Mid-term Evaluation

Although agriculture is the dominant sector of Uganda's economy, less than 30% of the total arable area is presently under cultivation. The country suffered considerably in the 1970s under the effects of political mismanagement, and the 1980s were characterised by efforts to restore the economy, which included large successive devaluations. In 1990 the Government of Uganda (GOU) obtained a major agricultural sector credit loan from IBRD, but inflation continued at high levels, estimated at 50% in 1992 (at the time of the MTE). The Southwestern Region, which borders Tanzania, Rwanda and Zaire, had a population in the 1991 census of 2.66 million people, in about 400,000 small farming units. The terrain ranges from rolling savannah to high mountains; rainfall tends to increase with altitude, and varies from less than 900 mm to more than 1350 mm in the highlands. Mixed farming predominates in the area, and three cropping regimes can be identified, depending on altitude. In the high altitude mountainous areas, which comprise steeply contoured hillsides, double cropping is common, and crops include maize, sorghum, millet, beans, sweet potatoes, and Irish potatoes. At mid altitudes similar crops are grown, but the main food crop is bananas, and coffee is an important cash crop. In the lower areas, where rainfall is least, the soils are quite poor and this area is mostly used for extensive grazing of cattle, sheep and goats, with very small household plots of food crops (bananas, cassava, maize, groundnuts and sweet potatoes).

Project design and objectives

IFAD's involvement in Uganda began in 1981, when food supplies were still severely disrupted after years of turmoil and destructive civil strife. As a result IFAD's emphasis was on the production of food crops and food security measures. Two ongoing projects in the mid 1980s, the Agricultural Reconstruction Programme and The Agricultural Development Project concentrated on input supplies and aimed at the rehabilitation of basic agricultural services and the provision of limited credit. The IFAD-initiated Southwest Region agricultural Rehabilitation Project (SWRARP) was intended to continue and enlarge these basic themes.

Target group

The population of the Region comprises six interrelated tribal groups, all speaking the same vernacular. The average household size is seven persons, and family groups operate smallholdings of 1-2.5 ha; there is an increasing problem of land scarcity with the result that farming methods are becoming more intensive and steeper land areas are being pressed into production. Income levels in the Region are about USD 100 per caput, which is less than half the national average. Although poor, nutrition is not a major problem throughout the Region. The target group consisted of virtually the whole population of the Region, as most people would benefit from the large road component by improved public transport, better marketing opportunities, and easier access to services, however the appraisal specifically estimated that over 28,000 farm families would benefit from the extension and research component, and over 270,000 farm families from the input supply component. Food crop production is predominantly undertaken by women; other tasks are collecting water and fuelwood and marketing. Since women constitute over 70 % of the agricultural workforce, they form the largest section of the target group.

Objectives and components

The main objectives of SWRARP were to: (i) increase food security and improve nutrition through increasing production; (ii) improve the incomes and living standards of small farmers in the Southwest Region, who constitute the large majority of the population; (iii) strengthen rural institutions; and (iv) provide the technological base to improve agricultural productivity and halt environmental degradation. These goals were to be achieved through support for adaptive research and extension, procurement and sale of inputs, rehabilitation of rural access roads, and support for project management including a small community development fund with a credit component. Total project base costs were estimated at USD 23.4 mn. and contingencies added a further USD 3.8 mn. The project consisted of six components, as follows:

Rural Access Roads USD 11.19 mn (48%)

Agricultural Input Supply USD 6.95 mn (30%)

Adaptive Research and Extension USD 3.36 mn (14%)

M & E USD 0.65 mn (2.6%)

Management Support USD 0.82 mn (3.5%)

Community Devt/Small-scale Enterprise Fund USD 0.40 mn (1.9%)

The rural access roads component was to be financed mostly by IDA (64%) and the GOU (20%), IFAD was more committed to the financing of the agricultural inputs (82%), the M&E (67%) and the management support activities (53%). The project duration was expected to be five years.

The rural roads programme was intended to provide engineering and financial support for a programme of rehabilitation and repairs to critical parts of the 2000 km of rural access roads not covered by other projects, and to provide strengthen the Ministry of Local Government's (MOLG) district-level road maintenance capacity (with equipment, tools, operating expenses and training). The identification of roads to be included in this programme was to be undertaken by means of an inventory and infrastructure evaluation study as part of the project start up. Criteria for selection included socio-economic aspects, the requirements of the agricultural extension and research programmes and linkages with other road rehabilitation projects; the work was due to begin in year two and be completed by year five. The agricultural inputs supply component was concerned with the procurement and sale of hand tools, seeds, bicycles, wheelbarrows, small threshers and flour mills and agricultural chemicals, and also with equipment (vehicles, materials etc.) needed to improve the marketing facilities in the Region.

