10. Investment in agriculture:
evolution and prospects

Technical background document
© FAO, 1996

1. Introduction

1.1 Recent development literature⁽¹⁾ has drawn attention to the possibility that there may have been complacency in recent years over world food supplies. During the past two decades, food output has risen faster than population growth; per caput production of cereals increased from 305 kg in 1970 to a peak of 342 kg in the mid-1980s, remaining at that level until recently, when it declined slightly. World cereal prices generally declined, interrupted twice briefly during the oil crises. The price decline occurred despite growing conversion of cereals into meat and milk in response to income growth, and despite a decline of total land devoted to basic food production in favour of expansion into higher-value crops and for non-agricultural uses. It is too early to judge whether the price increase of agricultural commodities, especially cereals, observed since mid-1995 is a temporary interlude, as happened often in the past, or signifies the halt or even reversal of the long-term trend towards lower prices.

1.2 There are, however, signs that despite the overall growth record worldwide, there has been underinvestment in agriculture over the past decade in some countries and regions, and in certain investment categories, which could constitute a risk to food supplies at a later date. The main evidence is as follows:

• The decline in international prices of major agricultural commodities and the return of positive real interest rates during the 1980s have slowed private investment. Statistics on agricultural capital formation in countries of the Organisation for Economic Co-operation and Development (OECD) suggest that disinvestment may have occurred in agriculture during the 1980s in several developed countries.
• The observed growth in agricultural output has been largely due to the spread of high-yielding varieties, which were developed two decades ago, and to associated factor use. The growth thus achieved has masked the effect of underinvestment in long-term assets, such as rural infrastructure, rural human capital and natural resource conservation.
• Although malnutrition affects a decreasing proportion of the population in developing countries, except in sub-Saharan Africa, it continues to be rampant in vast areas of the world, and the absolute number of malnourished people has dropped only slightly over the last 25 years.
• In some of the most productive agricultural regions of the world, little scope remains for increasing productivity economically by more factor inputs. Resistance to fertilizer and pesticide use is growing in areas of high agricultural intensity.
• In many developing countries, much public expenditure on agriculture is in the form of subsidies, leaving little public funding for the creation of new assets, for maintenance or for other growth-producing expenditure. The result is that many government agricultural support services barely function, rural roads are impassable for much of the year, farm machinery is inoperable and irrigation schemes have fallen into disrepair, while erosion, deforestation and salinization continue at ever-increasing rates.
• Starting in the second half of the 1980s, a marked decline has occurred in official development assistance (ODA) for agriculture. The great increase in foreign private-capital flows into a few developing countries has generally bypassed agriculture, which remains a relatively unattractive sector to private investment.
• There has been chronic underfunding of programmes of global and regional nature to underpin long-term agricultural growth and sustainability. Funding commitments for international agricultural research have levelled off and the growth of private financing for biotechnology research is heavily concentrated on the agricultural problems of developed countries. Although there is ample circumstantial evidence of increasing freshwater scarcity, of the desertification and salinization of formerly productive land and depletion of marine fisheries stock, far too little is understood about global trends in resource use and land potential, of the extent and implications of climate change and of the long-term impact on production and human welfare of reduced biological diversity. Of particular risk is the reliance of world food security on dangerously few plant and animal species and breeds.

1.3 Forecasts of world food supply and demand balances over the coming two to three decades differ widely, from the moderately optimistic (“Malthus must wait”) to the catastrophic (“Malthus is here”). Generally, they differ more over the future growth rates of agricultural output; that is, whether the agricultural performance of the last two decades can be repeated.

1.4 Few projections have examined the investment implications that must underlie any assumption on agricultural growth. Since the 1960s, economic growth models of the Harrod/Domar type have linked long-term growth of aggregate output to changes in capital stock, but a statistically robust relationship between agricultural investment and production is not easy to establish at the aggregate level (see Box 1).

1.5 Nevertheless, history shows that in market economies agricultural supply response to price incentives is typically good and can lead to cyclical over- and underinvestments. Price increases during major wars and other crises have seldom been of long duration. However, exclusive dependence on market forces cannot be an appropriate policy to handle food shortages, first because market mechanisms do not operate smoothly in all countries and situations, and second because permitting scarcity to provide adequate price incentives to producers may be intolerable to consumers with low purchasing power, and must be supplemented by proactive measures to relieve their suffering.

1.6 This paper attempts to make some general estimates of the types and volumes of investment⁽²⁾ which may be needed to secure the increases in basic food output required in developing countries over the next 20 years. The contributions required from the private and public sectors, and the role of external assistance, are discussed. Mention is also made of transboundary and global issues that will affect world food supply and its sustainability. It is important to note, however, that a discussion of the direct costs of providing targeted or emergency assistance to the undernourished, urgent as it is, does not fall within the scope of this paper. Similarly, the scope of the paper does not allow discussion in any detail of the special investment needs of the forestry and fisheries subsectors, which in some countries are of outstanding importance for food security and local welfare. The paper also cannot do full justice to the key problem of human-resource development, of which the relevance far exceeds agriculture, except in the context of extension.

1.7 The WAT2010 study projects agricultural output on the basis of expected effective demand, and of underlying production structure and natural resource endowment in different parts of the world. The result is a global model which can be used, as has been done below, as a first approach to the forecasting of agricultural investment requirements.

1.8 Other elements are needed to refine the estimate. As the world’s natural resources are increasingly stressed and good unused land is becoming scarce, agricultural growth will have to depend more and more on intensification, which is tantamount to investment. A persistently large number of food-insecure people do not have the means to purchase food; their demands add to that underpinning the WAT2010 model and pose an additional challenge to estimates of need and to the direction of investment.⁽³⁾ Furthermore, apart from supporting the growth of agricultural output, investment will also increasingly have to ensure the sustainability of resource use.

