Chapter 6
Economic and financial considerations

While the processes of land degradation are physical, and may be initiated or accelerated by human interference, the impact is ultimately viewed in financial and economic terms (Douglas 1992a). From the perspective of the individual household the financial costs may be felt directly. Lost production due to declining crop yields following soil degradation is a direct cost as are higher farming costs arising from increased application of fertilizer to compensate for declining soil fertility. Deforestation and deteriorating water supplies means more time and effort, especially for the female household members, has to go into collecting fuelwood and water. There is an opportunity cost to this as it decreases the amount of family labour available for on-farm productive activities.

From society's perspective the economic costs are just as real. Some such as the shortening in the economic life of a dam due to siltation, or increased costs of repairing damaged roads and bridges due to flashflooding are clear and can be quantified in cash terms. Others such as decline in the health and well being of the affected communities, are less easily costed although no less important (Douglas 1994).

Poverty and economic disadvantage

Poverty is the underlying cause of much of the land degradation within the Asia Pacific region. Lack of alternative income-generating activities (off- and non-farm) means that majority of the region's rural households remain dependent on small-scale farming and/or forestry activities for their livelihood. Within individual countries the indigenous and migrant population in the more remote upland and highland/mountain areas are generally very poor and often have a struggle to meet their basic survival needs. As a result they cannot afford to forego the chance of short-term production (e.g. growing of annual food crops on steep slopes) even when clearly non-sustainable, for the sake of long-term conservation benefits (e.g. planting tree crops which may not give any productive returns for several years).

While a range of soil and water conservation and agroforestry technologies have been developed for different agro-ecological zones, implementation requires substantial investments in labour, time, money and material resources, which many households do not have. Hence even when aware of the need to adopt specific sustainable farming practices, socio-economic constraints within their household circumstances prevent them from doing so.

Many conservation recommendations (e.g. terracing, alley cropping, reforestation) have high initial investment costs when compared to current land uses and the incremental development costs are beyond what many households can absorb.¹ There is generally a lack of spare cash within the household economy and access to low-cost credit is limited. Commercial banks, when present, are usually unwilling to lend money to those they perceive as having no collateral with which to secure a loan.

In whose interest?

The priority attached to SARM can be evaluated from the viewpoints of both the individual farm household and the wider society. The farm household is interested in the severity and frequency of an agricultural resource management problem in terms of its effect on the way the household seeks to meet its objectives and production targets. Decision-makers who are responsible to society at large (farmers and others) are concerned with the extent of a problem throughout an area, and the long-run interests of the present and future generations. Where a problem is of major concern to both the farm households and society at large, then there is mutual agreement to it being tackled as part of any development programme. Where the interests of the farm household and society diverge, two possibilities exist.

• If the problem is of interest to the farm household but not a concern of society at large then it becomes difficult to justify spending much government time and money on trying to solve it, given the need for local development initiatives to fit within a national development policy framework.
• Alternatively if the problem is in the interests of society but not a concern of the local farm households, the government has the choice of whether to alter the farming environment, including incentives, to bring farm household interests in line with those of society, or to leave conditions as they are (see box 22).

The concept of "in whose interest" is of growing concern in a number of countries within the region with regard to the issue of food security. This is because food crops increasingly give lower financial return to farmers than a variety of alternative cash crops. Thus as small-scale farming households become less subsistence and more commercially oriented there is a tendency for them to switch from food crops to cash crops, in some instances completely abandoning the growing of food crops, purchasing their food requirements with the proceeds from their sales of cash crops. From the perspective of national food security it is in society's interest that individual farm households continue to grow food crops. However, when viewed from the individual's perspective the household would be financially better off growing higher value cash crops. This divergence of interests between farmers and society has in developed countries, such as Japan and the European Union, been reconciled through the policy of using tax revenue from the non-agricultural sector to provide subsidies to farmers to continue to grow food (e.g. rice production in Japan and wheat/sugar beet in Europe). However this is not an option for most developing countries in the Asia Pacific region given the limited revenue budget resources available from non-agricultural sources.

Valuation of soil degradation

The negative consequences of soil degradation for agriculture are widely recognised, but until recently very few attempts have been made to estimate the magnitude of the costs involved (Bishop 1992). Economic losses arising from soil degradation may be divided into on-site and off-site costs.

On-site costs refer to the direct effects of soil degradation on the quality of the land resource itself. These may be expressed in terms of reduced agricultural productivity (declining crop yields, livestock losses), in terms of increased costs to sustain production levels (additional fertilizer inputs, feed purchases etc) or the costs of restoring degraded soil to its former productive condition (drainage, leaching and gypsum application to correct salinisation and waterlogging/opportunity cost of putting the land under fallow).

Off-site costs refer to the indirect effects of soil degradation, and usually take the form of externalities. Most off-site costs are traced to the effects of silt, soil nutrients or agro-chemical products washed into surface water or leached into subterranean aquifers by rainfall and irrigation run-off. The deposition of sediment over lowland paddy fields, and the downstream sedimentation of reservoirs, hydroelectric facilities or irrigation channels, as a result of poor management of upland agricultural areas, are typical negative externalities (Bishop 1992).

A range of analytical techniques can be used to evaluate the impacts of soil degradation in terms of economic costs and benefits (see box 23 and Dixon et al 1989, Stocking et al 1990, Winpenny 1991). There is also a growing body of papers and case studies that illustrate use of such techniques (e.g. Dixon et al 1990). The most detailed studies are usually of small field plots with the focus on analysing the on-site impacts, particularly the effect of soil loss on production (e.g. Sessay and Stocking 1992). Lack of physical data generally precludes an assessment of the off-site effects (Bishop 1992).

It is harder to estimate the costs and benefits of soil conservation on a project, regional or national level. Most such studies are based on an inadequate database, make wide-ranging assumptions and accept that the findings are approximate in nature (e.g. Finney and Western 1986, Boj� 1987, FAO 1994a). Also, significant errors can arise when data and techniques obtained at the level of individual field plots are extrapolated on a broader scale (Stocking 1987). Despite their limitations such studies provide a qualitative indication of the extent of the costs of soil degradation. For instance in Indonesia the annual depletion of soil fertility has been calculated as 4% of the value of crop production, or as large as the annual increase in production (Repetto et al 1989, Magrath and Arens 1989). A recent study of the effect of land degradation in South Asia concluded that (FAO 1994a):

Soil - an economic asset

Soil is one of the most essential natural resources for agricultural production. Yet agricultural land use often results in soil degradation and reduced productivity. This raises the question, Is soil a renewable or non renewable resource?

Under natural systems any soil lost by erosion is largely replenished by the ongoing processes of soil formation. Sustainable agriculture is based on the premise that soil is a renewable resource and recognises that there is a threshold level below which resource use renders it nonrenewable. Many of the land uses and management practices followed by farmers in the Asia Pacific region appear to cross this threshold level. Because they effectively treat soil as a nonrenewable resource they are unsustainable (Anderson and Thampapillai 1990).

