Madagascar - Agricultural Sector Adjustment Credit

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 13081 PERFORMANCE AUDIT REPORT MADAGASCAR AGRICULTURAL SECTOR ADJUSTMENT CREDIT (IDA CREDIT 1691-MAG AND SFA CREDIT A-16 MAG) MAY 23, 1994 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. Fiscal Year of Borrower Weights and Measures January 1st - December 31st Metric System ACRONYMS AfDB - African Development Bank CASA - Cr6dit A l'ajustement du secteur agricole (Agricultural Sector Adjustment Credit) CASEP - Cr6dit A l'ajustement du secteur public (Public Sector Adjustment Credit) CASI - Cr6dit A l'ajustement du secteur industriel (Industrial Assistance Project) CASPIC - Cr6dit A l'ajustement du secteur des politiques industrielles et commerciales (Industry and Trade Policy Adjustment Credit) IDA - International Development Association OGL - Open General Licence Facility (same as SILI) SFA - Special Facility for Africa SILI - Sistme d'importation liberalis66 (same as OGL) THE WORLD BANK FOR OFFICIAL USE ONLY Washington, D.C. 20433 U.S.A. Office of Director-General Operations Evaluation May 23, 1994 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Performance Audit Report on Madagascar Agricultural Sector Adjustment Credit (IDA Cr. 1691-MAG and SFA Cr. A-16-MAG) Attached is the Performance Audit Report on Madagascar Agricultural Sector Adjustment Credit (IDA Cr. 1691-MAG and SFA Cr. A-16-MAG) prepared by the Operations Evaluation Department. The audit concurs with the conclusion in the Project Completion Report (PCR) that the overall assessment of the Credit is satisfactory, and its achievements are likely to be sustainable. The Credit included no institutional development objectives. The adjustment operation was consistent with both Government policies and Bank support strategy. The rice liberalization component, the core of the reform program, was, in the event, successfully implemented and generated a reduction in budgetary subsidies. In contrast, compliance with other policy changes was partial, and the rationale and justification for releasing the Credit's second tranche are questioned. Eventually, however, these other policy changes were either effected under, or rendered irrelevant by, the two subsequent adjustment credits (Industry and Trade Policy; and Public Sector). The audit discusses the retrogression in price liberalization in the second year of implementation, the problems among the three last adjustment operations and between them and investment projects, the justification for second tranche release, and the unfortunate lack of "safety net" arrangements to protect the poor. Robert Picciotto by Hans-Eberhard Kopp Attachment  FOR OFFICIAL USE ONLY PERFORMANCE AUDIT REPORT MADAGASCAR AGRICULTURAL SECTOR ADJUSTMENT CREDIT (IDA CREDIT 1691-MAG AND SFA CREDIT A-16 MAG) TABLE OF CONTENTS Page No. PREFACE ........................................................ i BASIC DATA SHEET ....................................................iii EVALUATION SUMMARY .............................................. vii 1 BACKGROUND ..................................................... 1 2. OBJECTIVES ....................................................... 2 3. IMPLEMENTATION AND OUTCOMES ................................... 2 Rice Liberalization ................................................. 2 Ancillary Policy Reforms ............................................. 5 Overall Performance of the Agricultural Sector ............................. 6 4. OVERALL ASSESSMENT ............................................. 6 5. ISSUES AND LESSONS ................................................ 7 Rice: Success and Relapse ........................................... 7 Adjustment: the sequencing problems from IDA's side ........................ 8 The ancillary conditions and the release of the second tranche .................. 9 Adjustment, rice liberalization, and the poor .............................. 10 ATTACHMENT 1: PCR Evaluation Summary ................................. 13 ATTACHMENT 2: Comments from the Ministry of Agriculture and Rural Development ...................................... 19 This report was prepared by Jos6 Olivares (Task Manager), who audited the project in February, 1993. Pilar Barquero provided administrative support. 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.  PERFORMANCE AUDIT REPORT MADAGASCAR AGRICULTURAL SECTOR ADJUSTMENT CREDIT (IDA CREDIT 1691-MAG AND SFA CREDIT A-16-MAG) PREFACE This is the Performance Audit Report (PAR) on the Madagascar Agricultural Sector Adjustment Credit, for US$60 million equivalent, approved on May 8, 1986, signed on May 20, 1986, and declared effective on November 18, 1986. The second tranche was released in May 1989, two years late. The original Closing Date was extended twice to June 30, 1990. The PAR is based on the Project Completion Report (PCR)' prepared by the Africa Regional Office; the President's Report; the Credit Agreement; supervision reports; correspondence between the Bank and the Borrower; consultant reports and internal Bank studies; memoranda; and interviews of Bank staff, Government officials and private operators. An OED mission visited Madagascar in March 1993. The mission discussed the genesis, implementation and impact of the Credit with current and former officials in the Ministries of Agriculture, Planning, and Finance, the Central Bank, and the Prime Minister's Office. It also interviewed producers and importers of agricultural inputs and commodities. The PCR is complete and informative. The PAR builds on its findings and explores the wider context of the operation, particularly regarding certain areas which the PCR did not satisfactorily resolve. The audit discusses the retrogression in rice liberalization in the second year of implementation; the problems among the three last adjustment operations and between them and investment projects; the justification for second tranche release; and the effects of adjustment on the poor. The audit concurs with the PCR's rating of the project. Thus, the outcome of operation is rated as satisfactory, and its achievements sustainable. The Credit succeeded in the institutional development objectives. Following OED procedures, copies of the draft PAR were sent to the Government for comments on February 8, 1994. The comments received from the Ministry of Agriculture and Rural Development are included as Attachment 2. The PCR was prepared by the Africa Regional Office and is scheduled to be released to the Executive Directors and the President in January 5, 1994. The Governments was sent Parts I and III of the PCR on March 26, 1992. No comments were received, despite follow-up from the Resident Mission.  PERFORMANCE AUDIT REPORT MADAGASCAR AGRICULTURAL SECTOR ADJUSTMENT CREDIT (IDA CREDIT 1691-MAG AND SFA CREDIT A-16-MAG) BASIC DATA SHEET CREDIT POSITION (Amounts in US$ Million) Credit Original Disbursed Cancelled ReRald Outstanding IDA 1691-MAG (SDR) 19.0 19.0 0.0 0.0 19.0 SFA A-16-MAG (SDR) 31.0 31.0 0.0 0.0 31.0 Sub-total (SDR) 50.0 50.0 0.0 0.0 50.0 Japanese grant (YEN) 600.0 n.a. n.a. n.a. n.a. German contribution (DM) 10.0 n.a. n.a. n.a. n.a. CUMULATIVE ESTIMATED AND ACTUAL DISBURSEMENTS FY87 FY88 FY89 FY90 FY91 IDA 1691-MAG (SDR) Planned 10.26 19.00 19.00 19.00 19.00 Actual 3.29 9.32 10.40 16.71 19.00 SFA A-16-MAG (SDR) Planned 16.74 31.00 31.00 31.00 31.00 Actual 5.80 12.98 15.18 24.21 31.00 Total (SDR) Planned 27.00 50.00 50.00 50.00 50.00 Actual 9.09 22.30 25.58 40.92 50.00 Actual as % of Planned 34 45 51 82 100 Japanese grant (Yen) German contribution (DM) n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. n.a. - Iv - PROGRAM DATES Original Actual Initial Project Brief - May 17, 1984 Final Project Brief - May 14, 1985 Negotiations - Dec. 23, 1985 - March 7, 1986 Letter of Development Policy - March 24, 1986 Board Approval May 8, 1986 Signing of Credit Agreements Effectiveness May 20, 1986 Second Tranche Release June 30, 1986 November 18, 1986 Closing Date March 31, 1987 May 1, 1989 June 30, 1988 June 30, 1990 STAFF INPUTS (staffweeks) Fiscal Year 82 83 84 85 86 87 88 89 90 91 92 93 TOTAL Project Preparation and Pre-appraisal 3.3 47.6 29.9 48.8 129.6 Appraisal 32.3 12.6 47.7 92.6 Negotiations 17.1 17.1 Supervision 2.3 40.3 34.6 15.5 0.2 92.9 PCR 5.0 10.0 15.0 Total 3.3 47.6 62.2 61.4 67.1 40.3 34.6 15.5 0.2 0.0 5.0 10.0 347.2 MISSION DATA Month/Year No. of No. of Staff Report Ratings Weeks Persons Weeks Date Identification April 1984 2.5 3 8 Pre-appraisal June 1984 2.5 4 8 July 16, 1984 Preparation September 1984 1 1 1 October 5, 1984 October 1984 1.5 5 8 December 5, 1984 February 1985 1.5 1 2 April 3, 1985 Appraisal May/June 1985 3.5 4 14 June 26, 1985 (issues paper) Post-Appraisal September 1985 2.5 3 8 October 16, 1985 September 1985 2 2 4 n.a. Supervision July 1986 2 2 4 July 22, 1986 n.a. October 1986 1 1 1 November 3, 1986 n.a. January 1987 2 3 6 March 12, 1987 2 May 1987 3 1 2 June 22, 1987 n.a. July 1987 3 3 6 August 11, 1987 2 October 1987 1.5 2 3 November 30, 1987 2 February 1988 3 2 3 June 20, 1988 n.a. July 1988 2 2 3 August 12, 1988 2 November 1988 3 1 1 n.a. n.a. Completion November 1992 2 1 2 March 1993 Borrower/Executing Agency: Democratic Republic of Madagascar  - vii - PERFORMANCE AUDIT REPORT MADAGASCAR AGRICULTURAL SECTOR ADJUSTMENT CREDIT (IDA CREDIT 1691-MAG AND SFA CREDIT A-16 MAG) EVALUATION SUMMARY Background 1. The Agricultural Sector Adjustment had been liberalized in 1983. In late 1985, Credit (CASA), which provided US$ 60 the President of Madagascar and the Bank's million for balance of payments support, was Regional Vice President agreed that the the second in a series of four adjustment Government would free rice trade in those operations to liberalize product and factor areas and the Bank would help in setting up markets and increase efficiency in resource a buffer rice stock. During the first project allocation. It followed the 1985 Industrial year (1986-1987), the Government effective- Assistance Credit, and was followed by the ly implemented the rice agreement. The Industry and Trade Policy Adjustment Credit results appeared by the book: abundance of (CASPIC) in 1987, and Public Sector Adjust- rice in the markets, lower and more stable ment Credit (CASEP) in 1988. CASA was consumer prices, higher producer prices. approved on May 8, 1986. The second tranche was released in May 1, 1989, two 4. In the second project year, however, years late. The Closing Date was extended Government did not adhere to the rice twice, to June 30, 1990. The audit finds the agreement. Due to political considerations PCR description and analyses of develop- (paras 10 - 11), imports increased beyond ments and achievements accurate and can- the levels agreed; stocks were sold at low did. Its Evaluation Summary is attached for prices and even during the harvest period. easy reference (Attachment 1). Stop-gap agreements with the Bank were signed and then breached. In the third Objectives project year, the agreed modus operandi was resumed. 