Malawi - Food security report

Report No. 8151-MAI Malawi Food Security Report June 27,1990 Agriculture Operations Division Southern Africa Department Africa Region FOR OFFICIAL USE ONLY Document of the World Bank This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Kwacha/US$ -- Average for Year 1975 .864 1976 .913 1977 .903 1978 ^844 1979 .817 1980 .812 1981 .895 1982 1.056 1983 1.175 1984 1.413 1985 1.719 1986 1.861 1987 2.209 1988 2.561 1989 2.822 Source: IMF - International Financial Statistics.  FOR OMCIAL USE ONLY Preface This report is part of a special initiative by the World Bank Africa Region, in conjunction with the Food and Agriculture Organization and the World Food Program, to develop food security strategies for several African countries. This report presents the findings of a mission led by V. R. Southworth that visited Malawi between February 22 and March 16, 1989. The report draws on the work of several individuals, including: J. de Beyer, A. Duncan, P. Fleming, J. Meindertsma, B. Mlingi, I. Moreithi, and K. Tuinenberg. The principal author of the report was R. Christiansen. The work of these individuals could not have been completed without the cooperation of the Government of Malawi. This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  ACRONYMS ADD Agricultural Development District ASA Annual Survey of Agriculture CDA Community Development Assistant DC District Commissioners EMU Emergency Management Unit EP&D Economic Planning and Development Division FACU Food Aid Coordination Unit FHA Farm Home Assistant FSNU Food Security and Nutrition Unit HIID Harvard International Institute for Development HCW Homecraft Workers IDA International Development Association JOCMR Joint Operations Committee for Mozambican Refugees MHCN Malawi Maternal and Child Nutrition NDPRC National Disaster Preparedness and Relief Cammittee NEWS National Early Warning Systems NRDP National Rural Development Program NSC National Seed Company NSSA National Sample Survey of Agriculture OPC Office of the President and Cabinet PHAM Private Hospital Association of Malawi SACA Smallholder Agricultural Credit Administration SFFRFM Smallholder Farmer Fertilizer Revolving Fund of Malawi SGR Strategic Grain Reserve IPF World Food Program  TABLE OF CONTENTS EXECUTIVE SUMMARY . . . . . . . . . . . . . .. . .i I. FOOD SECURITY ISSUES IN MALAWI . . . . . . . . . . . . . . . . . . . 1 A. Background . . . . . . . . . . . . . . . . . . . . . . . . . 1 B. Structural Adjustment and Smallholder AgricuLture . . . 1 C. Smallholder Production and Social Indicators . . . . . . . . . 4 D. Organization of Report ... . . . . . . . . . . . . . . . . . 6 II. FOOD AVAILABILITY AND MALNUTRITION IN HALAWI . . . . . . . . . . . . 7 A. Trends in Aggregate Food PIoductior . . . . . . . . . . . . . . 7 B. Evidence of Household Food Insecurity . . . . . . . . . . . . . 11 C. Causes of Malnutrition Among Children . . . . . . . . . . . . . 17 D. Conclusions ...... ........ 20 III. AGRICULTURAL PRODUCTION, INCOME, AND FOOD AVAILABILITY . . . . . . 22 A. Constraints to Increased Household Agricultural Production . . 23 B. Off-Farm Sources of Income and Household Food Security . . . . 30 C. Commercialization of Smallholder Agriculture . . . . . . . . . . 31 D. The Role of Women in Household Food Security . . . . . . . . . 37 E. Conclusions and Implications . . . . . . . . . . . . . . . . . 38 IV. POLICIES AFFECTING HOUSEHOLD FOOD SECURITY . . . . . . . . . . . . . 40 A. Agricultural Credit and Extension . . . . . . . . . . . . . . . 40 B. Input and Output Pricing . . . . . . . . . . . . . . . . . . . 42 C. ADMARC, The Private Sector, and Agricultural Marketing . . . . 50 D. Land Policies and Restrictions on Smallholder Production . . . 59 E. Operation of the Strategic Grain Reserve . . . . . . . . . . . 60 F. Conclusions and Policy Recommendations . . . . . . . . . . . . 61 V. GOVERNMENT INSTITUTIONS AND FOOD SECURITY . . . . . . . . . . . . . . 63 A. Food Security and Nutrition Unit . . . . . . . . . . . . . . . 63 B. National Disaster Preparedness and Relief Committee . . . . . . 64 C. Food Aid Coordination Unit . . . . . . . . . . . . . . . . . . 65 D. Joint Operations Committee for Mozambican Refugees . . . . . . 65 E. Ministry of Agriculture . . . . . . . . . . . . . . . . . . . . 66 F. National Early Warning System . . . . . . . . . . . . . . . . . 67 G. Government Nutrition Interventions . . . . . . . . . . . . . . 68 H. Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . 70 VI. CONCLUSIONS AND POLICY RECOMMENDATIONS . . . . . . . . . . . . . . . 73 BIBLIOGRAPHY . . . . . . . . . . . . . . . . . .. . 77 ANNEX 1: ADMARC Purchases of Smallholder Production of Tobacco, Groundnuts. and Cotton, 1972-1988 .................... 83 ANNEX 2: Prevalence of Malnutrition by District . . . . . . . . . . . 84 ANNEX 3: Households Depleting Maize Stocks . . . . . . . . . . . . . 86 ANNEX 4: Inter-Regional Maize Price Differentiation . . . . . . . . . 68 ANNEX 5: Comparison of SFFRPM and OPTICHEM Costs and Prices for 1988/89 98  LIST OF TABLES Table 1.1: The Ratio of Smallbolder Producer Prices to International Prices for Malawi, Kenya, and Tanzania, 1970-1986 . . . . . . . . . 3 Table 1.2s Trends in GDP and the Value of Agricultural Output, 1980-89 . 5 Table 1.3: Basic Social indicators for Kenya, Malawi and Tanzania . . . 6 Table 2.1: Estimated Maize Producti.on and ADMARC's Net Sales and Net Exports . . .. ... . . . . . . . . 8 Table 2.2: Net Cereal Imports and Self-Sufficiency Ratios for Kenya, Malawi, and Tanzania, 1961-87 ... . . . . . . . . 9 Table 2.3: Apparent Utilization of Maize and Food Crops . . . . . . . . 10 Table 2.4: Prevalence of Child Malnutrition in Malavi as Estimated from Various Surveys ....... . ......... 11 Table 2.5: Prevalence of Stunting, Wasting, and Underweight in Malavi by Age in Immediate Post Harvest Period, 1980 . . . . . . . . 12 Table 2.6t Prevalence of Severe Malnutrition . . . . . . . . . . . . . . 13 Table 2.7: Prevalence of Malnutrition in Zomba South . . . . . . . . . . 13 Table 2.8: Prevalence of Malnutrition by Age Group in Zomba South . . . 