Poor land use may be due to a lack of information or a misperception of the gravity of soil degradation on the part of the land users. In many cases though, farm households simply value the short-term profits obtained from activities which degrade the soil more highly than they value the long-term benefits of soil conservation (Bishop 1992). Such behaviour is not necessarily irrational. As one report puts it (Bishop 1995),

On the other hand, as is especially true of most Asian and Pacific countries, arable land is neither limitless nor costless to obtain; hence some form of conservation is warranted. As with any economic asset, determination of an optimal rate of exploitation depends ultimately on a comparison of the benefits of conservation to potential returns from other investments and activities. Farmers may be justified in liquidating the capital value of soil fertility if the profits derived from non-sustainable agriculture yield a higher economic rate of return in some other enterprise than in soil conservation (Bishop 1995).

Shifting cultivation is a rational economic activity where unexploited fertile land is readily available. The costs (primarily household labour) of opening up new fields being far less than those that would be involved in sustaining yields in the old fields, i.e. to combat soil degradation and weed competition. Similarly farmers with insecure tenure (e.g. short-term tenants, squatters) are following a rational economic strategy when they seek to "liquidate the capital value" of their soil fertility by using exploitative forms of crop and livestock production for short-term profit. Where land is scarce and there is security of tenure, it becomes rational for farmers to actively protect the soil as an economic asset. This fits the hypothesis that population growth (through its impact on land availability) will drive technological change due to the need to increase productive capacity per unit of land area (Boserup 1965).

A recent review of economic decision models concluded that some depletion of soil fertility can be justified on economic grounds (Bishop 1995). The efficient or `optimal' rate of depletion can be defined as the point where the costs and benefits of soil conservation are exactly balanced (in marginal, present value terms). While the costs of soil conservation are easily determined, the benefits are often ambiguous and depend on a number of factors. In general, the benefits may be expressed in terms of the value of increased future crop yields, relative to yields on degraded soils (the on-site impact), plus the value of any off-site costs avoided (e.g. sedimentation, siltation and pollution) (Bishop 1995).

Not all soils have the same asset value

In considering soil as an economic asset it should be remembered that not all soils are equal i.e. have the same productive potential (see chapter 4). The impact of erosion on yield is very much site- and soil-specific, and to a lesser extent crop-specific (Stocking 1984). High levels of topsoil erosion have little effect on yields where the soil is deep and has appreciable reserves of weatherable minerals through the profile (e.g. volcanic ash soils). In such a situation the on-site costs of erosion (loss of yield) may be relatively low although off-site costs (e.g. downstream siltation and flood damage) may be high. For such soils this begs the question as to who would benefit from any soil conservation programme and therefore who should bear the costs (i.e. the farmer or society).

Many tropical soils show very high rates of initial yield loss, decelerating as erosion progresses (Stocking 1984). This occurs where, as is common in the tropics, the bulk of the nutrients are concentrated in the top few centimetres. Hence a small amount of topsoil loss may lead to a very rapid and marked decline in yield (particularly acute for coarse textured highly leached soils). Not only is it very serious that yields plummet with only a small amount of erosion, but reduced soil productivity provides less vegetation cover enabling the erosion rate itself to accelerate. In such a situation there are tangible on-site benefits to the farmer in preventing erosion, and their quantification will help justify the costs of adopting appropriate conservation practices.

Inefficient land use

Farmers may not choose an economically optimal rate of soil degradation for a number of reasons. The prevalence of market, policy and institutional failures means that farmers do not always take into account the full costs of soil degradation to society. Such failures distort economic incentives leading farmers to deplete soil assets at an economically inefficient rate. The underlying causes of inefficient land use may be grouped into the following categories (Bishop 1995):

• The presence of non-marketed and uncompensated external impacts;
• High rates of time preference that diminish the present value of future yield losses;
• The availability of technical substitutes for natural soil fertility and alternative assets;
• Inappropriate policy incentives which inadvertently discourage soil conservation; and
• Other technical and economic constraints which prevent farmers from adopting soil conserving practices.

External impacts

External impacts or externalities are costs or benefits which are not reflected in market prices. A typical negative externality resulting from soil erosion on crop or forest land is the sedimentation of downstream reservoirs, hydroelectric facilities or irrigation channels. The protection of watersheds provided by well-managed tree plantations, orchards and other perennial crops are an example of a positive externality. Such off-site costs and benefits are not reflected in the prices of agricultural outputs, nor in farmer decision-making, but are an integral part of the economic cont-ribution made by agriculture and forestry. Because the externalities escape the arena of existing markets their effects are rarely documented and are often difficult to measure (Bishop 1995).

Time preference

Time preference refers to the simple fact that most people prefer current income to future income. Economists use the discount rate to determine time preferences when comparing present and future costs and benefits. Private individuals (e.g. farmers) are often presumed to have a high degree of time preference, and thus employ higher discount rates on average than society as a whole. For this reason society can be expected to ascribe a higher value to future crop yields foregone due to soil degradation than will farmers. Society is also likely to be more concerned about long-run stability, sustainability and equity in agriculture, all of which may depend in some measure on conservation efforts. Hence a socially optimal level of soil depletion is usually below the level tolerated by farmers (Bishop 1995).

Farmers will not all display the same time preference. Private discount rates and patterns of resource use will vary with the level of household income, food security and access to opportunities for investment. High rates of private time preference may be associated with extreme poverty, when immediate subsistence is uncertain. Land tenure problems can also engender high time preference rates, wherever insecure land use rights or shared access to scarce resources discourage investment and prudent exploitation (Magrath 1989a, Southgate 1988). A farming systems research project in the Eastern Visayas region in the Philippines found that the private discount rates of farmers were high (between 40-70%) and this was the main determinant of their willingness to adopt soil loss prevention measures which had medium to longer term payoff (O'Brien 1991).

Substitutes

In SARM technical innovation is largely devoted to devising substitutes for, or increasing the productivity, of scarce agricultural resources. Fertile land is considered an essential resource, particularly in the developing countries of the region where subsistence food production is still important for the livelihood of the bulk of the rural population. The prominent role of agriculture in national welfare in such countries justifies concern about the possible lack of substitutes for natural soil fertility, and the scarcity of alternative economic opportunities (Bishop 1995).

Policy incentives

Most Asia-Pacific countries have instituted many policies affecting agriculture, including measures which stimulate production, others which dampen output, and a number which influence the way crops are grown. Many have significant impacts on land use and soil conservation practices, because of the way they modify relative returns to certain crops, inputs or methods of cultivation. Policies may aggravate the problem of excessive soil degradation, or alleviate it.

