Agricultural Diversification and People's Irrigation Project in the North (1994) - IOE
Agricultural Diversification and People's Irrigation Project in the North (1994)
Mid-term evaluation
Project design and objectives
Country context. During the past three decades, Thailand has enjoyed a consistently high rate of economic growth which has pushed it into the group of middle income developing countries, with a per capita GDP in 1991 of USD 1 650. However, the agricultural sector has been contributing a steadily decreasing proportion of the GDP, going from 32% in 1965 to 16% in 1986 to 13% in 1991, though it still accounts for around 30% of total exports by value. In terms of labour employment, agriculture accounted for 60% of the total labour force in 1992 compared with 75% in 1986. In spite of this relative trend, agriculture continues to be the basis of livelihood for the majority of the population and to play a very important role in the Thai economy.
Poverty incidence in Thailand has diminished steadily over the last decades, from 60% in the 1960s to 18% in 1990. While diminishing overall, however, it is increasingly localized in rural areas and especially in the agricultural sector. Farmers are the lowest income group, with 22% below the poverty line (reaching 29% in the North East). Income disparities are also significant within the rural population, with the lowest 40% of farmers in each region earning less as a group than the highest 5%.
The project is located in Thailand's Northern Region, and project activities were concentrated on the upper north, which was a priority area also for the government Poverty Alleviation Program. The SAR states that about 64% of the population in this area lived in absolute poverty, defined as a per capita annual income of less than USD 148.
Project rationale and objectives
The Agricultural Development and People's Irrigation Project in the North (ADPIP) is to establish a programme for technical and institutional strengthening of People's Irrigation Associations (PIA) systems and to promote inter-PIA cooperation for Operation and Maintenance (O&M) of subproject irrigation systems. It is to increase the income and improve the well-being of the poorest segments of the rural population through improved production efficiency and agricultural diversification. An important element is the beneficiary participation in the planning, financing, execution and O&M of the proposed irrigation improvement works. The government is expected to preserve the independence and autonomy of the PIAs due to their importance for strength and stability of rural production in the Northern Region.
The project, with a total investment of USD 18.3 million, is to finance the first phase of a long-term programme which would improve irrigation for about 6 000 ha through the construction of small dams, and the improvement of some existing PIA weirs and canal systems. The principal components are: (i) Irrigation subprojects (USD 11.3 m), comprising construction and improvement of small irrigation systems and establishment of additional Mobile Campaign Units (MCUs); (ii) Pioneer Watershed Component (USD 1.1 m) where the project is to develop an integrated whole basin approach; and (iii) Institutional development and technical support (USD 3.1 m).
Project financing came from an IFAD loan of USD 10.0 million, a loan from the OPEC Fund (USD 3 million), and from a government contribution of approximately USD 5.3 million (of which USD 2.3 million was the estimated in-kind beneficiary contribution).
The project was considered a first-phase pilot operation in an effort to establish long-term assistance to PIAs. Lessons learned from previous IFAD financing in Thailand had pointed out the need to link project activities to existing farmer groups. It was assumed that by focusing the project on PIAs, an indigenous form of efficient grassroots farmer organization would be strengthened. RID would be able to effectively collaborate with PIAs, and that PIAs would be able to preserve their autonomy and efficiency with project intervention. These were, however, cited as project risks in the SAR.
Decentralization of project management and close collaboration between several different Ministry of Agriculture and Cooperatives (MOAC) departments were key elements of project design, and it was assumed, in spite of a highly centralized system of government administration, that this system of project management could work. However, this was also cited as a project risk in the SAR.
The investments in dams, weirs and canal improvements would, as mentioned above, secure the water supply to irrigated areas. By increasing production and thereby raising income, it was assumed that there would be a decrease in cultivation of upland areas by irrigated farmers. This was considered an unquantified benefit of the project.
Reprogramming. At appraisal, it was agreed that the cost of construction would not exceed "about USD 2 000" per irrigated hectare. However, during the initial stage of implementation the feasibility studies carried out showed that this condition had to be modified due to significant cost increases. In 1989 it was thus agreed to consider an average figure of USD 2 850 per ha to be acceptable within the existing language of the Loan Agreement. It was decided to concentrate on the five least expensive subprojects and stop further exploratory work on the more expensive subprojects. The change from the 13 subproject areas of the original plan to only five which were actually implemented, substantially reduced the total area irrigated by the project.
