Document of The World Bank FOR OFFICIAL USE ONLY Report No. 16529 MOZ MOZAMBIQUE AGRICULTTRAL SECTOR MEMORANDUM VOLUME II: MAIN REPORT April 30, 1997 5Z C Africa Region Agriculture and Environment K ; CURRENCY EQUIVALENTS Currency Unit = Metical (Mt) US$1.00 = Mtl 1,378.00 in February 1997 FISCAL YEAR April I - March 31 WEIGHTS AND MEASURES Metric System Vice President: Mr. Callisto E. Madavo, AFR Country Director: Ms. Phyllis Pomerantz, AFCO2 Technical Manager: Ms. Sushma Ganguly, AFTA I Task Team Leader: Mr. Peter G. Moll, AFTAI ii ACRONYMS AGRICOM - Agricultural Marketing IAM - Instituto do Algodao de Enterprise Mo9ambique (Cotton Institute of ARA - Regional Water Board Mozambique) CDA - Center for Agricultural ICM - Instituto de Cereais de Mo,ambique Documentation (Cereals Institute of Mozambique) CFA - Agricultural Training Center IDA - International Development CGIP - General Coordinator for Integrated Association Projects INDER - Rural Development Institute CITES -Convention on International Trade INIA -National Agricultural Research in Endangered Species Institute CNP - National Planning Commission INIVE - National Veterinary Institute CPFA - Provincial Center for Agricultural IPA - Animal Production Institute Training IPP - Import Parity Price DINAGECA - National Directorate for ISNAR - International Service for National Geography and Cadastre Agricultural Research DINAP - National Directorate for Livestock JC - Joint Venture Company DINAP- National Directorate for Rural MOA - Ministry of Agriculture (and Development Fisheries) DPA - Provincial Department of MOF - Ministry of Finance Agriculture Mt - Metical DPCCN - Departamento de Preven,cao e MT - metric ton Combate as Calamidades Naturais NCP - National Commission on Price and (Department of Prevention of and Wages Combat against Natural Disasters) NGO - Non-governmental agencies DRC - Domestic Resource Cost Coefficient NEPC- Net Effective Protection Coefficient EFA - Education for All NNPC - Net Nominal Protection EPC - Effective Protection Coefficient Coefficient ESRP - Economic and Social Rehabilitation NPC -Nominal Protection Coefficient Program PDP - Priority Districts Program EWS - Early Warning System PESU - Emergency Seeds and Tools FFADR - Fund for Rural and Agricultural Program Development PTIP -Triennial Program of Public FFDHA - Fund for Irrigation Development Investment FRELIMO - Frente para a Liberta,ao do PUPI - Small Unit for Intensive Production Mozambique (Mozambique Liberation PRE - Economic Rehabilitation Program Front) RENAMO - Mozambican Resistance Front GEF - Global Environment Facility ROCS 11- Second Roads and Coastal GPIE - Agency for Foreign Investment Shipping Project Promotion SEHA - Secretariat of Irrigation GUC - General Union of Cooperatives SOE - State-owned enterprises iii TABLE OF CONTENTS CHAPTER 1. BACKGROUND ............................................1 A. The Legacy of Colonialism ............................................ 1 B. The Legacy of Socialism .......2....................................2 C. The Legacy of the War .......5....................................5 D. Economic Recovery ............................................6 CHAPTER 2. CURRENT ECONOMIC FRAMEWORK ............................................9 A. The Economic Rehabilitation Program ............................................9 B. Economic Reforms and Performance . ........................................... 0 C. Agriculture and the Macroeconomy ........................................... 15 The Shrimp and Prawn Industry ........................................... 19 CHAPTER 3. THE RURAL ECONOMY ............. .............................. 23 A. Natural Resources ........................................... 23 B. Land Use ........................................... 34 C. Rural Population and Labor Availability ........................................... 42 D. Rural Poverty and Gender Issues ........................................... 44 Rural Poverty ........................................... 44 Gender Issues in the Rural Areas ........................................... 51 E. Rural Social Infrastructure ........................................... 53 F. Rural Physical Infrastructure ........................................... 54 CHAPTER 4. THE INSTITUTIONAL FRAMEWORK ........................................... 61 A. The Legislative Framework ........................................... 61 The Land Law ........................................... 61 Water ........................................... 66 Marketing, Prices and Incentives ........................................... 66 Taxation, Tariffs and Licensing ........................................... 71 B. The Government as it Affects Agriculture .................. ......................... 73 The Agencies Dealing With Agriculture ................. .......................... 73 Agricultural Research ........................................... 74 Agricultural Extension ........................................... 75 Health and Nutrition ........................................... 77 Education ........................................... 79 C. The Financial System ........................................... 81 D. The Inputs Supply Markets ........................................... 83 E. Food Aid ........................................... 85 CHAPTER 5. THE AGRICULTURAL ECONOMY ........................................... 89 A. Aggregate Demand ........................................... 89 B. Aggregate Supply ........................................... 92 C. Comparative Advantage ........................................... 94 D. Subsectoral Conditions and Prospects .............. ............................. 100 Food Crops ............................................100 Cashew ............................................ 105 Cotton ............................................ 110 Citrus ............................................ 115 iv Other crops ......................................................... 117 CHAPTER 6. A STRATEGY FOR AGRICULTURAL DEVELOPMENT IN MOZAMBIQUE ......................................................... 119 A. Introduction ......................................................... 119 B. Previous Sectoral Reviews ......................................................... 119 C. Government Policy Statements ......................................................... 120 D. Current Status of Mozambican Agriculture ......................................................... 121 E. Major findings of this Memorandum ......................................................... 122 (a) Raising SmallholderAgricultural Production is an Effective Way of Addressing Poverty ......................................................... 122 (b) Smallholders are the Most Efficient Farmers .............................................. 125 (c) The Private Trading System is Recovering .................................................. 126 (d) Integration in regional grain markets will dominate long-distance internal (north-south) trade ......................................................... 127 (e) Food Security is Attainable Without Government-Directed FoodAid ........ 128 (/) Local Cashew Processing Can Compete at World Standards ...................... 129 F. Constraints on Agricultural Growth ......................................................... 130 G. A Vision of Mozambique's Agricultural Development ....................................... 133 H. An Agricultural Development Strategy for Government ..................................... 138 (a) Principles of Government Intervention ........................................................ 139 (b) The Incentive Framework ......................................................... 140 (c) Rural Transport and Communication Infrastructure .144 (d) Public Services: Research and Extension .................................................... 145 (e) Regulation ......................................................... 146 69 Public Services: Health, Education, Water .................................................. 147 (g) The Public Investment Program ................................................ ......... 147 (h) The Roles of IDA, other International Agencies, Donors, and NGOs ............ 148 (i) Timing of Operations and a Short-Term Action Plan ................................... 150 0) Timing of Operations: Matrix ......................................................... 152 REFERENCES ......................................................... 153 APPENDICES ......................................................... 161 v LIST OF TABLES Table 1. Annual Percentage Changes in Real Gross Output, 1987-96 .............................1 ]0 Table 2. Percentage Annual Growth Rates of Marketed Agricultural Output, 1989-1996 ........................................................ 15 Table 3. Regional Distribution of Production in 1992/93 ................................................ 16 Table 4. Output of the Mozambican Shrimp and Prawn Fishing Industry ....................... 19 Table 5. Land Potential and Current Use ........................................................ 24 Table 6. Current Land Availability ........................................................ 25 Table 7. Land Suitable for Irrigation ........................................................ 26 Table 8. Area under Cultivation and Percentage of Cultivated Area Planted to Specific Crops, 1992 ........................................................ 34 Table 9. Cultivated Areas by Producer Category ........................................................ 37 Table 10. Smallholder Sector Characteristics ........................................................ 38 Table I ]. Total and Marketed Production, 1990/91 ........................................................ 39 Table 12. Estimates of Average Yields ........................................................ 41 Table 13. Estimated Population, 1992-2025 ........................................................ 43 Table 14. Basic Social Indicators ........................................................ 46 Table 15. Characteristics of Female- and Male-headed Households ............................... 52 Table 16. Second Roads and Coastal Shipping Project: Phase One Trunk Road Rehabilitation ........................................................ 56 Table 17. Price Regime of Major Agricultural Commodities, 1993 ................................ 68 Table 18. Determination of on-farm maize prices based on import parity ...................... 69 Table 19. Protection and Profitability of Major Crops ..................................................... 70 Table 20. Total agricultural lending, 1988-92 ........................................................ 81 Table 21. Cereal Food Aid Imports 1980-1993 ........................................................ 85 Table 22. Yields and Domestic Resource Cost Coefficients ............................................ 96 Table 23. Regional Competitiveness of (Unimproved) Maize Production: Domestic Resource Cost Coefficients and Farmer Earnings Per Day ........................ 97 Table 24. Domestic Resource Cost Coefficients, Smallholder Cotton in Cabo Delgado Province, With and Without Fertilizers ........................................................ 98 Table 25. Alternative Approaches to Fulfilling the Functions of a Grain Marketing Parastatal ........................................................ 103 Table 26. Farmer Price as a Share of Cashew Export Earnings ..................................... 107 Table 27. Competitiveness of Local Processing and of Raw Nut Exports ..................... 108 Table 28. Protection Required by Cashew Processors ................................................... 109 Table 29. Seed Cotton: Area Sown, Yields and Output, by Farm Type, 1993 ...............l II vi LIST OF FIGURES Figure 1. Money supply growth and inflation, 1990-96 ................................................... 11 Figure 2. The real exchange rate, 1990-96 ........................................................ 12 Figure 3. Agricultural sector productivity ........................................................ 17 Figure 4. Commodity composition of exports, 1996 ........................................................ 18 Figure 5. Commodity exports, 1989-96 ........................................................ 18 Figure 6. Livestock ownership indices, 1995 ........................................................ 29 Figure 7. Percent of farm households which do not sell any farm output ........................ 47 Figure 8. Percentage of farm households without non-agricultural income, 1994 .......... 47 Figure 9. Patients per hospital staff, by province, 1992 ................................................... 48 Figure 10. Intra-hospital infant mortality, 1992 ........................................................ 48 Figure 11. Access to safe drinking water, by province .................................................... 49 Figure 12. Schooling of household heads, Monapo district (Nampula) ........................... 49 Figure 13. Enrollment of children in primary school, by province .................................. 50 Figure 14. Mean months of food insecurity, by farm size, as reported by rural households, 1994 ........................................................ 51 Figure 15. GNP and roads, all countries, 1993 ........................................................ 55 Figure 16. Feeder road rehabilitation ........................................................ 57 Figure 17. Road length and area under cultivation, by province ...................................... 58 Figure 18. Real prices, 1976-92 ........................................................ 67 Figure 19. Marketed agricultural production, 1975-94 .................................................... 67 Figure 20. Infant mortality rates, Sub-Saharan countries .................................... ............ 78 Figure 21. Stunting of under-5-year-olds, 1995 ........................................................ 79 Figure 22. Real interest rates for agriculture, and the implicit subsidy for agriculture as a percentage of total government expenditure, 1988-1994 .................. 82 Figure 23. Food aid maize as a percentage of total cereals available .............................. 86 Figure 24. Real price of maize, Maputo, 1990-1995 ........................................................ 90 Figure 25. Marketed output of raw cashew in Mozambique, 1955-96 ........................... 105 Figure 26. Seed cotton production, by farm type, 1955-94 ............................................ 111 vii LIST OF APPENDICES APPENDIX 1: TABLES ........................................................ 161 Appendix Table I. Selected Macroeconomic Indicators ................................................ 161 Appendix Table II. Commodity Composition of Exports, 1989-96 ............................... 162 Appendix Table III. Shrimp and Prawn Exports by Mozambique, 1990-1995 .............. 163 Appendix Table IV. Nominal and Real Producer Prices, 1976-92 ................................ 164 Appendix Table V. Marketed Agricultural Production, All Crops, 1975-94 ................. 165 Appendix Table VI. Sources of Income of Rural Households ....................................... 166 Appendix Table VII. Marketing of Principal Food Crops: Percent of Farming Households which Sell Maize, Rice, etc . ........................................................ 167 Appendix Table VIII. Percent of Households which Own Livestock, and Mean Numbers Owned ........................................................ 168 Appendix Table IX. Distribution of Land Area Possessed by Households .................... 169 Appendix Table X. Classification of Vulnerability of Communities, by Diet ............... 170 Appendix Table XI. Mean Months of Food Insecurity Reported by Rural Households ........................................................ 171 Appendix Table XII. Percentage Moderately and Severely Undernourished Under Five Year Old Children, in Terms of Wastedness, General Malnutrition and Stundedness ........................................................ 172 Appendix Table XIII. Land Concessions, 1986-93 ........................................................ 173 Appendix Table XIV. Land Tenure Among Smallholders: Sources of Acquisition of Machambas ........................................................ 174 Appendix Table XV. Ownership, Renting and Use of Tools of Agricultural Production ........................................................ 175 Appendix Table XVI. Marketed Production, Processing, and Export of Raw Cashew in Mozambique from 1980 to 1994 ........................................................ 176 Appendix Table XVII. Price Structure and Profit Margins of Cashew Nuts, 1993-1995 ........................................................ 177 Appendix Table XVIII. Competitiveness of Local Processing and of Raw Nut Exports ........................................................ 178 Appendix Table XIX. Seed Cotton: Area Sown and Output, by Firm and Farm Type (Outgrowers, Medium, Large-Scale), 1993 ..................................................... 180 APPENDIX 2: Domestic Resource Cost Coefficients ................................................ 181 viii ACKNOWLEDGEMENTS An initial mission to Mozambique took place in November 1993, led by Mr. Jean-Paul Chausse (then AF6AE). Participants included: Ms. Paola Agostini (forestry), Mr. Amos Ben Mayor (extension and research), Ms. Christine de Voest (food aid and food security), Mr. Pierre Haas (farmers' associations), Mr. Peter G. Moll (comparative advantage and farm efficiency), Mr. Augusto Moreno (the institutional environment), Mr. Raymond Noronha (land management), Ms. Adelina Paiva (pricing and incentive policy), Mr. Gale Rozell (privatization and the private sector environment), Mr. Lloyd Strachan (input supply and marketing), Mr. Jorge Tembe (farming systems and sub-sector notes). The first draft was prepared by Mr. Jean-Paul Chausse; the report was completed by Mr. Peter G. Moll. Thanks are expressed to Messrs. Jean- Paul Chausse, Inacio Manecas and Luigi Marcuccio for their assistance and to Mr. James Coates for his guidance. The Peer Reviewers were Mr. Graeme Donovan and Mr. Gene Tidrick. The work was carried out under the supervision of Mr. Jean-Paul Chausse (then Chief of the AF 1AE Division), and subsequently under Ms. Sushma Ganguly (Technical Manager for AFTAI, Agriculture and Environment, Eastern and Southern Africa). The Country Director is Ms. Phyllis Pomerantz. ix Chapter 1. BACKGROUND 1. Mozambique's structural development problems are deeply rooted in its colonial past and the impact of this colonial legacy on post-independence policies. Economic and social dualism under the colonial power, Portugal, a long war of independence, post-independence policies aimed at promoting the fast government-led industrialization of the economy and a profound restructuring of a "backward" peasant society imposed devastating human and economic costs on the country. A. The Legacy of Colonialism 2. The problems facing Mozambique since independence are intertwined with its colonial history because of the particular nature of Portuguese colonial administration. Portuguese economic activity has from the outset been based on various forms of forced labor. The colonial system rested at first on slavery-based plantations and after the abolition of slavery in 1869, a system of forced and contract labor continued to guarantee an ample supply of cheap labor and raw material for Portuguese enterprises. Local populations were required to supply contract labor to former owners and there was extensive use of forced labor for plantation and public works. Taxes were levied to induce populations to seek employment on plantations and in mines, in Mozambique or in neighboring countries, in order to eam the cash needed for tax payment. This system helped force rural populations into production of cash cops such as cotton and rice. Mozambique developed essentially as a supplier of agricultural raw material to Portugal (cashew, copra, cotton, tea), though there was some industrial development, essentially for the import substitution of basic consumer goods. Owing to its unique geographical position, Mozambique also generated substantial income from transit services for South Africa and land- locked British colonies to the West. Over time, the Mozambican industrial sector became dependent on South Africa for spare part and services. The sources of foreign exchange (agricultural exports, transit services and workers remittances) and the trading partners (Portugal, South Africa and the neighboring British colonies) were therefore few. This excessive external dependence became a severe liability after independence, with the drastic deterioration of Mozambique's relations with its traditional partners. 3. After the Second World War, there was a renewed push by Portugal to extend its occupation of the overseas territories. Thousands of peasants and unskilled workers were sent to Mozambique to work as agricultural settlers or fill public and service sector jobs that in other colonial systems were usually open to the local population. Between 1940 and 1974, the Portuguese population rose from 27,000 to 200,000. There was a lack of investment in social infrastructure, particularly in rural areas. Africans were denied access to education (only mission schools provided schooling for a few Mozambicans) and the economy became totally dependent on a non-African labor force for all but the most menial tasks. The mass exodus of expatriate settlers at independence (about 90 percent departed within 18 months) left Mozambique confronted with a critical shortage of skilled labor: medical personnel, teachers, managers, technicians, traders and semi-skilled workers. 4. Colonial Mozambican agriculture had three distinct sectors: large plantations, medium scale settler farms and small African family farms. The commercial sector included 2,000 farms, including the few large plantations, and cultivated about 175,000 hectares or around 7 l percent of the country's total cultivated area of about 2.5 million hectares. The average size of commercial farms' concessions (not freehold but long-term usufruct rights that could be transferred and inherited) was about 45 hectares. The farms produced mainly export crops (cotton, tea, cashew, coconuts and sugar) and also food for the market. There were about 1.5 million African smallholder farms cultivating 2.3 million hectares (1.5 hectares per farm). They practiced shifting cultivation, essentially for subsistence on land allocated through the customary land tenure system. The colonial government did not provide support services to smallholders who, because of the lack of resources and secured markets followed a risk-averting strategy, producing a low but reliable output. However, the smallholder sector was not technologically static. Work on commercial farms was a source of exposure to new opportunities. Also, the rural commercial network provided consumer goods and agricultural inputs, marketing services for output and some (very limited) credit. This sector consisted primarily of Portuguese and Asian merchants and was a vital component of the rural economy. The smallholder sector, in spite of being undercapitalized, not having access to essential services and being subject to unfavorable terms of trade, produced in the early 1970's an estimated 70 percent of total agricultural production and contributed more to total exports than the commercial farms (30 percent against 27 percent). 5. It is important to note the lack of interconnections between the regions of the country during the colonial era. Taken as a whole, Mozambique has ample land and food production potential to be self-sufficient. However, this fact obscures the regional differences persisting from the colonial period to the present. Though there is ample grain production potential in the North and Central highland regions, these areas have never been large suppliers to the large population centers of the arid South. Marketing networks to move surplus grains in this way were never developed, and the South has historically always been a food deficit region, exporting labor and transport services in return for food needs. Thus, development of the country's natural comparative advantage in food production will require not only the rehabilitation of infrastructure and marketing systems destroyed by the protracted war, but the development of entirely new interregional linkages and marketing channels. B. The Legacy of Socialism 6. After 10 years of nationalist war, Mozambique won its independence in 1975. FRELIMO, the Frente para a Libertasao do Mozambique (Mozambique Liberation Front), assumed control of a country where the administration and most of the economic infrastructure was left abandoned after the mass exodus of the Portuguese population. During the first few years of independence, the Government's primary concern centered on avoiding the collapse of economic activity and preserving urban food supplies, foreign exchange earnings and jobs. In line with the prevailing Marxist economic philosophy of the ruling party there was a strong tendency toward centralized control of large production units in a system envisioned as an industrialized command economy. In the agricultural sector, the State took over the 2,000 abandoned commercial farms. Due to the acute shortage of technical and management skills, settler farms were regrouped into larger units. Parastatals were set up to undertake the marketing and trade functions left vacant by the departing traders. This initial drive for State control, partly dictated by urgent practical considerations, was further reinforced by ideological factors, underpinning a socialist development strategy based on a rapid "modernization" of the Mozambican society. Following the third FRELIMO Congress in 1977, the Government consolidated its role in the economic and social life of the country. Under a centralized management system, output targets 2 and resource allocation became increasingly determined by administrative fiat. Prices were fixed for most commodities, at all stages of the production and distribution chain, with little reference to markets. External dependence was increased further by the adoption of a development strategy based on heavy (public) investments in imported capital intensive technologies. The foreign exchange and technical assistance requirements for the operation and maintenance of these investments further aggravated the country's dependence on external resources. In addition, the low capacity of most of these investments to generate the expected value added and foreign exchange earnings quickly created severe balance of payments problems and made the development strategy unsustainable. 7. In the agricultural sector, absolute priority was given to the development of the State Farm Sector as the most effective means - through rapid intensification, mechanization and expansion of the area under cultivation - to quickly increase production and provide employment opportunities. Massive purchases of 1,200 tractors and 500 combines were made in 1977, with additional purchases in the following years. Between 1978 and 1982, 90 percent of Government's investment in agriculture was allocated to the state farm sector. The latter's performance was, however, disappointing. State farms relied on mechanized and input intensive technologies which required experienced managers and technicians, who were in critically short supply, and substantial imports of equipment and inputs. The central planning process imposed excessive rigidity on a production process - agriculture - that requires great flexibility to adjust to the uncertainties of farming. The Government lacked the capacity to plan and coordinate the various support services necessary for these complex enterprises, and inputs were often not available at the right time to sustain efficient production. Production targets were generally set without taking agronomic or financial viability into consideration. Worker motivation was low. Although state farms' yields were higher than in the smallholder sector, their contribution to the economy was low and sometimes negative, on account of technologies not well suited to the country's endowment and the low efficiency of their operations. As a result, many state farms were not even able to cover their operating costs and accumulated large debts to the (public) banking sector. 8. The departure of the Portuguese traders left large areas without any network for the distribution of goods and marketing of produce. The void was filled by the creation of a network of government stores (lojas do povo) while the remaining private traders were authorized to operate at retail level. The government retained a monopoly at the wholesale level, however, and marketing margins were tightly regulated. Exclusive responsibility for crop marketing was granted to parastatals. In 1981, all marketing of crops (except cashew and cotton) and the distribution of goods to rural areas were consolidated under AGRICOM, a state enterprise under the Ministry of Internal Commerce. In 1994 AGRICOM was abolished and replaced by the Instituto de Cereais de Mo~:ambique (ICM), which took over the staff and physical assets of AGRICOM. The ICM purchased maize from farmers and from the commercial trade network and sold it chiefly to aid donors such as the World Food Programme. 9. As in colonial times, the new Government considered the smallholder sector - which accounted for 95 percent of the total land under cultivation and for a slightly lower share of total agricultural production - as economically inefficient and a constraint to the "modernization" of rural life. Government's strategy to "socialize the countryside" focused on grouping smallholders in "Communal Villages" along the lines of the Ujaama village developed in neighboring Tanzania. Besides offering a tool for the collective organization of smallholder production, the communal villages also facilitated political mobilization and provided a focal point for the delivery of social and agricultural services to scattered rural communities. People 3 mobilized in communal villages had to accommodate Government-sponsored schemes for collective work. The stated objective was to speed up the transfer and adoption of improved technologies for increasing the productivity and production of smallholders. 10. The impact of Government's rural socialization policy was uneven but had some positive results. It helped supply education and health services to rural populations in an effective, low- cost manner; the effort was ultimately unsuccessful, however, on account of the war, and infant mortality fell by only 15 percent (from 171 per 1000 in 1970 to 146 in 1993) compared with the drop of 29 percent for Sub-Saharan Africa (from 141 to 100) . The Government's policy promoted the emancipation of women by weakening some traditional male-dominated social structures. The efforts in literacy and schooling were considerable. For instance, the percentage of primary-age children registered in school rose from 47 percent in 1970 to 60 percent in 1992.2 11. The negative results of the rural socialization policy were far greater. Stable patterns of rural subsistence were replaced by poorly managed and unsustainable enterprises. Government- appointed managers had little experience and the supply of inputs necessary to support intensified agriculture did not materialize. The high concentration of farming families reduced the possibility of shifting cultivation and led to soil degradation while fuelwood scarcity increased rapidly around the villages. Currently, fifteen years after the villagization drive, many of these villages still remain, despite the fact that they were not barred from leaving. Although their survival is partly due to the increased security they provided to populations during the war, it is also evidence of the useful role they played in the provision of essential services. Collective production schemes, however, have been abandoned and families cultivate their own plots, albeit sometimes at considerable distance from the villages. The socialization of rural life, besides its negative impact on the management of the natural resource base and on the production of the smallholder sector, also alienated part of the rural population, as people lost land rights they had maintained during the colonial period and traditional leaders were stripped of their authority by the new Government. Discontent fueled underground domestic opposition and provided an opening for the destabilization strategies of the then hostile Rhodesian and South African regimes. 12. Agricultural cooperatives were established on farms abandoned by their pre- independence owners. While present in almost all regions of the country, these cooperatives are concentrated in the southern provinces and in particular in Maputo and Gaza. Their importance has markedly declined over the years and they are currently responsible for less than I percent of the total marketed output. Many of these cooperatives have spontaneously evolved into service cooperatives. Most have survived or emerged as a result of external support and are still significantly dependent on this support. The survival and importance of this sector lies essentially in its potential role in promoting the access of smallholders to basic services, as a representation of smallholder sector's interests, and as a countervailing power in the negotiations with the Government and other types of farns. ISource: World Development Indicators 1995, Table 27. This is the under-two mortality rate per one thousand live births. 2 Source: World Development Indicators 1995, Table 28. 4 C. The Legacy of the War 13. After a dramatic fall in economic activity following the abrupt departure of the colonists (1975-77), there was a partial recovery in economic activity and agricultural production until 1981, by which time production in many sectors had recovered to 1975 levels. (These were still well below 1970 levels.) The guerrilla war waged by the Mozambican Resistance Front (RENAMO), supported by Rhodesia in response to Mozambique's adherence to the UN sanctions following the unilateral declaration of independence in 1965, intensified after the independence of Zimbabwe in 1980, when RENAMO support was assumed by South Africa. The combined effect of the war, misguided development policies and a severe drought in 1982/83 were dramatic, amounting to a virtual collapse of the Mozambican economy. Over the 1981-86 period, GDP fell by an estimated 30 percent (and GDP per capita by about 45 percent); in real terms, GDP fell at an annual rate of about 6 percent. Exports declined by 75 percent. With fiscal and foreign exchange resources declining with the level of economic activity and escalating military expenditures3 the Government's capacity to provide even basic services became severely restricted. 14. The impact of the war on the agricultural sector was devastating. Insecurity in the countryside forced a massive displacement of populations toward safe areas near the cities, along the coast, and to neighboring countries. By 1990, there were about 1.0 million dead and close to 5.0 million displaced persons. A large part of the country's economic infrastructure (roads and railways, marketing depots and shops) had been destroyed. The number of private traders declined from about 6,000 at independence to fewer than 2,000 in 1990. The conflict also destroyed or forced the closure of about 1,000 clinics and health posts, and 3,000 schools. Some 400 teachers were killed. The destruction of the transport infrastructure and the pervasive insecurity cut producers off from markets for their output, inputs and essential consumer goods. Agricultural GDP fell by 5 percent p.a. in real terms. By 1986, cotton and sugar production had collapsed to 9 percent and 16 percent respectively of their 1980 levels. Production of tea and cashew, which together contributed about a third of the country's total exports, fell to 30 percent of 1980 levels. Livestock herds and wildlife were decimated. Although it is difficult to assess the total food production of smallholders, it is estimated that the overall food production may have fallen by as much as 30 percent between 1981 and 1986 as smallholders focused on ensuring their household food security and survival. The marketed production of maize and rice declined by about 50 percent. As a result, there was a severe deterioration of the food availability situation at the household level, both urban and rural. The country became dependent on food aid to meet the minimum needs of its population. 15. In 1983, at the Third Party Congress, the Government recognized that its policies were a contributing factor in the economic collapse and that, among the farming population, small family plots were the preferred and most efficient units. One of the first steps toward the liberalization of pricing and marketing arrangements was to free fruit and vegetable prices from administrative controls in 1985. This liberalization resulted in a substantial increase in In 1993, total recurrent budgetary expenditure for the Army, excluding expenditures in the Ministry of Security and Interior, were 280 billion meticais. This figure is one-and-one-quarter times larger than the combined current budgetary expenditures for health and education (221.0 billion meticais), and constituted some 24% of current budgetary expenditures and 5.1% of GDP. Military expenditure increased to 337.0 billion meticais in nominal terms in 1994 (compared to 284.7 billion meticais for health and education), but these now amounted to only 3.9% of GDP. 5 Box 1. Impact of the war on the environment The war has forced rural populations to seek refuge around the cities, and in narrow strips of land along the protected railway corridors and the coast, including the easily accessible islands. In these areas, population densities have vastly exceeded the carrying capacity of the natural resource base. For example, the Manica District had a population of more than 200,000 at the end of 1992, compared with fewer than 30,000 at the beginning of the 1980s. Soils have been heavily over-used and deforestation for fuelwood has led to serious soil erosion and the possibility of land slides during heavy rains. Coastal and island environments are being heavily degraded with wide-spread "dunification" and destruction of mangroves. Land conflicts are also frequent in the areas where displaced populations have sought refuge and try to eke out their minimum food requirements from the small plots allocated by local councils. On the other hand, the war forced the abandonment of economic activity from large tracts of land, which as a result of having been left idle may have recovered some of their fertility. In areas without active military operations, wildlife may also have been given a respite from the dangers of human advances. production, in particular in the relatively secure "Green Zones" surrounding the principal urban centers. The success of this initial experiment paved the way for the more radical reforms undertaken after 1987 to gradually move toward a market-based economy. D. Economic recovery 16. Mozambique has made significant progress in the management of its economy since 1987. Large devaluations have substantially reduced the overvaluation of the metical, and improved the overall competitiveness of the economy. There has been progress in liberalizing external trade and rationalizing the tariff structure. A major restructuring of the banking system has been initiated. Lending to non-performing state enterprises has declined significantly and increased resources should now be available to finance private investments. Great strides were taken to dismantle price controls, with the proportion of GDP subject to price controls dropping from 70 percent to less than 10 percent since 1987. However, fiscal deficits and inflation are still high and domestic savings and investments are negligible. Import, export and domestic trade licensing are still pervasive and there remains substantial room for improvement in the coherence and transparency of the trade regime. Government's intervention in the market remains pervasive as well. Price controls are still extensive: 20 agricultural products and most inputs are still subject to some kind of administrative prices. The producer prices of essential agricultural products are effectively depressed by Govemment's policies, by leakages in the distribution of emergency food aid, and by current policy with regard to the sale of "commercial" food aid. The monopsony situation enjoyed by the cashew agroprocessing enterprises up to mid-1995 imposed heavy costs on farmers. The ban on exports of the raw product has been scrapped in favor of a 20 percent tariff which is barely an improvement from the point of view of the peasant farmer at whose expense this unnecessary protectionist policy is pursued. The ineffectual regulation mechanisms applied to the cotton monopsonies (Jvcs) result in unnecessary implicit taxation of smaliholders. 6 17. Further efforts are required to consolidate macroeconomic stabilization and increase the amount and efficiency of investment. Macroeconomic management will need to focus on maintaining a realistic exchange rate and reducing fiscal deficits and inflation. Fiscal policy should enlarge the tax base, reduce tax distortions, improve collection, continue the rationalization of expenditures, and improve the management of foreign aid. Monetary policy should focus on the strengthening of the Central Bank, the improvement of adequate monetary policy instruments, the restructuring of the banking sector to restore monetary control and reduce the rate of inflation, and the establishment of an effective domestic payments system. Forceful public enterprise reform is urgently needed as certain of these enterprises are a drain on Government and banking sector resources. 18. Trade and market liberalization must be vigorously pursued. In particular, remaining Government involvement in the setting of agricultural prices must be removed. It is clear that in nearly all cases the interventions work to the disadvantage of smallholders even where they may be the intended beneficiaries. This is particularly true, for example, where floor prices in staple crops operate as defacto maximum prices at the farm gate. Interventions on food security grounds, and particularly the rules and procedures for acquiring and disposing of food aid, must be done in such a way as to create positive (rather than negative) price and market incentives for the agricultural sector. Reducing Government's intervention in the marketing system should be viewed as a complement to, and not a substitute for, investments in infrastructure, farmer education, etc. Farmers will respond more rapidly to incentives if the physical and institutional rural infrastructure is good. Policy reforms and investments need to go hand in hand if the desired agricultural supply response is to materialize. 19. Urgent measures must be taken to improve the private business environment. In spite of Mozambique's evident advantages and resources, the interest of investors, and of the much needed foreign investors in particular, cannot be taken for granted, as investors make a decision only after a careful comparison of how the business environment compares with that of other countries competing for their business. Mozambique suffers from a series of structural handicaps that hamper its ability to attract foreign investors. The lack of a skilled labor force, which reduces labor productivity and flexibility; the dearth of middle and upper level managerial talent; and a relatively poor physical (transport, storage, telecommunication) and institutional (banking and other services) infrastructure reduce Mozambique's attractiveness. 20. Removal of government intervention in those areas where it is counterproductive must not, however, prevent recognition of the legitimate and essential role of government in the promotion of long term sustainable private sector growth. Among the necessary conditions for growth which the private sector cannot be expected to provide under any reasonable circumstances are investments in public health, education, and the infrastructure needed to support a modern economy. In the agricultural sector the public sector is key to the development of not only educational systems but also research and extension to the smallholder sector which comprises the vast majority of Mozambican producers. 21. These constraints to development will continue for some time and will be corrected only slowly. They make it all the more urgent for the Government to correct the imperfections in the investment climate that are within its control or influence. Of critical importance are the current complexity and lack of transparency of the legal and institutional environment, which make it difficult to clearly grasp the rules of the game and to ensure that they will be consistently applied and upheld. This is particularly the case with respect to access to natural resources (mineral resources, tourism sites, arable land) for which long-term, undisputed access is necessary. Of equal importance, although more difficult to overcome, is the Government's still prevailing 7 ambivalence toward the private sector, which translates into a strong propensity to regulate private activity and foreign investments, with constant pressure for Mozambican interests to be partners of any new investment. Removing price and trade controls should allow private agents to compete with parastatals and among themselves on a fair footing, reduce costs, and improve farmers' incentives. 8 Chapter 2. CURRENT ECONOMIC FRAMEWORK A. The Economic Rehabilitation Program 22. In 1987, the Government launched a structural adjustment program - the Economic Rehabilitation Program (ERP) - which was supported by Bank policy dialogue and lending, and the IMF structural adjustment facilities (sAF and ESAF). The ERP, renamed the Economic and Social Rehabilitation Program (ESRP) in 1989 to emphasize the social aspect of the adjustment program, aimed at stemming the economic decline, reducing macroeconomic imbalances, and reducing the role of the state in production and in regulating private economic activity. On the macroeconomic side, the ERP/ESRP has been focused on creating a sound macroeconomic environment - a competitive exchange rate, sustainable budgetary deficits, and low inflation- favorable to private saving and investment. The Government is gradually pulling back from direct intervention in the economy and improving its capacity to provide basic services to private businesses. The price system was progressively liberalized; the exchange rate system has progressed from a severely overvalued official rate to a market-determined rate now; the trade and banking systems were streamlined, permitting the entry of two foreign banks; a policy of maintaining positive real interest rates was implemented; and the effort to restructure public enterprises resulted in the privatization, conversion to joint ventures, or leasing of 500 small and medium-sized public enterprises, concentrated in the manufacturing, commerce and fishing sectors. A new, more liberal foreign investment law was enacted in June 1993. 23. One fundamental shift in development strategy has been to view agriculture not as a backward sector but as the central engine of growth and the primary mean for reducing poverty. The new agricultural development strategy focuses on increasing incentives to production in the smallholder ("family") sector through the gradual liberalization of marketing and prices and improved access to productive assets and basic services. 24. Considerable progress has been achieved under the ERP/ESRP, as is shown in Table 1. After a 30 percent decline between 1981 and 1986, the GDP grew at an annual rate of 5.4 percent between 1987 and 1989. Economic growth slowed in 1990-91 and GDP fell by 2.7 percent in 1992 owing to the drought which severely affected agricultural production. In 1993, the Mozambican economy grew by 14 percent, the highest rate since independence, reflecting good rains and improved security, but also the liberalization of the economy and increasing demand for services. Per capita income in real terms increased for the first time since 1989. Growth rates continued to be buoyant, on the whole: 7.7 percent in 1994, 2.1 percent in 1995, and 8.7 percent in 1996. Agriculture has grown swiftly (21 percent in 1993, and 5.0-9.4 percent in 1994- 96). as increased agricultural marketing and improved security led to more commercial activity in the rural areas. Notwithstanding the gains achieved since 1987, Mozambique's economy continues to be hampered by less than satisfactory economic policies. Further work remains to be done to reinforce macroeconomic stability and complete the transition from a centrally planned to a market economy. 9 Table 1. Annual Percentage Changes in Real Gross Output, 1987-96 1987-89 1990 1991 1992 1993 1994 1995 1996 Agriculture 6.2 1.1 1.9 -7.5 21.3 5.0 6.9 9.4 Industry and fishing 7.6 -8.3 -0.5 -9.5 -6.7 -3.3 16.3 11.6 Construction -4.5 1.5 3.0 -1.8 7.0 7.5 7.0 8.0 Transport 2.1 19.3 13.2 13.6 16.7 5.8 13.3 7.6 Commerce 3.5 2.5 2.0 4.0 16.7 5.8 -12.1 7.4 Gross output 5.4 0.9 2.7 -2.7 14.3 7.7 2.1 8.7 ............................... ........................................................ . ...... ............... . .............. .................. . .............................. ............................................................... Agriculture's share of GDP - 38 36 31 31 29 31 32 Sources: 1987-94: Southem Africa Department (1995), Annex A.5. 1995-96: Republica de Movambique (1997). Shares from World Bank (1996a), Annex 1. p 1. aIncludes inputs (value added by sector not available) B. Economic Reforms and Performance 25. Fiscal policy. Under the ERP/ESRP, the Government attempted to reduce the country's fiscal deficits. The emphasis was on increasing fiscal revenues through expanding the tax base and improving collection perfonnance. The government succeeded moderately: fiscal revenues increased from 13 percent of GDP in 1986 to 20 percent in 1993, falling slightly to 18 percent by 1996. Total expenditures at first rose from 42 percent of GDP in 1986 to 51 percent in 1990, but fell by 1996 to 35 percent4 . Aid came to play an increasing role. External grants and loans were 3 and 5 percent respectively of GDP in 1986, and peaked at 21 and 9 percent respectively in 1994, before falling to 11 and 8 percent in 1996. 26. Inflation and the real interest rate. With increased fiscal and monetary discipline, the inflation rate was brought down from more than 160 percent in 1987 to 33 percent in 1991.5 But Figure I shows that inflation ballooned to 63 percent in 1994, before falling to 17 percent in 1996. The 1992-5 inflationary spurt was accompanied by a bubble in M2, which varied between 37 percent and 63 percent, before being forced down to about 22 percent in 1996. These percentages may be overestimated due to the likely underestimation of GDP. See the data series and sources in the Appendix Tables. 10 70 60 i 50 - 40 20 10 0 1990 1991 1992 1993 1994 1995 1996 F- 4 - M2 Growth Inflation Figure 1. Money supply growth and inflation, 1990-966 27. The real interest rates faced by agricultural borrowers were strongly negative in the 1980s and even in 1990 stood at -27 percent. Between 1989 and 1994 the bank rediscount rate was raised steadily, from 10 percent (1989) to 70 percent (1995). But the rediscount rate and the lending rates to enterprises were one step behind the inflation rate, so that the real interest rate faced by firms was most often a small negative. As the inflation rate fell, the real rediscount rate rose to 14 percent in 1995 but then fell to 8 percent in 1996. It is anticipated that real interest rates will continue to decline.7 28. Investment. Total investment increased from 46 percent of GDP in 1990 to a spectacular, and not really credible, 60 percent in 1994 and 48 percent in 1996. The increase is due to: (a) understatement of GDP (unofficial foreign trade, poorly measured agricultural production); (b) improving conditions for private investment; (c) increased donor contributions; and (d) confusion of investment with consumption, such as certain nonperforming loans, which were recorded as investment but are properly viewed as consumption. The inaccuracy of the data precludes any sentient judgments about the efficiency of the country's investments at the present time. 6 Sources: see the Appendix Tables. 7A further decline would be of especial importance for small borrowers Until recently the banks required them to pay the interest in full in advance, and repay their capital in equal monthly installments. With more frequent amortization and with higher listed interest rates, the effective interest rate increases exponentially. If in 1995 a small firm took out a loan at 40 percent (below 1995 inflation of about 55 percent) and had to pay the interest immediately and the capital in monthly installments, the effective interest rate was 178 percent, making the real interest rate 126 percent. 8 See the data series and sources in the Appendix Tables. 11 100 1 Real e.r. 9o 0 index (US$ per 80 local currency 70 unit) 60 1990 1992 1994 1996 Figure 2. The real exchange rate, 1990-969 29. Exchange Rate and Trade Policies. A remarkable achievement of the ESRP has been the substantial improvement in foreign exchange rate policy. The grossly overvalued official exchange rate was devalued from Mt 929/USD in 1989 to Mt 3,723/USD in 1993, and to Mt 11,378/USD in February 1997). The parallel market premium for foreign exchange decreased from an average of 2,100 percent in 1989 to 10 percent in late 1994, where it has remained up to the present. There proved to be enough price "stickiness" for the nominal devaluation to have real effects: the real exchange rate index depreciated from 100 in 1990 to 60 in 1993, though it subsequently rose to 69 (Figure 2). The international competitiveness of the Mozambican economy has improved markedly, incentives for rent-seeking activities have been curtailed, and the allocation of foreign exchange was moved from an administrative to a market-based mechanism. Trade reforms were undertaken. The Govemment rationalized the tariff structure, reducing the number and dispersion of rates and assigning these rates according to objective criteria. The tariff code was updated and simplified in 1991 and again in November 1996, so that there are now five rates of between zero and 35 percent. Export duties, including those for agricultural products (except cashew) were reduced to 0.5 percent and then to zero. To become an exporter one now has to do no more than fill in a simple form. For registration as an importer, however, there is still an array of requirements to be fulfilled (amounting to a check on creditworthiness), which slows down the process of licensing and creates scope for opportunistic action by the administrators. One remaining problem is that licensing procedures restrict entry and competition in the domestic trade of agricultural commodities. 30. Resource Mobilization and Growth Prospects. Gross domestic saving was negative in 1990, and has since become positive but remained low (7 percent in 1993, 5 percent in 1994, 12 percent in 1995). With lower inflation and positive real interest rates, and growth of disposable income, the savings rate is expected to rise. However, as remarked above, Mozambique's economic recovery program has been dependent on a substantial flow of foreign aid. The bulk of resources for domestic investment comes from the extemal sector. In 1996, official capital grants totaled some 19 percent of GDP (see para. 25). This dependency on foreign aid raises concerns about the prospects for growth and sustainability. Continued public investments will be needed to rehabilitate the country's basic infrastructure, and substantial private investments will be necessary to renew and expand an ageing productive capital stock. In the medium term, Sources: see the Appendix Tables. 12 however, the economy will not be in a position to generate sufficient domestic savings to sustain the required rate of investment, even assuming accelerated economic growth, nor the foreign exchange receipts necessary to finance essential imports. Mozambique will therefore remain dependent on foreign aid. But long-term economic growth will primarily be based on private investment and, given the low domestic saving rate, on a large inflow of private foreign capital. The importance of foreign investment is underscored not only by the severe constraints on domestic savings and the prospects of declining foreign aid, but also as a source of the technologies and management skills essential to growth that are in short supply. 31. Since external aid is set to decrease in the medium term, it is essential to ensure that it is used more efficiently. Improved management of foreign aid should bring about significant additional benefits. There are currently more than 500 donor assisted projects in Mozambique. Most are donor-driven and heavily dependent on long-term foreign technical assistance. Many are implemented outside of existing institutions. They crowd out local capacity-building. Project fragmentation also imposes a heavy burden on local technical and managerial skills. This fragmented, donor-financed approach needs to give way to integrated sector-based programs such as those in health and road-building. The program approach implies the preparation, by Government and local stakeholders, of coherent national policies and strategies and of sector-wide programs that can be supported by all donors. Examples of this are the Priority Districts Program (PDP), and the so-called Pre-Program. 32. Liberalization. Sustained growth requires a high rate of investment and an efficient allocation of available resources. This, in turn, requires getting the price signals right and creating a climate that allows businesses to respond to those signals. Macroeconomic reform provides the essential framework for private investment but the deregulation of domestic prices and markets is also necessary to create a supportive business environment. In 1987, the commercial sector was extensively state-owned and there were pervasive price controls. This sector produced according to the dictates of the national plan; large subsidies, amounting to 8 percent of GDP, were provided through the budget or via credit from the (state-owned) banking system. Since then, there has been a gradual price liberalization. The prices of a number of essential consumer and intermediate goods are still determined administratively, but only a few products and services are currently subject to a fixed price regime: transport, communications, oil products, utilities, housing rents, medical services and four basic food staples - maize, maize flour, wheat flour and bread. The other main crops are now subject only to minimum producer prices (and the minimum price for maize was suspended during the 1995/6 season owing to the glut in the southern African region). The Government wants to keep the consumer price of those food staples affordable to poor consumers. In the case of yellow maize, widely available in plentiful quantities from food aid, the consumer price has been kept significantly below border-parity price for humanitarian reasons. However, this has also depressed the price of (white) maize produced locally and has a negative impact on domestic production and farmer incomes. l 33. Business Environment. Improving macroeconomic policies and the rehabilitation of the country's physical infrastructure should provide a firm basis for rapid private sector development. Mozambique has assets that offer significant potential for private investments: its location next to higher-income land-locked countries and the availability of good ports (Maputo and Beira) with established transportation corridors to their borders; extensive foreign investor relationships with Mozambican interests both in the private and public sectors; extensive natural 0 Note that wholesale and retail margins uncontrolled, but the price of cotton ginning is set. 13 resources, including arable land, minerals, wildlife, timber, fisheries, and tourism sites. The attraction of these assets is demonstrated by the interest shown by potential investors, particularly in the mineral, energy, fishing, and tourism sectors. However, actual - as opposed to proposed - investment has been slow to materialize. The restoration, at last, of a credible, lasting peace; a new investment code, and the divestiture of the ubiquitous State enterprises should now remove the remaining obstacles to private sector development. The Investment Code of 1993 is a step in the right direction as it provides a straightforward incentive package and equal treatment for local and foreign investors. However, a number of cumbersome features still remain (registration, licenses, etc.) that raise costs and create uncertainty for investors. The rate of approval of foreign investments has increased since its introduction, but the increase is due more to the new-found political stability than to the effects of the 1993 Code. 34. Privatization has proceeded swiftly. Some 500 small-and medium scale enterprises have been privatized through ownership transfer, management contracts, conversion into joint ventures, or leasing arrangements. In addition, 29 large-scale companies have been privatized since the state firm privatization legislation was passed in 1991. Almost all offers in the privatization of large-scale companies have received an acceptable private sector response. On the other hand, the State continues to play a role in some sectors (transport, banking ) and certain large state-owned enterprises such as the electricity company continue to impose a heavy drain on Government's budget. In the agricultural sector, practically all of the 82 state farms covering 520,000 ha have been privatized defacto; they have either been officially distributed to private interests, or been taken over by former employees, or been reclaimed by local populations with traditional use rights. To date, only a fraction of these have been divested de jure. Nearly all the major agro-industrial parastatals, including the 14 cashew processing factories, have been privatized. Only a few sugar estates (e.g. Incomati Sugar) remain in state hands, and these are in the process of being sold. 35. There remain, however, a number of constraints on private investments in the agricultural sector, in particular for investors outside the smallholder sector. Mozambican property law remains restrictive. According to the Constitution, all land belongs to the State. The law permits the State to lease land for up to 50 years, but does not allow for freehold ownership. The right of private parties to transfer leasing rights and title to the buildings established on the land is subject to discretionary approval. In addition, traditional rights apply to all land outside of the urban areas, and leases granted by the State to private investors without prior consultation with traditional right-holders do not offer long-term tenure security to the lease holders. Thus the incentive of private entrepreneurs to invest and their ability to mobilize credit is reduced. The low level of land taxes, and the even lower level of their actual collection, has also led to growing pressure from Mozambican nationals to have (low-cost) access to agricultural land, often not for productive but for speculative purposes. 36. Foreign Private Investment. The Center for Promotion of Investment (CPI) reports that since 1985 it has authorized projects representing nearly $1.2 billion, mainly in agriculture, agro-industry, tourism and mining. Though there is inevitably a gap between approved and actual investments, the gap has been particularly large in the case of Mozambique. This reflects uncertainties related to the peace process, as well as difficulties in the legal and regulatory environment for private investors. One of the state banks has been privatized and another is in the process of privatization. 14 C. Agriculture and the Macroeconomy 37. During the first half of the 1980's, agricultural output fell sharply as a result of the spreading armed insurgency, rigid administered prices, and the institutional neglect, if not repression, of the smallholder sector. The latter, without access to basic services, and heavily taxed by a vastly overvalued exchange rate and official producer prices, was also increasingly cut off from markets to sell their products or buy essential consumer goods. As a result, the marketed production of smallholders plummeted as they adopted a strategy aimed essentially at food self-sufficiency and survival. Under the ERP, Government strategy to revitalize the agricultural sector has been essentially based on the improving macroeconomic enviromnent and the gradual liberalization of the pricing and marketing policies. 12 38. Table 2 presents growth rates of marketed agricultural output for the period 1989-1995 Subject to the natural variability imposed by the rainfall regime, it took some time for the liberalization and reforrns under the ERP to become wide and deep enough to make a difference. After a satisfactory 4.0 percent growth rate in 1989, agricultural production grew at only 1.1 percent in 1990, and dropped by -4.0 percent and - 11 percent in 1991 and 1992, respectively, mostly on account of the extended drought. When rains returned to nornal in 1993, marketed output by the sector grew by a substantial 21 percent; it grew by 5.0 percent in 1994, 6.9 percent in 1995 and 9.4 percent in 1996. The behavior of the crop sector is very similar to the overall sector pattern. Table 2. Percentage Growth Rates of Agricultural Output, 1989-1996 2 1989 1990 1991 1992 1993 1994 1995 1996 Crops N/A N/A N/A -11.2 22.7 4.9 7.0 9.8 Livestock N/A N/A N/A -13.5 3.0 7.8 4.3 3.3 Sector 4.0 1.1 4.0 -11.3 21.4 5.0 6.9 9.4 Source: National Plan Commission, Revised February 2, 1995; Republica de Mocambique (1995b), pp. 24, 29 and 3 1; and Republica de Mocambique (1997), table "Quadro Global da Agricultura e Pecudria". a Production of the commercial sector plus marketed production of the smallholder sector plus "autoconsumo" or non-marketed production of the smallholder sector. 