Technical Paper N. 11


Guidelines for Planning Irrigation
and Drainage Investment Projects


Table of Contents


Objective of the Guidelines



1. Irrigation in the Context of Water Resources Management

2. Irrigation Types and the Issue of Scale

3. Irrigation, Food Supply and Drought

4. Effective Implementation

5. Fiscal Sustainability

6. Water User's Associations

7. Social and Environmental Aspects

8. Choice of Technology

9. The Drainage Dilemma

10. Implications for the Planning Process

Further reading




Publicly financed irrigation and drainage investment projects have too often performed poorly. Sometimes the reasons have been unforeseeable or unavoidable. But in some cases shortcomings were because planners gave inadequate consideration to institutional constraints or to the practical problems of subsequent implementation, or because there was insufficient commitment by governments or users to the developments proposed. Delay, dilapidation, waste of scarce water and adverse social and environmental impacts have been among the familiar consequences. Lessons have been learned from these setbacks, however. These Guidelines give prominence to the planning approaches which have evolved, and are still evolving, to avoid future difficulties. The publication stresses sounder formulation of irrigation and drainage investment strategies, improved conceptualization of project options, and building stronger participation and commitment into the detailed planning process. The intended users are FAO Investment Centre staff, trainees and consultants, as well as local planning groups set up by governments to prepare investment proposals. Some of the material may also be useful to consulting firms and financing institutions involved in planning or appraising such projects. The following Investment Centre Technical Papers complement and should be read in conjunction with these Guidelines: Guidelines for the Design of Agricultural Investment Projects (No. 7, revised 1995); Financial Analysis in Agricultural Project Preparation (No. 8, 1991); and Guidelines on Sociological Analysis in Agricultural Project Design (No. 9, 1992).

Objective of the Guidelines

The objective of these Guidelines is to summarize present thinking and practice, and to assist practitioners to plan irrigation investment projects and programmes that will realize and sustain their full potential. They cover the whole investment planning process, from formulation of subsectoral strategies, to conceptualization of project options, and detailed planning of the preferred option(s).


Why New Guidelines?

Irrigated agriculture has made a major contribution to food production and food security throughout the world: without irrigation much of the impressive growth in agricultural productivity over the last 50 years could not have been achieved. Nevertheless it is widely accepted that the overall performance of irrigation and drainage (also implying reclamation and water control) investments has too often fallen short of the expectations of planners, governments and financing institutions alike (Ref: Report No. 13676, A Review of World Bank Experience in Irrigation, Operations Evaluation Department, World Bank, Washington DC, 1994).

 In 1992, the World Bank's Portfolio Management Task Force reported that major problems that constrain the performance of World Bank financed investment projects in various sectors, are: 

The World Bank Task Force stressed that the ultimate success of a project is to a significant degree determined by what happens in the "upstream" planning process, and that many implementation problems can clearly be traced to deficiencies here. These findings are now considered generally valid by other international financing institutions and donors. The lessons learned regarding these problems, and the new approaches that have been developed to tackle them, therefore make new guidelines essential and timely. 


Since the ultimate success of an investment is largely determined by the quality of the upstream process of planning, it is pertinent to examine recent lessons learned from experience and their implications for future planning. The main lessons are that water is an increasingly scarce resource for which there are many competing demands that are more profitable, socially and economically, than irrigation; also that low world prices for basic food and fibre crops, together with typically high development costs, have recently made new irrigation development increasingly difficult to justify. Even so, world food supply will depend even more on irrigation in the next century than in the present. Future irrigation investment must therefore focus on lower cost solutions, both for new development and for rehabilitation, on making better use of existing irrigation facilities and on increasing output value per unit of water used. Planners should seek to establish the conditions that will promote this focus. 


In recent years, water issues have been the focus of increasing international concern and debate (1). More than two thirds of the water withdrawn from the earth's rivers is used for irrigated agriculture; in developing countries the proportion is even higher - more than 80 percent. But agriculture is a relatively low-value and often highly subsidized water user. Competition for water with other sectors is already constraining economic development in many countries; as populations expand and economies grow this competition will intensify, as will conflicts between water users, or between countries where such competition transcends international borders.

Cities and industries can afford to pay more for, and earn a higher economic rate of return from, a unit of water than agriculture. Hence governments and financing institutions are being forced to reconsider the economic, social and environmental implications of investment in irrigation. As a result, it is likely that in future the water sector will be less dominated by irrigation, and in some countries water formerly used for agriculture is already being reallocated for higher-value uses.

The 1993 World Bank policy paper Water Resources Management crystallized much of the thinking of governments and financing institutions with regard to the overall management of water resources. It called for new approaches, including demand management, - that is to say the use of economic, legal, institutional and other policy interventions to influence the demand for water in order to improve the efficiency of its use. In countries with significant water management problems, the international financing institutions increasingly require the preparation of inter-sectoral water resources management strategies to guide the lending programme in the water sector, as a precondition to lending for irrigation. The implication is that loans for irrigation development will not be made where this will prejudice other more profitable or socially desirable uses of water.

These Guidelines, however, start from the assumption that water policy reviews have indicated that irrigation is a justifiable option within the context of a country's overall water resources management strategy, and that investment finance could be made available for its development. The Guidelines therefore do not address the principles and processes involved in water resources management strategy formulation, which are well covered elsewhere(2). The need for project planning to be in strict conformity with such strategies nevertheless cannot be overemphasized.


Irrigation Typologies

With about 250 million hectares irrigated throughout the world in vastly different climatic and socio-political environments, some categorization of irrigation may be thought desirable. Numerous typologies are commonly used, such as system size, the nature of the water source, and whether schemes are operated publicly or privately. Definition by size presents difficulties on a global level, since, for example, what might be considered large-scale in some countries in Sub-Saharan Africa would be considered as only small or medium-scale in South Asia. Furthermore, many of the problems confronting publicly financed irrigation transcend scale, and some attempts at categorization have confused "small-scale" with "traditional" or "informal" irrigation. Definition by the type or nature of the water source does not recognize the very different characteristics of major public surface water schemes based on dams in the USA, for example, and small community-managed tank irrigation schemes in Sri Lanka.

