Ghana - Agricultural Services Rehabilitation Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 16857 PERFORMANCE AUDIT REPORT GHANA AGRICULTURAL SERVICES REHABILITATION PROJECT (CREDIT 1801-GII) July 1, 1997 Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. Currency Equivalents (annual averages): Name of Currency: Cedi 1986 US$1.00 151 (appraisal) 1987-93 US$1.00 422 (average) 1994 US$1.00 915 (completion) Abbreviations and Acronyms ADC Agricultural Development Corporation (South Korea) AfDB African Development Bank AESD Agricultural Engineering Services Department APCC Agricultural Policy Coordination Committee ASRP Agricultural Services Rehabilitation Project CRI Crops Research Institute ERP Economic Recovery Program FAO Food and Agriculture Organization of the United Nations GIDA Ghana Irrigation Development Authority GOG Government of Ghana ICR Implementation Completion Report KfW Kreditanstalt fur Wiedaraufbau MEC Monitoring, Evaluation, and Coordination Division MOFA Ministry of Food and Agriculture MTE Midterm Evaluation MTR Midterm Review NGO Nongovernmental organization O&M Operation and Maintenance OED Operations Evaluation Department PAR Performance Audit Report PCU Project Coordination Unit PPMED Policy, Planning, Monitoring, and Evaluation Department SARI Savannah Agricultural Research Institute SDR Special Drawing Rights SOE State Operating Enterprise T&V Training and visit (Extension System) TA Technical Assistance UNDP United Nations Development Programme URADEP Upper Region Agricultural Development Project VSD Veterinary Services Department VORADEP Volta Region Agricultural Development Project Fiscal Year: January 1 to December 31 Weights and Measures: Metric System Director-General, Operations Evaluation : Mr. Robert Picciotto Director, Operations Evaluation Department : Ms. Elizabeth McAllister Division Chief, Agriculture and Human Development Division : Mr. Roger Slade Task Manager : Mr. E. B. Rice FOR OFFICIAL USE ONLY The World Bank Washington, D.C. 20433 U.S.A. Office of the Director-General Operations Evaluation July 1, 1997 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT Subject: Performance Audit Report on Ghana Agricultural Services Rehabilitation Project (Credit 1801-GH) Attached is the Performance Audit Report on Ghana, Agricultural Services Rehabilitation project (ASRP), prepared by the Operations Evaluation Department. ASRP was supported by a credit for US$17 million equivalent approved in FY87. The credit was closed in FY94, after one extension of the closing date, and a small undisbursed balance was canceled. The project had parallel financing from the African Development Bank and German KfW, which provided specialized commodity imports to support the reform program, and from UNDP, which financed technical assistance for planning. The main objective of ASRP was to initiate a self sustaining process of rehabilitation in the Ministry of Food and Agriculture (MOFA) so that it could give more effective support to the agricultural sector. In particular, it aimed at strengthening the institutional capacity of MOFA to formulate and implement agricultural policies, plans and programs, and extending agricultural policy reforms already underway that would eliminate subsidies and divest the ministry of its commercial services, including fertilizer marketing, tractor hire services and veterinary services. The project's investment components were aimed at restoring the capacity of four of the ministry's primary services: planning, research, extension and irrigation. The project was also to finance preparation of a series of follow-on Bank operations in agriculture. Despite a slow start, Bank disbursements lagged only a little behind the expected profile and most project components were completed on schedule. There were serious shortfalls in physical progress toward the research and irrigation area targets, and the planning component produced the expected reports but mostly without involving ministry staff. In contrast, the extension component got off to a quick start and piloted training and visit (T&V) methods that were later expanded to the national level under one of the follow-on projects. Government embraced the policy agenda and maintained the agreed schedule for staging reforms. An inter-ministerial coordinating committee was formed, which attracted high-level attention and achieved a consensus on key sector issues. The results for the capacity-building exercises highlighted in the appraisal report were disappointing in three of the four services-especially in the planning department. The project had only a minor impact on the research department through the rice and cotton research programs, and failed to reorient the agenda of the irrigation department to small scale, participatory schemes. The extension component was the only unqualified success, fulfilling the modest role of a pilot for the new national system. Nevertheless, there were impressive gains in terms of overall institution-building, through the formation of the coordinating committee, the preparation of a medium-term sector development strategy, This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. 2 and the generation of five successor projects which would continue the work at the subsector level. The policy reforms aimed at removing subsidies and divesting state enterprises and were carried out as agreed. They were not followed immediately by the projected recovery of private enterprise because the preconditions for successful privatization were not satisfied. The overall outcome of ASRP is marginally satisfactory, a qualified rating reflecting the failure to make significant progress to capacity-building targets specified in the appraisal report. Institutional development itself can be rated modest, reflecting in part the positive effects of the T&V pilot. Both ratings are lower than those given in the ICR, which was less concerned about the lack of progress toward specified institutional targets. Yet, if capacity building is taken seriously ex ante, it must be taken seriously expost. The audit accepts the ICR ratings for sustainability as likely and Bank performance as satisfactory, though there were serious lapses in supervision. Borrower performance is rated satisfactory, giving special weight to government's success in pushing through the tough policy agenda. Lessons that are drawn from this project include: (i) subsidy removal and divestiture opens the way for privatization of services, but other preconditions must be met before the private sector responds; (ii) the unsophisticated technical messages provided through the T&V extension system can have a positive impact on relatively primitive cropping systems; and (iii) Bank indifference during supervision to M&E designs proposed at appraisal can be fatal to their chances of successful implementation. Attachment Contents Principal Ratings and Key Staff Responsible for Project.................................................... 3 Preface..... .................................................... 5 1. Introduction ................................................... 7 2. Project Objectives And Design A. Objectives ................................................... 8 B. Design...................................................... 8 3. Implementation ........................................ ....10 4. Outcomes ..................................................... 12 5. Findings and Lessons A. Findings.................................................... 22 B. Lessons .................................................... 26 Annexes A. Basic Data Sheet ......................... .............. 27 B. Project Costs................................................. 29 C. Information Requirements under ASRP ................................ 31 D. Comments from Ghana's Ministry of Food and Agriculture ......... .......... 33 E. Comments from the Africa Regional Office.. .............................35 Map Agricultural Services Rehabilitation Project (IBRD 26697R)  3 Principal Ratings Principal Ratings Outcome Marginally Satisfactory Sustainability Likely Institutional Development Modest Bank Performance Satisfactory Borrower Performance Satisfactory Key Staff Responsible for Project Director Division Chief Task Manager Appraisal B. Alisbah A.K. Seth J.C. Cole Implementation C. Koch-Weser A.K. Seth, L. Campbell, E. Lim J.M. Joyce T. Haile-Mariam, C.G. Ranade Completion 0. Lafourcade C.C. Cook G. Alibaruho ICR Prepared by: G. Alibaruho  5 Preface This is a Performance Audit Report (PAR) on the Agricultural Services Rehabilitation Project, Ghana, involving a Bank credit in the amount of SDR 13.3 million (US$17.0 million equivalent). The credit was approved on June 30, 1987. It was closed on June 30, 1994 after one extension of the closing date. The small undisbursed balance of about SDR 24,000 was canceled. The PAR is based on the Implementation Completion Report (ICR) prepared by the Africa Regional Office and submitted to the Board on May 12, 1995, Government's subsequent comments on the ICR, the Staff Appraisal Report, the President's Report, the Credit documents, a study of project files, and discussions with Bank staff. An Operations Evaluation Department (OED) mission visited Ghana in May-June 1996 and discussed the effectiveness of the Bank's assistance and project execution with the Ministry of Food and Agriculture (MOFA) and other relevant agencies. MOFA's kind cooperation and valuable assistance in the preparation of this report are gratefully acknowledged. The ICR provides a good account of the operation of the project and the performance of the Bank and project executing authorities, although it makes several factual errors. The discussions in the ICR of capacity building in MOFA's Policy, Planning, Monitoring, and Evaluation Department, and of the privatization campaign, have been expanded in this report. Following standard OED procedures, copies of an initial and a revised draft were sent to the government for comment. The comments received from MOFA on the revised draft are reproduced as annex D. Outstanding differences with the Region are indicated in the text, and the Region's most recent comments comprise annex E.  7 1. Introduction 1.1 In April 1983 the Government of Ghana (GOG) launched a series of major economic reforms identified as the Economic Recovery Program (ERP). In response, the Bank approved several short term credits to provide critically needed imports and immediate rehabilitation works for key sectors, and stepped up its economic and sector work. Bank policy advice and mobilization of donor assistance played a key role in defining the GOG's policy framework that underpinned economic recovery. The GOG's program was then supported by a series of Structural Adjustment Credits. As part of its sector work, the Bank in 1984 undertook an Agriculture Sector Review, which in a report issued the next year outlined the main elements of a strategy to promote agricultural development. 