlands 15 26,888 31,707 59.4% 1,473 26.8% 6 23,393 3.4% Mbeya 23 10,969 4,116 15.3% 238 4.3% 1 7,499 2.9% Iringa 20 8,030 14,090 22.4% 730 13.3% 5 1,233 0.5% Ruvuma 11 7,455 9,220 16.9% 23 0.4% 14,661 12.2% Rukwa 9 434 4,281 4.8% 482 8.8% Lake Victoria Basin 48 4,594 2,858 7.6% 823 15.0% 23,944 1.9% Mwanza 91 1,566 1,231 2.8% 177 3.2% 3,109 0.8% Mara 41 475 567 1.1% 320 5.8% Shinyanga 34 1,770 837 2.6% 283 5.2% 14,204 4.2% Kagera 47 783 223 1.0% 43 0.8% 6,631 2.3% Total 25 45,564 53,123 100.0% 5,489 100.0% 12 126,570 2.8% Source: FAO/World Bank 1987; World Bank 1983. Figure 29 Figure 30 Fertilizer consumption in Nigeria by region, 1984 Fertilizer use in Malawi by region, 1981 Total = 580,000 metric tons Percent of total households Southern X States-12% States-3 Source: Lele, Oyejide, et a). 1989. Karonga Mzuzu Kasur,au Salima Lilongwe Liwondex lantyre Ngbu Northern Central m Southern region, coupled with the failure to intensify agriculture there, has reached the point where it no longer Source: National Statistical Office 1984b. acts as a positive inducement to intensify production but rather has begun a downward spiral of declining fertility, policy as they do with high population densities: The still declining input use, and declining output. An unfortunate inadequate access to sources of cash, credit (less than 20 omission from the Boserup hypothesis is the effects of percent of all small farmers receive credit), and purchased inadequate public policy. In Malawi, the problems arising inputs have stifled the autonomous movement toward from population pressure have as much to do with poor intensification. 37 Conclusion Whether higher population densities are an important aid the fact that there is less congruence between land to development or a hindrance will remain an intensely potential and population densities due to factors such as complex and highly controversial issue. Boserup provides disease, cultural barriers to migration, and colonial an intellectual justification for high population densities; a patterns of investments in infrastructure. powerful body of opinion in Africa believes that higher 4. The political and welfare considerations of including the population densities are necessary and desirable for future largest proportion of people in the growth process have development. Fertility is highly valued culturally at the local influenced past patterns of public policy toward regional level, and children are seen as assets in labor and development. These considerations have been insurance for old age; many social and cultural factors that addressed differently in various countries. Only in Kenya resist empirical analysis will shape a country's movement and to a lesser extent in Cameroon did they achieve toward more intensive, productive, and tenable use of land. broad-based growth by using their regional comparative In this paper we have shown that: advantages. Elsewhere policies resulted in considerable I. Data on some of the most basic facts needed to plan redistribution in the sources of production and perhaps agricultural development are scarce in Africa. They raise helped to commercialize agriculture. Countries will need more questions than they answer. to make difficult choices in the future to realize growth. 2. Targeting policies and investments in the areas of high 5. The policy-led process of intensification conceived here productive potential and high population densities offers is different than the autonomous intensification envis- the greatest scope for achieving growth in the short and aged by Boserup. Its implications are outlined in the medium run. Summary and Policy Recommendations section and will 3. Achievement of this objective is complicated in Africa by not be repeated here. 38 -------- ---- ------------ -- -- -- -- -------- --- - 8 o o o - - - --- -- C- _ °o-n-C cC X' 0) SCC C S o 8~C- CoCO s a > X oe o 8 8>>8 e S s1 O * oe :. ..S3 o 4_ c Ec -8°°8 8 m8 u CCC . CoC* - C oC-&o! 0- ° g 2 a o mo g - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - o- - --- ------- ---- ---- --- -- -------- g 0 Ss0^wnf {- ~ ~~ *e CC COOC CO 7CO CM Co CCCwS , D ww E >uc E <§6oO 0g - C~~~~C O o: N _ t=C CWCCtO CC 7 gC r h8 8 8-,° CO C " _ . _ > -8 2 ; ° S sr to °~~~~~~~~ICCCo 8 r v O w > q o0 o0 o o c X c r n 8 4 _- ICa RS <~~~~~C COCC Co EoC COC IV 8 0 - CAC Cv8>EnOC' Co gi c , {_gc CCn< t > v _c> CO >, )>o O o_c o °ogga ° 'C o~c C ' naB gunI > "n 2 ~ ~ ~ O CO COCo Co C)^ CC Co CO COa CO CoE X8o _ _ c ° Co COC. S 8g 0 n 0 ro ( >b-h-gc ao snO CO CO CoO O > O 00O v ffi o E ^ R S r * X r 0 N X 8 x < 6 X ~~~~~~OCei8oCn8°° 0 °°°>8°o Co CA* C Co e E0 .¢ <° C BO _O CO C E<°aa28io oC o Coooo°°°°8°°° ^1 ~ ~ ~ ~ CC C COCO E1C 22o o °C Co Co ee' C) o ;0E :82t E e CL cq~~~~~~~~CC . 0 ~ Co C.~COCC cm Table 3 Tonnage marketed by product and by province, 1985 (in metric tons) PROOUCTION IN (TRIC TONNES Cocoa Coffee Coffee Cotton Tobacco Maize Sorgi,i Rice Cassava Yams Taro GrourutPlantain Bananas white Beans 011 Palm Arabica Robusta Potatoes 'O)) liters THE NORTH Far North 0 0 C 30,573 0 2,159 4,993 13,170 7,081 2,354 North 0 0 0 48,523 0 3,291 2,617 490 6,542 8W Adasaaoa 0 0 5,780 0 95 23,214 5,662 25,584 2,062 Su-Total 0 0 5,780 79,101 95 28,684 13,272 25,584 15,68, TOPICAL RAINFOREST East 6, 840 0 22,810 0 1,754 5,548 35,711 2,608 3,532 48,518 20,103 590 Center 45,880 0 7, 540 0 125 2,995 81,379 5,194 9,524 3,243 70,568 55,680 2,853 South 19,960 0 210 0 0 751 15,058 3,069 807 11,077 2,908 247 St-Total 72,680 0 30,560 0 1,879 9,294 132,149 5,194 15,200 7,582 130,164 78,690 3,690 WESTERN LEUNDS Llttoral 5,580 0 37,420 0 0 1,208 26,656 1,186 1,316 098 22,924 5,302 4,984 Swtlest 35,020 0 12,900 0 0 3,744 147,039 4,301 11,691 1,260 123.725 50,869 6 774 Sib-total 40,600 0 50,320 0 0 4,952 173,695 5,307 13,997 2,258 146,649 56,171 11,758 WESTERN HILMANOS Northeest 140 16,160 4,870 0 0 36,164 1,059 17,020 61,101 6,903 9,965 4,915 79,609 55,939 14,142 10,616 10,086 West 580 19,240 27,300 0 0 16,463 23,740 5,111 4,920 1,663 42,258 22,893 3,592 5,475 1,982 Sib-total 720 35,400 32,170 0 0 52,627 1,359 84,841 12,014 14,885 6,578 121,867 78,822 17,734 16,091 12,068 TOTAL 114,000 35,400 118,830 79,101 1,974 95,537 14,331 0 416,268 22,515 43,092 32,103 409,679 213,692 17,734 16,091 27,516 Gnrce: Agrioultural Census 1984 Table 4 Percentage of harvest marketed, 1985 Area -I OF PRODUCTION kARKETED IN A5TRIC TONNES - - - --- - Cilttvated Cocoa Coffee Coffee Cotton Tobacco Maize Sorpm Rlce Cassava Yams Taro GradrutPlantain Bananas White beans Oil Palm Arabica Rbousta Potatoes ' lO liters (Hectares) NORTH Far North 411,700 91.71 31-8X 3.5X 47.31 50.4x 24.0x North 1503 OW 99.33 25.10 6.40 28.7x 34.8X 24.4x Adamaoa 83,000 199.00 95.00 53.60 25.40 32.10 66.50 Sb-Total 644,709 TFOPICAL RAINFOiREST East 142,300 1O.0X 1O.03 96.48 21.00 18.1% 26.5x 37.90 33.63 47.30 20.08 Center 261,600 100.0S 192.0 73.58 19.40 21.80 28.00 33.70 17.53 41.20 48.09 12.1X .South 114.500 100.0% 199.0o 19.70 11.70 36.80 12.10 19.40 13.4X 6.0r Su-Total 518,400 WESTERN LIANDS Littoral 81,500 100.00 103.00 17.58 27.28 12.8S 11.49 25.80 36.10 11.51 50.70 Southwest 200,500 1O.0 100 o0 33.4X 48.4x 38.4X 23.7x 5s.00 50.59 30.19 40.3S Sub-total 282,000 WESTERN HIGHLANOS Northwest 229,100 19.00 loO.00 100 00 21.40 74.60 83.29 55.s8 34.90 25.03 41.9x 50.10 43.65 54.9X 48.3X 51.20 West 292,600 100.0S 100.00 1W.X0 14.6X 27.19 13.4X 12.2 15.79 33.3x 14.60 23.40 33.9S 42.8X Stb-total 521,700 TOTAL 1,966,800 Source: Agrlcultural Cansus 1984 40 Table 5 Rural per capita agricultural income, by department, 1985 Department CASH CROPS INCOME FOOD CROPS INCOME TOTAL CROPS INCOME SIARE OF CASH SHARE OF TOTAL RURAL POPULATION RURAL PER CAPITA PROVINCE ('000 CFA) ('000 CFA ) ('000 CFA) CROP INCOME INCOME 1985 INCOME 1985 1 Logone et Chari 417,888 417,888 0.0% 0.3% 228,656 1,828 2 Mayo Sava 348,870 95,467 444,337 0.5% 0.3% 185,681 2,393 3 Mayo Tsanaga 927,862 267,495 1,195,357 1.3% 0.8% 341,464 3,501 4 Damare 369,337 36,008 405,343 0.5% 0.3% 326,e32 1,241 s Mayo Danai 1,262,862 996,783 2,259,646 1.8% 1.6% 255,103 8,858 6 Kaele 1,047,156 192,832 1,239,988 1.S% 0.9% 191,057 6,490 FAR NORTH 3,956,087 2,006,471 5,962,558 5.7% 4.1% 1,528,563 3,901 7 Vayo Louti 1,310,776 310,540 1,621,316 1.9% 3.3% 197,079 8,227 8 Benoue 2,086,557 512,057 2,598,614 3.c% 1.8% 171,238 15,175 9 Faro 40,756 848,266 889,022 0.1% 0.6% 50,080 17,752 10 Mayo Rey 2,834,930 343,571 3,178,501 4.1% 2.2% 76,955 41,303 NORTH 6,273,019 2,014,434 8,287,453 9.0% 5.8% 49E,352 16,730 11 Faro Edecn 334,915 334,915 0.0% 0.2% 39,346 8,512 12 Vina 383,661 383,661 0.0% 0.3% 77,486 4,951 13 Mayo Banyo 176,811 406,966 583,777 0.3% 0.4% 82,743 7,055 14 Djekem 127,860 127,860 0.0% 0.1% 26,254 4,870 15 Mbere 69,645 4,258,465 4,328,110 0.1% 3.0% 93,996 46,046 ADAMAOUA 246,456 5,511,867 5,758,323 0.4% 4.0% 319,825 18,005 16 Lomedjerem 133,685 3,075,708 3,209,393 0.2% 2.2% 95,036 33,770 17 Kadel 1,359,579 1,482,457 2,842,036 1.9% 2.0% 119,662 23,751 18 Haut Nyong 1,829,202 1,533,286 3,362,488 2.6% 2.3% 100,621 33,417 19 Boumbae Ngoko 482,576 278,296 760,872 0.7% 0.5% 38,415 19,807 EAST 3,805,042 6,369,747 10,174,789 5.4% 7.1% 353,734 28,764 20 Mbam 5,436,397 2,574,990 8,011,387 7.8% 5.6% 202,662 39,531 21 Haute Sanaga 543,892 475,464 1,019,356 0.8% 0.7% 36,267 28,107 22 Lekie 7,295,808 3,922,837 11,218,645 10.4% 7.8% 280,892 43,001 23 Mefou 1,803,712 3,100,905 4,904,617 2.6% 3.4% 130,947 37,455 24 Nyong et Yfoumou 737,945 251,356 989,301 1.1% 0.7% 48,077 20,577 25 Mfoundi S3,698 211,454 265,152 0.1% 0.2% 42,357 6,260 26 Nyong et Kelle 1,139,840 2,796,410 3,936,250 1.6% 2.7% 75,662 52,024 27 Nyong et Soo 2,314,203 802,398 3,116,601 3.3% 2.2% 67,745 46,005 CENTER 19,325,495 14,135,814 33,461,309 27.6% 23.3% 864,609 38,701 28 Ocean 805,684 631,951 1,437,635 1.2% 1.0% 64,539 26,360 29 Ntem 3,418,171 975,305 4,393,476 4.9% 3.1% 127,513 34,455 30 Oja et Lobo 3,890,304 631,009 4,521,313 5.6% 3.1% 114,382 39,528 0.0% SOUTH 8,114,159 2,238,26s 10,352,424 11.6% 7.2% 296,434 34,923 3i Metchum 977,282 1,427,150 2,404,432 1.4% 1.7% 179,571 13,390 32 Ndonga Mentoum 656,390 2,451,280 3,107,670 0.9% 2.2% 240,688 12,912 33 Momo 154,563 1,216,031 1,370,594 0.2% 1.0% 105,389 13,005 34 Mezam 1,442,757 8,783,434 10,226,191 2.1% 7.1% 258,638 39,539 35 Mbui 749,736 1,417,179 2,166,915 1.1% 1.5X 210,331 10,302 NORTHWEST 3,980,728 15,295,074 19,275,802 5.7% 13.4% 994,617 19,380 36 Noun 382,559 2,741,061 3,123,620 0.5% 2.2% 193,016 36,383 37 Bamboutos 1,370,786 1,309,542 2,680,328 2.0% 1.9% 180,058 14,886 38 Menoua 1,921,357 668,501 2,589,858 2.7% 1.8% 296,329 8,740 39 Mifi 454,698 795,258 1,249,956 0.6% 0.9% 166,246 7,519 W0 Haut Nkam 103,851 261,966 365,817 0.1% 0.3% 71,196 5,138 t Nde 524,965 422,143 947,108 0.8% 0.7% 45,514 20,809 WEST 4,758,216 6,198,471 10,956,687 6.8% 7.6% 952,359 11,505 42 Manyu 1,412,102 3,820,424 5,032,526 2.0% 3.5% 108,358 46,444 43 Ndian 723,307 4,478,864 5,202,171 1.0% 3.6% 41,843 124,326 44 Meme 11,896,492 4,994,578 16,891,070 17.0% 11.7% 200,010 84,461 45 Fako 972,147 3,116,783 4,088,930 1.4% 2.8% 102,503 39,891 SOUTHWEST 15,004,048 16,210,649 31,214,697 21.4% 21.7% 452,714 68,950 48 Youngo 3,280,089 1,348,824 4,628,913 4.7% 3.2% 115,602 40,042 47 Nkam 21,768 1,098,438 1,120,206 0.0% 0.8% 34,180 32,774 48 Sanaga Maritime 1,206,857 1,184,258 2,391,115 1.7% 1.7% 85,096 28,099 49 Wouri 212,064 212,064 0.0% 0.1% 67,240 3,1S4 LITTORAL 4,508,714 3,843,584 8,352,298 6.4% 5.8% 302,118 27,646 TOTAL 89,971,964 73,824,376 143,796,340 100% 100% 6,560,325 21,919 41 Source: BCEOM Inventory of Feeder Roads, 1985. Cited in Gaviria, 1988. Rural Population calculated from Sixth Plan, 1985. Table 6 Population density, proportion of land cultivated, and ratios of farms using purchased inputs REGION Powolatlon .Prooprtlon Farms Ratio of Farms Farms Ratlo of Farms Province Density of Land Purchasing Purchasing Seeds Using Using Fertl- 1986 Cultivated Seeds to Total Farms Ferti lzer lizer to Total Far North 50.4 12.0D 103,400 39% 182,900 68% North 9.0 2.21 40,500 42X 61.100 63% Adamaoua 6.8 1.3X 12,700 24X 13,400 25X THE NORTH 16.8 3.9% 156,600 37% 257,400 61% East 4.4 1.3% 27,500 41% 17,700 27% Center 25.4 3.8% 90,700 56X 4,700 3% South 8.6 2.4% 24,100 44% 400 1% TOPICAL RAINFOREST 11.7 2.3% 142,300 50% 22,800 8% Littoral 83.0 4.01 43,900 69% 29,300 46X Source: Southwest 33.1 8.01 48,600 661 17.400 24% Ministry of Siculan Diagnostic, WESTERN LOWLANDS 55.4 6.2% 92,500 67% 46,700 34% Population data from Sixth Plan, 1986. Northwest 70.6 13.2% 92,800 71% 58,400 45X Use of Modern Inputs data from West 95.8 21.1% 121,600 77% 126,700 801 Agricultural Census, 1984. WESTERN HIGHLANDS 81.8 16.7% 214,400 74% 185,100 64% TOTAL 22.4 4.21 605,800 54% 512,000 45% Table 7 Transport owned N&unber of Farms Owning Transport by Type of Trarsport Owred N.nber of Farms that lsed Tractors and Carts Province Total Farms Farms OvinIg TYPE OF TRANSFtX. T OWNED Trarsport 1/ Cart Bicycle M4otorcycle Automoblle TrickALorry Tractor Cart rumber/(percent) 2/ THE NORTH Extreme North 285,400 101,300 4,100 95,200 17,500 400 3W 6,900 23,300 (35.5) (1.4) (33.4) (6.1) (0.1) (0.1) (2.4) (8.2) North 98,700 32,700 4,800 29,800 3,900 300 100 6,900 8,800 (33.1) (4.9) (30.2) (4.0) (0.3) (0.1) (7.0) (8.9) Adamaoua 55,600 11,000 2,600 6,700 4,000 1,100 200 2,800 2,900 (19.8) (4.7) (12.1) (7.2) (2.0) (0.4) (5.0) (5.2) 9SLb-Total 439,700 145,000 11,500 131,700 25,400 1,800 600 16,600 35,000 (33.0) (2.6) (30.0) (5.8) (0.4) (0.1) (3.8) (8.0) TROPICAL East 66,700 15,200 2,700 10,900 3,600 900 500 1,100 2,700 RAINFOREST (22.8) (4.0) (16.3) (5.4) (1.3) (0.7) (1.6) (4.0) Central 162,900 27,600 13,400 15,900 9,400 2,800 BOO 2,000 17,800 (16.9) (8.2) (9.8) (5.8) (1.7) (0.4) (1.2) (10.9) South 55,100 9,600 6,500 3,700 5,600 700 200 <100 7,600 (17.4) (11.8) (6.7) (10.2) (1.3) (0.4) (13.8) Sub-Total 284,700 52,400 22,600 30,500 18,600 4,4W 1,300 3,100 28,100 (18.4) (7.9) (10.7) (6.5) (1.5) (0.5) (1.1) (9.9) WESTERN Llttoral 65,400 9,000 8,500 4,900 2,200 1,700 600 <100 9,900 LOWL-ANDS (13.8) (13.0) (7.5) (3.4) (2.6) (0.9) (15.1) Southwest 74,600 14,000 8,300 7,000 5,000 3,600 700 <100 14,500 (18.8) (11.1) (9.4) (6.7) (4.8) (0.9) (19.4) Sb-Total 140,000 23,000 16,800 11,900 7,200 5,300 1,300 24,400 (16.4) (12.0) (8.5) (5.1) (3.8) (0.9) (17.4) WESTERN Northwest 131,800 25,400 400 18,400 4,500 3,500 1,600 1,200 2,300 (19.3) (0.3) (14.0) (3.4) (2.7) (1.2) (0.9) (1.7) HIGH1LANDS West 159,300 51,600 24,000 30,400 19,500 4,700 1,100 100 34,500 (32.4) (15.1) (19.1) (12.2) (3.0) (0.7) (0.1) (21.7) Sub-Total 291,100 77,000 24,400 48,800 24,000 8,200 2,700 1,300 36,800 (26.5) (8.4) (16.8) (8.2) (2.8) (0.9) (0.4) (12.6) TOTAL TRADITIONAL 1,155,500 297,400 75,300 222,900 75,200 19,700 5,900 21,000 124,300 (25.7) (6.5) (19.3) (6.5) (1.7) (0.5) (1.8) (10.8) 42 1/ Parts do not sun to totals due to nultIple counts 2/ Percentages expressed In terms of total farms and sl-iwn In parentheses. Source: 1984 Agricultural Census; cited In Gavirla, 1988. Table 8 Farming method used to cultivate fields and province (first crop cycle only) Hand Forms Province Only Tractors Cattle Donkeyx With Crops 1/ ------------------------------ number/percent 2/ -------------------------- THE NORTH Far North 188,800 8,0oo 83,800 8,000 265,100 (83.6) (2.8) (31.6) (2.3) (100.0) North 38,700 8,800 48,300 4,500 98,300 (40.2) (7.1) (48.0) (4.7) (100.0) Adamaoua 48,900 2,800 4,100 100 53,900 (87.0) (5.2) (7.8) (0.2) (100.0) Sub-Total 254,200 18,500 134,000 10,800 415,300 (81.2) (4.0) (32.3) (2.8) (100.0) TROPICAL RAINFOREST East 86,700 1,000 3/ 3/ 68,700 (98.6) (1.5) (100.0) Central 180,000 2,000 3/ 3/ 182,000 (98.8) (1.2) (100.0) South 55,000 3/ 3/ 3/ 55,000 (100.0) (100.0) Sub-Total 280,700 3,000 283,700 (98.9) (1.1) (100.0) WESTERN LOWLANDS Littoral 84,000 3/ 3/ 3/ 84,000 (100.0) (100.0) Southwest 73,500 3/ 3/ 3/ 73,500 (100.0) (100.0) Sub-Total 137,600 137,500 (100.0) (100.0) WESTERN HIGHLANDS Northwest 128,900 1,200 100 400 130,800 (98.7) (0.9) (0.1) (0.3) (100.0) West 158,700 3/ 3/ 3/ 188,700 (100.0) (100.0) Sub-Total 287,800 289,300 (99.4) (100.0) Total Traditional 980,000 20,700 134,100 11,000 1,125,800 (85.3) (1.8) (11.9) (1.0) (100.0) I/ Includes only forms with first cycle crops. SOURCE: 1984 AGRICULTURAL CENSUS 2/ Percentages shown in parentheses. 