From the technical viewpoint, a further distinction can be made between rice schemes, which comprise more than half of the world's irrigation, and non-rice schemes, because of their fundamentally different characteristics. Since the rice plant tolerates waterlogging and needs much more water to thrive than almost all other of the major irrigated crops, it is dominant where water is cheap and plentiful, notably in the humid eastern side of South Asia eastward through to Japan and Indonesia. Non-rice projects are generally found in the drier or cooler parts of the developing world. The design and operation of a rice-growing irrigation system is significantly different from that for other crops: rice fields are waterlogged to reduce the weeding requirements, whereas crops such as wheat, maize and especially cotton will die under these conditions(3). Once the crop is established rice schemes usually receive a small but continuous flow to maintain flooded conditions; field-to-field irrigation is acceptable because a down-catchment farmer will often use what an up-catchment farmer wastes. Irrigation systems designed for other crops do not usually suit rice very well, and vice-versa.

For the purpose of these Guidelines, irrigation could perhaps be categorized globally as either public or private, ie by the degree of end-users' commitment of resources to, and control over the operation of, the system, rather than the usual categorization by scale. Even this is difficult to define precisely, since the share of public versus private resources can vary widely between schemes. Nevertheless, public irrigation is defined here as any irrigation in which government has the dominant financial interest or management responsibility/control. Public irrigation may range in size from schemes of hundreds of thousands of hectares, down to very small schemes of 10 ha or less; but in each case these are initiated and developed under public authority and control, and operated and maintained in the same way. This definition includes for example state-owned large-scale estates (e.g. for sugar cane production), joint ventures between government and quasi-government financial institutions, and large-scale through to small-scale smallholder irrigation schemes set up under government authority.

Conversely, private irrigation can be defined as any irrigation in which farmers (or a private sector group) have the dominant financial interest or management responsibility/control. It includes: 

The most important differences between public and private irrigation as defined above are that: 

Public irrigation therefore tends to be supply-driven and may incorporate political or social objectives, while private irrigation is demand-driven and reflects financial objectives or, at times, the survival strategy of the poor. These characteristics impinge on many of the basic decisions for development planning and imply fundamental influences on the investment approach. The features that make successful private irrigation self-sustaining should if possible be emulated in planning public irrigation investments. Thus farmers must be involved in the planning decisions, contribute at least a part of the capital costs, and accept full responsibility for operation and maintenance (O&M). As will be seen, the planning and investment trend in publicly financed irrigation is to emulate those characteristics of private irrigation that make it generally self-sustaining.

The Issue of Scale

According to a World Bank review of its experience in irrigation(5), there is no evidence to suggest that small-scale irrigation is more or less likely than large-scale to achieve success, judged in terms of sustained economic internal rate of return. Nevertheless, it can be argued that where irrigation institutions - public or private - are still relatively weak, where there is a lack of capacity to plan, implement, operate and manage large schemes, attention should focus on smaller developments. Smaller schemes are more conducive to farmer management and control, and market limitations for the crops produced often make such schemes the only viable choice. On the other hand, there are many examples of the development of small public irrigation systems, scattered over a wide area, that have overstretched the logistical and staffing capabilities of irrigation agencies and have eventually failed. In theory, larger developments should encourage more Government support, attract better management, be easier to organize, and therefore enjoy better prospects for sustainability.

There are numerous other arguments for and against large or small irrigation schemes: for example, the obvious economies of scale and multiplier effects of large schemes (see Box I-1). Many of the arguments are valid in some countries, for certain irrigation types, but not in others. Thus, generalization should be avoided and the issue of scale should be approached considering the individual circumstances of the project and institutional capacities in the country concerned. As will be seen, there are more important issues than scale: the overwhelming experience is that what is important in predisposing irrigation to success is the extent to which it enjoys the commitment of stakeholders(6) to good engineering design, quality construction, efficient operation and adequate and timely maintenance. 

Box I-1. Large versus Small Irrigation Schemes

Large Scale

Small Scale


Engineering economies of scale usually possible, hence, potentially lower unit costs.

Governments more disposed to take the actions necessary to ensure that project succeeds.

Economies of scale result in cost-effective provision of extension services and social/economic infrastructure.

Greater regional impact of secondary benefits.

Easier physical planning of contiguous blocks than scattered areas.



Demand for high level professional skills and institutional capacity in planning, implementing, operating and maintaining. 

Relatively complex organization and management requirements; scope for farmer management limited to tertiary system, hence greater recurrent cost burden to government or other central authority (which may offset potential economies of scale). 

Longer period required to bring complete project into production. 

Greater potential for irreversible adverse environmental and social impacts, such as displacement of settlements or disruption of wildlife habitats.


Usually less exacting technical demands for high level professional skills for planning, implementing, operating and maintaining.

Greater opportunity for farmers to participate in planning, financing, implementing, operating and maintaining.

Better adapted to supplying local markets with (high value) horticultural products without depressing prices.

Relatively simple organization and management.

Often quick yielding.

Smaller risk of irreversible adverse environmental and social impacts


Diseconomies of scale sometimes result in relatively longer period required to plan and implement (per ha developed). 

Fragmented distribution results in more difficult logistics for implementation, extension coverage and provision of social and economic infrastructure.






Options and Alternatives for Food Supply

As populations in some developing countries continue to grow faster than increases in food production, the options for meeting the consequent incremental demand for food need to be considered. These are discussed below. 

Rainfed Production of Food as an Alternative to Irrigation

Although between 30 and 40 percent of the world's food at present comes from the irrigated 20 percent of total cultivated land, before contemplating further irrigation development the potential for increased food output from rainfed areas should be considered. There may be prospects for obtaining sustainable production increases under rainfed conditions through relatively simple low cost technologies: for example improved in situ water conservation techniques(7), and the adoption of integrated pest management and integrated plant nutrient management strategies. However where land resources are scarce, further area expansion of rainfed food production could increasingly involve more marginal areas, with a risk of increased deforestation, soil erosion and general land degradation. In the less well-endowed areas particularly, the potential for stabilization or intensification of existing rainfed production by increased use of agrochemical inputs is also technically limited: either the possible gains have already been achieved, or they are unlikely to be achieved because of aversion by farmers to the known risks of investing in improved inputs in marginal rainfall areas. 

