D#ocume The Worl FOR OFFICIAL Report No. 9875 PROJECT PERFORMANCE AUDIT REPORT ETHIOPIA GRAIN STORAGE AND MARKETING PROJECT (CREDIT 789-ET) SECOND AGRICULTURAL MINIMUM PACKAGE PROJECT (CREDIT 1088-ET) SEPTEMBER 3, 1991 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. EXCHANGE RATES Appraisal Year: US$1.00 Birr 2.07 Implementation Period: US$1.00 = Birr 2.07 Cownletion Year: US$1.00 = Birr 2.07 GOVERNMENT FISCAL YEAR July 8 - July 7 ABBFEVIATIONS ADD - Agricultural Development Department ADS - Agricultural Development Service AEZ - Agro-ecological Zone AIDB - Agricultural and Industrial Development Bank AISCO - Agricultural Inputs Supply Corporation AMC - Agricultural Marketing Corporation ARDU - Arsi Rural Development Unit BCSC - Basic Commodities Supply Corporation CADU - Chilalo Agricultural Development Unit CIU - Crop Information Unit CSA - Central Statistical Authority CSU - Crop Surveillance Unit DAP - Diammonium Phosphate DCA - Development Credit Agreement EBCA - Ethiopian Building Construction Authority ECDOD - EEC - European Economic Commission EGA - Ethiopian Grain Agency EGB - Ethiopian Grain Board EGC - Ethiopian Grain Corporation EOPEC - Ethiopian Oilseeds and Pulses Export Corporation EPID - Extension and Project Implementation Department ESC - Ethiopian Seed Corporation ETCA - Ethiopian Transport and Construction Authority FAO - Food and Agricultural Organization of the United Nations GDP - Gross Domestic Product GOE - Government of Ethiopia GOSE - Government of Socialist Ethiopia IAR - Institute of Agricultural Research IDA - International Development Association IFAD - International Fund for Agricultural Development MDT - Ministry of Domestic Trade MIIF - Market Infrastructure Improvement Fund MOA - Ministry of Agriculture MPP I - First Minimum Package Project MPP II - Second Minimum Package Project PA - Peasant Association PADEP - Peasant Agricultural Development Program/Project PCR - Project Completion Report RMEA - Resident Mission in Eastern Africa RPA - Rural Projects Agency RRC - Relief and Rehabilitation Commission SAR - Staff Appraisal Report SC - Service Cooperatives SDR - Special Drawing Rights SIDA - Swedish International Development Association SORADEP - Southern Rangelands Agricultural Development Project SWC - Soil and Water Conservation TA - Technical Assistance T&V - Training and Visit Extension 4TSU - Technical Support Unit USAID - United States Agency for International Development WADU - Wolaita Agricultural Development Unit WFP - World Food Program THE WORLD BANK FOR OFFICIAL USE ONLY Washington, D.C 20433 U.S A. Office of Director-General Operatens Evilukaion September 3, 1991 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Project Performance Audit Report on Ethiopia Grain Storage and Marketing Project (Credit 789-ET) and Second Agricultural Minimum Package Proiect (Credit 1088-ET) Attached, for information, is a copy of a report entitled "Project Performance Audit Report on Ethiopia Grain Storage and Marketing Project (Credit 789-ET and Second Agricultural Minimum Package Project (Credit 1088-ET)" prepared by the Operations Evaluation Department. Attachment 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. FOR OFFICIAL USE ONLY PROJECT PERFORMANCE AUDIT REPORT ETHIOPIA GRAIN STORAGE AND MARKETING PROJECT (CREDIT 789-ET) SECOND AGRICULTURAL MINIMUM PACKAGE PROJECT (CREDIT 1088-ET) TABLE OF CONTENTS Page No. Preface . . . . . . . .i.................. Basic Data Sheets ...... . . . . . . . . . . . . .iii Evaluation Summary . . . .... . . .. . . . ..ix I. BACKGROUND .1........ ... . . . . . I II. THE PROJECTS. . ............ . . . . . .I A. Grain Storage and Marketing Project . . . 1 B. Second Agricultural Minimum Package Project 4 III. PROJECT OUTCOMES .... . . . . . . . . . 7 IV. FINDINGS AND ISSUES . . .... . . . . . . . . 8 A. Farmer Incentives . . . . . . . . . . . . 8 B. Bank Anquiescence in Counter Productive Policies. . . . . ....... . . .10 C. Storage Issues . . . . . . . . . . . . . 12 D. Aspects of Project Supervision . . . . . 14 E. Postponed Board Presentation . 15 Map - IBRD No. 22996 - May 1991 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. PROJECT PERFORMANCE AUDIT REPORT ETHIOPIA GRAIN STORAGE AND MARKETING PROJECT (CREDIT 789-ET) SECOND AGRICULTURAL MINIMUM PACKAGE PROJECT (CREDIT 1088-ET) PREFACE This is a Project Performance Audit Report (PPAR) of the Grain Storage and Marketing Project (Credit 789-ET) involving a credit in the total amount of US$24.0 million; and the Second Agricultural Minimum Package Project (Credit 1088-ET) with a credit amounting to US$40.0 million. Credit 789-ET was approved in September 1976 and closed after two extensions of the Closing Date, fully disbursed on September 30, 1985. Credit 1088-ET was approved in December 1980 and closed on June 30, 1985 after two extensions of the Closing Date, and a cancellation of US$2.1 million due to misprocurement. The PPAR is based on two Project Completion Reports (PCR) issued in 1989 and 1990 1/ respectively by the Eastern Africa/Afric- Regional Office, on the Staff Appraisal Reports, the President's Reports, Cre Lt documents, and a review of project files. Staff involved with the projects were interviewed. An OED mission visited Ethiopia in January/February 1991. The mission had discussions with officials of the Ministries of Agriculture and Domestic Trade and agencies connected with the projects. Field trips to meet participating farmers and visit project facilities were undertaken. The kind cooperation and valuable assistance given by all persons met is gratefully acknowledged. The PCRs provide a fair account and assessment of project experience, and discuss the performance of the Bank and project executing agencies. This PPAR, prepared several years after completion of the projects, re-evaluates project outcome and assesses the Bank's sectoral approach, aspects of pricing policies, and farmer incentives. The draft PPAR was s the Borrower for comments but none were received. 1/ Report Numbers 7835 dated June 21, 1989; and 8627 dated May 8, 1990. - iii - PROJECT PERFORMANCE AUDIT REPORT ETHIOPIA GRAIN STORAGE AND MARKETING PROJECT (CREDIT 789-ET) BASIC DATA SHEET KEY PROJECT DATA Actual as Z Appraisal of Appraisal Item Estimate Actual Estimate Total project cost (US$ million) 34.6 30.4 88 Credit Amount (US$ million) 240 24.0 100 Disbursed 24.0 24.0 100 Cancelled - - - Date physical components completed 12/81 11/84 Proportion completed by above date (%) 100 59 59 Economic Rate of Return (%) 18 Negative - Financial Performance - Institutional Performance Good Poor CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS FY78 FY79 FY80 FY81 FY82 FY83 FY84 FY85 Appraisal estimate (US$ M) 7.9 13.2 18.7 24.0 Actual (US$ M) - 8.6 9.4 11.7 14.1 15.6 16.8 24.0 Actual as % of Appraisal (%) 0 65 50 49 59 65 70 100 Date of last disbursement: PROJECT DATES Actual or Original Plan Revision Estimated Actual First Mention in Files or Timetable 09/7, Government Application 10/74 Appraisal - - Negotiations 06/76 06/76 (aborted) 02/77 Board Approval 09/76 09/76 04/78 Credit Agreement ,ate - - 06/78 Effectiveness Date - - 09/78 Credit Closing Date 12/81 12/84 09/85 - iv - STAFF INPUTS (staff weeks) FY73 FY74 FY75 FY76 FY77 FY78 FY79 FY80 FY81 FY82 FY83 FY84 FY85 FY86 Total Ident./ 4.0 20.2 2.6 26.8 Prep. App. 39.5 56.5 1.2 97.2 Neg. 26.6 26.6 Sup. 29.3 11.2 22.1 40.4 22.5 8.1 7.8 8.7 1.1 151.2 Sub- total 4.0 20.2 41.2 56.6 27.8 29.3 11.2 22.1 40.4 22.5 8.1 7.8 8.7 1.1 301.8 MISSION DATA Date No. of Specializations Performance Ifs of (mo/vr) Persons Represented Rating Trend PtJm ia Lb iC /4 Appraisal 07/76 4 Ag