39. Marketed volumes have grown faster than the totals mentioned in Table 2. For instance, 13 marketed food crop output grew by some 34 percent in 1996 . Marketings of export crops, on the other hand, declined in 1993 (-19 percent, led by a dramatic -56 percent fall in cashew nut 14 production), but recovered in 1994 (21 percent), in 1995 (7.5 percent) and in 1996 (23 percent) The leader was cashew, which grew at 14 percent in 1995 and 99 percent in 1996. Copra performed well in 1993 (39 percent) and 1994 (27 percent), but then declined (5.2 percent in 1995, -16 percent in 1996). Sugar production grew swiftly, at 34 percent in 1995 and 49 percent in 1996. 12 Estimates of total agricultural production are presented in Chapter 3. 13 Republica de Mocambique (1995b), p. 30, and RepuTblica de Mocambique (1997), p. 6. 14 Republica de Mocambique (1995b), p. 29, and Republica de Mosambique (1997), p. 6. 15 Table 3. Regional Distribution of Production in 1992/93 (tons) Region Maize Cassava Sorghum Rice Beans Northern 181 2,480 77 22 46 Central 287 723 59 31 19 Southern 65 313 7 13 15 Total 533 3,500 143 66 79 Source: National Plan Commission, Revised February 2, 1995. Regions: Northem: Cabo Delgado, Nampula, Niassa; Central: Zambezia, Tete, Manica, Sofala; Southerm: Inhambane, Gaza and Maputo. 40. There is little doubt about the strong short-run response of agricultural production to real price increases. The real price of beans rose 100 percent between its nadir in 1986 and 1992, and the marketed output of beans increased correspondingly by about 300 percent. The real price of maize increased by 50 percent between 1986 and 1992, and marketed output increased by about 1 5 100 percent . This supply response was achieved in spite of punishing structural physical and institutional constraints. Indeed, the removal of physical and institutional barriers to production and resource flows have received much less attention than the improvement of price incentives and the dismantling of administrative controls. There are several reasons for the heavy emphasis on price incentives. First is the sheer magnitude of the distortions between administered and market prices, and the fact that it was in Government's power to decide the changes. Second, the war prevented the Government from undertaking any significant program of infrastructure construction or maintenance, or the private traders from reestablishing their network and stores, in most of the countryside. Finally, the stabilization program imposed a severe constraint on Government's public expenditures and investment programn. 41. Figure 3 presents the evolution of productivity of the agricultural sector as a whole. The results must be interpreted with extreme caution as tne numbers - agricultural GDP, the CPI and the rural population - come from different sources and cannot be regarded as any better than rough estimates. It appears that the productivity of the sector increased somewhat between 1987 and 1989, as one would expect, since the agricultural markets were being freed up during that period. Since then there have been fluctuations, for instance a dip in 1992 owing to the drought, but no obvious trend. Similarly, in the case of foodgrain production per rural inhabitant, there was a deep reduction in the 1992/3 production season. Preliminary figures for the 1995/6 season, in which the rains were excellent, suggest further increases in per capita foodgrain output. 5 See the discussion about prices and incentives in Chapter 5. 16 60000 - 70 U, 0 ~~~~~~~~~~~~60 ~ 50 _ 40 & 250000 - 30 0 ~~~~~~~~~~~20 20 40000 0Co 1987 1988 1989 1990 1991 1992 1993 1994 1995 Figure 3. Agricultural sector productivity: Agricultural GDP in constant 1990 Meticais per rural person, and foodgrain production (kg) per rural person16 42. Further progress in the short tern requires the pursuit of incentive reforns and the removal of the most binding physical and institutional constraints: the rebuilding of the rural physical infrastructure; the restoration of security and the resettlement of displaced population in productive activities in rural areas; and the effective liberalization and decentralization of economic decision-making. In the longer term, sustainable development will require establishment of a viable national transport system for rural/urban movements of food and industrial crops as well as marketing inputs and consumer goods in return. Establishment of a national research and extension system will also be key. Exports 43. Total registered exports of all sectors (agriculture and other) grew from $105 million in 1989 to $226 million in 1996, with an average growth rate of about 12 percent p.a. Agriculture and renewable natural resource exports constituted 76 percent of all exports. Prawns constitute the single largest export item in Mozambique, accounting for over a third of total registered exports (Figure 4). 16 Sources: Population from UNICEF, total GDP from International Financial Statistics On-Line, ratio of agricultural GDP to total GDP from World Bank (1 996a), cPI from International Financial Statistics On- Line. The official exchange rate for 1990 (period average) was Mt929/$, so that the productivity levels are about $50 per rural inhabitant. Foodgrain production (white maize, rice, sorghum and millet): from Tschirley et al. (1996), Table 1, p. 4; the table year "1992" refers to the production season April 1992 to March 1993. 17 See further detail in the Appendix Tables. 17 Note: Total exports in 1996 Shrimp & were US$226 million. Prawns 35.5% Miscellaneous 31.7% Petroleu ~Cashew Nuts 1.2% Copra Sugar Cotton 19.2% 1.1% 5.7% 5.7% Figure 4. Commodity composition of exports, 1996 (percentages) 18 44. Export crops contribute some 33 percent of total exports (Figure 5). The largest export crop was cashew nuts which comprised almost one half of total exports up to 1992. The sub- sector declined dramatically in the subsequent two years, falling to just two percent of total exports in 1994, but rose to 19 percent of exports in 1996. The second largest export crop is cotton, with an average value of exports close to 10 million dollars in the period 1989-1996. Sugar is in third place, with average annual exports valued at 13 million dollars in 1996. Copra and citrus are the other exported crops of any importance. Timber exports accounted for 8 million dollars in 1996, though this figure is likely to be significantly under-reported. Crop exports without cashew nuts grew, albeit irregularly, from $18 million in 1989 to $32 million in 1996, or some 8.4 percent p.a. on average. 240 200 0 Other U Petroleum 160 mill. 120 _____ Cotton so S~~~~~~~ Cashew Nut 4 I . I {o w I | | l _ 1 g s1 Prawns _ _ 1l I I £ 1-||iRl 1989 1991 1993 1995 Figure 5. Commodity exports, 1989-96 19 cotSource: Bank of Mozambique, Ministry of Commerce: Planning Commission, January 1995; and 1995-6 data from Bank of Mozaibique statistics. See detailed data and sources in the Appendix Tables. 1 8 The shrimp and prawn industry20 45. Whereas subsistence fishing has always been present, large-scale shrimp and prawn fishing dates back only to the late 1950s, when commercial operations commenced under the colonial order. All activities within the sector (i.e. fishing, the service industries, sales and distribution) were managed by the Portuguese or other foreign interests. The collapse of the sector was therefore inevitable when the Portuguese emigrated massively after independence. The Mozambican rejuvenation of the fishery sector began in 1976, with the formation of a fisheries administration within the Ministry of Industry and Commerce. In the early 1980s, the fisheries administration implemented a basic reorganization of the sector, which was reflected in steadily increasing catch figures. In 1993 the licensing of foreign firms was phased out, and all fishing firms operating along the Mozambican coast are now national firms or joint ventures (at least 51 percent Mozambican). 46. Shrimps and prawns account for 80 percent of the fish production of Mozambique, and provide around 50 percent of export revenues. Output has fluctuated in the last five years at around 10,000 metric tons, while the exports in value have grown by 40 percent between 1990 and 1995 (Table 4). Mozambique's exports account for about one-half of one percent of world shrimp and prawn exports. Table 4. Output of the Mozambican shrimp and prawn fishing industry 1990 1991 1992 1993 1994 1995 Production (MT) 10,539 11,249 10,018 10,990 9,500a 10,200' % of world production 0,41 0,40 0,35 0,38 N/A N/A. Value of exports ('000 us$) 50,349 62,981 65,985 70,386 N/A 84,000^ Sources: FAO (1993), unless otherwise indicated. Note that these figures, and in particular the export ones, are underestimated, owing to clandestine trade. a From: Estimates by Luigi Marcuccio on the basis of the data available in Republica de Mocambique (1995b). 47. The Secretaria de Estado das Pescas (SEP), since 1985 directly responsible to the Government, is in charge of the overall organization and development of the sea fishery sector, while the freshwater fisheries are under the Ministry of Agriculture. The SEP coordinates the Instituto de Investigacao Pesqueira (IIP) and the Institute for the Development of Small Scale Fisheries. The IIP, headquartered in Maputo with branch laboratories in Beira and Quelimane, was created 15 years ago as a Portuguese "overseas mission". All its staff have a thesis degree acquired after five years of study in general biology, with a marine biology orientation in the final year, and some have received training in Cuba and Norway. The IIP conducts and coordinates research in the fields of marine and fisheries sciences, and concentrates on the large- Source: Bank of Mozambique, Ministry of Commerce: Planning Commission, January 1995; 1995 and 1996 data from Bank of Mozambique tables. See detailed data in the Appendix Tables. The category "Other" includes tea, copra, citrus, coal, and miscellaneous items. Sources: Repablica de Mocambique (1995b), FAO (1990), FAO (1993), Economist Intelligence Unit (1996), Shrimp Notes: A Market News Analysis, Hernandez (1991), and World Bank (1991). 19 scale sector. The Institute for the Development of Small Scale Fisheries, whose mandate is mainly one of extension and coordination, has not followed a comprehensive plan, nor has it succeeded in coordinating the various projects implemented in different parts of the country in, for example, fish processing and community development. Finally, there is a Department of Biology at the Eduardo Mondlane University, that undertakes teaching and research activities in marine and fisheries biology. 48. The industry has a dualistic structure. A few large enterprises account for around 97 percent of the total catch. EMOPESCA, a parastatal, is responsible for the large shallow-water shrimp industry and accounts for around 30 percent of the total catch. Other important enterprises include PESCAMAR, a joint venture between Mozanbique and Spain, EFRIPEL, a joint venture between Mozambique and Japan, and MOSOPESCA, a joint venture with Russia. These enterprises are equipped with 50 m trawlers which have deep freezing facilities, and cater exclusively to the export market, in particular to Japan, Spain and France. Some 40,000 fisher families account for the remaining 3 percent of domestic production. They produce both for their own consumption and for the domestic market. Their fishing boats are three to eight meters long, most of them unrechanized, but often equipped with sails. 49. The industry is viable and growing. Regarding the export market, Mozambican shrimps and prawns can access the EU market at preferential prices and without duties and quotas: in 1995 the average export price was us$9 for prawns and us$ for shrimps. The world market prices for these commodities are expected to move upward, because of the us embargo of exports from countries which are not equipped with turtle excluder devices, which began effective as of May 1, 1996. Mozambique, although it is subject to the embargo, should not be particularly affected by it, since its prawns and shrimp exports to the us were negligible with respect to total: however, the other 34 embargoed countries could re-direct their exports to Japan and Europe, thus increasing the competition in these markets. The lack of infrastructure is currently the major hindrance to the growth of the domestic market. From the supply side, recent assessment indicate that, while the current production of prawns cannot be increased in the short run, shrimp production is at the moment only 37 percent of the total sustainable catch. There are good prospects also in shrimp and prawn farming: there are more than 170,000 hectares of tidal swampland, larvae are locally available, there is a fair amount of skilled labor, and the ecological and climate conditions are appropriate. Under a UNDP/FAO pilot project for coastal shrimp culture, 33,000 hectares have been identified as potential sites for shrimp farms, with an anticipated production increase of a 1,200 metric tons and the creation of 80,000-90,000 jobs, taking into account the indirect effects. The French Ministry of Cooperation has been involved in two pilot projects (one near Beira and the other near Quelimane) financed by the Caisse Fran,aise de Cooperation Economique. The establishment of shrimp farms could also encourage the establishment of a canning industry, which the Government is trying to encourage through various incentive schemes and tax breaks. 50. The Government intends to achieve the objective of a growth rate of eight percent in the shrimp and prawn industry in 1996, and to increase the labor absorption capability of the industry, which currently accounts for less than 2% of the agricultural labor force, by targeting artisanal fishermen. Announced measures include: * strengthening the Fundo de Fomento Pesqueiro ability to cater to the credit needs of artisanal fishermen; * rationalizing the legislation in the industry by issuing a Regulamento da Pesca Maritima and revise the sanitary regulations; 20 * privatizing the parastatals involved in the fisheries sector; and * revising the quota system, which discriminates against artisanal fishermen. 21 Chapter 3. THE RURAL ECONOMY A. Natural Resources 51. Mozambique has a rich natural resource base. Its total land area extends over 786,389 sq. km with an additional 13,000 sq. km of inland waters, including part of Lake Malawi. It has an extensive coast line of more than 2,000 km on the Indian Ocean and borders Tanzania, Malawi, Zambia, Zimbabwe, South Africa, and Swaziland. A large, flat coastal plain - wider in the Southern part of the country - rises steadily to the mountainous areas to the West where altitudes reach up to 1,000 meters. The climate of Mozambique is influenced by the inter- Tropical Convergence Zone and the trade winds from the Indian Ocean. Average annual rainfall ranges from 400-800 mm in the South, which suffers from prolonged drought periods, to 1,000- 1,500 mm in the North, which generally benefits from a well-defined and reliable rainy season. Mozambique's river systems flow generally east to the Indian Ocean. Three major systems, the Zambezi in the center of the country and the Limpopo and the Save further to the South, drain the Central African plateau. In the Northern part, the Rovuma and the Lorio drain the area to the East of the Rift Valley. The soils are generally of mediocre quality. Those with the better agricultural potential are found on the plateau and highlands in the North and West. Except in the extensive alluvial plains along the main rivers, the soils of the low-lying coastal plain are sandy and of low fertility. Their poor water retention capacity accentuates drought risks in the South. The predominant vegetation of the plateau in the Northern and Western parts of the country is an open woodland, or Miombo forest, which is prone to tse-tse fly infestation. The Miombo landscape is interspersed by patches of denser forests, usually at higher altitudes, in Niassa, Manica and Sofala provinces. In the South, the vegetation becomes increasingly open and the Miombo gives way to open, dry savannah, free of tse-tse infestation and offering pastures of relatively good quality, but often with limited water availability. 52. Agro-ecological Zones. The climate and physical features of Mozambique divide the country into two main agro-ecological zones, north and south of the Save River. More intensive cultivation is possible in the north because of higher, more regular rainfall and better soils. In the drier South extensive agriculture prevails, except along the coast and in pockets of irrigated land. 53. Most Type 1 areas (intensive and diversified agriculture) are in the North, concentrated in pockets with good rainfall and soils in part of Niassa, Nampula, Manica, and upper Zambezia provinces. These areas have a wet upland tropical climate and can support high value cash crops such as coffee, tea and tobacco as well as food crops such as maize, sorghum, rice, and cassava. Small pockets of Type I areas can also be found in the South along the border with Swaziland and South Africa, where citrus and other fruits can be produced and dairy farming was once important. The smallholder farmers in Type I areas typically practice slash-and-burn agriculture, with some inter-cropping with legumes. Cotton and cashew are also grown in these areas. Nampula and Cabo Delgado provinces are the main cashew producing areas in the north. The North has limited livestock potential due to heavy tsetse fly infestation. 54. Areas of intensive and semi-intensive agriculture (Types 2 and 3) are well represented in Zambezia, Sofala, Niassa, Nampula and Cabo Delgado provinces, and along the coastal strip in the south. These areas can support a wide range of rainfed crops. Areas of semi-extensive and 23 extensive potential (Types 4 and 5) are prevalent in most of the South (Maputo, Gaza, and Inhambane provinces). (a) Land 55. Table 5 suggests that Mozambique has ample cultivable land resources to accommodate the needs of its current and projected future population. The average population density is 20/sq. km and that of the rural population 16/sq.km. With an arable area of about 36-40 million ha, depending on the criteria used, and about 3 million farm families, there are some 12-13 ha per family. These are aggregate figures, however; they conceal important regional differences in land availability and in market access. Table 5. Land Potential and Current Use Land Resources Percentage of (104 sq.km) Land Area Total Country Area 79.9 Inland water Land Area 78.6 100 Of which: Urban 0.8 1 Parks and Reserves 10.1 13 Unsuitable for Agriculture 15.6 20 Forestrya 16.2 2 1 Cultivable 35.9 46 Source: Ministry of Agriculture. ' Suitability for agriculture not determined. 56. Table 6 indicates that, although there is ample unused cultivable land in the aggregate, most of it is concentrated in the Central and Northern area of the country. In the three southern provinces - Maputo, Gaza and Inhambane - there is already a land constraint, under the assumption that about 4 ha of fallow are needed per hectare of food crop to maintain soil fertility under traditional farming practices. In these areas, traditional long fallow farming systems are no longer able to sustain soil fertility. The farmers' own response to the degradation of soil fertility is not documented. Since without technical innovation yields should start to decline soon, these areas should be targeted for urgent strengthening of extension services. In addition, these provinces are currently the object of strong interest on the part of commercial farmning and speculative interests, domestic and foreign. Owing to the deficient property rights regime (see Chapter 4), there are conflicts between smallholders, the traditional rightholders to the land, and the promoters of new farming ventures. Another area of concern is Nampula, where the only unused land available is some 1.6 million ha in "severely limited" areas. 57. In the short term, measures are needed to avoid an aggravation of these conflicts and protect the rights of smallholders. The Government could direct new investors toward areas where land is readily available. The provinces of Niassa, Cabo Delgado and Tete have the 24 greatest amounts of land avaiiable, with 8.5 million, 3.9 million, and 5.0 million hectares respectively.2" Hence in the long term, an "extensification" strategy is possible. But most of this land is far from town centers, roads and other infrastructure, and is not available for immediate development. Therefore, the conventional wisdom that Mozambique has ample cultivable land resources must be put into a more proper time perspective; even though there are ample land resources, not everyone has access to unlimited land in the short term. In addition, the southern three provinces as well as Nampula province present an immediate problem of non-sustainable use which has the potential for adverse long run consequences. Table 6. Current Land Availability Province Smallholdings Area Required Cultivable Area Available Land ('000) (million ha)a (million ha) b (million ha) Niassa 186 1.2 9.7 8.5 C.Delgado 308 1.7 5.6 3.9 Nampula 720 4.2 5.8 1.6 Zambezia 736 2.8 7.9 5.1 Tete 205 0.1 6.2 5.0 Manica 118 0.7 3.8 3.1 Sofala 238 1.4 2.6 1.3 Inhambane 305 2.2 1.0 0 Gaza 223 1.2 0.4 0 Maputo 25 0.6 0.7 0.05 Total 3,010 17 44 27 Source: Moll (1993), in tum from the Early Warning System and other sources. a Area required: cultivated + fallow estimated at 4 ha per ha of food crop. b Cultivable land is the sum of land types I, II and II determined using criteria followed by USDA (United States Department of Agriculture): class I = few limitations, II = moderate limitations, Ill = severe limitations. Hence the discrepancy between total cultivable land of 44 million ha in this table and the 36 million ha in Table 5. 58. Landmines. Over a period of 20 years, in the course of the war of independence and the subsequent civil war, landmines became an important part of the strategy of all the warring parties. They were frequently placed along road and access ways. The number of mines placed is unknown.22 Worst affected were Maputo, Zambezia and Tete provinces, as well as the western parts of Cabo Delgado and the central parts of Nampula.23 The number of people killed or maimed is large but is not known with any precision. The number of landmine incidents peaked at about 300 in 1994, as people were returning home from internal or external exile; the number of incidents fell to about 200 in 1995 and to double digits in 1996, as the return of the exiles was completed. The decline in incidents was also due to the process of mine removal, and to a 21 On the assumption that "cultivable land" is land of USDA types I, II and III. These figures would be 5.8 million, 2.7 million and 2.6 million ha respectively if only USDA types I ("few limitations") and II ("moderate limitations") were considered. 22 Estimates of the number of mines ranged as high as two million. See UNDP (1 996), p. 1. 23 See the maps in UNOHAC (1994), p. 24f. 25 learning process as communities identified the most dangerous areas. Apart from the toll in human suffering, the landmines have reduced the effective supply of arable land. In some cases, entire communities were deterred from retuming to their traditional lands until the demining was effected. The UNDP/DHA Accelerated Demining Program had, as of July 1996, cleared 6,145 km of roadways, and in so doing had destroyed some 40,000 anti-tank, anti-personnel and anti-group mines24. At the present rate of mine clearance, the North of the country will be practically free of landmines within three years, and the South in five years. Among the areas which urgently need mine clearance still are the interior of Zambezia and the western part of Maputo province. (b) Irrigation Potential 59. For the country as a whole, preliminary evaluations (FAO/INIA, 1980) indicate a potential of about three million hectares suitable for irrigation. These estimates are based on a limited number of soil surveys and are very approximate; data need to be improved by systematic studies. In some cases (Maputo and Gaza provinces), on-going studies seem to indicate that these figures may in fact have to be substantially reduced. Despite their imprecision, these estimates indicate that Mozambique has considerable irrigation potential. The greatest part of the irrigable land is concentrated in the central and northern parts of the country, with the Zambezi river basin representing about 60 percent of the total. The Southern part of the country (Maputo, Gaza and Inhambane Provinces), which has the greatest need of irrigated agriculture on account of its low and erratic rainfall, has less irrigation potential: there are fewer than 300,000 ha of potentially irrigable land and this figure may well be in excess of the possibility of use for agriculture of available water resources when requirements for urban consumption (in particular those of Maputo) are taken into consideration. Table 7. Land Suitable for Irrigation (ha) Watershed Irrigable Area Developed Area Irrigated Area Maputo 2,000 500 150 Umbeluzi 33,000 2,000 1,800 Incomati 110,000 18,000 5,000 Limpopo 150,000 40,000 24,000 Save 220,000 400 150 Buzi 36,000 6,100 1,700 Pungoe 77,000 9,700 600 Zambeze 2,200,000 12,000 1,600 Other North Basins 600,000 5,300 2,500 Total 3,428,000 93,700 37,500 Source: FAO/INIA, 1980, MacDonald et a]. (1990), Table 2.3, page 2-7, and Mihajiovich and Gomes (1986). 24 See UNDP (1996). 26 60. At present, there are about 93,700 ha equipped for irrigation, most of it under large (>I OOha) systems, including 36,000 ha in sugar estates. Therefore, less than 5 percent of the total irrigable potential is currently developed. The performance of the large perimeter sector has been extremely poor: only 40 percent of the equipped area (37,000 ha) is actually irrigated due to security problems and the poor operation and maintenance of infrastructure. Of the 36,000 ha of sugar schemes, only 5,500 ha are being operated due to the low standards of existing infrastructure, the lack of organization and organization and acute shortages of equipment, inputs and skilled manpower. Almost none of the existing dams benefited from prior environmental impact assessments; there are worries about the sustainability of these dams, since their impact on the environment (salinization, salinity intrusion, loss of floodplains, mangrove destruction) is in many cases becoming greater than, or a threat to, the benefits they bring. 61. The estimates of irrigation potential may need to be revised sharply downward, as Mozambique's rivers are being used extensively by South Africa and Zimbabwe upstream, so that several of the rivers are drying up. There is very little water in the Olifants River (which joins the Limpopo on Mozambican soil). The Rio Maputo is backed up with 30 km of saline water. Furthermore, there is a backup of 85 km in the Pungoe (Sofala province).25 This is an important international issue on which an accord should be sought; donors could play a role in bringing the parties to the bargaining table. 62. In light of Mozambique's experience with large scale irrigation, embarking on new investment in major irrigation schemes would best be avoided. Large scale irrigation is very capital intensive and requires a long gestation period. It also relies essentially on public investment and requires sophisticated management to achieve the expected increase in productivity. Where possible, the rehabilitation of large scale irrigation infrastructure would provide much higher and prompter returns. However, even such rehabilitation should be considered with caution. It still represents substantial, essentially public, investments and requires sophisticated management. Furthermore, some of the irrigation schemes have been re- occupied by smallholders - sometimes the original land users before they had been removed in the process of collectivization - and coordination among these large numbers of people, many of whom lack the financing and skills to run a sophisticated large-perimeter scheme, may not be economically worth while.26 63. The development of the country's large irrigation potential should essentially be based on simple, low-cost small-scale irrigation technology (run-of-the-river schemes, small pumps taking water from the extensive network of rivers or wells). This approach would not only be much cheaper; it would also rely essentially on private investment and management.27 64. There are already many initiatives in small-scale irrigation in Mozambique. In Manica province, for example, irrigation is based either on the derivation of water by (groups of) farmers directly from the rivers (Sussundenga District) or on the use of small earthen dams scattered throughout the central part of the province. The total number of dams is 192, but only a third of these are in good condition. Their average irrigation potential varies from three to five 25 Personal communication, Robert Clement-Jones. 26 Careful study is needed to determine which of the schemes are worth rehabilitating. Among those that have been re-occupied are the Lower Incomati-Palmeira-Marracuene are and the Xai-Xai area (sOGREAH 1993, Annex A9-2, p. 6 and p. 1 1) 27 No studies comparing the costs and benefits of small-scale versus large-scale irrigation could be located. 27 hectares.28 In the Zonas Verdes (Green Zones) around the cities of Maputo, Beira and Quelimane there are highly successful smallholder-operated irrigation systems operating on a drainage basis. 65. The many river valleys with peat deposits in Mozambique (machongos) present significant opportunities for smallholder development. Though these systems require careful and regular drainage, there is a tradition among the peasantry in their management. Large investments in drainage could, however, only be supported by the cultivation of high-value crops, which in tum will have to await broader changes in the entire agricultural, marketing and financial system. (c) Livestock 66. The contribution of livestock to the national economy was always relatively low. Even in 1980/81, which represents the peak level of livestock population after independence, production did not exceed 5 percent of total agricultural production. It then substantially decreased over the last ten to fifteen years due to the joint effects of the war, severe drought periods, and the collapse of the State farm system. It has been estimated that, between 1982 and 1992, the national herd decreased by 80 percent for cattle, 60 percent for small ruminants (whose number is likely to be substantially under-reported), and 30 percent for pigs. In the same period, average per capita consumption of animal protein has decreased by 46 percent, and the contribution of domestic production to it has fallen from 70 percent to 36 percent; present per capita consumption appears to be very low (about 1.2 kg meat and 2.4 kg milk per year). 67. Aggregate measures of livestock ownership in the smallholder sector, for eight selected districts, are presented in Figure 6. The measures are the sums of indices determined on the basis of local prices; the details are given in footnote 30. There is tremendous inequality in livestock holdings. The mean in the district of Massingir (Gaza province) is 586 points, roughly the equivalent of one head of cattle and nine pigs, whereas the median is only 2, the equivalent of two hens. Most households do not practise animal husbandry at all.29 The exception to this is in the district of Panda (Inhambane), where livestock production is practised on a small scale by most of the population, with roughly a third keeping one to five goats and or pigs, and a small number practising commercial livestock production on a larger scale. The median livestock ownership index for the eight districts amounted to less than a goat. 28 Bastiaansen (1991), p. 13. 29 These findings are broadly in accord with the national data from Ministerio da Agricultura (1994). See the Appendix Tables for further detail. 28 1000 100 10 Median I , 0 0 C Figure 6. Livestock ownership indices, eight selected districts, 1995 30 68. The national herd is mainly concentrated in the Southern area. As with crops, the smallholder sector is predominant; it is estimated to have 75 percent of cattle, 93 percent of small ruminants, and 60 percent of the national pig herd. The smallholder sector's low productivity is due to the production systems utilized (agro-pastoralism and mixed farming systems) and poor sanitary, feeding, and management conditions. 69. Mozambique has extensive range resources, which offer a major potential for future expansion. Roughage is abundant, and the significant crop potential should contribute a large amount of crop by-products as well. The main sources for concentrate feed (oil extraction cake, copra, cotton, groundnut, molasses and citrus pulp) face competition from export possibilities. The potential to increase meat output is important, as very few inputs are currently used. The good genetic potential of the existing breeds for meat production and animal traction constitutes a significant advantage. Local cattle, goats and sheep are well adapted to the natural conditions, including periodic droughts. 70. Three important epidemic diseases (Rinderpest, Contagious Bovine Pneumonia, "Peste des Petits Ruminants") are not present in the country. Tsetse fly infestation and ticks present the most serious sanitary constraints to cattle development, while pig rearing is hampered by the African Swine Fever; all three are mostly prevalent in the Central and Northern parts of the country. Tsetse infests approximately two-thirds of the country and evidence suggests that a southward advance is in progress. Tick-borne diseases are of particular importance to exotic breeds and their crosses. Veterinary services used to be provided free of charge to the smallholder sector, but reduced security in the r-ural areas, the destruction of infrastructure and vehicles, and budget constraints have led to their deterioration and neglect. Current budgetary constraints make it unlikely that such services can be restored in the required scale. 71. Animal and meat marketing used to be controlled by GAPECOM, a parastatal. Private traders are beginning to emerge as marketing is liberalized. The main problems today are the 30 Source: Mddecins Sans Fronti&res (1996), Table 4, p. 55. Index weights, based on local prices: Poultry 1, goat = 10, pig = 20, donkey = 50, head of cattle = 400. Data collected in November 1995. Massingir & Chigubo (Gaza), Mabote & Panda (Inhambane), Guro & Macossa (Manica), Milange (Zamb&zia), Mecula (Niassa). 29 decrease in the animal herd, and frequent thefts and illegal slaughter stimulated by the high price of meat. Milk collection and processing was completely disrupted by the war. 72. In terms of resource availability and market potential, the first priority should be given to cattle, with the triple objective of meat and milk production and animal traction. The limited number of reproductive females is now the binding constraint for livestock development in Mozambique. The high price of meat mentioned above is a powerful incentive for improving stock quality and reproductive efficiency. In the short term, efforts could also be placed on small ruminant and pig development, which have a faster rate of herd growth and benefit more directly the rural poor, including the poorest segments of Mozambican society. (d) Forests and Wildlife3 73. Forests. Forested areas cover about 55.0 million ha, or 70 percent of Mozambique's total land area. Out of this total, about 10 million hectares (18 percent) are classified as protected areas (parks and reserves) and about 20 million hectares (36 percent) are presumed to have some timber production potential. There are about 500,000 ha of mangroves along the coastal line. Some 19 million hectares are estimated to have commercially marketable wood. A national satellite inventory of biomass resources, plus three detailed woodfuel inventories around the major cities to assess the extent of the forest cover and locate areas with significant productive potential were recently completed. Extensive ground truthing and field work verification of the satellite image interpretation was conducted. These results will allow the Government to focus its efforts on the protection of environmentally sensitive areas and the sustainable management of areas of greatest productive potential. 74. Forest resources are currently the main source of domestic energy (an estimated 85 percent of the total) and traditional building materials. They also are an important source of raw material for the timber industry and for export. Apart from wood materials, forests provide a habitat for wild flora and fauna; game meat, fruits and berries; grazing and fodder; honey and beeswax; skins and hunting trophies; and medicinal plants, in addition to a number of endemic species that are important on biodiversity conservation grounds. 75. The productivity and potential maximum allowable extraction of Mozambique's natural forests is not known with any degree of certainty. Estimates range from about 500,000 m3/year (FAO, 1982) to 1.1 million m3/year (Kir, 1986). This would be considerably higher than the maximum volume ever logged in the country and several times the annual volume logged over the last few years, which saw a decline from 186,000 m3 in 1981 to 71,000 m in 1990. 76. It is likely, however, that these official figures grossly underestimate actual logging. Indeed, most of the logging of natural forests has taken place in areas out of the effective control of the Forest administration and most likely has gone unreported. There is currently strong interest from domestic and foreign private investors. Logging and export of tropical timber would represent significant sources of foreign exchange for the country; based on productivity and revenues in other African countries, the potential value of exports has been roughly estimated at USD 20-30 million per year (100,000 m at USD 200-300/m ). However, strong protective measures should be immediately put in place for protecting existing resources so that 31 A more detailed discussion of the environrnental issues related to Mozambique's natural resource base, soils, water and irrigation, livestock, forests and wildlife is presented in AFI AE, Country Environmental Strategy Paper for the Republic of Mozambique, Report No. 12666 MOZ, December 18, 1994. 30 they be managed on a sustainable basis. This will involve detailed inventories and the preparation of management plans. It will also require full participation of the local populations since it is clear that without their cooperation and support, adequate enforcement is impossible. In addition, there are about 42,000 ha of industrial and fuelwood plantations containing mostly exotic species: Eucalyptus (43 percent); Pinus (55 percent); and Casuarina (2 percent). Industrial plantations are in the process of being privatized. 77. Even though forest resources comprise a vast range of species, only a small number of them are actually exploited and marketed domestically and externally. The most valuable species are estimated to represent between 10- 15 percent of the total. Ebony, rosewood and umbila, in particular, have a high value in external markets. So, even though total extraction today is well below the potential maximum allowable extraction, several species may be in danger of overexploitation. Given the impossibility of inventories or management plans during the war, the availability of the resource has not been taken into consideration in deciding the quantities authorized for annual felling. 78. The timber trade is a major source of foreign exchange for Mozambique. In view of inadequate controls, however, there is a danger that the country will be stripped of its finest hardwoods, with negative effects both on the environment and on the long-term sustainability of the export effort and little actual benefits for the country. Conditions would seem to still be in place for the recovery and sustainable expansion of the commercial timber sector, following the decline in recent years caused by the war and the obsolescence of equipment. 79. Biomass Fuels. Energy use in Mozambique, which rose only slightly between 1965 and 1989, consists primarily of household energy consumption, which accounts for two-thirds of the total. Due primarily to the shortage of foreign exchange, which made LPG, kerosene and electricity too unreliable and expensive to use, many urban households began using fuelwood and charcoal in the 1980's. Accordingly, the share of commercial sources in total energy consumption has dropped from 20 percent to 7 percent in the last 10 years. Given the country's socio-economic situation, woodfuel and charcoal will continue to be the main source of energy for a long time to come. 80. Even though the available forest stocks would suffice to meet the bulk of household fuel needs in years to come if current trends continue, it is estimated that by the year 2000 the supply of biomass fuels may become constrained at the regional level. Sizable regional imbalances are already being encountered in some areas - above all, Maputo, which according to surveys conducted has been undersupplied with biomass fuels since at least 1985. Nampula and Tete provinces are also now experiencing biomass fuels shortages even though overall provincial balances indicate they should have no supply problems overall. Thus, there can be no doubt that household consumers may face fuel shortages at the local level even if the biomass supply at the national level is still sufficient to meet the country's requirements. 81. Starting in the late 1970's, three large fuelwood plantations with a combined area of 7,900 hectares were established nearthe cities of Maputo (1977), Beira and Nampula (1981) in an effort to compensate for the anticipated deficit. The output of these plantations, however, fell far short of the anticipated yields because of bad species choice and high mortality rates, inadequate funding for management, armed attacks, and brush fires. 82. Production and marketing of biomass fuels differs in scattered rural settlements, denser rural settlements, and areas of urban attraction. In the countryside, women and girls traditionally collect branches and wood from fallen and dead trees when it is dry and ready to burn. The environmental damage is thus almost nil or even positive, as burnable materials are removed 31 from the forest. As rural population densities increase, the most immediate consequence is the increase in the time women and girls must spend looking for adequate fuelwood. Since their daily time is already overly committed for food production and home-care activities (§ 142), increasing fuel collection time must come out of either cropping (with consequent reduction in crop production) or husbandry (with consequent reduction in the nutritional and health status of the family and, particularly, infants). 83. In contrast, collecting woodfuel and producing charcoal are local industries, employing mainly men, and these mainly from the periurban areas. Trees are felled for these purposes, which is already causing environmental degradation, mainly around cities, villages and refugee camps, along the main transport corridors, and in the mangrove areas in the coast. Charcoal production, because of the energy lost in the process, increases the pressure on the forest resource, but since it is cheaper to transport, enables a larger area to be exploited. 84. There is no Govemment regulation or constraint on biomass fuel extraction in Mozambique, and cutting licenses and fees are only enforced to a limited extent32. A free market for woodfuel and charcoal has developed, and farmers have often sold wood as a substitute for cash crops when crop yields are low, or when the security situation did not permit them to stay in the rural areas through the cropping period. The only real constraint is the naturally high transport cost, compounded by the deteriorated state of infrastructure and the transport fleet. The Biomass Energy Unit estimated in 1992 that the transport cost of a truckload of woodfuel over a distance of 30-40 km cost about two-and-a-half times the worth of the wood itself. 85. Wildlife. There is a wide variety of wild animals in Mozambique. It is well known that the total biomass and maybe even the number of species suffered greatly from the recent war, but war activities did not allow a methodical evaluation of the state of animal biodiversity, censuses of large animals, or re-estimates of biomass to be performed. The larger faunal species concentrate in the provinces of Sofala, Tete, Niassa, and Cabo Delgado. About eight species of mammals are currently noted as rare, threatened or endangered, namely, the black and white rhinos, roan antelope, giraffe, dugong, cheetah, tsessebe, Johnstone's wildebeest, and elephant. There is little knowledge about smaller species. There are four main uses of wildlife resources: game consumption; wildlife breeding; tourism use; and commercial hunting. 86. Game consumption. Game meat is one of the principal sources of animal protein for the majority of the rural population, and an important supplementary source for many in the urban and periurban areas. It is estimated that wildlife provides an important source of protein to about 70 percent of the Mozambican population; it has been roughly estimated that the value of game meat consumed each year, at import substitution prices, is worth between USD 465-930 million. Skins, bones, and other byproducts (including valuable ones, like ivory tusks and rhino horns) constitute important raw materials for both domestic use and manufacturing, and a ready source of cash for farmers and hunters. 87. Wildlife breeding. Commercial breeding of animals has increased rapidly in Southern Africa during the last decade. Game farms and ranches have a number of advantages over domesticated cattle: they use a varied range of grazing, are resistant to local disease and pests, and require little or no free water sources. There is an increasing and extremely lucrative 32 In practice the Provincial Forestry and Wildlife Service does not have the capacity to enforce cutting licenses. A nominal control is effected by control posts on the main roads feeding into the main cities, but in practice it is not very effective. The amount paid as cutting fee is less than 1% of the final worth of the wood or charcoal, insufficient to encourage tree-planting or conservation, or to finance better enforcement. 32 market for both game meat and skins, and for live wild animals, which could be supplied in a sustainable way. Today, the only such game farms in Mozambique are four crocodile farms, while ranchers in South Africa, Zimbabwe and Kenya raise many different species. 88. Tourism. Wildlife is also a potential source of income from tourism, including non- consumptive safaris and ecotourism. Mozambique has considerable tourist potential from the combination of beaches, wildlife, and beautiful countryside. Wildlife-related tourism income is estimated at USD 500 million a year in Kenya and USD 100 million in Zimbabwe. During the war, wild animals were killed indiscriminately, and as a result total biomass was severely depleted, and many species, including most of the charismatic large fauna, are on the verge of extinction. 89. Commercial hunting. For many years, wildlife was so abundant that commercial hunting went uncontrolled. The Government created a parastatal company, EMOFAUNA, which systematically "mined" the biomass of buffalo and other edible large mammals for both domestic consumption and export. Today the biomass available cannot sustain such activities. There is, however, a ready international market for prize and trophy hunting, and a few commercial hunting companies are currently operating in Mozambique. Care has to be taken to ensure that the state receives its proper share of the hunting and trophy fees and private concessions, and that local populations get involved in, and benefit from, in wildlife management and harvesting to ensure the sustainability of the wildlife and the proper distribution of benefits derived. The experience of the CAMPFIRE program in Zimbabwe in this regard is worth transferring to Mozambique. 90. Protection is required for preserving and developing Mozambique's wildlife potential. Mozambique has signed the Convention on International Trade in Endangered Species (CITES). It has also agreed to the banning of international ivory trade, thus prohibiting hunting elephants for sport. But the national system of forest and wildlife protected areas has been a shambles since the war. Park and reserve limits are not well defined. Staff in the field is almost nil, except for the Bazaruto National Park and the Maputo Elephant Reserve, where there is a certain, though inadequate, presence. The school for game wardens and scouts in Gorongosa has been closed ever since it was overrun during the war. Formally, protected areas include four national parks with some 16,200 km2; five wildlife reserves33 with 19,300 kM2; and thirteen game reserves34 with 56,700 km2 (plus fourteen forest reserves with 5,300 km2). In total, these areas cover 11.5 percent of the national territory. Together, they compose a protected wildlife system larger than Kenya's, South Africa's or Zimbabwe's. However, little is known about their current situation. The Government and the World Bank, through funding from the Global Environment Facility (GEF), are finalizing preparation of the Transborder National Parks projects, which should permit the rehabilitation and reactivation of the national protected areas system, while introducing the CAMPFIRE population participation principles into Mozambique. 33 Niassa (Niassa), Gile (Zambezia), Romene (Inhambane), Maputo (Maputo) and Marromeu (Sofala). 34 Gorongosa (in Sofala Province); Zinave (Gaza, Inhambane); Bahine (Gaza), and Bazaruto (Inhambane). 33 B. Land Use (a) Structure of Production 91. Agriculture is the most important sector of the Mozambican economy. It accounts for 40 to 50 percent of GDP, employs more than 80 percent of the labor force, and contributes more than 80 percent of its foreign exchange earnings. Compared with the sectoral shares just after independence (1980), when agriculture accounted for about 45 percent of total GDP, there has been a "reversal" of the norrnal development pattern, with agriculture's share growing as other sectors collapsed faster than agriculture. This reflects the collapse of the industrial sector because of the war and economic policies and also the relative resilience of subsistence agriculture due to the country's favorable endowment of natural resources. Mozambique is rich in fertile soils and has a wide range of agroecological conditions and abundant water resources. This potential permits diverse agro-pastoral farming systems to prosper when the rains are good, and to endure when they are not. This enormous potential is largely unrealized. It is estimated that about 2.9 million hectares were under cultivation in 1992, over half of them are on the more fertile soils of the Northern provinces. This represented less than 4 percent of Mozambique's total land area and 8 percent of its estimated total arable land. Table 8. Area under Cultivation (ha) and Percentage of Cultivated Area Planted to Specific Crops, 1992 Area Food Crops Cash cultivated Cassava Sorghum Maize Rice Beans crops Province (ha) (%) (%) Niassa 220,000 11 15 47 1 24 2 Cabo Del. 297,000 41 16 18 3 12 10 Nampula 830,000 56 14 11 3 8 11 Zambdzia 435,000 41 9 28 9 6 7 Tete 160,000 1 22 51 - 12 14 Manica 120,000 1 18 72 - 1 8 Sofala 176,000 7 26 39 11 8 9 Inhambane 337,000 19 8 31 1 15 26 Gaza 183,000 18 2 50 4 15 11 Maputo 156,000 5 2 64 5 15 9 Total 2,914,000 32 13 31 4 11 9 Sources: Based on satellite surveys of land-use cover, carried out under the Early Warning System (EWS) of the Ministry of Agriculture. (b) Main Farming Systems 92. Overall crop and animal distribution. In the North, cassava and sorghum are the dominant food crops at low and medium altitudes, and maize at higher elevations. Maize, 34 cassava and rice are the dominant crops in the Central provinces. In the South, maize is the main rainfed food crop, with mixtures of groundnut, beans and cassava, while rice and vegetables are grown as irrigated crops. The principal cash crops are cotton, cashew and coconuts. Cotton is widely grown at medium altitudes in the North. Cashew and coconuts are grown along the coast: coconut along the whole coastal belt and cashew over large areas in the North and within a narrow coastal strip toward the South (more than 60 percent of the trees are in Nampula Province). Tea is confined to the higher altitudes of Western Zambezia. Sugar cane is grown in irrigated estates in the Center and South. The cattle population is concentrated in the tsetse free Southern region and in localized areas in the North and West. Sheep, goats and pigs are raised throughout the country (the former particularly in the Western provinces) and poultry mostly around the cities. 93. The Northern Farming System. In the north, farmers adopt a diversified agro-forestry strategy based on a range of cereals (maize, sorghum, rice) as well as root crops (cassava), fruits (cashew) and vegetables, with wildlife and forest products (honey, berries) providing an important source of food. Farmers practice shifting ("slash-and-burn") agriculture with long fallow periods. Seeds are usually planted with a hoe among the remaining tree stumps. The major crops are maize (in the more humid areas closer to the coast), sorghum, millet and cassava, with lesser amounts of rice groundnut and legumes. In general, cereals are intercropped with beans and cowpeas. Sweet potatoes, peas and papaya are cultivated for home consumption, often in separate garden plots. The chief cash crops are cotton and cashew, with groundnuts, tomatoes and onions playing a lesser role. Cashew nuts are widely used for home consumption and for making alcoholic beverages. Cashew production has major growth potential in the short and longer term. Cotton production is located in Cabo Delgado and Nampula provinces. Production fell dramatically after independence, because of its negative socio-political connotation, but also because of the collapse of the ginning infrastructure. Recently, the World Bank has been supporting its rehabilitation and expansion through the Agricultural Services Rehabilitation and Development project (Credit 2337-Moz). 94. The Highland farming system. Farmers traditionally practiced shifting cultivation with a long cropping period of between 8 and 12 years and a fallow period of 20 years or more. In some villages, lands once farmed have never been cropped again. When the population was forcibly concentrated in permanent villages in the late 1960's, the cultural practices required to maintain soil fertility in a semi-permanent production system were not introduced. Nevertheless, due to natural high fertility and favorable climate, farmers still manage to maintain production levels. Most families cultivate two types of plot, one in rainfed conditions cropped in the rainy season, and one in a valley bottom cropped in the dry season on residual moisture supplemented by hand-watering from streams. The main food crops are maize and beans, and the main cash crops, beans and potatoes. Cassava is also grown for food security reasons, and many fruit and vegetable crops are grown for home consumption. 95. The Southern farming system. An extensive agro-pastoral strategy evolved in this region to provide an adequate response to the frequent droughts. Farms are in general composed of several plots, often quite a distance apart. Upland areas (serras) have light sandy soils and support rainfed cereals (millet, sorghum, maize). Maize is the principal cereal, except in the driest areas of inland Gaza province, where millet and sorghum are dominant, and along a coastal strip between Beira and Quelimane, where rice is the principal crop. On account of low tsetse fly infestation and tree cover, cattle and animal traction were introduced long ago and smallholders usually destump their fields for plowing. Near the coast and along rivers, land is scarce and fallow periods are short. Along rivers or in depressions where good soil and moisture 35 collect, maize and rice are planted. Small-scale irrigation along river banks support vegetable and tubers (potatoes). Machongos, coastal depressions with rich peat soils, are found near the mouth of the Incomati, Limpopo and Zambezi rivers. They present an important agricultural potential, although their soils are fragile and require very careful management. Cattle are used for plowing and providing milk, and as a store of wealth and a security for hard times, as the extensive network of permanent rivers permit access to enough water even during severe droughts. The number of cattle, however, fell dramatically as a result of the civil war. Potential for future growth 96. Mozambique has enough soil and water resources to support a volume of production much larger than today's. In the case of food crops, the production of maize and oilseeds (groundnuts, sunflower) could increase significantly, by expansion of areas in the short term and by increasing yields in the longer term. Maize can be grown to advantage in the highlands due to appropriate natural conditions (temperature and rainfall), and relatively cheap traction, and in the south due to the availability of markets and relatively cheap transport. Prospects for maize in the northeastern region would evolve as convenient and cheap transport links the area with the larger urban or export markets. The DRCs"5 indicate that such marketing patterns could be viable in some cases even under present conditions given appropriate food aid and pricing policies. In many areas of the country, production of rice and of cassava could go up substantially as well. Resources for the production of beans and potatoes in the highlands are quite abundant, the main constraint being access to markets. 97. In the case of cash crops, cashew nut production could grow substantially in the short term through better husbandry of existing trees, rehabilitation of poor stands and semi- abandoned groves, and use of some minimal inputs as a result of the much higher prices producers would fetch as a result of trade liberalization. In the longer term, the enhanced profitability of cashew nut production should lead to new plantations of pest-tolerant and pest- resistant varieties. Areas for the expansion of copra production would also seem to be relatively abundant, albeit less so than for cashew trees. The potential for expansion of tea and sugar cane production would seem to be more limited than that for cashew and copra; most increases in production in the short term would come from the rehabilitation of tea groves; and the rehabilitation of the sugarcane plantations, sugar mills, and supporting irrigation systems. 98. The profitability of cotton production improved considerably between 1993 and 1996 on account of the improvement in intemational prices. Furthermore, the competitiveness of the cotton sector could be enhanced by both the introduction of cultivars with a higher processing conversion factor, reducing ginning costs, and more frequent updates of cotton procurement prices to keep pace with inflation. 99. In the case of livestock, Mozambique has a number of advantages, mainly for meat production from ruminants. The first is abundant range resources, and the significant agricultural potential should allow a large use of crop byproducts, such as oilseed cake, molasses and citrus pulp. The second is the good genetic potential of the existing breeds for meat production and animal traction. Local ruminant breeds are well adapted to local conditions and have proved resilient under drought conditions. The third is the existence of an experienced commercial beef farm sector and a skilled Livestock Service in the Ministry of Agriculture. 35 See Chapter 5 for a presentation of the domestic resource cost coefficients. 36 100. Animal populations today are well below the levels achieved immediately after independence, as a result of both the war and the collapse of the state farm system. The total number of cattle dropped from almost 1.5 million head in 1974 to 0.25 million in 1992; while the number of cows dropped from 533,000 to 86,000. In small ruminants, the figures for goats and sheep dropped from 500,000 and 130,000 head, respectively, to 169,000 and 23,000 head. The species relatively least affected was pigs, which dropped "only" by 30 percent, from 214,000 to 150,000 head in the same period. Production levels per head are not known to any precision, but all estimates yield extremely low figures, with sanitary conditions a major limiting factor. Thus, just recovering past levels would provide a major expansion of the livestock sub-sector. Over the longer term, the abundant availability of pastures, and then the increasing availability of crop byproducts and feed supplements, would allow a much larger herd to be achieved. 101. The main constraint for such expansion, however, is the limited number of reproductive females. Assuming optimal development on the basis of existing females only, the cattle herd could grow at some 2.7 to 3.8 percent p.a. If 1,000 cows were imported each year for the smallholder sector and another 1,000 for the commercial sector, rates of growth would go up to 3.9 to 5.2 percent p.a. The recovery of the grain and oilseed crop production should generate additional byproducts, which should support the development of pigs and poultry. (c) Main Categories of Producers 102. Agricultural production is carried out by two main categories of producers: smallholders (normally called "family sector" in Mozambique) and large-scale (often called "commercial") producers, which in turn include private and state farms as well as joint venture companies (JvCs), and cooperatives. National agricultural production is dominated by the smallholder sector in terms of both areas cultivated and total production. The table below presents the evolution of cultivated areas by category of producers between 1983 and 1992. It illustrates the gradual demise of the state sector; the virtual disappearance of the cooperative sector; the appearance since the mid- 1980s of JvCs; the gradual increase in the private commercial sector, and the relative stability of areas cultivated by smallholders. Table 9. Cultivated Areas by Producer Category ('000 ha) Category 1983 1986 1992 Smallholder Sector n.a. n.a 2,780 Commercial Sector 41.2 55.0 68 Cooperatives 7.9 4.7 3 State Farms 118.6 66.9 35 Joint Ventures - 1.9 25 Sub-Total 167.7 128.5 120 Total 2,900 Source: Ministry of Agriculture and Fisheries. 103. In 1992, the smallholder sector accounted for over 95 percent of the total land under cultivation, 75 percent of all cattle, 93 percent of small ruminants, and 60 percent of pigs: It is 37 comprised of about 2 to 2.5 million family units farrning 0.5 to 3 hectares each, with an average farmed area of 1.0 ha.36 The size of plots is basically determined by the available family labor, with a lesser effect of other factors such as location, ecological conditions, presence of irrigation, and distance to markets. For example, in the family cycle in the highlands, the mean farm size for the first five years of married life is about 0.7 ha on good soils. As children are born and reach working age, the farm size increases to 1.2 ha, including 0.2 ha of vegetable garden. The aggregate family's farm size may expand even further, to 2.0 ha of good soils when the husband marries a second wife; as discussed elsewhere, however, from the farm management and decision-making points of view such expansion is better defined as a new farm. The mean cropped area per family in Zambezia in 1993 (0.8 ha) was lower than in other provinces because large parts near the coast are irrigable, while the relatively high average in Inhambane (1.4 ha) is due to poor rainfall. As the 1980's evolved, average family plots had decreased, particularly in the south, partly due to the war particularly through its effects on rural security and the availability of animal traction. For example, the areas cultivated in the three southern provinces in 1993 range from 1.0 to 1.4 ha, whereas a 1969 survey had showed them to be between 1.8 and 2.2 ha. Table 10. Smallholder Sector Characteristics Provinces Farms % Total Area % ('000) ('000 ha) Niassa 172 6 220 8 Cabo Delgado 273 10 295 11 Nampula 707 26 796 29 Zambezia 630 23 424 15 Tete 152 5 150 5 Manica 100 3 139 4 Sofala 185 7 157 6 Inhambane 248 9 334 12 Gaza 188 7 179 7 Maputo 111 4 85 3 Total 2,766 100 2,779 100 104. The geographic distribution of smallholder farms is presented in Table 10. Nampula and Zambezia together account for about half of the country's total number of smallholder farms and total cultivated area. This is less than at the end of the 1970s (about 1.3 ha), the reduction being due to the impact of the war on the availability of labor and on access to the fields. 105. Table 11 presents estimates of the share of smallholders in the total and marketed production of major crops. For example, in 1994 smallholders accounted for close to 100 percent of the total and marketed production of most food crops, and for 87 percent of rice. They also accounted for 100 percent of cashew production; almost half of cotton production, and 79 percent of copra. 36 See the Appendix Tables for further detail on farm sizes. 38 Table 11. Total and Marketed Production (1990/91) Total Production Marketed Production Total Smallholders Smallholder Total Smallholders Smallholder Share Share (MT) (MT) (%) (MT) (MT) (%) Maize 327 314 96 84 73 87 Beans 78 77 99 14 13 93 Rice 56 49 87 20 6 30 Sorghum 155 154 99 1 1 100 Cassava 370 369 100 24 24 100 Cotton 30 14 47 Cashew 23 23 100 Copra 27 19 79 Citrus 26 2 8 Chicken 2 - 0 Beef 2 1 50 106. Private "commercial" farmers are distinct from smallholders in that they employ salaried labor and pay rent to the State for their land concession. The farms vary greatly in size, from 10 to several hundred hectares and produce almost solely for the market. In spite of severe difficulties in obtaining inputs and credit, the private commercial sector has shown a surprising resilience and dynamism throughout the 1980s. In 1983, they accounted for less than 2 percent of the total cultivated area, but for 4 percent of marketed production and 14 percent of exports. Their current number is unknown but has increased markedly in recent years, in particular with the privatization of State farms. They are currently active in the production of fruits and vegetables as well as cotton, tea and sugar. They will have an important role in the introduction of new technologies and the opening of new markets to smallholders. 