Ghana - Agricultural Sector Investment Project

Document of The World Bank FOR OFFICIAL USE ONLY C/--- ) (CC- n 6-/I Report No. 12222-GH STAFF APPRAISAL REPORT REPUBLIC OF GHANA AGRICULTURAL SECTOR INVESTM PROJECT NOVEMBER 10, 1993 MI CROGRAPHICS Report No: 12222 GH Type: SAR Agricultural Operations Division Western Africa Department IV This document has a redricted distibudion and may be used by redpients only in the performance of their official duties It contents may tol otherwise be disclosed wthout World Bank authorzaton.  CURRENCY EOUIVALENTS Currency Unit = Cedi US$1.00 = Cedi 630 Cedi 1.00 = US$ 0.00159 WEIGHTS AND MEASURES 1 metric ton (m ton) = 2,205 pounds (ib) I hectare (ha) = 2,47 acres (ac) 1 kilometer (km) = 0.62 miles (ml) I meter (m) = 3.28 feet (ft) LIST OF ABBREVIATIONS AND ACRONYMS ADRA Adventist Development and Relief Agency AGSAC Agricultural Sector Adjustment Credit ASRP Agricultural Services Rehabilitation Project CEDEP Center for the Development of People CIDA Canadian International Development Agency WCOBOD Ghana Cocoa Board CUSO Canadian University Students Overseas DA District Assemibly DCA Development Credit Agreement DFR Department of Feeder Roads DP8U District Planning and Budgeting Unit DWM December 31st Women's Movement EAP Environmental Action Plan EIRR Economic Internal Rate of Return EPC Environmental Protection Council ERP Economic Recovery Program FY Financial or Fiscal Year GAPYOD Ghana Association of Private Voluntary Organizations in Development GDP Gross Domestic Product GEPC Ghana Export promotion Council GFDC Ghana Food Distribution Corporation GIDA Ghana Irrigation Development Authority GOG Government of Ghana GTZ German TechnicaL Cooperation Agency IAPSO Inter-Agency Procurement Services Office ICOUR Irrigation Company of the Upper Region ICR Implementation Completion Review MLG Ministry of Local Government and Rural Development MOFA Ministry of Food and Agriculture MRH Ministry of Roads and Highways MTADS Medium-Term Agricultural Devetopment Strategy NTR Mid-Term Review NORRIP Northern Region Rural Integrated Project NPV Net Present Value ODA Overseas Development Administration of UK PANSCAD Program of Actions to Mitigate the Social Costs of Adjustment and Development PCC Project Coordination Committee PCW ProJect Coordinating Unit PNDC Provisional National Defence Council PPMED Policy Planning, Monitoring, and Evaluation Department PU Project Unit SOE Statement of Expenditures SAC Structural Adjustment Credit SPAC Subproject Approval Committee TSA Technical Support Agency URADEP Upper Region Agricultural Development Project VORADEP Volta River Agricultural Development Project FISCAL YEAR JANUARY 1 - DECEMBER 31  FOR OMCIL4 USE ONLY REPUBLIC OF GHANA AGRICULTURAL SECTOR INVESTMENT PROJECT TABLE OF CONTENTS Page No. CREDff AND PROJECT SUMMARY I. INTRODUCTION ........................................ I H. BACKGROUND ......................................... 2 A. The Economic Setting ..................................... 2 B. The Rural Sector . ...................................... ? C. IDA Assistance Strategy ................................... 9 D. Bank Group Operations . .................................. 10 E. Rationale for IDA Involvement ......... .................... 11 F. Environmental Issues .1................................... 1 G. Lessons Learned from Previous IDA Involvement ..... ............ 12 IM. THE PROJECT ........................................ 13 A. Project Objectives and Scope ............................... 13 B. Summary Description ................................. 13 C. Detailed Features .................................. 14 IV. PROJECT COSTS AND FINANCING .......................... 16 A. Cost Estimates .................................. 16 B. Proposed Financing Plan ................................. 17 C. Financial Implications for Government ......... ............... 18 D. Procurement ................................. 19 E. Disbursement ................................. 22 F. Auditing and Reporing ................................. 24 This repoat is based on the fndings of a pre.appruisa1 mission that visited Ghna in November 1992 wnd Febmaiy 1993, and an appaisal mission that took place in May/June 1993. 'he pr-appmisal minssion was comprised of Mmes. and Mess. P. Werbrouck (Mission Leader - Sr. Agicultuml Economist), L. Campbell (Sr. Rural Infastucturc Engineer), C. Ranade (Sr. Agriultunrl Economist), H. Weindler (Disbursement Officer), 0. Vela (Procurement Analyst), J. Silverman (Public Sector Managent Specialist) of IDA, and Dr. S. K. Dapuah (Economist), A. Mensa-Bonsu (Irrigation Engineer), J. Cofr-Agnam (Planner), S. Asuming-Bmpong (Agricultural Economis), and S.A. Gumah. Kofi Ahaliga, B. Lugu-Zuri, and Ben Wu-eku (Farmers). Dudng thc appraisal mission Mesms R. Polson and K. Atta-Kmh joined the tm. Mr. Hans Binswanger was the lead advisor. Mr. Owusu-Ansah and R. Poison worked on the economic and inancal analysis and the cost tables. Messrs. J. Joyce and E. Lim are the mantaging Division Chief and Department Director, respectively, for this operation. This document has a restricted distribution and may be used by recipients only in the performat -e~ of their official duties. Its contents may not otherwise be disclosed without W#orld Bank authorization.|  - ii - Page No.- V. PROJECT MANAGEMENT AND IMPLEMENTATION ....... ...., 25 A. Management .25 B. Eligibility Criteria .28 C. Beneficiary Contributions .30 D. The Subproject Processing Cycle .32 E. Training .34 F. Project Implementation Rate ....... ................. ...... 34 G. Monitoring, Evaluation and Project Impact Studies .35 H. Project Supervision .36 VI. PROJECT JUSTIFICATION, BENEFITS AND RISKS .37 A. Project Justification .37 B. Determination of Benefits and Costs .38 C. Risks .41 D. Environmental Effects .42 E. Impact on Women .42 F. Poverty Alleviation .42 VW. ASSURANCES AND RECOMMENDATION ..................... 43 ANNEXES Annex 1: Investment Component Description ...... ................ 44 Annex 2: Project Costs and Benefits ........ .................... 51 Annex 3: Project Implementation Schedule ...... .................. 62 Annex 4: Schedule of Disbursements ........ ................... 64 Annex 5: Project Supervision ........... ..................... 65 Annex 6: Key Monitoring Actions ......... .................... 69 Annex 7: Technical Assistance and Training ....... ................ 73 Annex 8: Enviromnental Problems and Mitigation Measures .... ......... 74 Annex 9: Local Communities, Government Service Structures and Non-Government Organizations ...... ................ 76 Annex 10: Documents Available in Project File ......... ............ 81 MAP: IBRD No. 25215  - iii - REPUBLIC OF GHANA AGRICULTURAL SECrOR INVESTMENT PROJECT CREDIT AND PRO.IECT SUMMARY Borrower: The Republic of Ghana Beneficiaries: The Ministry of Food and Agriculture (MOFA), District Authorities, and Rural Community Organizations Amount: SDR 15.3 million (US$21.5 million equivalent) Terms: Standard, with 40 years maturity Project Description: The project would increase the flow of technical and financial resources for rural development in response to needs expressed at the village and district levels, and consequently increase agricultural productivity and access to markets, and strengthen the communities' capacity to develop, manage and maintain their investments using the resources of NGOs and local consulting agencies. The project would finance the following investments for the benefit of rural communities: (i) construction and rehabilitation of small-scale uater schemes operated by farmer associations (11 percent of project costs); (ii) construction and rehabilitation of rural markets for district authorities and communities which have a cost-sharing and maintenance agreement with traders and market users (44 percent); (iii) roads to Improve access to markets (17 percent); and (iv) small-scale storage and village food-processing units for rural associations (2 percent). The project would also finance project-related technical assistance, monitoring and training in irrigation, marketing and food-processing management, as well as monitoring, evaluation, beneficiary assessment and impact studies (26 percent). Eligible subproject costs would not exceed US$150,000. Eligible beneficiaries would be community associations and district authorities. Beneficiaries would contribute between 10 and 25 percent of the capital costs in cash or kind (e.g., labor) and would be responsible for all operation, maintenance, and equipment renewal costs. Costs of assets, which have to be replaced in a relatively short period, would be fully recovered for income-generating projects benefitting selective groups of people. Beneficiaries would receive technical assistance from eligible technical support agencies such as NGOs and consulting agencies under technical assistance agreements with the communities. These agencies would also carry out project-related training, extension and follow-up, depending on the need. The financing of the activities of the technical support agencies would be related to the investment costs and the nature of the investment. The Ministry of Food and Agriculture (1MOFA) would be responsible for overall project management and inter-ministerial coordination, and implementation would be the responsibility of a Project Unit (PU) staffed with fixed-term employees experienced in private sector and project management. Initially, PU would establish regional offices in Accra, in the South, and Tamale, in the North. Regional Subproject Approval  - iv - Committees (SPAC) would assist PU with the review of subproject applications. IDA would appoint a full-time supervisor at the Resident Mission to ensure operational liaison with PU. Decentralization of procurement is an important feature of project design. Beneficiaries would manage the procurement process with assistance from technical support agencies and in accordance with rules established in an Implementation Manual and agreed with IDA. MOFA would recruit auditors and MOFA's Policy Planning, Monitoring and Evaluation Department (PPMED) would carry out beneficiary assessment and development impact monitoring. Project BenBefits: The main project benefits would be: (i) improved links between production and markets leading to more efficient agricultural marketing and reduced post-harvest losses; (ii) strengthened rural associations and local authorities; (iii) increased irrigated land and agricultural production; and (iv) more equitable regional income distribution. Most beneficiaries would be rural people with lower-than-average incomes; and women, market traders, vegetable producers, transporters by head- load, and food processors. The project would also provide the framework for further donor and government assistance to irrigation, marketing and rural infrastructure. Risks: The four main risks and mitigation measures are: (i) poor capacity of government and PU to cope with so many small subprojects-PU would be responsible for day- to-day operations, have experienced private and public sector management staff, receive support from qualified NGOs and consultants, and would be equipped with computer technology; (ii) the subproject approval process may become overly complex through donor and government procedures-the project implementation manual would have simple but detailed procedures, PU would hire influential farmers as part-time project facilitators, and IDA would appoint a project supervisor at the Resident Mission to ensure liaison with PU; (iii) funds might be misused-adherence to eligibility criteria, strict and regular control by independent auditors, and public sanctions against misuse of funds, including loss of eligibility for further project support, would limit such risk; and (iv) the lack of capacity to prepare subprojects-PU would establish and yearly update a list of capable technical support agencies to assist the communities. Sensitivity on project costs and benefits show switching values for project costs +36.2 percent and benefits - 26.6 percent for an opportunity cost of capital of 12 percent. Economic Rate of Return: 23.3 percent Alleviation: The project would have a 25 percent poverty alleviation component. The intended project beneficiaries have a significantly larger proportion of poor people than the country's population as a whole. Poverty in Ghana is predominantly a rural phenomenon and 43 percent of the rural inhabitants live below the poverty line. Categorv: B  PrOlect COSt Summary % Of FOregin LQoC Total FE Base ........ (US$ million) ......... (%) Ws 1. INVESTMENT CATEGORY A. Civil Works 0.2 11.4 11.6 2 54 B. Goods/Supplies/Equipment 1.7 2.7 4.4 36 20 C. Studies and Technical Assistance 0.5 3.2 3.7 14 17 D. Refinancing PPF _ 0.8 16 51 7 Total Investment Costs 3.2 18.1 21.3 15 98 II. RECURRENT COSTS 0.1 0.3 0.4 10 2 III. TOTAL BASELINE COSTS 18.4 21.7 15 100 Physical Contingencies 0.3 1.8 2.1 15 10 Price Contingencies 0.2 1.47 TOTAL PROJECT COSTS1j/ 3.8 21.6 25.4 15 117 I/ Including duties and taxes of US$1.62 million. Proposed Flnancin! Plan by Project Component IDA Local Govt Ii Communities ............... .(US$ million). Investment Component A. Water Schemes 2.5 0.2 - 2.7 B. Markets 8.4 2.7 - 11.1 C. Access Roads 4.0 0.4 - 4.4 D. Storage/Processing 0.3 0.1 - 0.4 Studies and Technical Assistance 2.7 0.1 - 2.8 Project Unit 2.1 - 0.3 2.4 Refinancing PPF I 5 OA TOTAL 1/ 21.5 3.6 0.3 2S.4 I/ Includes duties and taxes of US$1.62 million.  - vi - Prose n Finacing Plan by Foreign Exchange and Local Costs Foreign Local Total . (US$ million). IDA 3.8 17.7 21.5 Local Government and Communities 0.0 3.6 3.6 Central Government 0 u 03 TOTAL 1 3.8 21.6 25.4 1/ Includes duties and taxes of US$1.62 million. Estimated Schedule of Disbursements ................... US$ million). IDA Fiscal Ye 1994 199S 1996 1997 1998 1999 Annual 2.3 2.3 3.5 5.5 5.5 2.4 Cumulative 2.3 4.6 8.1 13.6 19.1 21.5  REPUBLIC OF GHANA AGRICULTURAL SECTOR INVESTMENT PROJECT STAFF APPRAISAL REPORT 1. INTRODUCTION 1.i In 1990, the Government of Ghana prepared a Medium-Term Agricultural Development Strategy (MTADS), with assistance from IDA. The strategy focused on the institutional and infrastructure needs for more rapid growth in agriculture, and highlighted the main policy changes necessary to support such growth. Since publication of the MTADS, IDA assistance has been given for several parts of the program, covering agricultural research, extension, livestock development, environmental protection and, througi. a 1992 Agricultural Sector Adjustment Credit, the creation of a more liberal trading and marketing system. The Government has now requested IDA to continue its support for rural development by financing investments in irrigation, rural access roads, market structures and small-scale agro-processing. 1.2 In recent years, the Ministry of Food and Agriculture (MOFA) has prepared several studies and organized workshops on irrigation development and agricultural marketing. All these studies have indicated the Importance of small-scale rural infrastructure and post-harvest processing in increasing rural productivity, incomes and employment. They have also highlighted the difficulty of planning such smail-scale investments and budgeting for their implementation, and have recommended that the most effective approach is to work through District Assemblies, farmer groups and rural communities in elaborating and implementing them. Such an approach is consistent with PNDC Law 207 (1988), which allows for increasing administrative decentralization and delegation of authority to local governmen. and assigns to District Assemblies development functions in a variety of crucial areas, including road, market and water infrastructure, education, health, town and country planning and agriculture. 1.3 Experience with the IDA-financed Ghana Community Secondary Schools Project and the PAMSCAD 1/ initiative indicates that rural communities and local governments are quite capable of planning, impiementing and maintaining rural infrastructure projects as long as they have a significant voice in the decision-making, planning and implementation process. Moreover, local communities have the capacity to contribute financially and in kind towards the construction of such infrastructure. Consequently, following a Government request, IDA organized in August 1992 a joint mission with a project preparation task force composed of staff of the Ministries of Agriculture and Local Government, local consultants and farmer/community leaders to design a project which would finance small-scale rural investments and would strengthen the capacity of local communities to identify, appraise and manage such investments. The project was appraised by IDA in June 1993. I/ PAMSCAD: Progra of Actions to Mitigate the Social CoStS of Adjustment in Development.  - 2 - II. BACKGROUND A. The Economic Setting 2.1 Between 1980 and 1983, Ghana's GDP declined at an average annual rate of S percent, following a period of negligible growth during the 1970s. During that period, the currency was significantly overvalued, agricultural producer prices were low and Government controls and participation in the economy excessive. Agriculture, representing the bulk of GDP, was virtually neglected resulting in a steady decline in the production of food. Cocoa, the leading export crop, was subject to high net taxat.on and poor producer incentives, leading to a dramatic and steady decline in both production and export. The decline in food production and exports (made worse by drought and bush fires in 1982-83) contributed to food shortages, galloping inflation and a severe balance of payments crisis. 2.2 Starting in 1983, the Government introduced a series of far-reaching policy reforms under the Economic Recovery Program (ERP). These included devaluation of the Cedi, liberalization of internal and external trade and reduction of subsidies and price controls. As a result, the performance of the economy improved steadily up till 1991. Real GDP grew at an annual average rate of 5.4 percent between 1983 and 1991, and the rate of inflation dropped from 40 percent in 1987 to 18 percent in 1991. A sizeable overall balance of payments surplus was maintained between 1987 and 1991. 2.3 In 1992 and 1993, the pace of reform and the growth of the economy slowed somewhat. A succession of adverse developments in 1992, some of which were related to the political transition to civilian rule, caused a drop in the real growth rate to 3.9 percent. Low world market prices of cocoa, gold and timber affected export earnings, and the Government substantially increased civil servant wages, creating an unsustainable budget deficit. In the 1993 budget, however, significant corrective measures were introduced, aimed at restoring macroeconomic stability. The Government has increased petroleum taxes, accelerated the divestiture program and intersfile tax collkction measures. Public sector wages and salaries are expected to be kept at current levels, and the improved fiscal position will permit the Government to make repayments to the banking system, thereby improving the availability of credit for private investment. During the first quarter of 1993, consumer prices rose by 13 percent and the cedi depreciated by 15 percent. But, reflecting the impact of budgetary measures, the exchange rate and inflation have stabilized since then. GDP growth is expected to recover to 4.4 percent in 1993 and to reach 5.1 percent in 1994. B. The Rural Sector Performance of the Agricultural Sector 2.4 Agriculture remains the dominant sector of the Ghanaian economy, contributing nearly half of the GDP and about three quarters of export earnings. It provides a livelihood for about 70 percent of the population. The nascent industr.al sector depends on agriculture as a source of raw materials. Agriculture is an important source of public revenues and has an important impact on both public and private savings. Food prices account for a significant  - 3 - component of household expenditures. Hence, agriculture also has substantial influence on real wages and is a key determinant of overall macro-economic performance. 2.5 Agriculture was severely hampered by the adverse economic conditions of the 1970s and early 1980s. As a result, the per capita food production index declined from 100 in 1974-76 to 62 in 1983. Following the introduction of the ERP, there was some improvement in agricultural performance. Agricultural output as measured by sectoral GDP grew in real terms for five successive years from 1983, and in 1988 achieved its highest level since 1974. Nevertheless, sectoral growth, at around 2 percent per year since 1983, has still lagged significantly behind that of the economy as a whole, and the share of agriculture in total GDP has fallen from 55 percent in 1982 to around 47 percent at present. Characteristies of Ghanaian Agriculture 2.6 The total land area of Ghana is about 22.4 million ha, of which 1 percent is cultivated and 7 percent is under perennial tree crops (cocoa, oil palm, and rubber). Ghana's agriculture is predominantly sma1lholder, traditional and rain-fed. The mean farm size is less than 1.6 ha (4 acres). Small- and medium-sized farms of up to 10 ha account for 95 percent of all cultivated land. Of the total 2.37 million farms operated in Ghana by smallholders, about one quarter produce mainly for subsistence, about 55 percent sell up to half of their produce and less than a quarter market more than half of their produce. This signifies the importance of smallholders for family, regional and national food security and for achieving sustained growth in agricultural production. 2.7 Common factors in Ghanaian agriculture are the use of bush fallow to restore soil fertility, mixed cropping to minimize risks and, in the north of the country, the widespread integration of livestock into farming systems. Women head 30 percent of rural households and are responsible for about 70 percent of the total food production. They often make decisions about the type and area of crop to be planted and have a key role in weeding, harvesting and processing. Even in male-headed households, women take care of small livestock. There is little use of purchased inputs and land preparation is manual in most areas, though ox-traction is important in the North. Mechanization is used only by a few large enterprises, and irrigated agriculture is poorly developed, covering only 9,000 ha throughout the country. 2.8 Traditiornal farming systems have developed over time as adaptations to the six major agro-ecological zones in W3hana. These zones are: Rainfall Area Percent of Zone (mM g.a.) ('000 ha) Total Area Rain Forest 2,200 750 3 Deciduous Forest 1,500 740 3 Transition 1,300 6,630 28 Guinea Savawna 1,100 14,790 63 Sudan Savanna 1,000 190 1 Coastal Savanna 800 580 2 Total Z2360 100 Source: MOPAIFAO Ptoject GHA/84/003  - 4- 2.9 In the two forest zones, tree crops are predominant with cocoa, oil palm, coffee and rubber being of particular importance. Food crops in this area are mainly inter-cropped mixtures of maize, plantain, cocoyam and cassava. The transition zone is characterized by mixed or sole cropping of maize, legumes, cocoyam or yam, with tobacco and cotton being the predominant cash crops. Cotton and tobacco are also important in the savanna zones, where food crops are mainly sorghum, maize, cowpeas and yam in the wetter areas, whilst sorghum, millet and cowpeas predominate in the drier North. Rice is locally important in all zones, being grown in seasonally-flooded valley bottoms. In the 1970s, cocoa contributed almost 30 percent of agricultural GDP, but this has declined significantly in recent years. The gap has been filled by roots and tubers, plantain and to some extent cereals. At present, the breakdown of agricultural GDP is approximately 60 percent roots, tubers and plantain; 13 percent cocoa; 9 percent livestock and fisheries; 7 percent cereals; and 11 percent forestry and miscellaneous products. Growth Potential 2.10 In general, crop yields in Ghana are very low, and large increases could be achieved through simple changes in crop husbandry such as contour plowing, timely planting and weeding, plant spacing, composting, manuring and fertilizer application. In Ghana, as elsewhere in Africa, however, traditional methods of production are slow to change. Such resistance to change is the result of physical isolation, aversion to risk and a history of government intervention which has effectively taxed away the potential for capital accumulation in the farming sector. But it is also the result of poorly organized government- run technical support services and neglected rural infrastructure, constrained by over- centralized decision-making. The lack of rural infrastructure is a particularly critical constraint to growth, because it isolates rural communities from the marketing system and imposes very high costs on rural labor. With only 3,300 km of Ghana's 21,300 km of feeder roads currently usable in the wet season, farmers spend much of their time head-loading commodities from field to home and from village to market. 2.11 The need for such change was the underlying theme of the MTADS, which set out a number of strategic priorities for achieving higher growth in the sector. A target growth rate of 4 percent per annum was established as a benchmark, but to achieve this a far nmore rigorous definition than in the past was needed of the role and responsibilities of the public sector. This was long overdue. Agriculture had lagged behind other sectors in being liberalized from government interference under the ERP. By the late 1980s, for example, the Government still controlled trade in many basic commodities, it distributed agricultural inputs, produced seed, set minimum prices and ran large strategic storage and agro-processing facilities. Recent policy decisions have mandated the withdrawal of the public sector from most of these activities, leaving it to concentrate its limited financial and administrative resources on activities where public intervention is more clearly needed, that is in providing better infrastructure and services and in monitoring and regulating the use of natural resources. Given these changes, the prospects for more rapid growth in agricultural output are better than they have been at any time in the last 20 years, provided an adequate flow and effective allocation of resources can be maintained for investment in rural areas.  -5- The Irrlgation Sub-Sector Za 2.12 Poor hydrology and relatively flat topography hamper the development of irrigation in Ghana. The bulk of surface water cannot be used for irrigation in a cost-effective way, while average water yields of shallow tubewells in pilot projects have been found to be low and only useable for drinking water and micro-scale irrigation. The physical potential for irrigation in the whole country has been estimated to be about 120,000 ha, but the economic potential is significantly less than that. The prospects for cost-effective development are in low-lying valley bottoms and river flood plains and on levels commanded by existing dams. 2.13 The Ghana Irrigation Development Authority (GIDA) is the main irrigation develepment institution and operates under MOFA as a semi-autonomous body with four regional representatives. Since 1977, GIDA has built, and continues to manage most irrigation schemes, although autonomous operating authorities have been set up for some larger schemes. Local authorities operate small water projects. In the North, the Small Irrigation Division of the Upper Region Agricultural Development Project (URADEP), the Irrigation Company of the Upper Region (ICOUR) and some NGOs assist local communities to develop irrigation. Regional MOFA offices are responsible for extension services. Where crop husbandry practices for rainfed and irrigated crops differ, GIDA assists farmers and MOFA staff with technical advice on irrig..ted crops. 2.14 Most of the irrigation in Ghana, totalling about 7000 ha, has been developed and is managed by GIDA. GIDA's schemes vary in size from 100 to 2,500 ha and most of them are beset with problems. Poor site selection, inappropriate design and unsuitable contracting arrangements have led to unacceptably high costs. Poor extension services have resulted in low yields, while land tenure problems, high maintenance costs and a dependence on GIDA for land preparation have resulted in a lack of farmer interest. This has been exacerbated by a lack of farmer participation in irrigation design and management. 2.15 Private irrigation is very limited in scope. Small dams and dugouts for drinking water have been constructed in the northern regions, and these sometimes allow small-scale irrigation for traditional vegetable production on small plots. Some private irrigation schemes for commercial crops have also been constructed along river banks, using low-lift pumps, and a number of community-owned valley bottoms in the South have shown potential for rice under increased water control. Additional studies to identify valley-bottom schemes need to be carried out, however. 2.16 The main lessons which stand out from Ghana's past experience in irrigation relate to participation and management. Most of the problems arise from a lack of farmer involvement in design and implementation and from a lack of group cohesion to ensure cost-effective water distribution. Irrigation in Ghana has been shown to be costly under any circumstances, and even the most cost-effective schemes are likely to be justified only where rainfed farming is difficult and valuable dry-season crops can be grown. For its future development strategy, therefore, the Govermnent has chosen, first, to exploit existing investments by reducing operating costs and completing irrigation schemes wherever economically justified and, second, to focus on developing micro- and small-scale irrigation schemes, which can show ai Details on inigation issues are in Worling Paper No 3.  -6 - good economic returns and can be managed by the farmers themselves through rural farmers' groups. The best prospects lie in developing fadama schemes with shallow tubewells and in areas subject to seasonal flooding, where construction of simple hinds and structures would enable some degree of water control. High priority is also needed for training in irrigation techniques. Agricultural Marketing, Storagie and Processing a/ 2.17 Marketing Intitutions, The two principal public sector institutions involved in agricultural marketing are the Ghana Food Distribution Corporation (GFDC) and the COCOBOD. The GFDC has in the past been engaged in buying and storing cereals and has also attempted to assist the private sector by testing and developing technology for on- and off-farm storage. COCOBOD has traditionally held a monopoly of cocoa marketing. The Ghana Export Promotion Council (GEPC) is also involved in agriculture in a small way by promoting export marketing of non-traditional commodities through interaction with chambers of commerce, exporters' associations and the business community at large. Recent decisions by government, supported by IDA under the AGSAC (1992), have been directed at eliminating the involvement of GFDC in commodity storage and marketing and at introducing licensed private traders into domestic cocoa buying. With the exception of the export of cocoa, the role of the state in agricultural marketing (both inputs and outputs) is now being reduced to negligible proportions. 2.18 Several institutions and NGOs are involved in promoting small-scale processing. The Food Research Institute organizes training programs for artisanal fish and cassava processing. The Ghana Regional Appropriate Technology Service (GRATIS) of the Ministry of Industry, Science and Technology have set up operational centers in Tamale, Tema, Kumasi and Bolgatanga. They aim at transferring technology to the informal industrial sectors, training technicians from local technical institutes and providing manufacturing, repair and service facilities for small-scale industry in the rural areas. The Science Import Substitution (SIS) enterprise is manufacturing food processing equipment such as corn mills, gari graters, and oil presses. The National Council on Women in Development and the 31st December Women's Movement undertake projects to demonstrate to women various opportunities for income generation in small-scale agro-processing. 2.19 There are few livestock marketing institutions. The Meat Marketing Board has been abolished, and no significant cattle owners' association exists, though there is a Poultry Council and Feed Manufacturers' Association which promotes its members' interests. Within MOFA, the Animal Health and Production Department is responsible for animal disease control, meat inspection, drug control, breeding farms and nutrition and range management, all of which have some bearing on the marketing of livestock. Three main fishermen's organizations promote the interests of fishermen: (i) the National Fisheries Association; (ii) the National Council of Inland Fishermen; and (iii) the Ghana National Canoe Fishermen's Council. The Department of Fisheries (MOFA) is promoting sustainable marine fisheries and aquaculture through research and extension, which covers aspects related to the processing and marketing of fish. 3/ Details on markets, storage and processing are in Working Papers No 4 and 6.  -7 - 2.20 Small-scale Marketing. Storage and Processing. Most agricultural marketing in Ghana is conducted by small-scale operators. The principal agents are women who bring the farm output and food products by headload to rural markets. A dense network of private wholesalers further distributes the merchandise. Cotton, tobacco, oil palm and rubber marketing is organized by the processors in an effort to vertically integrate marketing and processing. Smallholders also sell oil palm and rubber to traditional processors or small traders. Sugarcane farmers market cane to small-scale sugar syrup processors. 2.21 Rural food markets lack basic infrastructure such as paved floors, cover protection, water supply, refuse collection, storage, etc. Poor market conditions increase produce spoilage, transport and marketing costs, and health hazards, and they adversely affect prices. Poor market access roads and marketing information diminish market integration and increase marketing costs. Livestock markets are in open areas without watering, kraaling or animal loading infrastructure. Insufficient landings on the Volta Lake are a major fish marketing constraint. 2.22 Storage is a major constraint to effective marketing, both in the urban centers and on the farm. A long tradition of public purchasing of grains and a lack of capital for municipalities and private traders have resulted in a completely inadequate storage infrastructure. Few commercial farmers have installed grain stores, and urban wholesalers usually store produce only for short periods. Lack of suitable storage for the various tiers of the marketing chain (on-farm, local and regional wholesale markets) is a crucial impediment to efficient marketing, creates strong seasonal price fluctuations, discourages commercial production, increases post-harvest losses and wastage, and generates multiple small transactions inflating handling, transport and marketing costs. 2.23 Women are responsible for most food processing, using simple methods to make locally-consumed products. Small-scale food processing includes processing of palm kernel seeds and shea nuts into oil and butter; cassava into gari, starch, dough, pellets; maize into maize meal and fermented dough; and fish into smoked and dried products. The traditional processing sector is severely undercapitalized, resulting in low labor productivity and poor product quality. Most slaughterhouses lack hygienic slaughtering facilities: animals are slaughtered on the floor in overcrowded spaces without adequate water supply. Fish processing (drying, smoking) is very labor intensive. Modernization of equipment is needed to increase productivity levels, but access by women to credit for this purpose is very limited. Rural Roads 41 2.24 The severely deteriorated condition of the feeder and market access road network is a major obstacle to increasing farm production. Two thirds of the feeder road network remains in bad condition due to past neglect. This has led to frequent disruptions of goods and passenger transport in most rural areas of Ghana, with scores of farm areas losing access to markets. Feeder and market access road rehabilitation and maintenance using labor-based small- and medium-sized private contractors is critically needed. Rural Ghana resembles a 'footpath economy' with women spending substantial amounts of time transporting 4/ Details on Rural Road Infstmueture are in Working Paper No 5.  - 8 - merchandise, fuel wood and water on their heads. At the present rate of funding (about US$5 million per annum) Ghana's feeder roads rehabilitation would take more than 30 years. 2.25 The responsibility for planning, construction and maintenance of feeder roads lies with the Department of Feeder Roads (DFR) under the Ministry of Roads and Highways (MRH). Under the Government's decentralization policy, DFR has proposed to create 110 District Offices. DFR is considerably under-staffed, especially in middle-management grades. It urgendy needs to recruit and train qualified staff or consultants including planners, engineers, designers, economists and supervisors. Strengthening the management capacities of the Department of Feeder Roads (DFR) to plan and supervise rehabilitation and maintenance works in the future is the highest priority. Particular emphasis is being placed on the "rural access and mobility approach", targeting poverty alleviation in rural areas by combining low- cost non-motorized transport technology (bicycles, bicycle trailers, hand-propelled farm carts) and low-cost rural road construction ("one-blade roads"). This will help to raise rural productivity and incomes and involve NGOs and women's organizations in design and implementation. The IDA-funded feeder road program for 1990-2000 comprises 18,000 km of rehabilitation, construction of new roads, repair of 1,200 water crossings and development of a sustainable maintenance system. local Government Decentralization 5/ 2.26 Ghana is decentralizing its government and entrusting development activities to the 1 10 newly-created District Assemblies, which are becoming increasingly responsible for planning, budgeting and revenue collection. Key constraints to improving local government management are: (i) lack of financial resources and qualified staff; (ii) centralized control by ministries over their district staff; and (iii) inadequate logistical support to the districts. A basic feature of the 1988 decentralization law (Law 207) is to integrate all agencies of line ministries at the local level under the authority of the district administration. All District Assemblies were elected in early 1989 and can now take decisions on a wide range of issues. While management problems at the planning, budgeting, fiscal and project implementation stages are being solved, MOFA is also implementing its own internal decentralization program. 2.27 District Assemblies. A key element in decentralization is that districts are planning and managing their own budgets and staff. District Assemblies have established sub- committees responsible for: social services, infrastructure, finance and administration, education, justice and security, economic development (including agriculture), environmental management, and women's affairs. They have also created District Development Planning and Budgeting Units (DPBUs) and Regional Coordination Councils. To prepare their budget, District Assemblies receive support from: (i) technical officers assigned by the line ministries; (ii) budget officers to assist DPBUs; and (iii) Regional Mobile Planning Teams. 2.28 Reform of revenue collection and control is a second key element of decentralization. Fiscal decentralization means: (i) revenue sharing between central and local government; (ii) delegating some taxing capacity to the districts; (iii) delegating procurement to local tender boards; (iv) introducing less cumbersome spending controls and procedures; and (v) allowing 5/ Details on the organization of local govenment are in Workdng Paper No 2.  -9 - districts to attract external aid through direct channels. One important element of fiscal decentralization has been the development of a composite budget: a compilation of the district's 22 decentralized central government department budgets, together with a separate listing of the recurrent and development budgets to be funded from the District Assemblies' own general revenues. This composite budget is to give each District Assembly a broad view of the total resources likely to be available for development activities in the district. Local authorities procure goods and services within the district using their own resources. All procurement has to be channeled through the District Tender Boards. Where central government's financial resources are involved, payment procedures through the Ministry of Finance and Economic Planning become overly complex. 2.29 Decentralization is not without implementation problems. In reality, few districts have development plans or budgets, and those few mostly lack funds for implementation. Revenue sharing between central and local governments is still at the initial stages and districts have very few resources which can be taxed to provide District Assemblies with revenue. Previous constraints are still valid: (i) the increased number of districts has created needs for additional qualified staff; (ii) the roles and responsibilities of the ministries, departments and agencies with regard to decentralization have not yet been clearly defined; and (iii) weaknesses at the center in designing and implementing effective planning and budgeting systems persist and cause difficulties at local level. Local and central authorities, however, are designing simpler management methods, which will improve coordination, provide greater autonomy for the districts and increase accountability. 2.30 Minist of Food and Agriculture. MOFA has been a leader among various ministries in decentralizing its operations to the regional and district levels. The Ministry's main concern is to shift functional responsibility for formulating district and regional agricultural development plans and for making budget allocations to the district level. Furthernore, MOFA has taken five specific actions to adapt its operations to the decentralization program. First, the Government's decentralization program calls for planning departments to be developed in each ministry. MOFA's Policy Planning, Monitoring and Evaluation Department (PPMED) has made substantial progress in strengthening its planning, technical support, monitoring, evaluation, and reporting functions. Secondly, MOFA has appointed district coordinators for the agricultural sector in each district to liaise on all agricultural policies and programs with the Economic Sub-committee of the District Assemblies and with the DPBU's. Thirdly, it has established District Agricultural Advisory Committees consisting of members of the agricultural departments and select district farmers. And fourthly, the regional agricultural management team has prepared a regionally-oordinated set of district development plans, and MOFA has set up a training program for regional planners. C. IDA Assistance Stratel 2.31 Over the past decade, Ghana has implemented one of the most successful adjustment programs in Sub-Saharan Africa. During that time, under the Economic Recovery Program (ERP), the Government has dismantled price, distribution and import licensing controls, liberalized interest rates and foreign exchange, eliminated subsidies and improved the tax structure and collection. It has also reformed the banking system, improved the allocation of public expenditures and launched a program of public service retrenchment and divestiture of  - 10- state-owned enterprises. The ERP has been supported by a series of IDA Structural and Sectoral Adjustment Credits and by co-financing from bilateral and multilateral sources. In addition, an extensive program of support has been mounted in all major sectors concentrating on investment and institution-building. Looking to the future, it is now possible to foresee an end to adjustment lending and a shift towards sector investment lending in support of an accelerated growth strategy. The outline of this strategy has been set out in the recent report 'Ghana: 2000 and Beyond". 2.32 IDA support for agriculture in recent years has been fully consistent with this general strategy and with the priorities set out in the MTADS. These were: (i) privatizing commodity buying, trading, processing, seed production, input supply and some technical (e.g., veterinary) services; (ii) setting clear objectives in research, extension, animal and plant disease control; (iii) strengthening institutions to provide these services and to regulate natural resource use (forests, soils); and (iv) concentrating on low-level infrastructure investment (small-scale irrigation, feeder roads, rural markets) and decentralizing investment decisions to respond more rapidly to the needs of rural communities. All of these priorities except the final one have been supported by adjustment and investment operations in the last two years. Since 1990, there have been five new IDA operations in agriculture, one Sector Adjustment Credit concerned with the withdrawal of the State from marketing and input supply; three credits supporting agricultural services in the public sector (Research, Extension and Livestock); and one credit (Environment) supporting public institutions charged with regulating and managing natural resources. The credit now proposed (Agriculture Sector Investment Project) is designed to provide resources for low-level investment in rural infrastructure. D. Bank Group Operations 2.33 From 1962 to June 30, 1993, IBRD and IDA lent to Ghana a total amount of US$2.7 billion, of which US$1.7 billion was for investment lending. Of the total lending, agriculture took up 10.8 percent or US$296.5 million; transport 13.8 percent; energy 11.9 percent; urban infrastructure 6.5 percent; industry 5.3 percent; education 5.2 percent; environment 2.1 percent, technical assistance 1.9 percent; water supply and sanitation 1.9 percent; health 1.7 percent; and telecommunications 1.5 percent. Policy-based lending amounted to 37.6 percent. During thb 'ast ten years, US$1 billion has been lent for policy-related adjustments, including the AGSAC of US$80 million in FY92. 2.34 During the last five years IDA has lent in the agricultural sector (net of cancellations) a total amount of US$151.4 million, including the National Livestock Services Project (FY93 - US$22.5 million); the National Agricultural Extension Project (FY92-30.4 million); the National Agricultural Research Project (FY91-US$22 million); the Agricultural Diversification Project (FY91-US$16.5 million); the Rural Finance Project (FY89-US$20 million); and the Cocoa Rehabilitation Project (FY88-US$40 million).  - I1 - E. Rationale for IDA Involvement 2.35 There is an urgent need to increase the flow of resources for infrastructure development at village and district level and to change the process by which they are allocated. This is a particularly good time for IDA to take such an initiative, since the Government has embarked upon an administrative decentralization program aimed at allocating planning, budgetary and implementation responsibilities to local authorities. The 1988 "Local Government Law" referred to earlier established 110 District Assemblies and devolved the responsibilities of 22 Central Government Departmnents. Ghana's policy objectives with respect to this decentralization are essentially three-fold: (i) to increase direct popular participation in development decision-making; (ii) to increase the revenue base and efficiency of revenue collection at local levels in support of development activities; and (iii) to improve managerial and admninistrative efficiency. IDA's involvement, as envisaged in this project, would promote such decentralization, develop a coherent strategy for farmer participation in investment and lead the way towards locally-managed agricultural marketing and rural infrastructure improvement. F. Environmental Issues 2.36 A growing concern for environmental issues led the Ghana Government to prepare a national Environmental Action Plan (EAP) in 1990. The Environmental Protection Council (EPC) is generating widespread awareness of the importance of environmental issues and a consensus on the need for effective action. The IDA-financed Environmental Resource Management Project is carrying out studies and actions to prevent soil degradation and erosion, deforestation and forest degradation, and degradation of the coastal zone environment. 2.37 The project now proposed supports EAP goals. Small-scale irrigation would increase agricultural intensification and thereby relieve pressure to exploit fragile marginal soils, and the impact of investment in rural market facilities would be highly positive in terms of health and waste reduction. There are, however, some possible environmental problems for which mitigation measures will be needed. They include soil erosion during construction, water quality degradation through agricultural intensification and water-borne diseases. The significance of the possible environmental impact of each proposed community project will have to be evaluated during its design and mitigation measures taken. To meet this need: (i) consultants preparing subprojects would include an envirommental assessment in the project preparation studies; (ii) the local health services would thoroughly inform local communities of the health hazards involved in constructing water reservoirs and markets and, if necessary, carry out health hazard eradication campaigns; (iii) project staff would help the beneficiary communities to take preventive measures such as introducing biological defense mechanisms, perimeter fencing, farmer training in the use of chemicals, imposing strict limits on land excavation; and (iv) EPC would carry out regional studies and actions to increase community environmental awareness.  - 12 - G. Lessons from Previous IDA Involvement 2.38 The lessons learned from IDA's activities in Ghana are country, sector and project specific. 2.39 Country-specific Lessons. The 1991 and 1993 Country Implementation and Strategy Reviews recommend the following measures to improve project implementation and increase impact: (i) to encourage infrastructure investments that support private sector activity and alleviate poverty; (ii) to address project implementation capacity constraints from the outset of the project; (iii) to design projects with ease of implementation in mind, i.e., few project components, one implementing agency, few conditions and a timetable taking into account local conditions; (iv) to strengthen project implementing agencies to improve accounting and procurement; (v) to use local instead of foreign consulting services; and (vi) to increase aid coordination. 2.40 Secor-specific Lessons. The 1991 Agricultural Portfolio Implementation Review indicated the following main shortcomings in project implementation and solutions: (i) Inadequate Implementation capacity-solutions would be to improve staff training and recruitment, and to merge agencies with overlapping responsibilities; to decentralize management, introduce a participatory approach to project management; and to increase accountability for performance; (ii) project complexity-project design should be simplified to fit implementation capacity and to minimize risk of failure; (iii) procurement delays-delays can be reduced by promoting the use of sample bid documents and contracts; preparing procurement plans and schedules; assigning procurement officers; and promoting consistency amongst projects and sectors in procurement procedures; (iv) training-the demand for external training is high, training is expensive and absence of staff during training adds to project implementation problems. The solution is to bring trainers to Ghana instead of sending staff abroad; and (v) inadequate counterpart funding-project design should be more realistic about the availability of counterpart funds, and counterpart fund deposits in project accounts should be released in advance, or kept in a central pooled account. 2.41 Project Specific Lessons. Two projects, similar to the one now proposed, have been implemented in Ghana. The Program of Actions to Mitigate the Social Costs of Adjustment in Development (PAMSCAD), supported by 14 donors under the social dimension of adjustment initiative, has financed, among other things, local initiatives and small-scale poverty-related projects in rural areas. The IDA-financed Ghana Community Secondary Schools Construction project finances two-thirds of the local communities' school construction. The lessons learned from experience with those two projects are summarized below and are incorporated into design and implementation of the project. 2.42 The 1990 PAMSCAD mid-term review mission clearly identifies management priorities and implementation bottlenecks. Its recommendations include: (i) to simplify the project's operational scope which covers nine sectors involving 13 different national implementation agencies/departments; (ii) to devolve responsibility for initiating subprojects to the district level, with responsibility for approval at the regional level and responsibility for supervision and monitoring at the national level; (iii) to reinforce efforts to upgrade district- level planning, implementation and management capacity; (iv) to place responsibility for leadership and policy oversight for PAMSCAD under one PNDC Secretary; (v) to focus on a  - 13 - limited number of subprojects with a high payoff, to target community-initiative projects to the poorest areas, and to permit a lower level of contribution from the poorest communities; (vi) to reduce the number of donors for each project; (vii) to avoid creation of new institutions, which absorb considerable human and financial resources; (viii) to disseminate information on project activities to the rural areas; and (ix) to introduce an extensive monitoring and evaluation system of beneficiaries, costs and resource mobilization by communities. 2.43 The IDA-financed education project funds secondary schools through matching grants. Lessons learned from this project are the following: (i) simple design of schools and uniform cost standards and contributions have made project implementation easier to administer; (ii) the total community effort should be kept low in absolute terms; (iii) independent audits and strict cost control by IDA staff are crucial in keeping costs in line with the private sector; and (iv) where District Assemblies have participated financially in school construction, implementation has been slower than where community groups have taken full financial responsibility for their contribution. 2.44 The above lessons would be addressed under the proposed project as follows: (i) implementation capacity would be built up through preparation and pilot activities and recruitment of professional private sector managers. Project Unit (PU) managers would be empowered with sufficient authority to react flexibly to rural needs. PU would establish regional offices and employ farmers as facilitators to respond more rapidly to local needs; (ii) project components would be limited to investments, feasibility studies and capacity building closely related to implementation; (iii) the project would be subject to an annual review; (iv) no large procurement contracts are anticipated; and (v) the project would only support investments with significant local participation, which would decrease dependance on central funding and administration. 111. THEPROJET A. PBroect Objectives and Scope 3.1 The main project objective is to increase the flow of technical and financial resources for rural development in response to perceived needs at the village and district levels. Related objectives are: to increase agricultural productivity and access to markets; and to strengthen the communities' capacity to develop, manage and maintain their investments using the resources of NGOs, local consultants, and public institutions. B. Summary Description Investment Component 3.2 The project would finance civil works and equipment to upgrade small-scale water schemes, rural markets, market access roads and small-scale agro-processing. Local  - 14- communities would identify subprojects, participate in their design and development and contribute in cash or kind towards their implementation and maintenance. Total costs of this component are estimated at US$18.6 million. Sudies and T_chnical Assistance 3.3 The project would finance the services of technical support agencies (TSA) to help local communities to formulate and implement local investments. NGOs and local consulting agencies would assist the beneficiaries with feasibility studies, management services and training in irrigation, rural business and local government management. MOFA's Policy Planning, Monitoring and Evaluation Department would monitor project impact and performance. Total costs of this component are estimated at US$2.8 million. Project Implementation 3.4 A Project Unit (PU) within the Ministry of Food and Agriculture (MOFA) would manage the project. MOFA would recruit key PU staff with private sector experience and delegate project implementation responsibility to PU, which would manage the project according to a detailed project implementation manual prepared in consultation with IDA. Initially, PU would establish offices in Accra and Tamale. It would have a limited number of administrative and technical staff, recruit part-time local consultants to carry out technical studies and supervision, and employ farmer facilitators to ensure promotion and liaison with the rural communities. Total costs of this component are estimated at US$2.4 million. C. Detailed Features Investment Component 6/ 3.5 Small-scale wat.-r schemes (water schemes) would support irrigated agriculture as well as potable water supply for livestock and humans. They include water conservation structures such as small dams and dug-out reservoirs, stream diversions, valley bottom schemes, pump schemes, shallow ground-water development, and other water supply and harvesting systems. This type of small-scale community investment would increase yields and dry season output of high value vegetables. Local communities have difficulty obtaining finance for these small schemes because the public sector has emphasized large-scale irrigation and the banking system does not yet provide financing for small schemes. NGOs and other technical support agencies would assist farmers to design, construct and manage their schemes and to accumulate funds to operate and renew the equipment or to expand the association's activities. The beneficiaries' initial contribution for infrastructure works would be at least 10 percent of the investment cost in labor or cash. Other contributions would depend on the type of investment. The project would finance approximately 500 ha of irrigation (20-30 subprojects) for an estimated amount of US$2.7 million. O/ A morm detailed desription of the investmet component is in Annex I and in the corrsponding working paperi.  - 15 - 3.6 Marketing infrastructure includes construction and rehabilitation of markets with storage facilities, accommodation, commercial vehicle access, sanitary facilities and management services. Improved marketing infrastructure will encourage trade and generate revenues for district authorities. Traders' associations and district authorities would be required to have an agreement on market design, financing, management, and cost recovery methods. Project funds would be used mainly for construction and rehabilitation of basic market infrastructure such as pavement, drainage, coverage, warehouses, sanitary facilities, and water supply. Traders and District Assemblies would contribute at least 25 percent of subproject costs by financing, among other things, market stores, stalls and parking. The project would support about 110 markets for an estimated total cost of US$11.1 million. 3.7 Roads to improve access to markets would be "one blade" roads carrying less than 10 vehicles per day, except on market days. Subprojects would include rehabilitation of existing roads, spot improvements, construction of culverts or small bridges, and the opening of new roads. All construction would use labor-intensive methods. The communities' minimum contribution would be 10 percent of investment costs plus full maintenance costs. The project would finance 375 km of access roads and other road works for an estimated US$4.4 million. 3.8 Storage and food processing includes village-and on-farm storage and processing such as village abattoirs, gari processors, palm oil crushers, shea-butter processors and similar community investments. Farmers in remote areas have great difficulty in obtaining formal bank credit for setting up storage and processing activities. On-farm storage would increase incentives for traders to buy from the villages, while post-harvest processing would free female labor from head-loading and manual processing to make time for more productive activities in health, education and agriculture. NGOs and technical support agencies would assist communities to realize subprojects. The costs of processing equipment would be fully recovered from the beneficiaries over a period of three years. The project would finance, on a pilot basis, about 20 initiatives for a total estimated cost of US$0.4 million. tudies and Technical Assistance Coinonen 3.9 The project would also finance feasibility studies, management services, training in irrigation, rural business and market management and food processing, as well as monitoring, evaluation, social, economic and environmental impact studies. Total costs of this component are estimated at US$2.8 million. 3.10 Feasibility Studies. Management Services and Training. NGOs and consulting agencies would carry out feasibility studies and provide management services and training. To be eligible for project support, the technical support agencies would need to be pre- qualified by PU and have an agreement with beneficiaries stipulating their terms of reference, expected output, as well as financial implications. 3.11 Monitoring, Evaluation, and Proiect Impact Studies. The project would use the services of the Policy Monitoring and Evaluation Department (PPMED) and participant observers to obtain feedback on the project's impact on local communities and efficiency to reach its objectives. 3.12 Environmental Studies. Some subprojects may require environmental assessments. As environmental assessment capacity in Ghana is yet to be developed, PU and the  - 16 - Environmental Protection Council (EPC) would finance consultants to carry out assessments for particular projects. To reduce the costs of the individual environmental studies, EPC would carry out a general watershed master plan study in the northern regions. IV. PROJECT COSTS AND FINANCING A. Cost Estimates 4.1 Total project costs are estimated at US$25.4 million (Cedis 22.2 billion) inclusive of duties and taxes in the amount of US$1.62 million (Cedis 1.3 billion). Of this 15 percent or US$3.8 million are foreign exchange costs. Investment costs (including technical assistance) amount to 98 percent and incremental recurrent costs to 2 percent of base costs. Total contingencies are 17 percent of base costs, with physical contingencies accounting for 10 percent and price contingencies for 7 percent, calculated on base costs plus physical contingencies. The appraisal exchange rate of Cedis 630 to the US Dollar, declining to Cedis 810 in 1998, and the following assumptions on the domestic and international annual rates of inflation, were used to derive the total project cost estimate. Annual Inflation Rates and Exchange Rate Assumptions International Domestic Exchange Rate (%) (%) Cedis/US$ 1994 1.2 15.0 717 1995 2.4 10.0 771 1996 3.2 5.0 784 1997 3.3 5.0 796 1998 3.3 5.0 810 4.2 The following table gives a summary of project costs by component. Additional summary tables on project costs by year, by component and summary accounts are presented in Annex 2.  - 17 - Project Cost SummarY Foreign Loal Toaw FE % of ........ (US$ million) (....... B ase 1. INVESTMENT CAT1EGORY A. Civil Works 0.2 11.4 11.6 2 54 B. Goods/Supplies/Equipment 1.7 2.7 4.4 36 20 C. Studies and Technical Assistance 0.5 3.2 3.7 14 17 D. Refinancing PPF j 0 Lk 51 7 Total Investment Costs 3.2 18.1 21.3 15 98 II. RECURRENT COSTS 0.1 0.3 0.4 10 2 MII. TOTAL BASELINE COSTS 184 21. 15 IQQ Physical Contingencies 0.3 1.8 2.1 15 10 Price Contingencies 0.2 LI 17 12 7 TOTAL PROJECT COSTS I/ 3.8 21.6 25.4 15 117 1/ Including duties and taxes of US$1.62 million. B. Proposed Financing Plan 4.3 The proposed financing plan would involve the provision of funds by IDA, the Ghana Government and the local communities. Of the total project cost of US$25.4 million, the IDA credit would finance US$21.5 million (85%), the Government US$0.3 million (l%), and the local communities US$3.6 million (14%). Net of taxes and duties IDA's financing would be 90% of project costs. 4.4 The IDA Credit would cover 100 percent of foreign exchange costs and 82 percent of local costs. The local community and government contribution would cover 12 percent of local costs, plus duties and taxes (6%). Recurrent costs are PU's operational costs and amount to two percent of total costs, including contingencies. IDA would fund about 79 percent of PU's recurrent costs, which amount to about US$0.32 million over the four-year project period. The fnancing plan is set out below.  - 18- Propgsed Financing Plan by Project Component JLcA Govt, Total Communities .......... .(US$ million). Investment Component A. WaterSchemes 2.5 0.2 - 2.7 B. Markets 8.4 2.7 - 11.1 C. Access Roads 4.0 0.4 - 4.4 D. Storage/Processing 0.3 0.1 - 0.4 Studies and Technical Assistance 2.7 0.1 - 2.8 Project Unit 2.1 - 0.3 2.4 Refinandng PPF l.S 0Q1 0 1.6 TOTAL 1/ 21.5 3.6 0.3 25.4 1/ Includes duties and taxes of US$1.62 million. Proposd Fninacing Plan byl ForeigLf Exchange and LoA"! Costs Foreign Local Total ................ (US$ million). IDA 3.8 17.7 21.5 Local Government and Communities 0.0 3.6 3.6 Central Government u 0.3 0 TOTAL 1 3.8 21.6 25.4 1/ Includes duties and taxes of US$1.62 million. C. Financial Implications for Government 4.5 Incremental recurrent expenditure resulting from the project would amount to about US$ 0.4 million. The central government's contribution is limited to the financing of PU operating costs, as local communities, including local government, would provide all other counterpart funds. The central government contributions to the project would average 85 million cedis per year, which represents 1.2 percent of the annual budget of the Ministry of Food and Agriculture.  - 19- D. Procurement 4.6 The table below summarizes the project elements and their estimated costs and proposed methods of procurement. Summary of Proposed Procurement Arrangements (US$ million equivalent) Project Element Procurement Method LCB Local Other Total Shopping 1. Civil Works a) Water schemes 0.8 1.2 2.0 (0.7) (1.1) (1.8) b) Rural markets 3.3 5.0 8.3 (2.5) (3.81 (6.3J c) Access roads 2.0 1.3 3.3 (1.9) (1.1) (3.0) d) Storage/Processing 0.1 0.1 (0.1) (0.1) 2. Equlpment/SuppileslMatefials a) Rural Investments 0.3 4.7 6.0 (0.3) (3.8) (4.1) b) PU 0.3 0.3 (0.3) (0.3) 3. Consulting Services 4.4 4.4 (4.1) (4.1) 4. PU Operatng Costs 0.4 0.4 (0.3) (0.3) 5. PPF 1.6 1.6 (1.5) 71.5) TOTAL 6.5 12.5 6.4 25.4 (5.5) (10. 1J (5.9) (21.5) Note: Figures in parentheses are the respective amounts financed by the Credit LCB covers stndard end sinmlified LCB procedures, as well as the Distrct Tender Boad procedures 4.7 There would be no ICB procurement, since all works are scattered and in remote places and overseas contractors are unlikely to be interested in bidding. Contract amounts would, on average, not exceed US$40,000. Procurement of standardized goods or materials would not be possible, as all goods and equipment would be tailored to beneficiaries' needs. Moreover, some works would be pre-financed end e^ .ic d out uY 'l -b muom4wimau ius uLemseives, and reimbursed by PU. Procurement would follow the District Tender Boards procedures, amended to enable foreign firms to participate and to ensure that contracts are awarded to the  - 20 - lowest responsive bidder. A distinction is to be made between the procurement processes applicable to the PU and those applicable to subprojects. Procurement for PU 4.8 Procurement of Goods. Vehicles and Equipment. PU would need to procure vehicles, fuel, furniture and office equipment. Total expenditure would not exceed $300,000 for the whole project period. PU would procure these items through local or international shopping with price quotations from a least three suppliers, or through the Inter-Agency Procurement Services Office (JAPSO). Procurement above $40,000 per contract, if any, would be through Local Competitive Bidding (LCB). 4.9 ODerating Costs. About US$400,000 worth of supplies and services would be procured through local shopping procedures except for vehicle operating costs and government contracted supplies and services. 4.10 Recruitment of Consultants. To recruit consultants and local technical assistance, PU will follow IDA guidelines. Detailed procedures are explained in the Implementation Manual. Procurement for Subprojects 4.11 Procurement of Works. Beneficiaries would carry out procurement for subprojects, and PU would assist and monitor the process. Works may be carried out by the beneficiaries themselves or by contractors. (a) Where the beneficiaries carry out the works themselves, the project would only disburse for physical works completed and certified for quality by a PU- appointed supervising engineer, as set out in the detailed design and bill of quantities. (b) Where contractors are involved, the following rules would apply: (i) District Authorities would engage contractors according to local government procurement rules. These rules require that, whenever local government finances a project from its own resources, procurement is to be done in accordance with procedures of the District Tender Board. These include: (a) advertisement of tenders in newspapers, on radio and public notice boards; (b) detailed procedures on bid opening, evaluation and award; (c) composition of the technical evaluation team; and (d) award to the lowest evaluated responsive bidder. The procedures, which are consistent with IDA guidelines, are set out in detail in the Implementation Manual. (ii) Local Community Groups would engage contractors according to the rules in para 4.12 below. The government tender procedures would not apply in this case.  - 21 - 4.12 There are three thresholds in the procurement process for works: (a) contracts of ug to $40.000 up to an aggregate amount of US$7.5 million would constitute the bulk of the contracts. PU would approve financing according to cost estimates and bills of quantity in a detailed engineering study, certified by a PU-appointed engineer. The beneficiaries would then engage contractors through local shopping or district tender procurement procedures; (b) contracts of more than $40,Q00 but less than $150.000 up to an aggregate amount of US$6.5 million. The beneficiaries would launch a tender and advertise in local newspapers to ensure wide participation and select the lowest responsive bidder. The technical support agencies (TSAs) would assist beneficiaries in evaluation and selection of contractors. PU would verify the award process before disbursing. The threshold for IDA prior review is US$75,000. Where contracts exceed this amount, the resident project supervisor would review contract award procedures and contracts. PU has prepared simplified procurement documents and contracts for use by village community organizations, associations, local goverament and TSAs on subprojects below $150,000 and are contained in the Implemetation Manual; and (e) contracts $150.000 or more. If such contracts occur, the beneficiaries would follow the IDA standard sample bidding documents. Selection and contracts would be approved by the resident project supervisor before disbursement. 4.13 Procurement of Goods and Equipment. Procurement of goods and equipment such as building materials, pumps, fuel, rental for earth moving equipment, etc., would follow the procedures below. (a) Purchases of building materials and supplies. rental equipment and operational expenditures of less than $5.000 up to an aggregate amount of US$1.2 million would be made by the beneficiary community after local consultation of district or village suppliers (ocal shopping). If only one supplier is available, the beneficiaries may purchase from that supplier or choose to consult suppliers in other villages/districts. (b) Purchases of $5.000 or more and less than $40.000 up to an aggregate amount of US$3.5 million would be made by the beneficiaries on the basis of price quotations of at least three suppliers. PU would audit such purchases ex-post and compare prices with prevailing market prices. PU may declare mis- procurement and decline reimbursement of expenditures if the agreed procedures have not been followed or may waive the procedure in exceptional cases, such as lack of interested suppliers. (c) Purchases of $40.000 or more up to an aggregate amount of US$0.8 million would be made by the beneficiaries through contracts awarded on the basis of competitive bidding. Beneficiaries would request approval from PU before the purchase. PU would verify the proposed prices and quality with  - 22 - prevailing market prices before approval. The threshold for IDA prior review would be US$40,000. The award procedure of contracts exceeding US$40,000 would be reviewed by the resident project supervisor. 4.14 Contracting of ConsultantslTSAs. As not many consultants live in rural areas, beneficiaries may find it difficult to attract engineers to enter into a bidding process. Direct contracting may be the simplest way to get a consultant to work on a low-cost subproject, but it may also lower quality and increase fees. For this reason, PU would make a list of acceptable TSAs and their fees available to the local communities. To generate this list, PU would launch a yearly bidding process, by which interested consulting firms would inform PU of their services, which regions they cover, their fees and their professional references. PU would retain firms acceptable to IDA on the basis of their technical and financial proposals and their work experience. Beneficiaries would be able to chose a firm on the list and would share in the costs of the studies. PU has already pre-qualified NGOs and consulting agencies, acceptable to IDA, as technical support agencies for the first year of the project. 4.15 The PU would apply the following consultant recruitment procedures: (a) Contracts of less than $10.000: the beneficiaries would select three consultants from the list of pre-qualified firms, PU would invite the firms to quote a lump sum price and award the contract to the lowest bidder. (b) Contracts of $10,000 or more: PU would invite technical and financial proposals from at least three consultants, selected by the beneficiaries, and award the contract to the highest ranked bidder, in accordance with the ranking system presented in the Implementation Manual. The resident project supervisor would review contract awards beyond US$15,000. 4.16 Procurement Assistance and Supervision. Technical support agencies and PU would assist beneficiaries to comply with the established procurement rules. Project procurement supervision would consist of PU procurement reporting and supervision, verification by the IDA resident project supervisor, and procurement audits. PU would verify whether beneficiaries follow competitive procedures and report on procurement matters in the semi- annual reports. The resident project supervisor would carry out a detailed review of the first ten subprojects as well as 10 percent of the procurement on a selective basis. MOFA would appoint a consultant to verify whether beneficiaries follow district and project procurement procedures in each of the first two subprojects entered into in respect of each category and thereafter, on a sample basis, throughout the implementation of the project, in agreement with IDA. It was agreed during negotiations that MOFA would appoint a procurement auditing consultant by August 31, 1994 (para 7.1). E. Disbursement 4.17 The IDA credit of US$21.5 million would be disbursed over a period of 5 years, beginning in FY94. The disbursement schedule is presented in Annex 4. The disbursement pace is more rapid than the historical disbursement profile for Ghana's agricultural projects, because local demand for this project is estimated to be significantly larger than the project  *23 - amount. Disbursements, however, would also depend on the management capacity of PU and the capacity of the local communities to prepare subprojects. Disbursements would be slow during the first year, but would build up rapidly afterwards. Summary Disbursement Schedule Category Amount Allocated IDA Financed ('000 US$) (%) Foreign LQol Civil works 10,000 100 90 Equipment, supplies for communities 3,800 100 90 Consulting services 4,000 100 100 Project preparation facility 1,500 PU: (a) operating costs 7/ 300 85 until 31/12/96 and 75 thereafter (b) Vehicles, goods and equipment 250 100 90 Unallocated 1,650 T-OTAL 21,500 4.18 PU would disburse as much as possible for works carried out by the beneficiaries after verification by a PU-appointed engineer and through direct payments to contractors, consultants and suppliers. Where this is not practical, the following procedures would apply: (a) District Authorities would pre-finance up to US$5,000 of all subproject costs. PU would reimburse District Authorities for expenditures up to the agreed amount, on the basis of fully-documented statements of expenditure (SOE). (b) Disbursement for village community or association subprojects would be determined in the subproject financing agreement and implementation schedule. The following guidelines would apply: (i) As much as possible, PU would disburse on the basis of statements of expenditure after the counterpart contribution in kind (abor, materials) has been provided; ;/ OpenUing cost am PU's expcndibtrue for vehicle operaion and raintenae, travel, ofrice supplies and matercs, and salaries for temporaiy contractWal saff.  - 24 - (ii) where this is not possible, but a technical support agency is involved, project fundsl/ would be advanced to the beneficiary associations' bank account where withdrawals would be sanctioned by at least two community leaders (double signature account). Advances would not exceed US$5,000 in cedi equivalent, and would only be provided after the beneficiaries have carried out a number of preparatory activities. (c) To avoid misuse of advances, PU would use severe sanctions such as loss of eligibility for further project support, publicity and legal action. 4.19 A US$1.5 million Special Account covering four months of eligible expenditures and two cedi Project Accounts (one in Accra and one in Tamale) would constitute the main project revolving funds. The initial deposit of the Special Account would be US$500,000, increased to US$1 million at the opening of the Tamale office, and increased further in accordance with need. The accounts would be opened at a commercial bank in Ghana on terms and conditions acceptable to IDA. It was agreed during negotiations that MOFA would establish and maintain a Proiect Account and that the deposit by the Government of a US$50.000 equivalent into the account would be a condition for credit effectiveness (paras 7.2 and 7.11). The Tamale project account would be opened when the Tamale office is about to be established. PU would disburse through the Project Accounts, which would be replenished with funds from the Special Account against documentation of actual expenditures. 4.20 Disbursements for expenditures under contracts valued below US$75,000 equivalent would be made against SOEs. IDA would replenish the Special Account upon review and approval of SOEs and documentation by IDA's resident project supervisor. PU would retain all documentation and IDA would replenish strictly against summary sheets approved by the resident project supervisor. Wherever possible, withdrawal applications would be aggregated in amounts of not less than US$100,000. PU would be responsible for preparing withdrawal applications and have sole responsibility for submitting withdrawal applications to IDA. PU would not transfer funds from the Special Account to the Project Accounts to provide subproject advances. This would limit PU's capacity to advance funds, but decrease risks. F. Auditing and Reporting 4.21 In addition to the procurement and disbursement supervision mentioned in paras 4.16 and 4.20 above, all project accounts, including the Special Account and SOEs would be audited annually by independent auditors, acceptable to IDA. Auditors would prepare, in addition to the auditors reports and statements of accounts, a long report including financial statements and a management letter. PU would submit the auditors reports within six months of the end of the Government fiscal year. 4.22 PU would also prepare semi-annual reports covering details on subproject and TSA activities, procurement, disbursements, key monitoring and management efficiency indicators, financial projections, and the results of project impact studies. PU would submit the reports 81 Project funds originate from the project accounts, not from the IDA Special Accounts.  - 25 - financial projections, and the results of project impact studies. PU would submit the reports to IDA by April 30 and October 31 of each year. The detailed format of the reports is described in the Implementation Manual. 4.23 The Government would also prepare a Mid-Term Review report by April 30, 1996 and a Implementation Completion Review report and provide it to IDA within six months of the closing date. Procurement and disbursement supervision, financial auditing, and semi- annual reorting requirements were agreed on during negotiations (para 7.3). V. PROJECT MANAGEMENT AND IMPLEMENTAT10N 5.1 This chapter describes the management arrangements for the project, subproject eligibility criteria, beneficiary contributions, subproject processing, training, monitoring and supervision. The Project Implementation Manual provides more detailed informnation on implementation issues. A. Management Prjct Management 5.2 The Ministry of Food and Agriculture (MOFA) would be responsible for oveiall project implementation. MOFA would delegate full project authority to a Project Unit (PU) established within MOFA. A Project Coordinating Committee (PCC) composed of delegates of MOFA, MLG, MFEP, Technical Support Agencies, Farner Organizations, and other representative institutions would supervise implementation. Regional Subproject Approval Comnmittees (SPAC) would assist regional PU offices to review and decide on subprojects. A resident IDA project supervisor would ensure operational liaison with PU. A detailed project organization and management chart is presented below. Project Coordinating Committee (PCC) 5.3 MOFA would establish a PCC composed of representatives of MOFA, Ministry of Local Government, Ministry of Finance and Economic Planning, NGOs, National Farmer Association, and other representative institutions. PCC would be an advisory body under MOFA's leadership. The Committee would meet twice a year to: (i) discuss the semi-annual progress report and the workplans for the next 12 months; (ii) solve coordination problems which may arise; and (iii) advise PU on general implementation issues. The maintenance of PCC has been agreed upon during negotiations (para 7.4).  