8

GLOBAL CHALLENGES, POTENTIALS AND PRIORITIES

Widespread and severe household food insecurity occurs in many farming systems and this is expected to remain a major concern during coming decades. Indeed, available projections1 point to a probable failure to meet the commitments made by the international community at the 1996 World Food Summit and in the Millennium Declaration to halve hunger and poverty by 2015. In all regions, poverty and household food insecurity are more prevalent and more severe in rural areas than in urban areas. Futhermore, food insecurity and poverty are interrelated. Poor households lack the purchasing power to ensure adequate nutrition, while food insecurity in itself causes poverty. Malnourished people, because of reduced intellectual and physical capacities, often suffer more illness, and have lower labour productivity and weak educational achievements. Whilst there is an urgent need for transitory measures to safeguard minimum nutritional and livelihood standards (e.g. safety nets), long-term programmes are required to support a sustainable process of pro-poor economic growth and development.

In this Chapter various ways of closing the hunger and poverty gap, which exists between current projections (a business as usual scenario) and the international development goals of halving hunger and poverty by 2015, are examined. The Chapter commences by grouping the 72 farming systems identified in the six regions into eight broader categories. The relative importance of household strategies for escaping from poverty is then examined in relation to each category, in terms of: intensification; diversification; increased operated farm size; increased off-farm income; and exit from agriculture within the farming system. In order to shed further light on strategies for poverty reduction, the farming systems are then re-grouped according to: (i) resource potential, and (ii) agricultural intensity (as an indicator of access to services).

Secondly, a global strategy for hunger and poverty reduction is defined, focusing on five areas; the reform of the policy and institutional environment, measures related to efficient markets, improving the availability of information and human capital, effective technology deployment and the application of sustainable natural resource management.

System Categories, endowments and household strategies

Farming System Categories

The 72 farming systems identified in the six developing regions can be grouped into eight major categories (see Chapter 1 and Annex 2), based on the characteristics described in Chapters 2 to 7, in order to facilitate comparison and integration of individual system priorities into an overall global strategy for poverty reduction. These eight system categories, are: (i) irrigated farming systems; (ii) wetland rice based farming systems; (iii) rainfed farming systems in humid (and subhumid) areas; (iv) rainfed farming systems in steep and highland areas; (v) rainfed farming systems in dry or cold areas; (vi) dualistic farming systems with both large-scale commercial and smallholder farms; (vii) coastal artisanal fishing mixed farming systems; and (viii) urban based farming systems. Except in the case of the dualistic category, these system types are dominated by smallholder producers.

The quality of underlying resources varies widely between farming system categories and differences can be noted even within a single farming system. Farming systems endowed with irrigation or humid climates normally have higher agricultural potential than systems in dryland areas. In addition, access to agricultural services - especially markets - has a major influence on the development opportunities of a farming system. Figure 8.1 shows the relationship of seven of the eight system categories in terms of these two important variables.

The domains of each system category are relatively large because of the heterogeneity of their constituent farming systems and differences between farm household environments within a single system. In this connection, dualistic systems would tend to cover the entire Figure and thus were omitted.

As can be seen from Figure 8.1 (and as discussed in earlier chapters), the irrigated, wetland rice based and, to a lesser degree, rainfed humid system categories tend to have relatively high average household resource endowments. However, their access to agricultural support services and markets varies markedly - from poor (towards the upper left corner of the Figure, e.g. Gezira Irrigation Scheme in Sudan) to good (in the upper right, e.g. Lower Nile Valley irrigation areas in Egypt).

In contrast, often because of poor infrastructure, rainfed highland farming systems have restricted access to services but resource endowments can vary from fairly good (in the upper left corner, e.g. Upland Intensive Farming System in China) to poor (in the lower left corner, e.g. High Altitude Mixed Farming System in the Central Andes). Rainfed farming systems in dry or cold areas tend to have poor resource endowments and poor access to services (e.g. Pastoral Systems in all regions). On the other hand, Coastal Artisanal Farming Systems often have good access to services but the underlying resource base varies, placing the domain of this system on the right hand side of the Figure. Finally, Urban Based Farming Systems typically have good access to markets, but their resource base is typically quite restricted.

The eight categories of farming system are further compared in Table 8.1, which shows the areas of total land, cultivated land and irrigated land, agricultural population and market surplus. The six irrigated and rice based wetland systems2 contain an agricultural population of nearly 900 million people with some 170 million ha of cultivated land, of which nearly two-thirds is irrigated. There are three major classes of smallholder rainfed farming system (in humid, highland or dry/cold areas), which together contain an agricultural population of more than 1400 million people with around 540 million ha of cultivated land. Dualistic systems comprising farms of mixed size contain a further 190 million farm people with a cultivated area of 11 million ha. Finally, two further minor classes of smallholder system -four coastal artisanal fishing mixed and six urban based systems - contain a combined total of about 100 million people. These eight categories of farming system are described in the following paragraphs.

The three smallholder irrigated farming systems are dependent on large-scale irrigation schemes dominated by small-scale farming. Although this category contains only about 30 million women, men and children who farm about 15 million ha of irrigated land, it is important for national food security and export earnings in many countries. Although huge investments have been made in large irrigation schemes, most still face the challenge of improving water use efficiency in coming decades. In addition, many are now facing financial difficulties; notably in Sub-Saharan Africa, Eastern Europe and Central Asia. Environmental sustainability is also a concern in a number of cases3. Outside the context of large-scale schemes and wetland rice systems, small-scale irrigation supplements the livelihoods of many millions of rainfed farm households, and is considered an integral part of the particular farming system in which it is located.

The three wetland rice based farming systems of East and South Asia, which include a substantial proportion of irrigated land, support an agricultural population of around 860 million. Although bunded rice cultivation is the distinguishing characteristic of these systems, a wide range of other food and cash crops are produced and poultry and livestock are raised for home consumption and sale. These systems depend on the monsoon, but nearly 60 percent of the cultivated land is equipped with irrigation facilities. Relatively little grazing or forest land remains - almost half of land is under annual or permanent crops - and these systems suffer from intense human pressure on the natural resources base, with 5.5 persons per ha of cultivated land. There is a moderate degree of market surplus; these systems underly national food security in most Asian countries.

The 11 rainfed humid farming systems are based on smallholder cultivation of root crops, cereals or tree crops. They often contain an important component of livestock and support an agricultural population of approximately 400 million. There is little irrigation. Pressure on land is typically moderate - only 2.5 persons per cultivated ha on average - although there are some areas of intense pressure. There is a moderate, but varying, degree of market development and substantial opportunities for further development.