2. The core of the reform program was to eliminate direct Government intervention 5. In the event, rice liberalization was market in the rice market. Other objectives most successful. Rice is freely traded; im- included improving pricing and marketing for ports have been drastically reduced and major crops; eliminating subsidies on agri- when needed, carried out by private import- cultural inputs and transferring their supply ers. Administrative restrictions to private to the private sector; and eliminating from sector operations are minimal. Rice markets the public investment program seven pro- became more efficient. Price differentials jects IDA had found questionable. now reflect transport costs and the cost of storage and risk; price differentials between Implementation and Outcomes paddy and rice have been reduced. Private mills are rapidly replacing state-owned com- 3. Except for the two largest rice-pro- panies, and several firms are exporting luxury ducing areas marketing of paddy and rice rice qualities. Prices to producers increased - vii - substantially, as did for those consumers who Overall Assessment before the reform had had access to subsi- dized rice distribution. 8. The Credit outcome is satisfactory. The core policy reform, rice liberalization, 6. In contrast, achievements on the had been well identified and prepared. It other policy reforms were limited. (i) The proved successful, increased producer prices seven investment projects which the Bank and reduced Government subsidies. Its had objected were implemented, raising the success facilitated the acceptance by Govern- issue of how far the Bank can go in criticiz- ment of further adjustment. Rice reform is ing investments financed by other multina- deemed to be sustainable. In contrast, the tional organizations. (ii) The special fee on other policy changes were either effected export crops for parastatals was reduced, and under, or rendered irrelevant by, CASPIC the price of coffee raised, but increases in and CASEP. The justification for releasing coffee exports, were dampened by the col- the Credit's second tranche is discussed by lapse in international coffee markets in 1989. the audit. (iii) The study on administrative constraints to agricultural exports was completed; dia- 9. As an unenvisaged result of its own logue on this matter was furthered under the success, CASA jeopardized the operation two subsequent adjustment credits. (iv) The and sustainability of two irrigation projects program for promoting private sector partici- supported by IDA under earlier credits, pation in agricultural marketing was never whose implementing agencies had enjoyed prepared; the Bank felt the issue would be monopsonistic rents on rice marketing. better handled under CASPIC or CASEP. (v) The study on local government finances Issues and Lessons was completed, but the action plan was never prepared. (vi) The study on export Rice: Success and Relapse crop stabilization funds became irrelevant since Government (under CASPIC) abol- 10. Why would the Government abandon ished the monopolies on coffee, cloves and a policy which had demonstrated its effec- pepper exports. (vii) For similar reasons, tiveness? The Government stated it has the study on edible oils was obsolete by the privatized the official rice import activities. time it was issued; the action plan was not What this actually meant was that rice dona- prepared. (viii) New seed quarantine regu- tions were channeled to a commercial outfit lations were adopted, and the Government of the ruling Party, thus technically placing stopped supplying fertilizer and veterinary them outside the terms of the rice agree- products; limited progress was achieved in ment. The Bank did not accept this argu- the removal of fertilizer subsidies. ment and charged the Government with breach of faith. But this leads to the para- 7. After ten years of stabilization and dox of a Government abandoning a success- adjustment, agricultural sector performance ful program for short term political gains. remains disappointing. Between 1984-1990, the crop sub-sector grew at 1.5% per annum. 11. In the audit's view, what happened Growth came mostly from rice, the commod- was the result of the power struggles fre- ity benefitting the most under adjustment. quently observed in adjustment operations. Traditional export crops, not liberalized and Up to late 1985, political forces against facing poor prospects in international mar- liberalization held the upper ground. But kets, stagnated or regressed. they were upstaged by the agreement be- tween the President and the Bank. They -ix - were able to recoup and, on grounds of the 15. There is a need for careful coordina- forthcoming elections, mount a counter- tion between successive adjustment opera- offensive the following year. But their tions, and between adjustment and invest- policies failed, eventually reinforcing the ment operations, the more so if they are President's hand. handled by different divisions in the Bank. If adjustment will affect the fundamentals on 12. At the end of 1988, market prices which investment projects depend, the scope climbed. IDA allowed use of CASA funds or design of the affected projects started be to import 30,000 tons of rice, conditioned on revised at the time of the appraisal of the a minimum disposal price. In January 1989, adjustment operations. active Bank micro-management of rice im- ports ended. Government stopped operating The release of the second tranche the buffer stock in 1990. 16. There was little to show as compli- 13. Two lessons are derived from this ance with second tranche conditions (para experience. First, convincing key political 6), but the Bank, released it anyway. It is players of the merits of adjustment as a obvious from the context and details that the viable policy and political path is essential. success of the overall adjustment process in And second, adjustment processes should be Madagascar, had outweighed the rather timid sequenced so as to place the earlier benefits CASA conditionality. The release occurred of reform in the hands of those who favor one year after the Board had approved the adjustment. release of the second tranche for the Argen- tina Agricultural Sector Loan on rather Adjustment: the sequencing problems from the similar grounds. It is too early to tell wheth- Bank's side er the "Argentina principle" has become the de facto Bank policy for releasing tranches in 14. Problems arose among the adjust- not-fully-compliant operations within overall ment operations, and between them and satisfactory adjustment processes; the neces- agricultural investment projects. First, the sary evidence will come from future perfor- major devaluation concurrent with Board mance audits. approval of CASPIC reduced demand for fertilizers imports funded under CASA. Adjustment, rice liberalization, and the poor Second, CASPIC introduced an easier, non- discretionary licensing procedure for alloca- 17. Before adjustment, real incomes and tion of foreign exchange while CASA used a consumption had been boosted to unsustain- cumbersome bidding procedure. This slowed able levels. In relative terms, the rice subsi- down CASA disbursements and led to a two- dy benefitted the urban poor the most, but year delay in releasing its second tranche. consumers outside the official distribution CASA disbursement mechanisms were modi- channels gained no benefit. Stabilization fied only when this tranche was released, a and adjustment reduced both aggregate serious misjudgment on IDA's part. And demand and the rice subsidy. The latter, third, some parastatals implementing agricul- plus liberalization, increased consumer prices tural investment projects were privatized or and led to a reduction in family incomes and dismantled under CASEP without consulta- in rice consumption at all levels. In hind- tion with IDA's agricultural division, jeopar- sight, CASA, might have incorporated a dizing the operation and sustainability of safety net for the poorest. these projects.  -1- PERFORMANCE AUDIT REPORT MADAGASCAR AGRICULTURAL SECTOR ADJUSTMENT CREDIT (IDA CREDIT 1691-MAG AND SFA CREDIT A-16 MAG) 1. BACKGROUND 1.1 The Agricultural Sector Adjustment Credit (CASA ') was the second in a series of four IDA-supported adjustment operations to liberalize product and factor markets and increase efficiency in resource allocation, and thus lay the foundation for the recovery of the Malagasy economy, which had deteriorated rapidly since the early 1980s. Initially, the Government was not ready to implement a broad-based structural adjustment effort. Instead, it adopted a sector-by- sector strategy. This approach was supported by two IDA sector adjustment credits, the Industrial Assistance Project (CASI; Credit 1541-MAG') in 1985 and the Agricultural Sector Adjustment Credit (CASA) in 1986. Building on the success of the latter, adjustment gained momentum. The Government, recognizing that sectoral policy improvements by themselves were unlikely to lead to significant supply response, moved to a broader agenda of reforms supported successively by the Industry and Trade Policy Adjustment Credit (CASPIC; Credit 1834-MAG) in 1987, and the Public Sector Adjustment Credit (CASEP; Credit 1941-MAG) in 1988.' 1.2 CASA was supported by an IDA Credit (Cr. 1691, for USD19 million equivalent), an SFA Credit (Cr. A-16, for USD31 million equivalent), a Japanese Grant for JPY 600 million, and a German contribution of DM10 million. All together, they provided for balance of payments support of approximately US$ 60 million, in two tranches. The IDA and SFA credits were approved on May 8, 1986, signed on May 20, and declared effective on November 18, 1986, four- and-a-half months after the anticipated date. The first tranche was released on effectiveness. The second tranche, originally scheduled for March 1987, was actually released on May 1st, 1989. The original Closing Date was extended twice, to June 30, 1990. 1.3 The PCR' presents a good summary of developments under the Credit and of its main achievements. The audit finds its description and analyses accurate and candid. Therefore, they are not repeated here; the PCR's Evaluation Summary is attached for easy reference (Attachment 1). The audit text, however, summarily presents the main features of the project and The terms "the Credit" and "CASA", the Credit's French acronym, will be used interchangeably through this report to refer to the Agricultural Sector Adjustment Credit. 2 This early operation was audited in 1992. Report No 10405, of February 27, 1992, includes both the PAR and the PCR. 'The former is already closed; its PCR was issued as Report No 10540, April 15, 1992; the PAR is currently under processing. The latter is still under implementation. 4 PCR on Madagascar Agricultural Sector Adjustment Credit (IDA Credit 1691-MAG and SFA Credit A-16-MAG); Report NO 12544 of November 24, 1993. -2- developments through its processing as required to introduce the discussions and make the text self-standing. 2. OBJECTIVES 2.1 CASA supported four major policy objectives. First, to progressively reduce the need for rice imports by eliminating direct Government intervention in the domestic rice market. Second, to improve pricing and marketing systems for major export and industrial crops. Third, to gradually eliminate subsidies on agricultural inputs and transfer the responsibility for input supply to the private sector. And fourth, to improve the quality of public investment in agriculture by eliminating from the public investment program seven projects IDA had found questionable. 2.2 If grouped by importance, however, CASA had two main objectives. The core objective was the liberalization of rice pricing and marketing, including an end to Government rice imports. All others can be grouped as ancillary objectives, which covered a range of desirable but not critical improvements in several areas, like eliminating subsidies to agricultural inputs, reducing the parastatals' special fee (commission) on export crops, adopting a new methodology (acceptable to IDA) for computing the producer price for coffee, and re-studying questionable public investment projects, plus others to be defined during project implementation as a result of studies to be funded from the Credit. 3. IMPLEMENTATION AND OUTCOMES Rice Liberalization 3.1 Marketing of paddy and rice had been liberalized in 1983, except in the two largest rice- producing areas (Lac Alaotra and Marovoay). In spite of intensive Bank pressure during preparation and appraisal, Government refused to let them go free. Technically, the grounds for retaining them under control were "food security"; Government wanted to have assured access to rice stocks for distribution to the urban population. But there was a strong political element as well. It took a lengthy meeting held in late 1985 between the President of Madagascar and IDA's Regional Vice President to forge an agreement, the so-called "rice compact" whereby the Government would free rice trade in those areas and IDA would help in setting up a buffer rice stock (stock tampon)'. A description of the "Rice Compact" and the operation of the buffer stock is pre-cited in Box 1. Since appraisal had already been carried out, the technical and legal details of the buffer stock had to be worked out during Negotiations, which had to last for three months. -3- BOX 1. THE RICE COMPACT AND THE BUFFER RICE STOCK 1. As part ofthe Agreement between President Ratsiraks and Mr. Jaycox on November 7, 1985, the Bank agreed to "encourage other donors to finance the ewtablishment of a rice buffer stock and participate on a "last resort" basis ourselves, as part of a global "rice compact' between the Government and the donor community. The IDA participation is needed to ensure an IDA role in management of the rice import program and release of imports into the local market... .We intend that the buffer stock will be a transitional device to deal with Government's political problems and not a permanent scheme, although we cannot require Goverment to agree to its transitional character now." 2. The content of the Rice Compact, above all of the buffer stock, was worked out during negotiations of CASA. Underlying the agreement was acceptance of Government's monopoly on importing rice until 1990, when self- sufficiency was to be attained and rice-impcirts wsould no longer be needed, During this period, the Rice Compact was to provide a better approach to managing rice imports than in the past. The essential element in the ST %as an automatic mechanism for putting imported rice on the markets when the market price exceeded the "trigger price." In addition, sales from the stock were to occur only during the pre-harvest period, or 'soudure"-from November through the end of April-when rice is usually in short supply and expensive. 3. The buffer stock (or Stock Tampon) was to be multipurpose: (i) a stabilization stock based on the trigger-mechanism (ii) a stock to meet the needs of the oficial circuit (iii) a catastrophe stock 4. Specific measures to implement the buffer stock were spelled out in the agreement expressed in a side letter of the Minister of Finance, of March 24, 1986V. The Most important of these proved to be: (i) Government would set up a "Special Cell" (that is a special office) responsible for developing rice- management policy and procedures, plus implementation and coordination of the operators in the production-marketing and import-marketing chains. To this end, the Cell would be empowered to enter into contracts and act as the Government's agent in the management of rice imports, whether commercial or through foreign aid. The Cell would be responsible for the maintenance, disposition, and periodic replenishment of the ST. (ii) Before October 31 of each year, Government would agree with the World Bank on total rice imports for the succeeding year, part of which would be used for the official circuit, part for the buffer stock. Government was also to periodically inform the Bank of the evolution and projections for basic buffer stock parameters. (iii) Part of the stock could be used to help victims of natural catastrophes. (iv) The first trigger-price for Antananarivo was FMg 480/kg ($0.62 at the overvalued exchange rate). Once market-prices reached FMg 480, sales from the buffer stock would be made at FMg 450. In initial operations the rice was sold in minimum lots of one ton from several outlets. (v) The Government agreed to open the two reserved areas to private trade by April 15, 1986. 1/ The 'Rice Compact* involves material in three documents: Development Credit Agreement, for Credit 1691-MAO, May 20, 1986, Sections 3.03 and 3.07 Madagascar, Cr6dit d'Ajuatement Sectoriel Agricola, Prock4-verbal des NEgocilions, du 12 d6cembre, 1985 au 16 janvier, 1986, Program global da gestion da riz, pp. 5-6; letter and annex of the Minister of Finance to the International Development Association, March 24, 1986, re: Madagascar, Agriculture Sector Adjustment Project. Source. Attachment from AF3AG, Evaluation of the "Buffer Stock', February 1991. Draft, paras 2.2 - 2.S. -4- 3.2 During the first project year (1986-1987), the Government effectively implemented the rice agreement. The results appeared by the book: abundance of rice in the markets, lower and more stable consumer prices, higher producer prices. 3.3 In the second project year (1987-1988), however, the Government discontinued adherence to the rice liberalization program and reverted to the pre-adjustment mode of importing excessive rice stocks and dumping them on the market at the most inappropriate moments and prices. Imports increased well above the level agreed with IDA under the rice agreement. Bank missions constantly called this non-compliance to the attention of the Government to no avail; stop-gap agreements were signed and then breached. In the third project year, the agreed modus operandi was resumed, but the incongruence between the Government's satisfaction with the positive first-year results and its relapse in year two is not easily explained. 3.4 The rice pricing and liberalization component was most successful. Today, the commodity is freely traded; need for imports has been drastically reduced - last year to zero - and when needed, imports have been carried out by private importers. 