15 Table 2.9s Percent of Children Underweight by Age in Hzimba, Dowa, and Mwanza Districts, 1987 . . . . . . . . . . . . . . . . . . . 16 Table 2.10: Percent of Children under 5 Years of Age Suffering from Malnutrition as Measured by Weight for Age . . . . . . . . . 16 Table 2.11: Relative Ranking of Malnutrition Among ADDs . . . . . . . . . 17 Table 2.12: Fulfillment of Energy Needs Under Conditions of Food Shortage . . . . . ... .19 Table 3.1: Selected Population Densities in Sub-Saharan Africa 23 Table 3.2: Distribution of Households by ADD and Area Under Cultivation,1980181. . . .. ........... . 24 Table 3.3: Distribution of Agricultural Holdings by Size, 1987/88 . . 24 Table 3.4: Maize Yields by Agro-Ecological Zone and Holding Size . . . . 25 Table 3.5: Sale of Hybrid Maize Seed by ADMARC, Area Planted and Proportion of Total Maize Area . . . . . . . . . . . . . . 26 Table 3.6: Average Fertilizer Use in Selected African Countries . . . . 27 Table 3.7: Percentage of Households Using Fertilizer by Holding Size and ADD, 1984/85 . . . . . . . . . . . . . . . . . . . . . . 27 Table 3.8: Percent of Households Depleting Food Stocks by Region, Area Under Cultivation, and Time of Year for 1980/81 ... . . . . 28 Table 3.9: Main Source of Additional Supplies of Maize/Cassava for Food Deficit Households by Area and Holding Size . . . . . . . . . 32 Table 3.10: Ganyu Labor Days and Other Off-Farm Working Days in 1989 Kawinga Sample of Rural Households . . . . . . . . . . . . . 38 Table 4.1: Percentage of Households Receiving Credit in 1986/87 by ADD . 40 Table 4.2: Relative Cost of Comonly Applied Nutrients in East Africa . 43 Table 4.3: Transport As a Share of Marketing and C.I.F. Fertilizer Cost 43 Table 4.4: Benefit/Cost Ratios for Fertilizer Use on Maize, 1986187- 1988/89 .......... ... .......... 44 Table 4.5: Stock Flow Estimates for the SFPRFM, 1988/89-1992/93 . . . . 49 Table 4.6: Number of New Vehicle Registrations, 1978-1988 . . . . . . . . 53 Table 4.7 t Simulated Effects of Maize Price Differentiation on Household Food Security by Holding Size and Agro-Ecological Area . . . . . . 56  List of Tables (con't) Table 4.8: Simulated Effects of Maize Price Differentiation on Gross Margins Per Day of Labor by Holding Size and Agro-Ecological Area. . . . . . . 5 7 Table 4.9: Estimated Share of ADMARC Purchases at MK 24.60 . . . . . . . 58 Table 4.10: Impact of Differential Pricing on ADHARC's Maize Trading Account . . . . . . . . . . . . . . . . . . . . . . . . . . . 59  EXECUTIVE SUMMARY This report examines the extent and causes of malnutrition and food insecurity in Halawi. Although the report attempts to be comprehensive, it must be acknowledged that due to the complexity of the subject, it is impossible for a singLe report to be exhaustive in its coverage of the materials relating to food security even in a single country. With respect to the extent of malnutrition there can be little question that malnutrition is a serious problem in Malawi-- aoproximately half of the children under five Years of age are stunted due to chronic malnutrition. Evidence of malnutrition appears as early as 6 to 11 months of age, when the stunting rate is 34.1 percent. Among two-year old children (24 to 35 months), two-thirds are stunted. The causes of malnutrition are complex, but can be characterized as an interaction of the nature of the diet, disease, feeding practices, and-- most importantly--the availabil.L.y cf food at the household level. This report concentrates on the issues of household food availability and mechanisms for increasing the amount of food available at the individual household level. Recognition of the large number of factors that affect household food security and the level of malnutrition requires that a strategy for improving household food 6ecurity cannot be based solely on increased production of food and/or export crons. Instead a combination of efforts that will: increase household agricultural production, improve education on topics related to nutrition, expand supplementary feeding interventions, improve health services, and increase off-farm employment opportunities. This is not to deny, however, the fundamental importance of increasing the productivity of small-scale farmers as a means of increasing income and food security. The major recommendations that emerge from this report are divided into those affecting production and those concerning health and nutrition interventions. In brief, it is recommended thats o Government establish goals for improving the availability of credit to food insecure households and revising the technical advice provided by the extension service to the very small scale producers; o Government develop a program to improve the marketing of agricultural inputs which includes greater availability of fertilizer in remote areas, private sector sales of fertilizer, cash sales of fertilizer, and greater availability of fertilizer in small bags; o a thorough review of thes operations of SFPRFH be completed and that at a minimum, this review should cover thes (i) adequacy of financing arrangements, (ii) availability of grant-aid fertilizer, (iii) management operations, and (iv) scope of operations;  - ii - o Government work with donors to establish a long-term plan for increasing smallholder production of burley tobacco that will set specific goals in terms of the number of smallholders that will produce burley and the nature and level of assistance that the donors will provide; o Governments ti) establish guidelines for the operation of a more flexible price stabilization system, (ii) define the development functions that ADMARC is to prov'.de on behalf of the Government, (iii) set guidelines on how these functions are to be fulfilled, and (iv) determine what reimbursement ADMARC will receive from Government for performing these functions; o Government in cooperation with donors complete an evaluation of the domestic transport sector and design a program for enhancing rural transport services; o Government review the constraints placed upon activities in the informal sector and implement a liberalization program as a