Changes in land use patterns can arise directly and intentionally through policies affecting user rights, farm land prices or conservation incentives (e.g. land taxes or subsidies). In many cases the effect of policy on soil conservation efforts is incidental. For example, subsidies for inorganic fertilizers can artificially reduce the private costs of soil degradation as they cheapen the perceived cost of substitutes for natural fertility (Barbier 1990). Similarly price supports and export subsidies for certain crops can lead to cultivation of marginal or vulnerable lands, which would be better left under pasture or woodland. In addition to agricultural policy, other economic policies can also have profound effects on land use. Virtually any policy which distorts the market prices of agricultural inputs and outputs can alter incentives for soil conservation (Bishop 1995).

Source Magalinao and de Guzman 1995

Other factors

Soil conservation requires access to labour, capital (including land, equipment and materials, or the funds to obtain them) and information (technology). Poor farmers often lack access to one or more of these inputs, preventing them from adopting conservation measures. They may fail to perceive the gravity of soil degradation or lack information about available soil conservation measures. Even when they know of appropriate technologies farmers may lack access to sufficient labour to undertake soil conservation measures on their own. They may also suffer limited access to capital with which to hire additional manpower or purchase any tools required (Bishop 1995). For instance the best time to construct or maintain soil conservation works, or plant vegetative conservation barriers, is shortly after the start of the rainy season, when the soils are softened and vegetation cover is still light. But this is also the moment of peak labour demand for field preparation and planting. The true opportunity cost of soil conservation is thus often higher than at first appears, when considered in relation to other demands on farmers' resources.

Land values

One direct method for valuing soil degradation is to compare the sale or rental price of plots which differ only in the extent of their degradation status. In principle the difference in productive capacity is reflected in the price people are prepared to pay, indicating the present value of net returns over time (Bishop 1992). In practice this method is applicable only where land markets are well developed, and price data available. It also assumes that the degradation status will be a primary factor in determining the land's value. It may understate the full cost of soil degradation to society, as it does not take into account off-site costs.

Valuing soil productivity

Soil degradation affects agricultural productivity directly - for example when erosion washes away or buries crops - or indirectly, due to changes in soil properties. A method for valuing these on-site costs is to estimate farm revenues foregone due to soil loss or reduced top soil depth (Bishop 1992). This approach relies on estimates of the impact of erosion on crop or livestock yields, combined with farm budget data.

However, yield is a proxy indicator of soil productivity and the link between soil degradation and yields of crops or livestock is not well defined (see chapter4). Again this method takes no account of the off-site costs.

Valuing foregone revenue can be used, not only for estimating the financial impact on individual households, but also for its economic impact on the wider society. In a study of the costs of soil erosion in Java (Magrath and Arens 1989), the discounted present value of current and future net farm income foregone due to annual soil loss was evaluated at US$68 per hectare. In aggregate terms this was equivalent to about 3% of agricultural GDP.

Replacement and restoration cost

Another way of measuring the on-site cost of soil degradation is to estimate the cost of additional inputs required to compensate for reduced soil fertility. This may include increased labour inputs, or increased application of fertilizer to compensate for the loss of plant nutrients due to erosion, leaching and volatization, or removed in crop residues. Some studies have attempted to assign replacement values to lost nutrients (e.g. box 24).

There are also direct costs associated with the reclamation and rehabilitation of land already subject to physical soil degradation. This includes the use of labour inputs, machinery, equipment and/or materials to restore the land to a condition in which it can again be used for productive purposes e.g. the plugging and/or filling in of gullies, deep ripping to break up compacted subsoil horizons etc.

Likewise, there are direct costs associated with the reclamation and rehabilitation of irrigated land degraded by waterlogging and salinisation. Restoring the productivity of such soils may require investing in improved surface and subsurface drainage (canals, pipe drains) as well as soil amendments for eliminating sodicity (FAO 1990c).

The on-site impact of soil degradation may not be fully determined by the replacement/restoration cost approach. For example, nutrient losses do not reflect the effects of erosion on soil structure or depth, which are also important determinants of soil productivity (Stocking 1984). The approach also ignores off-site costs.

On-site and off-site costs

The direct costs of undertaking a SARM programme are readily identifiable e.g. staff salaries, consultantancy fees, buildings, equipment, vehicles, financial incentives (food for work, free inputs, cash payments) etc. When seeking to justify such expenditure it is worth considering both the on-site and off-site economic consequences of allowing land degradation to continue.

Economic losses arising from land degradation may be divided into on-site and off-site costs.

On-site costs result from the direct effects of degradation on the quality of the natural resources (soil, vegetation, water etc) used by farmers. These are expressed in terms of reduced agricultural and/or forestry productivity notably declining crop yields, reduced livestock carrying capacity and a decreasing supply of forest products. In addition there may be an increase in costs to sustain existing production levels (additional fertilizer inputs, feed purchases etc). In some instances farmers may have to change their cropping system in response to soil and/or water degradation by planting crops (usually of lower economic value) with less demanding nutrient and water requirements .

Estimating the off-site costs of soil degradation involves a different set of techniques to those used for on-site costs (Bishop 1992). Some costs can be quantified in economic terms and, particularly in developed countries, attempts are being made to assign values to these as the basis for developing `polluter pays' policies (Winpenny 1991). Others are less easily quantified and require qualitative social value judgements e.g. good health, bio-diversity, scenic views etc cannot in themselves be quantified in monetary terms.

The off-site costs associated with agriculture generally arise from the negative impact of runoff from crop lands, pastures or plantations, on downstream land and water users. Increased costs may be associated with changes in water quantity or quality. Typically a higher proportion of rainfall will be lost as rapid surface runoff under agriculture than natural vegetation. Downstream costs of runoff changes will be associated with increased flood damage to fields, buildings, roads etc, as well as a reduction in the seasonal quantity and reliability of streamflows used for irrigation, livestock and domestic purposes (Douglas 1994; 1996).

High levels of soil loss under agriculture leads to increased sediment load and heavy deposits of silt downstream. This will increase costs associated with keeping irrigation and navigation channels open. Siltation also reduces the life span and storage capacity of reservoirs, resulting in diminished benefits from hydro-electric power generation and gravity-fed irrigation systems. Furthermore siltation increases the turbidity of public water supply, requiring increased filtering and reducing equipment life in water treatment plants.

Agricultural practices also affect surface and groundwater quality. Residues from fertilizer and pesticide products, or from farm wastes, may be carried in surface runoff or leached into groundwater resources polluting potable water supplies. Excess nutrients can contribute to eutrophication of surface water bodies and to the growth of weeds in canals and watercourses (FAO 1990c).

The expected reduction in off-site costs has often been used as the primary justification for investing in watershed management programmes, particularly when an area is upstream of an irrigation or hydroelectric dam. Where the emphasis is on reducing off-site costs, much of a watershed management programme's investment goes into installing physical structures (check dams, silt traps, gully plugs etc) designed to reduce sediment flows. This ignores the fact that the off-site costs resulting from siltation may be a much lower order of value than the on-site costs related to productivity losses (see box 25).