The 10 smallholder rainfed highland farming systems in steep and highland areas contain an agricultural population of more than 500 million. In most cases these are diversified mixed crop-livestock systems, which were traditionally oriented to subsistence and sustainable resource management. However, these days they are characterised by intense population pressure on the resources base, which is often quite poor - averaging 3.5 persons per cultivated ha, aggravated by heavy grazing pressure on the four-fifths of the land which is not cultivated. Given the lack of road access and other infrastructure, the level of integration with the market is often low.

The 19 smallholder rainfed dry/cold farming systems in dry or cold low potential areas, cover an enormous land area - around 3.5 billion ha - but support a relatively modest agricultural population of around 500 million. These lower potential systems are generally based on mixed crop-livestock or pastoral activities, merging eventually into sparse and often dispersed systems with very low current productivity or potential because of environmental constraints to production. Market development in these extremely low potential areas is limited.

The 16 dualistic farming systems are characterised by significant contrast, i.e. a mix of large, often commercial, farms together with smallholder farms. This category contains an agricultural population of nearly 200 million and more than 400 million ha of cultivated land in a variety of ecologies, and exhibits diverse production patterns. Such systems are prevalent in Eastern Europe, Central Asia and Latin America, but can also be found in Africa. All except one are predominantly rainfed systems - the exception being the Irrigated Farming System in Eastern Europe and Central Asia, which is dominated by medium and large farms.

Four coastal artisanal fishing mixed farming systems have been defined4. The crop component of these systems is important for household food security, but the principal livelihood is inshore fishing, with a rapid growth in aquaculture in many parts of the world. Because of infertile soils crop yields are often low. The few areas with fertile soil often face serious risks of storms and floods - as occurs around the Bay of Bengal. Many systems include some tree crop production (e.g. coconut and cashew) and small livestock, especially goats, and poultry.

Six urban based farming systems have been identified, that are dynamic, market-driven systems, typically focused on intensive, high-value horticultural and animal production. Urbanisation is increasing everywhere and these farming systems will assume even greater importance in the future.

The analysis of the eight global categories of farming systems suggests significant differences in the relative importance of the strategies available to poor households for reducing poverty and hunger. Derived from expert judgement, Table 8.2 shows the relative importance of these household strategies - which also correspond to rural development strategies on a wider scale. The relative potential impacts suggest an optimal mix of strategies, required for closing the gap between the `business as usual' projections of slow poverty reduction and the goal of halving hunger and poverty by 2015 in each of the major farming systems.

The irrigated farming systems have relatively low incidence of poverty and hunger, and are fortunate in having good prospects for halving existing poverty. It is judged that three-quarters of the reduction is poverty will be derived from farm improvement (intensification, diversification and increases in operated area). Fully one-third will be derived from production intensification - because growing conditions are so favourable in these schemes, productivity increases will arise from improved water, soil fertility and crop management - and diversification, often involving export cash crops with established marketing channels, will provide a further 30 percent. Nevertheless, increases in off-farm income, from both local and seasonal migration, are expected to contribute nearly 20 percent of aggregate reduction of poverty.

The wetland rice based farming systems will be developed largely through farm improvement, which will contribute some 60 percent of the reduction of poverty - notably through diversification of crop, livestock and fish production. As one aspect of diversification, local processing of farm produce will also add value and income. Intensification is also important (especially in countries such as Cambodia and Myanmar), although less so than the expected one-third contribution from the increase in non-farm income. The importance of off-farm income reflects the strength of the rural off-farm economy in China, which accounts for nearly half of the farmers in this system category.

The rainfed humid farming systems depend, to a significant degree, on all five household strategies for the halving of poverty. Among these strategies, diversification will contribute more than one-quarter of poverty reduction and increased off-farm income slightly less than one-quarter. In these crop-livestock systems, livestock will play a major role in diversification; and small scale farmer managed irrigation will contribute to both intensification and diversification.

The rainfed highland farming systems offer fewer prospects for farm improvement, which accounts for about 40 percent of poverty reduction. The driving forces for poverty reduction will be increases in off-farm income and emigration (exit from agriculture), which both contribute about one third. Diversification, especially to high value products with relatively low transport and marketing costs, will also contribute significantly to poverty reduction.

The rainfed dry/cold farming systems have a similar pattern of poverty reduction to the highland systems, because of their low agricultural potential and the poor marketing infrastructure. Livestock and irrigation development will play important roles.

The dualistic farming systems will derive poverty reduction from all five household strategies, with more than two-thirds arising from farm improvement. Both intensification and diversification will be supported by irrigation and market development. Increases in operated farm size will contribute around one-fifth of aggregate poverty reduction.

The coastal artisanal fishing mixed farming systems are heavily dependent on increased off-farm income for poverty reduction, with substantial contributions from diversification and exit from agriculture.

The urban based farming systems also depend for poverty reduction on increased off-farm income and diversification, with significant contributions from intensification and increased size of business.

Implications of farming system resource endowments

Both existing livelihood levels and the potential for future improvement depend upon the quality and availability of natural resources5. The resource base of a farming system is best conceptualised as the average resource endowment of typical farm households, measured according to their productive potential when using existing technologies. With few exceptions, all farming systems can be classified in this manner, by taking into account both average farm size and the quality of natural resources. When approached in this way, resource potential can be viewed as a continuum running from systems situated in arid areas to those located in fertile, irrigated conditions.

Examples of farming systems in low potential areas are: (i) the Agro-Pastoral Millet/Sorghum System in Africa; (ii) the Rainfed Mixed System in South Asia; and (iii) the High Altitude Mixed (Central Andes) System in Latin America. Farming systems in high potential areas include: (i) all Irrigated Systems; (ii) the Cereal-Root Crop Mixed System in the moist savannah of West and Central Africa; (iii) the Tree Crop System in East Asia and Pacific; and (iv) the Extensive Mixed (Cerrados and Llanos) System in Latin America. A summary of the contrasting characteristics of high and low potential farming systems situated at the two opposite ends of the spectrum is presented in Table 8.3.

More than 60 percent of the agricultural population of developing regions - or around 1450 million people - live in systems with relatively high potential for increased productivity from the viewpoint of resources. Farming systems in higher potential areas have greater average population densities and generally enjoy better economic and social infrastructure than low potential areas. However, these areas are not without constraints: in fact, many environmental and socio-economic factors impede agricultural development and extreme fragmentation of holdings may also be a problem. Moreover, contrary to popular wisdom, a significant proportion of the world's poor and hungry people is also found in areas with higher potential.

Farming systems in low potential areas with low or erratic rainfall and poor soil fertility, tend to have relatively few agricultural development opportunities, and farmers may be more concerned with minimising risk than maximising food production or profit. Low potential areas tend to face higher risks, particularly in the distribution and amount of water available for farming needs. Other natural disasters, such as locust attacks, may also be important. Unfortunately, the range of options open to smallholders for livelihood improvement from off-farm employment in these low potential areas is also more restricted than in high potential areas as a result of limited market development.

