Malawi - Agricultural sector memorandum : strategy options in the 1990s (Vol. 2 of 2) : Main report

Iepb,Drl No. I 211)5-NMAI Malawi AriculIturai Sector Memioran dum: Strategy Options in the 1 990s (Iii -I\, W'It Vi auns) \'uIun1,.' I1: Md.iii I&pnrl( Mard,.i It, 11195 SOlm lw0ni Afri'r i \ I. )i paI lin.11 Du-i-ei t of the1 World Ban Document of the World Bank  ACRONYMS ADB African Development Bank ADD Agriculture Developmcnt Division ADMARC Agriculture Development and Marketing Corporation ASAC Agricultural Sector Adjustmnent Crcdit ASC Agriculture Sciences Committec ASM Agricultural Sector Memorandum ASP Agriculture Services Project ASSC AgricultLurc Sector Steering Commitcc CEM Country Economic Memorandum CIMMYT International Maize and Wheat Improvcment Center CMA Credit and Marketing Assistant DAET Department of Agriculture Extension and Training DAHI Department of Animal Health and Industry DAP Diammonium phospaic (fertilizer) DAR Departmnnt of Agricultural Rcscarch DEMATT Development of Malawian Traders Trust DEVPOL Malawi Govcrnment Statement of Developmcnt Policy DLV Department of Lands and Valtation DOI Department of Irrigation DREA Department of Research and Environmental Affairs DRIMP DistricL Roads Improvement and Maintenancc Program EC European Community EEC European Economic Community EEST Estate Extension Service Trust EP&D Economic Planning and Development EPA Extension Pianning Area FA Field Assistant FAO Food and Agriculture Organization FSRP Fertilizer Subsidy Removal Program GDP Gross Domestic Product GOM Government of Malawi IDA International Dcvelopment Association IDF Institutional Development Fund IFAD International Fund for Agriculture Development IMF International Monetary Fund LREP Land Resources Evaluation Project M&E Monitoring and Evaluation MFNR Ministry of Forestry and Natural Resources MIWR Ministry of Irrigation and Water Resources MMFSP Malawi Mudzi Financial Services Project MOALD Ministry of Agriculture and Livestock Development MOE Ministry of Education MOU Memorandum of Understanding MRFC Malawi Rural Finance Company NDDF Northern Division Dark Fired (tobacco) NEAP National Environmental Action Plan NEWS National Early Warning System NGO Non-Govemmental Organization NRDP National Rural Development Project NSCM Nationai Seed Company of Malawi ODA Overseas Development Association PIB Professional Interviewing Board PIPA Public Investment Program for Agriculture PSIP Public Sector Investment Program RDP Rural Development Projects RFSP Rural Financial Services Project SACA Smallholder Agricultural Credit Administration SDD-) Southern Division Dark Fired (tobacco) SEDOM Small Enterprise Development Organization of Malawi SFFRFM Smallholder Farmers Fertilizer Revolving Fund of Malawi S5R Strategic Grain Reserves SSA Sub-Saharan Africa TAMA Tobacco Association of Malawi TRF Tea Research Foundation UNDP United Nations Development Program  PREFACE This report is based on the findings of the main Agricultural Sector Memorandum (ASM) mission in March/April 1992. During subsequent missions to Malawi involving the World Bank's assistance program to Malawi, various ASM team members have carried out follow-up work and discussions with Government and donor officials in order to enhance the operational usefulness of the ASM for both Government and the World Bank (as well as to contribute to the design of several IDA- financed investment programs during the ASM period). These discussions have included fruitful reviews of selected draft ASM working papers, the building blocks of the ASM. The IDA missions have been led by R. Anson, also Task Manager for the ASM. ASM team members on behalf of the World Bank include: V. Scarborough, K. Simler, S. Hiwa, G. Donovan, A. Conroy, L. Schwartz, N. Okidegbe, R. Noronha, S. Rietbergen, E. Thigpen, D. Larson, D. Ng'oag'la, D. Fullerton. A. Yates, D. Jansen, 1. Hayes and S. Jaffee. Hallie Eakin contributed editorial assistance and Ella Hornsby/Wendy Wiltshire provided efficient word processing assistance. Useful cross-country experience and constructive suggestions to enhance the contents and presentation of the ASM were provided by P. Hazell, ASM Lead Advisor (formerly World Bank staff, currently with IFPRIp, J. Coates (AF2AE) and K. Saito (AF5AG), Bank staff peer reviewers. Valuable contributions have been made by numerous Goverrment and donor officials during the course of preparing the draft ASM. MOALD's Planning Division and its staff have been the main counterparts. They arranged an Agricultural Strategy Workshop in March 1992 at the time of initiating work on the ASM, and the ASM Strategy Workshop to review the initial draft ASM, held in Lilongwe on December 3, 1993. The ASM findings and strategy options were also presented and discussed at the Malawi Consortium Group meeting held in Paris on December 20-21, 1993. Building on the ASM work and important developments in the sector, in 1994 Government (with consultant/donor assistance) carried out several studies which contributed to key policy decisions and actions taken in 1994, including: a smallholder credit repayment study; a fertilizer policy study; and a review of the tobacco quota system. The relevant findings and recommendations of these studies are incorporated in this report. Building on the ASM and other recent sector studies carried out by ADB, USAID and EC, in 1994, MOALD prepared a consolidated national Agricultural and Livestock Development Strategy and Action Plan, under the leadership of Mr. A. Gomani, Chief Planning Officer of MOALD, supported by a Task Force led by G. Jere (Deputy Chief Planning Officer). This draft document was discussed at a Government/donor/private sector Agricultural and Livestock Strategy Symposium, held on November 28-30, 1994. Currently, Government is fmalizing its Strategy Document and Action Plan as a reference document for implementation and periodic updating. This final ASM incorporates the relevant comments and suggestions provided by Government in the discussion held on July 12, 1994, and relevant points arising from the Agricultural and Livestock Strategy Symposium held in late 1994. While the data collection and analysis were carried out previoulsy during 1992 and 1993, the ASM findings and recommendations take into account developments up to February 1995. The ASM also complements the valuable contributions being made by other donor agencies, especially the African Development Bank-financed Malawi Agriculural Sector Study, and recent studies financed by USAID, EEC, UNDP, UKIODA and FAO.  MALAWI AGRICULTURAL SECTOR MEMORANDUM TABLE OF CONTENTS Page No. VOLUMEJ I: OVERVIEW VOLUME II: MAIN REPORT 1. INTRODUCTION AND BACKGROUND ............. ............. 1 A. Introduction ........... ..................... 1 B. Objectives of the Study ................................ I C. Main Audience and Participatory Approach to Preparation .... ........ 2 D. Context of Study and Linkage with Other Related Studies ............ 2 E. Organization of the Report . ................................. 3 F. Country Overview . ..................................... 4 G. Macroeconomic Envirornent ........ ..................... 4 2. THE AGRICULTURAL SECTOR, POVERTY, AND THE NON-FARM ECONOMY .10 A. Overview .. 10 B. Resource Base and Agricultural Potential .. 10 C. Structure of the Agricultural Sector ..13 L The Smallholder Subsector ........................... 13 2. The Estate Subsector . .............................. 16 D. Poverty and the Human Resource Base in Malawi . . 16 E. Agriculture's Linkages with the Non-Fanr Sector ....... .......... 19 F. Government Objectives and Strategies .. 22 G. Institutional Arrangements ................................ 23 3. PERFORMANCE OF AGRICULTURAL SECTOR 1980-1994 .... ........ 26 A. Overview ..................................... 26 1. Growth: Sources and Constraints ........ ............... 27 B. Structural Changes andi Performance: 1980-1994 ................. 28 1. The Estate Subsector . ............................. 29 2. The Smallholder Sector . ............................. 31 C. The Implications of the 1980s for Growth and Poverty Alleviation ....... 39 1. Income Effects of Estate Expansion .......... .... 39 2. Income Effects of Smallholder Burley and Hybrid Maize Production . 41 3. Multiplier Effects ................................ 42 4. Limits to Growth and Adverse Effects of the 1980s ........... 42  4. THE SMALLHOLDER SUBSECTOR ............................ 44 A. Overview .44 B. Land and Natural Resources .44 C. Agricultural Inputs .50 D. Pricing and Marketing .65 E. Technology Generation and Transfer .82 F. Rural Finance .91 G. Targeted Interventions .................................. 95 H. Evaluation of Policy and Technology Options .98 S. THE ESTATE SUBSECTOR ..107 A. Constraints .107 B. Strategy Options .112 6. CROSS-SECTORAL CONSTRAINTS AND STRATEGY OPTIONS .. 116 A. Agricultural Diversification ......................... . .116 B. Human Resources .136 C. Rural Infrastructure .138 D. Public Sector Expenditure and Management .140 7. CONCLUSIONS AND RECOMMENDED OPTIONS .150 A. Overview .150 B. Summary of Principal Constraints .152 C. Recommended Strategies and Options .157 TEXT TABLES 2.1 Land Use in Malawi .11 2.2 Relative Importance of Smallholder & Estate Subsectors: Selected Indicators for 1992 .................................... 13 2.3 Typology of Smallholders on Malawi .............................. 14 2.4 Percentage of Maize Produced by Smallholders with Varying Farm Sizes ... .... 15 2.5 Typology of Estates ......................................... 17 3.1 Annual Agricultural GDP Growth Rates (1973-1994) .................... 26 3.2 Leasehold Estate Expansion 1970-1993 ............................. 29 3.3 Nominal Protection Coefficients on Smallholder Export Crop Production ... ..... 32 3.4 Real ADMARC Producer Prices .35 4.1 Higher-Yielding Input Sales to the Smallholding SubSector ................ 50 4.2 Impact of Subsidy Removal on Profitability of Fertilized Maize Production ...... 54 4.3 Indicative Impact of Increasing Technical Efficiency of Fertilized Maize Production . 64 4.4 Ratio of ADMARC Producer Prices to Maize Producer Price; .... .......... 66 4.5 ADMARC Purciases as a Share of Production ........................ 70 ii  4.6 Projected Annual Per Capita Growth Rates in Smaliholder Subsector Under Five Policy Scenarios ................................. 102 4.7 Selected Variables from Aggregation of Linear Progranmming Models for Base Year and Three Future Policy Scenarios ..106 6.1 Agriculture Subsectors And the Agriculture Diversification Project . .118 6.2 Indicative Diversification Cormmodities and Key Characteristics, Market Prospects and Likely Participants ..119 6.3 Summary Indicators for Activities Analyzed (by Activity) . .122 6.4 Summary Indicators for Activities Analyzed (by Sector) . .124 6.5 Tobacco Exports Relative to Total Exports Under Different Indicative Scenarios . . . 134 7.1 Binding Constraints: Smallholder Subsector .......................... 153 7.2 Agriculture Strategy Options ..................... 16G TEXT BOXES 1.1 Malawi's Terms of Trade Trends. 9 2.1 Productivity, Nutrition, and Malawi's Labar Force of the Future .18 2.2 Farm-Nonfarm Linkages: Some Cross-Country Comparisons .21 4.1 Natural Resource Management and Sustainable Agriculture in the NEAP and the FPR 47 4.2 The World Burley Tobacco Market and Malawi's Prospects .74 4.3 Burley Tobacco Reform Options .75 4.4 Hybrid Flint Maize Story .83 4.5 Fertilizer-Maize Recommendations: Current Practice and Untapped Potential for improved Efficiency 86 6.1 Diversification: Insights from Selected African Experiences .117 STATISfICAL ANNEX: TABLES AND FIGURES/DIAGRAMS F-we 1. Household Characteristics and Maize Production/Consumption Relationships ..... 1 2. Incremental Return to Nitrogen Use on Maize ........................ 2 1. Smallholder Hectarage of Local Maize, Hybrid Maize and Groundnuts ('000). 3 2. Average Smallholder Yields (based on MOALD crop production estimates) ...... 4 Tables 1. Urban and Rural Population Growth, 1966-1987 ....................... 5 2. Key Macroeconomic Indicators, 1987-95 ............................ 6 3. GDP by Sector at 1978 Factor Cost, 1978-1990 ....................... 7 4. Land Resources, Customary Land and Population by Region ............... 8 5. Agroclimatic Suitability for Rainfed Production. 9 6. Volume of Agricultural Exports .10 7. Value and Average Price of Agricultural Exports ...................... 10 8. Summary Data on Value of Agricultural Exports ....................... 11 iii  9. Aggregated Household Production Levels by Holding Size Calculated from the 1987188 Annual Survey of Agriculture ............................. 12 10. Cropping Pattern by Holding Sizc 1987/88 ASA ....................... 13 11. Smallholder Crop Hectarage Estimates ............................. 14 12. Smallholder Cropping Mix, 1992/93 (Chart) ......................... 15 13. Smallholder Crop Production Estimates ............................. 16 14. Annual and Three-Year Average Smallholder Yields for Selected Crops ... ..... 17 15. Cropping Mix by Estate Size Categories ............................ 17 16. Estate Cropping Mix (Chart) ................................. 18 17. Estate Crop Exports as a Share of Total Malawi Exports .................. 19 18. Human Resource Indicators, 1990 . ............................... 20 19. Social Indicators of Development, 1990 .21 20. Labor Force, Functional Distribution of Income and Income per Worker by Sector . 22 21. Average Budget Shares and Estimated Marginal Budget Shares and Expenditure Elasticities .................................. 23 22.(a) Agriculture GDP in 1978 Constant Terms ........................... 24 22.(b) Agricultural GDP (Smallholder and Estate Subsector Shares) .25 23. Burley Tobacco Prices ................................. 26 24. Export Crop Production from the Estate Subsector, 1974/75-1987/88 .... ...... 27 25. Changes in Land Tenure, 1964-1988 . .......................... .... 28 26. Wage and Salary Employment in the Estate Subsector, 1969-1992 .... ........ 29 27. Ratio of ADMARC Producer prices to Maize Producer Prices .............. 30 28. Nominal and Real Smaliholder Producer Prices ........................ 31 29. ADMARC Maize Trading Account .31 30. ADMARC Profit and Loss Data ................................. 31 31. Fertilizer Sales to Smallholder Farmers ............................. 32 32. Real Fertilizer Prices ................................. 33 33. Estimate of Gross Income Effects of Selling Burley to Estates and to Auction Floors 34 34. Estimated Incremental Income Effect of "Burley Boom", 1980-92 .... ......... 34 35. Estimated Average and Total Tenant Incomes ......................... 35 36. Average Smallholder and Tenant Gross Margins (1989) .................. 35 37. Indicative Average Real Income for Tenants, Rural Wage Laborers and Smallholders 36 38. Indices of Tenant Profits, Minimum Rural Wages and Smallholder profitability ... . 36 39. Gross Margins on Maize Production at ADMARC Prices (trial yields) ... ...... 37 40. Gross Margins on Maize Production at ADMARC Prices (survey yields) . ....... 37 41. Real Gross Margins in Maize Production (trial yield assumptions) ............ 37 42. Real Gross Margins in Maize Production (survey yield assumptions) .... ...... 37 43. Fertilizer vs. Seed Subsidy: Cost-Effectiveness ....................... 38 44. Indigenous Fruit Trees Found on Farms in Lilongwe and Mchinji Districts and Their Reported Uses ................................. 39 45. SACA: Seasonal Credit Uptake and Participation 1982/83-1990/91 .... ........ 40 46. Nominal and Real (1980) Commercial Bank Interest Rates ................. 41 47. Cross Country Comparison of Burley Tobacco Yields .................... 41 48. Licenced Weight of Burley Tobacco by Category ...................... 42 49. The Costs and Returns to Agricultural Production on Different Types of Tobacco Enterprise ................................. 43 50. Regional Market prospects .................. 44 51. Malawi Indicative Scenarios for Horticulture Exports .................... 45 52. Medium Tenn Public Sector Investment Program for the MCA .............. 