107. The State farms are the juridical heirs to the large private estates abandoned at independence. In 1983, the State farn sector accounted for about 6 percent of the total cultivated area, 43 percent of the total marketed surplus, and 66 percent of total agricultural exports. From then on, State farm output declined sharply due to the country's mounting budgetary difficulties, combined with poor management and the growing insurgency. By 1986, the area cultivated by State farms had declined by 50 percent. In recent years, a massive formal and informal distribution of State farm holdings to private individuals and companies has taken place in the context of the privatization program. According to the most recent estimates, some 25 State farms have been officially transferred to private interests. The remaining public agricultural enterprises are tree crop plantations or are associated with agro-industrial complexes (citrus, tea, coconut, sugar). Most are practically abandoned (tea estates) or vastly under-utilized (only 7,000 of a potential 52,000 ha of sugar cane was harvested in 1992/93). 108. Joint venture companies (JVCs) were created as successors to Government's estates, in particular in the cotton and tree crop sectors. First established in 1990, the most important of these are the three cotton companies, LOMACO (a joint venture of Lonrho and the government), SODAM, (a joint venture of Joao Ferreira dos Santos and the government) and SAMO (a joint venture of Entreposto and the government) . These companies have a monopoly on marketing 39 cotton within their specified area and are responsible for extension advice and input supply to producers. (d) Technology and Productivity 109. As noted above, smallholders typically practice long fallow shifting cultivation, with cultivation periods extending from five to twelve years, and fallow periods varying from three or four years in areas where population pressures are increasing, to several decades in low population density areas such as large parts of Niassa province. In most parts of Mozambique, smallholder families cultivate two types of farms, the field (machamba) in the hot, rainy season, and the garden (horta) in the dry season. The latter is smaller and often located in a valley bottom (machongo) with residual moisture, or irrigated manually. 110. The technology is rudimentary. Smallholders use manual methods and simple tools such as the axe (machado), the billhook (catana), and the hoe (enxada). T'he average farm household owns five to six manual tools, and four to five containers for storing produce37 The areas are manually cleared and burnt, the ash serving to enrich the soil for the first planting. Normally they use no input other than seeds. Only seven percent of smallholders use purchased inputs, such as fertilizer and insecticides,38 except in limited areas of semi-intensive cotton cultivation, where they are usually provided by the relevant cotton ginning firm. Improved varieties are typically not used, except for the free seed and tool distribution program carried out by Government to help in the resettlement of populations displaced by the war. It appears that there has been little technological advance in the smallholder sector over the last two decades. In some areas, smallholders' production strategies may even have regressed toward pure subsistence agriculture because of the lack of access to markets and inputs brought about by the war (for example, the near total disappearance of animal traction in the southern provinces). 111. Smallholders raise livestock usually in very small herds, on a completely extensive pattern, with exclusive use of natural pastures, and using no inputs except for the mandatory vaccinations and free tick baths. Lack of supplementary feeding may lead to weight losses in the dry season of up to 20 percent of the maximum weight gained in the rainy season. Animal traction was relatively widely used in the South and in localized areas of higher elevation in the northern and western region of the country, where tsetse fly infestation is not a problem. With the decimation of livestock during the war, that is seldom the case. In turn, lack of cattle means manure is not used either, nor are practices such as mulching and composting common. 112. The smallholder sector has been neglected by Government's support services and its access to markets was severely restricted by insecurity and massive destruction of transport infrastructure. Until recently, agricultural research had not been directed at smallholders' needs and productivity; these had no access to extension advice except for some limited services provided by NGOs. Few smallholders had access to credit, and most of those who do get in-kind distribution of insecticides and spraying media from the cotton gins which, on account of their monopolistic rights, are able to recover most of the cost. 37 Ministerio da Agricultura (1994), as quoted in Desai (1996), p. 14. See the Appendix Tables for detail by region and tool type. 3S Desai (1996), p. 14. 40 113. The poor technological and productivity status of the smallholder sector is compounded by low producer prices, resulting in the past from the overvalued exchange rate and Government intervention and today from inappropriate pricing policy for food aid and "leakage" of food aid into commercial channels. Similarly, poor marketing channels (or complete lack of) are further obstacles for technological renovation in the smallholder sector. Adoption of new technology, if at all, is more likely in crops which are traded (totally or at the margin), such as the cash crops or maize. It is much more unlikely in crops which are not or seldom traded, like sorghum, millet, or cassava. Thus, crops like cashew nuts, cotton and maize are the best bets for introducing new varieties or production technologies - e.g., drought-resistant improved maize varieties. 114. Yields and productivity are low. Table 12 presents estimates of average yields in the smallholder sector. Table 12. Estimates of Average Yields (t/ha) Crop Average Comparative Sub-Saharan Yield a Zimbabwe b Africa b Cassava 5.0 - 7.0 4.0 7.5 Sorghum 0.3 - 0.6 0.6 0.8 Maize 0.3-1.3 1.4 1.2 Rice 0.5-1.0 2.8 1.6 Beans 0.2 - 0.4 0.7 0.7 Cotton 0.3 - 0.6 1.2 0.8 Cashew 0.1 - 0.2 ---- ---- a Varies according to agroecological conditions. bFAO Production Year Book. 115. Farmers store food crops such as maize and beans in storage huts or bins made of wicker or adobe. Maize is often stored as whole ears and left on top of elevated racks, being shelled only as needed. Cassava and potatoes are harvested as needed for consumption or sale and are thus not stored. 116. Commercial farms and joint ventures use a more capital and input-intensive technology, availing themselves of improved seeds, purchased inputs including fertilizers, agro-chemicals, and veterinary products, and some machinery. State farms were intended to be similar, but socialistic resource allocation and misplaced concepts about what being modern meant led them to have (and use) more machinery than appropriate at prevailing input prices; it has been estimated that their added value in foreign exchange terms was negative. This was compounded by poor management, leading to poor maintenance and ensuing idling of large parts of the machinery fleet. Yields are higher (one is tempted to say "normal") in commercial farms, but were often very poor in the state farm system. 117. With low labor opportunity cost and wages, and import parity prices for machinery and imported inputs (at today's exchange rate, and subject to import tariffs of only 10 percent), manual, labor intensive production methods will enjoy an advantage for years to come. Since the industrial sector is not likely to grow quickly enough to create labor scarcity, the market and opportunity cost of labor is likely to remain low. Land is becoming scarcer in the southern provinces and in pockets around cities like Lichinga and Nampula, but elsewhere in the country it does not yet have a market value. Therefore smallholders are likely to hold a comparative 41 advantage over commercial farmers for several decades, except in activities requiring unusual skills or precision, such as banana exporting, or those subject to forward linkages such as cotton. DRC calculations indicate, as LOMACO officers themselves suggested, that smallholder outgrowers are more efficient in producing cotton than the LOMACO estates, which are run by the firm's employees using capital-intensive methods and have yields five times higher. There exists enormous scope for an expansion of production through outgrower arrangements. An indication of the superior efficiency of smallholders is given by the experience of LOMACO in market gardening: the firm used to produce tomatoes, onions and cabbages for the Maputo market, but with the growth of small producers this had to be discontinued (Aube 1995, p. 9). 118. Similarly, in most of the country apart from the South, land-extensive technologies make more economic sense than land-saving technologies such as improved seeds and chemical inputs, even if both have been found to raise the output of smallholders significantly. However, it is important to realize that relative prices are not the only relevant factor in choice of technology - risk management being perhaps as important in Mozambique. Therefore, while drought- resistant varieties reduce risks associated with dry spells - the most likely risk faced by agriculture in Mozambique - and may thus be adopted, chemical inputs increase them, making their adoption less likely. Improved seeds are also labor-saving devices, since they substitute for labor required to open, cultivate and weed a larger plot to be otherwise sown with traditional seeds - and labor is a scarce resource in smallholder agriculture. Given the state of research today and the fact that several improved (open-pollinated) varieties are available, there is reason for optimism that improved varieties for several crops will be adopted. As markets develop and the transport system improves, the possibility of selling surplus crops will provide an additional incentive for their adoption. C. Rural Population and Labor Availability 119. There is considerable uncertainty over the current population data and much more still with regard to projections. The last population census was carried out in 1980. Current projections, based on extrapolations of the 1980 census data and an overall population growth rate of 2.6 percent p.a., indicate that the total resident population of Mozambique was about 17.5 million in 1995, on the assumption that the 1.5 million refugees in neighboring countries had returned. There is great uncertainty with regard to the relative proportion of rural and urban populations. In 1980, about 85 percent of the total population lived in rural areas but the war and a succession of natural calamities have had an enormous impact on the fabric of rural society and on settlement patterns over the last decade. They added a massive impetus to the normal processes of population concentration and urbanization, as people were driven into secure areas in the rail corridors and around provincial and district towns. It is estimated that today more than 25 percent of the total population lives in or around urban areas and that two-thirds of these live in a 50 km-wide strip along the coast, where population density exceeds 120/km2. The table below presents estimates for 1992 and 1995, and projections for 2000 and 2025. The projections assume that most refugees in neighboring countries had returned home by 1995, boosting the resident population to 17.5 million. The overall annual population growth rate of 2.6 percent yields a total population of 20 million by year 2000 and of close to 30 million by year 2025. 42 Table 13. Estimated Population 1992-2025 1992 1995 2000 2025 Total 14.8 17.5 19.9 30.0 Of which: Urban 3.7 4.5 6.0 10.0 Rural 11.1 13.0 13.9 20.0 Source: UNICEF. 120. The war created two groups of populations whose behavior will have a considerable impact on post-emergency strategies at both the national and the family level. About 6.5 million people (about half of the estimated total population) were displaced (deslocados) or directly affected by violence and famine, and 1.5 million persons found refuge in neighboring countries (refugiados). The distinction between these two categories and the populations considered "non- affected" are in a sense artificial, as the internally displaced and the "local" populations have had to develop integrated strategies for resolving common problems and mitigating common hardship. Limited surveys indicate that most refugiados have returned to Mozambique. Practically all of those who were crammed into the 500 deslocados camps throughout the country have gone back to their places of origin. The resettlement program was a remarkable achievement attesting to the increasing prospects for a lasting peace. 121. Most families adopted a risk-minimizing strategy, sending some members back to re- establish farms while others stayed behind to retain urban incomes and benefit from food distributions. It appears that some of these will remain permanently divided between city and countryside. The war also exposed many millions of rural dwellers to urban life and this will contribute to long-term urbanization because many families were displaced for years and established new lives; this applies particularly to the younger generation which is reluctant to move to an unknown and unattractive rural lifestyle. Thus it is unlikely that there will be any net urban-rural outmigration: many refugees have moved directly to the cities where income opportunities are better, and many displaced people around cities, especially the younger ones, have stayed put. 122. The long-term settlement pattern will be driven by population growth and density, the economic potential of each region, sociological factors - the "culture" and local institutions (such as land tenure systems) and policy induced factors (incentives and infrastructure). Historically, the rich natural resource base of the Zambezi and Limpopo River valleys has attracted economic activity and settlement. The Portuguese developed the coastal zones and built five railway lines that cross the country from east to west. Consequently, more than 75 percent of total population clustered along the coast, the main rivers and railway lines. These areas will continue to be magnets for populations. The rehabilitation of existing infrastructure will allow populations to gradually re-establish settlement patterns similar to those existing before the war, with an increased concentration in urban and coastal areas. The Government can play a role in influencing future settlement patterns by directing economic activity and new settlement to areas so far by-passed by development, in order to relieve population pressure in urban areas and regions where the density of population is already a threat to the sustainable use of the natural resource base. Its main instruments in opening new development poles are the provision of basic transport infrastructure and economic incentives to private entrepreneurs. In 43 order to help the process, Government should let a land market develop and allow it to determine the real value of land on the basis of its opportunity cost. Government will also have a critical role to play in containing development in ecologically sensitive areas. Due consideration should be given to ecological factors in the creation of major new transport infrastructure and opening up new areas. 12'3. Table 13 indicates that the urbanization rate is likely to increase from its current 25 percent to 30 percent by the year 2000. This has profound implications for the development of the urban and rural economy and for the Government's strategy. In the urban sector, in the short and medium term, urban unemployment and poverty is likely to rise from the already high current level. Sustained productivity gains will also be required in agriculture to meet an increasing urban demand for food. It is therefore imperative that the Government make an all- out push for (private) productive investments and job creation both in the urban formal and informal sectors, and where most of the potential for growth and job/income generation lies: the rural sector and agriculture. In addition, due to severe budgetary constraints, urban social services (education, health, housing, water), which are meeting less than half the current demand, will also grow more slowly than the population. The Government will have to adopt determined cost recovery strategies, balanced with the necessary safety nets, and carefully weigh the necessity of catering to the minimum needs of the urban population with the need to improve the standard of living in rural areas for slowing migration to the cities. D. Rural Poverty and Gender Issues (a) Ruralpoverty 124. The only poverty line that has been calculated for Mozambique is that of del Ninno and Sahn (1993) for Maputo in 1992.39 They placed the poverty line at the level of expenditure per month that would ensure, given typical consumption patterns, a daily food intake of 2,500 Kcal. This corresponded to Mt3 1,904, or $13.50 per month. 33 percent of Maputo households (40 percent of the population) fell below this line. They also defined a line for the "ultra-poor", namely that expenditure level which guaranteed an intake of 2,000 Kcal. This corresponded to Mt20,15 8, or $8.50 per month. 10 percent of households and 13 percent of the Maputo population were ultra-poor. The two most important determinants of poverty were the level of education of the household head and the household size. Some 46 percent of the children in Maputo lived in poor households. The figure is higher than the share of poor households, reflecting the finding that larger households, and particularly those with more children, are more likely to fall below the poverty line. Virtually no household heads with at least Grade 10 education were poor, while 45 percent of those with Grade 2 or less were poor. Some 60 percent of people in households sized 16 or more were poor.40 The authors concluded that the government and the donor community had kept poverty at relatively moderate levels. 125. Most likely the extent of poverty and ultra-poverty was worse in other cities and in rural areas. Regular waged employment, functional social services, opportunities for informal sector 39 The FSD/Cornell survey was fielded between October 1991 and April 1992. The nominal official exchange rate in mid-January was Mtl900/$, and this was about 25% overvalued. 40 del Ninno and Sahn (1993), p. 14. 44 work, and government and NGO distributions were more readily available in Maputo than elsewhere. The war displaced millions of rural households and destroyed the bulk of the social services, including a third of all rural clinics and two-thirds of all primary schools. In addition, people dependent upon rainfed agriculture are vulnerable to drought and natural disasters. 126. Given the lack of a national poverty line measure, one could alternatively define poverty as the lack of a specified combination of conveniences which our society has come to consider essential to human dignity. This basket of services often includes access to clean water, access to primary education, access to medical care, and access to enough food calories for health and comfort. By any reasonable standard incorporating arbitrarily defined levels of the latter variables, the vast majority of Mozambicans are poor. About three-quarters of the population lacks access to safe water.4' 55 percent of men and 79 percent of women (rural and urban) are illiterate.42 Chronic malnutrition affects 27 percent of children under the age of five.43 About 60 percent of the population, rural and urban, lacks access to health services." Infant mortality (146 per thousand live births) and under five mortality (282 per thousand live births)45 are much higher than the sub-Saharan African averages of 99 and 172. Maternal mortality (1I,100 per hundred thousand live births)46 is one-third higher than the average for sub-Saharan Africa. Reflecting the lack of access - urban and rural - to medical care, the average life expectancy is 46 years.47 127. In the middle 1960s and at independence, Mozambique's social indicators were among the worst in the world. The level of human resources and social development was extremely low, even compared with the standards of other low-income countries (Table 14). FRELIMO gave high priority to education and health, and indicators started to improve in the early 1 980s. With the intensification of the war and the economic collapse most of these early gains were lost and, despite significant achievements in some areas like primary education and literacy, the human resource picture today remains bleak, as is shown in Table 14. 41 World Bank (1996a), p. 2. 42 World Development Indicators, 1995, Table 1. No rural/urban breakdown is available. Data refer to 1990. 43 In the sense of having a weight-for-age Z-score less than -2.0 (Inquerito de Indicadores Multiplos (1996), quoted in Desai (1996), Table XV). No rural/urban breakdown is available. Data refer to 1995. 44 World Bank (1996a), p. 2. 45 World Development Indicators, 1995, Table 27. No rural/urban breakdown is available. Data refer to 1993. 46 World Bank (1996a), p. 2. 47 World Development Indicators, 1995, Table I. No rural/urban breakdown is available. Data refer to 1993. 45 Table 14. Basic Social Indicators Mozambique Low income countries 1965 1995 1995 GDP per capita (1995 $) - 80 460 Life Expectancy (years) 41 47 63 Infant Mortality 179 146 58 Adult Literacy (%) 7 40 66 Source: World Development Indicators 1995; the World Bank's Africa Live Data Base. 128. Mozambique's per capita GDP is one of the lowest in the world, and while the precise numbers may well be questioned, owing to the poor state of the statistics, the country's social indicators are consistent with the view that Mozambique is one of the poorest countries. Its adult literacy rate (40 percent) lags well behind those of the bulk of low-income countries in the world (66 percent). Its infant mortality rate (estimated at 146 in 1993) is among the highest in the world, and correspondingly its life expectancy is one of the lowest, at 47 years. 129. It will be shown in the following paragraphs that a strategy of raising the farming output of smallholders is an effective way of reducing poverty. First it is shown that most Mozambican people are directly dependent upon smallholder agriculture; then that smallholders are among the poorest people in the country; and finally that raising smallholder production would reduce poverty. 130. As of 1993, 69 percent of the Mozambican population lived in rural areas and small rural villages. 48 This figure has risen since, as refugees and deslocados returned home; many observers estimate that some 80 percent are now rural. Of these rural people, about 90 percent have their livelihood in smallholder agriculture.49 Multiplying, about 72 percent of all Mozambicans are directly dependent upon smallholder agriculture. 131. Smallholders are among the poorest in Mozambican society. While many urban people, particularly the deslocados, are poor, the effects of deprivation, isolation and vulnerability in rural areas are generally worse. As of 1994, 68 percent of smallholders had no income from marketing their own production. 50 As is shown in Figure 7, the degree of market integration is low in all regions, with between 67 and 73 percent of households selling no produce at all. 48 World Development Indicators, 1995, Table 31. Data refer to 1993. 49 Desai (1996) reports that about 5 percent of rural households do not have "cultivated land" (2.5 percent in the North, 1.7 percent in the Central region, and 8.0 percent in the South) (p. 20). He also reports (Table XI), that 91.5 percent of rural households have some income from agriculture, fishing or livestock. 50 Ministerio da Agricultura e Pescas (1994), as quoted by Desai (1996), Table 6, p. 22. 46 80 60 40 20 0 North Center South Figure 7. Percent of farm households which do not sell any farm outputs, 132. Rural people are vulnerable to climatic uncertainty, and while all occupations, urban and rural, are subject to uncertainty (of demand, of employment, of prices, etc.), rural people have less scope for risk-reducing diversification.52 In 1994, most smallholders had no off-farm income (77 percent53). Non-agricultural income sources provide diversification against risk. Opportunities for such diversification are fewer in the more distant, rural areas, as is shown in Figure 8. 53 percent of farm households in the South of the country have no non-agricultural income source, while in the Center and the North the figure rises to over 80 percent. This is partly due to their lack of appropriate skills and partly to the costs of transport, which are high relative to incomes. 100 80 60 North Center South Figure 8. Percentage of farm households without non-agricultural income, 199454 133. Rural people have less access to medical services than do urban people. In urban areas, there were, as of 1992, 1.83 hospital beds per 1000 people in urban areas, and 0.48 in rural areas.55 Reflecting the urban bias of medical spending, medical staff are more thinly spread in rural areas. Figure 9 shows that in Maputo City there were about 20 patients for each medical staff member employed, whereas in the remainder of Maputo province there were some 43. In other largely rural provinces, the patient/medical staff ratios were also high - Zambezia (50), Nampula (42), Cabo Delgado (39). 5' Source: Desai (1996), Table 6, p. 22. The data apply to 1994. See the Appendix Tables for further detail. 52 In farming areas the problem is covariant risk. Demand for all services and goods falls if the rain fails. Covariant risk is less of a problem in urban areas. 53 Ministerio da Agricultura e Pescas (1994), as quoted in Desai (1996), Table 6, p. 22. 54 Source: Ministerio da Agricultura e Pescas (1994), as quoted by Desai (1996), Table 6, p. 22. 55 Ministerio da Saude (1994), p. 33, Table 3.1.2. 47 Niassa ,.= .. . Cabo Delgado I I Nampula I I Tete _3 Man ca Mn I _ -|l 111 Sofalae Inhambane m ! 1 1D Gaza Maputo Prov _ MaputoCty G__tl 0 10 20 30 40 50 Figure 9. Patients per hospital staff, by province, 199256 Infant mortality is worse in rural than in urban areas. One indicator is intra-hospital infant mortality, which varies from 1.4 percent in Maputo City to 4.3 percent in Cabo Delgado. The contrast between Maputo City and the remainder of Maputo province is stark: 1.4 versus 2.7 percent. Niassa _ _ ' _ '_'_I_ Cabo Delgado t_I_l_I Nampula Zambezia I I Tete Man:c a I Sofala Inhambane I I Gaza __ I MaputoProv. . C Maputo City _ r 0 1 2 3 4 5 Figure 10. Intra-hospital infant mortality (percent), 1992"7 134. Fewer people have access to safe water in rural than in urban areas. Figure 11 shows the percentages, by province, of people who have access to safe water. The highest, at 97 percent, is Maputo City. The remainder of Maputo province lags behind the city, with 64 percent. Sofala (87 percent) and Gaza (81 percent) are high because the samples were taken largely from the cities of Beira and Xai-Xai. The samples from Niassa (58 percent), Inhambane (27 percent) and Zambezia (38 percent) reflect more closely the situation in rural areas, which is considerably poorer than in Maputo City. 56 Source: Ministerio da Saude (1994), Table 3.2.5, p. 55. Hospital staff refer to the sum of doctors, nurses and technical staff. 57 Source: Ministerio da Saude (1994), Table 4.3.3, p. 85. Note that this is different from "infant mortality", which is usually defined as the rate of death per thousand of infants under two or five. 48 2 40 60 I g te. safe drnkn water, by p Furtermoe, mstgural peoplechav to gaet watrinfargawayer(32 percetoviner 10mtesawy,2 percent over 500m)59, the burden of which falls upon the women of the household. 135. Illiteracy is higher in rural areas than in urban areas. In some r-ural areas, as many as 87 percent of household heads are illiterate, as is shown in Figure 12, whereas the country average (rural and urban) is about 60 percent.60 Some Primary Primary completed 41% 9% Secondary No chooling 47% Figure 12. Schooling of household heads, Monapo district (Nampula)61 136. Access to education is more difficult in rural than in urban areas. For the country as a whole (rural and urban), the enrollment rate of children between six and eleven years is 52 percent, whereas for rural areas alone, the rate is 45 percent. The breakdown by province, rural and urban, is given in Figure 13. In most provinces, the rural-urban difference is prominent, particularly in Nampula, where the rural enrollment rate is only 23 percent, while the total enrollment rate (rural and urban) is 48 percent, indicating that the urban enrollment rate is of the order of 70 percent. 58Source: Multiple Indicator Cluster Survey, Table 8, p. 25. 59 Multiple Indicator Cluster Survey (1996), quoted in Desai (1996), Table XVI. 60No rural/urban breakdown is available. Data refer to 1995. Male: 42% illiterate, female 77%. In 1990: overall illiteracy was 67 percent ( World Development Indicators, 1995, Table 1). 6 Source: Borges and Doto (1996, p. 29ff). Counting those with incomplete primary as illiterate gives a total of 87 percent male illiteracy. The data refer to 1995. 49 _ _ X.. Rural only IOUrban + Rural 0 20 40 60 80 Percentage of children 6-1 0111 enrolled Figure 13. Enrollment of children in primary school, by province6Q 137. A survey by Medecins Sans Frontieres in November 1995 showed that about 1.5 percent of rural households in six selected districts had not eaten at all the day before, and about 15 percent had eaten no staples, viz. only vegetables, fruit or a wild food the day before.63 Government agricultural surveys reveal that the average rural household experiences "food insecurity" for 3.7 months in the year.64 This food insecurity is quite general, appearing in all regions (North, Central, South), across all household configurations (number of children, number of adults), and whether or not the family has some off-farm income, and irrespective of the household's access to land. There are two exacerbating factors: landlessness and gender. Rural landless households in the Center and the South of the country experience almost six months of "food insecurity" per year, whereas rural households with access to land average 3.5 to 5.5 months of food insecurity. Notably, there appears to be no link between farm size and food insecurity, given that the household has some access to land. There appears to be more food insecurity in rural areas of the South than in the Center or North, and this may have to do with low rainfall. 62 Rural data from Multiple Indicator Cluster Survey (1996), quoted in Desai (1996), Table XIII, referring to children between the ages of 6 and 10. Country-wide data (rural and urban) from Multiple Indicator Cluster Survey (1996), Table 7, p. 23, referring to children between the ages of 6 and 11. 63 Medecins Sans Frontieres (1996), p. 57, Table 12. See further detail in the Appendix Tables. 64 Ministdrio da Agricultura e Pescas (1994), quoted in Desai (1996), Table XIV. "Food insecurity" is here a subjective judgment of the respondent. 50 0 E 2 _ 0 <1 1-1.5 1.5-2 '2-3 >3 Land possessed by household (ha) |ONOrth Central ISouth Figure 14. Mean months of food insecurity, by farm size, as reported by rural households, 199465 The other exacerbating factor is gender: female-headed households experience more food insecurity. The average for the country is 3.5 months for male-headed households and 4.3 months for female-headed households. 138. Raising smallholder agricultural production would reduce some aspects of poverty directly. If smallholders faced higher output prices due to the reduction of export taxes or due to more efficient marketing (arising from better roads) their incomes would rise immediately; the stronger incentives would induce a supply response and incomes would rise further in subsequent years. Demand for primary education and elementary medical services would rise, and this in turn would tend to reduce the infant mortality rate and raise the life expectancy. With their basic food needs covered, smallholders would be better able to smooth their income inter- seasonally and inter-annually, resulting in less food insecurity. 139. The government's infrastructural interventions in aid of agricultural production would also reduce poverty. Restoring the feeder roads and installing culverts, for example, would lower transport costs and improve people's access to medical services and education. In turn the increase in basic education would facilitate technological transfer which would raise agricultural output and incomes in the medium term. Low-cost transport services would also result in an increase in rural demand for urban-produced commodities, resulting in improved incentives for increasing agricultural output. 140. Thus a strategy aimed at smallholders would simultaneously address the largest group of those in poverty and those of the poor whose situation is the most precarious. (b) Gender Issues in the RuralAreas 141. Traditionally, rural Mozambican men spend long spells (2 to 3, even 5 years) working as miners or industrial workers in South Africa. This plays a significant role in the generation of cash and savings at the family level (and of remittances, at the national level). While in Mozambique, they often retain the same role, working in urban areas or as laborers in mines or commercial farns. While in the rural areas, their main roles are productive, being involved in 65 Source: Ministerio da Agricultural e Pescas (1994), as quoted in Desai (1996), Table XIV. See further detail on food insecurity in the Appendix Tables. 51 land clearing; land preparation; and hunting for meat supply, as well as in construction of houses, livestock pens, and fences. In most ethnic groups, men make all managerial decisions regarding both the household and farm activities, even if they are away or abroad; and are the sole handlers of cash assets. 142. Rural women play basically three roles: productive; family care; and household tendering. In their productive role, they are in charge of producing food for the household, usually in two places: grain or tuber crops in the field, and vegetables in enclosed gardens. Except for land clearing (normally a male responsibility) and male help in land preparation, women are responsible for all crop husbandry, harvesting and, when relevant, home processing of food crops. In their familial role, they are in charge of nurturing the family, including preparing food for the family group, feeding and otherwise taking care of the children (with the help of elder sisters, if available), and providing health and other support to everyone. In their household role, they are responsible for water and fuelwood gathering, house cleaning, and clothes washing and mending. Time studies reveal daily schedules of 14 to 16 hours, of which the two largest time claimants are cropping and water supply, each with three to four hours a day.66 143. Between 19 and 25 percent of rural households are headed by women (Desai, 1996, p. 24). Female-headed households have smaller endowments than male-headed households in termns of land, education and agricultural implements, as is shown in Table 15. Female-headed households cultivate 1.2 ha on average, while male-headed households cultivate 1.9 ha; the difference is less on a per-adult basis (since male-headed households are larger) but still noticeable: 0.75 ha vs. 0.87 ha. As a consequence of their lower endowments (and smaller size) fewer female-headed households have non-agricultural earnings than male-headed households (10 vs. 26 percent), and are consequently more vulnerable to risk. Female-headed households are also more likely than male-headed households to suffer from food insecurity. The average rural female-headed household experiences food insecurity for 4.3 months during the year, compared to 3.5 months for male-headed households. Table 15. Characteristics of female- and male-headed households Item Female-headed Male-headed Area cultivated by household 1.2 1.9 Area cultivated per adult (ha) 0.75 0.87 Mean number of manual tools owned 3.7 6.3 Mean months of food insecurity 4.3 3.5 % of households which hire labor 12 21 % of heads with no schooling 80 45 Receipt of non-agricultural income (%) 10 26 Sources: Ministerio da Agricultura (1994), and Multiple Indicator Cluster Survey, as quoted in Desai (1996), Tables III-VII, VIII, XIV, 13 and 12. 66 Similar studies carried out in neighboring Tanzania have concluded that a major reason for infant malnutrition is the lack of mothers' time to give them more than 2 feedings a day, while infants require 4 to 5 to get the adequate amount of calories. 52 144. Polygamy is still frequent in rural Mozambique, but each wife is set up as a separate household within the family compound, with her own productive, reproductive and household responsibilities. Therefore, for all practical purposes, including farming practices and agricultural extension, as well as child rearing and health concerns, each wife's activities must be counted as a separate household or farm. E. Rural Social Infrastructure 145. There are three main types of "traditional" social organization in rural areas: (i) mutual aid groups (often groups within the clan); (ii) associations of paid laborers (e.g. groups of herders); and (iii) rotating saving-credit associations (Xitique), which most often involve women. These forms of organization are closely linked with the traditional authorities and cultural arrangements. They are not static and have shown considerable capacity to survive and adapt to external factors. 146. In addition there are at least three kinds of formal organization: (i) state-created cooperatives, (ii) "associations" linked to the Ministry of Agriculture, and (iii) associations of producers linked to the outgrower activities of the JVCS.67 147. The cooperatives were established on farms abandoned by the colonists or in the context of Government's "villagization" program. Even at their peak, they covered only 12,000 ha and accounted for less than 1.5 percent of marketed production. By 1986, cooperatives were already down to a third of that. Those which survived are centered on the provision of input distribution and marketing services to their members. The most successful among them are those in the periurban areas devoted to the production and marketing of fruits and vegetables (see Box 2). The future importance of the cooperative sector lies essentially in its potential role for promoting smallholder access to basic services (inputs, credit) and representing their interests in the negotiations with the other actors in the development process. 148. Given the bad name the cooperatives acquired, the Ministry of Agriculture pursued a policy of creating "associations" after 1986. These were often linked to projects run by or with the assistance of the Ministry. For instance, there were associations of irrigation farmers, or peasant associations linked to the Casas Agrarias at Ch6kwe. 149. Some of the JVCs have active policies of involvement of the outgrowers in certain decisions. For instance, LOMACO has encouraged the villages surrounding its outgrower schemes to form associations which both represent the outgrowers' concerns to the firm and which exert discipline over the activities of the outgrowers. 6 Source: Haas (1993). 53 Box 2. The Green Zone Women's Cooperative Movement The Maputo Green Zone Women's Cooperative Movement has its roots in Government's promotion of "people plots" to ensure the provision of fresh vegetables to the country's main cities in spite of the war. The General Union of Cooperatives (GUC) was later formed by the individual cooperatives to help members increase production and improve their lives. The movement has been a major success and has become a symbol of what women can achieve through adequate organization. There are now about 180 cooperatives covering 1,000 (mostly irrigated) hectares. There are about 5,500 members, of which 95 percent are women. They have established consumer cooperatives, training centers and creches. In response to changing opportunities, the cooperatives are shifting from basic vegetables to higher value products such as fruits and chicken. The GUC is emerging as a powerful force in the general democratization of the Mozambican society, and in protecting its members from outside threats. 150. Apart from the Green Zones associations, the Mozambican peasantry may be characterized as lacking "social capital". There do not yet exist strong peasant associations with the capacity to intervene in factor or input markets. This is a constraint in all lines of smallholder production, but especially in cotton, where the use of credit for insecticides is essential for profitability. F. Rural Physical Infrastructure 151. The national network of public roads is managed by the National Directorate of Roads and Bridges (DNEP) and comprises some 5,300 km of paved roads and 23,900 km of earth/gravel roads,68 of which 13,873 km are tertiary roads.69 Only about 30 percent of the network is in good condition, and 15 percent is not transitable on a regular basis.70 (This is an improvement on the situation in 1994 when only 18 percent was in good condition.) The feeder roads are in poor condition and many of them are impassable. As a result, the operating costs for road users are high and vehicular traffic is low and sporadic on much of the network. This situation is due to the lack of adequate funding for road maintenance in the past, as well as the abandonment of systematic maintenance due to the concentration on new construction in the 1 970s and the security problems of the 1980s. 152. The intercountry comparison in Figure 15 helps to put the Mozambican infrastructure situation in context. There is a strong association between GNP per capita and the amount of paved roads. The causality is presumably two-way: richer people wish to consume more road services, and better roads reduce transport costs, resulting in gains from trade, Notably, the ex- Soviet countries such as Lithuania are all placed well above the regression line, reflecting the 68 World Bank (1994b), p. 2. 69 Direccao Nacional de Estradas e Pontes (1996), para. 16.5. 70 Direccao Nacional de Estradas e Pontes (1995), p. 6. Data refer to December 1995. "Good condition" is taken to be the sum of the categories "good" and "fair". 54 slump in their national output during the post-socialist transition. Small island countries such as Hong Kong have far fewer roads than countries of the same wealth; conversely, countries with a long geographical structure and with low but evenly spread densities of population, tend to have more roads. Mozambique, with a per capita GNP of (very roughly) 2 percent of that of the United States, has an amount of paved roads equivalent, on average, to that of a country with double its wealth, viz. that of a country with 4 percent of the per capita GNP of the USA. The differential may be explained by Mozambique's elongated structure and the relatively even spread of its population across the land mass. Another factor is that the Portuguese settlers undertook the original road-building program for trading and military reasons at a time when Mozambique was much richer. 100,000 ; - d ; . ; ~~~~~~~I RFL!AI4 3-- o~~~~~~~~~T U L TI IS* O . , + _ _ .1 CO _ _ _ _ _ _ _ _ _ _ _ _ I..o ___ - |~ F G K}NG i a , _ _ _ - _ _ _ _ i : i_ _ 10 1 10 100 PPP estimate of GNP (US=100, 1993) Figure 15. GNP and roads, all countries, 1993." 153. Road rehabilitation is being carried out in a single integrated, holistic project, supported by donors and Government. The project, the Second Roads and Coastal Shipping Project (ROCS 11), started in 1994 and is to run until 2000. The total cost is projected at USD815 million, of which the share of IDA is 188 million. It comprises (a) emergency rehabilitation of 11,700 km of unpaved roads to open access in all ten provinces; (b) rehabilitation of 3,450 km of priority trunk roads; (c) labor based reconstruction of 3,250 km of feeder roads; and (d) current routine and periodic maintenance of that portion of the network that is in good or fair condition. The main trunk roads slated for upgrading in the first phase of the project are shown in Table 16. 7' Data source: World Development Indicators 1995. Table 30: PPP estimates of GNP per capita, Table 32: Paved road density. 55 Table 16. Second Roads and Coastal Shipping Project: Phase One Trunk Road Rehabilitation Province From - to Length (km) Cabo Delgado Pemba-Montepuez 199 Nampula Nampula-Nacala 199 Nampula-Angoche 170 Zambezia Quelimane-Namacurra 69 Tete Vanduzi-Changara 260 Manica Inchope-Machipanda 152 Sofala Beira-Inchope 134 Gaza Xai-Xai-Choengoene- 75 Chibuto Maputo Marracuene-Manhica 49 Maputo-Ressano Garcia 89 Boane-Moamba-Sabie 39 Maputo-Namaacha 75 Total 1510 Source: World Bank (1994b), p. 42. 154. Part of the ROCS II, the UNDP-coordinated Feeder Road Program, is undertaking labor based reconstruction of feeder roads in 40 priority districts. "Feeder roads" here are understood as secondary and tertiary roads. Funding comes from UNDP, ASDI/SIDA, Swiss Development Cooperation, GTZ, ODA, IFAD, USAID, the EU and other organizations. The total amount of money committed for the period 1989-2000 is USD 70 million.72 During 1995 a total of 487 km of feeder roads were rehabilitated, of which 318 km were full gravel and 169 km spot improvement. The progress of the project is shown in Figure 16. The total output of the Feeder Roads Program so far is 1,783 km of rehabilitated roads, of which 1,358 km have come under regular maintenance.73 The project is thus roughly half-way to its objective of 3,250 km. 72 Direccao Nacional de Estradas e Pontes (1996), Table 4. 73 Direccao Nacional de Estradas e Pontes (1996), para. 16.3. 56 500 400 Km. 300 - 200 - 100 1989 1990 1991 1992 1993 1994 1995 Figure 16. Feeder road rehabilitation (km. per year)74 155. Figure 17 contrasts the amount of rural roadways and the area of cultivation, by province. As one would expect, there is a close relationship between the two variables, since each creates demand for the other. On average, one kilometer of roadway corresponds to about 100 ha of cultivated land. In the case of Nampula, the most agriculturally important province, there are about 200 ha of cultivated land per kilometer of roadway. If a rural road-building project were undertaken, the highest-return investments would probably be in that province. 75 74 Source: Direccao Nacional de Estradas e Pontes (1996), Table 5b. Data refer to the sum of full gravelling and spot improvement. 75 Part of the reason Nampula has fewer roads per cultivated hectares is that certain key trunk roads linking it to Zambezia province have not yet been rehabilitated. But even when these roads are rehabilitated, Nampula will still be under-served compared to the other provinces, in the sense of having a higher ratio of cultivated area (and output) to road length (viz. a ratio about 1.6 times that of the other provinces compared to about twice as high without this correction). 57 10000 8000 6000 4000 2000 0 o c o ' a @J , , 0 o E N E L) 0 I Road Length (km) U Area under Cultivation (x100 ha) Figure 17. Road length and area under cultivation, by province76 156. From the point of view of supporting agriculture and the rural economy, the highest- return investment is to rehabilitate all-weather access under the existing rural network and then maintain it. The highest priority in rehabilitation, from the agricultural point of view, should be given to Nampula province. The main constraint to the rural road rehabilitation program is the lack of resources for maintenance." This concerns, among other things, trained personnel, and in particular lengthmen, to undertake the maintenance tasks required. This constraint is being addressed through training courses in the Second Roads and Coastal Shipping Project and in the Feeder Road Program. For the next five-year period it will not be possible to increase the amount of road rehabilitation and maintenance beyond that which is already planned, and it will not be possible to engage in any substantial amount of road-building (which in any case bears lower rates of return than rehabiitation and maintenance of existing roads). Nevertheless there can be reallocations of resources within the program. One reallocation which might be investigated is to concentrate more resources on feeder roads in Nampula province, which appears to be under-served relative to the other provinces. Another possible reallocation is to spend more on farm to market tracks and culverts in rural areas. 157. Roads drive agricultural growth by reducing the costs of road transport and increasing the availability of transport, which in a virtuous cycle creates new markets which in turn result in increased demand for transport services. Research in other countries turns up elasticities of agricultural output with respect to road investments of as high as 20 percent.78 The returns to 76 Road length from Direccao Nacional de Estradas e Pontes (1996), p. 9. Length includes primary, secondary and tertiary roads, for a country total of 26,285 km. Area under cultivation (x I 00 ha), is from 1992 data from satellite surveys of land-use cover, carried out under the Early Warning System of the Ministry of Agriculture; see Table 8. " The sma1l size of the local road construction industry presents another constraint. The industry is growing, partly due to support from ROCS 11. 78 See Binswanger et al. (1993), p. 364. The data are time-series observations of districts of India, 1971- 81. Antle (1983) finds elasticities of agricultural output to infrastructure investment of between 0.20 and 0.25 in an intercountry model. With household-level data from India, Antle (1984) finds an elasticity of output with respect to distance to the nearest road of -0.056. 58 maintenance, given that a network already exists, are much higher than this. While the Second Roads and Coastal Shipping Project will result in sufficient distance and length in the major trunk roads to permit the linkage of markets between north and south, there will still be a need in the future to improve market access in rural areas, and to build more feeder roads, farm to market tracks, and culverts. Future transport planning should look into the possibility of support to a "spot improvement" program for "farm to municipio" roads, in those provinces where this is feasible. In this context, it is possible that a series of small bridges in rural areas to help people ford rivers during the rainy season would bring higher returns than more feeder roads or bulldozing of tracks. It is essential, therefore, that the current plans for road-building take the need for access road improvement into consideration. The planning of rural road rehabilitation by the National Directorate of Roads and Bridges should be closely linked with the Ministry of Agriculture's plans for promotion of development. 158. A new way of addressing the problem of infrastructure at the local level could be a rural investment fund (RIF) of the kind that was started in Zambia in 1996. Local communities would apply to the fund for matching grants for specific purposes, typically farm to market tracks, culverts or other items of social infrastructure. The justification for the subsidy is that these items of infrastructure would not ordinarily be undertaken by private agents. It may be more efficient to permit decision-making at the community rather than the ministerial or provincial level, as local communities would be best equipped to evaluate the benefits arising from competing infrastructural investments. For instance, some communities might prefer dams over tracks or culverts, and this possibility should be allowed for in the design of the rural investment fund. 159. Another important component in the effort to provide long-distance transport linkages is coastal shipping. As a part of the First Roads and Coastal Shipping Project (not to be confused with the Second, treated above), the state-owned shipping company has been privatized and the firm must now compete against other firms which can hire its shipping capacity. One private company has succeeded in using a foreign vessel which plies longer East African routes and which now stops at Mozambican ports as well. These measures have resulted in a real reduction in the price of coastal shipping. In terms of the same project, several secondary ports are to be developed, two of them in Nampula province. 59 I Chapter 4. THE INSTITUTIONAL FRAMEWORK A. The Legislative Framework (a) The Land Law79 160. The basic legislation regarding land rights is contained in the Constitutions of 1975 and 1990, the Land Law 6/79 (1979), and the Land Regulations (Decree No. 16/87, enacted in 1987). Both Constitutions state that "the land and natural resources in the soil and subsoil ... are the property of the State". It also states that "In the People's Republic of Mozambique land cannot be sold or in any way alienated, rented, mortgaged, or pawned." The Land Law of 1979, with the clarifications and rules contained in the Regulations of 1987, govern both the family and commercial sectors. They are comprehensive documents which set out general principles; provisions relating to the use of land for agricultural and non-agricultural purposes; principles relating to protected zones; the legal competence of different authorities to approve of land use; the transmission and extinction of rights of use; the creation and maintenance of the National Cadastre and Registration. Customary landholders are recognized in this legislation though no specific rights are conferred. 161. In addition, Decree No. 21/89 was enacted to specifically cover the divestiture of state farms. The State would retain the ownership of the land, but would sell the "enterprises", buildings, equipment, infrastructure, and crops. There are provisions for bids, valuation, public auction, reserve prices, and bid deposit. In practice, however, this Decree has been honored mostly in the breach. 162. Management of Protected Areas. The Ministry of Agriculture, through its Directorate for Forests and Wildlife, is responsible for managing the areas gazetted as protected (national parks, game reserves, coastal reserves). The combined effect of the war, limited budgets, and inadequate personnel, however, have resulted in minimal activity in the field. Only two national parks have today a minimal semblance of protection activities - in one case, with funding from the South African chapter of WWF. The proposed GEF Transborder National Parks project, scheduled for appraisal in FY96, should provide resources for recruiting, training, equipping, and fielding additional personnel in several locations. 163. Information and Documentation Services. The Ministry of Agriculture houses and runs the Center for Agricultural Documentation, CDA, which leads a network of documentation centers in every department and institute in the Ministry, plus the Eduardo Mondlane University and INDER. FAO has provided technical and financial support for bibliographical procedures, equipment, and training. 164. Land Allocation and Registration. Only three categories of authorities are legally competent to allocate land. They are the cities' Executive Councils, which deal with urban lands; the districts' Executive Councils, which deal with rural lands; and the Council of Ministers, which deals with land issues at the national level and which is empowered to overrule 79 For further detail see Noronha (1994) and World Bank (1996c). 61 the former if required to defend the national patrimony. The National Planning! Commission (CNP) is empowered to submit for the consideration of the Council of Ministers proposals for the creation, alteration or extinction of land rights. 165. The Ministry of Agriculture, through its National Directorate for Geography and Cadastre (DINAGECA), is charged with the management and maintenance of the National Land Cadastre; administers the State Patrimony; can grant concessions for the use of land for agricultural purposes; and has the right to intervene in the creation or extinction of protected areas. The Ministry of Agriculture can grant concessions for agriculture up to 2,500 ha; for livestock, up to 5,000 ha; and for silviculture, up to 10,000 ha; Provincial Governors can grant similar concessions up to 10 percent of the above amounts. 166. Members of the smallholder sector have their right to occupy and use land guaranteed by the Constitution. Thus, they are not required to register the lands they use though they can often obtain a Certificate of Family Occupation or a Title of Use and Exploitation if they choose. All other applications for agricultural exploitation of land must be submitted to DINAGECA, which is also responsible for producing official land maps, subdividing the national land into lots, issuing land titles, and maintaining the national land registry. Normally, the registration process takes at least one year and is required of private or commercial farmers. The failure by DINAGECA to investigate properly whether the lands being applied for are actually occupied or used by smallholders (including the fallow period in their extended fallow cropping system) has often resulted in the loss of land by smallholders without compensation. In actual practice, the non- requirement of registration has left them without protection in the process to register new land allocations to the commercial sector. In some areas of high land pressure, smallholders have resorted to registering their land holdings as a defensive measure. Implications for tenure security and investment 167. Most smallholders are unable to rely on the law to get access to land, protect themselves from eviction and resolve land disputes. They lack the resources to employ formal procedures to obtain titles or certificates of occupancy or to the courts to resolve disputes. Instead they rely on community-based systems of land administration (traditional authorities, associations, cooperatives, and other emerging community formations) which are not always legally recognized. These community-based systems have no formal voice in land titling and state farm divestiture. 168. While displaced persons have often been allocated land, their claim to the land is insecure. The rights of women are often governed by contradictory and changing customs and laws and are also insecure. Since the law does not recognize the many transfers of land use rights in the market, those purchasing these rights face great insecurity. 169. Since 1992 large land concessions have been granted both by the provincial governments and at the national level. Though the official repositories of land concession documents are not sufficiently maintained to permit precise verification, there is evidence - on the basis of conservative calculations -that the central government has granted over seven million ha for agricultural, hunting and mining purposes, viz. 19 percent of the total arable area of the country of 35.9 million ha. If to these are added provincial concessions and reported cases of concessions, the total rises to eight million ha, or 22 percent of all arable. If we add state farms and existing private agricultural enterprises, the total is 9.1 million ha, or 25 percent of the arable 62 80 land area . Since 1993 there have been further concessions, so these numbers should be taken as a lower bound. 170. Some of these land concessions and titles have created very large private sector farms, despite the lack of evidence that large farms are more efficient than smallholders. A disturbing feature of the land allocation process is that the allocations tend to be made in areas where land is scarcest, raising the specter of dispossession of traditional holders and land conflicts in the future. Between 1990 and 1993, 637,000 ha of the 724,000 ha of concessions recorded by DINAGECA were granted in Maputo province. Smallholders in Maputo province (and Gaza and Inhambane) are already in a position where they do not have enough land to maintain their soil fertility by fallowing. 171. Many of these concessions, if not all, contain at least some areas currently occupied by smallholders or capable of being exploited by them. Given the lack of titling or registration requirements for smallholders, they are in many cases in danger of losing customary rights to land use. The titling has been justified on the grounds that the areas are allegedly "empty" or "unused". Yet DINAGECA does not have the personnel to go out to all the areas and do a thorough inspection in loco, so that the allegation of "empty land" is questionable. Furthermore, there is reason why some areas are unused: during the war people had to flee their traditional lands, leaving them temporarily unoccupied. Upon returning they lacked the wherewithal to farm the areas that they used to. Outside investors have on occasion taken advantage of the temporary and artificial vacuum to stake their new claims. 172. Furthermore, the titling procedure has enabled people in effect to obtain large tracts of land virtually gratis. The negligibly low price gives an incentive to people to accumulate land in the hope of selling once the legal environment has changed. The low price discourages the efficient use of land. The land tax, presently $0.50 per hectare, does nothing to correct this distortion. In any case, fewer than 30 percent of "privados" are actually paying their land taxes. 173. Defacto practically all the state farms have been subdivided and occupied by various claimants, usually in locally negotiated processes with little transparency. Distributionally unfair - and possibly economically inefficient - though the process was, these de facto allocations could not be revoked without provoking political crises. The government will have little choice in the end but to rubber-stamp the existing pattern of de facto ownership. In the meantime, the delay in deciding on the ultimate fate of the state farms creates insecurity for the current occupants and likely discourages long-term investment by them. Suggestions For Land Policy 174. Land policy can be implemented consistently only if there is a broad political consensus on its appropriateness to the country. While a consensus is emerging at the technical level, a political consensus is still needed. The following suggestions provide a basis for a viable land policy: Locally controlled land administration. Local community-based land administration could be given legal recognition.81 These communities may include traditional communities or 80 Sources: Myers, Eliseu and Nhachungue (1993), Myers (1994). See the Appendix Tables for further detail, methodology and sources. See also the discussion in World Bank (1996c). 63 arrangements based on cooperatives, religious groups, and other groups and associations. These groups, whether traditional or newly formed, should be given legal standing and have sole rights to land allocation in their area. To strengthen the rights of women, such communities could be encouraged to include women in their leadership structures. The state could assist the communities in delineating their boundaries. Conflicts among community members, or between communities, which cannot be resolved at the local level, may be referred to higher-level arbitrage institutions, such as locality land committees. These land committees could include members from all stakeholders, including the communities, the private sector, women, displaced persons, etc. Only conflicts which cannot be resolved by the land committees should be referred to the judiciary. * Private rights to land. In high population density zones with land conflicts, farmers could be given secure and tradable private rights to land, on either a leasehold or a freehold basis. Individual, corporate and foreign ownership could be permitted. Titling on demand could be replaced by area-based titling, in which all claims to land within the specified area should be adjudicated simultaneously. * Transfer from community-based to private ownership or leasehold. Communities wishing to cede land under their control to private control by outsiders could be enabled to do so by consensus or majority vote. * Divestiture of state farms. Where titles have been granted to large farmers for areas partially occupied by smallholders, the titles could be reissued excluding the smallholders areas, for which certificates could be issued. It has been alleged, for example, that certain citrus developments resulted in the ejection of smallholder families; a process of redress is needed so that the aggrieved parties can regain their lands - after which, of course, the citrus developers would be at liberty to negotiate with them and continue the development at an agreed price. * Settlement policy. New land could be opened up in Tete, Niassa and Cabo Delgado by building roads which would result in spontaneous settlement. Employment and foreign exchange earnings could be increased by permitting foreign and large-scale investors to settle in these parts. In all cases, it would be advisable for settlers and investors to negotiate first with the individuals and communities which claim ownership of the land. Priority in road-building and provision of other infrastructure could go to schemes which favor employment and smallholder production, e.g., nucleus estates with outgrowers. It should not be necessary, nor would it be efficient, to introduce special tax breaks and subsidies to colonize the "frontier". * Land taxes. It would be ideal to raise the land tax rate to a level which discourages speculation and encourages the rational use of land. Land taxes levied on the unimproved portion of land values do not distort work incentives and so constitute the ideal tax; but legions of economists since Henry George have so far failed to convince even a single government to impose a land tax equal to the land rent, owing to farmer resistance. It is feasible (and commonly observed), however, to have a light rate of tax paid to local authorities for use in local projects which have a bearing on the welfare of the rural farmer. Beside inducing more efficient land usage, this strategy would foster decentralization of economic decision making, a valid goal in itself. At least one-quarter of all land in smallholder possession was obtained by allocation from a traditional authority. See Appendix Tables for detail on the sources of acquisition of land by smallholders. 