E N V I R O N M E N T - z F 3 D E P A R T M E N T _ ^ P A P E R S Paper No. 003 PARTICIPATION SERIES Participation in Irrigation Ruth Meinzen-Dick Richard Reidinger Andrew Manzardo February 1995 A iFILECOPY Environmentally Sustainable Development The World Bank ESD Environment Department Papers Participation Series ENV-001 Participation in Education Nat J. Colletta Gillian Perkins ENV-002 Participation in Water & Sanitation Gabrielle Watson N. Vijay Jagannathan ENV-003 Participation in Irrigation Ruth Meinzen-Dick Richard Reidinger Andrew Manzardo ENV-004 Participation in Social Funds Mary Schmidt Alexandre Marc Forthcoming: Participation in Country Economic Dan R. Aronson, and Sector Work Ellen Tynan Designing Community Based Deepa Narayan Development Participation in Country Andrew Norton Poverty Assessments Thomas Stephens Participation and Indigenous Peoples Shelton H. Davis Lars T. Soeftestad Participation in Forest and Ajit Banerjee Conservation Management Gabriel Campbell Maria C. Cruz Shelton H. Davis Augusta Molnar Intermediary Nongovernmental Tom Carroll Organizations Mary Schmidt Tony Bebbington Copies are available from the World Bank's Environment Department, Social Policy & Resettlement Division. ~A Social Policy and Resettlement Division Participation in Irrigation Ruth Meinzen-Dick Richard Reidinger Andrew Manzardo February 1995 Papers in this series are not formal publications of the World Bank. They are circulated to encourage thought and discussion. The use and citation of this paper should take this into account. The views expressed are those of the authors and should not be attributed to the World Bank. Acronyms and Abbreviations ha Hectares NGO Nongovernmental Organization O&M Operation and Maintenance TM Task Manager WUA Water Users' Association Contents Acknowledgments iii Executive Summary 1 1. Benefits and Costs of Participation 5 Benefits of Participation Costs of Participation Limits to Participation 2. The Participants: Roles and Incentives 11 World Bank Task Managers Policymakers Agency Staff Farmers Other Stakeholders 3. Organizing Farmer Participation: Lessons Learned 19 Structured Organizations Institutional Organizers Agency Interaction 4. Building Participation into the Project Cycle 27 Project Identification Preparation Appraisal Supervision Evaluation Participation Series Boxes Box 1: Cost Savings and Increased Performance through Participation 6 Box 2: The Philippines: Developing a Service Orientation through Participation 9 Box 3: Mexico: Rapid Change in a Crisis Situation 13 Box 4: How to Foster Participation 15 Box 5: Senegal: Creating Incentives for Farmer Participation 16 Box 6: Pakistan: Failure to Empower User Groups for Ongoing Participation 22 Box 7: The United States: Lessons about Scale and Sustainability of Farmer Management 25 Box 8: Nepal: Generating Demand from Farmers 28 Box 9: Nepal: Bank and Government Partnership for Participation 31 References 35 Acknowledgments This is one of a series of papers written as Policy (HRO), and from support from the inputs to the World Bank's Participation Swedish Intemational Development Agency Sourcebook. These papers were themselves (SIDA) and the German Gesellschaft fur produced in a participatory way. Topics were Technische Zusammenarbeit (GTZ). selected by a technical committee chaired by Bhuvan Bhatnagar; and the production of this The following people generously have shared paper was guided by a steering committee their time, expertise, and lessons learned in consisting of participation practitioners with developing participatory. irrigation projects: knowledge of the irrigation sector. Richard Ashok Subramanian, Bhuvan Bhatnagar, Reidinger convened the steering committee. Masood Ahmad, Michael Cernea, William Johnson, William Jones, Hyko Laeyendecker, This series builds on the work of a participation Joma Mohamadi, Van Tuu Nguyen, Keith learning group which was led over three years Oblitas, Stan Peabody, Herve Plusquellec, by David Beckman and Aubrey Williams. It Mikitaro Shobayashi, Jose Simas, Mark has benefitted from financial support from the Svendsen, and Hope Thavaraj. Gloria Davis World Bank's Vice Presidencies for Environ- and Ted Howard provi,ded editorial comment mentally Sustainable Development (ESD) and and assistance. Human Resources Development & Operations Executive Summary The irrigation sector provides a rich source of farmers provide closer supervision of staff than experience and lessons in user participation. distant agency supervisors; breakages are Participation by farmers in system design and reduced because farmers feel a greater sense of management helps to ensure sustainability of ownership. As a result of more timely water the system, reduces the public expenditure delivery and repairs, farmers' yields are higher. burden, and improves efficiency, equity and standards of service. Mobilizing support at all Reduced Government Expenditure levels and establishing the participatory One of the most noted effects (although this has process, however, involves costs; it also nothing to do with farmers motives for partici- demands knowledge of the incentives facing pation) is the reduction in government staff each group of stakeholders, and of the essen- and expenditure requirements, due to farmer tial elements in building effective users' management and contributions of cash, labor organizations. and materials. Farmers' associations have proved more effective collectors of user fees Benefits than government agencies. It is not unusual for farmers to be willing to pay more than the Attempts to increase user participation have original user rates after transfer of the system been spurred by the poor performance-in to their control. However, increased collection terms of efficiency, equity, cost recovery and of fees does not motivate farmer participation. accountability-of many large scale irrigation Participation must also result in direct benefits systems managed by government agencies. to participants. Greater participation by farmers, through water users' associations, has helped to overcome Sustainability many of these problems and produced substan- Sustdingbirity tial benefits. Building irrigation systems which are wanted, supported and owned by users themselves Improved Performnance of Systems provides the best assurance of sustainability. Physical and fiscal sustainability of the irriga- The overriding reason for increasing partici- tion system beyond the project is enhanced pation is to develop better projects. Clear when operation and maintenance costs are met gains in efficiency and in the standard of from user fees rather than high levels of service are achieved when design and man- government subsidy. agement of the irrigation system are trans- ferred to farmers. Design of the system Equity benefits from local knowledge. Farmers have a direct incentive, and the means, to minimize More equitable organizational arrangements a drec inentve,andthemeas, o mnimze and water deliver have been noted when costs as well as improve the service: users' andcwato ry he are foted A associations can reduce labor costs by paying participatory approaches are followed. A lower wages than government agencies; local contributing factor is the socioeconomic status Participation Series Participation in Irrigation of the leadership, which tends to be closer to policy-makers and agency staff as well as that of the ordinary member, involving more farmers and other water users is essential for tenants and small farmers than in non-partici- successful participatory projects and involves patory systems. paying close attention to the incentives relevant to each group. The greatest receptivity to Spillover Effects participation is often found in crisis situations, when management problems or revenue drains The transformation of water users from benefi- are most apparent. ciaries to partners in irrigation development In building the confidence of policymakers and can have a widespread impact, as farmers senior agency staff, pilot projects have been become trained and organized. It can increase used effectively to demonstrate the capacity for local capacity to coordinate input supplies, for farmer management, potential improvement in example, and to deal with other government system performance, potential saving in agencies involved in rural development. government expenditure and improvement in cost recovery rates. Building alliances with Costs and Risks supportive individuals in government has also been effective and has been facilitated by Efforts to introduce participation are not participatory economic and sector work, by without costs for mobilizing field staff, training enabling task managers to spend several years and organizing farmers and carrying out working in a country, and supporting them socioeconomic research. Nevertheless, these with good social analysis. additional costs are usually offset by subse- quent savings in construction costs and higher Project implementation rests ultimately with loan repayment rates. agency staff. Internalizing support for partici- pation within irrigation agencies often involves A bigger problem can be the additional time structural changes, to link agency budgets needed to establish a participatory approach firmly to farmer contributions instead of and get the project off the ground, especially in government allocations, and to promote a more the absence of existing local democratic institu- service oriented approach. Since agency staff tions. Developing farmer organizations is often typically come from engineering backgrounds a slow process, less under the project's control and are not oriented toward dealing with than constructing dams or delivery structures. farmers, incentives for them to support farmer Once the participatory approach has been participation need to be backed up by training established, however, it is not unusual for programs. Study tours to farmer managed participation actually to reduce the implemen- irrigation districts can be particularly useful, tation period. The typical causes of delay in not only for their demonstration effect but also implementing non-participatory projects- in raising the prestige of participation, expos- difficulties in negotiating rights of way, and ing staff to new possibilities and creating a obstruction by farmers or local politicians-are bond amongst participants. eliminated by effective participatory processes. The strongest opposition to farmer participa- Conditions for Success tion is often encountered at the field technical level, especially where civil service unions are Mobilizing Support among Policymakers strong. When field staff perceive the proposed and Agency Staff changes as a threat to their jobs and livelihood, these vested interests can retard or even User participation changes but does not elimi- sabotage participatory projects. Clear directives nate the role of government agencies in irriga- are needed from policymakers, supported by tion development. Building support from performance measures linked to bonuses and 2 Environment Department Papers Executive Summary promotions, to encourage greater accountabil- capacity developed, when they are irivolved ity to the farmers. The new ethos can only from the beginning in decisions on system develop gradually. Sudden cuts into the status design, and when their organization has full quo should be avoided and the composition of ownership and management control of the staff allowed to change gradually. system. It is essential, for example, that special- ized staff be selected by and accountable to the Building Effective Farmers' Organizations farmers' organization, even if they have been Teams of trained specialists acting as commu- trained by government agencies. nity organizers have proved to be the most successful catalysts in participatory irrigation To be successful, farmer organizations must projects. Wherever possible, existing organiza- interact constructively with government agen- tional capacity should be built upon. In cases of cies and technical experts. This relationship very hierarchical social structure and inequi- works best when uniform rules are established, table distribution of assets, it may be unrealistic and supported by government regulation, for to expect fully democratic local organizations. the turnover of responsibility throughout the To control vested interests, the varying incen- project. Building the necessary organizational tives of different categories of farmers should capacity for this turnover involves training be identified and accounted, for in project farmers for a variety of new functions, from design (for example, in defining water rights), basic literacy, accounting, how to hold meetings, along with the resulting problems of achieving how to deal with agencies, to legal regulations, collective action. and even computer applications, as well as water management and operation of equipment. Appropriate incentives are needed if farmers are actively to support the users' associations Fundamental in meeting all these conditions, a which are essential channels for participation, strong and transparent legal framework for the and to assume the additional costs in time, organization is needed from the outset, provid- materials and fees. The most important of these ing farmers with rights and beriefits as well as incentives are improved irrigation services, and duties and responsibilities. This framework a voice in management decisions through a should also be flexible enough to allow farmers users' organization which is fully accountable to evolve their own organizational structure, to its members. The support of farmers is most and to permit the organization's responsibilities likely to be sustained, and organizational to grow in line with its capacity. Participation Series 3 1. Benefits and Costs of Participation Benefits of Participation benefits of participatory irrigation operations. One of the most noted effects (although this has Types of Benefits nothing to do with farmers' motives for partici- The overriding reason for increasing participa- pation) is the reduction in govemment expen- tion in irrigation is to develop better projects. diture due to farmer management and contri- Popular participation is a way to increase the butions. For example, in Nepal, contributions proplab piipabion . i a way jecto ireae wanthei by Farmers' Irrigation Associations saved 15 proabilitya ,o buildin poet paaeo a percent of capital costs on projects. In Paki- ways that people can and wil manage. Sustainability of the irrigation system beyond stan, more than 17,000 WUAs were organized the project is enhanced if payment for opera- under the On-Farm Water Management tion and maintenance (O&M) expenditures is Projects and they agreed to contribute labor, collected from the users rather than being materials, and cash worth 10 to 25 percent of dependent on high levels of government the costs of watercourse lining. Recovery rates subsidy which cannot be provided over the were 90 percent in the first year, and more than long run. Sustainability also is enhanced if 70 percent the second year, though sustaining farmers and stakeholders participate in the such resource mobilization proved problematic project and see it as their own, rather than as (Byrnes, 1992:52). In the Philippines, efforts to somecthng exteitasy imposed. strengthen farmers' organizations and turn irrigation systems over to farmer management resulted in reductions of staff and other costs in Participation of users in managing antd maintaining the Natonal Irrigaton Authority, in higher usater facilities and .operahons brings many farmer equity contributions, and in enhanced beiefits. Participation in planning. operating, and fee collection. As a result, when subsidies to maintaining irrigation... increases the likelihood the Authority were eliminated, the agency was that these will be uell maintained and contribute to able to meet full operating expenses, including communiity cohes-ion and empouwment in ways salaries (Svendsen, 1992). that can spread to other development activities.... Depending on the social context and local condi- Reductions in costs to government have tions, snch participation can progressively increase . . . in int;ensity over the project cydefrom consultation significant advantages, and may be explicit at the design stage to actual O&M of some parts of objectives of Bank-assisted operations. How- the system. ever, they do not tell the full story, particularly World Bank Water Resources from the standpoint of irrigation performance Management Policy Paper and social costs and benefits. If government (1993: 55) expenditures are reduced by eliminating essential functions which are not taken up by others, then the performance of irrigation Evidence of Benefits systems will deteriorate and the value of Many of the benefits of participation are quite investments will be lost. Altematively, f real but difficult to