3/ Less than 100 farms. Table 9 Breakdown ot planned regional investments, by province, 1971-1986 (in million FCFA) PROVINCE THIRD PLAN FWRTH PLAN FIFTH PLAN Planned % Per Plarred X Per Plamed % Per Spending of Total CapIta SperdIng of Total Capita Spwnding of Total Capita (In CFA) (In CFA) (In CFA) NORTH 84,336 24.3% 37,768 51,219 9.3% 20,829 254,000 16.4% 92,112 EAST 8,478 2.4% 23,291 9,768 1.8% 23 038 63,000 4.1% 132,381 CENTRAL-SOUTH 83,147 24.0% 55,729 91,221 16.5% 53,221 471,000 30.4% 218,167 LITTORAL 117,610 33.9% 125,786 301,373 54.7% 267,887 416,000 26.8% 247,973 SOUTHEST 22,326 6.4% 36,010 89,429 16.2% 129,046 124,000 8.0% 150,358 NORTHWEST 6,886 2.0% 7,019 2,745 0.5% 2,558 60,000 3.9% 49,120 WEST 24,142 7.0% 23,303 5,439 1.0% 4,653 163,000 10.5% 122,529 CAMEROON 346,925 100.0% 45,285 551,194 100.0% 63,670 1,551,000 100.0% 148,472 Source: The World Bank, Agrlwltural Sector Review, 1986. Note: Flgures in 1980/81 FCFA. Does not Include recurrent expenditure. 43 Annex 2: Tanzania Table 1 Population density and per capita agricultural land by region, 1986 and 2000 AREA - --- ------- Pnilation ('00) ---------…-- ----- Land ('000 Ha.) ---- C- 0.'tivatEo : P,pulation ' Per Capita Region Total X of Urban X Rural Rate of Total Total Cultivated Cultivable as percent of Derity Land 1988 1/ Total 1988 i/ 1,988 Growth 1/ 2000 2/ 3/ 1970 3/ 1970 3/ Cultivable 1 1,988 1,988 Dar-es-Salaarr 1,361 6X 1,361 0 J9.8X 2,708 139 n.a. r:.a. n.a. 977 n.a. NORTHEAST HIGHLANDS 2,461 11% 97 96% 3,646 9,535 340 8,889 3.8%: 26 3.6 Arusha 1,352 6% 103% 3.8% 1,S92 8,210 165 7,775 2.1%1 16 5.8 Klilmanjaro 1,109 5% 97 91X 2.1X 1,744 1,325 175 1,114 15.781 84 1.0 COASTAL BELT 4,681 21% 237 95X 7,038 21,260 1,336 10,501 12.7Xj 221 2.2 Coast 638 3% 10% 2.1% 913 3,255 255 2,643 9.681 20 : 4.1 LlIrdI 647 3% 42 94% 2.0% 956 6,604 148 5/ 1,221 5/ 12.1%: 10 1 1.9 Mtwara 889 4% 77 91% 1.4X 1,391 1,671 213 5/ 1.758 5/ 12.1%X 53 ' 2.0 Tanga 1,284 6% 1O% 2.1% 1,9f9 2,668 340 2,118 16.1%: 48 : 1.6 Morogoro 1,223 5% 118 90% 2.69 1,809 7,062 380 2,761 13.8%X 17: 2.3 CENTRAL AND WESTERN 3,921 17% 464 88% 6,036 20,384 849 10,122 8.4%: 19 : 2.6 Doroma 1,238 5% 204 84% 2.43 1,866 4,131 265 3,511 7.5%X 30 1 2.8 Slnrglda 792 4% 81 90% 2.5% 1,163 4,°34 160 2,960 5.4X' 16 : 3.7 TaDora 1,036 5% 94 91% 2.4% 1,761 7,f,15 144 6/ 2,440 6/ 5.9%: 14 2.4 Klgoma 855 4X 85 90% 2.5% 1,246 3,704 280 1,211 23.1%1 23 j 1.4 SOUTHERN HIGHLANDS 4,163 18% 417 90% 5,971 24,950 695 10,033 6.9X1 17 j 2.4 I;beya 1,476 7% 153 90% 3.1X 2,138 6,035 255 3,518 7.2%: 24 i 2.4 Iringa 1,209 5% 85 4/ 93% 2.7% 1,753 5,685 240 3,343 7.2%i 21 2.8 Rkuuaa 703 3% 87 89X 3.4% 1,105 6,367 120 1,572 7/ 7.693 12 : 2.0 kukwa 695 3% 92 87% 4.38 975 6,894 80 6/ 1,600 6/ 5.0%: 10; 2.3 LAKE VICTORIA BASIN 5,948 26% 217 96% 8,898 12,096 1,245 9,165 13.6%X 49 : 1.5 Mwanza 1,878 8% 100% 2.68 ,,770 1,368 410 1,382 29.71 95i 0.7 Mara 971 4% 69 93% 2.9X 1,372 2,176 205 2,137 9.6%1 45 i 2.2 Shinyarga 1,773 8% 101 94% 2.9% 2,665 5,076 340 3,315 10.3%: 35 | 1.9 Kagera 1,326 6% 47 96% 2.78 2,091 2,846 290 2,331 12.4%1 47 1.8 Total Malnland 22,535 1OO% 2,793 88% 2.8% 34,297 88,324 4,465 48,710 9.2%X 260 2.2 Sources: 1/ 1988 Population Censis, Preliminary Results. Bureau of StatistIcs. 5/ Derivec; tltal figzre of 360,090 la. g:ven for Mtwara and Lindi, combined. 2/ By calculatlon, using 3.181%. From The Demograpny of Tarnzanla 6/ Cerived total figpre of 2250.01 given for Tabora and Rukwa, combined. 3/ 1970 Statistical Abstract. Clted In 1974 Agricl.ltural and Rkural Sector Study, 7/ From Vin Valthiznur, "Ar, Asses &nt of Land Resources for Rainfed Maize, World Bar4k, Vol. 111, Table 23. '0ultivable" dces not Irolure forests. Wheat and Rice in Tanzania." Soutl;rn Er V 9opartnont, Tin florid Bank. 4/ By calculation; initlal censoc report In error. .ine, 1988. Original flgu9-e 160,000 ha. Table 2 Share of marketed production, by region (selected years) AREA Export Production Food Production Region Crops Share Crops Share Dar-es-Saleas NORTHEAST HIGHLANDS By RegIon Arusha Coffee A. 15% Maize 297 Who~at 75% Kilimanjaro Coffee A. 36X COASTAL BELT Coast Cashews 27% Lindi Cashews 22% Cassava 51% Mtwara Cashews 42% Cattle 28% Tanga Cat .Ie 20% Morogoro CENTRAL AND WESTERN PLATEAU Dodoma Maize 13% Scrpgn 43% Singida Tabora Rice 25X KIgoma SOUTHERN HIGHLANDS II2eya Coffee A. 218 Rice, 528 I r I nga Coffee A.21S Maieze 12% Sources: Ruvusa Coffee A. 15% MaZa 12% Rikwa aZe, 11% W ize data fro2 f&AOB. Based cc six-year 11976-1982) oean shares of Nr tcrass. Cassaoa data from NC. Based on six-year (1976-1982) mean share af marketed prodution. weat data from fM8/w. Based on six-year (1976-1982) mean share of marketeo prordctlon. LAKE VICTORIA BASIN .ROe data from W.O.A./DSt (Price Policy Recarumdatlons for 1981-2). Based on five-year (n97a5-198) purchases of rice ard paddy (converted at rate of 95%). Mwanza Cotton 38X Cassava 21X sorfw Oate from MC. Based on six-year (1976-1982) mean share of marketed proeactlon. Mara Cattle Coffe data from TnOe. Shares represent TCM wercrases of Arabica anr Robusta for 19S/86. Shlnyanga Cotton 38% Rica 12% Cattle data from Statistlcal Abstract, 1973-1679. Total Indigeous cattle holdirgs en Kagera Coffee R. °J9Xlar9-scaIe farms (lo. of cattle less sinifilant for 'UJamea rateory. cotton data from TC18 (as cIted In Agr. sector Renrot 1987). Based m share of pu,rchased sew castes i IS894/85. *Casme data from Tnae (as cited in FAO, om. cit). Based o sax-year (1S80-1888) mean share 44 shara of sarketed prodhctloo. If' Mz ~~ g tOeon8 m N.008 o8 8 P Eg 228 PI ---- --- --- --- ---- --- --- --- - f0- N 0 0 0 0 9E~Z ----------------------- - -- - 8 - g - - Ul) Z~~~~~~~(000'1 )r D iAN.04cO8M° -8 8 _ ,- V 0'gsS _ _- -- - -- - -- - - -- - -- - - 0 0 .888o>X,or|t . n< _ _ < . . .r 8 (-40 C') ---- 'A W 0 ------------------------ 00-- - - - - - -> - - - > E g ° - - - - - 00 _______s 08 8e -ac, srec4--- ~ |8- M88gHH§M"t g X(0 N Ns > e ~z 8o°e 80'rA°r ° ot~0 04XRg04'0 40 - o8wrO8r°8v oOO, |~~~~~~~~~~~~~~~~N 8 0008S4 CO >'A C)0 0 0 N. 00 0 oor°0°8lr°8OSXr r i-g G!i~ ~ ~ ~ 0 04o8o<8988< >XSiv gieEU 0 80Sv E o8suooz88s8ar 8 : '3A0000'A00008o- Ee 00000088 0 ° Zo 8 02 9u0n 8nr 8 t r~~~~~~~' 'A r0 ' 04/ )c -F 8° o888o8 i*-s 0 i nc -g-cR,i8 8Os8<8C N. 44iggO-zRFRR__5 rl <72r 8_88r 8 82 0 oWK co~~~~ ~ ~ ~ l 80 0400 - - * a)s rs >cs ;t_* * N. S~ Q g$88 --- 8 -)° O8 18r 3gr ta << n=. ez _ _X<-88ir-8^ 80080 00008 O88 o 0) 8o8Q 0 8 8 8r 8 E ~~~~~~~~~~~~~~~~~~~8 8 a8 8 8QC88g8E8Rgi a,0 .°r mov o rr r>° c c - 0 o a40 00c Table 5 Shares of NMC purchases by region, 1970/71-1987/881 1970/71 197172 1972n3 1973/74 1974/75 1975/76 1976/77 1977/78 1978/79 1979/80 1980/81 1981/82 1982/83 1983/84 1984/85 1985/86 l9S6/87 1987/88 CstJLAU 0.08 0.0 0.0 0.OS 0.0 1.6S 1.4X 1.0S 0.4S 0.0 0.0 0.08 0.1S 0.18 0.08 0.0 0.08 0.08 Uorogoro 3.68 9.18 9.08 7.38 4.2S 11.58 5.38 6.6S 2.18 0.78 0.98 0.58 0.38 0.18 0.78 0.48 0.4X 0.68 Targa 0.58 0.28 0.08 0.08 15.98 22.28 12.08 3.3S 3.38 0.28 0.18 1.58 3.48 1.4S 3.45 0.38 2.18 2.28 Utwara 0.08 0.08 0.08 0.08 0.08 3.08 2.58 0.8X 0.58 0.08 0.28 0.08 0.08 0.18 0.08 0.38 0.18 1.18 Llndi 0.0 0.08 0.08 0.08 0.08 1.38 1.68 1.48 0.98 0.08 0.2S 0.48 0.5 0.OX 0.18 0.48 0.48 0.58 Arusha 24.28 17.78 18.18 9.58 12.18 11.18 34.9S 31.58 31.78 29.4S 20.0S 3.68 1.4X 8.98 3.38 20.28 26.98 25.18 KilimanJaro 8.68 6.78 11.18 8.1X 20.18 5.3S 3.5X 10.48 6.28 3.78 0.2S 0.0% 068 0. 18 O. 18 0.48 0.08 0.28 Dodma 31.48 36.38 50.88 46.78 00.8 6.6S 6.68 8.18 16.78 16.88 3.28 4.98 1.78 7.48 1.38 6.78 4.18 5.6S Sirigida 2.98 2.38 0.78 2.28 0.0 0.5 0.68 0.58 1.68 0.48 0.58 0.38 0.08 0.18 0.18 2.88 2.78 2.88 Tabora 0.78 2.18 0.58 0.08 0.08 0.18 2.08 4.9X 2.6S 3.0S 2.98 1.3S 2.38 0.48 0.48 0.88 1.38 1.7S Klgoa 0.18 0.08 0..OS 0.08 0.08 0.2 0.48 0.48 0.58 0.28 0.28 0.58 0.48 0.48 0.28 0.18 0.58 0.48 Rubta 0.08 O.8 0.08 0.0 2.98 3.38 6.88 3.98 2.48 9.9S 21.5X 17.88 20.58 14.38 18.58 16.48 16.38 20.58 UWanza 0.78 0.2n 0.28 0.58 00.8 3.28 0.88 1.18 1.98 1.28 0.08 0.08 0.08 O.OS 0.08 2.8S 1.28 0.28 Mara 5.4S 3.7X 3.48 8.48 7.18 1.28 3.48 2.58 1.98 1.98 0.18 0.38 0.18 1.58 0.28 0.08 0.28 0.38 Shinyanga 0. 18 3.38 0.08 0.08 0.0 0.88 O.08 1.18 1.18 0.75 0.2S 0.3S 0.38 0.18 0.28 1.58 2.48 3.1S Kagera 0.08 0.0 O..0 0.08 0.08 0.2S 0.6X 0.6X 0.4S 0.4S 0.18 0.08 0.18 0.68 0.08 0. 1S 0.68 0.38 rirnga 19.68 17.9X 7.78 15.2X 17.28 11.5 8.5S 9.5X 12.48 16.38 26.38 37.08 30.48 35.48 25.78 21.38 21.18 11.28 Uoeya 1.3S 05.8 0.18 1.98 2.98 2.48 3.28 5.38 3.18 4.08 6.58 8.08 11.08 10.98 8.28 9.08 6.88 4.7S RAua 0.98 0. 08 0.58 0.18 17.68 13.98 5.88 7.32 10.48 11.08 17.08 23.68 26.58 18.18 37.68 16.38 12.98 19.58 Source: 197u/71-1979/80 data from Ministry of Agriculture, 'Price Policy RecorTendatIons for the 1981-82 Agricultural Price Review, 1980 1980/81-1987/88 data froa Goverrnw,t of Tanzania, Mln. of Ag. & Livestock DevelopIct, 'Annal Revlew of Maize, Rice and Wheat, '1987 Notes: 1/ Shares were calculated assuaing that blarks eq.al zero purchases, whilch ray not be a correct assustion. Table 6 NMC sales of maize by region, 1978/79-1987/88 ('000 mt) 1978/79 1979/80 1980/81 1981/82 1982/83 1983/84 1984/85 1987/88 Cst/DSM 88.0 107.0 133.0 137.0 127.8 127.9 135.2 100.8 MDrogoro 3.0 8.0 8.0 9.0 4.8 4.8 3.0 5.4 Tanga 15.0 26.0 31.0 13.0 3.7 3.7 8.4 5.4 Mtwara 4.0 4.0 7.0 6.0 3.5 3.6 3.2 3.1 LIMidl 3.0 5.0 6.0 3.0 2.4 2.4 1.1 2.4 Arusha 4.0 16.0 17.0 10.0 7.9 7.9 6.0 6.8 Killmanjaro 2.0 8.0 10.0 8.0 2.0 2.1 6.5 2.0 Dodoma 5.0 15.0 16.0 30.0 21.6 21.8 11.2 20.5 Singida 2.0 7.0 2.0 2.0 0.7 0.7 2.0 5.0 Tabora 1.0 2.0 5.0 9.0 2.4 2.4 6.8 3.0 Klgona 0.0 1.0 1.0 2.0 1.4 1.4 1.8 1.3 Rukwa 1.0 2.0 4.0 4.0 0.5 0.5 4.2 3.1 lanza 6.0 5.0 15.0 19.0 5.6 5.6 8.3 10.2 Mara 2.0 3.0 11.0 11.0 6.1 6.0 4.8 6.8 Sh irryaiga 1.0 2.0 12.0 14.0 6.0 6.1 7.5 6.8 Kagera 13.0 2.0 7.0 4.0 2.8 2.9 3.4 6.5 Irlnga 4.0 7.0 6.0 6.0 5.2 5.1 5.7 2.7 Mteya 2.0 2.0 3.0 1.0 3.3 3.3 1.1 0.7 Rmuma 0.0 1.0 1.0 1.0 0.6 0.6 1.3 0.7 Total 156.0 223.0 295.0 287.0 208.3 208.8 221.6 193.2 Source: Goverrinnt of Tanzanla, Min. of Ag. & Llvestock 1987/88 data from "Arrual Review of Malze, Rice and Wheat" 1987. Other data from 'Price Policy Recoamindatlon: Maize, Rlce & Wheat," varIcus years 46 4' ooooooooooooooooooooooo C 0 - - 8 0 -_ C.8B" 0 ^".r¢gm."rMM"_XUS. . . . . . . . . O. 93 I -:2~~~~~~~~~~~~~~~- ow 0_N ___N_N_ ...... .. .. ....... ~ ~ 5 .,z. .2 -~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~F 94 oog! R-|Rn i S _4o0 _ "8M 88UMoSgX 0. 9gARg^gcg az $ S il°OiW$Xio|pffit:ffiE ' l 0~~~~~~~ >>t,. g8igSol U in. W 0rooWCotflWtQCOCOU40sos>N _1 c P:zi i~~~~~~~~~~~~~0 ) 00000 O z°anS8np~-ing ~~~~~~~~~ -~~~~~~~~~~~~~~~~~~~~~~ I~~~- - - ----- --- -- -- l CD Y S _ X F ~~~~~~~~~~~~~~~~~~~~~~~~t- C -8 . o . s . . o . i . . . . (0 D' ._ --C-- ---oco -- C---- - ( - '- C C - c- - - -- - - _2c <_ rr- m a . na 7c >cc nc /c44C4 4-4 o 4 (3 2.; 44 c384)4g _ 8 ) C a=-------------- C . . O CNr)f_lccB X c ~ ~ ~ ~ ~ ~ ~ ~ ~ -' 4 000( ~ )O'0 o. 0oo o o oc o o o I0 CrrCP-0.(0(0(00C0040)0 _) CC, ~~- 6(0)0)00(00(40)000.04(000.40. .a). . . .. . cc - - - - - C - v° c -c c-c?o<&e,t??ro ,?;g t 0000 040)0)0)0) (Cl~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~0 ! 0s ?o _" ? > >D X ~~~~~~~~~~~~~~~~~~~~~~~0)). - )00. a u f 044) 0 Zc 4- o.3~~~~~~~~~~~~~~~~~~C v r>>onoycBol rl>-rox7?"n g-r i Table 10 Public expenditure 10TAL EXPENDITURE BY SECTOR (In killor,s of Current Tanzania Shillings) : ~~~~~~~~~~~~~~~~~~TOTAL EXPENLTITLR5E BY SECTOR , TOTAL AGR/ LOUSA- HEALTH RURAL ROADS REkAINDER Total Agr/Lvstk Educ. Health Water Roads Remalnder LVSTK TION WATER 1974/5 1,121.7 69.1 275.0 184.1 42.1 77.7 482.9 1974/5 1O.09 6.2S 24.52 16.48 3.88 6.98 43.1X 1975/6 1,086.4 76.2 318.3 208.4 43.4 57.6 390.8 1975/6 100.0S 7.0S 29.39 19.2X 4.09 5.39 36.0% 1976/7 1,231.5 88.3 424.7 258.8 49.7 55.9 361.0 1976/7 100.09 7.2S 34.5X 21.08 4.09 4.59 29.39 1977/8 1,683.9 98.5 692.8 331.2 91.2 86.7 394.5 1977/8 103.09 5.88 41.19 19.78 5.49 5.18 23.49 1978/9 1,637.7 100.1 683.1 316.3 101.2 108.1 348.9 1S7°/9 100.09 6.1X 41.71 19.39 B.29 6.69 21.39 1979/80 1,959.5 121.3 850.1 355.7 108.7 113.7 430.4 1979/80 100.09 6.29 43.4% 18.23 5.5X 5.89 22.09 1980/1 2,294.7 130.6 932.1 397.7 128.0 122.2 605.5 1980/1 100.09 5.7S 40.69 17.39 6.89 5.3S 26.49 1981/2 2,608.5 148.0 1,090.3 455.5 139.2 135.2 664.4 1981/2 10O.09 5.7S 41.88 17.59 5.29 5.2S 25.5% NORTHEAST HIGHLANDS AS PERCENT OF TOTAL suTHERN H1IADS AS PERCENT OF TOTAL Total Agr/Lvstk Educ. Hsalth Water Roads Remalnder Total Agr/Lvstk Etbc. Health Water Roads ReamIander 1974/5 12.09 13.99 14.49 13.59 14.79 8.98 11.3S 1974/5 17.1% 22.19 17.18 16.39 14.39 16.09 17.39 1975/6 12.1X 15.18 11.58 12.78 14.39 8.78 13.0 1975/6 17.4X 21.58 17.19 16.7S 12.98 20.79 17.79 1976/7 12.11 13.0X 12.69 12.6S 13.7% 7.7S 12.39 1976/7 17.1X 19.69 16.9X 15.89 13.39 18.19 18.49 1977i8 12.0S 15.1X 12.2X 11.1S 16.4X 9.09 13.09 1977/8 17.1X 18.4X 17.6X 16.09 13.98 17.5X 17.8X t973/9 12.59 15.48 13.68 10.98 10.49 13.09 14.89 1978/9 18.69 21.39 18.6S 19.29 13.68 20.58 18.98 '979/80 12.59 16.09 12.59 11.98 10.39 12.89 15.59 1979/80 19.8% 17 69 19.89 18.99 14.68 20.09 23.09 198C/1 12.2S 20.39 11.48 12.49 12.59 12.59 13.79 1980/1 '.8 8Y 18 69 19 3X 16.8X 14.1X 20.4S 20.5S Is81/2 12.59 20.29 12.29 12.18 12.78 12.69 13.79 1981/2 19.39 17.98 20.19 18.39 14.99 19.89 20 18 COASTAL BELT AS PERCENTAGE OF TOTAL LAKE VICTORIA BASIN AS PERCENT OF TOTAL Total Agr/Lvstk Educ. He6ltn Water Roads Remainder Total Agr/Lvstk Eroc. Health Water Roads Remialnder 1974/5 25.69 14.69 21.19 27.59 25.49 29.78 27.99 V974/5 20.59 23.99 24.39 17.29 18.89 19.69 19.19 1975/6 25.29 19.98 23.0X 25.29 24.09 29.58 26.98 1975/6 20.89 24.99 26.19 17.89 20.59 19.49 17.09 1976/7 24.69 18.68 26.6S 25.99 25.69 29.59 21.49 1976/7 21.91 27.29 22.98 20.19 22.1X 18.19 21.09 1977/8 24.79 19.6X 23.09 28 7X 23.79 29.68 24.23 1977/8 22.23 25.98 25.19 19.0 21.48 19.68 18.9X 1978/9 25.49 19.39 24.2X 27.29 31.19 29.99 23.58 1978/9 23.&2 26.09 24.S9 21.9X 20.98 16.48 24.68 1978/80 24.79 20.49 23.6S 26.79 30.49 28.19 23.2X 1979/80 22.7X 21.69 23.6X 22.89 20.68 19.89 21.19 180/11 25.29 18.58 24.39 26.98 29.09 28.49 24.59 1980/1 23.7X 23.6X 25.09 32.29 20.19 19.19 23.69 1981/2 25.89 20.68 23.98 27.78 28.29 29.19 26.98 1981/2 2. 09 22.49 24.89 21.78 20.89 19.68 21.69 CENTRAL AND WESTERN PLATEAU AS PERCENT OF TOTAL Total AGr/Lvstk Educ. Health Water Roads Remalnder 1974/5 18.59 23.0X 18.98 18.19 28.89 15.29 17.29 SOURCE: An AnalysIs of &uogertary Allocationa by M. Schluter, 1982 1975/6 18.09 16.88 18.98 18.29 28.29 17.09 17.78 anrJ Estlmates of PlkAlc ExpendIture Supply Votes (Reglonal), 1984-1985. 1976/7 18.39 19.78 16.69 17.78 25.49 17.99 19.19 1977/8 18.69 19.19 18.49 17.32 24.69 14.59 18.99 1978/9 18.69 16.49 18.19 20.3X 22.49 19.49 16.49 1979/80 19.32 23.239 19.98 19.2% 22.89 18.49 15.59 1980/1 19.29 18.08 19.59 20.89 22.98 18.59 16.79 1981/2 19.32 18.98 19.09 20.3X 23.3X 18.98 17.78 Table 11 Enrollment in primary school by region, 1978, and percent of children ages 5-14 enrolled NREA Enrollment 1978 Children 5-14 Children 5-14 Total Percent Region orlbic Private Total Rural Only Urban Only Chlldren EnrolIerJ Males Feaales Total Males Females lotal Age 5-14 Dar-es-Salaam 99,055 6,034 105,089 8,551 8,311 16,862 83,834 90,894 1741728 191,536 55S NCRTI-EAST HlGHLANDS 349,088 0 349,088 263,119 255,622 518,741 15,175 15,821 3u, 96 549.737 645 Arusha 137,733 0 137,7^3 127,572 120,215 247,787 7,720 8,075 15,795 263,582 52a Klllmanjaro 211,355 0 211,355 135,547 135,407 270,954 7.455 7,746 15,201 286,155 749 COASTAL 8ELT E56,852 5,537 662,389 456,544 445,842 902,386 54,780 5, I61 2,941 1,015,327 65X Coast 986,894 813 97,707 63,397 60,588 124,575 4,301 4,614 8.315 133,490 73S irdl 13/ 86,805 203 87,008 60.'74 59,010 11 ,0t4 6, 09 6,466 12,74 131,556 6r: Mtwara 132,765 4,202 136,967 88,275 87,159 175,434 11,060 11,356 22,416 197,856 69M Tarna 184,629 319 184,948 134,349 131,366 255,415 17,793 18,985 36,778 302,193 618 Morogoro 155,759 0 155,759 110,159 107,719 217.878 15.618 16.740 32,358 258,236 62X CENTRAL AND WESTERN PLATEAU 487,575 22,443 510,018 393,336 376,906 770,302 39,932 39,E63 79,595 349,897 608 onoma 169,965 5,854 175,819 137,815 127,482 265,297 13,225 10,951 21,176 286,473 Si% SirTlda 97,561 6,241 103.802 79,339 77,565 156,3904 9,165 7,471 16, 636 173,540 0X3 Tabora 106,537 1J,348 116,885 96,447 92,726 189,173 12,053 12,775 24,843 214,016 SSX Klgma 113,512 0 113,512 79,785 79,133 158,926 8,474 8,466 16.940 175, 868 65X SOUT,ERN HIGHI.ANDS 562,392 31,690 594,082 401,168 400,627 801,795 33,559 37,055 70,614 872,409 683 MI,eya 202,365 11,382 213,387 142,294 140,993 283,287 11,763 13,185 24,928 308,215 695 Iringa 179,251 9,421 188,672 127,599 129,089 256,688 9,927 11,301 21,208 277,896 68s Ruvma 112,706 202 112,908 74,450 74,361 148,81i 5,143 5,492 0,635 159,446 71X Rkdwa 13/ 68,430 10,685 79,115 56.825 56,184 113, 09 6,746 7,097 13.843 126,852 62S LAKE VlCTORIA BASIN 757,022 14,885 771,907 617,249 614,238 1,231,487 31,535 33,754 65,289 1,296,776 60X Mwanza 242,732 10,136 252.868 188,349 186,153 374,506 15,400 16,336 31,745 406,247 62 Mara 159,247 0 159.247 103,872 104,019 207,891 6,371 7,102 13.473 221.364 72X Shinyarga 195,707 4,749 200,456 190,790 136,270 380,860 6,553 6,921 3.474 394,334 519 Kagera 159,336 0 159,336 134,238 133,996 268,234 3,202 3,395 6,597 274,831 583 lotal 2,911,984 80,589 2.992,573 2,140,027 2,101,546 4,241,573 258,815 275,348 53J. 163 4,775,736 63s 49 SOURCE: lanzanla Central Bureau of Statlstics, 1979. 