A few localized examples of co-management
Back to contents - Previous file - Next file
In the late 1890s, the Norwegian government enacted special legislation for the Lofoten fishery (in the north of the country) where the Arctic cod has its spawning grounds. This was because, due to growing numbers of participating fishermen, this fishery had become seriously crowded. As in Japan, this legislation actually delegated responsibility for the regulation of the fishery to the fishermen themselves. Following one account, 'Special district committees of fishermen representing different gear groups were set up to make the rules for the fishery, such as: allowable fishing times; which gear is allowed on which fishing grounds; and how much space should be reserved for certain gears such as handlines, gillnets, longlines, seines. In addition to elected fishermen inspectors, a public enforcement agency was established to assure that the rules initiated by the fishermen were being obeyed' (Jentoft, 1989: 141). This system of comanagement under which the fishermen of Lofoten have been given exclusive rights by the government still prevails today and, according to Svein Jentoft, it can be considered a success: it has worked well for a long period of time and the fishermen concerned take it for granted (ibid. 140-1, 153).
In the Alanya fishery in Turkey, as we have already pointed out (see above, Chapter 10), there are precise rules governing access to fishing sites and ensuring equitable rotation among the different rights-holders. The effectiveness of these rules is all the greater as they require extensive knowledge of local conditions which can be acquired only through prolonged experience. In the words of Ostrom: 'Mapping this set of fishing sites, such that one boat's fishing activities would not reduce the migration of fish to other locations, would have been a daunting challenge had it not been for the extensive time-and-place information provided by the fishers and their willingness to experiment for a decade with various maps and systems' (Ostrom, 1990: 20).
What needs to be noted in the context of the present discussion is that the Alanya co operative claims a legal status on the basis of a broad interpretation of the Aquatic Resources Act which states that co-operatives have jurisdiction over 'local arrangements' (Berkes, 1986: 222). Such legal status, Ostrom thinks, adds legitimizing power to the authority of the Alanya fishermen's association to make and enforce local fishing regulations (Ostrom, 1990: 20). Yet, it should be emphasized that such legal status is noticeably looser than that enjoyed by the Japanese FCA and, therefore, the Alanya experience is only an imperfect illustration of what a co-management approach can achieve. In actual fact, and in contrast to coastal lagoon fisheries which are leased by the State and operated by private interests or co-operatives (in these fisheries, there is thus a sound legal basis for the regulation of access to the resource), other coastal fisheries are legally open-access territories. Consequently, the Alanya co-operative has no legal authority to restrict membership or to act as an exclusive organization gathering all of the fishermen under its umbrella (Berkes, 1986: 221-2). There is thus a profound ambiguity in the Turkish legislation which is hardly congenial to effective decentralized management of fish resources.
Located in Mindanao island in the southern Philippines, the Panguil Bay area is the source of some of the most valuable species of shrimp and crustaceans in the country. About forty-seven rivers and tributaries flow into the bay which has a coastline spanning 116 km. and a population of 450,000 inhabitants. Mangrove destruction and overfishing have contributed heavily to the decline of this abundant resource base during the past decade: less than one-third of the mangrove forests remain and, while fishing boats in the bay doubled between 1985 and 1991, the total catch dropped by 75 per cent (FAO, 1993: 146 7). In an attempt to combat problems resulting from environmental degradation, resource depletion, low productivity, and poverty, the Philippine government, supported by the Asian Development Bank and the Overseas Cooperation Fund of Japan, established the Fisheries Sector Programme (FSP) in 1989. Being one of the priority areas selected for improved coastal resource management, Panguil Bay has been made a testing ground for a new approach to CPR management based on site-specific planning and participation of local governments, NGOs and fishing associations. Note that, under the new Local Government Code, the management of municipal waters falls under the authority of the local government units and that the coverage of these waters has been expanded from 7 to 15 km. from the shoreline (Cure, 1994: 3).