quantify. However, there government expenditure is replaced by farm- are a number of studies that document the Participation Series 5 Participation in Irrigation Box 1 Cost Savings and Increased Performance through Participation The Philippines Both cost savings and efficiency related to participation have been documented in the Philippines. Svendsen (1992) shows that, in five major irrigation systems, equity of water deliveries increased, particu- larly in the dry season, so that available water was spread over more of the service area while yields increased. Bagadion and Korten (1991: 90) report that mean equity contributions on participatory systems were 357 pesos/ha compared to 54 pesos/ha on non-participatory systems. A study by NIACONSULT (1993a) found that, in 1991, National Irrigation Systems which adopted farmer participation had signifi- cantly higher collection efficiencies for irrigation service fees (74 percent vs. 45 percent for non-participa- tory systems); lower recurrent maintenance costs (1.8 vs. 4.6 pesos/ha); lower personnel costs (260 vs. 463 pesos/ha); higher dry season benefitted area (77 percent vs. 65 percent); and higher dry season rice yields (4650 vs. 4150 kilograms/ha). Taking costs and labor contributions into account, farmers' net income per month increased from 764 to 1,149 pesos after participation was introduced on three systems. These gains were most dramatic for tail end farmers who saw major improvements in the equity of water deliveries with participation. Senegal Further evidence of the effectiveness of farmer participation comes from Senegal. Before devolution of responsibilities from the government agency to farmer organizations, farmers paid little if anything toward O&M, and the irrigation agency (SAED) conducted maintenance and paid for electricity on an irregular basis. This created unreliable irrigation services, system breakdowns, and overpumping that led to salinization. - After taking over the systems, farmers began to pay. two to four times the former irrigation rates to cover full O&M, including electricity costs and a fund for replacement of equipment. Increases in performance of water deliveries resulted. At the same time, electricity costs were reduced by half through more careful monitoring of pumping because farmers had both the incentive and the ability to control the amount of overpumping (Nguyen, personal communication). costs or performance of the system, there may there is some question of underfunding of be no impact on net social benefits. However, O&M, the system has performed adequately gains in net social benefits are possible, when for more than twenty years since turnover. participation leads to improved output of the system, or when farmers deliver the same level Another source of benefits is that local farmers of service at lower cost than government can provide greater supervision of staff and agencies. contractors than distant agency supervisors. This was seen as a major reason for the reduc- Both cost savings and performance increases tion in pumping costs in Senegal where irriga- have been documented when system manage- tion agency employees often turned the pumps ment is transferred to farmers. Lower costs on and left for other activities, while the farmers' derive from the farmers' ability to hire labor at pump operators would be fired for not staying lower salaries, on a temporary basis, or without with the pumps. In the Philippines, after farmer the full benefits packages often required for associations agreed to pay back a portion of civil service employees. For example, construction costs, they monitored the use of Svendsen and Vermillion (1994) show that after materials and contractor activities much more the Columbia River Basin Project in the United closely, and even contracted some activities States was transferred from the Bureau of themselves to save money. In participatory Reclamation to farmer management, real per systems, farmers have a sense of ownership of acre irrigation assessments were 78 percent of facilities and are therefore likely to be more the level under Bureau management, in part careful about not damaging the structures, thus because the farmers' associations sought other incurring lower O&M costs (NIACONSULT, sources of revenue to meet expenses. Although 1993a; Merrey and Murray-Rust, 1991). 6 Environment Department Papers Benefits and Costs Other Benefits activities, such as education, health, or market- In addition to the quantifiable gains from ing. In Nepal, some Farmers' Irrigation Asso- farmer participation, numerous less tangible ciations for tubewells have taken steps to advantages have also been identified. These provide clean drinking water for village include the use of local knowledge to improve schools. It is clear that these associations have system design ando oeration. According to given the people a feeling of empowerment system design2, and opertion. Accordng to with respect to the government. The empower- Kortncipatory oneproaches inkthey Phlesones of ment of water users, and their transformation participatory approaches in the Philippines is from beneficiaries to partners in irrigation that "involving farmers in the planning, design, development can have a widespread impact, and construction of irrigation system promotes farmer satisfaction with the physical facilities including greater accountability and develop- and is a useful way to strengthen irrigation ment of a service orientation within govern- organizations, which then can become the ment agencies. managers of the new or improved systems." Costs of Participation More equitable organizational arrangements, as well as more equitable water delivery, have Resource Requirements been noted when participatory approaches are Serious, effective commitment to increasing followed. In the Philippines, organizational participation in irrigation operations is not leadership in participatory systems included without cost, both for the projects and for the more tenants and small farmers, whereas in Bank. Resources need to be made available for non-participatory systems the leaders tended to training and organizing. It can require millons be wealthier. According to Bagadion and of dollars to mobilize the field staff of institu- Korten (1991:95) "these differences indicated tional organizers, train farmers, agency staff, that in the associations of participatory sys- and consultants to design appropriate partici- tems, the socioeconomic status of the leader- and conduce researchci- ship was closer to that of the ordinary member patory approaches, and conduct research to than was the case for non-participatory sys- document learning processes. Extra vehicles tems." This may be a contributory factor to the may be needed to ensure that staff can meet more equitable water distribution in participa- with farmers on a regular basis. However, tory systems noted above. these costs are a relatively small fraction of total lending in an irrigation project (see Table Spillover Effects 1). In the Philippines, institutional develop- Participatory irrigation operations can have ment costs for preparing socioeconomic pro- positive spin-off effects into other sectors. As files, fielding and supervising community farmers become trained and organized, they organizers, and providing training to organiza- may assume responsibility for other activities tions, totalled $25 per ha, approximately 3 such as coordination of input supplies. The percent of construction costs. Savings in exposure to external agencies in the project construction costs on participatory projects often increases local capacity to deal with made up 90 percent of this institutional outlay; agencies involved in other rural development equity contributions, plus higher loan repay- Table 1 Sample of Institutional Development Costs in Recent Participatory Projects Year Country Project US$ million % Total Project 1991 Mexico Irrigation and Drainage Sector 24.0 1.9 1994 Mexico On-Farm and Minor Irrigation Networks Improvement 37.0 6.0 1994 India Tamil Nadu Water Resources Consolidation (estimated) 1.6 0.4 Participation Series 7 Participation in Irrigation ment rates, provided a positive rate of return Expertise on institutional investments in participatory Preparation, appraisal, and supervision mis- projects from the perspective of the National sions must include expertise on water users Irrigation Authority (Bagadion and Korten, associations (WUAs) and participation. Funds 1991). for such expertise may be difficult to obtain, Time Requirements especially for the critical up-front costs in project preparation (though potential sources A resource constraint more serious than are discussed below). These costs can, how- funding is the time required to establish a ever, pay off with better project implementa- participatory approach. If the country is not tion and performance. For example, the Nepal familiar with or receptive to user participation, Irrigation Line of Credit obtained $2.5 million the first time cost is the lead time TMs need to in Technical Assistance from the United gain support for the approach. This is made Nations Development Programme (UNDP). particularly difficult when performance is This infusion of expertise enabled the project to judged on how rapidly and smoothly a project show substantial results within two years proceeds from identification through appraisal despite major economic and political upheav- to signing. Successful TMs who have increased als. the level of participation in their operations have learned to exhibit a great deal of patience, Costs of Non-Participation persistence, and personal commitment to this There are also significant costs to excluding approach. farmer participation in operations. In the 1975 Philippines Tarlac Irrigation Project, for ex- Within the project's life, it also takes time for Plisting communalrr rgation systems paricpatonto ak hod,spra,adso arnple, existing communal irrigation systems participation to take hold, spread, and show already irrigated more than 2,600 ha within the effects. Developing farmer organizations is areaot new prethe proec ma no often a slower process, less under the project's preovisons friolvin Thes pexting co control than constructing dams or delivery provisions for civolng these exasteg commu- strucures To eal ith tis isue,the amil nal irrigation societies. A few years later, the strucWtures Tesourdealswit thisoissu th ojectmi project completion report found that because of Nadu Water Resources Consolidation Project the neglect of their societies, farmers within the planned for a seven-year project rhe tn a communal irrigation systems had refused to mcor typic calcfivye pcsWe in mos join the project. This led to a shortfall of almost economic calculations "time is money," the 10 percent of the area the Tarlac project in- increased emphasis on sustainability of irriga- tend o oe asea to arlo f potential tion projects has shown the value of a longer- benefits. term approach. Indeed, once the participatory approach has been established, participation Even when farmers are willing to accept an actually may reduce the implementation irrigation project, they are much less likely to period. In the Philippines, for example, partici- pay fees to cover full O&M if they have not patory projects were extended an average of been involved in planning the project, are not two years, compared to five years in non- satisfied with service, or do not have some participatory projects. The latter were attrib- understanding of how project and O&M funds uted to "many delays in project construction are spent. Thus, non-participatory projects due to interference of farmers who were not usually have difficulty in meeting cost recovery informed or consulted on what was con- objectives. This can lead to subsidies for structed, difficulty in the right-of-way negotia- irrigation, which place a strain on government tions, and the meddling of local politicians. revenues, and underfunding of O&M, which is The problems were almost nil in projects with recognized as one of the most widespread participatory elements." (NIACONSULT, causes of poor irrigafion system performance. 1994b:14) 8 Environment Department Papers Benefits and Costs Box 2 The Philippines: Developing a Service Orientation through Participation The first and best documented nationwide program to build participation in as a comerstone of irrigation policy occurred in the Philippines. Beginning in 1976, a Ford Foundation supported pilot project worked with the National Irrigation Authority (NIA) to formally tum over responsibility, ownership, and manage- ment of small-scale communal irrigation systems to organized farmer associations. This approach was expanded to cover all communal systems in 1980, and even extended to large-scale national irrigation systems. Among other factors, legislation passed in 1974 to make NIA financially autonomous was crucial to the authority's active support for farmer organizations. As a financially independent agency, NIA's subsidies were phased out, and all expenditures, including staff salaries, had to be met from irrigation service fees. This created powerful incentives for the agency to devolve recurrent O&M to farmers and increase collec- tion of irrigation fees. The latter, in tum, required improving irrigation service so that farmers would be willing to pay. Korten and Siy (1989) point out that although the concepts behind farmer participation were simple, establishing them in a large technical bureaucracy involved a major effort. The primary factor in this success was the decision on the part of policymakers to support the participatory approach fully and make the institutional and legal changes necessary to apply it. This decision was supported by workshops, training programs, and materials dissemination to develop widespread understanding and support at the policy level for the lessons that were emerging from these pilot activities. The gradual "leam by doing" approach to methodology development and the movement of this methodology by NIA toward national acceptance are two features of its strategy. Given modifications for differences in culture and terrain, many aspects of the NIA strategy have been borrowed by other countries, including: * the requirement of legal recognition of user groups prior to their active collaboration with government; * the use of in-house community organizers; * performance measures for irrigation personnel which encourage greater accountability to the farmers they served; * increased participation of farmers in key decisions and in up-front planning and development of the physical systems; * development of cost recovery mechanisms which made farmers more responsible and instilled a sense of collective ownership of systems; * development of budget systems which can be adjusted to be responsive to clients; and * development of mechanisms for systems which allow the gradual refinement of their methods and procedures. Through these changes, NIA has evolved from an agency primarily concemed with construction to one committed to developing farmer irrigation associations and supporting their management capacities once projects are completed. NIA's success is an example of what can be accomplished when the government has the will to empower farmers and carries out policies accordingly. Participation Series 9 Participation in Irrigation Limits to Participation examples of farmer managed systems, each covering thousands of hectares in Colombia, Trade-Offs Argentina, Mexico, and even Nepal, which demonstrate that farmers can hire their own Participation is not a panacea. Nor are participat- engineering expertise and operate large areas ing farmers selfless. On occasion, there are themselves. In Nepal, traditional farmer man- conflicts and trade-offs between increasing aged schemes frequently cover several thousand participation and other project objectives. Build- hectares, without government involvement in ing active participation can interfere with timely construction, maintenance or management. In execution of projects (although it also can have Mexico, for example, economies of scale have the opposite effect), and conflict with project even been found that make farmer managed components such as rotational water supply (for irrigation districts more viable financially if they example, in the Haryana m project in India). cover more than 5,000 ha. The size now consid- ered ideal in Mexico is 8,000 ha. Similarly, the Participation may also be at odds with equity merger of traditional irrigation associations in objectives if some groups have more influence Mendoza, Argentina, into organizations covering than others. Issues arise because powerful 5,000 to 10,000 ha increased their financial farmers are likely to have a more influential voice viability. than other participants within irrigation associa- tions, and because agencies and other organiza- In these cases, the government agency