means of encouraging a multi-channel marketing system and the expansion of off-farm employment opportunities; o Government in cc-peration with donors design and implement a program that will relieve the demands placed on women's time, especially during the planting season; such interventions should include, but not be limited to: (i) improved availability of clean water supplies in villages, (ii) expanded rural health care facilities, (iii) improved access to modern food processing technology; o the Food Security and Nutrition Unit conducts a review of the objectives and implementation of the various programs dealing with food security issues with a view to coordinating their activities and establishing a training program for staff; o Government in cooperation with donors pursue an expansion of feeding programs for children and pregnant women which includes an education component along the lines of those being implemented by the WFP; and o Government and donors assign a high priority to establishing small-scale production units for likuni phala using locally procured materials.  I. FOOD SECURITY ISSUES IN MALAWI A. Background 1.01 During the 1970s Malawi's agricultu. sector and economy as a whole displayed one of the fastest growth rates for GDP and exports in sub-Saharan Africa. Between 1965 and 1980 the average annual growth of GDP was 5.8 percent, while that of exports, chiefly tobacco, tea, and sugar, was 4.1 percent (World Bank, 1989b). The principal force behind this impressive performance was the rapid expansion of estate production of tobacco and, to a lesser extent, sugar and tea. The strong performance of estate agriculture was, in large part. the result of preferential policy treatment accorded the estates. Although the estate-oriented development strategy was very successful as a means of generating growth, especially of exports, the benefits of this growth tended to be concentrated in the estate subsector without much spillover to the much larger smallholder subsector (see Kydd and Christiansen 1982; Christiansen and Kydd 1990; Lele 1989). The emphasis on development of the estate subsector contributed to the weak production performance of the smallholder sector throughout the 1970s and 1980s. This weakness has been a significant factor in compounding the problems of food insecurity at the household level, because productivity increases in smallholder agriculture have not been sufficient to allow food production to keep pace with population growth. 1.02 Despite the success of estate agriculture during the 1970s, the subsector suffered from a number of weaknesses, e.g., many estates were under- capitalized and suffered from weak management. These weaknesses contributed to the financial crisis and consolia _ion of the estate-based strategy in 1979-80 (Christiansen and Kydd 1990). The problems of the estate sector were exacerbated by several serious external shocks, the combination of which caused the Government to seek assistance from the IMF and World Bank through a structural adjustment program. All of these problems had consequences for smallholder employment, agricultural productivity, and household food security. B. Structural Adiustment and Smallholder Agriculture 1.03 In broad terms the structural adjustment program in Malawi has focused ont (i) restoring macroeconomic balance, (ii) improving agricultural pricing policy, and (iii) increasing the efficiency of the marketing parastatal, ADMARC. The reforms were necessary to address Malavi's eroding economic performance, which was due to a combination of external shocks and domestic policies. External factors included falling world prices for exports, rising oil prices, higher interest rates on external debt, and the increasing disrupticn of external transport routes through Mozambique, all of which contributed to a sustained decline in Malawi's terms of trade. Among the domestic factors contributing to weak economic performance were: structural weaknesses in the estate sector, the public sector deficits, and the lack of growth in the emallholder sector.  - 2 - 1.04 The reforms implemented under Malawv's structural adjustmcnt process were supported by three SALs, successive stand-by operations, and an extended facility from the IMF. These adjustment packages included macroeconomic measures such as improved public sector resource management that required reductions in budget deficits, strict credit and monetary policies, and external sector reforms. Smallholder Pricina Policy 1.05 The objective of the pricing policy reforms has been to stimulate smallholder production of erport crops by paying farmers prices that are closer to export parity. For maize, Malawi swings between being a net importer and a net exporter. and the objective is to set a maaize price that allows national maize self-sufficiency. For cotton and tobacco, the prices paid by ADYARC are the only ones available to smallholders, as ADMARC has a legal monopsony in these crops. Fcr all other crops, including maize, the prices represent a guaranteed minimum price because ADMARC is obligated to buy all quantities offered at that price.l/ 1.06 The implementation of the revised pricing policy has yielded mixed results. In general, since 1981 smallholders have received a higher share of the world price for some export crops (cotton and groundnuts), however, this trend has been hindered due to declining terms of trade which are, in turn, partially the result of the deteriorating external transport links (see Table 1.1). An exception to this trend has been the producer price of dark-fired tobacco. Although in recent years the producer price as a share of the price received at auction has risen, the price paid to the producer is still relatively low. The impact of higher prices on smallholder production, however, has been mixed. Although individual crops show considerable responsiveness to changes in relative prices, a combination of technical constraints and a shortage of arable land have hindered an increase in aggregate smallholder production in response to higher prices. 