Interim evaluation

Ghana has a total area of about 240 000 km2, of which the Upper East Region (UER) covers about 8 800 (3.7 % ) in the Northeast corner. The total population with annual growth rate of 3%, has been estimated at 17.5 million in 1996, of which an estimated 1.1 million or 1.15 million in 1998 live in UER. The rate of population growth is about 3% per annum. Population density averages 63/km2, with considerable regional variations. The UER's mean population density is 125 people/km2 and in rural areas is about 100 people/km2. UER is predominantly rural (87%). The country can be divided into three main ecological zones, going from south to north on the basis of soils, climate and vegetation. These are the rainforest, the transitional and the savannah zone, which includes UER. Despite the predominantly gentle slopes, about 70% of the country is subject to moderate or severe erosion, UER being, one of the most erosion prone regions. In the Upper East, rainfall over the past 40 years has averaged 1 044 mm, suitable for a single wet season crop. The very short wet period is marked by confusing fluctuations in arrival time, duration and intensity of rainfall. This, together with the poor water retention capacity of the soils, cause large inter-year variations in the production potential. Administratively, the region is divided into six districts which correspond roughly with the main tribal groupings. The area is served by a network of access roads, the condition of which has been classified as fair to good.

Project design and objectives

The project area is characterised by a high and expanding rural population; by the unreliability of its rainfall and susceptibility to drought; and by the inexorable rate of degradation of its environment. The farming systems used were still largely based on the traditional shifting cultivation approach of using long season crops combined with the bush-fallow system to maintain soil fertility. However, over time fallow periods are constrained by population pressure on the land. Only by developing a location specific programme of carefully designed and supervised trials would the problems of environmental degradation be adequately addressed. There is a need for providing irrigation of high value crops as well as for livestock. During the 1950's and 1960's many dams have outlived their 20-25 years design life, leading to their deterioration aggravated by lack of maintenance. Hence, the rationale for dam rehabilitation and dugouts construction.

Target group

The target group for the project is defined as those sections of the rural population who are most at risk from malnutrition, ill health and general low quality of life. Rural people are drawn into ‘at risk' category due to a combination of factors including lack of resources and access to adequate basic services and facilities. Absolute poverty was defined as "that income below which a nutritionally adequate diet plus essential non-food requirements is non affordable." For Ghana, on a whole, it was estimated that in 1986 67% of smallholders, defined as those farming less than 4 acres (1.6 ha), were living in poverty i.e. below the Basic Need Income (BNI). Given the lower productivity of UER upland farms, a more representative figure for UER would be 6 acres (2.4 ha). Hence, virtually the whole of the smallholder population of the project would fall within the target group on these general criteria. However, within a generally impoverished rural society, there are those families, including women-headed households, who faced substantially greater hardships than their neighbours. Accordingly, the determinants of access to land and of notional farm income have been used to establish a framework within which specific identification of the more disadvantaged families may be made and project resources be channelled to them.

On the basis therefore, of a holding of 2 ha or less and of the inadequacy of alternative employment, it is estimated that some 70 000 farm families or some 66% of the rural population are severely disadvantaged and would constitute the target group. Project initiatives would directly benefit about 40 000 families or about 60% of the target group through the proposed crop and livestock development and the promotion of income generating activities (IGAs).

An estimated 11 000 households or 20% of the rural population, in UER are headed by women. In reality, the number of women who cultivate is probably higher than official estimates due to seasonal emigration of mates, operation of separate accounting households within polygamous marriages and non-residential marriages. Although women do not have rights to own land, they are invariably allocated plots. In all households, women provide labour for their husbands' fields and cultivate their own land in both the compound and bush farms. For women-headed households, the major constraint is lack of labour for the range of heavier farm tasks.

Objectives and components

Objectives: The principal objectives of the project were to: (a) increase food production and household incomes of poor rural families; (b) establish mechanisms for environmental protection and improvement;

(c) strengthen formal and informal beneficiaries organisations; and (d) improve the economic status of women beneficiaries through the provision of credit and support for IGAs.