46 iv  MAIN WORKING PAPERS (AVAILABLE ON REQUEST) 1. Sources of Growth in Malawi: Past Trends and Future Prospects (by K. Simler, Consultant) 2. Prospects for Expanded Burley Production in Malawi (by E. Thigpen and D. Larson, World Bank) 3. Poverty Profile of Rural Households in Malawi: A Summary of Recent Findings (by K. Simler and A. Quisumbing, Consultant and World Bank, respectively) 4. Economic Reform and Agricultural Strategy (by G. Donovan, World Bank) 5. Household Expenditure Behavior and Farm-Non-Farm Linkages in Rural Malawi (by K. Simler, Consultant) 6. Inputs Sector: Fertilizer and Seeds (by A. Conroy, Consultant) 7. Agricultural Pricing and Marketing Issues (by V. Scarborough, World Bank) 8. Agricultural Diversification: Issues and Strategy Options (by D. Jansen and I. Hayes, with inputs from L. Schwartz, Consultants, and MOA Diversification Working Group) 9. Public Expenditure Review of the Agricultural Sector: Issues and Options (by S. Hiwa) 10. Coping Strategies for Smallholder Households and Scope for Targeted Interventions to Improve Food Security (by K. Simler, Consultantj 11. Available Technology for Malawian Smallholders (by S. Carr, Consultant) 12. Agricultural Market Liberalization and Technological Innovation in Malawi: Modeling Responses and Outcomes in the Smallholder Subsector (by K. Simler, Consultant) OTHER WORKING PAPERS (AVAILABLE ON REQUEST) 1. Land Quality and Strategy Implications (by A. Yates, World Bank) 2. Women in Agriculture: Issues and Strategies (by L. Schwartz, Consultant) 3. Agricultural Research (by A. Yates, World Bank) 4. Agricultural Extension: Issues and Strategies (by L. Schwartz, Consultant) 5. Land Issues and Strategies (by R. Noronha, World Bank) 6. Estate Sector Issues and Strategy Options: Case Studies of Burley Estates in Malawi (by D. Ng'ong'la, Bunda College and Consultant) 7. Estate Sector Issues and Strategy Options (by L. Schwartz, Consultant) 8. Natural Resource and Environmental Management Issues and Strategies (by S. Rietbergen, World Bank) REFERENCES MAP: IBRD No. 24575 V  CHAPIER 1 INTRODUCTION AND BACKGROUND A. INTRODUCTION 1.1 Over the past 14 years, the Government of Malawi (GOM), in collaboration with donor agencies, introduced an ambitious range of agricultural policy reforms and investment projects. As a result, the rural economy is beginning to be transformed. The Agricultural Sector Memorandum (ASM) strives to build on these efforts. It looks at recent reforms and performance with a view to determining what is needed to stimulate further sustainable economic growth and reduce poverty. It takes a strategic approach and focuses on identifying key constraints to the realization of these two objectives. In doing so, it asks the following questions: (a) What are the key features of Malawi's economic structure that will inform the updating of an agricultural strategy? (b) How did the agricultural sector perfonn during the 1980s and what is its future role in accelerating growth and reducing poverty? (c) What are the main factors underlying Malawi's agricultural performance since 1980, including structural changes, policies and public investment? (d) What are the main constraints to sustained agricultural growth and poverty reduction? (e) What are the main policy, institutional and investment options that could help remove these constraints? (t) What are the likely outcomes of implementing such strategy options? B. OBJECTIVES OF THE STUDY 1.2 The object of this study is to assist the GOM in identifying the constraints to increased and sustainable agricultural growth and rural poverty reduction and to define various options for dealing with these constraints. In turn, the aim is to help the Government formulate, in close collaboration with participating donors, an updated agricultural strategy for Malawi that will facilitate and accelerate a transfornation of the agricultural sector and the economy. Such a strategy will strive to promote rapid and widespread growth in incomes and reduce poverty in a manner that is environmentally sustainable, while increasing the diversity of agricultural production. To achieve this, land and labor productivity for all farmers must be increased, more higher value crops must be cultivated, commodities must be diversified away from maize and tobacco, and production and agro-processing activities must become more efficient and profitable. The recent droughts (1992193 and 1993/94 and possible drought in 1994/95) also highlight the need to promote more drought-resistant food varieties and drought risk management strategies. 1  C. MAIN AUDIENCE AND PARTICIPATORY APPROACH TO PREPARATION 1.3 The ASM's main audience is comprised of Government involved in the agricultural sector (especially the Ministries of Agriculture and Livestock Development, Forestry and Natural Resources, Irrigation and Water Development, Research and Environment, Lands, Economic Planning and Development and Treasury) and donors. Other parties which may also be interested include other sectoral Ministries involved in rural areas, NGOs, banks and entrepreneurs involved in Malawi's agricultural sector. In preparing the ASM, emphasis was placed on working with Government officials and donors, to ensure that the most vital issues and options are examined. The IDA team worked especially closely with the Ministry of Agriculture and Livestock Development (MOALD) counterparts in its Planning Division. During the initial ASM mission in March 1992, IDA and MOALD convened a workshop on agricultural strategy to review the main issues and options. This workshop was attended by senior officials from MOALD (headquarters staff and the Agriculture Development Division (ADD) program managers), Finance, Economic Plaraing and Development (EP&D) and by representatives from the main donor agencies in Malawi. The workshop served as a forum for the exchange of views and cnabled a consensus to be formed in many areas. In addition, various draft working papers, the building blocks for the ASM, were discussed with GOM and donors in Lilongwe in November 1992 and May 1993. In early December 1993, MOALD convened an ASM/agricultural strategy workshop attended by the various Government ministries and donors to review and discuss the draft findings and reconmmendations. At the Consortium Group (CG) meeting in Paris (December 20 and 21, 1993), there was a special session on agricultural strategy options, which was based on the ASM recommendations and outcome of the workshop in early December 1993. D. CONTEXT OF STUDY AND LINKAGE WITH RELATED STUDIES 1.4 Over the past five years, ffie Government, with support from various donors, has implemented significant agricultural reforms and investments through the Agricultural Sector Adjustment Credit (ASAC), complemented by the USAID-financed Agricultural Sector Adjustment Program (ASAP, and recent-launched ASAP II) and numerous investment projects within the framework of the National Rural Development Program (NRDP). The Governnent and donors have also conducted various studies involving the sector.' When the ASM was conceived in late 1991, the MOALD began to In mid-1991, UNDP stared to prepare the UNDP Fifth Country Program, which included various backgroLnd studies on die agricultural and non-agricultural sectors as an input to the proposed invcstmne prograns. In latc 1991. WFAD-fiuded consultants carried out a country strategy mission to help fonnulate a framework for future IFAD assistance to Malawis agricultural sector. The main output of this latter mission was a draft report which emphasizes a smaliholder-based strategy. In late 1992, the African Development Bank (ADB)-flnded consultants (CODA and Partners) initiated a Malawi Agricultural Scctor Study (MASS). which is providing a comprehensive inventory and assessment of the agricultural sector, MASS aunepts to emphasize a pipeline of invesuTent projects. ADB consultants prepared a draft report in November 1993. and then finalized it in March 1993. USAID has carried out recendy various studies examining specific aspects of the agricultural sector to support the implementation of the ASAP and the design of the follow-up proposed ASAP 1 (especially the progress of the smaliholder burley pmgram, nmarket liberalization involving produce and ipts, and agricularal diversification). In late 1992. a World Bank and Government team initiated a Forcstry Policy Rcview assessing the various issues and options involving the natural resource sector. This draft report is currendy under preparation. In mid-1993. in anticipation of resuning development assistance to Malawi. the EC-funded consultants initiated an agricultural study to help set the priorities for future EC assistance. In mid-1993. Governuent started to prepare a National Environmental Action Plan (NEAP). which is expected to provide the framework for a national Environmental Support Pmgram. Govenment completed the NEAP in June. 1994. In August 1993, a World Bank ten initiated work on a Malawv Country Economic Memorandum (CEM) which is addressing somc of the broader macro-economic and multi-sectoral 2  prepare, with IDA assistance, two major sectoral investment programs-the Agricultural Services Project (ASP) and the Rural Financial Services Project (RFSP). These two have also provided the framework for two IFAD-funded projects--the ASP Food Security Project and the Mudzi Financial Services Project. It was decided that an assessment of recent reforms and a consolidation of findings from numerous studies would be useful in the preparation of these projects. Concerted efforts have therefore been made to coordinate with other relevant studies to ensure complementarity. 1.5 In mid-1994 the MOALD, using a Task Force approach, formulated an Agricultural and Livestock Development Strategy and Action Plan. The newly elected Government, under the leadership of the MOALD, convened a national agricultural and livestock strategy symposium on November 28-30, 1994 to present and discuss their landmark document with donors/NGOs, the private sector (including representatives of the smallholder and estate subsectors) and representatives of various political parties. The main goal was to develop a general consensus on the most important constraints to increased agricultural growth and poverty alleviation and the most appropriate strategies and supporting actions for lifting these. The new Govermnent's genuine commitment to addressing Malawi's pervase poverty was reflected by the Government's willingness to addrece frontally key structural issues. Based on a large degree of consensus on the document and the successful outcome of the Symposium, MOALD is now finalizing the document as a reference document for implementation. In addition, it would provide a basis for the formulation of a Malawi Agricultural and Livestock Investment Program (MALSIP), which Government intends to use as a framework for focusing its operational strategies and coordinating multi-donor assistance in the agricultural and livestock sector. The ASM was used by the Task Force as an input to a Malawian-led process, which culminated in the Strategy Symposium and initial conceptualization of the MALSIP. The final ASM supports the main thrust of MOALD's Strategy Document and incorporates the relevant strategy points arising from the Strategy Symposium. 1.6 In sumnary, this sectoral study oudines a broad strategic framework that focuses on key constraints and policies along with institutional and investment options for overcoming these constraints. It has provided the analytical underpinning for the four projects described above which are being financed by the ADB, WFAD and IDA, and has provided inputs to GOM's Strategy and Action Plan and Strategy Symposium. Finally, the ASM places the analysis of the agricultural sector within the framework of the macroeconomic and multi-sectoral assessment being undertaken for the Malawi Country Economic Memorandum (currently being finalized). E. ORGANIZATION OF THE REPORT 1.7 The structure of the report has been designed to reflect a comprehensive approach to addressing the underlying constraints of the agricultural sector. Chapters 1 and 2 provide background information on the agricultural sector and poverty, and how these are both linked to macroeconomic growth and productivity in the non-farm sector. Chapter 3 details the policy and performance issues in the agricultural sector, paving the way for the next chapters which elaborate on the current constraints to growth and poverty alleviation, and strategies to eliminate them. Chapters 4, 5 and 6 constins and srategy options. In addition, Govcnmem has carried out various studies addressing specific issues. Ibe ASM lea has coordinated and exchanged daft working papers with the above- mentioned teams to minimize duplication of work and ensare complementariy among the various studies. 3  describe strategies for the smallholder and estate subsectors as well as those that apply to both (on agricultural diversification. human resource and infrastructure development). Chapter 7 summarizes the recommendations, consolidating the main points from the previous chapters. 1.8 The detailed analyses underlying this study are contained in 20 working papers prepared by the ASM team in collaboration with MOALD. These working papers are listed in the Table of Contents, and are available on request. Because these working papers were prepared in 1992 and 1993, the data presented in the ASM refers to the latest available at that time. Nevertheless, the final ASM (Volumes I and Volume II) reflects recent and significant macro-economic and sectoral policy and strategy developments occurring through early 1995, but emphasize the medium term strategy imnplications for the agricultural sector. F. COUNTRY OVERVIEW 1.9 Malawi, a small, land-locked country in Africa, bordered by fanzania, Mozambique and Z:ambia, is one of the poorest countries in the world. The economy depends on agriculture and production and is heavily concentrated on maize and tobacco; crop yields remain below potential and agricultural production and marketing are inefficient. Governmrent policies and investments with respect to the sector have varied over the three decades since independence and have often undermined progress. Evidence suggests that recent reforms may somewhat transforn the rural economy, but constraints to growdt and poverty reduction remain. This indicates the need for an updated agricultural strategy which addresses the constraints and links the sector to national economic and social development. 1.10 Due to the country's land-locked position and narrow resource base, the economy is extremely vulnerable to external shocks. Social indicators and the human resource base have not improved over the past decade and compare unfavorably with those of other Sub-Saharan countries. The population is estimated at 9.1 million (1992), and growing at an annual rate of about 3.6% (see Table 1 for population distribution and growth rates). The country has one of the highest population densities in Africa, with 96 inhabitants per square kilometer of land and 176 inhabitants per square kilometer of arable land. Also, civil war in Mozambique has driven more than a million refugees into Malawi, placing an additional burden on its environrmental and financial resources. 1.11 Notwithstanding its discouraging social and economic indicators, Malawi has the potential to achieve sustainable annual growth rates of 4%-5% by deepening and broadening the macroeconomic and sectoral reforms initiated in Llte 1980s. G. MACROECONOMIC ENVIRONMENT Backglround to Policy Reform 1.12 During the first 15 years after independence in 1964, the economy exhibited impressive growth rates, with real GDP rising at an average annual rate of 5.8% between 1965 and 1980 and real per capita income growing by 3 % per year. This growti was largely driven by the expansion of estate agriculture. Over two decades, Malawi's economy was open and world market prices for its major exports (tobacco, tea, coffee, and cotton) were relatively high, real GDP more than doubled, 4  and real per capita income grew by 3% per year. In the mid-1970s, however, falling world prices for the country's exports (especially for tobacco and tea), rising oil prices and the disruption of transport networks through Mozambique led to a rapid decline in Malawi's terms of trade, for example, a fall of 28% between 1978 and 1981. This situation was intensified by drought in 1979/80, the consequent need to import maize, the maturation of external debt and rising interest rates. These factors led to an economic crisis, manifest in negative GDP growth rates in 1980 and 1981; similarly, during the late 1970s, the slump in growth rates was accompanied by a sharp deterioration in macroeconomic balances. Current account deficits and budget deficits increased to nearly 25 % (in 1979) and 16% (in 1981) of GDP respectively (see Table 2 for further details on macroeconomic indicators). The 1980s and Structural Adjustment 1.13 In 1981, in an attempt to restore macroeconomic stability and growth, Malawi embarked on a package of structural reforms, supported over the decade by successive IMF stand-by operations and IDA structural and sector adjustment loans.2 Between 1982 and 1985, the economy grew at a rate of 4.1%, a positive trade balance was achieved, and the current account and budget deficits were reduceC. However, in 1986, further economic setbacks were experienced following an increase in external transport costs, a large influx of refugees from Mozambique and a continued deterioration in the terms of trade. ThiLs led to further adjustment measures based on the strategy of "growth through poverty reduction" that aimed at removing constraints to sustainable growth, while simultaneously addressing the roots of pervasive poverty. The attempt to liberalize the economy was comprised of reforms to redress the policy bias against smallholders, removing constraints on private sector participation in all areas and liberalizing external sector policies. 1.14 In the first half of the adjustment period, while devaluations were being inplemented, the trade regime became more restrictive as exchange controls on imnports were introduced to ration the limited foreign exchange available. However, between 1988 and 1991, under the Industrial and Trade Policy Adjustment Program, imports were liberalized. Prior approval by the Reserve Bank for use of foreign exchange for imports was eliminated in early 1994; at the same time, the exchange rate was liberalized further under a new, largely market based, system. Under this new system, the exchange rate is determined by a system of Dutch auction that takes place every week. Moreover, the buying and selling exchange rates offered by the commercial banks during the rest of the week are also free of any restrictions. It is expected that this flexible, new exchange rate system will help to increase supply response in the agriculture sector in the following ways: (a) by increasing the supply and improving the allocation of foreign exchange to the most efficient users, the import of inputs will be facilitaed; (b) by raising or maintaining real prices and the real effective exchange rates, and by enhancing incentives for the agricultural exports and products; and (c) by improving the terms of trade between the agricultural and non-agricultural sectors, given the dominant role of agricultural exports on rural household incomes. 2 The World Bank has fnanced scren adjustment operations in M31awi. duhee Stuctural Adjustennt Loans (SALs) in FY81. 84 and 86. a fertilizcr loan in FY83, an industuy/ntade policy operation in FY88, an Agricultural Sector Adjustment Cmdit (ASAC) in FY90 an die Entrepreneurship Development and Drought Rccovery Program (EDDRP) in 1992. 