Argentina - Agricultural Sector Loan Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 11925 PERFORMANCE AUDIT REPORT ARGENTINA AGRICULTURAL SECTOR LOAN (LOAN 2675-AR) MAY 28, 1993 Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. CURRENCY UNIT - AUSTRAL (A) Rate at Appraisal: US$1 = A 1.0 Rate at Completion: US$1 = A 261.0 ARGENTINE REPUBLIC FISCAL YEAR January 1 - December 31 GLOSSARY OF ABBREVIATIONS AGR Agriculture and Rural Development Department BCRA Central Bank of Argentina BNA Banco de la Naci6n Argentina CEM Country Economic Memorandum GOA Government of Argentina INTA National Institute of Agricultural Technology ME Ministry of Economy PRONAGRO National Agricultural Development Program SAGyP Secretariat of Agriculture, Livestock and Fisheries SOE Statement of Expenditure UNDP United Nations Development Programme FOR OFFICIAL USE ONLY THE WORLD BANK Washington, D.C. 20433 U.S.A. Office of Director-General Operations Evaluation May 28, 1993 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Performance Audit Report on Argentina Agricultural Sector Loan (Loan 2675-AR) Attached is the Performance Audit Report on Argentina Agricultural Sector Loan (Loan 2675-AR) prepared by the Operations Evaluation Department. The audit concurs with the conclusion of the Project Completion Report that the overall assessment of the Loan is unsatisfactory on the grounds that if failed to introduce the agreed policy changes and did not generate the anticipated production response. But the audit puts the "unsatisfactory" rating in context by noting that the loan's resource transfer objectives helped to shore up the country's external accounts until the crisis was overcome in 1990, and stressing the achievements reached under the technical assistance component and their contribution to the current lending program. The audit attributes the Loan's failure in introducing policy changes to deficiencies in Loan identification and appraisal, particularly the failure to appreciate the lack of ownership and commitment outside the Ministry of Agriculture. 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. г � , ь FOR OFFICIAL USE ONLY PERFORMANCE AUDIT REPORT ARGENTINA AGRICULTURAL SECTOR LOAN (LOAN 2675-AR) TABLE OF CONTENTS PaLe No. PREFA CE ................................................ BASIC DATA SHEET ...................................... EVALUATION SUMMARY ................................. v 1. BACKGROUND ........................................... I The Agricultural Sector ...................................... 1 Agricultural Institutions and the Policy Relationship ................. 2 Bank Group Operations in Argentina ............................ 3 The Project Completion Report ................................ 3 Ii. IDENTIFICATION, PREPARATION AND APPRAISAL ........... 4 Introduction ............................................... 4 Loan Preparation and Appraisal ................................. 5 Loan Size ................................................. 7 Negotiations and Approval .................................... 8 111. IMPLEMENTATION ....................................... 9 Release of the Second Tranche ................................ 10 Disbursement, Procurement and Audit ........................... 11 Compliance with Loan Conditions .............................. 11 Institutional Performance ..................................... 11 IV . IM PA CT ................................................. 12 V. ISSUES AND LESSONS ..................................... 14 The Origins of the Loan ...................................... 14 Weaknesses of the Loan's Policy Package ....................... 19 Events Surrounding the Second Tranche Release ................... 21 VI. M AIN LESSONS ........................................... 24 This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. -2- FIGURES 1. Export Taxes on Grains as of June Each Year .............. ...... 27 2. Argentina: Current Land Tax Rate Per Province (%)................ 29 TABLES 1. Argentina: Structural Adjustment Lending ................. ....... 31 2. Argentina Agricultural Sector Loan (Loan 2675-AR)................. 33 PERFORMANCE AUDIT REPORT ARGENTINA AGRICULTURAL SECTOR LOAN (LOAN 2675-AR) PREFACE This is the Performance Audit Report (PAR) on the Argentina Agricultural Sector Loan. Loan 2675-AR, for US$350 million equivalent, approved on April 3, 1986; signed on June 5, 1986; and declared effective on July 22, 1986. The second tranche of USD 170.5m was released on June 2, 1988, 18 months later than anticipated. The original Closing Date of June 30, 1988 was extended to June 30, 1989 to permit completion of the national cadastre funded under the Loan. The final disbursement was made on June 23, 1989 and an amount of US$0.04 million equivalent was canceled. The PAR is based on the Project Completion Report prepared by the Latin America and the Caribbean Regional Office-; the President's Reportv; the Loan Agreement; supervision reports, correspondence between the Bank and the Borrower, consultant reports and internal Bank studies, memoranda and other materials in the Bank files; interviews with Government officials and Bank staff; and transcripts of the relevant Board meetings. An OED mission visited Argentina in November 1992. The mission discussed project genesis, implementation and impact with current and former officials in the Ministry of Economy; the Secretariat of Agriculture, Livestock and Fisheries; the Secretariat of Public Revenue (Treasury); Central Bank; and National Institute of Agricultural Technology (INTA). It also interviewed the President of the largest farmer/landowner association (Sociedad Rural), and Secretaries, Directors and staff of the Government of the Buenos Aires Province. Their kind cooperation and valuable assistance in elucidating some of the complex issues associated to the loan is gratefully acknowledged. The Project Completion Report is of outstanding quality, candid and informative. It is exhaustive in analysis and discussion of the background, justification and effect of the ASAL, and critical of the Region's handling of both appraisal and implementation. The PAR builds on its findings and explores the wider context of the loan, in terms of Government policy and internal Bank dynamics. It addresses the antecedents of loan design and administration, and identifies the contradictions and shortcomings which eventually doomed the loan. The PCR assessed the outcome of the loan as unsatisfactory on the grounds that it failed to introduce the required policy changes and it did not generate the expected production response. The audit confirms this assessment, but qualifies the "unsatisfactory" rating by noting that the resource transfer objectives, underrated by the PCR, proved important in facilitating the transition in the country's external accounts until the economic crisis was overcome in 1990. The audit also elucidates the events surrounding the release of the second tranche. The audit confirms, and amplifies, the 1/ Project Completion Report: Agricultural Sector Loan (Ln. 2675-AR), Report No. 9913, September 26, 1991. 