1.2 The strategy signaled a shift of attention back to the neglected small farm sector, after two decades of government investment in state farm and cooperative enterprises, large irrigation schemes and public marketing agencies. The short term emphasis of the new strategy was on improving farmer incentives and the availability of inputs to exploit existing capacity, while rehabilitating road infrastructure and transport. The medium term emphasis was on strengthening the capacity of public sector institutions to plan, coordinate and implement programs within the sector. Physical facilities and working conditions in the then Ministry of Agriculture and its ten regional offices had deteriorated to the point where buildings were dilapidated, office supplies were often unavailable, and most field services were immobilized. 1.3 At the time of the Sector Review, one of two area agricultural development projects financed by the Bank had just been completed (the Upper Region Agricultural Development Project [URADEP]), and another was at midterm (the Volta Region Agricultural Development Project [VORADEP]). These were representative of a large group of donor-supported activities in the rural areas-initiated both before and during ERP-which were either area or service- based but uncoordinated among themselves and designed around individual donor-determined strategies for small-scale farm development. The lack of coherence was most obvious in the proliferation of extension services: the Ministry's specialized departments were each armed with their own extension capability for promoting programs and independent services were created for the area development schemes. The largest extension service belonged to the Ghana Cocoa Marketing Board and was dedicated exclusively to this premier export crop. 8 2. Project Objectives And Design A. Objectives 2.1 The basic objective of the Agricultural Services Rehabilitation Project (ASRP), approved in 1987, was to initiate a self sustaining process of rehabilitation so that the Ministry of Food and Agriculture (MOFA) could give more effective support to agriculture. In particular, it aimed at (i) strengthening the institutional capacity of MOFA to formulate and implement agricultural policies and programs; (ii) extending the agricultural policy reforms already underway, particularly with respect to elimination of subsidies and divestiture of the ministry's commercial services; and (iii) rationalizing and improving efficiency in the delivery of those services which would remain in the public domain. The Bank's funds of US$17 million equivalent (SDR13.3 million) were complemented by substantial parallel financing from the African Development Bank (AfDB)/Fund (US$25 million) and German KfW (US$8.5 million) for specialized commodity imports supporting the agricultural policy reforms-in particular, as originally conceived, fertilizer (AfDB) and veterinary supplies (KfW). The Bank's funds were also complemented by a United Nations Development Programme (UNDP) grant of US$1.5 million for technical assistance (TA) to MOFA. B. Design 2.2 The project included a wide array of policy conditions, pilot investment operations and TA activities. On the policy side, government agreed to (i) progressively remove all subsidies on fertilizer over four years; (ii) start charging full costs for drugs administered by the Veterinary Services Department (VSD); and (iii) privatize several services and agencies under the MOFA umbrella, in particular fertilizer trade, tractor hire operations, import and sale of veterinary drugs (and some vaccines), and most if not all of a group of 18 state owned enterprises belonging to the ministry. These substantial steps toward liberalization of markets would complement major reforms already underway in foreign exchange rates and crop prices, all aiming to restore profitability to those agricultural subsectors which could effectively compete in international markets (one of the implicit targets was to shift farm enterprises from nontradable to tradable products). By agreeing to the full set of these politically sensitive measures, government signaled its acceptance of abandoning the command economy put in place during the previous two decades. 2.3 The TA and investment operations were initially aimed at restoring the capacity of three of the primary ministerial services: agricultural research, agricultural extension, and irrigation. The emphasis was on low-cost cropping systems. Research, for example, would focus on the cultivation of dryland cotton in the north and river valley paddy in the center and south, both activities focused on the needs of small-scale farmers. Details of the individual components are discussed in chapter 4 on outcomes. The TA had two other major targets. One was to build a capacity within MOFA to develop its own policies, programs and projects, supported by efficient monitoring and evaluation services. The existing Economic Research and Planning Services unit 1. Agricultural research was not under the control of MOFA, but the Council on Scientific and Industrial Research of the Ministry of Industry, Science and Technology. 9 in the ministry would be remodeled, expanded and upgraded to become the Policy, Planning, Monitoring, and Evaluation Department (PPMED). This component would be financed mostly by the UNDP grant, and carried out by a consultant team selected and administered by the Food and Agriculture Organization (FAO). The second was to finance a series of studies that would lead to subsequent operations for the Bank and other donors. Agricultural credit, research, extension and food security were the four subsectors specified during appraisal. 2.4 From the perspective of most ministry staff, the immediate objective of ASRP was simply the equipment and modernization of their offices: air conditioners and fax machines in public rooms, vehicles for extension and veterinary staff, computers where needed and particularly for PPMED, etc. A Project Coordination Unit (PCU) was established in MOFA, and somewhat later a procurement/accounting specialist (a Ghanaian from VORADEP) was appointed to the unit to help manage the gradual build-up of project-identified goods and services. In 1989 the Bank agreed to allow several other ministry departments to be included under the ASRP umbrella, including in particular crop services and fisheries. In this sense, ASRP provided a pool of funds to facilitate the upgrading of the ministry and to provide staff with acceptable working conditions. 10 3. Implementation 3.1 The IDA Credit was made effective in December 1987, three months later than planned. Operations got off to a slow start. Supervision Reports and the Implementation Completion Report (ICR) attribute initial delays to PCU staff being overwhelmed by the Bank's procurement, disbursement and reporting requirements. But, as stated in government's section of the ICR, the PCU was on a "learning curve" and these problems were eventually overcome. Subsequently, many of the procurement tasks under already approved budgets were turned over to the implementing agencies, and progress accelerated. Some activities moved quickly out in front-extension in four pilot regions gained and never lost the lead-and others lagged behind-the most troublesome to Bank supervision was the small scale pilot irrigation component. But overall progress was reasonable, Bank disbursements lagged only a little behind the expected profile, most project components were completed on schedule, and actual costs at completion of US$56 million were only five percent over target. The next paragraphs describe especially difficult implementation problems. 3.2 Small Scale Pilot Irrigation. This component was intended to upgrade the capacity of the Ghana Irrigation Development Authority (GIDA) and shift its priorities from large, capital intensive schemes that generated losses each year to small schemes costing a maximum of US$5,000/ha with substantial farmer participation in management and cost recovery. It was the most expensive component in the original cost tables, responsible for 29 percent of total baseline projections, and ended with a slightly higher share of actual costs (32 percent, excluding parallel financing; see annex B). 3.3 Three factors explain the delays in carrying out this component. First, the process of selecting a "twin" for GIDA under the TA program took twice the time expected, partly because final approval of the South Korean Agricultural Development Corporation (ADC) was interrupted by persons in government who complained that a Chinese agency had not been short- listed. In fact, the Korean team arrived in late 1988 even before the contract was signed, a risky maneuver that nevertheless allowed them to start mobilizing resources and help catch up with the original schedule. Second, the process of identifying six sites suitable for either pumping or gravity-feed during wet and dry season irrigations took GIDA and its Korean advisors almost two years longer than anticipated. Government insisted the selected sites be representative of all the agroecological zones of the country, a condition that led to field visits to over 100 sites, with identification of 72 of them. This group was eventually cut to 14, of which 12 were fully prepared and reduced again to a short list of 6. The third reason was that GIDA determined that average costs at all short-listed sites were going to exceed the US$5,000 ceiling and, in the view of the Bank (but not of GIDA), undermine the economic prospects of the schemes. The Bank eventually approved two sites, where worked commenced in 1992 three years behind schedule. The two schemes were eventually brought into production in the main wet season of 1994. By that time the Bank had shifted its attention to a completely different low-cost irrigation strategy-that had not been anticipated at appraisal (and is described in chapter 4 under low risk water conservation). 3.4 Cotton Research. This small component (4.5 percent of appraisal costs) was managed by the Savannah Agricultural Research Institute (SARI, formerly the Nyankpala Agricultural Experiment Station). SARI had (and has) no cotton breeding capability, and its research trials 11 were intended only to develop improved small scale, nonmechanized cultivation practices using varieties already imported from C6te d'Ivoire and other West African countries by cotton- ginning companies-which contracted with small scale producers for provision of inputs and delivery of seed cotton. SARI carried out trials at its main station near Tamale and three substations. But it was unable to advance to farmers' fields, ostensibly because the ginneries did not encourage their contract farmers to dedicate portions of their fields to this experimental work. Some observers claim that the SARI officers in charge could have been more aggressive in developing a farmer outreach program. Nevertheless, when ASRP was completed in 1994 the new cotton practices recommended by SARI had still not been tested under field conditions, and the suitability of the recommendations for widespread commercial application had still to be confirmed. 