Interim Evaluation

In accordance with the Evaluation Policy of the International Fund for Agricultural Development (IFAD), the Office of Evaluation (OE) conducted an Interim Evaluation of the Upper West Agricultural Development Project (UWADEP) in Ghana in May-June 2005, given the interest of both the Government of Ghana and IFAD's Western and Central Africa Division to proceed with further investments in the area. This evaluation adopts the standardized IFAD methodological framework for project evaluations.

IFAD projects in Northern Ghana. Northern Ghana consists of three regions, Upper East (UER), Upper West (UWR) and the Northern Region (NR). IFAD has projects in each of them: Land Conservation and Smallholder Rehabilitation Project (LACOSREP), UWADEP, and Northern Region Poverty Reduction Programme (NORPREP) for the Northern Region. UWADEP originated from a FAO General Identification mission in 1993, was appraised by an IFAD team in mid-1995 and modelled on the example of the existing LACOSREP I. UWADEP was approved by the IFAD Executive Board in September 1995, became effective in March 1996 and closed in December 2004. The total project cost is USD 11.3 million, out of which IFAD provided a loan for USD 10.0 million. As of August 2004, the disbursement rate reached 94.64 % of the total loan amount. IFAD is the only international financier of the project, which was supervised by the United Nations Office for Project Services (UNOPS) for most of its implementation.

Macro-economic and poverty indicators. Located in West Africa, Ghana has an estimated population of 20.5 million, of which 63% rural. The structure of the economy is characterised by a large (in relative terms) services sector (42% of the total GDP, compared to 34% for agriculture and 24% for industry). It has an annual GDP per capita of USD 304 and GDP growth has averaged 1.8% in the last ten years (i.e., below population growth) although this has accelerated recently. Agriculture continues to be the mainstay of the economy, employing about 60% of the labour force. Ghana is classified as 131st out of 175 countries, by the UNDP Human Development Index (2004). The percentage of households below USD 1 per day has been estimated at 44.8%, and the percentage of poor households according to a national poverty line at nearly 40% (World Bank, World Development Indicators, 2004).

Agricultural policy. Agriculture contributes to ensuring food security, provides raw materials for local industries, generates foreign exchange, and provides employment and incomes for most of the population. Agriculturally-dependent rural households (72% of the population) form the largest potential domestic market for output from other sectors of the economy. Recent agricultural policy in Ghana is reflected in the following key documents: (i) Medium-Term Agriculture Development Programme (MTDP) (1991-2000); (ii) Accelerated Agricultural Growth and Development Strategy (AAGDS); (iii) the Food and Agricultural Sector Development Policy (FASDEP) (2002); and (iv) the Ghana Poverty Reduction Strategy (GPRS) (2002-2004), currently under revision.

The Ghana Poverty Reduction Strategy (GPRS). GPRS (2003) was revised in 2005 to focus on the identification of vulnerable groups and possible risk-reduction strategies. It recognises that rural farmers and fishermen are particularly at risk and specifically mentions Northern Ghana as a locus of perennial food deficits. Women are identified as particularly discriminated in this context and instruments to promote gender equality are emphasised. The document also notes the importance of environmental factors in increasing vulnerability. However, it does not put forward concrete measures to reduce risk in agriculture.

The rationale for UWADEP was summarised at appraisal as:

• strong demand for dam rehabilitation in rural communities;

• the potential for Water User Associations (WUAs) to be sustained and assure food security in the region; and

• build on existing credit experience to establish effective mechanisms for rural financial institutions.

Strategic thrust. UWADEP was intended to improve food security and increase the income of smallholders. Its components include:

• capacity building to strengthen project delivery and management skills of key implementing agencies;

• water resources development (rehabilitation of dams, formation and support to WUAs, catchment area protection, and demonstration/promotion of manually-operated tube wells);

• agricultural development, including farmer training and demonstrations, support to technology generation and research/studies, marketing and processing, and livestock development;

• promotion of income-generating activities through the supply of rural financial services; and

• rural infrastructure comprising rural road rehabilitation and the construction of hand-dug wells for drinking water and latrines.