Hence, even though irrigation development cannot, and perhaps should not, be relied upon to meet the entire future increase in demand for food, supply can be expected to depend to an even larger extent on irrigation in the next century than it has in this. 

New Irrigation Development

Increased production through new irrigation development is nevertheless increasingly difficult to justify economically for the production of basic foods, because of the decline in world market prices for these crops(8) and typically high per hectare capital costs (see Box I-2, also Annex 2 of Investment Centre Technical Paper 5(9)). The situation may change in the longer term if, as world population grows, the demand for food begins to outstrip supply. In this case prices might reasonably be expected to approach the marginal cost of irrigated production, and the use of current World Bank price projections for project analyses may be inappropriate(10). Indeed, should world market prices for basic food crops show signs of recovery, this could significantly alter the profitability of production of such crops under irrigation. Nevertheless, for the foreseeable future any expansion of irrigation for the production of basic foods will only be possible if substantial reductions in per hectare capital costs of new development can be achieved. 

In many countries, however, the better irrigation sites are already developed, and hence new projects could be expected to cost even more per hectare than those developed in the past. New irrigation development in these countries may therefore increasingly be justified only for the production of relatively high value crops - for which markets and marketing are often constraints - rather than for basic foods. In this situation markets, as much as the availability of suitable sites, will determine the pace of investment in new irrigation, unless lower cost technologies can be devised and introduced. This is today's challenge to irrigation engineers. 

Box I-2. Asian Food Production in the 1990s 

The introduction and rapid spread of high-yielding rice and wheat varieties combined with heavy investment in irrigation and rapid growth in fertilizer use in the late 1960s and the 1970s resulted in strong growth in output of these crops in Asia. For rice, the rate of growth of yields increased from 1.7% per annum during 1958-66 (before the spread of modern technology) to 2.9% during 1974-82. However, growth in rice yield, the primary contributor to rice output growth throughout these periods, has slowed to 1.9% annually since the early 1980s. 

Area expansion contributed about one-third of Asian rice output growth during 1966-74, but little after that. The annual growth rate in rice output therefore declined in the 1980s, from 3.1% in 1974-82, to 2.2% during the period beginning in 1982. Similar trends have occurred with wheat output. 

Reductions in the amounts of new investments in irrigation have been dramatic. Aggregate lending and assistance for irrigation in Asia in the 1970s and 1980s by four major financial institutionsa/ reached its peak in real terms in 1977-79. By the mid-1980s it was less than 50% of the 1977-79 level. What has caused this decline in investment? Contributing factors include the large public and foreign debt loads carried by most of the agriculturally based economies in the region, the declining share of unexploited irrigation development potential in many countries in the region, inter-sectoral competition for water, and increasing stringent project evaluation in response to political resistance from environmental interests and those displaced or otherwise negatively affected by irrigation development. 

However, the main reasons for declining investment are the increasing real costs per hectare of new irrigation development and decline in world rice and wheat prices. Rosegrant and Svendsen presented real capital costs for construction of new irrigation systems in five countries in South and Southeast Asia over the period 1966-88, the unweighted average for which increased by a factor of 2.5, from US$1,744 to US$4,385 per ha, over the period. The real price of rice and wheat over this period was halved, representing a swing of a factor of 5 in the ratio of costs to benefits. 

a/ World Bank, Asian Development Bank, Japanese Overseas Economic Cooperation Fund, and US Agency for International Development. 

Source: Mark W. Rosegrant and Mark Svendsen (1993), Asian Food Production in the 1990s: Irrigation Investment and Management Policy, "Food Policy", February 1993.


Intensification of Existing Irrigation Systems

Given that the cost of new irrigation development for food production is increasingly difficult to justify, and that many existing irrigation systems perform below potential, it is logical to consider intensification and increased output from existing systems. The investment emphasis in recent years has therefore shifted towards improving the latter, taking advantage of sunk costs to achieve incremental production at low incremental cost. It is important here to note the distinction that has been made between an endless cycle of rehabilitation, which is simply deferred maintenance, and upgrading, which involves making existing schemes work better. Upgrading usually calls for engineering, economic and sociological analysis to arrive at solutions. Sectoral loans aimed at such improvements have become an important part of the portfolios of the financing institutions, often linked to system management transfer to the users (see Chapter 7). 

Low Cost Irrigation

The above discussion focuses on irrigation development in formal systems and takes no account of the existence, in various parts of the world, of large areas of informal or traditional irrigation. 

These, by definition, have been developed on the initiative of farmers rather than governments, and have continued their existence in the same way. Traditional irrigation systems are often characterized by poor water control, and consequent low cropping intensities and yields. In many cases improved water control can be achieved at comparatively low cost, and is often easily justified by the incremental production that can be achieved as a result. Thus, given that in some countries the area under traditional irrigation far exceeds the area under formal irrigation, the scope for obtaining increased food production from these systems could be significant. The identification of opportunities for such improvements may therefore be a priority for planners. However, it must also be noted that the most important feature of these systems is local initiative, responsibility and control; proposed improvements should avoid inadvertent transfer of responsibility to government. 

Apart from traditional irrigation systems, other opportunities exist for low cost irrigation, particularly for localized irrigation, including systems based on the use of clay pots for the storage and gradual release of irrigation water. These, and other similar devices, often bring nutritional benefits to local communities because they are generally used for fruit and vegetable production. They make efficient use of scarce water, but are in general unsuited to large-scale food production(11)

Irrigation and Protection from Drought

In many regions of the world the major river systems have their headwaters in high rainfall or snowmelt areas and flows are relatively insensitive to droughts in agricultural areas downstream. Here the value of irrigation in "drought-proofing", by creating greater yield stability and out-of-season food production, is undoubted. 