Components. To achieve these objectives the project included the following components: a) Water Resource Development (31% of total cost) - ensuring water supplied to about 350 ha of existing and 205 ha of newly irrigated lands through the rehabilitation of existing small dams or reservoirs and the formation and strengthening of WUAs; b) Land Conservation and Catchment Area Protection (9% of total cost) - focusing on protection and stabilisation of the catchment areas of dams through contour bunding, planting grass on bunds and on livestock exclusion zones, planting trees, hedges and fodder crops and the development of a comprehensive environmental protection policy for the area; c) Group Formation and Development (7% of total cost) - promotion and training of small homogenous producer groups as vehicles for the delivery of project goods and services to the target group; d) Rural Credit (28% of total cost) - finance of crop production and agriculture-related small-scale enterprises, mainly targeted to women heads of households and women's groups; and e) Institutional Strengthening (16% of total cost) - directed towards various MOFA regional departments of agricultural services and other GOG agencies.

Expected effects and assumptions

Project interventions were expected to result into:

increased agricultural production in the UER from wet season cultivation through generation and adoption of appropriate technical packages (improved seeds, early maturing varieties, fertilisers etc);

increased production of high value crops (onions, tomatoes and peppers) from dry season farming due to dam rehabilitation and expansion in command areas;

improvement of livestock production through provision of drinking water (dugouts), expanding grazing areas, and selecting optimum genetical mix;

enhanced equity for rural women, particularly in allocation of additional farm land accruing from the rehabilitation of dams;

improved capacity for income generation activities through provision of rural credit to women members of credit groups;

integration of the project management and activities into the regional and district structures and in the process support the decentralisation policy; and

enhancing sustainability of project benefits, for example with Water Users Associations (WUAs) taking a leading role in catchment area protection and credit groups in repayment of loans.

The main assumptions for such expectations are:

technical packages are readily available both for wet and dry season crops and that the capacity to deliver these messages to the farmers is at hand;

the physical supply of agricultural inputs is not constrained, and that the input/output price ratio will continue to be favourable for the producer to adopt and apply the recommended technical packages;

IDA will remain committed to the dam rehabilitation programme and WUAs will be established and fully participate in catchment area protection, provide labour to the construction works and install cost recovery mechanisms for post rehabilitation maintenance of the sites;

an effective credit delivery mechanism will be established, such administrative structure will be highly disciplined. Credit groups will be formed, properly animated and developed and members promptly repay their loans; in addition an operational revolving credit fund will be established from group savings and loan repayments; and

the project management and activities will be fully integrated into the regional (and district) government structures.

Evaluation

On projection completion, both the government and IFAD indicated willingness to continue their co-operation for the development of a highly depressed, impoverished and food insecure region. Consequently, an Interim Evaluation Mission (IEM) was sent in February-March 1998 to evaluate the effectiveness of project interventions, identify implementation constraints and from thereon, formulate clear sets of recommendation and lessons learnt to assist in defining the scope of the next phase. The mission divided into four teams on management, credit, farming systems and community development. These teams conducted extensive interviews with regional and district authorities, project personnel, beneficiaries, officers and members of WUAs and credit groups, as well as the officials of the participating banks (PBs). Appropriate sets of questions were designed for the interviews. In as much as possible, the sample of interviewees was drawn at random.

Implementation context

Poverty in Upper East. The characteristics of UER, such as high rural population and the reliance of the vast majority of the people on agriculture, especially food crop production for its subsistence, correspond with the attributes of poverty in Ghana. In terms of both incidence and depth of poverty, the region is among the poorest of the country. Indeed, there are indications that while poverty was, on average, decreasing in other areas in the 70's and 80's, in the Upper East, it was increasing.

The first years of the project implementation were severely affected by the stabilisation and structural adjustment policies, especially pressures to reduce fiscal deficit. Before 1996, a major problem for the project implementation was the inadequate and untimely release of the government's counterpart funds. The fact that the government did not pay its co-financing share of the project operating costs, severely disrupted and delayed the project activities, particularly in the critical area of dam rehabilitation.

The economic policies of the country have not solved the problem of inflation. High inflation rates have been especially harmful for the project credit component, since inflation rate is reflected in the cost of borrowing. The high interest rates have depressed poor farmers demand for credit and combined with low farm gate prices have affected negatively the loan recovery.

