First Eastern Zone Agricultural Project (1997) - IOE
First Eastern Zone Agricultural Project (1997)
Interim evaluation
The EZ comprises the project area of FEZAP, the population in 1991 was estimated at 191 000 and the target group comprised a total of 21 730 households (131 000 people). The latter were defined as farm families with less than 3.0 ha and incomes of no more than the lowest paid government employees (105 USD per capita). Irrigation schemes to be rehabilitated would be selected where 50% of the households would have wetland less than 0.3 hectare.
The Project sets out to improve the living standards, cash incomes and household food security of the small farmers with holdings below 3.0 hectare. Five project components were included in the original SAR design: (i) Renewable natural resource development (RNR; (ii) Small-scale irrigation; (iii) Participatory community development; (iv) Credit; and (v) Institutional support. The RNR components comprised support for research, extension, and the Natural Resources Management Unit.
Expected effects and assumptions
The RGOB policies for rural development, expressed through the "Five Year Plans" aim to reduce income disparities, curb the drift from rural to urban areas and protect the environment. Soil degradation and erosion arise if households are forced to continue extensive cultivation systems. But relevant production technology must be first generated to permit an agricultural intensification to proceed that generates higher land and labour productivity. The generation of such technology is constrained by the wide variation in cropping suitability, microclimate and market access.
The path to progress is to design the processes where the emerging capabilities in technology generation are supported by the building-up of institutions and methodologies for decentralised decision making. Rural communities are best served when their own location specific priorities guide research as well as project supported service provision. The shift from directed service provision in research and extension, and from central to decentralised decision making for rural communities, in determining priorities and annual work programmes is a tall order not just in Eastern Bhutan. To achieve success, issues must be correctly defined: methodologies must be made transparent, processes of solutions can then be more easily agreed upon. This was not the case in the design of the FEZAP, in two central areas, in the generation of production technology and in participation: outcomes then were not in line with needs or expectations.
First, the SAR underplayed the importance of designing capabilities to generate processes and production technology relevant to the great variation in location specificity. With the benefit of hindsight, it is fair to state that the principal cause of FEZAP's problems rests in the ambiguity created in the SAR as to the Project's proper nature. The SAR under project risks, stated: "Technically no unusual risks are involved as the Project would promote proven technology. The farmers are conservation minded and have demonstrated their willingness to adopt new and financially viable technical recommendations." The only exception in the SAR concerned essential oils, where reference was made to the risk that high quality standards required for premium export prices may not be maintained.
The ambiguity about the true nature of the Project was compounded by the rationale provided by the SAR for the farming systems component. The farming systems research was not described as an effort that primarily would generate useful technology or relevant recommendations by involving farmers.. Instead, this component would identify constraints and development opportunities for the various farming systems. The point was never driven home that farmers need to fully participate in research and trial design, if efforts are to become productive and generate results that are ultimately adopted by farmers.
The Project should have been presented as a research and development operation in RNR of national interest. It should have been presented as a Project that would generate and test different technologies and village level interventions, and where extension would support farmers' capacity to experiment, not least because of the wide variation in agro-ecology and farming conditions. Standard messages would be adapted. Technology and recommendations would be forthcoming, aligned with the great disparity in farming conditions.
Second, the bias in favour of merely extending already proven messages was compounded. since the RGOB Blue Book in redefining the SAR proposed participatory community development component, removed the participatory elements. The rural awareness and empowerment elements were removed: this component was narrowly redefined as organising farmers to receive technical extension messages, again reflecting the incorrect notions that adequate technology was available and that extension is a simple one way diffusion process.
The key to economic, social and environmental benefits from research investments and subsequent development and extension efforts is to design technologies which farmers can access, and choose to use. Institutional capabilities for R&D, especially a well directed process supported by effective training delivery, are required to reach this end. The final design for FEZAP excluded this necessary institution building.
Purpose: The evaluation sets out to understand to what extent MoA programme managers, project and Dzongkhag staff review implementation progress against the physical targets set. Moreover, are they also seeking the data that show impact at the village and farm level? Finally, are they trained and encouraged not only to explore causes to variations in performance, but also to seek to address these underlying causes?