1.07 The experience with maize serves to highlight the price responsiveness of individual crop output. In 1981/82, the Government raised maize prices by 60 percent in response to food shortages. This price rise resulted in a large increase in ADMARC's maize purchases and a stagnation in groundnut and tobacco sales, as farmers shifted cropping patterns to take advantage of the shift in relative prices (see Annex 1). The resulting market surpluses not only allowed ADMARC to fill the newly established Strategic Grain Reserve, but also forced it to store, at considerable cost, the stocks that were in excess of its domestic needs and export opportunities. Faced with ADMARC's growing financial difficulties, Government pricing decisions for 1984/85-1986187 were designed to shift production toward export crops, particularly groundnuts. Farmers again responded to the change in relative prices by increasing 1. Producer prices are announced in September, well before the planting season (November-December), and are held constant over the entire season. The same price is paid by ADMARC at all its markets, no matter the distance from consumption centers or export points.  - 3 - marketed output of groundnuts, from 9,900 tons in 1984/85 to 44,800 tons in 1987/88. Over the same periad, marketed maize declined from 296,000 tons to 59,600 tons.21 Table 1.1 The Ratio of Smallholder Producer Prices to International Prices for Malavi. Kenya. and Tanzania. 1970-1986 ---Kenya--- --------Halawi --- ---Tanzania--- Dk-Fired Year Coffee Tea Tobacco Groundnut Cotton Tobacco Cotton 1970 0.91 0.60 0.22 0.69 0.24 0.43 0.72 1971 0.90 0.67 0.25 0.63 0.21 0.50 0.61 1972 0.98 0.63 0.23 0.70 0.22 0.46 0.57 1973 0.96 0.60 0.22 0.51 0.13 0.44 0.35 .974 0.97 0.55 0.23 0.28 0.15 0.42 0.32 1975 1.01 0.63 0.22 0.50 0.19 0.47 0.51 1976 0.85 0.57 0.21 0.51 0.13 0.40 0.41 1977 0.92 0.70 0.26 0.41 0.16 0.42 0.45 1978 0.94 0.64 0.26 0.41 0.17 0.47 0.55 1979 0.93 0.66 0.24 0.72 0.17 0.37 0.51 1980 0.98 0.76 0.23 0.84 0.14 0.35 0.52 1981 0.84 0.62 0.19 0.61 0.14 0.33 0.61 1982 0.83 0.56 0.24 1.28 0.17 0.30 0.73 1983 0.90 0.98 0.23 1.4' 0.17 0.38 0.67 1984 0.80 0.66 0.25 1.40 0.18 0.27 0.65 1985 0.88 0.76 0.21 1.23 0.22 0.36 1.07 1986 0.79 0.69 0.22 N.A. 0.33 0.32 1.11 --------------------------------------------------------------- Source: Christiansen and Southworth 1988. 1.08 The response of smallholders to the changes in output pricing policy indicates that there are other constraints affecting agricultural production. For example, the farmerls ability to increase production in response to changing prices hinges critically on timely access to inputs, the availability of appropriate technology, and reliable output marketing arrangements. In Halawi, these nonprice factors currently appear to be the binding constraints to increased productivity. As discussed below, the land constraint in the smallholder subsector prevents most farmers from expanding areas under cultivation. 2. The size of the decline in ADMARC purchases of maize in 1987/88 is exaggerated because private sector trading was liberalized that year and a considerable but undetermined amount of maize was sold through private channels.  - 4 - 1.09 The same factors that prevent a significant aggregate output response also affect food security. Most notably, the low produictivity that characterizes smallholder production hinders off-farm employment opportunities, increases in food production, and the generation of income from other crops. 1.10 A final element of the reform of smallholder pricing policy was the effort launched in the mid-1,A80s to eliminate subsidies on smallholder fertilizer. A phased four-year program to reduce the subsidy and encourage the use of high analysis fertilizer was initiated as part of the Government's structural adjustment program and was supported with assistance from USAID. The program, however, was suspended by the Government before completion because of concerns about the impact of fertilizer prices on the ability of the smallholder to produce adequate quantities of marketed maize. The Government has indicated its intention eventually to remove the subsidy, but at a pace that takes into account other factors influencing the cost of fertilizer, most notably high external transport costs. Private Sector Participation in Agricultural Marketing 1.11 A critical element of the liberalization program was the effort to encourage private sector marketing of all smallholder crops except tobacco and cotton. The objective of this measure was to create an environment conducive to the development of a multi-channel marketing system where private traders could compete with ADMARC in the buying and selling of smallholder crops. The initial steps taken by the Government involved a simple clarification of the legal status of private trade in food crops, as ADMARC has a monopsony only in cotton and tobacco. The clarification, even though it changed nothing legally, was intended to send a signal that the Government was encouraging the development of a parallel private marketing system. The experience with the introduction of private trade raises a number of questions concerning the role of ADMARC, how to achieve price stability, and the impact of pricing policy in a mixed marketing system comprised of the Government and the private sector. These questions are discussed in Chapter IV. C. Smallholder Production and Social Indicators 1.12 Although attempts to foster growth in the economy, including the smallholder agricultural sector, experienced some success prior to 1986, population pressure on supplies of arable land and the lack of productivity increases in the smallhold c sector have caused production and real per capita income to stagnate since the mid-1980s (see Table 1.2). 