This cost increase was also reflected in a reallocation of the IFAD loan proceeds among the financial categories in 1992, mainly in favour of dam construction costs. Project implementation was delayed by about two years due to shortage of counterpart funds, shortage of construction materials such as cement and steel and escalation of construction costs. As a result of this delay in implementation, the project received a one-year extension of the loan closing date to 31 December 1994.
Overall, ADPIP can be qualified as a successful project and there is no doubt of the project's popularity with the beneficiaries, technical (if not financial) soundness, and reasonable level of management. PIAs and farmers are clearly satisfied with these developments, as well as with the formalization of the PIA structures. Data on changes in yields of the most important crops are incomplete and sometimes contradictory, but there seems to be a modest increase in yields for most of the crops. It was estimated that total household income of irrigation beneficiaries would increase by 26%. However, the degree accorded to this overall project success depends on how one chooses to read and interpret the project's original objectives and goals.
The question that arises in ADPIP regards the existence of two alternative views of the basic nature of the project. Is it a people-centred, participatory project mainly intent on working with PIAs to improve their existing irrigation systems and supplying them with training and support as well as dry season water, or is it an infrastructure development project whose goal is to build high quality dams at a reasonable price? Is the main project objective concerned with the concept of grassroots human development, or rather with the successful completion of large engineering works? The answer to these questions apparently depends on who is being asked. According to IFAD, the emphasis is on the first, while according to the Royal Irrigation Department, the principal implementing agency, it would rather be on the second. As will be seen below, this has certain implications when assessing the degree of project success.
Irrigation. The major development objective of ADPIP was to establish a programme for technical and institutional strengthening of PIAs and their irrigation systems. In pursuing this goal, a contrast of methods and approaches emerged between IFAD on the one hand and the World Bank and RTG on the other. What began as a people-centred, institution-building, small-scale irrigation and crop production project gradually became a project which would spend nearly 80% of available funds on large engineering works in the form of medium-sized dams aimed at ensuring dry season water availability to PIAs.
The dam construction activity has been very successful, in spite of spiralling costs (from USD 2 000/ha to USD 3 620/ha), shrinking area of coverage (from 6 000 ha to 3 876) and number of beneficiaries (from 8 650 households to 4 174). At the time of the evaluation, three of the five subprojects included in the project, Mae Kham Pong, Mae Phrik and Huai Dua, were complete, and the first two had been operational since 1992. At Huai Dua, reservoir filling was awaiting clearance from Royal Forestry Department (RFD) for submergence. Of the remaining two dams, Mae Kon was 90% complete and Mae Laeng Luang nearly 60%. Weir replacement and weir and canal improvement have been completed as planned. Supervision missions have indicated satisfaction with the performance of the engineering companies contracted for execution of the irrigation investment works, especially in terms of quality of the work done.
Subprojects were to be implemented by RID and the concerned PIAs as a unit. Local PIA leaders and farmers were to be consulted and to agree jointly to undertake their subproject, to request it officially and to be directly involved in implementation, as well as contributing 20% of capital costs in kind (labour) or in cash. It was observed that RID has not followed this approach during implementation. Although PIAs were involved from the beginning, this involvement was limited mainly to the formation of registered groups and the establishment of agreements for the management of water after dam completion. Very little PIA involvement appears to have taken place during construction of the dams and concrete weirs. Almost no PIA contribution to capital costs in any form was found to have taken place, with limited labour contributions being the only exception.
The key stumbling block of many dam projects, the water distribution and management system, is already in place through the PIA system, and this is the key factor which makes ADPIP such a promising project. The capacity to divide and keep track of water, to prevent and punish theft or abuse, to carry out accurate accounting of water credit and debit, to share in a way that allows all users to maintain a reasonable agricultural calendar, etc., are all present and applicable to the new system in much the same form as to the old.
In the view of the interviewed farmers, the project is very positive, the expected benefits are clear, the costs to them are low and there are virtually no risks. Production (and incomes) will surely rise as dry season water becomes available, and crop diversification for the dry season cultivation is being encouraged by the Department of Agricultural Extension (DOAE).
Thus by RID criteria the project is very successful. The availability of dry season water will improve irrigated farmers' living conditions and raise area agricultural production. From the point of view of the "people-centred" project, the project has been successful in terms of PIA strengthening and institutional development despite an approach quite different from that prescribed in the SAR. The very fact of formalizing the PIAs and organizing them into Inter-PIA Committees has apparently given significantly greater importance to the existing associations in the eyes of their members, strengthening cohesion and effectiveness.