Data have been obtained with which to understand: (i) farmers' constraints and their coping strategies; and (ii) how the Project has managed to assist farmers to overcome or mitigate their constraints? The OE conducted one survey in preparation of the evaluation (wealth ranking and poverty profile), the FEZAP conducted the PBME; moreover, the mission conducted a PRA and an informal survey of irrigation schemes.
Wealth Ranking and Poverty Profile: Precise data on the size distribution of resource endowments were not available at the time of design. The SAR presumed that 80% of the households would possess livestock and almost all households owned some land. More precise and up-to-date data sets were required for the IE. The First Phase IE mission which visited the FEZAP project area in Bhutan in October - November 1995, explored a methodology for wealth ranking.
Distinctions between wealthy and resource poor households were explored with structured but open ended questionnaires, in four villages, two in Khar gewog in Pemagatshel and two villages in Toetse gewog in Trashiyangtse. Some 40 farmers classified as poor were interviewed to determine if the use of the farmers' direct ranking correspond well with an "actual" or objective ranking of their assets (owned wetland, dryland and livestock). This testing confirmed that the direct method of ranking farmers can be used as a shortcut to identify those that are resource poor.
The PRA undertaken by the mission confirmed the validity of the Phase I IE wealth ranking. The PRA was conducted in three villages (in Khar, Lhuntse, and in Menjabi gewogs). It confirmed that the absolutely poor are those who borrow food and money in order to survive the hungry season. Data on food insecurity and on entitlements of productive assets represent powerful indicators for poverty.
The PBME: The FEZAP replicated the wealth ranking in its own survey, the PBME survey, on a wider scale. The PBME survey, conducted in two gewogs in each of Pemagatshel, Lhuntse and Trashigang Dzonkhags, sampled 477 households. FEZAP should be praised for issuing its report in advance of the mission's arrival. Yet, much important data remained to be analysed by the IE on return to Rome (Chapter 5).
The Zonal Administrations were abolished shortly after FEZAP became effective (October 1992). Responsibility for project implementation shifted to the six eastern Dzongkhag Administrations, supported by a new Project Facilitation Office (PFO). Project activities became part of the individual Dzongkhag development plans. Important changes were: (i) the resources and activities of the Community Development Division and the Natural Resources Management Unit were transferred to the training and extension programme to be undertaken by the PFO and the District Administrations; (ii) the irrigation programme was extended to include new schemes (up to 400 ha); (iii) a crop promotion programme was introduced; (iv) financial support was given to a breeding programme for Mithun (draught) cattle and to fodder development; and (v) the credit component was taken out (grant funding was available from another donor).
Progress: The FEZAP organisation and management structure has evolved satisfactorily once the reorganisation of the administrative structures was completed and roles and responsibilities of the PFO had been re-established. The task is huge of operating through six different Dzonkhag administrations in vast and remote areas.
Satisfactory relationships between the PFO and Dzongkhag staff have been reached.
Physical progress has surpassed targets in many areas, over the last two years. Physical progress across the crop and livestock sub-components (all part of the RNR component), implemented through the Dzongkhag administrations on the whole has been quite satisfactory. FEZAP is funding the livestock stations at Lingmethang and Arong, both have been fully equipped. The irrigation component is making steady progress; unit costs have been lower than expected, and disbursements are less than planned, but physical progress in terms of construction activities ranges between 20%- 45% of targets in the 7th FYP.
Research is undertaken by the RNRRC, but progress has been stymied by insufficiencies in direction, in the building-up of capabilities, and in transport equipment. Support for the PFO is well on target, with civil works and equipment all being provided. Yet, provisions for vehicles and training are far below requirements.
Constraints: FEZAP has continued to operate under three critical handicaps. First, the misconception based on the notion of "proven" technology led to a minimal appreciation of the need of direction in research: capabilities were not put in place to generate relevant improved technology to farmers. Proven messages were available for some produce, for potatoes and vegetables; some of the local cereal varieties were of reasonable quality in some niches and altitudes. But the actual extent of missing knowledge and insufficient capabilities increasingly have become visible.