1.13 Of particular concern has been the performance of the smallholder sub- sector. The average annual increase in the real value of smallholder production between 1980 and 1987 was 1.23 percent, a rate tlat did not keep pace with population growth estimated to be 3.8 percent between 1980 and 1987 (World Bank 1989b). As a result, smallholder production of food crops in per capita terms has been declining since the mid-1970s. This trend has received comparatively little attention, in part, because Malawi has continued to generate a marketed surplus of maize. The marketed surplus and occasional exports of maize have earned Malawi the reputation of attaining national food  self-sufficiency. Apparent food security at the national level, however, has masked widespread food insecuritv at the household level. As will be detailed in Chapter II of this report, over half of the children under five years of age suffer from malnutrition. The rate of infant mortality is such that 15 percent of children die before reachint one year of age (World Bank 1989b). This is the fourth highest rate of infant mortality in the world. With respect to virtually all other basic social indicators, Malawi is behind the other countries in East Africa that it is frequently compared to (see Table 1.3). At the same time, per capita income in Malawi is among the lowest in East Africa. For example, per capita income in Malawi is $150; in Tanzania, $180; in Zambia, $250; and in Kenya, $330 (World Bank 1989b). Table 1.2: Trends in GDP and the Value of Agricultural Output, 1980-89 Real Percentage Percentage Small- Percentage GDP in Change in Total Change in holder Change in Year 1978 MK Real GDP Agric Total Agric. Aaric. SR Apric. 1980 835.7 ---- 284.1 ---- 231.2 ---- 1981 792.3 -5.19 260.7 -8.24 210.6 -8.91 1982 813.9 2.73 277.6 6.48 215.9 2.52 1983 844.6 3.77 289.9 4.43 223.8 3.66 1984 890.5 5.43 306.5 5.73 240.9 7.64 1985 931.4 4.59 308.0 0.49 242.0 0.46 1986 929.1 -0.25 309.9 0.62 244.5 1.03 1987 952.8 2.55 320.4 3.39 250.3 2.37 1988 984.0 3.27 331.9 3.59 256.8 2.60 1989 1024.8 4.15 349.4 5.27 264.3 2.92 Sources Malawi Government, Economic Reports, Various Years 1.14 The failure of smallholder production, particularly smallholder food output, to keep pace with population increases stems from: (i) the serious land shortage faced by the majority of smallholders; (ii) the low technology agriculture practiced by most smallholders; and (iii) weaknesses in the supply of agricultural services. The average per capita availability of arable land in 1985 was 0.64 hectares and is estimated to decline to 0.26 hectares by the year 2000. In addition, at present less than a quarter of smallholders apply fertilizer to anS crop and the adoption rate for high yielding varieties of maize has never exceeded 10 percent. The low adoption rate for improved varieties of maize is, in turn, related to several other factors. These will be examined in detail in Chapter IV.  -6- Table 1.3: Basic Social Indicators for Kenya. Malawi and Tanzania Kenya Malawi Tanzania ---------------------------------------------- Life Expectancy (in years) 1965 45 39 43 1986 54 46 52 Infant Mortality Rate 1965 112 199 138 ('000) 1986 91 156 110 Child Death Rate 1965 25 55 29 ('000) 1985 16 35 22 Population per Physician 1965 12.8 46.9 21.7 ('000) 1981 10.2 53.0 19.8 Percent Enrollment in 1965 54' 442 32Z Primary School 1984 97? 62Z 872 Percent Encollment in 1965 42 22 22 Secondary School 1984 192 42 3Z -----------------------------------------------------------------------__--- Source: World Bank, World Development Report. 1986-87. D. Organization of Report 1.15 The situation in Malawi presents an especially difficult one for donors and the Government. Although Halawi has a good record of implementing policy reforms, evidence of broad based, sustainable growth is hard to find. Of particular cencern is the evident decline in per capita food availability and real per capita income. Responding to these concerns, this report will examine the problem of food insecurity in the broader context of an agricu,tural strategy for Malawi. Although the issues of food security and smallholder agricultural production are linked, as will be shown in this report, simply increasing food availability is not sufficient for solving the malnutrition problems that characterize the population in Halawi. Therefore, in order to understand the nature of food insecurity in Malawi, it is necessary to begin with an examination of the nature and causes of malnutrition and then examine the links with food availability. With this in mind, this report begins--in Chapters 1I and 111--with an examination of the evidence, characteristics, and causes of malnutrition and food insecurity in Malawi. Chapter IV analyzes the relevant policies affecting food security, including input and output pricing policy and the role of the marketing board. Chapter V examines the government institutions that are designated to deal with issues related to food security. Chapter VI presents a strategy for improving food security along with the policy options available to Government and donors.  -7- II. FOOD AVAILABILITY AND MALNUTRITION IN MALAWI 2.01 Given the critical importance of food security for economic growth, it is especially important to understand the nature, extent, and causes of malnutrition, which is the primary direct manifestation of food insecurity. Therefore, this chapter examines the existing evidence from Malawi on: (i) trends in aggregate food production, (ii) household food availability, and (iii) the prevalence and causes of malnutrition. 2.02 The results that emerge from an examination of data on malnutrition and food availability are difficult to fit into a single comprehensive hypothesis about causes of malnutrition and food insecurity. It is clear, however, that the pattern of malnutrition that is observed is not caused by one or two simple factors, e.g., population pressure on land. Instead, malnutrition is the product of a complicated interaction of social, cultural, and economic factors. For example, although there is evidence of a link between holding size and the prevalence of malnutrition in households, the beneficial effects of a household gaining access to more land can be overwhelmed by the gender of the household head, the education level of the mother, the types of crops grown, and/or the amount of the staple crop that is sold. Of particular interest is the preliminary evidence that indicates that male control of cash crop income and female