'Statlstlcal Abstract, 1973-1970.' Ministry of PlTr9r.; a d 0n)M. I c AffaIrs. "ar es Salaam. Annex 3: Senegal Table 1 Population density, land use, and per capita agricultural land use by region, 1985 (hectares per person) LAND (o00 Hectares ) Irncludes Woodlareis REGION POPLLATIDN ('Ff14) 1otal 1PUpUlItior0 . Unoer I 9nused tut 1, unoer X Availaole X 3 Ffh WITA AtR9TIU.14L U&Ai, Total 1/ As S of Rural X Total 2/ Area 3/ De-sIty Cult. of Potent. of IWoodis and of Ciltivable of Ojltlvable 1985 Cjltlvable 1985 Total 1985 Aural 2000 j 1985 4/ total jClt. 5/ total Forest totall Land total Total riural Total Pop. per/sq.km: Poo. Poo. 2D0: DAa(ASi 1,470 233 221 15X 2,290 55 2,673 3 53 6 6 1131 6 111 15 271: 0.01 0.07 C.01 GIUJ T BASIN 3,153 493 2,856 91X 4,912 6,409 49 1,987 3131 1,200 191: 2,155 3431 5,342 8331 1.69 1.87 1.09 THIES 860 133 774 98% 1,340 660, 130, 361 553, 9 131 98 15t, 468 71X, 0.54 i.60, 035 DIOJBEL 504 83 444 88% 785 436 116 311 71X1 0 031 39 9Sl 350 80X1 o2-9 0.79 O 45 KAOLAD( at FATICK 1,289 203 1,173 SU1 2,008 2,394 | 54j 910 3831 240 101 716 3031 1,868 7831 1 45 1.59 0.93 j LDJGA 500 83 465 933 779 2,9191 17 405 1431 951 3331 1,302 453: 2,658 913S 5.32 5.72 ! 3.41 DUTLYING REGINS 1,855 293 1,535 833 2,890 13,208 1 14 1 622 5X 692 5ss 3,825 2931 5,139 3931 2.77 3.35 1.78 | ZIGUINCHOR at KJOLDA 880 143 730 83S 1,371 2.835, 31! 296 1031 454 16Si 685 2431 1,435 5131 1.63 1.96 1 05 SAINT LDUIS 610 96 531 873 950 4,413 1 14 1 110 231 104 2S1 1,572 3631 1,786 4031 2.93 3.37 1 3987 TA)BAt0Ufl 365 63 274 753 569 5,96001 61 218 431 134 231 1,568 2631 1.918 3231 5.25 7.01, 3.37 TOTAL SENEGAL 6,478 1D30 4,340 673 10,093 19,672 26 2,612 1231 1,898 1031 5,986 3031 10.496 5331 1.62 2.42 1.04 S603SS: 1/ From 'VII Plan de Develrppe&t Ecrmiuje et Social: 1985/1989. Sltuatlio oe L'Ecwrie Seregalatse, Strategle Je Developuement. Minlstere de Plan et oe Ia Cooperation. p.19, Table 4. 2/ 1985 Populatlon projected at rate of 3.03 for all reglons. 3/ From VIl Plan. See Note 1/ above. V From Situaticn Ecrmlo,je du Senegal 19832, Dlrectlon Statistiie, et Rapoort Aruel Direction Eaux, Forets et Chasses, 1978. Land Under Ojltivatlon Defined as "Terres AgrIcoles: siperficles cultives.' 5/ Potentlally Oultlvable doflned as Terres lrutilisees et susceptibles d'utilization agricole ou forestlere.' See Note 4/ for srurce. Table 2 Average annual rainfall by region, 1960-1983 (in mm) YEAR SENEGAL DAKAR ZlIGiJl I NDO/ DlIOiJiBEL ST. LOUIS LOiJA TAiMBACOUNJIA KAOLACK/ THIES KOLDA FAT ICK 1960/61 643 582 1,079 739 379 523 602 601 640 1961/62 664 586 1,254 566 371 448 789 664 635 1962/63 694 577 1,319 621 264 346 862 592 969 1963/64 665 547 1,219 579 382 451 943 644 556 1964/65 757 531 1,310 726 369 495 1,024 876 727 1965/66 680 400 1,458 563 438 449 939 655 544 1966/67 738 515 1,251 604 416 371 1,235 981 530 1967/68 880 918 1,560 858 342 667 964 907 828 1968/69 432 208 830 340 276 237 792 441 330 1969/70 660 687 1,198 571 426 372 745 654 624 1970/71 513 196 1,136 386 243 285 690 482 684 ! 1971/72 607 410 983 564 283 296 1,225 771 327 1972/73 349 120 702 410 118 205 622 415 202 1973/74 565 964 1,118 307 197 272 723 464 476 1974/75 583 367 1,110 538 229 341 957 564 555 1975/76 645 675 1,322 453 302 267 783 694 668 1976/77 573 392 1,282 443 260 284 970 540 415 1977/78 415 152 813 302 159 250 932 415 290 1978/79 601 269 1,258 571 281 331 575 941 580 1979/80 482 260 968 478 227 247 691 571 412 1980/81 436 378 760 349 237 327 609 436 394 1981/82 563 339 1,108 437 263 356 878 599 528 ! 1982/83 492 311 1,072 388 198 324 736 584 321, 1983/84 313 119 723 197 157 182 515 355 255 irEAN 581 438 1118 500 284 347 825 585 520 STO. DEV. 132 229 229 155 89 112 190 178 188 C.V. () 23 52 21 31 32 32 23 30 36 Rate of Growth -4.09 t -1.5 * -3.2 -3.1 -2.7 -1.2 * -1.5 * -3.1 .~~~ ~ ~ ~ ~ ~~~~~ … …… … = c = = _ = = = =I== = = = - = = =====_====== Source: Ministere di Developpement Rural/DGPA * Slr Iflcant a 1X. S S ificant at 5X. 50 - - - - - - - - - - - - - - - - - - - - - - - - - - - -->o > - ,-- - - - - - - - - - - - F, 0o; S~~ §Z;R§- A g ¢8og!§- -g F EBg§Mg 2 - ------ - 9- - -- ..- 6 iLIC ... -- ------------------------------------------------------ 3- - - - - - - - - -- - - -- -- - -- -- -- - -- -- LJ 00000000~~~~~~~~~~~~ 000000;-I0000 oo6 ~ 0 0 tE E SW,S° 8 8= 69 s 2 8 Fe ,o2 ig 8k8-a ........i;Ci SS~~~~~~..... .... --o-oo-o-oooooow - - - - - - - - - - - -- - - - - - - - - - - - - - - - - -- - _ n o-flnc nrrON CN -U a r 9 o m A! r88E m~~~~~ Fl s°4 s Loa c9-o FF MP >|F g;s,-;-,, Eo 2j § A L3 9 A 99 5 9 9 e 3 K0 E _ =0o~------;-"o -t-<--">> - 8c4too0l2E=O<>0-g= C L .0 ~ ~ ~ ~ a- _ _o-oooo-ooooo 088 ¢E9St;RS8 - 9 4-c wE >4e o Nr¢UH U-MM0fX"2g ~ ~ C 0 - 4~~~~~~~~~~~~~~~~i-- -- - - - - - - - - - - - - - - - - - - - -E NO gMMSaar-M¢0r 988S zc>oS e _ o e MXMMoN¢8X"""S"R¢^rMMUSrSoNN N N dl X Ro8S0@nS9xoRr_3%83g20Gi0o- N a-0U 0g H _ _ -o oo o oo oo o oo oo o o o o oo o_ ___ ____ 8_ o: '- -a--'>- N-'-t -''o 880°w 0°°8s°f88e' o°o 0 88X8888oXXoON8o88sX°N 2 80gR>-'@-W¢>SS8e 4t X vawu;*>ovrocc_ --- g8ga -o0egPiG8s3*2i8S>Fr E X~~~ o O Wss-Sws0'ecoSw-s eD * -~J) 4 84N5 88 4 fl0g4-.<3 8 °~~~~~c antr---nsNsSg o oooooo oo ccccc oooooooocccccocccc Co cc CCC cc c - c-CCoccc CJcr 80 ----c 0 --- - - - _ - - - - W~~~~~~c ccsS_>00SA° ccS C o SSr-ooo U) 8E "cU ._ ° 't>>8¢w>g¢esi3----govoov g _ _~~~~~~~~- ----------------------------------- E ~ P8 ~,8-gs0m8>¢-iEsAo> &3- X ;A~,W6~ 8 X>Xt> _ 2 __ xa iESE trUtFurtstS0 _._ .... _ _ .... Q aio..- ~~~ ~ ~ ~ ~ .. -~~~~~~~~~~~ ~~~~~ 0)0)~~~~~~~~~- --------------------------------- - - - - - - - - - s s ~~- ts "Z' ZS ~3 8 s c -2 cc- N- -~~~~~~~~~~~~~-- -- - - - - -- - - - - -- - - - - - --- -- - -- -- --- s~ ~ CC .)cNC cc cc-Moooooacor cgEz o-wa ciojSiGcc oocooooo 0o-ooo3 .c E , Cnj0r>00uX>>o0cscrJ>o,lo0g0o0e oo FarF>0rs0ggSrcctrrciaDicacEgno s_. . . .. . . . . . . . . . . C cc - ~~- -O-- -N- - - -c- S ,~~~~~~~~~~ cc CoOOC*lcDOU)CtoC OC>.ONOC ccc N-COcO _v_r-C-O - te- vt._s .. l;wiw ~oS~-ow E - - mcc NcCC CJ - 8 UG_a cccc OOcccc 0siccSg - cc CV 0c _)c CC 2° c N-ccN t-N-E cccc>r~ccn; a i @i 8rc_- N ccc., -9 ccccccccccmcccccc*-c z2 A c ------------------- cco- 2 F3 5 N6 m s Ef c - ----C) c -- CU ,~6: -- - - - - - - - --- --3-----c-. --CV-. ccCa,cca...c.c.c.a.o-.r.-.-.-.c CU CC 5~~~~~~~CC)~~~~CS5COCtC-C5cc5CV-~~~~.. .. .....N5 C 9 "; ~~~2 ~ ~ N ccC0C C, cccr CU - 0 --- --- -~~CC5N a- o-~C,0-~- -- -- ------------------------------------------- . -.c.o. cc.-.-.-.cn.-I-. . . . I . no,.N WS --- ---------------------- K Fa°0*8G;880^SssSA(oSENSEESt~~~- _ M S,;59g . . . . .Y .s .<> .L .- .;( t t S £ - - - -o -t -C - -r - - - - - 3.aw Ew | S 8 ~--~ Y 2 2 g _ > Y y i = |< | ;9 R T2 m mm 8 c ' a S S i | < , _ q 23 Y Y | Y = z K = . 2 2 < S-r ° o z o > 8 v r 3 t 0 r r~~~~~~~~~~~~~~~~~~~~~~~~-------------- RJ v° o t'R) -- -- -- ------------ -Sr qSa,( ,u O 1 N oi > o Le xr wbr X (D-v7 . . t/ < ° ° 8 > v r ffi t: t4 ? n 'q l. c.e , g g ov ;e 0 X E X > E M I_ t m o o ou - g r- g r W? c,i~F .L - - - - - - - -0 - t2_C 0 ' ) D..... Dr0°00°005CXr>;-tt CD -- --------------------------------------------a----------e°> t c8 42 CO ~ ~~~~~~ ~~~~ ... .g .S .) IL 3U S 8s0 ?E° isSr wgbC O) 0c 0 _ _ _ _ 8 w @ o o o o O O O _ O - ~~~~~~~~~~~W~. 7; Rv Ei- .--0 g 0 qgg ° s -- ------------------------------ar -c c?8----E-------¢%& °Y < XgMo S Mov O~~~~~~~- - - -l - - -? - ->° - - - - U Q O>.Oa ozz3a,a 0 3EaO0 fiS aiO~vrW??8v8e? snar-t$U Q~~~~~ . . .? . r° .c . . . . . .w . . . g, . . . .3 . F38. 8 < ci3 S3°;8g;§oo°ss°;o8gooo~>is?$8ow °o< - 8 P 1 t ^>igg3gtgei"U7iaq0RUN"iFY>soA o0= 6 3 00c3gSo@3sCop^0r-c3rg-8sa3tor?a s go _~ ~ ~ ~~ ~~~P _______________________ __ ____.. _5f- --i ~bi ^iW ot -l -f -3 -R -3 -t -8 -$ -0 -i -° -° -) - - -; - -> ,C - -a -3 -g - - - - -_ -! - - - @ ._ W _________~~~M ~n - - - - ._ --- --------------------------------------------------- -r- Annex 4: Kenya Table 1 Per capita arable land, 1985 and 2000 Province : - Populatlon , Total Arable Popuilatlon Per Capita Arabie Land Total As X of Rural Total Area as X of Density 1985 1985 2000 1985 1/ Total 1985 2/ 2000 1/ ('000 ha) Total: 1985 Total Rural Total Nalrobl 1,092 5X na 1,886 68 1,596 Central 3,094 15X 2,924 5,346 1,317 60X, 235 0.25 0.27 0.15 Coast 1,771 9X 1,234 3,060 8,304 41X' 21 1.94 2.78 1.12 Eastern 3,587 18X 3,279 6,198 15,576 25X' 23 1.08 1.18 0.62 No. Eastern 493* 2X 414 852 12,690 4 Nyanza 3,487 17X 3,213 6,025 1,253 80X' 278 1 0.29 0.31 0.17 Rift Valley 4,273 * 21X 3,827 7,384 16,388 31X' 26 1.17 1.31 0.68 Western 2,417 * 12X 2,278 4,176 820 72X, 295 0.24 0.26 0.14 TOTAL 20,200X) 1(X) 17,153 34,927 58,416 26%-, 36 0.73 0.86 0.42 Source: Population Statistics: Republic of Kenya, Central Bureau of Statistics, Vol. Ii, Analytical Report, p. 1, Table 1.1 Agricultural Land Statistics: Farm Management Hancbook of Kenya Vol. II, as reported in ISNAR. Notes: 1/ Assumes a 4.0X Popilation Growth Rate. 2/ CalIcted using 1979 Census figures for Urban Centers wIth popuIlatlIn above 2,000 (Table 1.2, p. 5) *MInor computational errors In the lIne. Original (incorrect) totals are used. Errors In total amcunts i.e to rounding. 54 Table 2 Land classification by district - Land Ouatlty Huid Se.i4nid Transitlonal & Sub-mid Transitional Seml-arld PROVINCE DISTRICT AREA Hlgh % of Mdldim % of Low % of ARABLE ARABLE (Sq. Km.) Potential Total Potential Total Potential Total LAND AS % ('00 ha.) OF TOTAL NAIROBI 684 CENTRAL Klambu 2,448 778 54.7% 33.1% 174 12.2% 1,422 58.1% Klrinyaga 1,437 285 29.8% 69.6% 5 0.5% 955 66.5% &iranga 2,476 961 53.2% 847 46.8% 1,808 73.0% Nyandarua 3,528 763 36.6% 1225 58.8% 97 4.7% 2,085 59.1% Nyerl 3,284 695 43.7% 685 43.1% 209 13.2% 1,589 48.4% SUB-TOTAL 13,173 3482 44.3% 3892 49.5% 485 6.2% 7,859 59.7% COAST Kilifl 12,414 2541 35.7% 4572 64.3% 7,113 57.3% Kwaie 8,257 235 3.2% 1850 25.3% 5228 71.5% 7,313 88.6% Lamu 6,506 3887 70.5% 1630 29.5% 5,517 84.8% wfb4asa 210 Talta/Tavbta 16,959 40 0.7% 663 11.3% 5139 88.0% 5,842 34.4% Tana Rlver 38,694 418 4.9% 8132 95.1% 8,550 22.1% SUB-TOTAL 83,040 275 0.8% 9359 27.3% 24701 71.9% 34,335 41.3% EASTERN Embu 2,714 161 8.0% 639 31.7% 1213 60.3% 2,013 74.2% Islol 25,605 Kltul 29,388 2902 14.5% 17162 85.5% 20,064 68.3% Machakos 14,178 131 1.2% 3526 31.3% 7616 67.6% 11,273 79.5% Marsabit 73.952 Maru 9,922 743 14.0% 2127 40.0% 2447 46.0% 5,317 53.6% SUB-TOTAL 155,759 1035 2.7% 9194 23.8% 28438 73.5% 38,667 24.8% NORTH EASTERN Garissa 43,931 Mandera 26,470 Wajir 56,501 SLBTOTAL 126,902 NYANZA Kisil 2,196 1914 99.4% 11 0.6% 1,925 87.7% Kisumu 2,093 605 37.9% 992 62.1% 1,597 76.3% Slaya 2,522 985 47.8% 1054 51.2% 20 1.0% 2,059 81.6% South Nyanza 5.714 2033 45.2% 2091 46.5% 375 8.3% 4,499 78.7% SIB-TOTAL 12,525 5537 54.9% 4148 41.2% 395 3.9% 10,080 80.5% RIFT VALLEY BarIngo 9,885 207 2.9% 1769 24.6% 5209 72.5% 7,185 72.7% Elgeyo Marakwet 2,279 603 41.5% 501 34.5% 350 24.1% 1,454 63.8% KajIado 19,605 3 0.1% 308 9.2% 3019 90.7% 3,330 17.0% Kericho 3,931 2553 75.6% 801 23.7% 21 0.6% 3,375 85.9% Laikipia 9,718 75 0.9% 1255 15.5% 6757 83.6% 8,087 83.2% Nakuru 5,769 1138 30.3% 1540 41.1% 1073 28.6% 3,751 65.0% Nandi 2,745 1136 59.0% 790 41.0% 1,926 70.2% Narok 16.115 2179 18.4% 3256 27.4% 6438 54.2% 11,873 73.7% Samburu 17,521 Trans Nzoia 2,078 344 22.1% 1206 77.4% 9 0.6% 1,559 75.0% Turkana 61,768 Uasin Gisl- 3,378 328 11.8% 2453 83.2% 2,781 82.3% West Pokot 9.090 522 10.8% 846 17.4% 3487 71.8% 4,855 53.4% SU8-TOTAL 163,883 9115 18.2% 14725 29.3% 26363 52.5% 50,203 30.6% WESTERN Bagona 3,077 1210 60.7% 782 39.3% 1,992 64.7% Bsusia 1,626 927 68.7% 422 31.3% 1,349 83.0% Kakamega 3,495 1918 75.3% 630 24.7% 2,548 72.9% SIB-TOTAL 8,196 4055 68.9% 1634 31.1% 5,889 71.9% TOTAL 564,162 23500 16.0% 43152 29.3% 80382 54.7% 147,034 26.1% Source: Jaetzold and Schmidt, 1982, as reported in ISNAR, 1986. 55 Table 3 Population, area, and arable land by province and district PROVINICE DISTRICT AREA POPLLATION POPLLATION % CHANGE POPULATION DENSITY DENSITY ARABLE ARABLE PER CAPITA ARABLE LAND (Sq. Km.) 1969 1979 1969-79 2000 1/ 1969 1979 LAND AS % 1979 2000 ('00 ha.) OF TOTAL (ha./pers.) (ha./pers.) 1/ NAIR0BI 884 509,286 827,775 62.5% 1,836,307 745 1210 CENTRAL Klambu 2,448 475,576 686, 290 44.3% 1,53,896 194 280 1,422 58.1% 0.21 0.09 Kirlnyaga 1,437 218,988 291,431 34.3% 664,104 151 203 955 66.5% 0.33 0.14 kuranga 2,476 445,310 648,333 45.6% 1,477,401 180 262 1,808 73.0% 0.28 0.12 Nyandarua 3,528 178,928 233,302 31.9% 531,641 50 66 2,085 59.1% 0.89 0.39 Nyerl 3,284 360,845 486,477 34.8% 1,108,568 110 148 1,589 48.4% 0.33 0.14 SUB-TOTAL 13,173 1,675,647 2,345,833 40.0% 5,345,609 127 178 7,859 59.7% 0.34 0.15 COAST KIlifI 12,414 307,568 430,986 40.1% 982,117 25 35 7,113 57.3% 1.65 0.72 Kwale 8,257 205,602 288,363 40.3% 657,112 25 35 7,313 88 6% 2 54 1.11 Lamu 6,506 22,401 42,299 88.8% 96,390 3 7 5,517 84.8% 13 04 5.72 Mombasa 210 247,073 341,148 38.1% 777,397 1177 1625 0.OO Talta/Taveta 16,959 110,742 147,597 33.3% 336,339 7 9 5,842 34.4% 3.96 1.74 Tarna Rlver 38,694 50,696 92,401 82.3% 210,560 1 2 8,550 22.1% 9.25 4.06 SLB-TOTAL 83,040 944,082 1,342,794 42.2% 3,059,916 11 16 34,335 41.3% 2.56 1.12 EASTERN Embu 2,714 178,912 263,173 47.1% 599,710 66 97 2,013 74.2% 0.76 0.34 Islol 25,605 30,135 43,478 44.3% 99,076 1 2 0.00 Kitul 29,388 342,953 464,283 35.4% 1,057,993 12 16 20,064 68.3% 4.32 1.90 Machakos 14,178 707,214 1,022,522 44.6% 2,330 090 50 72 11,273 79.5% 1.10 0.48 Marsabit 73,952 51,581 96,216 86.5% 219,254 1 1 0.0o Meru 9,922 596,509 830,179 39.2% 1,891,785 60 84 5,317 53.6% 0.64 0.28 SUB-TOTAL 155,759 1,907,301 2,719,851 42.6% 6,197,910 12 17 38,667 24.8% 1.42 0.62 NORTH EASTE Garissa 43,931 64,521 128,867 99.7% 293,658 1 3 Ma!ndera 26,470 95,006 105,609 11.2% 240,658 4 4 WlaJIr 56,501 86,230 139,319 61.6% 317,476 2 2 SUB-TOTAL 126,902 245,757 373,787 52.1% 851,774 2 3 NYANZA Kisli 2,196 675,041 869,512 28.8% 1,981,416 307 396 1,925 87.7% 0.22 0.10 Klsumu 2,093 400 .643 482,327 20.4% 1,099,111 191 230 1,597 76.3% 0.33 0.15 Slaya 2,522 383,186 474,515 23.8% 1,081,312 152 186 2,059 81.6% 0.43 0.19 South Nyanza 5,714 663,173 817,601 23.3% 1,863,123 116 143 4,499 78.7% 0.55 0.24 StE-TOTAL 12,525 2,122,045 2,643,958 24.6% 6,024,963 169 211 10,080 80.5% 0.38 0.17 RIFT VALLEY BarIngo 9,885 151,741 203,793 26.0% 464,397 16 21 7,185 72.7% 3.53 1.55 Elgeyo Narak 2,279 159,265 148,868 -6.5% 339,236 70 65 1,454 63.8% 0.98 0.43 KaJlado 19,605 85,903 149,005 73.5% 339,548 4 8 3,330 17.0% 2.23 0.98 Ker cIo 3,931 479,135 633,348 32.2% 1,443,253 122 161 3,375 85.9% 0.53 0.23 Lalkipla 9,718 66,506 134.524 102.3% 306,549 7 14 8,087 83.2% 6.01 2.64 Nakuru 5,769 290,853 522,709 79.7% 1,191,133 50 91 3,751 65.0% 0.72 0.31 Nandl 2,745 209,OS8 299,319 43.2% 682,079 76 109 1, 926 70.2% 0.64 0.28 Narok 16,115 125,219 210,306 68.0% 479,239 8 13 11,873 73.7% 5.65 2.48 Sanburu 17,521 69,519 76,908 10.6% 175,255 4 4 0.00 Trans Nzoia 2,078 124,361 259,503 108.7% 591,347 60 125 1,559 75.0% 0.60 0.26 Turkana 61,768 165,225 142,702 -13.6% 325,185 3 2 0.00 Uasin Gishu 3,378 191,036 300,766 57.4% 685,376 57 89 2,781 82.3% 0.92 0.41 West Pokot 9,090 82,458 158,652 92.4% 361,531 9 17 4,855 53.4% 3.06 1.34 SIX-TOTAL 163,883 2,210,289 3,240,402 * 46.6% 7,384,125 13 20 50,203 .30.6% 1.55 0.68 WESTERN 6ungoa 3,077 345,226 503,935 46.0% 1,148,351 112 164 1,992 64.7% : 0.40 0.17 Busia 1,626 200,486 297,841 48.6% 678,711 123 183 1,349 83.0% 0.45 0.20 Kakaoega 3,495 782,586 1,030,887 31.7% 2,349,152 224 295 2,548 72.9% 0.25 0.11 SU6-TOTAL 8,196 1,328,298 1,832,663 * 38.0% 4,176,214 162 224 5,889 71.9% 0.32 0.14 TOTAL 564,162 10,942,705 15,327,064 * 40.1% 34,926,824 19 27 147,034 26.1% 0.96 0.42 SUIRCE: Populatlon Statistics: Repub IIc of Kenya, Central Bureau of Statistics, Vol. I I, AnalytIcal Report, p. 1, Table 1.1 Agricultural Land Statistics: Faro Management Handbook of Kenya Vol. II, as reported in ISNAR. 