3.5 Privatization of Fertilizer Marketing. A special unit in the Crop Services Department of MOFA was created in 1988 to manage removal of the fertilizer subsidy and privatization. Initially, the unit received large numbers of applications from prospective wholesalers and retailers who wished to enter the trade. But the applicants were ignorant of the probable impact of the removal of the subsidy on fertilizer demand, as well as of the difficulties of handling, stocking and storing fertilizers from one season to the next. Unit staff traveled frequently throughout the country promoting the program and educating the actual and potential applicants. But their efforts were largely frustrated. Within two years practically all the retailers had withdrawn their applications and, on the wholesaling side, only four companies have replaced government and this group is itself dominated by one among them, which controls over 50 percent of all imports. 12 4. Outcomes 4.1 The level of success of the multiplicity of components varies widely, and any overall rating depends on the weighing of the individual parts. The ICR rates the overall outcome as satisfactory, with sustainability as likely, institutional development as substantial and Bank and borrower performance also as satisfactory. The audit does not support all of those ratings, as discussed at the end of this chapter. One reason for the downgrading is the almost complete failure to progress toward one of the prominent goals of the project, which was to create an in- house capacity in PPMED to develop policies, programs and projects. The first section below discusses this component in detail and the subject is referred to again at the end of chapter 5 on lessons. 4.2 PPMED. Measured by the installation of computers, training in word processing and other computer applications, expansion of the statistical data base and a facility for rapid 2 retrieval, this component could be called a success. But the good news ends there. The main failure was the lack of involvement of PPMED professional staff in most of the important planning and monitoring operations that were carried out under the project, the opposite of what had been intended. 4.3 This component must be weighted heavily in any ex post assessment of results, especially of institutional development. It is the leading component of Schedule 2, the Description of the Project, in the Development Credit Agreement (page 14): Part A: Policy Formulation, Planning, Monitoring and Evaluation 1. Reorganization and strengthening of PPMED to enable it to: (i) develop sustainable policies, monitor their impact and take the lead in the development of an agricultural sector investment program including the direction of donor support for the program; (ii)... (emphasis added). 4.4 The outcome was poor on all of these implicit indicators. FAO makes the same point in its Terminal Report for the UNDP project: "the staff of PPMED has been distanced from much of the real policy work of the department"3 and it then concludes: "It is strongly recommended that Government undertake a long-term steady and sustained effort to build PPMED into an institution capable of providing timely and relevant policy advice and analysis pertaining to agriculture and food policies"4 which was the premise of the UNDP operation designed ten years earlier. 2. PPMED also takes pride in having substantially cut the turnaround time for delivering its market price reports. 3. FAO. AG:DP/GHA/86/008 Terminal Report. Strengthening Agricultural Planning in Support ofthe Government's Agricultural Sector Rehabilitation Programme, Ghana, Project Findings and Recommendations. Rome, 1995, page 28. 4. Op. cit., page 30. 13 4.5 The fault can be attributed to several overlapping factors: an extremely weak team fielded by the consulting firm without adequate support from its Washington headquarters or surveillance by Bank supervision; the assumption of the directorship of PPMED by a well- regarded academic, who appears not to have given sufficient priority to staff upgrading, opting rather for contracting out most local assignments to better-trained university consultants; and Bank supervision staff who became increasingly involved in high-level negotiations with government on subsequent reforms and adjustment operations, considerations which created an urgency to get jobs done and crowded out any original intentions to use the project as a training ground for the young professionals of PPMED. The quality of that professional cadre was relatively weak to begin with, a fact recognized by all. But by in-house and overseas short term training the intention was to gradually upgrade those skills. Nevertheless major opportunities were missed to involve the staff in key assignments.5 4.6 The record was not uniformly poor, of course. In fact in the arena of upgrading the statistical data base, by, among other activities, broadening and strengthening the sample survey, the consultant team carried on the good work initiated during an earlier FAO operation. Also, the reports produced by the university-based staff were usually of good quality. Finally, PPMED dispatched staff to take up residence in other departments and all regional offices, a welcome and unprecedented move to strengthen inter-office coordination. 4.7 The signature document produced by PPMED during the project period was the Medium Term Agricultural Development Strategy issued in 1990. This should have been a high profile exercise for staff of the newly created Policy Planning and Analysis Division of PPMED, but it was not involved at all (with the exception of the director). Government recruited Ghanaian and expatriate consultants to carry out most of the work, supported by Bank staff. 4.8 Other joint products of this period were the master plans and/or preparation reports for a long series of projects subsequently approved by the Bank's Board: a Rural Finance Project (1989), National Agricultural Research Project (1991), National Agricultural Extension Project (1992), National Livestock Services Project (1993) and Fisheries Subsector Capacity Building Project (1995). PPMED's Project Planning and Budgets Division was created in 1987, along with the Policy Planning and Analysis Division, with the intention of involving staff in developing projects. But none was involved-by the director or the Bank-in the detailed preparation work of any of these Bank-financed projects. This was not a case of indifference by Bank staff to getting the divisional staff included: it was mostly the result of lack of interest, by the American consultant assigned to that division, and, again, failure of departmental leadership to insist on an inclusive arrangement. 4.9 Another example of this almost universal exclusion of PPMED staff from priority activities was the experience of the Monitoring, Evaluation and Coordination Division (MEC) with preparation of a Midterm Evaluation Report (MTE), issued in 1991. In this case, MEC staff 5. In Annex D the government takes exception to the criticisms of the PPMED component in a draft of this report. Some of those criticisms have been removed from the final text. The draft had rated the project outcome as unsatisfactory, but that rating has been raised in the final report. 6. Although one of the Ghanaian consultant teams' major efforts, supported by PPMED's Director and the American consultant firm, to develop an input-output model for Ghana's agricultural sector, was never successfully implemented in computable form or transferred to PPMED and was thus largely wasted. FAO had been critical of this initiative, on grounds that it required too many resources to develop and maintain. 14 helped prepare the terms of reference for the MTE, but became progressively disappointed by its lack of involvement in the field work and eventually quit participating. The MTE was issued over the name of the principal Ghanaian consultant, with only incidental reference to MEC. It was used by the Bank as a basic input into the Midterm Review (MTR) of 1991, in which, again, MEC played a marginal role.7 Other elements of a depressing monitoring and evaluation (M&E) experience are discussed in chapter 5. 4.10 The ICR reported in 1995 that the government intended to maintain apropriate funding for Operation and Maintenance (O&M) of PPMED in the post project period. None of the follow-on subsector specific Bank projects provided funds for generalized institutional support to PPMED,9 and UNDP/FAO has not followed-up with another operation. At audit in 1996 there was criticism within PPMED that the O&M budget was in fact deficient and PPMED was unable to keep up especially with improvements in computerization10 and staff replacement. 4.11 Agricultural Policy Coordination Committee (APCC). While the project failed its commitment to PPMED, it made a strong contribution to establishing a coordination mechanism at a higher-level of policy development. A key institutional change envisaged under and supported by ASRP to help improve sectoral policy, expenditure and investment decisions was the creation of APCC, comprising senior membership from all major ministries and agencies in- the sector. The APCC was established and is functioning effectively in coordinating sector-wide policies and public expenditure plans. The committee is chaired by the Ministry of Finance and Economic Planning and serviced by MOFA as secretariat. Bank staff argue that, despite the prominence given to PPMED at appraisal, ASRP can be credited with the higher-level attention paid to policy matters and the consensus that was formed on critical issues, and that these gains were of equal or greater importance to sector developments in the early years of ERP. Success here easily compensates for the disappointments at PPMED. 4.12 Extension. In contrast to the planning component, the extension component accomplished its primary objective. This was to introduce a "modified" version of the training and visit (T&V) extension system to a small number of regions to serve as pilotsl1 for a 7. Government (MOFA) had carried out two earlier "midterm reviews" in 1989 and 1990, the first of which led to the expansion of project coverage that year (para. 2.4). The Bank felt the 1990 MTR was of inadequate depth, and called for another exercise preceded by the MTE. MEC was not involved in any of the three MTRs, and, as discussed in the text, only marginally involved in the MTE. 8. In fact, the ICR completely misjudges the performance of the PPMED component. The following leading sentence from para. 3.06 is misleading in the extreme: Capacity Building in MOFA: The more successful aspects of the ASRP include the strengthening of capacity for: policy analysis and decision-making processes; formulation of sector strategy and programs; project coordination, implementation, monitoring and evaluation. 