Completion Evaluation

The project area, in the northern part of Uganda, covers about 25% of the total land area, or approximately 59,000 square km, in seven administrative districts. The terrain is flat to undulating with numerous low-lying areas around lakes and swamps which were estimated to cover 3.6 million ha and to provide as much as 60% of the country's fish. The whole region is in the tropical savannah belt, where mean annual rainfall varies from 1000-1500 mm: in the south a bimodal rainfall pattern permits double cropping, but in the north the rainfall is unimodal and only cropping season is possible. Except for the cracking clay soils in areas of impeded drainage near lake Kyoga, the land lends itself to cultivation with ox ploughs and simple hand tools. The agricultural pattern throughout the area is quite homogeneous, with cotton as the main cash crop, and maize, millet, groundnuts, cassava, beans, sorghum and banana as food crops. There is a large cattle population in the region, over one million head, or more than one third of the national herd. Land ownership is vested in the tribal clan, which allocates usufruct rights to individual clan members, hence there is no absolute security of land tenure. Grazing areas and water resources are communally owned.

Project design and objectives

After many years of turmoil and civil unrest the production systems of much of Uganda had been completely devastated; in 1981 IFAD funded the Agricultural Reconstruction Programme (ARP) with the objective of assisting the Government of Uganda (GOU) in its recovery programme by providing basic agricultural inputs and services to north and east Uganda. The agricultural Development Project (ADP) was designed as the second phase of this emergency programme with particular emphasis on smallholder food production capacity.

Target group

The three tribal groups in the area are Bantu, Nilotic and Nilo-Hamitic in origin. At appraisal in 1984 the population was estimated at 2.6 million; the target group consisted of the 387,000 families who were either small-scale farmers, livestock owners or fishermen, on the assumption that the economic levels of this entire group were about the same. However, this also included about 15% of the population who were handicapped, and had little potential to participate in this type of project. The unrest of the recent past had caused the status of women to change rapidly for the better, and the rapid rise of the AIDS epidemic (10% of the population and 20% of the male population were estimated as HIV positive) also meant that there were a significant number of women-headed households. Traditionally, men exert total control over land through clan and household leadership, and have authority over all economic decisions in the household. However, as agricultural activities decreased as a result of the socio-political unrest, so that crop inputs were no longer available and cattle rustling was rife, then women's ability to be income earners came to the fore, increasing their decision making role in the household. (At the same time their work load also increased, since they were responsible for most of the work on the farms.)

Objectives and components

The objectives of the ADP were to increase food production, upgrade the health of the livestock population, increase the fish catch and provide a basis for the long term development of agriculture in Uganda through the strengthening and improvement of institutions and by carrying out adaptive trials. Through this process the intention was to improve the incomes and food security of the project's target group of poor households. The objectives were to be attained by (i) the provision of basic agricultural, fisheries and livestock inputs, (ii) support for adaptive research, tsetse fly control programmes and various studies, and (iii) institutional development covering extension, training, a pilot T & V extension system and support for a monitoring and evaluation unit. The project's total base cost was estimated as USD 23.9 mn, plus an allowance of USD 7.4 mn for contingencies; the component analysis was:

Input Supplies	USD 17.7 mn (74%)
Research & Technology Dev.	USD 2.5 mn (10%)
Institutional Development	USD 3.7 mn (16%)

IFAD's share of the funding was 46% of project costs, IDA 33% and GOU the balance of 21%.

Input Supplies, by far the largest component at 74 % of the funding, included, for crop production, hand tools, ox-carts and ploughs, small flour mills, seeds, bicycles, wheel-barrows, workshop equipment, spare parts and chemicals and forestry nursery supplies. For livestock, supplies were vaccines, veterinary drugs and equipment, acaricides, spare parts, vehicles and a mobile laboratory. Fishery inputs to be provided were outboard motors, spares, fish nets and accessories, twine, boat building materials and vehicles/. Support for the Research and Technology component included surveys to generate new information for planning and extension, and an adaptive research programme carried out on farmers fields. Tsetse fly control was to be tested using traps (this was later extended to a full catching programme), and fish catches and fish populations were to be studied in lake Kyoga. Fish preservation was to be addressed by demonstrations of a chorkor kiln smoking house. The Institutional Development component covered the provision of funds for civil works, building renovation, office and field equipment, audio visual aids, survey equipment, technical assistance, training and some operating costs. Agricultural extension was to be strengthened by rehabilitating training centres and introducing a T & V scheme in one of the districts. A Monitoring and Evaluation Unit (MEU) was to be established and equipped within the project management unit. Staff training would be included in most activities (including overseas training courses) and incentives would be paid to staff based on performance and responsibilities.