1.9 On the other hand, population growth is slowing down worldwide, and with rising incomes, effective demand for food is growing less than proportionally. Technology continues to advance and tends to reduce investment needs per unit of output. The implications for net requirements in agricultural investment are not entirely clear, especially in the absence of systematic resource accounting. What is certain, is that the composition and quality of investment must change drastically to reflect the new economic environment which has taken shape over the past 20 years, which has been characterized by structural adjustment, the opening up of the world economy, and changing demands made upon the world food production system.

2. Categories of agricultural investment

2.1 Future avenues for growth in food supply will be different from those exploited in the past. Dwindling reserves of unused land with high-production potential, plateauing of productivity in areas where modern technologies have already been introduced, and degeneration of the resource base in areas of both high and low potential, pose a new set of development challenges. Less will be gained in future from introducing new individual inputs (modern varieties, inorganic fertilizers, insecticides). The so-called green revolution, though not yet finished, has passed its peak contribution to the growth of world food supplies. More will depend in future on optimization of the mix of technologies, on sustaining their effectiveness, and on adapting this mix to the capabilities of less well-endowed land and poorer farmers. In agronomic terms, more intensive but sustainable production systems will increasingly take over from further application of green revolution approaches or the mere expansion of cultivated area as the main avenues of growth.

2.2 Such a shift has clear implications for the categories of supporting investment that will be needed in the future.⁴

• The new technology will have to be researched and extended. Increased emphasis will be needed on system intensification, adapted to optimizing returns to the factor of production (land, water, labour, capital) that is critical in each setting. Means to generate and communicate new technology will have to adapt to the challenge posed by greater agro-ecological and socio-economic diversity. But at the same time, opportunities should not be neglected to pursue further input-dominated (green revolution) approaches, to safeguard these gains by maintenance research, and to exploit innovations coming from biotechnology for the benefit of all categories of farmer.
• Farm-level changes will be required for intensification to be absorbed. These will range from expanded investments in irrigation and drainage, through rain-fed land and long-term soil improvements, to new types of machinery, new livestock breeds or additional commitments of farmers’ labour. Changes will be more location specific. If output is to keep pace with population growth, the emphasis should increasingly be on securing a substantial contribution from all potential actors involved, ranging from mechanized commercial farmers to subsistence hill dwellers.
• Upstream and downstream support will be needed. For intensification to proceed under ever-increasing diverse conditions – sustainably, and on the scale required – the supply systems for inputs, seeds, machines, equipment and services will have to grow in proportion. The same applies to storage, processing, markets, transport systems and the other downstream links between producers and consumers, increasing numbers of whom will live in large cities, often distant from sources of food supply.
• Facilitating, largely public investments will be required to complement and to stimulate the flow of private investment. These include development of economic infrastructure such as roads and communication systems, which allow producers to access new markets, as well as strengthened social-support services, notably in the areas of health, education and welfare, which are prerequisites if rural people in many developing countries are to take advantage of new opportunities to earn and produce.

2.3 The economic importance and the significance for growth in food output of some of the key items among those listed above are discussed in more detail in the sections that follow.

Intensification

Irrigation and more efficient water use

2.4 Worldwide, some 250 million ha receive irrigation. Irrigation serves 18 percent of all arable and permanent cropland but there are wide regional variations, with Asia at 35 percent, sub-Saharan Africa at 6 percent and Latin America at 11 percent. New irrigation has been added at an average rate of 3 million ha annually over the past two decades, of which 87 percent is in Asia. At present, irrigation development has slowed substantially compared with the 1960s. Internationally funded irrigation development has also declined with, for example, World Bank lending to irrigation having been halved from US$2 billion in 1980 to US$1 billion in 1993.

2.5 Large-scale projects with full water control face increasing obstacles. Such projects have been criticized for causing environmental damage, social inequities and showing little respect for the rights of traditional land users. Greater use of irrigation can also lead to increased water-borne health risks, notably schistosomiasis and malaria. Such public health concerns need to be addressed, initially in the planning and design phases of irrigation schemes, and then by the routine monitoring of the health and nutrition impacts of the scheme on the population.

2.6 In sub-Saharan Africa, less than 20 percent of the theoretical irrigation potential of 35.4 million hectares⁽⁵⁾ has been developed, and the low economic feasibility of irrigation is the greatest obstacle to its expansion. This is because of higher initial investment costs than elsewhere, low cropping intensities, market limitations for high-value crops, competition from low-cost imported rice, high transport costs due to poor roads (major consumption centres are often located on the coast and supplied more cheaply from overseas), and the lack of an irrigation tradition.

2.7 In Asia, the average investment cost of new irrigation has doubled in the last ten years, during which time international rice prices have declined. Rates of return on new irrigation infrastructure are falling in consequence. In some cases such investment can be justified partly by the development of hydroelectricity or aquifer replenishment, which benefits water users outside the command area proper; but there is no doubt that despite the multipurpose nature of some irrigation investments, the opportunities for new large-scale irrigation development have diminished, and the continuing downward price trend of basic staples has made irrigation increasingly uneconomic for such crops.

2.8 If opportunities for large-scale irrigation expansion under formal schemes appear at present to be less abundant than in the past, there are substantial opportunities for modernizing and making better use of existing irrigation. Rehabilitation can be achieved at a fraction of the cost of new irrigation development, and can make substantial contributions to agricultural growth if coupled with better management, adequate extension, input supply and marketing. In most irrigation areas in developing countries, system efficiencies could be greatly increased as an alternative to new irrigation development (Rosegrant and Svendsen, 1993).