It is often taken as axiomatic that soil conservation projects reduce downstream sedimentation and the supply of agri-chemical and other pollutants associated with the transport of the finest sediment fractions; but in large catchments reduced sediment loads may not be realised for decades or even centuries (see Mahmood 1987, Doolette and Magrath 1990, Dickinson 1995).

In practice the primary economic justification for undertaking a catchment management programme will usually come from the benefits to be gained by reducing on-site costs. Any potential to reduce off-site costs should be seen as a bonus, i.e. a secondary justification, and perhaps one for deciding the development priority between two competing project areas.

On-site and off-site benefits

SARM programmes can be expected to provide a range of both on-site and off-site benefits through the promotion of improved land use practices at the field, farm and community levels. On-site benefits, notably from the maintenance and enhancement of soil productivity, come from sustaining and increasing the productive output of various agricultural and forestry enterprises (crop, livestock and tree production). Off-site benefits derive primarily from the preservation, or improvement, of the value of downstream land and water resources and the avoidance of rehabilitation costs.

The following are the type of on-site benefits that could be expected (after Winpenny 1991):

• Avoided incurring the costs that would result from unchecked land degradation;
1. Prevention of crop production losses from decrease in soil productivity (due to soil erosion, loss of soil depth and declining fertility);
2. Prevention of land going out of production through gully erosion;
3. Alternatively saved the cost of fertiliser that would have to be purchased to maintain yields on eroded soils;
• Value of enhanced food and cash crop production through increased crop yields resulting from the `ecological benefits' of improved land husbandry (increased soil organic matter, improved topsoil structure, better soil moisture retention, micro-climate amelioration etc);
• Value of enhanced livestock products (meat, milk, wool, dung) from restored or improved pasture, better use of crop residues (for feed and bedding), or from growing fodder species (grasses, shrubs and trees) in contour hedgerows;
• Value of wood and non-wood products to be obtained from increased tree planting and sustained management of natural forest areas (timber, poles, fuelwood, forage, fruit etc).

The following are the type of off-site benefits that could be expected (after Winpenny 1991):

• Irrigation benefits:
1. value of crops preserved through reduction of sedimentation in reservoirs and channels;
2. reduced cost of maintaining and cleaning reservoirs, channels, and works;
3. output saved from preserving existing water regime;
• Hydroelectric power benefits:
1. by avoiding reservoir siltation, extending the life of a hydropower scheme, especially its ability to generate dry-season power;
2. avoiding cost of raising level of dam, with all that implies for extra inundation;
3. avoiding cost of alternative generating capacity;
4. plus savings in repairs and cleaning of turbines and intake works;
• Flood damage avoided; alternatively, savings in cost of flood prevention works, or reduced cost of roads and bridge maintenance;
• Gains to fisheries;
1. less silting and turbidity in reservoirs and rivers and more even year round flows;
2. avoided damage to productivity of coastal waters and mangrove systems.
• Navigation benefits:
1. from more predictable river channels;
2. reduced dredging costs, both in river system and inshore coastal waters;
• Benefits to domestic water quantity and quality; avoidance of need to resite intake works through river silting and changes in channels;
• Tourism and recreation benefits preserved and enhanced.

Once the benefits are identified attempts can be made to quantify them by assigning them economic values. This generally will be easier for the on-site benefits where monetary values can be assigned for commodities with a market price. This may be less easy for many of the off-site benefits where valuation in economic terms may have to be based on the principle of society's `willingness to pay' for the use, or protection, of a particular resource (Stocking et al 1990, Winpenny 1991).

Benefits of "soil harvesting"

Cross slope soil conservation barriers (both physical structures and vegetative measures) can capture and accumulate silt suspended in runoff from upstream sources. Such eroded sediments contain higher concentrations of soil nutrients than the soils from which they come, and the sites in which they accumulate likewise have better soil moisture conditions. The additional increase in yield from the trapped silt is distinct from the benefits associated with retaining the top soil in situ. Such soil harvesting may be an incidental benefit of soil conservation measures or in some cases the main aim of farmers' efforts. In parts of India farmers are known to build silt harvesting structures across gullies in order to create plots suitable for growing crops in otherwise marginal areas.

Similarly in hilly areas they may induce erosion in the upper portions of their land in order to concentrate the soil in the lower part (Kerr and Sanghi 1992). In both cases there are direct financial benefits to farmers from allowing erosion to take place in one area so as to enrich another.

There is also a positive side to the high sediment levels in many Asian rivers. A significant proportion of the sand used for building purposes within the region is eroded material from river beds. The methods used for sand extraction varies from individuals with a shovel and basket, to larger operations involving suction pumps and floating dredgers. Whatever method is used sand extraction provides valuable non-farm employment for many rural people. Should it be physically possible to stop all soil erosion then this source of easily obtainable building material could become depleted.

Costs of SARM

In addition to the costs allocated within a SARM project other costs associated with implementing particular conservation recommendations should be provided for. Of particular concern is the short-term opportunity land costs taken out of crop or livestock production. Similarly the initial reduction in yield following construction of some forms of terraces (i.e. when crops are planted in subsoil exposed during excavation). Few small-scale farmers are in a position to forego short-term production for the sake of possible benefits in the future (Douglas 1994).

Many conservation farm plans call for some 5-10% of a holding to be set aside for graded terraces, waterways etc. Farm households with only small holdings, struggling to produce sufficient food for themselves, cannot afford to take part of their land out of food crop production to accommodate physical conservation structures; the lost production could mean the difference in the next cropping season between meeting the family's food requirements or going hungry.

This also applies in the case of biological conservation measures such as hedgerows and grass strips. Whereas these may ultimately contribute to total farm production (by providing fodder, green manure, fuelwood etc) during their initial establishment there is the foregone cost of lost production for the one or more years it takes for the productive benefits from the hedgerow or grass strip to be realised.

Who incurs the cost, who gets the benefit?

Many watershed management programmes involve the closing of areas to livestock or other uses to allow degraded grazing lands and woodlands to regenerate. The declaration of protected watershed areas may also involve the banning of any form of annual crop production (see box 26) in upper watershed areas. Tree planting programmes may also have unforeseen costs - for instance by encroaching into grazing lands or resulting in the loss of traditional non forest products when mono-plantations replace what, although degraded, was a bio-diverse natural area. Such programmes can prove unpopular with the affected communities. Whereas some within the community may benefit, for instance those with land which can be irrigated from the `protected' water sources, many could prove to be worse off.

The effect of the closing of traditional grazing lands may be to force livestock owners to increase grazing pressure on the remaining rangelands, and with a reduction in fodder availability to decrease overall livestock productivity. Switching to zero grazing, using cut and carry feeds grown in farmers fields, is only an option for those with land, and may significantly increase the workload of women who are are traditionally responsible in many Asian countries for fodder collection as opposed to herding free grazing animals. Following the implementation of one watershed management programme in India it was found that the women's workload in terms of collecting and transporting fodder had increased by 127% (Arya and Samra 1995).