However, the question of the PIA "sense of ownership" of project engineering works remains uncertain. This was to be given to PIAs through their involvement in preparation and in investment costs through labour contribution, and especially through the handover of complete responsibility for O&M of all irrigation works following completion, including dams. Instead, there was no PIA participation in subproject selection and design, very little PIA contribution to construction, and RID has stated its intention of keeping O&M of large project works under its own control.
ADPIP was intended apply participatory development through the use of PIAs. However, this focus on people shrank as the project progressed towards implementation, to the point of becoming only a minor aspect of an infrastructure project whose major objective was dam construction. While people's participation cannot be said to have been a major aspect of the dam construction activity, the institutional development component as applied to the PIAs appears to have been conducted with extensive interaction between the project implementation staff and the leaders and members of the PIAs. The very acceptance of the indigenous water management system as the basis for water management after dam construction is in itself a form of involvement of beneficiaries. Even this, however, is not evidence of genuine people's participation, which implies beneficiaries having a decisive role in project selection, design, implementation and evaluation. The relationship here still tends to be one of donor and passive recipient rather than facilitator and activist farmer groups. More time will be necessary before real participatory development has a chance to be tried with PIAs in the upper North.
Thus it would appear that when taking PIA support and institutional development as the first priority for ADPIP, success is tempered by the loss of opportunities for further strengthening of PIAs. While it was clear that RID worked hard at supporting PIAs, the overall non-interventionist intention of the project irrigation component was not being effectively applied by RID. Dam and concrete weir construction was an engineering activity isolated from the control or participation of PIAs, and there was never any clear intention on the part of RID to involve PIAs in the planning or execution of these interventions. O&M of weirs and canals will apparently be partly shared with PIAs, but dam O&M is to be exclusively the responsibility of RID, as it always has been in previous projects of this scale. In addition, a precedent was meant to be set by this project for the handover of irrigation works to water-users' associations on future projects, but this was not done. Thus the project set rather a negative precedent which seems to indicate that in spite of donor insistence, RID is not yet ready to transfer control of its irrigation works to private users.
Inter-PIA Committees have been set up in all the subproject basins under the guidance of the MCUs, and they are reportedly meeting regularly in anticipation of project dam operation. Reactions from both leaders and members of these new organizations were unanimously positive regarding their usefulness and their potential effectiveness.
Under the project, the system of MCUs would be strengthened. These multidisciplinary teams would link PIAs/farmers, RID and other agencies, and provide technical assistance and guidance to PIAs. However, although the SAR called for six new MCUs to be set up, as well as the inclusion of DOAE and RFD staff in all MCUs, not one new MCU was set up under the project, nor has it been possible to include other agencies in these RID units (although collaboration between MCU staff and local staff of the agencies has reportedly increased). The result is that project PIAs are receiving much less support and assistance than foreseen at appraisal.
Sustainability of irrigation interventions will depend mainly on the availability of RID funds to continue O&M of larger project works and continued interaction of PIAs in sharing reservoir water. The irrigation construction part of the project has a high degree of replicability in the context of existing PIA systems in northern Thailand. The degree of replicability of the PIA model of farmer-built and managed irrigation systems is more questionable, but some aspects of PIA organization could no doubt be applied irrigation schemes were they do not exist.
Crop diversification. This component fits in well with both existing DOAE activities and current farmer practice: when dry season water is available, farmers seek the most profitable crop to plant. In the wet season, the vast majority still plant exclusively rice in the lowlands, but on hillsides they are willing to grow anything which will increase their incomes, and are therefore quite open to diversification as it may be encouraged by the project. The confusion and unreliability of the figures available on crops and yields makes it difficult to judge the extent of diversification, but it was clear that farmers will respond more readily to market indications than to research results in their choices. Either way, dry season cropping will continue to move toward greater diversification according to comparative advantage, and DOAE seems ready to support farmers in their search for the most advantageous crops.