1/ AssLnes a 4.0% Population Growth Rate. *Mlnor computatlonal errors In the line. Original (incorrect) totals are used. 56 Table 4 Maize area., production and yields by province lUiIT 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 lEAN SHPE MOIATH AREA 00 OHecta es AREA Rlft Valley 169.9 120.8 151.4 164.9 207.1 200.9 268.2 303.3 261.5 252.9 321.4 345.4 350.1 350.0 360.0 256.3 27% 8.iX Western 141.2 134.3 132.7 139.8 128.8 137.3 108.3 163.4 130.2 148.3 186.4 199.0 209.4 187.0 202.3 156.6 17% 3.3% Nyanza 100.4 129.5 138.7 98.4 58.2 95.6 110.3 127.0 131.9 119.9 189.7 189.2 201.9 208.9 112.9 134.2 14% 5.4% Central 93.6 98.5 119.2 130.5 117.4 98.1 86.4 99.6 94.3 97.8 102.7 102.6 108.6 192.7 75.7 107.8 11% 40.4% Eastern 194.1 182.0 219.7 223.6 230.2 194.2 218.9 242.7 214.9 290.3 294.2 319.5 318.5 293.4 193.1 242.0 26% 4.3% Coast 40.2 42.9 24.5 22.8 22.0 52.9 60.9 66.0 41.9 28.8 41.7 63.1 40.0 60.2 55.3 44.2 5% 3.45 TOTAL 739.4 708.0 786.2 780.0 763.7 779.0 853.0 1,302.0 874.7 938.0 1,136.1 1,218.8 1,228.5 '.308.2 999.2 941.0 10% 4.6% PR1CT I ON 8330 tornes PROiT I ON0 Rlft Valley 336.7 305.5 328.5 376.-7 577.8 799.9 733.4 808,8 785.8 642.4 699.1 993.5 448.5 982.1 587.9 38% 1o.C0 Western 391.4 362.8 239.0 251.8 231.9 370.7 292.4 441.2 175.8 260.4 335.5 477.7 288.0 407.1 301.7 19% 7.0% Nyanza 93.1 291.4 249.6 177.1 106.6 136.5 267.6 328.5 263.3 250.9 389.9 437.7 355.1 290.5 242.5 16% 9.0% Central 126.3 205.8 253.5 179.9 221.4 227.5 233.5 233.3 206.8 171.0 188.4 282.5 72.7 204.3 187.1 12% 2.2% Eastern 124.7 262.1 237.3 290.8 248.6 96.1 139.2 196.4 273.2 253.0 136.4 280.0 257.9 176.1 198.1 13X 1.0% Coast 36.2 64.3 22.0 20.5 28.0 57.1 81.9 71.3 34.2 25.9 18.0 113.6 25.7 74.0 44.0 3% 2.8: TOTAL 1,108.4 1,491.1 1,329.9 1,296.8 1,414.3 1,687.8 1,748.0 2,079.5 1,739.1 1,603.6 1,767.3 2,585.0 1,447.9 2,134.0 1,562.2 103% 5.2% YIEL ES tones/ha YIELD Rift Valley 2.0 2.5 2.2 2.3 2.8 4.0 2.7 2.7 3.0 2.5 2.2 2.9 1.2 1.7 2.3 Western 2.8 2.7 1.8 1.8 1.8 2.7 2.7 2.7 1.4 1.8 1.8 2.4 1.5 1.5 2.0 Nyanza 0.9 2.3 1.8 1.8 1.8 1.4 2.4 2.6 2.0 2.1 2.1 2.3 1.7 1.2 1.8 Central 1.3 2.1 2.1 1.4 1.9 2.3 2.7 2.3 2.2 1.7 1.8 2.8 0.4 0.4 1.7 Eastern 0.6 1.4 1.1 1.3 1.1 8.5 0.6 0.8 1.3 0.9 0.5 0.9 0.9 0.9 0.8 Coast 0.9 1.5 0.9 0.8 1.3 1.1 1.3 1.1 0.8 0.9 0.4 1.8 0.4 0.4 o.9 TOTAL 1.5 2.1 1.1 1.7 1.9 2.2 2.0 2.1 2.0 1.7 1.6 2.1 1.1 1.1 1.6 SOICE: Ministry of Agricwlture Sproadsheets. Note: in 1983 no district agricultural reIorts were sbeiittad to the Ministry. Tha data given for 1985 Is provisional ard only for long rains. Table 5 NCPB purchases of maize by province, 1970/71-1986-87 (in '000 90 kg bags) Provirce 1970/71 1971/72 1972/73 1973/74 1974/75 1975/76 1976M 1977/78 1978/79 1979/80 1980/81 1981/82 1982/83 1983/84 1984/85 1985/86 1986/87 Rift Val Ie) 1,,926 2,412 3,226 2,386 2,609 3,703 3,000 1,562 1,714 1,063 2,997 5,214 4,882 4,164 6,781 6,165 Western 835 642 1,113 1,401 1,674 1,923 1,909 918 570 314 897 1,369 1,066 833 1,480 1,033 Nyanza 20 172 401 140 234 395 673 166 47 36 349 605 566 509 570 55 Eastern 13 177 390 23 93 7 487 35 93 55 4 465 370 0 169 19 Central 5 132 214 142 378 144 240 31 94 1 5 84 104 22 255 284 Coast o 0 0 0 0 0 0 0 0 0 0 0 0 0 1 3 Total 2,.799 3,536 5,345 4,092 4,988 6,171 6,337 2,713 2,519 1,46% 4,251 7,739 6,963 5,528 4,219 9,236 8,059 Total 252 318 481 368 449 555 570 244 227 132 383 696 627 498 380 831 725 (in o Metric tons) Swourca: 1970/71 - 1983/84 data from NCF. Statitics Diyision. 1985/86 - 1986/87 data from Coopers anrd Lybrarnd, NCPB Reorganisation Study, 1987. Table 6 Shares of INCPB purchases of maize by province, 1970/71-1986/87 Province 1970/71 1971/72 1972/73 1973/74 1974/75 1975/76 1976/771977/78 1978/79 1979/80 1980/81 1981/82 1982/83 1983/84 Rift Valley 68.8% 68.2% 60.4% 58.3% 52.3% 60.0% 47.3% 57.6% 68.1% 72.4% 70.5% 67.4% 69.8% 75.3% Western 29.8% 18.2% 20.8% 34.2% 33.6% 31.2% 30.1% 33.8% 22.6% 21.4% 21.1% 17.7% 15.3% 15.1% Nyanza 0.7% 4.9% 7.5% 3.4% 4.7% 6.4% 10.6% 6.1% 1.9% 2.5% 8.2% 7.8% 8.1% 9.2% Eastern 0.5% 5.0% 7.3% 0.6% 1.9% 0.1% 7.7% 1.3% 3.7% 3.7% 0.1% 6,0% 5.3% 0.0% Central 0.2% 3.7% 4.0% 3.5% 7.6% 2.3% 3.8% 1.1% 3.7% 0.1% 0.1% 1.1% 1.5% 0.4% Coast 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.4% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% 0.0% Saorce: 1970/71 - 1983/84 data froc NCPB Statltics Division. 1985/86 - 196/87 data from Coopers and Lybrand, NOB Reorganiza 57 Table 7 Coffee area, production, and yield, 1981/82 PROY INCE SUALLHOLDERS ESTATES TOTAL Area Productlon Yield Area Productlon Yield PERCENT OF (1 0Xtia) (100t) (kg/ha) (lOWha) (1000t) (kg/ha) OUTPUT EASTERN 45 18.9 420 2.2 1.8 818 23.8X CENTRAL 38.6 27.9 723 24.1 31 1286 67.8S RIFT VALLEY 0.4 0.1 250 7.3 1.6 219 2.OY NYANZA 8.6 4 465 - - 4.6X WESTERN 4.5 1.6 356 - - 1.81 COAST 0.4 0.1 250 - - 0.1S TOTAL 97.5 52.5 538 33.6 34.4 1024 1W0.OS SCURCE: Coffee Board of Kenya/J. de Graaff, 1988. Table 8 Tea area, production, and yield by province, 1973-82 ITEM LNIT 1973/74 1974/75 1975f76 1976M 1977/78 1978/79 1979/80 1980/81 1981/82 WAAN SHARE Rate of Growth CENTRAL Area ha 15066 16052 16942 18028 19579 20465 21036 21783 22283 19026 421 5.1% Prodaction (Made Tea) Kg. 7024201 8059358 9480532 14705004 16953929 18869771 15214224 16187859 18164785 13819518 50s 11.8% Yleld Kg./ha 466.2 502.1 560.5 815.7 850.6 922.1 723.2 743.1 815.2 711 (Klaabu, Mirang'a, Iyarl, Klrlr'yanga) EASTERN Area ha 5so9 5597 6308 7050 7439 7568 7754 7954 8112 6977 15% 5.89 Prod±atlan (Wade Tea) Kg. 2058834 2250306 2711529 3826925 4071862 5319557 4461672 3995793 4893553 3732203 14% 10.99 Yield Kg./ha 411.0 402.1 429.9 542.8 547.3 702.9 575.4 502.4 603.2 524 (Eabu, Weru) NYANZA Area ha 6944 7566 8185 9202 9844 10329 11257 11928 12405 9740 219 7.4% Prodatlon (Made Tea) Kg. 2826258 3065542 4012989 5234196 6863523 7375623 5540765 6309173 6508734 5304089 198 10.8% Yield Kg9/a 407.0 405.2 490.3 568.8 697.2 714.1 492.2 528.9 524.7 536 (Klsll, Sotik) RIFT VALLEY Area ha 6119 6625 6831 7698 8235 8743 9501 9994 9855 8178 18% 6.5% Prodetion (Made Tea) Kg. 2341638 2629797 3055189 4345238 5512701 6920267 4733610 5184526 5234213 4338581 16% 10.7% Yield Kg./ha 382.7 397.0 447.3 564.5 669.4 688.6 498.2 518.8 530.1 522 (KeridW/hes8plr, Nandl/ Lessos, E/aralt/oratganl) WESTERN Area ha 1246 1365 1474 1650 1764 1849 1872 1972 2038 1692 49 6.19 Prodxticon (Made Tea) Kg. 267604 258000 304856 417069 4848s9 572935 604141 575887 556650 505325 29 11.79 Yield Kg./ha 214.8 189.0 206.8 252.8 274.9 309.9 322.7 291.9 273.1 260 (Kakameg) TOTAL Area ha 34384 37205 39740 43628 46861 48954 51420 53631 54693 45613 1809 6.3x Prodxtin (Made Tea) Kg. 14518585 16263003 19581145 28519436 33584674 38158153 30554412 32250647 35347935 27641999 180% 11.32 Yield Kg./ha 422.2 437.1 492.7 653.7 716.7 779.5 594.2 601.3 646.3 668 oAla81: Compl I ed from KTXT A jal Reports. 58 Table 9 Cotton production by province, 1974-85 (1) (bales) YEAR WESTERN NYANZA EASTERW COAST(3) TOTAL PROVINCE (2) PROVINCE (2) CENTRAL FROVINCE PROVINCE 74/75 12,784 3,73i 6,838 5,371 28,535 75/76 14,186 6,401 4,091 6,854 31,532 76/77 11,482 4,762 11,398 7,105 34,747 77/78 12,227 10,596 18,357 5,707 46,867 78g79 15,196 12,252 29,593 5,138 62,179 79/80 16,514 12,413 17,325 4,998 51,250 80/81 10,422 13,642 15,282 7,642 46,988 81/82 10,144 11,180 12,356 7,877 41,557 82/83 4,668 9,144 21,420 6,821 42,053 83/84 (4) 4,711 4,733 11,880 7,703 29,027 84/85 (5) 6,135 6,700 35,000 15,984 63,819 85/86 4,600 (6) 6,500 - 18,500 - YEAR WESTERN NYANZA EASTERN/ COAST(3) TOTAL NYANZA/iESTERN PROVINCE PROV I NCE CENTRAL PROV I NCE PROVINCES X OF TOTAL(2) X Of TOTAL(2) X OF TOTAL X OF TOTAL X OF TOTAL 74/75 44.80 13.10 23.26 18.82 100 57.90 75/76 44.99 20.30 12.97 21.74 100 65.29 76/T7 33.04 13.70 32.80 20.45 100 46.75 77/78 26.09 22.61 39.17 12.18 100 48.70 78/79 24.44 19.70 47.59 8.26 100 44.14 79/80 32.22 24.22 33.80 9.75 100 56.44 80/81 22.18 29.03 32.52 16.26 100 51.21 81/82 24.41 26.90 29.73 18.95 100 51.31 82/83 11.10 21.74 50.94 16.22 100 32.84 83/84 (4) 16.23 16.31 40.93 28.54 100 32.54 84/85 (5) 9.61 10.50 - 25.05 100 20.11 SOXJRCE: SUPERVISION REPORTS, JJNE 14,1984 AND 1986 NOTES: (1) BASED ON CLSMB ESTIMATES; A DEGREE OF OVERLAP OCCRS BETWEEN YEARS AND REGIONS. (2) SOfE RIFT VALLEY PRODUCTION INCLIDED IN WESTERN AND/OR NYANZE PARTICUIARLY IN LATER YEARS. (3) IRRIGATED AND RAINFED PROOUCTION COMBINED FOR YEARS BEFORE 1983/84. (4) AT LEAST 20000 BALES POTENTIAL LOST TO RIOUGIT. (5) ESTIMATED VALUES; RIFT VALLEY PROODCTION OF 35 INCLUED IN WESTERN PROVINCE. (6) AN ADDITIOML 1000 BALES IS EXPECTED FROM TDE RIFT VALLEY. (5) ESTIMATED VALUES; RIFT VALLEY PRODUCTION OF 35 INCLUDED IN WESTERN PROVINCE. (6) AN AODITIONAL 1000 BALES IS EXPECTED FROM THE RIFT VALLEY. Table 10 Table 11 Geographical distribution of Zebu cattle, 1978-82 ('000 head) Geographical distribution of grade cattle, 1977-82 ('000 head) PROVINCE 1978 1979 1990 1991 1992 kEAN WAD PROViNCE 1977 1978 1979 1980 19fi 19f2 EA4 HfEAD Rift Valley 3.739 3279 317 2,939 3.290 3.264 Rift Valley 544 557 511 954 972 911 932 (9 of Total) 44.15 44.01 Vex.91 36.21 38.21 40.2S (s of Total 50.6S 49.51 47.51 46.51 45.99 42.71 46.6Ti Eastern 1.452 1.224 1 50 1.30 1,349 1,397 Central 415 435 491 575 595 847 555 (E of TotaI) 17.21 16.51 19 3 16.79 15.99 19.89 (S of Total 39.71 398.6 39.19 40.91 40.6S 44.69 40.91 Nfyanza I1,19 234 1,493 1,494 1,92 1.199 Eastern 61 77 92 91 99 115 87 (a of Total) i3.19 4.51 1.718 18.79 i8.91 14.89 Is of Total 5.71 67.8 7.8S 5.8S 6S.9 6.11 69.4 North Eastern 920 929 910 Bw 930 939 lity-a 39 33 39 41 42 47 39 It of Total) 10.91 11.11 9'.9° 1903 9.°6 15.32 (I of Total 2.89 2.91 3.32 2.91 2.9S 2.5S 2.81 Coast 390 1.074 519 707 999 999 Western 11 11 13 40 41 59 29 (9sf Tota 3 4.35 14.55 6.2S 9.91 7.9 89.21 (1 of Total 1.01 1.91 1.i1 2.91 2.98 3.0S 2.11 Westemr 722 547 653 634 704 652 Coast "1 13 14 16 19 22 16 (1 of Total) 8.51 7.4S 7.9S 9.1X 8.2S 8.0S (S of Total 1.01 1.25 1.21 1.11 1.32 1.21 1.2 Contral 165 151 95 99 12 127 North, Eastern (C of Total) 2.01 2.0S 1.21 1.11 1.65 1.91 (1 of Total 0.9 5.91 O.OS O.OS 0.OS 0.OS 0.91 TOTAL s.460 7.426 8.218 7,846 8,616 8.113 TOTAL 1,072 1,126 1.180 1,407 1,465 1,898 1,358 Sourse: Anlsal Pro3utl3n Olvisin Arnual Rsrts. Reported in Kenya StatistIcal Abstract, vario years. 59 Table 12 Quantity and value of Inputs purchased and used by smaliholders, by type of output, by province, 1978 (in '000 ksh and '000 kg) INWUT COAST S of EASTERN X of CENTRAL X of RIFT VALLEY 0 of NYANZA : of WESTERN S of TOTAL Total Totai Total Total Total Total FERTILIZER amtl ;ty 132.7 0.1S 18,378.2 20.68 55,021.8 E1.6T 8,190.9 9.2S 4,380.1 4.93 3,220.1 3.63 89'323.8 ValLe 278.4 0.33 13,172.3 14.03 55,296.0 58.63 14,402.5 15.3S 6,919.1 7.33 4,280.2 4.5S 94,348.5 CIltivated Area ('ODO ha.) V 232.2 7.23 680.4 21.53 472.6 14.73 792.6 24.63 651.6 20.33 376.7 11.73 3,218.1 Kg. Fort. Per Ha. 0.57 26.62 116.42 10.33 6.72 8.55 27.77 SPPAYS Cuant Ity 0.1 0. 6 4,463.9 80.13 2,621.7 35.3S 89.3 1.2S 224.7 3.10S 32.9 0.4X 7,432.5 Value 12.4 0.63 5,600.8 22.33 15,135.3 60.23 629.2 2.53 3,595.2 14.33 15.9 0.63 25:123.8 nOTE I WUr (Seeds) i Oatity 0.0 9O.0 147.7 3.83 2,945.7 75.1S 341.4 8.73 7.2 0.23 479.5 12 2S 3,921.5 Value 0.0 0.03 717.8 2.96 21,882.7 89.1S 1,808.3 7.4X 28.6 0.13 119.7 0.53 24,537.1 FEEDO OmantIty 12.5 0.03 1,291.7 6.23 12,548.3 60.53 8,441.8 31.13 85.1 0.43 365.4 1.83 20,744.8 Value 4.6 0.63 1,503.1 4.03 19,233.9 50.73 16,612.3 43.83 58.5 0.23 521.9 1.43 37,934.3 144CHINERY CDHTRACT Value 391.6 132.3 .983.6 14,082.3 0.0 0.0 15,539.8 WAGES (Iecl. In kInd)I Qiantity 2,249.7 2.53 2,153.1 5.4X 23,165.2 58.3S 5,837.1 14.73 5,917.3 14.93 426.6 1.13 39,749.0 Valual 2,152.7 2.43 14,385.9 13.53 53,600.4 43.51 22,801.8 16.63 27,995.0 20.43 10,103.9 7.4S 137,039.7 Avg. age 1/ 0.96 8.68 2.57 3.91 4.73 23.7 3.4 SE6D: Integratad Rural Sirveys, 1976-1979. kinistry of Eccroelc Plarning and Development, Kerrya. Table 10.3, p. 108 Note: 1/ SIallholder area for 1978. lntagaterd Rural Surveys, 1378-1979. Table 14.1, p.142. V/ Calculated. Table 13 Wage labor, earnings, and per capita income by province' (currency unit = '000 Kenya pounds, CUrrent) GROM2 RATE PROVINCE 19S 1970 1971 872 1973 1974 1975 1976 IST7 1978 1979 1980 1981 1982 1983 1984 NKMI.NL REAL Kober E loyed 163,615 184,002 178,149 192,279 203,443 226,959 218,589 230,269 235,465 244,431 260,822 274,209 284,534 291,327 309,815 315,701 4.43 4.43 Earnings/Ircem 72504 73510 88180 97062 188159 123209 1 '426 168428 183161 203629 244134 287715 349609 382808 421134 464411 13.2S 1.2S Per CaPlta/Ealoyed 443 448 495 505 517 542 647 723 778 833 936 1049 1229 1313 1359 1471 8.83 -3.2S CEFNTRAL Huce Eplayed 93,80 98.738 112,991 116.269 122,263 133,235 123,992 133,5a8 143,687 137,612 145,801 149,555 152,557 153,451 155,808 1568 55 3.23 3.23 Earnirgl oe/Ir 13948 16859 20873 20989 2383 28282 31508 41811 47551, 50814 59109 71311 81573 83517 95046 114757 13.9s 1.9s Per CapitaiEployed 149 171 183 181 193 212 254 313 331 369 409 477 535 544 811 733 10.73 -1.33 NYANZA Nuder Eplayed 45,722 46,578 48.859 51,511 51,923 59,985 61,728 63,432 64,753 67,301 71,996 74,516 T7,019 80,443 90,453 93,702 4.73 4.73 Earnlrns/lIe 7773 W 10544 11702 12776 18252 19154 23788 3068 36399 42340 47915 46394 52139 65495 72135 15.13 3.13 Per Canlta/Ewloed 170 223 218 227 246 304 310 375 474 541 588 643 602 . 648 724 770 10.43 -1.63 .ESTERN Hr Employed 18,761 19,837 20,929 22.142 24,495 34,758 36,745 38,184 44,598 42,465 48,019 49,466 52,322 53,294 56,624 61,915 8.3% 8.3S EarnIngs/lnre 3,921 4,395 4832 5665 7341 8879 11354 14363 18113 18748 21848 28809 35736 36793 44218 50290 17.73 5.73 Per Caplta/Employed 209 222 231 259 300 278 309 397 406 441 475 582 683 690 781 812 9.4x -2.63 COAST f4er Employed 84,526 86,574 89,908 8S.925 89,363 100,522 101,813 105,855 113,833 122,878 132,040 1S,286 139,521 140,592 140,918 142,419 4.13 4.13 Earnings/lncoe 22,301 22,873 25,817 2S.E60 31,418 37,989 43,554 51,294 57,966 70,366 80,931 91,385 109,384 118,541 122,610 139,672 13.53 1.5 Per CapIta/Employed 264 284 285 235 352 378 428 485 5S9 574 812 856 784 843 870 981 9.4X -2.63 RIFT VALLEY N&.er Eeiloyed 178 949 194.312 191,694 195,985 214,646 208,178 209,847 2159,25 225,798 221,133 234,375 232,848 230,221 228 143 241.356 242,517 1.93 1.93 Earniras/lin 22.725 26,157 27,470 31,548 35,451 38,838 44,827 55,342 62.715 68,368 74,828 90,518 188,970 116,659 136,228 149,545 12.93 0.90 Per CapIta/EWIWysW 127 142 143 161 165 187 214 255 278 309 316 389 473 516 564 617 11.1% -0.93 EASTERN 8&er Employed 39,219 41,562 45,700 48,a49 -1,807 58,791 62,195 66,450 70,110 71,003 76,001 80,572 80,463 83,456 89,104 92,305 5.73 5.7S Earningsr/ir 6,313 7,134 7,318 12,501 1-,549 17,039 18,448 24,380 28,667 31,346 37,121 41,929 50,902 55,428 65,098 70,104 16.63 4.6S Per CapIta/EIloyed 161 172 10 256 231 280 297 367 409 441 488 520 633 664 731 759 10.9S -1.13 N TH-EASTEFR Huttr E Iywed 2,622 2,878 2,958 2,917 3.193 3,835 4,172 4,827 4,652 4,941 5,253 * 5,51 7,672 9,325 9,402 9,451 9.03 9.0X EarnirgVlsro SW 668 716 828 5S 1,177 2, 05 2,210 2,593 3,157 3,537 4,538 6,114 7,359 8.323 8,776 19.83 7.83 Per Caplta/Eepleys 225 232 242 283 27 307 491 458 557 639 673 825 797 789 8a5 929 10.83 -1.23 TOTAL ?&dar Ecloyed 627,214 644,481 681,188 719,777 761,375 826,2-- 819,086 857,538 902,896 911,561 972,307 1,005,753 1,024,309 1,038,031 1,093,278 1,114,855 3.9S 3.93 Earningsr/Iae 150,074 161,998 185,420 208,854 231,169 274,385 312,320 379,614 431,434 482,824 563,5O9 E64,121 788,692 653,044 958,222 1,069,689 13.86 1.83 Per Caplta/Ee1loyed 239 251 288 287 304 332 381 443 478 530 580 660 770 822 876 960 9.9S -2.13 C.P.I. (19M.108) 2/ 100.0 101.8 105.5 112.2 122.8 144.4 71.9 191.5 219.9 257.2 277.8 316.0 353.3 425.7 474.6 522.7 - 12.03 S1.8MES: 1/ Statlstical Abstract, Central ursaeu of Statistics, Rapjollc of Kaba. Years 1978, 1982, ard 1985. 