responsibility for food can work to the detriment of household food security when tobacco is the cash crop that is being produced. Such a wide range of factors indicate that policy interventions intended to improve household food security and thereby reduce t)-e prevalence of malnutrition will need to focus on a broader range of topics than a simple increase in food availability. A. Trends in Aggregate Food Production 2.03 In Malawi, national food self-sufficiency has been assessed in terms of the ability of the agricultural sector to produce a marketed surplus of maize for sale to ADMARC. Implicitly this approach assumed that food security would be achieved if maize were available to meet the effective demand for the commodity. It was the economy's success in generating a marketed surplus which fulfilled the effective demand of food deficit households--and even allowed occasional exports of maize--that earned Malawi the reputation of having achieved food security through national food self-sufficiency (see Table 2.1). 2.04 The goal of generating marketable maize surpluses has figured prominently in determining the official producer price for maize. Increases in the announced ADMARC producer price for maize were often made in reaction to declines in ADMARC purchases of maize during the previous season. (For example, the increases in producer prices for maize for the 1981182 and 1988/89 crop years were made in response to declines in ADMARC's purchases of maize in the previous year). Halawi's success in producing marketed surpluses of maize in most years can be seen in a comparison of food self-sufficiency ratios for it and other countries (see Table 2.2). The mean value of Malawi's  - 8 - food self-sufficiency ratio is higher than either Kenya's or Tanzania's and is characterized by greater stability, as indicated by a smaller coefficient of variation.3/ Table 2.1t Estimated Maize Production and ADMARC's Net Sales and Net Exports Crop Estimated ADMARC Year Production Purchases Sales -'__________-- 000 MTT- 1980/81 1,237 137 73 1981182 1,244 246 190 1982/83 1,369 245 86 1983/84 1,398 297 164 1984/85 1,355 272 286 1985/86 1,295 111 130 1986/87 1,218 113 N.A. 1987188 1,427 59 N.A. 1988/89 1,520 122 N.A. --------------------------------------------- Source: Christiansen and Southworth 1988 2.05 Although the marketed volume of maize in most years has been sufficient to meet the effective demand for maize, it has not been enough to achieve food self-sufficiency in the commonly used sense of the expression, i.e., domestic production of sufficient quantities of food to meet the nutritional requirements of the population. With the present balance of population between age groups and gender and 75 percent of the national calorie requirement from maize, an average of 180 kilogram per person per year is needed to meet WH0 estimates of calorie needs. This level of consumption in turn requires field production of 219 kilograms per capita per year. Since 1980/81, the amount of maize produced and consumed has fluctuated between 136 and 198 kilograms per person per year, leaving a substantial deficit even at the national level. To meet 75 percent of the WH0 estimate of calorie needs of the population recorded in the 1987 census would require the consumption of 1.427 million tons of maize and a production of 1.74 million tons. The latter is some 350,000 tons above the peak harvest of 1983/84 (World Bank 1989a, pp.5-6). 2.06 In addition to concerns about the adequacy of national maize supplies, there are concerns about the long-ternm trends in food availability. Although considerable emphasis has been paid to periods in which per capita domestic production of maize has declined, e.g., 1984-87, maize production is less significant than trends in the apparent utilization of maize and, more 3. The food self-sufficiency ratio is the proportion of total domestic consumption of a commodity that is produced domestically.  Table 2.2: Net Cereal ImDorts and Self-Sufficiency Ratios For Kenya. Malawi. and Tanzania. 1961-86 -- -Kenya--- -- -Malawi--- ---Tanzania--- Net Net Net Imports S.S. Imports S.S. Imports S.S. Year ('000 HT) Ratio ('000 MT) Ratio (t000 MT) Ratio 1961 69.9 0.96 -7.8 1.01 110.9 0.90 1962 -17.2 1.01 -4.7 1.01 129.4 0.89 1963 -114.1 1.06 -3.7 1.00 79.2 0.95 1964 -51.7 1.03 -5.5 1.01 6.2 0.99 1965 57.2 0.97 16.2 0.98 61.7 0.95 1966 140.5 0.94 32.4 1.03 66.6 0.95 1967 -174.5 1.08 80.9 1.06 52.1 0.96 1968 -327.3 1.14 75.0 1.07 32.4 0.97 1969 -234.4 1.09 32.2 1.03 16.8 0.99 1970 -76.1 1.03 11.1 0.90 36.1 0.98 1971 27.0 0.99 16.5 0.99 14.5 0.99 1972 -10.1 1.00 15.9 1.01 173.5 0.89 1973 -184.1 1.07 26.6 1.02 27.0 0.98 1974 -64.5 1.02 25.2 1.02 430.0 0.77 1975 -35.9 1.01 32.0 0.97 452.7 0.83 1976 -107.3 1.03 38.3 0.97 97.3 0.96 1977 23.8 0.99 14.9 0.99 134.8 0.95 1978 72.3 0.97 4.2 1.00 76.2 0.97 1979 -101.6 1.04 7.3 1.00 -55.9 1.02 1980 386.4 0.85 23.8 0.98 395.1 0.89 1981 143.0 0.94 66.3 0.96 265.9 0.92 1982 264.0 0.92 22.7 0.99 324.5 0.90 1983 35.7 0.99 00.8 1.07 235.6 0.92 1984 508.0 0.77 08.3 1.07 266.3 0.93 1985 244.6 0.93 88.4 1.06 402.0 0.90 1986 -38.7 1.01 58.1 1.04 222.4 0.94 1987 -5.6 1.00 10.6 0.99 73.7 0.98 Source: MADIA database importantly, the apparent utilization of maize plus other food crops.4/ Because either groundnuts or pulses can be easily substituted for the production of maize, trends in the production of a single crop may not accurately reflect total food availability. For example, between 1984 and 1987 per capita maize production is reported to have declined from 204 to 160 kilograms per capita (Malawi Government, Department of Economic Planning and Development 1989, p.6). This trend was evident largely in terms of marketed production and was partially offset by increased production of groundnuts, 4. Apparent utilization is defined as the sum of domestic production and net imports. Food crops include maize, millet/sorghum, paddy/rice. wheat, pulses, cassava, and groundnuts.  - 10 - since ADMARC's purchases of groundnut increased from 9,900 metric tons in 1984 to 44,800 metric tons in 1987. Apart from such seasonal fluctuations in food crop production, chiefly due to changes in the relative prices of cash crops and the cost of inputs, the long-term per capita availability of food as measured in caloric value has displayed a significant downward trend since 1972 (see Table 2.3). 51 Table 2.32 ApDarent Utilization of Maize and Food Croos -I Zs ------- ---------- Fd Crop.--- Dal ly Total Per Capita Total Per Capita P*r Capita Year Ct) (ka) (Mt) (ks) Calories 1961 642,660 188 888,917 238 2,846 1962 670,006 186 866,840 241 2,372 1963 670,806 154 776,069 210 2,664 1964 711,900 187 923,u83 248 2,392 1966 717,671 184 986,60 239 2,859 1996 715,966 178 961,216 287 2,820 1967 924,616 224 1,159,669 262 2,7" 1966 764,286 16 1,404,392 247 2,426 1969 798,871 184 1,646,857 240 2,36t 1970 912,842 188 1,642,988 285 2,867 1971 967,664 217 1,29,969 283 2,793 1972 1,011,261 217 1,384,719 296 2,811 1978 969,822 207 1,846,261 261 2,758 1974 9s9,56 195 1,858,828 265 2,599 1976 826,769 167 1,181,262 225 2,207 1976 899,659 16 1,279,4u8 2a8 2,881 1977 1,65,86 191 1,467,178 265 2,597 1970 1,140,100 201 1,581,924 279 2,782 1979 1,116,800 190 1,520,W62 269 2,544 1no 048,166 156 1,847,898 228 2,179 1961 1,662,066 169 1,486,482 28s 2,808 1982 1,188,164 177 1,557,154 248 2,880 1968 1,019,476 154 1,428,288 216 2,117 1964 668,867 127 1,849,454 197 1,986 1965 976,858 1S8 1,416,973 201 1,969 1986 948,087 186 1,846,481 18s 1,612 Source: FAO and MADIA database. 2.07 The evidence presented here indicates that Malawi has been successful in achieving large marketed surpluses, and therefore in achieving the appearance of national food security, however, these surpluses have only served to meet the effective demand for maize. The volume of national maize production has not met the requirements of the national population as a whole. As will be demonstrated, the failure to meet food requiremonts at the household level has contributed to serious malnutrition problems. 5. These other food crops include millet/sorghum, paddy/rice, wheat, pulses, cassava, and groundnuts.  - 11 - B. Evidence of Household Food Insecurity 2.08 The primary evidence of food insecurity in Malavi and its principal manifestation--malnutrition--comes from two main sources: (il the National Sample Survey of Agriculture (NSSA) conducted in 1980181. and (ii) other surveys of malnutrition. Although there has been a limited amount of research on malnutrition in the adult population (Beckerson 1983), the vast majority of research has concentrated on the population betneen birth and five years of age. The results of these surveys, described in Table 2.4, demonstrate a consensus that the extent and severity of child malnutrition is a serious problem in Malavi. Table 2.4: Prevalence of Child Malnutrition in Malawi as Estimated from Various Survels Percentage Percentage Survey Geographic Underweight Stunted Source Season Coverage (802 Median) (2 S.D.) 1. NSSA August 1981 National 272 562 March 1982 National 352 562 2. CSR Oct-Dec 1982 6 Districts 24? 522 3. MOH July-Aug 1983 National 5OZ - 4. MOH/UNICEF 5. WHO Aug. 1984 6 Districts in 59S - 3 Regions 6. Kurth/CSR March 1985 Likom Is'land 38? 362 7. Red Cross/CSR May 1985 Mzimba (1 Vil) 262 602 Ntcheu (2 Vil) 321 552 Mulanje (1 Vil) 262 602 ---------------------------------------------------------------------__------_ Sources (Centre for Social Research, University of Malawi 1988a, p. 18) 2.09 To date the most comprehensive and reliable survey to include questions about child malnutrition is the 1980/81 NSSA, which collected data on weight-for-age, height-for-age, and weight-for-height for the under five age group.6/ The main conclusions that emerge from the post harvest enumeration conducted by the NSSA with regard to child malnutrition are presented in Table 2.5. 6. The 1983 Ministry of Health (MOH) survey referred to in Table 2.4 provided national coverage, but is difficult to interpret because of the nature and timing of the sample. The sample consisted of patients at selected clinics in 22 of the country's 24 districts. This sample introduced a significant upward bias in the inferences about the extent of child malnutrition. On the other hand, the survey was conducted between June and August 1983. Because this is the immediate post-harvest period, it introduced a downward bias into the data with respect to the level of malnutrition.  - 12 - Table 2.5: Prevalence of Stuntina, Wasting. and Underweight in Malawl by Age in Immediate Post Harvest Period. 1980 Percent for Children with: ------------------------------------ Child's Age (Mos) Stuntina Wastini Underweimht 0- 2 9.OZ 3- 5 19.3 6 - 11 34.12 2.32 30.0 12 - 23 54.6 3.5 35.2 24 - 35 65.2 1.2 28.4 36 - 47 60.7 0.2 22.8 48 - 60 60.5 0.3 25.3 All Ages 56.4 1.6 27.4 ---------------------------------------------------------- Source: (Centre for Social Research 1988a, p. 19) 2.10 The rate of stunting, which refers to a height that is less than approximately 97 percent of all children in a well-nourished population of the same age, is 56.4 percent. The rate of wasting, which refers to a weight that is less than approximately 97 percent all children in a well-nourished population of the same height, is 1.6 percent. The high rate of stunting indicates that malnutrition in Malawi is of a chronic nature as opposed to occasional severe episodes of malnutrition and/or disease that result in vasting.7/ Because these data were collected in August. which is within one or two months of the harvest depending on the part of the country, they underestimate the level of malnutrition.8/ Therefore, it is not surprising that the NSSA's pre-harvest estimate of malnutrition (defined in terms of weight for age) indicates that a higher percentage of children (34.6 percent) in the under-five group are underweight than did the post harvest estimate (see Table 2.6). 7. In addition to the conclusions regarding stunting and wasting, a combined biological measure of both stunting and wasting, and therefore a more general measure of malnutrition, was collected, i.e., the number of children that are underweight. By this measure over a quarter (27.4 percent) of Malawil's under five population suffers from malnutrition. 