64 1 75. The following are some recommendations for immediate action: * The first priority is to accelerate the process of political consensus-building. Support for the process is needed from the highest level of government, and resources should be made available by the donor community for this purpose. Especially important in this regard is that there be a process of consultation with local communities, particularly rural communities. * Until such time as a new land law is in place, the negligible cost of land to well-placed individuals invites abuse. To stem this possibility, a moratorium could be placed on all titling on demand. * For areas where title has been granted or applied for which include settlements of smallholders a moratorium could be imposed upon evictions of smallholders. The Land Commission and the revision of the land law 176. In September 1995 the Ministry of Agriculture and Fisheries issued the Politica Nacional de Terras e Estrategia de Implementagdo (hereafter referred to as the National Land Policy document), in response to the growing concern of the public at large that the old land law was out of step with the new realities of Mozambique. The National Land Policy was a considerable improvement upon the existing land law in that it (a) explicitly recognized customary tenure; (b) required that outsiders who wish to obtain access to land must negotiate with the customary right-holders to these resources; and (c) recommended that communities be permitted to receive compensation, financial or other, for land which they temporarily release to outsiders as concessions. It is still too early to say, however, whether the new land law will incorporate these progressive recommendations. 177. The government proceeded to establish a timetable of activities which would eventually issue in the formulation of a new land law. These included upgrading the Land Commission of the Ministry of Agriculture to an Inter-Ministerial Land Commission; a process of discussion and consultation with local communities; provincial seminars and a national seminar; ongoing work on legislative revision; and several studies. A detailed work program for the Land Commission was developed (Comissao de Terras, 1996). The Interim Program was to run for 1996 and the National Land Program from 1997-9. The Interim Program started with a six- month period of discussion and consultation with local communities, provincial governments and representatives of other national interests concerning the National Land Policy document. A draft land law was discussed at a National Land Conference in July 1996. Studies were approved and financed, on land conflict, community-based land registration, regional-level land management systems, the capacities of DINAGECA and the Registo Predial (the urban registry office), and the future permanent land/natural resource management institution. 178. The judicial group of the Land Commission will continue the process of revising the land law until it is presented to the National Assembly in 1997. rNIA will undertake a National Land Use Plan, with FAO assistance. Support for the legislative review process has been obtained from a range of donors, including IDA. 65 (b) Water 179. The Water Law defines water as a publicly-owned resource; provides a general management policy for those resources; determines the framework for protecting, using, developing, and taxing water resources; and defines the role of the Government in the field of public water resources. The main institutions with regard to water are: * The National Water Board (Direc,co Nacional de Aguas) part of the Ministry of Construction and Water. * The Secretariat of State for Agricultural Hydraulics (SEHA) which oversees large scale irrigation schemes. a Regional Water Boards 180. The Law envisages the creation of Regional Water Boards (ARAs) to: participate in the preparation, implementation and revision of each basin's hydrology use plan; administer and control the public water domain; maintain a register of waters of private use; introduce and charge water user fees and taxes for using infrastructure built by the public sector; design, build and operate hydraulic works on own account, and approve hydraulic works by others and inspect construction; provide technical services; and collect and keep hydrological statistics. In addition, ARAs should also promote the establishment of water user associations. 181. Management of Irrigation Systems. Two parastatals are in charge of operating and maintaining the two largest irrigation schemes: SIREMO for Chokw6, and SRBL for Xai-xai. Both have now practically stopped all their activities, owing to a number of problems, including shortage of qualified staff and machinery; management problems; the impossibility of recovering water costs and long delays in budgetary allocations. As a result, the schemes are rapidly deteriorating, with problems of salinization, alkalization, and waterlogging. 182. The schemes irrigating the three sugar estates (with a total of 36,000 ha, of which only 5,400 ha were being operated in 1992) are managed by the estates themselves. These, together with smaller schemes, suffer from the same problems mentioned above. As a result, only a small fraction of the total area equipped is actually under irrigation, water delivery is not organized in a rational manner, and maintenance is almost nil. 183. In 1992, the Secretary of State for Irrigation was relieved of his position and the Secretariat itself (SEHA) was put under the direct supervision of the Minister of Agriculture. In the same year, a General Coordinator of Integrated Projects (CGPI) was appointed, subordinated to the Minister, who took over from SEHA the operation of most irrigation projects. (c) Marketing, Pricing and Incentives 184. Prices and Incentives. After independence, producer prices for agricultural commodities were administratively fixed on a pan-territorial basis. After an increase in 1977, the nominal prices of most commodities were kept constant until 1982. Between 1980 and 1986 inflation accelerated and real prices declined. In the case of beans and rice, the real 1986 price was lower than it had been in 1977, as is shown in Figure 18. 82 See the discussion about small-scale vs. large-scale irrigation schemes in Chapter 3. 66 16 12 8 4 u * , , . .. . . . . . . . . . . . 1976 1980 1984 1988 1992 --+-Maize - Rice Beans Figure 18. Real prices, 1976-92, constant 1980 Meticais per kg.83 The real price fall between 1980 and 1986 was even more accentuated for export crops. The real price of cashew fell by 35 percent, and that of cotton by more than 50 percent. Falling real producer prices were mirrored by a sharp drop in agricultural production, as is shown in Figure 19. In the case of beans, for instance, marketings reached their nadir in 1985, but with the rise in price from 1986 on marketings improved, attaining the record level of 23,000 tons in 1993. 160 25 140 120 -20 .~100 E ~~~~~~~~~~~15 80 i 60 1l m 40 20 0 0 1975 1980 1985 1990 1995 -4-Maize -4-Rice -+--Beans Figure 19. Marketed agricultural production, 1975-94 ('000 metric tons)84 The first step in the realigning of real prices and incentives was the large devaluation in 1987, with concurrent large increases in nominal producer prices (between 300 and 600 percent). These increases translated into real price increases of 30 percent to 1 00 percent depending on the crop, with the largest increases benefiting the export crops. After this initial burst, periodic devaluations have induced a gradual real depreciation of the exchange rate which has contributed to an increase in nominal producer prices. Real agricultural prices showed little trend until 1992, 83 Data: Dorosh and Bernier (1994). The evolution of the nominal and the real prices of food and export crops during the 1977-1992 period are presented in the Appendix Tables. 8 Data: Informacao Estatistica, CNP: various years. See the Appendix Tables for further detail. 67 when they increased again significantly. Higher real producer prices were accompanied by a restoration of agricultural production between 1986 and 1992. The increase in production was particularly large for the main export crops and for maize. Marketed production of maize increased by 165 percent between 1986 and 1991. 185. While real producer prices were being raised, the number of agricultural commodities subject to administered prices was reduced from 44 in 1986 to less than 20 in 1990. The fixed producer prices were replaced with minimum producer prices. Currently nine smallholder crops are subject to these minima (Table 17). The government's justification is to protect them from predatory practices by monopsonistic traders. The minimum prices are calculated by the Ministry of Agriculture on the basis of estimates of "fair" marketing/processing costs and farm production costs, and are approved by the National Commission on Wages and Prices (CNP). Recently, official minimum producer prices have been increased in line with inflation and depreciation of the metical. However, they are still below the import parity price in most 85 cases. Table 17. Price Regime of Major Agricultural Commodities (1993) Minimum Producer Price Maximum Consumer Conditioned Price Price Beans Yellow maize Rice Groundnut Yellow maize flour Milk Sunflower Wheat flour Sugar Cashew nut Bread Beef Cotton Whole rice Cooking oil Mafurra Tea Paddy Frozen fish Tobacco Fertilizers White maize Pesticides Agr. implements Note: Conditioned prices are prices set by firms on a cost basis and subject to ex-post review and approval. 186. It is difficult to see any value in a single national minimum price given that demand and supply change over seasons, over years, and across space. The official minima are constantly to be out of step with market realities. The agency setting the minima cannot know in advance whether, for instance, maize will be in surplus or deficit in any given year. Under free market conditions, the price at the farm level would differ vastly depending on whether there was a regional deficit or surplus. Table 18 shows that the variation of the maize price between surplus and deficit years can be enormous. In a deficit year, a farmer in Western Nampula could get Mt440/kg or Mt25 1/ kg for maize sent to the Maputo market; in a surplus year she could get only Mtl 05 or less. If the government agency set the minimum price "somewhere between", it would 85 It is argued elsewhere that these minima are partially effective - even though the government does not intervene in the market to defend the price - because traders are sometimes prosecuted for buying at lower prices. See para. 186. 68 not benefit farmers at all in deficit years and would harm them in surplus years when the traders would not buy at sub-minimum prices for fear of penalties - which have sometimes been applied, in the form of jail sentences and fines (Coulter, 1995, p. 37). There are also reports of traders having their businesses closed because they purchased farm products at less than the minimum price (Ministerio da Agricultura e Pescas, 1995e, p. 58). A minimum price for a product characterized by great output variability makes sense only if the govemment has the institutions and the resources to defend the price floor by becoming the buyer of last resort. Such a role is, however, not on the cards, for reasons which are discussed in Chapter 5. Table 18. Determination of On-Farm Maize Prices Based on Import Parity Location and transport assumptions Deficit Year Surplus Year I Price to farmer (Mt/kg) in Western 440 105 Nampula province, assuming shipping to Maputo by sea 2 Price to farmer (Mt/kg) in Western 251 -84 Nampula province, assuming shipping to Maputo by truck 3 Price to farmer (Mt/kg) in Manica 1046 661 province, assuming shipping to Maputo by truck Source: Coulter, 1995, Appendix 4. Assumptions: All prices refer to early 1995. Exchange rate Mt9,000=$1US. Importation from Gulf in deficit years. Competition with South African maize exports to Maputo in surplus years. 187. One of the main benefits claimed for minimum prices is that they reduce risk, thereby promoting production. The manner of their administration in Mozambique has sometimes increased rather than decreased uncertainty. For example, in 1994, the announcement of the minimum price for maize was delayed until April 20 (after the harvest), and the anticipated price of Mt700/kg was reduced to Mt500/kg (Ministerio da Agricultura e Pescas, 1995e, p. 57). 188. Further, the minimum prices are defacto interpreted/used as official/maximum prices by traders, thus acting as a tax on producers. To the extent that the minima are enforced, they serve to close down markets. They would best be abolished in favor of strengthening the existing market information system (SIMA). SIMA is an initiative aimed at the development of agricultural marketing in Mozambique, and is part of the USAID-funded and MSU-supported Food Security Project. It collects price data for 30 product classes in 25 localities around the country (Coulter, 1995, p. 18). Information is fed to users in several ways: through information bulletins, the local media fax service, and through the public broadcasting service, though the latter has encountered implementation problems. There is clearly a strong demand for this sort of information. In other African countries the broadcasting of market information is an unusually effective tool with which the public sector can influence the performance of food markets. It is to be regretted that progress with radio broadcasts has been halting as this could become the farmer's best defence against local monopsony and collusion. 69 189. The only consumer prices controlled by the Government are those of foodstuffs imported and distributed under the commercial (and not emergency) food aid program. The maximum prices are set much below import parity, in line with the humanitarian objectives of donors. The Government contends that this does not depress the price of substitution crops, but there is growing evidence that it does and that it represents a major disincentive for domestic production (page 87). The simplest solution would be to auction the commercial food aid grain instead of setting a price. 190. Input subsidies. The prices of some of the main inputs and agricultural equipment are currently heavily subsidized: seeds, implements, insecticides, fuel. Others are sold through private marketing channels at import parity prices. The subsidies are justified as necessary due to the inability of many farmers to acquire the cash to pay for them. During the resettlement period (especially 1992-94) newly resettled families, or deslocados, were dependent on the provision of AgPaks, containing minimal requirements for cultivation, to re-establish themselves as self-sustaining producers, but these distributions have since been wound down (see page 84). There is no objection to such distributions in the context of emergency assistance. As long-term development policy, however, seed subsidies are a low-return investment and possibly worse. Even if the seeds are improved, open-pollinated, drought resistant varieties capable of being reproduced by the small holders themselves - the standard justification for their subsidized distribution - it would be better to cost them fully. For if the improved seeds are a profitable investment, smallholders will find the means to purchase them. The income distributional impact of subsidies is likely to be disequalizing as larger farmners are better placed to take advantage of subsidized materials. Cheap distributions of seeds also create an attitude of dependence on the part of farmers, and create demands for the continuation of the subsidies beyond their planned time span. 191. Table 19 presents the impact of the current price regime on the protection and financial profitability of agricultural production. It indicates that the production of the main crops is still substantially explicitly and implicitly taxed: net protection coefficients (NPCS)86 and effective protection coefficients (EPCs) below unity due to inefficiencies arising from the trading system and the taxation and price regime. This taxation includes the overvaluation of the metical ( 10 percent in 1996) and depresses farm incomes and the potential supply response. Table 19. Protection and Profitability of Major Crops, 1993 Nominal Effective Domestic Financial Protection Protection Resource Cost Profitability Coefficient Coefficient Coefficient in 1993 (NNPC=I) Meticaisg Maizea 0.8 0.8 1.3 Beansb 0.6 --- 0.8 GroundnutC 0.6 --- 0.4 Cashewd 1.0 --- 0.3 10,800 Mt/day Cottone 0.7 1.0 0.9 4000 Mt/day Paddyf 1.02 --- 0.7 NPC: Ratio of the domestic price, given the intervention, and the border price in the absence of intervention. If the NPC is less than unity, the product is taxed. EPC: Ratio of value added in domestic prices to the value added in border prices. 70 Source: Moll (1993). a Small farm using local seed; yield 0.4 t/ha; no purchased inputs. bYield 0.6 t/ha; excludes irrigation costs. c Yield 0.25 t/ha; small farrn; rainfed; no purchased inputs. dYield 52 tlha. 47 days worked at 3150 Mt/day (viz. 50 percent above the minimum wage in 1993 of mt 21 00/day). 1500 Mt farmgate price. Outgrower production; 78 family labor days at Mt 31 50/day (viz. 50% above the minimum wage); tield 500/ha Yield 7.0 t/ha; large farm; irrigated. g Exchange rate in September 1993: Mt4,400/$. Unofficial rate: MT6.000/$. (d) Taxation, tariffs and licensing 192. Company tax. Up to 1994 the tax rate on profits was 50 percent throughout. In 1995 it was changed to: 35 percent in agriculture, 40 percent in industry and 45 percent in trade and services (Menestrel 1995, p. 87). While the general lowering of the tax rates is probably to be applauded, their differentiation by sector is arbitrary and since it distorts incentives it is economically inefficient; a single rate - say 40 percent - would have been better. The matter is not of great importance for agricultural policy because farmers worldwide pay trivial amounts of tax and it is unlikely that their payment rates would be any better in Mozambique. 193. The foreign trade regime. Import and export taxes were revised in June 1991 and again in November 1996, as part of the Third Economic Recovery Credit (see Ministerio da Agricultura e Pescas, 1 995b). Both export and import duties decreased. Export duties for all products but cashew are now zero. As less developed countries go, Mozambique has a relatively flat customs duty structure. As a general rule, essential goods such as medicines pay no duty, the rate for raw materials is 2.5 percent, for capital goods and spare parts is five percent, and for internediate goods is 7.5 percent. Consumption items are taxed at 35 percent or higher.87 The new 1996 regime eliminates practically all the scope for discretion. In addition there are charged a general customs 87 The Governrment proposed, during the negotiations of the Third Economic Recovery Credit in 1996, to apply a seasonal tariff to five agricultural products (maize, rice, beans, groundnuts and sugar). The intention, presumably, was to have a high tariff during the harvest season and a low tariff in the "hungry" season. This would cut off imports and protect producers during the harvest season and help consumers during the "hungry"season. There are several'objections to this arrangement. (i) The repeated changes in the tariff, and the dates selected for so doing, would become the object of speculation and would provide scope for patronage. (ii) This rough attempt at price stabilization would rob farmers and wholesalers of the incentives for inter-seasonal storage. The market would itself develop a certain degree of price stabilization - particularly if the government helps with on-farm storage technologies through the research and extension system - and the government is unlikely, through blunt tariff instruments, to be able to improve on private market outcomes. (iii) High tariffs are harmful to welfare as a whole because there are "deadweight losses" involved in the process of redistributing from consumers to producers. (iv) There is no obvious reason why the producers "need" a high tariff. If the problem is dumping, then this would best be addressed using the methods of the World Trade Organization, which would require an objective definition of "dumping" and would issue in a ban on imports from the relevant country. If the problem is long-term, permanent protection in neighboring countries, then no tariff is called for because it would not improve on the free trade outcome - the tariff would merely redistribute from consumers to producers and entail unnecessary deadweight costs. 71 charge (emolumentos gerais aduaneiros) on all products but those associated with the emergency or defence. and a consumption tax (imposto de consumo) of between on certain products. 194. A better arrangement would be to abolish the consumption tax and have a single rate of customs payable on all products without exception or exemption. Real resources would be saved by the government (fewer customs staff) and by the private sector (shorter delays, fewer bribes, and less work "proving" that a particular item falls under a particular category). 195. Turnover taxes and licensing. The imposto de circulaqdo (circulation tax), a tax on turnover, applies to all transactions. The rate is usually 5 percent for wholesalers and 10 percent for retailers (but I percent and 3 percent respectively for maize - Coulter, 1995, p. 18). The imposto de consumo (consumption tax), also imposed on turnover, is applied only once, at the production stage or upon entry into the country, and applies to only certain products. Rates are 20 percent and more. The imposto de circula do is cascading and, to the extent that it is enforced, penalizes the act of exchange, favoring backward and forward mergers. A superior altemative would be a value added tax (VAT), which has been introduced in some less developed countries despite the high accounting standards required; the Government plans to introduce a VAT in 1998. In the meantime, given the difficulties of verification in rural Mozambique, a better alternative would be to abolish the imposto de circulaqdo for rural traders, which would result in only minimal tax losses since it is poorly enforced. There may be an overall gain for the government because the costs of inspection would be reduced. The imposto de circula9do would be retained, then, only for wholesalers, who are more easily monitored, and the cascading nature of the tax would be eliminated. 196. Trading licenses. Obtaining a wholesaler's license is an arduous process, the most time- consuming step being the securing of a police record. While Ministry of Commerce officials are convinced that the clearance does not take longer than a few weeks, there is plenty of anecdotal evidence that it can take many months. The current law has a provision requiring the applicant to show a bank account containing a certain amount of money. As of early 1995, certain applicants had to show a bank statement with a balance of Mt 27 million, worth $3,375 at the time. The Ministry contends that it is no longer enforcing this requirement; but there is considerable confusion as to what rules apply. It is possible that there are officials who are exploiting the confusion for their own advantage. It is difficult to see why the licensing requirements should do any more than ensure that the licensee fulfills her tax obligations and obtain statistics; the bank statements can only complicate the application process and serve no useful purpose. It is recommended that this extraneous requirement be scrapped in favor of simple licensing. The Ministry of Commerce is presently developing a revised law, but copies of the draft are not yet available. 197. Similar barriers exist in the case of retailing. Retail licenses are approved by the provincial governors. The cost of obtaining one was some $50-90 in November 1993. As of November 1995, the cost was Mt400,2 10, or about US$ 40. Applicants for a retail license must submit a list of documents, including a police security check that must come from Maputo and which, according to interviews with licensed traders, can take several months. As is the case with wholesalers, there is confusion over whether applicants must submit a bank account statement showing a specified balance. Senior officials in the Ministry contend the rule is not being applied, but at the local level it appears that it is; sometimes US$1000 is required and 88 "Comercio Privado: Comissao Perrnanente da Assambleia Popular, Lei No. 7/79; Diploma Ministerial No. 47/80", 2a Edigao, 1989. 72 double that in Manica. This requirement obliges the trader to borrow money, frequently from a wholesaler, creating future obligations in the process. The effects of these regulations, combined with confusion about their application, are to discourage registration by traders who have already entered the business, and to reduce the incentives for entry. Better would be a simple license entailing only a single payment without further paperwork. B. The Government as it Affects Agriculture (a) The Agencies Dealing With Agriculture The Ministry of Agriculture 198. The Ministry of Agriculture and Fisheries (MOA) is responsible for overall agricultural development in Mozambique: planning, management of state farms, formulation of pricing proposals, preparation and implementation of agricultural development projects, and collection of agricultural statistics and documentation. The Ministry is headed by the Minister and one Vice-Minister, and consists of National Directorates, National Institutes, and Provincial Directorates. 199. Two State Secretaries are responsible for the overall planning and direct management of Irrigation and Cotton. The National Directorates define policies and programs for Crops, Livestock, Forestry and Wildlife, Rural Development, and Geography and Cadastre. The Institutes are responsible for research (Crops, Sugar, Cotton, Livestock, and Veterinary). 200. The MOA is represented in each province by a Provincial Directorate. These are responsible for the implementation and control of programs and projects, in regular liaison with the provincial and district level authorities. Sub-sector service staff are within a double subordination mechanism (dupla subordinaqdo), answering to both the Provincial Director, and the central sub-sector Directorate. The Rural Development Institute (INDER) 201. Established in 1990, INDER is responsible for rural development policies and programs. Its President has the rank of Minister and reports to the Cabinet. It has three Directorates: Economy and Rural Sociology, Rural Engineering, and Regional Planning; two administrative departments, and a data and documentation center, affiliated to the MOA's CDA. As an independent institute rather than a ministry, INDER functions as a coordinator of the line ministries involved in rural development. These include especially the MOA, the Ministry of Construction and Water, the Ministry of Commerce, and the social sector ministries. INDER coordinates with the Ministry of Cooperation in organizing donor support; with the Ministry of Finance in deciding on public investments in the rural areas; and with the National Planning Commission (CNP) regarding fiscal, investment, and macroeconomic matters. 202. INDER has a leading role in implementing the Priority District Program (PDP), which receives support from IDA (Rural Rehabilitation Project, Credit 2479-MOz), DANIDA, UNICEF, GTZ, the Italian Govemment, and other donors. In its work INDER relates to many NGOs already working in the rural areas, and actively seeks to involve them in the implementation of its own programs. The Agricultural Training Center (CFA) 73 203. CFA is the leading non-formal training institution for rural development and other rural- related activities. It was created by MOA in 1984 to train rural development agents (and, later on. extension agents), mainly at the provincial and district level. In June 1992, it had 14 professional staff, including two expatriates, and 28 administrative and support staff. The CFA has excellent physical facilities; vehicles; and computing, audio-visual and printing equipment; and the capacity for producing training materials and providing distance learning services. The CFA supports and supervises Provincial Centers for Agricultural Training (CPFAs), which exist in every province except Maputo. At its Maputo headquarters, it has departments of Training, Communications, and Administration. Main courses are on district administration; rural physical planning; and social action and rural extension, plus training trainers from the CPFAs and other training centers. Dissemination is provided through an Information Bulletin and a rural development and extension Journal (published quarterly and every 4 months, respectively); a radio program (5 minutes twice a week); production of training materials; and distance learning. 204. Skills Availability. A problem faced by the Government in general, the three agencies above included, is the lack of technically skilled personnel. According to a 1990 survey, only half of the 247 central government heads surveyed had completed secondary school, and only 74 of the 206 district directors surveyed had completed primary school. In 1989, the technical base of the MOA consisted of 83 university graduates (18 agronomists, 36 veterinarians, 13 economists, and 16 of other professions). The root source of the problem is that Mozambique does not produce enough secondary, let alone university, graduates to meet its needs for high- level manpower. Gross enrollment ratios for secondary and tertiary education (7.0 percent and 0.2 percent, respectively) are but a fraction of those in the rest of Africa. Major remedial training programs, both domestically and abroad, are trying to alleviate the situation, but again the lack of personnel with even the basic knowledge level to be trainable are similarly scarce, and many such programs end up training the same persons repeatedly. (b) Agricultural Research 205. The Ministry of Agriculture has three research institutes. The National Agricultural Research Institute (INIA) is responsible for food crops research; it also carries out resource inventories, surveys and evaluations. It has four technical departments dealing with Agriculture and Farming Systems, Soils and Water, Plant Protection, and Botany. Many experimental stations are spread over the country. The INIA has 34 researchers (12 of which are Mozambican nationals at B.Sc. level) and a total staff of 400. 206. The Animal Production Institute (IPA) is responsible for research in nutrition, breeding, and herd management. It has departments on Nutrition and Forage, Animal Reproduction, Breeding and Selection, Farming Systems, and Artificial Insemination. Research in genetic improvement takes place in three research stations at Chobela, Mazimochopes and Ang6nia. It has a staff of 13 university graduates, 17 medium-level specialists, and 335 support staff. 207. The National Veterinary Research Institute (INIVE) is responsible for the diagnosis of animal diseases, studies on epidemiology, food hygiene control, production of vaccines and biological products, quality control of these, and training of laboratory technicians at basic and medium level. It has departments on Diagnostics and Research; Vaccines and Biological Products, Food Quality Control, and DrugsNaccines Control. It supervises the provincial 74 laboratories, and has a staff of 18 university graduates, 28 medium-level technicians, and 50 basic technicians. 208. Many factors have negatively affected agricultural research in Mozambique. Overriding them all was the war, which made travel and operations difficult if not altogether impossible; and led to significantly reduced area coverage. Research results were communicated poorly, leading to an image problem, which in turn led to little recognition, low prestige and low salaries. Qualified researchers are inundated by bureaucratic, administrative, and logistic tasks, leaving little time for actual research. In addition, budgetary allocations were low. 209. Within the frame of the 1992 national policy guidelines for the agricultural sector, the MOA's research policy objectives are to generate and place at the disposal of producers, especially family farmers, appropriate genetic materials and technologies that permit a continuous and sustained increase in yield. This implies: improving the national genetic bank through collection and import of germnoplasm; developing higher-yielding, pest- and drought- resistant varieties for different agro-ecological zones, with priority to basic food staples; and developing technologies especially designed for family farmers, including cultural practices that increase productivity and conserve soil fertility. In turn, this requires the rehabilitation of the research station network to provide adequate coverage to the different agro-ecological zones and their respective production systems. 210. With the assistance of ISNAR89, INIA has defined a Medium-Term Strategy and Plan, and prepared a 1994-2000 Research Operational Plan. These state that INIA will pursue the national agricultural research policy objectives through applied and adaptive research rather than basic research. The emphasis will be on selecting and crossing crop varieties and lines, testing promising varieties on-station and on-farm under different agro-ecological conditions, adapting promising varieties to the farmers' conditions, and producing breeder and pre-basic seed material. Top priority will be given to maize, cassava, sorghum, rice, sweet potato, cowpea, bean, millet, cashew and cotton. 211. In the longer term, INIA and the other research institutes must rely on graduates from the national university for trained staff. Currently, the Faculty of Agronomy and Forestry Engineering of Eduardo Mondlane University is itself understaffed and also has fewer applicants than it has the capacity to train. Low faculty salaries are a major problem in staff retention, and the greater perceived ability of other courses of study to result in subsequent employment is a problem in attracting students. Efforts to strengthen the University and to improve direct links to the research institutes will be important in establishing a viable national research and extension network. (c) Agricultural Extension 212. The National Directorate for Rural Development (DNDR) is responsible for rural extension services. It was created in 1986, and was the first national public extension organization to be established. It has two departments: Planning, and Rural Extension. In mid- 1993, it had 32 staff members (of which 19 were university graduates) at the national level; 25 medium-level and 5 university-level staff members at the provincial level; and some 400 medium- and basic-level technicians at the district level. As of end- 1994, there were around 670 89 ISNAR is the International Service for National Agricultural Research, a member of the CGIAR. 75 poorly-equipped field extension workers (550 in public projects; 120 with other organizations) to attend more than two million family farms. Preliminary assessments point to country needs four times as large. Most extension workers are basic-level technicians, and there are few qualified subject matter specialists. 213. Most extension activities are funded by external financiers (IDA, IFAD, bilateral donors), supervised by the DNDR and its provincial services, and executed through the Provincial Departments of Agriculture (DPAS). The financing organizations usually recruit and appoint project coordinators, who are accountable to them, and these recruit local personnel to man the provincial services, which are accountable to the project coordinators. Under the system of double subordination, each DPA is accountable to both the Ministry of Agriculture and the Provincial Governor, each of which may decide on locations, priorities, and allocation of resources; this leads to some confusion and inefficiency. The official extension model adopted by Government and funded by IDA and IFAD is Training and Visit (T&V). 214. The content of messages for farmers is not necessarily the best available or the most relevant. Sometimes there are variations in extension messages for farmers in the same areas. They are impaired because of the lack of a formnal procedure for their selection and formulation, and of a formal linkage between research and extension; any actual linkage happens only between individuals and as a result of friendship or serendipitous contacts. Extension messages were often formulated by DNDR or CFA staff, some of which did not have adequate technical skills, or by donor project personnel, without technical input from researchers. Some were generated by INIA staff, but haphazardly because the responsibility of the task is not clearly defined. But research has produced results which would be useful for farmers, and extension has feedback to provide to researchers. Thus, increasing efforts are being made to more closely integrate extension with research as a major source of new messages for the farmers, and extension with the farmers, to make these more relevant to their needs and constraints. 215. Some 120 extension workers are engaged in the private and joint-venture sectors. The cotton companies are maintaining extension services mainly along the West Africanfili&re approach, i.e., a comprehensive, one-crop technical assistance system, including productive technical packages and upstream and downstream activities. One joint-venture cotton company (LOMACO, in Cabo Delgado) is trying to introduce a more systemic, longer-term approach on the whole farm as a unit of concern. 216. Veterinary Services. The National Directorate for Livestock (DINAP) is responsible for defining policies and directing programs in the livestock sub-sector through two technical departments: Animal Health, and Animal Production. The former comprises three divisions: Epidemiology; Tsetse and Trypanosomiasis; and Public Health and Hygiene. At the provincial level, work is performed by the Servico Provincial da Pecuaria (Livestock Provincial Service). Their main task is carrying out vaccination programs and dip tank anti-tick baths. Extension services as such are not performed, but the provision of veterinary services at dip tanks provides some opportunities for extension to the family sector livestock owners. There are few follow-up visits after vaccination campaigns and other routine activities. During the war, only 25 (24.5 percent) of all veterinarians were involved in activities at the provincial level, and 77 (75.5 percent) were working in Maputo. 217. At present, the highest priority is given to vaccination and control of livestock movements and quarantine. In 1992, vaccination campaigns reached the following percentages of the national targets: anthrax (67 percent), blackquarter (60 percent), brucellosis (12 percent), foot & mouth disease (36 percent), rabies (54 percent), and Newcastle (53 percent). These 76 figures are low, since it is considered that between 75-80 percent coverage is necessary to protect the herd from epidemics. Control of ticks is done by the use of acaricides either by dipping or spraying. In 1992, only 275 (39 percent) out of 700 dip tank installations were in operation, and the annual number of immersions per animal per year was found to be between 10-12, compared with 24 in 1971. The last tsetse control campaign was carried out in 1974 by the then republics of South Africa and Rhodesia, including the use of residual insecticides (DDT). At the moment, tsetse surveys are carried out in areas where there are plans for livestock development. In the commercial sector, preventive chemotherapy is used for trypanosomiasis; in the family sector, only sick animals are treated. During the war, rules for treatment of cattle to be moved from infested to free zones were seldom enforced. Taking advantage of Mozambique's livestock potential is highly dependent on a successful control, and in some cases eradication, of diseases. Among the priority activities are: * tick and tick-borne disease control * cattle screening for tuberculosis and brucellosis * in some areas, vaccination against anthrax, blackleg, and rabies * tsetse fly control and trypanosomiasis chemotherapy and prophylaxis 218. The main constraints faced by the veterinary services are the inadequate number of subject matter specialists and other well-trained staff for field work; transport facilities; and inadequate supply of drugs, acaricides and veterinary equipment, which are all imported. At the same time, changes have to be introduced in the relations with farmers in the field; at present, most of the contacts are restricted to cattle owners, and veterinary services are provided without taking into consideration the rest of their exploitation. Needed improvements are: to enable field staff to effectively provide extension services to livestock owners; to increase their effectiveness; and to set up a monitoring system to measure their input and impact, including a herd monitoring element. Even though for the time being veterinary services for the family sector should remain free of charge, the introduction of cost recovery mechanisms, and their integration into revolving funds, should be considered. (d) HealthI and Nutrition 219. Children bore the brunt of Mozambique's social and economic dislocation during the war period. The infant mortality rate and the "under 5" mortality rate, at 146 and 282 per 90 -aaa fia iue2 thousand live births respectively , are among the highest in sub-Saharan Africa. Figure 20 shows the changes in infant mortality in Mozambique and in Sub-Saharan Africa between 1970 and 1993. Infant mortality fell practically everywhere, the mean drop being from 141 to 100 per thousand in the period (29 percent). Mozambique ran against the trend, with a much smaller decline of 15 percent (from 171 to 146). 90 World Development Indicators, Table 27. Data refer to 1993. 77 200 1 t_ Mozambique 160 . 120 Sub-Saharan~ *a Africa (mean) 80* 1970 1993 Figure 20. Infant mortality rates, Mozambique and Sub-Saharan Africa91 The maternal mortality rate, at 1,100 per hundred thousand births, is among the highest in the world. These figures are national averages and indications are that the situation is worse in rural areas which were the most affected by the war. 220. It was shown in Chapter 3 that rural people have less access to medical services than do urban people. This is partly due to an urban bias in government medical expenditures, as well as to low effective demand for medical services in rural areas, and to the destruction of one-third of rural clinics. 221. The war drastically reduced the food security of millions of households as they were forced out of their farms and left without income-earning opportunities. It was shown in Chapter 3 that the average rural household suffers food insecurity for 3.7 months in the year. Child malnutrition continues to be a serious problem. Weight for height, usually called "acute malnutrition", indicates the degree of vulnerability and wastedness. By this measure, five percent of under-fives have moderate and three percent severe malnutrition.92 Zambezia province is the worst, with 12 percent moderate and 10 percent severe malnutrition. A measure of malnutrition of long duration in the past is given by the rate of stunting, or height for age. By this measure, all areas of Mozambique are seriously affected, rural and urban, with stunting rates varying from 34 in Maputo province to 74 in Zambdzia (Figure 21). l World Development Indicators, Table 27. This is the under-two mortality rate per one thousand live births. 92 Multiple Indicator Cluster Survey (1996), p. 21. "Moderate" is a measure between two and three standard deviations below the intemational mean; "severe' is over three standard deviations below. See also the Appendix Tables for further detail. 78 C- Nip = = =N FS = under=5-year!oldsI 111! 0 20 40 60 80 Percentage with height for age more than 2 standard deviations below norrnal Figure 21. Stunting of under-5-year-olds, 199593 222. Four policy approaches could help to reduce the infant mortality rate and the amount of acute malnutrition. The first is better roads, which would reduce transport costs and enable rural dwellers to diversify their income sources and market more produce, which would improve their ability to smooth their consumption across years and increase their ability to purchase medical services. The second is the extension of basic medical services, and in particular to communicate knowledge of the emergency sugar-and-water approach to diarrhoea. The third is simplified registration for traders, which would improve rural dwellers' incentives for participation in markets. The fourth is revamped research extension services which would help to raise farm productivity and incomes. (e) Education 223. The Government made strenuous efforts after independence to make basic education available to all. It launched a massive literacy drive through primary education and adult literacy campaigns. As a result, the overall population literacy rate jumped from 7 percent in 1975 to 20 percent in 1980. The net enrollment rate in primary school rose to 49 percent in 1980 and 55 percent in 1985. Progress has been slower since as the war, droughts and the general economic decline have affected the whole educational system. The overall literacy rate is still only 40 percent (50 percent for low-income Africa) and primary school enrollment rates have actually fallen. By far the most affected was the primary school level, in particular in rural areas that were most vulnerable to attacks. Of the 5,750 primary schools that existed in 1980, only 3,500 still operated in 1989. By comparison, the number of secondary schools, located mostly in urban areas. expanded from 19 to 39 during the same period. In 1991, only 35 percent of the 7-year olds enrolled in primary schools and, although the figures are not available, enrollment was again heavily skewed in favor of urban areas and it is estimated that fewer than 20 percent of rural children registered for first grade. The performance of primary schools is low because of low qualifications and morale among teachers and severely restricted resources. Over 1975-89, the per capita spending on primary education declined from $2 to $1 and in 1992 the non-salary expenditures were about 15 cents, far below the average of $1 for low-income Africa (and the $5 recommended minimum). The primary school drop-out rates are extremely high: less than 50 percent of the entering cohorts graduate from primary school. Of those graduating, fewer than 20,000 enter secondary school and not more than 1,000 graduate. Although this is an 93 Source: Multiple Indicator Cluster Survey (1996), p. 22, Table 6. 79 improvement over the pre-independence numbers (about 500 mostly Portuguese children), it represents less than 0.5 percent of the annual cohort. 224. The government always recognized the importance of women's education for the economic and social development of the country. Women play a leading role in agricultural production and there is also a strong link between women's education, the health of children and the overall birth rate. Studies have shown that national education can help improve nutrition, is correlated with birth weight, and can significantly affect the number of births per woman. The record of female school admission has been relatively good. Girls in primary education rose from 35 to 42 percent between 1975 and 1992 and in the latter year, girls represented 37 percent of secondary pupils. However, female literacy is still only 21 percent compared to 45 percent for males, with particularly low levels in rural areas (11 percent against 43 percent of urban women). However, women are still heavily discriminated against with regard to access to education, especially in rural areas where conservative attitudes and women's importance in the economic and family life combine against their participating in educational programs. This is in particular the case for adult education programs where relevance is a major issue. Adult education seldom addresses the real needs of rural women and classes often clashes with more important productive and domestic tasks. 225. The current inefficiency of Mozambique's education system has two basic causes. First, in spite of Government's stated commitment to primary education, sector funding is heavily skewed in favor of urban areas (Maputo in particular) and higher education: In 1990, primary education which represents more than 90 percent of the students, received less than 35 percent of Government's education recurrent expenditures and less than 8 percent of the sector investment allocation (in primary education, salary expenditures represented 99 percent of all expenditures). The second factor affecting efficiency was the relative lack of relevance of the current school curricula for most children's socio-economic environment. Following the Jomtien Declaration, the Government has reaffirmed its policy of Education for All (EFA) by year 2000. This policy, supported by all donors including IDA, involves an in-depth review of the entire education system and deep changes in its focus, organization and curricula. A key element of the EFA approaches to build upon the experience of an on-going "School-Community linkages" pilot project in Xai-Xai, in the Gaza Province the local community is called upon not only to build the necessary infrastructure and to bear part of the school operating cost but also to participate in the management of the school and the design of its curriculum to match local conditions and needs. This program will also emphasize programs of basic education for girls/women as an essential step toward fostering their key role in the production process and the reduction of poverty. 226. The role of education will be crucial in the evolution of Mozambican agriculture. Research in other countries has repeatedly shown a strong link between education and the capacity to adopt technological innovations. The "critical mass" of schooling required is typically four years. With the recovery after the war, enrollment rates in Mozambique have once again risen - for 6- 1 0 year olds, from 24 percent in 1991 to 45 percent in 1994.94 Much more than this will be needed to achieve technological progress, and particularly in educationally backward but agriculturally rich areas such as Nampula, where the enrollment rates were II percent in 1991 and 23 percent in 1993. Source: Inquerito Demografico Nacional (1991), and Multiple Indicator Cluster Survey (1995), as quoted by Desai (1996), Table XIII. 80 C. The Financial System 227. There are no special banks, funds, or other agencies lending specifically for agriculture. Agriculture is eligible to draw credit from the same banks as anybody else though there has been very little such lending. Historically, smallholders obtained informal credit from the network of rural traders and shopkeepers. Reestablishment of the marketing system will therefore be important in providing informal credit as well as marketing of agricultural products and consumer goods in rural areas. 228. Mozambique has three banks which lend to the agricultural sector: the Banco Comercial de Moacambique (Central Bank of Mozambique), the Banco Popular de Desenvolvimento (BPD, Popular Development Bank), and the Banco Standard Totta. Of the three, only the last is private. Negotiations are under way for the privatization of the two state banks. The Banco Standard Totta lends very little to the agricultural sector, accounting for about I percent of all agricultural lending during the 1 990s95 The bulk of agricultural lending was undertaken by the Banco Comercial. Table 20. Total agricultural lending, 1988-92a Nominal In 1988 Meticais In US $ (Mt Io0) (Mt 109) b (millions) c 1988 91 91 200 1989 139 99 189 1990 159 79 169 1991 204 75 135 1992 243 58 94 Source: Abreu and Baltazar (1993), Appendix. ; "Agriculture" here includes agro-industry and agricultural marketing. b Using the CPI. c Using end June exchange rate from International Financial Statistics. 229. Table 20 shows that total lending to the agricultural sector decreased sharply between 1988 and 1992, from roughly US$ 200 million to US$ 94 million. Here "agricultural lending" is taken to include agro-industry and agricultural marketing, as well as lending for farming operations. Probably the farming component of lending was only a small proportion of the total96, although we do not have data to confirm this. 230. In the past, most agricultural lending was channeled through the Banco Comercial, much of it to four state enterprises, Caju de Mo,ambique (the cashew processor), Emocha (the tea processor), Incomati (an irrigation project), and AGRICOM. Lending to these four firms held Abreu and Baltazar (1993), p. 37, Table 15. 96 Coates (1994, p. 92), for instance, reports that in Tanzania loans for marketing outweighed loans for agricultural production six to one. 81 constant in real terms between 1989 and 1992, although their share in all agricultural lending rose from 14 percent to 26 percent in the period.97 They have since all been privatized.98 231. In the socialist period the nominal interest rates charged by the banks were kept down by edict. There were as many as 12 different rates for various branches of agricultural activity. With the economic reforms, these were reduced to four rates in 1992, and by 1993 the rates were unified with those of other sectors.99 The nominal interest rates were raised and by 1993 the real interest rate was positive for the first time (Figure 22). The implicit subsidy to the agricultural sector fell from 13 percent of total government revenues in 1988 to about two percent in 1993, and has fallen further since. 5 14 0 ° E C,. -10- --15- ( -20 1 .~-25- ~,-30 35 1988 1989 1990 1991 1992 1993 1994 Figure 22. Real interest rates for agriculture, and the implicit subsidy for agriculture as a percentage of total government expenditure100 232. Credit outside the banking sector is limited, partly due to the lack of a long-standing tradition of rural finance organized by smallholder organizations, and partly due to the dislocation of the war which upset social ties. An exception to the rule is given by the Uniao Geral de Cooperativas, which channels credit from the BPD to its member cooperatives, in the form of physical inputs. In 1992 the monthly average of credit disbursed was of the order of $100,000.'°1 Certain NGOs also supply production credit, but the extent of their activities is unknown. 233. Actions required. It is important to reinforce the macroeconomic stability which has been brought about. It is urgent that the government complete the privatization of the one remaining state-owned bank, so as to assist the development of a competitive banking sector. Abreu and Baltazar (1993), p. 26, Table 6. 8 Get more recent data on their borrowing. Abreu and Baltazar (1993), p. 32. 100 Sources: Abreu and Baltazar (1993), Table 14, p. 35, and Appendix; Southern African Department (1995), Annex A.5. 1993-4 interest rates from Banco de Mocambique (1994). For calculating the implicit subsidy, a "normnal" real interest rate of 5 percent was assumed, and the difference between this and the actual real interest rate was multiplied by the stock of credit. 101 Abreu and Baltazar (1993), p. 40. 82 Adjusting the land law to facilitate land transactions would enable swifter expansion of credit to farmers with title to their land. For the large-scale farming sector (and also for the marketing and processing subsectors) there would then be no need for special government-run arrangements for credit. For smallholders, the government can encourage the growth of rural financial institutions. There are examples in other countries of successful small-scale enterprise banks which lend in small amounts on the basis of savings deposits and which rely on group solidarity to guarantee repayment - for example, the Grameen banks, the Cooperatives d'epargne et de credit in Togo, and the Credit Union Movement in Cameroon. Donor financial and technical assistance should be provided at the initial stage of institution-building and training. These efforts will surely fail, however, if inflation rates continue to be high and unpredictable. Finally, the government can help to reestablish the marketing system in the rural areas by a road repair and building program; this will rejuvenate the tradition of credit supplied to farmers by traders and shopkeepers. D. The Inputs Supply Markets (a) Seeds 234. Following independence, input policy focused on providing state farms with ample input and equipment supplies; smaller amounts were distributed to the cooperative and private sectors, and none to the family sector. Import orders were placed with international firms represented in Maputo; two parastatals then imported and distributed the products. The collapse of the state farm sector during the 1980's and their recent divestiture made governmental intervention in this regard unnecessary. The largest users of agrochemicals today are the three large cotton- producing joint venture companies, which import directly. Smaller users are the sugar and citrus producers, rice, tomatoes, and other vegetables. The closing down of most parastatal retail outlets has forced small consumers to purchase inputs directly from the parastatal's headquarters office in Maputo, or to import them informally from Zimbabwe or South Africa. 235. The Government has an installed capacity in INIVE to produce livestock vaccines, but budgetary constraint in official vaccination programs and retraction of private livestock production has led to paralysis of most production. SEMOC, a joint Government (80 percent) and Swedish (Svalof TA; 20 percent) venture, has been both a significant producer and an importer and distributor of improved seeds. 236. Before 1987, seeds and hand tools were distributed through the trader network. As the war weakened the traders' presence in the rural areas, AGRICOM increasingly substituted for them. AGRICOM was replaced by the ICM (Instituto de Cereais de Moqambique) in 1994, whose future is in question owing to a lack of donor interest. The two parastatals involved with input and implement production and distribution (BOROR and Agro-Alfa) are currently in the process of being privatized, and Boror Commercial is no longer active. Thus there is no longer a government presence in the commercial input supply market - nor has it been replaced by a retail network of agrochemicals. 102 See Gurgand and Yaron (1993) for an excellent exposition of the characteristics of successful rural financial institutions in Africa. 83 237. On a non-commercial basis, large volumes of seeds and tools were distributed to the family sector by the Emergency Seeds and Tools Program (PESU). This Program was started in 1987 in response to the large number of family sector producers who, due to war or drought, had lost access to sufficient seeds and tools to guarantee their survival. The Program was expanded in 1992 as the Peace Accord opened RENAMO-controlled areas and encouraged displaced families in other parts of the country to return to their lands. At its height an estimated 1.2 million families received seeds and tools in the context of the resettlement program. These and other free distributions by donors and NGOs were a potential obstacle both for the "graduation" of farmers receiving them and for the development of trader network and a market in seeds and tools. All these programs have now been phased out, barring one final IDA-financed program, valued at $5 million in total, to have World Vision distribute AgPaks in Sofala and Zambezia provinces. (b) Fertilizer and Chemicals 238. In the early 1980s Mozambican agriculture, in particular the state farm sector, was consuming between 40,000 and 60,000 MT of fertilizer per year, and 2-3 million liters of pesticides. Imports were done by the parastatal INTERQUIMICA, and distribution was effected by the parastatal BOROR. Both have since been privatized, and all firms are free to make their own import arrangements. Demand for fertilizer fell dramatically to approximately 8,000 MT in 1993.103 As the new private sector firms developed, the demand for fertilizer rose again to about 15,000 MT in 1996.104 Most large-scale users make their own import arrangements. 239. The three cotton JVCs (Joao Ferreira dos Santos, LOMACO, and Entreposto) dominate the market for most agrochemicals, especially pesticides. They import directly as well as draw down inputs from the KRIi program, a large agricultural supply program of the Japanese government. Only a small number of retail outlets outside of Maputo sell fertilizers and agricultural chemicals, representing a serious constraint on input use by small and medium-sized farmers. The Sociedade Mocambicana de Produtos Agroquimicos is one of the suppliers for small-scale farmers. This distribution problem will be resolved over time as demand increases and more private retailers stock the products, but a supply constraint is likely to persist for several years, delaying recovery of production of the private sector. 240. The KRII program has been in effect since 1986. At its peak it ran to some X900 million ($9 million) per year, and has since fallen to some $3-4 million.'05 The inputs supplied include fertilizers, pesticides and equipment. It is not clear that smallholder cotton producers are receiving the full benefit from the KRII inputs (pesticides). There is no public reporting on these subsidies, nor of how they are used. Consideration could be given to auctioning the products and using the funds, in a transparent manner, for development activities which would benefit smallholders more directly. 103 Strachan (1994), p. 64. 104 Source: interviews with the major importers and users (INTERQUiMICA, LOMACO, Joao Ferreira dos Santos, Entreposto, Citrinos de Maputo, Citrinos de Manica, Sociedade Mogambicana de Produtos Agroquimicos, Tabacos de Maputo, Acucareira de Mocambique, and Agribuzi. 05 Information from the Sociedade Mocambicana de Produtos Agroquimicos, Lda. 84 E. Food Aid 241. Before independence, Mozambique was a net exporter of agricultural commodities, including food crops. The departure of the Portuguese, the Government's marketing policies and the war resulted in a dramatic decline of overall and marketed food production. Food aid imports grew rapidly. Throughout the 1980s, close to 90 percent of all marketed grain was imported, most as food aid. The 1991/92 drought further reduced domestic production and between 1990 and 1992 maize imports doubled. Table 21. Cereal Food Aid Imports 1980-1993 ('000 tons) Year Maize Maize Maize Wheat Rice Total Commercial Emergency Total Cereals 1980 10 110 12 132 1987 71.9 62.4 134.3 103.8 74.4 312.5 1988 201.4 180.9 382.4 127.7 61.7 571.8 1989 143.1 151.4 294.5 109.1 73.4 477.0 1990 79.7 158.8 238.5 101.5 30.0 370.0 1991 162.0 286.3 348.3 82.6 47.6 478.4 1992 253.3 309.9 563.3 67.7 56.8 687.8 1993 93.0 390.8 482.8 38.6 22.1 543.5 1994 77.0 208.1 285.1 57.0 28.5 370.6 1995 70.2 123.9 194.1 103.1 50.0 347.2 Source: 1980-93: Ministry of Commerce. 1994-5: Tschirley et al. (1996), Table 1, p. 4. 242. Since the peak in food aid shipments in 1993, there has been a marked decrease in all categories of food aid, both commercial and emergency. Local production picked up between 1993 and 1996, so that food aid as a percentage of total cereals available is currently at less than 15 percent (Figure 23). A major change in one of the components of food aid - commercial food aid - is that from 1996 it was auctioned off upon arrival, thereby reducing its negative impact upon the local price of cereals. 85 80 ~-20 1989190 199011 199112 199213 199314 199415 Figure 23. Food aid maize as a percentage of total cereals available106 243. The Government's food aid policy was justified in a war-time setting. The main objective of food aid was to bridge the gap between domestic production and food needs and to ensure a safety net for impoverished urban and rural populations without sufficient production or income to meet basic calorie needs. Other objectives were to save foreign exchange and provide Government with budgetary resources through the sale of a portion of food aid to finance development activities. Without question, food aid prevented the starvation of possibly millions of people. 244. Mozambique is now no longer at war and has staked its future development on a market economy and strong growth in the agricultural sector. It is therefore important that food aid not undermine these objectives. Favorable agroclimatic resources, at least in the northern two-thirds of the country, suggest that Mozambique could feed itself and be an important food exporter in regional markets. However, even under optimistic assumptions, Mozambique will not be able to achieve a satisfactory level of food security within the next decade and will need assistance to close the projected gap between domestic production and basic needs. At least part of this assistance will come from food aid. Hence the challenge is now to ensure that future food aid programs do not impede the recovery of domestic production. Institutional Framework for Food Aid 245. Food aid enters Mozambique through two different programs: "commercial" food aid and "emergency" food aid. Under the first program, food aid is used to close the gap between domestic demand and supply in a situation of extreme scarcity of foreign exchange. It amounts to general balance of payments support offered in kind instead of in cash. Under the emergency program, food aid is provided with the specific purpose of assisting populations in time of famine brought about by the war or natural disasters. Emergency food aid is distributed free, or under food-for-work programs, to the target populations. 