The Pioneer Watershed Component was to be designed following a "whole basin approach" which would integrate activities in watershed forest areas, under exclusive jurisdiction of RFD, with activities in the irrigated valleys which were being entirely managed by RID. RFD was given principal responsibility for implementing the Pioneer Watershed Component. Clear guidelines were set out in the SAR for the interagency coordination. It was assumed that RFD and RID had the ability to work together to carrying out a unified intervention in the whole river basin. RFD was to be the lead implementing agency, but the Department of Land Development (DLD) was to take the lead in planning the component because of its expertise in land use planning. It was also indicated that RFD staff were to help PIA leaders tackle watershed management issues. This participation never took place to any significant degree, and the positive link was never established.
RFD kept the watershed component very much in its own hands, operating it in near isolation, directly out of Bangkok under its own budget line from the loan. Planning of the component was given to the Project Consultant who presented a component that was far too costly. In the end, RFD implemented only the part of the project which fit in most easily with its previous experience: reforestation with teak along with the building of forest Field Stations within the watershed (1 000 ha). Some other soil and water conservation techniques and planting of fruit trees on sloping land was encouraged. There was no integration of valley-dwellers into the watershed management activities, and no coordination between them and the hill-dwellers, neither were PIAs involved. The participatory, people-oriented spirit of the design was lost during implementation. Unfortunately, the Mae Lai Watershed represents a situation that is almost unique in the Northern Region, wherefore few lessons from Mae Lai will be applicable to other ADPIP subproject watersheds.
Institutional Strengthening. Subproject preparation as well as financing related studies and pilot projects by the consultant engineers was part of the institutional development component. The work was reportedly carried out skilfully by the Project Consultant, but it remains unclear why this activity was considered to be part of the project's institutional development component. The only institution to derive a clear benefit from this part of the loan was the Project Consultant.
Training progressed well and both farmers and officials expressed their satisfaction with the training they received. To date, the training programme has never undergone an evaluation, and the MTE recommended that such an evaluation be carried out.
The project was intended to finance costs of studies, pilot projects and other related research. In close collaboration with the International Irrigation Management Institute (IIMI), studies were to establish an action-research programme to monitor the workings of the PIAs with reference to their enlarged role under the project. However, no contact was established with IIMI and an interesting opportunity for international collaboration in monitoring the workings of the PIAs was therefore lost. Only by 1993 were two studies finalized and judged satisfactory by the CI, one on cash crop development and one on women.
A critical aspect of the project, is the institutional support and development of the PIAs in the context of newly secure wet and dry season irrigation water supplies resulting from the project. The project has also been successful in strengthening the independent PIAs and in promoting inter-PIA cooperation. But the question of the PIA "sense of ownership" of project engineering works remains uncertain. Support will terminate just at the time when the project activity of support and assistance to the PIAs would become most important, i.e., when the dam construction is finished and water begins to arrive. Thus, ironically, the project ends at the moment when IFAD's catalytic role in encouraging RTG/PIA collaboration and strengthening PIA performance should become most important, and the further support to PIAs is left in the hands of a government department (RID) which has often been more interested in engineering than human aspects of water distribution.
The project has served to develop stronger links between RID and PIAs, this appears to be having a beneficial and invigorating effect on these organizations, in spite of SAR predictions to the contrary. PIAs have been quick to adapt to changing circumstances. The attention which they have received from the project, as focal points for all project intervention and information regarding new reservoir facilities and agricultural diversification, has given them a new authority in the eyes of their members and other villagers.
Monitoring and Evaluation (M&E). Although some attempts were made to formalize the M&E of the project, none of the proposals were finalized and no separate M&E unit was ever set up as foreseen in the SAR. This meant that the responsibility for project M&E remained with the Office of Agricultural Economics (OAE) in RID, which already had responsibility for M&E of other projects. In spite of hard work on the part of the responsible staff, this situation has significantly handicapped project M&E activities.
Progress reports were furnished throughout the project implementation period, although with occasional delays. To improve this situation, the CI proposed some simplifications in the requirements and standardization in reporting tables. Although this proposal resulted in a more regular stream of progress reports, simplifying the requirements for reporting may have had an adverse effect on the quality of the information provided to the CI and IFAD.
The OAE prepared baseline studies for the subproject areas. The Supervision Mission (SM) judged these to be insufficient and requested OAE to provide information on the methodology used for the surveys. It was also concerned that pre-project irrigation areas may have been overestimated and with-project area underestimated. There is no evidence that these reports were resubmitted and some of the defects identified at the time are also found in the incremental benefit studies and the evaluations of subproject areas that were subsequently produced.