To wit: (i) proven technology and recommendations were not - and are not - available for improved maize and rice varieties across the agro-ecologies; (ii) fertiliser and manuring recommendations for cereals are too high and not used by most farmers; (iii) the irrigation systems could not generate expected productivity: agro-economic parameters, geotechnical considerations and soil instability were not sufficiently considered in design; maintenance costs become too high in relation to overall earnings of farmers, and capacity to contribute cash and labour; (iv) soil erosion has become a major issue not least on wetland, but relevant approaches and messages for soil conservation have not been generated; mere propagation of vetiver grass is insufficient; (v) the relevancy, and productivity of pasture and fodder production is not confirmed, and little progress in found for fodder trees: (vi) farmers' adoption of AI, and natural insemination for improved breeds is uncertain in the absence of a subsidy; (vii) capacity utilisation at the Lingmethang farm is low in producing improved poultry; unit costs of feed are too high in relation to selling price, and experiments need to be designed in decentralising poultry multiplication to the village level are required; and (viii) a study of pig production and feed under village conditions is underway, but results are not complete and not sufficiently monitored.
Second, the misconception as to proven technology was accentuated because of the biases against fact finding about farmers' conditions and reflection: (i) insufficiently participative methodology and procedures also within the commodity driven research system; (ii) limited use made of the technical assistance provided; and (iii) an inward looking M&E unit, developing neither any indicators for progress, adoption and constraints at the gewog, WUA and farm levels, nor any updated farm budgets.
Third, the removal by the RGOB of the directly participatory elements included under the SAR design meant that FEZAP did not have at hand the overall capabilities, resources and methodology with which to train EAs to empower farmers and encourage experimentation in group mobilisation and farmer led diffusion methods. Moreover, the importance was not appreciated of supporting the GYT mechanism with which to encourage feed back from the gewogs so that RGOB and FEZAP resources could provide services in line with local priorities.
Staff skills, staff positions, technical assistance were defined narrowly. The capabilities set-up were aligned with an extension project: i.e. the dissemination of proven technology and farming and irrigation systems. The importance of creating capabilities for diagnosis, research and technology generation was not set out: what was not clearly understood could not be resolutely acted upon.
Addressing design weaknesses: the FEZAP staff should be praised. They increasingly realised that they needed far more research support to generate productive messages that would respond to farmers' constraints, but that this support was not forthcoming. The PFO and RNRRC staff should be much acknowledged for their efforts in reducing uncertainty and re-directing the Project through: (i) the two village level studies, conducted with SNV support, on adoption behaviour for maize and rice, which defined a proper methodology for obtaining farmers' screening criteria; (ii) the training workshops with extension agents in supporting them with methodology to undertake experiments with farmers; (iii) the undertaking of the far ranging PBME survey; (iv) supporting a pilot project to develop community based land use development plans; and (v) applying the requirement that farmer communities themselves request assistance with irrigation scheme rehabilitation.
Effects assessment and sustainability
Growth has taken place across the project area, especially farmers with more land are benefiting from increased specialisation in production and trade in horticulture and vegetable, but the resource poor households are far from sharing proportionately in these benefits. Overall land distribution across the Eastern Zone is unequal rather than equal. Although land poor but labour rich households rent in wetland and the distribution of owned and operated land is more equal than that of owned land (Chapter 5), the limited assets of land and livestock contribute to low productivity of labour and high degrees of food insecurity. Resource poor farmers have little choice but to increase the input of the only resource they control, their family labour. They benefit only to a smaller degree in the overall growth related mainly to marketing of cash crops. Moreover, prospects for off-farm cash income to women through weaving are bleak because of the competition from cheap manufactured Indian cotton cloth. The IE Phase I survey showed that of the resource poor farmers, two thirds are food insecure for more than three months; of the latter, half had not repaid their borrowed food even after the harvest.
Declining physical resource endowments increase already high labour loads of women: the Project has not directly eased women' labour constraints . Distances involved in herding cattle and to collect fuelwood are increasing. Labour productivity declines further when mineral fertiliser or organic manure is not available. The effects of peak labour loads and prolonged food security is associated with high incidence of malnutrition, and with illiteracy, of adults and children.