The experience is apparently a success story. Thus, we hear that: the decentralization, training and involvement of local law enforcement is credited with an impressive record in 1990: local forces confiscated or destroyed about 1,600 filter nets, apprehended more than 60 violators, seized 30 scissor nets and uprooted more than 200 net posts in Panguil Bay . . . In addition to communities enforcing regulations aimed at protecting their resources, local fishing associations are constructing, protecting and managing artificial reef sites in the mouth of the bay to replace coral reefs destroyed by dynamite fishing. Some municipalities and communities are developing territorial use rights in fisheries, delineating zones for specific fishing gears and establishing areas for seaweed, mussel and oyster cultivation. Coastal inhabitants are reforesting 600 ha. of open mud-flats through community-based contracts. Individual families are receiving certificates of stewardship to increase land tenure security and use rights to both these reforested mangrove forests and existing forests. At the same time, mangrove zoning is resulting in commercial production zones, buffer zones, limited use zones and strict nature reserves to improve management of the mangrove resources. (FAO, 1993: 47).
It needs to be pointed out, however, that the success achieved by this co-management project has probably much to do with the fact that, so far, the measures taken to protect local fish resources have been essentially directed against external intruders. A complete test of the ability of coastal communities to play their part in a co management mechanism requires that they successfully participate in the managementand not only in allocation through exclusionary arrangements of their resources, implying that they help design, monitor, and enforce rules of restraint to be applied to their own members (personal communication of Rolf Willmann).
Co-management prospects for inroad-based user organizations
In the Canadian Atlantic fishery, the so-called Maritime Fishermen's Union (MFU) got started in the mid-1970s as a militant fishermen's organization that wanted to break the dominance of inshore fishermen by big commercial companies and to defend their way of life. Made up of fishermen who mainly own and operate fishing vessels less than 13.7 metres in length, it covers a very long coastline comprising three maritime provinces on the east coast of Canada (Newfoundland where 70 per cent of the cod fishery is found is excluded, however). The fishermen concerned rely primarily on lobster fishing but they also catch other species. Their communities have usually been quite successful in regulating access to the fishing grounds for lobsters as well as in enforcing state-engineered conservation measures for that species (minimum legal size of lobsters allowed to be caught). Such a success, it seems, can be largely ascribed to the favourable characteristics of the resource (it is well localized) as well as to the fact that there is no competition from outsiders (in particular, from big commercial companies).
The MFU intends to strengthen itself with a view to co-managing other fish resources which are more problematic than lobsters. It appears that, after the complete failure of its management policy which resulted in the collapse of the cod and the herring fisheries (the latter of which has since recovered following the assignment of a much lower share of the total quota to large industrial fleets), the Canadian government is now willing to deal with organizations such as MFU in order to delegate to them the tasks of distributing the fishing licences and enforcing specific regional quotas set on the basis of conservation considerations.
It is a noteworthy feature of the present fishing scene that broad-based fishermen organizations have also arisen in developing countries such as Chile (Conapach, or the Confederacion Nacional de Pescadores Artesanales de Chile), India (the National Fishworkers Forum), Senegal (CNPS, or the Collectif National des Pêcheurs Sénégalais), and the Philippines (Kammppi, Bigkis-Lakas Pilipinas, Pamalakaya). These organizations are first and foremost concerned about the rapid degradation of the fish resources which are the mainstay of their members' livelihood. Their objectives are close to those pursued by the MFU in Canada. In the Philippines, for instance, the struggle of fishermen organizations is presently focused on the need for institutionalizing the participation of fishermen in coastal resource management through the creation of resource management councils in every municipality (Cure, 1994: 4). In the near future, it will be interesting to see to what extent they succeed in playing an effective role as comanagers of coastal fish resources, assuming of course that their respective governments agree to enter into co-management agreements with them.
India: co-management as a
In Chapter 11 (sect. 1), we offered the reader a rather detailed review of the forest policy of the government of India during both the colonial and the post-independence eras. At the end of this review, we quoted the so-called Report of Committee on Forest and Tribals in India (1982) in which the authors advocated a major turn towards a management strategy in which user groups were to play a much more active role than in the past. Article (xi) under the heading 'Forest Policy' (pare. 5) thus reads:
Tribal and local organisations, may be made use of if in good shape and after revitalisation if not in good shape, for management of protected and village forests for commercial, social and farm forestry purposes.
Further on, under the heading 'Management System' (pare. 23), we find the following set of recommendations:
(i) The crux of the problem of forest management lies in the need for integration of tribal and forest economies. The relationship between forest managers and tribals should be one of partnership. This will be possible if an identity-interest between the forest department and tribals is created.