2/ President's Report: Agricultural Sector Loan (Ln. 2675-AR), Report No. P-4161-AR, March 14, 1986. - 11 - achievements reached under the technical assistance component of the loan, and their subsequent contribution to the lending program. Copies of the draft PAR were sent to the Borrower and its executing agencies, as well as to the International Monetary Fund for comments but none were received. - iii - PERFORMANCE AUDIT REPORT ARGENTINA AGRICULTURAL SECTOR LOAN (LOAN 2675-AR) BASIC DATA SHEET KEY PROJECT DATA Original Disbursed Cancelled Repaid Outstanding Loan 2675-AR US$350M US$349.96M 0.04M 29.17M 320.79M Original Loan Dates Initiating Memorandum May 13, 1985 Letter of Development Policy February 14, 1986 Negotiations December, 1985 Board Approval April 3, 1986 Loan Agreement June 5, 1986 Effectiveness May 1986 July 22, 1986 First Tranche Release Prior June 30, 1986 September 24, 1986 Second Tranche Release January 1987 June 2, 1988 Loan Closing June 30, 1988 June 30, 1989 Actual Completion June 30, 1988 June 30, 1989 CUMULATIVE LOAN DISBURSEMENT US$MILLION 1986 1987 1988 1989 (a) Planned Policy Component 170.5 341.0 - Studies/Institutional Support 1.5 5.9 9.0 Total 172.0 346.9 350.0 (b) Actual Policy Component - 170.50 341.00 Studies/Institutional Support 7.3 6.80 8.96 Total 170.83 347.80 349.96 - iv - MISSION DATA Month/Year No. of No. of Manweeks Date of Specialization Performance Type of Weeks Persons Report Representedl/ Rating / Problems3/ Exploration 07/84 1 2 2 08/10/84 E,E - - Preparation 11/84 2 3 6 11/30/84 E,E,A - - Appraisal 06/85 3 6 18 07/12/85 E,E,E,F.F.A - - Post-Appraisal 08/85 1 4 4 09/18/85 E,A,F,R - - Supervision4/ 05/86 2 1 2 07/)9/86 E 1 - Supervision 10/86 1 1 1 12/12/86 E 1 F Supervision5/ 11/86 1 1 1 12/05/86 E - - STAFF INPUTS (Staff weeks) FY 85 86 87 88 89 90 91 Total Pre-appraisal 42.3 42.3 Appraisal 13.1 17.0 30.1 Negotiation 3.0 3.0 Supervision 3.5 8.5 33.4 .7 14.3 19.1 79.5 Other 10.6 16.6 .3 27.5 Total 66.0 40.1 8.5 33.7 .7 14.3 19.1 182.4 / A= Agriculturalist; E= Economics; F= Finance Specialist; R= Resources Planning Specialist. 2/ 1= Problem-free or minor problems; 2= Moderate problems; 3= Major problems. 3/ F- Financial. 4/ Pre-effectiveness. 5/ No Form 590 or BTO prepared. Supervision of ASL was part of a multi-tasked mission. PERFORMANCE AUDIT REPORT ARGENTINA AGRICULTURAL SECTOR LOAN (LOAN 2675-AR) EVALUATION SUMMARY Introduction 1. In December 1983, a democratic Gov- for the land tax, complementary studies and ernment took office after 8 years of military rule. institutional support. Several other adjustment The economy was in disarray and foreign com- loans have followed it. mercial bank lending had ceased. In December 1984, the Fund approved a Standby Arrange- 4. Preparation was inadequate. It was ment; in June 1985, the Government launched effected very rapidly (in slightly over one year the Austral Plan, which delivered stability for through appraisal) with only one preparation more than a year but eventually collapsed. It mission. Major aspects were not properly ana- was replaced by the Spring Plan of August 1988, lyzed, namely, (i) the overall tax regime and the which in turn lasted only a few months. changes being introduced by the Government; (ii) the changes in the structure and subjects of 2. Agriculture is crucial for Argentina's sectoral taxation as a result of the proposed tax development, contributing some 15% of GDP, swap; (iii) the price and income distribution 75% of foreign exchange earnings, and 14-20% effects of the reduction in export taxes; and (iv) of federal tax revenues. Through taxes on the institutional arrangements and disbursement commodity exports and input imports, agriculture procedures. Neither farmers and landowners production was discouraged and faced distorted nor the Treasury were given guarantees that incentives. Use of yield-augmenting inputs was their situation would not be negatively affected low, and the sector held significant unexploited by the agreed tax swap. The loan's tax policy potential. The negative protection was designed package run counter to the fiscal measures being to satisfy non-agricultural objectives, i.e., to keep introduced under the Fund's guidance. The food prices low and to protect industry. political viability of the proposed federal land tax was not assessed at appraisal. Objectives 5. Appraisal faced multiple problems, 3. The Agricultural Sector Loan (ASAL) which should have alerted the Bank about the was the first adjustment operation in Argentina, poor preparation and insufficient Government and the first operation in the agricultural sector ownership. Indications that the loan concept in eight years. Its main objectives were: (i) to was unviable and might not work, including a reduce export taxes on grains and oilseeds; (ii) negative reaction from the Legal Department, to replace lost revenue through the introduction were ignored. of a federal land tax; (iii) to modify regulations and tariffs on input import; and (iv) to introduce 6. The loan included the enactment of fiscal measures to compensate for foregone fiscal legislation (the proposed federal land tax act) as revenues. The USD 350m loan would be dis- a condition for second tranche release, but the bursed against general imports. The ASAL also Bank noted that Government was confident of financed the preparation of the technical bases getting legislative approval. The Bank does not -vi - normally include completion of anticipated agro-chemicals, instead of removing their 10% legislative action as a condition for tranche import surcharge, Government increased it to release, for the Executive Branch of many 15% as part of Fund conditionality. Another governments cannot guarantee the delivery of condition for second tranche release had become decisions of the Legislative Branch. unattainable. 7. The Bank attempted to determine the 12. Release of the second tranche had size of the loan on the basis of the fiscal cost of originally been anticipated for January 1987. adjustment. The loan size was set at USD 200m, With two major conditions now out of reach, similar to the estimated transient fiscal revenue such release became a complex process. Mean- shortfall. The Fund, however, rejected the while, the second tranche funds had become notion of substituting borrowed funds for fiscal linked to the financial package agreed between revenue, even in an interim period. The loan the Government and the commercial banks; and size was subsequently raised to USD 300m, on disbursements from the commercial banks be- portfolio grounds, and then to USD 350m, on came linked to specified levels of Bank disburse- consideration of foreign exchange needs. ments. Pressure from several quarters to release the tranche rose. Implementation Experience 13. In late 1987, the Bank accepted a 8. The Loan was approved in April 1986. package of fiscal measures recently submitted to It took almost six months to sort out disburse- Congress as a substitute for the land tax and for ment procedures and the role of the Central revenue foregone due to export tax reductions Bank. The Loan was then declared effective, and Management recommended a waiver of prompting the release of the first tranche. coiiditionality to the Board. This was the first time the Board was involved in a tranche release 9. Legislation to establish the federal land decision. tax stalled in Congress. The fall in international commodity prices sapped the Government's 14. After three hours of acrimonious de- resolve to push for the draft legislation. Thus, bate, the waiver was approved, but the high rate one critical condition for second tranche release of abstention (6 chairs, with 40% of voting could not be met. power) suggests that cancellation of the loan might have been considered. The debate cen- 10. In 1986 and 1987, reductions in tax rates tered on whether an adjustment operation on grain and oilseeds exports had exceeded should be analyzed as a single contract (and those required under the loan agreement. Such remedies taken if loan conditions were not reductions, however, had been prompted by the satisfied) or as a part of a series of operations fall in international prices, and not by a commit- (and consideration given to whether the country ment to the loan's policy package. After the was doing enough to adjust itself) in deciding on Spring Plan, the process was reversed, and possible remedies. The release had secured the before loan closing grain export taxes were resource transfer objective while abandoning a raised to levels even higher than those prevailing major policy reform objective. before loan processing (Figure 1). 