The adaptive research and extension component was intended to establish adaptive research programmes at selected stations, and included an outreach programme for farmers. Emphasis was to be placed on the major food crops of the Region, with soil conservation and fertility improvement being priority factors. The Regional Home Economics Units of the MOA would be integrated into the extension component in order to provide a channel to and from women's groups so that women's needs could be catered for in the programmes developed. Technical Assistance and training would also be provided under this component. As most institutions were moribund as a result of the civil disturbances, institutional strengthening was required in a number of ways to assist project implementation and sustainability. Management support was to be supplied to the road maintenance units of MOLG, the extension and research service, the MOA and the Monitoring and Evaluation Unit, which would be established in the Regional headquarters of the MOA. In addition community development through the formation of appropriate groups and the promotion of small-scale entrepreneurial activities (which included the provision of medium term credit) would be undertaken by seconded staff from the Ministry of Community Development; the project would support the organisation and training of these groups and would provide small equipment and materials to assist with the start-up cost of entrepreneurs.

Expected effects and assumptions

Project benefits were expected to flow from expansion of the cropped area, the increased availability of inputs, increased productivity (from extension and research activities) and improved market access as a result of road rehabilitation. Most benefits to farm families would be in the form of better food security (incremental food production was estimated to reach 120, 000 mt by year 5). Transport costs were expected to decline by between USD 0.05-0.20 per mt per km, giving the possibility of higher farmgate prices. Cash incomes were expected to rise by about USD 117 per farm family (up 74%) for food secure families and by about USD 83 (up 36%) for farm families in food insecure areas/. The most important assumptions made in the project design concerned the GOU's ability to (i) procure the large quantities of inputs needed and (ii) to complete the tendering procedures required for large (road) contracts in a timely manner.

Evaluation

The MTE mission took place in October and November 1992; the multidisciplinary mission met with central government agencies and other donors in Kampala, then split into four groups for field visits to cover (i) adaptive research and extension, (ii) community development and women (iii) roads, buildings and credit, and (iv) management and institutions. Discussions were held with project staff, government field staff involved in the project, farmers and other beneficiaries. Final meetings took place to discuss the findings both with the project management in the Region and also the government agencies in Kampala.

Implementation context

Many external factors, some related to the country's efforts at restructuring and rehabilitation following 15 years of turmoil, had a delaying effect on the project. Among these were:

• the government's decentralisation programme, which required time-consuming reorganisation of the project management structure;
• the civil service reform, which prevented the realisation of certain objectives of institutional strengthening and rehabilitation;
• the varying performances of the Cooperative Bank and the Uganda Commercial Bank (UCB) over time, requiring repeated reassessment of possibilities for collaboration in the pilot credit scheme and causing delays in scheme start-up;
• the rapid devaluation of the Ugandan Shilling, from U sh 60 per USD at appraisal in 1987 to U sh 1200 per USD at the time of the MTE, with its effect on project local costs and the counterpart funding burden on GOU;
• the 1989 collapse of coffee prices (the main export crop) and consequent reduction of government revenues;
• the war in Rwanda, and its impact on the southernmost districts of the project area.

Project achievements

Due to excessive delays at start-up and during implementation, the project was at least two years behind schedule. The reasons for this were processing and procurement delays, delays in establishing institutional operations and snail-paced disbursement, due in part to inadequate provision of local counterpart funding. The project's largest component, rural roads rehabilitation, was yet to begin, and the most optimistic date for start-up was spring 1993, a full four years behind schedule. However, project implementation appeared to be taking a turn for the better, partly as a result of the recent arrivals of a new Financial Controller and a new Project Coordinator (PC). The new PC had already started streamlining management and coordination activities, and financial controls had improved significantly, with the management information system developed by the M&E Unit being effectively utilised for management purposes. However, under the present circumstances SWRARP could not meet the closing date of June 1994.