3.5 Rice markets have become more efficient'. Price differentials between heretofore unconnected markets today reflect transport costs; price differentials between seasons reflect cost of storage and risk; and price differentials between paddy and rice have been reduced. For example, for Lac Alaotra, the largest rice producing area in the country, the above-mentioned study found reasonable price margins for spatial arbitrage; prices in Lac Alaotra are well integrated with prices in Antananarivo and Toamasina - the major consumer and exporting markets, respectively; transfer to urban consumption centers at costs lower than three years earlier; reduction of margins between paddy and rice due to pressure by private - and usually small-scale - rice mills (d&cortiqueries). Market situation was found to be healthy and highly competitive; no cartel has been formed among the multiple private traders purchasing and processing rice in the Lac area - in spite of a few attempts to create one. Private mills are rapidly replacing state-owned companies (whose share in the market has fallen from 60% in 1990 to 10% in 1992). Several firms are exporting luxury rice qualities. Administrative restrictions on private sector operations were found to be minimal. 3.6 At the same time, rice liberalization resulted in a net increase in consumer prices in real terms for consumers who had had access to subsidized rice distribution before the reform. Open market prices in 1987 were 75% higher in real terms than what the official price had been in 1982. Favored urban consumers, particularly in the four largest cities, were affected; the poorest 25% of households in Antananarivo were specially hit'. This issue is discussed in further detail in the Issues and Lessons Section. 3.7 Even more important, success in the first year of liberalization was instrumental in convincing Government of the welfare gains which could be reaped from adjustment, and opened 6 J.P. Azam et al. 9tude du marche du RIz A Madagascar, CERDI/DAI/SOA.TEG, provisional report, December 1992, prepared for the Ministry of Agriculture. 7 Dorosh, PA; R.E. Bernier and A.H. Sarris, Macroeconomic Adjustment and the Poor the Case of Madagascar, Cornell Food and Nutrition Policy Program, Monograph 9, December 1990. -5- the - heretofore closed - way to the two additional adjustment operations already mentioned , which have also proved successful. Ancillary Policy Reforms 3.8 In contrast to the success in rice liberalization, achievements on the ancillary policy reforms were more limited and less effective. (i) The seven investment projects IDA had considered as dubious and in need to be reassessed (para 2.1) were in the event implemented. The most controversial case involved Antalaha, an oil palm scheme which the African Development Bank had already appraised and agreed to finance. AfDB strongly opposed any attempt at changing the design or timing of the project. A cashew processing plant, also financed by the AfDB, was built as well. This raises the issue of how far IDA can go in criticizing investments being financed by other multinational organizations, which have criteria, thresholds, and lending programs which do not necessarily correspond to those of IDA. 3.9 (ii) As agreed, the Government reduced the special fee on export crops for parastatals, and raised substantially the price of coffee before adopting IDA-agreed methodology; the potential effects of improved coffee producer prices on production and exports, however, were dampened by the collapse in the international coffee agreement in 1989. (iii) The study on administrative constraints to agricultural exports was completed as planned; further dialogue on this matter was pursued under CASPIC and CASEP rather that under CASA. (iv) The program for the promotion of private sector participation in the marketing of agricultural products was never prepared, and IDA dropped this requirement under CASA in October 1987 since several decisions had already been made or were being envisaged, and the Bank felt the issue would be better handled under CASPIC or CASEP. 3.10 (v) The study on local government finances was completed, but the action plan required for second tranche release was never prepared. At the time, IDA felt the issue was politically too sensitive and that it was better to defer it until the then on-going elections were over. But no concrete decision has since be made, and a number of critical issues were still pending at the time of the audit mission. (vi) The study on export crop stabilization funds was eventually completed, but by the time it came out it was largely irrelevant since Government had made the decision (under CASPIC) to abolish the state monopoly on coffee, cloves and pepper exports. As a result of this and other measures supported by IDA under CASPIC and CASEP, Government amply exceeded the modest conditionality of CASA concerning export crops. (vii) Largely for similar reasons, the study on edible oils was similarly irrelevant by the time it was issued. IDA decided that it was best to rely on the trade liberalization, export promotion, and parastatal restructuring measures supported under CASPIC and CASEP. The action plan required for second tranche release was never prepared. 3.11 (viii) New seed quarantine regulations were adopted. The Ministry of Agriculture disengaged from direct fertilizer distribution, and the state monopoly on veterinary products was dismantled. Mechanisms for foreign exchange allocation to private fertilizer importers were put This assessment was unanimously stated by both Bank personnel and relevant Government officers interviewed during the audit mission. -6- in place. But only mixed progress was achieved in the removal of fertilizer subsidies; at the time of the audit visit, donated fertilizers were still allocated administratively to established operators, and pricing remained an issue. Overall Performance of the Agricultural Sector 3.12 On the whole, after ten years of stabilization and adjustment, the overall performance of the agricultural sector shows progress, but remains nonetheless disappointing. During the 1984- 1990 period, the crop sub-sector had an average growth of only 1.5% per annum, which came mostly from rice, the commodity benefitting the most under adjustment. Traditional export crops, not liberalized under CASA and facing poor prospects in the international markets, stagnated or regressed. This poor sub-sectoral performance is attributed to remaining price and market distortions, and deficiencies on support services and infrastructure. Most of the former are being addressed under the subsequent adjustment operations; some of the latter, under investment projects. 4. OVERALL ASSESSMENT 4.1 Both the PCR and the audit rate the Credit outcome as satisfactory. The core policy reform objective, rice liberalization, had been well identified and thoroughly prepared. The component proved quite successful and resulted in increased producer prices, improved supplies, and reduced Government subsidies'. Its evident success played a major role in the acceptance by Government of economy-wise adjustments. The best tribute to the widespread national satisfaction is that in the 1992 presidential election campaign, a propitious time for criticizing incumbents, none of the candidates described the rice policy in negative terms or included restoration of price or marketing controls in his plank. In this sense, the rice reform achievement is deemed not only satisfactory but also sustainable. 4.2 As an unenvisaged result of its own success in rice liberalization, CASA jeopardized the operation and sustainability of two irrigation projects supported by IDA under earlier credits, whose implementing agencies perceived substantial monopsonistic rents on rice marketing. Rice marketing liberalization put an end to those rents. . 4.3 In contrast, there is no similar evidence that the ancillary policy changes had been properly prepared or fully accepted by the Government. Compliance with them was poor, and the rationale and justification for releasing the Credit's second tranche (on the basis of such compliance) are discussed below in the Issues and Lessons Section. Eventually, however, these 9 This non-threatening nature of the policy reforms has also been observed as a feature of other successful agricultural adjustment operations, such as the Brazil (Agricultural) Credit and Marketing Reform Project, PAR, under processing, and the Tanzania Agricultural Adjustment Credit, PCR under preparation. 10 These two projects, Mangoky (Credit 881-MAG) and Lac Alaotra Intensification (Credit 1137-MAG), were eventually given the coup de grace by CASEP. A detailed discussion of negative interrelations between adjustment operations and investment operations is presented in Madagascar, Madagascar. Performance Audit on Five Investment Credits, Report No 12646 of December 30, 1993. -7- ancillary policy changes were either effected under, or rendered irrelevant by, the two subsequent adjustment credits, namely CASPIC and CASEP. 5. ISSUES AND LESSONS Rice: Success and Relapse 5.1 Why did the Government abandon in the second year a policy which had just demonstrated its effectiveness to everybody's satisfaction (para 3.1 to 3.3). According to the officials interviewed at audit, the Government acted on the bona fide belief that it had privatized the official rice import activities - a proper action under the rice agreement. What this actually meant, however, was that official rice donations from friendly governments, like North Korea, which up to then had been received and handled directly by the Government - and were thus under the terms of the rice agreement - were instead channeled to/through a commercial outfit of the ruling Partyn - thus technically placing them outside the terms of the rice agreement. IDA did not accept this argument and charged the Government with breach of faith. But this leads to the paradox of a Government allegedly abandoning a successful program for short term political gains. 