15. Procurement was straightforward, but 11. Initial moves to liberalize imports of disbursements were delayed. Audit compliance tractors and machinery were effected, but com- was poor, but eventually clean audits were pletion of the process was delayed by the Gover- received after intense Bank support. As men- nment's reluctance to act in the face of de- tioned above, compliance with loan conditionali- pressed conditions in the domestic industry. On ty was mixed. - VHi - 16. The preparation of the technical bases raw hides. There is no decision to suppress of the federal land tax, as well as the five studies them, and no attempt has been made to amend and two institutional support programs, in con- the Customs Law. The current Government is trast, were successfully completed, with funds putting in place a single federal tax system based administered by the UNDP. The Closing Date on efficient taxes, eliminating sector-specific of the loan was extended by one year to permit taxes. Neither grain export taxes nor land taxes completion of the technical bases of the land tax. was ever mentioned again in the Bank-Govern- ment dialogue. Results Findings and Lessons 17. Achievements under the loan were mixed: (i) albeit the second tranche was released 21. The wisdom of having ever made the 18 months later than originally planned, the ASAL has often been questioned. However, intended resource transfers were effected; (ii) even in retrospect, there were good reasons to the technical studies were effectively carried out support the operation given the potential of and have proven useful in several respects; and agriculture, the attractiveness of the policy (iii) the original policy package was either not package, the resumption of lending to Argenti- implemented or reversed before the loan's na, and the quality of the Borrower's economic Closing Date. Since structural changes are at and agricultural teams. the essence of an adjustment operation, the loan in the aggregate must be deemed a failure. 22. A corollary question is whether it was judicious to start adjusting such a distorted 18. Argentina was short of foreign resources economy through the "back door" rather than throughout the loan's disbursement period; the with a SAL. But at the time, the question never disbursement of ASAL's second tranche was a arose. Adjustment was a new lending instru- major element in securing a steady flow of new ment, and the wisdom of having a macroeco- money in 1988 and helped to shore up Argenti- nomic program in place before having a SECAL na's external accounts until the crisis was over- only evolved later. The Bank did not have a come in 1990. Importation of tractors and full-fledged strategy for adjustment in Argentina. machinery was liberalized as agreed. The techni- On the Government's side, there was no capaci- cal studies were properly implemented and made ty or political support to proceed with a full- important contributions to the policy dialogue fledged SAL. So, addressing major sector distor- with Argentina and to the Bank's lending pro- tions appeared as a plausible way to start the gram. adjustment process in Argentina. On the other hand, macro-economic variables had not been 19. In contrast, export taxes, which were stabilized and a resilient framework was not in initially reduced to the agreed levels and beyond, place. To a certain extent, the failure of the were later raised to levels higher than those ASAL must be placed in the context of the prevailing before loan processing. And, as collapse of the whole stabilization program. mentioned above, the federal land tax was not introduced. 23. With hindsight, most of the loan's problems can be attributed to inadequate identi- Sustainability fication, preparation and appraisal. In turn, these can be mostly attributed to two causal 20. The loan's policy package faltered and factors, one on the Bank's side and one on the was never revived. Grain export taxes remain on Borrower's side. On the Bank's side, the onus the books. Their level is currently zero for most of handling the first adjustment operation ever grains, but taxes remain in effect for oilseeds and in Argentina was put on the Agriculture Divi- - Viii - sion. This was a poor decision, for sector divi- farmer and landowner associations or the Con- sions lack the economy-wide overview that gress. The Bank required that a draft land tax adjustment operations demand. The failure of act be sent to Congress before Board presenta- the Programs Department to provide such tion, but decided against appraising the political support during identification and preparation viability of the package. (iii) The loan was to may have been due to the focus of economic effect a tax swap, whereby the new federal land work on growth rather than on adjustment, and tax would substitute for the export taxes. But to a temporary shortage of specialized staff. such a swap had not been negotiated with the These problems were later addressed by the two other interested parties (Treasury, and the 1987 reorganization, when program and sector farmers and landowners), nor were these offered divisions were put under the same Department; guarantees that adequate revenue would contin- most country operations divisions carry now the ue flowing in, and that the export taxes would main responsibility for macro-economic aspects not be reinstated at a later date, respectively. of SECALs. On the Government's side, no This should no longer be a problem; current consensus existed on the tax swap issue. The appraisal procedures require an analysis of Ministry of Economy did not get into the details whether policy change proposals are politically of the work being done by the Secretariat for viable. Agriculture. Although weekly briefed by the Secretary, senior staff in the Ministry did not 25. Since the loan was among the early crop alert the Secretary of Finance of the tax swap of sector adjustment lending, many of the les- proposed under the loan, nor did they alert the sons it yields have already been internalized in Secretary of Agriculture of the diverging objec- the Bank and are now covered by revised proce- tives of the land tax and the on-going tax reform. dures. The most important lessons are the need to confirm Government ownership and assess political viability of the agreed policy packages; 24. The loan's policy package had important to have a clear understanding with the Fund on deficiencies: (i) The reduction in export taxes the overall direction of adjustment and particular addressed only part of the issue, for such taxes tools to be used; to have a stabilization program had been used interchangeably with changes in before proceeding with sector adjustment; to the nominal exchange rate, but the Bank has no have clearer rules on administration of condi- mandate on the latter. The Bank's tax prescrip- tionality in adjustment operations; and to avoid tion ran counter