Poverty reduction in low resource potential systems therefore, often depends upon seasonal or permanent migration to seek employment as labourers in wealthier systems, or to urban areas. The need for alternative livelihoods is striking: exit from agriculture contributes almost one half of the poverty reduction potential and increased off-farm income around one quarter. In both cases, a substantial proportion of seasonal and permanent migration will be long distance and sometimes across national boundaries. Nevertheless, intensification, diversification and increased farm or herd size together contribute about one third to aggregate poverty reduction, largely centred on livestock or irrigation development. Relationships between household improvement strategies and resource potential are encapsulated in Table 8.4 below.

Farming systems in the high potential areas, by contrast, depend principally on diversification and off-farm income for poverty reduction, and together these factors contribute nearly 60 percent of the improvement potential. Intensification will also make a substantial contribution, but may be constrained by land fragmentation in some cases. The linkages to poverty reduction resemble those for the irrigated, wetland rice based and rainfed humid systems. Overall, high potential systems offer the principal hope for expanded food production in the future, and will thus justify a considerable proportion of rural investment funds targeted at expanding global food production.

Implications of agricultural intensity and access to services

Within the range of farming systems there is a gradation in the level of production intensity which is closely related to access to agricultural support services (see Table 8.5). Low intensity systems - including agro-pastoral, pastoral and sparse agriculture systems - support 350 million people but have scattered populations, extensive land use practices, low levels of input use and little market surplus. These systems tend to coincide with the low resource potential areas discussed above.

Medium intensity farming systems have evolved into two distinct sub-types - food oriented and market oriented - depending on the forces involved. Where population pressure on resources has been the dominant factor, systems have adjusted towards intensive food crop production - including root crops, plantains and cereals - often within a mixed crop-livestock system. Systems of this type cover more than 2000 million ha and support an agricultural population of about 950 million people; they include many of those systems most associated with rural poverty. A moderate level of market development is evident, associated with higher intensity of external input use and greater intensity of land use.

The second sub-type - market-oriented medium intensity - contains fewer systems and development has been driven principally by the existence of readily accessible market opportunities, whether local (typically horticultural) or international (often based on tree and industrial crops). With an agricultural population of only 100 million people, these systems are less important in terms of poverty reduction than they are in terms of export earnings.

High intensity systems will be a major factor in increasing global food security. These include irrigated and wetland rice systems (where more than 50 percent of the cultivated land is irrigated), as well as systems with reasonable resources and infrastructure. Despite a high population pressure of almost five persons per cultivated ha, and small average farm sizes, these systems have an advanced level of market development with significant external input use and a substantial amount of irrigation - allowing them to generate a surplus for sale to urban areas and export markets.

The relative importance of potential strategies for poverty reduction by system intensity is shown in Table 8.6. The low intensity farming systems resemble, from the viewpoint of sources of reduction of poverty, the farming systems in the low potential areas (see above). Whilst farm improvement (intensification, diversification and increase in farm size) will account for around 40 percent of poverty reduction, the greatest driving forces are exit from agriculture and increase in off-farm income.

The food-oriented medium intensity farming systems will depend for poverty reduction almost equally on farm improvement and alternative livelihoods (increased off-farm income and exit from agriculture). Logically, among the farm improvement group, diversification away from an emphasis on food production will be the most important source of poverty reduction. The commercial medium intensity farming systems resemble the food-oriented systems, except that farm improvement will contribute around 60 percent of poverty reduction. Interestingly, both intensification and diversification will play important roles.

The high intensity farming systems resemble the commercial systems, except that diversification will be relatively more important, and intensification relatively less important, presumably because substantial intensification gains in the high intensity systems have already been realised. Farm improvement will contribute around 60 percent of poverty reduction, but off-farm income is also expected to play a major role in poverty reduction.

Overall importance of different household strategies for poverty reduction

While the above depiction is inevitably a simplification of reality, it provides an insight into the components of a global strategy that are likely to be most fruitful for poverty reduction within any given type of system, and hence can indicate approaches - whether national or international - that will most strongly justify investment. Farm improvement will be, overall, the most important source of poverty reduction; and for higher potential systems, at least, the main engine whereby the international development goals of halving hunger and poverty will be realised. Within farm improvement, on-farm diversification will be the most important source of poverty reduction, while intensification has an important contribution to make in a more restricted range of systems. Nevertheless, intensification and diversification often proceed in parallel. Among alternative livelihoods, off-farm income will be the most important source of poverty reduction, with a contribution nearly as large as diversification. The following paragraphs examine the roles of each of the five strategies with respect to poverty reduction.

Further intensification of existing production patterns will be a significant source of poverty reduction in four farming system categories (see Table 8.2), and will be dominant in the development of irrigated and dualistic farming systems. It is of medium importance in rainfed humid and wetland rice systems and of relatively minor significance in two other system categories. As might be expected, it is of medium importance in high potential systems overall yet, counter-intuitively, is a higher priority in commercial medium intensity farming systems than in food crop or low intensity farming systems.

Diversification is rated as being an important means of reducing poverty in all categories of farming system. On-farm diversification entails an increasing emphasis on non-traditional cash crops - particularly vegetables, fruits, spices and colorants, livestock products and aquaculture - especially those that are labour intensive. In addition, agro-processing and other post-harvest activities that add value, are judged to be a promising development avenue. Diversification is expected to be most successful in high intensity farming systems in areas of high potential.

The opportunity to increase farm size seems to be a significant option only for poor households in rainfed humid farming systems, where pressure on land is only moderate and land development potential is good, and especially in dualistic farming systems. Most other categories of system have little short-term opportunity for expanding agricultural areas and any increase in farm size will depend on farm land consolidation. Nevertheless, highly intensive systems in East Asia and, to a certain extent South Asia, have reached such small average farm sizes that some land consolidation may be essential if production intensification programmes are to be effective. Some potential to expand the operations of poor urban based farm households also exists.

The existing trend to increased off-farm income is expected to continue, particularly among poorer farmers. The future exploitation of this source of income is anticipated to be a key source of poverty reduction in all systems, but is of particular importance in the highland system category, as well as in rainfed systems in dry and cold areas. Because of excessive land division and fragmentation caused by high population density, land resources available to farmers are often too restricted to support viable modern farming techniques. The ultimate solution for many families is to move away from farming and seek alternative employment, often through migration to larger urban areas, but sometimes by relocating to frontier areas (e.g. the Extensive Mixed System in Latin America or the Cereal-Root Crop Mixed System of West Africa). This strategy of exit from agriculture within the system is important in low potential areas as a whole, especially where few opportunities exist for supplementing household income with local off-farm employment. It is expected to be particularly prevalent in the rainfed humid, highland and dry/cold categories of farming system, as well as in the coastal artisanal fishing category.