5.2 In the audit's view, what happened was the result of the kind of power struggles between groups for and against reform frequently observed in adjustment operations. Up to late 1985, political forces against liberalization - of any kind, including rice liberalization - held the upper ground. But they were surprised and upstaged by the unexpected November 1985 agreement between the President and IDA's Vice President (para 3.1), and did not have the strength to oppose implementation of the rice agreement during its first year. But they were able to recoup and, or grounds of the forthcoming elections, mount a counter-offensive the following year, using the ruling Party's commercial branch as a faqade to bypass the agreement. But the policies they were reintroducing were doomed to fail; when they did, they reinforced the President's hand, and privatization - and not only of rice - became official policy for the whole economy. 5.3 That year passed and a new year came. The 1988 harvest was lower than the 1987's as a result of both poor climatic conditions and the negative price incentives. At the end of 1988, market prices began climbing rapidly. The Government made a plea for utilizing funds from the CASA credit to import 30,000 tons of rice. IDA agreed, but conditioned its approval on a floor disposal price. A Memorandum of Understanding signed with the Minister of Agriculture in January 1989 put an end to active Bank micro-management of rice imports in Madagascar. Government stopped operating the rice buffer stock in 1990, and relies now on the private sector for rice imports needed to complement official grants (if any). No imports were effected in 1992. 5.4 As mentioned above (paras 3.7 and 4.1), the evident success of adjustment and policy reform expanded the constituency for reform. The President and the Governing Council - then including representatives of all political parties - agreed to proceed with discussions with IDA which eventually led to the next two adjustment operations (CASPIC and CASEP). 1 PROCOOPS, a cooperative wholesale supplier owned by the President's AREMA political party. (PCR, op.clt., para 6.6). -8- 5.5 Two lessons can be derived from this experience. The first is the importance of convincing the key political players of the effectiveness of adjustment as a viable policy and political path, which in turn leads to commitment and, ultimately, to ownership. The second is that adjustment processes should be sequenced so as to place the early benefits of reform in the hands of those who favor adjustment, thus reinforcing their power base within and without Government and making the next steps likelier. A critical element in designing and sequencing successful adjustment processes is therefore to start with those policy changes more likely to generate easily observable and widely benefitting effects". Adjustment. the sequencing problems from IDA's side 5.6 But adjustment is not a smooth, ballet-like sequence of well-planned, well-tuned events. As mentioned in para 1.1, the Malagasy adjustment process started tepidly in 1985 with the Industrial Assistance Credit (CASI), which basically financed studies. CASA followed in 1986, and its successful rice agreement opened the way to CASPIC (1987) and CASEP (1988), which were handled by different Bank Divisions. 5.7 Problems between the successive adjustment operations, and particularly between CASA and CASPIC, and between CASEP and several investment projects handled by the agricultural operations division, became evident early on three fronts. First, Board approval of CASPIC was concurrent with the major June 1987 devaluation (not an "official" condition for Board presentation). The devaluation resulted in a sharp increase in the domestic price of imported fertilizer, which reduced farmer demand for the fertilizers being imported under CASA. Second, CASPIC introduced a non-discretionary open general licensing procedure for allocation of foreign exchange (OGL, or in French, SILI), while foreign resources provided under CASA were allocated through a lengthy and cumbersome bidding procedure. This resulted in the slowdown of CASA disbursements, as a significant fraction of the foreign exchange demand was transferred to the OGL facility. The discriminatory competition from CASPIC contributed to a two-year delay in releasing CASA's second tranche. Disbursement mechanisms for CASA were modified to match those for CASPIC only when CASA's second tranche was released. This proved a serious misjudgment on IDA's part during the processing of CASPIC. And third, some parastatals executing agricultural investment projects were privatized or dismantled under CASEP without consultation with IDA's agricultural division. This proved seriously detrimental to the operation and sustainability of these investment projects. This third area of conflict is analyzed and documented in the performance audit report of five investment operations "; the negative effects of CASA itself on two irrigation projects has already been mentioned (para 4.2). 5.8 The lesson here is the need for a more careful coordination and coherence between successive adjustment operations, the more careful when they are handled by different divisions in IDA. Ideally, the Government and IDA should agree beforehand on a basic framework for 12 A similar effect was observed in the adjustment process in Tanzania. In that country, the "initial success" element was the expansion in availability of imported goods in general, and of "incentive goods" in the rural areas, in particular. What makes the Tanzania case even more noteworthy is that the policy change which brought this change about was entirely designed and implemented locally, without Bank input. D. Mans, The Adjustment Process in Tanzania, AF2DR, draft, March 1993. 13 Report No 12646, op. clL -9- adjustment, identifying the intended policy changes and anticipated IDA support operations and their sequencing. Similar care must be exerted on the coordination between adjustment operations and investment operations, irrespective of sector. This is particularly serious when adjustment aims at eliminating or privatizing the parastatals which were executing previous IDA- supported investment projects, or when liberalization will destroy the monopolistic rents many parastatals use to finance their business. If adjustment operations are going to affect the fundamentals (price, marketing and trade policy, institutional structures, financial flows) on which investment projects depend, explicit consideration should be given during this preparation and appraisal to revising the scope or design of the investment projects which might be affected by such changes, including their outright cancellation. The ancillary conditions and the release of the second tranche 5.9 As discussed in paras 3.8 to 3.11 above, the Government had little to show as compliance with the Credit-specific conditions for second tranche release (see Box 2), but IDA decided to release it anyway. Did IDA act irresponsibly in releasing the second tranche without solid demonstration that the Government had actually complied with the conditionality agreed? Or are there alternative ways of viewing the situation? As it happens, the release of the second tranche of the Argentina Agricultural Sector Loan had already laid the ground for just such an alternative view. 5.10 Almost exactly one year earlier, the Board had debated whether to waive the two major conditions for the release of the second tranche of the Argentina Agricultural Sector Loan". The discussion had focused on the nature of conditionality in adjustment lending and how cases of non- or partial compliance should be handled. The debate showed a diversity of opinions, but on the essentials chairs basically split into two groups. The first group considered conditionality of an individual loan or credit as definite covenants in a freely agreed, self-standing contract, whose compliance was a contract condition and without which the contract had to be canceled. They tended to see the issue of non-compliance on "project" grounds, asking themselves whether the individual operation should go forth or be canceled. 5.11 The second group, in contrast, considered adjustment to be a longer-term process, supported by IDA through a series of operations, and whose individual elements (i.e., each loan/credit agreement) should be evaluated not only in and by themselves but also as a part of such series. Thus, failure to fulfill the conditions of an individual operation had to be seen in the context of whether the overall adjustment process was on track. They tended to see the issue of non-compliance on "country" grounds, asking themselves whether the country was doing enough in adjusting itself and, if so, whether continued Bank support was warranted. The waiver for Argentina was approved. 