3' 13 (INF) 95 Yearbock. NOTE: Earr;rgs or Wges aNer all cash payents, IrrcWOinrg osc satary, cost of ilyvrg alloarxm, 9rofit bowus, together with tYe Value of ratlons ard free bard, and an astlmate of tre amloyer's contriltulor toward housIng. Earnings as sron In thls 60 sectlon are lar than the estleate of factor i,ce goirg to empdovees 1.ecause they excluoe pansl, eovloyers cntrlbutlio to the Natlonal Searity Furd f oprivate Dr3ovldn r fuds and persowal emolu ts for the arred forces. Earnings In the rural non-agriculture sector are excluedo. Table 14 Expenditure on main services by province, 1970-1984 (in thousand Kenya pounds, current) 'GROTH RATES PROV8INC 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1960 1981 1982 1983 NDAINAL REAL CENTRAL 591 732 1,688 1,079 1,319 1,191 1,569 2,119 3,5:4 4,076 4,0,38 6,043 6,949 6,143 18.7X 6.2% WESTERN 184 283 435 260 399 568 438 524 712 862 922 1,562 1,416 2,996 17.4X 4.99 EASTERN 397 805 708 589 861 758 1,054 1,117 1,561 1,759 2,341 3,849 3,034 3,386 16.19 3.6X NYANZA 352 320 990 313 441 640 741 799 1,106 1,322 1,623 1,504 1.987 2,036 14.49 1.99 RIFT VALLEY 852 1,207 1,472 1,213 1,540 1,712 1,964 1,886 2,435 2,696 2,675 3,818 4,906 3,924 11.79 -0.8X NGRTH-iEASTERN 71 117 249 82 111 106 6,600 7,145 143 195 238 275 472 502 11.6X -0.9X COAST 379 399 537 326 589 833 702 543 508 730 925 1,326 1,087 1,254 9.49 -3.19 AVERAGE 404 552 868 552 751 901 1,867 2,019 1,417 1,663 1,823 2,625 2,836 2,892 14.29 1.79 TOTAL 2,826 3,883 6,079 3,862 5,260 5,608 13,068 14,133 10,129 11,640 12,7932 1,377 19,851 20,241 14.99 2.49 C.P.I. (1970=100) 100 104 110 121 142 169 188 216 253 273 311 348 419 467 - 12.59 SOiJRCE : Statistical Abstract, 1978, 1982, and 1985 Editians. Central Bureau of Statistics, Kenya. IFS, (IIF), 1985 Edition for C.P.I. Irdex. NOTE: All figures listed as 'provisloIal' except years 1970/1973. Table 15 Percentage distribution of households by distance to water source in dry season by province DISTANCE COAST EASTERN CENTRAL R I FT NYANZA WESTERN AVERAGE On HoldIng 28.4 27.3 67.5 62.1 41.3 65.5 50.7 O - 1 Km 12.8 37.7 20.7 15.1 26.8 22.9 23.8 1 - 2 KIm 29.7 15.2 10.3 9.6 19.9 9.1 14.2 2 - 4 KIl 16.2 11.9 1.5 7.9 10.3 1.7 7.5 4- 8 Km 8.3 6.9 0.0 4.3 1.7 0.8 3.1 Over 8 Km 4.6 1.0 0.0 1.0 0.0 0.0 0.7 Average Distance 2.7 1.8 0.9 2.1 1.4 1.0 1.7 SOJRCE: Integrated Rural Surveys, 1976-79: Basic Report. Table 16 Health services available by province, 1978-1984 PROVINCE 1978 1979 1980 1981 1982 1983 1984 96 i PM0V1ICE 1978 1979 1S80 m91 19S2 19843 1S04 kiAN P4AI FMI CENTRA4L HosPitals 26 28 17 17 17 17 17 20 Hospltals 45 47 46 45 45 43 43 45 HBolth Centers 2 2 8 8 8 7 8 8 Heoath Centers 31 41 36 38 36 45 41 39 flspeRnarles 113 112 61 62 62 71 85 81 Dispensarles 158 154 175 1830 180 207 193 178 ea= per 1lD0.1 I/ 479 56 Su8 720 585 534 5S8 571 1eds per 1DD.D00 1a8 185 185 179 174 160 158 174 MAS5T RIFT VALLEY K= Itals 23 23 23 24 25 25 25 24 6i,o tals so 52 51 52 51 50 58 51 isalth Canters 18 18 19 22 23 27 Z0 92 inalth Centers 8 65 72 86 88 79 82 77 Dlspoor les 129 137 133 133 133 133 '42 135 0i9Dsm rles 311 317 331 383 3E8 39? 408 353 Beds per 100.003 188 183 183 211 196 161 1278 160 8s per 100,00 1oD 147 147 147 138 132 141 145 EASTERN N6YAet Hospitals 27 27 27 27 28 31 31 28 4rspItal3 39 32 34 38 36 31 28 33 9oIth OCmters 20 29 2S 27 33 71 39 85 Health enters 30 37 R9 89 43 49 55 42 Dlspisarles 201 191 197 193 193 191 227 13 Dlspaneres 144 112 133 142 144 181 15 142 Beds per lo,OCo 127 128 128 138 128 118 125 127 Beds per 1lo,D.D 137 148 148 19 144 120 117 131 N13T11 EASTERN WESTEMN 96opitals 3 3 3 3 3 3 3 3 Haspitals 10 1S 15 15 IS 16 1S 6s Health Ceters 3 6 4 3 4 6 8 S Health Centers 31 35 37 39 39 34 34 38 Dlspenarles 16 18 18 18 17 21 21 18 Dlsperarles 31 47 34 39 38 47 48 41 Beds per lo,OOD 115 90 so 91 84 79 85 91 Beds par 100,00 127 140 140 138 135 125 122 131 TOTAL Hospitals 225 228 210 221 220 218 213 220 H8alth Ceiters 201 233 241 282 278 288 293 256 Dispensarles 113 1088 1087 1130 1135 1213 1273 1147 8 per 1w100,OO 16S 175 174 177 171 15S 158 16B S0IC: 8lnlstry of HeaIth. Health InforetIr SystWe. NOTE: Statistical errors from orInpI MON qatatlcns In Statistical Abstract, years 1979-1985 61 Table 17 Primary school enrollment by province, 1968-84 PROVINCE 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 RIFT VALLEY 174,597 183.233 202,992 228,797 250,975 279,119 454,185 495,699 530,646 556,948 571,667 NORTH-EASTERN 2,389 3,301 3,432 4,668 5,048 6,377 7,200 6,965 7,507 9,234 9,487 NYANZA 221,138 206,462 234,012 248,990 269,764 291,128 562,511 602,695 550,580 554,450 518,346 WESTERN 145,932 169,930 201,787 200,708 234,900 245,847 401,475 431,259 446,185 447,281 415,894 COAST 71,642 76,805 83,983 87,445 96,102 103,107 149,778 156,927 160,156 163,225 170,664 EASTERN 242,059 269,652 289,867 315,454 339,582 370,555 515,624 545,877 543,222 572,635 601,851 CENTRAL 296,863 311,970 349,378 371,913 407,762 443,509 531,675 558,133 571,583 584,734 616,064 NAIROBI 55,060 60,944 61,238 67,523 71,786 76,375 83,430 83,400 84,738 86,342 91,540 TOTAL 1,209,680 1,282,297 1,427,589 1,525,498 1,675,919 1,816,017 2,705,878 2,881,155 2,894,617 2,974,849 2,994,894 - GRWT RATES - PROVINCE 1979 1980 1981 1982* 1983' 1984' X SHARE NOMINAL POPLLATION ACTUAL 2/ RIFT VALLEY 706,262 781,847 826,481 859,425 931,468 959,224 18.82 11.8X 4.72 7.1% NORTH-EASTERN 10,590 12,171 12,109 14,097 15,456 16,284 0.3X 11.1X 5.2S 5.92 NYANZA 767,249 785,537 777,413 814,010 835,762 833,067 19.02 9.92 2.52 7.42 WESTERN 539,946 569,057 573,280 587,982 611,096 615,243 14.32 9.3X 3.82 5.52 COAST 210,328 230,221 242,432 254,888 273,174 281,867 5.92 9.2x 4.22 5.02 EASTERN 706,654 752,844 748,142 768,958 807,902 812,751 19.2X 8.12 4.3X 3.8S CENTRAL 663,015 696,968 699,039 715,238 741,258 750 373 19.52 6.02 4.0% 2.02 NAIRO6I 94,202 97,984 102,266 105,549 107,706 110,902 3.02 4.22 6.32 -2.02 TOTAL 3,698,246 3,926,629 3,981,162 4,120,145 4,323,822 4,380,232 100.02 8.8X 4.02 4.82 SOLIRCE: Ministry of Edicatlon. Reported In Kenya Statistical Abstract, varlous Issues. NOTES 1/ Calculated from 1969 and 1979 Population Census. 2/ Actual growth here defIned as rate of groyth In enrol ment above rate of gr-mth In population. 'ProvlslonaI Table 18 Secondary school enrollment by province, 1972-1984 (in thousands) PROVINCE 1972 1973 1974 1975 1976 1977 1978 1979 1980 CENTRAL 37.1 41.7 51.4 55.6 66.7 78.2 87.4 94.6 105.8 COAST 14.4 15.4 17.4 16.4 16.5 19.0 19.6 20.7 23.1 EASTERN 23.5 24.7 35.4 38.5 45.1 54.5 63.2 67.9 71.1 NAIROBI 25.3 25.6 21.1 22.9 28.0 29.5 31.4 30.0 30.4 NO. EASTERN 0.3 0.4 0.4 0.5 0.6 0.7 0.7 1.0 1.5 NYANZA 22.3 23.0 28.8 31.8 46.1 50.5 62.4 65.8 70.9 RIFT VALLEY 20.4 22.3 21.4 29.2 34.5 40.8 46.7 53.4 58.8 WESTERN 18.6 21.7 20.2 31.9 42.8 47.1 50.3 51.0 54.3 TOTAL 161.9 174.8 195.8 226.8 280.4 320.3 361.7 384.4 415.9 -GROWTH RATES- PROVINCE 1981 1982 1983 1984 X SHARE NCMINAL 1/ POPULATION ACTUAL 1/ CENTRAL 92.4 98.8 119.5 124.8 24.1X 9.81 4.00X 5.8X COAST 27.3 29.2 28.6 26.9 6.32 6.22 4.22X 2.02 EASTERN 66.0 70.1 84.4 89.6 16.82 10.81 4.26X 6.51 NAIROBI 32.2 34.4 38.5 31.7 8.72 3.61 6.251 -2.71 NO. EASTERN 1.3 1.4 1.5 1.3 0.32 14.02 5.21X 8.81 NYANZA 77.6 83.0 86.9 80.8 16.71 12.32 2.46X 9.82 RIFT VALLEY 60.1 64.3 68.6 80.8 13.7X 12.02 4.661 7.3X WESTERN 53.1 56.8 65.7 75.0 13.5X 11.11 3.801 7.3X TOTAL 409.9 438.4 493.7 510.9 100.02 9.92 4.011 5.9X SOLUCE: MInistry of Edicatlion. Reported In Kernya Statistical Abstract, varoius Issues. NOTE: V 'Noinal grmth rate of secondary school enrollment calculated from MinIstry flgures; 'Actual' growth rates Is derived from nomlnal' less rate of populatlon growth. 62 Annex 5: Malawi Table 1 Population density, land use, and per capita agricultural land by region, 1987 and 2000 R'GION , - POPULATION ('000) H LAND ('000 Flectares) ' :Populatlon PER CAPITA AGRICLLTUIAL LAND Agr. Dev. Dlv. Total 1/ As S of Rjral 2/ X RuraI Total 3/ Total 4/ Cultivated Diltlvable S Forest I Det ilty Total Rural Total District 1987 Total 1987 1987 2000 1985 1985 5/ 1965 6/ Cultivable Reserves 7/: 1987 1 Pop. Pop. PoP. !oers/sq.kmi 1987 1987 2000 NMI1HERN 907.0 119 840.0 93S 1,211.4 2,691 344 1,236 46X 180 34 1.36 1.47 1.02 KarogagADD 243.9 39 222.1 911 334.5 646 44 269 429 38 1.10 1.21 0.80 Chltlpa 96.8 19 92.4 95S 134.6 350 153 449 28: 1.58 1.66 1.14 Karorga 147.1 29 129.6 889 199.9 296 116 399 501 0.79 0.90 0.58 kzuzu AOO 663.1 89 618.0 93X 876,9 2,045 300 967 47X j 321 1.46 1.57 1.10 Whata Bay 136.0 29 130.1 96X 197.7 409 114 289 331 0.84 0.83 0.58 Rumphl 94.7 19 87.8 939 117.0 595 136 232 16: 1.44 1.55 1.16 Vzlaba 432.4 51 399.7 929 562.2 1,041 717 699 42 : 1.66 1.79 1.28 CEi RAL 3,116.2 399 2,683.6 869 5,007.3 3,559 1,110 2,250 639 245 : 88i 0.72 0.84 0.45 Kasrugu ADD 1,013.9 139 912.2 909 1,607.9 1,593 478 985 629 : 64: 0.97 1.08 0.61 Kasurgu 322.9 49 286.9 899 454.7 788 462 59X ' 41: 1.43 1.61 1.02 Ichlnjl 248.2 3S 226.0 919 370.5 336 200 60S : 74: 0.81 0.89 0.54 Ntchlsl 120.7 2S 108.7 909 204.8 166 126 769 | 73 | 1.04 1.15 0.61 DOwa 322.1 49 290.3 909 577.9 304 198 659 : 1069 0.61 0.68 0.34 LlI oneADD 1,756.9 229 1,475.4 841 2,870.7 1,321 500 832 639 i 133 , 0.47 0.56 0.29 i llM.'e 986.4 12S 780.7 79S 1,644.4 616 414 679 : 160 ' 0.42 0.53 0.25 Dedza 410.9 59 370.6 909 697.5 362 229 639 : 113 0.56 0.62 0.33 Ntdeu 359.6 51 326.6 919 528.8 342 189 559 j 105 j 0.53 0.58 0.36 Salima ADD 345.4 49 297.2 869 528.7 646 133 433 679 : 54 : 1.25 1.46 0.82 Nkthtakota 157.1 29 128.9 829 220.3 426 248 589 1 37 : 1.58 1.92 1.12 Sal lIa 8/, 188.3 29 167.4 899 308.4 220 185 849 j 868 0.98 1.10 0.60 S5UTHERN 3,959.5 50S 3,468.4 88S 5,411.8 * 3,175 755 1,823 57S 291 125 0.46 0.53 0.34 Llwon)de AD) 1,446.7 189 1,360.3 949 1,957.0 1,482 369 1,032 709 98 : 0.71 0.76 0.53 Vanqochl 495.9 69 475.6 969 593.7 627 404 641 ' 79 ' 0.81 0.85 0.68 iachlnga 514.6 61 488.9 959 671.1 596 488 829 : 86 1 0.95 1.00 0.73 Zomba 438.2 59 399.2 919 692.2 258 140 549 j 170 | 0.32 0.35 0.20 Blantyre ADO 1,989.7 259 1,633.6 829 2,859.6 1,024 289 450 44X , 194 , 0.23 0.28 0.16 ChIlradzulu 210.7 39 205.2 97S 346.4 77 31 419 : 275 1 0.15 0.15 0.09 Blantyre 587.9 71 266.4 459 802.0 201 81 409 ' 292 ' 0.14 0.30 0.10 keanza 121.3 2S 114.4 949 140.7 230 84 379 j 53 0.69 0.74 0.60 Thyolo 431.5 59 412.9 98X 632.1 172 47 279 1 252 0.11 0.11 0.07 MilanJe 638.3 81 618.7 979 938.4 345 206 60X 1 185 1 0.32 0.33 0.22 Hgabu A8O 521.1 79 486.1 939 694.2 670 96 341 519 : 78 1 0.65 0.70 0.49 Chlkwawa 319.8 49 300.4 949 381.5 476 233 499 j 67: 0.73 0.77 0.61 NsanJe 201.3 3X 185.4 929 312.7 194 109 569 j 104 1 0.54 0.59 0.35 Total 7,982.7 1009 6,990.2 889 11,830.5 9,425 2,208 5,309 569 716 85 0.67 0.76 0.46 Soarces: / Malawil Porulation anid HisirK Census 1987: Prel lminary Report. Natlonal Statistical Office: Zomba, 1988. 2/ Riural poculation proJected from 1977 Census data (by district) to 1985 at 2.49499 p.a.. Rate of growth derived from ialael Population i'ensus, 1977: Analytical Report. Vol. 11. p. 115, Table 9.1 which gives urban populatlon at 8.59 In 1977 and rcuihly 259 in 2000. See also M POP file. Data not yet available from the 1987 Population Census. 3/ Malawl Populatlon Census 1977: Vol. 11 (N.S.O./Zormba) p. 168, Table A.9.43. 4/ Laid data from Malawi Population Census 1977, Analytical Report, Vol. 1., Table 2.4 5/ Cultivated land calculated from 1984/85 M.O.A. Crop Estimates for total customary hectarage, plus area under tobacco estates in 1985, by ADD, from Deloitte Haskins and Sel 1, 1986, 'Proposed Extenslo and Trainilr Service for the Estate Sub-Sector,' reported In IBRO Malawl Land Policy Study, AprIl 1987, p.21 Table 3.2 6/ Cultivable land figires from 1965, Oepartnent of Agriculture estimates, PjbLlsIe in Compendlum of Agricultural Statistics, 1977. (NSO: Zomba), Table 1. (Corv. to Ha. at 2.47) Arable land estimates are gewailly more coaiervative than the fig-ires given above; the by Office of President has cited 199 In 'SAL IV: A Proposal...' for arable laid, and the World Bark has alternately cited 38 cuiltlvabe (1981 Development of the Agricultural Sector Report) and core recently 22S without forests, 629 with (Laid Policy Study 1987. p.7). Elsewhere Wkaridawirl ad Phirl, 'Laid Pollcy Study' (Jan. 1987) cite the figure of 37S arable as a natlonal average. We use the fig res above as they represent official goverrment data and are more dIssa~eggeated, to the district level, this despite that they may he overlnflated. 7/ Fully gazetted forest reserves. Office of the President and Cablnet, 'Statement of Development Pollcles 1987-1998." p.38 Table 5.1 8/ The use of Agriculturai Development Divlslonis (ADO's) as sut-4heads for districts Is useful because nuch of the avallabia data, Ie Natioial Sanple Survey 1980/81 data, Is only given by ADD. However, in soms cases, such as Sal Ina District, It appears that district boundrles are nDt strictly observed. * Crepatloial errors In the iir. Original (irocrrect) runbers are used. 63 N CD D~ D EL C, CD (P~~~~~~~~~~~~~~~~~~EtC a CD ~ ~~~~~~~~~~~~~~~~~~~~------------ C----------------------------------- -- -- - C'C 3~~~~~~~~-- ----- - 8 ~~~~~~~~PC 0 28 0C ~~~ ~~~~~ .~~~~~~C.COC C,)- -- - -- - - - -- -- --- - - --C- - - - -- - -- - - - -- - - - - --(--P 8 88 -- C~~~~~~~~~~~~~~ ~~~ C 51 o.~~~~~~~~-- ------------- -----t Ce O Cr CC- -. a -~ - - -- - -- - -- - -- - -- - -- - -- - -- - -- - -- - - - M .-. C- CD0 CD C~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~D -- 0 3~~~~~~~~~~~~~~~~------------ is Z8 5 ~ ~ ~ ~ ~ ~ ~ ~ 0 C) a r0 8E8 8 8 ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ 8~~~~~2 e a CC a a a~~~~~~h C,c~. 0~ ssss 8. 8 - -o8-5- -8- 0000 o 8 o 0 0 ooo ~~~~~~~~~~ ------ -- - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - --.- - - - - - - - - - - - - - - - :_23 o~ T _ 0000 i 00.0 Cr~j Table 4 ADMARC maize purchases and sales by region Marketinrg 1970r71 1971/72 1972/73 1973/74 1974/75 1975/76 1976/77 1977/78 1978/79 1979/80 1980/81 1981/82 1982/83 1983/84 1984/85 1985/88 1986/87 1987/88 Year (5/10/87) Pu;rchases Northarn 63.6% 62.1% 43.68 39.18 16.08 15.68 13.58 15.69 17.8% 18.48 22.7% 25.98 19.86 18.98 15.78 25.09 43.88 43.88 Central 6.88 18.9% 27.5% 52.81 66.8% 72.28 69.28 71.9X 63.3% 70.78 68.8X 69.58 58.18 62.58 51.28 55.6% 49.08 37.4% Sa&twern 29.68 19.08 28.88 8.1% 17.28 12.28 17.38 12.58 18.98 10.98 8.5Y 4.68 22.19 18.66 33. 1% 19.3% 7.28 18.8% Sales Northern 11.18 24.09 8.68 6.78 Central 20.28 25.7% 14.58 21.8% SoJutw3rn 68.78 50.28 76.9X 71.58 Swrces: 1970/71-79/80 data from C. Ranadce, Fleldtrip (6/86) mimeographAed sheetss. 1980/81 - 87/88 data from Dloitte, Harskins & Sel Is, ADIARC Organization and Managesmnt Review, 1987. Table 5 ADMARC maize purchases and sales by region ('000 mt) k8arketinr 197071 1971/72 1972/73 1973/74 1974/75 1975/76 1976M 1977/78 1978n9 1979/80 1980/81 1981/82 1982/83 1983/84 1984/851985/86 1986/87 1987/88 Year (5/10/87) Purchases Total 3.2 5.5 7.0 10.8 26.4 18.3 36.5 58.5 79.9 71.2 91.7 136.5 246.1 244.8 296.4 272.7 111.3 56.9 Northern 2.0 3.4 3.1 4.2 4.2 2.9 4.9 9.1 14.3 13.1 20.8 35.4 48.8 46.3 49.5 68.3 48.8 24.9 Ccntral 0.2 1.0 1.9 5.7 17.6 13.2 25.2 42.0 50.6 50.3 63.1 94.8 142.9 152.9 151.8 151.7 54.5 21.3 Southern 0.9 1.1 2.0 0.9 4.5 2.2 6.3 7.3 15.1 7.8 7.8 6.3 54.4 45.6 98.1 52.7 8.0 10.7 Sales Total 125.3 47.0 100.7 187.8 Northern 13.9 11.3 8.7 12.6 Central 25.3 12.1 14.6 40.9 Southprn 86.1 23.6 77.4 134.3 Souroes: 1970/71-79/80 data from C. Ranade, Fledtrip (6/86) mimeographed sheetss. 1980/81 - 87/83 data from Deloitte, Haskins & Seals, ADMARC Organization and Management Revle., 1987. Table 6 Groundnuts (area in '000 ha; production in '000 mt; and yields in kg/ha) FREG I ON A..o. 1984 1985 1986 1987 AREA PROD. YIELD AREA PRDD. YIELD AREA FSDO. YIELD AREA PROD. YIELD NORTHN 10.4 4.2 0.40 9.8 4.1 0.42 17.0 7.7 0.45 14.6 5.6 0.38 Karoga AM 1.1 0.4 0.39 1.2 8.5 0.44 1.6 0.7 0.44 2.2 0.9 0.40 - ADD 9.2 3.7 0.41 8.6 3.6 8.42 15.4 7.0 0.46 12.4 4.7 0.38 CENTRAL 98.5 48.5 0.49 136.8 75.8 6.55 163.8 68.8 0.42 131.5 58.8 0.45 Xasugj AaM 55.3 26.6 0.48 64.9 35.1 0.54 83.3 34.7 0.42 53.0 22.8 0.43 Ll Ia ADa 40.1 19.7 0.49 65.3 35.6 0.65 75.5 31.2 0.41 72.4 31.9 0.44 SalIa ADm 3.1 2.2 0.72 6.6 4.9 0.74 5.O 2.9 0.59 6.1 4.1 0.67 JlfTIERM 27.2 9.5 0.35 29.6 8.3 0.28 29.2 11.5 0.39 30.1 12.3 0.41 Llb AM 16.9 6.4 0.39 15.3 4.3 0.29 15.9 8.6 0.42 21.5 9.0 0.42 Blantyre AmD 9.8 2.9 0.30 13.7 3.6 0.26 12.4 4.7 0.39 7.7 2.9 0.37 NgaI AM O.5 0.2 0.43 0.6 0.4 0.56 0.9 0.2 0.27 0.9 0.4 0.49 Scurae: Ministry of Agi6clture Crco Estleate Spread,ets. Total 136.0 62.2 0.46 176.3 88.2 0.so 239.8 88.0 0.42 176.2 76.7 0.44 Table 7 Customary tobacco by region, 1984/85 to 1987188 (area in '000 ha; production in '000 mt; and yields in kg/ha) RE810N 1984 1985 1986 1987 A.D.D. AREA PR03. YIELD AREA PR3O. YIELD AREA ?RoD. YIELD AREA PROD. YIELD NORTERfN 0.9 0.3 0.32 0.4 0.2 0.37 0.5 0.3 0.52 0.0 0.3 6.49 o sto arY KaroZ AmD 0.0 0.0 0.48 0.0 0.0 0.50 0.0 0.1 1.47 0.0 0.1 1.58 a2 AMW 0.9 0.3 0.32 0.4 0.2 0.36 0.5 0.2 0.44 0.5 0.2 0.46 CENTRAL 41.7 18.1 0.43 34.0 14.6 0.43 30.9 12.8 0.41 22.5 8.2 0.35 CatocAry 6asLo6 amD 18.1 6.5 0.36 16.9 6.2 0.36 14.9 5.1 0.34 8.9 2.4 0.27 LI laie ADD 23.6 11.6 0.49 17.1 8.4 0.49 '0.8 7.6 0.48 13.6 5.7 0.42 Sal In AODD 0 0.0 0. 