Ec, FA Supervision 1 07/78 1 Ag Ec 1 2 Supervision 2 01/79 1 Ag Ec 1 2 Supervision 3 08/79 1 FA 2 3 PFT Supervision 4 04/80 3 Ag Ec, FA, Eng 2 2 TFMP Supervision 5 11/80 4 Eng, Agr, FA, Ag Ec 2 2 TMP Supervision 6 03/81 4 Eng, Agr, FA, Ag Ec 2 3 TMFP Supervision 7 06/81 2 Eng, Ag Ec 2 3 TMFP Super4ision 8 11/81 1 Eng 2 2 TFMP Supervision 9 07/82 2 Eng, Ag Ec 2 2 MFP Supervision 10 09/83 1 Ag Ec 2 1 OTMP Supervision 11 02/84 1 Ag Ec 2 2 OPM Supervision 12 01/85 2 FA, Arch 3 2 PO Supervision 13 03/85 1 FA 4 2 PO Completion 02/87 2 Ag Ec, FA OTHER PROJECT DATA Borrower Government of Socialist Ethiopia Executing Agencies Agricultural Marketing Corporation, Ethiopian Grain Agency, Crop Information Unit Follow-on Project None /a Ag Ec - Agricultural Economist; FA - Financial Analyst; Agr * Agriculturist; Eng = Engineer; Arch - Architect. /b 1 - Problem free or minor problems; 2 - Moderate problems; 3 - Major problems. Lc Before 1986: 1 - Improving; 2 - Stationary; 3 - Deteriorating. After 1986: relating to Available funds; Project Management; Overall status. /d F - Financial; M - Managerial; T - Technical; P - Political; 0 - Ot'er. PROJECT PERFORMANCE AUDIT REPORT ETHIOPIA SECOND AGRICULTURAL MINIMUM PACKAGE PROJECT (CREDIT 1088-ET) BASIC DATA SHEET KEY PROJECT DATA Actual as Z Appraisal of Appraisal Item Estimate Actual Estimate Total project cost (US$ million) 77.2 79.2 103 Credit Amount (US$ million)/ 40.0 37.9 95 Cofinancing Total (US$ million) 18.9 11.4 60 IFAD 18.0 10.4 58 SIDA 0.9 1.0 111 Date physical components completed 12/82 06/86 Proportion completed by above date (%) 100 95 95 Economic Rate of Return (%) - 2 - Financial Performance - Fair Institutional Performance Fair CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS Jt FY81 FY82 FY83 FY84 FY85 FY86 Appraisal estimate (US$ M) 10.2 38.8 45.9 Actual (US$ M) 0.4 7.5 25.6 33.6 41.0 45.0 Actual as 2 of Appraisal (Z) 4 19 56 73 89 98 Date of last disbursement: PROJECT DATES Actual or Original Plan Revision Estimated Actual First Mention in Files or Timetable - Government Application - 12/75 Negotiations - - 04/80 Board Approval 06/80 - 12/80 Credit Agreement Date 07/80 - 02/81 Effectiveness Date - - 09/81 Credit Closing Date 06/83 06/84 06/85 - vi - STAFF INPUTS (staff weeks) FY72 FY73 FY74 FY75 FY76 FY77 FY78 FY79 FY80 FY81 FY82 M83 FY84 M85 fY86 FY87 FY88 Total Preapp. .1 - .2 .9 17.1 21.4 - - 5.5 .1 - - - - - - - 45.3 App. - - - - - 32.6 54.1 24.9 31.5 .3 - - - - - - - 143.5 Neg. - - - - - - - 13.7 33.2 5.3 - - - - - - - 52.2 Sup. - - - - - - - - - 23.1 44.1 65.3 29.4 26.6 5.4 10.3 1.3 205.5 Other - - - - - .6 .9 .5 .8 - .7 .4 - - - - - 4.0 Total .1 - .2 .9 17.1 54.7 55.0 39.2 71.0 28.8 44.8 65.6 29.5 26.6 5.4 10.3 1.3 450.5 MISSION DATA Date No. of Specializations Performance 7WW C (mo/yr) Persons Represented Rating Trend Btd_RM &C / La Preparation 09/75 1 a - - - Preparation 10/75 7 a-b - - - Preparation 10/76 1 a - - - Preappraisal 02/77 2 a-c - - - Appraisal 04/77 5 a-d-i-e - - - (Plus 5 from USAID & SIDA) Supervision 01/81 2 a-c - - - Supervision 02/81 1 a(2)-c - - - Supervision 02/81 1 e - - - Supervision 03/81 1 e - - - Supervision 03/81 2 d-a - - - Supervision 05/81 2 a-d - - - Supervision 07/81 1 f - - - Supervision 08/81 3 a-f-e - - - Supervision 09/81 1 d - - - Supervision 10/81 3 a-e-d - - - Supervision 01/82 2 a-g - - - Supervision 03/82 1 f - - - Supervision 05/82 1 e - - - Supervision 06/82 1 a - - - Supervision 09/82 1 e - - - Supervision 10/82 2 a - - - Supervision 11/82 2 d-g 2 1 M,T,O Supervision 02/83 2 a-e 2 1 M,T,O Supervision 09/83 1 d - - - Supervision 12/83 e - - Supervision 03/84 3 e(2)-d 2 2 M,0 Supervision 07/84 3 e(2)-d 2 1 M,T,0 Supervision 10/84 3 d-e-g 2 2 M,T,O Supervisiin 03/85 2 d-h 2 3 M,T,O - vii - OTHER PROJECT DATA Borrower Government of Socialist Ethiopia Executing Agencies Ministry of Agriculture Ethiopian Seed Corporation L_& Equivalent to SDR 32.0 million. Lk SDR 0.9 million cancelled. Includes disbursements from the IFAD loan of SDR 13.9 million as joint financing. le a - Agricultural Economist; b - Engineer; c - Economist; d - Financial Analyst; e - Agriculturalist; f - Livestock Specialist; g - Architect; h - Extension Specialist; i - Rural Development Specialist. Ld 1 - Minor problems; 2 - Moderate problems; 3 - Major problems. Le 1 - Improving; 2 - Stationary; 3 - Deteriorating. ff M - Managerial; T Technical; 0 - Other. - ix - PROJECT PERFORMANCE AUDIT REPORT ETHIOPIA GRAIN STORAGE AND MARKETING PROJECT (CREDIT 789-ET) SECOND AGRICULTURAL MINIMUM PACKAGE PROJECT (CREDIT 1088-ET) EVALUATION SUMMARY Introduction 1. Of Ethiopia's total land area between years, and at relating of 122.6 million hectares, only about geographical price differences 10 percent is producing crops. closely to costs. Agricultural production techniques followed traditional lines until very 3. The Second Aricultural Minimum recently. Use of modern inputs such Package Project (MPP II) was designed as fertilizer, improved seeds, and to bring basic agricultural inputs pesticides was negligible until the and services to large numbers of sixties. It was argued that farmers small farmers. It was an extension had little interest in adopting and consolidation of the MPP I modern technologies due to the share- project but the scope of minimum cropping land tenure system under the package activities was substantially previous regime. Following the 1975 broader and included virtually all land reform measures, expectations of agriculturally productive weredas modernizing the agricultural sector (districts) throughout the country. were great. Providing additional Whereas MPP I had concentrated on the incentives for production increases high potential portions of 280 received special attention as well as weredas, MPP 11 concentrated on 440 improving the farmers' access to weredas and was expected to be modern inputs. implemented in two years. Oblectives Implementation Experience 2. Tie objective of the Grain 4. Progress in the case of the Storage project was to support the Grain Storage Project was slow establishment of an ordetly grain throughout, and the project took marketing and input distribution almost twice as long to implement as system in the country through the estimated in the SAR. Construction development of GOSE's grain marketing delays were considerable, and infrastructure and reform of GOSE procurement of equipment was slow and policy towards the grain market. In troublesome. The GOSE's policy of order to permit handling of grain at socializing grain trade and the intended volume, the project was emphasizing consumer interests during to provide GOSE's grain marketing implementation conflicted with the organization with additional principles established at appraisal, infrastructure, in particular grain and the institutional arrangements warehouses, and transport. On the were altered to conform with the new policy side, the project aimed at policy. The technical assistance stabilizing prices over seasons and -x - program was not implemented and only However, following the completion of a few expatriates were hired. the project, GOSE made further changes in pricing and marketing 5. In the case of the MPP II, policies which improved grain imp amentation was slow for a variety production and produced some returns of reasons, including inadequate on the investments of the project. funding of preparation work, delays Recent liberalization of trade in procurement, late arrival of activities have led to the limited numbers of technical underutilization of Bank financed assistance personnel, and poor storage capacity. communications, with some differences of opinion