2.9 The transfer of irrigation management from government institutions to farmers’ groups and users’ associations has been shown to increase efficiency, lead to more realistic water pricing to recover a higher share of operation and maintenance costs, and to improve scheme sustainability.⁽⁶⁾ This is an area where positive advances can be, and are being made.

2.10 Simple engineering techniques are often suitable for development by rural communities themselves under food-for-work programmes, or on a cost-sharing basis with government or non-governmental organizations (NGOs), or by private operators. Small-scale private pump irrigation from streams, motorized, hand- or animal-powered waterlifting from shallow wells, and valley bottom development under collective self-help schemes, have all been taken up spontaneously and with success in many countries. Much of the investment cost is in the form of local labour, which can generate multiplier effects at the local level. Indigenous populations have also practised various forms of water harvesting and soil moisture conservation for centuries. Understanding these traditional techniques of water management is a promising starting-point for developing irrigation support which is sensitive to local preferences and skills, and avoids the pitfalls of externally imposed, large centrally planned schemes. Aid agencies have in recent years perfected rural appraisal techniques and participatory planning methods which can help design socially and economically acceptable improvement programmes in this field.

2.11 Much scope also exists for in situ moisture conservation and water harvesting, either of which can expand and stabilize crop yields and contribute to aquifer replenishment. But additional research on the economic feasibility of water harvesting is needed in various agro-ecological environments (Reij, Mulder and Begemann, 1988), as many projects of this type are grant financed, and reported investment costs vary widely and are often too high to provide an acceptable return.

2.12 An important issue is the increasing competition for water between agriculture and other sectors, especially in Asia and in the Near East. Debate on how to cope with inevitable future water scarcity often leads to a recommendation to establish and improve water markets as a means of raising allocative efficiency. Small gains in water use and efficiency in irrigation, the developing world’s largest consumer of water (80 percent), would release large volumes of water for non-agricultural sectors, where they compete. The institution of effective water markets normally requires rehabilitation of irrigation schemes, the transfer of management to users, and investment in conveyance and metering systems, as well as effective law enforcement. In countries where water is still available at relatively low cost, the social benefits from better allocation may not justify these measures (Rosegrant and Binswanger, 1994). However, the growing scarcity of water will make such property rights an increasingly feasible alternative to actual water-allocation practices, as has been demonstrated, for instance, in Chile and in the western United States.

Land development and improvement

2.13 The WAT2010 study projects that one-fifth of agricultural growth during the next two decades will come from the exploitation of new land and four-fifths from agricultural intensification. The spread of farming into new, more fragile areas will not stop, because people will often have no choice but to settle on increasingly marginal land, clear primary forests, deplete the soils rapidly and cause damage to the environment. The social cost of their output is high in such situations, and their labour productivity low. Agricultural practices in such circumstances are dictated more by survival needs than by financial or economic reckoning.

2.14 The public sector can contain this trend, inter alia, by improving the security of tenure of individuals and communities. Experience shows that free access is the most damaging way of exploiting a natural resource and will eventually lead to its destruction. Legalization of land occupation, and giving incentives to squatters to protect the remaining reserve land, along with the provision of services (roads, extension and planting materials to allow for sustainable cropping), can help to stabilize the situation.

2.15 The alternative to opening up new areas is land improvement. Sustainable increases in land productivity can be made through investment. The publication of research results in the Machakos district in Kenya over the last 60 years has shown the potential for a reversal of land degradation under population pressure, given a favourable institutional and market environment (Tiffen, Mortimore and Gichuki, 1994).

2.16 Means to improve land and agricultural productivity may involve changes in tillage and soil cover to improve rainfall penetration, improved management of crop residues and soil organic matter, water harvesting, drainage, liming and rock phosphate applications to remedy chemical imbalances, as well as mechanical measures such as land levelling, terracing and bunding for runoff and erosion control. Most such improvements require investment which can be highly economical. However, in some cases government participation in costs may be needed to lever change, because part of the return consists of public-good or intergenerational benefit, and private returns alone are insufficient to attract farmers into land improvement and better land use practices.

2.17 Land reform is a sensitive issue in many countries. However, except where well-developed commercial agro-industries exist, smallholdings tend to be more intensively exploited than large holdings. When large holdings are intensively managed they may depend on large numbers of hired labourers, which can lead to a high cost of supervision and to industrial dispute (Lipton, 1995). Land redistribution through market-assisted land reforms⁽⁷⁾ could thus make a significant contribution to long-term agricultural growth in many less-developed countries and take pressure off marginal land.

Purchased inputs

2.18 Increased emphasis on efficient, safe, sustainable, complementary use of fertilizers, pesticides, plant growth regulators, veterinary drugs and vaccines will call for continuing investments. Purchased input use requires prior investment in manufacturing plants, storage and distribution infrastructure and ties up substantial amounts of working capital. Mineral fertilizer in particular, of which the world uses 140 million tonnes (nutrient equivalent) annually, half of it in developing countries, has been a major source of agricultural growth for the past century.

2.19 Whether fertilizer can repeat its past contribution to agricultural growth is not clear. In high-potential areas, e.g. China, central Luzon in the Philippines and the Punjab in India, incremental cereal/fertilizer volume ratios are declining. The phasing down of input subsidies in various developing countries is further forcing economies on farmers. In developed countries consumption has been declining for some time, due to environmental concerns and area reduction. The process has reached a point where concern has been raised in some circles about the underuse of plant nutrients in relation to yield potential.