AGRICULTURAL SECTOR INVESTMENT PROJECr PROJECr MANAGEMENT ORGANIZATION CHART MIISTY OF F[OOD AND AGRICULTURE PROJECT UNff ~~~~~~~~4 Ch PU ACCRA | PUTSI AM DISTRICT AUTHORITIES ASSOCIATIONS VILLAGE GROUPS l  - 27 - rject Unit 5.4 PU is part of MOFA and would be responsible for implementing the project. PU was established in November 1992 as a unit within MOFA, building upon the experience of the existing Project Coordinating Unit (PCU), which has received intensive training and gained significant experience in project management and IDA procedures. PU's activities would be monitored by PPMED. 5.5 PU Tasks. PU's tasks would be to: (i) or ganize and control project operations; (ii) elaborate work plans and cash flow projections; (iii) manage project funds; (iv) organize project accounting in cooperation with MOFA's PCU; (v) elaborate and review standard project documents and procedures; (vi) manage project information and promotion campaigns to local communities; (vii) review subproject proposals; (viii) organize and report on Subproject Approval Committee meetings (para 5.9); (ix) review qualifications and evaluate supporting agencies and consultants; (x) introduce a subproject cost data monitoring system; and (xi) report regularly to MOFA, and semi-annually to PCC and IDA. 5.6 PU Staffing and Management. The Project Director would staff the PU wi local consultants under fixed-term employment contracts. Some staff would have experience with World Bank project management procedures. PU's full-time staff would be limited to ten professionals experienced in providing services to the rural sector and ten support staff. PU would hire other skills whenever needed, including short-term irrigation and rural road engineers, and market development and management specialists. PU would also employ influential farmers as part-time project facilitators in each region to assist the local communities with subproject identification, application procedures, and to solve implementation problems. In every district, the farmer facilitators would explain the project objectives, eligibility criteria, and application procedures to the District Assembly, associations, and community leaders. PU accounting and IDA disbursement management would be done by MOFA's PCU accountant, until PU's workload justifies the recruitment of a full-time PU accountant. The maintenance of PU during the proiect period has been areed during negotiations (para 7.4). 5.7 Logistics and Overheads. PU would establish offices in Accra and Tamale. The head of the Tamale office would report to Accra, but would have the same regional subproject approval powers as the head of the Accra office. PU would control its overheads, which should not exceed 10 percent of project disbursements after the first project year. This percentage sbould decrease to 5 percent after three years. 5.8 Relationship with the Ministry of Food and Agriculture. MOFA would delegate responsibility for project implementation and management to PU, including authority to approve and implement subprojects, to recruit auditors and to monitor and assess project implementation and impact through PPMED. The Minister would appoint the Project Director and preside over the PCC meetings. PU would carry out the project in accordance with its terms of reference and the Implementation Manual. It would report regularly to the Ministry. To preserve PU's management flexibility, the Director would be responsible for selecting and managing all PU fixed-term staff and consultants. Employment conditions, staff numbers, qualifications and terms of reference of key staff would be agreed on a yearly basis with the Minister and IDA. The respective roles of the Ministry and PU are contained in a Memorandum of Understanding signed by both parties and acceptable to IDA.  - 28 - Subgroject Approval Committee (SPAC) 5.9 SPAC's main task would be to help PU in the evaluation of subproject proposals estimated to cost more than US$75,000. SPAC would decide upon eligibility subprojects and their beneficiaries. SPAC would provide a second and decisive opinion on the acceptability of subprojects in case PU has reservations concerning subproject feasibility. The committees would be composed of, at maximum, nine experts appointed by MOFA, MLG, MFEP, DFR, Ministry of Works and Housing, representatives of the technical support agencies, and representatives of farmer and traders' organizations (National Farmers Association, women's associations). At least two committee members would be women. PU would propose the composition of each SPAC to the Minister and the regional Deputy Ministers of Food and Agriculture. PU would convene the committee meetings. Each subproject approval would require a simple majority of the committee members present at the meeting to be in favor of the project. The maintenance of Subproiect Approval Committees was agreed during negiotiations (para 7.4). B. Eligibilitv Criteria 5.10 The eligibility criteria define the type of subprojects the credit would finance, the local communities which would benefit from project funds, and the eligibility of technical support agencies. The eligibility criteria for subprojects. beneficiaries and technical support agencies were agred during negotiations (para 7.5). Subprjed Eligibility 5.11 Eligible subprojects would be: (i) small-scale water schemes; (ii) construction and rehabilitation of market infrastructure; (iii) construction and rehabilitation of roads to improve access to markets; and (iv) small-scale storage and food processing infrastructure. Projects involving resettlement would not be eligible for financing. The environmental impact of subprojects would be studied, and environmental safeguards built into subproject design. 5.12 Small-Scale Water Schemes. The project would finance small-scale water schemes if: (i) schemes are managed and maintained by village community groups or associations; (ii) beneficiaries provide evidence of usufruct rights on the land; (iii) the subproject is technically feasible and construction costs and equipment purchases do not exceed US$150,000; (iv) costs per irrigable hectare do not exceed US$5,000; (v) fuiancial rates of return to subprojects exceed 10 percent; (vi) the subproject does not create environmental hazards or water shortages elsewhere; and (vii) beneficiaries have an assistance agreement with a technical support agency. 5.13 Markets. The project would finance market infrastructure if: (i) district authorities, trader associations and local communities reach agreement on market design, layout, cost sharing, responsibilities for maintenance, cost recovery methods, and user fees; (ii) market design includes sufficient toilets, water supply and refuse collection; and (iii) except for regional wholesale markets, subproject costs do not exceed US$150,000.  - 29 - 5.14 Road. The project would finance roads to improve access to markets if: (i) district authorities and beneficiary communities propose an acceptable maintenance plan; (ii) the investments are technically feasible at pre-established construction standards and a maximum cost of US$10,000 per kcm; (iii) subproject costs do not exceed US$150,000; and (iv) the roads complement the feeder roads system targeted for development by the Department of Feeder Roads (DFR) and are situated in high and medium priority areas designated by DFR, or PU is satisfied as to the economic viability of the proposed road. 5.15 Storage and Agro-Prgcessing Units. The project would finance storage and agro- processing infrastructure and equipment if: (i) there is sufficient primary produce and market opportunity; (ii) storage units are technically feasible at a capacity below 25 percent of storable community production; (iii) there is agreement amongst beneficiaries on procedures for access by members and on user fees for cost recovery; (iv) the cost per unit does not exceed US$50,000; (v) financial rates of return to investment exceed 10 percent; and (vi) associations have an assistance agreement with a technical support agency which has agreed to assist the beneficiaries with cost recovery. EgBdilt ofBeneSlcdares 5.16 The project beneficiaries would be local communities, not individuals. Provision of project funds to associations would follow stricter eligibility criteria than provision of project funds to village community organizations or District Assemblies. (a) District Assemblies would be eligible if their budget includes provision for the proposed subproject and if they provide sufficient evidence that adequate financial provision has been made for the beneficiary contribution; (b) Village Community Organizations would be eligible if they demonstrate: (i) common activities; (ii) documentation on the way the community is organized (eaders, membership records, etc.); and (iii) the capacity to provide their matching contribution and to manage and maintain the proposed subproject; (c) Associations would be eligible if: (i) the District Executive Committee confims that members form a group with a common agricultural purpose and provides facts supporting this confimation; (ii) the association shows that it has had business activities for more than one year and could deliver the promised contribution; (iii) it has a minimum of 10 members with a clearly identified leader subject to membership control and of which the chairperson, the secretary and the treasurer and at least two other members would enter into an agreement with PU; and (iv) it intends to enter into a technical assistance agreement with an acceptable technical support agency. E1ligibility of Technical Support Agencies 5.17 Support agencies such as NO0s, private consulting firms and institutions would help the communities to implement and manage the projects. The agencies would be eligible for project support if they have the required experience and managerial capacity to assist the  - 30 - communities and enter into a technical assistance agreement 2/ with the beneficiaries. The agencies' eligibility for project support would be reviewed on an annual basis. Every year, PU would review the pre-qualified agencies and invite technical and financial proposals from other potential technical support agencies and pre-qualify them for eligibility in agreement with IDA. PU intends to limit project support for eligible supporting agencies to a maximum 15 percent of the total subproject investment costs, and to 20 percent of the yearly subproject operating costs. The project would also introduce specific contract terms for technical assistance agreements with supporting agencies covering several initiatives. C. Beneficiary Contributions and Financing Percentages 5.18 The beneficiaries would match the project grants in cash, kind, or labor. The local communities' contribution would depend on the type of investment. For market-related infrastructure, both District Assemblies and trader organizations would contribute. Moveable pumps for water schemes and equipment for agro-processing subprojects would require full cost recovery from the beneficiaries through the technical support agencies. Tbe beneficiary contribuin-n cotL_ _taon (para conriutonsan cst ecvey= methods have been agreed upon during negotain(pa 7.6). 5.19 Subproject Identification. Beneficiaries would identify the subproject and complete the project identification forms in a manner acceptable to PU. To stimulate correct completion of the subproject identification forms, PU would pay a nominal fee for every acceptable subproject identification form submitted by the local communities. This would enable communities to remunerate skilled people to assist them with completion of the form.10/ The fee would only be paid upon signature of the financing agreement with the beneficiaries. 5.20 Feasibililt Stdej. When the subproject cost estimate exceeds US$25,000, a feasibility study would need to be done. The beneficiaries would contribute 5 percent of the feasibility study costs. 5.21 Investments. Several contribution percentages would apply: (a) Water Schemes. Beneficiary contributions towards small-scale water schemes would be at least 10 percent of the subproject cost, plus full operation and maintenance costs. For moveable pumps, the association would provide PU with an initial outlay of 25 percent of the pumping equipment costs. Subsequently, the association would contribute 25 percent per year (including interest) during the next three years. 2/ More detailed eligibility criteria and a sample technical assistance agreement are in the Implementation Manual. 10/ The completion of the idenification fonn is very important for subprojects below US$25,000, as no feasibility study is requited for such small subprojects. The identification forn is the basic information PU will use to evaluate and eventually approve the subproject.  - 31 - (b) Market Infrastructure. Beneficiary contributions (District Assembly and traders) for market infrastructure would be 25 percent. (c) Access PR,oa. Beneficiaries would contribute at least 10 percent. (d) Agro-processing and Storage Units. Beneficiaries would contribute 10 percent towards the infrastructure costs, and initially 25 percent plus 25 percent per year (including interest) during the three following years for the equipment. 5.22 Technical Assistance. The beneficiaries would pay 15 percent of the technical assistance and training costs of the technical support agencies. Beneficiary Contributions Investmnt Categories Beneficiary Contribution Comments Small-scale water schemes 10% For moveable pumps: 25% l _______________________ ______________________ plus 75% over 3 years Markets 25% Access roads 10% Storage and processing 10% for infrastructure For equipment: 25% plus 75% over three years Feasibility studies 5% Technical assistance from 15% technical support agencies I_ I 5.23 Cost Recovery. A distinction has to be made between investments which would benefit the whole community such as markets, roads, village storage, etc. and investments which would benefit selected groups in the community (some types of irrigation and agro-processing). Cost recovery from projects benefitting whole communities would be done indireely by the District Assembly through user fees, taxes, market tolls, and other forms of taxation. For small water schemes and agro-processing projects which generate income and benefit selective groups of people, costs of assets which have to replaced in a relatively short period and which have an income-generating purpose, such as moveable pumps and processing equipment, would be recovered directly from the beneficiaries. Such cost recovery would enhance the credit- worthiness of the beneficiaries. 5.24 Costs from income-generating subprojects would be recovered as follows: the beneficiaries would pay 25 percent of the equipment costs up-front to PU. Seventy-five percent plus interest woutd be paid by die beneficiaries over three years. Beneficiaries would be charged an interest rate equal to at least the interest rate for term deposits at the Agricultural Development Bank plus 5 percent. Fifty percent of the amount recovered would be paid to PU, to expand PU's  - 32 - financing base. Ten percent would go to the District Assembly and 40 percent would be used to finance social, economic, and infrastructure operations beneficial to the local community, and agreed upon with the community. The technical support agencies would be contractually obliged to ensure that such yearly (or monthly) contributions take place and, to this end, would train associations in cost recovery through user fees. D. The SubproJeet Processing Cycle 5.25 The normal subproject processing cycle would consist of five steps. The potential beneficiaries would: (i) identify subprojects; (ii) present the subprojects to the District Executive Committee for evaluation; (iii) present the subproject to PU (or Subproject Approval Committee) for approval; (iv) carry out the feasibility studies; and (v) implement the subproject. Technical support agencies could assist the beneficiaries during the process. In general, PU would approve financing for all projects below US$75,000. Projects above US$75,000 would be approved by the Subproject Approval Committee. To control and monitor subproject implementation PU would recruit technical supervisors, MOFA would engage a procurement review consultant and monitor the project through PPMED, and IDA would appoint a resident project supervisor at the Bank's Resident Mission. Tbe subprQiect processing cyl was ard during negotiations (para 7.7). 