Project area. The UWR is situated in the northwest corner of Ghana, with an estimated total population of 580 000, of which about 90% is rural. The average population density is 29.8 persons/km2, about one fourth that of the UER. UWR is one of the poorest regions of Ghana, with an annual per capita GDP of USD 170 and on most social indicators, the most neglected region of the country. Infant mortality, seasonal hunger, cash incomes, school attendance, transport networks are weaker in UWR than other regions. Despite this, UWR has benefited from very few targeted development projects, and until recent years, not many NGOs were operating there. UWR has abundant land both for crops and livestock and much lower population density, but access to markets and off-farm opportunities is constrained by poorly maintained feeder roads and lack of transportation services.

Components of UWADEP at appraisal were as follows: (i) agriculture (33% of total base costs); (ii) water resources (17%); (iii) rural roads (16%); (iv) credit (22%); (v) community and women's development (6%); and (vi) project support unit (7%). The target group is poor farmers, which represent up to 80% of the population of UWR, some 20 000 households according to the 1995 appraisal. Overall responsibility of the project was with MoFA, with the Chief Director responsible for policy direction and the provision of counterpart funds at regional level. A Project Support Unit (PSU) was to be established within the Regional Office of Agriculture to assure project programming, prepare work programmes and budgets. Community mobilisation was to be the responsibility of an officer seconded to the PSU to establish linkages with the Community Development Department and Women In Agricultural Development, to assure the emphasis on gender.

Implementation

Rural credit for income-generating activities. The project reached 5 805 beneficiaries in 379 groups for a total of circa ¢ 5.6 billion, equivalent to USD 640 000 (about 60% of target). Even though women constituted 56% of this number they received only 47% of total loan amount. The largest portion (64%) of the disbursements went to farming which is dominated by the males. Loans for income-generating activities such as trading and food processing, reserved for women, followed with only 24%.

Dams, irrigation, water and roads. After the end of extensions, the project finally closed, leaving irrigation infrastructure (dams and canals) incomplete on several sites. The draft Project Completion Report (PCR) states that 41.5 ha are under dry season cropping and 154 ha available for irrigation against an Appraisal Report (AR) target of 220 ha (70%). The Ghana Irrigation Development Authority (GIDA) Technical Review on completion (June 2005) mentions the same figure of 154 ha, but cautions that certain areas are not under the command from the canals, which means that hand watering or pumping has to be used. Farmers still largely depend on hand-dug wells, as was the case before the project rehabilitations took place. The evaluation team could find only 23 ha. of additional irrigable area resulting from the project (figures obtained from physical observations and cross-checked with WUA secretaries/chairmen in each case). Prescribed sanctions were not applied to defaulting contractors in several instances by PSU. Generally, no laboratory investigations for quality assessments of the work were conducted. This is a major problem, as many structures suffer from poor quality, and thus uncertain sustainability. After the two extensions, the project finally closed in 2004, leaving irrigation infrastructure incomplete on several projects. Through the Ministry of Food and Agriculture (MoFA), additional funds were provided from Food and Agriculture Budget Support (FABS) to cover the outstanding works. The AR mentions involvement and training of WUAs, as well as specifically-targeted interventions for women. In the AR, 40 hand-dug wells were planned at the dam sites to provide safe drinking water to participating communities. About 35 of these were sunk - 12 in 2004 alone, by an NGO (PRONET, Wa). The target for feeder road infrastructure in the AR was 140 km and this has generally been exceeded.