Consideration of short and long-term costs and benefits

Economic and financial appraisal normally involves assessing proposed improvements in monetary terms and in comparison with, for instance, the present performance of a farm household system. This is effectively a before-and-after approach and will yield valuable information related to the short-term productive effects of a recommendation i.e. will this recommendation produce higher crop yields, milk and meat production etc. compared to the present situation, and if there is an increase in the production costs is this offset by a much larger increase in the benefits (i.e. a high marginal rate of return).

However the long-term sustainability issue which requires consideration of the with-and-without scenario should also be considered. This requires predicting future performance depending on whether or not the recommendation(s) are adopted. For instance if present farm management practices lead to soil erosion and fertility decline in the croplands and overgrazing in the rangelands, then in five years' time the present performance of a farm household system can be expected to have deteriorated. Crop yields may have declined (or require higher levels of inputs to sustain production) and there may be a similar decline in the quantity and quality of the outputs from the livestock enterprises. However, had the farm household adopted SARM practices, then over the same five-year period they may have been able to sustain, or preferably increase, the overall performance of their system. The with-and-without analysis allows a comparison between the two options to reveal the longer term costs and benefits of staying with the present system or making improvements to it.

The before-and-after analysis, with its emphasis on the short-term costs and benefits, is likely to produce the type of information which influences decision-makers at the farm household level. On the other hand, the with-and-without analysis may provide information that can be used to justify to decision-makers, at the community and society levels, the need for particular interventions because they are in the long-term interests of the wider society.

In most cases the before-and-after and with-and-without analysis should provide a similar assessment of the suitability of a particular improvement. However in some cases this may reveal that, in the short-term the costs exceed the benefits, but in the long-term the situation is the reverse. In such a situation two options are open to planners - one is to recommend a policy of short-term subsidies or incentives to improve the cost benefit ratio, the other is to look for alternative improvements where long-term benefits can be ensured with practices that offer short-term benefits. The latter is usually the more cost effective and sustainable option (Douglas 1992).

Cost benefit analysis

Cost benefit analysis (CBA) is seen by a number of authors as a useful tool for the appraisal and evaluation of soil conservation projects (Boj� 1986a, 1986b, 1987, Dixon et al 1989, Stocking et al 1990, Winpenny 1991). CBA has been described as:

The costs and benefits of implementing a conservation activity will occur at different points in time. For instance there are costs associated with establishing a contour hedgerow (including the opportunity cost of the land in which it is planted), but the benefits in the form of fodder, green manure, fuelwood, higher crop yields etc will not be realised until at least 2-3 years later. Also the costs and benefits of a conservation project may be distributed unequally over different individuals at any single point in time. For instance restricting the growing of annual crops by hill farmers in the catchment area of a lowland rice scheme, as it will be several years before they get any benefits from growing perennial crops instead.

CBA provides a technique for evaluating costs and benefits both over time and between beneficiaries/interest groups. It allows choices between different development options to allow the selection of those that achieve the desired aims at lowest cost or that achieve the highest level of goal achievement with the resources available (Stocking et al 1990). It allows policy makers to address the different time horizons of individual farm households and society with regard to acceptable returns to investment in land improvements (i.e. need for short-term returns without compromising future returns). There is still a need for further experience with its use for evaluating different SARM options. Particularly with regard to how to value the long-term benefits of conserving soil productivity. This is needed to justify the costs of project intervention when some conservation benefits may only be realised long after the normal life span (3-5 years) of a project.

Targeting of economic resources

Recognising differences in the economic asset value of soils with different bio-physical potentials brings to the fore a development dilemma: should scarce funds be directed to high potential or low potential areas?

From an economic perspective returns will be greater in high potential areas where the opportunity exists for sustaining high levels of production, with less risk and lower cost. From a social welfare perspective the needs may be greater in low potential (marginal) areas, where poverty levels are usually higher and the opportunities for alternative livelihood systems lower. Such a situation requires a cost benefit analysis that addresses both the economic and social dimensions.

In discussing dryland degradation, Dixon et al (1989) invokes the analogy of the wartime medical concept of triage, as a way of deciding where to concentrate scarce resources. Under the triage system war casualties were divided into one of three groups on the basis of a very quick evaluation. Those patients in the worst condition with little chance of recovery, even with intense treatment, were made as comfortable as possible without expending scarce resources that would probably be ineffective anyway. A second group included those in reasonably good condition and needing little or no immediate attention. The third group include those who, with immediate help, could recover and survive; this group received the most attention. In this manner, scarce medical resources were allocated to maximize overall benefits.

Given limits to the financial and manpower resources available from government and donor sources, this technique could be used to target those areas where a SARM project could have the most cost-effective impact. Its application means recognising that land in good condition may need little or no additional attention, i.e. no current need for external project intervention. Whereas land already so severely degraded, as to have little remaining productive potential, should be either left untreated, even if further degradation will continue, or simply fenced off and allowed to recover by natural means as best it can. Instead the bulk of the resources should go to areas where timely intervention can arrest existing degradation before it has become severe, or prevent it from occurring in the first place where land use practices are clearly non-sustainable. There may be objections to policies allowing degradation to continue in particular areas. However undesirable such policies might appear, the alternatives are often no better. Saving one area may take scarce resources away from another with higher potential economic returns.

Financial considerations at the farm household level

Financial considerations will figure highly in the decision-making process within small-scale farming households. Specifically a farm household will seek to maximize its present and future well-being by allocating its resources to those productive activities which can be expected to give the highest returns with the minimum of risk. These resources will include the household's access to farm, range and forest lands, accumulated capital goods, financial resources (cash or credit), family labour and the skills and knowledge of its individual members. The household's decisions on which particular activities to pursue at any one time are also influenced by its perceptions of the local constraints and potential opportunities.

A household's future well-being depends on its ability to maintain and expand its options, particularly through investment in land (and farm) improvements and education (learning about new techniques, sending the children to school to gain qualifications for skilled off-farm employment). Which options are followed will depend on culture and tradition, access to productive resources, past experience with, and investment in, productive enterprises, and the way the household interacts with the economy at large. These include options for wage employment and wage rates, markets for agricultural goods and prices, credit opportunities and interest rates (Hunter 1993).

Except in very remote and isolated areas very few small-scale farm households in the Asia Pacific region are these days purely subsistence producers. Most have the option to engage in some form of market production and wage employment. Likewise several of the items that contribute to the household's well-being are purchased. Some farm production is dictated by culture and tradition (e.g. production of certain preferred staple foods). However there will be competition for the remainder of the households time and resources from a number of alternative on-farm as well as off- and non-farm activities. The final selection depends on the household's perceptions as to the relative costs and benefits of each.