(ii) Forestry development programmes should aim at internalising its components into the rural production system as a whole.
(iii) The management should ensure strong backward and forward linkages between forestry and other development sectors on the local, regional, State and national levels.
(iv) Tribals should be inducted into a more constructive role of forestry. They should be employed in forest service at different levels by imparting specialised training.
(v) Forest management practices need to be modulated to be able to generate employment all the year round for prevention of migration and sustained supply of raw materials for the requirements of agriculture and industry.
(vi) The role of forester needs to be reappraised. The new emphasis should be on forester as an extension agent advising the owners or the management personnel of village, communal private and other forests for undertaking scientific forestry.
(vii) The transformation from conservation to development forestry should be induced through community forestry. Forestry activities should be carried out by many, often local institutions, rather than by a single forest department. In other words, a meta-management system should be applied rather than super-management.
(viii) The course from departmental production forestry to broad-based community forestry can be made smoother through public participation. At the State-level, a broad-based body comprised of officials, technical experts, academics, leaders of public (particularly tribals) opinion should be built up. At the forest, divisional and ITDP levels, advisory committees representing forest interests like government departments, statutory bodies and forest dwellers to review, formulate programmes and oversee their implementation should be set up. Similar committees should be set up at development block level.
(ix) The respective roles of the Forest Department, the Forest Development Corporations, the Tribal Development Corporations should be clearly spelt out. The Forest Department might be the apex agency for formulation of policies and programmes as well as for supervision of their implementation. Execution of programmes may be entrusted to field level corporate organisations like the Forest Development Corporations as well as local representative institutions.
Finally, under the final heading of the Report, 'Legislation', it is recommended that new laws be passed in order to 'strengthen the symbiotic relationship between forests and tribals' (pare. 24). Under pare. 25, more precise principles are laid down, in particular:
After some hesitationapparently due to deep-rooted mistrust of some officials in the users' ability to manage their resources effectivelythe Government of India decided to experiment with at least some aspects of the new strategy suggested in the above report. Institutional structures similar to districts are thus currently being tested for forest management in many Indian states through joint forest management (JFM) projects. JFM typically involves user groups (generally villages) forming societies and reaching agreements with the Forest Department regarding the management of local forests (Moench, 1992: A-11). It is incumbent upon the Forest Department to set broad boundary conditionsfor example, grazing and cutting limitationsand it is its commitment to turn over 25 per cent of the timber harvest and all minor forest products to the local society. In return, the latter body agrees to design and enforce a management system which meets its needs within the aforementioned boundary conditions.
According to Moench, the limited experiences that exist with JFM tend to suggest that it will only function well in certain circumstances. JFM appears to work successfully in West Bengal (bear in mind the successful Arabari experiment detailed in Chapter 12), in Orissa, while in Haryana, 'the greatest potential for success appears to be in situations where relatively small socially homogeneous villages are the primary users of clearly defined forest areas and have a high level of dependence on those forests' (Moench, 1992: A-l I ). It is apparently in the first of these states that the new policy has yielded the most convincing results and the fact that West Bengal has benefited from a longer period of experience with joint management than other Indian states no doubt accounts for this relative success. In the words of Dolly Arora: 'If JFM has been far more effective in West Bengal than in several other states, it is noteworthy that participation there preceded the adoption of JFM rules, and that too in a big way. There were no less than 1,200 protection committees already in operation in various parts of the state' (Arora, 1994: 696). (Note the striking similarity between these conditions of successful collective action and those mentioned in Chapter 12.) In other states (bear in mind, however, the noticeable exception of Uttarakhand in northern Uttar Pradesh which has been referred to in Chapter 11), the bureaucratization of the programme of participation seems to be a problem. As a matter of fact, the concern of forest officials for attaining high targets in terms of formation of forest protection committees has sometimes resulted in the formation of a large number of committees that either exist only on paper or are manipulated by a few powerful persons in the area (ibid. 694, 696).