Agricultural extension and seed production. Prior to UWADEP, the Seed Growers Association (SGA) in the region had only 12 growers who had previously produced seed, mostly maize, under contract to seed supply firms. Considerable progress was made prior to decentralisation, with the number of growers increasing from 12 to 60 by 1999. Numbers have since reduced, with 30 successfully producing seed in 2004. On-farm demonstrations. Demonstrations of the application of single super phosphate (SSP) to groundnut were promoted throughout the project life. The project worked with a total of 3 600 farmers in 300 groups, 75% of the target of 4 800 farmers in 400 groups. Support and promotion of animal traction. The implementation progress of the animal traction component has been very modest at field level, with only 130 carts, 65 ploughs and 65 ridgers made available to 260 individuals, training conducted and refurbishment of one blacksmith's workshop at Tarsaw.

On-farm adaptive research. The on-farm adaptive research component had two major activities: infrastructural support to the establishment of a research station in the UWR and implementation of research in support of improved crop production systems for the region. A permanent base was established for the Savannah Agricultural Research Institute (SARI) at Dokpong Agricultural Station at Wa, with the construction of an office block with laboratory facilities, accommodation for ten staff and a large conference hall. Initially SARI conducted a wide range of trials attempting to cover the issues identified during the initial project planning workshop held in 1995. Research subsequently focussed on two main areas; improvement of soil fertility, and identification of improved varieties of the major legume and cereal crops. Introduction of improved varieties of cowpea, soybean, sorghum, maize and rice have generated much enthusiasm among farmers, however access to certified seed remains a problem for small-scale farmers.

Upgrading of local livestock. During 1998 and 1999, 242 improved rams and 90 bucks were distributed to farmers in the five districts. High levels of mortality led to concerns over the approach and the programme was suspended for two years. It was restarted in 2002, with a refocus on beneficiaries with sufficient resources to be able to afford improved housing, feed and recommended health measures. Some 219 (50%) offspring of the improved rams and 123 offspring of improved bucks have been recovered for redistribution. Poultry and guinea-fowl. Between 1998 and 2004 some 13 640 keets were imported from Belgium. The average survival rate for the first two batches was 84% (M&E unit, 2002). Keets were sold from the 1998 batch with 24% subsidy, but from later batches at full cost recovery. Beneficiaries report that the imported guinea fowls were robust and about twice the size of local birds. They fetch ¢35,000 as compared to about ¢20,000 for local birds. Training and support to Community Livestock Workers (CLWs). The CLW scheme recruited and trained 150 volunteers and provided them with basic livestock kits to help improve livestock health and nutrition within project groups and in time, within their whole community. Groups report significant reductions in mortality rates with the instigation of recommended management and health care.

Performance

Relevance. The overall objective of UWADEP was to empower rural populations living in poverty to access improved technology services and credit. There is no doubt that the overall and specific objectives are relevant to these rural communities which depend almost entirely on agriculture. Ghana has signed up to various international undertakings to reduce poverty and UWADEP has this as its direct focus. Despite its commitment to donor strategies on poverty reduction, the government of Ghana has yet to make available additional funding for the development of these regions, which makes IFAD's approach all the more relevant. However, allocation of resources within sub-components between infrastructure development (SARI research station, animal traction centre and Livestock holding centre), training, demonstrations and practical interventions had limited relevance. As an example, the focus on training within the animal traction component seems misplaced in the context of working with groups of farmers who already own bullocks and donkeys.

Effectiveness has been low for the rural finance component: transaction costs are high for both banks and clients and credit discipline has been weak, with the exception of Sonzelle Rural Bank (RB). In the agricultural development component, AT interventions were limited in spite of high farmer demand. Areas of investigation in adaptive research have been limited, surprisingly excluding soil and water conservation. Research has ignored traditional farmers' crops such as yams (without practically any linkage with IFAD's Roots and Tubers Programme) and the spontaneous development of hand-watered gardens along rivers. Improvement of small ruminant and poultry/guinea fowl breeds has proved popular, in spite of an outgrower scheme that prioritised the better off, with delays to the servicing of poor farmers. Supervision of dam construction has been very weak and characterised by intra-government disputes.