0.0 . 0.6 0.80 0.1 0.0 0.50 0.0 0.0 0.40 S6J11f18 4.3 2.0 0.47 3.4 1.3 0.37 1.7 0.6 0.36 1.7 3.6 0.47 Ostosary Liwa1e AOD 2.4 1.0 0.42 1.9 0.6 0.31 1.1 0.4 0.32 1.2 0.5 0.45 Blantyre AmD 1.9 1.0 0.53 1.5 0.7 0.46 0.6 0.3 0.45 0.5 0.3 0.52 NgabLa AM 0.0 0.0 0.88 0.8 8.0 8.00 0.0 8.0 0.00 0.0 8.0 8.80 SaLrce: Ministry of Agriculture Crop Estlates Spreu ca8ts. Total 46.9 20.4 0.43 37.9 16.0 0.42 33.2 13.7 0.41 24.3 9.3 0.38 D__t___ry I ,_-_11_I_ 165 s s~~~~~~~~~~~~~~~~~t - - - - - - - - - - - - - - - - - - - - - - - - >~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~ -i - - - - - - - - - - - M ~ ~ ~ ~ ~ ~ ~ ~ ~ ~~~-- ----------------------------- C 7 A 7 A O 2 _ 7 7; 7 t t @ ¢1 @ _ > - 7 _ . . . . . . . . . .I . . . . ..Es. o_ _ _ . _i _ --- W7 7 co70_@" ---- -_--______-------------______--_-_______ E Ga8 8 ,G°- - - - - - G-A87D707 - g C7 _ 9 B s; 60 8; o 97 G7s E g a7 F G07 G7 0 s > U7 ~C @ - < 0 7A A C; A 0 A 0 A_ _ _ A. _ _ _ . _ _ _ _ _ __ _ _ __ _ _ __ _-_-_ -_-_-_--_-_-_-__-_-_-__-_-_-_--_-_-_--_-_-.- @ * i7ffi7 7 (o>707R07t U7 ._ _ _ 020C82G._o,~~~~~~ ----,----------E--------------------- w 70ViiU8G3 GgQ _ g0 077~OO§7CU=> 67g0°°7_^ > g _ __ __ __ __ _ _ _ c 8V G 7 |a7;7 7 s >N(NoC3G 8E. 0 D )- 3 )g 1_ 3l0eG GD oz n 9 ° ° O R r 8 G0O 8 GN O G U, o Gi _ O-8 .- 0~~~K - - - - - -----------o 0 < couraosuvgS¢s . _ _ iD : ._ ooooooooo~~~~~~~~~~~~~~~~-o------ . . . X¢ I$ i | iii 11 @* l! I|@; twM3SiE1; gI- - -- - - mm m m 9 2 2 - - - - - - - - - - - - - - - - - - - - - 0 tt0C R F£t __ ffi = e ~~~~~~~~> C 0 ma- -4 .0 A 0R gw DiHHWS gH @ g° ° R @s-t >|ii;!a ti! i5RiC :° D>. 3 -gC t 2°8 tD B ~ ~o o ~ gce g o oR oz g q g oML *@ ,Re r ^7 g O _ __ _ _ tD . ____ __ _ _ __ _ __ _ __ __ _ _____ __D___CD _ a SS o88 sso --S------° - -- <--S----------w-- R-- -- - --- - --E _~~~~~~~~~~~~~~~~~~~~~~~~~ - _ _ _ _g 8tt@- I D _ w R § < 0 w _ _ _ = @ __ _ 2 _ _ _ _ _ _ __ X~~~~~~~~~~~~CL 0 cD ~~~~~~~~~~~~t - - - --( --- -- - - - - - - - - - - - --------------- E ---- i s > @ ., < 0 ~~~~~~~~~~~o 3 9 .0 Mi o oOo c 9srn ~~~~~~~~~~~o qg i 'RE R R Rg q f i0 O ~ , n X X _ < < S a)a - ---------- <0 20 gB g 8 zr o c - m z~~~~0 C 3-------------------------------------- ------ _ t o o i o _ _ H 2 H a o 9 < ^ o S o 8 = n - _ ~~~~~~~~~~~~~~~~~~~~~~~~~~~ ~~ - a S z ~ 9'F9 1 8 : I mn | | w 9> ,,,, ; ~~~~~~ ->, 8 8 |g | < | "9 9~~~ 89 'o9 ~~~9 Z9 Ei3 ji | a 9 8 8 X 8 > < CDB . 9 < 0 E > SO) X s >8 ~, 8 = ,< w > S 0 Co ° 9 8 X 0 o l < W iF 8~~~~~~~~~~~~~~~~--------------------- ----_ ---------- ---- z o~~~~~~~~~~~~~~~~P 9_9l- uXnX8XX8 93 > | - [ | DB | | tS ;2- 9^ ~ '''Cl c g* o g9 9|9 3 rs-c ) iS j e_ -e88 oa t+8 cow 8 °>Otn °. ..... . ..... 8. L 1 [ 8s~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~I:1 @~ ~ ~ ~~ ~~D2~ !, - " -& - i------ --------------------------------------------------~~~ ~~ -----I - - ------ -- --- ---------- - -- -- -- -- -- -------- - --- -- ---- ---------_ ¢) L _ |? ZS g2 ) -E 2 AD c 8 R 3!0; l cliB l r9 X OOO ,O p BS;OWjB O>RlEeR«e S8 U U) _ w O CMi_S'' 'Ft'N' sX ovu r>8" i 13 g =E c ° L , 8| X n O f ,Baw X i0 ig]] -° j BF ,fiJ 3.OgSi i i ° : OD w c >u a_ 5 _9 , _ i _ 0 QXg v @ 8 g <- o F >- >> 01 _ i8~~~~~~~~~~~F _'M . .Ja Mg .8i i8.28 .gt .u .-w . . .E_... Table 4 Capital expenditures in the agricultural sector by states I SECOND PLAN THIRD PLAN FLURTH PLAN STATES 1 1970-74 1975/76- 1979/80 1981-85 I (Actual) (Actual ) (Actual) M.N % of Total N/ha | M.N X of Total N/ha M.N % of Total N/ha NORTHERN 56 18.1 1.91 240 13.1 8.18 1472 18.1 50.15 Bauchl 3 15.8 0.61 22 7.4 4.46 251 19.6 50.91 Bormo 8 17.7 0.67 37 10.7 4.12 180 11.4 20.07 Kadina 8 8.3 1.53 66 18.3 12.62 307 19.2 58.70 Kano 31 26.7 9.42 74 16.3 22.49 407 18.5 123.71 Sokoto 8 18.6 1.15 1 41 11.1 5.92 327 22.3 47.19 MIDCLE BELT j 25 13.8 1.03 i 188 11.5 7.73 j 697 12.3 28.67 Beruje 4 12.9 1.17 i 60 16.6 17.49 228 19.1 66.47 Gwgola 1 5 18.5 .70 37 12.1 5.20 121 9.3 16.99 Kar 1 5 9.3 1.10 26 5.8 5.74 1 97 8.8 21.41 Nlger 6 18.8 1.18 31 12.9 6.11 109 10.9 21.50 P1ateau 5 13.5 1.2 1 34 12.0 8.17 142 12.9 34.13 S91THERN 73 14.2 5.10 368 10.1 25.70 1258 9.0 87.85 Anaawra 1 7 18.9 5.43 1 27 9.4 20.93 138 9.2 106.98 Bendel i 10 10.1 3.41 : 39 7.4 13.31 1 159 7.6 54.27 Cross Rlver 1 11 15.5 5.37 1 52 15.3 25.37 144 11.1 70.24 1lmo i 5 20.0 5.75 40 10.3 45.98 210 14.0 241.38 Lagos i 10 11.6 38.46 19 3.9 73.08 125 6.4 480.77 0gn 1 5 19.2 3.85 36 11.5 27.69 100 10.0 76.92 QIdJo : 6 19.4 3.97 1 62 15.1 41.06 169 12.1 111.92 OY° i 11 19.6 3.96 43 12.2 15.47 100 6.3 35.97 Rlvers i 8 9.4 6.02 50 9.5 37.59 113 7.1 84.96 SoLrces Nigeria, Seccdx, Third and Fouirth National Develo,Tet Plans. Table 5 Federal aliocations and independent revenues of the Table 6 states, 1981-85 Primary and secondary education, 1978 Federal Independent Total PRLIARY SCHOOLS SECONDARY SCHOOLS Allocations Revenues Revenues Noober Enrollcent of ToteS Number Enrollment S of Total --------_--… Percent of---------------- f Schools *000 Students Population of Schools '000 Students Pop.ltio. Nm. Total Nm. Total Nm. NORTHERN STATES 14.172 3,204 11.1 272 103 0.4 Bauchi 2,477 399 11.3 48 13 0.4 ROono 2,428 693 15.9 59 14 0.3 Northern States 7,628 92.0 663 8.0 8,291 Kduna 2.857 845 14.2 74 35 .6 Bauchi 1.302 93.9 84 6.1 1,386 Kano 3,032 843 10.0 33 20 0.2 Borno 1,5l 0 94.7 87 5.3 1,637 Sokoto 3,378 424 6.4 58 21 0.3 Kaduna 1,624 90.7 166 9.3 1,790 Kano 1,357 86.9 204 13.1 1,561 MIDDLE BELT STATES 9.205 28786 19.3 434 164 1.1 Sokoto 1 795 93.6 122 6.4 I 917 ~~~~Banue 2.786 866 24.6 183 45 1.3 Sokoto 1,195 93.6 122 6.4 1,917 Gongola 2,224 473 12.5 44 20 0.5 Ro-r. 1.414 308 23.7 105 61 2.3 IMiddle Belt states 6,121 93.2 444 6.8 6,565 Niger 1.133 320 18.5 27 11 0.6 Benue 1,247 93.9 81 6.1 1,328 Plateau 1.648 539 18.4 75 27 0.9 Gongola 1,408 94.2 87 5.8 1,495 X-ara 1,176 94.0 75 6.0 1,251 SOUTHERN STATES 14.092 6.759 17.9 2.200 1,332 3.5 Niger 1,076 93.7 72 6.3 1,148 Anambra 1,926 962 18.5 370 146 2.8 Plateau 1,214 90.4 129 9.6 1.343 Bendel 1.690 836 23.4 267 185 5.2 Cross River 1,693 851 16.9 210 105 2.1 Southern 8tates 13,287 81.7 2,982 18.3 1,269 Imo 1.946 1.025 19.3 350 251 4.7 AnamSbara 1,385 82.8 2872 2 .2 1,69 Lgos 712 465 20.6 125 154 6.8 A-mbara 1,385 82.8 287 17.2 1,672 Ogun 1.226 330 13.6 151 73 3.3 Bendel 2.075 89.7 238 10.3 2,313 Ondo 1,500 478 12.1 252 139 3.3 Cross River 1,423 91.2 137 8.8 1,560 Oyo 2. 475 1,282 17.0 378 204 2. 7 Imo 1,550 84.3 288 15.7 1,838 Rivers 924 510 20.5 97 75 3.0 Lagos 948 42.6 1,278 57.4 2,226 ogun 1,006 87.4 145 12.6 1,151 ALL NIGERIA 37,469 12,749 15.8 2,906 1,599 2.0 Ondo 1,239 86.2 198 13.8 1,437 OYo 1,738 89.8 197 10.2 1,935 source: Nigeria, Fourth National Development Plan. River. 1,923 90.0 214 10.0 2,137 So.rce: Nigeria, Fourth National Development Plan. 70 Table 7 Table 8 Hospital facilities, 1979/80 Rural and urban water supply, 1978 NUMBER OF POPULATION PERCENT OF PERCENT OF HOSPITAL BEDS PER BED RURAL POPULATION URBAN POPULATION SERVED SERVED NORTHERN STATES 11,174 2,577 Bauchi 1,111 3,173 NORHERN STATES 19 67 Borno 1,455 2,988 Borno 0 70 Kaduna 4,178 1,422 Kaduna 13 31 Kano 2,944 2,852 Sokoto 39 100 Sokoto 1,486 4,429 MIDDLE BELT STATES 9,683 1,493 MIDDLE BELT STATES 27 69 Benue 88 90 Benue 1,640 2,146 Gonue 2 31 Gongola 2,148 1,759 Gongola 2 31 Kwara 2,391 1,040 Plateau 0 83 Niger 1,381 1,254 Plateau 2,123 1,384 SOUTHERN STATES 25 79 SOUTHERN STATES 48,809 772 Anambra 84 37 Anambra 7,140 730 Cross River 8 85 Bendel 6,626 539 Imo 20 100 Cross River 5,429 929 Lagos 4 94 Imo 5,546 960 ogun 14 100 Lagos 5,244 432 Oyo N/A 79a/ Ogun 2,978 755Rivers Ondo 6,874 576 73 Oyo 6,265 1,206 AVERAGE OF ABOVE STATES 24 Rivers 2,707 921 ALL NIGERIA 69,666 1,161 Refers to percent of total population served (i.e., urban and rural). Source: Nigeria, Fourth National Development Plan. Source: Nigeria, Fourth National Development Plan. Table 9 Indicators of child under-nutrition Percent of Children below 2 Standard Deviations of Weight_to-Height Indicator Urban Rural Total North 25.4 20.9 24.3 Bauchi 21.9 23.9 22.3 Borno 25.9 15.4 24.3 Kaduna 20.6 15.0 19.1 Kano 32.1 22.2 29.0 Sonoto 25.6 23.8 25.1 Middle Belt 26.1 18.0 23.9 Benue 23.0 16.2 21.5 Gangola 28.2 28.1 28.2 Kwale 36.3 10.2 29.3 Niger 19.4 15.7 17.7 Plateau 21.6 20.6 21.3 South 12.6 18.7 14.4 Anambara 14.9 20.2 16.8 Bendel 11.9 23.5 14.4 Cross River 9.3 8.0 8.6 Imo 10.1 18.5 12.5 Lagos 3.4 19.4 6.2 Source of Basic Data: F.O.S., "The Health of Nigerians 1983/84: Health and ogun 13.4 13.2 13.3 Nutrition Status Survey (A module of the National Integrated Survey of ondo 12.7 15.3 13.8 Householders (NISH) April 1983- March 1984),' Lagos, September 1985. Oyo 5.5 34.1 11.5 Rivers 26.4 23.9 25.7 All Nigeria 20.9 19.1 20.4 71 Annex 7: MADIA Tables rable 1 Computation of per capita land availability using FAO and government data EAST AFRICA WEST AFRICA ITM YEAR KENYA MALAWI TANZANIA CAIROON NIGERIA SENEGAL Lk,C (In CC ha.) Tctal Lard Area Natlcial 1985 56,416 9,428 88,366 46,540 90,241 19,672 FAO Yearbook V 1984 56,925 9,408 88,604 46,944 91,077 19,200 A.ca thder Cultlvation IbtloaI 2,577 3,639 4,465 6,830 12,542 2,612 (as 8 of total) 58 396 58 15% 14% 138 FAD Yearbock 14/ 1S8 2,335 2,345 5,190 6,965 31,035 5,225 (as S of total) 4S 258 6S 151 348 27% FAO Atlas 15/ 1980 4,400 2,500 9,200 7,700 32,300 5,200 (as X of total) 88 27S 108 160 35S 27% 'Arable Land FA0 (UnadJLsteo) 16/ 1985 1,S50 2,320 4,130 5,910 28,500 5,220 (as 5 of total) 38 258 5% 138 31% 27S FAo (AdJusted) 17/ 1984 6,075 6,085 42,785 32,165 46,235 5,942 (as 8 of total) 118 658 48% ;9, 51% 312 FAO Atlas (Potentlally Clltlvable) 18/ 1980 6,708 4,190 36,608 31,508 47,93) 9,700 (as 8 of total) 128 44% 418 678 53% 515 Natlcnal Arable Estimate 1985 14,703 5,280 49,100 34,905 67,951 10,481 (as 8 of total) 285 561 561 75% 75% 53S (I1 'OOo) Initlal (Cais of 19EO's) 10,942 4,040 12,313 na 55,670 na Presat Total NatiTaal (Cuei of 1970s) 15,327 5,547 17,036 7,761 na 5,069 Natlonal (jrrent Estlaate) 1985 20,20 7,208 21,383 10,130 936 125 6,478 FAQ 4/ 1985 20,632 6,944 722499 9,873 95.198 6,444 i8R0 5/ 1985 20,000 7,000 22,000 10,000 100,000 7,000 N3tl Rual 16,593 6,276 18,389 6,469 67,288 4,340 (as 8 of total) 828 878 888 1648 708 67% FAQ 7/ 1985 16,242 5,440 18,574 6,036 634844 5,121 (as 8 of total) 708 788 838 1318 678 798 IBM /a 1985 16,0005 6,160 18,920 5,8 08 70,0060 4,480 (3s 5 of total) 808 88S 86S 5!8 708 648 Pr~cjcted Total 2000 KP teiTal 37,505 11,783 34,066 16,682 140,220 10,053 ItPM 10/ 366C00 11,000 37,000 17,000 163,000 10,000 Pro-jqctad ir,aal 2083 0,ttiaonal 26,103 8,837 25,073 8,34l1 T7,121 5,955 (as 8 of total) 708 758 74% 50% 55% 519% PER CAPITA LA. AVAILABILITY Total Land Per CapIta AvallabIllty 8atliali Data 1965 5.16 2.32 7.18 ERR1 1.62 E R 1985 2.79 1.31 4.13 4.59 0.94 3.04 2086 1.50 0.80 2.59 2.79 0.64 1.95 ArabIe Lara Per Capita AvaIlablilty Natlaljal Data 1935 1.34 1.31 3.99 ERR 1.22 ERR 1985 0.73 0.73 2.30 3.45 0.71 1.62 2000 0.39 0.45 1.44 2.09 0.48 1.04 Aerable Land Per CapFAta AvaIlabIlIlty Natlcral Data (Ftural Po8i.latIo1) 1985 0.69 0.84 2.67 5.40 1.01 2.41 2000 0.56 0.80 1.93 4.18 0.88 1.76 Arable Landl Per Capita AvaIlabIlIlty FAO Atlas (Iand)/i8W (P0O.) Data 1985 0.33 0.59 1.63 3.13 0.50 1.51 210)0 8,19 0.37 0.99 1.5 0.29 0.87 ArabIe Lard Par CapIta AvaIlablIty FAO Yearbock Definitlon 1985 0.09 0.33 0.18 0.60 0.30 0.81 Sairces: See Tables 2 and 3. 72 Table 2 Population projections, and urban/rural growth, 1985-2000 Kenya Malawi -- Population in Thousands -- -- Population in Thousands -- Year Total Urban Rural X Urban r*ar Total Urban Rural % Urban 1985 22,200 4,094 18,108 18.4% 1986 7,200 893 6,307 12.4% 1988 22,990 4,383 18,606 19.1% 1986 7,440 967 6,473 13.0% 1987 23,808 4,693 19,116 19.7% 1987 7,889 1,047 6,642 13.6% 1988 24,656 5,026 19,630 20.4% 1988 7,945 1,134 6,811 14.3= 1989 25,532 5,380 20,152 21.1% 1989 8,211 1,228 6,982 15.0% 1990 26,440 5,760 20,680 21.8% 1990 8,486 1,330 7,15S 15.7% 1991 27,381 6,187 21,214 22.6% 1991 8,768 1,441 7,327 16.4% 1992 28,356 6,603 21,752 23.3% 1992 9,061 1,560 7,501 17.2% 1993 29,364 7,069 22,294 24.1% 1993 9,363 1,690 7,674 18.0% 1994 30,408 7,689 22,839 24.9% 1994 9,678 1,830 7,846 18.9% 1995 31,490 8,104 23,386 25.7% 1996 9,999 1,982 8,017 19.8% 1996 32,611 8,677 23,934 26.6% 1996 10,333 2,146 8,186 20.8% 1997 33,771 9,290 24,481 27.6% 1997 10,678 2,326 8,363 21.8% 1998 34,972 9,946 26,026 28.4% 1998 11,034 2,518 8,516 22.8% 1999 36,217 10,849 25,68e 29.4% 1999 11,402 2,726 8,676 23.9% 2000 37,505 11,402 26,103 30.4% 2000 11,783 2,953 8,830 25.1% Calculate Growth Rates Calculate Growth Rates 1985-2000 1979-2000 1972-1986 1985-2000 1977-2000 1979-19S5 Pop. Growth Rate: 3.68% 3.583 4.3blS 8.37X Pop. Growth Rate: 3.34% 3.34% 3.33% 3 ?1% Urbanization Rate: 3.39% 3.39% Urbanization Rate: 4.8CX 4.80% Urban Growth Rate: '.07X Urban Growth Rate: 8.30% Rural Growth Rate: 2.47% Rural Growth Rate: 2.27X Sources: Growth Rate calculated from 1985 and 1987 Economic Sources:Growth Rate calculated from 1977 Census figures, Vol.I7 Survey figures given in Table 2, for 198C and 2000; Urbanization Rate calculated from 1977 Census figu,res of Urbanization Rate calculated from 1979 Census figure 8.6x in 1977 and 25% in 2000. and 2000 figure (30,4%) given in 1986 Economic Survey. Tanzania Cameroon Year -- Population in Thousands -- -- Population in Thousands -- Total Urban Rural % Urban Year Total Urban Rural % Urban 1985 21,383 2,994 18,389 14.0% 1956 10,130 3,659 e,671 36.1% 1986 22,067 3,223 18,844 14.6X 1986 10,467 3,761 8,896 38.0% 1987 22,773 3,470 19,303 16.2% 1987 10,796 3,975 6,820 36.8% 1988 23,502 3,736 19,786 15.9% 1988 11,144 4,201 8,943 37.7% 1989 24,264 4,023 20,232 16.6% 1989 11,504 4,440 7,064 38.6% 199 25.030 4,331 20,699 17.3% 1990 11,875 4,693 7,183 39.5% 1991 25,831 4,663 21,168 18.1% 1991 12,259 4,960 7,299 40.5% 1992 286,58 6,020 21,637 18.8% 1992 12,865 5,242 7,413 41.4% 1993 27,511 6,405 22,106 19.6% 1993 13,063 5,540 7,524 42.4% 1994 28,391 5,82G 22,572 20.5% 1994 13,486 5,856 7,631 43.4% 1995 29,300 8,286 23,034 21.4% 1996 13,921 8,188 7,733 44.5% 1996 30,237 6,746 23,491 22.3% 1996 14,371 6,540 7,831 45.6% 1997 31,206 7,263 23,942 23.3% 1997 14,836 5,912 7,923 48.8% 1998 32,204 7,820 24,383 24.3% 1998 16,314 7,305 8,009 47.7% 1999 33,234 8,420 24,814 25.3% 1999 15,809 7,720 8,088 48.8% 2000 34,298 9,065 25,232 26.4% 2000 18,319 8,160 8,160 50.0% Calculate Growth Rates C-l culate Growth Rates 1985-2000 1977-2000 1985-198S 1985-2000 1976-2000 1973-1984 1981-91 Pop. Crowth Rate: 3.20% 3.34% 3.33X 3.31% Pop. Growth Rate: 3.23X 3.23X 2.b2% Urbanization Rate: 4.33% 4.33% Urbanization Rat-: 2.38X 2.38% 2.37% 2.34% 1.39% Urban Growth Rate: 7.87% Urban Growth Rate: F.S9% Rural Growth Rate: 2.13% Rural Growth Rate: 1.45% Sources: Pop. Growth Rate from The Demography of Tanzania, p. 231. Sources: Population Growth Rate from Sixth Plan (1988-1991), p. S. Urbanization Rate calculated from WDR figures of 8% in 1985 Urbanization Rate calculated from 1985 figure (Sixth Plan; and 14X in 1985. p. 3) and World Bank estimates for 2000 (Country Economic Memorandum, 1987; p. 18). 73 Population projections, and urban/rural growth, 1985-2000 Nigeria Senegal -- Population in Thousands -- -- Population in Thousands -- Year Total Urban Rural % Urban Year Total Urban Rural % Urban 1986 96,125 28,838 67,288 30.0% 1985 6,478 2,164 4,314 33.4% 1989 98,576 30,383 68,192 30.8% 1988 6,672 2,261 4,412 33.9% 1987 101,088 32,011 69,077 31.7% 1987 6,873 2,362 4,510 34.4% 1988 103,686 33,726 69,938 32.5% 1988 7,079 2,469 4,610 34.9% 1989 106,307 35,534 70,773 33.4% 1989 7,291 2,579 4,712 36.4% 1990 109,017 37,438 71,679 34.3% 1990 7,510 2,695 4,815 36.9% 1991 111,796 39,444 72,352 35.3% 1991 7,735 2,818 4,919 36.4% 1992 114,646 41,568 73,088 36.2% 1992 7,967 2,943 6,024 36.9% 1993 117,698 43,785 73,783 37.2% 1993 8,206 3,075 5,131 37.5% 1994 120,665 46,131 74,433 38.3% 1994 8,463 3,213 6,239 38.0% 1996 123,638 48,603 75,035 39.3% 1995 8,706 3,368 6,348 38.6% 1996 126,790 61,208 75,582 40.4% 1996 8,967 3,609 6,469 39.1% 1997 130,022 63,962 76,069 41.6% 1997 9,236 3,B68 6,670 39.7% 1998 133,336 66,843 76,492 42.6% 1998 9,614 3,831 6,683 40.3% 1999 136,735 59,890 76,846 43.8% 1999 9,799 4,003 6,796 40.9% 2000 140,220 63,099 77,121 45.0% 2000 10,093 4,183 6,910 41.4% Calculate Growth Rates Calculate Growth Rates 198b-2000 1985-1985 1972-1982 1986-2000 1978-2000 1976-1984 Pop. Growth Rate: 3.00X 3.00X Pop. Growth Rate: 2.S5X 2.65X Urbanization Rate: 1.4SX 1.456 3.82X Urbanization Rat.: 2.74X 2.74X Urban Growth Rate: 4.49X Urban Growth Rate: s.38s Rural Growth Rate: 2.12X Rural Crowth Rate: 0.91X Sourcsces Population Growth Rate derived from National Population Population Growth Rate from WDR 1987. Commission figures for 1986 and 2000, cited in Lelc et al, Urbanization rate derived from WDR 1987 estimate of 27X 'Nig;ria's Economic Development ...' April 1988 draft, in 1985 and 38X for 1986. Note that projecting the Urbanization rate derived from WDR 1987 estimnte of 30X for Government's rate (1972-1982) would yield 58X urban by 1985, and Nigeria: Basic Economic Report,' Aug. 1981 for 2000 the year 2000. figure of 456. Table 3 Population pressure and deforestation, 1974-1984 (as percentage of total forest area) Country Per Capita Arable TROPICAL FOREST STLOJY 1/ FAO PRCOUCTION YEARBOOK 2/ Land (Ha/Person) In 'GOO As a % In '000 As a % Rural Total Hectares of Total Hectares of Total Malawi 0.53 0.48 1,200 24% 450 9% Nigeria 1.01 0.71 2,850 16% 2,700 15% Senegal 1.02 0.70 500 8% 308 5% Kenya 0.86 0.73 190 5% 270 7% Cameroon 5.23 3.34 800 3% 983 4% Tanzania 2.59 2.30 100 0% 1,063 2% Sources: 1/ Forest Resources of Tropical Africa, Part I. Table 6d, P. 88. lncludes closed broadleaved, conlferouis and bamboo forests. 