continues tions are often more effective if they can draw to operate the dams and regulate river flows, and upon the contacts and resources of the wealthier some irrigation systems have not been found members. While inequities cannot be eliminated, suitable for transfer because of technical, eco- efforts can be made to structure organizations to nomic, or political problems. Establishing and include all users and build transparency within enforcing water rights, constructing and operat. WUAs, especially in resource management. ing major facilities and river basin systems, and Defining organizational membership to indude mediating allocation among irrigation systems marginal groups such as resettled farmers, female and between irrigation and other sectoral water headed households, and tenant farmers can also uses remains a crucial role for governments. improve equity. Resource mobilization, particu- Finally, there will continue to be a key role for larly to pay for long-term capital investments, technical expertise and for government agencies may be more equitable if assessed based on land to ensure the water supply, particularly in the ownership, rather than cultivation of the land. main system of larger irrigation networks. In cases of very hierarchical social structure and Role of Government inequitable distribution of assets (for example, As earlier examples illustrate, user participation Sindh in Pakistan) it may be unrealistic to expect changes but does not eliminate the role of gover- fully equitable and democratic local organiza- ment in water resource management. Govem- tions. Therefore, the Bank and government need ment pohymakers and agen staff must to recognize their role inen controlling veste agec stfms to recognize their role in controlfng vested actively support the participatory process if it is interests and acting as advocates r . to take hold. Active user participation can be The differential incentives of different categories of farmers ("head"/"tail', small/large) should be perceived as a threat to some governments, and offamers("hea"/tin" plrogrm d n ( may be difficult or impossible to establish if there recognized and dealt with In program design (for are few democratic institutions. There also may instance, in defining water rights), along with the be important regional differences within the resulting problems of achieving collective action. country, as in Colombia or Mexico. Wel de- Scale signed operations seek to identify the appropriate balance between agencies and farmers, and to It is often claimed that farmers can manage small engage a wide ranging set of actors in developing schemes but not large ones. But there are many a participatory approach to irrigation. 10 Environment Department Papers 2. The Participants: Roles and Incentives The issue of whose participation is to be sought management for participation is critical to is central to all participatory operations. The provide resources for participatory project World Bank's policy on Water Resources development, negotiate with borrowers to Management emphasizes participation by ensure participation, and ensure continuity in water users, which in irrigation projects approach when TMs change. With the many generally means farmers. However, there are demands of project and sector work, it is other important stakeholders, such as govern- difficult for TMs to add one more priority ment policymakers, agency staff, other water activity. Management support is one incentive users, people adversely affected by projects, but there are others. Those who have placed and the Bank itself. To be successful in increas- emphasis on participation in their operations, ing farmer participation, projects need to whether to empower water users or develop ensure that these other actors are involved in more sustainable projects, have found that and supportive of participatory processes. The despite the fru,strations this is one of the most following section highlights the roles of each of satisfying areas of their work. these key actors and their incentives for in- creasing participation. Policymakers World Bank Task Managers Convincing Policymakers Whlespnin the water Experience has shown that the participatory While supprt for paricipationapproach will not work without a firm commit- resources sector has been growing and the ment.rom p y ke wthous pafricionu number of TMs actively supporting participa- in planning, operating, and maintaining irriga- tion in irrigation has increased since the 1970s, in works. opicymakr mustalso bercom- it is still far from universal. TMs opposed to tion works. Polcymakers must also be com- patiiato ar fre to igor i, anhs mitted to empowering farmers and to adopting participation arefrti not te policies, institutional changes, and a legal who support participation may not receive full framework to back them up. support from their management. This results in implementation strategies such as settling for S i s "priipto on pae"o .loiggvm Strong inital support from pohlcymakers (such "partecipation on paper" or allowng govern- as that found in Mexico at the Presidential ments to establish participation programslee)ncuasqikapiatoofnbig which are focused too narrowly to have much level) encourages quick application of enabBng benefit. Neither of these approaches has much policies and regulations. Bank support for sustainability once the project has ended. participation also has proven critical in legiti- emizing this approach among other donors, for example in irrigation projects in Nepal and the Successful participatory activities require a Philippines (NIACONSULT, 1993b). Unless the serious commitment of time and effort on the borrowes hAdOprior exprinc with part~~~~~ ~ ofTsadsno.Bn aaeeti borrower has had prior experience with parto s and seni Bank anagmers participation, in many cases support will be order tois ishw governnt staf and fe only lukewarm. In such cases, a pilot program that this is an important issue that will be followed up. The support of higher levels of can be used to build confidence among policymakers. The Philippines experience Participation Series 11 Participation in Irrigation (Korten and Siy, 1989) shows that proper to cover O&M, necessary maintenance is not education through seminars, frequent interim carried out, and systems deteriorate. Crises in reports to policymakers, and pilot programs all debt repayment add to external pressures in can be used to build support. such situations. Inadequate agency manage- ment also creates pressures for change from Strategies to convince policymakers to support farmers. participatory approaches can come from cultivating a wide range of contacts and by Rehabilitation projects or irrigation sector loans building alliances among supportive which require some form of cost recovery have policymakers. Political connections can pay off provided valuable opportunities to build a if appropriate questions are asked in parlia- serious commitment to participation in irriga- ment or statements are included in speeches. tion. Raising irrigation fees high enough to In Turkey, the executive summary of a Bank meet full O&M costs (usually 2 percent of report calling for greater participation was . . translated~~ ~ int Tukshad,aeav ,abet original infrastructure costs per year, in real paanslatdintoTrliameant madeavailableterms) is often politically unpopular, and the all members of parhament. higher fees may be impossible to collect unless Invitations for the Bank to comment on na- irrigation service is improved dramatically. tional or provincial water policies have, in Devolving responsibility for O&M to farmers, some cases, led to incorporation of participa- at least of tertiary structures, is also a means of tory ideas. Higher-level discussions between reducing government expenditures and in- Bank management and political leaders such as creasing cost recovery rates and the quality of ministers also can be influential in promoting irrigation service. participation. Identifying and working with young "rising stars" can pay off as they move Policymakers in finance departments are often into positions of greater power. Participatory more conscious of the unsustainable revenue economic and sector work can be used to drain of O&M subsidies than those in irrigation identify existing patterns of participation to departments who generally have little if any build upon, and can lead to participatory role in collecting fees and whose budgets are projects. Many of these strategies require not affected by levels of cost recovery. Build- detailed knowledge of the country and people ing support for participatory approaches in government. This is facilitated when TMs among finance officials can be valuable, but this are able to spend several years working in a is sthl "external" pressure from the standpoint country and are supported by good social of those in irrigation departments who will analysis. need to implement the policies. Structural changes to curtail subsidies and link irrigation Crisis Opportunities budgets to fee collection (through financially Many of the successes with participatory autonomous irrigation agencies, as was done in irrigation operations have arisen out of crisis the Philippines) are often necessary for irriga- situations, in which "business as usual" was tion agencies to internalize support for partici- untenable. During a period of expansion in pation (Small and Carruthers, 1991). When irrigation, the availability of project funds and agency budgets are linked to farmer contribu- attention given to construction are likely to be tions rather than government allocations, the great, and deficiencies in performance may not agency is more likely to develop a service be apparent. But as funding for new construc- orientation and value farmer participation. tion becomes harder to obtain, government Without such structural changes in the imple- subsidies for recurrent O&M may also become menting agencies, participation becomes an more difficult to sustain. Where the level and extra responsibility, which will not receive collection rates of water charges are inadequate attention beyond project completion, if at all. 12 Environment Department Papers Roles and Incentives Box 3 Mexico: Rapid Change in a Crisis Situation Mexico experienced rapid and widespread incorporation of user participation in the irrigation sector. The objective was to make the national irrigation system financially self-sufficient as well as to obtain full cost recovery over time for major works already constructed. The cornerstone of this policy was the transfer of irrigation management to Water User Organizations. Crisis situations in irrigation system financing and management provided the impetus for sweeping changes. By the end of the 1980s, an estimated 1.5 million ha (out of 6.1 million) of irrigated land went out of irrigated production because of lack of funding for completion of infrastructure and O&M. Bank management was influential in pointing out the need and direction for change, and the Bank provided a loan (co-financed by the Inter-American Development Bank) for the Irrigation and Drainage Sector Project. The three pillars of this project were decentralization and transfer of irrigation districts to Water User Organizations, self-sufficiency in fee collection to cover full O&M costs, and efficiency in budget alloca- tion. The transfer has been done in two stages. The first gives producers, organized in Water User Organiza- tions (covering 5,000 to 18,000 ha), responsibility for operation and maintenance of large lateral canals and drains. In the second, these organizations take responsibility for the main irrigation and drainage canals and the machinery and equipment required for O&M through the creation of an enterprise or Sociedad. Farmer groups are set up as organizations, rather than less formal associations, so that, under Mexican law, they can operate as legal entities and obtain loans. These organizations are meant to become finan- cially self-sufficient through collection of water charges. Each organization hires a professional team to carry out O&M, including a manager and a group of water masters (one for approximately 3,000 ha) and a chief of maintenance (all graduate engineers) as well as their support staff. The process of transferring management to farmers already has exceeded targets. Since 1991, thirty-three irrigation districts covering 2.3 million ha have been transferred, and an additional eleven are in process. O&M cost recovery rates have increased from 18 to 78 percent. Many irrigation districts have increased their fees 400 to 600 percent since farmers took over management. The main reason for the success in Mexico is the strong commitment by the central government. This translates not only into pressure on the National Water Commission to carry out the program (to the extent of firing powerful individuals who opposed the changes) but also into reforms of the Agrarian and Water Laws. These give holders more security and encourage them to make investments for long-term improve- ments in the field level delivery systems. They also separate land and water rights, and provide for water rights to users through their associations, thus creating an incentive to use water efficiently. To educate farmers to the changes and convince them to support the program, Mexico relied heavily on mass media campaigns prepared by communications specialists from FAO, along with universities and industry. These were followed by detailed training of the staff of farmer organizations in, among other subjects, computer applications and use of maintenance machinery. Districts in the best financial condi- tion were transferred first (after deferred maintenance was done) to ensure a successful start and build confidence. "Citizenship participation" was a theme, with transparent management and farmers assessing fees through their organizations. On the agency side, building a service concept was stressed so that the remaining functions would be carried out dependably. The Mexican example shows how rapidly change can be made when there is a clear plan and vigorous government support. It also shows the relative importance of creating the proper environment for partici- pation at the policy level. Participation Series 13 Participation in Irrigation Agency Staff Water Management projects originally gave responsibility for organizing WUAs for water- However supportive Bank staff and national course improvement to the Agriculture Depart- policymakers may be of farmer participation, ment because the Provincial Irrigation Depart- project implementation ultimately rests with ments were unwilling to become involved. agency staff. It is crucial to ensure that staff at However, subsequent efforts to expand the all levels work to strengthen farmer participa- involvement of WUAs into O&M of lower level tion. Moreover, because irrigation agencies are irrigation systems faced problems in coordina- likely to continue to have roles in water re- tion between the Irrigation and Agriculture source management which require interaction Departments. Efforts to ensure that the agen- with farmer associations, their ongoing support cies involved in implementing irrigation of participation should be secured beyond the projects take part in participatory operations life of any single project. In many cases, this are likely to pay off in the long run. requires bureaucratic reorientation toward a more service oriented approach within the Study Tours agency (Uphoff, Ramamurthy, and Steiner, Tours to areas with active participatory irriga- 1991). Securing this support requires structure tion projects have been found to be effective in of incentives for agency staff, training pro- convincing agency staff of the value and grams, and the personal involvement of TMs. feasibility of participatory approaches. In Turkey, for example, the World Bank sup- Identifying Allies ported study tours to Spain and Mexico for the Irrigation agency staff (who typically come from provincial governors and staff of the two engineering backgrounds favoring design, agencies involved in irrigation. In Ecuador, construction, and technical standards over study tours were arranged to farmer managed operation, maintenance, and adaptability to local irrigation districts in Colombia. Lower level conditions) often are disinterested initially in staff can be taken to visit farmer managed farmer participation. They may even see it as systems within the country to see how they "interference" in "proper" management of sys- operate. Farmer-to-farmer study tours also tems. Nevertheless, through wide ranging have been very successful in training and discussions within the agency, TMs can identify capacity building among WUAs in Nepal and. individuals who are receptive or supportive of elsewhere. participatory approaches. These champions of participation are important because they con- In addition to their demonstration