Ghana is divided into ten administrative regions. From the late 1980s GOG began implementing a decentralisation policy for government structures, entrusting development activities to the regions which are further sub-divided into 110 districts. The UER is comprised of six districts. Each district has a District Assembly (DA) now taking decisions on a wide range of issues and planning and managing their own budgets and staff. Towards the end of project implementation in July 1997, decision was taken to further decentralise the Ministry of Food Agriculture (MOFA) to the district level.

Project achievements

Most of the dams had been rehabilitated although minor works and/or corrections were needed at some sites. IDA expects all works to be completed prior to the 1998 rainy season. Because of some changes in the dam sites to be rehabilitated, total command area (CA) was reduced from 555 ha to about 525 ha. IDA's latest estimations indicate that the reservoirs as designed and constructed will command an irrigable area of about 463 ha. This figure ignores the considerable amount of irrigation occurring, at many sites, outside the "official" CA. However, the IEM applied the SAR's gross water requirement to actual storage capacity data to relate available water to CA as determined by IDA. The actual total storage for the water reservoirs was theoretically enough, using this methodology, for over 640 ha of irrigated vegetables, well over the designed 463 ha.

It is apparent that labour-intensive rehabilitation costs about 71% of rehabilitation carried out by contractors. The real difference may be less as supervision costs using contractors may be less. On the other hand, during periods of rapid currency devaluations, the real cost of machinery and imported consumable items will increase at a higher rate than labour, thus giving labour-intensive methods a further advantage.

The design of the catchment area protection was unreasonably optimistic in expecting communities to leave catchment areas of 200 ha uncultivated and to expect the levels of co-operation visualised. The prescribed protected area was reduced to a more manageable 20 ha and eventually to 10ha. Nineteen damsite nurseries have been established and about 631 000 seedlings of various kinds of trees have been produced for planting in the catchment areas. About 147 000 seedlings have been distributed for protection of 235 ha at 39 damsites. The rest were given to farmers to plant around their homesteads.

Prior to the rehabilitation of Manga Research Station (MRS) there were few relevant recommendations available to MOFA staff for dissemination to UER farmers. As a consequence of subsequent work by researchers working out of MRS and the efforts of CSD many varieties/recommendations, have been released or produced. To the end of 1997 some 34 trials involving varieties, soil fertility and crop husbandry practices have been carried out. Nineteen of these have been published. Each "trial" is usually repeated at many farmer-owned sites over several seasons. For instance, 113 farmers reportedly had been selected in readiness for the 1996 wet season trials.

The Research Extension Linkage Committee (RELC) was used by UER since 1996. Under RELC methodology, a "grass-roots" approach has been adopted with AEAs discussing and prioritising problems with farmers and other stakeholders at each Agricultural Extension Agents (AEA's) sub-operation area initially. The deliberations at district level are further vetted by SMSs. A zonal workshop encompassing UER and Upper West Region (UWR), with RELC as governing body, is held to review previous outputs assess priorities and resources available and finalise programmes.

Due to strong local demand and the experiences of IFAD-supported SRDP, Sahelian sheep were distributed to upgrade local stock. 190 Sahelian rams had been purchased to the end of 1997 (PY 6) out of the project set target of 500 the IEM's overall assessment is that the much higher price obtained for crossbred Sahelian sheep compared to local sheep at the same age, together with the popularity of the crossbred indicates that there is a valid case for promoting the crossbred.

Since project inception the following activities have been implemented by the Crop Service Department (CSD): a) control of four significant pest outbreaks covering nearly 600 ha, in a range of crops including sorghum, millet, maize, rice and onion; b) supervision of and assistance to 109 contractor seed growers and dealers in production techniques and marketing; that about 330t of certified seed were produced and marketed; c) training of FLSs in integrated pest management in the major staple crops; d) about 600 groups with 9 000 members were trained, in insecticides safe use and handling; and e) routine pest and disease monitoring in the UER.

All 400 Community Livestock Workers (CLWs) have been trained and 160 have been provided with bicycles and basic veterinary kits. The CLWs dispense drugs, on a cash basis, provide some primary health care and act as a link between the group and district services. Most of the CLWs attention is given to his group and therefore direct services go to a small proportion of the community.