246. Emergency food aid. Initially the Departamento de Prevenqao e Combate as Calamidades Naturais (DPCCN) managed all the emergency food aid efforts. Since as much as a third of the food was diverted to commercial channels, the donors increasingly bypassed DPCCN and relied on NGOs. By November 1993, DPCCN was handling only approximately 40 percent of all emergency distributions, with UNILOG and the NGOs each accounting for about 30 percent. ° Numerator: Yellow maize for commercial and emergency food aid. Denominator: All food aid (maize, rice, wheat), plus local production (white maize, rice, sorghum, millet), plus commercial cereal imports (mostly rice and wheat flour). Source: Tschirley et al. (1996), Table 1, p. 4. 86 247. What rationale is there for the continuation of emergency food aid? The World Food Program explained its activities in 1995 as follows. The maize requirement of the Mozambican population, based upon was for 1.5 million tons, but total output was only about I million tons due to the drought which hit the South of the country but not the Center or the North. The World Food Program and other donors purchased about 50,000 tons locally, some of it from the ICM, which in turn bought from traders in the North. The donors purchased another 80,000 tons from South Africa, raising the total of emergency food aid to about 130,000 tons. Thus emergency food distributions amounted to some nine percent of the country's food requirement. (Some of the remaining deficit was covered by commercial food aid imports.) The emergency maize was distributed in rural areas in the South by NGOs. So as to avoid competition with the market, the intention was to target families in particular need with ration cards, but due to the sensitivities involved the distributions were sometimes general. Given that most rural people are extremely poor and droughts are frequent, the World Food Program anticipates working in Mozambique for many years. 248. Even if the beneficiaries are targeted , the free distribution of maize, maintained over a long period of years, discourages maize production activity because it prevents prices from rising to levels which would become strong signals. Free distributions inhibit the expansion of trade. The expectation of free food in the event of a drought reduces incentives for saving, for improved on-farm storage , and for the creation of other insurance mechanisms . There is no longer an emergency situation as the war has ended and the refugees have returned. Even the droughts have a large degree of predictability. The problem is not one of unforeseeable calamity which justifies international and government hand-outs; it is one of poverty, which continued hand-outs are converting into aid dependency. In other war-torn parts of the world there are destitute people in with a stronger claim on limited donor emergency resources. Drawing down the emergency food program brings risks of leaving some people in real distress, but these risks can be reduced if the program is closed over a period of time and if NGOs are selected for distribution on the basis of their capacity to target appropriately. 249. Commercial Food Aid. Commercial food aid has represented the larger proportion of total food aid except during the 1992-93 drought years. All commercial (maize) food aid used to be sold to officially appointed wholesalers (consignees) at an official price. The consignees undertook to distribute the maize in designated provincial and district capitals through the local retail network. The proceeds from the sales to consignees accrued to Government for general budgetary support. This resulted in serious disincentive effects, as the consignee price was pitched low: in 1993 the price of yellow maize food aid to consignees stood at a mere 42 percent of the maize import parity price and 75 percent of the official producer price for white maize. This low price for yellow maize in turn influenced the price of white maize, the local staple. Recent studies indicate that the premium that consumers will pay for white maize does not exceed 30-40 percent in the main urban markets, and that yellow maize has become a close substitute for white maize (Minist6rio da Agricultura e Pescas, 1993). 250. With the introduction of auctions of commercial food aid in 1996, the disincentive effect has decreased but is still present. These food imports likely result in greater imports of food than would have been the case without Government intervention. The unpredictability of commercial One observer noted that farmers in the district of Marigue held stores of sorghum for two years- personal communication, Mr. Philip Clarke, World Food Program. l 8 For instance, outcome-contingent loans, used by farmers in some parts of Nigeria (Udry, 1990). 87 food aid impons obstructs the planning process of private traders, so that the development of commercial trade links is impeded. 251. Short-term food security would better be promoted by encouraging international trade and the development of long-term storage. Merchants could benefit from advance warning about potentially bad harvests, using prediction based on the El Ninio-Southern Oscillation phenomenon109. Also, the efficiency of the ports and the domestic marketing system (wholesaling, retailing and transport logistics) could be improved, thereby reducing the import- export parity band from a potential $54 to $14 per ton of maize.110 With better informnation and efficient port handling, private imports of maize would prevent sudden rises in food prices. 109 ENSO-based forecasts are reliable for up to six months ahead of the wet season. The technical expertise is available at South African universities. See Gibberd et al. (1995), p. 26. 110 In a year in which South Africa has a deficit of maize, the wholesale price in Maputo would be $233/ton with current efficiency levels and $179 with improved efficiency. In a year in which South Africa is in surplus, the Mozambique price would be $177 with current arrangements and $163 with improved efficiency. The price swings are reduced from $54 (=233-179) to $14 (=177-163). See Coulter (1995), Appendix 4. 88 Chapter 5. THE AGRICULTURAL ECONOMY A. Aggregate Demand (a) Domestic market 252. Even though there are constraints imposed by the low level of incomes and by the distance between production and consumption centers, domestic markets provide a significant margin to accommodate expanded food crop production. The situation for grains and other crop foods; industrial raw materials; and meat and milk products is analyzed below. 253. Grains and other crop foods. The overall level of demand is limited by the low level of incomes. Income per capita in Mozambique is one of the lowest in the world. The drastic drop in production from 1975 to 1986"', compounded by periodic droughts, has required massive food aid, which is distributed free ("emergency" food aid), at prices well below import parity levels (regular or "commercial" food aid) or in between (food stolen from official aid shipments). This has boosted current consumption to a level unsustainable at market-clearing prices. Even then, there is a wide gap between current domestic production and demand at market prices which would allow for a significant expansion of production. 254. The most important influences upon the long-term growth in the demand for food are population growth, income growth and the income elasticity of demand. Future population growth has been estimated at 3.3 percent. I2 The growth of per capita private consumption has been estimated at 2.2 percent. 13 A not unreasonable assumption regarding the income elasticity of food demand is 0.7.114 Then - on the assumption of constant food prices - we can expect food demand in Mozambique to rise by about 3.8 percent per annum. 15 See Chapter 4 for figures on marketed output. World Development Indicators 1995, Table 25. 113 Estimate from World Bank (1996a), Annex 1, p. 1, referring to 1997. Tsakok (1990), Table D.2, p. 255. This is the median of results for less-developed countries including India, Bangladesh, Philippines and Ghana, for the period 1950-1970. Coates (1994) finds estimates of between 0.95 and 1.05 for Tanzania in 1992, but these are unusually high. By definition, the income elasticity of demand C = {AD/AY}.{Y/D}, where Y is income and D is demand. Let P be population. Rearranging, D=YE = [(Yj P)+ PIE = (2.2 + 3.3) * 0.7 =3.85 89 255. If food supply did not rise at least as fast as this, food prices would likely rise. In fact, between 1992 and 1995 food supply increased at about 6 percent per annum, although this period should not be taken as typical, since much of the growth was due to the reintegration of millions of deslocados, and to the recovery after the reduction of food aid and the development of trading links. The real price of food staples was brought down suddenly in early 1993, on account of massive food aid shipments. The price remained at this historical low - about 50-60 percent of the price between 1990 and 1992- until late 1995, when it started once again to reach its historical levels. 160 * 120 CD $ 40,+ 80 40 0D .', . . . . . . . . O O _ _ N N X~ X' t 0 £ e S ew 0 0 Za 0 e wOao Figure 24. Real price of maize, Maputo, 1990-1995116 256. However, there is a spatial problem. The marketable surpluses of the most important staples, maize and cassava, as well as beans, are produced primarily in the North, which lacks major urban centers, while most demand for these commodities is in the South, and 117 especially in Maputo . Thus, the South is and will remain a net "importer" of grain (particularly maize), and will need supplies from outside it. It is worth noting that this has always been the case from colonial times to the present. The South has exported labor and transport and tourism services in exchange for food and other consumer goods. 257. From a closed-economy point of view, the "natural" food supplier for the South would be the Northern provinces, but certain conditions must be met for the North to have a competitive advantage to supply it. The most important condition is that the price of maize in the urban areas must increase substantially to permit the Northern provinces to be able to ship maize to the South. Yellow maize would have to be priced at least at its import parity price equivalent for white maize producers in most of the North (apart from Niassa and parts of Tete) 18 to receive a price that would just cover the cost of production in their respective provinces' But at the current subsidized price levels, the North cannot compete. The North's cost 116 Source: 1993-1995: Donovan (1996), Appendix C, pp. 253-256. 1990-1993: SIMA data. 117 The four Northern provinces (Cabo Delgado, Niassa, Nampula and Zambezia) account for 60% of marketed maize (1990, most recent good crop with provincial data). The two Southern cities of Maputo and Matuba represent 40% of the population of the larger towns and cities in the country. 118 This assumes an import parity price of US$170/ton for commercial yellow maize, and a white-over- yellow maize premium of 15%. Historically, the premium has been 35%, but increasing consumer use of yellow maize in urban areas plus the fact that yellow maize presents lower bulking costs and predictable quality will create a downward pressure on the white/yellow maize premium over time (Strachan (1994: 50-53). 90 disadvantage is compounded by the progressive acceptance in the South of yellow maize as a staple by the populations currently supplied with food aid products, while the North produces white maize. 258. Industrial raw materials. The only crop in Mozambique that can qualify as production of raw materials for industrial processing (except for export crops, which are analyzed below) is cotton. Currently, textile production is extremely low, but it is clear that substantial domestic demand exists for cotton clothing. Current policies favor cotton producers (JVCs) but little attention has been given to their natural complement, textile production. Illegal imports of textiles are widespread, and the evidence suggests that the resultant effective protection coefficient is about -100 percent. The government should adopt firmer measures to ensure that the (already fairly low) tariffs on textile imports are paid by all importers. 259. Milk and meat products. The income elasticity for livestock products is typically well above unity in very poor countries.119 There is substantial domestic demand for milk and meat products, evidence of which are the (often illegal) imports of dairy products from South Africa. Part of the reason for the imports is that the war decimated the national herd of both livestock and wildlife (a significant source of meat supplies). The ready markets for both milk and meat products will provide an important market opportunity in the future. In contrast to food crops, however, the best areas for cattle raising are in the South, where demand also concentrates; the locational problem mentioned above for maize does not exist in the case of livestock. (b) Foreign markets 260. Mozambique faces "normal" export markets, in that access to them is not constrained by special agreements, quotas or the like. Being a small country, Mozambique's exports are not likely to have any effect on international markets or prices, with the possible exception of cashew nuts in the long term. Mozambique's share of world cashew exports is about 4 percent and its share of production about 8 percent (Kumar (1995), p. 14, p. 7). Cashew nut demand has recently grown at about seven percent annually (Kumar (1995), p. 14), and may even exceed this in the foreseeable future as sustained economic recovery is projected for the developed countries and new markets emerge in Asia, Latin America and Eastern Europe (Kumar 1995, pp. 26-33). Hence prices are expected to remain firm in spite of expected production increases in Mozambique and other countries. In the case of tea, copra, cotton and sisal, no persistent downward trend is envisaged. In particular, the price of cotton has recently risen and is likely to remain high for the foreseeable future. In the case of citrus, Mozambique is a highly competitive producer of grapefruit whose worldwide demand is increasing steadily. 261. The main problem with export crops lies not in the external markets but in poor domestic policies and the low operating efficiency of the processing industries. For example, to ensure the survival of the domestic cashew processing industry, Government used to prohibit the export of raw nuts, which in 1993 commanded a higher price than kernels in raw nut equivalent.. In April 1995 Government shifted to a 25 percent export tax on raw nut exports, which, as is shown elsewhere (page 107), is not needed by reasonably efficiently operating plants. On the issue of operating efficiency, farmers sell their produce to processing, trading or exporting firms 119 von Braun et al. (1991), p. 66. See also the elasticities for milk of 1.0 in Pakistan, and those for beef of 2.8 and more in Nigeria in the 1960s, Tsakok (1990), p. 255f. 91 whose processing facilities show serious deterioration (compounded by labor problems, at least in the case of cashew nut processing). This leads to reduced extraction rates and poor commodity quality. Both lead, in turn, to reduced foreign exchange revenues for the fraction exported, and low producer prices on account of both low extraction rates and low export prices . This is compounded by the monopsonistic situation prevailing in their domestic markets either by law, or de facto, or by contract in the case of the joint venture companies operating in cotton. Marketing and trade liberalization would go a long way towards increasing producer prices for these commodities (see also page 114). B. Aggregate Supply 262. One of the standard findings of the agricultural economic literature is that smallholder farmers respond favorably to price incentives. Accordingly the standard policy recommendation of the literature is that economic growth and rural welfare may me fostered through stripping away unnecessary taxation or price-reducing distortions. The finding is so common that it barely deserves restatement, except to get an idea of the relative magnitudes of the anticipated responses. In the case of Mozambique, the two areas of policy interest are: the sub-import parity prices of maize owing to food aid, and the tariff on the export of raw cashews. 263. There is no consensus on the magnitude of the price elasticity of food production, with estimates varying from near-zero to above 2.0, depending on the country, the period, and the methodology. The Mozambican data are too poor and the series too short to permit reliable estimates. It would not be unreasonable to use an elasticity of 0.3 as a working hypothesis.121 Then an increase of 40 percent in the maize price, arising from parity pricing, would lead to an increase of 12 percent in output. 264. The own-price elasticitiel of supply of export crops are generally high, frequently above unity. For instance, Coates (1994), using Tanzanian data for 1969-91, finds an elasticity of supply of export crops of 1.2. Hilmarsson's (1995) review of supply elasticities for cashew concludes with a short-run elasticity of 0.25 and a long-run elasticity of 1.25 (p. 30). Full liberalization of the cashew market, including open domestic marketing and the free export of raw nuts, would permit a farmgate price of at least 50 percent of the FOB price, that is, more than double its 1993/94 price and half again as much as its 1994/95 price in real terms (cf. Table 26). This would generate a major cash injection into the rural areas, alleviating poverty, providing cash for inputs and consumer goods, and providing an additional incentive for the establishment For example, the low technical performance of the cashew processing industry has severely penalized cashew producers. The kernel yield (18.7% of raw nut weight, compared with 23-25% in India) and the proportion of whole kernels (less than 50%; whole kernels command prices up to 50% higher than broken ones) are both extremely low. As a result, Mozarnbique obtains far lower export earnings per ton of processed nuts than its major competitors (e.g., 27% less than India and 10% less than Tanzania). Compounded by the prohibition on exporting raw nuts, Mozambican producers are the lowest paid in the world; at the official exchange rate, their share of FOB prices ranged between 30-40%, dropping to less than 20% in 1993/94, compared with an average of about 50% in neighboring Tanzania. 121 Tsakok (1990, p. 250ff) lists many estimates. Consider the following selection of short-run supply elasticities: 0.39 for rice in Thailand, 1951-65; 0.10 for wheat in India, 1950-67; 0.53 for barley in India, 1951-64; 0.95 for maize in Kenya, 1950-69; 0.94 for millet in Syria, 1947-60. For Tanzania, Coates (1994, p. 116) funds an own-price elasticity of food supply of 0.34, and an elasticity of food output with respect to the price of export crops of 1.08. 92 of the rural trader network. Over the medium and longer term, production and export receipts could treble to 100,000 tons and US$ 65 million per annum, respectively, and farmers' incomes would increase eight-fold.122 (a) Short Term Supply Response: Sources 265. In the short term, production increases are expected to come from the smallholder sector, stemming from the concurrent effects of re-occupation of former crop lands abandoned during the war and the short-term increases in the productivity of female labor (for food crops) and the liberalization of cashew marketing and trading policies (for export crops). 266. There are several reasons for looking to the smallholder sector for a supply response. First, because this sector operates some 95 percent of arable land and generates some 90 percent of total value of production; therefore, it is the only sector with capacity to effect a quantitatively significant impact on agricultural production. None of the four types of farm in the commercial sector has such capacity. Second, because it operates on a self-subsistence mode; any increases in production translate into much larger, more than proportional increases in the marketed production. In addition, development of the family sector would permit the ready re-absorption of returning deslocados and refugiados, create substantial employment, alleviate poverty, and improve the food security situation of the majority of the country's households. Third, the smallholder sector has been shown to be a more efficient producer of most agricultural crops than the large-scale sector.123 267. On the re-occupation of former crop lands abandoned during the war, deslocados and rejugiados have been spontaneously returning to their old traditional lands (in some cases, to other lands in addition or instead); the re-settlement is expected to be completed by the end of 1997. Returnees start farning lands abandoned during the war, replicating the current smallholder sector in areas from which they had been absent for several years. This horizontal expansion of the smallholder sector should provide an increase in total and marketed production proportional to the number of returning families which get settled. 268. On the short-term increases in the productivity of female labor, the smallholder sector does not face a land constraint in the short run. The binding constraint at this stage is the productivity of female labor, which carries out most agricultural work. Therefore, in the short term, the fastest way to increase agricultural production from existing (and returning) farming families is to increase the time women have available for cropping. The main competing claim is for water gathering (3-4 hours per day), so provision of water supply points at the village level is the easiest labor-augmenting investment. There is a good institutional base and experience in installing water points in villages. A second way to increase labor productivity in the short term is the introduction of improved, open-pollinated varieties. It has been estimated that currently available improved varieties can increase current crop yields by some 50 percent. The current practice of distributing seeds (and tools) to returning farmers and farmers in temporary distress provides an expedient mechanism for such introduction. As with water supply, the institutional base and ample experience in this regard already exist. 122 See Hilmarsson (1995) for detail on methodology and data. 123 See the discussion on comparative advantage, p. 94. 93 (b) Long Term Supply Response: Sources 269. In the longer term, i.e., after the once-for-all effects of the short-termn measures mentioned above are exhausted, additional increases in labor productivity would come from new, low-input technologies and farming systems to be adapted or developed, and from the re- introduction (where there was such a tradition) and introduction (elsewhere) of draft animals in tsetse fly free areas. Experience throughout Africa in the introduction of animal draft in areas where this was not a tradition is mixed. Given the fact the Mozambican cattle herd fell from 113,000 in 1970 to fewer than 30,000 today, the scope for successful projects is considerable. 270. Mozambique has excellent opportunities for expanding agricultural production in the long run. In contrast to most African countries, it has a relatively ample supply of land, much of which is of good quality. It has adequate water resources, both in surface water - some of which are currently utilized for power generation and irrigation, and groundwater, particularly in the coastal plains. It has good locations for wildlife development, both for consumptive and non- consumptive purposes. Labor is abundant in absolute terms, albeit poorly skilled and not abundant enough to sustain a much larger production level at the current levels of technology. It has ample pasture resources for much larger livestock production, even though the Northern two- thirds of the country are infested with tsetse fly. Research and extension facilities and activities are limited, however, and infrastructure and processing facilities are mostly in very bad shape. Marketing suffers from the collapse of the state-owned and controlled official marketing system, while the private sector network is just beginning to develop. 271. In the long term, therefore, Mozambique can potentially generate sustained annual growth, variously estimated at 4 to 6 percent per annum. Achieving it presents a major challenge. It will require major investments in infrastructure and transport facilities; the strengthening of the research and extension systems, and the development of a strong private marketing network. It will require the rehabilitation of the existing irrigation systems and of agro-industrial processing facilities. And it will require a conducive macroeconomic and sectoral policy environment. 272. Such an accelerated growth rate would result initially from short-term increases in the productivity of female labor and the re-occupation of former crop lands abandoned during the war. Then, it would come from the introduction of new, simple technologies with minimal input requirements, and the re-introduction of animal traction in those parts of the country where there was such a tradition. Further development would then require additional capital investments; increasing utilization of purchased inputs; and massive investment in infrastructure. C. Comparative Advantage 273. The domestic resource cost (DRC) coefficient approach is used to investigate the comparative advantage of different farm activities and models. The assumptions used in these calculations are discussed in detail in the preparatory document (Moll, 1993) and are listed in the notes of Table 22. Two aspects deserve special mention: the costing of labor and of land. The calculations were done with a shadow wage set alternately at 50 percent and 150 percent of the rural minimum wage, so as to determine the lower and the upper bound, respectively, of the DRC. As far as land is concerned, most Mozambican peasants practice shifting agriculture and are able to obtain land at zero cost barring the labor expense of clearing it. Hence for rainfed production by smallholders, the land cost was taken to be the clearing expense. Large-scale operators, 94 however, incur real costs in addition to clearance: they have to secure their land rights by searching for unoccupied land and by having the land surveyed. Frequently further unofficial payments have to be made. In the absence of reliable rental or purchase information in Mozambique, the shadow price was set at 50 percent of the price of land in the Zimbabwe African Purchase Areas. The coefficient of 50 percent was selected arbitrarily, on the grounds that population-land pressures are lower in Mozambique than in Zimbabwe, and communication and transport are more expensive. In the case of (large-perimeter) irrigated agriculture, no distinction was made between smallholders and large-scale producers in respect of the land cost, since in both categories considerable investments have to be made in irrigation works and maintenance. 274. Food crops. Table 22 summarizes the available information on domestic resource 124 coefficients for food crops . Rainfed cassava production by smallholders shows up as strongly competitive, with a DRC of between 0.15 and 0.3, depending on whether the shadow price of family labor is assumed to be 50 percent or 150 percent of the minimum wage. Under normal circumstances, cassava would be moved to cities even at long distances. The same is true for groundnuts. Maize production by smallholders with improved seeds is highly competitive (DRC=0.6 to 0.8), while that produced using local seeds is less so (DRC=0.9 to 1.3). Maize production by medium-scale farmers was unprofitable at the prices reigning in 1993 (DRC= 1.2), and is certainly less efficient than maize production by smallholders (DRC = 0.6 to 0.8). Irrigated sunflower is another crop for which Mozambican agriculture is competitive, particularly if produced by smallholders. Medium- and larger-scale farmers have significantly higher DRCS (0.9 and 1.1 vs. 0.6-0.7) due to the higher cost of their mechanized operations, including harvesting, and, to a lesser degree, higher administrative and (notional) owner-operator costs. Irrigated bean production by smallholders who rent some mechanized services appears reasonably competitive only if depreciation charges for the irrigation works are not included in the calculation. If they are, such bean production becomes significantly non-competitive (DRC=] .2 to 1.5); productivity increases through improved seeds and chemical inputs may be essential for a competitive bean production under irrigation. Appendix 2 defines the DRC. The main assumptions are listed at the bottom of Table 22. It is worth highlighting that labor in the smallholder sector was valued alternatively at 0.5 times and 1.5 times the minimum wage so as to give lower and upper bounds. The exchange rate assumed was midway between the official and that unofficial rates. Given that Mozambique was receiving large quantities of foreign assistance in 1993, these exchange rates were overvalued. In the future as aid declines the exchange rate will depreciate so that all the DRC figures in Table 22 will decrease. 95 Table 22. Yields and Domestic Resource Coefficients (DRCs) CROP Farm Yield Irri- DRC, wage DRC, wage Border' Border' 1993 Size (tons gated? = 50% of = 150% of parity price, parity price, Mini- (S, M, per minimum minimum at city at farm mum L) ha) (MT/kg), (MT/kg), price 1993 1993 Maize, local seed, 1993 S 0.4 N 0.9 1.3 880 540 425 price Maize, Matuba, 1993 price S 0.8 N 0.6 0.8 880 540 425 Maize, local seed, 1996 S 0.4 N 0.8 1.1 880 540 425 price Maize, Matuba, 1996 price S 0.8 N 0.5 0.6 880 540 425 Maize, 1993 prices M 1.4 N 1.2 1.2 880 540 425 Maize, 1993 prices M 2.0 Y 2.0 2.0 880 540 425 Groundnuts S 0.25 N 0.3 0.4 2200 1600 990 Cassava S 6 N 0.15 0.3 1100 770 None Sugar Cane L 5.5d y 1.1 1.1 1400 320 None Rice L 7.0 Y 0.7 0.7 935b 570b 580b Sunflower' L 2.5 Y 1.1 1.1 1730 1200 540 Sunflower' M 2.5 Y 0.9 0.9 1730 1200 540 Sunflowerg S 2.0 Y 0.6 0.7 1730 1200 540 Beans S 0.6 Y 1.2 1.5 2600 1700 1050 Citrus M 25 Y 0.2 0.2c 1900 1400 None Cotton, 1993 prices S 0.5 N 0.4 0.9 - - - Cotton, 1995 prices S 0.5 N 0.15 0.3 - - - Cotton, 1993 prices L 2.2 N 1.2 1.2 - - - Cotton, 1995 prices L 2.2 N 0.6 0.6 - - - Cashew (current) S - N 0.3 0.7 - - - Cashew (improved) S - N 0.1 0.3 - - - a Smallholder = 2-6 ha, Medium =50 ha, Large = 200-500 ha. bBorder parity price of shelled rice, at city: MTI700. Multiply by processing cost factor of 0.55 to get unshelled rice price ("em casca"). Other prices: also unshelled. Yield reported as unshelled. ' Citrus: The low DRC assumes a c.i.f. price of $500/ton, using the schedules in the World Bank publication "Price Prospects". The National Irrigation Development Master Plan (SOGREAH 1993) equivalent is about $300/ton, perhaps because poorer quality citrus were expected; at this price the DRC is 0.4. dRaw sugar, i.e. after processing. ' Import parity in all cases (food crops, sugar cane, rice, sunflower, beans, citrus). Export parities were not calculated for the export crops. g Improved (forecast) models of sunflower, viz., with higher levels of inputs and yields. Assumptions: 200 km from city: opportunity cost of family/unskilled labor = 50 percent or 150 percent of minimum wage: shadow exchange rate (mr5200/$) midway between the official and the informal rates. The official rate in October 1993 was MT4400/$.) Land for smallholders was uncosted but land clearance costs were included. Land for large-scale producers and irrigation lands is costed at half of Zimbabwean land values. For irrigated crops. the costs include a charge for depreciation of the irrigation works of $8000/ha spread over two annual crops for 30 years. Sources: Moll (1993), Table 8. Farm models were adapted from SOGREAH (1993). Irrigation construction costs from Barghouti and Le Moigne (1990), p. 10. 275. Regional competitiveness of maize. A more detailed analysis of the impact of transport costs on the competitiveness of smallholders in different regions was carried out (Table 23). If the opportunity cost of labor is estimated at 50 percent of the minimum wage, all regions 96 in the country can produce maize for sale in the Maputo market competitively. If the opportunity cost of labor is instead 1.5 times the minimum wage, as assumed above, then only the Manica highlands, which enjoy higher yields, can grow maize for the Maputo market competitively. With improvements in the transport sector, maize for the Maputo market could be grown efficiently in Tete as well; with introduction of existing improved varieties, DRCs would fall even further, enabling longer-distance transport, which would further improve the possibilities of expansion in Tete and in Niassa. Table 23. Regional Competitiveness of (unimproved) Maize Production: Domestic Resource Coefficients (DRCs) And Farmer Earnings Per Day in 1993 Meticais Route Road Ocean Yield (tons Existing transport & Improved transport & distance transport cost per ha) margins margins (km) ($/ton) DRCa MT/day DRC MT/day Maputo 50 0 0.4 1.4, 0.8 950 1.3, 0.7 1400 Gaza-Maputo 200 0 0.4 1.5, 0.9 500 1.4, 0.8 1000 Chimoio-Beira- 200 25 0.8 0.9, 0.6 2600 0.8, 0.5 3500 Maputo Tete-Beira- 600 25 0.8 1.3, 0.9 200 1.0, 0.7 1600 Maputo C.Del-Pemba- 300 40 0.8 1.1, 0.7 1300 0.9, 0.6 2400 Maputo Method: Five routes are selected. The first is "Maputo", viz., a 50-kilometer distance from the farm to the capital by road. The second is "Gaza-Maputo", a 200-kilometer distance. The third is from the capital of Manica (Chimoio) by road to Beira, and thence by ship to Maputo. The fourth is from Tete province to the Beira port by road, about 600 km in all, and thence to Maputo. The fifth is from the interior of Cabo Delgado to the Pemba port and then by ship to Maputo. Assumptions: Unimproved seed and methods; labor days 78; maize price, f.o.b. Mexican Gulf is $109; ocean transport to Maputo is $70; the maize price at the city wholesaler is border parity, i.e., c.i.f. plus port and other import charges. Mozambican road transport is Mt500/tonlkm (unimproved) or Mt400 (improved); Mozambican ocean transport is $40 for Pemba-Maputo and $25 for Beira-Maputo; in addition to transport costs, margins of 20 percent (unimproved) or IS percent (improved) are charged bv both the retailer and the wholesaler. ' First DRC assumes a wage 150 percent of the minimum (Mt3,150), the second 50 percent of the minimum (Mtl.050/day). b Official exchange rate, October 1993: MT4,400/$. Shadow exchange rate used: MtS,200/$, see Table 22. 276. Export crops: cashew. Production of raw cashew nuts enjoys a very strong comparative advantage, with DRCs of 0.3 to 0.7 for normal, standard production modes, depending on the assumed price of family labor (50 percent or 150 percent of the minimum wage). Improved production modes are even more competitive, with DRCS of 0.1 to 0.3. If exports were freely permitted, farmgate prices would rise considerably - to almost double the official prices. In contrast, the processing of nuts by several Mozambican factories carries a strongly negative effect, "creating" negative added value and thus wasting scarce resources; the high cost of processing and low percent of whole kernels extracted are to blame. Only 43 percent of exported kernels are whole, whereas in Tanzania the figure is 55 percent, in Brazil 70 percent and India 83 percent (Hilmarsson 1995, p. 20). The implicit subsidy to the processors is 97 paid by taxpayers and donors, whose monies were in the past channeled to Caju de Mocambique in the form of loans which could not be repaid. 277. Cotton. The competitiveness of cotton varies with its volatile international price. In 1993, at a cotton fiber price of $1.28/kg, the DRC for outgrowers was between 0.4 and 0.9, depending on the costing of family labor (50 percent or 150 percent of the minimum wage). Estate cotton production was uncompetitive, with a DRC of 1.2, indicating that the cotton firms spent $1.2 of domestic resources in order to generate $1 of foreign exchange (excluding the cost of destumping, long since fully depreciated). The cotton fiber price rose to $1 .76/kg in 1994 and $2.20/kg in 1995, making both forms of production substantially more competitive (DRC = 0. 15- 0.3 for smallholders and 0.6 for large-scale producers). A recent study (Fok 1995) indicates that Mozambican cotton productivity could be greatly improved given adequate extension, input supply, and varietal innovation. The average outgrower yield in Mozambique is between 300 and 400 kg/ha, equivalent to the yield in West Africa in 1960, which has since risen to over 1000 kg/ha. 278. Land-saving technological innovations are most likely to be adopted in areas where land is scarce and its price is rising. Given that Mozambique is one of Africa's most land- abundant countries, it would be surprising if in the land-abundant north peasants were to start using chemical fertilizers for field crops. This hypothesis is borne out by simulations of DRCS of smaliholder cotton-growers in Cabo Delgado. The details are summarized in Table 24. The base case is the current arrangement of smallholder cotton in the interior of Cabo Delgado province: the peasants clear the land by burning and use a rotation which includes cotton for the first two years before moving on to new land. Pesticides are applied but not fertilizer. As before, the shadow price of labor is taken to be alternately 50 percent and 150 percent of the minimum wage, so as to bracket the "true" labor cost. It is assumed that for every two years of good rainfall there is one drought year, during which yields fall from the usual 500 kg/ha to 250 kg/ha. Table 24. Domestic resource coefficients, smallholder cotton in Cabo Delgado province, with and without fertilizers Shadow price of labor, as a percentage of the minimum wage. 50% 150% Base case: without fertilizers Year of good rainfall 0.2 0.5 Drought year 0.8 1.8 2 good, Idrought 0.3 0.8 Simulation: with fertilizers Year of good rainfall 0.2 0.3 Drought year -2.9 -4.7 2 good, Idrought 0.8 1.2 Source: Moll (1993), and further calculations by the author. Data from LOMACO (Cabo Delgado) and published sources. Assumptions: Fertilizer application is 200 kg of N+P+K per hectare. Yields of seed cotton in a good year: 500 kg per hectare without and 1 100 kg with fertilizer; 98 yields 250 kg with and without fertilizer in a drought year. Price of cotton fiber: mean of 1993 price ($1210/tonne) and 1995 price ($2200/tonne). The no-fertilizer case incorporates the cost of land clearance. 279. The simulation is done assuming conditions similar to those in West Africa: the fertilizer applied is 200 kg/ha, and the yield with good rains is 1100 kg/ha. During a drought the yield falls to 250 kg/ha, as in the case without fertilizer. 280. The results are instructive. In a year of good rainfall, the use of fertilizer is superior (DRC=0.2-0.3, depending on the shadow wage) to the maintenance of fertility by shifting cultivation (DRC=0.2-0.5). Drought years, however, are a disaster for fertilizer users; the DRC iS negative, showing that the value is actually subtracted. If one year in three is a drought, on average, the final DRC estimate for shifting cultivation of cotton is 0.3 to 0.8, depending on the shadow price of labor, while the DRC for fertilized cotton is 0.8 to 1.2. Furthermore, these DRCs reflect the ex post situation. Ex ante, the asset-poor peasant is risk averse, and would probably choose shifting cultivation over fertilizer usage even if their efficiency were approximately equal as measured by the DRC. 281. Citrus production, in contrast to cotton, is highly competitive with a DRC of 0.2 even after accounting for depreciation costs for irrigation works. As is stressed elsewhere (page 115), substantial scope for increased production exists and some of this could be undertaken by smallholders in outgrower schemes. 282. Farm size and productive efficiency. In the crops for which data are available (maize, cotton, and sunflower), smallbolders are more efficient producers than medium- and large-scale operators. This is even at the current low technological level; it is presumed that their competitiveness could be further increased by the adoption of already available technologies improving the productivity of labor. This relative advantage comes from their low opportunity cost of labor, which leads them to employ labor-intensive methods. The relatively high economic cost of machinery and of the owner-operators' time make medium and larger operations less efficient. These, however, were efficient in a few cases, with domestic resource coefficients below unity. The relative efficiency of smallholders reflects as well the country's relative factor endowment, i.e., the low opportunity cost of labor and of land (except in localized areas in the south), and the high opportunity cost of imported inputs and machinery. Even without the present subsidy structure (loans with negative real interest rates), many of them would probably continue to survive. Hence, fostering development of the family sector, and channeling public investments like roads to the areas where they constitute the predominant production system is justified on both efficiency and equity grounds. At the same time, the wisdom of the current policy of fostering the development of commercial farmers using capital- intensive technologies through subsidies is doubtful. 283. Conclusions. The analyses above, the supporting tables and the working paper (Moll 1993) lead to several overall observations. * Assumptions on the opportunity cost of labor in the family sector are most influential in determining DRCs. For example, the DRC for maize grown in interior Cabo Delgado and sold in Maputo is 1.1 at the higher assumption (1.5 times the minimum wage) but only 0.7 at the lower one (0.5 times the minimum wage). * Improvements in the transport system and increase in competitiveness among traders only translate into a major improvement in DRCS for long hauls. For instance, the DRC for maize produced 50 km away from Maputo would barely fall from 1.4 under the current transport 99 system and marketing regime to 1.3 under the improved ones. For Tete, instead, some 600 km away, they would fall from 1.3 to 1.0, respectively. * Yield increases have a substantial effect on DRCs. In the case of cotton, for example, introducing improved cultivars would reduce the DRC from 0.8 to 0.5. Concentrating technological development on technologies which enhance the productivity of labor would reinforce Mozambique's strong comparative advantage in several crops. * In land-abundant areas of the country, the use of fertilizers with most smallholder crops is inefficient compared with the current alternative of shifting cultivation. This does not, of course, exclude the possibility that smallholders could succeed in certain specialized crops or specific technologies. * The comparative advantage of the family sector in areas where there is already a land constraint (e.g., most of Maputo province and along river basins in coastal areas in Gaza and Inhambane), where the opportunity cost of land is not negligible, should be expected to be lower than elsewhere. Since such areas are expected to progressively expand over the years, the comparative advantage of the family sector there should be studied in more detail, and the technologies required to sustain their competitiveness studied and disseminated. D. Subsectoral Conditions and Prospects 284. In this section the conditions and prospects of several of the major subsectors are described in more detail. The discussion starts with food crops, of which the most important is maize. It proceeds with export crops: cashew, cotton, citrus, and others. Finally the prospects of the livestock and dairy sectors are dealt with. (a) Food Crops 285. Marketing. AGRICOM, the Agricultural Marketing Enterprise, a parastatal, was created in 1981 out of recognition that neither the state farm system nor the remaining private trader system were meeting the family sector's marketing needs. The trader system had been virtually paralyzed by the abrupt departure of most of its Portuguese members at independence, and further weakened by destruction of trucks and shops because of the war. AGRICOM then provided marketing services and input supplies for a segment of the family sector, often working through mobile brigades under severe security conditions. For all practical purposes, AGRICOM was also the instrument for implementing and enforcing the Government's pricing policy. 286. The Government's pricing and marketing policy began a process of liberalization in 1983, beginning with the "freeing" of fruit and vegetable prices from bureaucratic control. This process was broadened to include far more commodities under the ERP. The gradual relaxation of crop pricing restrictions, plus the progressive recovery of the trader sector after years of war, weakened AGRICOM's monopsonistic and monopolistic position and eroded its earnings. This was compounded by progressive financial problems. 287. In early 1994, Government abolished AGRICOM and created the Instituto de Cereais de Mo!ambique (ICM), which took over the staff and physical assets of the organization, but not its accumulated debts, which were left with Government. The mandate of the ICM was to: act as buyer of last resort, paying the farmers the official pan-territorial minimum price, to manage strategic stocks to ensure food security; to contribute to the stabilization of producer and 100 consumer prices; and (although this has been accorded less importance) to provide inputs and disseminate storage technologies. The parent ministry of the ICM is the Ministry of Commerce. Staff numbers were cut from AGRICOM's 1600 in 1992 to 700 (Coulter 1995, p. 21). The ICM has sold off the vehicle fleet owned by AGRICOM. The lCM borrowed from banks at commercial interest rates (43 percent nominal) but did not succeed in obtaining donor funding. 288. In its first season of operation, the ICM purchased about one-quarter of the country's marketable surplus of maize (Coulter 1995, p. 22) from farmers and the commercial trade network. The main client was the World Food Programme. This helped to contribute to price rises in the latter half of 1994, a welcome development in the light of the production disincentives generated by food aid in recent years. But there were widespread problems of late payment of farmers (one complained of no payment six months after delivery) and there were allegations of fraud and theft. The ICM refused to provide any financial statements for 1994 or 1995, so that it is not possible to assess its financial viability. The position of the lCM is fragile because its role has essentially been that of a supplier to food aid donors, and this clientele is likely to diminish sharply in forthcoming years. So far the ICM has not established food security reserves. 289. The storage capacity of the ICM is underutilized. From August 1994 to May 1995 the institute bought 37,000 tons of maize; from August to November 1995 it bought 40,000 tons. Its total storage capacity is 120,000 tons. Some of the excess was rented to organizations such as the World Food Program. It appears that a considerable amount of storage could be sold without prejudice to the parastatal. Thus privatization could take the course of drawing down the assets of the organization and applying the proceeds to general govemment revenues. 290. Is there a rationale for the existence of a parastatal body such as the 1cM? Upon close examination of the purposes for which such a body might exist, it emerged that each objective can be achieved more effectively by a different kind of arrangement. These are summarized in Table 25. The first purpose - inter-annual price stabilization - is at the heart of the stated objectives of the ICM. But this could be more efficiently achieved by making the ports more efficient. The high cost of moving maize through the ports would result in large inter-annual price fluctuations. In years of excess supply, the local (Mozambican) price would have to fall very low before it became profitable to export maize to other countries. In years of poor rainfall, the local price would have to rise very high before it became profitable for merchants to import maize. By making the ports more efficient, this export-import parity band could be reduced by up to $40/tonne (Coulter 1995, p. 40). 291. The second intention of the ICM was to act as a wholesaler, intermediating between farmers and retailers on the one hand, and donor purchasers on the other. But competitive private traders could undertake this task, and would probably do so more efficiently than would a state-run body because they would be driven by the profit motive. There is no market failure which would justify state intervention. The very existence of AGRICOM and now ICM, however, has crowded out private wholesaling, and so special measures should be taken to reestablish private operators. This could be done by selling the ICM warehouse network and vehicle fleet. 292. A third objective for a grain marketing parastatal is to provide inter-seasonal storage. This would smooth prices between the harvest period when prices are low and the following growing season when prices rise owing to low supplies. The argument is that a grain parastatal (AGRICOM previously, and now ICM) could smooth inter-seasonal prices, improving welfare both for farmers, who can then sell at higher prices at harvest time, and for consumers, who are protected from price spikes in the growing season. 101 293. An unintended effect of the AGRICOM's purchasing at fixed prices immediately after the harvest was to decrease incentives for private inter-seasonal storage by farmers and traders. In so doing AGRICOM created a justification for its existence. A parastatal was "necessary" because there was little private storage, and the activities of the latter repressed the market further, and so on in a vicious cycle. 294. There were additional reasons for the low levels of private storage. Food aid shipments robbed traders and farmers of incentives for storage, because they eliminated inter- seasonal price movements. Food aid has disturbed traditional coping mechanisms. According to the GTZ missions in Manica and Sofala, farners are less cautious in selling their grain because they have been conditioned by the availability of food-aid grain. Development workers have repeatedly expressed their concern over food-aid dependency (Coulter 1995, p. 30). 295. Examination in the field has shown that Mozambican farmers are capable of storing maize (Coulter 1995, p. 30). Some 90 percent of farm households in the North and the Center have on-farm storage facilities, while fewer than 50 percent in the South do. During a period of scarcity in January 1995, maize farmers in Cabo Delgado sold maize from their stocks to petty traders who sold it in Nampula where prices had risen (Ministerio da Agricultura e Pescas, 1995e, p. 54). With appropriate incentives and advice farmers would be capable of storing substantial quantities for sale. If the financial system improves - which in turn would will occur only if inflation is reduced to 15 percent or less - traders will borrow money for storage of grain, as happens in Ethiopia and Ghana. By playing the market, traders would contribute to inter-seasonal price stabilization. Comparison with similar situations elsewhere in Africa suggests that the lion's share of storage will, however, be undertaken by farmers (Coulter 1995, p. 3 1). 296. A fourth objective for a grain marketing parastatal is to avoid major price surges. Food aid shipments are falling and in a short while food aid releases will no longer be an effective lever for regulating domestic price levels. The argument is that a grain parastatal could maintain food security reserves in the districts and release grain from these when the price reached a predetermined ceiling. Thus major price spikes would be eliminated. A related problem is that if the price surges in a city, traders move grain from rural districts to the city in the hope of fast gains, thereby raising the grain price in the rural districts and endangering food security. Well-timed releases from strategic stocks would therefore limit the out-movement of grain from rural areas. But there is a simple alternative, which consists of allowing traders to import maize from the world market. The resultant moderating effect on prices would quickly feed through into the supplying districts, and curtail the outflow of grain. 297. For all the four functions listed in Table 25, the "alternative approach" is less dependent upon political and bureaucratic processes, less institutionally demanding, and less demanding of subsidies than the alternative of creating a parastatal. As of late 1995, in any case, the ICM was acting simply as a wholesaler, purchasing in the main from traders, and it was not logistically feasible to ensure that the traders had passed on the ICM's allegedly good prices to the farrners. It is hard to imagine any justification for government ownership of a wholesaling business. 298. At least two options remain for the disposal of the lCM. The simplest and preferred option is to sell the ICM warehouse network and vehicle fleet in small lots to the highest bidder. This is the preferred option because it could be done quickly and because it would not require any purchaser to come up with the large sum of money which would be required in the event of sale en hloc. The second option is to sell the ICM as is, with a view to establishing a large 102 privately owned storage company which could assist in the standardization of Mozambican produce in ways acceptable to international buyers. Through the development of warranting services, the company could improve the access of traders to credit. This option is second best because it imposes a constraint, namely that a large amount of finance would have to be amassed by a prospective buyer, so that the process of sale could be delayed. In any case there is no guarantee that the new buyer will maintain the firm intact. The resources might be more efficiently used by being spread around a number of existing trading or other firms. In time to come, standardization and warranting services could be developed by smaller firms. There is no need to create a large privately-owned monopsony firm in order to develop these services. If there is an information problem or some other market failure preventing the rise of these services, it is best addressed at source. Table 25. Alternative Approaches to Fulfilling the Functions of a Grain Marketing Parastatal Function Alternative approach I Inter-annual price stabilization Make ports more efficient, and thereby reduce the import-export parity band by up to $40 per tonne 2 Wholesale procurement and trading Encourage development of competitive private trading 3 Inter-seasonal storage Encourage farners to store, with traders complementing 4 Avoidance of major price surges A liberal import policy Source: Coulter (1995), p. 40. 299. When food aid is eliminated, the warehouse network controlled by DPCCN (the agency in charge of emergency food distribution) should likewise be sold. In the aggregate, the facilities presently controlled by DPCCN, ICM and private warehousers are adequate to provide necessary capacity at the wholesale level. 300. Important market services, like market information and crop forecasting, are presently provided by the MOA. They are donor-funded, however, with no guarantee of long- term sustainability. Immediate decisions are necessary to secure their full institutionalization and development. 301. Long-distance trade, integration and tariff proposals. Long-distance trade is expected to play an increasing role in the future, particularly if the government succeeds in improving the efficiency of operation of the ports. It has been shown that, under reasonable assumptions concerning improvements in the legal regime, the marketing system and the ports, Mozambican maize would find its way to South Africa in years when the latter country is in deficit. Prior to independence, trading links between the north and the south of Mozambique did exist. Today even bulky products such as cassava sometimes find their way by truck from the north to the south of the country. However, for the northern provinces (Cabo Delgado, the 125 See the extended discussion in Coulter (1996), Appendix 4. 103 western parts of Nampula province, Tete) to produce cereals for Maputo is unlikely to be more than a marginal proposition 26; they would do better to export to Tanzania, Malawi and Zimbabwe.127 The volume of trade would be greatest with countries to the west, on account of the rail link from Nacala to Malawi. The north of Mozambique could also export to markets further afield, via the ports of Nacala and possibly Pemba, to, for instance, Sudan and Kenya whose growing seasons do not coincide with those of Mozambique. Areas which are less distant, such as the Manica highlands, could produce maize and other products for the Maputo market. 302. In the light of these findings about likely future trading patterns, is there any justification for a special import tariff on, say, cereals, in order to encourage north-south integration and promote self-sufficiency in cereals?128 This seems unlikely. The optimal amount of north-south integration is that which would arise in a context where there were no important market failures. Currently there is at least one129 market failure: the inadequate state of the road system. The amount of north-south trade is consequently suboptimal. But the best way to address this problem is through improving the road system, which is already being done through the ROCS 11 program.130 Using a special import tariff would be a blunt and inaccurate instrument. While it would increase the quantity of north-south trade, it would impose unnecessary costs in so doing, by raising the cost of maize for consumers and encouraging the cultivation of maize in unsuitable parts of the south. 303. Concerning cereals self-sufficiency, these findings indicate that this objective would be expensive to attain. Due to the elongated north-south structure of the country, and the obvious opportunities for regional trade at both ends, self-sufficiency would require major departures from the patterns that the economy, unfettered, would "naturally" produce. The objective would probably require a substantially increased tariff, which would entail the costs mentioned above. Self-sufficiency is an impossible goal given Mozambique's drought-prone agriculture; with a probability of one, the country would from time to time be forced to import. In any case self-sufficiency and household food security are not coterminous. It could be argued that the government's highest goal should be householdfood security, which in turn depends upon many other factors such as household incomes and the capacity of households to smooth their incomes by crop storage or financial instruments. An expensive program in food self- 126 The calculations were done both under current price assumptions and under assumptions of realistic price reductions arising from improved shipping, improved port operation, and reduced marketing margins due to the secular increase in trade. The findings as reported in this paragraph reflect the calculations on the basis of assumptions of long-term price reductions. ]27 Koester (1986) showed that maize production in northern Mozambique is negatively correlated with that in southern Tanzania and more or less independent of that of Malawi and Zimbabwe. 128 This argument was invoked during the course of discussions with the government in 1996 concerning the Third Economic Recovery Credit. It is sometimes stated in graphic terms: Why should the north of the country be exporting maize when the south is being forced to imnport and rely on food aid? 129 Up to 1994 there was another problem: the high cost of north-south sea transport, owing to the monopoly status of the government-owned shipping company. The sea transport market has since been opened and there is now an additional private sector competitor. Prices have correspondingly fallen. 130 See Chapter 3 for further detail. 104 sufficiency would not necessarily improve the food security position of the most vulnerable households.13' (b) Cashew 304. The cashew industry is one of the most important components of the economy of Mozambique. It is an important source of income for over a million smallholder farmers. In the 1970s, Mozambique was the world's biggest exporter of cashews, capturing 40 percent of the world market. Due to the war and the disorganization of the industry the situation deteriorated to the point where Mozambican exports lag behind those of India and Brazil. Marketed output of raw cashew peaked at 200,000 tons in 1972, and thereafter fell steadily, owing to the war and to unfavorable pricing policies, reaching its nadir of about 25,000 tons in 1982/3. In the last ten years marketed output has fluctuated between 30,000 and 50,000 tons per annum. Marketed output increased to about 60,000 tons in 1995/6 (Figure 25), owing partly to the decrease in the tariff on exports of raw cashews to 20 percent in that season.132 250 200 - `150 Independence 100 50 0 Figur 25 Marketed outpu of raw cashew in Moamique, 1955-9 133 . N t t - h .- r . siy F a F, te - a o 1Figure 25. Marketed output of raw cashew in Mozambique, 1955-961nd 133~~~~~~~~~~~~~~~~3 Milon.bia Thesehelltw1 ilo indsr 1992d riseing ten yert3 S$10 million,i 99u torte patialitngoth Mozambique's total foreign exchange revenues today. If in addition the peasant farmers were 131 See the section on Food Aid in Chapter 4 of this volume for further detail. 132 Data on the production, processing and export of cashew in Mozamnbique are presented in the Appendix Tables. 133 Source: Hilmarsson (1995), p.4, in turn from the Secretariado de Estado de Caju. 134 Hilmarsson (1995), p. 22. In 1971, foreign exchange earnings from cashew alone were some $151 million. These fell to $ 16 million in 1992, rising to $30 million in 1994 due to the partial lifting of the export ban on the raw product. Meanwhile total foreign exchange revenues were $164 million in 1994 (Hilmarsson, 1995, p. 5). 105 permitted higher prices, so that they supply much more raw cashews to the market, then foreign exchange revenues could reach $230 million in ten to fifteen years. 306. It is widely accepted that the most important objective for the cashew sector is to attain a high rate of growth. The expansion of the supply of the raw nut to the processors could result in massive increases in earnings by peasants, with important multiplier effects throughout the rural economy. Increased output by the processors would mean increased value added for the economy. The increase in the exports of the processed good overseas would provide increased foreign exchange. There could be a complementary process of skill acquisition by the managements and the workers of the processing industry. Employment in the processing industry would rise. Another major area for development is in processor-grower linkages. Processors could find it in their interest to provide peasants in the vicinity with technical assistance and inputs so as to improve the quality of the nuts they supply. There is no question that it is in the long-term interest of the growers and of society at large to have a vibrant and growing processing industry. 307. The avenue that the government has chosen to achieve its growth objectives in agriculture is to use the private sector and rely on private initiative. The government has accepted that direct government ownership and control stifles incentives and mires industrial producers in bureaucracy. The government has also accepted that it will be essential to encourage private marketing, by removing the barriers to licensing of traders and by building roads. On all of this there is a broad consensus. There is still debate respecting one policy issue: to make the cashew industry expand, should the cashew processors be protected in some way? Or should they, like other exporters, be subject to the discipline of international competition? This section concentrates exclusively on this aspect. 308. In Mozambique in the past the cashew processors used to be protected by a ban on exports of the raw nut. This meant that farmers (apanhadores) and traders could sell only through channels which led to the fourteen cashew processors in Mozambique, who would then prepare the product for table consumption and export it in processed form. It was a boon for the processors because they were not required to compete with traders who wished to export the raw nut to India or elsewhere. The prices which the processors paid to the local traders were lower, and the prices received by the apanhadores were reduced. This also meant that the farrners did not supply as many cashews to the market. So farmers' incomes were reduced both because their prices were lower and because the overall size of the raw cashew market was smaller. In effect, the export ban redistributed wealth from the apanhadores to the Mozambican processors and their employees, while also causing "deadweight" losses to society. 309. The export ban continued until 1992, when a complex licensing system was put in place enabling a small number of operators to export a proportion of the crop (never more than one-third) in raw form. Finally in the 1994/5 season the government removed the quantitative restrictions on exports of the raw nut in favor of a 26 percent export tax. Subsequently it reduced the tax to 20 percent in 1995/6 and to 14 percent in 1996/7. 310. The data suggest that the removal of the export ban and the reduction in the export tax have fed through to improved farngate prices. As is shown by Table 26, during the period 1987 to 1992 the farm gate price was between 30 and 54 percent of the world price. This ratio fell to about 20 percent during the period 1992/3- 1993/4.135 Then with the reduction in the export tax the percentage rose to 27 percent in 1994/5, 44 percent in 1995/6 and 55 percent in 35 In Tanzania, most farmers at the same time enjoyed a price of around 50 percent of the world price. 