effect, tours vince others and provide ongoing support for raise the prestige of participation, expose staff necessary changes between visits from Bank staff. to new ideas and possibilities, and create a bond of common experience among the partici- Receptivity to farmer participation is often pants. However, to be effective, study tours more prevalent in agriculture, rural develop- must be planned carefully and tailored to the ment, or on-farm development agencies than it experience of the participants. Selecting sites is in irrigation departments because the former with similar agro-ecological environments, are more accustomed to dealing with farmers. cropping patterns, irrigation technology, and Thus, building farmer organizations and cultural backgrounds will make it easier for the supporting other types of participatory activi- agency staff to identify with the experience. ties through such agencies may seem an attractive option, particularly when irrigation Supportfrom Field Staff agencies are unresponsive. The problem with The strongest opposition to participatory such an approach is that it does not build approaches is usually encountered at the field experience with participation within the technical level, from the heads of large irriga- irrigation agency. In Pakistan, the On-Farm tion districts to local overseers, ditchriders, and 14 Environment Department Papers Roles and Incentives gatemen. This is especially likely where civil service unions are strong. Initially, field staff Advantages for Field Staff may feel that participation makes the job more Rehabilitation or modernization of systems in difficult and decreases their authority. They preparation for transferring them to farmers may also believe that the transparency of may provide a more professionally satisfying participatory management will reduce their task for staff than the lower-level O&M work rents from a project. Lower level field staff, being turned over to farmers. In many cases, such as ditchriders, pump operators, and as field level staff gain experience with partici- gatemen, see farmer participation as a threat to patory methods, they come to appreciate relief their jobs and their livelihoods. Without clear from constant individual demands of farmers directives from supervisors and policymakers, as regular channels for communication develop vested interests at the field level can retard or (Merrey and Murray-Rust, 1991). Selecting even sabotage participatory projects. Direc- areas with successful participatory approaches tives should be supported by performance as training sites enhances the prestige of measures linked to bonuses, raises, and promo- working with farmers. Over time, the vocabu- tions which encourage irrigation personnel to lary and approach of those working within the have greater accountability to the farmers they system unconsciously begins to change. serve. Gradually a new ethos develops within the field level staff, one which will spread through- Opposition from agency staff is minimized if out the system both through participatory increased farmer participation does not cut development training programs and as person- heavily into the current status quo. For ex- nel are transferred to other projects. ample, if the absolute O&M budget is not diminished as a result of farmer participation Farners in O&M, or if money is shifted from activities for which farmers are taking responsibility to Building Confidence activities which are not yet being conducted. Farmers generally respond positively to any Opposition also can be avoided by trying to system which they feel gives them a reasonable ensure that staff made redundant by participa- assurance of getting water while taking their tion find employment elsewhere in the system, interests and values into account. Although the and slowly changing the composition of the main purpose of a participatory project is to staff to enable the agency to assume a new establish such a system, the willingness of service role. farmers to participate should not be taken for granted. A lack of confidence in the govern- Box 4 ment and a fear that they will be handed more How to Foster Participation than they can manage physically or financially often makes farmers reluctant partners at first. 1. Help people reflect on their own conditions. Confidence grows once a system is developed 2. Speak their language. that ensures their direct involvement and some 3. Live (or at least spend time) with them. degree of control while providing benefits that 4. Take their interests and values into account. allow tasks to be carried out at reasonable cost. 5. Respect them as individuals. Special efforts to gain the support of commu- 6. Find ways to get each individual to have a nity leaders can contribute to success, but these stake. should not.be at the expense of more wide- 7. Train them. spread participation from all farmers. 8. Create appropriate reward structures. 9. Pay attention to detail. 10. Take adequate time. Danger of Exploitive Approaches 11. Do your homework. Many participatory program failures grow out Thomas Dichter (1992) of a government's view of farmers as a re- source. In such an environment, farmers' Participation Series 15 Participation in Irrigation associations are viewed as providers of labor for so organizations are formed which are quite bene- construction and maintenance or for cash support ficial to governument but provide no long-term which can be tapped as necessary to supplement benefits for the farmers. These organizations are tightening budgets. What the association looks - not sustainable and generally die soon after the like from the farmers' point of view is never asked, short-term resources of a project disappear. Box 5 Senegal: Creating Incentives for Farmer Participation Irrigation systems in Senegal were plagued with problems of poor. financing and O&M in SAED (the agency providing irrigation services), resulting in poor agricultural production and frequent need for- system rehabilitation. As the government began to withdraw from provision of many agricultural ser- vices, transfer of irrigation system management to farmers was included in SAED's terms of reference. However, early efforts to establish village units showed little success because of unclear plans, timetables, and provisions for the transfer, and a lack of control or incentives for the farmers. Economic Interest Groups set up as legal entities empowered to obtain credit, purchase inputs, and provide services were more successful but had little role in irrigation. Recognizing these problems, the 1990 Bank-assisted Irrigation IV Project required total transfer of O&M to farmers as a precondition for financing rehabilitation works. Getting farmers to agree to take over these systemns required considerable negotiation, particularly because they perceived existing irrigation fees as too high, let alone the additibnal costs and responsibilities of covering full O&M. Increasing the control of farmers over irrigation services was the key factor which convinced them to agree to assume responsibility for system management. Farmers demanded the right to hire their own staff, choosing SAED operators only if they had performed well, and even then reducing their salaries from the full civil service package. The formal agreement with SAED specified that farmers would be responsible for full O&M, cost recovery, and compliance with cropping plans. SAED would transfer ownership of rehabilitated, operating systems with their infrastructure, and thereafter could enter the schemes only with farmers' permission. Manuals were prepared for each system which specified the technical requirements for O&M, require- ments for administrative and financial management, and agroeconomic requirements of the cropping plan and agricultural intensification. These helped farmers to understand and become more confident of their abilities to manage the systems. In addition, training in basic literacy as well as technical and financial skills for organizational leaders was incorporated, with farmers willing to pay at least part of the costs. Although the long-term sustainability of the project is difficult to assess, the accomplishments to date are impressive. Before the transfer, SAED assessed fees covered only 17 to 21 percent of maintenance and replacement costs, and collected less than a quarter of that (Nguyen, 1994). As a result, maintenance suffered, electricity often was not paid for, and system reliability was very poor. After the transfer, farmers paid fees four times as high, covering not only O&M (including full electricity) but a replacement fund for capital. The benefit to the farmers was greatly improved irrigation reliability. Moreover, because they were able to monitor the pump operators and had an incentive to save on energy costs, electricity requirements were reduced by half. Savings to the government were apparent, as SAED's personnel were reduced enormously when farmers took over functions. The government's support for withdrawal of the state from direct provision of services, and the poor performance of systems which made officials and farmers willing to try new approaches, were clear factors in the success of this management transfer. However, the project also was able to build on the institutional capital of Village Units, Economic Interest Groups, and other organizations developed during the 1980s, and on a civic tradition of citizens asserting their rights. 16 Environment Department Papers Roles and Incentives Costs to Farmers sible. In Morocco, for example, irrigation services When seen from the farmers' side, participatory provided by the agency generally have been very projects usually demand increasing contributions good, but the systems cannot supply water if of cash and time. The "transa'ction costs" for even a single support of the flume canals breaks. farmers of calling and attending meetings, taking While the agency frequently takes a week or over management functions, and negotiahing more to repair the supports, a farmers' organiza- over~~~~~~~ maaemn f.,.ins anheoitnton has found it can operate the necessary conflicts among themselves are often significant tio nthas nd itpcan ote s the n s y considerations in addition to the increased eqipmen a repair tespotawh a day. irrigation fees required. Farmers may be willing I v to assume these costs, particularly if service from organization, which now may consider contracts the agency is very poor, buit this may require a to do such maintenance for neighboring organiza- great deal of persuasion if existing irrigation tions (Plusquellec, 1994). In Mexico, timely service is minimally adequate. In the long run, to availabilty of funds for maintenance was a be successful, farmers have to perceive their problem under government management, but organizations as being purposeful and worth the has been improved under farmer management by effort and cost it takes to maintain them. Systems requiring fee payments from farmers in advance must be structured in ways that allow both the of the season. Farmers are able and willing to govtbernmenctuand the farm tohreceive aow goo pay in advance ibecause of credit availability and government and the farrner to receive a good asrane tha th ,ilgtwtr share of the benefits they require. The best a y systems. establish a fair and stable balance. System rehabilitation and upgrading of physical Incentives for Farmners facilities are perhaps the most common incentives for farmer organizations to participate in Bank- Just as incentives for agency staff to involve assisted activities (for example, Pakistan On- farmers (such as financial autonomy and appro- Farm Water Management). Rehabilitation or priate performance criteria) are critical to their upgrading fits easily with most project designs, participation, so too are incentives for farmers and is often sufficient to encourage farmers to essential if they are to participate in desired ways. pledge payments, agree to take over maintenance Farmers must see a return on their investment of functions, or even contribute toward improve- time, materials and fees, though that return may ments. The problem is that such incentives do be in the form of better irrigation facilities, not sustain farmer participation, particularly after services, or voice in system management. The the facilites have been improved and start particular incentives which will be valued most decaying again. One-shot incentives may be by farmers vary among locations and types of useful to generate interest, but sustainable farmers. The most appropriate set of incentives participation requires longer lasting incentives. for each project can be identified best by asking farmers which changes in their systems are most Reduction in costs of irrigation services is another important to them. potential incentive for farmers participation. Irrigation fees may be reduced for those farmers Better water supply is a strong incentive for who agree to take over certain tasks, particularly farmers. However, using this as the only incen- tertiary level maintenance. In practice, costs to tive is risky because the project may not deliver farmers often increase by up to 500 percent when improved water supply or be enough of an they take over functions, especially if proper improvement to satisfy farmers. In addition, the O&M is undertaken, because the activities no link between farmer participation and improved longer receive govermment subsidies. Although delivery may not be clear. such increases in farmers' fees occurred after transfer of O&M in both Mexico and Senegal Another fanner incentive is more rapid response collection rates were higher than under govern- to system breakdowns. If breakdowns are a ment management because the farmers them- critical problem, this may be easier to establish if selves had set the level of fees and controlled farmers, rather than a distant agency, are respon- their use. Participation Series 17 Participation in Irrigation Establishing appropriate incentives for pilot nesia. The Mexican campaign for system projects is particularly difficult. On the one tumover is built on existing concepts of citizen- hand, special subsidies or enhanced services to ship participation. the first groups to organize usually are not sustainable when the pilot is replicated and Other Stakeholders participation spreads over a large area. On the other hand, if farmers are asked to pay full While farmer participation is becoming more O&M and take over irrigation functions, and common in irrigation operations, participation they see other irrigated areas continue to of other stakeholders at the local level rarely is receive the same benefits without assuming the solicited. These include non-agricultural users extra costs, there is no incentive to participate. of water (particularly for domestic and live- Appropriate combinations of incentives which stock purposes) and those adversely affected can be replicated over the entire irrigation by the project (particularly those who are to be system or the sector must be identified. These resettled). Participation of these other stake- include improved performance and benefits, holders is called for increasingly as the Bank's cost sharing of investments and matching policies on water resources and resettlement grants, reductions in fees paid to the govern- are applied. The former advocates integrated ment, and other resources for active farmers' planning of water resources, including all organizations. sectors, while the latter pays attention to involving those adversely affected by projects Ownership of the irrigation system has been to ensure that their needs are met. found to be a powerful incentive for farmers to participate in its management and upkeep. Diversity of Interests People take care of what they own because they Involving other local stakeholders is compli- know the benefits or losses will come to them, cated othey oal areholdffferent not to an external agency. Thus, many projects cated because they often are from different (for example, in the Philippines, Indonesia, and social groups than the irrigated farmers. Nepal) have included formal "tunover" Women, including those from landless and ceremonies, in which farmer ownership and non-farm households, are the primary non- responsibility for managing the systems are farmer group affected by domestic water formally acknowledged. Whether or not there supply. Those with strong interests in water is a formal ceremony or certificate of owner- for livestock also may include women as well ship, farmers will not develop a sense of as members of ethnic groups other than those ownership over an irrigation system (or part of in the farming community. Households to be it) unless they have some degree of actual resettled frequently are from different ethnic control. This, in turn, requires that farmers groups or areas and may lack traditions of actively be involved in decisionmaking about irrigated agriculture. Irrigation agency staff their systems. Related to this is the incentive of have even less experience with such groups greater influence over water allocation. In than they do with irrigated farmers. To in- areas of water scarcity or competition between crease participation by these stakeholders may uses, farmers are often anxious to have a say in require work with other government agencies the amount of water they receive. In other or NGOs with more background on domestic contexts; greater influence over the timing of water supply, livestock services, or general water deliveries (within or between seasons) is social services, and more experience and important. contact with the local people most concemed. Stakeholder workshops and other structured Finally, cultural values can provide an induce- discussions should be more regularly included, ment for participation. Traditional pattems of but most Bank-assisted irrigation projects have cooperation have been used for collective had relatively little experience with such action in system O&M in Sri Lanka and Indo- approaches. 