2.20 There is still room for more fertilizer use on different crops in many developing countries, even though high-yielding varieties and associated irrigation, which favour higher fertilizer application, are expanding at a slower pace than in the past. Changing agricultural production patterns towards higher-value crops will increase demand for various types of purchased inputs.

2.21 In sub-Saharan Africa, where fertilizer use is still very low, consumption is hampered by high distribution costs, the lack of markets for output, lack of a domestic fertilizer industry and poor yield response, and the high risk of using fertilizer in traditional agricultural settings.

2.22 As the economic returns from conventional inputs decline and environmental concerns grow, more important than additional input use will be their more judicious application.⁽⁸⁾ New technologies for integrated pest management (IPM) and integrated plant nutrient management (IPNM), and the obsolescence of some inputs as their effects become achievable through biotechnology, will lead to narrower segmented markets which will be a major challenge to the agrochemical industry. Products which could fill crucial niches in the new integrated strategies for sustainability may become orphaned because they cannot be privately profitable to a potential manufacturer.

2.23 Good soil structure is often as important as good nutrient availability. Soil enrichment and texture maintenance with organic matter through fallow and alley cropping, livestock/agriculture integration or similar means of increasing the organic matter in the soil, are common recommendations by research institutes, but are not always seen as financially rewarding by farmers. More investment is needed in applied and adaptive research to offer farmers a wider menu of options to manage soil fertility adequately as well as to control pests and disease.

Mechanization

2.24 Mechanization has a dramatic potential to increase labour productivity. It can also greatly increase farm sizes where land is not a constraint. Agriculture in developing countries depends to a surprising extent on mechanization. According to FAO statistics, an additional 200 000 tractors on average have been used in farming systems annually during the last two decades, and it appears that there are some 5.5 million tractors in use in less-developed countries. Mechanization is attractive where it replaces repetitive, high energy-consuming and exhausting human labour or animal traction (Binswanger and Donovan, 1988), and it is highly responsive to labour cost changes. Because it is a relatively expensive item in the farm budget, often exceeding that of fertilizer, mechanization is also strongly responsive to output prices. Purchases of tractors and water pumps have fallen dramatically in recent years in some regions, especially in Latin America and in the Near East and North Africa. On the other hand, overinvestment has occurred in the past in countries with high administered farm prices (Western Europe, Republic of Korea and Japan), where subsidized mechanization credit was available (Pakistan), and where government-managed tractor hiring schemes existed (some African countries).

2.25 It may reasonably be expected that due to the rising cost of mechanized services in relation to agricultural prices, mechanization will become less economical and force adjustments in mechanization density and efficiency, although a ratchet effect is likely to prevent any return to premechanization farming methods. Nevertheless there is large scope for more use of renewable energy sources at village level. Biogas and wind energy have great potential in some areas for crop drying, primary processing, water heating, water pumping and cooking. Research should help further to develop appropriate methods and related equipment.

2.26 Animal traction is an attractive alternative to motorization for augmenting farm energy for agricultural intensification. Nearly two-thirds of all draught animals in the developing world are concentrated in South Asia, a problem area for food security. With rising labour costs tractorization is gradually replacing animal traction in many areas, and increasing amounts of farm residues are used off the farm and are unavailable for animal feed or soil enrichment. This means that while mechanization might not be a constraint to intensification, a soil-fertility constraint could emerge which should be closely monitored.

2.27 In parts of the African savannah zone there are free grazing reserves, and the potential for increasing agricultural output by expanding animal traction seems immense where it is not limited by other factors, in particular marketing constraints and animal disease incidence. Where animal traction is feasible, important investment needs for technology testing, training and acquisition of draught animals, for veterinary services and for manufacturing and purchase of draught implements may be expected.

Post-production operations

2.28 Post-production operations (handling, storage, transport, marketing and processing) are likely to show dramatic growth in agriculture-related investment in the future, fuelled by a combination of continued urban population increase and income growth, especially in East Asia. By the year 2010 the urban population in developing countries will have doubled to nearly 2.7 billion, from 1.4 billion in 1990.

2.29 The problems of urban food supply, in particular to future megacities, have received less public attention than those of food production at the farm level. Investment in farm-to-market roads, transport services, rural and urban wholesale and retail markets, slaughter facilities and dairy plants, food, feedgrain and oil mills, and dry and cold storage space, could easily be of the same order in some regions as for primary food production. Post-production operations could become a major limiting factor to investments in primary production. Unless these problems are addressed, the social costs of traffic congestion, produce downgrading and spoilage, food and water contamination, income losses to farmers and traders, and the higher cost to consumers, as well as lost time and inconvenience, could be staggering.

2.30 Investments in promoting better food quality and safety at all stages in the food chain are usually quickly recovered through reduced food losses and better consumer acceptance. Therefore most investment may be expected from the private sector. To the public authorities, the provision of urban public-marketing infrastructure and institutional capacity for its planning, implementation and regulation will be a major challenge. Developing and strengthening effective regulatory and quality assurance systems for food control can be particularly beneficial for protecting consumers and promoting trade in food. A key problem is the provision of space in urban areas for physical transactions in competition with other demands. The integration of food marketing and processing,⁽⁹⁾ including street vending of prepared food, into the fabric of urban economic activities must be through a combination of market-driven land allocation and mandated urban zoning. Many such activities will be forced away from city centres to more peripheral areas, a process that has taken place in industrialized countries during recent decades. It will have important repercussions on traffic flows and public and private investment in marketplaces, transport systems, roads, habitats and utilities. It will also influence the degree of competition and concentration of marketing activities and of living costs in urban areas.