2/ FAO Production Yearbood, Vol. 39. 74 Notes 1. The six countries selected for analysis (Kenya, Malawi, and principle of farming is to change the natural system into one Tanzania in East Africa, and Cameroon, Nigeria, and Senegal in which produces more of the goods desired by man. The man- West Africa) collectively account for 40 percent of the population made system is an artificial construction which requires continu- of Sub-Saharan Africa and 50 percent of its GNP. They cover ous economic inputs obtained from the environment to maintain almost all the ecological zones in Africa, ranging from the Sahelian its output level. Farming thus implies the abolition of an and Guinea Savannah zones in the North to the equatorial rain unproductive 'steady state' in favor of a man-created, more forest in the South, and including the volcanic, humid, and semi- productive but unstable 'state,' and much of the farm input humid highlands of East and West Africa. Taken together, the six (tillage, fertilizers, weeding, etc.) is nothing but an effort to grow almost all the major crops of Africa, including tea, coffee, prevent the new state from declining towards an unproductive cocoa, tobacco, cotton, groundnuts, cashews, sisal, sugar, maize, low-level steady state" (Ruthenberg 1980, p. 9). Increasing the sorghum, millet, and rice. They include two oil-exporting and four intensity of cultivation increases the relative instability in the oil-importing countries, two land-surplus and four land-short ecosystem. The danger of instability is that if sufficient inputs are countries. Despite their diverse physical characteristics, and not maintained (or invested) over time the plot will return not although they have followed different policy paths and achieved merely to its former low-productivity state, but to a state of lower different outcomes, the six countries have enough in common to potential, as is evidenced by "desertification" of marginal lands. permit fruitful comparison. MADIA is a REPAC-(Research Approval 6. There are many cases where population growth, rather than Committee) funded research project approved in June of 1984. increasing capital accumulation, has depressed savings and The MADIA study has the active support of seven donor agencies diverted investment away from production to consumption. See from Denmark, France, the Federal Republic of Germany, Sweden, for instance Ruttan 1984. the United Kingdom, the United States, and the Commission of 7. For a more detailed discussion of the role of ethnicity on the the European Communities. making of agricultural policy, see Lele and Hanak, eds., The Politics 2. Ruthenberg (1983, p. 15) defines the R-value (or "intensity of of Agricultural Policy, forthcoming. rotation") as 8. Note that in Nigeria a large work on land potential has been R 00 = Y, toocompleted for the North-Central plains. See Ministry of Overseas R- '4t Development 1979. Yf + Y' 9. A good source of further reading on interactions between ecology and development economics can be found in H. Daly where: Yt = years cultivated (1989). Yf= years fallow 10. Figure of 19 percent cited in "SAL IV: Adjustment with Growth and Development," Malawi Government (Office of the Thus, if a plot were cultivated for 3 years continuously and then President/Ministry of Finance) Special Studies Document 1986/2 left fallow for the next 7 years, the R-value would equal 30. (January, 1987), p.vii.; figure of 56 percent arable is cited in Malawi Similarly, annual cropping without fallow would have an R-value of Population Census 1977: Analytical Report, Vol. 1. National Statistical 100, and growing more than one crop per year each year would Office (Zomba: 1984), p.3. have an R-value above 100. 11. See for instance the "sources of growth" analysis in jammeh 3. The production of flue-cured tobacco is considered harmful and Lele 1988. to the environment insofar as the treatment process consumes a 12. In Cameroon, regional demographic surveys were under- fair amount of wood and contributes to pressure on wood taken from 1960-65, but the first full national census was in 1976. resources. However, the effects are occurring through the expan- Likewise in Senegal, the first complete national census in 1976 sion of estates, bypassing smaliholders from potential sales. For was predated only by an administrative census in 1960 and a a more thorough critique of tobacco production on the environ- demographic survey in 1961 (Domschke and Goyer 1986). ment, see Boehnert, 1988. 13. An intermediary step in the "normal" trajectory of intensi- 4. While recognizing the fundamental importance of irrigation, fication includes significant rural-to-urban migration as the however, the MADIA study documents the extent to which the productivity of labor decreases. Boserup writes that possibilities for small-scale irrigation, whether developed by farmers by using traditional means or the more modern tubewells . . . people in rural areas, instead of voluntarily accepting the and valley bottom development schemes, are unexploited harder toil of a more intensive agriculture, will seek to relative to the complex and capital-intensive large-scale irrigation. obtain more remunerative and less arduous work in non- Not only have governments shown frequent preference for such agricultural occupations. (Boserup 1965, p. 118.) irrigation but donors have provided large support for it. Examples 14. We were fortunate to receive a significant contribution to include the Bura irrigation scheme in Kenya (at the cost of $25,000 this section from G.M. Higgins, who helped to draft the FAO/ per hectare), the River Basin Development Authorities in Nigeria UNFPA/IIASA study. We are grateful for his reviewing this section (at the cost of between $35,000 and $100,000 per hectare), the and making helpful suggestions on the original manuscript. SAED irrigation schemes in the Fleuve (the cost of which is 15. Higgins, G.M./UNFPA/IIASA 1982. Three levels of input use unknown but estimated by FAO at $50,000 per hectare), and the are assumed in the FAO/IIASA analysis to calculate the kilocalorie SEMRY projects in Northern Cameroon ($13,000 per hectare). Each production frontier: exemplifies inappropriate technocratic approaches that donors a) low level assumes only land and labor, and no soil supported because of historical political involvement without conservation; regard to the development of the appropriate capacity for their b) intermediate level assumes improved hand tools and/or draft management. Important exceptions to this are the World Bank's implements, some fertilizer and pesticide application, moderate support for tubewells and surface irrigation on Kebrija in soil conservation, and a cultivation mix of improved and tradi- Northern Nigeria and the valley bottom development in tional crops; and Cameroon. c) high level assumes "complete mechanization, full use of 5. Initially, Ruthenberg argues, land is at low productivity but in genetic material," necessary farm chemicals, soil conservation equilibrium. To increase the land's current productivity is to risk measures, and cultivation of "only the most calorie (protein) jeopardizing its future productivity. He observes, "the basic productive crops on all potentially cultivable rainfed lands." 75 16. Several assumptions implicit in the FAO/lIASA analysis are provide ample soil cover curtailing soil erosion which normally masked by giving results in terms of sustainable populations. For sets in after trees are felled for saw milling and for other every potential population that can be sustained (at given levels purposes. As trees are kept short by constant picking, it was of input use), decisions have been made regarding optimal land expected that the tea zones would act as buffer zones and use with respect to crops, consumer preferences, minimum calorie trespassers into forests are easily sighted from considerable requirements, and response coefficients. These variables are used distances" (Weehly Review, 1989). to calculate a production possibility frontier in kilocalories, based 23. It is now well established that symbiotic root microorgan- on agroclimatic and soil constraints. The assumptions remain isms (Rhizobium, Frankia, and mycorrhizal fungi) can effectively largely hidden as the study lists only the end result: sustainable contribute to tree productivity in marginal climatic and edaphic population figures. conditions. Since significant advances have been made recently in 17. In a recently published Ph.D. dissertation, Boehnert (1988) the manipulation of the microorganisms, it is not possible to notes that "the increasing population is pressing more and more contemplate their use in the field .... A number of trees have the people into the arid and semi-arid areas. With them they bring potential for fixing atmospheric nitrogen through their symbiotic their traditional farming practice, used in wetter and cooler areas associations with Rhizobium (leguminous trees) or Frankia (nitrogen- with a different soil structure. For example, deep ploughing with fixing nonleguminous plants, now dubbed actinorrhizal plants). heavy farm equipment and the custom of keeping the soil Promoting the nitrogen fixation capacity of these trees through cultivated and open most of the year." inoculation with the proper symbiotic microorganisms or through 18. Crops can be high value in terms of relative price, but may selection of the plant host is an elegant approach to making the not yield higher returns if yields are low. Cassava is considered a forest ecosystem self-sufficient in nitrogen (Gorse and Steeds low value crop, but returns are higher than cocoa in Nigeria 1985, p. 54). because of its high yields and the lower yields of aging cocoa 24. See for instance Forest Resources Crisis in the Third World, trees that are nearing the end of the 20-year productive cycle. Proceedings from the Conference, September 6-8, 1986. Sahabat 19. One exception is the volcanic soils, found in highlands such Alam Malaysia (Penang: 1987). For a more optimistic scenario, see as in Kenya and Western Cameroon, which are deep and remain Anderson 1987. highly productive year after year. 25. Although in absolute terms the Rift Valley province contains 20. The importance of wood as a source of fuel is nicely more high potential land (911,500 hectares) than either the illustrated by the fact that the cuisine of the Sahelian and Central, Western, or Nyanza provinces, the relevant proportion of Sudanian zones consists mainly of simmered stews, sauces, and high potential to total land is much lower-only 6 percent as grain porridges, whose preparation requires slow cooking and a compared to about 25 percent in the Central province. The lower great deal of wood using the traditional "three-stone" stove proportion of high potential land, the large tracts of medium and (Gorse and Steeds 1985, p. 29). low potential land, and the inclusion of nomadic peoples in the 21. The critical position of Nigeria and Malawi is confirmed by equation-such as the Turkana and the Masai (who constitute just other available evidence. The FAO Production Yearbook also has under 10 percent of the Rift Valley population)-may help explain data on area under forest/woodlands that suggest a positive the appearance of a more abundant supply of arable land in the relationship between diminishing area under forests and wood- Rift Valley whereas its high potential districts are equally densely lands and population densities. The area under forests decreased populated. by 15 percent, for instance, in Nigeria during the 1974-84 period. 26. The land survey was published in 1965, and subsequently In Malawi and Kenya, also characterized by high population republished in 1985 (Stobbs and leffers 1985). These figures are pressures, forest area is listed as decreasing by 7 and 9 percent, also cited by the government in 1977 Compendium of Agricultural respectively (see table A.3, annex). The figures are slightly less for Statistics. Arable land estimates are generally more conservative Senegal (5 percent), Cameroon (4 percent), and Tanzania (2 than the figures given above; the Office of the President, for percent). example, has cited the figure of 19 percent arable in "SAL IV: A 22. A recent article in Kenya's Weekly Review presents the Proposal..." while the World Bank has alternately cited 38 government position on the new Nyayo tea zones as "an percent cultivable ("1981 Development of the Agricultural Sector outstanding example of President Daniel Arap Moi's commitment Report"') and more recently 22 percent without forests, 62 percent to environmental conservation. Inaugurated by the president with (Land Policy Study 19871 p.7). Elsewhere Mkandawiri and himself in 1984, it was billed as one of the most effective means Phiri, "Land Policy Study" (1987) cite the figure of 37 percent of protecting and conserving Kenya's forests against wanton arable as a national average. We use the first set of figures as they destruction through illegal human settlements. Tea is planted in represent official government data and are more disaggregated a thin strip of land adjacent to gazetted forests. The tea bushes (to the district level), although they may be high. 76 Bibliography Ake, Claude. 1987. "Sustaining Development on the Indigenous." Lele, Uma. Forthcoming. "Agricultural Development in Africa: The University of Port Harcourt, Nigeria (December). Lessons from the MADIA Study." Brussels: The Courier/EEC. Anderson. Dennis. 1987. "The Economics of Afforestation: A Case July. Study in Africa." World Bank Occasional Paper No. I. . 1988a. "Agricultural Growth, Domestic Policies, the External Baltimore: johns Hopkins University Press. Environment and Assistance to Africa: Lessons of a Quarter Binswanger, Hans, and Prabhu Pingali. 1988. "Technological Century," in Colleen Roberts, ed, Trade, Aid, and Policy Reform: Priorities for Farming in Sub-Saharan Africa." Research Proceedings of the Eighth Agriculture Sector Symposium. Washing- Observer 3, No. 1, January. ton D.C.: \World Bank. Boserup, Ester. 1965. The Conditions of Agricultural Growth: The . 1988b. "Empowering Africa's Rural Poor: Problems and Economics of Agrarian Change under Population Pressure. New Prospects in Agricultural Development," in lohn P Lewis, ed. York: Aldine Publishing Company. Strengthening the Poor: What Have We Learned? Overseas . 1981. Population and Technological Change: A Study of Long-Term Development Council: Washington, D.C. Trends. Chicago: The University of Chicago Press. 1. 989. "Sources of Growth in East African Agriculture." The Boehnert, Joachim. 1988. Agroforestry in Agricultural Education with a World Bank Economic Review 3 (lanuary: 119-144. Focus on the Practical Implementation. TROPICAL AGRICULTURE -, William Kinsey, and Antonia Obeya. 1989. "Building #2. Dr. Doris Knuth, ed. Gaimersheim, West Germany: Agricultural Research Capacity in Africa: Lessons from Verlag Josef Margraf Scientific Books. MADIA Countries." MADIA Paper.. Daly, Herman E. 1989. "Steady-State Versus Growth Economics: , Nicolas van de Walle, and Mathurin Gbetibouo. 