Role of agriculture within the household economy

A key question that has to be asked in any SARM programme is, how dependent is the household on its farming activities for meeting its priority goals? It is common to find that the off-farm activities by some or all of the household members, contribute significantly to the overall household economy. For a growing number of rural households farming may already be of secondary importance.

There is an opportunity cost associated with a household's factors of production, particularly in the case of labour, time, capital and academic and vocational skills. A household will take this into consideration in deciding whether more benefits could be gained by using these for increasing production on-farm or for obtaining income from off-farm activities. Any SARM recommendation will be evaluated not just in terms of the effect on farming activities but whether it will conflict with any off-farm activities. Suggested improvements must not impose unacceptable costs in terms of foregone benefits or should offer significantly higher returns than could be obtained from other, particularly off-farm, activities.

In many countries the potential returns to small scale farming are low. As a result those household members with the best prospects for wage employment, usually the young to middle-aged men, will attempt to enter the labour market. Since the available jobs are often some distance from the farm, they become migrant workers and no longer participate directly in on-farm subsistence or market production. They remit cash to meet household welfare needs, which may also help pay for farm inputs. Older men return to the farm to enjoy the fruits of their earlier remittances. Women tend to remain on the farm because of child-bearing and their lower comparative advantage in the labour market. In such circumstances the function of the farm is not income generation, or food production, but provision of a rent-free place to live with free (or at worst low cost) fuel and food and maybe communal pastures on which to keep cattle, sheep and goats in which remittances have been invested. Under these circumstances households have little incentive to adopt either production or conservation oriented improvements that would require more resources (especially labour) to be devoted to farming. There is a high opportunity cost to a household using its labour and other resources for conservation (Stocking and Abel 1992).

In some situations it has been observed that where a household is able to meet its basic welfare needs from off-farm sources, there may be a deliberate change in land use away from the growing of annual food crops to perennial cash crops, such as fruit trees. This occurs in areas with a steadily ageing farming population, due to the younger generation moving off the land as is happening in countries such as Malaysia and the Philippines (personal observation). Annual crops, such as paddy rice, have high seasonal labour demands, which may be beyond the resources of the old and very young within the household to sustain. However, once perennial tree crops have been planted the annual labour demands are usually much lower. In such a situation it becomes a rational strategy to change the households farming system, in response to a shortage of able bodied labour. Although this entails foregoing short-term returns (loss of food production) it is a long-term production investment on the assumption of future, lower cost, benefits.

If the enterprises and farm management practices within a household's existing farming system enable it to presently meet its goals there may be little incentive to change even if, to the soil conservation specialist, the system is not sustainable. In such a situation the motivation to change and adopt SARM practices would be if the recommendations offered tangible benefits, such as lower capital and labour input costs, enabling previous targets to be maintained or reduced without impairing a household's ability to meet its goals. A participant (Sabati Solomona) in a FAO farming systems workshop in the Pacific noted (quoted in FAO/IRETA 1995):

Market opportunities

The decision by an individual household on what to produce, when and how much, if any, to sell, will depend on the market opportunities, and recent experience of seasonal and annual price variations. It will also be based on the differential between the producer and consumer prices, government price guarantees and risks associated with relying on cash income for food purchases versus subsistence production. Similarly decisions on which, if any, inputs to use will be influenced by their cost and local availability.

Experience shows that smallholder farmers, even when primarily engaged in subsistence production, are very responsive to market conditions, especially price. Commodity prices will determine many of their production strategies, leading at times to environmentally poor land use practices, such as continuous mono-cropping of high value crops, where farmers find the recommended crop rotations financially unattractive.

A major problem for farm households is that they lack the ability to predict future market demand for specific commodities. Thus they face a risk when expanding production of an existing crop, or planting a new one, with the primary intention of selling the increased production.

In Asia most countries have significant local markets to stimulate and consume a significant amount of any increased cash crop production. With rising incomes particularly in the urban business sector, this has stimulated increased demand for a range of agricultural commodities, such as fruit, vegetables and meat, as people aspire to eat more than the basic traditional staples. There is thus a growing market for horticultural and livestock products within Asia. However the market is not unlimited and there are many examples of farmers' responding to a perceived market demand and expanding their cash crop production only to fail to get the increased cash returns. Often, as the increased production reaches the market, prices fall as shortage has changed to surplus.

In the late 80s and early 90s citrus production was heavily promoted in parts of the red soils region of China. Large areas of citrus orchards were established with the aid of a World Bank project. Once all the new trees started bearing fruits the supply significantly exceeded demand and prices slumped.

A similar situation is starting to appear in Northern Vietnam with the overproduction of litchi. One of the major problems now facing the FAO/UNDP FARM project site in Ha Bac Province is how to expand the market for the surplus production from the recently established litchi orchards on the plots allocated to individual households. Suggestions have ranged from introducing new cultivars (which fruit either earlier or later than the present ones so as to lengthen the fresh fruit season) to finding ways to extend the `shelf life' of the harvested product (cold stores, drying, processing into juice).

In the Pacific the internal market for agricultural products is small due to the small size of the populations in each of the island countries. As a result most cash crops have been produced for the export market. This has made the economies of these countries and the market opportunities for individual farm households very dependent on global demand and supply. Thus the history of cash crop production in the Pacific is characterised by a series of boom and bust cycles. Pacific island cash crops like coconut, cocoa, ginger, vanilla have all gone through such cycles. Tonga has identified a seasonal market niche in Japan for squash, and farmers have responded to this in a big way to the extent that currently there is overproduction and some 50% of the crop has to be dumped either because it is not of the right quality or because it exceeds the quantity required by the exporters (Umar personal communication).

A key point that should be considered when promoting increased commercial as opposed to subsistence production is that with annual crops farmers have more flexibility to change their production strategy from year to year in response to fluctuating market opportunities. Commercial farming systems based on perennial crops, particularly tree crops, have far less flexibility and may represent a more risky investment.

Labour costs and returns

Labour is an inextricable part of SARM. All forms of conservation require it, mechanical and physical works the most, but biological conservation with its intensification of land use also makes high labour demands. Yet the labour requirements of conservation activities are all too often ignored or undervalued in the design of conservation schemes (Stocking and Abel 1992). There is generally no abundance of labour within small-scale farm households nor is their labour necessarily `free' (see box 28). Whereas there may appear to be underemployment in rural areas, in practice during the planting and harvesting seasons labour is scarce and carries an appreciable opportunity cost (Southgate 1988).

Soil conservation recommendations, especially those with rather long payoff times and requiring heavy labour inputs, commonly fail to be adopted, not because farmers fail to recognise the need, but because the benefits compare unfavourably with the opportunity cost of the labour that would have to be redirected to these activities. Labour intensive conservation activities are generally inappropriate for small-scale farm households where labour is commonly in short supply. Instead the need is for developing labour saving, low maintenance conservation technologies or developing farming practices that have worthwhile production benefits while at the same time being conservation effective. Alternatively innovative forms of labour mobilisation and organisation may have to be developed at the community level (Hunter 1993).