to the Fund's, whose main defining results of legislative action as tranche concern at the time was reduction of the fiscal condition. Further, the Bank should not accept deficit. (ii) The Government believed it had "passive" cross-conditionality on its disburse- the political strength and proper design to ments; and should heed more carefully the introduce a federal land tax (which past Govern- available indicators or warnings that selected ments had failed to do), but did not convince the policy packages may not be feasible. PERFORMANCE AUDIT REPORT ARGENTINA AGRICULTURAL SECTOR LOAN (LOAN 2675-AR) I. BACKGROUND 1.1 Argentina was one of the richest countries in the world at the onset of the century, but by 1980 its per capita income had fallen far below that of many countries with similar resource endowment. The regressive process started in the 1930s, when the Depression curtailed agricultural exports. From then on, development was based on an import substitution strategy. Industrialization was financed largely by transfers from agriculture. Governments alternatively emphasized autonomous development or a market-oriented open economy, and these reversals damaged the productive and financial structure. Five changes of leadership between 1979 and 1983 and the unstable international environment eventually led to the cessation of foreign commercial bank lending. 1.2 President Alfonsfn took office in December 1983, ending eight years of military rule. In December 1984, the Fund approved a 15-month Standby Arrangement to lower the rate of inflation and to enable Argentina to meet its external obligations while initiating domestic recovery. To widen and deepen the change, the Government introduced the Austral Plan in June 1985, supported by a revised arrangement with the Fund and commercial bank creditors. In addition, the Government formulated a medium-term growth strategy for 1985-1989, based on improving investment climate and restoring export incentives. The overall program received wide political and social support. 1.3 The Plan Austral, which delivered stability for more than one year, ran into serious problems in August 1986 and eventually collapsed. It was replaced by the Plan Primavera (Spring Plan) of August 1988, which in turn lasted only a few months. Its collapse in March 1989, shortly before the end of the Alfonsin Government, led to the dismissal of the economic team which had been in place through most of the loan's life. The Agricultural Sector 1.4 Agriculture and agro-industry have been crucial in Argentina's economic development. Crop production grew rapidly from the early 1970s. Primary exports rose from 70% of total exports in the late 1970s to above 80% in 1983/84. Argentina was, at that time, supplying around 12% of annual world exports of maize and sorghum and about 5% of wheat, ranking respectively second and fifth among world suppliers. 1.5 At the time of loan preparation (1984/5), the agricultural sector was contributing about 15% of GDP, 17% of employment, 75% of foreign exchange earnings and between 14% and 20% of the Federal Government's tax revenues. Cereals, cultivated under rainfed conditions and with little use of fertilizer, were the main crop activity. The Bank's 1984 Country Economic Memorandum (CEM; Report 4979, June 1984) found there was substantial scope for greater use of fertilizer and other inputs, and for investment in infrastructure and equipment. -2- 1.6 Agriculture bore the brunt of the import substitution strategy. Agricultural exports received a lower effective exchange rate than other exports, received prices substantially reduced by export taxesy, and were charged inflated prices for their inputs. Price discrimination focussed primarily on the traditional agricultural exports, but its consequences affected the entire sector. The combination of direct and indirect nominal protection was strongly negative for the period 1960-1985: agriculture was discriminated against by between 40-50%. 1.7 These discriminatory policies were meant to satisfy non-agricultural objectives: (a) maintaining low food prices in urban markets; (b) protecting domestic industry; and (c) raising fiscal revenues. Between 1981 and 1983, export taxes rose steadily. They collected about US$1 billion per year in 1983 and 1984 and around US$1.3 billion in 1985 (some 1.8% of the GDP). By 1983 they represented about 20% of Federal fiscal revenues, being the third most important source of revenue after value-added and fuel taxes. In 1985, most agricultural exports were taxed at rates up to 25%. Soybeans, maize and wheat accounted for over 90% of export tax revenues. Buenos Aires, Santa F6 and C6rdoba provinces contributed over 75% of those revenues. 1.8 In Argentina, export taxes have some desirable characteristics: (a) they are easy to collect; (b) tax rates are set by the Executive without congressional approval, and (c) they are politically attractive in reducing consumer prices and protecting processors of agricultural raw materials. On the negative side, they are an impediment to agricultural growth because they: (a) penalize productive investment; (b) create unfavorable input/output price relationships; (c) cause intra-sectoral distortions in resource allocation; and (d) cause domestic resource allocation to be out of alignment with international prices, and hinder adjustments in response to changes in market signals. 1.9 Agricultural investment and exports had also been discouraged by an over-valued currency. Usually, export taxes had been used to soften the impact of changes in the nominal exchange rate and to capture windfall benefits associated to nominal devaluations. They were lowered when the currency was over-valued, and increased following nominal devaluations. 1.10 Argentine farmers were discouraged from adopting high-yield technologies associated with the use of fertilizers, agrochemicals and machinery. The combination of a negative protection rate for agriculture and high positive protection for industry rendered most inputs and machinery too expensive for farmers to use in adequate quantities. As a result of the high cost imposed by industrial protection, Argentina is a striking example of a relatively advanced agricul- tural producer with a generally low level of agricultural input usage. Agricultural Institutions and the Policy Relationship 1.11 The Secretariat for Agriculture, Livestock and Fisheries (SAGyP) is under the Ministry of Economy (ME) and focuses mainly on technical issues. On policy matters, it normally acts as technical adviser rather than policy-maker. As a result, sectoral policies have often been subordinated to macroeconomic, industrial and trade policies, and to short-term adjustment needs. Other Secretariats under ME are Treasury, Industry and Commerce, and Public Revenues. 1/ Called retenciones in Argentina. -3- Frequent turn-over of the Secretaries of Agriculture had led to discontinuities and even inconsistency in agricultural policy, which contrasted with the constancy of export taxes. Bank Group Operations In Argentina 1.12 Operations in the agricultural sector had been limited to three projects, namely: the Balcarce Livestock Development Project (Ln. 505-AR, of 1967), designed to encourage the adoption of new technology in pasture production and animal health; the Agricultural Credit Project (Ln. 1564-AR, of 1978), to provide medium and long-term credit for on-farm investments; and the Grain Storage Project (Ln. 1521-AR, approved in 1978 and reformulated in 1983) to expand national grain storage capacity . Between 1979 and 1985, lending comprised ten non- agricultural loans totalling US$1.045 billion, on infrastructure projects and credit for the industrial and hydrocarbon sectors. 