The implementation of the project was affected by a number of factors (see Implementation Context below), which led to the project being reformulated in 1991. The modified project kept the original components but added a pilot credit scheme, ox-cultivation, cassava mosaic disease (CMD) control and a women's programme. All this was still within the original total budget. The pilot credit component was to enable farmers to purchase the inputs which were being supplied, and also to help rural entrepreneurs: credit was to be provided to 3,000 farmers in groups and societies of between 10-30 members each, identified as having "high potential", and was linked to the uptake of inputs, extension and technology. It was also specified that women should make up at least 50% of the beneficiaries. The cost allocation for the credit component was USD 1.8 mn, of which USD 1.4 mn was for the establishment of a revolving fund. The CMD programme was added after a devastating outbreak of the disease virtually wiped out the crop and caused widespread famine in one of the project districts. The objective was to import clean/uninfected planting material, multiply this at selected centres and distribute it to farmers for further multiplication and dissemination; extension services were also to be strengthened for this operation, which was designed to cover an area of 5,000 ha.

Expected effects and assumptions

The inputs component would have a direct effect on production, the incremental benefit from the 1.9 million hoes and 30,000 new or repaired ox-ploughs would allow the area cultivated to be extended and the improved seeds, pesticides and storage chemicals should lead to increased yields; it was estimated at appraisal that this would result in production worth an addition USD 9.3 mn annually. The livestock inputs would improve the health and numbers of the herd, and result in increases in milk and meat production worth USD 0.82 mn annually; fish offtake was predicted to rise significantly, giving an extra USD 12.7 annually. Other indirect benefits would come from the institutional strengthening and the various adaptive trials undertaken, and the social infrastructure would be improved by the group formation exercises. Overall theses changes were expected to result in enhanced food security in the whole region and higher farm income levels. A Credit Needs Survey was carried out during project design and this suggested that credit was not perceived by the farmers as being necessary, hence it was assumed that there was no need to include a credit component in the design. Under ARP the programme distributed the inputs to cooperative stores at set fairly low prices, whilst for ADP the distribution was stopped at three regional warehouses, from where inputs could be purchased wholesale by cooperatives and traders; it was assumed that this method would (i) resuscitate the private sector traders, but without leading to exploitation of the farmers and (ii) help to establish realistic market prices for the various inputs.

Evaluation

The multidisciplinary Completion Evaluation mission (CEM) undertook its work in Uganda in November and December 1994. The mission had an extensive field itinerary, holding discussions with the ADP implementing agencies, farmers' groups, and individual beneficiaries. Part of the project area could not be visited because of security problems. The mission was able to assess the reports and other documents produced on the project, and presented preliminary findings to a wrap-up meeting chaired by the Permanent Secretary.

Implementation Context

The entire cycle of implementation of this project was dominated by severe civil unrest. Under these circumstances the start-up of ADP was delayed one year to May 1986, and the project implementation period expanded from four to seven and a half years. The plan of implementation was repeatedly modified to keep abreast of the changing map of insecurity and working conditions, hence the geographic location of project activities and the duration of component implementation underwent significant changes during the life of the project. The project was also restructured in 1991, adding the new components described above; at loan closure 76% of the funds allocated had been utilised.

Project achievements

Input Supplies. With the exception of a few items, procurement whether depicted in volume or value, has exceeded considerably appraisal estimates. However, the CEM was unable to obtain accurate data on the final amounts of inputs eventually distributed to farmers, and it was not possible to estimate losses due to looting and diversion with any degree of accuracy. (The results from a Baseline Survey (1986) and Repeat Benchmark Survey (1988) indicated that there had been significant penetration of ADP inputs even in the very early stages of the project and field investigations during the course of the CEM indicated that the uptake of inputs in the later stages of the project may have been considerably higher.) Production and productivity increases could also not really be assessed for individual farmers, and the effects of the credit component on input uptake could similarly not be gauged. In addition there were some delays incurred by the procurement methods used and also in ensuring the quality and acceptability of the tools to be supplied. Livestock Inputs. In spite of severe constraints it had been reported that 50% of cattle had been vaccinated against rinderpest and CBPP every year of the project, which is a praiseworthy achievement, and largely the result of project impetus. Vaccination of poultry against Newcastle Disease had, however, been disappointing.