Women in resource poor households may sleep only three hours per night during the peak labour season, and the incidence of diarrhoea and malnutrition increases. The frequency of children under five years of age, exposed to chronic malnutrition (height for age) in the EZ was recorded at a far too high rate of 64% in 1989. Stunting levels of children probably still remain far too high, and recent research confirm that future mental capacity is then impaired.
Farmers with little land, in addition, are less literate and their children are more likely to drop out of school. The illiteracy of those most resource poor is as high as 80%; moreover, 36% of the children in the poorest group are dropping out; compared to 22%, and 17% in the next two size groups. Productivity in farming is negatively affected by such low levels of literacy and schooling.
Current inequities in distribution of land and access to production technology can be justified neither in terms of equity, nor growth. Those owning little land are highly efficient in their resource use; their cereal yields are higher, and they contribute less to soil erosion in relative terms. The latter may be explained since ownership, compared to tenancy, is normally associated with more incentive to invest labour in soil conservation and erosion control.
Perpetuated literacy and malnutrition for the already food insecure and most resource poor farmers should be a cause of concern. Their situation is further worsened because of the RGOB legislation to prohibit the use of traditional tsheri. The damage of crops by wild animals is already unacceptably high and negates much of the overall RNR efforts: losses are probably increasing when the prohibition to use tsheri land is enforced: wild animals may come closer to the settlements.
Main issues and recommendations
Two critical features stand out in this evaluation. First, the ineffectiveness of FEZAP as a vehicle for extending proven technology because of its weaknesses in design; and second its praiseworthy efforts in addressing the original shortcomings. The knowledge basis has begun to be slowly built-up to permit system design and extension recommendations to spring first from a diagnosis of the constraints that affect the target groups, and then a pursuit of experimentation and testing of viable solutions accepted by local farmer communities. But these efforts, or their impact, will remain marginal and not be properly supported unless the entire Project is correctly defined. It must be understood that: (i) technology generation through research and experimentation at the village and farm level needs to precede extension; and (ii) this process needs to be supported by the creation of institutional capabilities for the delivery of services to the rural communities.
Vicious circles perpetuate poverty and food insecurity in the EZ and a two pronged strategy is called for. The concerns of those which have already a surplus of marketable produce need to be addressed.
Nonetheless, for the RGOB it is neither equitable nor efficient to permit the resource poor farmers to fall further behind. To begin with, technology needs to be developed through on-farm and village level trials that suits the particular need and niche of resource poor households. It has to meet the needs of food security, be land saving, generate higher labour productivity and reduce incidence of erosion. A properly designed R&D project can and should be an effective vehicle for poverty alleviation: this was not the case so far under FEZAP.
The demand of smaller farmers for cereal varieties that meet their need of food security (shorter maturities) and for relevant manuring recommendations must be met. About 40% of all households have no cattle; on average, probably about 10% of overall farmers, or 30% of the food insecure households, have neither pigs nor poultry. Improved breeds of poultry and pigs normally represent a highly productive investment for the resource poor, generate animal protein, a decent return on labour and reduce overall risk. But such breeds are made available by FEZAP only in very limited numbers. Recommendations for feed and maintenance under village level conditions are to be developed through adaptive research and testing.
Second, the institutional component will have to ensure that: (i) mineral fertiliser is more accessible to permit smaller farmers to realise their comparative advantage in production (yields); (ii) a strong extension programme is carried out especially for vegetables to permit smaller farmers to progress beyond cereal production and share in the benefits of vegetable production; the latter compared to perennial crops (oranges and apples) is neutral to scale, may contribute to less erosion, and generate earlier needed cash; (iii) infrastructure for the marketing is available; (iv) farmer-to- farmer extension approaches are used to increase coverage and increase FEZAP cost-effectiveness; and (v) group credit is becoming available to permit smaller farmers to diversify their enterprises and reduce co-variant risk.
But smaller farmers are exposed to the inequities of the current land distribution which is unequal rather than equal. For this reason, a priority is to: (i) make more wetland with improved irrigation available to smaller farmers by ensuring that inequities in the present land distribution, especially of wetland are removed; and (ii) target the rehabilitation of irrigation schemes to those wetlands where 75% of all farmers have landholdings below 0. 3 ha.; per se this should reduce the incidence of soil erosion.