1.13 The ASAL was the first of a series of adjustment operations in Argentina. It was followed by the (First) Trade Policy and Export Diversification Loan (Ln. 2815-AR, of May 1987, for USD 500m, fully disbursed); the Banking Sector Loan (Ln. 2923, of March 1988, for USD 400m; the Loan Agreement was never signed); the Second Trade Policy Loan (Ln. 2996, of October 1988, for USD 300m, fully disbursed); the Provincial Development Loan (Ln. 3280-AR, of December 1990, for USD 200m); the Public Enterprise Reform Adjustment Loan (Ln. 3291- AR, of February 1991, for USD 300m); and the Public Sector Reform Loan (Ln. 3394-AR, of July 1991, for USD 325m). For the last three, only the first tranche has been released. These adjustment loans were accompanied by a series of technical assistance loans, including the Public Sector Management Loan (Ln. 2712-AR, of June 1986, for USD 18.5m; fully disbursed, which included a component on tax policy); and the Tax Administration Technical Assistance Loan (Ln. 3015-AR, of January 1989, for USD 6.5m); the Public Enterprise Reform Technical Assistance Loan (Ln. 3292-AR, of February 1991, for USD 23m); the Public Sector Reform Technical Assistance Loan (Ln. 3362-AR, of June 1991, for USD 23m); and the Second Tax Administration (Ln. 3460-AR, of April 1992, for 20m), all four under implementation (Table 1). 1.14 The ASAL permitted rebuilding the agricultural sector project pipeline and, thus, it gave origin to the Agricultural Credit II Project (Ln. 2970-AR, of June 1988, for USD 106.5m) and the Agricultural Services and Institutional Development Project (Ln. 3297-AR, of February 1991, for USD 33.5m), both under implementation. The Project Completion Report 1.15 A PCR for this project was issued on September 26, 1991 (Report No 9913). The review carried out by OED found the PCR to be of outstanding quality, candid and informative. It contains detailed analysis of the background, justification and effect of the ASAL, and is critical of the Region's handling of both appraisal and implementation. A reading of the PCR thus provides a wealth of details on the Loan. The first three sections of the audit make extensive use of the PCR. 2/ All three are now completed. PARs were issued on the first two (OED Reports No 3729, and 3699, respectively). A PCR was issued for the third one (Report No 9346). -4- 1.16 But the PCR, by its very nature, focuses on the loan itself, and on its achievements and shortcomings in terms of the policy package it supported. In doing so, it leaves the reader with major questions, such as why did the Government and the Bank get into a non-starter position in the first place, and why did Bank staff recommend release of the loan's second tranche. The audit had to explore the wider context of the loan, both in terms of Government policy development and of internal Bank dynamics, particularly on adjustment lending. As a result, the audit describes the rationale for the loan and identifies more bluntly the internal Government contradictions and Bank processing shortcomings which eventually doomed it. In terms of impact, the PCR assesses the loan as unsatisfactory as it failed to introduce the required policy changes and generate the expected production reaction. While the audit confirms this assessment, it puts the "unsatisfactory" rating in context by noting that the loan's resource transfer objectives, underrated by the PCR, proved important and shored up the country at a difficult time, thus facilitating its transition towards financial and economic success. The audit also discusses in detail the events surrounding the release of the second tranche and the conditionality management policy and cofinancing policy issues it raised. Finally, the audit stresses the achievements under the technical assistance component of the loan and their contribution to the current lending program. II. IDENTIFICATION, PREPARATION AND APPRAISAL Introduction 2.1 Several considerations spurred development of this loan. First, policy-based lending was a rather new lending instrument in the Bank; the dual possibilities of rapid disbursements supporting major policy change while alleviating acute foreign exchange shortages made such lending attractive. Second, a new democracy in Argentina, plus the Baker Plan announced in October, 1985 to assist debtor nations develop their way out of debt, lent added impetus. Third, lending in Argentina had been modest and sporadic and the Region wanted to develop a lending pipeline; agriculture seemed to offer the readiest opportunity. And fourth, the 1984 CEM, which had recommended policy changes similar to those proposed by the SAGyP, had been well received by the Government. To a significant extent, the CEM's chapter on agriculture set the policy issue and loan concept, and largely became its blueprint and justification. 2.2 The 1984 CEM had identified three policy areas as crucial to economic stabilization: (a) the public sector deficit, (b) export promotion; and (c) weaknesses in the financial system. It recommended gradual replacement of agricultural export taxes and import tariffs on inputs by other revenue measures, preferably a federal land tax. But the CEM made no mention of the fact that the Government regularly made interchangeable use of export taxes and the exchange rate as tools to transfer agricultural income to other sectors, nor of the central significance of an adequate exchange rate policy to an export-dependent agricultural sector. 2.3 The Constitution reserves the right to tax land for the provinces. Nevertheless, various efforts had been made in earlier times to introduce a federal land tax, or an equivalent tax, or to introduce it disguised as a tax based on the worth of land or the stream of benefits flowing therefrom. The analytical basis for a federal land tax dates to the early 1960s. A draft -5- law of 1964 sought to reform farm taxation, but was not approved. In 1969, efforts were made to introduce a proportional land tax but this system was abandoned. In 1973, the Congress approved a tax on the potential net income of land but it was never implemented. A similar draft law in 1974 did not proceed. In general, provincial land tax administration was poor, collections low, and evasion widespread; only in three provinces were land taxes an important source of revenue. Loan Preparation and Appraisal 2.4 An exploratory mission in July 1984 identified the possibility of a fast-disbursing operation with policy conditionality to be aimed at promoting increased production and exports, and disbursements to finance imports of yield-augmenting inputs, particularly fertilizers and agrochemicals. (The main events in the Loan's history, plus correlated events in Argentina and elsewhere in the Bank and in the Fund, are summarily presented in Table 2). In September 1984, during the Annual Meetings, Government officials asked the Bank to finance a short-term agricultural recovery program, including measures to reduce distortions. Noting the Government's "long-term commitment" to eliminating export taxes on grains, the Issues Paper (November, 1984) outlined an operation based on export tax substitution and studies to guide medium-term policy and institutional reform. 