Efficiency has been explored for irrigation, through the comparison of construction costs per ha between the two phases of LACOSREP and with cost ranges of other organizations (cost-effectiveness analysis). In the case of the rural finance component, productivity and efficiency (as per Consultative Group to Assist the Poor (CGAP) definitions), are compared with regional benchmarks. Unit costs for irrigation and rural finance compare favourably with their peer observations (in the case of irrigation this is partly due to the devaluation of the cedi). It is however important to contrast low costs with limited achievements (for example both in terms of delays and poor infrastructure quality).

Impact

Methods. Findings are the result of triangulation between multiple sources: (i) quantitative surveys of beneficiaries and non-beneficiaries; (ii) a qualitative survey of five dam sites; (iii) an ad hoc survey of non-beneficiaries; (iv) participants' observations as recorded by mission members; and (v) secondary data.

Outreach and impact. Physical and financial assets of farmers assisted by the project are rising, but there is a general increase (i.e., also for non-project farmers) in access to material assets due to remittances. The choice of training topics has not always been appropriate, thus the impact on human capital was limited and limited attention has been devoted to human health hazards under the water management component, in spite of the potential risks of water-borne diseases. A very important and originally unplanned development took place when under-privileged groups (blind, disabled and single-mother) were the object of the project's assistance in Karni dam site. This happened also thanks to preparatory work carried out by an NGO. Attention to gender issues has been limited throughout the components. The present implementation does not ensure that benefits spread rapidly to the resource poor and women. If there is one message that emerged clearly from the field visits, it is that training, research and sensitisation has been over-emphasised to the detriment of assisting farmers with practical requests.

Innovation, scaling-up and replicability. Few elements of UWADEP can presently be considered an unalloyed success and therefore the desirability of replicating them in their present status is questionable. Clearly more dams would be desirable; one option is to use the NGO sector to deliver more modern dams. The introduction of improved breeds and seeds has proven partially successful, but could have achieved a wider diffusion if they had been more responsive to farmers' requests. Improved breeds of small ruminants and fowl have been successful in increasing incomes for a range of farmers and would seem suitable for replication.

Sustainability has been constrained in the rural finance component by below-market interest rates to final borrowers and poor credit discipline. Sustainability of the SGA is not assured because of the failure to link the group effectively with inventory credit and missing links in the seed chain between small farmer's demand and supply. In the irrigation component, the main issues are: completion of irrigation infrastructure through FABS, which, given the pace of project implementation during UWADEP must be of concern to communities, and quality of works (which is manifestly poor). Few WUAs have been in operation long enough to judge their sustainability, but in many cases these social groupings are robust because they have existed in a different form prior to UWADEP, managing the hand-irrigated land below the dams, sometimes for decades.

Performance of partners

IFAD. Two issues are critical in assessing the performance of IFAD; the acceptance of an appraisal document with features that were clearly inappropriate for UWR and an inadequate response to design and implementation problems which surfaced during the course of UWADEP. For all the participatory rhetoric, the design of UWADEP was top-down and failed to address the concerns of a high proportion of its target group. Design decisions in irrigation infrastructure should generally be flexible and appropriate to the actual environment, but those in the AR were too rigid, giving very little room for changes to suit individual sites.

UNOPS supervision seems to have been less accurate than in LACOSREP II (see the related evaluation for comparisons), the sister project in UER, in that many of the problems and inflated claims on achievements of UWADEP were not picked up. Given the time and resource costs of visiting individual communities to establish the exact implementation status, this may well be considered beyond the present resources and contractual arrangements between IFAD and UNOPS.

Government and its agencies. A characteristic feature of the PSU was its use of available MoFA staff, even when sector specialists were not present. Thus the PSU had no staff qualified to supervise infrastructural work, or to monitor and encourage gender sensitivity. There was no specialised staff member in the PSU dealing with credit. Changes in project management to introduce dynamic individuals took place late. The failure to engage with NGOs and the use of Agricultural Extension Agents (AEAs) to perform their tasks ‘after hours' is doubtful practice at best and hardly develops the grassroots and innovative approaches expected by IFAD. Project documentation was lamentable, and even after considerable work, the numerical data on implementation presented in this report must be treated with caution. The decision to use GIDA as a sole consultant for the irrigation infrastructure on the project was a very questionable one, causing many problems in the quality of project execution, as well as delays in implementation.