Research and Technology Development. Precise information on the adoption of recommendations from adaptive research and on-farm trials was unavailable but the CEM's impression was that, with the exception of a few recommendations, such as rouging of cassava plants affected by CMD, cassava spacing and the use of an improved maize variety, few farmers were using the recommendations as a matter of course. This was not because they were generally unaware of them, but because of lack of finance, physical shortages of improved seed varieties and lack of confidence in markets. The capacity to conduct adaptive research trials had been significantly enhanced. Because of insecurity, research into Tsetse Fly Control did not proceed in 1986-91, but after project reformulation and as security returned, communities assisted staff in setting and monitoring traps and large areas were cleared of tsetse fly. Similarly, after 1991 Fishery Surveys on Lake Kyoga resumed and were well conducted. They provided valuable information on fish catches, and confirmed that overfishing and the use of illegal nets was prevalent (but no action was taken on these findings). The project assisted in the control of water hyacinth/, by training staff in biological control methods, establishing a breeding centre for weevils, and providing tools for mechanical weed clearance. These initiatives, though meagre, were an important start to the national effort to control this weed.

Institutional Development. Surveys conducted to assess the performance of the extension services indicated that farmers in the one district selected for the pilot Training and Visit (T&V) scheme had adopted recommended husbandry practices and technologies more than those in the other six districts, which used traditional extension methods. The pilot extension system interacted with only a modest number of beneficiaries but whenever it did, changes in farmers' husbandry practices occurred and these contributed to increased production. The performance of the Media Resource Centres (MRC's) established under the project was satisfactory, however, in spite of the substantial amount of material produced, the CEM saw little evidence of it in the field; the lack of extension aids was a common complaint encountered. Without ADP's contribution, the training facilities in the project area, including the District Farm Institutes (DFIs), would not be functioning. All rehabilitation work and the necessary refurbishing had been completed and all the facilities were currently being utilised.

CMD control. Multiplication Centres (MCs) had been established in the designated areas, using CMD tolerant planting material. Although this constituted a positive step in combating CMD, varieties giving total resistance would have been preferable. The propagation and spread of CMD tolerant and resistant varieties in the project area is now being addressed by the Northern Uganda Reconstruction Project (NURP), as well as various NGOs.

The Credit Component was originally designed to operate in just three districts, but due to political pressure, was expanded to cover over 70% of the project area. This meant that administrative, technical, managerial and financial resources were spread very thinly. In spite of a concerted effort to extend credit to farmers, only about 40% of funds allocated to rural credit have, in fact, been distributed. The group formation task progressed more or less according to the programme, depending primarily on cooperative societies and female-oriented savings mobilisation schemes. At project completion the number of beneficiaries organised into groups (in particular women) had exceeded the targets set. The pilot credit component established an institutional/operational arrangement which though ad hoc in nature provided credit to farmers and rural entrepreneurs for two seasons. This structure left much to be desired and had it not been for the World Bank's project in the same area (NURP), it would not have been sustainable.

The original design advocated ensuring beneficiaries commitment to repayments through group training and dependence on peer pressure, however the recovery rates have been considerably lower than the 95% anticipated, in the range of 45-70%. The prevailing culture, that credit is merely financial support from the government, contributed to these low recoveries. To improve the situation the Credit Component Management Unit attempted to add collateral to the credit contract, but in the absence of land titles only moveable assets could be pledged. Notwithstanding these constraints, some useful experiences in rural credit delivery have been gained, particularly related to social organisation and group formation.

Counterpart funding was a chronic problem prior to 1991; the Treasury tended to disburse approximately 50% of requirements, but on occasion (1988/89) this fell as low as 8%. This situation was completely rectified in the later stages of the project (100% counterpart funding in 1993/94) by, firstly the donors agreeing to meet upto 75% of the local costs, and secondly by GOU allowing all the proceeds from the input sales to finance the counterpart allocation.