between GOSE agencies and 9. The MPP II, although originally Bank supervision staff. designed as a self-standing operation, had evolved by 1980 into Results an interim program, designed to strengthen agricultural 6. The improved production and infrastructure and lay the foundation marketing of cereals, pulses and for a countrywide agricultural oilseeds expected under the Grain production project. Preparation of Storage Proiect did not materialize the follow-on project was an integral due to unfavorable GOSE pricing part of MPP II objectives. However, policies which practically froze MPP II cannot be considered farmgate prices L_ the 1979 levels. successful if evaluated as an The recalculated rate of return is isolated effort to improve negativo. agricultural production and country economic performance. It can be said 7. The MPP II did not succeed in to have achieved a fair degree of expanding the use of modern inputs success if viewed as part of the and thereby increase agricultural continuous process of Bank production. In fact, fertilizer use involvement in the agricultural decreased during project sector in Ethiopia. In this context, implementation mainly because of an it was justified as an interim effort unfavorable government pricing policy to develop institutions that play a which did not offer incentives to key role in agricultural development farmers. However, with the recently and as a precursor to PADEP, a full introduced changes in GOSE policy fledged production project. regarding producer prices and liberalization of the grain trade, Findings and Lessons fertilizer use has expanded dramatically, made possible in part 10. The Grain Storage project because the project provided the proved once more that state necessary infrastructure, interference in agricultural Nevertheless, the recalculated ERR is marketing can lead to disincentives only 2 percent. resulting i. serious setbacks of agricultural marketable production. Sustainability 11. Insufficient understanding of 8. Given the disagreement between risks involved in volatile food GOSE and the Bank over pricing and markets led to an overestimation of marketing policies, the Grain Storage trader margins. Trader profits have project did not have any sustainable to be analyzed over a several year impact on agricultural production. period to arrive at sound judgments - xi - regarding margins. Experience from 16. On-farm storage losses are very Ethiopia indicates that trader net high, on average 25 percent of the profits in the long run may reach harvested foodstuffs and exceeding 1 about 5 percent of the turnover. million tons p.a. Despite having raised this issue in supervision 12. No effort was made to find out reports, the Bank has not provided from farmers what they consider any project funds to reduce these acceptable incentives which would losses. lead to increases in production of staples. 17. The efficient use of modern inputs is reduced because most local 13. Despite a clear understanding grain, oilseed, etc. varieties have a of GOSE's intentions regarding low response potential to fertilizer. nationalization of trade and emphasis on urban consumer protection as well 18. It is unreasonable to expect as serious reservations by supervision missions to succeed in knowledgeable Bank staff, the Bank bringing about major government agreed to drop a linkage of policy changes. More attention needs output:input prices from the (Grain to be given to including policy Storage) Credit Agreement. When the issues in high level Bank-Borrower output:input price ratio deteriorated discussions, or to suspending and/or during project implementation, the cancelling projects when the Bank had no legal instrument to conditions clearly preclude their rectify the situation and safeguard success. producer incentives. 14. The appraisal overlooked the need for ancillary facilities required to make grain storage fully Endnote: operational. The oversight led to considerable cost increases. P-oject Performance Audit Report, Ethiopia - First Agricultural Minimum 15. Losses due to primitive storage Package Project (Credit 416-ET), at wholesale level tended to be Report No. 3002, dated May 30, 1980. exaggerated in Bank calculations. A recent Ethiopian study points to annual losses reaching only 0.25 percent at trader level due to relatively rapid turnover of stored cereals. PROJECT PERFORMANCE AUDIT REPORT ETHIOPIA GRAIN STORAGE AND MARKETING PROJECT (CREDIT 789-ET) SECOND AGRICULTURAL MINIMUM PACKAGE PROJECT (CREDIT 1088-ET) I. BACKGROUND 1.1 The Grain Storage Project, as seen by the Bank, was intended to support the successful implementation of the Minimum Package Project II. During the implementation of the (first) Minimum Package Project (Credit 416-ET) as well as from the Bank-financed Wolamo Agricultural Development Project (Credit 169-ET) and the Swedish-assisted Chilalo Agricultural Development Project, it became obvious that unattractive producer prices, thought to be a consequence of unfair marketing practices by private traders, were one of the major constraints to the expansion of agricultural production. The Grain Storage Project was expected to permit a healthy competition between private and state trade, thus leading to increased producer prices, adequate production incentives, and increased use of modern inputs to be provided through the Minimum Package Project (MPP II). It is for that reason that the audit of these two projects has been combined in one report. II. THE PROJECTS A. GRAIN STORAGE AND MAPKETING PROJECT (CREDIT 789-ET) Project Design 2.1 The main project objectives were to assist Ethiopia in the more efficient handling and marketing of progressively increasing volumes of cereals, oilseeds and pulses, and the distribution of agricultural seasonal inputs (mainly fertilizers, seed and pesticides) in order to: (i) closely relate the seasonal and geographical price variations to cost and to stabilize prices between years; (ii) promote oound marketing practices including proper weighing, grading and stock management; (iii) assure timely and efficient supply of inputs to the ECDOD marketing centers/cooperatives and State farms and to remove produce in such a way as to avoid congestion and give good service to farmers; (iv) decrease the marketing margin between the producers and the wholesale level and thereby improve producer prices and incentives (in competition with private trade); -2- (v) assure proper food supply distribution. Although most of AMC's sales would be on a wholesale basis, it will also take action to reduce retail margins where they seem excessive; and (vi) cover its cost, including depreciation and interest on capital employed. The rate of interest was to be fixed at a level that would take into account the risk factor and the opportunity cost of capital in Ethiopia. 