In other areas subject to repeated and prolonged droughts, such as the Sahel, northeast Brazil or southern and eastern Africa, although at first sight there would appear to be no apparent alternative for improving local food security, irrigation does not always provide full insurance against drought. In much of Sub-Saharan Africa for example rivers and dams dry out and groundwater levels drop in years of recurrent drought. It could therefore be argued that in these circumstances irrigated agriculture is more vulnerable to drought than some less intensive forms of agriculture. Moreover, even where irrigation potential remains unexploited in these areas, its development cost nowadays will often only be justified by high value crops. These have limited markets and will bring primary benefits to only a few of the people normally at risk. Thus, despite its superficial appeal, irrigation development in these areas may not be a fully effective means to combat recurrent drought, rural poverty and food insecurity. There are unfortunately no easy, quick fixes for these problems. Further research is necessary, aimed at developing viable technical recommendations that take account of the recurrent drought cycle, including opportunities for non-farm rural employment. Policy assumptions that automatically equate irrigation development with the elimination of drought risks in such areas should be regarded with caution. 


Experience to date, well summarized in the 1992 report of the World Bank's Portfolio Management Task Force referred to earlier, makes it clear that a key condition for sustainable development impact from irrigation investments is implementability. This requires: 

Other factors, such as good technical design, sound construction and financial viability for the users are of course equally important. But experience indicates that in the past irrigation professionals have often underestimated the attention which also needs to be given to implementability. 

Implementation Capacity and External Technical Assistance

The conventional project identification/preparation approach of the past has often resulted in arriving at detailed project design only to find a mismatch with local capacity to implement it. Money and technical assistance has then been provided to bridge the gap. Technical assistance frequently then crowds out any local capacity, and may in effect substitute for local management rather than strengthen it, bringing no sustainable improvement. 

In the conventional planning process, detailed start-up and implementation plans have generally been considered as beyond the ambit of the identification/preparation teams work. That this was inappropriate is now clear and a new approach has been found necessary: projects should be planned to match local capacity for implementation, which implies that planning teams should first acquire a thorough appreciation of this capacity. If necessary the project scope and content may be reduced to match existing implementation capacity. Technical assistance can then be applied selectively, rather than indiscriminately as often in the past, for genuinely sustainable capacity to be built. The planning process should therefore give specific attention to an analysis of institutional capacity, and to providing a detailed programme to enable the implementers to prepare themselves for carrying out the tasks expected of them, once the project becomes effective. 

Participation, Ownership and Commitment

Successful implementation requires participation in the planning and implementation process by all stakeholders, in order to create a sense of ownership of, and consequent commitment to, the project. This requires that the project planning process should allow time for the borrower and users to participate in, or preferably drive, the planning process, and for any potential losers to have a substantive influence on decisions that affect their future. Ownership and commitment by the users are unlikely to be achieved unless they consider that the project would meet their felt needs(12) and they have a stake in the equity - that is, they share in or bear all of the investment costs. 

Building ownership and commitment through participation has often been difficult to achieve in the past. The conventional sequence of identification/preparation, carried out against tight deadlines by external planning teams, has seldom allowed time for genuine participation (which should go beyond mere consultation), either by government staff or farmers. On implementation, government irrigation engineers, for their part, have usually seen irrigation only from an engineering, rather than a farming or social, perspective. They have been reluctant to adopt participatory approaches with farmers, mainly because of a misplaced belief that farmers are unable to understand or make any contribution to technical matters, or because of concerns that participation might delay implementation or result in design changes that compromise the quality of the final product. 

Undue delays in project approval and implementation are undesirable, not least for the farmers; but taking time over stakeholders' participation in planning does not necessarily mean delay. It can often pay dividends, by preparing the implementers, ensuring smooth start-up, building farmers' commitment to change, and may ultimately lead to more rapid implementation and a more sustainable development impact. Experience has shown that the ultimate scheme design almost always benefits from involving the users in the planning process. Farmers, or at least those with some experience or knowledge of irrigation, from the poorest illiterate smallholder to the richest well-educated commercial farmer, usually have practical ideas of what works and what does not, from their detailed local knowledge of weather patterns, hydrology, soils, markets, and so on. Communities often have strong preferences regarding the nature and location of development that needs to influence planning, such as aligning a canal to avoid excavation in sacred ground. 

Participatory or consultative planning is essential in rehabilitation projects or the upgrading of traditional farmer-managed irrigation systems, to take advantage of the invaluable store of cross-disciplinary knowledge that farmers possess about the existing systems. Projects that involve the displacement and resettlement of people can only be planned and implemented effectively if those affected are involved in the planning process and their suggestions and concerns taken fully into account. 

As will be seen from Chapter 5, involving farmers in system design can also often result in significant cost savings, particularly if the farmers themselves are expected to take a share in the equity by contributing to the investment costs. Sound engineering is essential, but it can nonetheless take account of the farmers' experiences and preferences. Yet farmers, as a possible source of system design input, are still too often ignored by engineers, and as a result schemes are often inappropriately planned (see Box I-4). 

Box I-4. Second Approximations:
Unplanned Farmer Contributions to Irrigation Design

Farmers interviewed on the Kosinggolan Scheme of the Dumoga Irrigation Project in North Sulawesi, Indonesia frequently reported that during construction they had approached construction labourers or supervisors in the field to suggest changes and were usually told that the design had been established by the government and could not be changed. Often farmers relocated the construction markers when the crews had left. Others waited until construction was finished and the contractors had moved on before altering the structures. Altogether, 27 design alterations were identified in the sample blocks. Many cases involved multiple alterations that were interconnected. 

The most common kinds of alterations observed were channels being relocated, streams being diverted or ponded, project channels being abolished or not used, and channel offtakes or division points being relocated. Other actions included redirecting project channels into drains or streams, making new channels, adjusting division box gates to alter water divisions, making new flumes, destroying project flumes and lining channels. Several cases involved relocating channels to follow farm boundaries, to accommodate low water requirement crops or to continue to make use of pre-existing farmer-built structures such as small weirs, channels and ponds. 

The most frequent reasons reported by farmers for making design changes related to questions of conveyance and distribution efficiencies, farm boundaries and the conjunctive use of alternative water sources (in this case from natural waterways or from return flow). 

Source: Vermillion D.L., in Design Issues in Farmer-Managed Irrigation Systems, Proceedings of an International Workshop of the Farmer-Managed Irrigation Systems Network held at Chiang Mai, Thailand, December 1989. IIMI, Colombo, Sri Lanka (1990). 