5  Implications for the Agriculture Sector 1.15 The agriculture sector has been significantly affected by all these changes in macroeconomic policies and trends, and especially the macro-economic instability resulting in 1994 (discussed below). First, exchange rate realignments have maintained the profitability of agricultural exports vis-a-vis nontradable and imported goods, and given the importance of agricultural exports to Malawi's balance of payments position, such competitiveness and incentive structure for producers is vital. Second, the supply response of the sector, with respect to these price changes, has increased with the entry of smallholders into areas that were previously restricted, such as burley tobacco production. Since the beginning of adjustment reforms, nearly 25,000 smaliholders have entered burley production to take advantage of higher export prices. As a result of these changes in pricing, real prices in agriculture were maintained through the period of adjustment reforms, against a backdrop of a series of devaluations. Third, reforms in trade policies have reduced the implicit taxation of agriculture and the discrimination against exports by liberalizing imports and removing impediments to exports. The removal of import restrictions and greater efficiency in the allocation of foreign exchange have helped by increasing the supply of agricultural inputs. As a result, there was a substantial increase in the use, production of fertilizer in the agricultural sector despite the increase in their price relative to produce. Fourth, the increase in private sector participation in the imports and distribution of high yielding agricultural inputs has led to an increase in the supply and use of these inputs, benefitting the overall supply response in the agricultural sector. Fifth, the rationalization of input prices, such as reduced subsidies to fertilizers, that occurred under these policy reforms should have increased the efficiency in the utilization of inputs even while raising the costs of production. 1.16 Several aspects of these macroeconomic policies, however, may have impeded the supply response. While many estates and smallholders will benefit from enhanced profitability of exports, the rise in prices of imported goods has increased the costs for fertilizer, fuel, equipment and spare parts. More broadly, continued efforts to reduce the size of the budget deficit in recent years had adverse effects on the agricultural sector. To the extent such efforts have squeezed aggregate demand, or led to decreased subsidies for agricultural inputs like fertilizer and reduced public expenditure on the supply of vital public inputs such as extension services, these will negatively effect the supply of this sector. Moreover, given that the price of major crops lke maize has been controlled, these increases in input prices have led to a greater implicit taxation of the agricultural sector. In recent years, Government has been taking steps to liberalize the pricing and marketing system to help enhance producer incentives. In response to the sharp devaluation of the M Kw in 1994 (discussed below), in early 1995 GOM announced a significant increase in the minimum maize producer price for the upcoming harvest to restore producer incentives. However, there are concerns about the impact of this increase on the Government budget, inflation and on the poor. Macroeconomic Performance under Adjustment 1.17 Despite generally good macroeconomic management, evidenced by continued arrangements with the IMF, economic gains have remained unevenly distributed among the population and the supply response of the economy has remained below expectations. On one side, GDP growth rates have climbed to an average of 5% between 1990 and 1992 and fiscal deficits before grants have generally fallen to an average of 6% of GDP (the exception being the drought-affected rate of FY92/93 of 15.7%). However, successive devaluations, including 15% and 22% in March and June 1992, have helped maintain Malawi's external competitiveness. Still, there continue to be substantial areas of concern. The economy, for instance, continues to be subject to huge terms of trade shocks 6  and it is estimated that in dollar terms, these tenns of trade losses since 1991 were around US$200 million. Partly as a result of these declining terms of trade, the current account deficit on the balance- of-payments widened from 5.1% of GDP to above 14% in 1992, before declining to around 9% in 1993 (preliminary estimates). Such deficits have hampered the aggregate supply response by precipitating severe shortages of foreign exchange. 1.18 The macroeconomic instability and uncertainty during different phases of the adjustment period also were reflected in the drop in savings and investment. The national savings, which had reached 20% of GDP in 1979, plummeted to below 5% in 1992. Correspondingly, gross investment's share in GDP dropped from around 25% in 1980 to around 12% and 15% in 1986 and 1987, respectively. But, in keeping with the trend in many other adjusting countries, gross investment in Malawi recovered in recent years to around 19% in 1992.2 The distribution of aggregate investment, between public and private, has however varied over the adjustment period. Whereas up to the late 1980s the share of private investment accelerated upwards, it substantially tapered off in recent years. To the extent that investment in agriculture takes place mostly in the private sector, the sharp fall in the share of private investment in recent years will have adverse implications for the supply response in the future. 1.19 Inflation increased significantly in the first half of the 1980s, driven by a rapidly increasing money supply,4 as well as repeated devaluations. The average rate of inflation was abouL 10.8% per year between 1975 and 1982, whereas between 1983 and 1988, it was 18.7%. Inflation reached 22% in 1992, reflecting in large part the imnpact of two substantial devaluations. In August 1993, rediscount rates were increased from 20% to 23%, bringing interest rates to positive levels, curbing excess demand for domestic credit, and facilitating quicker repatriation of export revenues. However, financial markets are narrow and noncompetitive, and fimancial institutions have been able to profit without having to seek out new clients or channels of business. As a result, access to financial services is limited, especially for small enterprises and the poor. Although it is impossible to determine the extent to which rural farm and non-farm enterprises are constrained by a lack of credit, it is clear that with the development of productive new technologies and infrastructure and new opportunities created by policy changps, there wili be increased demand for credit. Thus, widening the financial services is an important task for the Govermnent. Recent Developments and Economic Outlook 1.20 Malawi's economy contracted by about 8% in 1992, primarily due to the drought and withdrawal of non-humanitarian aid by the donors at the May 1992 Consultadve Group meeting. The recent strains on the economy have resulted in a deterioration of macroeconomic indicators in 1992, with the 1992/93 fiscal deficit, before grants, rising to 15.7% of GDP (or 10% excluding drought- related operations) while the current-account deficit increased to 14.7% of GDP in FY92/93. Given the sharp reduction in balance-of-payments assistance, the budget deficit was financed by a large increase in bank credit to the Government which, combined with a concurrent substantial increase in TMhe prelininary estintes for 1993 show again a furtier drop in the share of irnestment to GDP to around 13%; this sharp drop was presumably caused by the severe shortage of foreign exchange and the continuing climate of political uncertainty. 4 A major reason for the expsion of money supply has been the rapid increase in public sector credit to finance large fiscal deficits. Ihc Govenmcnts share of total domestic assets rosc frm 28% in 1980 tO a peak of 60% in 1987. Subsequently, this declined m between 25%-33%. 7  MALAWI -- TERMS-OF-TRADE INDEX 105 75 815 Excludes Transport Cost Includes Transport Cost 65. 60 t 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 1992 1993 Diagram 1 credit to the private sector and a tight aggregate supply situation, created substantial pressure on prices and external reserves. Thus, inflation rose to an average of 22% in 1993 while external reserves declined from the equivalent of about 3 months' imnports at the end of 1991 to about one month at the end of 1992, with no improvement in 1993. 1.21 Despite confronting the worst economic setbacks since independence, Malawi managed to carry on with its adjustnent program. Economic recovery in 1993 was very good, with real growth estimated to reach nearly 11% from the low 1992 base. This is largely due to an excellent agricultural crop in 1992193, a tightening of the fiscal and monetary stance, and the positive response from bilateral donors and the EC to the outcome of the referendum on multiparty democracy. The maize harvest in 1993 was about 2 million MT, compared to the record crop of 1.6 million MT in 1991 and 660,000 MT in 1992. Largely as a consequence of another drought in 1993/94, real GDP growth rate dropped significantly in 1994 (to about -10.6%). This poor performance was also exacerbated by the collapse of the smallholder credit system, although the drought was the dominant cause of the poor harvest (of only 831,000 MTs of white maize). Moreover, given further deteriorating terms-of-trade, preliminary balance-of-payments projections also predict continuing large imbalances. Box 1.1 and Diagram 1 highlights Malawi's terms of trade trends and excessive dependency on tobacco exports. Another recent adverse development is that the share of aggregate investment in GDP has decreased by about 50%, from 19% in 1990 to a projected 10.3% in 1993. This also reflects a likely decreased investment trend in the agricultural sector, although data is too aggregated to determine the extent of the decrease involving the agricultural sector. 1.22 The single most important development since Malawi's independence in 1964 - the country's recent move to a multiparty democracy and the election of a new Government; this provides an opportunity for a thorough update of the country's assistance strategy. The effects of transition are not limited to political governance but also have wider consequences for economic management and growth prospects for one of the poorest countries in the world. Over the next few years, dependence 8  Box 1.1 Malawi: Terms of Trade Trends Malawi's export base, almost exclusively centered around tobacco (on average accounting for 70 percent of all cxport revcnues in the past four years), is one of the most concentrated in the world. In view of declining world market prices for primary commodities in general and for tobacco in particular, Malawi has been subject to tremendous terns-of-trade shocks over the past two years (Figure 1). In 1992 and 1993. thcse losses were equivalent to 4.8 and 6.7 percent of real gross national income, respectively; in dollar terms, the terms of trade loss over the past two years amounts to somc US$200 million. This terms-of-trade shock is comparable to the oil crises experienced by industrialized countries in the early seventies. Malawi' s medium-term balance-of- payment prospects will continue to be largely determined by its terms-of-trade. Over the next three years, unit values for Malawi's exports are expected to improve only moderately (at about 2 percent in US$ terms) because of the unfavorable oudook for world market prices of tobacco. on externa! aid will continue in order to sustain the adjustmnent process. While export volume increased by about 13% in 1993 (mainly attributable to tobacco), per unit values declined by a similar percentage. Given the bumper crop of burley tobacco, not only in Malawi but also in competing countries, auction prices are considerably lower than in previous years. The 1993 current account deficit is estimated at 8.7% of GDP. Subject to continued good progress with the political reforms, recent commitments by donors are likely to help meet these external financing requirements, at least until about mid-1994. This donor assistance will provide Malawi some welcome "breathing space" to focus on imnplementing the needed structural reforms at the macro and sectoral levels. 1.23 Macro-Economic Crisis in 1994 and Some Implications: When the new Government took office in May 1994, it inherited several daunting and immediate challenges. Fiscal and monetary discipline suffered in the months immediately preceding the May elections. As a consequence, the economic and financial stabilization efforts that started in early 1994 were not sustained. Instead, excessive borrowing by the Government from the banking system in the first two months of the fiscal year and significant exchange rate depreciation led to substantial inflationary pressures. These reduced the confidence of the private sector in the stability of the economy, increased uncertainties about instability of the relatively liberal economic environment, and exerted substantial further downward pressures on the value of the Kwacha. These developments were significantly worsened by the drought in 1994, which contributed largely to a significant drop in agricultural and economic growth (about -30% and -10.6%, respectively). These recent developments pose a short term challenge to and underscore the urgency of implementing the agricultural strategies outlined in this report. Restoring agricultural growth and addressing the economy's structural constraints will be a pre-requisite to achieving sustainable macro-economic growth. These recent developments also highlight the need for careful coordination and sequencing of macro and sectoral policies. 9  CHAPTER 2 THE AGRICULTURE SECTOR, POVERTY AND THE NON-FARM ECONOMY A. OVERVIEW 2.1 Agriculture is the backbone of the Malawi economy. It supports about 85% of the population. accounts for 35% of GDP, involves about 80% of the labor force and accounts for 90% of foreign exchange earnings (1991). The other key sectors provide the following contributions to GDP (1989): government services 15%. manufacturing 12% and distribution 11.5% (see Table 3 for historical details). 2.2 Given Malawi's predominantly agricultural economy, the driving forces of transformation in the 1990s will be rapid, widespread and sustainable growth in the agricultural sector, curtailment of population growth and improvement in human resources. Through the demand and supply linkages between the non-farm and farm sectors, agricultural growth will stimulate the rest of the economy, enabling it to absorb the excess labor of the farm sector. This chapter provides an analysis of the relationship between poverty, agriculture and the non-farm economy. B. RESOURCE BASE AND AGRICULTURAL POTENTIAL Land 2.3 Malawi's land area is 9.4 million hectares. En 1993, it was estimated that about 5.7 million hectares (or 60%) are suitable for agriculture, of which 4.55 million hectares are in the customary sector (or an average of 2.5 ha per household) and 0.55 million hectares are in the estate sector (or an average of about 21 ha per estate). It is estimated that about 4.6 million hectares are cultivated, including fallow land, leaving about 1.1 million hectares of land suitable for additional cultivation. Due to the absence of a reliable system for collecting and monitoring land use data, there is uncertainty and some debate over the accuracy of these figures. Table 2.1 summarizes three sources of land use data in Malawi. (Para. 4.17 provides further discussion of steps being taken to address this data problem and Table 4 provides additional details on available land use data). 2.4 Findings from the land suitability assessment carried out under the Land Resources and Environment Project (LREP) indicate that practically all land deemed suitable for cultivation in the farning sectors is presently cultivated. Most of the land still available for cultivation (0.6 of the 1.1 million hectares) is currently public land used for forest or game reserves. With little new land left to cultivate and a rapidly growing population, land scarcity has become an issue. In the smaliholder subsector, land scarcity varies considerably between regions. In the northern region, where rural population densities average 36/km,2 smallholders still maintain fallow periods, allowing soils to regenerate after cultivation. In the southern region, where population densities for all land reach 133/1kn2 and considerable amounts of customary land have been allocated to the estate subsector, smallholders have intensified agriculture by eliminating fallow periods, inter-cropping and relay cropping, and increasing the use of chemical fertilizer. In addition, pressure on the land has led to the expansion of cultivation into increasingly marginal areas, some of which are erosion-prone steep land 10  unsuitable for permanent agriculture. In the estate subsector, it has been estimated that about 16% of the area occupied is suitable for agriculture, but not cultivated.5 Table 2.1: Land Use in Malawi (millions of hectares) Area Suitable Area Available Land Classification Total Area for Agriculturc Arca Cultivated' for Expansion Public land 1.80 0.60 0.00 0.60 Estate land 1.20 0.55 0.30 0.25 Customary land 5.50 4.55 4.30 0.25 Other land 0.80 0.00 0.00 0.00 Total 9.30 5.70 4.60 1.10 Sources: The Estate Extension Services Trust: LREP as reported in Eschweiler H. (1993), 'Land Use in Malawi', Forestry Sector Review working paper: and Mkandawire. R., S. Jaffee and S. Bertoli (1990) 'Beyond Dualism: The Changing Face of the Leasehold Estate Subsector of Malawi' Institute for Development Anthropology, NY. It should be noted that recent work carried out by USAID in connection with the proposed ASAP II suggests that the customary sector may have additional unused agricultural land, larger than shown here, due to labor and other constraints to intensifying thcir production. Includes short-term fallow. Agro-Ecology 2.5 Malawi has a climate that only allows for one cropping season a year. It is endowed with moderately fertile soils, varied agro-ecological conditions, relatively reliable rainfall patterns and abundant water resources which permit a wide range of agricultural crops to be grown. Nearly all cultivation (about 99% in terms of land area) is rainfed, which makes Malawi subject to periodic droughts and fluctuations in production levels. Four main agro-ecological areas can be identified: (a) The S'hire Valley, which is hot and semi-arid, lies at 30-40 m above sea levei and has a mean annual rainfall of 500-750 mm; (b) The Lakeshore Plains, which lie at 450-600 m, are wam-to-hot and have a mean annual rainfall of 750-1000 mm. The soils in the above two zones are mostly deep, calcimophic alluvials and colluvials, with some hydromorphic and vertisols in low-lying areas; (c) The medium-altitude plateau, which covers about 75 % of the country, has an elevation of 800-1350 m.; mild to warm temperatures and a mean annual rainfall of 750-1000 mm. The higher parts of this zone are dominated by deep, well-drained latosols, while the lower parts are characterized by poorly drained hydromorphic soils (Table 5); (d) The high-altitude plateaux are between 1350 m and 3000 m in elevation with rolling hills to steep terrain; mild to cool temperatures and an average annual rainfall of between 1000 and 2000 mm. Soils are mainly lithosols and highly leached latosols. 