9. Though the National Agricultural Extension Project, the National Livestock Services Project and the Fisheries Subsector Capacity Building Project each provide support to PPMED specifically to oversee and support monitoring and evaluation activities in those subsectors. This distinction is important in understanding the Region's criticism (of para 4.10) in annex E (page 2 of the memo dated June 18, 1997). 10. Condemned to the PC "286" level of obsolescence, so to speak. The ministry budget provides recurrent costs, but not for replacements. 11. Actually T&V had already been "piloted," and "modified," in the URADEP and VORADEP programs. URADEP veterans think of ASRP and the National Extension Project as sequels to their own, earlier work in importing and domesticating T&V. 15 nationwide project that was already in the Bank's pipeline. Although Ghanaians assert that their modifications have substantially altered the style of the T&V system, in fact apart from doubling the interval between training sessions, and shifting from individual to group contacts, extension in Ghana is little different from traditional T&V. It is popular among the extension administration and field staff, who are conversant with the principles of T&V and committed to making it work. 4.13 One can argue that the decisive factor in explaining its popularity has been the increased mobility of the front-line staff and their supervisors. Indeed T&V had to compete with at least two other donor-supported extension-based programs that flourished in the late 1980s, one financed by the United States Agency for International Development (the Agricultural Productivity Promotion Programme) and another by the international nongovernmental organization (NGO) Global 2000, both of which put a premium on adequate mobility for extension field staff. Global 2000 had the extra feature of combining extension with credit, so that farmer adoption of technologies could be reinforced by the availability of subsidized loans, at least as long as the subsidies were continued. Initially, T&V and Global 2000 were on a collision course, because one of the fundamental principles of T&V is that the extension service should not deal with credit and inputs, whereas with Global 2000 the same staff were obliged to deliver technology, inputs and credit. The T&V model has prevailed in that competition. Front line agents support Global 2000 and other credit programs by sensitizing the farmers to the availability of special credit programs and helping select suitable borrowers, but they are no longer involved in lending and recovering funds. Within a year of ASRP startup government requested the Bank to approve an expansion of the extension pilot to one district in each of four additional regions. By the time the new national extension project became effective in 1993, most of the country had already at least begun the process of converting to T&V methods under a single unified extension program (cocoa continued to be the important exception 2). Given the pressures in the Bank to bring T&V to Ghana, the ultimate outcome might have been the same whether or not ASRP financed a "pilot" starting in 1988. But there is little doubt that the pilot did accelerate the process, and offered useful lessons that were incorporated in the preparation and appraisal of the national project. 4.14 Research. There were four subcomponents for research, one of which, TA to support intensified rice research, was dropped early in the project period. A related subcomponent was called the valley bottom rice program, in this case research dedicated to improving paddy cultivation in the small valleys of southern and central Ghana. This program was administered by the Crops Research Institute (CRI) headquartered at Kumasi, although at two of the four sites that were eventually developed CRI turned over management responsibility to allied research agencies. The objective of this component was to develop improved methods for managing stream and river water in these small valleys to secure one good paddy crop in the rainy season and a follow-on crop of cowpea or ratoon paddy where possible. Water management meant different things in different parts of the country: at the two sites near Kumasi (Besease, Aframso) the project used bunds to protect against floods and provide better drainage to remove excess rain water; at the single site near Tamale (Yipeligu) the project used bunds to conserve rain water. In each case, the research agency developed a small area donated by the local village to test new water management techniques along with improved varieties, tillage, rotations, pest 12. The ICR errs in stating that COCOBOD Teceived some ASRP extension finance. 16 management and other husbandry practices. Nearby farmers were formed into groups and encouraged to extend the proven results to neighboring fields.13 4.15 After five years the program has little to show. The total area exploited in and around the trial sites, instead of expanding to the full potential of the valleys,14 is shrinking and there is no sign of reversing that trend. Altogether the total area dedicated to these trials now amounts to less than ten hectares at each of the four sites.15 Given the earlier interest and concern shown for this activity, the present condition is disappointing.16 Part of the explanation is that the budget for the valley bottom rice component was never substantial, never included any promotional funding to support expansion onto farmer fields (i.e., it was conceived only as a research operation) and, at the Tamale site, was cut off after three years on the suggestion of a disillusioned Bank supervisor. As mentioned above (para. 3.4), the failure of the smallholder cotton research program to extend to farmer fields exhibits the same sense of inconsequential research results. 4.16 Finally, the research activities included preparation and appraisal of a national research plan and follow-up project. The Bank assigned a larger role to this component than to the two crop-specific activities, discussed above, for the long-term development of research services. Progress here overshadowed that other poor performance. Preparation was supported by the International Service for National Agricultural Research. Implementation of that follow-on project is proceeding well. The valley bottom rice research received only minor support from the follow-on project and, at least at the time of appraisal, the cotton research received no support at all. 4.17 Irrigation. The twinning arrangement between GIDA and ADC worked reasonably well. The ICR refers to complaints from GIDA staff over the lack of communication between Koreans and Ghanaians, and lack of full participation by Ghanaians in operational decisions. That attitude apparently reflected a vocal minority: most of the GIDA staff felt they benefited from the local and international training and on-the-ground demonstrations of water management techniques. Two activities which were embedded in the ADC contract were to carry out on-farm - demonstrations on four of GIDA's large scale irrigation schemes, and also to cooperate with GIDA in managing the small scale pilot irrigation project. 4.18 The latter was actually an integral part of the ADC training program: selection of the long and short lists of potential scheme sites, the preparation of engineering designs, and . implementation of the schemes were all conceived as training exercises to equip GIDA to continue at other sites on the list after departure of the Koreans. 4.19 The small scale pilot project was not successful. As mentioned above (para. 3.3), only two schemes were started under ASRP, and a third is presently being financed by another Bank project. There are no immediate plans to continue with the other schemes. The higher than expected costs convinced the Bank that at projected yield levels these small scale investments 13. At Yipeligu SARI never developed its "own" fields: the experiments were always run on farmer-operated fields. 14. The potential is approximately 100 ha at Besease and 500 ha at Afromso. 15. CRI is running micro-trials of about one hectare each in 40 locations, but this is entirely a research operation. 16. The ICR is more complimentary about the valley bottom program because it mixes it up with the low risk water conservation scheme described in para. 4.20. 17 could not be justified. Also, the rush in the last two years of the ADC contract to finalize work on at least two sites forced GIDA and the Koreans to telescope the original, drawn-out plans for participatory planning and execution. Of the two sites that were developed, the one dedicated to paddy (Kikam, 27 ha) is a failure and the other, upland maize site (Sata, 34 ha) is limping along having reached only 60 percent of the originally planned area of 56 ha. Kikam is a tubewell site near the coast, and at the time of the audit mission the pump was disactivated because farmers had refused to pay the costs. Even though the twinning program provided GIDA professional support which was put to good use, the fact that the small scale pilot program has at least for the time being been deferred means that the central objective of the twinning arrangement was not achieved. The US$6.4 million combined costs of the institutional arrangement and the pilot investments achieved organizational reform and improved skills at GIDA, but, in terms of this component's key physical target, resulted in establishing only 34 producing hectares-mostly maize-at Sata. 4.20 That disappointment is partly balanced by the much more dramatic results of an irrigation subcomponent that was not anticipated at appraisal, the low risk water conservation technology. This "innovation" of water harvesting actually has a long history in the Upper East Region (though not for paddy). It was picked up by the project in 1992 at the urging of a dynamic ex-official of MOFA who saw it as a much preferred alternative to the so-called low cost investments under the small scale pilot. Bank supervision swung its support behind this initiative, after the excessive costs of the GIDA small scale pilot were revealed. The technology is simple. Long bunds are built on contours on the seemingly flat, wide "valleys" of the northern plains, for the purpose of not only preventing sheet erosion but holding the rain runoff to preserve moisture during stressful periods, especially in the September flowering period, and hence to save the wet season paddy crop. Yields in erratic rainfall years are reported to have doubled and tripled over nonbunded fields. The bunds are punctuated by spillways and other drop structures which allow excess water to pass: otherwise the parallel bunds are spaced to keep a three to four inch depth of water between the bunds after the occasional torrential rains (with only a one inch drop in the depth of water from the uphill side of a bund to the downhill side of the next higher bund). This is the same principle as the water conservation orientation of the valley bottom rice research at the Tamale site, except the "low risk" scheme is being applied to thousands rather than ten hectares. The low risk sites are reckoned to cost only a fraction of costs of the small scale pilot irrigation schemes.19 The program is sponsored by the exofficial, hired back by the ministry on contract, together with the Agricultural Engineering Services Department (AESD) of MOFA. ASRP kicked this program off, although the real expansion, to around 1,500 ha total paddy in 1995 (and 500 ha of upland crops), came after the completion of that project. 