5.26 Step 1: Subproject Identification. Beneficiaries would need to call on qualified community members to write the initial subproject identification description. Consultants, local engineers and NGOs could help complete the specially-designed subproject identification forms. The forms would contain information on the promoters, the land tenure status, estimated subproject costs and the proposed contribution of the beneficiaries. The Implementation Manual contains sample subproject identification forms. 5.27 Step 2: Evaluation by the District Executive Committee (DEC). Associations and village community organizations would present the subproject to DEC, which would, within 30 days, determine whether or not: (i) the subproject appears technically feasible and belongs to an eligible category; (ii) the beneficiaries are an eligible group and are capable of providing the promised contribution; and (iii) the subproject identification forms reflect the truth. DEC would not bear any financial or civil responsibility in vetting the rroposals, but it would establish a track record with project management. Potential beneficiaries would be able to appeal a negative DEC-decision with PU. 5.28 Step 3: Approval by PU or SPAC. Recommended subprojects of less than US$25,000 would be presented to PU without a detailed feasibility study. PU would approve (or reject) the subprojects after (on-site) verification of data in the project identification form. Subprojects of more than US$25,000 would need a feasibility study. Subprojects of more than US$75,000 would need to be submitted for approval to the Subproject Approval Committee. SPAC would examine project and beneficiary eligibility and agree on the subproject implementation schedule and financing method. For subprojects exceeding US$150,000, approval from the Ministry of Food and Agriculture and IDA would be required.  - 33 - 5.29 Step 4: Feasibility Studies. Projects of more than US$25,000 would require feasibility studies in accordance with project standards. Specialized professionals, consulting firms or NMOs in association with the promoters would carry out these feasibility studies. For each subproject category, PU has designed acceptable feasibility study requirements and guidelines which are contained in the Implementation Manual. The requirements include a list of environmental impact mitigation measures. PU would review the feasibility studies for their accuracy, feasibility and subproject eligibility. Subgriect Processing Cysle Stp I--SUBPROJECT IDENTWICATION Beneficiaries idendfy subproject with help, of consultants, NGOs Output. identification form to District Executive Committee ,-tep2 EVA QUATION.BY DISTRJCT EXECUTIVE COMM0 1EE Output: a fetter rom DEC to PU copied to the Regional Deputy .Ministe of Agriculture' Step 3 APP>OVAL BY'PU or SPAC Itf subproect < $75,000: PU approves If subproject > $75,000: SPAC approves JutpUt: letter of approval from PU to beneficiaries recommending the feasiblity studies, or including a proposal for a financing agreement ' "S, tep 4 f FEASILflY S (for projects > $25,0W) Reruiitmeot of consultant, firm or NGO Co-ntrac approval by PU Output: feasibility/detailed engineering study to PU Step 5 SUPROJECT FINANCING AGREEMENT AND IMPLEMENTATION PU and beneficiaries draw up agreement on contributions, financing and implenientation schedules, procurement, disbursement, cost -recovery, and supervision.  - 34 - 5.30 Step 5: Subproject Implenentation. Having obtained PU's approval, the beneficiaries would start implementation in accordance with a financial assistance agreement. The agreement would describe the rights and obligations of all partners involved, cost recovery requirements, procurement methods, and include a subproject implementation schedule.llJ PU would help the beneficiaries with project implementation by financing the services of technical support agencies. Terms and conditions of financial assistanc agreements were agreed during negotiations (para 7.8). E. Training 5.31 The project would finance training in subproject management provided by technical support agencies. The training objectives, methods, and costs would be described in the technical assistance agreement between TSAs and beneficiaries. PU would not carry out any other training. PPMED would monitor training by TSAs. F. Project Implementation 5.32 PU's plan of action for the first year of the project has been drawn up and is attached to the implementation schedule in Annex 3. 5.33 Project preparation, financed by a Project Preparation Facility of US$1.5 million, focussed on: (i) staffing and equipping PU; (ii) launching pilot operations; (iii) refining eligibility criteria; (iv) pre-qualifying technical support agencies and consultants; (v) carrying out feasibility studies of some important subprojects; (vi) identifying members for the SPACs; (vii) preparing implementation documents; and (viii) preparing information campaigns. 5.34 Following appraisal, the project was launched by PU through: (i) workshops with NGOs, consultants, potential beneficiaries and District Authorities; (ii) pilot projects in nine regions; and (iii) identification and recruitment of suitable staff. 5.35 After project effectiveness, full implementation would require PU to: (i) organize regional launching workshops; (ii) establish a second regional office (Tamale); (iii) set up both SPACs; and (iv) launch project information campaigns in selected regions. In the meantime, subproject implementation would gain momentum. Although the project would operate in the whole country, PU would try to concentrate activities in the districts and regions which are better prepared. Progress would depend on the efficiency of the project facilitators, the relationships between PU and the technical support agencies and the capacity of the district authorities to prepare and approve subprojects. 5.36 Before effectiveness, the Government and IDA would organize a review to fine-tune the main implementation policies and to draw up an action plan to improve implementation. Annual project implementation review meetings would take place between the PCC, PU and IDA. These meetings would discuss past achievements and targets for the next twelve Il/ A sample fuuc agreemet is presented in the Implementation Manual.  - 35 - months. If any project components or regions are not performing, funds would be reallocated to components with demonstrated implementation and disbursement capabilities. During the annual reviews, agreement would be reached on the work program for the following year, and changes in implementation would be introduced on the basis of experience. A Mid-Term Review would take place not later than two years from credit effectiveness, and an Implementation Completion Review report would be prepared at the end of the project. G. Monltoring. Evaluation and Project Impact Studies 5.37 PU, PPMED and IDA would monitor project implementation through reports, monitoring studies, audits, supervision missions, and beneficiary assessments. PU would monitor progress of project activities including: (i) feasibility studies; (ii) infrastructure works; (iii) training and extension by technical support agencies; and (iv) subproject costs. PPMED would monitor PU performance and project impact on beneficiaries. IDA would monitor PU performance, compliance with procurement and disbursement procedures, and credit agreement legal covenants. Annex 6 describes the details of the project monitoring scheme. Proiect monitoring actions have been agreed unon during negotiations (para 7.9). 5.38 Monitoring of Project Activities and Costs. PU would be responsible for providing data on subproject implementation. Key monitoring indicators include: (i) information campaigns launched; (ii) subproject requests received; (iii) feasibility studies carried out; (iv) subprojects processed, financed and completed; (v) number of beneficiaries; and (vi) performance of technical support agencies. PU would prepare semi-annual reports covering the key indicators. The objective would be to focus strongly on subproject field implementation. PU would also set up a subproject cost monitoring system to develop standard costing of markets and engineering services. The objectives would be to keep subproject costs in line with private sector investment and engineering costs and to reduce project preparation time and costs. 5.39 Monitoring of Compliance with Procurement Procedures. To monitor the beneficiaries' compliance with procurement procedures, MOFA would engage a consultant who would review the first two subprojects of each category in detail and other subprojects on a sampling basis, in agreement with IDA. The IDA resident project supervisor would monitor compliance with procurement procedures for the first ten subprojects and 10 percent of the procurement later on. Financial auditors would also verify PU's compliance with procurement procedures. 5.40 Financial Monitoring. Annually, financial auditors would produce a full audit report on the project's financial management, including an audit of the Special Account, SOEs, and cost recovery. The IDA resident project supervisor would review SOEs on a continuous basis. 5.41 Monitoring of PU Performance. PPMED and IDA would monitor PU performance through key monitoring indicators and management efficiency indicators. These would aim to control PU's management and consulting costs and to avoid subproject preparation and administrative costs reaching unacceptable levels. Since the project objective would be to  - 36 - assist local communities, PU and subproject preparation costs should be no more than 25 percent of total investments (including preparation costs). 5.42 Monitoring of Project Impact on Beneficiaries. PPMED would monitor project impact through data collection and beneficiary assessment. (a) Data Collection. In order to evaluate whether the project reaches its objectives, PPMED would introduce a systematic data collection system on project impact. PPMED's regional units would collect baseline data on all subprojects. Before the investment takes place PPMED would make a report on agricultural practices, yields, income, market access, transport, community organization, processing and other relevant information such as the degree of poverty of the beneficiaries. During the project implementation period, PPMED would carry out several visits per year to the subproject a-,as to study changes in key variables and to evaluate impact on project enjectives, in particular the flow of financial and technical resources for infrastructure into rural areas, agricultural productivity, access to markets, the community's management capacities, and poverty reduction. (b) Beneficiary Assessment. PPMED would also consult the intended beneficiaries and incorporate their views in every aspect of project implementation. Beneficiary assessment uses qualitative research tools such as open-ended interviewing of a representative sanple of beneficiaries (focus group or one-on-one) as well as participant-observation techniques involving residence in a community of intended beneficiaries for a few weeks. PPMED would also try to obtain information on: (i) the quality of the investments/ works and the technical assistance to local communities; (ii) the maintenance carried out by the communities; and (iii) the operational problems which may arise during implementation. Costs of PPMED monitoring would not exceed 1.5 percent of investment costs. H. Project Supervision 5.43 IDA supervision would be closely linked to the project implementation schedule. IDA supervision would include: (i) continuous supervision and implementation assistance through the resident project supervisor at the Bank's Resident Mission; (ii) regular supervision and monitoring of key events through supervision missions; and (iii) yearly reviews of the work program, the project launching workshop, mid-term review and project completion review. The Government would participate in supervision through semi-annual progress reports and participation in supervision and review missions. Details are described in Annex 5 and in the Implementation Manual. 5.44 Regular Supervision. IDA would appoint a resident project supervisor at the Bank's Resident Mission in Ghana to supervise and assist with project implementation. Regular supervision would include review of progress reports, procurement, disbursement, and implemer.tation assistance. Project supervision missions would focus on physical  -37 - implementation, management performance and compliance with legal covenants. The key areas to supervise would be: (i) physical implementation: the physical project targets such as number and type of projects; number and type of beneficiaries; involvement of local communities; implementation rate; regional investment spread; (ii) management performance: project response to local demands; subproject lead times; procurement, disbursement and accounting practices; subproject cost control; and (iii) compliance with legal covenants. 5.45 The Project Launching Workshop would take place as soon as the conditions for effectiveness are about to be met. The workshop would focus on procurement guidelines, disbursement methods, management of the Special Account, audit requirements, project supervision and reviews, communications with IDA, project implementation, and other subjects of particular interest to the Borrower and PU. The First Year Implementation Reviw and subsequent yearly reviews would examine the overall and detailed performance of the project and, if necessary, make joint recommendations and take steps to improve project performance. The Mid-Term Review (MTR) would evaluate in-depth the physical and financial project implementation, and in particular its development impact. To this extent, the Government, and in particular PPMED, would prepare a mid-term report on the project to facilitate this review. If necessary, the MTR may introduce major design changes into the project. Terms of Reference for the MTR would be drawn up during project implementation at least three months before the time of the review and would be agreed upon with MOFA. The preMaration and organization of the Mid-Term Review not later than April 30. 1996 was agreed during negotiations (para 7.10). 5.46 An Implementation Completion Review (ICR) report summarizing the achievements of the project and its impact vis-a-vis its objectives would be submitted by the Government to IDA not later than six months after the completion of the project. The ICR would be carried out by the Government and IDA and would cover a full ex-post evaluation of the project. 5.47 The Government would: (i) participate in supervision through PU's half-yearly progress reports; (ii) assist all supervision missions by providing information on the project and organizing wrap-up meetings; and (iii) participate in all reviews: Project Launching Workshop, First Year Implementation Review, Mid-Term Review and Implementation Completion Review. VI. PRO.ECT JUSTIFICATION, BENEFlTS AND RISKS A. Project Justification 6.1 Growth in agricultural output over the last decade has been constrained by a low level of investment in the country's rural infrastructure system. Small-scale irrigation, necessary for increasing productivity of small producers, is underdeveloped. Over 80 percent of the rural market infrastructure is poorly constructed and lacks basic sanitation and health facilities. Access roads, which provide the critical linkages between rural markets in major production zones, remain inadequate. These constraints result in inefficient distribution and marketing of between 50 - 80 percent of the rural agricultural output, and losses to producers  - 38 - due to high post-harvest deterioration. They also directly affect the reliability of market information, leading to distortions in market prices and uncertainty in supply. 