5 Mkandawire e al. (1990). Op. C*.. 11  2.6 Within these four zones there is also considerable micro-variation. This highly diverse tropical and sub-tropical natural resource base offers many opportunities in terms of the range of crops that can be grown and farming systems that can be followed. These opportunities include a wide range of food and commercial field crops, orchard fruits and nuts, a high-value horticultural and floricultural crops, and trees grown on farm land that provide fuelwood and other products. This diversity also offers opportunities for increasing cropping intensity, particularly in the higher rainfall areas, through development of short duration varieties that fit into current farming systems to allow additional catch and inter-crops. Development of the latent irrigation potential will also open new opportunities for diversification and the development of more intensive high value cropping systems at an acceptable risk level. 2.7 Of the estimated 290,000 ha of potentially irrigable land, only about 25,000 ha is irrigated and most of this is in the estate subsector. Irrigation is almost entirely for sugar production in the Lower Shire Valley. The other important irrigated crops are rice (7,500 ha) and vegetables (3,500 ha). About 40% of Malawi's irrigable potential is found in lakeshore regions and in the Lower Shire. Expanded lakeshore irrigation could be based on lift or tubewell water extraction, lending itself to smallholder managed and maintained schemes, whereas in the Lower Shire Valley larger surface gravity flow schemes would probably dominate. A number of studies have quantified available water resources and identified potential irrigation locations. Clearly the development of latent irrigation potential should initially be focused in the high potential lakeshore and Lower Shire areas. However, irrigation development strategies that address issues related to salinity and the maintenance of ground water tables, and water-user participation in scheme development and maintenance, have not yet emerged. Composition of Production 2.8 Crop production accounts for about 91 %, livestock for 7%, and fisheries and forestry for less than 1 % of total estimated value of agricultural output. Exports primarily consist of tobacco (76% of total export revenue in 1991), tea (8%), sugar (6%) and coffee (2%) (see Tables 6 to 8). The estate sector contributes about 90% of the total value of exports, and this share increased over the 1980s. Maize, the dominant crop and staple food, occupies about 80% of cultivated land in the smallholder sector (see Tables 9 to 13 for a summnary of smallholder cropping patterns). The maize crop mainly consists of low-yielding local varieties, although in recent years the adoption of higher yielding hybrids has increased. Other important crops include groundnuts, cassava, cotton, rice, sorghum, millet, legumes and tree crops. Livestock includes cattle (Malawi Zebu), goats, sheep, pigs, chickens and rabbits. As smallholders generate surplus food crops, they also devote an increased share of resources to livestock activities, particularly smallstock. Over tine, livestock also will become an increasingly important contributor to future growth and diversification, given that increased incomes will lead to increased domestic consumption of livestock products (especially meat and milk, especially since Malawi's per capita consumption of these products is very low in absolute terms, relative to other developing countries). The small portion of agricultural GDP due to fisheries, like livestock, also underestirnates the importance of this subsector as a major source of protein for the population. Already, fish contributes about 70% cf animal protein consumption. Increased production of fish in Lake Malawi will need to balance the need to conserve the rich variety of fish resources, which in some areas are becoming overexploited. 12  C. STRUCTURE OF THE AGRICULTURAL SECTOR 2.9 Malawi's agricultural sector is composed of two main groups of farms: smallholders and estates. These groups are differentiated in terms of a variety of socioeconomic conditions as well as by regulations conceming production, marketing, pricing and land tenure. In Malawi, estates are defined as those farms which occupy leasehold or freehold land. In order to register as an estate, it is necessary to have a minimum of 10 ha. Thus the term estate in this context encompasses a range of enterprises ranging from farms of several hundred hectares to those with just ten. Smallholders have customary use-rights to the land they cultivate. Government policies restricting the rights of each subsector to grow tobacco, for example, have had a strong influence on the performance of each (see paras. 3.4, 3.28, 4.60, 4.83, 5.13 for more details). The relative contribution of the smallholder and estate subsectors is illustrated in Table 2.2. However, it should be noted that there is considerable variation within each of the two subsectors and that differences between them are becoming increasingly blurred. Thus, in formulating and implementing strategies, it is more useful to differentiate groups of farmers within each subsector. Figure 1 (Statistical Annex) shows the total population categorized according to six groups, based on characteristics such as farm size, maize deficit/surplus relationship to market, and cash incomes. A basis for classifying the different types of smallholders and estates and the strategy implications are discussed below and subsequent sections of this report. The Smallholder Subsector Table 2.2: Relative Importance of Smnalholder and Estate Subsectors: Selected Indicators for 1992 Subsector Contribution Contribution Contribution Agricultural Ha Cultivable Farmers to GDP to Ag. GDP to Exports Employment (million) ('000) Smaluholder 25% 65% 10% 90% 4.55 1,800 Estate 9% 35% 90% 10% 1.Z' 26* * 1993 estimate. Maybe as high as 30,000 (early 1994). 2.10 An estimated 1.8 million smallholder families farm under customary tenure, cultivating about 4.55 million hectares and producing about 80% of Malawi's food (mostly for subsistence) and 10% of its exports. About 25% of the smallholder households cultivate less than 0.5 ha, 55% less than 1 ha; 31% between 1 ha and 2 ha, and 14% more than 2 ha. Size of holding is inversely correlated with levels of food insecurity and poverty. 2.11 Typology of Smallholders There are significant differences between smaliholders in terms of their production technologies, constraints and capacity to respond to opportunities, and these differences highlight the need to tailor strategies and technologies accordingly. Table 2.3 provides a typology of smallholders according to cultivated area, region, average yields and the extent of their food surpluses/deficits.6 The three to four groups include: (a) Larger smallholders (about 14%) with enough land to produce a surplus; 6 For linber details. see Sinler, K. (1994) "Agriclazal Market IlberAlization and Tedmological lnnovation in Malawi: Modeling Responses and Outcomes in the Smnalholder Subsector'. ASM working paper. 13  (b) Smallholders (31 %) with enough land to produce enough to meet their food requirements and who could potentially become surplus producers; and (c) Chronically or sporadically food-deficit households (about 55%) for whom improved policies and technologies could enhance food security, although the majority would still need to supplement their incomes off-farm. 2.12 The MOALD's Annual Survey of Agriculture (1984/85) also found that farmers with more than 3 ha, representing only 4% of smallholders. used 35% of all the fertilizer sold to smaliholders, while those cultivating less than 1 ha, 55% of the smallholder population, consumed less than 5%- The 45 % of smallholding households which produce a food surplus generally have more than 1 ha and in aggregate produce about 75% of total maize, with average yields of 2400 kg/ha. This group could produce larger and more diversified surpluses and become prosperous small commercial farmers, if appropriate policies and supporting actions were implemented. Table 2.3: Typology of Smallholders in Malawi Household characteristics Area cultivated (ha) per household by ADD, North to South Karonga Mzuzu Kasungu LiIDngwe SnUmm Liwonde Blantyre Ngabu Smnllhiders with enough land to > 1.50 > 1.00 > 100 > 1.50 > 2.00 > 2.00 > 1.50 > 3.00 produce a surplus (15%)' SmalRbolders with enough land to 0.75-1.50 0.75-1.00 0.50-1.00 1.00-1.50 1.00-2.00 1.25-2.00 0.75-1.50 1.75-3.00 produce ondy own food requirements (22f)' Food-deficit households (63%)4: Sporadically deficit 0.5D-0.75 0.25-0.75 0.500.50 0.50-1100 0.50-1.00 0.50-1.25 0.50-1.50 1.00-1.75 Chronically derucit < 0.50 < 0.25 < 0.50 < 050 < 0.50 < 0.50 < 0.50 < 1.0 * Refers to rough estimate of proportion of total rural households. Source: MOALD 1987/88 National Sample Survey of Agriculture 2.13 The 55% of smallholders currently with less than I ha contribute about 25% of total maize production. Most of this group do not have access to the cash or credit with which to purchase higher yielding inputs, and produce average maize yields of about 800 kg/ha. These households go hungry and purchase maize for about four months of the year prior to harvest. To do so, they must find work away from their farms. This not only diminishes the labor invested in their own land, often at critical stages in the agricultural cycle when demand for hired labor is at its peak, but also the wages for such employment are low. Thus, strategies to increase food production among this group will enhance household food security and nutritional levels. Table 2.4 illustrates the relative importance of the different smallholder households in contributing to aggregate smallholder maize production. (Table 9 shows aggregate household production levels for food and cash crops by holding size). 2.14 Women and Female-Headed Households. Women comprise about 70% of full-time farmers and have primary responsibility for providing their family's food. About 30% of all rural households are headed by women, with the proportion headed by womeen increasing as farm size diminishes. Of households cultivating between 0.5 ha and 1.0 ha, some 34% are headed by women, whereas 42% of households with less than 0.5 ha are headed by women. Female household heads are either widowed, divorced or unmarried, or are the wives of men working elsewhere. Those in the first three categories are the poorest since they receive no cash remittances. To raise agricultural productivity in Malawi, special efforts must therefore be made to improve the access of female farmers (and especially female heads of households) to agricultural inputs and services. 14  Table 2.4: Percentage of Maize Produced by Smallholders with Varying Farm Sizes Category Small Medimn Large Total Holding Size (ha) 0 < 1 1 < 2 > 2 Number of households ('000) 874 484 203 1,560 Percentage of households 56 31 13 100 Production (% of Total) Local maize 31 40 29 100 Composite & hybrid maize 12 34 54 100 Total maize 28 39 33 100 Source: 1987/88 ASA. 2.15 Off-Own-Farm Employment. About 95 % of the agricultural labor force is employed in the smallholder subsector and most additions to the rapidly growing labor force have been absorbed by smalbholder agriculture to date. Off-own-farm employment is dominated by ganyu (casual, seasonal) labor, which accounts for around 70%-80% of rural labor.' The data available on the returns to such labor suggests that wage rates are highly variable, are not determined with reference to minimum wage legislation and are usually below the rural minimum wage, except during peak labor demand seasons, when they may briefly rise above it. Labor demand within the subsector is highly seasonal, with October to January being the peak months. For example, it has been estimated that although about 70% of smaliholder households hire some labor every year, most do so for very short periods. Only 5% of the total labor days expended in the smallholder subsector are supplied by hired labor.8 This limited demand for labor within the subsector is related to the limited quantity of labor intensive cash croDs grown and low inconmes, as well as the highly seasonal nature of agriculture in a single rainfall, unirrigated enviromnent. It implies a substantial unused seasonal labor reserve, particularly since non-farm rural activities are relatively under-developed in Malawi.' This reserve of labor, together with the land constraint and low incomes in the subsector, means that the smallholder subsector provides other sectors with an elastic supply of labor at very low wages. However, the demand for labor from other sectors, in particular the formal sector, remains weak. Hence the concentration of rural labor in the smallholding subsector. No change in this situation is envisaged in the foreseeable future. Thus agricultural intensification and diversification within the smallholder subsector are critical in increasing employment and income-generating opportunities. 7 Livingston, L. and S. Bose for GOM (1993:44). 'The Labor Market and Wages Policy in Malawi.' a MOALD Agro-Econonic Survey data and Sijm (1990), reviewed in Livingston and Bose (1993:47f8) op cdt.- 9 Reliable information is not availablc in the magnitude of rural Dn-farm employment in terms of its contribution to household incomes and labor absorptionL Various sudies esuimate that it can range from about 40-60% of total housebold incomne; while a large percentage. incomes levels are still very low in absolute tenns, thereby reflecting the acute land shboage and access to ius and tebnology. 15  The Estate Subsector 2.16 There are currently about 26,000 farms in the estate sector. Estates are highly variable in terms of size, degree of mechanization, and other factors. Table 2.5 presents a typology of estates by size and other characteristics. In 1989, about 68% of estates could be classified as small (i.e. less than 30 ha), 24% as medium (between 30 and 100 ha), and less than 8% as large. Nonetheless, the latter category occupied 58% of the estate land area and 68% of small estates accounted for only 20% of this land area. Estates currently occupy about 1.2 rnillion hectares. Of this, it is estimated that about 25% is cultivated (see Table 2.1) and an additional 250,000 ha could be cultivated. Estates produce mostly flue-cured and burley tobacco (about 40% of the estate cropped area) and maize (about 42%). Other crops grown by estates are tea, sugar, coffee and tree nuts (see Table 15 and 16 for cropping patterns by estate size). Estates contribute the major proportion of total exports, for example, about 90% in 1992 (see Table 17). They have easier access to fertilizer and credit and have security of tenure which motivates higher levels of investment and inputs (and thus higher yields) than are found in the smailholder sector. After the smallholder subsector, the estate subsector is the largest source of employment in Malawi, with tea and tobacco estates accounting for about 75% of subsectoral employment. There was a particularly rapid increase in the estate labor force between 1968 and 1977, when estate employment grew at 11% per annum. Since then, this growth has slowed and the share of agriculture in total formal sector employment has remained at around 50%. A In 1990, between 90,000 and 120,000 families cultivated land in the estate subsector as tenants, mostly on burley tobacco estates. About 160,000 permanent laborers were also employed in that year. In total, it is estinated that about 1 million estate workers and their dependants were supported through estate labor." 