8. Assuming that in most parts of the country harvesting begins between May and June, the data collection would have taken place after at least two months of relative food abundance. This period is long enough to allow recovery from malnutrition in the form of wasting that was suffered during the pre-harvest season.  - 13 - Table 2.6s Prevalence of Severe Malnutrition Percent deficient ---NSSA Sample Period--- in Weight for age Post Harvest Pre Harvest All cases 27.4 34.6 70-79.92 21.1 25.3 60-69.92 5.4 7.8 Under 602 0.9 1.5 ----------------------------------------------------------------_- Source: (Mlingi 1989) 2.11 The NSSA results are corroborated by a detailed study of the links between cash cropping, food security, and malnutrition (Peters and Herrera, 1988) in a small area of the Southern Region (Zomba South). The especially valuable part of the Peters and Herrera study is that three surveys were conducted over the course of the agricultural season, corresponding to the immediate post-harvest period (July), the planting period (October), and the hungry season (December/January). These results (see Table 2.7) confirm that large numbers of children, approximately two-thirds, suffer at least moderate malnutrition (as defined by a combination of anthropometric measures) in the post-harvest period and that this percentage increases to nearly three quarters (72.9) immediately prior to the harvest, April and May in the Zomba area. Table 2.7s Prevalence of Malnutrition in Zomba South Degree of Malnutrition October 1986 February 1987 July 1987 Percent of Population Surveyed Normal 34.3 27.1 32.1 Mild 50.9 51.4 49.7 Moderate 14.5 20.4 17.6 Severe 0.3 1.1 0.6 -------------------------------------------------------- Sourcet (Peters and Herrera 1988, Table 6.2) 2.12 The evidence of child malnutrition observed in these studies is corroborated by preliminary survey work done for the Malawi Maternal and Child Nutrition (MMCN) Study in three districts in the northern part of Malawi (Nkhata Bay, Rumphi, and Mzimba). In brief, the results indicate that, on average, the population is significantly below the norm in terms of both weight-for-age and height-for-age. For example, at the age ef 60 months the average child surveyed was 3.0 standard deviation below the norm for height- for-age (Centre for Social Research 1988b).  - 14 - 2.13 Although the level and extent of malnutrition indicated by these data is serious and disturbing at an absolute level, in fairness they need to be compared with other countries in the region. A limited comparison with other countries indicates that even in relative terms malnutrition is especially serious in Malawi. F_: axample, in Kenya the tate of malnutrition during the 1978179 pre-harvest period was 25.6 percent. compared to 34.6 in Malawi two years later (based on NSSA results). In Cameroon, the comparable malnutrition rate was 21.1 percent in 1978. The 56.4 percent stunting rate for Malawi based on the 1980/81 NSSA data is also considerably higher than the 34.1 percent recorded in Kenya for 1978/79 (Centre for Social Research 1908a). In at least approximate terms, this correlates with per capita income patterns. In Malawi, per capita income was $160 in 1987 (one of the lowest in Africa), whereas in Kenya it was $330 and in Cameroon it was $970 (World Bank 1989b). Malnutrition Amoni Children by Age 2.14 Data describing the pattern of malnutrition among children under five are valuable for designing and targeting interventions. Several studies have demonstrated age-specific patterns of malnutrition among children under five in Malawi. Again, NSSA data on malnutrition by age categories are the most comprehensive in terms of national coverage (see Table 2.5). The observed pattern shows that stunting rates increased with age. The peak stunting rate is 65.2 percent among the 29- to 35- month age group, whereas the underweight and wasting rates peaked during the 12- to 23- month age period with 3.5 percent wasting and 35.2 percent underweight. The high stunting rate among the 24-to 35- month age group is the result of the high underweight and wasting rates experienced during the developmentally critical 12- to 23- month. period. The wasting and underweight peaks are typically the result of poor weaning practices made worse by prolonged shortages of food and/or infections. These results are consistent with studies from other areas that demonstrate that the 12-to 24- month age group is the most nutritionally vulnerable group in communities suffering from food insecurity. The NSSA data on malnutrition among children are incomplete, however, because length measurements were not taken for the 0- to 5- month age group and therefore stunting ar.d wasting rates for this age group are not available. 2.15 The study by Peters and Herrera (1989) corroborates the results of the NSSA. These authors observed a high prevalence of malnutrition even among children under six months of age during the 'hungry' period, nearly two-thirds of these children suffered mild or moderate malnutrition (see Table 2.8). High rates of malnutrition within this age group indicate a high level of intrauterine growth retardation as well as poor antenatal and neonatal care. The study also reports an acceleration in the rates of stunting as children enter their second year of life, indicating that chronic food shortages affect the sample population and/or poor feeding habits even when and where food is available, during the immediate poat-harvest period. Again this effect is exacerbated during the pre-harvest period (February), when over 80 percent of children between 24 and 42 months are malnourished. 2.16 The prevalence of malnutrition among children in their second year of life is consistent with the results of other studies in Malawi. For example,  - 15 - the Center for Social Researct conducted a field survey between March 23 and May 17, 1987 involving 10 commnities in three districts: Mzimba, Dowa, and Mwanza (see Table 2.9).91 Table 2.8s Prevalence of Malnutrition by Age Group in Zomba South Age in Months Normal Mild Moderate Severe