GOU funding has proven to be a major delaying factor in implementation; at the time of the MTE the GOU had only provided 33% of the funds required by the loan agreement for the agricultural component. The procedure for releasing funds was complex and led to considerable delays. The situation has been worsened by government cash flow problems and the continuous devaluations and high inflation experienced during project implementation. The UCB, appointed as procurement agent, proved unable to fulfil this role without the benefit of a proper procurement unit. Procurement delays, such as the late arrival of vehicles, equipment and inputs, have severely delayed project implementation. According to UCB, the procedure for ICB procurement needed at least 10-12 months, but frequently took more time (up to 25 months in one case).

Feeder Roads and Buildings. There was no physical progress on the construction of feeder roads at the time of the MTE. Without the feeder road rehabilitation component, the full potential of the input distribution, and research and extension components cannot be realised. Long delays occurred in the appointment of consultant engineers, completion of the Roads Inventory Document and preparation of tender documents. The start of the actual road rehabilitation was four years behind schedule at the time of the MTE, and as a result actual cost estimates have risen so much that the target has been reduced from 2000 km to just 763 km. With the slowness of progress in the procedures to subcontract rehabilitation works, a plan was prepared for some road rehabilitation to be undertaken directly by the project. The project was also to provide funds for the renovation and construction of offices, staff housing, research station buildings and a warehouse for inputs. However, due to GOU liquidity constraints, the building programme has been divided into two phases; the first phase was near completion now, and phase two was expected to start in 1993.

Input Supply and Distribution: The objective of this component was to provide agro-inputs to small farmers in the Region in a timely and accessible manner. By the MTE, two procurements had been completed and over 90% of the inputs procured had been sold out of rented warehousing facilities. The combined third and fourth procurements were expected in early 1993. In order to promote input supply privatisation, market-based pricing of inputs and input retailer registration were included in project design. While the registration of input retailers by itself does not represent a major contribution to privatisation, mobilising and training retailers is an important contribution to sustainability of the service. The project's most important contribution, however, will come from its effect on the expansion of input and product markets.

Adaptive Research and Extension. The activities of the adaptive research unit have been satisfactorily implemented for five seasons; the programme has been able to recommend adoption by farmers of several improved varieties of Irish potato, wheat and maize, a chemical treatment for banana weevil, the use of multi-purpose trees in agro-forestry activities and low-cost seed storage facilities. The recommendations include the alternatives of low and high-input technologies. Outreach (on-farm) trials have also been carried out to test research station recommendations. The extension sub-component has been less active and further behind schedule due to poor mobility, lack of training and shortage of operational funds. However, a training programme addressing the needs of both extension staff and farmers was in place and a supervision mechanism had been developed, although to date, extension activities have been limited. An action plan to integrate SWRARP's extension activities with the National Extension Project supported by the World Bank was being prepared.

Monitoring and Evaluation. By MTE the monitoring and evaluation system was well established. Staff has been adequately trained, the MEU has received 19 months of apparently high quality technical assistance and transport and computer facilities are satisfactory. A number of useful surveys and activities have been implemented by MEU including a baseline and a mid-term household survey (which, however, are not structured in a way that permits direct evaluation of project impact), a well-prepared mid-term evaluation report, and quarterly input and marketing surveys. Because the day-to-day monitoring system was not having the anticipated effects, an "M&E awareness seminar" was organised, which improved understanding of the purpose of the monitoring system. There appeared to be a strong collaboration between the MEU and project management, with regular meetings held to discuss MEU findings and PMU needs: both sides appear satisfied with this interaction.

Community Development. This component has two parts, a grant fund of USD 95,000 to assist 40 groups selected by the project, and a loan fund of USD 500,000 for the establishment of a revolving credit fund to finance individual income-generating activities. This component suffered from a very delayed start, which, in addition to general project implementation delays, was largely due to its hasty inclusion in the project design after appraisal. Extensive group training activities have been carried out energetically and effectively, mostly through day courses. A Pilot Credit Scheme has been added during project implementation.

The Women Organisation Unit (WOU). Although not provided for at appraisal, the PMU established the WOU to facilitate the execution of women's activities in the project. The unit is very active, having organised over 140 day courses in various subjects, mobilised women and women's groups to take part in the on-farm adaptive trials of the ARU, assisted women's groups to participate in project input distribution, and developed a network of contacts for funding of women's small-scale income generating projects. In addition, the WOU monitors all other project units to promote the involvement of women in all SWRARP activities.