18 Environment Department Papers 3. Organizing Farmer Participation: Lessons Learned Experience has shown that participation tends zational capital and reduces the likelihood of to be more effective through organized groups success (Cemea and Meinzen-Dick, 1992). and that local organizations are crucial factors There may be situations in which existing in development efforts (Esman and Uphoff, organizations are inappropriate, either because 1984). Particularly in contexts with many they are not functioning adequately or are smallholders, organizations are vital intermedi- dominated by particular groups. However, aries between irrigation agencies and the even in these situations they should be exam- thousands of farmers in a system. Farmers ined to identify why they do not operate participate as individuals in privately owned effectively so that the new WUAs will avoid and operated irrigation systems, such as those mistakes. tubewells in India and Pakistan. However, when systems serve multiple farmers, organi- Purpose zations are needed to coordinate the many Organizations require purpose to be sustainable. users. Even in the case of private tubewells, While consultation with farmers' organizations the lack of a coordinating institution has in designing physical systems or irrigation allowed unreguated use resulting in ground- indsgngpycasstmorriain allowed unregulated use resulhng in ground- schedules can establish initial enthusiasm, water mining and other problems. Further- experience in Pakistan indicates that organiza- more, without such organizations to channel tions need continuing roles in system manage- participation and represent farmers, it becomes ment if participation is to continue. The pur- extremely difficult for officials to meet with all poses of these organizations must be established farmers, especially in a smallholder context. by the organizing staff through consultations Too often, the result is that only influenhal between farmers and technical staff. In develop- farmers can effectively participate. ing these objectives, attention must be paid to A number of key organiz'ational elements have the rights and the duties of the farmers, as well been idnumberiofiked insucce l farments partpave as to the issue of control over resources. How- bee indentified in successful farmer participa- ever, how an organization's activities are to be tion in irrigation. In the following discussion, carried out should be left to the farmers to these are grouped under the headings Struc- decide, with the organizer's help. Korten (1993: tured Organizations, Institutional Organizers, 2) states: "It is when irrigation organizations are and Interaction with Agencies. empowered to exercise management authority over such areas that local interest in, leadership Structured Organizations for, and capacity of such organizations flour- ishes." Organizations are a strategic resource for irrigation management; therefore, the act of Whether organizations for participation in organization is a capitalformation activity irrigation systems should focus on irrigation (Cemea, 1987). Wherever possible, existing activities alone or be multi-purpose in scope organizational capacity should be built upon. must be decided on a case-by-case basis. Failure to identify informal or indigenous Often, non-irrigation activities (such as control WUAs leads to costly disinvestment in organi- over fishing rights) are important sources of Participation Series 19 Participation in Irrigation financing for the organization or offer mem- associations much less control. Accountability bers important benefits (for example, coopera- is best when farmers hire, fire, pay salaries, and tive input supplies). Existing non-irrigation determine raises or bonuses for the staff. organizations may provide a basis for WUAs. However, assuming responsibility for too many Training non-irrigation activities can reduce the time Building organizational capacity requires and effort available for critical O&M. Perhaps training farmers for new functions. These the most appropriate rule on this issue is that include basic literacy, accounting, water farmers be allowed to take on additional inageme how to holmengs (in functions themselves without extra activities management, how to hold meetings (in funtions exthseralvs wposd (Cetrna activities Senegal), operation of equipment, and even being externay imposed (Cemea and computer applications (in Mexico). Learning Meinzen-Dick, 1992). how to deal with agencies and legal regulations is particularly important to incorporate into Specialized Staff training. In Senegal, for example, sample It is frequently assumed that farmers will, or contracts were provided to show organizations should, carry out activities such as routine how to get work from outside firms. Whether maintenance using cooperative labor. How- the training is provided by irrigation or other ever, it is often more efficient for the associa- government agencies, NGOs, or other farmers, tion to hire specialists who can reduce the it must be practical and tailored to the needs of transaction costs of notifying farmers of what the farmers and WUAs. needs to be done and by when, monitor who attends meetings, and settle confficts. These Legal Recognition of Farmner specialists may range from the common irriga- Organizations tors who clean channels and delver water in Creation of the legal structure should be part of South India, to pump operators in Senegal, to the formation process of farmer organizations, the engineers, accountants, and computer p roce of to oranizank opraor hire by fames asoiain in p .....articularly to enable them to operate bank Mpexico.s Firedfarmers' associations in C accounts, obtain loans, or bid on contracts for even hire lawyers. WUAs in Mauritania offer irrigation activities. The legal structure should farmers the option of doing manual work on be designed to allow as much flexibility as maintenance; those who choose to do so are g paid wages which they can use to offset their Model organizations established by outsiders, water charges. Professionalizing services is a hodel meaning and by designed, natual artof oganzatona devlopentand however well meaning and carefully designed, never take into account all of the farmers' own is particularly necessary as the management felt needs. As a consequence, they rarely work tasks become more complex and require well, nor do farmers identify with them as they specialed knowledge or equipment. do with organizations of their own design. In It is essential that these staff be selected by and Maharashtra, India, for example, the legal accountable to the farmers' association, even if framework for WUAs was based on the regula- they have been trained by goverunment agen- tion of formal Cooperative Societies and cies.yAs part ben irrigatn magoverment transfer bylaws which were too difficult for farmers to cies. As part of irrigation management hrans understand and restricted the activities societ- in Mexico and Senegal, farmers' organizations ie coud udertake, ee hntoeatvte were given the option of hiring the lower level would enhance the financial viabiois of the agency staff whose tasks the associations were organhans (ue l 1994; taking over. Those staff who had been respon- organzatons (Plusquellec, 1994; Moench, sive to farmers and done their jobs ween often 1994). Farmers should be allowed to determine were retained, but many others were replaced. the structure of their own organizations given were retain, but manyocthes were replced. agreed upon purposes. But flexibility should In~ ~~, Coomia fame asoiain wer require be retained in case either structure or purposes to retain agency staff within their irrigation neeto be modified as experience is accumu- districts, a requirement which gives these 20 Environment Department Papers Farmer Participation lated. Rules for resource mobilization and tion of water fees serves to improve the usage management, operations, and conflict resolu- and availability of water. Higher level organi- tion, along with the allocation of rights and zations also have taken over regulation of some responsibilities, must be specified clearly to the of the thornier issues of irrigation, including farmers (Ostrom, 1992; Tang, 1992). Written regulation of equity, participation in schedul- records of all decisions and actions, as well as ing decisions, settlement of disputes, regulation all agreements with the government staff, of livestock, use and maintenance of canal should be maintained by the members. roads, and specific maintenance activities at agreed upon levels. High expectations for the Membership Definition roles farmer organizations can play in system Experience has shown that the most successful management as they continue to develop are participatory irrigation organizations are those appropriate. based on hydro-geographic boundaries (rather than political distinctions) in which farmers are Institutional Organizers served by a common irrigation facility such as a turnout or night storage reservoir. Establishing Needfor Catalysts these boundaries may require reviewing land Organizations do not form spontaneously. records to identify members and enabling organi- Experience has shown that organization zations to receive their correct shares of revenues. building often requires outside individuals to Both the membership and leadership of participa- act as catalysts to initiate the process and link tory organizations should be lmited to water it to agency activities. In participatory projects, users, including explicit provision for involving teams of trained specialists acting as commu- small farmers, tenants, and marginal groups. nity or institutional organizers have proved to While not divorcing itself from local political and be the most successful catalysts. governmental leadership, participatory organiza- tions should be separate from political organiza- The organizational development staff are tions, with separately selected leadership. At the responsible for brokering agreements over the same time, an effort should be made to provide initial purposes of the farmer organization and local political leaders with information about for explaining the terms of the project to participatory irrigation and the role of user farmers before construction or rehabilitation groups, and with frequent updates about progress. begin. They must help farmers collect their contributions and complete whatever paper- Nested Enterprises work is necessary to gain project approval and In larger systems, the organization should be statutory or legal recognition. Should the handled in a graded fashion, beginning at the project not operate as agreed, they must act as lowest level (generally outlet or turnout groups), the farmers' communication channel until moving through intermediate organization higher level organizations are formed and can formation (tertiary and secondary committees) assume the role, or until the farmers are to higher level organizations (apex level com- sufficiently familiar with the system to do this mittees). The appropriate size of each group on their own. Organizers, in effect, act as low- will depend on average farm sizes, but the tech interactive communications media, process of structural design and determination broadcasting a series of focused messages to of purpose must come from the membership. individual farmers and groups. Increasing Responsibilities If farmers are to participate effectively in the As higher level farmer organizations have been planning and execution of physical works, it is put in charge of water fee collection. Rebates critical to have organizers and organizations in on the money collected are distributed accord- place before construction or rehabilitation ing to a formula; farmers may use their rebates begin. While the salaries, per diem, and for their own purposes. In some cases, collec- working expenses for institutional organizers Participation Series 21 Participation in Irrigation Box 6 Pakistan: Failure to Empower User Groups for Ongoing Participation Pakistan's irrigation program demonstrates what can happen when a government tries to use farmer associations to meet short-term goals without implementing policies which strengthen and empower them. Since independence, Pakistan has depended on large system irrigation. Although tradition and British law made farmers responsible for watercourse maintenance, there was, in fact, little cooperation among farmers for this task. As system deterioration cut water availability, farmers began depending more on tubewells to supplement and eventually replace much of the water they had been getting from the irriga- tion systems. By the late 1970s, policy began to shift from construction to rehabilitation of systems and to improvement of water management on existing projects down to the farm level. This shift was influenced by the United States Agency for International Development (USAID)/Pakistan On-Farm Water Management Pilot Program (1976-81) in which farmers supplied all unskilled labor for watercourse improvement while the government supplied only materials and technical assistance. Even the systems restored under this project quickly deteriorated, however, because farmers were not organized to continue post-construction O&M. A similar situation developed in the Punjab government's 1980 crash program for heavy cleaning of water- courses. The Bank became involved at the termination of the USAID sponsored effort through the On-Farm Water Management Program. The program worked through WUAs to mobilize labor and funds for improve- ment of watercourses, with watercourse lining serving as the major inducement to farmers. Socioeco- nomic criteria for selection of projects were developed. These included the empowerment of registered user groups and an agreement among farmers to improve and maintain their watercourse branches. The Bank required that each of the four provinces implementing the program establish a legal ordinance providing for the organization and registration of user groups at the outlet command (chak) level. The provincial Irrigation Departments showed little interest in working with farmer organizations, but an On-Farm Water Management Directorate was created in the Agriculture Department to provide training to user association leadership and field level assistance to help farmers organize. More than 17,000 user groups were registered. Although they performed well at the time of improvement, important questions have been raised about their sustained performance once the project was over (Byrnes, 1992). Since the user groups legally were organized and empowered only to support the improvement of the watercourses and provide ongoing maintenance, they had little purpose once the project was complete. While their agreement called for them to continue to maintain the watercourses, the farmers felt they were being used merely as a source of free labor. In their view, the benefit they received for maintaining the organizations and the watercourses was not worth the effort. The recent Irrigation Sector Review has highlighted increasing problems of the Irrigation Departments in managing canal irrigation systems and the unsustainable revenue drain caused by low cost recovery. The government now is discussing options for transferring management responsibility for O&M up to the distributary level, to federated farmer organizations. Although a pilot project for federating WUAs and expanding their roles may be incorporated into the project's next phase, the existing WUA structure does not provide much on which to build because these organizations have not had any ongoing role and most ceased to operate after watercourse lining. The Pakistan experience achieved considerable progress, if measured in terms of a one-shot contribution from farmers. At the same