2.5 In October 1984 and complementing the Austral Plan, SAGyP released a draft of the National Agricultural Development Program (PRONAGRO), whose main sectoral objective was increasing production, plus some consideration to regional development and natural resource conservation. It included proposals for the introduction of a federal land tax, but recommended maintaining export taxes, albeit at a lower level, for price stabilization purposes. The Government also proceeded to reduce export taxes on beef and grain. Although such reductions had other explanations, the Bank, which had not even been consulted, chose to interpret them as proof of the Government's commitment to the proposed policy changes. 2.6 The preparation mission (November 1984) recommended a sector loan (rather than an import financing or sector recovery loan) if a policy framework could be agreed with the Government. The loan would be around USD 200 million for two years. A second loan, or even a series of loans, was envisaged to complete the process. The preparation mission concluded that export taxes should be replaced as a source of fiscal revenue, and that a federal land tax was the best alternative, both on fiscal and on sectoral policy grounds. 2.7 The loan strategy proposed by the mission included: (a) total abolition of export taxes on agricultural commodities as soon as a viable revenue substitute could be found; (b) Government commitment to appropriate exchange rate policies beyond 1985 (the Standy Arrangement period); and (c) selective removal of restrictions on imported inputs. The Initiating Memorandum (IM) stated that reintroduction of export taxes would be by congressional approval only. 2.8 The following risks were acknowledged in the IM: (a) four years might be inadequate to secure the reforms; (b) future Fund-Argentina agreements might prove difficult, undermining assumptions about continuing macro-stability; (c) budgetary pressures might force the Government to raise export tax rates again; and (d) the US and the EEC might alter their agricultural support programs, thus undermining markets and prices for Argentina grain exports. -6- 2.9 The IM Review Meeting cast doubts on the practicality of the central loan concept in the Argentine context. The staff continued to assert that the time was opportune for tax reform, citing Government and farmer support. But since knowledge about the wider fiscal and legal context of the proposed tax changes was inadequate, and the magnitude of the tax changes proposed was substantial, the Meeting decided to include under the loan studies of a revised agricultural tax system. 2.10 The final loan design envisaged a two-year operation to be disbursed against general imports (negative list). The earlier concept of linking disbursements to a positive list of agricultural input imports was discarded because a concurrent Inter-American Development Bank loan would finance most of them. The idea of eliminating the export taxes was dropped; the Bank decided on a partial reduction, but did not support the idea of using export taxes as a price stabilization device, as stated in PRONAGRO. Reductions in export taxes announced in 1984 would be considered as sufficient for releasing the loan's first tranche. They would be further reduced over the loan period. Land tax introduction was expected to take about one year. Exchange rate requirements were very weak, just calling for commitment to the principles embodied in the Fund agreement for the duration of the loan. The strategy behind the loan assumed that the establishment of the land tax and the shift from indirect to direct taxation would lock in long-term change and prevent reinstatement of higher levels of export taxes once the loan was fully disbursed; hence, there was no conditionality barring the re-introduction of higher levels of export taxes. 2.11 But preparation had not been without problems. First, the preparation mission had run into difficult waters. The Ministry of Economy claimed it had not formally approved the mission, and refused to deal with it at an official level. Then, the much-publicized reductions recently effected on export tax rates were retracted during the mission in order to help offset a 6.8% devaluation carried out in late October under the Fund Standby Arrangement. Government officials stressed that these measures were temporary, and were necessary to alleviate an immediate fiscal crisis. 2.12 Second, despite the magnitude and fiscal implications of the loan, and the fact that the request for the loan had been signed by the Minister of Economy (March 18, 1985), the quality of the dialogue with the Government and the extent of commitment in Argentina remained unsatisfactory. Dialogue with the Ministry of Economy and its Secretariat of Finance remained minimal, and these continued to act unilaterally on matters central to the loan concept and objectives. Third, economic policy formulation and implementation were in disarray at the end of President Alfonsin's first year in office; the economic team was changed in February 1985. 2.13 Fourth, its prominence in the project concept notwithstanding, land tax legislation received little analytical attention until the final version of the Initiating Memorandum and Loan Committee discussion. And fifth, the Bank's hands were tied on the issue of the exchange rate, adverse movements in which could wipe out any achievements which the loan might bring. 2.14 These events should have led the Bank to consider halting loan processing until these issues were clarified, or stronger signs of Government commitment were forthcoming, maybe as a prelude to a re-assessment of the operation's feasibility. But it appears that the momentum of loan processing was too strong, and appraisal forged ahead. -7- 2.15 The Loan Committee commented that enactment of legislation as a tranche condition was unusual, but noted that the Government was fairly confident of getting legislative approval. The Committee considered a proposal to carry out an analysis of the loan's political viability, including the Congressional process and players, but decided against it. The Loan Committee questioned the wisdom of substituting an easily-collected tax at a time of very high inflation and economic deterioration. The Loan Committee endorsed the idea of having two tranches: the first, in mid-1986, for actions already taken, and a second, in January, 1987, for a 50% export tax reduction and introduction of a federal land tax. 2.16 During the appraisal mission (June 1985), the Government again raised export taxes (by 8-9%) and tariffs on imported inputs (by 10%). Fiscal revenues generated by export taxes rose by 35%. Nevertheless, the mission and the Bank once again concluded that these were temporary measures necessary to stabilize the economy and not the obvious challenge to the loan's tax strategy which they actually were. 2.17 The appraisal mission proposed several modifications to loan conditionality. The Bank decided to require presenting to Congress a draft land tax law acceptable to the Bank as a Board condition, and rolling back export taxes to their May 1985 levels and Congressional passage of the federal land tax as conditions for second tranche release. The Bank also decided that the land tax would not have to compensate fully for revenue lost from further export tax reductions. Again, questions about the feasibility of the program were raised in the Bank's internal review process, prompting a post-appraisal mission focussing on the practicality of the land tax implementation schedule. The post-appraisal mission (August 1985) concluded that the schedule discussed during appraisal was feasible. But while that conclusion was technically correct, the political obstacles remained formidable. 