In neighbouring Pakistan, in the West Frontier Province, a co-management experiment documented in Chapter 11 has also been started based on forestry co-operatives: to recall, these co-operatives manage a well-delimited forest area according to a management plan approved by the Forest Department and for the preparation of which technical assistance is provided to them. All costs related to maintenance and extraction of the trees are borne by the cooperatives that, in return, are authorized to retain at least 40 per cent of the revenue from the sale of trees. Unfortunately, the experiment does not seem to be promising owing to undue interference of the government and party politics which have the effect of demotivating the co-operative members.
Nepal: co-management as a new reality
In its forest management policy, the Nepalese Government has basically made the same mistakes as its Indian counterpart. Like the government of India, it has eventually come to question the premisses upon which its previous policy was grounded. However, it has shown quicker and stronger determination to reverse its previous policy (nationalization of all forest and waste lands, including village-controlled forests, in 1957) by making a radical shift towards people-based forest management. In 1978, indeed, the government of Nepal promulgated new regulations 'to enable substantial amounts of public forest land to be handed over to local communities to control and menage' (Arnold and Campbell, 1986: 431). As a matter of fact, it now became possible for the Forest Department to enter into agreements to transfer forest to village panchayats. The categories of forest that could be transferred to local community control are: (a) Panchayat Forests (PF), for the purpose of reafforestation in the interest of the village community; (b) Panchayat Protected Forests (PPF), for the purpose of protection and proper management; and (c) Contract Forests (CF), which could be awarded to either individuals or groups (ibid.).
During the early years, there was substantial resistance to authorizing the large-scale transfer of forest resources owing to the fears of many officials (who had been trained in the perspective of a centralized approach to resource management, as in so many other countries) that the local population would destroy them once government controls were lessened. In fact, the greatest barrier to community participation during the project's early years was the lack of widespread public knowledge of the purpose of the new strategy or of the details of managing a PPF (Arnold and Campbell, 1986: 444 5)
To help make the control transfer effective, the Community Forestry Development Project was established. This project was destined to support the three main elements of local management of forest resources: managed PPFs, planting of PFs, and production of seedlings for private planting. Forest nurseries were to be set up in all participating panchayats. Financing and training were to be provided for locally recruited panchayat forest foremen to run the nurseries and for panchayat forest watchers to protect the plantations and managed forests (in keeping with the tradition of forest watchers in former local forest management systems). Finally, a new cadre of forestry staffthe so-called Community Forestry Assistants (CFAs) were to offer technical assistance and advice at the panchayat and village level. Special attention was lent to providing a system of information and extension materials for communication and training at the grass-roots level (Arnold and Campbell, 1986: 439-40).
The process of establishing and operationalizing a PPF involves three important steps. First, the panchayat requests the government to hand over an area as PPF and the transfer is effected. Second, a panchayat forest committee is created and made to function (experience shows that it is important that committee leadership is kept separate from panchayat leadership). And, third, a management plan is drawn up which constitutes the legal document attesting the agreement between the government, the panchayat, and the user groups involved. It is the responsibility of the CFA to engage 'in a continual dialogue with the users, panchayat officials, and forest committee to arrive at a management system that best meets their needs'. The expertise of the CFA is especially important given the fact that the great diversity of forest types makes it difficult to lay down general management prescriptions that can be widely followed. As for the committee, it must not only support and supervise all community forestry activities but also ensure equitable distribution of products from the PFs and PPFs to all households in the beneficiary group (Arnold and Campbell, 1986: 440 2).
As hinted above, officials' fears have not been vindicated. In the rare instances where PF plantations have been destroyed, the community's behaviour has always resulted from the belief that the government intended to usurp their forest. This suspicion actually followed from the fact that when the area was surveyed the local community did not know that the purpose of the survey was to transfer the area to the local people. These exceptional cases excepted, if we believe Arnold and Campbell, the co-management approach followed by the government of Nepal since the late 1970s has been largely successful even though too ambitious initial targets have often caused achievements to lag behind expectations. The response of villagers was particularly enthusiastic when information has been well diffused, public discussion of the issues involved has been widespread, and benefits as well as responsibilities have been well specified by product and beneficiary (Arnold and Campbell, 1986: 440, 444-6). According to Arnold and Campbell again, something remarkable was 'how quickly group consensus on the value of establishing a PPF usually materialised when the actual provisions of specific management plans (spelling out group rules for protection, harvesting, and benefit sharing) were brought under group discussion' (ibid. 446). For the same authors, perhaps the most important lesson from the whole experiment is that 'communities themselves will take the responsibility for devising methods for solving the common property problem if they are given sufficient authority, information, and assistance in doing so' (ibid. 449, emphasis added).