In conventional financial analysis the costs and benefits of farm activities are expressed in monetary terms. In primarily subsistence farming cash may play only a small role in the decision-making process within the household economy. What is often of greater significance is returns to labour which may be a farm household's most limiting production factor. At first sight an activity that appears irrational in purely financial and economic terms may prove rational when analysed in terms of returns per unit of labour expended.

In areas where population densities are low it is common to find traditional farming systems whose component enterprises and management practices offer high returns per unit of labour. As population densities increase and land becomes scarcer this can lead to the adoption of farming systems which produce more but give a lower return per unit of labour. The returns per unit of labour from traditional shifting cultivation farming systems are generally higher than from more settled and intensive systems even though productivity per unit area may be lower.

Similarly extensive grazing is likely to give higher returns per unit of labour than stall feeding on a cut and carry basis. It is not uncommon to find that farmers are aware of quite sophisticated conservation techniques (such as mulching, ridging or terracing) or more intensive cultivation methods, but not using them because of their higher labour demand and the continuing adequacy of local land resources for the fallow periods of their existing systems (IFAD 1986).

When alternative economic opportunities, offering higher returns to labour, become available, more intensive farming systems may cease to be fully viable because they offer only low level equilibrium i.e. they give lower returns per day worked. Indicators of this may be a failure to undertake a second weeding or to continue maintenance of conservation measures (e.g. terraces), activities which were previously an integral part of the farming cycle. In some countries of Southeast Asia the result of changes in the opportunity cost of a household's labour is a growing phenomenon of `idle land', whereby previously intensively cultivated fields have been abandoned as off-farm employment opportunities have increased.

Acceptable rates of return

A key issue that needs to be considered in SARM is whether or not the returns to the factors of production (land, labour, capital, management skills etc), from adopting an improved land use enterprise or farm management practice, are attractive to a farm household, i.e. significantly better than their existing on-farm (and maybe off-farm) activities.

The key factor will be the minimum acceptable rate of return i.e. the rate at which farm households will consider it worthwhile to change and adopt a recommended improvement. This is the marginal rate of return at which the expected increase in benefits will more than justify any increase in the household's costs (e.g. higher inputs of labour, purchased seed and chemicals) and foregone benefits (e.g. land taken out of production). It is generally recognised that for the majority of situations the minimum rate of return acceptable to farmers will be between 50 and 100% (CIMMYT 1988).

If a technology is new to farmers (e.g. chemical weed control where farmers currently weed by hand) and requires that they learn some new skills, a 100% minimum rate of return is likely to be required. If it is an adaptive improvement representing a simple adjustment in a current farming practice (such as a different fertilizer rate for farmers who already use fertilizer), then a minimum rate of return as low as 50% may be acceptable. Unless capital is easily available and learning costs are low, it is unlikely that a rate of return below 50% will be accepted (CIMMYT 1988).

Insufficient attention appears to have been given to determining the minimum acceptable rates of return that farmers require before they will adopt improved land uses and management practices. Many conservation recommendations fail to be adopted because the rates are too low. An economic analysis of the use of Vetiver grass and contour bunds in soil conservation compared yield benefits to farmer costs per hectare treated (Magrath 1989). Estimated rates of return in net present value terms for the two technologies varied from 22% to 95%, depending on the assumed level of yield increase and the proportion of soil loss prevented. Care would need to be taken before promoting either technology to be sure that the actual yield levels that farmers could be expected to achieve would give rates of returns of at least 50%. However where the returns are perceived as worthwhile adoption may be rapid.

Consideration of risk and uncertainty

The decision-making process in small-scale farming households tends to be based on risk avoidance. Often this is because they cannot rely on the market for their supply of food, consumer goods, tools, equipment etc. and/or their farming systems do not generate sufficient cash to purchase them. The less commercialized the farming system is, the more important the on-farm production of food and other essential goods becomes to satisfy basic needs. For resource-poor farm households avoiding risk in the operation of their farming systems is a necessity.

Risk has three important implications for SARM (FAO 1990d).

• First, changes to existing farming systems that are proposed as improvements should be compatible with the farm households strategies for reducing risk. For instance before proposing that farmers should go for uniform early planting, the rationale behind their practice of staggered planting dates should be taken into account. Recommendations that do not take account of farmers' attempts to reduce risk have little chance of being adopted.
• Second, a study of the risks faced by farmers may suggest opportunities for developing recommendations to help stabilize farm production. Drought risk may be reduced with moisture conservation techniques (which will also conserve soil), and fruit tree losses due to high winds may be reduced by introducing varieties whose flowering and fruiting seasons do not coincide with the typhoon or cyclone season. A recommendation that does not necessarily increase average benefits may still be acceptable, if instead it introduces a greater degree of stability to the farming system, by reducing the year-to-year variability.
• Third, farm households make planning decisions on the basis of seeking to ensure their survival in the worst years, not just average ones. Therefore the amount that they are prepared to give up (in terms of average net benefits) to reduce the effects of an uncertain environment is a measure of their risk aversion. The degree of a household's risk aversion may depend on several factors but as a general rule, small-scale farming households are at least moderately averse to taking major risks. This is one reason why they prefer making adaptive improvements rather than radical changes to their existing farming systems.

Commodity and input prices

The decision by an individual household on what to produce, when and how much, if any to sell, will depend on the market opportunities, and recent experience of seasonal and annual price variations. It will also be based on the differential between the producer and consumer prices, government price guarantees and the risks associated with relying on cash income for food purchases versus subsistence production. Similarly decisions on which, if any, inputs to use will be influenced by their cost and local availability.

Experience shows that smallholder farmers, even when primarily engaged in subsistence production, are very responsive to market conditions, especially price. Commodity prices determine many of their production strategies, leading at times to environmentally poor land use practices, such as continuous mono-cropping of high value crops, where farmers find the recommended crop rotations financially unattractive.

Evidence from Indonesia suggests that under certain conditions changes in relative producer prices can affect choice of crops and farming system, thus impacting on degradation (Barbier 1988, 1989, 1990; Carson 1987). The relationships governing farmer responses to relative crop prices are complex, and depend on various factors such as household wealth and income, tenure security, attitudes to risk, access to off-farm employment, labour and capital constraints and intra-household allocation of land (Barbier 1991). However, farmers will respond to higher relative prices for erosive crops by seeking short-run economic returns from erosive crop cultivation at the expense of long-term land degradation. This result holds mainly for sedentary farmers cultivating rainfed plots in areas where agricultural extensification is reaching, or has already reached its limits.

In a few areas within the Asia Pacific region where there is the opportunity to expand the area under cultivation, farmers can be expected to open new fields when the returns from the new land exceed those from continuing to cultivate the existing lands (Southgate 1990). Higher relative prices and returns to erosive crops will not only accelerate degradation on existing land but, as a consequence, will also induce a more rapid expansion into new areas and increased land degradation (Barbier 1991).