Although exit from agriculture is more significant for low intensity systems, the regional analyses suggest that other household strategies are relatively little affected by the level of intensity of the farming system. Intensification, diversification, increase in farm size and off-farm income are all classified as having the same degree of importance at all four levels of intensification. Nevertheless, it is expected that diversification and increased off-farm income will generally be easier to achieve in high intensity systems, due to improved market and service development, while increased farm size may well be feasible only in lower intensity systems6.

A GLOBAL STRATEGY FOR REDUCTION OF HUNGER AND POVERTY

The principal regional initiatives highlighted as a result of the analysis of farming systems and presented in the preceding chapters, are summarised in Table 8.7. Although the titles of the initiatives may be similar, the actual components of the identified initiatives vary across the regions. The initiatives can all be related to the five broad areas of focus employed throughout the book: (i) policies, institutions and public goods; (ii) trade liberalisation and market development; (iii) information and human capital; (iv) science and technology; and (v) natural resources and climate.

Policies, institutions and public goods

There is a strong case for policy reform at national and global levels in order to support the sustainable development of farming systems. Reforms would include the creation of conducive macro-economic frameworks and expanded investment in public goods and services, in order to compensate for various market failures in research, infrastructure, etc. Many countries have embarked on reforms that have led to a less interventionist public sector, thus opening new opportunities for the provision to farmers of a wider range of services from multiple sources. What is needed now is to ensure that policies and institutions underpin smallholder development, as well as expanding the capacity of private sector service suppliers. This focus is particularly relevant to Africa and to Eastern Europe and Central Asia.

Most reviews of policies and institutions suggest that governments should complete the withdrawal of direct public support from viable commercial farming and privatise associated services, such as seed production and marketing. Nevertheless, it is clearly in the public interest that governments should continue to ensure reliable access by the small farmer sector to relevant public goods, as well as promoting the sustainable use of natural resources. Government efforts should be devoted to clear cases of public goods supply; including roads, education and health services, and research and extension facilities addressed to the needs of poor farmers and marginal areas (see Box 8.1). They should also concentrate on the enforcement of regulations (e.g. fair trade, food standards) with a particular emphasis on avoiding barriers to entry for new participants.

Some success has been achieved with the outsourcing of public service and infrastructure provision to private firms, NGOs and universities, thereby achieving efficiency gains. This type of innovation must include a strong element of local participation, which is critical for monitoring the private provision of such goods and services. Farmers' organisations and the private sector can play a key role in many areas, such as seed multiplication and varietal development (see Box 2.5 on Community-Based Seed Supply Systems). In addition, exporters' associations can often implement phytosanitary inspection. Moreover, research priorities and budgets can be managed through competitive bidding and public-private cost-sharing arrangements.

The five most important priority thrusts in reducing hunger and poverty, concerned with improvements to policy, institutional and public goods, are seen as:

• Establishing equitable, secure, transferable and flexible resource user rights;
• Providing sustainable infrastructure to poorly serviced areas;
• Supporting small-scale farmer managed irrigation schemes;
• Continuing agricultural policy reforms and strengthening meso-level institutions; and
• Strengthening targeted safety nets.

Establish equitable, secure, transferable and flexible resource user rights

The efficient and sustainable utilisation of both individual and common property resources requires clear, enforceable and transferable user rights; as well as functioning markets for those rights, plus taxation policies which encourage efficient resource use. Advisory services need to acquire more insight into the specific mechanisms whereby access to land, water, grazing rights, forest areas and other natural assets, are managed by communities.

The recognition of customary land rights - especially rights to common property resources - is critical to ensuring their productive and sustainable use in many systems with low population density. It would also permit communities operating under traditional land rights systems to arrange commercial relationships with agribusiness groups. Existing legislation, in attempting to protect such customary rights, often hampers arrangements for contracts and other legal agreements on land use7. Finally, poorly functioning or restricted land markets contribute to land fragmentation and ever decreasing average farm sizes. This is starting to become a significant contraint in the more intensive systems, and projected gains in production intensity, as well as diversification, are likely to be impeded if land market functioning is not facilitated.

In countries that have inherited dualistic agrarian systems, positive changes in land access can be encouraged through phasing out of subsidies to large-scale commercial farmers, as well as by applying fiscal disincentives to non-productive use of land. Development of land markets can often be advanced through the recognition of informal land rental and sharecropping transactions, facilitation of financing for land purchases and effective measures to protect women's (especially widows') access to land.

Provide sustainable infrastructure to poorly serviced areas

The rate at which farming systems develop towards market-based models is closely correlated with the availability of infrastructure. Road access and electrification have been particularly important in this regard and communications are of growing importance. Improving farmers' access to basic economic and social infrastructure reduces transaction costs. Agricultural knowledge underpins the improvement of farming systems and is usually associated with demand for improved educational services. Therefore, increased investment in rural infrastructure, especially when focused on transport and low cost communications for small-scale farmers, should be a high priority for governments.

With regard to the provision of infrastructure, the main problems in the past have been: (i) urban bias in public expenditure; (ii) inability of local governments to generate enough revenue for operation and maintenance; and (iii) lack of consideration of community priorities and participation in at the planning stage, leading to a passive dependence on government. Because of limited public investment capacity, careful targeting of infrastructure development is crucial in order to meet the needs of poor smallholders.

The reduction of poverty is often most difficult in areas of low potential where agricultural growth prospects are limited and where low population densities increase the per capita cost of service and infrastructure provision. As a consequence, relatively little public investment has been made in these areas, but the social justification is particularly high since private sector interest is difficult to attract.

Support small-scale farmer managed irrigation schemes

In most situations, policies for the development of water resources should give priority to small-scale, farmer managed irrigation, which is usually more sustainable and cost-effective than large-scale schemes. Moreover, rehabilitation of existing schemes should generally take precedence over construction of new schemes. In either case, the establishment of secure land and water rights and the building of effective user-driven local management institutions is an essential complement.

For existing large-scale schemes, the main thrust should be to make them more sustainable and competitive by encouraging greater farmer participation in their management, reducing O&M overheads and improving the efficiency of water supply and use. These changes involve the formation or strengthening of water users' associations, as well as reinforcing farmers' skills in business and on-farm water management.

Continue agricultural policy reform and strengthen meso-level institutions

Many developing countries have already been through a process of structural adjustment. The priority is to complete policy reforms already started by governments (especially in Eastern Europe and Central Asia), but at the same time, to safeguard poor rural communities and families whose vulnerability is often accentuated by adjustment measures - at least during the process of transition. This latter aim may require a combination of targeted short-term measures to assure access to adequate food, as well as structural steps to improve livelihoods through raising the productivity and risk bearing capacity of small farmers. In many systems, population pressure already exceeds theoretically sustainable levels and the main challenge is to identify alternative livelihood opportunities - either inside or outside the area. In some cases this may involve encouragement of out-migration, facilitated through appropriate skills training, capital endowment and the establishment of efficient land markets.