6.2 The proposed project would directly address these constraints and would have an impact on the rural economy through: (i) higher productivity growth in food crops through increased cropping intensity and more efficient use of available resources; (ii) reduction in post-harvest crop losses through improved storage and processing; (iii) increase in both product quality and shelf life; and (iv) a more efficient rural product distribution and marketing system. The project would increase rural income and employment through expanded rural investments and the direct participation and empowerment of local communities. B. Deteridination of Benefits and Costs ]Phv-sical Outpu 6.3 Small-Scale Irrigation. The project would finance the construction or rehabilitation of a total of 500 ha of small-scale irrigation schemes with direct impact on more than 2,000 farm households. Small-scale irrigation represents the most cost-effective way to increase production of mixed cropping, high-valued vegetables, and rice, and raise the incomes of subsistence producers. Small-scale irrigation schemes would contribute an additional 1,350 tons of rice (paddy), 920 tons of high-valued vegetables (tomatoes, peppers, leafy vegetables), and 180 tons of cereals and groundnuts annually at full project implementation. 6.4 Markets. The project would support rehabilitation of 110 medium-size markets comprising covered sheds, market stalls, locked storage, warehouses, and additional facilities for private sector investment in restaurants, day care and clinics. The rehabilitation of rural markets would improve sanitary conditions and reduce environmental hazards through better drainage and improved water supply and sanitation. The development of rural markets would directly impact over 5.7 million buyers/sellers in the rural economy each year, and would generate an estimated 4 billion cedis (US$6.6 million equivalent) in annual revenues for participating local communities. 6.5 Access Roads. The project would finance construction and rehabilitation of 375 km of access roads linking the high agricultural production zones to major marketing centers and the existing feeder road network. Access roads would complement the existing feeder roads targeted for improvement under the National Feeder Roads Rehabilitation and Maintenance Project. About 11,000 tons of incremental output per year would be produced from within the zone of influence of the roads by an estimated 21,000 households. Direct benefits to farmers are estimated at 1.3 billion cedis (US$2 million equivalent) per annum from the fourth year. Benefits to transporters (reduced cost of vehicle operations and maintenance, and increased volume of marketed surplus in accessible areas) is estimated at 148 million cedis (US$0.25 million equivalent) by the fourth year. 6.6 Agro-Processing. The project would finance investment in 20 community-based food processing activities (shea-nut, cassava and fish) on a pilot basis. The use of motorized processors would reduce cassava processing time by 99 percent, and increase recovery rates to 28 percent (compared to a recovery rate of 15 percent using traditional methods).  - 39 - Incremental processing capacity from investment in small-scale agro-processing is estimated at 965 tons of cassava, 1,210 tons of shea nut, and 250 tons of fish, annually. Benefits 6.7 Direct project benefits would be: (i) incremental output from farm production; (ii) direct revenue generation for communities; (iii) increase volume of marketed surplus and reduction in rural transportation costs; and (iv) higher product quality, and higher prices for processed agricultural commodities. Incremental output from small-scale irrigation is estimated at over 2,400 tons per annum, with an estimated gross value of 260 million cedis (US$0.4 million equivalent) at 1993 prices. Rental fees from the rehabilitation and management of a prototype rural market is used as a measure of traders' willingness to pay for improved markets and trading conditions and therefore represent the economic benefits accruing to local communities. Incremental revenue from a prototype rural market is estimated at 13 million cedis (US$ 21,000 equivalent) for the first year, and increasing to 19 million cedis (US$ 30,000 equivalent) by the fifth year of project implementation. Total annual incremental revenues from markets are estimated at 1.4 billion cedis (US$2.5 million equivalent) at full implementation. Marketed surplus from the zones of influence of 375 km of access roads is estimated at 4,000 tons, increasing to 11,000 tons by the fourth year. The incremental benefits at village level from agro-processing are estimated at over 3.3 million cedis per year by the third year of the investment and would reach as high as 11 million cedis per year. 6.8 All costs are estimated using post-1993 budget prices for goods, equipment, labor and other inputs obtained during appraisal. Costs for small-scale irrigation are based on GIDA estimates for reservoir rehabilitation, low-lift pumps and construction of tube-wells. The costs of markets are based on the average cost of upgrading/improving a prototype market serving a rural community of approximately 5,000 inhabitants and capable of accommodating 1,000 sellers and buyers during the peak weekly market day. Market costs do not include major canalization or construction. Cost data for access roads are from the Department of Feeder Roads, updated during appraisal for a 4-meter wide road. The cost of processing equipment for shea nut, cassava and fish drying are estimates from field data. The rural wage rate is estimated at 1,000 cedis per person-day. All agricultural outputs are conservatively priced at the farm gate, exclusive of transport and handling costs; prices of processed products are ex-processing units. Profitability 6.9 Financial Rate of Return. The financial rate of return is estimated at 19.1 percent and is presented in Annex 2, Table 6. Financial rates of return for individual project components are: small-scale irrigation, 19 percent; rural markets, 19.3 percent; and agro-processing, 21.8 percent (Annex 2, Tables 6,9 and 10). 6.10 Economic Analysis. The economic rate of return is estimated at 23.3 percent. Financial prices were converted to economic values by removing distortions in market prices of tradables and non-tradables, and adjusting for transfer payments, i.e., net of taxes and duties. Conversion factors used for major inputs and outputs are in Annex 2, Table 11. All  -40 - border prices are appropriately adjusted for domestic transport and handling costs to the project area. The opportunity cost of rural labor is estimated at 750 cedis per day, equivalent to 75 percent of the rural wage rate, which reflects the inclusion of family labor. The resulting economic evaluation reflects potential economy-wide gains from the project. Economic rates of return for individual project components are: small-scale irrigation, 21 percent; rural markets, 22 percent; access roads, 49 percent; and agro-processing, 23 percent. 6.11 Sensitivity Analysis. Sensitivity of the project to changes in costs and benefits was estimated using switching values (Annex 2, Table 8). Small-scale irrigation, rural markets and access roads can accommodate wide fluctuations in benefits and costs. Overall, the project would remain viable at an opportunity cost of capital of 12 percent even with a 27 percent reduction in benefits or a 36 percent increase in costs. Agro-processing is sensitive to a less than 10 percent change in either costs or benefits. Non-quantifiable benefits accruing to beneficiaries are not included in the above analysis. Summary of Flnancial and Economic Analysis Water R al Access Agro- Total Schemes 1- arkets Roads Processing Financial Rate of 19.0 19.3 not 21.8 19.1 Return (%) applicable Economic Rate of 21.5 22.0 48.8 26.0 23.3 Return (%) I NPV Economic 327,746 57,821 84,326 11,036 480,131 (@12% -'000 cedis) Sensitivity: EIRR decreases to 12% if Costs increase by 36.9% 28.0% 94.0% 7.7% 36.2 or Benefits decrease by 26.9% 21.9% 48.5 7.1 26.6 Developmental Impact 6.12 The developmental impact of the project would include: (i) strengthening capacity at the community level to implement projects; (ii) reduction of rural poverty through increased rural investment, growth and employment; (iii) transmission of better market information through improved rural markets and marketing systems; (iv) reduced food price fluctuations; and (v) improved rural hygiene. Because the project is demand-driven, it would respond directly to the needs of rural communities and would enhance their capacity to identify, prepare and implement development projects.  - 41- F1scal Impact 6.13 The project would complement direct budgetary support from the central government to local communities. Additionally, revenue from market rental would provide operational support to local governments as well as resources for investment in development-related activities. The ability to mobilize community-based res&.rces for local development would lead to a greater sense of local ownership ond the sustainability of the project. C. Risk 6.14 There are four main risks: institutional capacity, complexity of subproject approval process, misuse of funds, and lack of capacity to prepare subprojects. The first and main risk lies with the capacity of Government and PU to cope with so many small subprojects. Although the Government is already implementing two similar projects in the Education and Social Dimension of Adjustment (PAMSCAD) sectors, this project would be the first that would decentralize authority to disburse funds. A second risk is that the subproject approval process may become overly complex. Experience with the PAMSCAD project shows that bureaucracies (including donors) tend to introduce additional procedures into the approval, procurement and disbursement process, making these processes cumbersome and overly complex. A third risk is that funds might be misused in several ways: (i) as most subprojects would be highly visible, subprojects may be approved on the basis of political considerations other than on merit; (ii) feasibility studies and supervision of certain projects may be of lower quality than is required; and (iii) advances may be used for other purposes. The fourth risk is the lack of capacity to prepare subprojects. In normal circumstances, qualified consultants do not reside in the rural areas, and it may be difficult for the communities to recruit such consultants at a fair price. 6.15 Project design contains ' ieasures to mitigate these risks. Institutional capacity: PU would consist of experienced private and public sector staff including Agricultural Services Rehabilitation Project staff. The Credit Agreement would clearly stipulate the managerial independence of PU from the Ministry. PU would also be equipped with efficient and proven computer technology. Procedural complexity: to avoid excessive project complexity and paperwork, PU would hire influential farmers as part-time project facilitators and a participant observer. IDA would appoint a resident project supervisor. Project implementation problems would be solved through human contact instead of excessive document circulation, and a detailed Implementation Manual would be applied. Use of funds: misuse of funds would be limited through the adherence to eligibility criteria, and strict and regular control by independent auditors and the resident project supervisor. Moreover, sanctions against misuse of funds would be announced publicly and communities would, at least, lose eligibility for project support. Project studies. PU would provide a list of capable consultants to communities, as well as a list of reliable sources of technical support. A second-best alternative solution could be PU hiring a local consulting firm on a medium-term basis to assist communities.  - 42 - D. Environmental Effects 6.16 Construction of rural infrastructure such as small-scale irrigation works, access roads, markets, storage facilities, etc. could create localized environmental problems (erosion, contamination, water depletion, waste disposal, etc.). For example, small-scale water reservoirs (500 ha maximum under the project) would increase livestock drinking water supply but may create health and downstream water shortage problems. Construction of access roads in environmentally sensitive areas may cause lasting environmental damage. 6.17 Environmental aspects of all project activities would be carefully reviewed during subproject feasibility studies and appropriate mitigation measures designed for implementation under the project (see Annex 8: Environmental Problems and Mitigation Measures). Possible measures could include construction standards for access roads, local health authorities review of water reservoirs, sanitation standards for markets, etc. In addition, a water management study is to be undertaken within the Ghana Environmental Resource Management project to address the broader environmental aspects of water usage and disposal in the country. E. Impact on Women 6.18 This project could have a significant positive impact on women. First, market improvements mainly concern women's activities and income. Market modernization would supply market women with water, toilets and shelter, improving their health and the sanitary conditions of their children. Second, project support to food processing would mainly increase their income and self-reliance. Third, irrigation would benefit women's income considerably, as dry-season agriculture concerns vegetables, mainly produced and sold by women. Fourth, women transport most of the production by headload. Through market access roads, trucks would gain access to the production areas and women would be able to spend their time on more productive activities such as education, child care and agriculture than on transporting produce by headload. F. Poverty Alleviation 6.19 The recent Bank report "Ghana: 2000 and Beyond" describes how poverty in Ghana is predominantly a rural phenomenon. Over 43 percent of rural inhabitants live below the poverty line. The urban poor, much less numerous, are in general less poor than their rural counterparts. The proposed project would address rural poverty. The beneficiaries of small water schemes and agro-processing activities would be mainly farmers in isolated and poor rural areas of the savanna region, which accounts for 35 percent of the extreme poor compared with a population share of 12 percent. The project would also increase food security through irrigation and agro-processing among vulnerable groups, especially women and children, in the drought-stricken rural areas.  - 43 - VII. ASSURANCES ANI) RECQMMENDATION During negotiations, agreements were reached on the following . 7.1 Appointment of a procurement auditing consultant by August 31, 1994 (para 4.16 and Section 3.09 of the DCA); 7.2 Establishment and maintenance of a Project Account by MOPA (para 4.19 and Section 3.02 of the DCA); 7.3 Project procurement, disbursement, and financial auditing and reporting requirements, in particular the provision of semi-annual reports (para 4.22 and Sections 3.06, 3.09, 4.01 of the DCA); 7.4 Maintenance of the i sect 1Piit, the Subproject Approval Committee and the Project Coordinating CommitL -as 5.3, 5.6, 5.9 and Section 3.07 of the DCA); 7.5 Eligibility criteria for subprojects, beneficiaries, and technical support agencies (para 5.10 and Schedule 5, Parts A, B and C of the Development Credit Agreement - DCA); 7.6 Beneficiary contributions and cost recovery methods (para 5.18 and Schedule 5, Parts G and H of the DCA); 7.7 The subproject processing cycle (para 5.25 and Schedule 5, Part F of the DCA ); and 7.8 Terms and conditions of financial assistance agreements (para 5.30 and Section 3.04 and Schedule 5, Part I of the DCA); 7.9 Monitoring of project activities, costs, PU performance, and project impact on beneficiaries (para 5.37 and Section 3.07 of the DCA); and 7.10 Preparation and organization of a Mid-Term Review (para 5.45 and Section 3.08 of the DCA). Condition for Effectlveness 7.11 The deposit by the Government of US$50,000 equivalent into a project account (para 4.19 and Sectior. 5.01 of the DCA). Recommendation: 7.12 Based on the above, the proposed project is suitable for an IDA Credit of SDR 15.3 million (US$21.5 million equivalent) to the Government of the Republic of Ghana.  -44 - Annex I Page 1 of 7 GHANA AGRICULTURAL SECTOR INVESTMENT PROJECT TYPICAL INVESTMENTS 1. This annex provides: (t) a description of typical subprojects; (ii) project costs and benefits; and ffii) calculation of the financial and economic rate of returns. A. Small-Scale Water Schemes Decrp