Effectiveness of the M & E system

The M&E system has delivered a substantial volume of work, in spite of the security problems and shortages of staffing and equipment in the earlier years; in terms of institution building, it has been one of the more successful activities. The Monitoring and Evaluation Unit (MEU) not only provided information to the management of the project (although not always on a timely basis), but also assisted the government in designing sectoral and national development plans and has now become the core of the proposed MAAIF ministry-wide unit. However, its impact has been constrained by (i) an absence of specific targets in the project design documents, (ii) lack of baseline data, and (iii) lack of a unified and consolidated project management for a considerable part of the life of the project.

During the first few years of project implementation, monitoring of the T&V extension pilot activities was only partial and of limited value, because of failures to take advantage of some simple recordings, the poor presentation of results and the absence of a non-contact control farmer group. Incorporation of a control group would have provided a critical indicator of message diffusion, and hence the effectiveness of the delivery system. The MEU has also compiled very little information on livestock activities, because staff were not asked to provide the required information in their regular reports. This will make it difficult to gauge the extent to which livestock, tsetse and fisheries inputs and surveys contributed a lasting impact on food production and household earnings.

Effects assessment and sustainability

Crop Inputs. The final survey of the project showed that most of the gains in agricultural production have been the result of the expansion of cropped areas. The project provided substantial amounts of essential inputs at a period during which farmers had very little access to these materials, and also took prompt action in supplying disease free resistant varieties of cassava, which had an important effect on food availability to the rural population, particularly the poor. It was estimated by the CEM that over 60% of livestock farmers benefited directly from livestock inputs, but the effects of these veterinary inputs would have been greater if farmers had received husbandry and management information at the same time. After project reformulation more emphasis was given to small stock management but the project ended before much could be achieved in this respect. The overall effects of the fisheries inputs were reduced because they arrived late, but when they became available, fish catches improved significantly and family incomes clearly increased despite a reduction in fish prices, though precise figures were not available. The water hyacinth control measures had negligible effect due to the vast increase in the weed during project life. The tsetse fly control measures had a definite, though unquantifiable impact in 1991-93, and large areas were cleared of the menace of the flies.

The pilot credit programme has had some noted effects. Firstly, the emphasis on rural women, their organisation and enhancing their creditworthiness, especially in the Ugandan social context has been highly appreciated by the women themselves, and secondly, there are indications that credit has contributed to the increase in areas planted, restocking, and the expansion of markets for various agricultural and fisheries products. Overall, rural households have increased food self-sufficiency as a result of credit being available for the purchase of inputs. The project's support for women's groups has clearly had a positive impact on women through such activities as training, group mobilisation and credit, and has contributed to improving women's productive capacity and in maintaining their enhanced social status. The project's social organisation model was based on the largely non-functioning cooperative movement, which was a high risk strategy given the inheritance of the cooperatives, and groups were also formed hastily.

Although the years of unrest had caused a flattening of the socio-economic differentiation in the project area, the CEM perceived that there were still clearly underprivileged groups, determined by access to land, education and position in the clan. It was not clear who had actually benefited from the ADP, neither was this built into the M&E system as a measurement factor. Project awareness was, however, quite high (over 50%), although it was twice as high in men as in women. The CEM commented that project benefits appeared to have gravitated more towards those groups and individuals with greater influence in the communities, especially after the restructuring which took place in 1991, and led to much higher prices for many of the inputs. The T & V system was dropped in 1991, as it was seen that it was bringing main benefits to the more "progressive" farmers, which were also the larger and more influential farmers. The only specific targeting in the project was for women and credit, but even here the CEM noted that the funds appeared to go to the more influential groups. Nevertheless, group formation was quite spontaneous when there was a specific objective, and the project's training has reinforced group activities generally.

The provision of project financed vehicles, office equipment and some buildings provided immediate relief for the Ugandan administration in the area of rural development. In addition, extensive training has upgraded the capacity of large numbers of civil servants who, for the major part, remained in government service.