2.2 The project was part of Ethio-ia's 1977-81 program to develop its grain storage and marketing system, and included: i. For the Agricultural Marketing Corporation (AMC): (a) Design, construction and maintenance of: (i) storage facilities for produce and inputs with an aggregate capacity of about 350,000 metric tons; and (ii) a headquarters building. (b) Procurement and operation of vehicles, storage equipment and office equipment. (c) Strengthening of AMC management, operations and finances, including provision of permanent trading capital, training and support for the commercial, accounting, construction, transportation and planning activities. ii. For the Ethiopian Grain Agency (EGA) (a) Procurement and operation of vehicles and office equipment for the Department of Internal Marketing and Marketing Intelligence. (b) Strengthening of the management, operations and finances of EGA's Departments of Internal Marketing and Marketing Intelligence, including training and staff support. (c) Establishment of a Market Infrastructure Improvement Fund (MIIF) to be administered by EGA for the improvem-at of local markets, including provision of loans for such purpcses. iii. For the Crop Information Unit (CIU) (a) Vehicles and office equipment for use in its general operations. (b) Carrying out a study by CIU concerning the use of remote sensing on the basis of satellite images for crop forecasting in Ethiopia. (c) Strengthening of the management, operations and finances, including training and staffing. -3- Finance Plan 2.3 The project was expected to be implemented over a three and a half year period. Total project costs were estimated at appraisal to be US$34.6 million including taxes and duties with a foreign exchange component of US$12.41 million. The IDA Credit amounted to US$24.0 million covering all foreign exchange costs plus 57 percent of the local cost (net of taxes and duties). The economic rate of return (ERR) was estimated at appraisal to be 18 percent. Pre-implementation Processing 2.4 The project was pre-appraised in 1974 but, following the 1974 Revolution and the new government's changed policies, a re-appraisal had to be undertaken in 1975. Due to the interruption of the Addis Ababa-Djibouti railway line, on which most of the construction and other materials required by the project were to be transported, presentation of the project to the Bank's Board was postponed until April 1978. Proiect Implementation 2.5 Project start-up was largely problem free, although some delay occurred due to late payment of permanent state trading capital to AMC by the Ministry of Finance. It soon became apparent that AMC intended to concentrate store construction in the earlier years thus speeding up project implementation. Institutional development at AMC progressed well but it also became obvious that AMC was reluctant to hire any expatriate staff for key positions. 2.6 The physical infrastructure programwas revised. Due to aid received from other donors, the construction of warehouses was modified from 343,000 MT at 18 locations to 202,500 MT at 22 locations. However, costs increased, because of additional expenditures for ancillary facilities, site improvements, and inflation, with final expenditures of Br 28.7 million or Br 4.7 million above appraisal estimates. The Ethiopian Building and Construction Authority, a GOSE organization, was awarded some 60% of the contracts, with the remainder awarded under local competitive bidding. There were considerable delays due to lack of construction materials in the country and problems in design and quality control. The building program was completed in November 1984, though certain remedial works were still to be undertaken by the contractors prior to final payment. Construction of the headquarters building for the Agricultural Marketing Corporation (AMC) was also delayed by four years because of cost escalations. The Bank finally agreed to the increased allocation and the building was completed in October 1984. Delays were also encountered on the procurement and installation of the computer system due to the need for retendering following objections from some of the bidders over the procedure comparing received bids. 2.7 Institution Building. The project was to establish an institutional framework for intervening in the grain market. Policies and prices were to be formulated by the Ethiopian Grain Agency (EGA) and executed through AMC on the basis of data gathered by the Crop Information Unit (CIU). This framework fell apart early in the project. The EGA was not adequately staffed and was disbanded in 1980. Its functions spread to AMC, the Ministry of Domestic Trade (MDT) and -4- the Ethiopian Standards Institute. The transfer of AMC from MOA to MDT further weakened the production incentives aspect of its operations and emphasis was shifted to the consumer side. Equally discouraging was the little progress made in setting up the Crop Surveillance Unit in the Ministry of Agriculture (MOA). Dispute centered on how crop forecasting was to be organized. The technical assistance program contained in the appraisal report was not implemented and training of staff did not live up to appraisal targets. Only a few expatriate consultants were hired. 2.8 Policy Implementation. GOSE's policy of socializing the grain trade and the emphasis on consumer interests conflicted with the principles established at appraisal. According to the SAR, producer prices should provide adequate incentives to farmers and produce trade should be encouraged through Government credit and reasonable margins for produce trading. In the 1970s, with grain prices high on the free market and AMC procurement quotas small, the issue of production incentives did not arise. But as AMC increased its quotas in the early 1980's and purchased at prices lower than the free market, the issue of disincentives to farmers became serious. The Bank was thus party to this system under which farmers were obliged to sell increasing shares of their crops to government at below-market prices. Implementation of policies envisioned in the SAR was further weakened by the lack of a market intelligence and crop- forecasting function due to the dismantling of EGA. In addition, restrictions in movement and tightened control of the private sector discouraged produce trade. B. SECOND AGRICULTURAL MINIMUM PACKAGE PROJECT (CREDIT 1088-ET) Project Design 2.9 The project was designed to: (i) increase the productivity and incomes of small farmers by introducing them to fertilizer, improved seeds and other agricultural inputs; (ii) strengthen the institutional and technical framework to progressively expand and improve the minimum package program; and (iii) help to ensure long-term productivity of the land by expanding the Government's soil and water conservation efforts. 2.10 The project was to assist a two-year time-slice of GOSE's minimum package program and also to permit expansion of the program giving coverage to 440 of the country's 586 weredas (counties). In particular the project was to: (i) strengthen the Ministry of Agriculture and its field extension support to the minimum package program by providing staff and staff training, vehicles, buildings and equipment for the Ministry's headquarters and field offices; - 5 - (ii) provide, on credit, crop inputs including fertilizers, improved seeds, pesticides, and tools, and also animal husbandry inputs in selected areas. In particular, the project was to include only the annual incremental farm inputs in each of the years 1980/81 and 1981/82, using the level of inputs prevailing in 1977/78, the last year of MPP I, as a reference base for 1980/81; (iii) strengthen the Ethiopian Seed Corporation's capacity to produce basic seed, and the peasant associations' capacity to multiply and clean seeds, by providing technical assistance, buildings and equipment, and funds to help the Corporation meet its annual incremental operating costs; (iv) construct low-cost rural t-ads to support the expansion of the minimum package program; (v) strengthen the Institute of Agricultural Research by providing staff training, and conduct applied research and field demonstrations to help develop improved technical packages for farmers; and (vi) expand soil and water conservation efforts in badly degraded areas, and strengthen GOSE's capacity to implement further conservation work by providing field staff and equipment. Finance Plan 2.11 The project was expected to be implemented over a two-year period with three years allowed for disbursements. Total project costs were estimated at appraisal to be US$77.19 million including taxes and duties. The foreign exchange component of US$29.9 million represented about 39 percent of total estimated costs. The Credit of US$40 million was to finance 68 percent of the project's foreign exchange costs and 49 percent of the local costs, net of taxes. The International Fund for Agricultural Development (IFAD) allocated US$18.0 million to the project and the Swedish International Development Authority (SIDA) was to contribute US$0.9 million to the project costs. An economic rate of return was not estimated at appraisal on the curious pretext that "the project is a time-slice of an ongoing nationwide program." Nevertheless, the staff appraisal report (SARI affirms that a rate of return of 20 percent "is expected to be exceeded if the project as designed is implemented as projected." (SAR paras. 