At the same time, fluctuations in relative prices can increase the uncertainty and risk borne by, in particular, small-scale producers. Switching to and investing in new cropping systems and methods of cultivating involve high upfront costs and long pay back periods for small-scale farmers. Unless they can be assured that the relative prices and returns from the non-erosive systems will be sustained, these farmers may be less willing to invest in new, less-erosive cropping patterns and systems or in improvements to these systems where they already exist. Similarly small-scale producers may be less willing to invest some of the short-run profits from erosive cropping into expensive physical erosion control measures, such as bunds, contour ridging, bench terracing and so on, unless they can be sure that the high relative prices they receive for their crops today will also prevail in the future. Thus the price risk imposed by fluctuating relative prices may deter farmers from investing in SARM (Barbier 1991).

In Fujian Province in China, farmers receive high prices for longan fruits. As a result there has been a massive expansion of the area under longan orchards. Much of this new planting has taken place on degraded wasteland sites. Maintaining production on such sites requires massive injections of organic manure. An FAO project formulation mission (FAO/ADB 1993) determined that the greatest risk to the long-term sustainability of these orchards was an inadequate supply of organic manure within the farm household system.

Should farmers be unable, or unwilling, to continue applying the required quantities of organic manure the consequence would be, because of the inherent low fertility of the upland red soils, a decline in plant vigour and severe reduction of yield (and even death of the plant). Should yields fall to uneconomic levels, which they may well do on the most marginal former waste land sites, farmers may abandon the enterprise. In this case lack of terrace maintenance, combined with decreased ground cover, would lead to accelerating soil erosion thus very quickly recreating the type of waste lands the project was seeking to rehabilitate. Should over production of longan lead to a fall in the market price then the process of abandonment of orchards in the less favourable sites could accelerate.

Influenced by urban clientele, both affluent and poor, the governments of many Asian and Pacific countries have used price controls and other policy instruments to keep producer prices for food low. Receiving low prices for crops and livestock, small scale farmers are discouraged from investing in agricultural improvements. In particular they are less willing both to apply conservation measures to existing farmland and to clear new land for agricultural production (Southgate 1988).

Input levels

The inherent productive potential of land stems from the bio-physical conditions present, but actual production levels depend to a great extent on the level of management and inputs used. With good management and high levels of inputs, average crop yields may be higher in areas with lower bio-physical potential, than if the same crop is grown under a low input system, in a higher potential area. The level of input use is therefore a critical factor in any discussion on SARM.

In the context of SARM input levels are defined as follows (after FAO 1982, 1984):

• Low input systems: crop and livestock production primarily to meet the households subsistence needs. No use of purchased external inputs (i.e. no chemical fertilizers, pesticides or improved seeds). Cultivation done by hand or with traditional draught implements. Farmers may follow no specific conservation practices or use indigenous techniques.
• Intermediate input systems: significant percentage of the crop and livestock production goes to meet the households subsistence needs with the remainder being sold, may be some production of non-food cash crops. Limited use of purchased external inputs (i.e. some use of chemical fertilizers, pesticides and/or improved seeds but at suboptimum rates). Cultivation done by hand or with draught animals using improved tools and implements. Farmers may adopt some improved conservation practices or use sophisticated indigenous techniques.
• High input systems: bulk of the crop and livestock production is sold, only small percentage consumed on farm, may be significant production of non-food cash crops. Extensive use of purchased external inputs (i.e. chemical fertilizers, pesticides and improved seeds used at the recommended optimum rates). Cultivation done with the aid of draught animals and/or machinery using improved tools and implements. Farmers following the recommended land management and conservation practices.

Low to intermediate input systems are still the norm for much of the region. However, in the more developed economies of Asia, more farmers are operating at the intermediate level with some moving towards high input systems. A low input system is not necessarily unsustainable. Many traditional subsistence farming systems utilise indigenous techniques to control erosion and maintain soil fertility. In situations where land is not a scarce resource, these may still be adequate to sustain production, although at a low level per unit area, while providing worthwhile financial returns in relation to the investment of land and labour.

On the other hand high input systems are not necessarily sustainable and may cause their own environmental problems. The use of high levels of fertilizer and pesticides can lead to chemical pollution of nearby water resources. Poorly designed conservation structures and incorrect ploughing can lead to accelerated runoff and erosion. Moreover mechanisation and many of the purchased inputs are dependent on the use of fossil fuels and non-renewable mineral resources (e.g. phosphates).

Impacts of SARM

A recent review on the policies and practice for sustainability and self reliance in agriculture found that activities to promote SARM can bring economic, environmental and social benefits (Pretty 1995). It recognised three distinct types of agriculture, all of which can be found in the Asia Pacific region.

First, there are the industrialised agricultural systems of the developed countries, notably Japan, Australia and New Zealand. Second, the high input systems practised in the high potential green revolution areas favoured by good soils, reliable water and supporting services and infrastructure. Last, are the remaining agricultural and livelihood systems which can be characterised as low input, complex and diverse, with considerably lower yields.

The basic challenge for SARM for each system is quite different. In industrialised agriculture it is to reduce substantially input use and variable costs, so as to maintain profitability. Some fall in yield is likely to be acceptable given present production levels. In the green revolution areas, the challenge is to maintain yields at current levels while reducing environmental damage. In the diverse and complex systems the challenge is to increase yields per hectare while not damaging natural resources (Pretty 1995).

The evidence from farms and communities in many countries shows that SARM can be achieved in all three systems (after Pretty 1995):

• In the diverse, complex and "resource-poor" farming systems of the Asia Pacific region, farmers adopting SARM technologies have doubled or trebled crop yields, often with little or no use of external inputs (table 5);
• Iin the high input and generally irrigated green revolution areas, farmers adopting SARM technologies have maintained yields while substantially reducing inputs (table 6); and
• Iin the industrialised agricultural systems, a transition to sustainable agriculture could mean a fall in per hectare yields of 10-20% in the short-term, but with better levels of financial returns to farmers.

But farmers do not get more output from less input. They have to substitute knowledge, labour and management skills to make up for the foregone added values of external inputs.

Table 4. Impact of SARM technologies and practices in complex
and diverse agricultural systems

Source Table 7.5 in Pretty 1995

Table 5. Impact of sustainable agriculture in Green Revolution lands

Source Table 7.4 in Pretty 1995

¹ Alley cropping was one of the conservation farming practices tested by the IBSRAM ASIALAND sloping lands network. Reported results show that the initial establishment of the hedgerows incurred an additional cost of US$104 (in the Philippines) and US$58 (in Thailand) per hectare. This cost included: i) the labour needed for laying out the contour and planting the hedgerows; and ii) the cost of hedgerow materials (Magalinao and de Guzman 1995). Such costs would be beyond the means of most small-scale hillside farming households in both the Philippines and Thailand.