Overall assessment

The overall impact of UWADEP must be considered modest. The number of dams functional by the closure of the project was small and the irrigated area limited. Many of the dams have engineering and technical problems which bespeaks a lack of supervision. As a consequence it is hard to evaluate the increase in social capital and the strength of WUAs, because they have had a limited chance to operate. Much of the irrigated area claimed by UWADEP turned out to be well-based gardens that were in operation prior to the project and are of greatest use to the younger and stronger members of the community. Appraisal targets for women's involvement were almost never met. The majority of resources were spent on capital goods and central infrastructure with very little visible return and the agricultural components were unresponsive to the needs of actual farmers. In the case of animal traction, for example, expenditure on training and buildings far exceeded sums spent on getting implements into the community, despite the strongly voiced requests of those communities. Partnership with NGOs was weak. Credit was taken up and clearly valued by its recipients, but little emphasis on supporting rural finance institutions and cumbersome procedures meant that its impact was far less than its potential; moreover, most of the banks appear to be simply withdrawing rather than continuing with their clients. A fundamental problem was the copying of many features of LACOSREP without due consideration for the different ecological, economic, social and demographic characteristics of UWR, a design shortcoming for which IFAD must take some responsibility. However, many problems also arose from weak supervision and the MTR process. It would not be desirable to replicate or extend UWADEP in its present form. Further interventions in the area, in principle justified by the severity of poverty, require major changes in the project design.

Insights and recommendations

Insights

Further investment? If IFAD is to consider further investment in UWR, then the evaluation of UWADEP undoubtedly suggests major rethinking of structural elements in project design. A key problem is that projects with so many components, but no clear integrative strategy, are open to activities being carried out with no linkages, with consequently high management costs. The failure to link credit with other technical innovations suggests the added value that could be gained from a more coherent approach. The significance of delivering infrastructure in a timely manner cannot be overemphasised. No amount of training and sensitisation will compensate for this fundamental lacuna. Project design should consider sequencing much more carefully. Despite efforts to introduce greater clarity into contracting arrangements, the reality has been almost the reverse.

What type of project? In Ghana, in some sectors such as health and education, donors have started to fund multi-donor budget support initiatives. Although the pressure is not yet strong in the agriculture and water sector, the question is whether area development interventions deserve further funding. UWADEP illustrates that there is a very significant need on the ground for quite conventional projects that deliver services and inputs to impoverished farmers. It is yet to be demonstrated empirically that other formats can deliver the same benefits. Should IFAD decide to fund future investments, clearly a priority would be to improve the project concept and design. Another priority for IFAD and the Government would be to identify strategic partners. With its own resources only, IFAD's coverage will be inevitably limited.

Recommendations

Project design and partnership with innovators. UWADEP developed from the models of URADEP and LACOSREP, rather than being designed to address the specific situation of the UWR. This has been problematic, because its interventions are not necessarily those most valued by communities. Key areas of agricultural production such as yams and tree-crops were not considered, although these provide cash income and also help communities bridge the food gap experienced by cereal growers as well as reducing migration. In addition, the components were poorly integrated at the design level and continued largely to be managed separately, with consequently high transaction costs. NGOs are not necessarily better than their official counterparts, but major NGOs in UWR are rapidly developing capacity to cover many of the same areas as government-based extension. A future project design that does not recognise these realities would be highly problematic. There is a strong need to learn from "good practices" by other organisations, including particularly NGOs in UER and UWR. These practices should be documented and discussed not only at the regional level but also in national fora, perhaps taking the opportunity of the donor's coordination group in the agricultural sector.