106 1996/7. The real farm gate price has also moved in the favor of smallholders, as would be expected. Between the 1993/4 and 1994/5 seasons, the real price rose some 23 percent, and between 1994/5 and 1995/6 it rose 13 percent, before falling back 11 percent between 1995/6 and 1996/7. The overall improvement in the real price since the removal of the export restrictions was 25 percent. Table 26. Farmer Price as a Share of Cashew Export Earnings Year Farmgate price Average Farmer Share export of Mt/kg US$/kg Price Export Price (US$/kg) (%) 1987/88 105 0.20 0.66 30 1988/89 165 0.25 0.65 38 1989/90 200 0.23 0.72 32 1990/91 380 0.36 0.67 54 1991/92 460 0.24 0.55 44 1992/93 700 0.14 0.70 20 1993/94 1000 0.15 0.71 21 1994/95 2000 0.19 0.70 27 1995/96 3500 0.30 0.68 44 1996/97 4500 0.39 0.71 55 Source: 1987-1994: Chausse (1995), Table 5, Jaffee (1993); data from the Secretariado de Caju and the "Grupo de Trabalho sobre o Caju". a Converting Meticais at the official exchange rate. 311. Another defect of the raw cashew export ban was that it created incentives for illegal exports. Manipulations of the official numbers indicate a large and increasing gap between the cashews collected on the one hand and the sum of cashews processed and legal raw exports on the other. For instance, in 1992, 54,000 tons of cashew were marketed, of which 30,000 tons were processed and 6,000 tons were legally exported under a special licensing system.136 The discrepancy was 18,000 tons, of which there is no account. Probably some of this was exported illegally, and there may have been some understatement by the processors of their output so as to strengthen their case for industrial assistance. 312. The new private processors of cashew contended that they would need a period of time to restructure their plants before they would be in a position to compete with exporters of the raw product. But a detailed study in loco by a cashew trade expert (Kumar 1995) shows that a new plant using the existing technology would be capable of competing with raw cashew exporters, and that under conservative assumptions. Table 27 records the results of the analysis.137 136 See the detailed data in the Appendix Tables. 137 More detail is given in the Appendix Tables. 107 Table 27. Competitiveness of local processing and of raw nut exports Alternative price Resultant assumptions farm gate ($/ton) price FOB Factory (Mt/kg) gate I Current (1995) export of raw cashews from Mozambique, with: (a) total margin 50% 750 3789 (b) total margin 40%, assuming improvements in the 750 4546 trade system 2 Hypothetical model of a firn using the current capital- 550 4167 intensive technology 3 Hypothetical model of a firm using a labor-intensive 700 5304 technology 'Note: exchange rate was that of 8/31/95, Mt 10103 per dollar. Source: Kumar (1995), pp. 34-35, 63, combined with author's calculations. See further detail in the Appendix Tables. 313. The option of exporting raw cashews is examined in line 1. Given reasonable assumptions, the farrn gate price would be Mt 3789, using the August 1995 exchange rate. With an improved trading system, it could rise to Mt 4546 - well above the minimum farm gate price of Mt 3000/kg proposed by the government in August 1995. 314. In line 2 are presented the results of a hypothetical model of a factory using the current capital-intensive technology. Under realistic assumptions, and effective management, the farn gate price could be as high as Mt 4167, well above the likely equilibrium free market price if there were no hindrances to exports. 315. In line 3 are presented the results of a hypothetical model of a factory using a labor- intensive technology. Now, the farm gate price could rise as high as Mt 5304/kg, using the August 1995 exchange rate. 316. The policy consequence of the above discussion is that there is no need for protection for the cashew processors. A new plant is capable of competing with the next best alternative of exporting the raw product. The present plants are of course not new, and most require some rehabilitation, but the point is that rehabilitation of an existing plant is cheaper than building a new one, and they already have a trained labor force. Therefore with disciplined management the existing plants would be able to fare better than the hypothetical model in line 2 in Table-27. The point about disciplined management is crucial. As one experienced investigator noted, "After having analyzed the data provided by a plant in Beira as well as the most recent statistics of the Machava factory, one can conclude ... that the fact that competition was restricted amongst a few processors (unlike the 300 or so in India) led to a more relaxed management of these factories" (Kumar 1995, p. 43). 108 317. Another way of looking at the issue of protection for cashew processors is given in Table 28. A minimum level for the export price of raw nuts is $700/ton. We proceed to calculate the "level of protection required" by comparing this number with factory cost price or: [export price x kernel outturn - processing cost]. Two "inefficient" and three "efficient" scenarios are constructed. The most inefficient -the current situation - has a low export price of $4090/MT and a low kernel outturn of 18 percent. It would only be competitive if the cost of the raw nuts to the factory were $436/MT, so that it would need protection of 38 percent. The AlCaju kernel outturn of 20 percent is a slight improvement, but would still require 26 percent protection. The new Anglo-American investors are confident of attaining a higher output price of $4400 (due to higher quality nuts) and a 22 percent outturn and can compete with the raw nut exporters by paying $718 for the raw product. Another alternative is that given by Companhia Caju Beira whose lower processing costs would enable them to pay significantly higher prices for the raw material. The winner is the labor-intensive process referred to above, which could profitably pay $1000 per ton of raw material and still turn a profit. Indeed, even the processed product prices of $4090-$4950 assumed in Table 28 are modest compared with the numbers quoted in the international trade literature, giving rise to the question whether underinvoicing is occurring or whether further quality improvements could be achieved. At any rate it is clear that with effective management local processors would be capable of competing with raw nut exporters. This finding strengthens the case for a complete liberalization of the cashew nut market. Table 28. Protection required by cashew processors' Industry requiring Efficient industry, requiring protection no protection Raw b Anglo- Labor nut Current AlCaju C Ameri- C.C. inten- export can d Beira e sive f I . Export price ($/MT) (kernel) 4090 4090 4400 4400 4950 2. Kernel outturn £ 0.18 0.20 0.22 0.22 0.235 3. Export price ($/MT) of raw nut 700 736 818 968 968 1163 equivalent 4. Processing cost ($/MT) 300 300 250 150 150 5. Factory gate price ($/MT) (= row 436 518 718 818 1013 3 - row 4) 6. Domestic marketing cost 200 150 150 150 150 150 7. Farm gate price (= row 3 - row 4 500 286 368 568 668 863 - row 6) 8. Protection required (= [row 5 - - 0.38 0.26 -0.03 -0.17 -0.45 700] 700) 9. Producer price supportable by 0.71 0.39 0.45 0.81 0.95 1.23 industry - border price a Data on kernel prices, transformation ratios, and processing costs are based on industrial estimates, but no column should be taken as representing any company's official position. b Estimated present price structure, based on SimT 4090 kernel price from Euragel; 0.18 yield ratio from industry. c Assumes $/MT 4090 FOB price; data from Euragel (40 percent wholes), i.e. approximately 20 percent lower than the Indian FOB price and 7 percent lower than the Brazilian FOB price. 109 dAnglo-American proposal assumes slightly discounted Brazil CIF price for kernel (60 percent wholes): and Anglo- American/Oltramare processing costs. ' Comp. Caju Beira proposal assumes a small operation with lower processing costs. fData from Public Ledger's Commodity Week provided by Raghu Kumar. ' = kemel MT / raw nut MT Source: Computations by the World Bank mission. 318. On the basis of this analysis, we may conclude that the optimal form of assistance to cashew processing would be the provision of informnation, extension, and help with access to credit at market terms, rather than protectionist policies. During 1996 the International Development Association presented a proposal to the Ministry of Commerce for a plan of recovery for the cashew processing industry which would provide matching grants for technical assistance to improve the firms' management, technology, and access to markets. The proposal provides also for the issuance of loans, denominated in dollars, for purposes of rehabilitation.138 The government is presently conducting a study which will assist the further development of its policy in the cashew subsector. (c) Colton 319. A historical view of seed cotton production in Mozambique is given in Figure 26. Cotton was first produced in significant volumes in the early thirties, and until the mid- 1 960s practically all of it was produced by smallholders under outgrower arrangements. The large- scale (orprivado) sector became important in the 1960s and 1970s, and total production peaked at 144,000 tons in 1973. Output declined sharply with the war of independence. With the nationalization policy of the new government the decline continued. Total seed cotton production fell to its all-time low of 5,000 tons in 1985. Since then there has been some improvement, as certain of the outgrower arrangements have been rehabilitated and new private sector operators have arisen. Total output in 1994 was 46,000 tons. Clearly there is enormous scope for expansion simply by reattaining the previous levels of output. There are large areas in, for example, the Montepuez area of Cabo Delgado, where cotton used to be cultivated during the colonial period and which are now lying fallow. Bush encroachment is not serious. Most of these areas could readily be reestablished if the road network is rehabilitated. 3 Details are given in correspondence between the Ministry and Mr. A. Sowa, World Bank. 110 160000 1 40000 120000 o 2 80000 5 60000 0 40000 20000 Figure 26. Seed cotton production, by farm type, 1955-94 139 320. Smallholders under outgrowing arrangements continue to dominate cotton production in Mozambique, as is shown in Table 29. Financial calculations suggest that the profitability of outgrower arrangements is greater than that of "estate" farning (or "direct" farming, as it is called) by the ginneries. Yields in the smallholder/outgrower sector are very low, at about 400 kg/ha, because insecticide applications are sparing and infrequent. In addition, cultural practices are not well advanced and no fertilizer is used. Table 29. Seed cotton: area sown, yields and output, by farm type, 1993 a~ Farm type' Area sown Yield Output (tons) (ha) (tons/ha) Estate sections of Jvcs 8,000 1.6 13,000 Medium- & large-scale 7,000 1.2 7,500 Outgrowers a 63,000 0.4 26,000 Total 78,000 0.5 47,000 a Practically all outgrowers are smallholders. Source: Instituto de Algodao de Mocamnbique (1993), and Moll (1993), Table 2. See the Appendix Tables for further detail. Monopsonistic marketing arrangements 321. Several firms run large cotton outgrower systems as joint venture arrangements with the government. The largest of these are the multinationals Lonrho, Entreposto and Joao Ferreira 3 Data from Fok (1995), Annex 7, in turn from the Instituto de Algodao de Moqambique. 111 dos Santos.140 Smallholders associated with the largest joint venture, that of Joao Ferreira dos Santos, sow some 35,000 ha of cotton annually. 14 All three joint venture companies (JVcs) have exclusive rights to purchase cotton from smallholders in their so-called dreas de influencia, in return for which they are obliged to provide seeds, pesticides and extension services. The monopsony does not apply to other crops. The ginning prices, output purchase prices and input sale prices are fixed by the government. The dreas de influ6ncia are large: that of Lonrho is alternatively 400,000 ha (according to the managing director) or 1,400,000 ha (Myers and West, 1992). During 1995 the monopsony was restricted to smallholder sale of cotton, and farmers with more than 20 ha under cotton, even within the areas de influencia, were henceforth permitted to make whatever credit and marketing arrangements they saw fit. 322. The standard justification for the cotton marketing monopsonies is that due to the social structure and the land tenure arrangements, land cannot be used as collateral for loans of insecticide and improved seed. A cotton marketing monopsony solves the problem because the smallholders are docked, upon sale of the output, for all credit received. Under free markets, by contrast, smallholders face the temptation of opportunistic action at harvest time, namely to sell their cotton to traders and to shirk their repayment obligations to the provider of in-kind credit and extension. The monopsonies are controversial. Some observers (e.g. Jaffee, 1993) have urged that they be scrapped and that competitive trade be permitted. Using standard monopoly models, it is argued that it is in the interest of the cotton monopsonies to reduce the price paid to the outgrowers, raise the price of the inputs sold to them, and reduce the quality of service provided to them. Others (e.g. Moll 1993; Fok 1995) argue that marketing monopsonies in the smallholder areas should remain until such time as viable alternatives can be found to guarantee the continued existence of in-kind credit arrangements. 323. The monopsonies are a restraint on trade. The smallholders were never given the option of voting for their installation. Furthermore, Government pricing interventions, enforced through the Instituto de Algodao de Mocambique, are ineffective because the government lacks the information and skills necessary to conduct intricate negotiations with the JVCs over price- setting. Hence it is necessary to consider the alternatives. 324. One alternative to a monopsony arrangement is the development of peasant-run cooperatives. Cooperatives were an important engine of growth in agriculture in New Zealand, Australia, South Africa and Germany. Arguably a cooperative run by producers would have better information about the activities of its members than would a bank, and would thus overcome the problem of asymmetric information that is the prime justification for a marketing monopsony. Furthermore, a cooperative may enjoy cheap and efficient enforcement powers arising from personal contacts and the potential loss of reputation upon non-repayment. A third advantage of a cooperative is that any profits arising can be pumped back into capital goods for local communities -meeting halls, schools, etc. - which can help build loyalty and assure high rates of loan repayment. 325. The way forward is indicated by Francophone African cotton producers. In Mali, for instance, local cooperatives called Associations Villageoises were started in the mid-1970s (Fok 1995, Annex 32). They gradually took over various marketing functions, including responsibility for the distribution of credit for inputs. The mother company's marketing agents were retrained as literacy teachers so that the smallholders could acquire the minimum of accounting and literacy 140 For a complete listing of the Jvcs, see Fok (1995), Annex 8. 141 See the Appendix Tables for detail on areas sown, yields and output. 112 skills necessary for running a successful cooperative. Currently 80 percent of the cotton output of Mali is marketed by village associations. 326. The management and literacy requirements of cooperatives are high, however, and being vulnerable to manipulation by unscrupulous individuals and government officials, there are not many examples of cooperatives successfully providing substantial amounts of credit to smallholders in low-income Africa. A large system of cooperatives undertaking marketing and credit presupposes a long period of education and training, combined with the gradual accumulation of savings at the village level so that at least some significant part of the credit granted originates in savings by local people - for this would provide a powerful incentive for ensuring loan repayment. The Mozambican peasantry, in terms of education and savings, is well below this level. Furthermore, the unfortunate experience of cooperatives during the socialist period has made Mozambican smallholders suspicious of attempts to mobilize them. Hence this process of development is likely to take several years. If a start were soon made with expanding literacy, financial and administrative skills, cooperatives could in time replace some of the functions of the JVCs and the monopsony charters would gradually become redundant. 327. It is likely that if the monopsonies enjoyed by the JVCs were suddenly abolished, the cotton market in their areas of influence would collapse. Care should thus be used when bringing about changes. But an important qualification should be stressed: the only justification for the monopsony arrangement is that smallholders are better served than they would be under free-trade arrangements. Yet it is apparent that the JVCs are presently incapable of serving all smallholders in their areas of influence. The government regulators should therefore be charged with the responsibility of redrawing the boundaries of the areas of influence on a regular basis. Villages not being served fully by the JvCs should be excluded from the areas of influence and smallholders be permitted to sell their cotton to traders at will. 328. In the meantime the government, international donors and NGOs should work on alternative means of providing cotton-related services to smallholders outside the JvC areas of influence. The alternatives could consist of some combination of cooperatives and decentralized arrangements with sitting traders. This congeries of efforts should be viewed as a pilot scheme. If the pilot proves effective, the areas of influence of the Jvcs could be reduced gradually until such time as the marketing monopsonies can be abolished altogether. 329. Another direction for the future is given by the JVC Agri-Buzi, in the province of Sofala. It operates a cotton outgrower system but without a monopsonistic charter. The company is a joint venture between a South African company and the parastatal Companhia Industrial de Buzi. AGRI-BUZI provides seeds and insecticides on credit to medium-scale farmers (25-75 ha) or "privados", and gives them guarantees enabling them to obtain bank credit for the hire of harvesting labor (Fok, 1995, p. 33 & Annex 14). The firm does the ploughing and planting with tractors. It applies the insecticide treatments (about 8) by aeroplane, which is why it is more convenient to take on farmers with at least 25 hectares. The farmer's chief responsibilities are weeding and harvesting. The firm has no difficulty recouping its investments because since the number of farmers is small, the size of the commitments they have made is large, and the firm has a great deal of information about the progress of their crops. Deliberate reneging on a contract by the farmer would be extremely costly because she would be excluded from such contracts in the future. Control of the cotton price 330. As mentioned above, a minimum cotton price is set by the government each year. There are several objections to this process of regulation. First, the minimum cotton price does not 113 affect the farmers' incentive to produce because it is usually announced too late in the season, beyond the planning and sowing time (October to December). Second, the method of price regulation has penalized smaliholders because the price is set in nominal terms and does not take sufficiently into account the rise in inflation during the subsequent months. As a result the JVCs have often ended up with an unexpected bonanza at harvest time (Moll, 1993, p. 27). AGRI-BUZI, by contrast, protects its "privado" farmers against inflation by setting its cotton purchase price in US dollar terms, and applying the exchange rate as of the delivery date of the raw cotton (Fok 1995, p. 34; & Annex 14). It is recommended that the IAM and the JVCs move to some similar approach. 331. Third, and most importantly, the minimum price fails as a means of regulation of the cotton monopsonies, which would perhaps be its strongestjustification. Regulation is inefficient and the Mozambican government lacks the capacity to carry it out effectively. The political influence of the multinationals and the skills at their command likely tilt the regulatory system in their favor at the expense of the smallholders. The method for establishing the farmgate price is the residual one: the costs of the ginneries and JVCs are subtracted from the international price and the remainder goes to the smallholder. But price-setting on costs gives no incentive for cost reduction, and it does give incentives for opportunistic overstatement of costs. The regulators would need a high level of skill and rich sources of information in order to set prices which mimic efficient free- market prices. These skills and information conspicuous by their absence. Hence the control of the cotton price should be abandoned. 332. Even an uncontrolled monopsonistic ginnery has an interest in providing credit to smallholders in order to obtain supplies efficiently. Unsatisfactory though the monopsony be, the ginnery provides smallholders with one additional choice, namely to grow cotton, which can only improve their welfare compared with the alternative of having no such choice. Furthermore, smallholders typically find it easy to change from one crop to another; in Tanzania and Uganda, the measured elasticities of substitution are high. The best way of protecting the smallholders from abuse by the Jvcs would be the formation of associations or cooperatives which would enable them to represent their views and negotiate for improvements. Such associations already exist in the Montepuez area (Cabo Delgado) where LOMACO has a ginnery. These organizations could be strengthened and the concept exported to the other cotton areas. These efforts would be assisted if the government published a non-binding "reference price" or "indicative price" based on a suitable formula reflecting the international cotton fiber price. Active promotion of productivity in cotton 333. Government intervention via the IAM (Cotton Institute of Mozambique) is in most respects counterproductive, with administrative control of prices and markets dominating more productive investment and development activities. The government should adopt policies of active promotion of productivity in cotton. The government would be justified in requiring the JVCs to participate actively in certain of these areas of development because this is the quid pro quo for their monopsonistic charters. Especially noteworthy are the following areas of development: * Research. In the domain of research and innovation there are externalities that deprive private firms of the motivation to invest. The state could play a role by encouraging the JVCs to develop new seed varieties. The low ginning outturn of 34 percent in Mozambique (compared with 40 percent in Mali) is an indication of the need for improved varieties. * Animal traction and tse-tse fy control. It would probably be feasible to introduce animal traction in the cotton-producing areas of the north. Animal traction is widely used for weeding 114 in smallholder cotton in West Africa. A precondition would be a successful and low-cost tse- tse fly control program in the relevant localities.142 Reasonably successful programs of tse-tse control have been run in Central African Republic, Uganda and Cameroon.143 New techniques of eradication and control have been used in Kenya and Zimbabwe. A World Bank-funded program of trypanosomiasis eradication in Uganda using odor baits, targets and traps will cost some US $4.5 million to clear 9,000 square kilometers over a period of 10 years, or about $5 per hectare spread over 10 years. The program will, it is claimed, earn a return on investment of 36 percent (World Bank, 1990c, p. 13) . At so low a level of cost per unit area, it is not unrealistic to envisage a major tse-tse control program in Mozambique, working outwards from the tse-tse-free southern provinces and the highlands, and spreading gradually to some of the major areas of agricultural activity. * Improvement of the extension system. The quality of cotton produced by smallholders in Mozambique is low. Less than I percent of smallholder cotton is of quality "A" or better; at least 40 percent of the estate-produced cotton of three JVCs achieves this level (Fok 1995, p. 67). Training in techniques will be one of the chief means of improving the smallholders' performance. * Developing local associations of smailholders. A long-term objective should be to permit voluntary associations of smallholders to develop. Part of government educational policy could be focused on providing literacy training for adults in the cotton-growing areas, because the benefits from this would have important externalities for community development and, eventually, for the cotton industry. The ultimate objective is that the smallholder associations would take over some of the functions performed by the JVCs, such as credit and marketing, so that the JVCs could concentrate on ginning and research. (d) Citrus 334. Production of citrus was some 9,500 tons in 1994145. Some 80 percent of Mozambique's citrus exports are grapefruit, a product in which the country has a clear comparative advantage. World grapefruit exports are increasing at about 3 percent per year (Aube 1995, p. 5). In 1970, at its peak, the industry produced 769,000 cartons (about 12,000 tons) of grapefruit, 21 percent of Southern African citrus exports. Currently, however, little of the sector's productive capacity is actually in use and total output is stagnating. The current contribution of Mozambique to total citrus exports from Southern Africa is a minimal I percent. Among the reasons for the poor performance of the subsector is that three of the chief producers A three-year EEC project began in 1993 in Sofala and Tete provinces to control tse-tse. It is part of a regional program including Zimbabwe, Malawi and Zambia (AUSTRAL, 1993, p. 70). Note that the AUSTRAL report does not commit itself on the likely costs and benefits of a tse-tse control program. 143 Personal communication from Mr. Comelis de Haan, World Bank. The costs of (once-off and lasting) tse-tse eradication by ground spraying in Zimbabwe and Botswana in the 1980s were about S170-$235 (ILCA (1988), p. 332). The costs of aerial spraying were higher. The costs of tse-tse eradication by irnpregnated traps and screens in three African countries varied from $39 to $345 (op. cit., p. 333). See also Chadenga (n.d.) 145 Calculations based on data in Aube (1995), p. 14f. Assuming realistically that Citrinos de Manica produced 4,000 tons in 1994. 115 are state farmns which have not yet been privatized (Citrinos de Manica, Citrinos de Timangene and Citrinos de Mo,ambique). Citrinos de Timangene is almost completely inactive at present. Completion of the privatization process is urgent to avoid further deterioration of the estates, especially in the case of Citrinos de Mo,ambique. Substantial capacity for expansion exists on the basis of the holdings of the three state farms, even without developing new land. 335. The marketing of citrus in Mozambique is dominated by Outspan, a South African marketing company with an extensive distribution network in Europe and elsewhere. Outspan has recently acquired a proprietary section of the Maputo port which will be physically sealed off from the rest of the port for the sake of security. This will provide more than enough capacity for Mozambique's exports for years to come. Outspan sends technical advisors to its Mozambican member (as it does to its South African members) who advise on production and packing. The firmn also organizes training sessions and provides a complete set of technical protocols. 336. Other players in the international marketing business are Oceanic and Unifruco, both of which have well-established distribution chains in Europe. Oceanic is a German group based in Hamburg which operates in Southern Africa from its facilities in Cape Town. It possesses its own vessels. Zimbabwean producers sell their produce either through Outspan or through Oceanic. They are said to appreciate having some maneuvering margin vis-a-vis export organizations (Aub- '995, p. 42). Unifruco is a South African firm which possesses a statutory monopoly on the saies of deciduous fruit from that country. 337. The Outspan distribution chain is reasonably effective, though probably not as cheap or efficient as it would be if it were stripped of the monopoly powers granted to it in South African law. Also there is a noticeable lack of transparency concerning some of Outspan's cost elements, such as sea freight, which enter into the calculations of the costs of its services (Aube 1995, p. 27). 338. It is risky for Mozambique to be locked into the distribution chain of a single firm, for if the firm's market were challenged - perhaps by a competitor if Outspan's statutory monopoly were withdrawn - Mozambican producers could be left high and dry. Nevertheless, since Mozambican output is small, there is currently no justification for Mozambican players to attempt independent entry into these difficult foreign markets. 339. The following are some suggestions for the long-term development of the sector. * The privatization of the three state producers should be completed without delay so as to prevent further deterioration of the orchards. a An experiment is underway in South Africa to create an outgrower scheme in citrus. Performance has thus far been mixed but progress should be monitored to see whether a similar demonstration project should be undertaken in Mozambique. Another model may be given by the new enterprise Agrofarm which has 27 ha under cultivation (Aube 1995, p. 16). * The subsector would benefit from some form of economic observatory of the citrus and other fruit and vegetables markets. This activity would best be undertaken by a private association of producers. But the donors or the government could consider contributing a once-off subsidy to get it started, in return for the cooperation of the large producers in setting up an outgrower experiment. * The Beira port system could readily be improved, given effective management. Costs of transit through the Beira port are substantially higher than those through the Cape Town port (Z$4 per carton vs. Z$6.8-Z$7.5 according to Aube (1995), p. 45), despite South Africa's 116 higher labor costs. Furthermore, a severe security problem persists at Beira. Shipping operators would readily use refrigerated containers to compensate for the port's limited facilities, but are fearful of doing so due to theft. The port also has long delays which discourages vessels from calling. If the Beira port could be improved, Mozambique would have a second international marketer available in the form of Oceanic which moves Zimbabwean citrus through the Cape Town port. This would entail multiple synergetic effects on citrus from Manica. * Labor intensive vegetables such as French beans could be developed in outgrower schemes by large export companies (Aube 1995, p. 10). (e) Other crops 340. Seventy percent of the Mozambique coconut/copra industry is concentrated along a 50 kilometer-wide coastal strip in Zambezia province. The total are under coconut palms, as of 1992, was about 100,000 hectares, of which smallholders controlled some 60 percent. Although the smallholder copra is of low quality, it finds a ready market in South Africa where it is crushed and the crude oil is expelled and refined. The major industrial processor, Boror, virually collapsed in the late 1980s. Production is recovering now under private auspices. The firm MADAL has rehabilitated one of the processing plants and, in addition to running its own plantations, purchases some of its raw coconut from smallholders in the vicinity of its Zambezia plantations. Significant expansion of the subsector could be achieved through smallholder outgrower systems. Demand exists for fresh coconuts in South Africa, and the export of low- grade coconuts for crushing can be expanded. 341. Tea used to be produced in estate fashion by three tea companies in Gurue district, Zambezia province. At its height in the 1980s, some 17,000 tons of tea were produced annually. During the war, 13 of the 18 processing plants were destroyed, and production ceased altogether. One of the companies, EMOCHA, previously state-owned, is currently being privatized. At its peak it cultivated some 14,000 hectares of tea. There is no tradition of outgrower arrangements in tea production, but to judge by the experience in Zimbabwe and Kenya, ample opportunities exist for this form of development. 342. Before the war, some coffee used to be grown in small plantations (20 hectares) by private farmers, in the Gurue and Milange districts of Zambezia province. There is no tradition of smallholder production of coffee, though this could be encouraged through appropriate extension arrangements. 343. Tobacco is produced in Nampula, Cabo Delgado and Manica provinces. Smallholders specialize in burley tobacco, and large-scale farmers in Virginia tobacco which requires more precise management. Both Joao Ferreira dos Santos and LOMACO have outgrower arrangements with smallholders producing tobacco. 344. Mozambique used to export bananas to South Africa before independence, an activity which it never resumed even though demand has increased and South Africa now imports from the Caribbean. Mozambique also exported bananas to Portugal. It exported litchis, pineapples and mangoes to Europe before independence but this has not been restarted (Aube 1995, p. 8). An export potential also exists in papayas, French beans, asparagus and mushrooms. 117 345. Sugar. The production areas for sugar are in the valleys of the rivers Zambeze (Zambezia province), Pungoe (Sofala province), and Incomati (Maputo province). There are six sugar estates, of which only two are currently producing. One of the productive estates and one of the non-productive ones are still owned by the state; the remainder are in private hands. The total output of sugar in 1995/6 was 29,000 tons. 346. Currently there are no outgrower arrangements in the sugar subsector. In the past, some ten to twenty percent of Mozambique's sugar was grown by smallholders, usually without any credit or extension assistance from the large-scale producers. Outgrower projects have been successfully executed by large sugar producers and peasant households in Natal/KwaZulu in South Africa. The standard means of sugar production in the northeast of Brazil is through outgrower systems. Hence the capacity for outgrower arrangements is considerable. 118 Chapter 6. A STRATEGY FOR AGRICULTURAL DEVELOPMENT IN MOZAMBIQUE A. Introduction 347. This section summarizes the conclusions of the report and presents a strategy for government action in support of broad-based agricultural development. B. Previous Sectoral Reviews 348. The World Bank issued an Agricultural Sector Survey in 1988. It pointed out that despite generally favorable conditions for the agricultural sector, its performance had deteriorated sharply in recent years. As a result, food imports had increased rapidly while agricultural exports had fallen. Among the causes were climatic disasters and inappropriate economic policies. The government had artificially reduced agricultural prices, causing extreme shortages and illegal parallel markets in which prices far exceeded the official levels. 349. The report recommended that the state divest itself of its farming properties, which were much less efficient producers than the smallholder and large-scale private sectors, but stressed that "there is a critical role for Government in the provision of agricultural services such as research, extension, credit, education and training, and basic marketing infrastructure such as rural roads" (p. 80). The report noted that the government intended to promote the recovery of the agricultural sector by means of exchange rate and trade reforms, pricing policies, and fiscal and credit policies. The report also concluded that Mozambique would continue to need food aid to avoid mass starvation. 350. Between 1988 and 1996, the government made important strides in the directions recommended by the World Bank. It privatized practically all of its farming operations and all of its agro-processing establishments; declared itself in favor of smallholder development as the main engine of growth; instituted far-reaching exchange rate and trade reforms; replaced the state marketing agency AGRICOM with a much reduced Instituto de Cereais de Mocambique, and, while continuing to set "minimum prices", ceased to defend prices except in cotton. Its major failing, in terms of the recommendations made by the World Bank in 1988, was in the fiscal and credit areas: since government spending was not curtailed and the monetary supply was permitted to expand rapidly, inflation rates rose to 60 percent in 1994/5. Real interest rates for large firms were slightly negative, but effective real interest rates for small firms, who were required to pay loan interest in advance, were extremely high (over 100 percent in 1995). 351. Notably absent from the 1988 Survey was any serious treatment of the issue of land tenure. The country was embroiled in a civil war, mass starvation was threatening, and the omnipresence of the state in the production and marketing systems overshadowed concerns about land tenure. 119 C. Government Policy Statements 352. In 1995 the Mozambican government issued a document entitled "Agricultural Policy and Strategy for Implementation". It identified four main objectives for economic policy (Republica de Mocambique, 1995, p. 11, author's translation): "(a) food security; (b) sustainable economic growth; (c) reduction of unemployment rates; and (d) reduction of poverty levels." 353. From these overall economic objectives stemmed the government's objectives for agricultural development (loc. cit.): "The transformation of subsistence agriculture into an agriculture increasingly integrated into the functions of production, distribution and processing, such as to lead to: - the development of a self-sufficient family sector which contributes a surplus to the market; - the development of an efficient commercial sector which participates in rural development." 354. In order to attain these agricultural objectives, four fundamental principles were to be followed (Republica de Mocambique, 1995, p. 12): "(i) Sustainable use of natural resources ... based on the education and involvement of the rural population in the management and sustainable use of natural resources ... "(ii) The expansion of the capacity of production and agricultural productivity, based on the development of the family sector and the promotion of private and public investment. The development of agriculture and the achievement of food self- sufficiency depends fundamentally on the full involvement of the rural population in the search for local solutions which can result in an increase in production capacity and productivity and above all, the creation of self- employment. "(iii) Balanced institutional development, including the development of human resources, as well as training and incentives. In this context, the Agricultural Policy promotes the rehabilitation and development of agricultural schools ... and training centers for farmers which can increase self-employment and the establishment of centers for the demonstration of good agricultural practices in the rural context. "(iv) The recognition of the fundamental role of women in agricultural activity and in particular in integrated rural development. Women exercise a fundamental role in education and extension, and as a direct agent in development. It is in this context that the Agricultural Policy prioritizes the participation of women in programs of professional training, rural extension and specific projects in the activities of production and marketing." 120 355. The objectives and the four principles are strongly supported by the analysis contained in this document. At several points, however, the Government's statement fails to give an indication of the principle used for distinguishing between government actions and the actions of the private sector. Actions are recommended without specifying who will accomplish them. The well-known concepts of externalities and of public goods are not used to make these distinctions. 356. The government document strongly emphasizes extension, which is named as "the principal operational instrument of the Agricultural Policy" (p. 28). This is a strongly positive element in the document because extension services are a public good and have been shown in other contexts to have a high rate of return. Even here, however, the issue of a mixed public and private role in the provision of the service is not mentioned. 357. In what follows, the government's own statements of objectives will be used as a point of departure for developing an agricultural strategy, and an attempt will be made to use standard economic concepts in order to more clearly discern the specific role of government in an agricultural development strategy. These standard concepts include externalities, nonexcludability, market failure and public goods.'46 D. Current status of Mozambican agriculture 358. Agricultural output in Mozambique is increasin. FAO estimates put the 1996 grain harvest at 1.4 million tons, the highest in two decades. Agricultural marketing increased by an estimated 18 percent between 1995 and 1996. The good rains of 1995/6 helped, but there were other factors as well: the return of exiles and the return of deslocados to their farms, the sustained peace, the improvement in the trunk and secondary roads, and the increase in the price of maize at the farm gate owing to the substantial involvement of small traders in marketing. Significant increases in cropped area have been reported in some parts.148 359. On the policy side, Mozambique has made considerable progress towards the market orientation that the government has espoused since 1992. Although inflation rates fluctuated between 33 and 63 percent from 1990 to 1995, the effects of monetary tightening were felt in 1996, and the inflation rate dropped to 17 percent by the end of the year. Borrowers' interest rates were controlled until 1994, and they tended to follow the inflation rate, so that the real interest rate varied between -8 percent and 3 percent between 1990 and 1994. Real interest rates rose with the liberalization, and for small borrowers, owing to the greater risk involved, were sometimes over 100 percent in real terms. Real interest rates declined during 1996, and prime borrowers faced (nominal) interest rates of 25 percent. The official exchange rate was heavily overvalued in the early 1990s, but the freeing of currency movements and successive devaluations resulted by the mid-1990s in a negligible official-unofficial rate differential. Taking into account the recent improvements on the inflation front and the real depreciation of the exchange rate, the macroeconomic scene is set for further growth in the agriculture sector. 146 See paragraph 431 for further explanation of these concepts. 147 From: World Bank (1996b). 148 For instance, in the Monapo district of the northern province of Nampula, the area sown by the average household increased by some 80 percent between 1992/3 and 1994/5. See footnote 199 for details. No aggregate data are yet available on areas sown and yields for the 1995/6 season. 121 360. As far as product markets are concerned, the state-controlled system with fixed prices run by AGRICOM was dismantled, and AGRICOM was replaced by the ICM (Cereals Institute of Mozambique). While the ICM has a broad mandate of agricultural price stabilization, it has a very limited budget and is functioning as a regular wholesaler. The government sets minimum prices annually, often after the growing season has already started. Given the fact that inflation was high and unpredictable during the early 1 990s, these prices rarely corresponded to the observed market prices, but the government lacks the funds to defend the floor prices. 361. There remain certain constraints in the input markets. There is a minimum wage but it is rarely enforced in the agricultural sector and it is infrequently adjusted, so that its effects on factor choices are negligible. Capital markets are problematic for many farmers on account of the high effective real interest rates they are required to pay. Furthermore, land cannot be used as collateral because the land law forbids the sale of land, and there is insecurity for many smallholders in the face of land titling activity by outside agencies, firms and farmers. 362. The government has divested itself of practically all of its state farms and agro- processing facilities, the most notable of which was the cashew processor Caju de Mocambique. 363. Import tariffs, according to the new regime of November 1, 1996, are low, between zero and 35 percent. Export tariffs are zero but for the case of raw cashew nuts where the tariff was reduced from 20 percent in the 1995/6 season to 14 percent in the 1996/7 season. There is a sales tax (imposto de circulaado) of between 5 and 21 percent which impedes trade to the extent to which it is enforced. A more serious problem in the trading area, however, is the difficulty of entry into wholesaling and retailing, owing to the stringent legal requirements and to some misapplication of the rules by local officials. 364. Overall, taking into account the macroeconomic environment, the output and input markets and the marketing, trade and tax regimes, Mozambique has in recent years made substantial progress towards bringing about market-based growth. Its agricultural sector is well placed for expansion. E. Major Findings of the Memorandum 365. This memorandum has reviewed Mozambique's agricultural performance in some detail. Institutional and economic factors were analyzed in order to explain the successes and failures of past policy. Some key conclusions, which form the basis for subsequent recommendations on sectoral development policy, are set forth below. (a) Raising smallholder agricultural production is an effective way of addressing poverty 366. Poverty may be defined by reference to a poverty line, the level of which is necessarily arbitrary. A line frequently used in the African context is $1 equivalent per capita family income per day. The (unenforced) minimum wage in Mozambique is about $0.50/day; few rural employees receive wages higher than this. By the standard of $1lcapita/day, the vast majority of Mozambicans, rural and urban, would be classified as "poor". 367. An alternative definition of poverty is the lack of a specified combination of services which our society has come to consider as basic necessities. This basket of services often includes access to clean water, access to primary education, access to medical care, and access to 122 enough food calories for health and comfort (which in tum presumes a certain level of income). By any reasonable standard incorporating arbitrarily defined levels of the latter variables, the vast majority of Mozambicans are poor. Roughly three-quarters of the population lacks access to safe water. 49 42 percent of men and 77 percent of women (rural and urban) are illiterate. Chronic malnutrition affects 27 percent of children under the a e of five.'50 About 60 percent of the population, rural and urban, lacks access to health services. 51 Infant mortality (146 per thousand live births) and under five mortality (282 per thousand live births)152 are much higher than the sub-Saharan African averages of 99 and 172. Maternal mortality (1,100 per hundred thousand live births)'53 is one-third higher than the average for sub-Saharan Africa. Reflecting the lack of access - urban and rural - to medical care, the average life expectancy is 46 1 54 years. 368. The analysis has demonstrated that raising the farming output of smallholders is an effective way of reducing poverty. The argument is developed in three stages: smallholders constitute the vast majority of Mozambican society; smallholders are among the poorest people in the country; and raising smallholder production would reduce poverty directly, by raising incomes, and indirectly, by reducing the isolation of smallholder families. 369. As of 1993, 69 percent of the Mozambican population lived in rural areas and small rural villages.'55 This figure has probably risen since to about 80 percent, as refugees and deslocados returned home. Of these rural people, about 90 percent have their livelihood in smallholder agriculture.'56 About 72 percent of all Mozambicans are directly dependent upon smallholder agriculture. 370. Smallholders are among the poorest in Mozambican society. While many urban people, particularly the deslocados, are poor, the effects of deprivation, isolation and vulnerability in rural areas are generally worse. Rural people are vulnerable to climatic uncertainty, and while all occupations, urban and rural, are subject to uncertainty (of demand, employment, prices, etc.), rural people have less scope for risk-reducing diversification. The average rural household experiences "food insecurity" for 3.7 months in the year.'57 As of 1994, 77 percent158 of smallholders had no off-farm income, and 71 percent'59 had no marketable surplus. Rural people 49 Multiple Indicator Cluster Survey, Table 8, p. 25. 150 In the sense of having a weight-for-age Z-score less than -2.0 (Multiple Indicator Cluster Survey (1996), quoted in Desai (1996), Table XV). No rural/urban breakdown is available. Data refer to 1995. 151 World Bank (1996a), p.2. 152 World Development Indicators, 1995, Table 27. No ruralurban breakdown is available. Data refer to 1993. 153 World Bank (1996a), p. 2. 54 World Development Indicators, 1995, Table 1. No rural/urban breakdown is available. Data refer to 1993. 155 World Development Indicators, 1995, Table 31. Data refer to 1993. 156 Desai (1996) reports that about 5 percent of rural households do not have "cultivated land" (2.5 percent in the North, 1.7 percent in the Central region, and 8.0 percent in the South) (p. 20). He also reports (Table XI), that 91.5 percent of rural households have some income from agriculture, fishing or livestock. 157 Multiple Indicator Cluster Survey, (1996), quoted in Desai (1996), Table XIV. "Food insecurity" is here a subjective judgment of the respondent. 158 Ministerio da Agricultura e Pescas (1994), as quoted in Desai (1996), Table 6, p. 22. 59 Ministerio da Agricultura e Pescas (1994), as quoted by Desai (1996), Appendix Table IX. 123 have less access to medical services. In urban areas, there were, as of 1992, 1.83 hospital beds per 1000 people in urban areas, and 0.48 in rural areas.160 In Maputo City there were about 20 patients for each medical staff member employed, whereas in the remainder of Maputo province there were some 43. In other largely rural provinces, the patient/medical staff ratios were also high - Zambezia (50), Nampula (42), Cabo Delgado (39). 161 In rural areas, between 27 and 58 percent of people are served with safe water, whereas in urban areas the percentages are 81 to 97.162 Most rural people have to get water far away (32 percent over 100 meters away, 28 percent over 500m) , the burden of which falls upon the women of the household. In some rural areas, as many as 87 percent of men are illiterate.'64 Among children between six and ten years of age in rural areas, only 45 percent attend school.165 371. Raising smallholder agricultural production would reduce some aspects of poverty directly. Their incomes could be raised by increased yields, which in turn could be assured by improved agronomic research and dissemination. If smallholders faced higher output prices due to the reduction of export taxes or due to more efficient marketing (arising from better roads) their incomes would rise immediately; the stronger incentives would induce a supply response and incomes would rise further in subsequent years. Demand for primary education and elementary medical services would rise, and this in turn would tend to reduce the infant mortality rate and raise the life expectancy. With their basic food needs covered, smallholders would be better able to smooth their income inter-seasonally and inter-annually, resulting in less food insecurity. 372. The government's infrastructural interventions in aid of agricultural production would also reduce poverty. Rehabilitating the feeder roads and installing culverts, for example, would lower transport costs and improve people's access to medical services and education. Low-cost transport services would also result in an increase in rural demand for urban-produced commodities, resulting in improved incentives for increasing agricultural output. The government's investments in primary education would facilitate technological transfer, which would raise agricultural output and incomes in the medium term. They would also increase the access of rural dwellers to nonfarm earning opportunities, and increase their returns from such work. 166 373. Thus a strategy aimed at smallholders would simultaneously address the largest group of those in poverty and those of the poor whose situation is the most precarious. Fortunately an anti-poverty strategy aimed at smallholders is not hard to imagine; it arises naturally because smallholders are the most efficient farmers. 160 Ministerio da Saude (1994), p. 33, Table 3.1.2. 161 Ministerio da Sauide (1994), p. 33, Table 3.2.5, p. 55. Hospital staff refer to the sum of doctors, nurses and technical staff. 162 Multiple Indicator Cluster Survey, Table 8, p. 25. 163 Multiple Indicator Cluster Survey (1996), quoted in Desai (1996), Table XVI. 164 Borges and Doto (1995, p. 29ff). 165 Multiple Indicator Cluster Survey(1996), quoted in Desai (1996), Table XIII. Data refer to 1994. 166 This was shown for Tanzania by Collier et al. (1990). A person with no education had a 5 percent chance of obtaining nonfarm employment; literacy raised the chance to 11 percent; lower primary to 17 percent, upper primary to 31 percent, and secondary schooling to 74 percent. This is at the margin, of course; at the aggregate level an increase in schooling would lower these probabilities but total nonfarm employment would expand in turn. 124 (b) Smallholders are the most efficientfarmers 374. On efficiency grounds it is clear that it is in the smallholder sector that the country can find the highest return on its investments; this observation is supported by all measures of comparative advantage which have been obtained. Smallholders are efficient producers of food crops (maize, groundnuts, cassava, rice, sunflower, beans), cotton and cashew, with most domestic resource cost coefficients (DRCs)167 between 0.4 and 0.8, depending on the assumptions about the shadow wage and the output price. 68 In cotton, at the favorable 1996 international price, the smallholder DRC is 0.3, while that of large-scale producers is 0.6. In maize, the smallholder DRC is 0.6-0.8, while the large-scale producer's DRC is 1. I or 1.2. Irrigated sunflower seed production is most efficiently performed by smallholders with a small amount of (hired) machinery. The DRC is an attractive 0.7. Medium- and large-scale farms have significantly higher DRCS of 0.9 and 1.1, due to the higher cost of their mechanized operations. 375. These results are driven by the fact that wages in Mozambique are low, enabling the labor-intensive smallholder system to outcompete large-scale mechanized producers.169 On poverty and equity grounds it is equally clear that a smallholder focus provides the most effective way to promote growth while at the same time ensuring the participation and improvement of the people who most need it. 376. However, the analysis has also found that in many lines of production the large-scale sector, too, is profitable, although it generally does not have domestic resource cost coefficients as low as those of smallholders (see also §414). Therefore large-scale producers have their place in the economic framework, although there would be no rationale on efficiency or on distributional grounds for the government to direct special assistance to them. 377. Another reason for permitting the growth of a large-scale sector is that this is consistent with Mozambique's factor endowment, as one of the most land-rich countries in sub-Saharan Africa. If left to itself, the market would naturally generate a wide array of farm sizes; some smallholders would grow, in time, to medium and large sizes; and there would be new large- scale entrants. 378. It is recommended that the state provide public infrastructure and a conducive policy environment, the latter entailing clear definition of property rights, macroeconomic stability, and 167 The domestic resource cost coefficient (DRC) measures the cost, in terms of local resources, of obtaining one dollar's worth of foreign exchange, after excluding the effects of all policy distortions. A DRC of between zero and unity indicates that the project makes a net contribution to the economy. Given flexible markets, and no policy distortions, the project would be taken up by private agents driven by the profit motive. A DRC greater than unity indicates a project which is not worth its while in terms of its foreign exchange earnings. 168 Note that the DRCS were calculated with a shadow price of foreign exchange midway between the official and the parallel market rates. 169 This situation is not static. If the population growth rate of 3.3 percent p.a. falls, and if GDP growth rates are strong, and if (as is likely) there is a gradual process of rural-urban migration, wages will eventually be forced upwards, resulting in changes of technique (e.g. mechanization of milling and other processing, followed by mechanization of farming operations and an increase in the vehicles used at the farm level). But to judge by the experience of other African countries, e.g. Kenya, the increase in the rural wage may take decades to materialize; so that the relative efficiencies of smallholder and large-scale farming operations are not likely to change in the near future. 125 low tariffs. Such a policy environment would simultaneously encourage the smallholder sector and the large-scale sector, and any mix in between. (c) The private trading system is recovering 379. The private trading system was mostly destroyed by the war and discouraged by the government's panterritorial pricing policies exercised through AGRICOM. This study has shown that, since peace was reestablished in 1992, and since the withdrawal of the government-run marketing monopolies, private trading links have reestablished themselves in most parts of the country. There are, moreover, strong long-distance links between the south of the country and the center, and between parts of the north, though not yet between the south and the north. In 1993, shocks to the maize price in Maputo were not discernible in Chimoio (Manica province), 1200 km away; by 1995, Maputo price shocks were felt in Chimoio within a period of three weeks. 70 The differential between prices in Maputo and Chimoio fell from an average of about 150 Meticais in 1993 to about 50 Meticais in 1995.171 Hence it is likely that links will develop in due course between the north of the country and the south. Although the marketing margins are still high, they have been coming down and will continue to fall as the volume of trading increases. 380. The implication of this finding is that it would not be optimal for the government to mount major funding programs for the re-establishment of traders. '72 The best that such intervention could do would be to accelerate a process which is already under way. Small projects such as training in business operations for traders, or the construction of market arenas and group storage sheds by local communities would be more appropriate. But accelerated trader development could be achieved more advantageously by addressing directly the two main market failures which have constrained the development of the network hitherto: the inadequate road system and macroeconomic instability. The former raises the costs of trade and, particularly in the north, debars trade altogether with many villages. The latter results in high real interest rates which raise the cost of trade, as well as creating an atmosphere of uncertainty which deters traders from borrowing capital and banks from lending to them.173 381 The study also found that there are still many restrictions on the establishment and functioning of trading businesses and that the best avenue for action for the government in the marketing area would be the clarification of the business and legal environment. The study has also pointed out that trade in agricultural goods would be better stimulated through government spending on road maintenance and the construction of culverts than through direct interventions. 170 Donovan (1996), p. 222. 171 Donovan (1996), p. 185. 172 Interest in mounting such projects has been rekindled by the prospect of closure of many rural branches of the Banco Comercial de Mocambique and the Banco Popular de Desenvolvimento in the process of privatization. 173 An additional practical problem is that loan recovery is difficult in the case of small and petty traders. The experiment with the Fundo de Comercializac4o in 1993 went awry because most loans were not recouped. 126 (d) Integration in regional grain markets will dominate long-distance internal (north-south) trade 382. Long-distance trade is expected to play an increasing role in the future, particularly if the government succeeds in improving the efficiency of operation of the ports. It has been shown that, under reasonable assumptions concerning improvements in the legal regime, the marketing system and the ports, Mozambican maize would find its way to South Africa in years when the latter country is in deficit.174 Prior to independence, trading links between the north and the south of Mozambique did exist. Today even bulky products such as cassava sometimes find their way by truck from the north to the south of the country. However, for the northern provinces (Cabo Delgado, the westem parts of Nampula province, Tete) to produce cereals for 175 Maputo is unlikely to be more than a marginal proposition ; they would do better to export to Tanzania, Malawi and Zimbabwe.176 The volume of trade would be greatest with countries to the west, on account of the rail link from Nacala to Malawi. The north of Mozambique could also export to markets further afield, via the ports of Nacala and possibly Pemba, to, for instance, Sudan and Kenya whose growing seasons do not coincide with those of Mozambique. Areas which are less distant, such as the Manica highlands, could produce maize and other products for the Maputo market. 383. In light of these findings about likely future trading patterns, is there any justification for a special import tariff on, sa., cereals, in order to encourage north-south integration and promote self-sufficiency in cereals?M This seems unlikely. The optimal amount of north-south integration is that which would arise in a context where there were no market failures. Currently there is at least one 78 market failure: the inadequate state of the road system. The amount of north-south trade is consequently suboptimal. But the best way to address this problem is through improving the road system, which is already being done through the ROCS 11 program.179 Using an import tariff would be a blunt instrument. While it would increase the quantity of north-south trade, it would impose unnecessary costs in so doing, by raising the cost of maize for consumers and encouraging the cultivation of maize in unsuitable parts of the south. 384. Concerning cereals self-sufficiency, these findings indicate that this objective would be expensive to attain. Due to the elongated north-south structure of the country, and the obvious opportunities for regional trade at both ends, self-sufficiency would require major departures from the patterns that the economy, unfettered, would "naturally" produce. The objective would 174 See the extended discussion in Coulter (1996), Appendix 4. 175 The calculations were done both under current price assumptions and under assumptions of realistic price reductions arising from improved shipping, improved port operation, and reduced marketing margins due to the secular increase in trade. The findings as reported in this paragraph reflect the calculations on the basis of assumptions of long-term price reductions. 176 Koester (1986) showed that maize production in northern Mozambique is negatively correlated with that in southern Tanzania and more or less independent of that of Malawi and Zimbabwe. 177 This argument was invoked during the course of discussions with the government in 1996 concerning the Third Economic Recovery Credit. It is sometimes stated in graphic terms: Why should the north of the country be exporting maize when the south is being forced to import and rely on food aid? 178 Up to 1994 there was another problem: the high cost of north-south sea transport, owing to the monopoly status of the government-owned shipping company. The sea transport market has since been opened and there is now an additional private sector competitor. Prices have correspondingly fallen. 179 See §450 for further detail. 