2.31 A special problem in the provision of public services for food marketing and processing, which is not the case with agricultural production, is the diffusion of institutional responsibility for monitoring developments in the private sector, and for planning and implementing marketing improvements. These responsibilities may be divided between ministries of agriculture, industry, commerce, public works, the interior and the dependent municipalities, making concerted planning nearly impossible. Planning should as much as possible be devolved to local administrations (which it usually is) but should also be fully participatory, involving consumer groups, traders, city and district officials and farmers. The regulatory framework in food marketing must at once be flexible and adapted to the needs of small traders, as well as strictly enforced to contain the emergence of a large informal sector alienated from, and hostile to, the public administration, which would impose high social costs on the community.

2.32 In rural areas, key problems are the strengthening of farmers’ organizations to undertake their own marketing, input procurement and establishing community-based, viable financial intermediaries where private-sector operators find it too costly and risky to provide such services.

Rural infrastructure

2.33 Roads, electricity supplies, telecommunications and other infrastructure services are limited in all rural areas, although they are of key importance to stimulate agricultural investment and growth.⁽¹⁰⁾ Their scarity is partly due to the higher per caput costs of serving dispersed populations, buth also to a large extent to urban bias in the allocation of public funding. The problem is especially acute in Africa. Even Nigeria, the country with the highest road density in Africa, has barely reached the level of road coverage which existed in India in 1950. It has been argued that the lack of infrastructure, especially roads in sub-Saharan Africa, is so great, that for this reason alone food security in the next 20 years cannot depend on a green revolution modelled on the Asian example, but must be based on establishing self-contained agricultural production and consumption zones which can operate with a minimum of outside purchased inputs (Spencer, 1994).

2.34 Better communications are a key requirement. They reduce transport costs, increase competition, reduce marketing margins, and in this way can directly improve farm incomes and private investment opportunities. Furthermore communications help those who have been isolated to understand the value of technology and commerce. Health, clean water and education are additional stimulants to rural economic growth, and usually the first needs expressed by rural populations. Electrification can cause a quantum leap in the quality of life in rural areas, and trigger a multitude of private investment initiatives, especially in processing and artisanal activities. Of special importance is expenditure on rural health and education. Prevention and treatment of the most common childhood diseases and micronutrient supplementation are low-cost investments in human resources. Primary education, especially for girls, brings high economic returns, and is the greatest untapped rural productivity reserve (Psacharopoulos and Woodhall, 1985).

2.35 Lack of sustainability is a key problem with public infrastructure investments. Neglected maintenance of rural roads is common; nearly one-half of all roads in Africa are in need of rehabilitation. Periodic major rehabilitation uses funds which are then unavailable for expanding the network. Fuel taxes can be earmarked for road maintenance, although this may not greatly increase efficiency (World Bank, 1994a). More effective may be the handing over of the responsibility for infrastructure provision and maintenance to local government, together with the regular transfer of funds by the central authorities. When combined with participatory approaches to infrastructure planning, execution and maintenance, this can make access to infrastructure and services more equitable and increase the feeling of ownership by users. Other cost-effective ways to improve outside access for rural populations include investments in intermediate and intermodal transport (bicycle, oxcart, lorry and railway). Labour-intensive maintenance through private contractors can reduce wastage and provide a closer match between needs and delivery for rural infrastructure services, besides raising the incomes – and hence the capacity to buy food – of the rural poor.

2.36 Rural infrastructure, other than irrigation, is of key importance in stimulating private agricultural investment and growht. Infrastructurecan be highly attractive to investment, although much is not suitable for the private sector because of its nature as public goods. Of the US$200 billion invested in infrastructure in developing countries in 1993, only 7 percent originated from the private sector (World Bank, 1994a). Strained public finances and the need for greater efficiency impose a more differentiated view of infrastructure services. In nearly all developing (and developed) countries efforts are made to commercialize, decentralized and privatize such services with the aim of making them more self-supporting, and to minimize the number of those which must remain directly dependent on the central budget.

Technology generation and transfer

2.37 New technology has been the most consistent driving force behind agricultural growth. Evenson (1994) estimates that it has contributed from one-half to two-thirds of output gains over recent decades. Economic returns on investments in well-organized, well-funded and purposefully targeted technology generation are regularly over 20 percent, and often 30 to 40 percent or more.

2.38 Technology generation is shared between the public and the private sectors, with the share of the private sector tending to be greater in more developed countries. Private research is attracted to subsectors where markets for research results exist and can be privately appropriated. This is typically the case in countries with intellectual property right (IPR) protection for inputs such as agrochemicals, agricultural machinery and seeds to which the originator can restrict access by the manufacturing or breeding process. It is also favoured when a brand name can be established that commands consumer loyalty.

2.39 The benefits of technology gains are very unevenly distributed. In sub-Saharan Africa they have been limited mainly to export crops and, in some regions, hybrid maize. In South Asia, especially, input-intensive green revolution technology has reached almost all areas with the resource potential to support it. The expected adaptation of hybrid rice to tropical conditions could enable a further improvement in yields of 10 to 15 percent. In the longer term, the development of a new rice plant to replace the former International Rice Research Institute (IRRI) varieties holds promise of another 20 to 25 percent yield increase.

2.40 The future direction of research needs to reflect such differences.⁽¹¹⁾ In sub-Saharan Africa, and other generally poorly endowed areas, there is a deficit of smallholder-oriented research for resource-poor areas, using a systems approach. Green revolution research meanwhile needs to be maintained in high-potential areas, where more research funds must go towards tackling the yield plateaux and problems of resource degradation which farmers are now facing. These shifts will have important long-term investment and organizational implications. The more diffuse pattern of activities required by systems research calls for different staff skills and training, less research on station and more on farmers’ fields, and stronger interactions with extension. It is potentially more expensive. The change in scientific approach requires new forms of training both for farmers and researchers, plus a greater input from economists and social scientists.