2.18 As project processing gathered momentum, additional indications that the project might not achieve its objectives were ignored. First, export taxes, which had been reduced by the Government in 1984 without consulting the Bank and increased during the preparation mission, were again increased during the appraisal mission, raising doubts about Government commitment to one of the loan's central policy changes. Second, important personalities, including a former Regional Vice President of the Bank, warned Bank staff that a federal land tax was politically not feasible in Argentina. And third, the Bank's Legal Department commented negatively on the land tax proposals. A proper reading of these warnings should have led to a drastic re-design of the proposed loan or to halting its processing. Loan Size 2.19 No clear guidelines on how to determine the size of an adjustment loan existed at the time. The Region made an attempt to relate the size of the proposed loan to the fiscal cost of the proposed adjustment. The original policy proposal was for the grain export taxes to be dropped immediately, and later replaced by a federal land tax. Eventually total revenue generated by the latter would be equal to the foregone export tax revenue (i.e., providing revenue neutrality within the sector). But it was foreseen that it would take some 3-4 years (later re- estimated by Government at 12-18 months) for the land tax to be fully operational. Therefore, there would be a transient foregone revenue gap; the draft IM proposed a loan of USD 200 million to substitute for it. -8- 2.20 The Fund, however, firmly rejected the notion of substituting borrowed funds for fiscal revenues, even in an interim period. Once this anchoring mechanism was no longer available, loan size was determined on much softer grounds. First, the loan size was set at USD 300 million, equivalent to the average lending program envisaged for the next five years. Then, in view of re-estimated critically needed foreign exchange, it was raised to USD 350 million, almost doubling since the mission of November, 1984. 2.21 The Loan Committee believed that USD 250 million was a more appropriate loan size. It left the door open to a loan of up to USD 350 million, however, if appraisal showed that the operation would accelerate reform and if upcoming negotiations with the Fund and commercial banks were to bear fruit. 2.22 The Fund's rejection of the attempt to substitute borrowed funds for fiscal revenue lost or foregone quashed attempts to determine adjustment loan sizes in relation to the fiscal cost of adjustment. Current rules call for judgments about the size of an adjustment loan to be based mostly on the external resource gap (a criterion acceptable to the Fund), the size of the lending program, the strength of the reforms being supported and, in the case of IDA countries, the IDA fund allocation to the country. Negotiations and Approval 2.23 Negotiations were conditional on: (a) the presentation of the Government's macroeconomic and agricultural policy statement; (b) the presentation of the Government's proposal for alternative interim revenues; and (c) verification of federal land tax administrative capacity. The negotiations ran smoothly and were concluded on January 17, 1986, on the terms and conditions approved by the Loan Committee. At the specific request of senior management, the Minutes of Negotiations stated that enactment and initial implementation of the federal land tax would be criteria of satisfactory progress for second tranche release. 2.24 The loan would support the following policy reforms: (a) reduction in export taxes; (b) fiscal measures to maintain the fiscal deficit at satisfactory levels; (c) introduction of a federal land tax; (d) modified regulations and tariffs pertaining to the import of agricultural inputs, and (e) complementary studies and institutional support. 2.25 Loan conditions would be: (a) Board presentation would be contingent on the submission to Congress of a draft land tax law acceptable to the Bank; removal of the temporary increases in export taxes imposed in June, 1985; and presentation of a program, acceptable to the Bank, for preparation and implementation of the federal land tax; (b) the first tranche (of USD 170.5 million) would be released upon loan effectiveness; (c) the second tranche of USD 170.5 million would be disbursed following evidence of: reduction of the export tax rates on agricultural commodities and of the import tariffs on chemical inputs; fiscal measures to compensate for revenue losses; designation of tractors and agricultural machinery to the automatic entry classification; satisfactory progress in the execution of the agreed land tax implementation schedule; and progress in the studies to be financed under the loan. 2.26 An Executive Committee chaired by SAGyP was to oversee the preparation and implementation of the land tax and the sectoral tax studies. Provincial interests would be represented by the Federal Investment Council. -9- 2.27 The loan was approved by the Board on April 3, 1986, upon presentation of the draft Land Tax Act to Congress. But unknown to the Bank, the clauses concerning short-term implementation in 1986 and 1987 had been removed at the last minute. III. IMPLEMENTATION 3.1 The first tranche of US$170.5 million was released on September 24, 1986, five and a half months after Board presentation. 3.2 Legislation to establish a federal land tax stalled in congressional committees. Evidence suggests that the drop in international agricultural prices softened the Executive's willingness to pursue the project. Farmer organizations argued that the reductions in export taxes just compensated them for sharply lower international commodity prices, and could not be accepted as a quid pro quo for a federal land tax. A proposal by SAGyP to include elimination of export taxes in the draft land tax legislation was opposed by Treasury. 3.3 The resignation in late 1986 of the Secretary of Agriculture was the beginning of the end. The new Secretary expressed his support for the land tax proposal, but wanted to make it a replacement for all other sectoral taxation, a move likely to delay passage of the law and its regulation. In April 1987, on the anniversary of its submission, internal Congressional regulations led to the removal of the land tax act from active consideration 2'. 3.4 By late 1986, reductions in export tax rates on major agricultural and agro-industrial products had far exceeded those required under the Loan Agreement. Such reductions, however, had been prompted by the fall in international commodity prices to historical lows in real terms, and not by a commitment to the loan's policy conditions. After the Spring Plan of August, 1988 (para 1.3), the process was reversed. A dual exchange rate which discriminated against agriculture was introduced. After the exchange rate re-unification in May, 1989 (one month before loan closing), export tax rates were raised to levels even higher than those prevailing before loan processing. (The evolution of grain export taxes in Argentina between 1981 and 1992 is presented in Figure No 1). 3.5 The five studies and two programs of institutional support were successfully completed, with loan funds being administered by the United Nations Development Program. The Closing Date of the Loan was extended for one year (to June 30, 1989) to permit completion of the unified federal cadastre based on individual provincial cadastres. 