127 probably require a substantially increased tariff, which would entail the costs mentioned above. Self-sufficiency is an unrealistic goal given Mozambique's drought-prone agriculture; with a probability of one, the country would eventually be forced to import. In any case self- sufficiency would not guarantee household food security, to which we now turn. (e) Food security is attainable without continued government-directedfood aid 385. Price incentives and infrastructure are important for the growth of domestic food production, and in turn local food production is important for the attainment of food security. A controversial question, however, is that of direct food aid, through "commercial" food aid, as it has come to be called, and through "emergency" food distributions. Some think that direct food aid is an essential adjunct to market processes, in order to guarantee household food security and provide a response in times of crisis. This report, by contrast, has found that in recent years these direct food aid policies tended to undercut domestic production and hamper the development of a domestic marketing system, because commercial food aid used to be priced at less than import parity, and because emergency food aid leaked (and still leaks) into the market. The consignee price of commercial food aid (maize) stood at 80 percent of the market price in 1993, 30 percent in late 1994 and 50 percent in mid-1995.'80 Since 1996 commercial food aid has been auctioned upon arrival, thereby reducing its effect on the price of cereals. The effect is still negative, however, so that it would be better, from the point of view of local production, to phase it out altogether and leave imports up to private markets. 386. The extent of the disincentive for domestic production has been falling as the quantities of both commercial food aid and emergency food aid decreased markedly between 1993 and 1996. Commercial food aid peaked at 272,000 tons of maize in 1992/3, and fell to 54,000 tons in 1995/6; emergency food aid peaked at 389,000 tons of maize in 1992/3, falling to 40,000 tons in 1995/6.181 A straight-line projection would see no further aid by 1999. Since leakages of emergency food aid into the market are impossible to prevent, this aid should be phased out as soon as humanitarian considerations allow. 387. Long-run food security depends on adequate and stable national production; short-run food security depends on having a contingency plan in place to invoke in times of crisis. Given Mozambique's natural potential for efficient food production, the main requirement for long-run food security is appropriate price incentives and infrastructure. At the national level, development of high potential areas together with the marketing system needed to link producers with consumers will provide the national capacity for satisfying food needs while the increased incomes resulting from the development of export and domestically-oriented agriculture will enable the country to weather the inevitable fluctuations that can be expected in national output. The government can play a role by facilitating research on drought-resistant crops and by training extension officers to exploit this information. At the household level, both increased incomes and improved crop storage technologies will be key elements of a strong food security strategy. 180 From "Chegadas de ajuda alimentar comercial e de emergencia e pregos de milho, Maputo, Margo 1990-Janeiro 1996", by the Sistema de Informagao dos Mercados Agricolas (SIMA), in the Ministry of Agriculture. "8 Data from Ministry of Agriculture/Michigan State University studies, in turn from Ministry of Commerce, Ministry of Agriculture, FAO, World Food Program, and donors. 128 388. For short-run food security, since Mozambique is on the coast and has ready access to international markets, the country can draw on the international grain markets in the case of drastic shortfalls in supply. In order to prevent excessive price spikes, two actions could be taken: (a) advance warning could be given about potentially bad harvests, using prediction based on the El Niiio-Southern Oscillation phenomenon; and (b) the efficiency of the ports and the domestic marketing system (wholesaling, retailing and transport logistics) could be improved, thereby reducing the import-export parity band from a potential $54 to $14 per ton of maize. With better information and efficient port handling, traders would quickly import maize to satisfy local demand, and private markets would be sufficient to prevent price spikes. No government imports would be necessary. The problem of the income loss faced by drought- stricken farmers can be addressed by NGOs and the World Food Program, backed up by donors. Government initiatives in this area would not be cost-effective. NGOs will be more successful if provided with long lead climate forecasts. (t) Local cashew processing can compete at world standards 389. In the past, local cashew processors were protected through a ban on the export of the raw nut, and the entire marketing chain was controlled by a set of minimum prices. Not surprisingly, the Mozambican farm gate price was one of the lowest in the world, falling to 31 percent of export parity in 1992/3 and 20 percent in 1993/4. Since 1994 the marketing chain has gradually been opened to competition. For the 1994/5 season, the export ban was replaced by a tax of 26 percent on the raw nut, falling to 20 percent in 1995/6 and 14 percent in 1996/7. The farm gate price started to improve. Between 1993/4 and 1996/7 the real farm gate price rose 25 percent, and the share of the farm gate price in the export price rose from 21 to 55 percent.82 Using an "infant industry" argument, some of the domestic processors have claimed that they need continued protection, and that they bought the processing plants on the assumption that protection would continue indefinitely. 390. This analysis has shown, on the basis of a variety of data sources, that it is possible for cashew processors to run profitable operations even at a zero rate of protection. This can be achieved in two ways: processors using capital-intensive technologies (e.g. the Oltramare technology) can operate profitably provided they are managed more cost-effectively; or processors can switch to more labor-intensive approaches which are competitive because labor costs are low. Indeed there are already several new entrants into the processing business which are using more labor-intensive approaches than the traditional ones. 391. As mentioned above, the farm gate prices of cashews have improved, in large part owing to the increased competition in cashew trading and the decrease in the export tax. In addition to the rise in petty trade, many new medium-scale traders have entered the business which used to be dominated by a small number of large-scale traders. In addition, several processors have themselves started to trade cashew in search of secure supplies and better quality nuts. The abolition of the export ban contributed to this change in the market landscape by forcing the processors to search harder for nuts of the requisite quality and in the desired amounts. The 182 Data from the Secretaria do Estado do Caju. 129 increased competition will encourage increased quality consciousness on the part of traders and farmers alike - an awareness which the old controlled system stifled - and this will lead in due course to the adoption throughout the market of a multilevel grading system for raw cashews. There is therefore a strong rationale to continue to reduce the export tariff.'83 F. Constraints on Agricultural Growth 392. The constraints on the growth of the agricultural sector of Mozambique are briefly summarized here. They refer mainly to the scarcity of water; landmines; the deficient property rights regime; the inadequate transport and communications infrastructure; low levels of skill and technology; asset poverty; the lack of "social capital" and the absence of rural financial markets; distortions in price incentives due to minimum prices and food aid; and the lack of capacity of the Government apparatus. 393. Scarcity of water. Mozambique faces extreme water stress. Annual rainfall in the provinces of Gaza and Inhambane is less than 600 mm, and droughts occur periodically. Irrigation is required to reach significant levels of sustainable production. The main rivers are running very low owing to use upstream by South Africa and Zimbabwe. There are abundant groundwater resources underlying most of Mozambique's coastal plains, which would allow for significant small-scale (village and farm level) irrigation development. Many irrigation works have been constructed, but only one-third of the commanded areas actually get any irrigation. The need for rehabilitation and modernization of (some of) the existing systems is paramount. Current analyses (SOGREAH 1993) recommend against considering any new irrigation works in the foreseeable future, in view of the backlog of existing systems that need to be rehabilitated. The development of sound water resource management strategies is crucial for poverty reduction, economic growth and food security.'84 185 394. Landmines. During the war, both sides laid landmines' , often along road and access ways. The number of mines is not known accurately.186 Worst affected were Maputo, Zambezia and Tete provinces.187 Apart from the toll in human suffering, the effective supply of arable land has been reduced. In some cases, entire communities were deterred from returning to their traditional lands after the war. The UNDP/DHA Accelerated Demining Program has made significant progress in releasing this constraint on agricultural expansion. Among the areas which urgently need mine clearance still are the interior of Zambezia and the western part of Maputo province. At the present rate of mine clearance, practically all landmines which threaten human settlements and agricultural activities will have been removed in about five years in the South and about three years in the North. 183 The revenue implications of the reduction of the tariff are small, inasmuch as cashew exports accounted, in 1994, for only two percent of all exports. If the government wishes to increase its revenues, this would better be done by a low rate of tax on a wider base, for instance an increased import tariff on all goods, or a small tariff on all exports. 184 The National Water Resources Development Project will assist the Government in the management and conservation of its water resources. It will include institutional development, capacity building, local and international river basin studies, and critical investment in bulk water storage. 185 UNDP (1996), p. 1. 186 Estimates have ranged as high as two million landmines. 187 See the maps in UNOHAC (1994), p. 24f. 130 395. The deficient property rights regime. With five hectares of land per inhabitant (or 20 inhabitants per square kilometer), Mozambique is one of the continent's most land-abundant countries, once land quality is taken into account.188 Nevertheless land conflicts have been reported both in densely-populated areas such as the major river valleys, and in sparsely- populated areas such as the Matatuine district in Maputo province, and Monapo district in Nampula province.189 Smallholders, who have obtained rights in land through inheritance or through the customary tenure allocation system, have frequently been deprived of some of their land rights by outsiders who use legal instruments to obtain land titles, typically in the most productive areas.190 This insecurity of tenure of smallholders is mirrored by insecurity on the part of large-scale farming investors, tourist operators, and multinationals, whose legitimacy is frequently called into question and who fear reprisals by farmers or arbitrary cancellation of their titles. The lack of definition of the titling process is a constraint on smallholder production and on foreign investment alike. Furthermore, the Land Law No. 6 of 1979 prevents the sale of land, although buildings and land improvements may be sold. The absence of a legal land market, or a market in concessions, means that land per se cannot be mortgaged (though buildings can be). These restrictions are a constraint on investment. 396. Inadequate transport and communications infrastructure. The Roads and Coastal Shipping Projects I and II, which are supported by IDA and other donors, have concentrated mainly on primary and secondary roads. ROCS II, for instance, aims at major rehabilitation of 3,450 km of trunk roads, maintenance of 65,500 km of unpaved roads, emergency rehabilitation of 11,700 km of mainly unpaved roads, and labor-based building of 3,250 km of feeder roads (World Bank, 1994b, p. 10). The ROCS 11 also stresses the maintenance of the newly rehabilitated roads, which is appropriate, given the relatively low cost of maintenance compared with new building of roads. But further investments in rehabilitating tertiary and other feeder roads are required, and there is a need for culverts and farm to market tracks in rural areas. The lack of this infrastructure prevents consumer goods from reaching isolated rural populations, decreasing their incentives to produce surpluses for the market; and it raises the costs of transport which reduces farm gate prices, depressing the incentives of farmers further. 397. Low levels of skill and technology. Some 60 percent of the Mozambican population is illiterate, and in some rural areas male illiteracy is as high as 87 percent. The level of agricultural skills in the smallholder sector is correspondingly low. The technology is shifting cultivation with hand-hoes, with low levels of productivity. There is a need for massive investment in primary education in rural areas, which will show results in terms of productivity change and output. The research and extension systems also require major rehabilitation and upgrading. 398. Farmers ' asset poverty. During the war the Mozambican cattle herd was virtually eliminated. Farmers in the south and the highlands, where cattle were raised, thereby lost their 1 See World Bank (1996c). 189 For evidence on land conflicts, see the frequent references in the Mozambican press, as well as the outputs of the Ad Hoc Land Commission during the period 1992-3, and Borges and Doto (1995), McGregor (1995), Myers (1993), Myers and West (1993), Myers, Eliseu and Nhachungue (1993), Myers (1994), World Bank (1996c), Noronha (1994), and Tanner (1995). 190 The problem is exacerbated by two factors: (a) given the inadequate infrastructure, the demand for land by large-scale farmers and tourist operators is focused on areas close to the main roads and cities; and (b) landmines, which reduce the demand for land in the affected areas and channels the demand towards the already conflict-ridden areas around the cities. 131 wealth base, their draft power and their cheapest means of saving and capital accumulation. This makes the supply response to the incentives given by the new-found political stability and the freer markets less strong than otherwise; and it will constrain the development of rural financial institutions and cooperatives in the future. 399. Lack of social capital and the absence of ruralfinancial markets. The unfortunate history of cooperatives in Mozambique resulted in mistrust by farmers of efforts by the state to facilitate the creation of cooperatives. There are very few farmers' or peasants' organizations other than in the Zonas Verdes around Maputo. Such as there are have small membership bases, few resources and little access to organizational skills. As a result there are no farmers' organizations at present with the capacity to intervene in factor or input markets. This is a constraint in all lines of smallholder production, but especially in cotton, where the use of pesticides is essential for profitability. Furthermore, the ability of rural dwellers to accumulate savings and borrow resources is severely limited by the paucity of rural financial institutions. 400. Distortions in price incentives due to minimum prices andfood aid. Mozambique has largely moved away from central control of the economic system. Several prices and some marketing channels have been liberalized; some of the remaining controlled prices are being moved closer to their import-parity levels. Significant distortions remain, however, so that many producers do not receive proper market signals. Eight agricultural products are subject to minimum prices. 191 In early 1996, on account of the excellent harvests in the region, the government converted the minimum prices into non-binding "reference prices". The government has undertaken to abolish all minimum prices by July 1999. 192 Another problem area is food aid: emergency food aid tends to leak into the market, and although commercial food aid is now auctioned upon arrival, it still has a depressing effect upon prices, discouraging domestic food production and the national integration of maize markets. 401. Lack of capacity of the governmental apparatus. The capacity of the government to carry out its essential functions is limited, in terms of physical facilities, budgetary resources and human resources. One of the factors contributing to the limited government efficiency is the overreaching policies of the past, which forced the government to spread itself thinly over almost every conceivable field of intervention. There is a need for a major restructuring of the Ministry of Agriculture, focusing its mission narrowly on the key functions. Another of the factors is the dearth of skilled personnel. In order to continue operating even at the current levels of efficiency, the government is drafting into the Ministry of Agriculture and sector parastatals college students still in their third or even second year; this solution does not contribute new skills to government and jeopardizes the formnation of the future professionals. To overcome the lack of trained personnel, several donors have provided training programs, but, perversely, the dearth of skilled personnel in the Ministry of Agriculture has meant that a few staff are sent to these training opportunities and are seldom available in their bureaux. 191 See Table 17 of this Memorandum. 192 The minimum price for cotton should also be eliminated. Although the existence of the statutory cotton monopsonies (and, in their absence, the likely existence of natural monopsonies) provides a theoretical justification for the government's setting the cotton price, in practice the government's efforts in price determination have had a minimal impact and it is unlikely that this means of intervention is worth the administrative effort. See Chapter 5 for the details. 132 G. A Vision of Mozambique's Agricultural Development 402. The agricultural sector is recovering quickly, owing to the improvement in security conditions in rural areas, the return of the displaced populations to their traditional lands, the reduction of food aid and the liberalization of markets. Growth of agricultural GDP was 1.4 percent from 1980 to 1991 '93; after a decline of I percent in 1992, owing to the drought, it grew by 19 percent in 1993, 5 percent in 1994'9 and 1995 '5, and 6.2 percent in 1996196. These estimates, based upon district reports, are probably understated. Data from the Early Warning Unit and other sources suggest much larger increases; for instance, cereals output increased 9 percent between 1994 and 1995, and 33 percent between 1995 and 1996.197 Marketed agricultural output from the smallholder sector increased by 18 percent between 1995 and 1996.198 During the 1995/96 season, cashew production nearly doubled to about 60,000 tons, the highest level since 1981. There are ample resources for substantial expansion in the cashew sector, using simple and known technologies, and there are no demand or market constraints but for the export tariff on the raw product. The grain harvest, due to good rains, reached its post- war peak of 1.4 million tons during the 1995/6 season. Farmed areas per family have increased; for instance, in the district of Monapo, Nampula province, the farmed area per household increased from 2.1 ha in the 1992/3 season to 3.8 ha in the 1994/5 season. This was partly due to the return of displaced people, and partly to increases in the area farmed per adult, from 0.8 ha/adult in 1993 to 1.0 ha/adult in 1995.199 403. The benefits of the liberalization of pricing and marketing are being realized, particularly in the southern half of the country where the trading network has reestablished itself. A further cause of the recovery lies on the demand side. The quantities of food aid released on the market have fallen, the donors have shifted towards local purchases of maize rather than shipments from overseas, and commercial food aid is now being auctioned; all three factors raised the demand for local maize and drove cereal prices up. 404. These recovery-driven increases in output - given good rains - will continue for a time, possibly for another two years, as available labor and land resources are once again brought into use. Once output again attains the production possibility frontier, further growth will depend on additional horizontal expansion, on productivity improvements and on the expansion of demand. 93 World Development Indicators 1995, Table 2. 94 Southern Africa Department (1995), Annex, Part D: Annual Growth Rate (constant prices). World Bank (1996a), p. 4. 196 World Bank mission estimates. 197 Sources: information reported by Ministry of Agriculture/Michigan State University project, Working Paper 13, in turn from the Department of Food Security of the Ministry of Commerce, the FAO Production Yearbook, the World Food Programme database and the Early Warning Unit. 98 World Bank (1996a), p. 4. 99 Comparing the MSU survey of 1993 and Borges and Doto (1995), p.3 1. 133 Horizontal expansion 405. Horizontal expansion will arise naturally as smallholders and large-scale ("commercial") producers exploit the agricultural frontier in the northern half of the country, and particularly in the provinces of Niassa, Cabo Delgado, Zambezia and Tete. Using a few simple assumptions, it is possible to arrive at a rough estimate of the amount of land available for expansion in the future.200 We assume that the average rural household retains in fallow four times the area of land it cultivates, so as to maintain yields without fertilizer inputs. Then we calculate, for each of the provinces, how much cultivable land is currently in use and under fallow. Cultivable land beyond this total, less the area in national parks, is likely to be claimed and put to use at some time in the future. The calculations show that there is no "extra" land available in the three southernmost provinces (Maputo, Gaza and Inhambane) or in Nampula province - that is, yields cannot be maintained by fallowing alone, and the transition to a settled agriculture has already commenced. But in the other provinces there is a total of at least 14 million hectares of good land201 which is not currently under cultivation or intended as fallow. Most of this uncultivated land is to be found in the northern provinces of Niassa, Cabo Delgado, Zambezia and Tete. 406. Even if there were no urban-rural migration, and even if the population growth rate of 3.3202 percent were not to fall - both extreme assumptions - these 14 million hectares would be sufficient to absorb the increase in the rural population, on a shifting cultivation basis with no loss of fertility, for another 23 years. This process of horizontal expansion could be encouraged by the rehabilitation of the roads linking the major towns in the North, and would occur automatically without the government's adopting special settlement policies. Better roads in these parts would permit new farmers to settle, and enable sitting farmers to move from small plots to larger ones. In fact the growth process will be more like a patchwork quilt: some farmers will improve their productivity given their existing land, and others, particularly young farmers, will move to new land at the frontier. The process will not be smooth, and land shortages will develop in some areas, forcing farmers to gradually abandon shifting agriculture. This will be the case especially in the three southernmost provinces and in Nampula province. 407. A corollary of the above calculations is that land-saving technological improvements such as fertilizers are more likely to be adopted in the denser-populated areas of the South and Nampula. In the land-abundant North and Center, on the other hand, the introduction of fertilizers is likely to be less efficient, on the whole; in these areas certain types of pesticide and herbicide - which raise yields while reducing labor input per unit output - would have greater chances of adoption by farmers. It was shown, for instance, that cotton production by smallholders in Cabo Delgado is more efficiently done by clearing new land from time to time, than by using fertilizers to maintain fertility.203 200 See Moll (1993), Table F2, p. 13. 201 That is, land subject to "no limitations" or "moderate limitations" according to the USDA classification system. See Moll (1993), p. 13. 202 World Development Indicators 1995, Table 25. 203 See Chapter 5. Obviously the rule of thumb that land abundance militates against fertilizer adoption does not apply in every instance, since there are many other factors that go into determining efficiency and profitability. 134 408. This horizontal expansion and exploitation of the agricultural frontier will compete with scenic and wildlife uses. Mozambique has enormous areas whose obvious vocation is for wildlife development. Some of these areas are already included in the national park system, but others remain where the resources available, owing to swampy conditions or to the lack of drinking water, lend themselves better to wildlife development than to cropping or livestock. This natural vocation is reinforced by geographical location. Mozambique is adjacent to one of the foremost national parks in the world (Kruger National, in the Republic of South Africa), as well as others in Zimbabwe (Gona-Rhe-Su), Swaziland, and Northern Natal. Besides its natural tourist attraction for the landlocked Gauteng population, Mozambique's location would allow for its integration into international tourist circuits. As a result, the economic benefits for the national economy of scenic and wildlife development is anticipated to be several times larger per unit of (the relevant) land than its alternative use as cropland or pasture. Benefits would include generation of foreign exchange, employment and fiscal revenue, as well as increased meat supplies for the neighboring populations. For these benefits to be realized, Government action will be needed to define property rights and put in place institutions to ensure sustainable management of these natural resources. Productivity improvements 409. In the medium term, productivity improvements will be the main engine of growth, given appropriate public sector interventions. These improvements will arise in at least four areas: smallholder technology, large-scale farm technology, processing technology, and marketing. 410. Improving the technology and farm management skills of smallholder farmers is fundamental to increasing smallholder productivity. The key element of this improvement is the effective dissemination of new production options. An IDA-financed project with a large extension component initiated activities in 1992, in two provinces in the north (Agricultural Services Rehabilitation and Development Project). Despite the difficulties owing to the security situation in the early years of the project, a substantial contribution has been made to the improvement of productivity of small farmers. 411. Initially, smallholder output of annual food crops will be encouraged by simple labor productivity enhancements, such as seed spacing and conservation tillage, coupled with the reproduction of improved varieties of grain, legume and root crops. The research facilities have already produced some improved (open-pollinated) varieties, which have been shown to be more 204 profitable for farmers than using local unimproved seeds . If the research effort is well funded and organized, it can be anticipated that there will be significant further improvements in productivity. These will likely arise from some combination of: drought-resistant seed varieties (particularly for the drier regions of the South), seed varieties offering better yields (particularly for parts such as Zambezia where rainfall is high and consistent), pest and disease control, and the transition from shifting to settled agriculture. The latter transition is likely to occur in the more heavily populated southernmost provinces and in Nampula, and will entail the introduction of appropriate chemical and organic fertilizers. Export crop productivity, in particular, can be enhanced by treading well-known paths. In the area of cashew, for instance, productivity will respond to improved husbandry and, in due course, to replanting with mildew-tolerant, high- 204 Moll (1993). 135 yielding varieties and by increasing the availability of water. With the technology generation and transfer systems in place, the smallholders will start a (slow) process of capital accumulation which will allow them in the longer term to invest in land clearing, animal traction, on-farm storage facilities, and other capital goods. 412. In the past livestock contributed only 5 percent of the agricultural GDP. The gradual reemergence of the livestock herd will raise productivity per unit of labor, as cattle provide dung fertilizer and oxen provide more efficient traction. In the longer term, livestock accumulation is expected to play a major role in facilitating (and instrumenting) capital accumulation in the sector. This process will accelerate in time as the capacity of smallholders to save increases. This will occur initially in the tse-tse-free south and the highlands of Manica. It could be accelerated by judicious government interventions in restocking. Demand for livestock products is likely to be high (see para. 423) so that projects of this kind would earn high returns. As the population rises, a process of intensification will take over. Tree cover will recede, depriving the tse-tse of its habitat, and it will become possible to breed cattle in parts of the North as well. In this domain the government and the donor community could play an role, for instance in tse-tse control programs which have proved to be effective in East African countries. The downside of this mode of development is that the rural population will need altemative sources of energy as the tree cover recedes. There will also be increased soil erosion. This highlights the important role of the research and extension system in informing farmers about conservation approaches. 413. In the long run, the distinction between the family and commercial sectors will wither away as the increased market orientation of family farms gradually draws them into commercial production in addition to production for own-consumption. Fundamental to this process will be the development of both output and factor markets to allow smallholders adequate access to resources and inputs as well as the ability to sell their produce at market prices. 414. The large-scale ("commercial") fanning sector is set to play a significant, though lesser, role in the evolution of Mozambican agriculture. Although smallholder farming has superior (lower) domestic resource cost coefficients, some lines of large-scale farming are also efficient, on account of the low cost of land. The DRCs for large-scale farms are an estimated 0.6 for cotton (at the favorable 1995 prices), 0.2 for citrus, 0.8-1.1 for sunflower, and 0.7 for rice. Large-scale farmers are not efficient producers of maize, with a DRC of 1.1-1.2. However, when Mozambique's concessionary capital inflows decline in the future, the exchange rate will depreciate and all these DRCs will fall, and some large-scale farmers could profitably produce maize at international prices, although smallholder production will still be more efficient than large-scale production. 415. The large-scale sector is already providing services, such as processing (e.g. of cotton), transport and ploughing to smallholders; these interactions will intensify as smallholder output rises. The large-scale sector has expanded considerably in the past ten years through the process of privatization of the majority of the state and cooperative farms. The fact that these are now under private management will probably result in productivity improvements. The process will continue with the arrival of foreign capital and technology, and the capitalization of the existing farms. The large-scale sector is active in several irrigation schemes and could expand this activity to other irrigation schemes which are non-functional at present. High levels of investment will require a high level of tenure security. This points to the need for a conducive land regime, and clear enunciation of a private sector development policy.205 205 There is already in place a foreign investment policy. 136 416. There are already many cases in Mozambique in which the large-scale farming sector and smallholders operate in symbiosis. The cotton ginneries with their outgrowers are one example.206 The outgrower arrangement and its variants are often more efficient than large- or small-scale operations would be separately. The smallholders' lack of collateral is addressed by the estate or processor firm's capacity to provide a package including credit, extension and marketing, engendering a relationship from which the smallholder has no interest in reneging. At the same time, the unit cost of labor is lower for smallholders because there is no moral hazard problem which, in the case of a large hired labor force, would require significant investments in monitoring, management and union negotiations. There is considerable scope for recovering pre-independence levels of production by outgrower arrangements, for instance in cotton, tea, sugar and copra production. In time, further processor/marketer/smallholder interactions could develop, as they have in Kenya and other African countries, in dairying, flowers, citrus (especially grapefruit), specialty nuts, French beans and other products. Improvements in the Productivity of Trading, Marketing and Processing 417. Alongside the growth of large-scale farms there will also be the development of trading firms, input supply firms, and processing firms. A major requirement will be the reform and development of the banking sector, including capacity to provide finance and to operate a payments and resource transfer system (currently in process). 418. The commercial sector is expected to be involved primarily in agricultural production. Some commercial sector operations, such as the cotton ginneries, will in addition be involved in the supply of inputs, technology and credit to other farms in the commercial sector and to the smallholder sector; and in the processing and marketing of some commodities (e.g., cotton, tea, sisal, copra), on their own account and on behalf of outgrowers. 419. Improvements in processing efficiency will be an important source of growth. In tum these improvements will be driven by privatization and the liberalization of markets. An example is the cashew processing industry. With its many new entrants following privatization there will be improvements in processing efficiency and these will be transmitted upstream through the trading system to the farmer, and could induce further increases in farm marketing. 420. Another source of productivity improvement is the trading system. Although the southem half of the country is now effectively covered by trade links, the margins of traders are still high, often in excess of 100 percent207, owing to the small size of the majority of trading operations and to the thinness of the market. In time, the volume of trade will increase, the markets for transport and trade will become denser, and the margins will fall, so that farm gate prices rise, other things being equal. Public policy has a role here in providing infrastructure, particularly roads and telecommunications, so as to reduce the costs and risks of transport and facilitate entry into the business. 206 The PUPI (Pequena Unidade de Producao Intensiva = Small Unit for Intensive Production) is another model. This is one of the methods used by LOMACO, a cotton joint venture company. A plot of 60 hectares or more of the company's estate land is split up into 1-ha lots, each of which is assigned to a smallholder. The company does the ploughing, fertilizing, and pesticide application using mechanical means, and the smallholder is responsible for weeding and harvesting. The smallholder's revenue is docked for his/her share of the input costs. 207 Ministerio da Agricultura e Pescas (1995e), p. 59. The data refer to the 1994/5 season. 137 421. Inflows of foreign capital, particularly from South Africa, are already contributing to increases of efficiency and technology transfer in processing and trade. This process will doubtless continue, as witness the investment by South African firms in other countries such as Botswana. Foreign capital is also expected to play a leading role in developing wildlife reserves and ranches, private parks, and other tourism facilities, as well as some infrastructure on a build- and-operate; or build-operate-and-transfer basis. An explicit and conducive regime for foreign capital investments in Mozambique is required as well as reform and clarification of the land tenure law. Expansion of demand 422. Long-tern growth in the demand for food is determnined essentially by population growth, income growth and the income elasticity of demand. If future population growth is 3.3 20820 percent , per capita private consumption rises at 2.2 percent09, and the income elasticity of food demand is 0.7, then food demand in Mozambique is set to rise by about 3.9 percent per 210 211 annum. In fact, agricultural GDP has grown by over 6 percent per annum , on average, since the peace accords in 1992, though some of this is due to the catch-up after years of war, dislocation and unfavorable policies. 423. The demand for livestock products will probably rise more quickly than the demand for most food products, since its income elasticity, as calculated in other less developed countries, is well above 1.212 424. Mozambique's share of world commodity exports is low in all cases, so that demand will not be a constraint on expansion. In the case of processed cashews, Mozambique's share is 213 214 about 4 percent . World demand for cashews is growing at about 7 percent per annum so that prospects for Mozambican exports are excellent. H. An Agricultural Development Strategy for Government 425. In the light of the major findings of this Memorandum, it appears that a sound agricultural development strategy for the government would have three characteristics. (a) It would be based on the expansion and deepening of rural factor and input markets, bringing underutilized land and labor into production, facilitating the mobilization of new capital, introducing new technology into the agricultural sector, and raising productivity. It would not be based on the expansion of government controls or spending. (b) The process would be led by smallholders as opposed to large-scale farmers or multinationals. (c) It would be driven by an 201 World Development Indicators 1995, Table 25. 209 Estimate from World Bank (1996a), Annex 1, p. 1, referring to 1997. 210 This is on the assumption of a constant food price. 211 World Bank (I 996a), p. 4. Other data sources suggest larger increases in output between 1993 and 1996. See paragraph 402 for details. 212 von Braun et al. (1991), p. 66. See also the elasticities for milk of 1.0 in Pakistan, and those for beef of 2.8 and more in Nigeria in the 1960s, Tsakok (1990), p. 255f. 213 Kumar (1995), p. 14. Its share of production is 8 percent, ibid. p. 7. 214 Kumar (1995), p. 14. 138 appropriate regulatory framework, market-based incentive signals and infrastructure investments. 426. This overall perspective will be dealt with in more detail below. We start with the basic principles of government intervention in the economic process, before proceeding to discuss four areas of intervention in turn: the incentive framework, rural infrastructure, public services and regulation. (a) Principles of Government intervention 427. It is essential to try to identify a few principles for Government intervention in order to help with the task of prioritization. There are literally hundreds of things that any government could do, but the available finance and high-level skills cannot cover more than a tiny fraction of these. A judicious choice needs to be made from the long list of proposed actions in the Government's "Agricultural Policy and Strategy for Implementation" (1 995). Many ways could be found to sift through the multitude of possible proposals. One way is to select actions, policies or projects on the basis of their rate of return. Another is to try to identify the most important constraints on economic activity, and to see whether government intervention can alleviate these. A third approach - and the one favored here, for reasons given below - is to have the government intervene only, or primarily, in areas in which there is an obvious market failure which is susceptible of correction. In fact, these three approaches would arrive at similar rankings of the importance of different policies and projects, because any market failure is simultaneously a constraint on the activity of private actors, and lifting such constraints is likely to be a high-return intervention. Whereas having the government invest in activities in which the private sector unaided would itself engage, and where there are no obvious constraints or market failures, is likely to produce low returns because the government would merely be "crowding out" the private sector. 428. The principle followed here for distinguishing between the roles of government and of the private sector is that the government should intervene only in areas in which there is an obvious market failure. Market failures can arise from a number of different sources: positive or negative external effects of an economic agent upon other agents (e.g. air pollution, agronomic research, livestock disease); the inability to exclude individuals from the benefits conferred by a good or service (e.g. military defense, landmine removal, extension advice); the inability of agents to charge fees for goods or services (e.g. city roads); natural monopolies (e.g. the electricity grid); asymmetries of information (e.g. potentially dangerous agro-chemicals - see §462). One of the most notable market failures in poor societies is the inability of the state to enforce loan contracts; this compels banks to require collateral, which poor people cannot provide. 429. In all of these cases, the market left to itself would generate suboptimal outcomes, for instance underinvestment in research and infrastructure, a lack of extension advice, insufficient landmine removal, a dearth of good quality roads, an excess of livestock disease, or an inordinate amount of injury from agro-chemicals. Poor people typically underinvest in education, because even if their enhanced earnings after the investment would pay for the tuition, they are unable to present the collateral in order to get loans. In these cases, there is a rationale for some kind of government intervention. Depending on the specific case, this intervention might take the form of legislative enforcement, incentives, taxes, government provision or government operation. 139 430. Since the government's resources are limited, these are best concentrated in domains where the government has a comparative advantage, and these are generally to be found in appropriate interventions in clearly-defined instances of market failure. In markets which are competitive or at least contestable, the private sector is usually a more efficient provider of goods and services. In less developed economies, where the government's resources are limited and where high-level human capital is in short supply, it is usually counterproductive to have government involved in the production and marketing of private goods. Apart from its low efficiency in so doing, this also has the undesirable effect of diluting the high-level human capital available for the central "governing" work of government. There is no reason, however, why the private sector could not, under contract to Government, be brought in to produce public goods. 431. One way of summarizing these ideas is to say that the government should be active in the area of broadly defined "public goods". These are physical and institutional goods which are necessary for a well functioning economy but which the private sector does not typically produce in optimal amounts owing to one or other market failure. 432. From the large palette of actions which need to be taken, four stand out as priorities on account of their strategic importance for the sector and because they lay the groundwork for further developments in the future. Two are outside the agricultural sector and two are within it. The first is to ensure that the rate of inflation continues to fall. Rural financial markets will not develop without this step and foreign investment will be discouraged if the macroeconomic instability of 1990-5 returns. The second is to continue the projects for improving the road system, and to extend these further in the direction of feeder roads, culverts in rural areas, and farm to market tracks. This will pave the way for improvements in the efficiency of trading and transport, and in due course enable larger increases in supply. The third is to continue to reduce the tax on exports of raw cashew nuts. With lower taxes, supply will rise faster and there will be favorable effects in terms of rural demand for consumer goods. The fourth priority is to improve the land legislation, regulations and administration so as to guarantee security of tenure to smallholders and large-scale farmers. Without this improvement, investment could be held back and land conflicts may increase. 433. The timing of these actions is crucial, otherwise the value of many costly investments could be lost. Macroeconomic stability is an ongoing priority; and the legislative and regulatory changes required for security of tenure of smallholders and large-scale farmers could be effected within two years. With respect to the tax on the export of raw cashews, a study currently underway may provide insights on the optimal timing for cashew tax reform. Rehabilitation of the road system is under way, but the program needs to be extended in respect of feeder roads, culverts and market tracks; these improvements should logically accompany or precede the revamping of the agricultural extension service, so that the increased output can be marketed efficiently. (b) The incentive framework 434. The first order of business for government in the effort to achieve agricultural expansion is to create an enabling environment in which the private sector - that is, smallholders, large- scale farmers, traders, processors and bankers -will face market-determined economic incentives, so that the market acts as the main engine of growth. Attention should be paid to macroeconomic and fiscal policy; internal and external trade; the development of markets; 140 export crops; food aid; special incentives and subsidies; direct government involvement in production; and land tenure. 435. Macroeconomic policy: In the area of macroeconomic policy, it is essential that the progress attained in reducing inflation during 1996 be maintained. Without this, farmers will not have access to financial markets. While this is a matter of concern for all farning sectors, it is of especial importance for the "emergent" farmer group, viz. the ex-peasant farner who has titled land, some machinery and a few non-family employees; the effective real interest rates this group faced during 1995 were over 100 percent . It is also essential that the government's present commitment to a market-determined exchange rate be maintained. On the fiscal side, the sales tax ("circulation tax") should be abolished in rural areas because it has a cascading effect which raises the costs of and reduces the amount of trade. For the sake of the financing of local institutions, it could be replaced by a properly enforced land tax. The government plans to replace the imposto de circulaqdo with a value added tax in 1999. 436. External and internal trade: To minimize market distortions for a given revenue take, the tariff regime should be simple (with as few variations in rates as possible) and non- discretionary.216 The government has already made great strides with trade liberalization. As part of the Third Economic Recovery Credit (1996), the number of bands for customs duties were reduced from 12 to 5, with essential goods such as medicines paying no duty, and consumption items paying 35 percent, with other categories in between. The scope for discretion on the part of customs officials was also greatly reduced. 437. With respect to internal trade, there are still a number of bureaucratic obstacles to registration for wholesalers and retailers in agricultural factor and product markets, and in consumer markets. Some of these barriers to entry are due to the law itself (e.g. the requirements for bank accounts of a certain size, and police clearance) and can readily be altered. Others are due to misinterpretation of the law by minor officials (e.g. the requirement for the sexta classe), and would need public restatement of the rules as well as prosecution of officials who exceed them. It is likely that women will be among the chief beneficiaries of this legal clarification, 215 This is because small businesses borrowing from the banks have to pay their interest in full in advance. 216 The Government proposed, during the negotiations of the Third Economic Recovery Credit, to apply a seasonal tariff to five agricultural products (maize, rice, beans, groundnuts and sugar). The intention, presumably, was to have a high tariff during the harvest season and a low tariff in the "hungry" season. This would cut off imports and protect producers during the harvest season and help consumers during the "hungry" season. There are several objections to this arrangement. (i) The repeated changes in the tariff, and the dates selected for so doing, would become the object of speculation and would provide scope for patronage. (ii) This rough attempt at price stabilization would rob farmers and wholesalers of the incentives for inter-seasonal storage. The market would itself develop a certain degree of price stabilization - particularly if the government helps with on-farm storage technologies through the research and extension system - and the government is unlikely, through blunt tariff instruments, to be able to improve on private market outcomes. (iii) High tariffs are harmful to welfare as a whole because there are "deadweight losses" involved in the process of redistributing from consumers to producers. (iv) There is no obvious reason why the producers "need" a high tariff. If the problem is dumping, then this would best be addressed using the methods of the World Trade Organization, which would require an objective definition of 'dumping" and would issue in a ban on imports from the relevant country. If the problem is long-term, permanent protection in neighboring countries, then no tariff is called for because it would not improve on the free trade outcome - the tariff would merely redistribute from consumers to producers and entail unnecessary deadweight costs. See also the discussion on tariffs in Chapter 4. 141 since in Mozambique there is already a phenomenon of "market women" who deal particularly in food crop marketing. 438. Development of markets: The best way to ensure efficient incentives for agricultural production is for the government to abstain from its traditional interventions in pricing and marketing. Efforts at pan-territorial pricing, whether by AGRICOM or its successor, 1CM (Cereals Institute of Mozambique), have virtually always had counterproductive effects upon smallholders. Presently the ICM, due to lack of funding, is incapable of carrying out its mandate of price support and is acting as a wholesaler. Its operating efficiency as a wholesaler would likely improve if it were privatized, as the organization would then have a profit motive. Accordingly its warehouse network of 120,000 tons could be privatized without delay. The government is already considering restructuring it. 439. A prerequisite for the development of domestic markets and the growth of domestic production to satisfy domestic demand is the re-establishment of a system of domestic trade which provides the physical and economic linkages between rural producers and urban consumers. As pointed out earlier, a new network has gradually extended itself over the southern half of the country. As demand continues to rise in the future, and provided the road system is rehabilitated, the network will spread over the northern half of the country within the next three years. To achieve this, it will be sufficient for the government to reduce the registration requirements of traders; no special govermment interventions in the form of special loans for traders would be necessary. 440. The development and the dissemination of improved farmn level crop storage technologies, and the privatization of large, urban storage facilities, will be important as the government withdrawal from pan-territorial pricing causes greater price fluctuations at the farm gate. Improved on-farm storage will enable smallholders to both cope with, and take advantage of, these fluctuations in ways that will not only benefit them, but will reduce the need to develop off-farm storage capacity. 441. Export crops: Two sectors deserve special mention with respect to production and market incentives: cashew and cotton. Concerning cashew, the main challenge is to reduce further the export tax on the raw product. Studies have shown that it is possible for the processors to improve their efficiency so that they can compete on the world market without protection. There is vigorous competition for the raw cashew market among petty traders, large- scale traders and processors in rural areas of Nampula and Gaza, and there is already evidence that the export tariff reduction has had an impact on prices at the farm gate. The tariff reduction will benefit women disproportionately, since it is they who do most of the work in collecting cashew nuts. 442. Concerning cotton, there do not yet exist rural cooperative institutions which can provide credit to cotton smallholders. Unlike food crops, for which purchased inputs are not an absolute requirement, the profits from cotton are low without pesticides. The smallholders do not yet have the savings necessary for these purchases. Currently their only source of credit is the ginneries, several of which are joint-venture companies. The ginneries feel that they cannot enforce repayment of their loans unless they have the sole right to purchase cotton in their areas of operation. The government's task in this domain is to encourage the development of cooperative and other forms of credit with a view to the ultimate scrapping of the cotton monopsonies. If the efforts to promote rural financial markets are appropriately focused, they could be of especial assistance to rural women. 142 443. Food aid: Another aspect of the incentive framework concerns the effects of food aid. The humanitarian need for food aid in the early 1990s was inescapable -it prevented mass starvation and the death of probably millions. It has since become an obstacle to the development of a market-oriented smallholder sector. The leakage of emergency food aid into the market, and the importation of commercial food aid (albeit auctioned at arrival) tend to undercut both production for the market by smallholders, and the development of rural-urban market links. While a certain amount of food aid is still necessary at present, given the gap between national need and production, phasing it out in the near future is a prerequisite for development of existing national potential. 444. As Mozambique reduces its reliance on external sources of food, the government is working on the definition of a new food security policy. In the light of the discussion in this Memorandum, it appears that the policy ought to define the criteria for short term food security interventions at the national, provincial, district, and household levels, at times of a shortfall in supply. It should specify the relative roles of the Govemment, donors, NGOs and other actors in the field. 445. Special incentives and subsidies: From the set of principles outlined above there follow far-reaching conclusions about other sorts of government interventions such as subsidies. If private entrepreneurs, responding to market signals, are capable of operating and growing within this enabling environment, there is no need to provide them with subsidies except in two cases: when there is a clearly defined market failure which needs correction (and if alternative remedies such as taxes or prohibitions are judged unsuitable) or when there are distributional grounds (e.g. to ward off starvation, to alleviate the effects of poverty, to help women to obtain economic independence). 446. In the case of smallholders there may be distributional grounds in specific cases for subsidies. For instance, a rural investment fund (RIF) could be introduced, along the Zambian model, providing matching grants for local communities to build infrastructure. In view of the expected expansion of demand for livestock and dairy products, and the current inability of smallholders to borrow, the government could consider some form of restocking arrangement focused on building up smallholder herds. This might occur, for instance, in the context of a dairy program organized by a private sector processing company. Subsidies of this kind should be limited in time and have precise objectives. The prime task of the government with respect to the smallholder sector is to compensate for certain market failures which affect the smallholder sector especially seriously, and in particular the lack of smallholder-focused research and the lack of information and extension assistance. 447. Direct government production: Government involvement in agricultural production has decreased as nearly all the state farms have been sold. The few remaining state farms should be divested through sale, management contract, management employee buyout or other arrangements. Farms required for research or biodiversity conservation should be retained in public ownership, with proper arrangements made for their operation and management. The government has all but terminated its involvement in commodity processing and marketing, with the sale of the state cashew processor, among other properties. 448. There remain the joint venture companies such as the cotton producers/ginneries, in which the state has a 50 percent share. There is no market failure which demands the state's involvement in ginning. The involvement of Government in the firms' decision-making adds an inessential layer of bureaucracy and may induce managerial slack to the extent that the firms are protected from the risk of failure. Consequently it would be better not to contemplate any new 143 joint ventures.217 Divestiture of the government's share in the existing joint ventures should be discussed in tandem with the phasing-out of the monopsony arrangements in cotton (see §§ 442, 480). 449. Land tenure: The Land Law No. 6 of 1979 appears to be incompatible with security of tenure for smallholders and for the large-scale sector. The law also prevents the functioning of a land market and inhibits the use of mortgages. The government has undertaken to revise the land law and, in 1995, appointed an Inter-Ministerial Land Commission with a full-time secretariat to undertake the task of drafting a new law and regulations. The draft legislation of September 1996 makes certain improvements in the smallholder sector without providing sufficient guarantees of security, and it is not clear whether it would permit the functioning of a market in government-awarded concessions, so as to enable land mortgages. What is needed still is a mechanism to prevent outside land developers (farmers, tourist operators, etc.) from moving smallholders off land which the latter have acquired by customary tenure. One approach, suggested by the government's National Land Policy document of September 1995, is to require that outside developers negotiate with the occupants of the land, permitting the latter the right of veto; and to back this up with permission for the occupants to receive legal compensation, either in cash or in kind, from the would-be investor. In order to guarantee tenure security to small and large farmers, it will be necessary to introduce legislative changes, to adjust the accompanying regulations, and improve the implementation capacity of the government institutions dealing with land. (c) Rural transport and communication infrastructure 450. In Mozambique the needs for rural infrastructure are enormous, and as has been stressed in this Memorandum, roads are a primary driving force in the process of agricultural expansion. Research in other countries indicates elasticities of agricultural output with respect to road investments of as high as 20 percent, and the returns to rehabilitation and maintenance of existing road systems are higher still. To prepare for a take-off in marketing it is essential to continue with the rehabilitation of the existing feeder roads, as the Second Roads and Coastal Shipping Project is doing, and to build more culverts and farm to market tracks. In order to link up the northern and southern halves of the country the major trunk roads - currently in fairly good repair - will need constant maintenance. 451. One way of addressing the problem of infrastructure is through a rural investment fund (RIF) of the kind that was started in Zambia in 1996. Local communities would apply to the fund for matching grants for specific purposes, typically farm to market tracks, culverts or other items of social infrastructure. The justification for the subsidy is that these items of infrastructure would not ordinarily be undertaken by private agents. It may be more efficient to permit decision-making at the community rather than the ministerial or provincial level, as local communities would be best equipped to evaluate the benefits arising from competing infrastructural investments. For instance, some communities might prefer dams over tracks or culverts. and this possibility should be allowed for in the design of the rural investment fund. 217 Often multinational firms favor the joint venture concept because it provides them with security against arbitrary government actions. But then it would be optimal for the government to improve its economic policy-making so as to avoid interference and haphazard rule-setting. 144 452. One of the main contributions of the strategy of agriculture to the roads program will be to direct the investment program to those areas where agricultural development is most sustainable and appropriate. Research in other countries shows that farmers are highly responsive to land quality, that is, roads built in areas of high land quality (in terms of soil chemical composition, rainfall, inclination, etc.) attract farm settlement. On the other hand, roads built in areas with poor quality land are likely to have low economic/agricultural returns, while causing habitat fragmentation, threatening the viability of some populations.' It was pointed out that Nampula is infrastructurally under-served, given its agricultural output, compared with the other provinces 219; so that one of the priorities, within the existing road rehabilitation program, ought to be to concentrate rehabilitation and maintenance works on rural areas in that province. Associated with the road investment program, there should be a process of zoning in order to ensure the protection of parks, forest and wildlife areas. 453. Part of the process of improving the country's infrastructure is the demining program. This Memorandum has stressed that the remaining landmines will pose enormous costs in human suffering and are an important constraint on agricultural growth (§394). It is essential that the UNDP/DHA Accelerated Demining Program continue and complete its work. 454. Coastal shipping has been privatized and there is now more vigorous competition in this area. Communications and particularly the telephone system need to be extended to rural areas. (d) Public services: research and extension 455. Certain agriculture-related services are not provided in sufficient volume by private agents responding to private incentives. Research - at least for certain goods produced by smallholders - is a case in point. Private agents will not, for instance, do sufficient research on smallholder maize, because they will not be able to charge fees to the many users of their findings. On the other hand, hybrid maize varieties are routinely produced by major multinational firms and the market functions because the cultivars themselves are sold every season and cannot be reproduced. Similarly, private producers of agricultural hardware have an incentive to do research and produce better tools and machines, because their inventions are embodied in the objects that they sell; and they frequently provide technical advice to farmers as a means of marketing their goods. In these domains there is no need for government involvement other than for a regulatory function (see below). In sum, the government ought to play a major role in providing certain kinds of smallholder-oriented research. This does not imply that the government necessarily has to execute these projects. Some of them can be done in conjunction with the national university; others could be contracted out to private agents. Areas essential for government intervention include composite (or open-pollinated) grain crops, vegetatively propagated root crops, cassava, yams, potatoes, pulses (e.g. cowpeas), agronomic practices and national resource conservation. 218 Chomitz and Gray (1996) investigate the relationships between soil quality, deforestation, market access, tenure status, and road building in the case of Belize. They point out that in some parts of Belize, road building would be a "lose-lose" proposition, causing habitat fragmentation and providing low economic returns. 219 Part of the reason Nampula has fewer roads per cultivated hectares is that certain key trunk roads linking it to Zambdzia province have not yet been rehabilitated. But even when these roads are rehabilitated, Nampula will still be under-served compared to the other provinces, in the sense of having a higher ratio of cultivated area (and output) to road length. 