4.21 Policy Conditionality. Whereas performance of the investment programs was uneven, government's progress on the policy conditions was remarkably good, given the potential for political backlash. The audit mission was particularly sensitive to the issue of whether the Bank 17. The original contract was for three years (1988-91), and was extended for one year (1991/92). 18. Doubts have been raised whether the Koreans' style lent itself to leading GIDA toward more participation. ADC had been selected above candidates from India and the Philippines (and China) not because of its participatory skills but because it had a good track record and its position as an operating arm of a ministry matched the GIDA position exactly. 19. Officers involved with the water harvesting schemes claimed costs were as low as US$25/ha. GIDA told the audit mission that that real figure was closer to US$2,500. 18 bullied the government into accepting and implementing these reforms. It appears that the degree of arm-bending was small, and that in most cases the government embraced these actions as necessary to the overall recovery program. Government is more likely than the Bank now to call into question the timing of the actions, but that is a hindsight view and the schedules agreed in 1987 appear to have been endorsed by all parties. * Removal ofthe fertilizer subsidy. This was accomplished on schedule in three separate steps from 1988 to 1990. The biggest fertilizer price shock was absorbed when the cedi was devalued by 90 percent in 1984. But when ASRP was appraised in 1986, government still held a monopoly of fertilizer imports and sales, and charged the farmers only 30 percent of import and distribution costs. By 1990 that 70 percent subsidy had been reduced to zero. Removal of the subsidy was a precondition for the next step: privatization of the fertilizer trade. * Privatization offertilizer marketing. Here too government carried through with its commitment, to remove itself entirely from the trade (apart from leasing out most of its warehouses). It turned over the import/wholesale functions to the small group of firms mentioned in para. 3.5 in a series of steps taken between 1988 and 1991, and tried to encourage private retailers to fill the vacuum in the rural towns and villages. The results of the combination of the sharp and continuing reduction in the value of the cedi, subsidy removal, and the privatization program initially had traumatic effects on fertilizer usage, cropping patterns, areas sown and yields, raising questions about the appropriateness of the mechanism and scheduling of subsidy and privatization actions. In the last two years those downward trends have been reversed. Nevertheless, measured simply by fulfillment of agreements, performance against both of these targets was very satisfactory. * Privatization oftractor hire services. This was also carried out, in a series of steps that started with tractor services, then combine harvester services and finally land clearing. The existing AESD fleet was auctioned or retired. AESD feels the transformation has been a good thing for its professionalism. It has rid itself of the management of a legion of operators and mechanics, which was crowding out its engineering responsibilities while huge losses were being incurred. * Cost recovery from the sale ofveterinary supplies. This was carried out in principle, although charges have not been raised to fully cover costs and the purchasing power of the Revolving Fund that was established has eroded with rapid inflation of the cedi. KfW has objected to the management of-and to lax Bank supervision of the management of-the Fund, which benefited from the sale of the supplies imported under KfW's program (para. 4.31). * Privatization ofmost veterinary services. Here there has been little progress, due to the difficulty ministry veterinary staff anticipate after leaving the service in privately financing the establishment of rural clinics. Some private services have been successfully established at urban sites, where they support poultry and other high value enterprises and domestic pets. VSD still manages the veterinary clinics in rural areas. It does not intend to privatize vaccinations against most contagious livestock diseases. 19 VSD did get out of the import business, so that all drugs and vaccines are now imported by private enterprise (which sells to VSD what it needs).20 * Divestiture of State Operating Enterprises (SOEs). This function was removed from the project and turned over to the Divestiture Implementation Committee. Nevertheless some of the SOEs whose tenure was most strongly defended have been privatized, including the URADEP and VORADEP Farmers' Services Companies and the Ghana Cotton Company. 4.22 Other Activities. A number of other studies, seminars, investments and development programs were financed by ASRP, not all of them identified at appraisal. As suggested above, one of the substantial benefits of the project was the provision of a pool of funds to enable ministry staff to handle sudden, unexpected expenses or to pursue attractive new initiatives. These included: * a food security study, anticipated at appraisal, which was delayed by several years and ultimately set aside at the draft stage on the grounds that its main recommendations were politically unacceptable; * the training of staff, and the supply of vehicles and laboratory equipment for surveillance of locust and grasshopper outbreaks from neighboring countries; * the rehabilitation of hatchery/fish holding tanks, fishing inputs and specialized equipment for the Fishery Department's marine research vessels; and * the installation of cotton fiber testing equipment at SARI, which will help strengthen the functional bond between the research staff and the cotton companies. 4.23 Ratings. The audit provides grounds for some downgrading of the ratings of the ICR. The overall outcome is now rated only marginally satisfactory. Three of the four main capacity exercises had outcomes that fell well short of explicit appraisal targets, especially the work with PPMED but also with research and GIDA. The Operations Evaluation Department (OED) does not accept that the financing of a series of preparation reports for subsequent projects has enough 21 to do with capacity-building to warrant the higher rating. Although the ASRP reports prepared the way for capacity-building activities in the next generation of projects, that is not the same as building up the capacity in PPMED to prepare such reports. That was not done at all. 4.24 In irrigation, the main thrust of the twinning contract was to put GIDA in a position to continue to shift resources to small scale, low cost, "sustainable" investments, and this exercise was largely futile. There were welcome changes at GIDA-reorganizations, overseas training programs, etc.-that can be attributed to the twinning program. But that does not fully offset the poor results on the principal objective. Part of that intent was to enhance the participation of 20. Part I of the ICR errs in saying that the veterinary services have been privatized. 21. The ICR refers to the frequently mentioned claim that ASRP was the "seed" that produced the field of follow-on projects. That statement is correct, but in and of itself not a sign of capacity building. This argument applies specifically to the research component. The Region insists that disappointments with the two "very small" research activities were dwarfed in importance by successful preparation of a follow-up research project (see annex E, page 2 of the memo dated June 18, 1997). 20 irrigators in designing and managing their schemes, and improve cost recovery from large and small scale schemes alike. That did not happen either. On the two small scale sites developed, at Kikam farmers have not made their payments and have abandoned the canals and at Sika they feel GIDA is ignoring their interests. At Sika, GIDA did not have enough funds to extend the concrete lining on the main canal, despite an obvious and increasingly serious leakage problem. This has left the farmers with an unexpected and unwelcome maintenance burden. GIDA argues with good justification that it was never fully funded to carry out the ASRP assignments, especially after the IDA credit was closed. 4.25 Only in agricultural extension can one point to the successful achievement of the project's specific and explicit institution-building objectives. 4.26 The strong efforts by government to carry out the policy conditionalities enhance the image of this project. Of course, the impact of subsidy removal, cost recovery and divestiture on private sector behavior must be assessed to determine whether the reforms were fully effective. At appraisal, some reduction in fertilizer sales and usage was anticipated. But not the drop that occurred in the first three years. That was triggered mostly by the increase in the cedi price of imported fertilizer following devaluation, rather than by the removal of the subsidy. The trend, once observed, should have prompted a reassessment of the timing of the subsidy cuts so as not to aggravate a deteriorating situation. But the downward trend has now been reversed. Fertilizer sales, when measured by the nutrient content of the higher analysis fertilizers currently available, are regaining the level of the mid-1980s, suggesting that the long-run objectives of the reform will be reached. 4.27 The audit lowers the ICR's rating for institutional development from substantial to modest. Most of the explicit appraisal targets were missed, though some important other gains were recorded. Since the policy reforms are durable and the improvements that were made in the research, extension and irrigation services are likely to continue, the ICR's sustainability rating is left as likely. 