To wit, the benefits of a future programme to farmers in general and smaller farmers in particular will be compromised unless: (i) the current prohibition for farmers to use their tsheri land is redesigned; and (ii) effective measures are finally introduced to combat widespread damage to crops of wild animals. The widespread damage to crops by wild animals acts as a disincentive for parents to send their children to school: current high illiteracy rates are perpetuated.
Once the nature of the Project is recognised, the FEZAP Phase II Project should be redefined as a research and development project to: (i) devise extension recommendations for annual and perennial crops, for livestock and to design small scale irrigation systems to serve not least the resource-poor families in the different agro-ecologies; (ii) develop productive, sustainable and equitable approaches to overall land use and the maintenance of soil fertility and moisture; and (iii) devise and set in train a process of awareness and training in order for the relevant national institutions and other partners in development to appropriate the methodology of programming and executing research, development and ensuing extension activities through participatory processes.
This strategy translates into four sectoral objectives: (i) Renewable Natural Resource Development: crops, livestock and irrigation; (ii) small scale infrastructure and marketing support; (iii) credit; and (iv) institutional development. A final priority for a future FEZAP II is to co-operate closely with an organisations such as UNICEF and the Ministry of Health, to ensure that the health and nutrition dimensions are properly followed up on. Essential is to ensure that indicators for nutritional status are monitored, presented and discussed, for a representative sample of villages in the project area.
Renewable Natural Resource Development including Irrigation: the component entitled Renewable Natural Resource Development should include irrigation to become a component designed to improve farmers' overall systems which comprise both dry- and wetland. Farmers strive to cope with the demands for intensification, but without improved technology their productivity of efforts is falling behind. So far little has come of the approach advocated in the SAR with the recommendations to address soil erosion with emphasis on permanent cropping measures (pasture, trees, horticulture, strip farming, contour and bunding on cropped areas most at risk together with conservation of up-slope forested areas). Instead, soil erosion has become more serious, more on wetland and dryland than on shifting cultivation land.
Community based solutions to shifting cultivation should be sought for: moreover, it is not realistic to presume that permanent cultivation can or should become a universal solution.
In the irrigation sector, a two pronged strategy is called for that combines raising productivity and reaching more farmers, but at a lower RGOB cost per benefiting farmer.
The RGOB cannot afford in the longer run to provide continued subsides, beyond the regular first year "put right" support, to almost all completed irrigation schemes, even though maintenance requirements and repairs exceed the farmers ability to contribute. Farmers' capacity to contribute. maintain and self manage will increase only when productivity in total water use is improved. The first prong in this strategy comprises raising scheme productivity by improved design (emphasis on secondary and tertiary canals, geophysical factors, agronomy, e.g. crop rotation to reduce soil erosion). Criteria for selecting schemes for rehabilitation would be where: (i) market access for cash crops is ensured, which translates into high returns to farmers on cash and labour; (ii) a reasonably effective WUA has been already created; and (iii) the resource poor farmers share proportionately in the benefits that accrue (the targeting dimension).
The second prong of this strategy is to start a sub-programme, a pilot programme, to rehabilitate a subset of the large numbers of FMIS, which can be improved without large investment or high unit costs. This effort emphasises training of farmers that have been selected by the farmers themselves and through this training of trainers approach, the direct involvement of the irrigation sector staff can be reduced.
The experiment-demonstration element needs to be broadened and diversified. The extension of relevant messages, for instance, of the Integrated Pest Management (IPM) recommendations against fruitfly in Khar, needs to be made more effective. Alternative methods for farmer-to-farmer diffusion and group extension need to be explored.
Community based programmes: a community oriented approach offers much promise in promoting a sustainable pattern of improved land use, soil conservation and reduced erosion. Farmers' patterns of overall land use, interactions between different enterprises and factors explaining labour constraints need to be pursued at the local level. Interactions, constraints and feasible local institutional solutions emerge only through a dialogue with farmers and interest groups at the community level. The MoA LUPP Unit in Thimphu has undertaken a praiseworthy survey and developed a master plan for improved land use for Drametse gewog in Mongar. The FEZAP has built on this design. An experimental design for a community development plan has been drafted, oriented towards sustainable land use including social forestry; and necessary changes in land use regulation were to be approved in Thimphu. This orientation needs further support: technical solutions have to be designed through participatory methods, and incentives in undertaking conservation measures need to be properly explored. Pilot initiatives need to be launched to test technical solutions, approaches and acceptability at the local levels. As regards fodder trees, and living hedges, a research programme needs to be mounted, together with the Forest Services, without delay.