A more recent account by the World Bank is much less enthusiastic, however: panchayats 'gave the villages the most degraded lands, which required high investments for restoration and offered only delayed benefits'. According to this report, such an outcome is not surprising inasmuch as panchayats are 'large administrative units with little previous involvement in forestry' (world Bank, 1992: 143). Colchester has also recently expressed the opinion that the programme 'has not been without problems', particularly because the panchayat 'is too large a unit and too far removed from day to day decisions to effectively supervise and manage local forests' (Colchester, 1994: 90). As a consequence of this situation, the World Bank actually decided to support 'efforts to encourage management by smaller groups more closely associated with particular forest tracts and to give them responsibility for forests in good condition, as well as for degraded land' (World Bank, 1992: 143).
At this juncture, it is worth noting that co-management contracts can be struck between government and rural communities which are not focused on any single resource but encompass all of the resources located within the area of a given community. This more comprehensive approach has been adopted by the Government of Burkina Faso under its Programme National de Gestion des Terroirs Villageois (PNGTV).
The PNGTV operates in four stages described by Toulmin as follows:
Unfortunately, it is too early to assess this Programme as most of the pilot villages are still in the first two stages. Toulmin is nevertheless of the opinion that 'as a whole, the programme seems well thought out and a reasonable starting point for establishing local systems of resource management', all the more so as it 'includes as an important element, a programme of investment and resource improvement aimed at raising the productivity and sustainability of resources within the village lands' (Toulmin, 1991: 28).
In Chapter 12 (sect. 3), we briefly described the extremely instructive experience of the Gal Oya irrigation project in southern Sri Lanka. As it has been designed by the ARTI-Cornell team, this project can actually be viewed, at least in part, as an experiment with a co-management approach. This said, it is again Japan that is the best example of the co-management approach in irrigation matters. The Land Improvement Districts (LIDs), which are 'the culmination of historical experience of irrigation and agricultural development through co-operation through different eras', are the central institution governing water management in this country (Mitra, 1992: A-78). True, the LIDs are a much more complex affair than the former water users' organizations mainly because they follow a comprehensive approach to land development (including irrigation and drainage projects) with the direct support and close assistance of national and prefectural governments. Yet, basically, they are nothing but farmers' organizations, constituted as juridical persons, with irrigation associations (each of which is organized on the basis of a village) as their substructure (ibid. A-7880). In the following, we would nevertheless like to draw attention to another institutional approach to resource management which is much more centralized, yet does not fail to mobilize the efforts of resource users through various operational and organizational procedures.
To illustrate this bureaucratic approach to participation (this almost self-contradictory expression has been chosen to convey the rather paradoxical character of the organizations concerned), we refer below to two insightful and detailed studies of the management of irrigation systems in East Asia, particularly in Taiwan and South Korea (Wade, 1988b; 1990; Moore, 1989). These two countries are especially worth studying given their high rates of success in the management of irrigation water: Taiwan is thus considered a country possessing one of the world's most technically efficient irrigation systems, a remarkable achievement in view of the fact that most of its water control systems are gravity-flow systems without much storage (Levine, 1980).9 Worth pointing out is that, in the first of the aforementioned studies, the author was actually motivated by the desire to understand the reasons for India's poor performance with respect to irrigation management. Comparative institutional analysis was considered a fruitful way to get an answer to that question. As is evident from the account given below, the exercise was indeed quite conclusive.