time, it demonstrates that, to ensure sustainable farmer organizations, a working agreement and its legal framework must provide farmers with rights and benefits as well as duties and responsibilities. Organizations also require continued purposes which the members value; those created by projects only to provide cash and labor inputs will not continue to exist once the addi- tional resources of the project have dried up. 22 Environment Department Papers Farmer Participation generally do not require foreign exchange, * the interests of engineers and farmers are experience in the Philippines has shown that it often in conflict, and the engineer could be is risky to assign these costs to national coun- viewed as too biased to approach problems terpart funding in the projects. Shortages of with an open mind. counterpart funds due either to cash flow problems or lack of commitment in govern- At the same time, the engineers' technical ment agencies can hold up advance organiza- understanding of irrigation systems and what tional work, and, in consequence, disbursement they can deliver is valuable for involving of loans for construction work. In the Philip- farmers constructively in the project. If engi- pines, this problem in the Communal Irrigation neering staff are to conduct this critical task, an Development Project I resulted in 46 percent of appropriate incentive structure and training the loan being canceled despite three exten- program must be developed to facilitate the sions. Resolution came in the follow-on loan local organization, as was done in Mexico. when $1.8 million for organizers' costs were included in the loan amount (NIACONSULT, Extension Workers 1993b). Another potential source of organizers is the agriculture department of the national or local Selecting Organizers government. The advantages of using these One of the most important pre-project judg- extension workers are several: they are part of ments is the choice of a group organizer for the regular government staff; they have strong project. Organizers are important since they agricultural technical skills; and they generally act as the main line of communication between have had close interaction with farmers. The the project and the farmers until interaction has disadvantage is tied to the difficulty of trying progressed far enough for more direct commu- to fund and coordinate activities between nication to take place regularly. Inducement to agriculture and irrigation departments. participate, explanations of project approaches, and arrangement of activities fall to the organi- Irrigation Department Contracts zation team. Generally the work is difficult, In Nepal, organizers were hired by the project the conditions are hard, and the hours are long. on a contract basis. This enabled them to Great care is needed to find a good source of collect higher salaries in recognition of their individuals to carry out this demanding task. long hours and difficult conditions. Gradually, Many sources have been tried, including the Nepalese govemment regularized the engineering staff, extension workers, irrigation organizers and absorbed them into the Irriga- department contracts, NGOs and community tion Department, but this put them into posi- members. tions of dual loyalties, having to support their Department on the one hand and the farmers Engineering Staff on the other. A satisfactory approach still has One option that may appear attractive at first not been developed. However, it is clear that glance is to assign organizing work to the such a "cadre" of organizers is not the full engineering staff and their assistants. How- solution, Department engineers, who have the ever, this approach is limited because: real power and control in the farmers' view, must be directly supportive, knowledgeable * engineering staff have very little spare time and involved. if they are doing their jobs properly; NGOs * technical staff are relatively expensive and . . it iS not cost-effective for them to spend the Hrn raiesfo nNOi nte itime netcessarytoefgetiveforthe to kn rer whel possible source of well educated and trained eme necessarry tot itionoalrganizing community organizers. Dichter (1992) sug- properly; and gests, however, that the Bank needs "to go Participation Series 23 Participation in Irrigation more slowly and do its homework carefully project has used several types of organizers or when it looks for institutional partners for endeavored to compare them. Contrasts between participation." He advises not to take an projects and organizers may be more an expres- NGO's reputation for granted. Many NGOs sion of differences in the quality of projects than have their own agendas or lack sufficient staff of suitability of the individuals involved in the or technical ability in the field. Some NGOs work. Other cultural factors may be at play (for only provide short-term solutions to the example, one social or educational group may be staffing problem, since they may be ephemeral more suitable to an area than another), something and cannot be counted on for long-term sup- which should be looked into at the time of port. appraisal. Experience has shown that a confident person with good leadership ability, communica- Community Members tions skills, ability to listen, and grasp of at least Taking organizers from the community can basic technical aspects of irrigation can carry out provide a long-term solution for support to this type of work satisfactorily with training. farmers' organizations. Local participants should However, like any other project or agency staff, be selected carefully and trained well (NGOs can organizers need appropriate incentives, particu- provide such short-term training). However, this larly to keep them accountable to the farmers. approach has two serious drawbacks: Mechanisms for this include having local com- munities evaluate the perforrnance of the orga- * the selection process may make it appear nizers and allowing local communities to hire the that the project is favoring some farmers organizers directly if they prove useful. A pitfall over others (and sometimes those appear- is that organizers may have an incentive to keep ances are based in reality); and farmers dependent on them rather than working themselves out of their jobs. * village level workers often do not have enough authority to warrant serious Mass Media attention from technical staff, particularly In some settings, such as in Mexico, the indi- in early phases of projects. The addition of vidual community organizer can be supple- a senior supervisor to the organization mented through the use of radio or television team, however, could provide the needed to provide farmers with information about authority to locally hired field staff. irrigation organizations. Even when mass media are available, it is still likely that the Education field worker would be needed to help catalyze The educational level of organizers in various organization of the user groups and other projects has ranged from farmers with minimal institutions, and to help farmers file proper literacy through secondary school graduates to documents. However, their roles as providers university graduates. Experience has shown of information can be diminished greatly if a that each level has its own advantages. The strong communications program is developed. secondary school graduates, for example, are seen by farmers as being closer to them because Value of Organizers they are more like their sons and daughters. The cost of providing institutional organizers is University graduates, on the other hand, are substantial, but it has proved to be low relative more confident, better speakers and better at to total investment costs and the benefits of process documentation. farmer participation in the performance and ongoing maintenance of projects. It is even Skills Mix lower compared to the cost of the potential The value of utilizing one group in preference to failure of projects when participation does not another is difficult to determine since no single materialize or cannot be sustained. 24 Environment Department Papers Farmer Participation Box 7 The United States: Lessons about Scale and Sustainability of Farmer Management The U.S. Bureau of Reclamation's transfer of irrigation system management to irrigation districts provides valuable lessons on the potential scale and sustainability of farmer management. Svendsen and Vermillion's (1994) study of the Columbia River Basin Project documents the experience and performance of a major system (230,000 ha) transferred to farmer management. Because the Bureau of Reclamation required repayment guarantees from farmers, irrigation districts were established in 1939 and contracts for water sales and payments were signed between the Districts and the Bureau during the construction of the system. The Bureau continued to operate the system until 1969, when, after complex negotiations with the farmers, system management responsibilities were transferred. Irrigation service was of high quality both before and after the transfer. (The first few years after transfer showed some evidence of technical efficiency loss, ascribable to a learning period for the new manage- ment. This should be kept in mind when judging performance within a few years of management trans- fer.) O&M expenditures held constant, in real terms, although farmers' fees to the districts averaged only 78 percent of the fees paid to the Bureau. The difference was met by supplementary sources of income developed by the Districts and some reduction of operating expenses. While performance levels have been maintained for over thirty years, there are some indications that maintenance has not kept pace with system aging. The lessons Svendsen and Vermillion draw from this experience for developing countries relate both to enabling conditions and to the process of transfer. Enabling conditions they identify include: * a clear and consistent policy mandating irrigation management transfer; * financial autonomy for the managing agency; * a strong legal basis for farmer organizations; * secure and well specified water rights; * investment to bring physical systems up to standard before transfer; * a professional accounting and auditing system for transparency in financial management; and * a "partnership culture" between the agency and farmer organizations. Features of the process of transfer which they identify as contributing to the success include: * involving farmers from the earliest stages; * not expecting full cost recovery of capital and O&M, particularly in initial stages; * empowering farmers to negotiate with the agency; * using agency and farmer contracts to specify rights and obligations of each side; * training agency personnel to improve management and human relations skills; and * continuing agency activities in which they have a comparative advantage. To construct or rehabilitate a project without commitment of time and resources from farmers forgoes a valuable opportunity to build a participation partnership with those people most directly interested in the success of the project. To proceed with physical works in advance of farmer involvement may appear to keep projects proceeding smoothly, but it is likely to yield less sustainable projects over time. Participation Series 25 Participation in Irrigation Agency Interaction ground and an area of mutual interest in which the goals of each group can be established and To be successful, farmer organizations must activities can be carried out jointly. These interact with government agencies at some patterns serve to forward group activities and level. An open, constructive relationship on to mark the attainment of certain goals, permit- both sides is a hallmark of successful ting groups to build confidence in each other participatory programs, but it often requires while minimizing the anxiety caused by prior considerable time and effort to build. interactions. Stages of Interaction Joint activities, such as tertiary design or maintenance assessments, should be carried The organizational plan should allowsfor y out at agreed upon times with representatives farmer organizath to shoulder repnibilt from both sides utilzing set activities already already achieved. Technical training and established as part of the project design. A alpereadyon achieved. Technica tranin nd tgood example of a joint activity is the "walk supervision should be provided to the thouh inwihrpeetaie.fbt h organization and its leadership until the project farmers an the tenaservices phy go is satisfied that the farmers have the capacity to farough an the toeth ervto deer ine carry out tasks. Turnover of responsibilty ethrough a system together to determine should be the result of mutual satisfaction and c edaout tan io, o wiltake resposi shouldallow or a ommissoning eriodcarried out and, ff so, who will take responsi- should allow for a commissioning period biit fo. opeigth ok nte during which design or construction problems eile is .paticiptor desgn inowhic are identified and corrected. Uniform rules for mexmbes oarery go siand thica turnover should be established throughout the walketogetherroer arous determined projectand suported y govertrnen walk together over a previously determined project and supported by government layout to decide on modifications. Such regulation. "Special deals," such as engineers systems dca be modifieasines with performing services for individuals, should be agreemen both as. Farm riz- replacd by peratons though ormalagreement from both parties. Farmer organiz- replacdtbyioperationsnthroughsnorma ers should be present at these meetings. Great attention should be paid to the details of agreements and designs and to making written Fixed "patterns of action" are means of estab- records of these activities. lishing communications between the farmers and government staff. They provide a meeting 26 Enviromnent Department Papers 4. Building Participation into the Project Cycle Growing experience with farmer participation Pilot Project in irrigation systems has taught experienced Unless a government is familiar with participa- practitioners that the only hard and fast rule is tory methods and has implemented such that there are no fixed formulas for participa- . . . is .. tion. Participatory operations must be treated programs before, it is useful to initate partici- as a learning process, constantly monitored and patory projects in pilot form or to start in one as a leaming process, constantly monitored and area and expand as the techniques prove adapted to local conditions and changing themselves q b U .o thmselvesand government becomes confident circumstances. In what follows, strategies and of the approach. Because less is at stake in lessons on how to adapt and increase user pilot projects than in main projects (even of participation are organized according to the equivalent size), building participatory ap- project cycle. proaches as pilot efforts in larger projects may make governments and the Bank more willing Project Identification to try the approach. Completion of the pilot Building participation into operations from the can be set up to trigger expanded efforts beginning is desirable because it allows more automatically. Fine tuning the expanded time for the approach to take hold and for project plan should be the result of these "real other aspects of the project (from policy re- life experiments" (Cernea, 1992). However, it is forms to system design) to be made consistent not always necessary to start new pilot with user participation (Brown and Nooter, projects. Before launching new programs, 1992:43). This means engaging policymakers efforts should be made to look for participatory and agency staff in discussion of participation irrigation experiences sponsored by other at the earliest stages and identifying the most donors, NGOs, or indigenous farmer organiza- receptive indi-viduals as discussed above. As tions involved in irrigation management. Such one task manager has noted, "participation is programs also may be good sources of person- not a paragraph add on, nor a last minute nel with experience in participatory irrigation. injection. It has to be thought through carefully with the agency. If not, it means nothing." Working with Small-Scale Systems Pilot projects, while valuable for demonstrating Preparation the possibilities of participation, are often marginalized because they are seen as too small Participation should be integrated fully into or because they operate under specialized project preparation from the beginning if at all conditions. A stronger alternative in many possible. Indeed, if agency and user participa- countries is to develop participatory projects tion are not incorporated into priority selection working with larger numbers of small-scale or and design, but only in later phases of imple- community managed systems. These can be mentation, a project is not truly a "participatory separate projects or components of larger project" but a "project with participatory projects, such as the Irrigation Line of Credit components" (NIACONSULT, 1993b). pilot project under the Mahakali II and Participation Series 27 Participation in Irrigation Bhairahawa-Lumbini III Irrigation Projects in * irrigation agencies generally have less Nepal. Although it cost nearly $20 million, this involvement in small-scale systems and are officially was termed a pilot project attached to therefore not as threatened by strengthened another major project, because less would be at farmer organizations associated with them stake in a pilot than in a full project of equiva- than they might be in large-scale systems; lent size. In fact, the success of the participa- and tory pilot program made the main Bhairahawa- Lumbini III project look more successful. * it can cover a large irrigated area and test Advantages of this approach are that: the participatory approach under varying conditions (though many conditions * it builds on the experience small-scale facilitating cooperation in small-scale and systems usually have in farmer management; traditional systems may not apply in Box 8 Nepal: Generating Demand from Farmers In Nepal, the Bank has applied participatory methods to assist private, farmer managed schemes directly as a vehicle to improve irrigation performance and reduce the financial burden of irrigation development and operation by government. A system using farmers associations to construct, rehabilitate, and improve small- and medium-scale hill irrigation and shallow tubewell subprojects is being carried out with funding under Mahakali II and Bhairahawa-Lumbini III through the Irrigation Line of Credit (ILC) pilot program. Technical assistance has been initially funded by UNDP through two closely linked Irrigation Sector Support Projects, one execufed by the Bank and the other through the Asian Development Bank. ILC was large by pilot project standards, nearly $20 million in total, but as a pilot project it was flexible and could be modified to meet changing needs. It was designed as the first stage of a full-scale irrigation sector program based on farmer participation. Considerable attention was given to developing processes and procedures for participation along with technical designs for systems. To participate in ILC, farmer groups were required to form a legal farmer irrigation association, contribute cash and labor for construction costs, and assume full O&M responsibil- ity. In return, the associations received assistance in construction and provisional water rights. The organizations were structured to provide as much active participation and internal equity as possible by restricting membership and leadership to irrigators, requiring that the members appear in person before the district magistrate to enroll, and having the association president and project officer as signatories on the association bank account for capital expenditures. ILC provided substantial benefits in both crop production and govemment budget savings, and resulted in substantial behavior changes in farmers and the department, as well as related legal, organizational and political reforms. The subprojects were demand driven, and farmers were given an opportunity to determine the upgrading of irrigation systems in return for sharing costs of rehabilitation and taking responsibility for O&M. The project resulted in better service to farmers at lower cost and showed the importance of Bank timing in the creation of alliances with key government officials to initiate and implement policy and institutional shifts through popular participation (Gautam and Reidinger, 1993). Project developers recognized the need to demonstrate strong positive results within two years. This was achieved even though the first years of the project coincided with a period of political unrest and severe trade restrictions that made critical supplies unavailable. Even after the technical assistance ended and Bank involvement was reduced, the procedures and processes which were developed remained. Farmer participation is now difficult to "turn off" because its success has been demonstrated and there is demand for the project from the farmers' side. 28 Environment Department Papers The Project Cycle systems normally managed by agencies). * determining the nature of actual existing Coordinating larger projects in many, often irrigation practices; remote, locations presents difficulties, but it often forces the agency to rely more on * eliciting participation by farmers in the local input in developing each subcompo- identification and design of the project; and nent. * helping the borrower to develop a strategy Sweeping Changes for organizing and training farmers and establishing conmnunications between As mentioned above, because government is farmers and the technical staff. more receptive to new approaches during crisis situations such as occurred in Mexico in the The experience of specialists is important, but late 1980s, these may offer greater opportuni- circumstances in one country may dictate ties for building participatory operations substantially what can be borrowed from provided adequate stability exists. In these experience gained in another. It is useful to situations, it may be possible to incorporate combine international specialists who can bring more sweeping changes, such as devolution of ideas from a number of contexts with local irrigation management on a larger scale, experts and social scientists from universities without a multiyear pilot project. or other institutions who can provide more intimate knowledge of the local situation. Enabling Conditions Good field level social research is essential. The project preparation phase is also the Unless there is sufficient time and money for appropriate time to begin discussing legislative longer-term research, rapid appraisal tech- reforms or other enabling conditions needed to niques should be used for assessment of legitimize and strengthen the incentives for current institutional practices, attitudes, user participation. In this regard, legal recogni- women's roles, and willingness to participate in tion of farmer organizations and reform of O&M. However, institutional specialists must irrigation fees are most often the focus of not work in a vacuum. They must link plans to attention. In Mexico, however, reforms of land the engineering, financing, and broader institu- and water laws to provide secure water rights tional structure of the project. All disciplines and appropriate incentives for farmer invest- involved must have a core commitment to ment, including ownership of the systems participation. As Cernea (1992) says, "The themselves and water conservation, are receiv- elaboration of a social methodology for bottom- ing more attention. Other enabling provisions up planning requires the joint effort and to facilitate effective participation within integrated skills of professional researchers and WUAs include ensuring accountability to development practitioners; they must design members and restructuring agencies to become the 'software' together." financially autonomous. Legal requirements that the government transfer management of Consultancy Funds irrigation systems to farmers have been passed Funds available to TMs in project preparation in Colombia and Mexico. In addition to new are usually scarce and need to be spread among laws, the implementing regulations are critical engineering, environmental, economic, and to the effectiveness of these provisions. other areas of expertise. This may leave little money for anthropological, sociological, organi- Provisions for Participation zational, or legal consultants to design participa- Whether a TM is developing pilot projects or tory programs. Tapping UNDP, Population and changes in irrigation sector policy, assistance is Human Resources Development, trust funds, or needed to flesh out the nature and provisions for other sources of technical assistance can make user participation and organization. These available the necessary funds (often as grants, include: which governments accept more readily). In the Participation Series 29 i drrcipation in Irrigation Philippines, a $20,000 grant from the Swedish structure, and agency reforms. Pressure from the International Development Agency allowed a Bank may be useful if governments are to imple- local institutional consultant to work with the ment politically unpopular but necessary re- National Irrigation Authority Project Prepara- forms. However, conditions for signing, loan tion Team to ensure farmer input in identifying effectiveness, or disbursement will not in them- priorities. In Nepal, $2.5 million in technical selves lead to participatory projects without assistance from UNDP for the Irrigation Line of strong commitment from the policymakers and Credit proved extremely valuable not only in agency staff, as discussed above. technical feasibility studies for component subprojects, but in direct technical assistance to Design for Participation fill gaps, promote "learning by doing," and develop a program for training agency and The Bank has recognized the importance of farmer organization staff to implement the physical design for project hardware in irriga- approach. The TM coordinated the technical tion. Before a project can be approved, the assistance directly, which led to greater involve- detailed design of all major structures must be ment between the Bank and the Irrigation complete. A similar requirement would be Department and a more concerted effort to valuable for the design of a project's social involve farmers. Unfortunately, procedures and components. This approach would require schedules for such applications are time con- beginning social planning, selection, training, suming and often discourage TMs from pursu- and fielding of organizers, consultations with ing these resources. Another possibility is to use farmers, and even the early phases of organiza- funds from a previous sectoral or project loan tion before the project was approved officially. for preparation of follow-on participatory A detailed plan of action for organization projects (for example, in the Philippines Irriga- would be presented before the project was tion Operation Support Project I and II). submitted, specifying the plans and schedules for establishment of user groups, structures, Appraisal rights and obligations of farmer organizations, cost sharing and cost recovery mechanisms, support to communities in managing funds, One of the most important elements of success and patterns of action to carry the program in a project is working with government to through its developmental stages. generate a clear policy on participation, a strategy for its introduction, and a detailed In Nepal, early emphasis on developing clear program for interaction as the project is imple- procedures for participation paid important mented. In the appraisal phase, task managers dividends. Not only did these processes need to ensure that all of these elements are produce better results in the project, once they adequate and workable, that the borrower is were established within the irrigation agency committed to the approach, and that necessary and known to the farmers they became the preconditions are in place. standard operating procedure and continued beyond the phase of project technical assistance Conditionality and heavy Bank involvement. While the scope for shaping the approach of the Hard questions should be asked about how project is great during project preparation, formal each aspect will be implemented. If these are leverage in specifying conditionality is greatest not answered, appraisal is not complete. during the appraisal phase. If judiciously ap- However, this initial plan should be taken as plied, such leverage can faciltate the develop- tentative and modified as the project gains ment of an appropriate enabling environment for experience and dialogue with the farmers the success of the participatory approach, for progresses example through changes in legislation, pricing 30 Environment Department Papers The Project Cycle Farmer Input in Project Design farmer inputs come from a cross section of For sustainability in irrigation projects, systems potential users. The litmus test for farmers is must be built that people want and are able to their willingness to pay at least part of the use. To do this, people must be consulted from capital cost. This should be ascertained during the beginning of the project cycle, Interaction appraisal and, where feasible, the principle of with farmers can come at many levels and i "*no pay, no project" should be built in. To with forms, including information sharing, construct or rehabilitate a project without many fon, decisingaing,ran taring commitment of time and resources from consultation, decisionmaking, and taking farmers forgoes a valuable opportunity to build action. Normally, only the simplest forms ofpatcaio.Tpredwthhyclwrk interaction (information sharing and consulta- participation. To proceed with physical works tion) are involved in these initial contacts. in advance of farmer involvement may appear Since farmer irrigation organizations may not to keep projects proceeding smoothly, but it is be available at this stage, it should be the task likely to yield less sustainable projects over of the institutions specialist to hold some of time. these early consultations and ensure that Box 9 Nepal: Bank and Government Partnership for Participation The Bank has been integral to the development of the government's participatory approach to irrigation in Nepal. Although Nepal has a long tradition of user participation in irrigation, the use of these concepts for formal irrigation development was introduced through the pilot USAID sponsored Irrigation Management Program. The first full-scale implementation of this approach was through the Bank sponsored Mahakali Irrigation II Project which utilized user groups for participatory design and layout of tertiaries as well as many phases of O&M within the system. In Mahakali, outlet groups and tertiary committees were formed through the use of association organizers and, by 1992, farmers had organized secondary and system wide apex level associations that were involved in water fee collection and other aspects of system regulation. A version of this approach, modified for groundwater, was used in Bhairahawa-Lumbini III Project to promote the construction and operation of deep tubewells and to manage them after turnover. This project developed socioeconomic criteria for selection of projects including the formation of user groups, payments of shares of construction costs, and agreements to pay operating costs and to operate and maintain projects. Earlier phases of the project lacked the demand for farmer organization and turnover which is now being carried out, de facto, using canal lining as a quid pro quo. Not surprisingly, a marked improvement in water use and maintenance was evident soon after turnover took place. Legal recognition has been established for user groups and the registration process will soon be underway. The Nepalese projects adopted many of the features of projects in the Philippines, including: * increased participation of farmers in key decisions and in planning and development of the physical systems; * development of cost recovery mechanisms which instilled in farmers more responsibility and a sense of collective ownership of systems; * a well trained organization development unit able to live in the field and work intensively with farmers to help them form strong organizations; and * the development of mechanisms for a learning process which allow for gradual refinement of methods and procedures. The success of the participatory approach in these two projects has led the Nepalese to adapt the approach to two additional agency managed Bank projects, Narayani III and Sunsari-Morang II, although unfortu- nately this took place quite late in their project cycle. Participation Series 31 Participation in Irrigation Supervision restructure the project to include more farmer managernent. As previously noted, crisis Continued Involvement situations provide opportunities to build To be sustainable, participatory projects must participation at a time when the government is be treated as a continuous learning process by often relatively receptive to new approaches. Bank TMs, agency staff, and farmers. One-shot Once participatory approaches are established efforts, no matter how well designed, wiOl not in a country, other irrigation projects may be ensure sustainable institutions. The supervi- retrofitted t mahem more participatory, as sion phase is necessary to adapt plans to was done in the Mahakal project in Nepal. changing conditions, and often requires inten- . .i. sive guidance and dealing with practical Enforcing Conditonality problems. This is more difficult