1989. Issues for the Next Century." Paper prepared for the Hoover "Cotton in Africa: An Analysis of Differences in Perfor- Institution Conference on Population, Resources, and Envi- mance." MADIA Paper. ronment. February. _, Robert E. Christiansen. and Kundhavi Kadiresan. 1989. Diejomah, Vremudia P. 1988. "Population, Land Use, and Food "Issues in Fertilizer Policy in Africa: Lessons from Develop- Self-Sufficiency in Africa: An Overview." From Africa Popula- ment and Adjustment Lending, 1970-87." MADIA Paper. tion Conference, Vol. 3. Dakar. November. ._, Robert E. Christiansen, Paul Fishstein, and Mathurin Domschke, Eliane, and Doreen S. Goyer. 1986. The Handbook of Gbetibouo. 1989. "Planning for Food Security in Africa: National Population Censuses: Africa and Asia. New York: Green- Lessons and Policy Implications." MADIA Paper. wood Press. ___, and Manrnohan Agarwal. 1989. "Smallholder and Large- Food and Agriculture Organization. 1986a. Production Yearbook, 1985. Scale Agriculture: Are There Trade-Offs in Growth and Vol. 39. Rome. Equity?" MADIA Paper. . 1978. Report on the Agro-Ecological Zones Project. World Soils Matlon, Peter J. "The West African Semiarid Tropics," in John W Resources Report. No. 48, Vol. 1. Rome. Mellor, Christopher L. Delgado, and Malcom J. Blackie, eds., . 1986b. Atlas of African Agriculture. (African Agriculture: The Accelerating Food Production in Sub-Saharan Africa. Baltimore: Next Twenty-Five Years). Rome. The Johns Hopkins University Press. _\___ 1981. Forest Resources of Tropical Africa, Parts I and II. Technical Mellor. John and Bruce F. Johnston. 1984. "The World Food Report of the Forest Resources Assessment Project. FAO/ Equation: Interrelations Among Development, Employment, United Nations Environment Program. Rome. and Food Consumption." Journal of Economic Literature 22 Gorse, lean Eugene, and David R. Steeds. 1987. "Desertification in llune). the Sahelian and Sudanian Zones of West Africa." World Oram, Peter. 1987. "Africa: An International Food Research Bank Technical Paper No. 61. Strategy." Mimeo, International Food Policy Research Insti- Hayami, Yujiro, and Vernon W. Ruttan. 1985. Agricultural Development: tute, Washington D.C., October. An International Perspective. Baltimore: Johns Hopkins Univer- Pingali, Prabhu, Yves Bigot, and Hans P. Binswanger. 1987. sity Press. Agricultural Mechanization and the Evolution of Farming Systems in Harrison, Paul, and Earthscan. 1987. The Greening of Africa; Breaking Sub-Saharan Africa. Baltimore: The Johns Hopkins University Through in the Battle for Land and Food. London: Paladin Hans P Grafton Books. - and Hans P Binswanger. 1984. "Population Density and Higgins, G.M., et al. 1982. Potential Population Supporting Capacities of Farming Systems: The Chang ing Locu s of I nnovations and Lands in the Developing World: Technical Report of Project INT/75/ DC Chng. B nk O tober P13 "Land Resources for Populations of the Future." FAO/IIASA: D.C.: World Bank. October. Rome. Riddell, lames C., and Carol Dickerman. 1986. Country Profiles of Ho, Teresa 1. 1985. Population Growth and Agricultural Productivity in Land Tenure- Afrca 1986. Land Tenure Center, Universipy of Sub-Saharan Africa. World Bank, Population, Health and Wisconsi-Madison, April. Nutrition Department. Rosenzweig. M.R., Hans P Binswanger, and John McIntire. 1984. International Labour Office. 1986. Economicalfy Active Population: "From Land Abundance to Land Scarcity: The Effects of Estimates and Projections, 1950-2025, Vol. 11, Africa. Geneva. Popunation Growth onk PDroduction Paper ARU 28. Washing- lnternational Monetary Fund. International Financial Statistics. Wash- ton, D.C.: World Bank. November. ington, D.C.: Bureau of Statistics/[ME Various years. Ruttan, Vernon W. 1984. "Perspectives on Population and Develop- Joosten, J.H.L. 1962. "Wirtschaftliche and Agrarpolitische Aspekte ment." Indian Journal of Agricultural Economics 39, No. 4 Tropischer Landbausysteme" (mimeo). Institut Fur Landwirt- (October-December). schaftlische Betriebslehre, Go~ttingen. Srinivasan, TN. 1989. "Food Aid: A Cause of Development Failure Leach, Gerald, and Robin Mearns. 1988. Beyond the Fuelwood Crisis: or an Instrument for Success?" The World Bank Economic People, Land, and Trees in Africa. London: Earthscan Publica- Review 3 (January). tions Limited. 77 Wotid Bank. 1987. World Development Report. New York: Oxford . . Kenya Tea Development Authority Annual Report. Various years. University Press. Jaetzold and Schmidt, 1982. Farm Management Handbook of Kenya: .1988. Renewable Resource Management in Agriculture. Report Natural Conditions and Farm Management Information. Govern- 7345, June. ment of Kenya/Ministry of Agriculture and Germany Agency . 1984. World Development Report. New York: Oxford University for Technical Cooperation. Press. Shipton, Parker. 1987. "The Kenyan Land Tenure Reform: Misun- Zachariah, K.C., and M.T. Vu. 1988. World Population Projections, 1987- derstandings in the Public Creation of Private Property." 88 Edition: Short- and Long-Term Estimates. Baltimore: The Development Discussion Paper No. 239. Cambridge, MA: Johns Hopkins University Press. Harvard Institute for International Development, February. Weekly Review. 1989. "The Nyayo Tea Zones: The Struggle Against Cameroon Soil Erosion." Nairobi. 20 January. Chester, Lauren A. 1988. "Cameroon: Population Growth and Land World Bank. 1983. Land Issues Paper. Report No. 4391-KE. Resources: A Case Study." Population and Human Resources Department. Washington, D.C.: World Bank. Malawi November (Revised). Carr, S. 1. 1988. "Modification and Extension of the National Rural Gaviria, Juan. 1988. "Some Notes on Feeder Roads and Their Development Program." Prepared for Malawi: Symposium Impact on Regional Agricultural Performance in Cameroon." on Agricultural Policies for Growth and Development. World MADIA Paper. Bank. November. Government of Cameroon 1985. Ministry of Highways. "Inventory Deloitte, Haskins, and Sells. 1987. "ADMARC Organization and of Feeder Roads." under contract with BCEOM. Management Review." Government of Cameroon. Ministry of Agriculture. 1980. Bilan . 1986. "An Appraisal of the Proposed Extension and Training Diagnostic du Secteur Agricole au Cameroun. Yaounde, Camer- Service for the Estate Sub-Sector in Malawi." April. oon: Government of Cameroon. Mkandawire, Richard M., and Chimimba David Phiri. 1987. Assess- International Fertilizer Development Center IIFDC). 1986. "Camer- ment of Land Transfer From Smallholders to Estates. January. oon Fertilizer Sector Study." Muscle Shoals. Alabama: IFDC. National Statistical Office. 1977. Compendium of Agricultural Statistics, McHugh, Dermot. 1988. "Maize-Based Cropping Systems in the 1977. Zomba, April. Ndop Plain of the North West Province of Cameroon." Paper . 1984a. "Malawi Population Census, 1977, Analytical Report, prepared for Ministry of Higher Education and Scientific Volumes I and 11: Income and Expenditure, Crop Storage, Livestock, Research (MESRES) and USAID. Resources and Nutrition. Government Printer, Zomba, April. Ministere de L'Economie et du Plan. "V\ Plan Quinquennal de _ 1984b. "National Sample Survey of Agriculture 1980/81, Volumes Developpement Economique, Social et Culturel 1981-1986." 1, 11, and 111. Ministere du Plan et de l'Amenagement du Territoire. 1986. VI' _. 1985. Reported Employment and Earnings Annual Report. Zomba, Plan Quinquennal de Developpement Economique. Social et Culturel June. 1986-1991. . . 1988. "Malawi Population and Housing Census. 1987." World Bank. 1987. Cameroon Country Economic Memorandum. Report Preliminary Report, Zomba. No. 6395-CM. Washington, D.C. _ . Tobacco Control Commission. "Tobacco Control Commis- World Bank. 1986. Cameroon Agricultural Sector Review, Issues Paper. sion Circulars." Limbe: TCC. Various years. Draft. AFIAG Division, Washington, D.C. Office of the President and Cabinet Department of Economic Kenya Planning and Development. Statement of Development Policies K~enya 1987-1996. Government Printer, Zomba. de Graaff, lan. 1986. The Economics of Coffee. Wageningen, the Stobbs, A.R., and J.N.R. leffers. 1985. "Land Use Survey of Malawi, Netherlands: Pudoc Publishers. 1965-67. Land Resources Development Centre, England. Government of Kenya. Central Bureau of Statistics. 1977. Integrated Twyford, 1. T. 1988. "The Development of Smailholder Fertilizer Rural Survey, 1974-75: Basic Report. Ministry of Finance and Use in Malawi." Paper presented at FAO/FIAC meeting, Planning, March. Rome, April 26-29. _ . 1980. Agricultural Census of Large Farms, 1978. Ministry of World Bank. 1986. "Malawi Population Sector Review. Report No, 5648- Planning and Development. MAI, March 10. _ 1981a. The Integrated Rural Surveys, 1976-79: Basic Report. World Bank. 1987. "Malawi-Land Policy Study. Eastern and South- Ministry of Economic Planning and Development, ern Africa Division, April 24. November. 1981b. 1979 Population Census, Vol. I: Analytical Report. Ministry Nigeria of Finance and Planning, June. Federal Ministry of Science and Technology. 1985. "Impact of _ . Statistical Abstract. Ministry of Economic Planning and National Agricultural Research." September. Development. Nairobi. Various years. Federal Office of Statistics. 1983. Unpublished Ministry of Agricul- _ . Ministry of Agriculture and Livestock Development. 1985. ture Crop Area, Production, and Yield Spreadsheets. Kenya Agricultural Research Strategy and Plan: Priorities and Lele, Uma, Ademola Oyeiide, Vishva Bindlish, and Balu Bumb. Programs. ISNAR, The Hague, Netherlands. March. 1989. "Nigeria's Economic Development, Agriculture's Role . Ministry of Finance. 1987. National Cereals and Produce Board and World Bank's Assistance: Lessons for the Future." Reorganization Study. Technosynesis S.p.a. in association with MADIA Paper. Coopers and Lybrand Associates. Nairobi. October. Ministry of Overseas Development, Land Resources Development _ . 1983. "Grain Marketing Study: Interim Report." Booker Center. 1979. "Land Resources of Central Nigeria: Agricul- Agriculture International Ltd. in association with Githongo tural Development Possibilities." Surrey, England. and Associates. Nairobi. July. Nigeria Federal Census Office. 1963. Population Census of Nigeria, _ 1985. Economic Survey, 1985. Ministry of Planning and 1963. Lagos. National Development, Nairobi. World Bank. 1981. "Nigeria: Basic Economic Report." West Africa _ 1987. Economic Survey, 1987. Ministry of Planning and Programs 1, Report No. 3341-UNI. Washington, D.C. August. National Development, Nairobi. 78 Senegal Tknzania Jammeh, Sidi C. 1987- "State Intervention in Agricultural Pricing FAO/World Bank. 1987. "Tanzania Agricultural Sector Review and Marketing in Senegal." Ph.D. Dissertation for the lohns Mission: Annex." FAO/World Bank Cooperative Programme Hopkins University, Baltimore, November. Investment Center. Rome. July. Jammeh, Sidi C. and Uma Lele. 1988. "Building Agricultural Government of Tanzania, Bureau of Statistics. 1970. Statistical Research Capacity in Senegal." MADIA Draft Paper, June. Abstract 1970. Ministry of Planning and Economic Affairs. Dar Senegal Bureau National du Recensement. 1982. Recensement es Salaam. General de la Population d'Avril 1976. Resultats Definitifs, Donnees . 1981. 1978 Population Census. Dar es Salaam. Corrigees. Dakar. . 1989. 1988 Population Census: Preliminary Results. Dar es Sen6gal Direction d'Eaux, For6ts, et Chasses. 1978. Rapport Annuel. Salaam. Dakar. . 1983. Statistical Abstract, 1973-1979. Ministry of Planning and Senegal Direction de la Production Agricole. Rapport Annuel. Economic Affairs. Dar es Salaam. Various years. Dakar. . 1973. Bureau of Statistics/Demographic Unit. An Analysis of S6negal Direction de la Statistique. 1982. Situation Economique du the National Demographic Survey of Tanzania. Vol. VI: The Senegal. Dakar. Demography of Tanzania. Rushid A. Henin, ed. Senegal Ministere du Developpement Rural. 1987. Bilan de la . Central Statistical Bureau. 1969. 1967 Population Census. Dar Campagne de Production Agricole, 1986-1987. Dakar. es Salaam. Senegal Ministere du Plan et de la Cooperation 1985. Projet de VIl . Ministry of Agriculture. 1980. "Price Policy Recommenda- Plan de Developpement Economique et Social: 1985/1989. Dakar. tions for the 1981-1982 Agricultural Price Review." . 1981. Statistiques et Indicateurs des Regions du Senegal. Dakar. . 1987. "Annual Review of Maize, Wheat and Rice." August. "Kindoni, Kahama 'Lead in Population'." Tanzania Daily News, 16 Waterbury, lohn. 1986. "Agricultural Policy Making and Stagnation in March 1988. Senegal." MADIA Draft Paper, December. World Bank. 1983. Tanzania Agricultural Sector ReporL Washington, D.C. 79 ! ., THE MADLA STUDY Although many generalizations have been made about the agricultural crisis in Africa, relatively few detailed country and cross-country studies of African agriculture based on systematic data analysis have been conducted. Similarly, although foreign aid has constituted a large part of total government expenditures in Africa for close to fifteen years, there has been little analysis of the role of external assistance in African countries that goes beyond political criticism of official assistance or the alleged self- serving objectives of donors. The impetus for the study "Managing Agricultural Development in Africa" (MADIA) was to begin the process of filling this gap and to explain the nature and sources of the agricultural crisis, particularly the extent to which it originated in resource endow- ments, historical and contemporary events, external and internal policies, and the economic and political environment. The MADIA study involved detailed analysis of six African countries- Kenya, Malawi, Tanzania, Cameroon, Nigeria, and Senegal. In addition to the World Bank, seven donors, USA1D, UKODA, DANIDA, SIDA, the French and German governments, and the EEC participated in the study. The analysis of country policies and performance during the last 20-25 years was carried out with the benefit of substantial input from the governments and nationals of each of the countries represented. The study had three main areas of focus: (I) the relationship between domestic macroeconomic and agricultural policy and agricultural performance, (2) donors' role in the development of agriculture, and (3) the politics of agricultural policy. The MADIA study was the result of encouragement and support from many people. Anne Krueger, former Vice President for Economic Research Staff in the World Bank, encouraged the establishment of these studies on aid and development in 1984. Gregory Ingram, former Director of the Development Research Department, provided unstinting support for the study. During the reorganization of the World Bank in 1986, the strong support from Benjamin King, then acting Vice President for Economic Research Staff, proved invaluable. Barber Conable, President of the World Bank, and Mr. Edward V. K. Jaycox, Vice President for the Africa Region, have played a key role by ensuring support for the study's completion. as did Stanley Fischer, the Vice President for Development Economics. Yves Rovani, Director General of the Operations Evaluation Department, was particularly helpful as the MADIA study drew heavily on the works of OED. A special debt of gratitude is owed to the World Bank's Research Committee, which provided the initial funding for the study, and to the MADIA Steering Committee. In particular the strong support of the chair of the Steering Committee, Stephen O'Brien, has been of critical importance. Finally, without the active and continued encouragement of many African policymakers and donor officials, including numerous colleagues in the World Bank, this study would not have provided new perspectives. This support has taken the form of numerous reactions to written and oral presentations, and refinement of the analysis to identify the areas of consensus and continuing controversy. (D X The World Bank Headquarters 18 1 8 H Street, N.W. CD Washington, D.C. 20433. U.S.A. Telephone: (202) 477-1234 Facsimile: (202) 477-6391 Telex: WUI 64145 WORLDBANK O RCA 248423 WORLDBK Cable Address: INTBAFRAD WASHINGTONDC 0 European Office fD 66, avenue d'1ena 75116 Paris, France Telephone: (1) 40.69.30.00 Facsimile: (I) 47.20.19.66 Telex: 842-620628 Tokyo Office Kokusai Building 1-1, Marunouchi 3-chome Chiyoda-ku. Tokyo 100, Japan Telephone: 13) 214-5001 _. Facsimile: (3) 214-3657 ( Telex: 78 1-26838 5 :7 CD 00 0-8213- 1320-7