6.05, 6.06). Pre-implementation Processing 2.12 There was a delay in Credit Effectiveness due to belated appointment of technical assistance staff to key positions. Delays were further exacerbated by a six-months' postponement in presenting the Credit to the Board which caused uneasiness on the Ethiopian side and contributed to halting further preparatory work. -6- Project Implementation 2.13 After start-up project coverage was extended to further regions and more weredas in addition to those included in the appraisal. Major changes were also introduced for the seed component under the project and the introduction of a pilot training and visit (T&V) system, shifting of input supply responsibilities and the construction of self-help stores rather than Ministry of Agriculture (MOA) stores. 2.14 MOA was the lead agency for MPP II and was responsible for implementation of most components. The Ethiopian Seed Corporation (ESC) under the Ministry of State Farm Development, implemented the remaining activities. Other institutions which played important roles in the implementation of MPP II were the Institute of Agricultural Research (IAR); the Ethiopian Roads Authority which was, through the Ethiopian Transport and Construction Authority (ETCA), responsible for construction of some 200 km of rural roads, and the Agricultural and Industrial Development Bank (AIDB) which was responsible for providing credit to farmers for the purchase of inputs. The Agricultural Marketing Corporation (AMC) was originally intended to provide inputs. However, during implementation, AMC's input supply role was shifted to the newly established Agricultural Inputs Supply Corporation (AISCO), a semi-autonomous institution within MOA. 2.15 The MPP II project was to have been implemented over the two fiscal years 1980/81 and 1981/82. However, the project took longer to complete than expected. The original estimates were too optimistic given the experience under MPP I. Cautionary advice given during the review within the Bank should have been heeded. The SAR implementation schedule provided for the completion of physical works over 19 months. In practice, physical works within MOA took over 66 months to complete. The main cause for slow implementation of physical facilities was the slow start to construction due in part to the fact that no funding had been available for preparation work. Once construction had started, it proceeded with reasonable speed. Necessary plant and equipment was slow to arrive, and vehicle procurement was poorly coordinated. Also, it took a long time for TA posts to be filled. GOSE seemed reluctant to contract with a consulting firm to staff these posts. They relied instead on SIDA and on the Agricultural Development Service, through which the World Bank filled 6 posts. It was only towards the end of the project (November 1984) that GOSE entered into a contract with a consulting firm to fill three critical posts in MOA. Communication with the Bank was cumbersome and included differences of opinion between GOSE agencies and Bank supervision staff. 2.16 The national trials program for applied agricultural research suffered a delay of three and a half years because of MOA's reluctance to accept the appropriate technical assistance. In contrast to this component the soil and water conservation component progressed well, in part because of the timely employment of specialists under the TA provisions. - 7 - III. PROJECT OUTCOMES 3.1 The economic rate of return at project completion of the Grain Storage Prolect is estimated to be negative, compared with the SAP. estimate of 25 percent including buffer stocking and 18 percent excluding buffer stocking. The project did not have a positive impact on production incentives and grain production. During the project period, AMC purchase prices did not give a satisfactory return to fertilizer use and its consumption actually declined from a peak in the first year. Benefits from improved warehouses are estimated to be negligible and the effect on income redistribution was negative. Due to price controls and restrictions in movement, farmers got less for their produce (purchased by AMC) and consumers paid more (for open market grain). No benefits resulted from buffer stocking since AMC never operated any buffer stock scheme. Impact on consumer welfare is hard to judge, on balance there probably has been a measure of consumer protection in large urban centers such as Addis. However, the general quality of the grain distributed has deteriorated because of the elimination of quality grading. Returns to AMC's intervention in input distribution were negative on balance. 3.2 The MPP II project has succeeded in establishing the basic infrastructure necessary to deliver productivity-improving packages to farmers. An extension service, based at service cooperatives, has been established. The foundation for setting up an effective agricultural input supply corporation has been laid and a system of cleaning and distributing improved seeds has been developed. Most importantly. an effective pilot T&V extension project was implemented which has brought about significant improvements in "institutional understanding" of the role and value of extension. This increased acceptance has prepared the way for the establishment of a professional extension service in Ethiopia. The MOA capacity to identify, prepare and appraise development projects was improved through support to the MOA Project Preparation Department. 3.3 Production incentives deteriorated in the early 1980's. The price of fertilizer rose rapidly. This paralleled the imposition of the compulsory grain quota purchase system at low prices and its progressively rigid enforcement (official grain prices remained largely unchanged from 1979 on). Simultaneously, the distribution of consumer goods to the rural sector declined while prices rose. The expanding population resulted in new age groups entitled to land, which, combined with uncertainties about future collectivization, caused land tenure insecurity to emerge again as a disincentive to agricultural investment. Thus, the deteriorating policy environment worked against major production increases arising from project activities. Only recently have supply constraints for fertilizer been eased, and production incentives improved. Fertilizer use in 1986/87 had risen to record levels. The Economic Rate of Return (ERR) calculated for MPP Il is estimated at 2 percent. IV. FINDIV1'! AND ISSUES A. Farmer Incentives 4.1 From the onset of the Bank's involvement in assisting agricultural development in Ethiopia, the provision of adequate incentives to farmers was of paramount interest to the Bank. It was correctly assumed that the country had a considerable potential for boosting agricultural output through use of modern inputs such as improved seeds, fertilizers and pesticides. Up to 1974, during the pre-revolution days, the Bank considered the then existing land tenure system as a major disincentive to modern production techniques since tenant farmers, a large segment of the agricultural sector, had to share their harvests with the landlords on the one hand and carry the cost and risk of input use alone on the other. 