A Possible Role for Non-Governmental Organizations (NGOs)(13) in Participatory Development

Farmers may be as uninterested in participation as government irrigation bureaucracies, especially if in the past they have received free, though possibly unreliable, irrigation services and are now expected to bear more of the costs. They may often be suspicious of government officials, especially if they have been the losers as a result of incompetent or corrupt practices. Farmers may therefore require considerable persuasion to commit themselves to participate. 

Participation usually also requires behavioural change in irrigation bureaucracies, and until this is achieved they may not be in a position to implement participatory development. Some bureaucracies have successfully employed young graduates in social science to work directly with farmers to assist the latter to mobilize and organize themselves to participate in project planning and construction(13). However, for various reasons it is often difficult for irrigation bureaucracies to attract and retain such staff. Even if such staff can be recruited, farmers may still hesitate to cooperate fully with persons they regard as government agents. In this case what is often required is a non-governmental intermediary, to identify community needs and articulate them on behalf of the otherwise voiceless. While non-governmental organizations vary in their ability to work with the poor and to cooperate with government agencies, several of them have undertaken this role successfully. 

However, caution is necessary to avoid any suggestion that NGOs should replace more formal local institutions: in some cases, there would be advantages but in others such a move could be counter-productive. Instead, suitably qualified and motivated NGOs may be sub-contracted, either by local government(14) or through the farmers' own administrative structures(15), to provide technical assistance services to animate participation. They may also often assist in capacity building by training government staff in this role. 



The Need for Cost Recovery

Economic efficiency and fiscal sustainability demand that the capital costs of irrigation infrastructure should eventually be recovered from the users, in order to permit longer-term replication of investments. On equity grounds, it can also be argued that costs should be recovered, as farmers who are able to irrigate tend to be amongst the better off members of the rural population. In practice few countries have ever succeeded in recovering much more than the O&M costs of public irrigation directly, although indirect recoveries in the form of agricultural taxes or generally negative terms of trade for the subsector have in some cases been important (see Box I-5). 

Box I-5. Cost Recovery: Setting the Appropriate Level 

A 1986 World Bank Operations Evaluation Department (OED) report described serious cost recovery problems on the Muda project in Malaysia, where at the time of audit, water charges and land taxes remained far short of meeting O&M costs. The prospects for raising direct cost recoveries were considered poor, partly because of the heavy burden on farmers of zakat, a religious tithe, and a substantial sales tax collected from produce in the region. However, the audit report concluded that if the zakat, estimated in an FAO study to be 5-7 percent of gross farm income, as well as the indirect return to the government resulting from controlled prices were taken into account, the Muda farmers' combined payment of water charges, taxes, and the production tithe covered all the O&M costs plus 20 percent of capital costs (at 10 percent annual interest). 

Source: OED Report No 6233, World Bank Lending Conditionality: A Review of Cost Recovery in Irrigation Projects. World Bank, Washington DC, June 1986. 

It has also been argued that the complexity of some irrigation and drainage schemes justifies state intervention and subsidizing of part of the investment costs, without which some worthwhile projects may not have been constructed. In these cases capital costs not recovered may not really be subsidies if all the secondary benefits of irrigation development are taken into account. In a case study of the Muda Irrigation Project in Malaysia, for example, it was found that for every dollar of value added generated directly by the project, another 80 cents were generated downstream(16). Moreover, other researchers have argued that the multiplier effects of investment in agriculture in developing countries are generally greater than those associated with investment in other sectors(17). Nevertheless, the governments of many developing countries faced fiscal crises during the 1980s that focused their attention on the shortcomings of existing policies for financing irrigation, particularly with regard to the O&M costs. The general consensus now, among governments and financing institutions, is that users should pay all of the O&M costs and as much as possible of the capital costs. 

Cost Recovery, O&M and Water Charges

Problems with cost recovery and O&M form a vicious circle. Irrigators on public schemes are commonly reluctant to pay any charges that they are not forced to: poor collections lead to poor O&M, and an even greater reluctance to pay. It is also generally accepted that the standard tactic for dealing with poor O&M in the past - that of relying only on raising water charges - does not usually work, mainly because revenue from water charges (if they are collected) is often returned to the general treasury instead of being allocated to O&M. There are exceptions to this rule, for example on public schemes in Morocco, but in these cases increased water charges have been accompanied by improved service, hence greater willingness to pay on the part of the farmer. 

Any suggestion that irrigators on a public irrigation scheme cannot afford to pay even the O&M costs needs to be examined very critically. If irrigators cannot pay it can only be assumed that the scheme is either unviable - in which case the question should be asked why it was built or what can be done to make it viable - or unreliable, in which case measures should be taken to correct the situation. Moreover, there is an issue of equity involved in subsidizing some members of society by way of artificially cheap irrigation: in principle it may be a praiseworthy social objective, but with typically high unit costs for irrigation development, the social impact would in most cases be greater if any subsidy was spread more thinly over a higher proportion of the rural poor. Fiscal constraints in many developing countries simply do not permit subsidies of this kind anyway. 

The application and collection of water charges can be further complicated by various factors, including for example local customary law, or a fundamental belief in some countries that water should be free. There is also the difficulty of volumetric measurement. The first of these may be overcome to a certain extent by charging a "service" fee for irrigation. But this usually takes the form of an area or crop-based fee that provides no incentive for the efficient use of water and may thus contribute to wasteful usage. Even if it were possible to charge individual farmers for water on a volumetric basis, which it seldom is for most surface irrigation systems involving smallholders, setting an appropriate charge may present some difficulty because of local economic distortions (see Box I-5 above). Nevertheless, without some form of volumetric charging, individual irrigators have little incentive to make more efficient use of water. 

One solution to the problem of water charges is to devolve financial autonomy for O&M to users' groups, or to irrigation agencies dependent upon the users for finance. Metering of water supplied to larger groups, even if only approximate, is usually more technically feasible, and the group as a whole can then be charged. It becomes the group's responsibility to allocate water amongst its members and to recover the charges; the experience is that users' groups are more effective collectors of fees than government agencies. However, this solution will depend for success entirely on the cohesion of the group involved. Conditions for the sustainability of users' groups are discussed later. 