145 456. In order to effect long run technological change in the smallholder sector, an effective national research and extension system is essential. The system should be capable of doing adaptive research and variety testing on cultivars brought in from abroad. The current status of both the research and the extension systems is weak, as are the linkages between them. Experience of other countries in Africa and elsewhere demonstrates that the returns to investments in these areas are high, both in the short run (by "borrowing" and "verifying" technology) and in the long run (through applied research). Accordingly, research and extension merit attention, so that the basis for achieving these gains can be established. 457. Concerning research, the Government's strategy is outlined in the "Politica Agraria e Estrategia de lmplementavao". Strengthening of the national university is a prerequisite to adequately staffing the national agricultural research institutes. Training and retention of faculty and students are the most important factors here, as the current low levels of pay and training result in a high attrition rate. It is important, however, not to force development of the university system at the expense of the primary and secondary levels, since these levels provide the "raw material" for the university and provide target populations in rural areas and elsewhere with the education necessary to make full use of research and extension messages directed to them. 458. Strengthening the national research institutes, INIA in particular, is key, and the Government's strategy is appropriately focused on applied research in important staple crops. Screening and selection of existing varieties for local adaptation is a good tactic when coupled with on-farm trials and a good linkage with the extension system. Short season varieties of maize to reduce susceptibility to drought are important as are varieties with good storage characteristics and disease resistance. Similarly, there are strong reasons for emphasis on agronomic practices and soil and water conservation. 459. Development of the extension system to assist the smallholder sector will be important, as will the use of this system to provide feedback from the farm level to the research system. Strengthening these two-way linkages will be critical to maximizing the return to investment in the extension system. The government ought also to play a role in the provision of veterinary services in rangeland areas of difficult access, and the production or reproduction of basic genetic materials. The extension role could also encompass information about new technologies incorporating the use of fertilizers, focused upon the relative land-scarce areas in the South and in Nampula. This should be done as a complement and supplement to the actions of NGOs and the private sector, and not as a single, exclusive provider. Other challenges include ensuring adequate financing and adequate trained personnel. (e) Regulation 460. The classic role of government is to regulate the "rules of the game" and to seek ways of internalizing important positive and negative externalities. In the context of Mozambican agriculture, these areas of regulation include technology transfer, the use of dangerous chemicals, natural resource management, and zoning. 461. Technology development and transfer: The rules and procedures for importing new technologies need to be defined, particularly when the technologies are embodied in seeds or animals. This will raise, in turn, the issue of patents and royalties for trademarked genetic materials. Development of the national research and extension system will be important areas for government involvement. 146 462. The use of dangerous chemicals: The rules for the use of chemicals dangerous to workers and/or harmful to the environment need to be defined. The reason is that workers often lack information about the chemicals (e.g. pesticides) that they are required to handle. In the Philippines, for instance, it has been shown that the benefits of agro-chemicals were outweighed by the adverse health effects on the workers who handled them. It is the asymmetry of information that justifies some measure of state intervention. With respect to environmental harm, there may be external effects, for instance chemical wastes which are passed downstream after rains. The good done by regulation can, however, be more than undone by activist policies such as subsidies which encourage inefficient over-use of, for instance, pesticides. 463. Natural resource management: For the conservation of the country's natural resource base, the role of Government is unique. Even though the private sector, under clear rules and effective enforcement, can play a major role in this area, the lead, definition, and control must remain in public hands. Government should be the formulator and enforcer of policy, and the coordinator of foreign/donor aid and investments, but not necessarily the owner, developer, and/or manager of national parks and game reserves. Normally the policies for private sector intervention in this sector will be the same as for the rest of the economy. 464. If the aim of exploiting scenic and wildlife resources is to be achieved, it will be necessary to develop methodologies for involving local populations in joint management of multi-use lands, so that wildlife (centered in the lands that lend themselves for their growth) and crops and livestock (which will normally surround the reserved lands) merge harmoniously. By giving the local populations a stake and a role in wildlife development, such a strategy not only contributes to making wildlife development sustainable but also generates benefits for the local population (similar to the CAMPFIRE approach successfully tested and introduced in Zimbabwe.)221 (f) Public services: health, education, water 465. Effective public services in health, education and water supply are essential for rural development. Health services are needed for their own sake as well as for the improvements in labor productivity that they can bring. It has been stressed in this Memorandum that the technological change that will undergird future agricultural productivity growth is crucially dependent on the expansion of primary schooling education. 466. Access to clean water directly improves the quality of life of rural dwellers, but it also improves their health status, which raises their productivity. To this end, the National Water Resources Development Program, partly funded by IDA, has a rural water and sanitation component which aims to develop sustainable, demand-driven water supply management systems in rural towns. (g) The Public Investment Program 467. Investment in agriculture in 1992 was $62 million and has fallen steadily since. In 1995, public investment in agriculture totaled some $26 million, which corresponds to 10 percent 220 Pingali and Marquez (1994), and Antle and Pingali (1994). 221 This work is also supported by the Transborder National Parks project with funding from the GEF. 147 of the total Government expenditures, and less than one-half of investment in the social sectors ($57 million) and one-fifth that in infrastructure ($128 million). The share of agriculture spendinZ in the budget is scheduled to fall to 8 percent in 1996, 7 percent in 1997 and 6 percent in 1998.22 468. The government's investment program, the Programa Trienal de Investimento Publico (PTIP), ought to respond to the agenda outlined above. Concerning the distribution of investments within the agriculture portfolio, however, the largest claimant of investment funds was irrigation (26 percent), followed by extension (21 percent), food crops (16 percent), and capacity building (13 percent). Research received only 6.5 percent of the total. It was argued in this Memorandum that the emphasis on irrigation is overdone, given the fact that the Government's pro-smallholder policy is centered on labor-intensive and rain-fed smallholder agriculture, whereas large-perimeter irrigation is capital-intensive, and less than 40 percent of the area equipped for irrigation is actually in use. The emphasis on extension is appropriate, given its expected high returns. Furthermore, the money allocated to research is inadequate in the light of the importance of this function for technical change. More in line with the strategy outlined above would be an increase in research and extension investment and a decrease in irrigation investment. (h) The Roles of IDA, other InternationalAgencies, Donors, and NGOs 469. IDA has been a major supporter of the Government's Economic and Social Rehabilitation Program, and of its agriculture, rural development, and environment protection policies and initiatives. Currently, IDA is financing the Agricultural Services Rehabilitation and Development Project in the Northern regions, and the Rural Rehabilitation Project, centered mostly in the Center-North areas. The former supports the development of specific crops, particularly cotton, and includes support to the extension service and land reconnaissance and titling. The latter supports the rehabilitation of small-scale local infrastructure and the financing of local projects. A GEF-funded Transborder National Parks project has been approved, which will support the rehabilitation of some of the national parks and game reserves, severely destroyed by the war or their ensuing abandonment, and the establishment of compatible rules of operation to permit the eventual operational linkage of current and new parks in Mozambique with those in neighboring countries. 470. The current lending program includes an Environmental Support project, to assist the implementation of the National Environmental Action Plan formulated by the Government, with IDA support. 471. IDA was the executing agency for the nation-wide IFAD-funded Second Agricultural Rehabilitation Project (SRS-008-MoZ), which aimed at providing smallholder farmners and artisanal fishermen with inputs and means of production, and at helping the Government to integrate the provision of these inputs with improved extension services. At the request of the Government, the project was redesigned in 1989 to strengthen and integrate agricultural services in various sub-sectors (food crops, livestock, agroforestry, and other crops), and to provide tools and quality seeds. As originally conceived, the project was to be nation-wide, but due to 222 Repuiblica de Mocambique, 1996. Plano de Investimento Puiblico 1996-98. Ministerio do Plano e Finan9as, Direccao Nacional de Planificacao, p. 8. 148 implementation constraints it covered only six provinces (Maputo, Gaza, Inhambane, Manica, Nampula and Cabo Delgado). The project ended in January 1996. 472. A new IFAD-funded project is now under way in Niassa province. The objectives of the project are to improve employment, incomes and food security of the rural population. The project aims to (i) rehabilitate and develop the basic rural infrastructure; (ii) organize effective agricultural services, including extension and adaptive research; and (ii) strengthen institutional capacity for decentralized rural development and project implementation. The latter component includes tertiary roads and the major link to Nampula province. 473. Many bilateral and multilateral aid agencies operate in the agricultural and environment sectors in Mozambique. Due to a lack of coordination, their efforts are fragmented. Multiple donors fund multiple projects each, leading to a plethora of small projects, each calling for local managers, special procurement and accounting procedures, and the like. 474. There has been a recent move towards a more integrated approach to development support. There are two outstanding examples of this. The first is the Priority District Program (PDP). In 1989, the Government adopted the Priority District Program to re-establish the economic, social and administrative infrastructure of areas selected for their importance in terms of population and economic potential. The objective of the PDP was to rehabilitate the local transport network and essential social (education, health, water supply) and agricultural services in forty priority districts over a five-year period. The priority districts were selected largely on the basis of their agricultural potential. The emphasis of the PDP was on the restoration of smallholders' production and the creation of labor-intensive public works to provide employment and incomes. Another essential aspect of the PDP, in line with Government's decentralization strategy, was the strengthening of local governments in the planning, implementation and financing of economic and social development. The program was coordinated by the National Rural Development Institute (INDER) and was supported by most donors including IDA (Rural Rehabilitation Project, see para. 469). It was an ambitious program, but was one of critical importance for Government strategy for influencing medium and long- term regional development patterns. 475. The other example of an integrated approach is the Pre-Program, which was originally conceived of as a preliminary phase to eventually lead into a full-fledged integrated program of intervention on the agricultural sector. Originally developed by the Vice-Minister of Agriculture and led by the UNDP and the FAO, it soon attracted the support of IDA (several activities originally conceived as a part of the Rural Rehabilitation project were transferred to the Pre-Program) and other donors. As anticipated, the Pre-Program proved too large and complex for the limited implementation capabilities in the Government. Instead of being just a preliminary phase of something larger, it became in itself a full-fledged endeavor, which has demonstrated the advantages, and some of the challenges as well, of a jointly conceived, coordinated multi-donor effort in a country like Mozambique. 476. During 1995 and 1996 intense preparations began for the inauguration of a broad, integrated approach to development support, called the PROAGRI. Practically all the donors were involved, IDA accepting the role of donor of last resort and the FAO taking the function of coordinator on the side of the bilateral and multilateral institutions. The government took strong ownership of the concept from its inception. Plans for all the agricultural subsectors were developed, together with budgets. Institutional arrangements were given particular attention in a series of papers and high-level seminars. Appraisal is envisaged for 1997 and effectiveness for 1998. 149 477. NGOs have played a major role in providing support to communitv organization, input supply, technology transfer, and humanitarian aid. They have also played a significant role in supporting the development of small-scale projects and micro-enterprises; in protecting Mozambique's biodiversity; and in capacity building at the Government and university levels. They are expected to retain and intensify their role in these areas, particularly under the IDA- supported Rural Rehabilitation project, the GEF-funded Transborder National Parks project, and their eventual successors. (i) Timing of operations and a short-term action plan 478. It was stressed in the discussion above that there are four major priorities which the govemment could stress in its agricultural growth strategy. The first would be to reinforce macroeconomic stability. This is a precondition for the development of rural financial markets and significant increases in foreign investment. The second would be to pursue the present objectives with respect to upgrading the road system, and to extend further the efforts in respect of feeder roads, rural culverts, and farm to market tracks, possibly by means of a matching-grant Rural Investment Fund (RIF). These undertakings would reduce transport costs and help more smallholders to market a portion of their output.223 Through the second Roads and Coastal Shipping Project, these actions are already under way and are bearing fruit. Further extensions of the roads projects for the next ten years at least would be appropriate. The third leg of the strategy would be to continue the gradual reduction of the tax on exports of raw cashew nuts. The fourth priority would be to improve the land legislation, regulations and administration so as to guarantee security of tenure to smallholders and large-scale farmers. 479. Another short-term objective, in the area of trade, is to remove the bureaucratic obstacles to registration for wholesalers and retailers. This can be done within a few months. 480. Medium-term interventions include the following. e In the area of incentives, the govemment plans to eliminate all minimum pricing arrangements by July 1999. The Instituto de Cereais de Mogambique should be privatized. Development and dissemination of on-farm crop storage technologies is an important medium-term task for the research and extension agencies. A five-year crop storage program could be envisaged. * Food aid needs to be phased out. Even without the government's urgings, the donors may well withdraw all of their support in this area within another two years. In the interests of short-term food security, a unit could be created to obtain and disseminate long-lead climate forecasts, and the efficiency of the port system could be improved. * In the area of institution-building, the encouragement of the growth of rural financial institutions is an area where the government has a role as a facilitator. This is of particular importance in the smallholder cotton-growing areas, where the growth of smallholder- managed institutions is necessary for the withdrawal of the cotton-purchasing monopsonies held by the ginneries. These new financial institutions would take between five and ten years to work effectively. 223 Only 29 percent of all smallholders market any of their output at present. See Ministerio da Agricultura e Pescas (1994), as quoted by Desai (1996), Appendix Table IX. 150 * The government could consider selling off its 50 percent holdings in the cotton joint venture companies. * The research and extension services should be upgraded. As far as regulation is concerned, the rules and procedures for importing new technologies need to be defined. Methodologies for involving local populations in joint management of wildlife and scenic resources need to be developed and implemented. 151 #)Timing of operations: matrix -1997 1 19 98 19 999 20 00 III IVI I II.-III IVI I 11 III _IV I I 11 III IV PRIORITY ARiEAS Reinforce macroeconomic stability --- --- l---| Rehabilitating feeder road system I Reducing tax on raw cashew exports Security of tenure for farmers _ OTHER SHORT-TERM OBJECTIVES Ease registration for traders MEDIUM-TERM INTERVENTIONS INCENTIVE F RAMEWORK Eliminate minimum prices Privatize Instituto de Cereais de Moc,arnbique Develop crop storage technologies .,, __ Phase out food aidiI Disseminate long lead climate forecas,ts XX Improve efficiency of ports... INSTITUTION-BUILDING Facilitate smallholder financial institutions _ _ Withdraw JVC cotton monopsoniesXXX Sell government holding in jvcsEEX PUBLIC SERVICES Upgrade research and extension services__ Develop methodologies for wildlife mgt. _ _ _ 152 References Addison, A., and lain McDonald, 1995. "Rural Livelihoods and Poverty in Mozambique." Background document for "The Poverty Reduction Strategy for Mozambique" prepared by the Poverty Alleviation Unit - Ministry of Planning and Finance, February. AFIAE Division, n.d.. "Mozambique, Country Strategy, Agriculture and Environment." World Bank. Agostini, Paola, 1993. "The Forestry Sector." Working Paper prepared for the Agricultural Sector Memorandum. Southern Africa Department, Agriculture and Environment Division (AF6AE), World Bank, November. Agriculture and Environment Division, 1995. "Mozambique: Financial Aspects of the National Extension Program," Southern Africa Department, World Bank, May. Agriculture and Environment Division, 1995. "Mozarnbique: Preparing the National Extension Program," Southern Africa Department, World Bank, January. Agriculture and Environment Division, 1994. "Country Environmental Strategy Paper for the Republic of Mozambique," Southern Africa Department, World Bank, December. Antle, John M., 1983. Infrastructure and aggregate agricultural productivity: international evidence. Economic Development and Cultural Change 31 (April): 609-619. Antle, John M., 1984. Human capital, infrastructure, and the productivity of Indian rice farmers. - Journal of Development Economics 14: 163- 181. Antle, John M., and Prabhu L. Pingali, 1994. Pesticides, productivity, and farmer health: a Philippine case study. American Journal ofAgricultural Economics 76: 418-430. Aube, Thierry, 1995. "Strategy for the Promotion of Export of Fruit and Other Horticultural Products from Mozambique," Study for the World Bank, Southern Africa Department. First draft, April. 50 pages. AUSTRAL, 1993. Livestock sub-sector study: Draft Final Report. National Directory of Livestock. September. 120 pages. In conjunction with Louis Berger International, 100 Halsted St., CP 1947, Maputo. Bagchee, Aruna, 1994. "Agricultural Extension in Africa," World Bank Discussion Papers No. 231, Africa Technical Department Series, Washington, D.C. Banco de Mo,ambique, 1994. Boletim Estatistico Vol. 2, No. 6 (December). Divisao de Estatistica. Bastiaansen, A.P.M., 1991. Mission Report on Land and Water Management. Mozambique: Project MOZ/86/010: Formulation Mission. FAO. Maputo, July. Bell, Simon, 1996. Vital Statistics: A Publication of the World Bank's Mozambique Resident Mission, Vol. 3 No. 3, October quarter. Binswanger, Hans P., 1993. "Suggestions for Land Poiicy in Mozambique". Maputo, Mozambique: July. 6 pages. 153 Binswanger, Hans P., Shahidur R. Khandker, and Mark R. Rosenzweig, 1993. How infrastructure and financial institutions affect agricultural output and investment in India. Journal of Development Economics 41: 337-3 66. Borges, Klas Ernald, and Andrew Lazaro Doto, 1995. Social Survey: Uso e Gestdo de Terra do Sector Familiar, Distrito de Monapo, Provincia de Nampula: Relat6rio. PRDSA, Republica de Mocambique, Ministerio da Agricultura e Pescas, Direc,co Nacional de Geografia e Cadastro. Maputo, November. Boserup, Esther, 1965. The conditions of agricultural growth. Chicago: George Allen & Unwin. Chadenga, V., n.d. "Analysis of the Efficacy, Practicality and Cost of Various Tsetse and Trypanosomiasis Control Techniques as Used in Zimbabwe". Programme on the control of African animal trypanosomiasis and related developments. FAO Animal Production and Health Paper No. 100. Pages 105- 123. Copy supplied by Mr. Cornelis de Haan. Chausse, Jean-Paul, 1995. "Cashew Sector". Mimeo, World Bank, AFIAE. 10 pages. Chomitz, Kenneth M., and David A. Gray, 1996. Roads, land use, and deforestation: a spatial model applied to Belize. World Bank Economic Review 10/3: 487-512. Collier, Paul, Samir Radwan, Samuel Wangwe, and Albert Wagner, 1990. Labour and Poverty in Rural Tanzania Oxford: Clarendon Press. Comissao de Terras, 1996. Draft Work Plan 1996-1999. Maputo: Ministry of Agriculture and Fisheries. January 15. Commonwealth Secretariat, 1992. Mozambique: A Strategic Action Plan for the Rehabilitation of the Coconut Industry. Commonwealth Fund for Technical Co-operation: Industrial Development Unit. October. IDU/SCFM/15. Conway, John A., J. Coulter, and B. Trotter, 1994. "A Review of National Objectives for Establishing Food Security Reserves in Southern Africa - A Case Study in Mozambique." Natural Resources Institute, Kent, England, February. Coulter, Jonathan P., 1995. Maize Marketing and Pricing Study -Mozambique. Chatham, England: Overseas Development Administration, and Natural Resources Institute. April. Report No. R2247 (C). 69 pages. del Ninno, Carlo, and David E. Sahn, 1993. The determinants of poverty and income distribution in Maputo. Cornell University, Food and Nutrition Policy Program. April. Mimeo, 33 pages. de Voest, Christine, 1993. "Food Aid and Food Security". Working paper prepared for the Agricultural Sector Memorandum. Southern Africa Department, Agriculture and Environment Division (AF6AE), World Bank. Desai, Jaikishan, 1996. Mozambique: Rural Poverty Profile. Maputo: Ministry of Planning and Finance, Poverty Alleviation Unit. April. Direc,co Nacional de Estradas e Pontes, 1996. Feeder Roads Programme: Annual Progress Report 1995. Volume I - Text. Repuiblica de Mo9ambique, Ministerio das Obras Publicas e Habitacao. Maputo, February. Direc9ao Nacional de Estradas e Pontes, 1996. Trafego nas estradas de Mo ambique, 1995. Ministerio das Obras Puiblicas e Habita$ao, and Sweroad Mo9ambique. June. 154 Donovan, Cynthia, 1996. Effects of monetized food aid on local maize prices in Mozambique. Unpublished Ph.D. dissertation, Michigan State University, Department of Agricultural Economics. 270 pages. Dorosh, Paul A. and R. Bemier, 1994. Agricultural and Food Policy Issues in Mozambique: A Multi-Market Analysis. CFNPP Working Paper Number 93, March. Dorosh, Carlo del Ninno, and David E. Sahn, 1995. Poverty alleviation in Mozambique: a multi- market analysis of the role of food aid. Agricultural Economics 13: 89-99. Economist Intelligence Unit, 1996. Mozambique. Country Profile, 1995-96. Fok, Michel, 1995. The Cotton Sub-sector in Mozambique: Institutional Diversity, Performance and Prospectsfor Improvement. CIRAD-CA, Montpellier, France, May. Main volume: 127 pages. Volume of Annexes: 60 pages. Food and Agriculture Organization/INIA (Instituto Nacional de Investigagao Agraria), 1980. Mozambique: Land Use. Food and Agriculture Organization, 1990. Fisheries Country Profile 1990. Food and Agriculture Organization, 1993. Fisheries Yearbook 1993. Food and Agriculture Organization, 1995. "Mozambique, Food and Wildlife Strategy Development: Issues and Options," Working Paper CP MOz 37, Rome, Italy, February. Fortes, Carlos, Dannielson, T., Johannessen, B., and Dalton, M., 1994. "Report of the Evaluation Mission Regarding the Feeder Roads Programme," Final Report, Presented to the UNDP, Maputo, April 29. Friedman, Milton, 1968. The role of monetary policy. American Economic Review 58 (March): 1-17. Gibberd, V., John Rook, C.B. Sear, J.B. Williams, 1995. Drought risk management in Southern Africa: The potential of long lead climate forecasts for improved drought management. Draft 1.5, 13/11. November. Natural Resources Institute, Chatham Maritime, Kent, ME4 4TB, UK. Gilbert, Christopher L, 1993. "Domestic Price Stabilization Schemes for Developing Countries". In: Stijn Claessens and Ronald C. Duncan, eds. Managing Commodity Price Risk in Developing Countries. Baltimore: Johns Hopkins and World Bank. 466 pages. Pages 31-66. Haas, Pierre, 1993. Revue du secteur agricole: Renforcement des Organisations Paysannes. Mimeo, December 2. Working paper prepared for the Agricultural Sector Memorandum. Southern Africa Department, Agriculture and Environment Division (AF6AE), World Bank Hernandez, R.A., 1991. Bioeconomia del cultivo de camaron. Informe, Pilot Project for Coastal Shrimp Culture, Mozambique, MOZ/86/033, Roma, 1991. Hilmarsson, Hilmar, 1995. "Cashew Pricing and Marketing in Mozambique." Second Draft. Mimeo, 30 pages + 7-page questionnaire. ILCA, 1988. "Livestock Production in Tse-tse Affected Areas of Africa". Proceedings of an ILCVA/ILRAD meeting. ILRAD, Nairobi. Copy of a paper on costs supplied by Mr. Comelis de Haan. Inquerito de Indicadores Mudtiplos (Multiple Indicators Survey). Executed by UNICEF. Direccao Nacional de Estatistica. 1996. Data refer to July/August 1995. 155 Inquerito Nacional de Demografla 1991 (National Demographic Survey). Executed by the Direc,ao Nacional de Estatistica. Data refer to October 1991. Instituto de Algodao de Mo,ambique, 1993. "Campanha 92/93. Posicao ate 16.09.93." Mimeo, I page. By Mr. Eugenio Fazenda Gove. Instituto Nacional de Investigagao Agron6mica de Maputo (MOA), 1994. INIA "Research Operational Plan 1994-2000," January. Instituto Nacional de Investiga9ao Agron6mica de Maputo (MOA), 1993. INIA "INIA Medium- Term Strategy and Plan, Draft, May. Jackson, J.C. and A.P. Cheater, 1989. "Contract Farnning: Sugar, Tea and Cotton in Zimbabwe." Eastern Africa Economic Review 5: 97-107. Jaffee, Stephen, 1993. "Mission report on export-oriented agro-industries in Mozambique." Mimeo, 50 pages, World Bank, Division AF6AE. October. Koester, U., 1986. Regional Cooperation to Improve Food Security in Southern and Eastern African Countries. Washington: IFPRI, Research Report 53. Kumar, E. Raghu, 1995. "The Cashew Sector in Mozambique: Assessment and Perspectives". London: Mesira Limited. Prepared for AF1AE (Southern Africa Department), World Bank. 95 pages + 5 annexes, 1 vol. August. Kyle, Steven, 1991. "Economic Reform and Armed Conflict in Mozambique", World Development 19(6):637-649. Lele, Uma, and Manmohan Agarwal, 1989. Smallholder and large-scale agriculture in Africa: are there tradeoffs between growth and equity? MAIA Discussion Paper 6. Washington, D.C.: World Bank. 20 pages. Louis Berger International Inc. and Austral, C. P. Lda., 1993. "Livestock Sub-Sector Study," Draft Final Report, Maputo, Mozambique, September. MacDonald, Sir M. and Partners Ltd., and Hidroprojecto Consultores de Hidraulica e Salubridade, S.A., 1990. Sub-Saharan Africa Hydrological Assessment: SADCC Countries: Countrv Report: Mozambique. December. The World Bank and the United Nations Development Programme. McGregor, JoAnn, 1995. "Staking Their Claims: Land Disputes in Southern Mozambique". Oxford: St. Anthony's College. Mimeo, 25 pages. A University of Wisconsin-Madison and Ministry of Agriculture project. Medecins Sans Frontieres, 1996. "Household Food Security Surveys". Boletim da Seguranca Alimentar, Arz'mero Especial 38 (January): 47-57. Menestrel, H. Le, 1995. "Trade Facilitation," Report to the World Bank, Draft, Republic of Mozambique, Ministry of Agriculture, Maputo, April. 120 pages. Mihajiovich, D., and F. Gomes, 1986. Serie Terra eAgua do Instituto Nacional de Investigaqao Agron6mica: Areas de Rega: Inventdrio e Possibilidades Futuras. Maputo. Comunicacao No. 53. Ministerio da Agricultura e Pescas, 1993. "The Pricing and Distribution of Yellow Maize Food Aid in Mozambique: An Analysis of Alternatives," Executive Summary. Relat6rio Preliminar No. 16, Republica de Mocambique, September. 156 Ministerio da Agricultura e Pescas, 1994. Levantamento Agricola (Agricultural Survey). Direccao de Economia Agraria. Ministerio da Agricultura e Pescas, 1995a. "Politica Nacional de Terras e Estrategia de ImplementavAo." August. Mimeo, 19 pages. Minist6rio da Agricultura e Pescas, 1995b. "Revisao da pauta aduaneira: Proposta do Ministerio da Agricultura e Pescas. August. LD/DEA. Ministerio da Agricultura e Pescas, 1995c. "Proposta de politica de pre,os ao produtor para produtos agricolas". September. LD/DEA. Ministerio da Agricultura e Pescas, 1995d. "Politica agraria e estrategia de implementa,ao". Programa do Governo para 1995/1999. Projecto No. 109/CM/95 de 20 de Outubro. Ministerio da Agricultura e Pescas, 1995e. "Diagn6stico da estrutura, comportamento, e desempenho dos mercados alimentares rurais de Mo,ambique". Departamento de Estatistica e Projecto de Seguranca Alimentar da Direc,co de Economia Agraria, Equipa de Pesquisa. Relat6rio Preliminar No. 19, 4 de Julho. Ministerio da Educa,co, 1986. Atlas Geografico. Volume 1, Repuiblica Popular de Mocambique. Ministerio da Sa6de, 1994. Informa.do EstatisticaAnual - 1992. Departamento de Informacao para a Sauide, Direccao Nacional de Planifica,co e Cooperacao. Maputo, August. Ministry of Agriculture and Fisheries, 1995. "National Program of Forestry and Wildlife (1995- 2000)," Maputo, May. MOA/MSU/UA Research Team, 1992. Land access and land use among smallholders in Nanpula province. Maputo, Mozambique. Mimeo. Moll, Peter G., 1993. "Farm Efficiency in Mozamnbique." Working paper prepared for the Agricultural Sector Memorandum. Mimeo, AF6AE (Southern Africa Department, Agriculture and Environment Division), December 8. Moll, Peter G., 1994. "Contribution to the section on marketing in the Agricultural Sector Memorandum". Mimeo, World Bank (AF6AE, Southern Africa Department, Agriculture and Environment Division), March 31. 26 pages. Mozambique, Government of, 1994. "The Poverty Reduction Strategy for Mozambique," Draft for Discussion, Poverty Alleviation Unit, National Planning Unit, December. Mozambique, Government of, 1995. "Mozambique: Emergency and Drought Situation," March. Multiple Indicator Cluster Survey, Mozambique - 1995. Government of Mozambique and United Nations Children's Fund. Ministry of Planning and Finance, Department of Statistics; Section of Planning, Monitoring and Evaluation. 27 pp. + 7 annexes. Myers, Gregory W., 1993. "Land Tenure Issues in Post-War Mozambique: Constraints and Conflicts," Land Tenure Center, University of Wisconsin-Madison, April. Myers, Gregory W., 1994. "Competitive Rights, Competitive Claims: Land Access in Mozambique". Journal of Southern African Studies 20/4: 603-633. Myers, Gregory W., J. Eliseu and E. Nhachungue, 1993. Security and Conflict in Mozambique. Case Studies ofLandAccess in the Post War Period Land Tenure Center, University of Wisconsin-Madison. 157 Myers, Gregory W., and P. West, 1993. Land Tenure Security and State Farm Divestiture in Mozambique: Case Studies in Nhamatanda, Manica and Montepuez districts. Land Tenure Center Research Paper, No. 110. Madison: University of Wisconsin. National Plan Commission, 1995. Produqao e Valor Acrescentado. February 8, 1995. Newitt, Malyn, 1995. A History of Mozambique. Indiana University Press, Bloomington and Indianapolis. Noronha, Raymond, 1994. "Mozambique: Land Tenure and Use: Issues and Options". Mimeo, June. Working paper prepared for the Agricultural Sector Memorandum. Southern Africa Department, Agriculture and Environment Division (AF6AE), World Bank. Olivares, Jose, 1995. "Mozambique ASAP: Status Report," Memo to J.P. Chausse, March 17. Paiva, Adela B., 1993. "The Financing of the Agricultural Sector." Working paper prepared for the Agricultural Sector Memorandum. Southern Africa Department, Agriculture and Environment Division (AF6AE), World Bank. Paiva, Adelina B., and E. Luzio, 1994. "Mozambique: Public Investments in Agriculture, Preparing the Integrated National Investment Programme," Draft, Southern Africa Department, World Bank, October. Pingali, Prabhu L., and Cynthia B. Marquez, 1994. Pesticides and Philippine rice farmer health: a medical and economic analysis. American Journal ofAgricultural Economics 76: 587- 592. Republica de Mo9ambique, Conselho de Ministros, 1995. Politica agraria e estrategia de implementa,ao. Projecto No. 109/cM/95 de 20 de Outubro. Republica de Mocambique, 1995b. Plano Econ6mico e Social e Politica Or amental para 1996. Maputo: Novembro de 1995. Republica de Mogambique, 1996. Plano de Investimento Pzblico 1996-98. Ministerio do Plano e Finan9as, Direcgao Nacional de Planificagao. Republica de Mo9ambique, 1997. "Produ,cao e Valor Acrescentado: Valores de 1996 e Plano de 1997". Ministerio do Plano e Finan9as, Direc9ao Nacional de Planifica9ao. 25 de Mar9o de 1997. Rozell, Gale, 1993. "Private Sector Environment and Privatization". Working paper prepared for the Agricultural Sector Memorandum. Southern Africa Department, Agriculture and Environment Division (AF6AE), World Bank. SABAS, 1995. Approtech Products Limited, Toronto, Ontario, Canada; and Consultancy Services India Pvt. Ltd., Bombay, India: Consulting Engineers and Project Managers. 20 pages, company prospectus. Sahn, David E., and Carlo del Ninno, 1994. Living standards and the determinants of poverty and income distribution in Maputo, Mozambique. Working Paper 56. Cornell Food and Nutrition Policy Program. Shrimp Notes: A Market News Analysis, May 1 Oth and June 11 t, 1996. Secretaria de Estado de Hidriulica Agricola, 1993. National Irrigation Development Master Plan (NIDMP), Umbeluzi, Incomati, Limpopo, Buzi and Pungoe Basins: Final Report. April. Simpson, S. Rowton, 1976. Land law and registration. Cambridge University Press. 158 SOGREAH Ingenierie and HIDROGEST Consultoria e Gestao, 1993. National Irrigation Development Master Plan. Final Report. (Annex 5: Agronomy and Animal Production.) Southern Africa Department, 1995. "Memorandum of the President of the International Development Association to the Executive Directors on a Country Assistance Strategy of the World Bank Group for the Republic of Mozambique." June 6. 39 pages. Strachan, Lloyd W., 1994. "Agricultural Marketing and Input Supply." Working paper prepared for the Mozambique Agriculture Sector Memorandum, January. Mimeo, 75 pages. Southern Africa Department, Agriculture and Environment Division (AF6AE), World Bank. Tanner, Christopher, 1995. "Os Direitos das Comunidades acerca do Accesso e Uso de Terra e de Outros Recursos Naturais". Documento de Apoio, Projeto GEF - Areas Transfronteiras de Conserva,co e Capacita9ao Institucional. Cambridge SEPR Associates. September. Mimeo, 13 pages. Tembe, Jorge Mabay, 1993. "Os Sistemas de Producao do Sector Familiar em Mocambique," Paper produced for the Mozambique Agricultural Sector Memorandum, November. Southern Africa Department, Agriculture and Environment Division (AF6AE), World Bank. Tsakok, Isabelle, 1990. Agricultural Price Policy: A Practitioner 's Guide to Partial-Equilibrium Analysis. Ithaca and London: Cornell University Press. Tschirley, David, Cynthia Donovan, and Michael T. Weber, 1996. Food aid and food markets: lessons from Mozambique. Forthcoming, Food Policy. Udry, Christopher, 1990. Credit markets in northern Nigeria: credit as insurance in a rural economy. World Bank Economic Review 4/3: 251-269. UNDP/DAH (United Nations Development Program), 1996. Boletim de Desminagem. Departamento do Plano, PNUD/DAH Programa Acelerado de Desminagem. Edicao N° 9 (Junho/Julho). UNOHAC (United Nations Office for Humanitarian Assistance Coordination), 1994. Serie de Mapas do UNOHAC (UNOHAC Map Series). September. United Nations Department for Humanitarian Affairs. United Nations High Commissioner for Refugees, 1994. Reintegration Strategy," Draft Working Paper, Mozambique, December. Vietmeyer, Noel D., 1996. Lost crops of Africa. Volume 1: Grains. Washington, D.C.: National Academy Press. Board on Science and Technology for International Development, National Research Council. 250 pages. von Braun, Joachim, Hartwig de Haen, and Jurgen Blanken, 1991. Commercialization of agriculture under population pressure: Effects on productions, consumption and nutrition in Rwanda. IFPRI, Research Report 85. World Bank, 1988. Mozambique Agricultural Sector Survey. Agriculture Division, Southern Africa Department, Report No. 7094-Moz, May. World Bank, 1990a. Mozambique Restoring Rural Production and Trade, Volume I. Main Report. Agriculture Division, Southern Africa Department, Report No. 8370-Moz, October. World Bank, 1990b. Mozambique Restoring Rural Production and Trade, Volume 11: Statistical Appendices. Agriculture Division, Southern Africa Department, Report No. 8370-Moz, October. 159 World Bank, 1990c. Staff appraisal report: Uganda Livestock Services Project. Agriculture Operations Division, Eastern Africa Department. July 30. Report No. 8486-UG, CR 21 76-UG. World Bank, 1991. Fisheries and Aquaculture Research Capabilities and Needs in Africa: Studies of Kenya, Malawi, Mozambique, Zimbabwe, Mauritania, Morocco, and SenegaL World Bank Technical Paper No. 149, Fisheries Series. World Bank, 1993. Mozambique Capacity Building Study. Population and Human Resources Division, Southern Africa Department, Report No. I 000 I-MOZ, June. World Bank, 1994. Various documents providing statistics on cashew, Secretariado de Estado de Caju. World Bank, 1994b. StaffAppraisal Report: The Republic of Mozambique: Second Roads and Coastal Shipping Project. March 15. Report No. 125 80-MOZ. Infrastructure Operations Division, Southern Africa Department. World Bank, 1996a. Mozambique: Third Economic Recovery Credit: Initiating Memorandum. Mimeo, Southern Africa Department, World Bank, June 14, 1996. World Bank, 1996b. Mozambique: Economic Recovery Project Implementation Completion Mission, Back to Office Report, Rene Bonnel, May 1. World Bank, mimeo. World Bank, 1996c. Mozambique: Memorandum on Agricultural Land Issues. World Bank, Washington, D.C. Southern Africa department, AFIAE. Mimeo, April 29. World Development Indicators 1995. Washington, D.C.: World Bank. Machine-readable tabular system. World Food Programme, 1995. "Food Aid 1994/95 Mozambique," Maputo. Zartman, I. William, 1995. Collapsed States: The Disintegration and Restoration of Legitimate Authority. Lynne Reiner Publishers, Boulder. 160 Appendix 1: Tables Appendix Table I. Selected Macroeconomic Indicators 1989 1990 1991 1992 1993 1994 1995 1996 GDP (US$ mill., current prices)a - 1443 1433 1285 1467 1462 1484 1739 Real growth rate of GDP, market - 1.0 4.9 -0.8 19.3 5.0 1.4 6.4 prices' Growth of M2 (%) - 37 30 53 63 49 56 11 Government investmentlGDP (%) - 24 23 22 20 24 21 17 Private investment/GDP (%) - 22 26 31 40 36 33 31 Inflation rate (%)b - 49 33 45 42 63 55 17 Bank rediscount rate (%) 10 14 28 39 45 55 70 24.75 c Real rediscount rate (%) - -35 -5 -6 3 -8 14 7.75 Borrowers' interest rate (%)d 22 22 32 42 46 58 - - Real borrowers' interest rate (%) - -27 -1 -3 4 -5 - - Official/parallel exchange rate - - - 14 19 9 10 9 differential, July (%)e Real exchange rate (US$/Mt)f - 100 83 66 60 61 61 69 Exports (GNFS) (US$ mill) 200 229 310 304 312 341 411 479 Real growth of exports (%) h 5.8 12 28 -1.2 3.4 7.3 -2.2 16.5 Imports (GNFS) (US$ mill) 875 948 1001 970 1083 1155 965 973 Real growth of imports (%) 10.9 0.0 1.9 -5.9 8.5 4.1 -25 0.8 Resource balance (US$ mill) -675 -719 -691 -666 -771 -814 -553 -494 Net current transfers (US$ mill)' - 168 163 168 185 192 159 - Net pvte foreign invest. (US$ mill) 3.4 9.2 22.5 25.3 32.0 35.0 45 - Public finance: Current revenues (% of GDP) - 22 22 21 20 18 18 18 Current expenditures (% of GDP) - 26 22 24 21 23 17 - Current account surplus (% of GDP) - -3.3 -0.5 -3.1 -1.4 -5.2 1.2 - Capital expenditures (% of GDP) - 24 24 23 21 25 21 - Foreign financing (% of GDP) - 28 26 24 21 31 22 - Sources: 1989-1994: Southem Africa Department (1995), Annex A.5, unless otherwise stated. 1995: Wolrd Bank (1996a), Annex 1. Interest rates and bank rediscount rates from Banco de Moqambique (1994). 1996: Bell (1996), and documents from the World Bank/IMF mission. a The discrepancy between the levels and growth rates is present in the source documents. bCalculated from the consumer price index. 1996 inflation rate: annualized rate based upon the change in the current price index between December 1995 and August 1996; data from the Direccao Nacional do Piano. ' End of 1996. See Banco de Mo,ambique, Aviso No. Ol/GGBM/97. dAssumed to be an agricultural borrower (whose interest rate was lower) between 1989 and 1992. After 1992 no distinction was made between sectors. Interest rates were free after July 1994, and their levels are not known. Source: Banco de Mo,ambique (1994). ' Source for exchange rate differential: Bell (1996). f An increase in US$1Mt denotes appreciation. Method: MtlUS exchange rates and US and Mozainbique CPls. Use of the South African equivalents produced very similar results. Sources: CPls of SA & USA and Mt/$ and R/$ exchange rates from International Financial Statistics,. I GNFS = Goods and nonfactor services. hReal growth of exports, imports from: Southem African Department (1995), "Mozambique - National Accounts", Part D. iIncluding official current transfers. Source: World Bank (1996a), Annex 1, p. 2. 161 Appendix Table II. Commodity Composition of Exports, 1989-94 (in millions of U.S. dollars) Item 1989 1990 1991 1992 1993 1994 1995 1996 Total exports, f.o.b. 105.0 126.5 162.5 139.3 151.8 168.4 174.3 225.9 Prawns 39.3 43.4 60.8 54.5 88.8 82.7 73.1 79.7 Cashew nuts 20.1 14.3 16.0 17.6 8.2 3.3 9.5 12.0 Cotton 7.5 8.7 8.8 10.8 11.1 18.9 19.8 12.8 Tea 0.1 0.5 0.8 0.2 0.0 0.4 0.0 0.0 Sugar 5.3 7.9 9.8 6.7 0.0 11.0 7.3 12.8 Copra 1.9 2.6 4.7 4.2 2.5 3.4 6.1 2.4 Citrus 3.3 1.9 1.9 1.1 0.9 1.3 1.3 1.0 Petroleum 9.5 10.2 7.2 4.4 6.7 13.8 4.5 2.6 Coal 0.4 0.4 0.2 0.1 0.0 0.2 0.5 0.4 Miscellaneous 17.5 36.6 52.4 39.7 33.6 33.4 52.1 102.1 Source: Bank of Mozambique: Ministry of Commerce: Planning Commission, January 1995; and 1996 papers by National Directorate of Statistics, Informa,ao Estadistica, and the Bank of Mozambique. 162 Appendix Table m. Shrimp and prawn exports by Mozambique, 1990-1995 1990 1991 1992 1993 1994 1995 Exports (MT) 6,204 7,743 8,174 8,460 N/A N/A Exports (value in 1,000 us$) 50,349 62,981 65,985 70,386 N/A. 84,OOOa % of world exports (values) 0,45 0,51 0,52 0,54 N/A. N/A. % of world exports (quantity) 0,27 0,31 0,31 0,30 N/A. N/A Sources: FAO (1993), unless otherwise indicated. Note that these figures, and in particular the export ones, are underestimated, owing to clandestine trade. a From: Estimates by Luigi Marcuccio on the basis of the data available in Republica de Mo9ambique (1995). 163 Appendix Table IV. Nominal and Real Producer Prices, 1976-92 Maize Rice Beans Groundnut Sunflower Cotton Copra Nom Real Nom Real Nom Real Nom Real Nom Real Nom Real Nom Real 1976 2.5 2.6 5.0 5.3 6.5 6.9 8.5 9.0 7.0 7.4 6.5 6.9 3.2 3.4 1977 3.2 3.3 6.2 6.4 10.0 10.4 10.0 10.4 8.5 8.8 6.5 6.8 4.7 4.9 1978 3.2 3.3 6.2 6.4 10.0 10.3 10.0 10.3 8.5 8.7 6.5 6.7 4.8 4.9 1979 3.2 3.2 6.2 6.3 11.0 11.1 10.0 10.1 8.5 8.6 6.5 6.6 4.9 4.9 1980 4.0 4.0 6.2 6.2 15.0 15.0 10.0 10.0 8.5 8.5 11.0 11.0 5.0 5.0 1981 4.0 3.9 6.2 6.0 15.0 14.5 13.5 13.0 8.5 8.2 11.0 10.6 5.0 4.8 1982 6.0 4.9 10.0 8.2 15.0 12.3 15.0 12.3 10.5 8.6 11.0 9.0 5.0 4.1 1983 6.0 4.4 10.0 7.3 15.0 11.0 15.0 11.0 10.5 7.7 12.5 9.1 5.0 3.7 1984 6.0 3.8 10.0 6.4 15.0 9.5 15.0 9.5 10.5 6.7 12.5 8.0 5.5 3.5 1985 13.0 5.6 16.0 6.9 23.5 10.1 20.0 8.6 15.0 6.5 16.0 6.9 5.5 2.4 1986 13.0 4.0 16.0 5.0 23.5 7.3 20.0 6.2 15.0 4.7 16.0 5.0 5.5 1.7 1987 40.0 4.7 48.0 5.7 100 11.8 100 11.8 50.0 5.9 65.0 7.7 18.0 2.1- 1988 65.0 5.1 75.0 5.9 150 11.8 150 11.8 75.0 5.9 104 8.2 30.0 2.4 1989 110 6.2 145 8.1 230 12.9 255 14.3 130 7.3 175 9.0 100 5.6 1990 126 4.8 167 6.4 264 10.1 295 11.3 150 5.7 201 7.7 115 4.4 1991 190 5.4 256 7.2 400 11.3 440 12.4 250 7.1 320 9.0 165 4.7 1992 275 6.1 375 8.3 680 15.1 640 142 348 7.7 478 10.6 210 4.6 Source: Dorosh and Bernier (1994). Real prices are in constant 1980 Meticais. 164 Appendix Table V. Marketed agricultural production, all crops, 1975-94 ('000 metric tons) 1975 1980 1985 1990 1991 1992 1993 1994 Food crops Maize 95.0 65.0 68.9 97.0 74.0 67.8 142.7 148.8 Rice 94.0 43.6 33.4 25.5 23.9 8.0 17.8 26.8 Sorghum n.a. 0.6 2.1 1.4 2.6 1.3 2.4 2.1 Beans 14.8 9.6 4.6 16.6 14.2 12.4 23.3 15.7 Cassava n.a. 8.8 12.5 28.9 20.7 20.3 31.7 31.3 Groundnuts n.a. 6.3 2.3 5.1 6.6 8;6 14.9 9.2 Vegetables 4.0 6.4 39.6 34.6 35.1 35.5 42.5 44.1 Export crops Cashew 160.0 87.6 30.4 22.5 31.1 54.2 23.9 35.4 Copra 50.4 37.1 24.0 31.2 24.8 16.9 23.6 30.0 Cottonseed 52.0 64.9 6.0 29.7 40.0 49.8 47.0 49.5 Sisal 340.0 298.0 100.8 45.3 24.8 24.8 24.0 24.0 Tea 59.1 90.2 25.0 4.3 4.9 1.0 1.7 1.5 Other crops Sunflower n.a. 11.8 5.7 1.1 1.5 0.4 0.2 0.7 Sesame n.a. 0.0 0.3 n.a. n.a. n.a. n.a. n.a. Mafurraa n.a. 0.0 2.6 1.3 1.3 2.3 1.1 0.5 Tobacco n.a. 1.4 0.3 0.2 0.5 0.3 0.6 0.7 Tomato n.a. 6.4 13.4 19.6 12.4 10.4 16.6 16.8 aA tree crop, common in Inhambane. Its sweet-tasting fruit is eaten and the oil from the seed is used for tallow and soap. Source: Informacao Estatistica, CNP: Various years. 165 Appendix Table VI. Sources of income of rural households Province % of households which % without non- % with non- do not sell any farm agricultural income agricultural income output (1994) (1994) (1991) Niassa 70.8 83.5 9.6 Cabo Delgado 63.3 87.1 12.1 Nampula 59.3 76.8 9.3 Zambezia 60.7 87.4 20.3 Tete 79.8 73.1 42.5 Manica 66.3 87.9 29.7 Sofala 79.1 80.2 35.2 Inhambane 57.4 65.9 19.3 Gaza 78.2 76.9 32.5 Maputo 77.8 53.3 59.8 Total 68.3 77.2 24.1 Source: National Demographic Survey by DNE, Ministry of Planning and Finance, 1991; and Ministerio da Agricultura e Pescas (1994), as quoted in Desai (1996), Table 6, p. 22. 166 Appendix Table VII. Marketing of principal food crops: percent of farming households which sell maize, rice, etc. Maize Rice Tubers Sorghum Beans Peanuts Any food /Millet crop Regions North 17.9 13.6 3.5 0.8 21.8 24.1 33.3 Central 17.6 6.8 11.9 4.2 16.3 22.5 29.8 South 9.8 8.2 21.2 0.0 3.3 3.6 20.7 Land area possessed by the household (ha) <1.0 9.8 7.9 12.6 0.8 12.1 8.6 20.2 1.0-1.5 13.7 9.0 11.0 0.9 13.1 18.3 26.7 1.5-2.0 15.8 9.0 5.5 3.6 19.4 15.6 28.8 2.0-3.0 13.9 10.4 9.0 2.1 14.1 20.7 29.3 >3.0 25.6 12.2 11.4 5.4 14.0 16.5 40.9 Number of adults in the household 0 10.4 16.7 0.0 0.0 17.4 18.2 21.4 1 14.6 3.3 7.6 0.0 14.2 15.1 23.1 2 16.7 11.2 7.7 1.9 17.1 20.7 30.9 3 15.0 11.1 13.9 2.2 9.9 14.3 27.2 4 14.9 9.1 24.1 9.8 11.5 13.0 33.7 5+ 15.2 0.0 20.3 7.0 1.6 0.0 24.5 Gender of household head Male 15.9 9.8 10.3 3.1 14.8 17.2 29.7 Female 14.5 7.1 11.7 0.0 9.9 12.7 23.4 Household has off-farm income No 16.3 9.0 9.8 2.8 15.4 18.8 29.4 Yes 13.4 11.3 12.8 1.9 9.4 8.9 25.4 Total 15.6 9.4 10.5 2.6 14.0 16.5 28.6 Source: Ministerio da Agricultura e Pescas (1994), as quoted in Desai (1996), Table IX. 167 Appendix Table VIII. Percent of households which own livestock and mean numbers owned Mean number owned Percent of households which own at least one Goats, Goats, At least sheep, sheep one Region Cattle pigs Chicken Others Cattle pigs Chicken Others animal North 0.03 1.32 5.47 0.20 0.5 28.1 53.2 3.0 63.2 Center 0.39 1.58 5.66 0.18 6.4 27.6 51.6 2.6 63.6 South 0.55 1.98 4.60 0.17 12.4 32.3 52.1 3.3 64.0 Total 0.33 1.62 5.22 0.18 6.5 29.2 52.2 3.0 63.6 Source: Ministerio da Agricultura (1994), as quoted in Desai (1996), p. 17, Table 2. 168 Appendix Table IX. Distribution of land area possessed by households (ha) Mean Sample No land 0-1 1-1.5 1.5-2.0 2.0-3.0 >3.0 Total hectares size Region s North 1.6 18 19 16 23 22 100 2.0 807 Center 1.4 36 18 13 15 16 100 1.7 1103 South 4.8 29 18 8.8 17 23 100 2.0 839 Gender of household head Male 2.2 25 18 13 19 23 100 2.0 2232 Female 3.7 44 21 11 11 9.1 100 1.3 515 Number of adults in the household 0 2.5 46 20 10 11 10 100 1.2 80 1 4.2 42 23 12 9.6 9.4 100 1.3 427 2 2.1 28 19 14 19 17 100 1.8 1322 3 2.0 28 15 11 21 23 100 2.1 477 4 1.7 20 13 12 24 30 100 2.4 230 5+ 2.9 13 14 9.5 17 44 100 3.0 243 Household has off-farn income No 1.9 27 18 13 19 21 100 2.0 2121 Yes 4.5 33 19 13 16 16 100 1.7 627 Total 2.5 29 18 13 18 20 100 1.9 2749 Source: Ministerio da Agricultura (1994), as quoted by Desai (1996), Table 11. 169 Appendix Table X. Classification of vulnerability of communities, by diet, 1995 District Ate nothing Ate only fruit, the day before vegetables or wild (%) food the day before (%) Macossa (Manica) 1.5 8.1 Massingir (Gaza) 0.5 6.0 Chigubo (Gaza) 1.0 20.5 Mecula Sede (not rural, 0.0 5.7 Niassa) Milange-Mulumi 0.0 0.5 (Zambezia) Panda (Inhambane) 1.3 5.7 Mabote (Inhambane) 3.0 21.0 Guro (Manica) 1.5 22.1 Source: Medecins Sans Frontieres (1996), Table 12, p. 57. 170 Appendix Table XI. Mean months of food insecurity reported by rural households North Central South Overall Land area possessed by households (ha) No land 2.7 5.8 5.7 5.1 <1.0 2.5 3.4 4.8 3.7 1.0-1.5 2.4 3.4 4.9 3.6 1.5-2.0 2.4 3.2 4.6 3.2 2.0-3.0 2.2 3.6 5.5 3.7 >3.0 2.1 3.3 5.2 3.8 Number of adults in the household 0 2.4 5.5 6.5 5.5 1 2.0 3.6 5.3 3.7 2 2.4 3.3 4.7 3.3 3 2.3 3.7 5.4 4.0 4 2.3 3.1 5.0 4.0 5+ 2.4 4.1 4.6 4.1 Gender of household head Male 2.3 3.3 4.9 3.5 Female 2.5 3.8 5.5 4.3 Household has off-farm income No 2.3 3.2 5.1 3.5 Yes 2.5 4.3 5.0 4.2 Number of children 0-4 years old in the household 0 2.3 3.4 5.3 3.9 1 2.5 3.2 4.9 3.5 2 2.2 3.6 4.1 3.3 3+ 2.3 3.9 5.1 3.8 Total 2.3 3.4 5.0 3.7 Source: Ministerio da Agricultura e Pescas (1994), quoted in Desai (1996), Table XIV. 171 Appendix Table XII. Percentage moderately and severely undernourished under five year old children, in terms of wastedness (VW/H), general malnutrition (W/A) and stuntedness (WA) Total children Valid cases W/H (%) W/A (%) H/A (%) Province <5 yrs W/Ha W/A H/A Mod.' Severe Mod. Severe Mod. Severe b Maputo City 441 243 348 242 3 0 8 4 18 18 Maputo Prov. 373 176 303 176 3 0 15 2 19 15 Gaza 514 352 383 351 8 0 25 2 32 18 Inhambane 373 289 318 281 2 0 12 11 29 27 Sofala 489 384 420 378 8 0 14 9 34 19 Manica 425 250 295 247 3 3 20 4 20 32 Tete 505 395 436 380 3 0 21 10 23 35 Zambezia 189 132 143 127 12 10 22 28 21 53 Nampula 396 301 321 294 1 2 10 8 23 26 C. Delgado 417 285 351 265 4 1 22 8 25 33 Niassa 464 318 329 297 3 1 16 10 27 30 All 4586 3125 3647 3038 5 3 16 11 24 31 Source: Multiple Indicator Cluster Survey (1996), p. 22, Table 6. a W/H = weight for height, W/A = weight for age, H/A = height for age. b Moderate = between 2 and 3 standard deviations below the international mean. Severe: more than 3 standard deviations below the international mean. 172 Appendix Table XIII. Land concessions, 1986-93 a Central Provincial Reported cases Concessions Concessions of concessions Total A. Land Concessions. 1986- 199.3 Agriculture 792,000 606,000c no data d 1,398,000 Joint Ventures (direct) 92,000 e N/A N/A 92,000 Hunting 500,000 no data no data' 500,000 Tourism no data no data 300,000 300,000 Mining ' 5,597,000 no data 1 00,000 J 5,697,000 Subtotal 6,981,000 606,000 400,000 7,987,000 B. State farm sector k 600,000 C. Existing private aericultural 500,000 enterprises1 Total distribution of land 9,087,000 rights a The estimates are deliberately conservative, omitting (a) joint venture companies' marketing concession areas (see note e); (c) concessions in Renamo-administered areas for lack of information; and (d) forestry concessions of 1.5 million ha, including one of I million ha in Cabo Delgado, because most of these are drawn from areas suited for forestry and compete only minimally with smallholder agriculture. bReported in the Boletim (68,000) and by DINAGECA, Maputo (724,000). ' Partial data from selected districts in five provinces. dThere are unconfirmed reported cases amounting to 1.5 million ha, including estimates of concessions granted at the provincial level in Cabo Delgado, Tete, Zambezia, Sofala and Niassa. ' Ministry of Agriculture data. Only the "estate" or "direct farming" part is reported. The larger marketing concession areas (areas de influencia), amounting to some 2 million ha, are ignored because these relate to monopsony marketing rights and not to land per se. fHunting: Ministry of Agriculture data. It is assumed that the hunting concessions do not include park land or gazetted forest, i.e. that they are drawn exclusively from arable land where there is competition with smallholders. g There are unconfirmed reported cases amounting to 250,000 ha. Ignoring a concession of 50,000 ha to the religious organization Heaven on Earth. Assuming that 50 percent of all mining concessions are competing or could compete with smallholder occupation. See the discussion in the main text. Ignoring a pending application for 800,000 ha Assuming all the state farms have been divested in the form of medium- and large-scale concessions, viz. that smallholders have secured a negligible fraction of the total. There are only two cases where smallholders were included in the distribution of state farm land, and in each case they received only negligible amounts of land. In most cases they were prevented from participating in the distribution or lotteries of state land (Myers, 1996). 'Estimated from DINAGECA. Sources: Myers et al. (1993), and Myers (1994). Also see the discussion in World Bank (1996c). 173 Appendix Table XIV. Land tenure among smallholders: sources of acquisition of machambas (percent) Allocated by ... Purchased Traditional Formal Borrowed Occupied With Without Other Total authority authority title title North 11.4 3.6 8.2 48.7 0.6 24.1 3.5 100 Central 28.9 6.2 7.6 21.9 2.2 30.7 2.5 100 South 32.0 18.5 11.0 13.9 1.9 21.1 1.6 100 Total 24.5 9.1 8.8 27.8 1.6 25.8 2.5 100 Source: Ministerio da Agricultura (1994), as quoted in Desai (1996), Table 1, p. 9. 174 Appendix Table XV. Ownership, renting and use of tools of agricultural production Mean number of tools owned Percent using Manual Storage Mechanical Transport Mecha- Animal Any Sarnple tools containers tools tools nized tools draft rented size tool Regions: North 4.6 3.2 0.00 0.00 0.1 0.1 18 807 Central 5.8 4.1 0.01 0.11 0.5 3.8 7 1103 South 6.9 6.3 0.02 0.20 2.4 9.9 36 839 Land area possessed by household (ha) <1.0 4.7 3.7 0.01 0.3 0.5 2.2 20 771 1.0-1.5 5.1 4.3 0.01 0.09 1.2 2.6 20 496 1.5-2.0 5.6 4.1 0.00 0.03 0.3 3.5 14 345 2.0-3.0 6.2 5.2 0.01 0.09 1.0 4.1 19 483 >3.0 8.2 5.8 0.03 0.3 1.8 11 21 548 Number of adults in the household 0 4.6 3.5 0.00 0.01 0.0 6.3 16 80 1 3.7 3.2 0.00 0.03 0.5 2.4 22 427 2 5.2 4.1 0.00 0.04 0.3 2.7 15 1322 3 6.7 4.9 0.01 0.14 0.9 5.4 20 477 4 7.6 5.1 0.02 0.21 3.5 7.0 28 230 5+ 9.7 7.4 0,.04 0.45 3.3 15 30 243 Gender of household head Male 6.3 4.8 0.01 0.11 0.8 4.8 18 2232 Female 3.7 3.5 0.02 0.06 1.6 3.9 23 515 Household has off-farm income No 5.8 4.5 0.01 0.1 1.0 4.6 18 2121 Yes 5.6 4.7 0.02 0.12 0.8 4.6 23 627 Total 5.8 4.5 0.01 0.1 1.0 4.6 19 2749 Source: Ministerio da Agricultura (1994), as quoted in Desai (1996), Table VIII. 175 Appendix Table XVI. Marketed Production, Processing, and Export of Raw Cashew in Mozambique from 1980 to 1994 ('000 tons) Year Raw Cashew Processing Export of Raw Residual* Accumulated Cashew Residual 1980/81 91.5 83.2 0.0 8.3 8.3 1981/82 57.3 65.1 0.0 (7.8) 0.5 1982/83 17.6 24.2 0.0 (6.6) (6.1) 1983/84 25.3 21.9 0.0 3.4 (2.7) 1984/85 29.2 15.7 0.0 13.5 10.8 1985/86 40.1 24.8 0.0 15.3 26.1 1986/87 34.9 38.2 0.0 (3.3) 22.8 1987/88 44.4 39.3 0.0 5.1 27.9 1988/89 50.2 39.0 0.0 11.2 39.1 1989/90 22.1 20.8 0.0 1.3 40.4 1990/91 31.1 24.0 0.0 7.1 47.5 1991/92 54.0 29.7 6.0 18.3 65.8 1992/93 23.9 13.1 9.5 1.3 67.1 1993/94 35.4 6.0 21.6 7.8 74.9 Source: SSC, 1994 * Residual = Marketed Production - Processing - Export of Raw Cashew. 176 Appendix Table XVII. Price Structure and Profit Margins of Cashew Nuts, 1993-95 Price Structure 1993-94 Share of World Price Trader Margin as a (Mt/Ktg) (%) Share of Farm-gate Price (%) 1993/94: Farm-gate 700 15.5 ---- Retailer to Warehouse-gate 1,124 24.9 60.6 Wholesaler to Factory-gate 1,792 39.8 95.4 Wholesaler to Border 4,500 100.0 482.3 1994/95: Farm-gate 2,000 33.3 ---- Retailer to Warehouse-gate 2,500 41.7 25.0 Wholesaler to Factory-gate 2,900 48.3 20.0 Wholesaler to Border 6,000 100.0 175.0 Source: DEA/SSC, World Bank, 1995. 177 Appendix Table XVIII. Competitiveness of local processing and of raw nut exports Alternative price Resultant assumptions farm gate ($/ton) price FOB Factory (Mt/kg) a gate I Current (1995) export of raw cashews from Mozambique, with: (a) total margin 50%b 750 3789 (b) total margin 40%, assuming improvements in the 750 4546 trade system 2 Hypothetical model of a firm using the current capital- 550 4167 intensive technology. (Assumptions: kernel outtum 22%, interest rates 20%, wages $75/month, resultant internal rate of return 18%, trader margin 25%) 3 Hypothetical model of a firm using a labor-intensive 700 5304 technology, viz. steam cooking with blade cutting. (Assumptions: kernel outturn 23%, interest rates 20%, wages $75/month, resultant internal rate of return 22%, trader margin 25%) a Trader (=retailer+wholesaler) margin assumed to be 25% of the factory gate price. bThe total margin in neighbouring Tanzania is about 50%, viz. farmers get about 50% of the FOB price. Note: exchange rate was that of 8/31/95, Mt 10103 per dollar. Source: Kumar (1995), pp. 34-35, 63, combined with author's calculations. Explanation of Appendix Table XVIII 1. The option of exporting raw cashews is examined in line 1. Assuming a reasonable FOB price of $750/ton, and a ratio of farmgate to FOB of 50 percent (as is the case in Tanzania), the farm gate price would be Mt 3789, using the August 1995 exchange rate. If the efficiency of the trading system improves in time to come, and the total margin fell from 50 percent to 40 percent, the farm gate price could rise to Mt 4546. It is worth noting, incidentally, that both of these prices are in excess of the minimum farm gate price of Mt 3000/kg proposed by the government in August 1995. 2. In line 2 are presented the results of a hypothetical model of a factory using the current capital-intensive technology, viz. the hot oil bath with blade cutting. This is a model of a new factory, and the interest rate in real terms is set at a high 20 percent, so as to err on the conservative side. Wages are set at $75 per month, which are higher than current wage levels. The kernel outturn is set realistically at 22 percent, which is somewhat above the 18-20 percent of the Machava I factory of Caju de Mo~ambique (Kumar 1995, p. 42) but below the 23.4 178 percent outturn of a sample of nuts of the KJP group (Kumar 1995, p. 46). At a factory gate price of $550/ton, the operation is highly profitable with an internal rate of return of 18 percent. If the trader's margin is set at 25 percent of the factory gate price, the farm gate price would be Mt 4167, well above the likely equilibrium free market price if there were no hindrances to exports. 3. In line 3 are presented the results of a hypothetical model of a factory using a labor- intensive technology developed and used in India, viz. steam cooking with blade shelling. With a realistic kernel outturn of 23 percent, a high real interest rate of 20 percent, and wages at $75/month, the factory could purchase the raw material at $700/ton and still be highly profitable, with an internal rate of return of 22 percent. This approach could well be the technology of the future as it appears to be far more cost-effective than the traditional capital-intensive one. (Indeed it has been hinted that the reason the more labor-intensive and profitable technology has not been pursued is that there were special advantages attaching to the Oltramare machinery, such as the soft loans the manufacturer received from the credit guarantee agencies.) If the labor-intensive technology were to take root in Mozambique, the farm gate price would rise to something of the order of Mt 5304/kg, using the August 1995 exchange rate. 179 Appendix Table XIX. Seed cotton: area sown and output, by firm and farm tvpe (outgrowers, medium, large-scale), Mozambique, 1993 a~~~~~~~~~~~~~~~~~~~~ Firm name and farm typea Area sown Yield Output (tons) (ha) (tons/ha) LOMACO: Estate including PUPIb, Cabo Delgado 3,200 2.0 6,300 OutgrowersC, Cabo Delgado 6,900 0.6 4,100 Outgrowers, Nampula 3,300 0.5 1,600 Joao Ferreira dos Santosd: Outgrowers, Cuamba, Niassa 4,800 0.5 2,400 Outgrowers, Nampula 5,000 0.3 - Outgrowers, SODAN 35,000 0.4 12,000 Pinto outgrowers, Nampula 5,000 0.3 1,200 E.A.N./Foncar/N.G. (smallholders) 4,000 0.2 900 Entreposto (SAMO) outgrowers 6,600 0.35 2,200 Zamb6zia smaliholders & medium 3,000 0.3 900 Total: Estate sections ofjoint ventures 8,000 1.6 13,000 Medium- & large-scale farners 7,000 1.2 7,500 Outgrowers 63,000 0.4 26,000 Grand total 78,000 0.5 47,000 a Units with less than 2,000 ha sown are omitted for brevity. All units, whether outgrowers, medium- or large-scale, or joint ventures, are included in the totals. bPUPI = Pequena Unidade de Producao Intensa = Small intensive production unit, a hybrid of estate farming and outgrowing. ' Practically all outgrowers are smallholders. d Joint venture companies (LOMACO, aFs, Entreposto and others) are owned 50-50 by multinationals and the government. Source: Instituto de Algodao de Mocambique (1993), and Moll (1993). Table 2. 180 Appendix 2: Domestic Resource Cost Coefficients The measure of farm efficiency used here is the domestic resource cost coefficient (DRC). It is defined as I w, E DRC = i where Pb = economic border price of the commodity produced; Q = quantity of the commodity produced; wi = opportunity cost of the i-th primary factor of production; F1 = quantity of the i-th primary factor of production; Wj = world price equivalent of the i-th tradable input; and T, = quantity of thej-th tradable input. The numerator, which may be termed domestic costs, are non-tradables, viz. payments to capital, land and labor, plus the non-tradable components of inputs. The denominator is revenue less foreign exchange costs defined as the sum of expenses paid for the tradable component of inputs. Revenue is calculated using border prices. All costs are calculated excluding the effects of subsidies and taxes. The DRC then measures the cost in terms of local resources of obtaining one dollar's worth of foreign exchange. A DRC of between zero and unity indicates that the project makes a net contribution to the economy. Given flexible markets, and assuming that all the costs have been properly measured, projects of this kind would be taken up by agents driven by the profit motive. A DRC greater than unity indicates a project which is not worth its while in terms of its foreign exchange earnings. Under free market arrangements, these projects would be driven out by projects with DRCs less than unity. In the long run, ceteris paribus, there would be complete specialization in the product with the lowest DRC. 181 Appendix 2: Domestic Resource Cost Coefficients The measure of farm efficiency used here is the domestic resource cost coefficient (DRC). It is defined as DRC = I P Q-Ew," 7 where Pb = economic border price of the commodity produced; Q = quantity of the commodity produced; wi = opportunity cost of the i-th primary factor of production; F, = quantity of the i-th primary factor of production; wi = world price equivalent of the i-th tradable input; and Ti = quantity of thej-th tradable input. The numerator, which may be termed domestic costs, are non-tradables, viz. payments to capital, land and labor, plus the non-tradable components of inputs. The denominator is revenue less foreign exchange costs defined as the sum of expenses paid for the tradable component of inputs. Revenue is calculated using border prices. All costs are calculated excluding the effects of subsidies and taxes. The DRC then measures the cost in terms of local resources of obtaining one dollar's worth of foreign exchange. A DRC of between zero and unity indicates that the project makes a net contribution to the economy. Given flexible markets, and assuming that all the costs have been properly measured, projects of this kind would be taken up by agents driven by the profit motive. A DRC greater than unity indicates a project which is not worth its while in terms of its foreign exchange earnings. Under free market arrangements, these projects would be driven out by projects with DRCs less than unity. In the long run, ceteris paribus, there would be complete specialization in the product with the lowest DRC. c \mDz\agsecrem & molIl 52. 12/12/96 181