The reduction of biases against agriculture and rural development, particularly those that disadvantage the poor, is an important element of policy. Some of these biases have their origins in the goal of providing cheap food to cities; others are related to natural resource preservation and can have the effect of limiting access to resources by the poor more than the rich - one example is the case of forest regulations in the Highland Mixed Farming System in South Asia. It is not only policies per se that disadvantage the poor, but also the way in which they are implemented, as this takes place at the local level in a manner that often favours the powerful and discriminates against the poor.

Improved techniques must be developed to facilitate the assessment of risk in relation to policy and investment decisions, particularly in the light of the expected increase in climatic variability. All too often risk is not explicitly considered as a critical factor in small farmer decision making. Yet a wide range of decisions - including the choice of crop research priorities, the planning and provision of water and irrigation infrastructure, and the formulation of policies affecting land management - are all significantly affected by variability and risk. In addition, uptake of new technologies by small farmers is often inversely correlated to the perceived risks of adoption.

There is an increasing recognition that many traditional farming systems incorporate features that: (i) permit participants to reduce or share risks; (ii) make efficient use of resources; and (iii) resolve potential conflicts over resource allocation, while at the same time ensuring the long-term sustainability of limited natural resource endowments. This has been especially the case with pastoral and dryland farming systems, where attempts to replace traditional practices with new production technologies have often failed dismally, leading to increased poverty and breakdowns in existing resource management systems.

Little progress has been made in creating effective insurance mechanisms for small farmers producing in such difficult environments, and further work is needed on risk mitigation mechanisms that combine low per capita costs, acceptability to participants, and effectiveness in cushioning unpredictable shocks. A degree of insurance can, however, be built into farming systems, through diversification, inter-cropping, crop-livestock integration and the use of crop varieties resilient to stress conditions.

Strengthen targeted safety nets

As noted in the next section, the process of market-based development is likely to increase poverty - at least in the short to medium term - among many farm households. Governments have legitimate social responsibilities towards these households that transcend market functioning, particularly with respect to the protection of disadvantaged citizens (children, women, refugees, and those suffering from disasters). Even though social support programmes may raise issues of dependency, the problem of malnutrition and hunger must be addressed as a matter of urgency. Government intervention can complement market functions, on both humanitarian and economic grounds, instead of attempting to replace markets through general price supports, subsidies or tariffs. This complementarity could be achieved by providing carefully targeted support - for example through food security programmes in which communities are engaged in participatory planning to ensure adequate nutrition for all members. It could also involve adjustments in access to land and water resources, strategic changes in production systems (e.g. the addition of backyard vegetable production), and improvements in feeding habits and sanitation. Food assistance (e.g. school meals, food for work to create new infrastructure and productive assets such as small-scale irrigation) for vulnerable groups who are unable to translate their food needs into effective demand, is also of great importance, and can be designed to stimulate local markets.

Trade liberalisation and market development

At present, the movement towards reduced barriers to international trade is well established, although the pace of change is uncertain. This process will affect all but the most isolated producers in farming systems throughout the developing world. However, the type of changes induced - and their impacts on production, poverty and food security within individual systems - will depend on a variety of factors; most notably resource availability and market environment, future advances in technology (preservation techniques, transport, communications, etc.) and changes in consumer demand.

Evidence from a number of systems (Maize-Beans Farming System in Mesoamerica, Maize Mixed Farming System in Eastern Africa in the 1980s, and the Lowland Rice Farming System in East Asia) demonstrates that smallholder farmers can participate successfully in market-driven growth and significantly increase household income. Nevertheless, those farmers unable to adjust, due to a lack of resources or an unfavourable policy and institutional environment - such as are now evident in the Maize Mixed Farming System in Southern Africa - will remain dependent on traditional staples for cash income. As such, they face long-term declining incomes as reduced trade barriers and new technologies reinforce the downward trend in international prices of major food commodities.

Based on the analysis of farming systems in each of the developing regions, the five most important global priority thrusts concerned with maximising benefits from trade liberalisation and market development in order to reduce hunger and poverty, are seen as:

• Ensuring trade liberalisation is a two-way street;
• Focusing smallholders on labour-intensive or niche products;
• Satisfying household food security needs during transition;
• Fostering competitive business, especially small rural enterprises; and
• Supporting agricultural market development.

Ensure trade liberalisation is a two-way street

Although trade liberalisation is beneficial, even when enacted unilaterally, it is undeniable that those benefiting from reduced barriers to entry for agricultural products are principally urban populations, while those that benefit from reduced barriers to exports are principally agricultural producers. The considerable barriers still remaining to the entry of major categories of agricultural goods to markets in areas such as Europe, North America and Japan, mean that trade benefits to developing countries have so far largely been in the first group - reduced prices for urban consumers - thus threatening traditional producers with increased poverty as a result of a combination of lower prices for food staples and continuing barriers to diversification.

While it is true that trade barriers in industrialised countries have been reduced for many agricultural commodities, these have often been for products with only limited domestic production or those where the developed world has a strong technological comparative advantage. Protection levels for products such as sugar, where tropical countries may have a strong comparative advantage, are still high. Furthermore, the perception exists - fairly or otherwise - that the industrialised countries are quick to use a range of escape clauses related to phytosanitary, environmental or policy matters to justify extraordinary controls on the import of agricultural products that may threaten domestic business interests. Small developing countries simply do not have the expertise, financial resources or global political clout to seriously challenge these controls. Where complaints are made - as in the case of the European Union controls on banana importation - the perception is that settlements are arranged to meet the needs of the industrialised world, leaving developing country plaintiffs on the sidelines.

This imbalance, which is also apparent beyond agriculture in such areas as textiles, is contributing to a profound unease among many developing country governments concerning the wisdom of liberalisation and may ultimately result in a backlash against the entire globalisation process that goes beyond street protests at major economic fora. The breakdown of the globalisation process would be a great pity, as real long-term benefits are possible for even small-scale producers in developing countries, but a world economic order built on a structure and process that are widely perceived as unfair will not be durable. International agencies, and in particular governments of industrialised countries, must recognise the seriousness of this risk and act to reduce perceptions of inequality in the design and conduct of multilateral trade agreements.

Focus smallholders on labour-intensive or niche products

Although smallholders will inevitably wish to continue producing staple crops for food security and cultural reasons, major urban markets will generally find cheaper supplies from larger high-technology producers - often in other countries. However, as demand for specialist foods grows, and new post-harvest technologies improve product quality and extend the life and durability of perishable items, small producers will have increasing opportunities to achieve attractive cash incomes through the production of specialised products where diseconomies of scale exist (labour intensive and niche horticultural products, fruits, spices, ornamentals, organics etc.). Small farm size is not an impediment to accessing these types of markets, which are expected to grow rapidly in importance, especially if farmers can combine forces through voluntary associations. Even producers who are geographically isolated have the potential to participate in these opportunities, through a focus on very high value-to-weight products (colorants, extracts, essential oils, etc.).