2.17 A major difference between the smallholders and estate subsectors are the restrictions prescribed by Government policy regarding tobacco production. Until 1990/91, the production of burley and flue-cured tobacco was restricted to the estate subsector primarily through a system of quotas. Since then, smallholders have been allocated a small proportion of the national annual burley production quota, but the majority of burley, and all flue-cured tobacco, continue to be produced by the estate subsector. D. POVERTY AND HUMAN RESOURCE BASE IN MALAWI 2.18 Malawi's pervasive poverty is reflected in its low ranking among African countries with regard to numerous human resource indicators (see Table 18 and 19 for further details). The GOM recognizes that until actions are carried out to significantly improve the quality of life and enhance farmers' capacity to apply improved technologies and respond to opportunities, agricultural growth will continue to be inhibited (Box 2.1 highlights the inter-relationships between productivity and nutritional status in Malawi and other cross-country experiences). It is also recognized that there is an urgent need to reduce the population growth rate, one of the highest in the world, to enable real gains in per capita food consumption and incomes. to Livingston and Bose (1993). op. cit. it Jaffee, 5., 2. Mkandawire and S. Bertoli (1991:3). TMe Migrat Snailhoktdrs: Tenant and Laborer Participation, Remuneration and cial Wdfare within Malawi's Expanding Estate Subsector.' Institute for Devdopment Anthropology. NY. 16  Table 2.5: Typology of Estates Type or estate Size Main Initial Main Type of Major 11 Period Source of Occupation Management Capital Established Finance of Owner Investment Small-scale Graduated 10-30 Late 1980s Prior farm income Farmer Estate owner Ox-cart Smallholders or relative Small Entrepreneurial 10-30 Mid to late Business or Farner or Estate owner Ox-cart Estates 1980s employmcnt shopkeeper or relative baling jack Medium-scale (MS) MS Commercial 30-100 Early to Salary or business Fanner or Owner or Baling Estate mid-1980s civil servant hired jack employee tractor truck MS Undeveloped 30-100 Mid to Salary or business Civil scrvant Hired Baling Estate late 1980s or cmployee jack businessman Large-scale (LS) LS Private Estate 0> 10 Mid to Bank credit Farmer, civil Owner or Tractor late 1970s and salary servant or, hired truck businessman employee ploughs LS Corporate 0> 10 Mid to Bank credit + N.A. Hired Tractor Estate late 1970s ADMARC employee truck ploughs Source: Mkandawire et al (1990). Poverty Profile of Rural Households12 2.19 A recent World Bank study of growth and poverty reduction in Malawi constructed a poverty line based on minimum human nutrition requirements.3 Using the same definition of poverty here, households are defined as poor if they have insufficient income to purchase or produce 200 kg of maize per member. The incidence of poverty, thus defined, is very high in Malawi (about 55% of the total population). About 98% of the poor live in rural areas and most are smallholders (89%), agricultural laborers (wage laborers or tenants, 11% of total poor), and/or female heads of households (in both rural and urban areas). At least two-thirds of the total expenditure of poor households goes to meeting basic food needs, and the core poor may have expenditure elasticities greater than unity. 2.20 The rural poor are generally limited by having only small areas available for cultivation, almost no access to credit and extremely limited use of improved agricultural technologies such as 12 FulEther detais on this subjcwt arc provided by Sinler. K. and A. Quisuzbing (1993), 'Poverty Profile of Rural Households in Malawi: A Summy of Rcet Findings.' ASM Workin Paper. An updated assessmen of poverty is also uxderway: an imerim assessment is provided as part of dth fordtcoming County Econoniic Memoandum, "Progress Towards Povcrty Reduction in Malawi: An Evaluation.- Jamnary 1994, World Bank. 13 Wodd Bank (1990), 'Malawi: Growth Tbrough Poverty Reduction.' Repon No. 8151-MA, Wasington, D.C. 17  Box 2.1: Productivity, Nutrition, and Malawi's Labor Force of the Future Does belter nutrition raise farm productivity? While the common sense answer to that question would appear to bc "Ycs, of coursc," formal documcntation and quantification of the cffects of nutrition on farm productivity havc proven problematic, largely due to methodological difficulties. Several rccent studics, using data collected in Africa and Asia, have taken full account of these methodological issucs and have confirmed the expected rcsult that belter-nourished farm workers arc more productivc than thcir malnourished counterparts.' The studies havc shown that farm productivity, as measured by wages or own-farn output, is positively and significantly related to a number of nutritional indicators, including household caloric availability, individual caloric intakc, wcighl-for-height, and height. The clcar policy implication is that investments to improve health and nutrition, whilc desirablc in their own right, also serve as invesiments to improvc farm productivity and raise household incomes. Recent cvidencc (from an ethnically homogeneous sample) has also shown that attained hcight, which is a cumulativc mcasurc of good dicts and an absence of infecction in early childhood, has a strongcr cffect on adult productivity than short- and medium-term mcasures such as caloric intake and wcight-for-hcight. In Malawi over 60% of children under five years of age are nutritionally stunted, and will not attain their full potential height, nor their full potential productivity. Studies in Malawi have refutcd the notion that Malawians arc genetically short; in fact, children from relatively well-off households, wherc food is plentiful and health carc readily available, grow at rates comparable to thc NCHS/WHO standards.? Extending a similar quality of life to all children in Malawi will not only improve the well-being of the individuals, but will make the country richer as well. I The most analytically rigorous of these studies includc: Strauss, J. (1986), 'Docs Better Nutrition Raisc Farm Productivity'. Journal of Political Economy, Vol. 94, No. 2. pp. 297-320. Deolalikar, A. B. (1988). "Do Health and Nutrition Influence Labor Productivity in Agriculture? Econometric Estinates for Rural South India". Review of Economics and Statistics. Vol. 70Q No. 3. pp. 40K6-3. Sahn. D. E. and H. Aldeman (1988). "The Effects of Human Capita on Wages and the Detenninants of Labor Supply in a Developing Country'. Journal of I.rvelopnmnt Economics. Vol. 29, No. 2. pp. 157-84. Haddad, L. J. and H. E. Bouis (1991). "The Impact of Nutritional Status on Agricultural Productivity: Wage Evidence from dhe Philippines". Oxford Bulletin of Economics and Statistics, Vol. 53, No. 1, pp. 45-68. Higgins. P. and H. Aldceran (1993). 'Labor and Women's Nutrition: A Study of Energy Expenditure, Fertility, and Nutritional Status in Ghana'. Working Paper No. 37. Ithaca, NY: Corncll Food and Nutrition Policy Program. 2 Chimwaza, B. (1982), "Food and Nutrition in Malawi". Unpublished Ph.D. dissenation, Kings College, University of London and GOM (1991). 'Report on a Nutrition Survey of High Income Udban Children in Matawi'. Lilongwe: Food Sccurity and Nutrition Unit, Deparunent of Economic Planning and Development, Office of the Preidem and Cabinet. higher yielding hybrid seed and fertilizer. Poor households are slightly morc likely to participate in off-farm employment than the non-poor, and are much more likely to work at garzyu labor. But, the level of return to their labor is far below that of the non-poor for all types of off-fam employment. Further, they tend to have more members, and thus a higher proportion of dependants. 2.21 Often underemployed for much of the year, they also face severe labor shortages for their own farms during critical stages in the agricultural cycle such as planting and weeding. Labor constraints are further exacerbated by poor access to cooking fuel, safe water supplies, health services and other amenities. This extends the burden on available household labor in two ways. First, more time is required to obtain these household goods and services, which takes away from time available for 18  farming. Second, when not obtainable, health and nutrition suffer, further hindering the ability of the household to adequately perform necessary labor. Moreover, female-headed households are particularly prone to acute seasonal labor shortages, since only one adult is usually present. Where they have spouses working elsewhere who remit a portion of their earnings home, this constraint can be alleviated through hiring labor. However, a large number of females are unmarried, divorced, separated or widows, and do not receive regular remittances; it is this group which faces the most severe seasonal labor constraints. Heads of poor households also have considerably less formal education. E. AGRICULTURE' S LINKAGES WITH THE NON-FARM SECTOR 2.22 Since independence, agricultural strategy has primarily supported estate led development, while providing subsidies for maize production and consumption and public investment in the development of the smallholders sector. However, the latter has not responded to public investments as expected; subsidies have proven too burdensome on the budget to sustain. Moreover, the development of the estate subsector is now considered insufficient to generate growth and alleviate poverty in the rest of the economy. This disillusionment, coupled with severe land constraints facing the majority of smallholders, is leading some Govermnent officials and donors to turn to the non- agriculture sector (especially manufacturing and informal sectors). 2.23 While the non-agricultural sector will become increasingly important over time, five factors suggest that a continuing effort to increase agricultural productivity and incomes is also vital in terms of sequence: * A reported 98% of the poor are already located in the agricultural sector and increases in their productivity and income will therefore directly address poverty objectives; a The non-agriculture sector is currently very small (measured by cun.- ibution to GDP or employment) and industry employment elasticities suggest that very little of the incremental Malawi workforce (about 100,000 per year) will be absorbed by it; therefore, in the short-to- medium tenn. returns to labor will have to rise in agriculture in order to retain workers on the land. Increasing total employment in the non-farm sector is likely to be a slow process, making it important that it be given immediate attention. The non-farm sector will need to employ an increasing share of Malawi's labor force, and this can be achieved by promoting non-farm activities (especially in the informal sector) and raising rural incomes and purchasing power for goods and services provided by the non-farm sector (see Table 20); * Unless agricultural incomes increase for a broader section of the population, there will be little casn to buy the products and services of the non-agricultural sector and limited savings in the economy; * Given the relative size of the agricultural sector, increasing and diversifying agricultural exports and reducing food imports will be stabilizing influence in the economy; * It will be important to continue to invest in sustainable forms of agricultural growth in terms of natural resource utilization and conservation, as well as in an environment of shifting world markets. 19  2.24 To assess the effects of growth in the farn and non-farm sectors, the links between them should be analyzed. Recent research shows that, on average, rural households allocate 62% of their budgets to food (including food produced in household gardens)." One-third of total expenditure is on maize, with an additional 7% for roots and tubers. These high percentages are an indicator of low incomes. The expenditure elasticity of food (all food commodities taken together) is 0.88. Thus, food expenditure for most of the population is likely to increase a good deal with increases in income. 2.25 For the average household, almost all nonfoods are luxury items, except fuel and lighting, food processing, and sundries such as soap. For nonfoods, the highest budget share is allocated to clothing and footwear, which account for 10% of total expenditure on averagc. Overall, nonfood goods have the highest expenditure elasticities. These figures suggest that the absolute levels of poverty of the majority of the population result in a lack of effective demand for goods and services from the non-agricultural sector. (See Table 21 for a summary of the average budget shares, estimated marginal budget shares, and expenditure elasticities). 2.26 A simple two-sector, three-parameter, semi-input-output model based on the above data, suggests that a MK 1.00 increase in agricultural incomes could bc expected to produce an additional, indirect boost to the rural economy of about MK 0.50 (or a multiplier of 1.5). The multiplier is relatively large, because it is dominated by consumption linkages between the rural farm and non-farm sectors, as production linkages between these sectors are small. However, a significant finding of this analysis is that increases in the incomes of small smallholders (say, below 1 ha of cultivated area) have as strong a multiplier effect on the rural economy as a similar increase in the incomes of larger smallholders. Thus, policies can target poor households, so that the immediate impact on agricultural incomes is directed to those who need it most, and the same type of multiplier effect will be achieved for the local economy.' 2.27 The analysis also indicates which rural enterprises are likely to benefit most from increases in demand from improved agricultural income. These include: producers of clothing, footwear, and household durables (including construction); processors of oils and fats; and those that involve transportation services. These non-farm activities should be given special encouragement in any programs to promote rural enterprise. Further, present regulations on these industries should be reviewed to ensure that they are not inhibiting non-farm commerce. 2.28 The locally produced and consumed agricultural commodities which would be most affected by increased rural incomes are generally higher quality food items that have expenditure elasticities greater than one. These conunodities include eggs, oil and fats and pulses. In addition, fruits and vegetables have an estimated expenditure elasticity of approximately one, indicating that expenditure on these items would grow proportionately with income. As these agricultural commnodities are typically higher in value than staple foods, and are sometimes more labor-intensive in production, increased demand for these goods could represent expanded opportunities and incomes for smallholder households. 14 1Ihis amnlysis is based on data from the Food Security and Nutrition Monitoring Survcy. For further deails on famulnon-faxm linkages, see Sirnler, C. (1993a). "Household Expenditure Bchavior and Farm-Nonfarn Linkages in Rural Malawi.' ASM working paper. 15 If it can be shown hat nurket access is a problem even for high income households, the starenis of the preccding paragraph may need to be modified accordirgly. In any event, investments in markefting and transportation infrastrbcure would serve to improvc nmrkct access and cfficiency. accelerating the growdt ard development of both the fann and nonfami sectors of the rural economy. 20  Box 2.2: FanmlNon-Farm Linkages: Some Cross-Coustry Comparisons The same methodology used to study farm/non-farm linkages and agricultural growth multipliers in Malawi has been used to study a number of other countries in Africa and Asia.' A comparison of the findings provides a context for the study of Malawi. In general, the multipliers are higher (on the order of 50%) in Asia than in sub-Saharan Africa. The most prominent reason is that there are more backwards production linkages to agriculture in Asia. Fertilizer, equipment, cement, and building matcrials have been cited as the key domestic production inputs to agriculture in Asia, but Nigeria is probably the only Sub-Saharan country with the resources and internal market to develop a domestic fertilizer industry. Similarly, much of the technological change in Asian agriculture has involved irrigation, generating demand for pumps, piping, repair services, etc., while irrigation potential in sub-Saharan Africa is much more limited. Thus, forward linkages, especially food processing such as flour milling and oil extraction, arc thought to have greater potential for agriculture than backward linkages. Consumption links to non-fann activities are generally equal in Asian and sub-Saharan Africa, although African consumers tend to spcnd more of their average and marginal income on rurally- produced foods. The few cases where consumption patterns were lower in sub-Saharan Africa can be explained by poor transportation networks and greater dispersion of rural towns. Analysis of data from Sierra Leone and Nigeria produced estimated multipliers on the order of 1.5, about the same value as derived for Malawi. As in the Malawi case it is consumption linkrges, rather than inter-industry linkages, that account for the bulk of the growth multiplier. To deterniine whether smallholder strategies or estate strategies for growth would have the largest multiplier effect, a model was designed and applied to various regions in sub-Saharan Africa. It was found that either growth strategy produced multipliers in the range of 1.45 to 1.55. results which assume, for lack of data, that estate owners and smallholders have similar consumption patterns. A study of northern N.igeria that had such expenditure data for the smallholder subsector, indicated that smallholders with relatively large landholdings had higher marginal propensities to spend on rurally produced non-food items. The main policy conclusions drawn from the cross-country analysis that are germane to Malawi include the following: - Raising farm labor productivity is essential both to release labor to non-farm employment and to raise income levels to foster demand for the goods and services produced by the non-farm sector. - Commerce and services, rather than manufacturing, deserve emphasis because they are insulated from urban competition by their very nature, and can improve consumption patterns in the rural economy. - The development of rural towns, supported by physical infrastructure (especially adequate road or rail transport) is a key to providing larger markets and facilitating the operation of enterprises. - Human resource investment is critical, especially as services are expected to be a dynamic growth area. The role of women in food processing and preparation, trading, and many other services is of particular importance. t Haggblade, S. Hazell, P, Brown, J (t989). 