4.28 The audit accepts the ICR rating for Bank performance as satisfactory, although that hides some rather unimpressive contributions. The quality of the project at appraisal was acceptable. But supervision was divided not only between the Bank's headquarters and field office, but also between different substantive officers, such that there was no unifying vision or commanding view of the wide array of project activities. There were important lapses in supervision reporting. No one, for example, ever talked professionally with the consultant contract officer working on M&E during the two and a half years he was in Accra. So the frustrations of an incapacitated and irrelevant M&E service were not reported to the Bank. Little attention of any sort is given to PPMED in the succession of supervision reports, except for occasional references to meetings of the Tripartite Committee overseeing the PPMED component on behalf of the UNDP, FAO and the Bank. There is not even a hint of disappointment with the exclusion of PPMED staff from high profile activities. The comment in the completion mission's Aide Memoire-appended to the ICR-that actions taken after the 1991 MTR led to "substantial improvements in the implementation performance" is not supported by the evidence in the rest of the files. In fact, there is a gap in those files at Bank Headquarters and its Field Office after the MTR. For two years there was no full supervision 21 reporting. The many errors in the ICR text cited in footnotes in this audit report are indicative of 22 a general lack of detailed familiarity with project activity, at least in its closing years. 4.29 On the positive side, Bank staff used this project as one of their vehicles to expand the dialogue with senior government officials The increasing array of Bank-supported operations is evidence of an effective and growing partnership in Ghana's recovery program. Bank staff were concentrating on the larger scene, and apparently were prepared to dismiss poorly performing components. 4.30 The Region does not agree with OED's qualified rating of overall outcome as "marginally satisfactory." It maintains that OED has dwelled excessively on minor issues, while ignoring or downplaying substantial positive aspects-and ignoring as well earlier Regional comments. OED feels secure in its position that these "minor" issues, including capacity building in PPMED, dominated the appraisal process and cannot be dismissed now simply because the components did not work. The most recent Regional memorandum, taking issue with the alleged lack of balance in the report, is reproduced as annex E. It offers an interesting alternative perspective on interpreting project results. It could have acknowledged that OED has already made a substantial adjustment by raising the rating from "unsatisfactory": precisely in recognition of the Region's well-argued points.23 4.31 The government also did not agree with the earlier draft, although it concentrated its critical remarks on the sections referring to poor performance in managing the PPMED component. Many of those assertions were substantially modified in the subsequent draft of the report, although the overall critique of management was maintained. MOFA's formal comments on the redraft are short and non-confrontational (annex D). 4.32 Borrower performance is rated as satisfactory. It suffered from the same lack of vision and inattention to the grand design that weakened Bank performance. But government's efforts to implement the policy conditionality, despite the political risks, offset those failings. 4.33 AfDB and KfW The parallel financing was accomplished, though not without difficulty. The AfDB ultimately disbursed US$27 million, but only 23 percent went to fertilizer imports (against the appraisal projection of 80 percent). Other major items financed were: tires and tubes (34 percent), fishing gear and outboard motors (22 percent), and livestock inputs (6 percent). Of the total, US$6.2 million was disbursed through PCU as part of the project. The rest was assigned to the Bank of Ghana, which sold the commodities at auction. KfW disbursed all of the US$8.5 million programmed, though much more slowly than anticipated and in the face of increasing concern about the misuse and erosion of the Revolving Fund. 22. In the official files placed in the Bank's archives, there are seven folders covering the period 1985-September 1990, and one folder covering October 1990 to April 1994. 23. Three specific criticisms in the Region's recent memorandum are answered in footnotes in this text (footnotes 9, 21, and 26). The points made in the Region's earlier memorandum have already been answered, explicitly or implicitly, in arguments throughout the audit report. 22 5. Findings and Lessons A. Findings Privatization 5.1 The project provides three examples of schemes for privatization that did not work smoothly and where the results were less beneficial than anticipated at least for the early years: fertilizer trade, tractor hire services and veterinary services. In each case the intent was to remove the ministry from a service where in principle the private sector was expected to do better. The underlying notion was that government's intervention in these businesses in the previous decades was uniformly ill-advised. In each case, the Ghanaian examples show that government was in these businesses for good as well as bad reasons, and the reforms tried to go too fast, without establishing the preconditions for privatization. 5.2 For fertilizer, there were at least two factors that discouraged private merchants from taking over the retail functions. First, was the impact of devaluation and subsidy removal on demand for fertilizers. Clearly, this was not a propitious time for a prospective retailer to enter the fertilizer market. Some components of that market held up well: specialty crops such as tomato could clearly justify fertilization even at unsubsidized rates. But this was no longer the case for maize and other field crops. The input/output price ratios changed dramatically too. Previous to the reforms three 100 kg sacks of maize could buy twelve 50 kg sacks of the most popular fertilizer formula; in early 1996 they could buy two sacks. A recent study of the privatization scheme, carried out by the International Fertilizer Development Center-Africa, shows a "value cost ratio" for the popular 15-15-15 formula of 1.1, indicating that the yield response to fertilizer barely covers its cost.24 5.3 The second factor discouraging private sector involvement was that it did not have available storage to handle the stocks of fertilizer, particularly in the more remote areas which had previously been served by trucks dispatched from ministry storage depots. Fertilizer that could not be sold in the main growing season had to be stored until the next. Without proper stores, the fertilizer not only deteriorated but often damaged the walls of the rooms where it was kept (or blew them apart!) due to the alkalization of the atmosphere. Many first-time merchants are reported to have appealed to the ministry to reclaim the fertilizer, with no refunds, simply to protect the home storage space. The situation was made the more precarious by the absence of credit facilities to help merchants finance construction of appropriate storage. 5.4 For tractor hire services, much the same route was followed: the ministry abandoning the service but the private sector failing to take up the slack. In this case the results of privatization were not as dramatic, because most of the tractor fleet was already in private hands and most contract work was already at commercial rates. But the initial impact on tractor use was as severe as in fertilizer, since the imported price of replacement parts and tractors soared with devaluation and with them the price of contract work. Private contractors could not raise prices enough to 24. IFDC-Afica. Ghana Fertilizer Privatization Scheme: Private sector roles and public sector responsibilities in meeting needs offarmers. Fertilizer Sector Studies in Africa, vol. 5. 1995, page 42. 23 cover import costs, and many of them have simply retired ("put on blocks") most or all of their fleet. Bank staff say that is a good thing, that the fleet had been propped up by distorted prices. Some government officers called it a disaster. 5.5 The situation at the time of the audit in early 1996, especially in northern Ghana, was that large scale as well as small scale farmers were severely curtailing the use of both fertilizers and tractors, shifting to varieties that do not respond to high fertilizer doses, and to bullocks and hoes for clearing and cultivation. Dealers in hybrid maize seed could not sell their seed: farmers were requesting varieties that did not depend so much on fertilizer. In fact, many of the poorer areas of rural Ghana were regressing to subsistence farming: from maize and rice to sorghum, and from larger to smaller planted areas. Large scale rice farming had almost disappeared. At the Sata irrigation scheme, most of the 56 farmers could not afford tractor services in 1996 and wet season plantings had declined substantially.25 The ministry and NGOs were all recommending a shift to bullock power. 5.6 Though they may reflect appropriate adjustments to changing economic conditions, these dramatic shifts were not anticipated at appraisal-which looked forward to the creation of a vibrant, market-driven agriculture. The argument for subsidy removal and privatization was sound. It held that Ghana was still largely in a prefertilization, premechanization mode of production, and putting prices right at the early stage would pay off handsomely in developing efficient usage of these inputs as development proceeded.26 But the reforms in the short-term accelerated the contraction of the agricultural econony.27 This is not an apology for government services. But it does suggest that the drive for privatization should have been handled more slowly, absorbing first the full shock of the devaluation, then the gradual removal of subsidies, then the campaign to establish private sector competitors. The assumption that the new dealers could supply accessible-let alone remote-farming communities at affordable prices proved untenable. It is important to recall the reforms were staged, over three or four years. So the lessons here are to expect and plan for a reaction time that is considerably longer, and to promote privatization as vigorously as divestiture. 5.7 That is the progression that has been followed in veterinary services. As mentioned above (footnote 20), the ICR errs in claiming that these services have already been privatized. That is not yet the case, at least at the time of the audit mission. VSD still operates almost all of the rural clinics. It is trying to prepare a credit program with the Agricultural Development Bank that will support ministry veterinary officers who are willing to resign in order to start private practice, but need capital to get going. DVS will remain active until this conversion process is well underway. 25. The ridges prepared for maize at the Sata irrigation scheme cannot be reshaped easily by hoe or bullock. They were designed for tractor service. 