As a follow-up on the baseline studies, further studies were carried out by OAE on cash crop income and on farm income. Evaluation reports for each completed subproject are were produced by the OAE. The reports are not always mutually consistent and conclusions are based on a string of assumptions that are not always realistic. This points to a lack of fieldwork, and an overdependence on secondary and tertiary data.
No provision was made for beneficiary participation in project evaluation exercises. The household surveys are not geared to subjective assessments of project activities on the part of beneficiaries, nor has any systematic analysis of farmer perceptions of and attitudes toward ADPIP activities been undertaken. Furthermore, the evaluation reports do not provide an analysis of who benefitted or whether benefits reach the target group. However, it should be mentioned that none of this was requested in the SAR nor by the CI or IFAD.
RID should request OAE to continue monitoring of the subproject areas even after the loan is closed and the project ends because project benefits only begin to accrue after the termination of construction, and thus after project closing. In monitoring benefits from the project, more attention should be paid to off-farm income which is also important in understanding the opportunity cost of the extra labour needed for double or triple cropping made possible by project-supplied dry season water.
No targeting criteria were incorporated into project activities. Farmers in the project area were assumed to be generally poorer than the national average, and the criteria used to select subprojects to be financed were based almost entirely on technical issues. The particular watershed was selected because of technical criteria relating to degree of degradation and extent of illegal logging. Presently, there is very little apparent poverty in the subproject areas. For targeting of the poor IFAD should incorporate targeting mechanisms into its projects which are able to adjust with changing poverty distribution in the project area during the life of the project.
The issue of socio-economic differentiation was not touched upon during project preparation and appraisal. As the irrigation project was designed, direct benefits accrue only to members of the PIAs, and to a small extent to the families of labourers hired locally by the contractors for dam construction. The watershed component also produced direct benefits almost exclusively for the local labourers hired for reforestation. Some benefits to the general agricultural population can be expected from increased extension activities and encouragement of diversified agriculture.
The MTE mission's limited exploration of the issue in the field suggested that while there is an evident general situation of well-being and prosperity in all the project areas, a significant amount of differentiation exists. There is a sizable landless population in some of the subproject areas, which was reported as high as 20-30% by some village chiefs. These landless families generally either rent land (the most common case) or work as agricultural labourers on others' land or as wage labourers in town. This last group, apparently includes the poorest families in the village, and these are also the ones which are not involved in the PIA system and are therefore excluded from direct project benefits. The clearest indirect benefit should be an increase in dry season agricultural labour opportunities, which the SAR estimated at 18 600 man-months per year. However, in general, the improvements which the project irrigation and agricultural development will bring to landowning families will largely leave these families behind.
The families which rent paddy land apparently take part in the PIAs in which the land they rent is located. In cases encountered by the mission, these landless farmers were involved in most of their PIAs' activities, however, they do not participate in the election of PIA leaders nor the establishment of rules and regulations. This group would clearly be among the direct beneficiaries of the project. In one way, these farmers will increase their incomes even more from the dams than others: rent is based on rice production in most cases (usually equivalent to half the rice harvest), but does not change if the tenant manages to grow additional crops in the dry season. The new dry season irrigated production will be kept entirely by the tenant.
None of the project activities were specifically targeted at women, in spite of the equal sharing of roles and responsibilities in the house, and the active involvement of women in project activities. It is recommended that future IFAD projects should attempt to build in and monitor mechanisms for reaching women more directly and on a more equal footing with men. When project design specifies such things as people's participation and integration of women into project activities, specific modalities and monitoring mechanisms and indicators need to be included in the implementation guidelines in the appraisal report.
A Central Coordinating Committee (CCC) was set up under the Office of the Permanent Secretary and chaired by the Deputy Permanent Secretary, which was to be responsible for overall policy guidance for the project, coordination between the ten MOAC departments and agencies involved in implementation and supervision of project progress. The Office of Agricultural Economics (OAE) supplies secretariat support to the CCC. This committee has met regularly and meetings have been well-attended. The CCC has been instrumental in the coordination of project activities between the agencies. The main reason for its effectiveness has been the active involvement of the Deputy Permanent Secretary. Occasional problems persisted and the opinion is shared by many of those directly involved that the number of agencies may have been too high. Future projects may wish to concentrate on fewer departments to avoid this difficulty.