Effects assessment and sustainability

In view of the delay in project implementation, it was too early at the time of the MTE, to judge the extent to which the project was achieving its objectives. However, where feeder roads had been reopened and rehabilitated (by other projects), farm-gate prices had increased significantly, inputs had become more available on the local market, in part due to project importation, and a significant network of retailers was being consolidated. This argues for the continuation of the feeder road component, even though it is now totally unsynchronised with the other project activities. The adaptive research component has assisted in the introduction of improved varieties of various food crops and the control of some diseases, although, despite considerable investments in the extension service, more efforts are needed to disseminate varieties, improve farming practices and induce farmers to use modern inputs including fertilisers. A lack of credit has constrained farmers' ability to acquire inputs or adopt improved practices. The community development component has significantly assisted in the mobilisation of community groups and provided support to limited numbers of groups.

Targeting The assumption in the appraisal that all households in the Region should form the target group, because of the low income levels, was not confirmed by the baseline survey, which identified families with very small land holdings or no land at all. No suitable targeting mechanisms were included in the project design specifically to reach such groups. The criteria for road selection for rehabilitation did include social targeting, so these roads should be in the poorer areas of the Region. Lack of GOU funds has resulted in extension workers only being able to work in the location adjacent to their houses, which could lead to penalising of more remote areas.

Comparative Indicators Comparison of relevant indicators (these covered such items as meat consumption, radios, bicycles etc.) included in the baseline and mid-term surveys suggested a rise in household incomes. Such improvements cannot solely be attributed to the effects of the project, because the overriding factor has been an improvement in the security situation in the Region, but nonetheless a positive trend is indicated.

Effects on Women The project's main effect on women has been through the group formation schemes implemented by the Community Development Support Unit and the Women's Organisation Unit. In the project area women do not have ownership or inheritance rights to any type of property. The baseline survey found that FHHs (14% of the households in the project area), had less land and were generally poorer than other households; this section of the community is especially vulnerable. The self-help groups formed by women tended to be aimed at productive activities, including communal cultivation of members' fields; the project has assisted by providing training in a wide variety of subjects, from agriculture to child care. There is evidence that women's social position has also improved as these economic activities have become accepted, and that the groups are providing a platform to give women a stronger voice in community affairs.

Environmental Effects. The Extension Unit of the project has established soil conservation demonstration plots on farmers lands, but in general these have been poorly managed and used because of inadequate technical knowledge, and shortage of transport. However there is general awareness and concern about soil conservation matters. There is also a shortage of firewood in some areas of the project, with the result that animal manure, which was used as a soil conditioner, is being used for cooking. In some areas intensive farming methods, involving multiple cropping, are leading to depletion of the soil fertility.

Agricultural Production The field distribution of inputs commenced in 1990; at the time of the MTE nearly 50% of the total quantity of project inputs had been dispersed. Despite this, there were no indications of major improvements in production, as yet. This is probably because the extension messages which might influence seed choices and use of fertilisers are so far very limited: the factor which emerged from the latest survey as being the greatest influence was farm gate prices, which directly points to the need for improved marketing. The extension service has yet to deliver to farmers the technological packages developed by the adaptive research; the majority of farmers in the project area still believe that the only way to increase production is to expand the area planted. This indicates lack of awareness of the potential of the newer varieties and practices which have been developed. No revised yield data is as yet available for the project area, hence the potential impact of this component (or changes in incomes) could not be surmised. However, the evaluation noted that newer methods of row planting and conservation were being used in some instances, along with manual methods of weevil control in bananas, indicating that some extension messages had been received.

Roads Rehabilitation. Because this component has not started, there were no reported effects.

Community Development. Group mobilisation and training under the project have had a significant effect in the project areas. Membership of groups has increased to nearly 50% of women and 24% of men (results from the mid-term survey). In addition, the project has given increased support and training to existing groups. The fact that grant were available for community groups, once formed, may have been a stimulus to this development.

The issue of sustainability cannot be over-emphasised. Several of the project's activities will have no lasting impact on the Region unless effective ways are found to: pass project management responsibilities to the appropriate government offices; sustain incentive payments to ensure continuity of government services; maintain the quality of financial management and control after the termination of the technical assistance; continue the supply of inputs to project beneficiaries; and, finance and ensure the maintenance of rehabilitated roads.