4.2 Start up of the Minimum Package I Project coincided with the Ethiopian revolution. In light of the introduced land reform measures, acceptance of modern farming practices was encouraging. While harvests had no longer to be shared between producers and landowners, it became apparent that supporting services needed to be strengthened. Credit and supply of inputs were handled rather efficiently by the project authorities in charge of the MPP, CADU, and Wolamo projects, while marketing was singled out as showing serious deficiencies. By comparing farmgate prices with consumer prices, the Bank concluded that farmers marketing their produce were short changed, did not receive their fair share of consumer expenditures for foodstuffs, and thus continued to lack the necessary incentives to expand production. 4.3 Because of what the Bank considered unduly wide buying/selling margins as well as unreasonable seasonal food price fluctuations, the conclusion was reached in the early seventies, at the project preparation stage, that existing trade was inefficient, monopolistic, and above all, unfair to producers. Throughout the identification/preparation stage, the Bank was convinced that a Government marketing organization would successfully compete with private traders and provide farmers with attractive and relatively stable "floor prices" for cereals, pulses, and oil seeds. 4.4 In the audit's view, the Bank's approach to improving agricultural marketing though a grain storage project managed by a state agency was flawed in several aspects. First, by the early seventies when the grain storage project was prepared with Bank assistance, ample negative experience was available to the Bank about inefficiencies of state marketing organizations. The East rn Africa Regional Office in particular had gained this experience during impiamentation of projects in Tanzania, Malawi, and Burundi, to name only a few.1/ 1/ For instance, the supervision reports of the mid-seventies on Tanzania National Maize, Livestock, Cashewnuts, Tea, and Tobacco projects; Malawi Lilongwe, Karonga, and Shire Projects in regard to the Government trading/marketing organization ADMARC; and Burundi Coffee. - 9 - 4.5 Too much importance was also given to what were considered excessive handling margins and storage losses prior to appraisal. The appraisal mission, confronted with a Bank-prepared feasibility study, recognized some of the doubtful assumptions of project preparation but--difficult to understand-- accepted them despite their critical comments. The SAR 2/ described the marketing system then existing as "...an operation which, because of the uncertainties involved, carries with it high risks and the possibilities of either high profits or high losses." (The SAR cites "Marketing of Selected Agricultural Commodities in the Baco Area", W. Manig, November 1973). Discussing seasonal price fluctuations, the SAR concludes "...that the trading margin in addition to regular operating costs and a "normal" profit should thus include a risk premium which over time can be adjusted to equal the true financial cost of stockpiling. Such a premium currently charged may at times mistakenly be referred to as "excess profit", and as such be assumed to be available for redistribution to producers/consumers".3/ 4.6 The Bank was also well informed about the CADU experience in marketing operations at the time of appraisal, an experience demonstrating the fallacy of assuming excessive trader margins. Farmgate prices were raised by an average of Br 3.14/quintal (official wheat price 1980: Br 22/quintal), but at the same time CADU marketing operations suffered losses of up to Br. 2.29/quintal. The SAR comments on these operations by indicating that despite CADU controlling about 16 percent of total marketed production its influence on "the pricing and marketing practices has not been firmly established."A/ Nevertheless the Bank went ahead and expected proper incentives for the smallholder sector to come from a government marketing agency. 4.7 A review of project files gives no indication that the Bank at any time considered cooperative marketing as an alternative to state trading. From supervision reports of the First Wolamo Agricultural Development Project (Credit 169-ET) it was known to the Bank that cooperatives had played an important role, especially in the timely supply/distribution of inputs, but had suffered setbacks in produce marketing due to their relatively small size. The supervision reports pointed to the need for organizing these small cooperatives into cooperative unions to counteract the trust actions taken by the few large wholesale merchants at the Addis Ababa market. It should be noted that Ethiopia is in a much better position to promote cooperatives (service cooperatives) than many other African countries because of a 60 percent literacy rate. The Bank, as a matter of fact, had paved the way to cooperative development by assisting the construction of local godowns under the MPP I project. 4.8 In conclusion it can be said that the focus on improved marketing through a state organization, aiming at the provision of adequate incentives to smallholders, would not have worked even if the Government had not pursued a 2/ SAR, Annex 1, page 7, para. 6. 3/ SAR, Annex 1, p. 14, para. 26. 4/ SAR, Annex 1, page 20, para. 37. - 10 - pricing policy detrimental to production. In the absence of price incentives production remained stagnant at best and both projects failed. It is surprising that during all the years of project preparation and appraisal, the Bank had never tried to find out from the farmers what they considered an adequate price incentive.5/ B. Bank Acquiescence in Counter Productive Policies 4.9 Next to expected incentives through reduced trader margins as discussed above, the Bank anticipated substantial increases in modern seasonal input use through maintaining a favorable fertilizer:cereal price ratio. The draft Credit Agreement provided for linking cereal prices to the cost of fertilizer, a dubious enough idea to begin with. Facing stiff Ethiopian opposition to this covenant as being too complex to administer, the Bank retracted during negotiations and contented itself with an assurance that ... "The Borrower shall establish produce price ranges...to ensure: (i) adequate production incentives to .armers, taking into consideration, inter alia, the input costs of farmers and the cost of trai.sporting produce from farms to marketing centers,.. ."6! 4.10 Was the Bank correct in dropping a rather precise stipulation for a broad, general statement, and forgoing conditionality; is it only with the benefit of hindsight that it can be said the producer price structure was doomed from the onset? As project experience proved, GOSE never adhered to the principle agreed upon during negotiationa and the farmers' input cost:output price ratio deteriorated continuously during project implementation leading to declines in input use and consequently stagnation of agricultural production. 