2.41 Operational budgets per researcher in many developing countries have been declining in recent years. Due to the restriction this places on researchers’ access to their clients, much national research remains of little practical relevance. Effective links to extension and feedback cannot be properly established. National institutes have often been slow to adopt a client-oriented approach in research programming. International agricultural research centres (IARCs) and entities such as the Special Programme for African Agricultural Research (SPAAR) have made progress towards research on optimizing the use of existing water, land and purchased inputs and services; in systems research in resource-poor areas; in exploring the potential of new or orphan crops which have been bypassed because of concern with a limited number of crops and varieties benefiting from green revolution technology; and in capitalizing on indigenous knowledge. Most national agricultural research systems (NARS), however, remain slow and ill-equipped to follow this lead.

2.42 Umali (1992) estimated that worldwide, US$8.5 billion were spent on agricultural research in 1985, of which US$3.2 billion were in developing countries. More recent estimates are US$5 to $6 billion per annum spent by NARS in developing countries and, in addition, US$270 million through the Consultative Group on International Agricultural Research (CGIAR) system.

2.43 A target of 2 percent of agricultural gross domestic product (GDP) was suggested by the World Bank in 1981 for national research expenditure in developing countries, similar to the level of developed countries, which would bring total expenditure in developing countries to a level of some US$12 billion. It should be noted, however, that such targets are broad indications, which do not take into consideration the effectiveness of research, requirements at country level, and the critical mass of the research establishment in relation to the relevance of particular research lines.

2.44 Discussion continues over the future type and direction of research best able to promote food security, especially in the critical regions of South Asia and sub-Saharan Africa. Concerns have been expressed about the possibility of a technology gap in food production. In some areas in South Asia rice yields are advancing only slowly on research stations (Pingali, Moya and Velasco, 1990), and in the better areas such as the Indian Punjab or central Luzon, a yield plateau seems to have been reached.

2.45 A consensus seems to be forming that future research needs a carefully managed relative shift from commodity improvement towards resource management. This would be required in all developing regions that do not have the natural resource endowment, roads and irrigation infrastructure or market access to bring about a rapid new green revolution. More research than at present would thus be devoted to optimizing the use of available labour, water, land and purchased inputs and services, and would concentrate on resource-poor areas. It would be more systems-oriented, incorporate more orphan crops, capitalize on indigenous knowledge, and adopt a participatory approach to research programming.

2.46 Of key importance is the typically long lead time between research being undertaken, and the results of such research reaching farmers’ fields. Urgent and early thought must be given to identifying issues that are critical for food security in poorer countries, and accelerating the relevant research on these issues. If latent technology is there, much can be expected from the so-called Boserup effect;⁽¹²⁾ if not, a steady increasing stream of new technology needs to be funded through the IARC/NARS structure to form a technology reserve for uptake at the appropriate time.

2.47 According to FAO (1990), technology transfer in developing countries involves some 550 000 staff, most of them in public extension services, and costs about US$4.5 billion annually. Under the influence of structural adjustment and declining public funding, extension services have in recent years tended to shrink.

2.48 The economic returns obtainable from, and the issues connected with, technology transfer to a large extent mirror those of technology generation. In order to remain affordable under conditions of still, albeit slowly, growing rural populations, and at the same time carry the broader message resulting from the two-pronged research paradigm, extension services must change.

2.49 Traditionally, the number of extension workers for a given farming population has been used as a measure for an appropriate extension coverage. A typical goal was, for instance, one agent for every 500 farmers. On this basis the developing world would need 2.4 million extension agents, or over four times the present figure, of whom 80 percent would be in public service, to serve 1.2 billion farmers, costing nearly US$20 billion annually. This is clearly unrealistic. The key to future extension is cost recovery, cheaper delivery and greater relevance of the extension message.

2.50 Ensuring cost recovery calls for demand-driven extension services and farm-input supply. Such services typically develop with growing sophistication and differentiation of agricultural systems, and will become increasingly frequent in future, although at present they are still of limited importance in many developing countries. Privatization of extension is feasible where appropriation of returns from services rendered is possible. This is the case with many modern technologies, such as specialized agricultural information relating to high-value crops, hybrid seeds, agricultural machinery, agrochemicals, marketing and processing operations. Increasingly, assistance in bookkeeping and financial planning enter the range of required extension services, especially in transition countries. Modern technology services can be provided by input and output marketing and processing enterprises, which provide them as part of their marketing strategy, or by private consultants who provide pure advice for a fee. Intermediate forms are found where private services are partly subsidized by governments, or where governments subcontract extension to private agents. Another way to achieve greater cost recovery and self-financing of extension is through farmers’ associations, which provide extension against fees paid by their members. In other cases NGOs provide these services, and although cost recovery is usually not practised, the delivery of the services is often cheaper and more focused, and eases the financial burden to governments.

2.51 Discrete extension services are needed to provide adequate services in a differentiating farming environment. This contrasts with efforts to establish unified extension services in many developing countries. Such unification seems advisable in particular situations where highly homogeneous farming is practised and costs are contained by high farm densities, such as in some intensive irrigation areas in Asia. As a general rule, however, while retaining overall control of the quality of services, governments should actively seek opportunities for privatization, cost recovery, and devolvement of extension services as a means of reducing claims on their budgets and to improve the matching of services to patterns of demand. The public sector would then be able to concentrate on those services that are truly the domain of government: general agricultural information directed at resource-poor farmers who are not otherwise served.