Main issues and recommendations

The major recommendations put forward by the MTE were as follows:

• At this late stage the procurement system should be retained, bu there should be much greater collaboration between the project's input supply unit and the UCB to try and expedite procurement processes.
• The ARU should continue to concentrate priority crops and farming systems only. On-farm trials should be extended to areas where extension support is adequate.
• The extension service, with its limited resources, should give greater priiority to working in selected areas and should emphasise low-input technology packages, e.g no fertilisers, animal manure composts, row planting techniques etc.
• During the final year of the project the balance of five months of M&E related TA should be used to assist MEU with the design, implementation and analysis of a completion household survey and to provide guidance on the finalisation of the sustainability study.
• Participatory evaluation techniques, in which beneficiaries (and non-beneficiaries) are asked to give their own evaluation of project activities, need to be introduced into impact evaluation exercises such as the household surveys and planned beneficiary contact surveys.
• Strong emphasis should continue to be placed on group training activities as the most effective use of limited resources and limited training and extension staff in reaching the largest number of farming men and women. Ways should be explored for using existing networks and institutional resources to increase reach and coverage of group training.
• Within the 40 selected groups, further training should be given in basic planning and management to back up the grant funding and the future credit scheme. The groups and group members should be encouraged to express their needs and constraints more freely and democratically, having a voice in project decision-making where it concerns them rather than being mainly receivers of project advice and financial assistance. For these groups and for all group training, there is a strong need to promote the concept of self-help, and to attempt to reduce the mentality of dependency on projects and free goods.
• The SWRARP Pilot Credit Scheme should coordinate with and use the experience of current credit activities of other IFAD and IDA projects in Uganda, especially the ADP pilot credit scheme and the credit component of the Hoima project.
• The World Bank rule stipulating the purchase of generic products has partly resulted in a quantity of unsold chemicals as farmers are reluctant to purchase chemicals of unknown brands. SWRARP must step up its publicity campaign to familiarise farmers with the new chemicals. The availability of credit and demonstration activities by the extension service would help to promote the use of these chemicals.
• To enhance input supply privatisation, the government should contribute positively in three ways: (i) financial intermediation at all market levels, but particularly for the importers in terms of foreign currency and credit, and for farmers in terms of rural credit; (ii) phasing out of public sector importation of agricultural inputs. (In cases where grants of inputs are donated in kind to the country, their distribution should be market-price based to avoid subsidisation. It would be appropriate to auction these commodities to private dealers); (iii) drawing up and enforcing strict regulations on the agro-input trade to ensure quality to farmers, reduce hazards and protect the environment.
• Future variety trials should place additional stress on disease resistance, (especially with respect to Irish potato), as well as shorter maturity periods, low-input technology, and the study of the impact on banana yield and quality and on soil degradation of intercropping nutrient-depleting crops, such as maize, cassava and potatoes, under banana plantations.

Lessons learned

When the Co-operating Institution is also a co-financier, then different institutional mandates and methodological orientations can result in insufficient concern by supervision missions for the issues which are at the heart of IFAD's work in rural development. In the case of SWRARP, the World Bank's own role in the project (mainly concerned with the financing of physical infrastructure) and in Ugandan national development efforts in general (mainly concerned with macro-economic policy issues) affected the priorities given in supervision activities.

The SWRARP case shows that without detailed road design studies, the appraisal of this component is ineffective. The implementation schedule of the feeder road component was drawn up without a proper road inventory or feasibility studies, and the result has been overly optimistic planning and inappropriate phasing of other project components..

The component for women's activities was introduced by the project management, following successful group formation exercises. It is clear that women should have been a specific target group in the original design, and this points to the dangers of targeting being too general, as in this case, and missing not only an important group, but not identifying the greater development potential afforded by this group.

This project was severely affected by lack of or untimely release of local funds. It should have been clear from quite early on, by holding a dialogue with GOU, that this situation was going to affect the implementation of the project. There seems to be no point in merely stating that the government is not meeting the requirements of the loan agreement, a much more positive approach is required which might involve rescheduling project activities or agreeing to meet a larger proportion of the local funding costs.

Implementation schedules need to be based upon realistic estimates with regard to the time required for governments to complete procurements, if project phasing is not to be adversely affected. It appears that appraisals do not give this sufficient priority, or bother to check the actual procurement record with other projects or donors.

The community development component showed that the criteria for group selection for project assistance need to be developed with a clear understanding of who will be excluded. In an attempt to guarantee results, project designers tend to incorporate prohibitive conditions for group eligibility. Care must be taken to avoid excluding poorer groups which have the potential to become valid and viable organisations with the help of outside facilitators and trainers.

The most important stimulus to encourage the adoption by farmers of new crops or farming methods is the farmgate price. There is no point in developing the most complete extension messages and training extension agents unless the stimulus of prices is also correct.