The core programme would comprise a set of community development plans in each gewog: co-operative solutions to land use should be tested based on the local priorities. With an approach geared to needs, agro-economic potential and to tangible solutions by type of households, family income status and food security can be analysed with a set of indicators.
Small Scale Infrastructure and Marketing Support: The marketing of horticultural crops, vegetables, potatoes and oranges needs targeted support. Demand in India and Bangladesh provide the main impetus in marketed cash production. Markets are becoming more efficient, farmers even in distant communities are aware of border prices, but limited road access, storage and transport constrain supply response. Success in intensification necessarily requires higher margins for output to justify rising labour inputs, irrigation scheme maintenance and fertiliser application.
A Community Development Fund is proposed to propel the creation of small scale infrastructure, power tiller roads, hanging bridges, storage sheds at the road side. This infrastructure would be built on a self-help basis; farmers in Bumdeling gewog stating their need for such a road and a shed at the road side, were prepared to provide the necessary labour themselves. More often than not, such infrastructure is a requisite for a take-off in the production of cash crops for outlying gewogs.
The need of finding effective solutions to improving the delivery of mineral fertiliser falls under this component. The financing arrangements of the commission agents selling fertilisers need to be reviewed, so as to ensure that farmers actually obtain the means with which to improve soil fertility.
Credit: The RGOB and BDFC must ensure that the rural financial services are becoming more accessible and farmers' transactions costs are reduced. A pilot credit component should be designed to explore decentralised, self-governing financial services with autonomy vis-à-vis the rest of the Project. Typically, farmers' transactions costs for accessing BDFC credit are high, if not prohibitive. More easily accessible credit, with the use of group collateral to reduce risk, is fundamental in order to provide a take-off for intensified production where a market is already available. This component, must incorporate experimentation and testing in order for the BDFC and NWAB to merge and devise methods for setting up decentralised rural financial services. Rural financial services should reach first those gewogs where a sufficient loan volume can be foreseen. This is the rationale for the proposed Gewog Banking Units.
Institutional Support and Strengthening: the institutional support and strengthening component is expected to answer to three "transversal" objectives, namely: (i) analysis of the socio-economic environment and the building-up of a gewog based progress and performance monitoring system; the responsibility belongs to the M&E Unit and the Dzongkhag planning officers; (ii) promoting the research and development methods and extending testing and pilot efforts to support farmer group approaches to multiply: these responsibilities all fall under a strengthened extension component together with the RNRRC and the irrigation sections; and (iii) promoting participation to generate higher efficiency of resource allocation though feed back and aligning gewog level activities with local preferences. A seconded institutions specialist is required at the level of the PFO. The institution building component should be linked to the community development fund to permit the finance for self-help projects, to drive also the institution building efforts.
Reassessment of land use legislation: Present land use patterns and legislation need to be reassessed The current RGOB prohibition of shifting cultivation on tsheri land needs to be re-evaluated; the findings of this IE reconfirm the negative assessment of the current legislation by FAO in 1989. An argument can be made in favour of a land redistribution on five grounds, to: (i) increase overall production from a given area; (ii) improve food security and nutrition of the resource poor; (iii) provide better incentive to farmers in undertaking the additional labour required for soil conservation measures, since they own the land; (iv) improve parents' capacity to send their children to school and raise the literacy of their children; and (v) facilitate the uptake of technology by the next generation of farmers.
The argument has already been made: the LUPP study for Drametse argued for a redistribution of land, since one or two land owners in this gewog had very large areas of land.
Targeting: targeting mechanisms are most effective when technology is developed through on-farm and village level trials that suit the particular need and niche of resource poor households. It is for this reason that a properly designed R&D project can and should be an effective vehicle for poverty alleviation. A second priority is to ensure that inequities in the present land distribution, especially of wetland are removed. A third priority it to target the rehabilitation of irrigation schemes to those where 75% of all farmers have landholdings below 0.3 hectare.