2.52 Cost-effectiveness of extension delivery can also be enhanced through participatory extension methods and the use of modern modes of communication. Participatory approaches are vital for developing and disseminating extension messages. Participation is likely to require special training, a particular mindset of the agent, acceptance of a bottom-up flow of information, and a longer time to explore with farmers their preferences and to adapt research-generated technology packages to their needs. It will greatly reduce wastage in extension efforts that are not properly targeted to farmers’ expressed needs. Of particular importance is the adaptation of extension services to women. There is a widespread feminization of agriculture, especially in Africa and parts of Latin America, which calls for larger numbers of female extension workers, special advice on access to inputs and credit, which are less readily available to women, and on ways to overcome their lack of access to property rights and collateral. In South Asia, special financing intermediaries take care of women’s credit needs and anecdotal evidence has it that they are better credit risks than men, despite offering less formal collateral.

2.53 Modern modes of extension delivery could greatly reduce traditional face-to-face extension and the mobility of extension agents. Radio, television and video display can spread messages suitable for a general audience and leave more time for the agent to concentrate on individual farmers’ needs. These methods are extremely common in developed countries, and have gained acceptance in developing countries such as agricultural information through State television in India, Côte d’Ivoire and Brazil; videotaped messages in Brazil, Honduras, Mexico, Paraguay and Peru; and satellite systems for spreading agricultural information over large areas in Indonesia, the Philippines and West Africa.

2.54 Mass media methods can overcome illiteracy constraints which hold back many extension efforts. However, one of the principal means of raising the receptivity of rural populations to modern technology is broad-based education. As noted earlier (Paragraph 1.6), the subject of human-resource development in general is beyond the scope of this paper. Its importance for increasing human potential and raising awareness of options for improving rural life in all spheres cannot be overestimated. It has been consistently shown that farmers’ education is strongly and positively correlated with higher productivity and efficiency, and that primary education in rural areas tends to have higher returns than secondary and post-secondary education. Likewise, returns from educating girls are higher than from boys’ education, and sub-Saharan Africa shows the highest returns on primary and secondary education among all developing regions (Psacharopoulos, 1994).

2.56 Many of the investments required to support agricultural intensification raise transboundary issues; that is, the problems are beyond the means of one country to solve alone but once solved could bring benefits to many. Many interventions deal with global environmental challenges: for instance the Global Environmental Facility, which combats global warming, international water pollution, ozone depletion and the reduction of biological diversity, including such critical issues for agriculture as the narrowing genetic base of crop and livestock species. Other initiatives are the Mediterranean Technical Assistance Programme, the Mediterranean Coastal Cities Network, the Programme for Environmental Management and Protection of the Black Sea, and similar programmes for the Baltic Sea, the Aral Sea basin, the Caspian Sea, the Danube River basin and Lake Victoria. Many of the latter examples express concerns more specifically relevant to developed countries and the transitional economies of Eastern Europe.

2.57 Initiatives directed at dealing with specific transboundary problems in developing countries are also usually seriously underfunded. Examples are locust and desertification control, early warning systems for drought and food shortages, the Emergency Prevention System for Transboundary Animal and Plant Pests and Diseases (EMPRES), joint management of common fisheries and inland water resources between several countries, and post-disaster relief operations.

Regional Priorities

Asia

2.58 In much of the region inadequate agricultural research is becoming a bottleneck as yield growth is slowing and research stations have little unused technology to tackle either the many second-generation problems of degeneration in high-potential areas or the new, farmer-friendly systems needed to intensify agriculture on poorer land. Investment in technology generation and diffusion should be urgently accelerated to address issues of micronutrient deficiencies, reduced water availability for continuously irrigated rice and the like. Irrigation rehabilitation, coupled with more efficient and decentralized water management and the creation of water markets, is a high priority. At the same time, environmental rehabilitation and protection are required as rapid agricultural growth has often been at the expense of natural capital. Urban growth needs large investment in marketing infrastructure.

Africa

2.59 Rural infrastructure is badly lacking, especially in sub-Saharan Africa. Adequate infrastructure (e.g. irrigation and rural roads) is scarce and deteriorating. The region has the added disadvantage of relatively dispersed populations and limited irrigation potential. Heavy investments are needed
in rural communications infrastructure; irrigation rehabilitation and modernization; better exploitation of rainfall by simple and improved water capture and use of in situ water harvesting; land management and improvement; and education and health services. Research results and best practices, as yet untried on a large scale, are available for high-potential areas in Africa and can be adopted as soon as warranted by population density, accessibility of areas and development of markets. Systemic research for small, resource-poor farming systems, which form the majority in most African countries, continues to be needed to create new farming systems that are sustainable at higher population densities and that can support spontaneous intensification of African traditional agriculture.

Latin America

2.60 Land reserves are still available and appropriately scaled mechanized farming is one way to expand agriculture. The high degree of commercialization of farming points to a need for well-functioning rural financial markets. Land is also very inequitably distributed and small farmers are forced to cultivate marginal lands causing environmental damage (parts of southern Africa show similar patterns). A strengthening of land-reform initiatives, especially those that are market-based, seems to be needed to optimize development of the last remaining land reserves for agriculture. The expected high degree of urbanization (80 percent) would further suggest very substantial investment in marketing and processing infrastructure.

Public versus private investment

2.61 The overwhelming majority of investments envisaged consists of fixed or movable physical assets that are under individual control and can be privately profitable. At the other extreme are investments in public goods and services. These include health and education, regulation and control of private activities that could be prejudicial to the public interest and major investments, for instance in roads or dams, which are beyond the financing means of the private sector or which bring multiple benefits that cannot all be privately appropriated.