5.12 In the case of Madagascar, the "Argentina principle" was relevant. Although the argument was not couched in these terms, it is obvious from the context and the details that the 14 OED, Performance Audit Report on Argentina Agricultural Sector Loan, Report NO 11925, of May 28, 1993. The discussion referred to in the text is at paras 5.41 and 5.42. - 10 - Box 2: Schedule 4 of the Development Credit Agreement 7Ranche Release Actions referred to in paragraph 3 (b) (ii) of Schedule I to this Agreement (a) the Borrower has taken all measures necessary to ensure that no subsidies are paid for agricultural inputs with the cu:eption of pesticides; (b) the Borrower and the Association have agreed upon an action program for the implementation of the recommendations of the edible oil study and the Borrower has taken appropriate actions therefor, (c) the Borrower has adopted a public investment program in agriculture covering the years 1987 through 1989; such program to be acceptable to the Association; (d) the Borrower has carried out, in a manner acceptable to the Association, programs relating to: (i) the promotion of the role of commercial enterprises in the Importation and marketing of agricultural inputs and veterinary products; (it) improved pricing systems for major axport crops; and (iii) the Aic management program; (e) the Borrower has furnished to the Association its comments on, and an action plan for the implementation of the recommendaions, of the follwing studies: (i) the study on Madagascar's potential and comparative advantage in coffee production; and (ii) the study on eport arp administration, and (b) the Borrower has funished to the Association the results of the following studies, including an action plan for the implementation of the recommendations thereo( (i) the study on local government fnances; and (ii) the study refarred to in Section 5.01 (c) (i) of this Agreement. (Study on the disposition of the surplus generated by aport crops, including the future of the export crop stabilization overall success of the adjustment process in Madagascar, then well advanced, had surpassed the rather timid policy changes and studies that had been agreed on as conditions for the release of the second tranche of CASA. The staff thus confidently released the second tranche, a move which in retrospect looks as well justified as when it was adopted. 5.13 It is too early to tell whether the "Argentina principle" has become the de facto Bank policy for releasing tranches in not-fully-compliant operations within an overall satisfactory adjustment process. The necessary evidence will come from performance audits on additional projects experiencing this situation. Adjustment, rice liberalization, and the poor 5.14 A special monograph on adjustment and poverty prepared by the Cornell Food and Nutrition Policy Program illuminates the effects of adjustment on poverty. A special chapter is devoted to the main staple, rice. This section draws substantially from that monograph'. 5.15 In the late seventies and early eighties, real incomes and consumption had been artificially boosted to unsustainable levels through massive foreign borrowing, monetary expansion, and fiscal 15 Dorosh et al., op. ciL, pages xviii and xx, 73-74, 83, 96-97, 100, and 135. - 11 - deficits. Some policy interventions, particularly a subsidy on rice, was directly targeted to consumption; during its peak in 1981, the rice consumer subsidy reached FMG 1&5 billion, equivalent to 15% of Government revenue. The subsidy on rice was essentially a transfer from the budget to urban consumers in the four largest cities (Antananarivo, Toamasina, Mahajanga, and Fianarantsoa). Poor urban households, for whom rice expenditure accounted for 40-50% of total expenditure, benefitted the most in relative terms; higher income households, who purchased almost 50% more subsidized rice than the poor ones, benefitted the most in absolute terms ". Consumers outside the official distribution network, (i.e., the majority of consumers outside of the four largest cities), who bought their rice on the private market, gained no direct benefit from the subsidy; consumption of officially-distributed rice by rural households was almost non-existent. Actually, households without access to official rice faced market prices higher than those which would have existed without the subsidy ". 5.16 The succession of stabilization and adjustment reduced both aggregate demand and the rice subsidy. The loss of the rice subsidy, first, and the liberalization in rice pricing and marketing which followed, sharply increased consumer prices. Open market prices for rice in 1987 were 75% higher in real terms than what the official consumer price had been in 1982. As a result of both declining incomes and higher consumer prices, from 1982/3 to 1986/7 total expenditure fell for households in all four quartiles, declining by 18% for the poorest quartile and by 28% for the richest one. Households in all four expenditure quartiles devoted higher shares of total expenditure to rice as their total expenditure fell; per capita consumption of rice fell in all income quartiles as well ". 5.17 In short, stabilization and adjustment led to a reduction in family incomes and rice consumption at all levels. In the same 1982/3 to 1986/7 period, purchases of rice from the official distribution system fell for each and all four quartiles by around the same 50%. The richest 25% of urban households compensated by trebling their market purchases and thus reducing their total rice consumption by only 7%; the poorest 25%, in contrast, were specially hard hit: their per capita consumption of rice fell by 31% even while increasing their market purchases tenfold ". 5.18 CASA, however, did not incorporate any safety net for the poorest. It is difficult to see how it could have included one. First, stabilization and adjustment were trying to reduced aggregate demand, including fiscal spending; no domestic resources would have been available for such thing. Second, the pre-adjustment situation was not a "normal" one, from which the poor were dropping, but a bloated one, which the country could no longer sustain. And third, the provision of safety nets had not yet become an issue at the time. In hindsight, it lacks now 16 In 1982/83, the poorest 25% of households in Antananarivo purchased 105 kilograms of rice per person (90% of the rice they consumed = 116 kg p.c.) through the official distribution system. The richest 25% of households purchased 150 kilograms of rice per person (equal to 75% of their rice consumption = 201 kg p.c.) through the official system. IbL page 100 and table 32. 17 "Compared with a situation where the same amount of rice imports ws sold in the open marke die rice subsidy was an implicit income transfer that increased the demand for rice by favored conwmerm As a result total demand for rice was increased and the market-clearing price was higher." Ibid, page 97. 1o Ibid, pages xviii and 135. 19 Ibid, Table 32. - 12 - obvious that some safety net tools could have been considered. Self-targeting subsidies to protect the poorest, or food distribution to infants and children through the health and school systems, could have been introduced in the areas most likely to have been hit the hardest, and private voluntary organizations and targeted donor grants could have been explored. But they were not. - 13 - ATTACHMENT 1 Page 1 of 6 PERFORMANCE AUDIT REPORT MADAGASCAR AGRICULTURAL SECTOR ADJUSTMENT CREDIT (IDA Credit 1691-MAG and SFA Credit A-16-MAG) EVALUATION SUMMARY Background 1. The Agricultural Sector Adjustment Credit (ASAC), approved in May 1986, was the second in a series of four adjustment operations extended by IDA to Madagascar since 1985. Initially, the Government and the Bank adopted a cautious strategy and agreed to begin addressing the many structural distortions affecting the economy sector by sector, first in industry and then in agriculture. After ASAC, the adjustment process gained momentum and Government moved to a broader agenda of policy reforms supported by the Industry and Trade Policy Adjustment Credit (ITPAC) in 1987 and the Public Sector Adjustment Credit (PSAC) in 1988. 2. When ASAC was prepared in 1984-85, the agricultural sector was heavily taxed both directly and indirectly through price controls and currency overvaluation, and Government intervention dominated the internal and external marketing of major agricultural products and inputs. A large share of public investment in the agricultural sector went into large-scale and capital intensive operations managed by the state. As a result of Madagascar's balance of payments difficulties, agricultural inputs as well as consumer goods were extremely scarce in rural areas. Government had started to partially liberalize the domestic marketing of rice in 1983, but poor management of the rice imports had resulted in periodic rice shortages on urban markets and extremely high seasonal price fluctuations. Credit Objectives and Design 3. ASAC's objectives were: (i) to improve the returns on scarce resources in agriculture by financing from public sources only selected high priority investments; (ii) to put into place a coherent rice strategy to restore market confidence, improve productivity and reduce pressure on the balance of payments; (iii) to provide better incentives to farmers and improve marketing networks; and (iv) to help formulate consistent strategies for the-next phase of adjustment. The package of policy reforms supported by ASAC concerned mainly the management of rice imports after full liberalization of domestic rice trade, the removal of input subsidies and transfer of input distribution to commercial operators, improvements in the public investment program for agriculture, improved pricing of edible oils and export crops, and various studies and action plans to prepare a follow-up sector adjustment credit. 4. At the identification stage, a more ambitious program was envisaged, covering four key subsectors (rice, export crops, edible oils, beef). It was later decided that ASAC should concentrate on the completion of reforms in the rice subsector as a main priority, while gradually moving into other subsectors, such as edible oils and export crops, and improving overall resource allocation in - 14 - ATTACHMENT 1 Page 2 of 6 agriculture. 