26. Fertilizer was so cheap that some farmers had been using it in excess to kill weeds. 27. The Region says this is an untenable assertion (see annex E, page 2 of the memo dated June 18, 1997). Nevertheless it is an assertion made by all Ghanaian officers in interviews on this subject. The conclusion that the reforms moved too fast is one of the few points on which the author of government's critique of the first draft of the audit report agreed with OED's findings (see para 4.31). The Bank's consultant's report on the fertilizer trade documents and comments on the precipitous decline in fertilizer sales (see footnote 24). 24 T& V Extension 5.8 T&V has come under criticism recently for its simplistic approach to dealing with complex farming systems and with relatively sophisticated farmers. In most of rural Ghana, however, especially in the north, farming systems are still so primitive as to warrant an agenda of simple messages repeated often. This may not be the optimal approach to niche farming with high value crops, or to the development of new cropping systems. The valley bottom rice and low risk water harvesting campaigns are examples of promotional programs where T&V is inadequate, at least until the new systems have been established. What is needed at first in these instances is a collaborative effort bringing together farmers, extension agents and specialized services such as CRI and AESD to prime the pump, often with the support of subsidized equipment and/or credit. But in most of the villages visited during the audit mission these campaigns were not active, and the simple technologies that the T&V front line staff were showing groups and individual farmers seemed to be making a positive contribution. Of course the visits did not provide a test as to whether the new technologies were actually adopted. Also, there were signs at the time of the audit that government would not be able to maintain the high costs of the T&V mobile field cadre. If mobility is removed from T&V much of the value of this travel-intensive routine is lost. Monitoring and Evaluation 5.9 The Bank's plans for M&E coverage of this project were ignored. During appraisal in 1986 the Bank had sent one of its M&E experts to plan a system for assessing the impacts of the government's reform program in agriculture, in particular the effects on private sector production of the exchange rate and pricing reforms. His proposal is set forth in an office memorandum28 and his short list of essential indicators is repeated in an annex in the Staff Appraisal Report (see annex C of this audit). These include inputs, production and prices. In discussions during the audit mission with the first chief of MEC29 and with the consultant adviser on M&E, it was clear neither had been aware of the earlier plans or the agreed list of indicators. Bank supervision had apparently never followed up. In fact, the ability in MEC to deliver this type of information deteriorated. Shortly after project start-up, the director of PPMED instructed MEC to abandon the work it had been doing on farm budgets, so that even the modeling of input/output relationships was no longer available to provide the key indicators to be tracked. PPMED's large field enumeration team still collects production and price data, the former based on crop cuts. But in the absence of simultaneous questioning about input use, the production function relationships cannot be determined and the essential analytical feature of the indicator list becomes inoperative. MEC was never instructed to monitor ASRP, even though it was created under its umbrella. Its absence from working parties on the three MTRs (footnote 7) and the MTE (para. 4.9) are evidence of its marginal position. A recent MTR on the National Agricultural Extension Project makes the following comment, in a section entitled "The Need for Strengthening Monitoring and Evaluation," about the difficulty in getting PPMED support for an impact evaluation of extension: 28. Ronald Ng to Dennis Casley, Chief AGRME. "Ghana: Agricultural Sector Rehabilitation Project Back-to-Office Report." July 2, 1986. 29. Who subsequently joined the Bank. 25 But this would require the active involvement of PPMED in contrast to the laid back and passive role PPMED has played in monitoring and evaluating the impact of, not only this project, but also other projects managed by MOFA.o 5.10 It should be mentioned that the M&E expert on the consultant team resigned midway through his last year because he felt the team was simply not accomplishing any of its objectives, and he himself as well as MEC were not provided the support they needed from the PPMED directorship and the consultant team leader to mount an effective M&E program. Much of the responsibility for this poor outcome can be attributed to indifference by the Bank subsequent to appraisal. Capacity Building 5.11 In terms of enabling the ministry to initiate new activities and recover some of its earlier strengths the project was successful. The fact that it financed preparation of a series of new projects is an important outcome. What is much less impressive is the progress towards establishing a permanent capacity to carry on these operations in the absence of follow-on donor contributions. The same point was made above with respect to the most successful operation, which was the extension component (para. 5.8). Sustainability of the capacity building efforts at GIDA are also suspect, because the consultant services were mostly aimed at shifting GIDA to a small scale irrigation portfolio, and that has not yet happened. Also, the two research programs- in cotton and valley bottom rice-are feeble operations which do not appear able to survive without new external funding. 5.12 But the real failure in capacity building is the PPMED. This is not a criticism directed at PPMED staff, since most of them were prepared if not eager to get involved. Rather it was the management of PPMED which failed to mobilize those capabilities, and neither the consultant team nor the FAO and Bank supervision teams pushed in that direction.31 Since the first goal listed for the project was to create this capacity, the poor results are particularly worrisome for the Bank. 5.13 The PPMED failure here seems to be partly attributable to the pressure to get the big analytical and reporting jobs done, which did not allow for the patient work with in-service training to teach MOFA staff how to handle new jobs. Government and Bank supervision were eager to produce a Medium-Term Agricultural Development "Program" in 1990, at a time when it appeared that a sector program that would embody most of the investments projected in that document would be approved by the Bank. (It was not, and the document was relabeled "Strategy.") A year later, the Bank encouraged rapid progress with the MTE on ASRP, to serve 30. "Mid Term Review: National Agricultural Extension Project." Page 18. 31. There is ambiguity in the reports as to whether the consultant terms' shift from policy analysis and planning to data collection was deliberate. In the team leader's Terminal report prepared in 1992, he claims the Tripartite Review held in September 1990 highlighted that the original project focus had been altered "in order to concentrate on development of the PPMED database rather than moving immediately to strengthening planning capabilities as outlined in the Project Document. The change in focus was necessitated by the weak database of MOA which became a limiting factor in pursuance of original project objectives." (Terminal Report by the consultant for FAO; page 4). However in FAO's own Terminal Report dated 1995 it refers to an evaluation mission fielded in September 1991 that "strongly" recommended that a successor mission prepare for "providing activities to strengthen PPMED's institutional capabilities in policy analysis and investment project preparation" (op. cit. footnote 3). The latter suggests the swing away from these skills had gone too far. 26 as a background for the MTR which the Bank had scheduled for 1991. Since FAO was primarily responsible for managing the PPMED consultancy, the Bank had a reason for stepping back and letting FAO do its job. But that does not excuse the Bank's inaction in the face of clear evidence that the capacity-building initiatives planned for PPMED were not working. 5.14 The consultant firm must also bear some of the blame. Interviews with Ghanain and Bank staff suggest the team leader was inappropriate for the job, and did not demonstrate any clear vision of where he wanted to help steer PPMED staff other than making them computer literate, issuing their marketing reports in a timely fashion, and broadening the statistical service. Unfortunately, the consultant team member primarily responsible for statistical support fell ill and was evacuated. She was not replaced. The team leader and several short term visitors from the firm's headquarters were able to maintain momentum, but the team had other problems. As mentioned in the last section, the long term M&E advisor resigned in protest about the team's mismanagement. And the other long term member of the consultant team, who was advisor to the Project Planning and Budgets Division, was not involved in the processing of the new stream of Bank supported projects. This is one of the least successful TA experiences that OED has examined in agriculture. At the core of a capacity development component, its failure is all the more glaring. 5.15 This experience raises the question whether the Bank overestimates its capacity to support institutional development. In the Agricultural Sector Review of 1985 the Bank concluded that MOFA needed broad and sustainable reforms and the Bank was in a position to help across the spectrum. The question is whether those objectives were too ambitious, or whether parts of this project were just under-managed. B. Lessons 5.16 Subsidy removal and divestiture open the way for privatization of services, but other preconditions must be met before the private sector responds. Entrepreneurs are unlikely to rapidly assume these functions, and government must prepare the way. 5.17 The unsophisticated technical messages provided by the T&V extension system can have a positive impact on relatively primitive cropping systems. 5.18 Bank indifference during supervision to M&E designs proposed at appraisal can be fatal to the chances for successful implementation. 