Users' Contributions to Capital Costs

Apart from the obvious fiscal advantage, a contribution by users towards the capital cost of a new or rehabilitated scheme is an indication of demand and commitment. In effect, this is an investment in equity and hence the scheme becomes to some extent private, enhancing prospects for sustainability. For this to happen it requires that the users will be in a position to make such a contribution, and that there are no economic distortions in place that make it impossible for them to do so. 

It also presupposes that government is committed to recovering costs. Yet there are many examples around the developing world where governments and donors have adopted the view that users are too poor to make any capital contribution. In these cases the construction or rehabilitation of an irrigation scheme is often seen as a welfare project rather than as an investment project, and farmers are paid in cash or food to contribute labour to the construction of the scheme. There are no signs that such an approach engenders any sense of ownership or responsibility. Farmers are more likely to view construction simply as a source of off-farm employment, to regard the scheme as government infrastructure and ultimately to show little subsequent commitment to it. Moreover, the injection of large amounts of food into an area under food-for-work programmes can sometimes depress agricultural prices and affect other farmers' incomes. 

Even among very poor populations, individuals and communities have been willing and able to invest substantial amounts in cash and kind for projects that they consider are worthwhile. If nothing else, irrigators or prospective irrigators can contribute labour, even if only for a few hours a day, and provide locally available materials for construction. Unwillingness to contribute implies a lack of demand for the irrigation development proposed, hence a lack of commitment, which invariably leads to unwillingness to accept O&M costs. Thus the International Fund for Agricultural Developments (IFAD) irrigation and drainage investment strategy(18) requires that users should "contribute between 10-20 percent to the direct costs, preferably in kind or labour", and that they should "pay for the cost of irrigation equipment (cash/loans)". Experience suggests that this is not unreasonable, if the development proposed is an appropriate response to local demand for irrigation(19). On some public systems in Morocco, for example, up to 40 percent of the initial capital costs are recovered from farmers, and private irrigators throughout the world - including those on traditional farmer-managed irrigation schemes - are willing to pay up to 100 percent of the cost of their schemes. 

As mentioned, case studies have shown that requiring a capital cost contribution from farmers can result in significant overall savings if farmers themselves are involved in system design. For example, in the IFAD/World Bank-funded Communal Irrigation Development Project in the Philippines, actual costs were US$ 4,100 per hectare compared with the originally estimated US$ 7,000, not only because of the farmers' contribution of labour and materials, but also because of farmer-led design changes(20). Farmers will invariably seek the least cost solutions if they have to pay even a part of the costs. 

Land tenure problems can, however, be a constraint to users' participation in capital costs. Where farmers are unable to gain legal or even customary title to their land they are not normally willing to invest, unless they have confidence in long term usufructuary rights. Legislation may therefore be necessary in order that freehold tenure or long leases can be granted - for which reason, among others, government commitment to the proposed investment is essential. 



Water Users' Associations and Transfer of O&M Responsibility

Fiscal crises have in many cases forced governments to devolve financial and managerial responsibility for existing irrigation systems to the users - in effect to privatize them - and to ensure that users' groups on new schemes accept full responsibility for O&M from the outset. 

The degree of responsibility which water users' associations (WUAs) can be given for management of a system depends on its scale. It is obviously unrealistic to expect a WUA to take over full responsibility for a system that serves hundreds of thousands of hectares and which was previously operated (even if poorly) by a large irrigation agency. On the other hand there are other options available, such as users' representation, either directly or indirectly through apex WUA organizations, on the board of a financially autonomous irrigation authority, or for WUAs or their apex organizations to make a contract for the provision of irrigation services with the private sector. Smaller schemes, including their main water supply infrastructure, might on the other hand be managed entirely by a WUA. The objective in either case is greater user commitment, which can lead to more efficient use of the resources by helping to overcome many of the problems that public irrigation systems face, such as inequitable water distribution, corruption (see Box I-6), inefficiency and poor O&M. Attention is nowadays being focused on how to achieve this commitment, and to what extent WUAs can be assisted to form and to manage their own affairs. 

Box I-6. Corruption in Public Irrigation Schemes 

Social research and experience have shown that irrigation projects in some developing countries provide irrigation engineers and other operational personnel with opportunities to raise significant amounts of illicit revenue from the distribution of water and contracts, some of which may be redistributed to superior officers and politicians. Thus, in return for financial inducements, irrigation engineers will award contracts to high-priced or unqualified, incompetent contractors, and "turn a blind eye" to substandard work that saves costs for the contractor and increases his profit. The results of such corruption are not usually immediately apparent, but substandard work obviously has a detrimental impact on subsequent maintenance requirements and costs, contributes to the vicious circle of poor maintenance-poor cost recovery-poor maintenance, and hence has an obvious bearing on sustainability. Financial inducements may also be used to bribe ditch-riders and other operational personnel to enhance water supplies to one farmer, or a group of farmers, at the expense of others, usually the poorest and least powerful, which often means tail-enders. 

Corruption of this kind is considered to be one of the most important supply-side factors in the poor performance of public irrigation. It has been very difficult to control in the past because of lack of financial discipline and accountability within irrigation bureaucracies. 

Source: Wade R, The System of Administrative and Political Corruption: Canal Irrigation in South India. Journal of Development Studies 18(3): 287-328. 


Conditions for Sustainability of Water Users' Associations

Experience to date in the formation of WUAs and the transfer of irrigation O&M responsibilities to them has been uneven. The 1994 World Bank review of its experience in irrigation, referred to earlier, concluded that some WUAs have been stillborn, some have died in infancy, and some have lived on but performed no useful function. Byrnes(21) concluded that most WUAs in World Bank-assisted projects in Pakistan remain relatively weak. Meinzen-Dick et al.,(22) reached a number of conclusions regarding what leads to strong WUAs, the policy factors that can assist in the development of such organizations, and the implications for constructive interaction between irrigation agencies and WUAs, particularly for the transfer of irrigation responsibilities.