Dams are one strategy in a basket of options for improving household incomes. Other options have been discussed in the report and are presented later in this section.

Improve implementation support and monitoring and evaluation (M&E). The MTR was delivered late, but even so, it is very evident that most of its recommendations were not taken into consideration by the PSU. The PSU needs to be more proactive and so do UNOPS and IFAD to ensure follow-up. UWADEP appears at first sight to have a comprehensive record of its impact. However, many of the figures recorded turned out to be inaccurate. Record-keeping in UWADEP collaborating institutions, notably the banks, has also been weak.

Agriculture. Despite the rhetoric of participation and ‘demand-led', on-farm research as practised by the project is surprisingly similar to the testing of a package of improved technologies under farmer conditions. The almost universal use of organic manure by farmers, limited practice of composting and difficulty in affording and accessing fertilisers was not reflected in trial designs. More attention needs to be paid to women's crops (such as vegetables) and cropping systems, where there is little use of chemical fertilisers. Farmers' requirements with regard to animal traction tools (for example ridge weeders and robust furrow weeders) should be included in the selection of tools made available. Institutional arrangements for supply of tools at local level should be explored with NGOs. Exclusive focus on demonstrations of an application of SSP to groundnut led to limited results because of the unavailability of SSP and the farmers' inability to afford fertiliser inputs following withdrawal of the project. The introduction of 50% Sahelian improved stock would enable a greater number of farmers to take advantage of the intervention more quickly. As the 100% improved stocks do not survive for long under the traditional husbandry system, the approach would be more popular with the majority of farmers.

Additional areas for inclusion. Despite their importance in the local economy, their expanding production and their capacity to bridge the ‘hungry season' and the existence of the IFAD Roots and Tubers Programme, tuber crops have been virtually ignored. Yam and frafra potatoes have a good local market and in the case of yam, very good external markets. Agroforestry systems may be particularly suited to the region and help introduce permanent cropping alongside annual cropping to improve both production and income levels. Indigenous agroforestry systems based on the cultivation of yams and vegetables below certain trees, such as locust (Parkia biglobosa), neem (Azadirachata indica) and Acacia (Acacia albida) are a local response to soil fertility constraints and are worth investigating for wider application.

Rural finance. Financial sustainability depends very much on the participating banks profitability, which is based on financial income, operational cost and loan loss provision. As in several projects built around the ‘traditional' view, rural financial institutions in UWADEP have been seen as a mere conduit to provide an input (credit) to farmers. Institutional strengthening of financial institutions has not been seen as a priority. In spite of this, Sonzelle RB has been able to enforce good credit discipline and attain operational (although not financial) self-sufficiency. It is the focus on sustainable institutions that should characterise future interventions, either through a project component or a dedicated programme (in which case the issue of dovetailing with agricultural interventions in other projects becomes crucial). There is a wide range of credit facilities available when the District Assembly Fund, international NGO and local NGO activities are considered. Specifically, a large number of organisations are involved in inventory credit and credit for agricultural inputs and all offer credit at different rates, i.e., there is predatory competition in a poorly regulated environment. Without region-wide co-ordination, market led microfinance cannot succeed, as it will be out-competed by politically and socially-motivated interventions.

Water and infrastructure. Civil works on dams and irrigation infrastructure were the major cause of delays in project implementation. A quite different approach should be considered in future, with checks on quality mandatory at every step of implementation. Opening the consulting to a wide range of professional companies should be non-negotiable. Services should be segregated into the phases according to the project implementation schedule and contracts signed separately, whereby the contract for downstream works is not automatic, but subject to satisfactory performance in the first phase, if the same consultant is selected to undertake the assignment. For projects to be completed within the scheduled completion period, within the budget and to specification, stakeholders, comprising the project staff, consultants, DA, and the WUA should take active part in monitoring construction works. Finally, it should be recognised that irrigation infrastructure needs maintenance, some of which is beyond the capacity of the WUAs. This should be reflected by realistic budgeting.