It must first be noted that, in countries like South Korea and Taiwan, management of irrigation water is in the hands of catchment-based parastatal agencies which, despite their pseudo-democratic labelling (irrigation associations in Taiwan and farmland improvement associations in South Korea), do not involve formal mechanisms of accountability. In the two countries, an element of authoritarianism is undoubtedly present. The following statement, made with respect to Taiwan, is sufficiently clear in this regard:
Local Farmers' Associations and Irrigation Associations maintain, despite their formal nongovernment status, 'security' departments staffed by central Party or security personnel but financed by the Associations themselves. The utility of the Farmers' and Irrigation Associations as strategic institutions from which to maintain political surveillance at local level was enhanced by the policy towards 'private, associations pursued by the KMT in agriculture as in all spheres of public activity. Only officially recognised and registered associations may function. Those which are recognisedand closely controlled politicallyare given de facto monopolies. Where organisations do not exist, they are created on state initiative as a preemptive measure. The only farmers' organisations officially tolerated are the Farmers' and Irrigation Associations and the marketing 'cooperatives' mentioned above. (Moore, 1988. 132)
Against this background, a central lesson from an analysis of the functioning of irrigation associations in both Taiwan and South Korea is that even centralized organizational structures can perform well if they follow appropriate operational and organizational procedures that have the effect of establishing trust between farmers and irrigation staff and within the irrigation hierarchy itself. This is precisely where, according to Wade, the East Asian type of organization has proved so much superior to that found in South Asia in general, and in India in particular.
A key feature of the East Asian type of organization for water management is that methods are used with a view to creating a sense of common purpose and corporate identity within each irrigation agency (IA). This result is achieved to a large extent by obscuring the contractual nature of the employment relation between the IA and individual staff members, by setting objectives that can be shared by all members, by providing stable employment to officials, and by fixing pay scales which are not closely tied to hierarchical rank (so as to avoid 'the ingrained conflictualism' which is found between lower and higher staff in Indian irrigation departments).
A second feature, more directly relevant to our discussion in this chapter, is the high degree of staff involvement with farmers. In South Korea, for example, the lack of farmer participation in irrigation system management 'is partly offset by internalising farmers at the bottom of the formal management hierarchy itself, in the role of patroller'. At the same time as he is thus at the lowest level of the organizational structure, the patroller must be 'a farmer with land within the jurisdiction which he irrigates, so that he experiences irrigation problems at first hand' (wade, 1988b: 495). Note that he is nominated each year by the headmen of the villages within his jurisdiction and that, if the latter are unsatisfied with his way of handling the task, they nominate someone else. Most staff of the [As, even (male) clerks are expected and inclined to stop and chat with farmers whom they meet along the canals (for which rides they are provided with small motor cycles). In the words of Wade: 'The intensity of this local contact helps further to make up for the paucity of more formal channels of communication between farmers and staff, which is an expression of the authoritarian character of the South Korean political regime' (ibid.).
Recruitment and promotion procedures also play an important role inasmuch as they ensure that the senior-level staff are natives of the area in which they work. 'Hence the eyes of the irrigation staff are kept firmly on the locality, and identification between their interests and those of farmers is further encouraged' (Wade, 1988b: 495). Elaborating on this theme, Wade adds:
Local affiliation of the staff is important because it gives both sidesstaff and farmersa set of shared experiences. This directly assists a sense of mutual obligation between them; and also provides a basis for a shared set of beliefs according to which the existing order is fair and just, and every betrayal is perverse and unjustincluding betrayal of the irrigation agency's rules. This is a much more cost-effective method of avoiding free-rider problems than relying on a calculus of punishment. (Wade, 1988b: 495)
Much the same picture emerges from the situation in Taiwan where the staff of the irrigation associations are effectively linked to the local farmers on the one hand and to the national agencies on the other. For each rotation area, an irrigation group chief is elected to supervise water distribution and maintenance operations as well as to manage potential conflicts. In particular, they are in charge of closely monitoring the jointly hired common irrigators who have primary responsibility for the distribution of water and the guarding of the system against fraud and damage. As a matter of principle, these chiefs are local farmers and, to avoid undue interference of partisan politics, the process leading to their election is kept separate from elections for other offices. In addition, the irrigation staff themselves are typically recruited from local communities so as to ensure adequate incentives for effective management of the system. In the words of Moore:
The IAs are overwhelmingly staffed by people who were born in the locality, have lived there all their lives, and in many cases farm there. Further, IA staff are not sharply differentiated from their members in terms of education or income levels. I have a strong overall impression that IA staff are so much part of local society that they can neither easily escape uncomfortable censure if they are conspicuously seen to be performing poorly at their work, nor ignore representations made to them by members in the context of regular and frequent social interactions. (Moore, 1989: 1742)
There is obviously a direct parallel to be drawn between the above diagnosis and the finding reported in Chapter 6 according to which norms of reciprocity can be established by fostering communication in such a way as to transform other agents 'from mere strangers into real people'. The situation in India offers a striking contrast to that obtaining in East Asia since, in the former region, irrigation officers 'are normally rotated in and out of any one post every 18-24 months, and have no identification with the area of their responsibility' (Wade, 1988b: 495; see also 1982). Note that such high rates of mobility not only create low incentives for the effective management of irrigation systems, they also preclude any learning process since staff members are shifted to new assignments before they have become familiar with the setting in which they operate (Weissing and Ostrom, 1992). Moreover, the position of a 'guard' is usually held by a labourer who is directly supervised by an engineer, and both are part of the Operations and Management Division that holds a minor place in the (large) state-level irrigation department.