5. During project preparation, the politically sensitive question of rice marketing had given rise to intense debate between Government and the Bank. Government wanted to maintain its monopoly and price controls on the purchase of rice in the two surplus areas (Lac Alaotra and Marovoay), while the Bank considered the opening of these areas to private trade, as had already been done in the rest of the country, an essential condition. Shortly before negotiations, following a meeting between the President of Madagascar and the Regional Vice-President of the Bank, it was agreed that the Bank, with the assistance of other donors, would help Government set up a buffer stock of imported rice, provided Government would open up all areas to private trade and adopt appropriate practices for the counter-seasonal release of imported rice on the domestic market. Detailed operational arrangements for the management of the buffer stock of imported rice, which became the major component of ASAC, were worked out during negotiations. 6. The total funding for ASAC of about US$60 million equivalent (including grant financing) was to be disbursed in two tranches, the first upon effectiveness and the second upon satisfactory progress in implementing the agreed package of sectoral reforms. Financing was provided for US$42 million equivalent for agricultural inputs and, if needed, rice imports to complement the contributions of other donors, US$3 million for veterinary products, US$5 million for incentive goods (i.e. consumer durables to improve production incentives in rural areas) and US$10 million for tractor and transport equipment. Credit Implementation 7. ASAC was approved on May 8, 1986, after a last minute confirmation by the President of Madagascar of the opening up of all rice producing areas to private trade. Mainly due to administrative delays, the Credit did not become effective until November 18, 1986 (a delay of four months). First tranche disbursements took thirty-two months instead of the twelve to sixteen months originally anticipated, in part due to sluggish demand for commercial imports of fertilizers and other agricultural inputs, particularly after the mid-1987 devaluation. Moreover, the two subsequent Bank supported adjustment operations, ITPAC and PSAC, made foreign exchange available to banks and importers under administrative procedures which were less restrictive and time-consuming than those of ASAC. 8. The release of the second tranche, originally expected in 1987, took place on May 1, 1989. Besides slow disbursement of the first tranche, several factors contributed to this delay, including the large devaluation of mid-1987. This devaluation deepened and strengthened the agricultural market reform process, but also affected the pace and direction of the liberalization program and required some adjustments in the specific reforms agreed under ASAC. Despite an agreement to disburse ASAC's second tranche funds on the same terms as those applying to ITPAC and PSAC, full disbursement of the second tranche was not completed until October 1990. ASAC was closed on June 30, 1990, after three extensions of the original closing date of June 30, 1988. 9. Rice management program. The essential element of the rice management program agreed under ASAC was, in addition to the full internal trade liberalization, an automatic mechanism for putting imported rice on the domestic market whenever the market price exceeded a trigger price, set during ASAC negotiations at a level providing adequate incentives for private sector purchase and storage of domestic production. It had been further agreed that the official distribution of rice (i.e. - 15 - ATTACHMENT 1 Page 3 of 6 the Government operated ration scheme that provided rice at sub-market prices to urban residents), also based on imported rice, would be gradually phased out from 80,000 tons in 1986 to 20,000 tons in 1989, and that sales under official distribution would cover the full cost of imported rice at the prevailing exchange rate, including transportation, storage and distribution costs. An annual rice import program was to be submitted by Government to the Bank for approval. 10. In general terms, the rice management program was implemented as agreed under ASAC during the first 1986-87 season. The buffer stock went into operation in November 1986 at the beginning of the pre-harvest season, by selling imported rice at a fixed price in wholesale lots. This intervention was very successful; with a relatively modest volume of sales, it held rice prices throughout the pre-harvest season at a level only slightly higher than the sales price of the buffer stock. After the painful price fluctuations of the previous year, the stable prices brought about by the buffer stock were highly appreciated in Madagascar. The full liberalization of domestic rice trade in April 1986 had also resulted in intense competition for paddy purchases, which had pushed farmgate prices up and induced farmers to increase their paddy production for the following year. As a result of this, combined with favorable weather conditions, prospects for the 1987 harvest were quite good. 11. Starting in late 1987 and until 1989, however, rice imports were used to support Government's political interests. The rules agreed under ASAC were largely circumvented, and Government went ahead with imports that were in excess of actual needs for buffer stock replenishment. A semi-private agency (PROCOPS) imported rice commercially that was sold throughout the year to urban consumers at sub-market prices. This policy, which continued throughout 1988 and 1989, disrupted domestic rice trading activities and caused operating losses to traders who passed them on to farmers in the form of lower farmgate prices, thus contributing to a slow down of the growth of rice production in subsequent years. As the Credit Agreement only imposed limits on public sector imports, the Bank protested but could not take any legal action. 12. Government intervention in the domestic rice market was significantly reduced in 1990, when the private sector accounted for the first time for a large share of rice imports and the bufferstock was abolished in 1991. Government's stated policy is now to rely mainly on the private sector for rice imports needed to complement the official grants that it may receive from external donors and, except for special emergencies or disaster areas, to allocate external rice donations to commercial operators at a price reflecting their full commercial value. Rice remains nonetheless a highly political commodity, which entails significant uncertainty for commercial operators. 13. Export crops. Under ASAC, the producer price of coffee was raised significantly in 1986 and 1987, and the special rebate which discriminated in favor of parastatal enterprises in the collection, processing and storage of coffee was eliminated. These measures did not translate into any increase in coffee production and/or exports, because producer prices in real terms and relative to rice remained unattractive, and because of the dilapidated state of the road network and the absence of significant measures to improve traders' incentives. The latter were subsequently introduced under ITPAC and PSAC, with the abolition in 1988 of the state monopoly on exports of majors crops other than vanilla (i.e. coffee, clove and pepper). However, these measures came too late, as international coffee prices collapsed following the suspension of the International Coffee Agreement in 1989. 14. Edible oils. Government had agreed in late 1985 to an increase in the ex-factory price of groundnut and other edible oils, but not to the full liberalization of edible oils prices which the Bank - 16 - ATTACHMENT 1 Page 4 of 6 thought necessary. Instead, it had been agreed under ASAC that Government would prepare and agree with the Bank on an action program for the edible oils subsector and, prior to second tranche release, take appropriate steps for its implementation. The Bank initially insisted on this action program, the preparation of which was made difficult by the unclear allocation of responsibility and insufficient collaboration between the various ministries involved (agriculture, industry, commerce). After mid-1988, the Bank decided that it was best to rely instead on the general trade liberalization, export promotion and parastatal restructuring measures supported under ITPAC and PSAC. The edible oils subsector presently remains heavily dependent on donations of unrefined oil and/or imports, and the groundnut production, which had undergone a substantial decline throughout the late 1970s and early 1980s, is still stagnant. 15. Inputs pricing and distribution.: Under ASAC, the Ministry of Agriculture (MPARA) continued to disengage itself from the direct distribution of fertilizer and other chemical inputs, and leased most of its input stores to commercial operators, including parastatals. It had been agreed that input subsidies would be reduced to a maximum of 7% of c.i.f. by the time of ASAC effectiveness, and fully eliminated by second tranche release. However, the devaluation of mid-1987 translated into a sharp increase in the border price of imported inputs, at a time when abundant supplies of fertilizer were available on the domestic market, as a result of increased aid donations (representing about 50% of total supplies). At the insistence of the MPARA, the Bank shifted its position in 1988 from a policy of full import parity pricing of donated fertilizer to a policy which would instead allow the disposal of fertilizer through an auction system that would still promote the development of commercial channels. In 1988, MPARA finally opted for an administrative allocation system of donated fertilizer to established operators, at a price 20% to 30% below import parity. This policy is still in effect. 