2.62 Between the two, however, are mixed situations where investments need to be divided or shared between the private and public sectors, or where communities rather than individuals or companies have a role to play. As agriculture changes under the pressure of rising demand, new ways of linking public and private economic activity will have to evolve.

• Agricultural technology generation and diffusion need to be shared between the public and private sectors, with the private sector increasing its share in more advanced developing countries. The proportion of privately funded research is in the order of 30 to 40 percent of all research expenditure in developed countries (nearly two-thirds in the United States) and about 5 percent in most less-developed countries (Evenson, 1994). This has to do with markets for research results, and it can be expected that an increasing proportion of research activities mentioned in Paragraph 2.39 will move from the public to the private domain. On the other hand the public sector must, perforce, remain dominant in areas such as systems development and extension for the subsistence or resource-poor farmer, natural resource management and monitoring, and the understanding and containment of potentially negative effects of agricultural intensification.
• In irrigation, drainage and land improvement, public expenditure may be needed to stimulate private interest in taking responsibility for what should ultimately become communally or privately driven agricultural intensification. Such public expenditure should not be seen as mere subsidy. Often the changes encouraged will deal with important externalities which are of general public concern, giving benefits which only society can be expected to finance. It may also be fully justified economically under less narrowly defined methods of calculating costs and benefits, serving to bring financial returns from more sustainable intensification closer to the economic returns.
• Experience shows that for rural infrastructure and services to be sustainable it is essential to create a sense of ownership by the beneficiaries and communities concerned, and for them (as for irrigation) to assume the maximum responsibility for post-investment operation and maintenance.
• The changes in agricultural systems will aggravate the problems of orphan agrochemicals. The mutual suspicion which has dogged past attempts at cooperation between the private and public sectors in this field must be overcome. Otherwise moves towards more efficient integrated management of pests and soil nutrients will be increasingly frustrated by the lack of appropriately tailored external inputs. Analogous problems in biotechnology development must also be addressed if developing countries are to be able to make full use of the opportunities created.
• Investment in post-production processing, markets and distribution calls for consultative planning rather than joint investment between the public and private sectors. An efficient private sector will only be possible if investments by entrepreneurs match the location and timing of government investments in roads, markets and other facilitating infrastructure.

2.63 Two broad implications for governments can be drawn from this catalogue. The first is that more will need to be invested in the human skills and institutional capacities of governments, so as to interact fruitfully with the millions of private individuals on whose willingness to invest the growth in food supply ultimately depends. The second is that despite such government efforts within the agricultural sector, progress will be limited unless appropriate economic policy signals are also given at the national level to private individuals.

 [Next document]

Notes

⁽¹⁾Studies have been undertaken by Mitchell and Ingco (1993), Brown and Kane (1994), McCalla (1994) and Crosson and Anderson (1992), among others. A critical review of this and other literature on the subject is presented by Alexandratos and de Haen (1995).

Back to text

⁽²⁾ Investment is defined here in a broad sense as consumption foregone for the sake of future gain. The maturity of investment can range from one to over 20 years or into the next generation. In this sense, farm labour for bush clearing, a farmer raising a calf rather than slaughtering it, public expenditure on health and education, including extension, which enhace human capital, and soil liming and phosphating, which affect crop yields for more than one year at a time, are all termed investment for the purposes of this paper.

Back to text

⁽³⁾ The implication of a policy commintment to raise nutritional standards further for countries estimated inWAT2010 to be unable to reach the average dietary energy supply (DES) level of 2 700 Calories is the subjecto of the WFS companion paper 14Assessment of feasible progress in food security.

Back to text

⁽⁴⁾ See also WFS companion paper 6Lessons from the green revolution: towards a new green revolution. Back to text

⁽⁵⁾ Theoretically irrigable land is land that is both technically suitable for irrigation and in principle capable of being developed for irrigation at acceptable costs and returns. It is a somewhat imprecise working definition. Back to text

⁽⁶⁾  Recent findings by the World Bank’s Operations Evaluation Department (World Bank, 1994d) are that large irrigation projects often perform better than small ones due to economies of scale in engineering; overruns of construction schedules may improve performance because of economies of time (engineers’ learning process); better cost recovery does not necessarily improve operation and maintenance since higher water charges are seldom earmarked for operation and maintenance; what is more important than full cost recovery is financial autonomy of the water authority (after the state has assumed responsibility for capital costs) and a sense of system ownership by the users themselves.

Back to text

⁽⁷⁾  Market-based land reform is preferred to expropriation by law and redistribution through a State agency, despite the complexity of establishing market value. Land acquisition by smallholders would have to be assisted by suitable credit mechanisms to avoid an excessive initial debt-service burden, and be supported by a strengthening of agricultural services. For a more complete elaboration of the issue see WAT2010.

Back to text

⁽⁸⁾ See also WFS companion paper 11 Food production and environmental impact.

Back to text

⁽⁹⁾ See also WFS companion paper 8 Food for consumers: marketing, processing and distribution.

Back to text

⁽¹⁰⁾ See also WFS companion paper 3Socio-political and economic environment for food security.

Back to text

⁽¹¹) See also WFS companion paper 9Role of research in global food security and agricultural development.

Back to text

⁽¹²⁾ The model of Boserup (1993), prominent protagonist of population-induced agricultural innovation, claims that new technologies will be taken up when population pressure reaches a given point, but not before. According to this model it would be useless to extend new technologies among farmers when the situation is not ripe for their adoption.

Back to text