The project assisted in the rehabilitation (completed in 1996) of the Animal Traction Station at Doba (Kassina-Nankana district). In addition to catchment area protection, AESD has been involved in the following activities: a) instruction of 240 farmers in soil bunding and terrace bank construction using stones; b) 115 groups, consisting of 753 beneficiaries, have been trained in animal traction technology; and c) design and construction of onion storage facilities.

In terms of rural credit, the Agricultural Development Bank (ADB) disbursed about ¢ 2 billion which accounted for 60% of the total loan issued by all PBs. ADB provided credit to 784 groups covering 5 districts in the region which accounted for 68% of groups financed through institutional sources. In total 8 150 beneficiaries benefited from ADB financing. In contrast, rural banks participated in the project for the last two years. BESSFA covered 89 groups and provided ¢0.219 billion whereas NAARA covered 62 groups in both Builsa and Kassena-Nankana Districts. In total about 2 200 members benefited through RBs lending.

The activitywise credit flow 38% of total credit was disbursed for agriculture (both irrigated and rainfed). Among non-agricultural purposes, marketing was provided with as much as 40%, followed by IGA at 22%. Medium term credit which includes mainly bullock and donkey carts, equipment etc. absorbed an insignificant share. This emerging pattern indicates that beneficiaries preferred short-term credit for quicker repayment of loan rather than medium and longer term loans.

The low level of recovery (82%) contributed to the lower level of banks participation. The recovery performance of GCB was relatively better (92%) than ADB (81%). The reason for the better performance by GCB was its cluster approach of limited service area which facilitated better appraisal and follow-up. The limited outreach of ADB constrained its capability to monitor loans and improve recovery performance.

As against the target of 119 groups fixed in the annual budget prepared by PCU, the PCP could provide credit to 105 groups (88%). About 1570 target groups benefited under the programme about 77 % of the groups were exclusively women groups and 13% of them were mixed groups leaving the remaining 10% for men. Out of the total 15047 beneficiaries 68% were women. The volume of credit extended under PCP was ¢220 million during the project period which accounted for 6.5 percent of total disbursement.

The estimated average net family income per household of sample beneficiary during the pre linkage period based on primary data collected from the sample groups was ¢176 580 which was more than doubled during the post linkage period. In the post linkage period the net household income grew multifold in a positively correlated manner with group performance level. Thus, it was 261%, 195% and 116% for good, average and poor performers.

The total number of credit groups formed is 1264 and the number of WUAs formed at rehabilitated damsites is 27. There were 1 470 groups formed in agriculture and 400 in livestock. These constitute 518 of the livestock group target, 61% of the WUAs target, and 47% of the credit group target. No target in the design had been set for agricultural groups; instead the design specified that 3 000 farmers representatives would be formed. The number of contact farmers or representatives trained reached 334, or only 11% of the target.

Out of 44 damsites, 27 have some semblance of an association. Of these, only 17 have proper umbrella organisations in place and only 8 have their own bank accounts. The total number of members of these associations at various stages of development is 6 719, or 62% of the target. This membership figure is down from 8 990, which exceeded the target in 1997. In part the decline in membership can be explained by drought and the seasonal membership turnover at some damsites.

Women's historically lower involvement in gardening, explained by, for example, their limited access to land and their time and labour constraints, is reflected also in the current status of women's membership and positions in WUAs. In February 1998 there were about 2 566 women members (38 %) against 4151 men under the gardeners groups of WUAs. Women among fishermen and livestock owners are rare: 40 (6%) out of 618 fishermen and 125 (16 %) out of 769 livestock owners. At the entire WUA level 2 731 women make up 34 % against the total membership of 8104. In 1998 women gardeners membership outnumbered men only at 8 sites. As Table 5.4 illustrated, there are considerable variation in women's membership between the districts. At highest in Bongo 46 % of the WUA members are women, while women make up only 19 % of the all WUA members in Builsa.

All 270 Village Group Animators (VGAs) and 220 CLWs grassroots staff of DCD link groups with technical services. CLWs have been trained in animal health care. Out of a 400 total CLWs, 220 were trained in leadership and 200 in animal health; only 160 have received health kits. In the 1996 Mid-Term Survey, out of 91 VGAs, 86% were remunerated by in kind and 14% by cash from their groups. Only 71 of the VGAs and CLWs have received bicycles.