These were summarized as follows:

However, turning over management of public irrigation schemes is not merely a matter of consultation and forming WUAs, with government continuing to act as the prime mover. Power struggles, collusion and corruption may not always be eliminated by user participation and the creation of WUAs. Not too much should be expected from them, especially in the short term. Their creation requires a re-orientation of irrigation bureaucracies towards providing a service and creating an environment that facilitates the formation, by the users themselves, of sustainable WUAs. Sound social engineering is as necessary as good technical engineering. The ease with which sustainable WUAs will form, and successful transfer of responsibilities will then take place, will vary according to different physical, social and financial circumstances. There is no magic solution, no one set of rules can be applied, and considerable time and resources will have to be invested in learning how best to approach the process in each case. Investment project designs which provide flexibility and a progressive or pilot-led approach to transfer are more likely to lead to eventual success(23).



Adverse social and environmental impacts of irrigation investments have been many and varied. They include health impacts (malaria and schistosomiasis), and waterlogging and induced salinization(24). Displacement of people from dam sites with inadequate consultation and compensation has also been a major source of problems. Land acquisition and resettlement requirements have often caused delays to implementation or even cancellation of loan agreements. Although some would argue that on the whole the social and environmental disbenefits of irrigation are far outweighed by the benefits, there are a number of irrigation projects around the world that possibly would not have been built had the full negative effects been foreseen, costed and entered into a cost benefit analysis. 

Box I-7. Some Social and Environmental Issues in World Bank-Financed Irrigation Project Planning 

The 1994 World Bank review of its irrigation experience commented as follows on the coverage of gender issues in its sector work: 

"Irrigation affects men and women differently. Even if they have equal roles in agriculture, which they usually do not, women almost always have primary responsibility for such household tasks as food preparation, washing and providing drinking water. However, except for the most recent studies on Mexico and India, none of those (sector reports) analyzed was found to have addressed the subject." 

and on organization and management: 

"Coverage of management and organization was broad but generally superficial. It focused on management and organization of government institutions, occasionally on their relations with irrigators' organizations, and never on the irrigators' organizations themselves."

and on broader environmental issues: 

"Coverage of special areas of environmental impact has been poor and is still quite weak. This is the case for drainage, an important source of environmental troubles in numerous countries, and especially so for aquifer management, and the various dimensions of catchment management: deforestation, overgrazing, inappropriate farming, soil degradation, erosion, and silting." 

Source: Report No 13676: A Review of World Bank Experience in Irrigation. Operations Evaluation Department, World Bank, Washington DC (1994). 

Despite the years of experience and the lessons learned - and despite the existence of clear operational guidelines for dealing with social and environmental issues - governments, financing institutions, project planners and implementers have in the past often paid only lip service to the need for systematic problem identification, assessment and mitigation. In the past one of the reasons for this was that promoters and implementers of irrigation projects found that addressing such issues was an inconvenience, as well as a likely source of project delays or cancellations. And even though environmental legislation existed, environmental agencies generally did not have the teeth to implement regulations; if planners wished to strengthen these agencies in parallel with the formulation of irrigation investment proposals, it was already too late. 

The approaches and attitudes of governments and financial institutions have changed more recently. The consensus now is that social and environmental impact assessment is essential and as important as economic analysis in influencing the design of projects. 



Common sense dictates that the choice of technology for irrigation should be based on its appropriateness for the cropping patterns intended and should also consider cost-effectiveness. Irrigation engineers have in the past tended to overlook an additional need: for the technology also to be matched to the level of sophistication or operational capacity of the users. It has become increasingly obvious that the design process must start from a consideration of how the users will operate the system; this should then be designed to provide the optimum combination of efficiency in water use and cost effective operation and maintenance. Equally important, the designer must consider how the user will cultivate his land, and the implications that this may have for scheme layout. Thus it may be that the design which involves the lowest investment cost per hectare may not be the most cost effective solution if it also involves large numbers of staff for its operation, or if, because of operational difficulty, it cannot be utilized to capacity. On the other hand, a design to improve water use efficiency on a traditional irrigation system by the introduction of "modern" water control structures may not result in overall efficiency gains if the users reject the modern controls in favour of their traditional proportional dividers. 

The choice of technology, whether for new development or rehabilitation of existing schemes, has been the subject of much debate over the years. While most irrigation engineers would now agree that the starting point for design must be ease of operation, they still tend to polarize into two camps. One sees the problem largely as overcoming the hydraulic instability of extensively-gated manually operated systems; it sees the solution as the modernization of these systems, adding automatic downstream control structures and other feedback mechanisms designed to achieve hydraulic stability. The other accepts the reality of farmer damage in wet season drought and so favours designs based on cruder and more robust structures; the possibility of just-on-time, demand-based, delivery of water to crops is foregone, in the hope of preserving the civil works from interference(25)

Discussion of this issue is well covered elsewhere(26) and need not be continued here except to note the conclusion of the World Bank in its 1994 review of its experience in irrigation, that there is inconclusive evidence to favour one camp or the other. Both would agree on the need to eliminate anarchy and on the importance of flexibility of operation. 



The world is faced with a huge backlog in drainage requirements. Over the last quarter of a century water usage for irrigation has more or less doubled without a comparable increase in drainage capacity. In the longer run poor drainage is one of the most significant causes of reduced yields and of irrigated land going out of production, as shown, for instance, by the extent of saline and waterlogged areas in Pakistan. The stage has now been reached when it is necessary to correct the drainage omissions of the past. At the same time there is a need to improve water use efficiency to reduce the drainage demands of the future. 

It is necessary however to consider why, even where provision for drainage has been made in the past, it has often survived for only the first few years of a project's life. In most cases of poor scheme maintenance it is the drains that are allowed to deteriorate first. One of the main reasons is that within a year or two of construction tertiary drains are often cultivated over by the irrigators who are theoretically responsible for them. Secondary drains, which are usually the responsibility of the irrigation agency, are often also partially filled in by farmers to provide crossings or to pond water for other purposes. The main drains are therefore quickly rendered redundant. 

To improve the sustainability of drainage systems, channels should be limited to those which are essential; but these should be adequately maintained and defended against encroachment. Provision of crossings, each with adequate culvert capacity, is essential or obstruction by informal cultivator-constructed crossings will inevitably result(27). The challenge is to persuade farmers to accept the importance of drainage and to take responsibility for its maintenance. This further reinforces the need to promote participation and ownership by the users. 