Equally worthy of note is the fact that in East Asia the staff have a direct sense of dependence on the prosperity of farmers under their own system because the IA has its own revenue base (the water fees collected from the farmers) from which it must meet most of its operating costs (Wade, 1988b: 494). Thus, in Taiwan, fees paid by farmers account for one-third to three-quarters of the financing of the IAs (Moore, 1989). In the same connection, another feature that deserves to be strongly emphasized is the importance attached, even by higher-level authorities, to expeditious payment of irrigation fees by water users and the interpretation of any delay as a sign that some problem has arisen that needs to be addressed. As Moore puts it:
During the two periods of the year when fees are due, most of the IAs' institutional machinery appears to be devoted to completing collections as expeditiously as possible. The same working station staff who provide farmers with irrigation services come to them to encourage them to part with their money if there is any sign of delayed payment. Each working station is required to make daily telephone reports to superiors about collections in their area.... And the headquarters are obliged to report regularly to the Provincial Water Conservancy Bureau. At each level, delays in fee payment are taken as prima facie evidence of a problem which requires attention. (Moore, 1989: 1743)
This interactive process is actually part of an institutional structure that makes all staff members at every level subject to detailed annual job performance evaluations. Great concern is shown about these evaluations as they result in ratings which directly or indirectly affect salary increments, promotions, and access to additional resources (Moore, 1989: 1743). Furthermore, by pressuring relatively autonomous irrigation associations to perform well (for instance, in terms of their speed of fee collection), the Provincial Water Conservancy Bureau tends to stimulate competition among them and to link them together by passing information on best practices and rules (Weissing and Ostrom, 1992). All this is a far cry from the situation obtaining in India where the field guards employed by the irrigation department are supervised by a staff that has practically no incentive to make the system work effectively. The water users' bargaining position is made low to the extent that they cannot affect the performance of the field guards by refusing to pay irrigation fees or by delaying payments. Indeed, since these fees go to a Revenue Department rather than to the division in charge of irrigation operations and maintenance, whether water users pay their fees or not does not make any difference to the material situation of the field guards.