Nutrition Monitoring: a final priority for a future FEZAP II is to co-operate closely with an organisations such as UNICEF and the Ministry of Health, to ensure that the health and nutrition dimensions are properly followed up on. Essential is to ensure that indicators for nutritional status are monitored in a representative sample of villages in the project area.
(a) FEZAP has Four Essentials Lessons to be Borne in Mind for Research and Development Projects:
Need to recognise "research and development needs"
It is fundamental, before designing any project, to identify those recommendations that are available, i.e. sufficiently tried out and adapted to the point where they are suited for extension purposes, or have begun to be adopted. This means discerning also those that need further development, testing and adaptation to the conditions of the target area and target group, and those that should be discarded.
Too often, projects are designed with limited data and knowledge as to the degree of variation in farmers' productivity, entitlements, constraints and risk. What is not observed is not explained. The economic justification and analysis of benefits versus costs is further constrained by the emphasis on obtaining a quantifiable ERR. The latter accentuates the bias towards overestimating the extent of the actual knowledge and the usefulness of available or so called proven technology.
This means that projects too often are justified as extension projects when they should have been designed as research and development projects. The case for farming systems research and on-farm and village level testing that should drive the generation of improved production technology is then either not understood in-country, or it is underplayed.
A primary precondition for more appropriate project design is that more attention is placed on the diagnosis of research and development needs; it is essential that this process be conducted in representative locations together with the target group itself. By such diagnosis, farmers' actual preferences and community level constraints can be understood. Design of extension activities can be avoided that will later prove ill-adapted and not effective.
Knowledge generation and diffusion
In the first place, a research and development project has to be developed in terms of the methods of approach and not in terms of detailed activities and outcomes. First, results should be seen in terms of gaining knowledge and know-how, and not of material achievements. Second, against the overall project strategy, measurable objectives need to be specified to assist the local staff and the relevant institutions to acquire the necessary skills in this research and development. Thirdly, a series of indicators need to be designed against which acquisition of skills can be monitored. Fourthly, research and development take time: this dimension needs to be taken into account in order to secure continuity in the approach beyond project end.
An R&D approach for rural financial services
Extending provision of financial services into rural areas is vital to encourage intensification and diversification of farm enterprises. Alternative innovative models need to be tested so as to increase access to credit for the rural population. This is the case, especially when credit supply is not well adapted to demand, transaction costs of borrowers are prohibitively high, and there are no institutional solutions in the country that are backed by sufficient experience, and are established, recognised and replicable.
Innovative solutions are required for technical assistance
Innovative or alternative solutions need to be found for long-term technical assistance. A compromise is required between the need of reduced costs, as against securing continuing longer term benefits from the knowledge generation process. Long-term partnership arrangements with institutions specialising in research and development, with regular support missions, training courses and attendance at workshops, is one solution. Such "twinning" arrangements should be explored with both bilateral and multilateral donors.
(b) Lessons Learnt for Project Design in General
Aside from the lessons learnt concerning the need for making precise distinctions as to the actual knowledge available, and about proven technology, and the need to adopt a research and development methodology and clear-cut procedures at the design stage, FEZAP has seven lessons to impart as regards project design in general.
An understanding of initial socio-economic conditions
The project's acquaintance with the socio-economic conditions of the target group at the start was limited. The greater is this initial insufficiency, the higher is the probability that changes to the project concept will be needed in the course of execution. Moreover, the harder it will be to modify and adapt a project during its implementation. This is why it is essential, at the design stage, to analyse the target group's socio-economic as well as physical environment, and to conduct highly specific surveys to ascertain its constraints, preferences to contribute labour, and approval .
Low cost survey technology
Structured approaches are necessary for obtaining the necessary data. A low cost "survey technology", described in the IE report, is available to obtain better data on resource poor farmers. In preparation for the Interim Evaluation, a wealth ranking exercise was undertaken in the project area. Villagers themselves were asked to identify the resource poor households. The reliability of the results was confirmed by direct interviewing. The experience proved the benefits of using low cost methods for wealth ranking followed by subsequent targeted surveys. Once identified, the resource poor households are interviewed: feasible solutions to reduce constraints and food insecurity can then be discussed with them, and pursued.