The transition to market oriented production of competitive products is seen as a crucial step in rural economic development and poverty reduction. Not only are incomes generated in those households directly engaged in production, but employment is also created in packaging, transport and related marketing activities, as well as in service provision to successful producers.

Satisfy household food security needs during transition

The restructuring of farm activities to benefit from new market opportunities requires access to a minimum level of resources - natural, human and financial - as well as the ability to accept a certain level of risk in the transition process. In many systems, only a minority of small producers can meet these conditions without significant external assistance. However, with their example, as well as the assistance of larger producers who are often pioneers in innovative production and marketing of crops, the later entry of other, less well endowed, producers is more feasible. The activities of pioneers, whether large or small, is also often important in creating indirect employment and demand for supporting services.

However, if the number of potential pioneers is too restricted there may be insufficient output to create a critical market volume for these products and the development of market-oriented production may fail. Farming systems with a high level of chronic and widespread poverty will often face this barrier. Not only will producers be more concerned about family survival than cash generation, but local buyers and service providers will be few and the establishment of market mechanisms more difficult. Assisting the transition to market-oriented production in systems with high poverty-levels may, therefore, be possible only once food insecurity is reduced, and a minimum level of cash flow is available through the sale of surpluses from traditional crops.

In these systems, attention will need to be focused first upon increasing output of food staples, improved family nutrition, and food security. This will require further investment in appropriate genetic material, food storage structures, natural resource management and education. Careful consideration may also need to be given to the feasibility of facilitating the reduction in rural populations through the provision of skills training and other programmes to assist successful outmigration.

Foster competitive business, especially small rural enterprises

There is ample evidence that direct public participation in marketing operations is generally ineffective and may even yield negative returns. Experience has shown that market development occurs most easily where private individuals - both producers and traders - play a leading role in determining the functioning of markets. From the perspective of small producers who have very limited market power, a competitive market environment in which a number of buyers are active, is crucial. Small producers appear to benefit most strongly from the presence of small traders, processors and other buyers. Such small-scale players are usually unable to control prices in the way that larger enterprises can, and they are also more willing to purchase in small quantities from isolated locations. They may even be members of the same community in which the producer lives.

Helping develop a favourable marketing environment for small producers often means supporting small-scale entrepreneurs through such areas as financing, information, and simplification of bureaucratic procedures which only larger enterprises can manage. It also involves the improved provision of public goods (see previous section). Associations of small traders, processors and exporters can also provide important support to small entrepreneurs, reducing barriers to entry and ensuring that their needs are taken into consideration when developing or reviewing sectoral policies.

This is not to say, however, that larger agribusiness enterprises are necessarily a negative influence, especially if operating in a competitive environment. The contributory role of larger farmers and processors as pioneers, sources of technical advice and credit, buyers from outgrowers and as sources of wage labour, should not be forgotten. Small-scale traders - especially in the case of export crops - must generally sell-on their product to larger enterprises in the regional or national capital. In this situation, farmers' associations can play a role in helping small producers deal more equitably with larger enterprises by increasing volumes and strengthening their bargaining power.

Support agricultural market development

Although governments have little direct role in market operations, the transition of farming systems to market-oriented production can be greatly assisted through the creation of an appropriate enabling environment, including: (i) appropriate rural infrastructure, especially roads, electricity and telecommunications; (ii) commitment to a stable and balanced exchange rate; (iii) establishing and enforcing market standards for both inputs and outputs; (iv) avoiding taxation of agricultural output, focusing instead on under-utilised assets or income; and (v) monitoring and enforcement of trade agreements that have an effect on smaller producers.

Given the importance of agro-industry as a potential source of demand for rural production - and its potential to create employment - market development can also be assisted by removing disincentives to the location of processing and related operations in rural areas. Financial incentives tend to encourage short-term responses from agribusiness that result in unsustainable investments and withdrawal of operations once financing (or tax breaks) cease. However, governments can encourage sustainable rural investment through matching private investment with public infrastructure provision, training of employees, assistance to contracted suppliers, the reduction of red-tape required for investment, and by ensuring favourable long-term fiscal policies.

Finally, the public sector can support strengthened contractual arrangements between producers, intermediaries and processors. Such agreements may be for simple delivery of products, or may include co-operation throughout or in parts of the production process. Market development will be seriously impeded if these contractual arrangements are poorly defined or inadequately protected under law. Conflict resolution procedures need to be low-cost, timely and seen as fair by both parties, while enforcement of contracts should be guaranteed.

Information and human capital

Three important global priority thrusts, concerned with enhancing the benefits of agricultural information dissemination and use and developing human capacity, have been identified to reduce hunger and poverty:

• Ensuring wide availability of agricultural information;
• Providing broad, systems-oriented agricultural training, especially for women; and
• Strengthening off-farm vocational skills, especially for youths.

Ensure wide availability of agricultural information

Globalisation, urbanisation, and the accelerating pace of technological change, are all increasing knowledge requirements within farming systems in all regions. New approaches must be developed to support information flows between farmers and formal knowledge sources - such as research institutions and markets - as well as horizontally among farmers themselves. Small farms do not generally have the same degree of access to agricultural technology and market information as large farms. Thus, public investment in the dissemination of agricultural information is of particular relevance to small-scale farming systems. In some regions, information technology systems are already being adapted to respond to the information needs of small farmers (see Box 8.2), but further public support will be required before the full potential of information technology systems can be exploited, especially in remote areas and where poverty is prevalent. The need is to develop effective mechanisms for information generation, transfer and use by beneficiaries.

More thought must also be given to the long-term sustainability and relevance of agricultural information systems, which generally require significant levels of operating resources. There is little evidence that users are willing to pay enough to make such services self-supporting, while pressure on government recurrent budgets often means that services descend to `least-cost' solutions once external financing is withdrawn. This all too often results in out-of-date information that is of little relevance to the real needs of the users (e.g. market price information is given for international markets because it is available free, while data for nearby markets are omitted because they are too expensive to collect).

Provide broad, systems-oriented agricultural training, especially for women

Training and capacity building involve the empowerment of community members, enabling them to identify problems in a systems context, to analyse causes and effects, to assess options and to arrive at well-informed decisions in order to prepare to take responsibility for their own future by assuming ownership of the development process. This often requires the assistance of professional facilitators to ensure that all groups in the community have a chance to express their views, to help find appropriate technical solutions, formulate and implement community action plans, monitor results and assess outcomes. The implication of this approach is that extension services must reorient their operations by basing them on facilitative rather than prescriptive techniques; with community participation forming the keystone in determining priorities and testing possible solutions. This will involve the adoption and adaptation of participatory diagnostic tools and experiential learning methods. An essential concomitant is the formation and strengthening of common interest or resource user groups, that will strengthen the problem-solving capacities and self-reliance of rural communities. Implementation is likely to involve partnerships between governments, private sector or NGO service providers, civil society organisations and community-based groups.