'FarmlNon-Farm Linkages in Rural Sub-Sahamn Africa.' World Devdopment Report, 17:1173-1201. 21  2.29 In addition, the multiplier analysis examined the degree to which increased incomes through higher farm profits would lead to increased demand for locally-produced consumer commodities, which would, in turn, increase off-farm earnings and could be invested in agriculture. For example, a recent survey found that purchased agricultural inputs, area cultivated and farm profits were highest amongst female-headed households where husbands had gone to work in South Africa."5 The remittance was clearly the source of finance for inputs, allowing these households to overcome credit and liquidity constraints to satisfy their dtemand. Similar cross-subsidization of household enterprises is common among households that have both on- and off-farm income sources. However, the flow of funds in either direction is usually small due to the low levels of income from either source. F. GOVERNMENT OBJECTIVES AND TENTATIVE STRATEGIES 2.30 The objectives for Malawi's agricultural development are outlined in the Govermnent's Sttement of Development Policies (1987-1996). They are: * Improving and maintaining food self-sufficiency; * Expanding and diversifying agricultural exports, while conserving natural resources; * Raising farm incomes and promoting economic growth; * Improving social weltare. The newly elected Government appears to have embraced these broad objectives with a renewed commitment to make significant progress toward attaining them and in a manner which benefits a large proportion of the population. The Government is also pursuing these objectives within a broad framework of promoting private sector involvement; improving incentives through liberalizing pricing and marketing; increasing producer prices for export crops; and, rehabilitating rural infrastructure. The importance of improving food security and social welfare in the Govermuent's development objectives is demonstrated in the Food Security and Nutrition Policy Statement. '7 This emphasizes growth based on increased agricultural productivity, especially on holdings of less than 1 ha; increasing employment opportumties; improving human resource development; and increasing income transfers to the vulnerable in the short term, pending the achievement of longer-term objectives. MOALD's draft Agricultural and Livestock Development Strategy document outlines four main tdirsts to pursuing the above objectives in a manner which has components designed for all types of farms and households: increase the productivity of food crop production, and widen the range of food crops grown and marketed; smallholders will be given the opportunity to grow burley tobacco (with an indicative target of about 300,000 smallholders over the next 10 years); farmers in general, and large farms and estates in particular will be encouraged to diversify their activities to broaden the base, and increase the output of export and high value-added crops; expanding livestock activities and integrating them into existing farming systems. Government's draft strategy document gives emphasis to key commodities, complemented by sector-wide strategies aimed at providing the necessary policy, legislative and institutional framework, and supported by continuous generation and dissemination of appropriate and profitable technologies, the provision of the necessary inputs and services, and the improvement of human capital and physical infrastructure. 16 Pctcrs. P. and G. Herrexa (1989). 17 Office of the Prcsidet and Cabirct, Dcpautmcnt of Economic Planning and Development (1990). 'Food Stecrity and Nutrition Policy Saemcnt. Supplmemn to the Stament of Devdeopment Policies. 22  G. INSTITUTIONAL ARRANGEMENTS 2.31 The Ministry of Agriculture and Livestock Development (MOALD) plays a major role in formulating agricultural policy and delivery services. It is comprised of four technical departments (Agricultural Research, Agricultural Extension and Training, Animal Health and Industry and, until recently, Irrigation); a food and nutrition unit; a land resources and conservation branch, and smaliholder credit administration. In the current Public Sector Investment Program, (1993/94 to 1997/98) the Ministry is allocated about 13% of the total planned investment budget. 2.32 The agricultural field services of MOALD are organized into eight agricultural development divisions (ADD). These are Karonga and Mzuzu in the north, Kasungu, Lilongwe and Salima in the central region and Liwonde, Blantyre and Ngabu in the southern region. For delivery of services, each ADD is divided into two to five rural development projects (RDP) of which there are 30 in total. Each RDP is divided into extension planning areas (EPA) (total 146) and each EPA is made up of Sections where agricultural field staff demonstrate technologies to smallholders. 2-33 The Ministy of Forestry and Natural Resources (MFNR) includes five departments: Forestry, Fisheries, Geological Survey, Mines, and National Parks and Wildlife. The Forestry Department is the custodian of national forestry policy and executes the requirements of Malawi's Forest Act. The National Parks and Wildlife Department is also responsible for forest reserves. 2.34 The Department of Agricultural Research (DAR) conducts most national research except for that on tobacco, tea and tree nuts, which is undertaken by specialized units. DAR's activities are divided into five programs for cereals, grain legumes, oil seeds and fibers, livestock and pastures, and soils and land husbandry. In addition, it has an adaptive research and a technical services program. 2.35 Responsibility for smallholder agricultural extension rests with the Department of Agricultural Extension and Training (DAET). This includes a land resources and conservation branch which focuses on improving the use of land resources. Extension to the estate subsector is provided by the Estate Extension Service Trust (EEST), the Tea Research Foundation (TRF); the Tobacco Association of Malawi (TAMA) and other commodity specific organizations.'8 2.36 The Department of Animal Health and Industry (DAHI) is responsible for all aspects of animal production and disease control, including livestock marketing and the provision of improved livestock inputs, animal health research and disease investigation. The Department of Irrigation (DOI) is responsible for developing irrigation schemes and advising all other departments on irrigation activities. 2.37 The Agricultural Development and Marketing Corporation (ADMARC) is responsible for distributing fertilizer and hybrid, composite and improved local varieties of seed to smallholders, managing Malawi's strategic reserves of maize, and marketing strategic crops based on floor and ceiling prices. Until 1987, ADMARC had a monopoly on the purchase of all smallholder produce. Agricultural produce markets were then liberalized, but ADMARC continues to play a dominant role, especially in the maize market. The parastatal has also had an effective monopoly over retailing fertilizer and 90% of hybrid and composite seed sales to the smallholding subsector, although the Government also intends to liberalize these markets in 1993/94. 18 The EEST is fuided by EEC and ODA as initial grams, and a cess kvied on all tobacco growers, induding snialholders who do not benefit from EE5rs services. Tbc TRF and TAMA are funded by ihe industries thy service. 23  2.38 The Smalbolder Farmers Fertilizer Revolving Fund of Malawi (SFFRFM) was established in 1983, and is responsible for procuring fertilizers required by smallholders, managing the fertilizer buffer stock and facilitating the distribution of fertilizer provided under commodity aid agreements. SFFRFM used to operate as a separate unit within ADMARC, but was transformed into a trust corporation under Government control in 1988- Until 1990/91, the SFFRFM had a monopoly over fertilizer imports for smallholders. However, in that year, a private firm entered the market, and in 1993 Government adopted a Fertilizer Policy Paper which outlines plans to liberalize fertilizer imports and distribution. The current surplus of fertilizers probably will delay this aspect of the liberalization program. 2.39 The Smallholder Agricultural Credit Administration (SACA), established within MOALD in 1988, provides credit to smallholders. Field operations are managed by the ADDs, with specific staff assigned to credit activities. Extension staff have also been highly involved in credit processing, supervision and recovery. In 1994, SACA was converted into an autonomous Malawi Rural Finance Company (MRFC), which also led to the delinking of extension and credit field activities. Thus, responsibilities for credit will be transferred to credit and marketing assistants (CMAs), which will liberate an estimated 40% of extension agents' total working time. This time can then be spent on extensin!l activities. 2-40 With the establishment of the MRFC, this means that SACA is no longer a Government department; thus. it will have a conunercial orientation, and be able to mobilize savings and enharnce the sustainability of its operations. SACA started operating as such by the end of 1993. SACA/MRFC is the only formal institution providing credit to smallholders (mainly short-term), and it does so through the medium of farmers' clubs, of which 15,000 are operating, involving about 35% of the smallholders- 2.41 The Malawi Mudzi Fund Trust is based on the Grameen Bank of Bangladesh. It was established in 1988 to provide credit to resource-poor individuals for non-farm enterprises. Its lending operations rely on the group approach and joint liability- Experience to date suggests a need to restructure the Mudzi Fund to enable it to lend to the resource poor on a wider scale and to lend for both farming and non-farming activities. With IFAD assistance, Govermnent has designed and initiated a Mudzi Financial Services Project which will support the conversion of the Mudzi Fund into a "window" in the newly created MRFC. 2.42 In 1994 the newly elected Government established a new Ministry of Irrigation and Water Resources (MIWR) to give emphasis to developing Malawi's water resources as an important element of its poverty strategies. MOALD's Department of Irrigation is now transferred to this new Ministry. There is clear recognition of the need for MOALD and the MIWR to coordinate closely in ensuring sustainable and economically viable irrigation development. 2.43 The Private Sector. Unsubsidized fertilizer for the estate sector is suppiied by private companies through direct imports or from a limited number of private local companies (OPTICHEM and Norsk Hydro). Norsk Hydro also sells to smallholders. In fact, supplying smallholders with fertilizer has been becoming attractive to the private sector as subsidy levels are being phased out. Already, prices of calcium ammonium nitrate (CAN) and some tobacco compounds are sold to smallholders at prices (in 1993) below those charged by ADMARC. The recent devaluation of the Kwacha, however, is posing a transitional constraint to private sector participation, as in 1994 private sector import prices were substantially higher than ADMARC prices. 24  2.44 The supply of seeds and agro-chemicals to estates and smallholders originates in the private sector. The National Seed Company of Malawi (NSCM), in which Cargill International has a majority share, is the main supplier of certified hybrid tobacco and maize seed, 90% of which are then distributed by ADMARC, with the remainder being sold through private retailers. A second private firm, Lever Brothers, entered into the supply of hybrid maize seed in 1991192. However, the market remains far from competitive. A nwnber of private companies also supply imported pesticides and herbicides, which Malawi does not produce. Finally, since 1987, the private sector has played a substantial role in the marketing, processing and exporting of agricultural produce and products. 25  CHAPTER 3 PERFORMANCE OF THE AGRICULTURE SECTOR, 1980-1994 A. OVERVIEW 3.1 Agriculture has been an important contributor to economic growth and macroeconomic stability in Malawi. It is the major source of employment in the country, with over 80% of the labor force employed in the smallhoIder subsector and about I 1% in the estate subsector. Since Independence, growth in agricultural GDP has varied, averaging about 5% during the 1970s, 1% during the early 1980s, and 4.1 % during 1987-91. During 1973-94, annual growth in agricultural GDP has averaged about 1.75%. Table 3.1 shows the variability in the growth rates, which also reflects the effects of recent droughts in 1992 and 1994. During the same period the rural population has grown by more than 3% per year, indicating a real and tragic decline in agricultural GDP per capita. 3.2 Throughout the 1970s, 1980s and early 1990s, growth in the estate subsector has been significantly faster than in the smallholder subsector except in four years (1976, 1978, 1984 and 1991). For example, between 1973 and 1994, the estate subsector grew at an annual average rate of 5.56%, while the smallholder subsector grew at only 0.67%. Thus, the share of the estate subsector in total agricultural GDP rose steadily from 13% in 1973 to 30% in 1994. During the past 20 years, about 60% of agricultural growth has been in the estate subsector. Table 3.1 highlights these trends and the variability in agricultural growth rates, and Table 22 provides further details. Smallholder subsectoral growth rates have been low and highly variable, despite Govermment's stated policy emphasis on smallholder development. These subsectoral dynamics, with a rapidly growing estate subsector and stagnant smallholder subsector, are essential to understanding Malawi's agricultural growth constraints and future prospects. Table 3.1: Annual Agricultural GDP Growth Rates (1973-1994)' SECTOR 1973-94 1973-80 19887 1987-94 1987-94 1980-94 Estate 5.6% 8.6% 4.1% 5.3% 7.8% 5.1% Subsector Smallholder 0.7% 4.0% 1.8% -1.0% 4.1% 0.4% Subsector Agricultural 1.8% 4.7% 2.3% 0.7% 5.0% 1.6% Sector Source: World Bank, 'Malawi: Economic Reform and Agricultural Strategy' (with data for 1993 and 1994). * Excludes drought years 1992 and 1994. 26  Growth: Sources and Constraints 3.3 Much of the growth in the estate subsector has been the result of an increase in the area of land being cultivated under leasehold tenure; increased real world market burley tobacco prices between the mid 1970s and 1988; and increased cultivation of this crop. Until 1990f91, only estates were legally able to grow flue-cured and burley tobaccos and this policy has been one of the major causes of the difference between smaliholder and estate subsector performance. 3.4 The Tobacco Quota System. Under the Tobacco Act (1965) and the Special Crops Act (1965), the production of all types of tobacco was restricted to growers who were in possession of a license. Licenses to grow flue-cured and burley tobacco on estates are issued by the Ministry of Agriculture and Livestock Development and a quota stipulating how much an individual enterprise can produce is attached to the license. Growers apply directly to MOALD for licenses and quotas for these two types of tobacco. In the case of Southern Division Dark Fired (SDDF), Northern Division Dark Fired (NDDF), sun-air cured, Oriental tobaccos (and since 1990/91 smallholder burley tobacco), ADMARC is responsible for issuing licenses and quotas. However (with respect to NDDF, SDDF, sun-air cured and Oriental tobaccos), the quota system has become largely irrelevant since production is not meeting trade demand. Similarly, the vigor with which quotas have been applied by Govermnent has varied according the type of tobacco. 3.5 Annual production targets, or national quotas, are determined through consultation between Govermnent, the tobacco trade and tobacco growers' associations like the Tobacco Association of Malawi (TAMA). In actuality, demand on the part of tobacco buyers dominates this process. The allocation of these national quotas to individual growers is less transparent, however, and the criteria by which the MOALD issues such quotas are particularly murky. The enormous sensitivity surrounding the issue of quotas contributes to lack of clarity. 3.6 In the case of burley tobacco, where the effect of production restrictions on prices is probably the greatest, the allocation of quotas between individual growers is the mechanism by which economic rents from the rtsultant higher prices are distributed. Tobacco is bought and sold within the estate subsector and between estates and smiallholders in an attempt to ensure quotas are fulfilled. This trade effectively reallocates econonic rents beyond the quota system and as a result there is not necessarily any relationship between an estate's quota and its actual level of production. However, no formally recognized system of buying and selling quotas exists. 3.7 Estate Expansion. Aside from these legal advantages, the estates' ability to exploit the high world tobacco prices of the 1970s and early 1980s was also dependent on a supply of low cost labor and the availability of commercial bank financing. With increasing land shortages in the smallholder subsector, together with low crop prices, restrictions on the production of high value crops and limited higher yielding technologies, supplies of low cost labor from the smallholding subsector were assured. In addition, political pressure was placed on the commercial banking sector to expand credit supplies to tobacco estates, especially during the 1970s. 3.8 The average price of burley tobacco, in 1980 constant tenns, rose from MK 1.18 per kg in 1980 to z peak of ME 1.58 per kg in 1988 before falling again. In 1992 the average real price realized on the auction floors was MK 1.23 (see Table 23). Likewise, the quantity of burley tobacco sold at the auction floors in Malawi has increased every year since 1980 (except 1984), rising from 16,686 t in 1980 to 99,224 t in 1992 (see Table 24). These three factors (price, quantity of burley 27  and the number of registered estates) are inter-related, because smallholders, cultivating land under customary tenure, were not legally able to grow burley tobacco until 1990/91. Thus in order to obtain a license to grow the crop, many of the larger smaliholding farmers converted their customary land use rights to leasehold tenure, so as to gain access to the relatively high burley tobacco prices of the latter 1970s and 1980s." To be able to do this, it was necessary to have a minimum of 10 ha of land and, in some cases, members of an extended family pooled their land in order to register as an estate. 