The SAR gave overall responsibility for project implementation to the MOAC through the CCC, while RID was to be the lead agency for construction of reservoirs and improvements to PIA systems. During design it had been agreed that it was preferable for RID not to assume coordinating responsibilities. In spite of this, the increasing specialization of the project as an irrigation construction project and the concentration of funds in this department led to a situation where RID did in fact take on the role of coordinating agency. This also included a supervisory role in the implementation of the Pioneer Watershed Component, although main responsibility for this activity was vested in RFD.
The Cooperating Institution (CI) for this project has been the International Bank for Reconstruction and Development (IBRD - referred to in this report as the World Bank, or WB). The CI carried out both pre-appraisal and appraisal of the project and is supervising loan implementation. There have been thirteen supervision missions to the project since inception, including the one which coincided with the MTE. These have taken place at fairly regular intervals and have included an irrigation engineer on ten missions (the same one throughout), an environmentalist or watershed specialist on five, and an IFAD staff member on four (three different people).
In terms of continuity and regularity, the WB did an excellent job of project supervision, particularly concerning the technical aspects of physical implementation of the irrigation component. The frequent presence of environmental management specialists as of April 1991 added a positive input, though it came too late to have significant impact on implementation of the watershed component, which completed disbursement in 1992. The SMs showed flexibility in confronting problem situations, and took a number of decisions relating to implementation. Regarding the institutional side, however, some omissions occurred on the part of SMs.
The role played by the CI on this project also typifies some of the recurring problems which IFAD has encountered when collaborating with the WB or other large international financial institutions. The principal one is that these institutions tend to limit supervision to give insufficient attention to IFAD's specific mandate of poverty alleviation through participatory and sustainable development methodologies. The tendency, is to measure project progress and success only by checking physical completion against expenditure and specification adherence. This has the advantage of encouraging progress monitoring and the relation of expenditure to progress. However, very little consideration is given to such issues as targeting, beneficiary participation, women, organizations, and other human development issues. Even though the major declared aim of this project was to support and assist indigenous farmers' organizations, no social scientist or organizations specialist was included on any of the supervision missions.
To ensure greater IFAD specificity in project implementation, supervision must ensure that project design concepts, in addition to specific components, are implemented at the field level. Greater attention must be paid to monitoring and evaluation, which were unnecessarily weak in this project.
More attention needs to be paid to the qualitative aspects of WB supervision in future, possibly through specification of the specialists to be included on missions or other similar measures at the time of CI selection, and active involvement in defining the TOR for these missions. This would ensure that supervision is sufficiently in-depth to identify problems to be brought to the attention of project management or of IFAD.
Main issues and recommendations
In dealing with both the government and the CI, there is a need to keep a clear emphasis of IFAD's priorities (such as women, participation, poverty, etc.) throughout implementation, to avoid having them diluted or forgotten following appraisal.
More care must be taken in selecting the proper institution for project preparation. In this case, the preparation of the watershed component was done by a consulting company, who had neither experience nor expertise in this field.
In attempting to improve inter-agency collaboration and efficient project implementation, the importance of decentralization of control over project management to regional and local project staff is difficult to overemphasize.
The mission recommends that Mae Lai continue to be treated as pilot area and that efforts be made to develop mechanisms for combining valley and hillside activities such that a single agency or project entity can work in both areas simultaneously. This is essential if the migrating encroachers from the valley are to be actively involved in watershed protection, rehabilitation and conservation. Models for such a system should possibly be sought in the Northeast Region.
Existing supervision must ensure that project design concepts, in addition to specific components, are implemented at the field level. Greater attention must be paid to monitoring and evaluation, which were unnecessarily weak in this project.
RTG should request OAE to continue monitoring the subproject areas even after the loan is closed and the project ends. This is particularly important in the case of ADPIP because project benefits of this type of project only begin to accrue after the termination of construction, and thus after project closing.
In monitoring benefits from the project, more attention should be paid to off-farm income, which appears to be more and more important in farm economies as the general Thai economy expands and accelerates. This is also important in understanding the opportunity cost of the extra labour needed for double or triple cropping made possible by project-supplied dry season water.
To date, the extensive project training programme has never undergone an evaluation, and the MTE would recommend that such an evaluation be carried out in the near future to (i) ascertain the effectiveness of the training offered, and (ii) to identify further training needs.
Lessons learned
The dramatic cost increases which took place point to the importance of taking the necessary time during project preparation and appraisal to minimize oversights and approximations. In this context, IFAD should avoid putting excessive pressure on the appraisal process for the artificial time limit imposed by the desire to "get the project to the Board" at the earliest possible date.