With the decline in public resources for research and extension, many systems are chronically short of personnel who possess skills related to systems analysis, participatory methods and qualitative analysis techniques. The percentage of female professional staff is also deficient in most cases. To ensure that training of agricultural support professionals is broadened - particularly in systems concepts and participatory practice - the curricula of agricultural training institutions should be revised and continuing efforts made to recruit and train more women officers.

Strengthen off-farm vocational skills training, especially for youths

The farming system analyses presented in previous chapters clearly indicates that not all farm households can expect to escape from poverty through agricultural activities. In fact, a major percentage of youths will participate directly in off-farm income generating activities or exit from agriculture; few will remain employed solely in agriculture throughout their lives. Governments must strengthen vocational training aimed at alternative employment, whether within a rural setting or through migration to urban areas. In addition, schooling for younger rural children also needs strengthening, particularly at primary and secondary levels.

Science and technology

Most of the growth in food production during the past three decades has resulted from the adoption of productivity-boosting technology in areas of high agricultural potential - particularly those with relatively high and reliable rainfall or equipped with irrigation. A major challenge in the coming decades will be to generate technologies that contribute to increases in agricultural production and improvements in livelihoods in lower potential areas. For the longer term, there must be concerns about the heavy reliance of intensive agriculture on technologies that have inherently limited sustainability. `Factory' livestock farming systems are faced with enormous problems of organic waste disposal and are increasingly seen as sources of food safety problems. Inappropriate fertiliser application is leading to nitrate pollution of surface and groundwater resources, while the uncontrolled use of pesticides is creating enormous health and environmental hazards. Equally worrying is the progressive narrowing of the genetic breadth of farm crop and animal species.

The technical foundations for `modern' agriculture can no longer be taken for granted and there is a need to search for more sustainable strategies towards intensification. This task is urgent, given the limited extent to which thinking has been focused on more sustainable technologies for high-intensity farming, and the very long gestation period required to develop and disseminate new methods. The analyses of farming systems in this document suggest a number of important characteristics of technologies suitable for poor farmers. They also suggest some areas for technology development that offer opportunities for poverty reduction (see Box 8.3).

Five important global priority thrusts, concerned with targeting science and technology, have been identified to reduce hunger and poverty:

• Focusing technology through participatory research and development;
• Increasing land and labour productivity in high potential areas;
• Increasing labour productivity in low potential areas; and
• Promoting the adoption of biotechnology with safeguards.

Focus technology through participatory research and development

Except in regions where the Green Revolution has taken place, there are few opportunities for widespread dissemination of new agricultural technologies in the form of predetermined packages. Each farm differs in terms of its resource endowment; especially in the relation between land and labour resources, degree of access to input and output markets, and vulnerability to risk. Each household also differs in terms of needs and objectives; particularly in the extent to which farmers see production as contributing mainly to family food security as opposed to cash income. The need, therefore, is for participatory approaches to research and development that can engage farmers in diagnosing problems and in identifying possible solutions adapted to their particular circumstances. These approaches can also help to inform researchers of priority areas for investigation and enable them to better understand farmers' viewpoints and perceptions, thereby increasing the relevance of research.

Increase land and labour productivity in high potential areas

Technology does not move forward at an even pace. The major productivity gains made in rice and wheat, particularly in the 1970s and 1980s, have given way to a slower rate of growth in the last decade and it is still not clear how this trend towards stagnation can be reversed. Nevertheless, a number of important technological breakthroughs appear to be on the horizon - including `golden rice' and biological nitrogen fixation in non-legumes - that can be expected to have a significant impact on farming practices within the next three decades. In the meantime, substantial gains are possible from raising average yields towards the levels obtained by the best farmers, even under current technologies (see Box 8.4). It should also be remembered that there are traditional systems - older rice-based systems for example - that have remained highly productive because of skilled labour and management, despite continuous cropping for centuries. Nevertheless, decreasing average farm size and fragmentation of holdings is becoming an increasingly important constraint to development in a number of key Asian systems.

Different sets of technologies offer opportunities for growth in areas with high labour to land ratios, where there are often attractive options for the labour intensive production of high value commodities. A variety of high-value enterprises should be considered. Organic products offer one alternative (high value, high labour). Parts of many high-density systems are becoming peri-urban, providing good market potential in spite of population pressure - especially for vegetable growing and related agro-processing, small-scale dairying, small ruminant fattening and poultry production. Moreover, within these areas the abundant labour may be advantageously used to improve the productive capacity of land; for example through terracing or drainage.

Increase labour productivity in low potential areas

Areas of low agricultural potential often face more serious labour availability constraints than those related to land availability. Where productive activities have to be dispersed (e.g. transhumant herds, scattered production plots, food gathering), and where the functioning of the household requires lengthy journeys for basic inputs such as firewood or water, labour can rapidly become a key factor limiting output. Such limitations are sometimes exacerbated by out-migration or civil conflict, as has been the case in the Central Andean High Altitude Mixed Farming System or in many parts of Africa. Research on production technology has traditionally focused upon maximising returns to land rather than labour; replacing labour with capital only in more intensive systems (e.g. two-wheeled tractors in intensive rice systems). Yet there are good opportunities for improving labour productivity in many systems; examples include altering the time of land preparation from wet to dry season by using of cover crops to control weeds, by investing in draft power and by simple improvements to hand tools such as punch planters.

Where shortage of labour constrains production and incomes and there is easy access to more land, there is considerable interest in labour-saving technologies such as zero tillage and the use of draft animals for cultivation and haulage. Given the right stimulus, rural families will often convert labour resources into productive assets. Other opportunities for cutting cash expenditures without losing output include substitution of mineral nutrients by those from organic sources and changing from pesticide-based pest control systems to IPM.

Promote the adoption of biotechnology with safeguards.

Biotechnology offers a great potential for the customisation of new varieties of cash and food crops to specific farming systems and problems, if adequate safety standards are applied and provided that the new materials are affordable (see Box 8.5). The customisation of varieties, particularly those that can cope with abiotic stresses prevalent in low potential areas, has the potential to benefit poor farmers throughout the developing world.