3.9 As a result of this process, the area of land under leasehold tenure has grown from 79,000 ha in 1970 to about 255,800 ha in 1979 and 759,400 ha in 1989. It is further estimated that in 1993. leasehold estates occupy about 1,148,000 ha with the remaining 52,000 ha in the estate subsector being under freehold tenure (see Table 25 for details).20 Thus between 1979 and 1993 there has been an increase of nearly four and a half times the area of land held under leasehold tenure. Much of the increase in the area under leasehold is the result of transfers of land from the smailholder subsector, since most of the newv estates registered in the 1980s and 1990s are the result of smallholders changing the nature of their land tenure. This process of transferring land (and other resources) from the smallholder subsector to the estate one, together with the monopoly granted to the estates on the production of burley tobacco, is the major cause of the difference between the agricultural growth rates of the two subsectors. 3.10 Agricultural Intensification and Diversification. Conversely, little has been achieved in intensifying agricultural production until very recently and there has been an increased concentration in maize and tobacco production since 1980 (see Tables 11, 12, 13 and 24). Despite heavy public investment in agriculture (in the form of research, extension and marketing services for the smallholder subsector, and the provision of financial services to both subsectors) agricultural growth has been driven largely by the expansion of burley tobacco production and buoyant world burley tobacco markets. There are a number of factors which have inhibited increased productivity and diversification in agriculture. They include increased population pressure on the available cultivable land, which has led to reduced incomes; supply and demand constraints on the increased use of higher yielding technologies; Government pricing policies, which have favored food crop production over that of export crops in the smallholder subsector; world market shifts, which have favored the production of tobacco, in particular that of burley in the estate subsector; and a lack of incentives or ability within the estate subsector to increase productivity and operational efficiency. These constraints will be addressed in more detail in Chapters 4 through 6. B. STRUCTURAL CHANGES AND PERFORMANCE, 1980-1993 3.11 The dualistic structure of the agricultural sector has broken down over the 1980s, as a result of the creation of about 13,000 small estates, which resemble smallholder farms in many respects. The vast majority of new estates formed in the 1980s and early 1990s are less than 30 ha in size and 9 To obtain leasehold tenure of customary land. fanners have to obtain agreement from the local traditional authoaities concerned. before applying to the Department of Lands and Valuation for issuance of a lease. Once the lease has been issued by the Department of Lamds, farners have to apply to the MOALD for a quota to grow and sell burley tobacco. 20 Figures for the 1970 - 1989 period are drawn from Mkandawire et aL (1990) op. cit.. The 1993 figures have been estimated by the EEST. Beth sets of figures are baseu on a review of Ministry of Agriculture records. 28  have been registered as a result of converting customary use rights to leasehold tenure (para. 3.8).21 The following table illustrates these changes, in particular the decline in the average size of estates. Nevertheless, given the differences in the performance of the two subsectors over the 1980s, it is useful to assess what shifts in production have occurred in each. Table 3.2: Leasehold Estate Expansion, 1970-I93 Year No. of New Area Leased Mean Size Total No. of Total L/H Total Mcan LA! Estates (Ha) IH Estates Area (Ha) Size (Ha) To 1970 229 79,000 345 229 79,000 345 1970-1979 876 176,800 202 1,105 2S5.800 231 1980-1989 13,250 503,600 38 14,355 759,400 53 1990-1993 8,645 388.600 45 23,000' 1.148,000 50 Source: 1970-1989 Mkandawire et al. (1990); 1990-1993 EEST esdmatcs based on Ministry of Agriculture files. ' Recent estimates range from 26,000 to 30,000. The Estate Subsector 3.12 As noted earlier, growth in the estate subsector has exceeded that in the smallholder sector over most of the 1960s, 1970s and 1980s. The major changes within the estate subsector which occurred over the 1970s and 1980s and contributed to this growth, include the following: (a) Number of Leasehold Estates. The number of leasehold farms in the estate subsector has risen from an estimated 229 in 1970 to 1,321 in 1980 and to about 23,000 in 199331 Although the number of estates has been increasing throughout the last 20 years, there was a faster increase in the late 1970s; between 1980 and 1982 and between 1984 and 1988, following real increases in world prices for burley tobacco. The vast majonrty of estates which were established in the 1980s are small (less than 30 ha) and concentrate on burley tobacco and maize production. In 1989, it was estimated that 50% of cultivated estate land was planted with burley tobacco and 7,846 esates (or 67% of the total) had burley quotas.? In 1991/92, 19,020 estates had a license to grow burley.? Thus the expansion of the estate subsector, and of burley tobacco production in particular, has continued in the early 1990s. (b) Size of Estates. As a result of the large of number of small new enrants into the subsector the average size of leasehold estates has fallen from 345 ha in 1970, to 207 21 k Is esrmatd by thc EEST ta about 70% of the esta whih regisbe d in the 1980s, md at least 90% of those registered in the early 1990s. ae so-cae graduated smaioNdiags. 22 Fqigres for 1970-1989 are esmated by Mimndawire er al (1990) op. dt and for 1993 are estmated by the EEST. Reen estimaes are as high as 30.C0O estatcs. 23 gMkndawime t al (1990) op. cit. Recent research by the EEST has shown dot die number of estates which have a quota to grow burley tobacco, as recorded by the Tobacco Control Commis, is reLted to the previous year's esdmaes of thc mumber of estaes in Mkaldzwime es als work. 24 Tobacco Control Commissiot 29  ha i;: 1980 and 50 ha in 1993. Further, the average size of all estates (leasehold and freehold) is now 46 ha.' Thus, whilst the number of estates increased ten fold in the 1980s, the area of land under leasehold tenure only increased by a factor of 2.8, from 273,100 ha in 1980 to 759,400 ha in 1989. It is further estimated that about 1,148,000 ha of land is under leasehold tenure in 1993 (i.e. more than a four-fold increase over 1980 levels). The estate subsector as a whole (leasehold and freehold) has likewise expanded nearly four-fold between 1979 and 1993, from an estimated 308,000 ha in 1979, to 741,000 ha in 1989 and 1,200,000 ha in 1993. The slower pace of increase in area, as opposed to number of enterprises, is in large part due to the small size of most new estates, but also to the fact that the area under freehold land tenure has been declining since the latter 1960s. The area of land under freehold tenure was 147,500 ha in 1970, but has declined since then, stabilizing at around 52,000 ha since 1979. (c) Expanson of Estate Employment. With this transfer of land into the estate subsector, there has been a shift in labor resources between the subsectors over the last two decades. Although estimates of the number of tenants and permanent laborers employed in the estate subsector over the decade vary, the available data suggest that there were about 65,000 pernanent employees in the subsector in 1968; about 200,000 in 1978; between 330,000 and 425,000 in 1989 and about 500,000 in 1993 (see Table 26 for further details).6 In addition, one recent review of the labor market in Malawi estimated that employment on estates grew at a rate of 8% per annum during the 1980s, whilst that in the smallholder subsector grew at 2.3 %.2 (d) Increased Tobacco Production. With an increased focus on burley tobacco within the estate sector, tobacco production on estates has increased from 43,000 t in 1980 to over 96,000 t in 1990 and nearly 143,000 t in 1992 (see Table 24). This increased production, in combination with increased relative prices for burley, has meant that the contribution of estate tobacco sales to total export eamings rose from about 30% in 1980 to 45% in 1992.28 Malawi is now more susceptible to shifts in tobacco markets than it was during the late 1970s and early 1980s. 3.13 Summary of Estate Subsector Performance. Despite the growth and increased employment opporuities to which they have contributed, production in the estate sector is characterized by low yields; low rates of return to capital; low rates of labor remuneration; and a heavy dependence on one-crop, tobacco. Thus there appears to be potential for increased growth and poverty reduction in this sector. The constraints to realizing this potential will be addressed in the next chapter. Addressing these constraints in the 1990s is important, because, despite having led agricultural growth for a number of decades, it can be seen from the above analysis that the estate subsector has depended 25 Estatc Extension Scrvices Trust 26 Most burley tobacco in Malawi is grown by estate ants. Este owners provide tenants with land and inputs (fefdlizer, seed, tDols, barn consmiccion mterials) on credit, as well as food rations on credit in most cases. These advances are deducted from tic vajuc of the inants' tobacco when it is sold to the estte. n Livingstone. 1. and S. Bose, for GOM (1993:53). 'The Labor Market and Wages Policy in Malawi'. 28 f snulbolder production is added to hese figures. tobacco accounted for 49% of toal export earings in 1980-2 and 63 % in 1989-90 (Simier. K. (1993) 'Sources of Growth in Malawi: Past Trends and Future Pispects, Agricultural Sector Memorandum Woring Paper). 30  heavily on an expansion of area and buoyant world market tobacco prices and thus its growth is not sustainable. Increases in technical or operational efficiency have contributed very little, if anything, to this growth. Given that world tobacco market prospects are no longer so optimistic and that there is little room for increased area expansion, these sources of growth are unlikely to continue into the next century and it will become important for the estate subsector to become more diversified. The Smallholder Subsector 3.14 A number of important changes within the smallholder subsector also took place in the 1980s. The causes of these changes are more various than those which prompted the expansion of thte estate subsector and include shifts in Government's pricing, marketing and burley tobacco production policies; technological developments and public extension efforts to disseminate new technologies; increased land pressure; and increased supplies of credit and fertilizer. Two distinct periods can be identified with respect to these changes - the first being 1980/81-1986/87 and the second from 1986/87 to the present. Each era of change will be treated separately below, but one major transformation cuts across these historical periods - namely the increasing population pressure on the land. This has been caused not only by population growth and the influx of Mozambican refugees, but also by the transfer of customary land to the leasehold estate subsector and, until about 1976, to the public sector for game and forest reserves. As a result, the area under customary land tenure declined from an estimated 8.2 million ha in 1964, to 7.8 million in 1970 and 7.1 million in 1988 (see Table 25). In addition, recent estimates indicate that since 1989 a further 440,600 ha have been transferred to the leasehold estate subsector.29 This transfer of land out of the smallholding subsector, in combination with a high rate of population growth, led to a decline in the average size of landholdings in the subsector. While data on land use from different sources is inconsistent, one source suggests that the average size of smallholder farms declined from about 2 ha to 1 ha between 1970 and 1988.30 In addition, there has been increased cultivation of less fertile lands. Access to cultivable land is therefore a primary constraint on growth in the smallholder subsector. Any growth in the smallholder subsector will, as a result, have to come primarily from increases in land productivity and a shift to the production of higher value crops. 3.15 Changes in the Smallholder Subsector, 19801987. Under Malawi's first structural adjustment program (1981/82 to 1984185), there was an emphasis on increasing smallholder production of export crops. As a result, Government's pricing policy, which had hitherto had the effect of taxing smaliholder export crop production and subsidizing maize production and consumption, shifted in favor of export crops and away from food crops.3' Export parity pricing was almost attained for the major smallholder export crops by 1984/85, and the producer prices of export 29 Government of Malawi/Coda/Afuican Development Bank Agriculture Sector Study Interim Report, June 1993. 30 MOALD, National Agricultural Survey of Agriculture. 31 Most estate pmduce is narketed through international markeLs to which estates have free access. Thus estate production is only affected by Govemment pricing and mnarketing policies to the extent that market is linked to production quotas (i.e. flue-cured and busley tobacco) or that estates are growing the same crops as smaliholders (e.g.. maizre). The taxation of smaliholder export crop production and the use of revenues thus raised to subsidize ADMARC's food crop trading accounts and other development roles. as well as various estate and manufacturing invesmnts, is widely documented. See, for example, Kydd, J. and R. Christiansen (1982). 'Structural Change in Malawi Since Independence: Consequences of a Development Strategy Based on Large-Scale Agricultire' World Development 1; GOM (1986). 'Report on ADMARC Reconstruction', Govenment Prss, Zomba, and Sahn. D. and J. Arunlpegasam (1991) 'land Tenure. Dualism and Poverty in Malawi'. in Including the Poor, van der Gaag, J. and M. Lipon (eds), Johns Hopkins University Press. 31  crops relative to those of maize and other food crops increased (see Table 27). Real maize producer prices declined between 1982 and 1987 (see Table 28). 3.16 In addition, undcr the second structural adjustment program (1984/85 to 1986/87), the Government cmbarked on the removal of Fertilizer subsidies, and the aggregaite rate of subsidization fell from 30.5% in 1983/84 to 19.8%'X in 1987/88. Dcspite tlhis, sales of fertilizer to the smallholder sector contiinued to rise. Table 3.3: Nominal Protection Coefflicients on Snialliolder Export Crop Production* Year Rice Toibacco G'nuts icanis Cotton 1980/S1 0.21 0.16 0.32 0.31 0.47 1984/85 0.40 0.78 0.74 0.73 0.27 1987/88 0.20 0.26 0.87 0.68 0.50 1991/92 0.13 0.45 0.44 0.71 0.27 Source: ADMARC records anid National Statistical Office Statistical Bulletins. * NPCs have been calculated as a ratio of nominal producer prices toi cxport parity prices (adjusted to lfiringatc levcl), based on (a) ADMARC proxducer prices for Faya Gradc I rice; Chalimbana Gradc A groundnuts; white ilaIricot hbaNis; a weighted average of Northcrn Division Dark Fired tobacco prices for various Grndes and Gr.adc A coronn: (h) average export prices for all varieties and grades, except in the case of tohaccn, which is 'he average NDDF auction fouir price and beans, for whici average pulse cxport prices had to he used, and (c) in the absence of data on actual fob prices. niarketing cosis have heen assumed at 10% of the producer plricc for groundnuts; 20% for tobacco, beans and cottoni and 33% for ricc, following Salin e al. (1990:227), op. cil.. 3.17 These shifts in pricing policy, in combination with limited opportunities to expand the area of production and ADMARCs increasing inability to purchase all the maize offered to it, led to a decline in maize production between 1984 and 1987 (see Table 29). In particular, the area of land allocated to hybrid maize fell, and that allocated to groundnuts and pulses, which require few inputs, rose (see Table 11 and Diagram 1). The inability to sell mnaize was crucial to farmers' decisions concerning hybrid maize, because the types of hybrid available in the 1980s had characteristics which were not suited to on-farm processing or storage. Hybrid maize was therefore grown as a cash crop, rather than as a source of food. 3.18 Other shifts in smallholder production between 1981 and 1987, apart from the decline in hybrid maize production, include: (a) a decline in the production of composite nin.. Fe, for the same reasons as the decrease in hybrid maize production; (b) an increase in the production of local maize and groundnuts; (c) a decline in the production of tobacco (partiLularly the dark-fired varieties), which has been attributed to both relative prices and ar increasing shortage of fuelwood, which is necessary for processing of this crop; 32  (d) a decline in cotton production between 1984/85 and 1986(87, due to increased taxation of export crop production after that year and the relative attractiveness of food crop prices; (e) an increase in pulse production since 1984/85. This trend is more difficult to explain. After the liberalization of smallholder agricultural produce markets in 1987, the upward trend in pulse production can be explained by the huge increase in demand for pulses. with private traders paying farmers almost double the price that ADMARC was offering for this produce. However, the increase in pulse production began before 1987 and after 1984185, up to which year ADMARCs real producer prices for pulses had been increasing. The rise in pulse production between 1984/85 and 1986/87 was therefore more likely due to the general shift into food crop production as taxation on export crop production increased after 1984/85. In addition, the removal of subsidies on inorganic fertilizer and increased land pressures may have stimulated greater production of nitrogen fixing crops such as pulses. 