27 Annex A Basic Data Sheet AGRICULTURAL SERVICES REHABILITATION PROJECT (CREDIT 1801-GH) Key Project Data Appraisal Actual or Actual as % of Expectation Current Estimate Appraisal Estimate Total Project Costs (US$ m) 53.3 53.2 100 Loan Amount (US$ m) 17.0 17.0 100 Cancellation - .03 - Institutional Performance ... Modest Follow-on Operation Several Many ... Not available. - Not applicable. Cumulative Estimated and Actual Disbursements FY88 FY89 FY90 FY91 FY92 FY93 FY94 FY95 Total Appraisal Estimate (US$ m) 1.5 2.6- 8.3 12.9 16.1 17.0 - - 17.0 Actual (SDR m) 1.1 2.5 6.0 9.8 14.0 15.9 16.8 17.0 17.0 Actual as % of Appraisal 73 96 72 76 87 94 99 100 100 Date of Final Disbursement: September 8, 1994 - Not applicable. Project Dates Planned Actual Preappraisal 10/30/86 10/30/86 Appraisal 11/22/86 11/25/86 Negotiations 04/13/87 04/13/87 Board Presentation 05/21/87 05/21/87 Signing 06/22/87 06/22/87 Effectiveness 09/22/87 12/18/87 Project Completion 12/31/92 06/30/94 Loan Closing 06/30/93 06/30/94 Annex A 28 Staff Inputs (weeks) Stage ofProject Cycle Planned Revised Final Through appraisal ... . 181 Appraisal through Board Approval ... ... 38 Board approval through effectiveness ... ... 1 Supervision 124 235 174 Completion 20 20 24 Total ..... 418 ... Not available. Mission Data Date No. of Staff days Specializations Performance Rating Types ofproblems (month/year) persons in field represented " rating trend Through Appraisal 04/85-12/86 24 151 Re, Ag, Ex, Fa, - - - Ae, Me, Id Appraisal Through 01/87-05/87 9 21 Re, Fa, L, Ex, - - - Board Approval Ag, Ae Board Approval 06/87-12/87 2 9 Re, Ae - - - through Effectiveness Supervisionb 03/88-08/93 22 145 Re, Ag, Ex, Fa, 1-2 1 Implementation Ae Delays Completion 05/94-12/94 1 23 Ec 2 2 - - Not applicable. a. Re = Rural engineer; Ag = Agriculturalist; Ex = Extension specialist; Fa= Financial analyst; Ae = Agricultural economist; Ec = Economist; Me = Monitoring and evaluation; Id = Institutional development; L = Lawyer. b. Project files do not support clear identification of supervision missions during much of the project period. Responsibilities were split between Headquarters and the Field Office, and between individuals. Many visits were not formally reported in supervision format. This probably explains why the ICR did not list individual missions, as required. Much of the data on this row, taken from the ICR, cannot be confirmed. 29 Annex B Project Costs (US$ million) Appraisal Estimate Actual/Latest Estimates Local Foreign Local Foreign No. Item Cost Cost Total cost cost Total 1. PPMED 0.2 2.0 2.2 0.2 1.0 1.2 2. MOFA Head Office 0.0 0.0 0.0 0.2 0.8 1.0 3. Agricultural Research 0.4 2.6 3.0 0.6 2.0 2.6 4. Fisheries 0.0 0.0 0.0 0.1 1.3 1.4 5. Agricultural Extension 0.7 1.5 2.2 0.5 1.3 1.8 6. Irrigation Services 1.0 4.2 5.2 1.7 4.7 6.4 7. Veterinary Services 0.4 1.2 1.6 0.5 2.0 2.5 8. Plant Protection 0.1 0.7 0.8 0.1 0.4 0.5 9. CIDUa 0.1 0.3 0.4 0.1 0.6 0.7 10. VORADEP 0.1 0.7 0.8 0.2 0.6 0.8 11. PCU 0.5 0.7 1.2 0.3 0.5 0.8 12. Refunding of Project Preparation Advance 0.1 0.4 0.5 - 0.05 0.05 13. Contingencies 0.6 1.3 1.9 - - - 14. Vaccines and Drugs and KfW - 8.5 8.5 - 8.5 8.5 15. Fertilizers and Chemicals AfDB - 25.0 25.0 - 25.0 25.0 Totalsb 4.2 49.1 53.3 4.5 48.7 53.2 - Not applicable. a. CIDU = Crop Services Department of MOFA. Source: Implementation Completion Report, page 21, with corrections. Acul/aes siae 1 6 31 Annex C Information Requirements under ASRP for Monitoring Effects of the On-going Economic Recovery Program 1. Area of major crops for main and minor seasons, including pure, mixed-predominant, and mixed: subsidiary areas for each major crop (Source: Crop Statistics Section-ERPS). 2. Production of major crops for main and minor season (Source: Crop Statistics Section- ERPS). 3. Disposal of major food crops for subsistence and sale (Source: Crop Statistics Section- ERPS). 4. Regional food balance (production + import + food aid - exports demand) for major food crops (Source: Crop Statistics Section-ERPS: FAO Early Warning System). 5. Prices of major crops (monthly and annual averages) at farm gate, retail and wholesale, at rural, semi-urban and Accra markets (Source: Farm Management and Market Intelligence Sections-ERPS). 6. Prices of major cropping inputs (monthly and annual averages) at farm gate, retail and wholesale at rural, semi-urban, urban and Accra markets (Source: Market Intelligence Section-ERPS). 7. Agricultural input usage = amounts of composite fertilizer (15=15=1 5), TSP, urea and others, insecticides and herbicides. 8. Use of agricultural inputs = percentage of farmers using each kind of inputs e.g., fertilizers, insecticides, and herbicides (Source: Farm Management Section-ERPS). 9. Use of hired laborers = percentage of farms using hired laborers, amount (Mondays) used, and average wage rates (Source: Farm Management Section-ERPS). 10. Farm budgets of major farming systems--cash portion only, including non-paid labor input (Source: Farm Management Section-ERPS). 11. Adoption of modern agricultural practices = recommended varieties, levels of inputs, cultivation practices and post-harvest technologies (Source: Extension Service-MOA). 12. Percentage distribution of Household Income/Expenditure and Net Balance, by Rural, Semi- Urban and Urban Enumeration Areas with reference to Poverty Line and possibly by Occupation Group (Source: Statistical Services Board-ERPS). The responsibility for collating these items of information and for preparing the progress reports rests with the Policy and Planing Section in ERPS of the Ministry of Agriculture. Initially the reports should be submitted annually. Source: Staff Appraisal Report, Annex 3-1, Attachment 2, pages 56, 7. April 22, 1987  33 Annex D MINISTRY OF AGRICULTURE My Ref No. Your Ref No. Policy Planning Tel ACCRA 665421 Monitoring Cable: AGRICOSTATS & Evaluation ...................................... .............. .......... REPUBLIC OF GHANA P.O. Box .... ............. ......................................Accru DR ROGER SLADE, AGRICULTURE AND HUMAN DEVELOPMENT DIVISION, OPERATIONS EVALUATION DEPT. THE WORLD BANK WASHINGTON D.C. 20433 USA Dear Sir, RE: AGRICULTURAL SERVICES REHABILITATION PROJECT (CREDIT 1801 - GH) DRAFT PERFORMANCE REPORT Reference your Draft Performance Report on the Agricultural Services Rehabilitation Project (Credit 1801-Gh) which was received on March 10, 1997, I wish to indicate that we take particular note of the PPMED failure - the failure of the Agricultural Services Rehabilitation Project to,"create an in-house capacity in PPMED to develop policies, programs and projects". The Management of PPMED wants to assure you that the Department will ensure the active participation of its enthusiastic professional staff in all policy and project related issues affecting Ghana's agricultural sector. The Department hopes for the Bank's support in its efforts to assert her position. Yours faithfully, for: DIRECTOR LTC-r PLANNING, MONITORING AND EVALUATION SALOME DANSO (MS) ASSISTANT DIRECTOR  35 Annex E THE WORLD BANKIFC/M.I.G.A. OFFICE MEMORANDUM DATE: June 18, 1997 TO: Ms Elizabeth McAllister, Director, OED THROUGH: Mr. Callisto Madavo, Vice President, Africa Region FROM: Serge Michailof, Country Director, AFC10 EXTENSION: 3-3322 SUBJECT: Ghana Agricultural Services Rehabilitation Project OED Perfomance Audit Report Please find attached the Region's formal comments on the OED Performance Audit Report for the Ghana Agricultural Services Rehabilitation Project (Credit 1801-GH). cc: MessrsIMme: Slade, Rice (OED); Chausse, Schreiber (AFTA3): Keane (AFC 10) Annex E 36 THE WORLD BANKIFC/M.I.G.A. OFFICE MEMORANDUM DATE: June 18, 1997 TO: Mr. Serge Michailo , Country Director, AFC 10 FROM: Jean-Paul C , ask Manager, AFTA3 EXTENSION: 36758 SUSJECT: Ghana Agricultural Services Rehabilitation Project OED Perfomance Audit Report The revised final draft of the OED Performance Audit Report (PAR) on the Ghana Agricultural Services Rehabilitation Project (ASRP, Credit 1801-GH) does not adequately take into account the detailed comments provided by our staff (see memorandum from G. Schreiber through Jean Paul Chausse to Mr. Roger Slade, dated 27 January 1997) and the Region does not agree with OED's proposed rating for project outcome of "marginally satisfactory". OED's covering Memorandum to the Executive Directors and the President itself states that there "were impressive gains in terms of overall institution-building, through the formation of the [agricultural policy] coordinating committee, the preparation of a medium-term sector r'.velopment strategy and the generation of five successor projects which would continue the work [initiated under the ASRP] at the subsector level.... [and] the policy reforms .... were carried out as agreed." As the Region emphasized in its previous comments to OED, the PAR rating on overall project outcomes reflects an unreasonably narrow view of project objectives. Getting Government out of commodity production, trading and processing, liberalizing the marketing of fertilizers and veterinary drugs, and shifting the focus to rehabilitating and strengthening its basic agricultural services were important achievements. These have resulted in a leaner civil service, substantial savings of public financial resources (on subsidies and operating losses), and the gradual regeneration of effective basic services for small holders. While a number of critical statements contained in the previous draft have been removed or modified after the Region challenged their validity and provided evidence in support of its challenge, a number of other remain that are, in the Region's view, either unjustified, given undue weight in the overall context of the project's broad scope, or inadequately balanced by failure to provide recognition to positive and desired outcomes and achievements. Three examples will suffice to illustrate these points: 37 Annex E Paragraph 4.10 states that "none of the follow-on subsector ... projects provided funds for generalized institutional support to PPMED...". Footnote 9 to that paragraph notes that the subsequent Extension, Livestock and Fisheries Projects all provided support to PPMED. Since these are, in fact, the three main subsectors covered by the Ministry, the PAR criticism is misplaced. In discussing the research subsector, the PAR goes to great lengths in paragraphs 4.14 and 4.15 to point out disappointing results with two very small adaptive research activities funded under ASRP. Only after having established a basic tenor of failure, does the PAR then, in a short paragraph 4.15, mention that the project supported a much more important activity in this subsector, the preparation of a national research plan and follow-up project, and that this was successfully achieved. With regard to the gradual removal of the fertilizer subsidy and the privatization of fertilizer marketing, the PAR argues that this had "traumatic effects on fertilizer usage, cropping patterns, areas sown and yields" (paragraph 4.21, second bullet) and that "the reforms... accelerated the contraction of the agricultural economy" (paragraph 5.6). These are untenable assertions, as the Region has pointed out before. Moreover, the policy change has indeed resulted in a welcome shift from low- to high-analysis fertilizers. cc: Mr. Schreiber (AFTA3)  섭 」ULY 1 996   IMAGING Report No.: 16857 Type: PPAR