once the loan has been signed and the project is underway. Sus- A plan for organizational development must be pending disbursements is very difficult flexible should changes be required in the because of internal Bank pressures for chosen strategy. This requires follow-up smooth project execution and political refresher training for organizers and hiring of a repercussions from the borrower. Task coordinator who can determine when program managers therefore must depend on the changes are necessary, make appropriate commitment of policymakers and agency modifications, and retrain and reassign field staff to the participatory program. level staff. Carefully written records always should be kept of organization development Using Leverage activities and results at all levels. There is one type of conditionality that, if built into the project at appraisal, can be applied Where to Start during the supervision period and has proved From a pragmatic standpoint, it is useful to very useful in 6emonstrating the importance of a begin implementing participatory approaches commitment to participatory approaches. This where they have the greatest likelihood of involves a requirement that disbursements for success. On one hand, this may mean trans- particular project components (such as individual ferring management to farmers of the sys- irrigation districts or systems) be conditional tems in the best financial condition rather upon meeting agreed targets (such as for certain than in the poorest areas, where the addi- levels of O&M cost recovery or progress toward tional costs to farmers may be too great. On farmer management). Payments or suspension the other hand, farmers' interest in taking are triggered automatically by achievement or over management may be greatest in systems failure to achieve these targets. An advantage of where problems with agency management this approach is that it binds both Bank and are most apparent. borrower to a commitment to follow through on agreed upon objectives, even if the TM changes. Adding Participation During Because suspensions affect only portions of the Implementation project, the political repercussions are not as great. At the same time, suspension of one Though participation should be built into component shows the importance of meeting projects from the beginning, the supervision objectives for others. This was found in Colom- phase is not too late to introduce participation. bia where suspension of payments on one In Turkey, for example, project difficulties irrigation district for failure to turn management involved slow disbursements and failure to over to farmers led to redoubled efforts in other meet conditions for adequate O&M cost districts. recovery. These provided an opportunity to 32 Environrnent Department Papers The Project Cycle Following up on Participation information is collected on a regular basis can The attention TMs pay to the progress in user prove useful in focusing attention on participa- participation is crucial to demonstrate the tion during project evaluation. Indicators also seriousness of the Bank's commitmnent to partici- provide useful information on progress during pation. Otherwise, if only procurement, technical project implementation. standards, or economic returns receive attention The indicators of farmer participation that have during supervision missions, borrowers will been most commonly used are: assume that participation is less important. * number of organizations registered; Personal Approaches of Task Managers * farmers' contributions or fee collection The approach taken by TMs in project prepara- rates; and tion, appraisal, and supervision also sets important examples for participation. This * number of systems transferred to farmer includes spending time in the field. For ex- management. ample, Bank staff may meet with farmers alone, rather than only with agency staff, and in joint These indicators are quantifiable and the data meetings they may at times sit with farmers. they require are not too great. How well they Members of the mission who listen to and capture real participation depends on the involve policymakers, agency staff, and other context. While the registration of "paper stakeholders can similarly demonstrate com- organizations" may mean little in some situa- mnitment to participation throughout the tions (for example, in Pakistan's WUAs), in project. The key words for TMs engaged in Nepal the registration of an association and participatory projects are "collegial" and signing of an agreement to take over full O&M "catalytic," not "coercive" or "confrontive." could only happen after a substantial commit- ment of time and money on the part of the Evaluation farmers and the agency. Similarly, fee collec- tion rates by farmer associations may be a good Appropriate Evaluation Methodologies indicator of organizational strength and satisfaction, but government fee collection rates These are needed if the objective of increasing may reflect either farmer satisfaction with the user participation is to receive the same weight systern or strength of government enforcement. as technical performance, economic rate of Transfer of management to farmers is one of return, and cost recovery objectives. Without the strongest indicators of increased user these appropriate methods, while the effort participation (as, for example, in Mexico). that has gone into building participation may improve technical performance, it will not Detailed Indicators appear directly in the evaluation of the project. More detailed and process oriented indicators Indicators of Participation include whether farmer organizations have met, who attended meetings (among farmers Participation objectives and indicators of and with agency staff), whether records of success are often not clearly specified. While meetings and financial transactions are kept, those involved may have a good sense of the frequency and quality of maintenance whether or not projects are truly participatory, work, and the proportion of farmers who can this can be difficult to translate into data which identify association leaders or representatives. external evaluators can use. Specifying clear These require more information and time to indicators and ensuring that the necessary collect, and therefore may only be feasible on a Participation Series 33 Participation in Irrigation subsample of project areas. It is therefore and their expectations are raised. Tracking the critical that the sample of subsystems and number of meetings with farmers has similar farmers be truly random and not influenced by shortcomings, but staff familiarity with the project authorities. Contracting local universi- names and problems of the farmers is a useful ties or research organizations for such work is indicator. Farmers' formal evaluations of appropriate, especially where Bank staff cannot agency staff, indeed even the existence of move independently nor speak with farmers mechanisms for such evaluations, are signs of without a translator. participatory agencies. If agency revenues, farmers' payments, and service levels are Indicators for Agencies linked structurally (as in financially autono- Indicators can be used to evaluate whether mous agencies), fee collection rates are useful agencies are becoming more service oriented indicators. Without such a linkage, the Bank's and receptive to user participation. Logs of traditional preoccupation with cost recovery farmers' complaints are probably not a good linked levels as an indicator of sustainability indicator, for farmers may complain more as will not capture the real sustainability of they become more involved in management projects which meet the real needs of users. 34 Environment Department Papers References Bagadion, B. U., and F. F. Korten. 1991. Cemea, M. M. 1992. The Building Blocks of Developing Irrigators' Organizations: A Participation: Testing a Social Methodology. Learning Process Approach. In Putting People In Participatory Development and the World Bank: First: Sociological Variables in Rural Development, Potential Directionsfor Change, ed. B. Bhatnagar ed. M. M. Cernea. Second edition. U.K.: and A. C. Williams. World Bank Discussion Oxford University Press. Paper Number 183. Washington, D.C.: World Bank. Bardhan, P. 1993. Analytics of the Institutions of Informal Cooperation in Rural Development, Cemea, M. M. 1993. Culture and Organiza- World Development 21: 633-39. tion: The Social Sustainability of Induced Development. Sustainable Development 1: 18-29. Baum, W. C. 1982. The Project Cycle. Washing- ton, D.C.: World Bank. Cemea, M.M., and R. Meinzen-Dick. 1992. Design for Water Users' Associations: Organi- Bellekens, Y. 1986. An Impact Evaluation Study zational Characteristics. In Developing and on Water Usage and its Contribution to Agricul- Improving Irrigation and Drainage Systems. tural Production in Irrigation Projects in the World Bank Technical Paper Number 178. Philippines. Manila: Asian Development Bank. Washington, D.C.: World Bank. Bhatnagar, B., and A. C. Williams, eds. 1992. Coward, E. W. 1980. Management Themes in Participatory Development and the World Bank: Community Irrigation Systems. In Irrigation Potential Directionsfor Change. World Bank and Agricultural Development in Asia: Perspec- tivesfrom the Social Sciences, ed. E. W. Coward. Discussion Paper Number 183. Washington, Ithaca, NY: Cornell University Press. Coward, E. W. 1991. Planning Technical and Brown, E. P., and R. Nooter. 1992. Successful Social Change in Irrigated Areas. In Putting Small-scale Irrigation in the Sahel. World Bank People First: Sociological Variables in Rural Technical Paper Number 171. Washington, Development, ed. M. M. Cernea. Second Edi- D.C.: World Bank. tion, revised and enlarged. New York: Oxford Bymes, K. J. 1992. Water Users' Associations in University Press. World Bank-assisted Irrigation Projects in Paki- Dichter, T. 1992. De-mystifying Popular stan. World Bank Technical Paper Number 173. Participation: Institutional Mechanisms for Washington, D.C.: World Bank. Popular Participation. In Participatory Develop- ment and the World Bank: Potential Directionsfor Cemea, M. M. 1987. Farmer Organizations Change, ed. B. Bhatnagar, and A. C. Williams. and Institution-building for Sustainable Devel- World Bank Discussion Paper Number 183. opment. Regional Development Dialogue 8: 1-24. Washington, D.C.: World Bank. Participation Series 35 Participation in Irrigation Esman, M. J., and N. T. Uphoff. 1984. Local Merrey, D. J., and D. H. Murray-Rust. 1991. Organizations: Intermediaries in Rural Develop- Peoples' Participation in the Gal Oya Rehabili- ment. Ithaca, NY: Cornell University Press. tation Project as Viewed by Agency Personnel. In Farmers in the Management of Irrigation FAO. 1985. Participatory Experiences in Irriga- Systems. ed. K.K. Singh. New Delhi, India: tion Water Management. Proceedings of the Sterling Publishers Private Limited. expert consultation on irrigation water man- agement held in Yogyakarta and Bali, Indone- Manzardo, A. 1992. Small-scale Irrigation in sia, July 16-22, 1984. Rome: FAO. Himachal Pradesh: Project Completion and Recom- mended Further Activities. Educational Develop- Freeman, D. M. 1990. Designing Local Irriga- ment Center, Newton, MA. tion Organizations for Linking Water Demand with Supply. In Social, Economic and Institu- Manzardo, A. 1990. Farmer Organization in the tional Issues in Third World Irrigation Manage- Mahakali Irrigation Project: Development of Associa- ment, ed. R. K. Sampath and R. A. Young. tions and Field System Management. Boulder, CO: Westview Press. Mahendranagar, Nepal: Sir William Halcrow and Partners and Mahakali Irrigation Development Helmi, and D. Vermillion. 1990. Using Irriga- Board. tion Agency Staff as Institutional Organizers: The Small-systems Turnover Program in Moench, M. 1994. Farmers' Participation in Indonesia. In Role of Social Organizers in Water Management in Maharashtra. Report to Assisting Farmer-managed Irrigation Systems. ed. World Bank. Mimeograph. S. Manor, S. Patamakul, and M. Olin. Colombo, Sri Lanka: International Irrigation Nguyen V. T. 1994a. Water Users' Associations in Management Institute. the Irrigated Perimeters in Madagascar. World Bank, Washington D.C. Mimeograph. Korten, F. F. 1993. Correspondence to Bhuvan Bhatnagar. Nguyen V. T. 1994b. Water Users' Associations in the Left Bank Valley of the Senegal River. Korten, F. F., and R. Y. Siy, Jr., eds. 1989. World Bank, Washington D.C. Draft. Transforming a Bureaucracy: The Experience of the Philippine National Irrigation Administration. NIACONSULT Inc. 1993a. An Evaluation of the Ateneo de Manila University Press. Impact of Farmer's Participation on the National Irrigation System's Performance. Final Report. Manor, S., S. Patamatakul, and M. Olin, eds. Philippines: NIACONSULT Inc. 1990. Role of Social Organizers in Assisting Farmer-managed Irrigation Systems. Proceedings NIACONSULT Inc. 1993b. The Role of World of a Regional Workshop of the Farmer-Man- Bank in NIA's Participatory Programs. Final aged Irrigation Systems Network held at Khon Report. Philippines: NIACONSULT Inc. Kaen, Thailand, May 15-20, 1989. Colombo, Sri Lanka: International Irrigation Management OED. 1993. A Review of World Bank Irrigation Institute. Experience. 2 vols. Draft. World Bank, Wash- ington D.C. Manor, S., and J. Chambouleyron, ed. 1993. Performance Measurement in Farmer-managed Ostrom, E. 1990. Governing the Commons: The Irrigation Systems. Proceedings of an Intema- Evolution of Institutions for Collective Action. tional Workshop of the Farmer-Managed Cambridge, U.K.: Cambridge University Press. Irrigation Systems Network held at Mendoza, Argentina, November 12-15, 1991. Colombo, Ostrom, E. 1992. Crafting Institutions for Self- Sri Lanka: International Irrigation Manage- governing Irrigation Systems. San Francisco: ment Institute. Institute for Contemporary Studies. 36 Environment Department Papers Plusquellec, H. 1989. Two Irrigation Systems in Tang, S. Y. 1992. Institutions and Collective Colombia: Their Performance and Transfer of Action: Self-governance in Irrigation. San Fran- Management to Users' Associations. World Bank cisco: Institute for Contemporary Studies. Working Paper WPS 264. Washington, D.C.: World Bank. Uphoff, N., P. Ramamurthy, and R. Steiner. 1991. Managing Irrigation: Analyzing and Plusquellec, H. 1994. India, Morocco, Pakistan: Improving the Performance of Bureaucracies. New Mission from January 11-February 18, 1994: Back- Delhi, India: Sage Publication Inc. to-Office Report on Water Users' Associations. Mimeograph. Washington, D.C.: World Bank. Uphoff, N. 1986. Getting the Process Right: Improving International Irrigation Management Reidinger, R., and U. Gautam. 1992. Promot- with Farmer Participation. Boulder, CO: ing Private Irrigation Development: The Westview Press. Irrigation Sector Program Experience in Nepal. Irrigation and Drainage Session Proceedings of Uphoff, N. 1992. Learningfrom Gal Oya: Water Forum '92. Baltimore, MD: American Possibilities for Participatory Development and Society of Civil Engineers. Post-Newtonian Social Science. Ithaca, NY: Cornell University Press. Singh, K.K. ed. 1991. Farmers in the Manage- ment of Irrigation Systems. New Delhi, India: Uphoff, N. 1993. Grassroots Organizations and Sterling Publishers Private Limited. NGOs in Rural Development: Opportunities with Diminishing States and Expanding Siy, R. Y. 1982. Community Resource Manage- Markets. World Development 21(4): 607-622. ment: Lessonsfrom the Zanjera. Quezon City, Metro Manila: University of the Philippines Vermillion, D. L. 1991. The Turnover and Self- Press. management of Irrigation Institutions in Develop- ing Countries. Colombo, Sri Lanka: Interna- Small, L. E., and I Carruthers. 1991. Farmer- tional Irrigation Management Institute. financed Irrigation: The Economics of Reform. U.K.: Cambridge University Press. World Bank. 1992. World Bank Experience with Irrigation Development: Socio-economic, Institu- Svendsen, M. 1992. Assessing Direct Effects of tional and Technical Impact and Lessons. 6 vols. Policy Change on Irrigation System Performance: Washington, D.C.: World Bank. A Case Studyfrom the Philippines. IFPRI Work- ing Paper on Irrigation Performance Number 2. World Bank. 1993. Water Resources Manage- Washington, D.C.: International Food Policy ment: A World Bank Policy Paper. Washington, Research Institute. D.C.: World Bank. Svendsen, M., and D. Vermillion. 1994. Irriga- World Bank. 1993. Mexico: Review of Condi- tion Management Transfer in the Columbia River tionsfor Encouraging Private Investment in Basin (U.S.A.) Project. HMI Research Report. Irrigation. Report No. 11774-ME. Washington, Colombo, Sri Lanka: International Irrigation D.C.: World Bank. Management Institute. Participation Series 37 Environment Department The World Bank 1818 H Street, N.W Washington, D C 20433 202-473 3641 202 477 0565 FAX W Printed on 100% post-consumer recycled paper