3. Effects assessment and sustainability

On the assumption that most of the WUA members are beneficiaries, the number of gardener WUA members rose from 3 592 in the 1993/94 irrigation season to 8 990 in 1996/97 and dropped to 6 717 in 1997/98. The total number of WUA members in 1997/98, including livestock owners and fishermen was 8 156 in 1997/98.

In four of the earliest rehabilitated sites the number of farmers increased from 2 144 (in 1993/94, prior to rehabilitation) to 3 755 in the first year after rehabilitation - a rise of 75%. The number of females with access to irrigation doubted to 1 863. However, by 1996/97 the number of farmers had decreased to below the 1993/94 level, partly as a consequence of water shortages at some sites. The overall number of irrigators was 3 380, of whom about 44% were women. Net returns of USD 109 and USD 53 for plots of 0.05 ha of tomato and onion, respectively were computed. These returns are very substantial dry season period given the persuasive poverty in the UER.

At full production, project interventions would result in an incremental output of 1 839 t valued at

USD 476 300. The actual incremental production is difficult to assess because data on the pre-rehabilitation situation, if available, is often inconsistent and probably based on a single year's production. In relation to SAR estimates, output in PY 6 appears to be well short of expectations. The estimated figure of 367.6 t for incremental output in 1996/97 does not cover all the reservoirs.

The food-for-work programme distributed a total of 63 100 rations equivalent to 1.5 million person-days of work. Undoubtedly the cash paid and WFP rations represent a welcomed and much needed injection of funds and food into the communities.

Through MRS efforts suitable extra early varieties of maize have been identified and promoted for wet season farming. Significant amounts of maize are now cultivated in the UER as a result of project interventions. Three early varieties of millet, harvested at 70 to 75 days as against the normal 100 to 150 days, were released. The project estimates that up to 80% of millet currently grown are made up of these varieties. The release of two new varieties of soybean, a relatively new crop for UER, has increased its area.

The project has contributed significantly to sustainability by imparting training at all levels, and promoting participating institutions such as WUA, women credit groups, VGAs, CLWs, etc. Such efforts would be further augmented with the help of NGOs.

Most rehabilitated reservoirs now have WUAs. The strength of the WUAs varies widely. The project's considerable efforts in formalising them as legal entities accountable to their membership and the District Assemblies, represent a very positive step towards sustaining project interventions.

Sustainability

Insufficient farmer participation in on-farm development of the rehabilitated damsites jeopardises damsite impact and sustainability. During IEM visits to some damsites, tertiary canals were not well developed, land extensions showed little evidence of bunding from last year, and some dug wells are being used instead of canals in the command area. General reasons given for lack of on-farm development were: a) inadequate rainfall and low water levels; b) land disputes between owners and users; c) high costs to farmer of on-farm development; d) poor quality rehabilitation of the embankment; and e) silting caused by animal grazing in the catchment area.

Strengthening local capacity of WUAs to sustain the damsite

None of the established WUAs are registered as a cooperative legal entity nor are the laws in place to allow WUAs to own the damsites or even manage them. Not only policies still need to be developed for farmers management but WUAs need to be trained in how to demonstrate the capacity and competence to undertake such responsibility for public resource.

With NGO involvement, farmers have become more aware of the importance of tree planting for command area protection. Their impact is still on a pilot scale with only 5 NGOs participating in a few sites.

Sustainability of credit depends on profitability of the investments and debt recoveries. The project interventions were instrumental in creating opportunities for increased dry season farming, which has adequate returns to investments. It also supported technical package that would increase productivity in wet season farming of a range of arable crops. In addition, considerable efforts were exerted to improve livestock productivity and create opportunities to IGAs. These interventions were undertaken within a social and financial intermediation effort to enhance group animation and improve their institutional and financial capacity and commitment to repayment of loans. So far the levels of repayment attained are not adequate to ensure sustainability.

There are two opinions regarding the prospects for sustainability of the project activities in terms of managerial capacity. The integration of the project into the exiting management structure of the MOFA in UER and gradually into the district should create capacity, build experience and improve sustainability.

The decentralisation of MOFA functions enhances the autonomy of the district. It will not only bring the decision making closer to the beneficiaries, but will also increase their participation, which is conducive to sustainability. This is supported by government directive that 20% of resources at district level would be earmarked for agriculture; in addition, 5% would be for poverty reduction.