3.19 In 1986/87 a national crisis emerged in the form of a large maize deficit and groundnut surplus. The latter had to be exported at a loss, due to diffictilties in calculating in advance export parity prices for the specialist Malawian Chalimbana confectionery nut. In addition, large quantities of maize had to be imported, since production had declined and demand swelled, driven by the influx of Mozambican refugees. Simultaneously, world tobacco prices fell. The financial position of the marketing parastatal, ADMARC, became untenable in the face of these events (see Table 30). 3.20 This crisis is widely agreed to have been caused by a combination of: (i) external shocks - including the 1985 fall in world tobacco prices, the drought in the South, and the entry into Malawi of the refugees; (ii) inappropriate sequencing of marketing and input and output pricing reforms, such as the adoption of export parity pricing before the liberalization of export markets, which created a strain on ADMARC's cash-flow; the removal of subsidies on fertilizer and increase in the relative price of export crops prior to the availability of appropriate higher yielding food crop technologies, which led to substitutions between crops, rather than intensification of food crop production; (iii) an over- emphasis on pricing policy in a context of limited land and few technological options; and (iv) weaknesses within ADMARC, which meant that it did not have the resources necessary to react to supply and demand imbalances.32 3.21 Changes in the Smallholder Subsector, 1987-1994. Then in 1987, partly in response to this crisis, the Govermment liberalized smallholder agricultural produce markets; suspended the fertilizer subsidy removal program,33 and substantially increased the nominal producer price of maize, both absolutely and in relation to that of export crops (see Table 27). These moves, together with increased supplies of fertilizer, credit and hybrid maize seed, contributed to an estimated 69% increase in maize production between 1986/87 and 1992/93.34 Much of the increase in maize production since 31 See, for exanple. Harrigan. 1. (1988). 'Malawi: The Impact of Pricing Policy on Smahliolder Agriculture 1971-1988". Development Policy Review Vol 6, No 4. 33 'Ihe aggregate rate of fertilizer sub:;lizarion fell from 30.5% in 1983184 to 19.8% in 1986187 (Sahn D., J. Aruipragasam and L. Mcid (1990:120), 'Policy Reform and Poverty in Malawi: A Survey of a Decade of Experienrc". Coriell Food and Nutrition Policy Program, Monograph 7). 34 This cxpaiasion occurred despite declining real producer prices between 1989/90 and 1991/92 and falling real gross margins [or all maize enterprises between 1988/89 and 1991/92. 33  between 1987 and 1990 was the result of substituting hybrid varieties of maize for groundnuts. However, since 1990/91 rising maize production has increasingly resulted from the planting of hybrid maize varieties in place of local ones (see Diagram 1). The intensification of maize production, discernable in the late 1980s, was given an added boost in the early 1990s, with the introduction of smaliholder burley tobacco production which led to an increased ability to purchase hybrid seeds and fertilizer, and the release of some hard endospern hybrid maize varieties, which are more suited to smallholder processing and storage. Hybrid maize has now become a means of increasing food supplies, as well as of generating cash income. 3.22 While average yields of all other crops grown by smallholders have fluctuated or remained fairly constant over the 1980s and early 1990s, those of maize increased from about 1,000 kg/ha in 1986/87 to about 1,500 kg/ha in 1992/93. While this increase is not part of a sustained trend (see Diagram 2), the increase indicates that a number of Malawi farmers can increase their yields significantly. The big jump in yields essentially over a one year period was due to the big increase in the use of maize hybrid seeds, good weather conditions, drought-related assistance in the form of free seeds, and the temporary extra nitrogen available in the soil as a result of the drought in the previous year. Maize yields also fluctuated between 1986 and 1994. In the 1993194 crop season, crop estimates indicate that overall maize production dropped significantly (from 2.1 M MT to about 830,000), such that maize yields averaged less than 900 kgs/ha (due to a credit crunch, the resulting low usage of inputs, and erratic rainfall at a critical stage, with the last factor explaining about 65% of the drop in the previous year production). In the current 1994195 crop season, it appears that the delayed onset of the rains is likely to result in another low maize harvest and yields. Analysis of available maize production data indicates that the variability of maize production increased substantially during the last five years; the associated decline in maize productivity and increased variability of national production reflect two major constraints to maize production in Malawi: declining soil fertility and increased production variability due to drought (e.g., 1992, 1994 and possible low harvest in 1995). Thus, this demonstrates that the smallholder subsector has not yet achieved a sustained breakthrough in increasing maize yields. This report discusses the underlying reasons (apart from the weather factors) and how they could be addressed. 3123 This increase in maize yields in 1992/93 is the result of an increase in the proportion of hybrid maize planted from 3% of total maize area in 1986/87 to nearly 25% in 1992/93, as well as increased use of inorganic fertilizer throughout the 1980s and early 1990s, and the one-off extra availability of nitrogen in the soil. In tum, the first two factors are the result of technological changes; increased supplies of hybrid maize varieties, fertilizers and credit to the smallholder subsector; Government extension efforts; and changes in smallholder produce pricing and marketing policies, including the initiation of smallholder tobacco production in 1990/91. It is also important to recognize that the average yields of hybrid maize have been relatively stagnant (averaging about 2.5 MT/ha), well below attainable yields of about 4 MT/ha under smallholder conditions. This yield gap is partly due to deficient fertilizer reconmmendations and fanner crop husbandry practices. The drop in maize production is due to various factors as explained above, including a significant drop in the use of hybrid maize seed (dropping to about 10% of total maize area, after peaking at about 25% the previous year). The increases and declines in maize production can be traced to expanded farmer access to fertilizer, and their ability to purchase it (cash and/or credit, although fertilizer allocation has been linked to credit users). The majority of smallholders have not purchased hybrid seed because they cannot afford fertilizer. A small proportion of smallholders having fertile soils or farmyard manure could profitably use maize hybrid seed without fertilizer. Accordingly, the majority of smallholders farming on poorer soils have been constrained by a lack of purchasing power and access 34  to credit which have limited their uptake of fertilizer and the improved seeds. It is this lack of purchasing power which has resulted in stocks of hybrid seed being carried over from one season to the other in recent years. Table 3.4: Real ADMARC Producer Prices (1980 tambala/kg)* Yewr MCzc Rice Tobacco G'nuts Beans Cotton 198011 5.98 9.06 41.70 29.92 12.69 20.85 1984/S 7.02 9.79 58.70 40.27 23.01 26.46 1987/8 4.82 8.70 41.81 29.66 17.40 19.57 1991/2 5.72 7.51 48.27 21.56 14.05 17.33 Source: G Donovan (1993), 'Malawi: Structural Adjustment and Agriculture', ASM working paper. * Dcflated by Consumer Retail Price Index. Producer prices refer to Faya Grade I rice; Chalimbana Grade A groundnuts; white Haricot beans; a weighted averagc of Northem Division Dark Fired tobacco prices for various Grades and Grade A cotton. 3.24 Other significant changes in smallholder production patterns post-1987 are: (a) dramatic decline in groundnut production and in the area of land allocated to this crop; (b) decline in the area of land allocated to composite maize and, since 199L, to local maize, largely as a result of the increase in hybrid maize production, which began to substitute for local and composite maize after 1990; (c) steady increase in rice production, a market which is dominated by the private sector; (d) increase in the area of land allocated to tobacco after 1988/89, in particular of Northern Division Dark Fired and burley tobaccos; (e) increase in cotton production between 1986187 and 1990/91, followed by a decline in the area of land allocated to this crop, possibly as a result of hybrid maize substitution, (although this needs to be verified); (f) continued increase in pulse production; (g) rapid increase in fertilizer and hybrid maize seed sales and in credit disbursements, although there was a significant drop in these three factors in the 1993/94 crop season largely due to the poor smnallholder credit repayment in 1993 (discussed fuirther in para. 3.31). 35  Diagram 2: Average Smaliholder Yields (based on MOA crop production estimates) 1.8 1.4 1 . - MAIZE -a- RICE 0.1 0.8 - _ - \ / - TOBACCO 0.6 A COTTON 0.2 0~~~~~~~~~~ co co co oO, co so X ) a) 0)X 0) 0) co 0) o 0) 0 X X aX _ - _- _ - 0) n  3.25 The dramatic decline in groundnut production is attributed to the increased relative attractiveness of maize production post-1987, as a result of the shift in Government pricing policy following ADMARC's crisis in 1986/87. One of the reasons why Government's pricing policy has had this powerful effect is that an embargo has been intermittently placed on private traders engaging in both domestic purchases and the exportation of groundnuts. Government's justification for this embargo is that ADMARC has been unable to purchase sufficient supplies of groundnuts to ensure adequate seed supplies since the entrance of private traders into the smallholder agricultural produce markets; the latter have consistently offered farmers a higher price for their produce. ADMARC had therefore been unable to compete with the private sector in the groundnut market and resorted to appeals to the Government to reinstate its monopoly on the marketing of this crop. Consequently, despite liberalization, farmers have often only had access to ADMARC producer prices for this crop. These prices have fallen in real terms every year since 1983. 3.26 ADMARC's inability to compete with private traders in the groundnut market is the result of its failure to respond to a qualitative shift in the demand for and supply of Malawian groundnuts. In the 1960s, 1970s and first half of the 1980s, Malawi was a major exporter of large, high quality confectionery nuts, primarily to British markets. However, with geneti; drift the quality of supplies fell, and at the same time importers' demand shifted towards smaller, more consistently sized nuts, which are better suited to mechanical processing. Thus ADMARC's traditional buyers switched to US suppliers. The private trader demand for groundnuts illustrates that other export markets are available for the nuts currently grown in Malawi, although these have not been tapped by ADMARC. In addition, a new variety of nut (CT7 or ICGM S42) has been developed which better meets the quality demanded of major exporters as well as producing good yields. This, together with high private trader demand, offers the potential for a regeneration of Malawi's groundnut industry. However, the experience of the 1980s has shown that if taxation of smallholder farmers' production of export crops continues, through Government pricing and marketing policies, this industry's development will be cut short. 3.27 The increase in pulse and rice production since 1987 is widely agrced to be due to the entrance of private traders into these markets. Since 1987, private traders have consistently exceeded ADMARC's offers in paying farmers. (This means traders pay both farmers and ADMARC) for their rice and pulses. However, once again, the potential for increased incomes and export earnings provided by this trend has been limited by the Governent's embargo on private trade in beans in 1988 and its lirnitation on the level of all exports of food crops since then. 3.28 The rise in smallholder tobacco production since 1987 can in part be explained by the change in Government's policy which made it legal for the first time in Malawi's history for smallholders to produce burley tobacco. Under agreements with IDA and USAID, smallholders have been allocated a burley production quota, which has been increased every year since its inception, from 3 million kilograms in 1990/91 to 7 million kilograms in 1992/93 and a projected 15.0 million kilograms in 1994/95. It has been estimated that between 1990/91 and 1992/93 about 14 million kilograms of burley tobacco have been produced by smallholders under this program. 3.29 The reasons for increased production of NDDF are probably related to pricing policy, although reasons for the increase and subsequent decline in cotton production are less obvious. While real ADMARC producer prices for NDDF have increased since 1987, those for cotton have declined (see Tables 3.4 and 28). Howe-'er, cotton prices have fallen relative to maize prices since 1989 and substitution between these two crops may explain the shifts in cotton production (see Table 27). 37  3.30 ADMARC fertilizer sales have more than doubled during the 1980s, from 49,000 t in 1980 to 107,000 t in 1991, with a particularly rapid increase since 1987 (see Table 31).35 Similarly, between 1980/91 and 1991192 sales of hybrid maize seed increased by an average of nearly 16% per annum. These rates of growth compare favorably with those in neighboring countries36 and have taken place in a context of increasing real retail prices (see Table 32). It is therefore widely concluded that the level of fertilizer utilization by smallholders in the 1980s was principally limited by low supply.37 In the context of a supply constraint, fertilizer has been rationed through the credit system. About 65% of fertilizer is sold on credit through farners' clubs, and it is reported that obtaining supplies through cash purchasing, in a timely manner, is difficult. However, recent field surveys suggest that while the supply constraint may be binding for higher income groups of smallholders, effective demand and purchasing power constraints have limited (and continue to date) the use of hybrid seeds and fertilizer among lower income smallholders (see paras. 4.21-4.30). 3.31 The fact that the number of smallholder credit recipients increased by 15% per annum during the 1980s supports the argument that for many smallholders, fertilizer utilization is constrained by demand rather than supply factors. Credit supplies to the smallholder sector were increased during the 1980s, and particularly following the crisis in 1986/87, because it was realized that only the top 20% of smallholders were receiving such credit. Between 1982 and 1991, the proportion of smallholder households receiving credit from SACA rose from 12% to 23%. In addition, female borrowers increased from 15% of the total in 1982 to 30% in 1993. Total loans outstanding in 1991/92 amounted to MK 67.8 million, involving nearly 340,000 households. However, with the expansion of the SACA's disbursements, there has been a decline in recovery rates, from an average of about 96% between 1982/83 and 1984/85 to 81% between 1986/87 and 1990/91. Recovery rates have also continued to fall since then, although the very low repayment rates for the 1992/93 crop season (about 16%) can be attributed to exceptional circumstances.3' In part this decline in recovery can be attributed to a dropping of the 100% recovery rule which stipulates that credit clubs have to repay 100% of their loans in order to be eligible for fairther loans. The rule was reinstated in 1992, but while there is no doubt that increased credit supplies have contributed to the increased use of fertilizer and hybrid maize seed amongst smallholders, the sustainability of the credit supply system needs to be urgently addressed. 3.32 Summary of Smallholder Subsector Performance. The performance of the smallholder subsector has been worse than that of the estates. No sustained increase in the production of higher value, non-food crops has occurred (although there was a temporary move in this direction between 1984/85 and 1986187); average maize yields fell between 1980 and 1987, and although they have risen since then, only to drop again in the 1993/94 crop season, hybrid maize varieties only began to 35 Furher, Govermnens pricing and subsidization policies have encouraged the increased use of high analysis fertilizers. Consequendy salcs of nitrogen to the smaliholder sectior increased by about 23% a year between 1980 and 1992 and those of phosphate increased by ovcr 70% a year over die samne period. 36 Fertilizer consumption by the communal sector in Zimbabwe grew during the first five years of Inependence, but has declined in the last six years. In Kenya, the growth in fcrdlizer consumption by smallbolders avenged 8% per annum during the 1980s and in Tanzania it avcraged 4%. 37 See for example. Conroy. A. (1992) 'The Inputs Sector Fertilizer and Seed', Agricultural Sector Memorandum Worldng Paper. Sahn. D. and J. Arulpragasam (1991).