In the immediate future, some ADP activities have been incorporated into the IDA funded projects such as Northern Uganda Rehabilitation Project (NURP), Agricultural Extension Project (AEP) and Agricultural Research and Training Project (ARTP). While these will provide funding in the short to medium terms, they will not secure the sustainability of ADP activities in the long run. Currently GOU does not have enough financial resources to support project activities unless they are included in the Government's "core" programme or they become self-sustainable through privatisation and/or efficient cost recovery.

Main issues and recommendations

The major recommendations from the CEM were as follows:

• Policy dialogue with GOU should seek to ensure that (i) poverty alleviation constitutes an important criterion in the selection of core projects (i.e. those with a higher priority to receive GOU funds), and should address the issue of regularising input and product markets;
• The privatisation of the delivery system of input supply should be enhanced at all stages, starting from importation. The procurement should be undertaken by private established dealers, who would compete for the foreign exchange allocated for input supply.
• - There is a need of careful procurement planning. Procedures should be streamlined to ensure timely delivery of quality goods and services.
• Being the staple diet especially of the rural poor, developing and disseminating CMD resistant varieties of Cassava should have a high priority. The search for CMD resistant varieties with good yield potential, palatability and without additional processing costs should be continued.
• In the fisheries sub-sector policies and measures are required to ensure: (a) the control of overfishing, for example by the introduction of fishing nets with a proper mesh size; (b) the improvement of marketing outlets, particularly roads and means of transportation for fresh fish; (c) the preservation of excess fish through smoking and (d) the promotion of fish farming, by training farmers in pond management and feeding methods
• Tsetse fly constitute a very important health hazard especially to members of poor households and herders in the bush. In future projects located in areas infested with tsetse fly assistance should be provided to the Uganda Trypanosomiasis Research Organisation.
• With regard to water hyacinth control, the CEM recommends that it should be undertaken within a context of regional cooperation involving all countries in the Nile basin. It would be extremely difficult for IFAD to assist in a substantial way except, perhaps, in the training of technical staff.
• There is a need to properly assess the demand for agricultural inputs through farm surveys which would, in addition, include the farmers' preferences for various brand names. These need assessments should be undertaken in the context of an overall crop production and livestock survey.
• The GOU should adopt policies that would ensure the financial viability of credit operations. This means addressing the savings threshold necessary to obtain a loan, and the wide disparity between the interest paid on savings and interest charged on loans. In addition, ambiguity in the legal status of borrower groups, makes legal action for the collection of debts complicated.
• The future of some input supply activities has been adversely affected by the protection of inefficient local industry, which has led to disappointing results as local manufacturing (and distribution) capacity appears to be well below farmer's needs. There is a need to balance the promotion of local manufacturing with the welfare of the farmer.

Lessons learned

Counterpart funding was a major problem for ADP; the situation improved considerably after reformulation, when the GOU authorised the use of the proceeds from input sales to fund project activities. At appraisal a more realistic estimate of GOU's capacity to fund project activities should have been made, and this assessment should have been reviewed as part of the CI's role. In similar circumstances or projects more attempts at providing avenues for local financial support of project activities from within the project should be made.

Extension messages should be suited to the social and financial realities of communities in the project area. Farmers needs and priorities should first be assessed, and schemes determined to meet those needs - this did not happen under ADP. Until profitable cropping ventures are identified and efficient marketing channels are developed there are limited incentives for farmers to adopt improved technologies if they entail additional initial outlays. For example, the CMD resistant varieties were met with enthusiasm and similarly fisheries inputs, but pig raising was met with reservations.

The project's impact on livestock would have been greater if some attention had been paid to animal husbandry and management practices, at the same time as animal health interventions. Complementarity would have led to increased benefits, and could have been undertaken with very limited, if any additional costs. Small stock development was introduced in the later stages of the project and was particularly important for the poor and women (being an activity which required only a small amount of credit and with a fast turnaround); the design of future IFAD projects with livestock improvement components should not exclude interventions for small animals.

Where farm tools and implements are being supplied, they should be tested in the local environment, and certified as being acceptable by the relevant authorities; equally important is the certification of the quality and expiry dates of chemical inputs and drugs. Care should be taken that imported inputs, e.g. seeds and chemicals, are packaged in accordance with farmers needs, otherwise repackaging should be introduced.