IFAD has a recurrent problem with the large international financial institutions, in their role as Cooperating Institution, of receiving insufficient consideration of its very particular mandate and development specificity. It is worthwhile and important to give thoughtful consideration to selecting the best possible CI, especially if the project is innovative.
It is important to avoid simplistic approaches to an issue as complicated as beneficiary contribution to capital costs of a capital-intensive construction project. Assuming that large amounts of unskilled, "voluntary" group labour can be easily integrated into this type of activity is risky, and even more so is the assumption that farmers can be convinced to contribute out-of-pocket cash to pay for something like dam construction.
Similarly, it is optimistic to assume that a government department like RID, which has always maintained control over its investments, will be willing to hand over control of dams which cost them several million dollars, to farmers' groups over which they have a very limited control. This kind of attempt, to hand over full responsibility and ownership of project works to the beneficiaries to promote sustainability, should be carried out with smaller-scale investments to start with.
Targeting of the poor should not be assumed to take place from general figures of poverty levels in an entire region. IFAD should incorporate targeting mechanisms into its projects which are able (within limits) to adjust with changing poverty distribution in the project area during the life of the project. In the present case, poverty has hardly been an issue in the beneficiary population, so that most of the project efforts have been directed at helping the better-off members of the community (those with irrigated land) rather than the poorest, and no effort was ever made to look for the poorer PIAs in the region.
When project design specifies such things as people's participation and integration of women into project activities, specific modalities and monitoring mechanisms and indicators need to be included in the implementation guidelines in the appraisal report, just as they are for dam construction. Otherwise the chances of genuine implementation are very slim.
Collaboration between RID and RFD is essential for effective watershed management in situations where it is the farmers from the valley who are encroaching in the watershed, as in the case of the Mae Lai watershed.
In order for teams made up of staff from different agencies to be able to work together well in the field, it is necessary for the concerned agencies to be willing to put this staff entirely at the disposal of the team, including resources and time. Simply adding the obligation to participate in the team to an official's regular duties will not be sufficient for real teamwork, as sought by this project, to take place.
Possible future IFAD projects in Thailand
Since ADPIP was drawing to a close, the MTE mission also considered a limited number of possible follow-on IFAD projects in Thailand, using the experience acquired in this project:
(i) IFAD could consider a project which takes up where this project stops, by working with PIAs downstream of recently built RID small to medium-sized storage dams in the poorer valleys of the North. This would mean continuing the work of agricultural diversification, support for institutional strengthening of PIAs and IPIACs, and in particular working with PIAs and RID to develop a mechanism for the transfer of ownership (with or without payment) of these government-built irrigation works to the farmers who use them. Involvement of PIAs in watershed protection and rehabilitation would also be a part of this project. An explicit effort should be made to target the poorest PIAs, as well as trying to involve non-PIA poor farmers in the project activities.
(ii) A second possibility is to undertake a participatory watershed rehabilitation project in the Northeast. Many watersheds there suffer from encroachment similar to that occurring in the Mae Lai area. In these conditions, as would be appropriate in those of Mae Lai, productivity and environmental soundness could both be raised by combining the efforts of populations residing in the same sub-basin, including both those in the watershed forest area and those downstream.
(iii) A third type of project would be participatory hill tribe development in ecologically fragile areas on the northern borders. The project would target poor hill tribes living in Conservation Forest areas and practising shifting cultivation, encouraging them through a long process of group meetings and awareness-building to change their cultivation techniques in favour of ecologically more stable and economically advantageous systems. The project would aim at improving relations between RTG (in particular RFD) and downhill Thai farming communities on the one hand and the hill tribes on the other, reduce opium cultivation and promote forest conservation.
(iv) A final possibility was proposed by the project management and regards the financing of a similar medium-sized dam building project, with one major difference: in the case of the Community Irrigation Development Project (as it would be called), the dams would be built in valleys with no existing distribution systems, and the project would therefore attempt to reproduce the organization and structure of PIAs in areas where they have never existed. In the view of this mission, this is a rather risky proposition, though it is a challenging and interesting one for IFAD to consider. The four proposed dams would irrigate an estimated 15 000 ha in the Northern Region. It would important to determine the level of poverty in the proposed project areas, since the North is no longer a very poor region in Thailand.