4.11 Didn't the Bank have any possibilities of correctly assessing the likely developments in Ethiopia? And were these developments inevitable? The Revolutionary Government had pronounced its intentions on many occasions and the Bank was fully aware of them. The Bank was informed about GOSE objectives which, based on Bank experience in other countries,Z/ could be identified as economically unsound, a valuation free of any political judgment. So the Bank had adequate warning. 4.12 Encouraged by finding a government that immediately embarked on the land reform requested and recommended by the Bank for such a long time, no substantial resistance--in the form of conditionality for Bank lending operations--was put up by the Bank to avoid policies which were counterproductive to agricultural development, including the introduction of collective farming, 5/ The net farm-gate price increase under the CADU marketing arrangements would have represented less than 4% if full marketing costs had been deducted. 6/ Credit Agreement, Section 4.05. 7/ Tunisia, Cooperative Farming; Benin, Oil Palm; Tanzania, Geita Cotton. - 11 - fixed producer/input prices, state trading monopolies, preferential treatment of urban consumers versus producers, etc. 4.13 Did the Bank blindly stumble into something nobody could anticipate? The answer is in the negative since numerous warnings were raised within the Bank and also information from official sources gave the Bank adequate insight in regard to clearly-stated GOSE policies. For instance, at an early preparation stage, the Bank's Resident Mission in Eastern Africa (RMEA) objected to a parast .1 marketing monopoly whose activities "...tend to impede, rather than speed commercial development.8/ A number of Bank memos were equally to the point: "While the Bank might favor the retention of a strong private element, this may conflict with government views." (memo 02/19/75). "Producer and consumer interests can hardly be combined in 'one organization'." (memo 02/20/75). "I think we [the Bank] have to be very careful in our support of a new marketing arrangement and absolutely convinced that private trade is replaced by something which is more efficient otherwise the Bank will only have contributed to a setback for domestic production. (memo 04/17/75) "Unless a clear role for private trade is defined, announced and adhered to, AMC will be obliged to assume an even greater task which could become unmanageable." (Back-to-Office Report 08/01/75). "...we have little faith in the viability of the Government's policies for grain storage and marketing. We do not believe that they can be implemented under the proposed project." (memo 04/02/76). 4.14 It was not only Bank staff voicing concern about the likely problems of state marketing or the intentions of GOSE to monopolize food/agricultural marketing. Not only did GOSE's intention emerge during staff discussions with Ethiopian authorities but documents available to the Bank spelled out GOSE policies: "The time frame for the creation of a wholesale monopoly will be dictated by the capacity of AMC building up its organization." (Minutes of the Interministerial Meeting of 06/14-17/75. "The Government may, where necessary, engage in wholesale and retail business in order to stabilize prices particularly of basic consumption items, and thereby protect the interest of the masses." (Declaration on Economic Policy of Socialist Ethiopia; 02/07/75). All the documents/memos quoted are contained in the Bank's project files. 4.15 While GOSE in pursuance of its policy objectives focused on acquiring additional grain storage facilities to expand state trading without any trace of being in agreement with the Bank's idea of providing incentives to the farmers-- 8/ Bank memo dated July 11, 1973. - 12 - collective and state farming receiving priority--the Bank, without any evidence, still believed in free trade and expected higher farm incomes to prevail. Although politicel considerations never entered project discussions, the question needs to be asked why the Bank supported an economic decision when it knew that state trading monopolies had not worked in the past and were unlikely to work in the future. The audit wonders if the same positive project decision would have been taken if the Bank had to sl 3re project risks. In the case of Bank-assisted projects, however, repayment of the Credit rests squarely on the shoulders of the Borrower. C. Storage Issues 4.16 Credit 189-ET provided for the construction of large warehouses with a total storage capacity of 343,000 tons of which 292,000 tons was for grain/pulses/oilseeds storage and 51,000 tons for input distribution. Existing government and rented storage capacity was estimated by the appraisal mission to be around 143,600 tons, and private wholesale storage to approximate 467,500 tons (SAR, Annex 1, Table 7). Annual wholesale food trade was estimated at some 750,000 tons in the early seventies. Considering the fact that additional storage facilities existed at the small trader level, and that a minor part of the marketable surplus is not being sold immediately after harvest, justification for the addition of more storage capacity on economic grounds was doubtful, however desirable it might have been on political grounds. 4.17 Given the relatively large storage capacity already available, the Bank's support for further large storage expansion becomes questionable.9/ It is open to challenge also on having opted for the wrong priorities in regard to reducing storage losses. It should be noted that about 80-85 percent of all foodstuffs produced do not enter markets at urban centers but are consumed either on the farms or sold directly to the non-agricultural rural population. The quantities not entering urban markets can be deduced from the following table. Table 1 Ethiopian Cereal Froduction (million tons) Year 1979/80 1982/83 1983/84 1984/84 1985/86 1986/87 1987/88 1988/89 Production 6.5 6.7 5.5 4.2 4.8 5.6 5.9 6.5 Source: IBRD, Report No. 8062-ET. 2/ Following the recent liberalization of grain trade sizable AMC storage capacity remains unused. - 13 - 4.18 Considering that the Ethiopian rural population numbers about 40 million and that average annual per capita cereal consumption is around 140 kg, some 5.4 million tons of cereals are stored on the farms. According to Ethiopian experts, on-farm storage losses reach up to 50 percent of stored foodstuffs with a GOSE-estimated average of at least 25 percent. In contrast to these high losses, the appraisal of the Grain Storage projected assumed losses at deficient trader stores to be about 5 percent. However, more recent surveys mentioned in the PCR (para. 7.15) indicate that losses, even under primitive trader storage conditions, rarely exceed 0.25 percent, a consequence of traders being interested in relatively fast turnovers of their stocks. 4.19 When deciding to assist large-scale storage, the Bank opted to save about 50,000 tons annuallyiO/ instead of contributing to the reduction of losses amounting to about 1.3 million tons annually. Grain and other foodstuffs saved from losses would also considerably increase marketable surpluses and accomplish something both projects under review were unable to attain. 