Main issues and recommendations

The rehabilitation of remaining existing dams and construction of new small-scale irrigation dams should be further explored and special attention should be given to catchment area protection in the next phase.

With regard to dam construction, beneficiary communities should be involved at the conceptualisation and design stages as well as in construction through provision of labour. The proportion of manual labour to be used should be determined by the scope of the work and the labour availability at each site individually.

Provided reasonable guarantees of funds for wages and/or food-for-work are forthcoming and an adequately motivated agency with enough resources available is responsible for construction, their labour-intensive methods should be used where feasible.

A general recommendation is that IFAD should use existing institutes, which are, on the whole, well established, but whose field programmes especially in remote areas, would need considerable support, over and above the normal GOG fundings.

Emphasis should be given to the financial testing of crop packages and/or individual elements of packages. The MRSs agricultural economist should be given as much support as necessary to analyse the economics of various crops grown under irrigation and marketing intelligence particularly as irrigation production increases.

Given that the regional staff of the various MOFA will be the contact points between MOFA, as an umbrella organisation, and agricultural activities at the district and field levels, regional staff should be appropriately trained and provided with adequate operating means to provide the necessary linkages.

Formal credit institutions should continue to play an active role in the process of loan delivery to the target groups. More specifically: a) PBs should participate in selection of groups, animation and identifications of activities; b) PBs should follow-up and closely monitor, especially defaulting groups; c) Prompt payment must be encouraged through re-invigorating of the rebate scheme and increasing reward to 5%. In addition another 2 to 5% should be given to VGAs for helping groups to achieve full repayment; and d) Credit guarantee schemes need to be oriented towards smallholders by creating a reserve, for bad debts.

The mission observed that there is a need for establishing more effective financial and social, intermediation, in particular: a) The positive role played by NGOs in group formation and training should be supported by the project; and b) WUA are possible to reorganise and train as credit groups.

Following steps should be taken to strengthen the credit groups: a) The literacy programme must be directly integrated with the project activities; and b) Animation of groups particularly women groups must be further strengthened through providing adequate remuneration to VGAs. Record maintenance must be given greater emphasis.

Future project management and coordination should be integrated into the regional and district structures based on the successful experiences of LACOSREP. To ensure effectiveness of the external coordination mechanisms, the PCC should be maintained. It should be reconstructed to reflect decentralisation. IFAD should support in the next phase the institutions at the district level

To ensure financial control, project funds will be channelled through the District Director of Agriculture (DDAs) who will be held accountable for expenditures and financial returns. The financial controls should rest with the project financial management of the PCU as before.

It is necessary that in future the Monitoring and Evaluation Section (MES) should concentrate on assisting the implementing agencies to gather the needed information, without having the direct responsibility for the provision of data for physical monitoring. In this way, the MES would have sufficient time to perform its function of processing the data from IAs and keeping the PCU well informed.

To ensure that landuse plans are developed with the WUAs and that the interrelationships between catchment area protection, wet and dry season cropping and maintenance are clear to users, all stakeholders in the community need to be involved including chiefs, landowners, clan elders, and wet and dry season farmers and livestock owners when needs assessments are carried out.

More collaborative work needs to be done with NGOs. The need for technical input from NGOs for reaching women and the poor and for mobilising beneficiaries for community development activities could not be overemphasised. Three critical areas are needed: micro-finance training, catchment area protection, farmer management of irrigable land areas with succession of wet and dry season farming.

The issue of equitable and transparent land allocation needs even more serious addressing in the future. The guiding principle of land allocation should be transparency: a clear set of land allocation criteria and regular record keeping of actual land allocation, holders of plots and payment of fees. The WUAs themselves should remain responsible for the land allocation and be in a position to justify their decisions.

For enhancing women's participation in the future development, the project executors must pass the information on the dam rehabilitation and its implications to women directly and from the beginning. Special animation should be developed to involve the women, meaning they should be mobilised and given the necessary training to take decisions and not just take part in the labour.

As a means to empowering women in the irrigation management, provisions guaranteeing women's minimum representation in the WUAs should be included in the WUAs, or alternatively, in the DAs bye-laws.