Wade rightly lays stress on the fact that effective resource management by a bureaucratic agency depends on the perceived legitimacy of the irrigation staff's authority. In turn, this legitimacy 'depends on farmers' judgment of how competent the staff are and on how much they trust their good intentions'. To the extent that competence is difficult to assessparticularly so when the system is big'trust becomes the crucial factor' (Wade, 1988b: 497). What needs to be added now is that the amount of required trust can be reduced by designing physical infrastructures in such a manner that farmers' dependence on the allocation decisions of officials is significantly diminished. One design feature which is advisable from this standpoint is the 'online' or 'breakpoint' reservoir since it provides 'a clear hand-over point where the officials' jurisdiction ends and the farmers' jurisdiction begins'. The important thing, here, is that the reservoir be 'at a low enough level for farmers to see the stock of water which is "theirs", for water to be able to reach all the fields within a few hours, and for farmers to have a large (legitimate) hand in how it is allocated'. Another promising system is that experimented in the state of Gujarat (India). Under this system bulk amounts of water are sold to a tertiary distributarythat is, to all the farmers dependent on this distributory as a unitand the farmers themselves are entrusted with the responsibility of organizing the distribution of water and the collection of the fee (ibid. 496-7). This is in sharp contrast to many government-owned irrigation systems that have designed physical works without any concern as to how guards or farmers could observe activities at a low cost. Further, in most of these schemes (including those in India), the guards hired by a government agency are often given vast areas to cover and, for this purpose, they are not provided with even a bicycle, let alone a motorized vehicle, to travel to the canals themselves. (Weissing and Ostrom, 1992)
For a system of water management to work effectively, it is not only necessary that staff members have adequate incentives to perform their assigned tasks diligently, that user communities are made an integral part of the management system, and that physical infrastructures are properly designed so as to minimize the cost of fraud detection. It is also essential that a high degree of flexibility and autonomy be granted to decentralized management units so as to enable them to best adapt to local circumstances and variations. This is a point that Taiwanese authorities seem to have well understood. As a matter of fact, the Yun-Lin Irrigation Association which covers an area of 65,590 ha. of land has four regional management officers, forty-three working stations, 500 water groups, and 1,683 subwater subgroups based on rotational areas. What requires to be stressed is that each of the numerous (over thirty) systems in use in this command area has its own fee schedule (some systems have even more than one); 'each has a high degree of autonomy with respect to system operation, maintenance and improvement; each has specific water rights, usually based on historical development, which are respected in irrigation planning and operation, though they may be modified under emergency conditions' (Levine, 1978: 3, cited from Weissing and Ostrom, 1992).
In a recent paper in which the Indian and Japanese systems of water control and allocation have been compared, Mitra has argued that many efficiency losses and equity problems in the management of water in surface irrigation in India can be traced back to an administrative system 'devoid of organisational structure requiring people's participation' (Mitra, 1992: A-78). What is needed in India, and where India can learn from Japan, is a 'joint management' approach in which the Command Area Development Authorities (CADA) - an institutional innovation ushered in the mid-1970s with a view to improving farm-level irrigation water managementwould involve farmers in their development and management programme. This approach is not feasible, however, unless the attitude and perception of the irrigation departments and other state agencies like CADA undergo 'a sea-change' in order to enlist the active and willing co-operation of farmers (ibid. A-82). As a matter of fact,
The farmers' involvement and participation will not be forthcoming easily if the CADA is continued to be seen as a government programme imposed from the top. Under such circumstances farmers would not see CADA as a programme meant to benefit them or worthy of support. To ensure farmers' support, irrigation associations of the type of LIDs in Japan will have to be organised such that farmers are given the responsibility of irrigation management which brings about principle of equity as regard farmers' right to water. (Mitra, 1992: A-80)
The joint management approach envisioned by Mitra is one in which CADA would actually work as a liaison and catalytic agent between irrigation departments on the one hand and irrigation associations on the other. The latter would assume the responsibility for the distribution and utilization of the water at the tertiary level, maintain properly the distribution networks under their control, collect water dues from the members for payment to state authority, and adjudicate local disputes and resolve local conflicts. The effective working of such associations will admittedly take time since India does not have the background of Japan which has a very long history of water users' organizations (Mitra, 1992: A-81). In view of this lack of historical precedent in India, one may none the less wonder whether adoption of an organizational form more akin to the rather centralized pattern found in Taiwan and South Korea would not be more successful than emulation of the Japanese model in which people's participation has undoubtedly been more complete and the principles of co-management are more strictly or genuinely abided by.
Finally, it is of interest to note that a highly vulnerable point in many people-based irrigation projects is maintenance of the infrastructure: 'a pattern of neglect followed by repair and restoration is much more frequent than routine preventive mainten ance'. This may come as a surprise since the conventional wisdom has it that if farmers feel they truly 'own' a system, they will take good care of it (Bruns, 1993: 1843). A plausible explanation is that irrigators have a strong incentive to delay maintenance action until there is a need for major repairs and rehabilitation. This is because they usually have to pay the full costs of routine repairs while the government pays fully for major repairs and rehabilitation. Current policies regarding government assistance thus create perverse incentives that do not encourage farmers to be diligent about maintenance. Requiring local cost-sharing, even for relatively large repairs and improvements, would be a major step towards solving this important problem (ibid.: 1844-5).