To summarize, there is growing recognition that: 


Further Reading

[Editors Note: This publication is planned for revision in the near future. For an update on the World Banks current perspective on some of the issues raised above, the Investment Centre suggests readers look at these recent references: 



  1. The background to this is described, for example, in the World Bank Policy Paper Water Resources Management, World Bank, Washington DC (1993) and in Land and Water Bulletin 3 Water Sector Policy Review and Strategy Formulation: A General Framework (prepared jointly by the World Bank, UNDP and FAO), FAO, Rome (1995).

  2. See for example: Land and Water Bulletin 3, Water Sector Policy Review and Strategy Formulation: A General Framework (prepared jointly by the World Bank, UNDP and FAO), FAO, Rome (1995); Water Report 6 Methodology for Water Policy Review and Reform (Proceedings of the Expert Consultation on Water Policy and Reform - Rome, January 1995), FAO, Rome (1995); Irrigation and Drainage Paper 52 Reforming Water Resources Policy: A Guide to Methods, Processes and Practices, FAO, Rome (1995).

  3. Rice schemes nevertheless require adequate surface drainage, as total inundation of the crop leads to significant yield losses.

  4. e.g. drip or micro-jet.

  5. Report No 13676: A Review of World Bank Experience in Irrigation. Operations and Evaluation Department, World Bank, Washington DC (1994).

  6. The term stakeholders includes all individuals who may be positively or adversely affected by the project: government planning agencies (planning units, senior decision-makers, Ministers); government implementing and operating agencies (senior and middle level management of line ministries) who may be subsequently responsible for project implementation, operation and management; community-based organisations, including water users' associations (WUAs) or other farmers' organisations; individual farmers; public interest groups; non-governmental organisations; (NGOs) and private sector companies; financing institutions; and international project planning teams such as those provided by the Investment Centre.

  7. See Soils Bulletin 57: Soil and Water Conservation in Semi Arid Areas, Land and Water Division, FAO, Rome (1987); Investment Centre Technical Paper 10: Agricultural Investment to Promote Improved Capture and Use of Rainfall in Dryland Farming, FAO, Rome (1995); also Technical Paper 221: Conserving Soil Moisture and Fertility in the Warm Seasonally Dry Tropics, World Bank, Washington DC (1993).

  8. Paradoxically perhaps, the expansion of irrigation over the last 50 years has been a major factor in the decline in prices, since it has caused relatively strong growth in supply of rice and wheat compared with growth in demand.

  9. Investment Centre Technical Paper 5, Irrigation in Africa South of the Sahara, FAO, Rome (1986).

  10. Doubts over the use of World Bank price forecasts for food and fibre crops have often been expressed. However, in the absence of any better alternatives, the analyst can do no more than attempt best guesses based on the Bank's forecasts, or explore possible future differences between forecast and actual prices through sensitivity analysis.

  11. There are exceptions to this rule: one such system has been satisfactorily demonstrated for the irrigation of cassava in some Sahelian countries.

  12. This is the basis of genuine demand-driven development. Soliciting or orchestrating requests from farmers for government investments in irrigation is not - even if the prospective users promise or agree to make a contribution at some later date. Often farmers are driven by other motives in these circumstances, such as temporary wage employment on scheme construction, and they later lose interest in the irrigation scheme.

  13. For example, the National Irrigation Administration in the Philippines. See An Evaluation of NIA's Participatory Communal Programme, Public Intervention in Farmer-Managed Irrigation Systems, IIMI, Colombo (1987).

  14. e.g. village or district councils.

  15. e.g. water users' associations.

  16. Bell, Clive, Peter Hazell and Roger Slade, Project Evaluation in Regional Perspective. Johns Hopkins University Press (1982).

  17. See Vollrath T.L., The Role of Agriculture and its Prerequisites in Economic Development: A Vision for Foreign Development Assistance. In: Food Policy 1994 19 (5) 469-478.

  18. Drawn from Irrigation and Drainage Cluster - Module: The Role of Water Users Associations, IFAD, Rome (Draft 20/10/94).

  19. Evaluation of the Special Programme for African Countries Affected by Drought and Desertification: Thematic Study on Small-Scale Irrigation and Water Control Activities (Main Report No. 98/073 IFAD-SSA), IFAD 1998

  20. From An Evaluation of NIAs Participatory Communal Programme : Public Intervention in Farmer-Managed Irrigation Irrigation Systems, IIMI, Colombo (1987). Other experiences from different countries are summarised in Robert Yoder and Juanita Thurston (eds), Design Issues in Farmer-Managed Irrigation Systems, IIMI, Colombo (1990).

  21. Kerry J Byrnes, World Bank Technical Paper Number 173, Water UsersAssociations in World Bank-Assisted Irrigation Projects in Pakistan, World Bank, Washington DC, 1992.

  22. Meinzen-Dick R. et al., Sustainable Water User Associations: Lessons from a Literature Review. Paper prepared for World Bank Water Resources Seminar, 1994.

  23. Practitioners are referred to Orstrom, E., Crafting Institutions - Self-Governing Irrigation Systems, ICS Press, San Francisco, California (1992) which covers some practical planning principles which can be applied in most cases. Also Yoder, R., Locally Managed Irrigation Systems - Essential Tasks for Assistance, Management Transfer and Turnover Programmes, IIMI, Colombo, Sri Lanka (1994).

  24. According to the 1990 FAO report An International Action Programme for Water and Sustainable Agricultural Development: A Strategy for the Implementation of the Mar del Plata Action Plan of the 1990s, 20 to 30 million hectares (or about 10 percent of the world's irrigation) is severely affected by salinity and an additional 60-80 million hectares are affected to some extent.

  25. Burns R., Irrigated Rice Culture in Monsoon Asia: The Search for an Effective Water Control Technology, World Development XXI, (May 1993), pp771-789.

  26. e.g. in World Bank Technical Paper 246, Modern Water Control in Irrigation(1994) by Plusqullec et al., World Bank Technical Paper 256, Design and Operation of Smallholder Irrigation in South Asia (1995) by Donald Campbell, and numerous IIMI publications.

  27. Campbell D., Design and Operation of Smallholder Irrigation in South Asia, World Bank Technical Paper 256, World Bank, Washington DC (1995).