16. Public investment program for agriculture. Seven projects of doubtful economic value included in the agricultural Public Investment Program (PIP) had been questioned by the Bank during ASAC post-appraisal, and it had been agreed that they would be reexamined. Three were rural development and/or irrigation projects which were either dropped from the PIP or reshaped in consultation with the Bank. The most contentious projects were agro-industrial investments in oil crops (oil palm, soya and copra) and cashew which were both of doubtful economic viability and inconsistent with Government's stated policy of disengagement from direct production and marketing activities. Two of them (oil palm and cashew) were new projects with an ensured source of external finance, which were executed as originally planned in the PIP, despite Bank suggestions to the contrary. 17. In 1987, several new externally-financed irrigation projects were introduced in the PIP which, as pointed out by ASAC supervision missions, were inconsistent with the strategy agreed between Government and the Bank. An agreement was reached to drop some of these projects, and reassess the others. From 1988 onwards, dialogue between Government and the Bank on the PIP has been pursued in the broader context of ITPAC and PSAC. Credit Impact 18. Most of the impact came from the policy reforms introduced under ASAC, in particular from the rice management program which was the core of this operation. During the 1986-90 period, domestic paddy production increased at an average rate of close to 3.5% per year. Although this growth was aided by favorable weather and has slowed down since then, with a resulting average - 17 - ATTACHMENT 1 Page 5 of 6 growth rate of only around 2.5% per year for the whole period 1986-92, it nonetheless represents a significant improvement over the stagnation which had prevailed throughout the 1970s and early 1980s. Rice imports, which exceeded 200,000 tons per year on average during 1980-83, have stabilized at an average level of around 70,000 tons per year over the 1987-90 period. 19. Although rice consumer prices became significantly higher in real terms with the liberalization of domestic markets, most urban consumers have benefited from the liberalization of internal rice trade and, at least during the 1986-87 pre-harvest season, from the successful operation of the buffer stock. Official rice prices were hardly enforced before their elimination, and Government's attempts to control the domestic market together with poor timing of the release of imports had translated until mid-1986 into wild fluctuations of consumer prices and periodic rice shortages which badly hurt consumers. Thereafter, rice became available in adequate quantity throughout the year, and inter- seasonal price fluctuations have remained in a reasonable range throughout the second half of the 1980s and early 1990s; these gains have, however, been partly set back from mid-1991 onwards, with the onset of political turmoil and recurring foreign exchange shortages. A significant proportion of the population, both in urban and in rural areas, nonetheless continues to live below poverty level, thus facing serious difficulties in securing adequate food supplies. Specific measures are needed to help these vulnerable groups, and the Bank has started in late 1988 to support Government's efforts to address this problem under the Economic Management and Social Action Project (EMSAP). Also, a Food Security and Nutrition Project was approved in 1993. 20. The direct and immediate impact of ASAC on agricultural production other than rice was marginal. Increases in the producer price of coffee supported by ASAC in 1986 and 1987, although substantial, did not translate into increased production and/or exports, mainly due to the dilapidated state of the transport network in coffee producing areas, the absence of significant measures to improve traders incentives and the collapse in international coffee prices. The edible oils subsector remains heavily dependent on oil donations or imports. 21. On the whole, the performance of the agricultural sector in Madagascar has shown some progress but remains nonetheless disappointing. During the 1984-90 period, the agricultural sector grew at an average rate of 2.6% annually, which compares favorably with the performance of the late 1970s and early 1980s but still remains modest and masks a wide disparity in performance among subsectors. Most of that growth originates from the fisheries and to a lesser extent the livestock sectors, which together grew on average 4.9% annually between 1984 and 1989. By contrast, the performance of the crop sub-sector has been much more modest, with an average growth of only 1.5% per year during the 1984-89 period, which is well below the 3% population growth rate. Within the crop subsector, most of the growth has come from rice and to a lesser extent from other foodcrops. Traditional exports crops (coffee, vanilla, cloves), faced with increased international competition, have stagnated or regressed, and this has not been compensated for by the emergence of alternative export crops. 22. In order to sustain and expand agricultural growth, improvements will be needed simultaneously on several of these fronts. Appropriate and consistently followed macro-economic policies (exchange rate management, tariff and trade policies, fiscal regime, etc.) will also be essential for preserving and building upon the gains which have been made under ASAC and other adjustment operations. Political stability and consensus is an obvious pre-requisite for achieving this, together with well conceived and coherent support from external aid agencies. - 18 - ATrACHMENT 1 Page 6 of 6 Conclusions and Main Lessons 23. The main lessons learnt from the experience gained under ASAC can be summarized as follows: (a) the adoption of a phased and gradual approach to sectoral adjustment, as has been the case under ASAC, has its merits but nonetheless results in substantial foregone benefits; in Madagascar, the external liberalization of major export crops other than vanilla (i.e. coffee, clove, pepper) occurred only at a late stage in the adjustment process, at a time when world coffee prices had already collapsed to a level that considerably reduced private sector incentives to engage in this sector; (b) the Bank should be firmer in reacting to serious lapses in program implementation. Government's renewed intervention in the rice market in the 1987-89 period and its going ahead with investments outside the agreed upon PIP are cases in point; (c) the trends in paddy production since 1986, point out the benefits which can be drawn from a sustained policy of minimal market intervention; (d) special programs of targeted food subsidies should have been put in place simultaneously with the establishment of the buffer stock; this would have helped ease the transition for the poorest segments of the urban population, particularly after the mid-1987 devaluation; (e) greater coherence between the strategies followed by Government and the various external donors or lending agencies is needed in order to improve the PIP process; and (f) in the case of commodities for which grants in kind represent a major share of domestic supplies, as was the case under ASAC for fertilizers and edible oils, pricing policies need to be worked out carefully, taking into account the potentially distortional effects of this type of aid on prices and the absorptive capacity of the market. - 19 - ATTACHMENT 2 Translation REPUBLIC OF MADAGASCAR Tanindrazana-Fahafahana-Fahamarinana Ministry of Agriculture Antananarivo, March 28, 1994 and Rural Development General Secretariat The Minister of Agriculture and Rural Development to Chief of the Agriculture and Human Development Division, Operations Evaluation Department 1818 H Street, N.W.Washington, DC, 20433 REFERENCE: Performance Audit Report Agricultural Sector Adjustment Credit (IDA Credit 1691-MAG and SFA Credit A-16 MAG) Your letter dated February 8, 1994. Sir, In response to your letter dated February 8, 1994, I have the honor to let you know that the above mentioned Performance Audit Report has received the full attention of our Department. Reviewing the Report allowed us to appreciate the impact of the Project in the rural development sector. The experience acquired during project implementation has enabled a better orientation of Ministry's new policy e.g. in the areas of liberalization and marketing of agricultural products. To sum up, I would like to confirm that my Department does not have any additional observations on the Report. Sincerely, For the Minister of Agriculture and Rural Development Charles D. Basendratsirofo Secretary General    IBRD 20035R 411.• 12· 12' Antsiranana MADAGASCAR R P.red roads fl.tes g~ee~ndes A »-ete ods äpeopj ecess/b/es en tonse s~sso. NOS-DE Amb Iob. -- Non-permanent road, himorino~ Nte-; se/som nades Ckom. l fer YAania Aira b nort m ., Antortkh Natienol capiteo , de, puft Moholnga PC0 BM-rs1.on Biottirao Mo,orooy Md.pikony Mor Anotebo~ny I MaeotoenTsa toor ,Soonronro-ivongo Anddomno 4..bdiF*totd o Andr-mon* F.oovo Arono Maatorofenobe Main,i,mo /gmbotondroz a aosoroAn .Toamasino ...t An n Ivv TsronomW=mond.d i Ml>Whib'~ny o O~m.ongo iM +,Ma~ S vi rAndr saVFon B-lo-TsribAn Mfond-oo F ror Aca Vnka Antsirabg 4 M oonoro Morondono Fonde ao 4--o1 A.nbolofinondrobono Tnbos ra oyVonks Ambohimanhosoo Mananjary Fionarantsoa Mo- . ongo 4. / A,r,haloo Ankok.bo Vokipaffb KILOMETERS 0 s0 100 150 200 Von s 0 50 :00 Sökorehe o vondio Farofogno Toliary Vrp.~ro Toiroky - Sud OM.dgy / ZAII,. TANZANIA Alneiena, ej=ae urZAMB ©°"""'00"B.r O$* Moan04n-n Bök.l.. enZAMB> gs. ZIMBABWIEr Antonononod enr ee mes t nt meson Ambo0ser F-m.er.n.cpon _~ ýp lagnaro .>- . 0" - d'50V 0" ~ .-SOUTH " c rAFNR ICA1 JUNE 1990