Institution building and capabilities
The creation of capabilities should be part of a sector strategy with which to raise productivity in the systems for the delivery of services to the rural population. But the creation of capabilities for generating better performance within and across projects is a neglected dimension at time of design. At the stage of design, a strategy for training, for creating capabilities, a management structure, and financial resources for training must be formulated. Training should focus on: (i) the processes required for generating knowledge from diagnosis, experiments, testing and monitoring; (ii) disseminating the knowledge generated from the R&D among local staff and institutional partners; and (iii) evaluation.
Extension activities
Conventional, ministry driven transfer of technology oriented agricultural extension has failed to promote rural development in particular in those regions defined as agro-ecologically diverse, resource poor and risk prone. It is recognised that farmer-led approaches better integrate research and extension functions drawing upon knowledge and research capacities of local communities and combining them with those in formal research and development organisations. Local capacity for experimentation must be supported so as to permit technology to be adapted and further disseminated.
Present diffusion models need review. The possibility of using farmers as part time lower level extension workers (after necessary training) need to be explored. The use of village level extension workers offers promise in terms of a cheaper, more effective channel to encourage farmers' experimentation, relevant feedback, and dissemination of extension messages in the remote villages. Such farmer-extension agents would not be directly paid by Government, but could be assisted in kind through free inputs, training, etc., and villagers could compensate them for foregone earnings (labour lost).
Land use and conservation
Project based lending normally fails to generate productive and sustainable land use that reduces soil erosion. Projects concerned with the development of land and water resources should identify and address cases to already visible erosion. Soil erosion needs to be analysed by type of land, and by type of management and cropping system. The current IE has found that soil erosion on small owned plots is less than on larger holdings. Smaller farmers who own their land, cultivate more intensively, and are likely to generate less erosion. Such findings should lead to reflection and further exploration. In future projects, land use and soil erosion should be monitored by type of land and size of entitlements of farmers.
Neglect of small livestock
Much attention is normally given to large stock across projects in contrast to small stock, pigs and poultry. Pigs and small poultry under right conditions provide high returns to female household labour and provide animal protein as well as cash. Research support for small stock is often minimal, and this area of intervention is one of the most promising in assisting the IFAD target group.
More attention to animal feed
In the light of the tight feed situation, continued acquisition of more cattle (even if improved) gradually becomes counter-productive.
Compensating measures are needed to increase feed while reducing useless cattle units.
(c) Lessons learnt as regards project operation
FEZAP offers three important lessons for project implementation in general:
Beneficiary participation and tenancy
In sites where tenancy rights are uncertain, voluntary labour contributions are not forthcoming for channel rehabilitation. Tenants are reluctant to participate/contribute to channel rehabilitation, since their short-term tenure preclude reaping their longer-term benefits.
Implementation is delayed and disputes need early arbitration. Other channels that are not affected by unequal ownership patterns and/or disputes should be given chronological priority.
The introduction of new water users' groups has often resulted in conflict and lack of participation, as has happened in some other Asian countries. Areas of conflict between different users should be monitored to explore more efficient and equitable mechanisms for water sharing, which then can be supported by project interventions.
Beneficiary participation: labour constraints
The management unit of the Project should maintain close, ongoing co-ordination with the district authorities to ensure that the demands for beneficiary labour contributions are not excessive and do not interfere with agricultural and other income-generating activities.
Full account must be taken of higher on-farm labour requirements resulting from project initiatives to promote infrastructure and increased cropping intensities. The RGOB still practices a system of taxing rural households through labour contributions: the timing and the extent of such labour needs to fit into the seasonal farming calendar so as to not worsen food insecurity.
M&E
The evaluation and support provided by IFAD and the Co-operating Institution are essential for developing the capacities of the monitoring and evaluation units. The competence of these units should be related to job descriptions and activities, be specified at the outset and be upgraded over time. The units should not be permitted to be overloaded with demands for its services. Otherwise, their ability to gather and analyse information, and to develop its own capacity for reflection and adaptation will suffer.