4.20 The need for improving on-farm storage was raised already during supervision of MPP I. Traditional storage is very primitive and cereals, oilseeds, and pulses are subjected to a relatively high humidity during part of the year, leading to mouldy grains, serious insect infestation, etc., compounded by high losses due to rodents. Limited studies about proper low cost storage construction (on-farm) have been undertaken by Ethiopia with outside assistance but more work, especially for different ecological regions, needs to be done. 4.21 Appraisal of the project's storage facilities was incomplete because ancillary facilities were omitted from project design, i.e. offices, storage rooms for empty bags, fumigants, cleaning sheds, guard houses, fencing. All these facilities had to be added to the project-financed storage facilities, leading eventually to substantial cost overruns--of already underestimated warehouses costs. The following table indicates the magnitude of costs omitted by the appraisal. Table 2 Construction Cost of a 5,000 t Capacity Warehouse (1981 costs) Warehouse Br 650,739.50 2 room office Br 30,671.20 Bagstore Br 36,553.10 Fumigation room Br 9,374.50 Cleaning Shed Br 13,272.75 Guard Housing, Fencing Br 37,354.70 Total Br 777,938.75 Source: AMC Ancillary facilities = 16% of total cost. 10/ The 50,000 tons correspond to the high appraisal estimate of 5 percent losses. According to recent surveys, the losses would be 2,500 tons. -14 - D. Aspects of Project Supervision 4.22 The two projects teach interesting lessons on how the Bank is reacting to recommendations made by supervision missions and what it is expecting from supervision missions. 4.23 Preparation, appraisal, and implementation of the two projects under review provide evidence that supervision mission reports do not get the attention required. As mentioned above, supervision of the Wolamo Project (Credit 169-ET) highlighted the properly-functioning project-established service cooperatives as far as input supplies were concerned and recommended the early founding of cooperative unions to strengthen the marketing side of these small cooperatives. No indication could be found in Bank files that cooperatives were ever considered an alternative to state marketing. The unsuccessful supervision efforts to draw the Bank's attention to on-farm storage losses have been discussed above. The MPP II project had a sub-component providing farmer credit for the purchase of insect/rodent control materials for use on farms but did not support any research into on-farm storage construction or finance of such facilities. 4.24 Supervision of the MPP I project raised the issues of the unsatisfactory response potential of traditional grain varieties to fertilizer application as early as 1974. This problem was highlighted in several other supervision reports. The supervision report of the July-August mission of the Grain Storage project states in Annex 6, page 8: "According to several recent reports by consultants to USAID, Ethiopia's research establishment is unable to produce a stream of fertilizer-responsive cereal varieties, without which the continued profitability of the fertilizer strategy cannot be maintained." This was more than a year prior to the project's appraisal updating missions. The approved project did not include measures to improve the situation. 4.25 The RMEA supervision mission of September/October 1981 raised the problem of the MPP projects once more: "In the high production areas the lack of response to fertilizer is linked to the varieties now in use. Given the potential of these favored districts, it would appear that a special research effort is justified to identify cultivars which are capable of responding to higher levels of nitrogen than those currently in use, which would result in substantial grain surpluses becoming available from these intrinsically fertile zones." The Bank is presently considering a national seeds project in Ethiopia but the issue of cultivars showing better response to fertilizer application remains unresolved. 4.26 While supervision did not succeed in bringing about Bank action on some important issues, Bank/management expected supervision missions to successfully resolve major polizy issues, e.g. the Government pricing policy, when the same subject could not be settled successfully during Credit negotiations. 4.27 The Bank's optimism on possible accomplishments by supervision missions requires some rethinking. Since the majority of the Bank's customers have by now a long-established working relationship with the Bank, they are fully - 15 - aware of the Bank's hierarchical structure. Recognizing that most members of supervision missions are technical staff, relatively on the lower rungs of the Bank establishment, they have rarely nn opportunity to discuss their findings with ministers or permanent secretar es or their equivalents. Discussion of major policy issues is thus confined to exchanges of views with some lower- ranking GOSE officials, and Bank staff are in no position to know if their suggestions, or criticism of policy issues, are or will be conveyed to the Borrower's decision makers. 4.28 Equally unsatisfactory is the communication in writing following supervision missions. Usually the Division Chiefs are responsible for summarizing the supervision findings and recommendations in letters which may or may not be addressed to or received by high level government officials. 4.29 During the implementation of the two projects under review there were several encounters between senior Bank management and high level GOSE officials especially at the Bank's Annual Meetings. Minutes of these meetings give no indication that the pricing policl,, which was particularly important for the success or failure of these audited projects, was discussed. Of course supervision missions failed in getting GOSE to change this rather delicate policy issue. A le&son emerges from this experience: putting unresolved major, project related policy issues on the terms of reference of supervision missions is doomed to fail; putting the issues on the agenda of senior management/borrower discussions offers better chances. E. Postponed Board Presentation 4.30 Board discussion of the Grain Storage Project was postponed by the Bank for more than a year giving rise to some serious complaints from the Borrower. When Bank management withheld Board presentation of the project at short notice, the Borrower suspected the Bank to have given in to outside political pressure, a very serious accusation in light of the Bank's bylaws specifically excluding political considerations in its decision-making process. 4.31 It was for that reason that the audit paid special attention to this issue. From all information contained in Bank files and from interviews conducted it has become apparent that the sole reason for the management decision was a concern that (i) the blasting of bridges of the Addis/Djibouti railway might have led to serious shipping delays of warehouse construction materials and (ii) these delays would lead to substantial cost overruns. The accusation that the Bank succumbed to outside political pressure can also be refuted by the fact that the Bank member which Ethiopia suspected of intervening actually welcomed the proposed project in a letter to the Bank's Ethiopian desk. E ETH l0 PIA GRAIN STORAGE AND MARKETING PROJECT AREA WHE'E PEOPLE ARE IN ROADS JND'R CONSTIUC'ION N NEED OF lIOOD A5SISTANCE TOTAL POPUILATION By PRVINCE *A5d bI,'eb A~, INMILL IONS = = GAv A WEATHIR EST-MTED POPuAtIO)N BY PROvlNCE NEEDING FOOD Å$C IN MLLIONS .I INTERINAONAL IRPORTS NAT] ZNAL CAP TAL SPoTs ® PROVINCE CAMTAL .O.DS PROVINCE BOUNDARIES ASPHTALT ALL W"ATKR 212.12.,02022 ..&. AsHL ALWAHEI- INTERNAT;ONA, BOUNDARIES GRAVEL.ALL WEATHER - RURALANDIORD0Ry WEATER WAREHOUSES ARSOMGIE ASELÅ L~2 DM AUMLE DEUE MARKOS 5 U D A N K~t oo EE Omed e DmE DAWA w - DJIBOUTI EFESON KOMBOL /MEKE ebe MEN 3 xce GNANRET Ÿ NEKEMTE DD5 ABABA S O m A I A h 02* AioTæo b Mo 'W W 0 2M ,A UGANDA r 'rL APRIL 1991