To the extent that the lead in biotechnology applications will lie with the private sector, deliberate measures will have to be taken at the international level to create incentives for expanded research on themes relevant to poor farmers in developing countries - whose needs would otherwise be by-passed. Such measures could include international public funding for biotechnology research and development focused on improving the performance of tropical food crops as well as livestock. There is also a need for safeguards and regulations to reduce risk to indigenous genetic materials from contamination, and to guard against further erosion of genetic resources of farm crops and livestock. The regulatory mechanisms, including regional agreements and risk assessment protocols, need to be put in place as a matter of urgency in order to permit safe movement of useful genetic materials into and within developing regions.

Natural resources and climate

Increasing pressure on the use of scarce land and water resources, accelerating environmental degradation, and the possibility of climatic change are challenging the sustainability of farming systems in all regions, even those with low population densities. Four important global priority thrusts have been identified, concerned with achieving more sustainable and productive use of natural resources and minimising adverse climatic effects:

• Focusing on the sustainability of natural resource use;
• Recapitalising soil fertility;
• Improving water resources management; and
• Increasing capacity to respond to climatic changes.

Focus on the sustainability of natural resource use

There is now a heightened awareness, among both the public at large and farmers in general, of the need to conserve and productively manage natural resources. Public interest in industrialised countries assigns a high priority to the maintenance of natural resources for future generations and to reducing global environmental damage. This provides the main justification for international public funding towards the development and promotion of methods that maintain or enhance natural capital and increase global environmental benefits (e.g. increased carbon sequestration through higher soil organic matter levels, reduced production of greenhouse gases and less pollution of international waters, etc.).

Declining productivity and farm incomes on degraded lands have highlighted the need for farmers to improve the management of natural resources. Improved land management can be stimulated by the promotion of practices that not only generate environmental benefits, but also rapidly yield tangible returns. Thus, public research and extension should focus on measures that increase farm incomes whilst also conserving and enhancing the condition of natural resources. These measures are typified by minimum tillage technologies and integrated plant nutrient management, which simultaneously reduce production costs while improving in situ retention of moisture and raising soil fertility, thereby raising yields, reducing yield variability and cutting erosion.

In many situations, conservation agriculture, involving reduced tillage, offers promising possibilities for increasing labour productivity and the efficiency of input use while simultaneously reducing moisture stress. Conservation agriculture has been promoted in a number of farming systems, and its performance in Latin America - and more recently in Africa - has been promising.

Recapitalise soil fertility

The loss of soil fertility, with associated stagnation of productivity, cuts across most farming systems in all regions. The loss is particularly acute in irrigated wheat and rice-based farming systems, some rainfed farming systems (e.g. the Maize Mixed Farming System in Africa) and highland farming systems (e.g. the High Altitude Mixed (Central Andes) Farming System). Because of declining commodity prices, exchange rate adjustments and reduced subsidies, application of mineral fertiliser on staple crops has frequently become unprofitable and its utilisation by small farmers in most developing countries has fallen sharply. Farmers have not yet fully compensated for this loss through greater use of organic sources of nutrients, more balanced use of nutrient applications or adoption of soil conservation practices to minimise erosion. As the greatest concentration of nutrients occurs in surface soils, the degradation and loss of topsoil can severely reduce yields. The cost of investment in preventative measures is far less than the costs and time required to restore the original level of soil fertility once erosion has occurred.

Priority initiatives to rectify the current situation include: (i) greater use of green manures, enriched fallows and other sources of organic materials, including composting; (ii) expanded use of biological nitrogen fixation; (iii) better integration of crops and livestock; (iv) wider adoption of inter-cropping systems; (v) expansion of silvo-pastoral systems, especially on steeper slopes; and (vi) improvement of fertiliser import and distribution facilities and services with a view to reducing the farmgate price of fertilisers.

Most of these initiatives place little reliance on external inputs - and hence on input supply and financing services - and some have the advantage of enabling farmers to increase the productivity of their land through converting labour resources, which often have low opportunity cost in the off-season, into productive assets. This type of investment can play a extremely important role in building-up soil fertility, improving land management (e.g. terracing steep hillsides, draining bottomlands) and intensifying land use (e.g. planting tree crops, building fishponds). The creation of a policy environment that gives smallholders confidence in the future of farming, including reasonable security of land tenure, is an important trigger for this sort of initiative.

Improve water resources management

Although the adverse consequences of water shortages are more evident in some regions than others, water supply constraints are important in specific farming systems in all parts of the developing world, e.g. the marginal drylands of Latin America or the agro-pastoral systems of Sub-Saharan Africa. In most cases, the rising demand for water for domestic and industrial purposes associated with urbanisation will greatly intensify the competition for available fresh water. Where farming systems are rainfed, strategies must focus on improving the capture of rainfall and the utilisation of soil moisture. Possible measures include: (i) making minimum and no-tillage technologies accessible to small farmers, as done successfully in Brazil; (ii) providing true-breeding, short-season crop cultivars and drought resistant fodder species that can maintain livestock populations through dry periods; and (iii) expanding water-harvesting efforts for small-scale capture and utilisation of run-off, as pioneered successfully in such countries as India and Niger.10

Poor water use efficiency is often the result of water being considered a low value or free public good. For irrigated farming systems, changes are needed in the key areas of water and rural energy pricing policy and strengthening local management of irrigation infrastructure; both of which are important elements in increasing the technical efficiency of water use.

Increase capacity to respond to climatic changes

Changes in the frequency of extreme climatic events - whether it be temperatures, precipitation or atmospheric events - and shifts in agro-ecological conditions including those triggered by rising sea levels are likely to alter farming practices fundamentally in some vulnerable areas, such as coastal areas, semiarid zones and steep lands. Droughts, floods and hurricanes or typhoons are all expected to become much more frequent.

A better understanding of the probable nature and impact of climatic changes is urgently needed, and appropriate adjustments of agricultural policies and projects are required to mitigate adverse effects. The development of watershed protection and anti-desertification measures is likely to take on a greater urgency. It is also necessary to establish a greater capacity, both nationally and internationally, to respond effectively to damaging weather events, such as floods and droughts, to minimise their long-term impact on resource management and rural livelihoods.

1 FAO 2000a.

2 One irrigated farming system in Eastern Europe and Central Asia has relatively large farms and, for the purpose of the present discussion, is included in the category of dualistic systems.

3 Box 2.6 recounts the poor water management, input supply and yields in the Gezira scheme. It is also expected that some

of the large scale schemes in Eastern Europe and Central Asia which are not viable will be reduced in size or even closed.

4 The Coastal Plantation and Mixed Farming System in Latin America has been characterised as a dualistic system.

5 Jazairy et al 1992.

6 It should be remembered that increased farm size refers to small farmers only. More intensive systems may see larger farmers expanding as they buy out sub-marginal producers, but offer little potential to smaller producers.

7 Tanner 2001.

8 Abstracted from Case Study 1, Annex 1.

9 Abstracted from Case Study 2, Annex 1.

10 See Case Study 4, Annex 1.