3.6 Tractors and agricultural machinery were transferred from the "prohibited import" list to the "prior approval" category as a condition of loan effectiveness. However, their inclusion on the "automatic entry" list for second tranche compliance was delayed by depressed conditions 3/ A draft law which is not enacted within one year of submission to Congress is automatically removed from active consideration; any further processing requires re-submission. Bank staff seemed unaware of this regulation, and of the fact that the draft land tax act was no longer being considered by Congress. - 10 - in the domestic industry. The Government had also agreed to remove a 10% import surcharge on agricultural chemicals; by late 1987, it instead increased it to 15% as a part of Fund conditionality. Deficit reduction requirements under a new Fund agreement in early 1988 caused the Government to retain the surcharge. Release of the Second Tranche 3.7 The second tranche release (USD 170.5 million) had been anticipated for January 1987. This was overly optimistic 5. It was impossible to get a federal land tax law enacted, and the Government refused to remove the import surcharge on agrochemicals. The Government was also late in assigning tractors to the "automatic entry" category. 3.8 Doubts about the original proposal began to resurface 1. Staff at the Department's and Region's front offices argued that implementation of the federal land tax had not been a major objective of the loan. However, there are some 47 references to it in the President's Report, and the central structural aim of the loan clearly was to "improve incentives necessary for use of agricultural resources to increase production and exports". 3.9 In late 1987, the RVP agreed to accept a package of fiscal measures recently submitted to Congress as a substitute for the federal land tax and an alternative replacement source for the loss of revenue resulting from export tax rate reductions. 3.10 Release of the second tranche required now a waiver of the two unfulfilled conditions (land tax and import surcharge). Bank management submitted to the Board a request based on the acceptance of alternative fiscal measures and on the understanding that the relaxation of import restrictions on tractors and agricultural machinery was a significant move in trade policy reform. The request acknowledged that agricultural sector reform had made progress, as set out in a letter from the Government to the Bank (April 28, 1988). 3.11 Board discussion revealed the controversial nature of loan conditionality administration and of the case at hand. While the waivers were eventually approved, the high rate of abstentions indicated that cancellation of the second tranche might have been a better decision. The second tranche was released on June 2, 1988. The tranche's release secured the resource transfer objective but meant abandoning a major policy reform objective. 3.12 As a result of experience under this and other adjustment operations, the Bank no longer includes completion of anticipated legislative action as a condition for tranche release, for the Executive Branch of most governments cannot commit themselves to deliver on decisions to be adopted by the Legislative Branch. 4/ Experience worldwide suggests that not less than one year is usually required for policy changes associated with second tranche release to be implemented. 5/ These discussions followed the Bank's reorganization in 1987, which brought to the table a new group of Bank managers. - 11 - Disbursement, Procurement and Audit 3.13 The loan used standard Statements of Expenditure (SOE) for loan disbursement under Category I (totalling USD 341.0 million), but difficulties arose even prior to first tranche release. The Central Bank (BCRA) was the designated authority for SOE preparation, but BCRA balked at assuming this authority, having been excluded by the Ministry of Economy from participating in the preparation or negotiation of the loan. The process of sorting this out delayed first tranche disbursement. 3.14 The BCRA proceeded to prepare the SOEs for both tranches. Obstacles were encountered by Bank supervision. The BCRA was unwilling to allow access to supporting documents. High level official intervention resolved the matter, enabling the Bank to undertake the appropriate checks with satisfactory results. 3.15 Procurement of goods against general imports was straightforward. Audit compliance under the loan was poor, with clean audit reports for 1986-1988 still not received by mid-1990. The BCRA would not allow auditors of the Tribunal de Cuentas de la Naci6n (TCN) access to commercial bank records. The TCN was obliged to issue a disclaimer of opinion on loan audits. The ME refused to release the disclaimed audits to the Bank, in the fear that the Bank might seek restitution of loan funds disbursed. A new audit was effected and a clean opinion accepted by the Bank in August, 1990. Compliance with Loan Conditions 3.16 Compliance with first tranche conditions was satisfactory, though delayed. On second tranche, compliance was mixed. Reductions in export taxes proved ephemeral. The original federal land tax condition was not complied with. To compensate for lost export tax revenues, the Bank accepted other measures already taken as fulfilling loan conditions. Maintenance of an appropriate exchange rate was not a condition for tranche release, but it was part of the Government's commitment "to maintain an exchange rate competitive for exports" (Letter of Development Policy). In this respect, the introduction of a dual rate under the Spring Plan of August 1988 violated the spirit of this commitment. Institutional Performance 3.17 The World Bank The process of re-orientating agricultural taxation from export taxes to a federal land tax was not well thought out. The loan was prepared and appraised at a time when resource transfer and an enhanced lending program to Argentina competed in urgency with policy reform. 3.18 Loan conditionality conflicted with Fund stabilization regimes, and involved the Government in a conflict of interest. Bank's experience with sector lending has shown that these must fit into a broader program of adjustment to have a chance of success and sustainability. 3.19 Preparation should have been supported by more rigorous economic and sector work. Vital aspects of the loan's legal, fiscal, political and institutional environment were poorly understood. Advice from the Fund, the Bank's own review staff and experts retained to examine loan options and feasibility went largely unheeded. The Bank gave the principal role in the - 12 - planning and implementation of the reform program to SAGyP, although SAGyP had very little control over the policy and resources affecting the sector. 3.20 The Bank should have made a greater effort to establish sustainability of export tax reductions. SAGyP and two of the four major farmer groups already had reached broad agreement on the need to lock in export tax rate reductions and remove their discretionary use. The Bank's Working Paper No.10, as part of loan appraisal, states: "This is an opportune moment to abolish export duties since the decline in world prices will inevitably force their elimination or substantial reduction". The Loan Agreement, however, did not include a Government commitment to eliminate ME's legal right to set these taxes without recourse to Congress. 3.21 The Borrower. SAGyP was the chief proponent and supporter of the proposed reforms. The ME seems to have had little interest in structural reform and was ambivalent about losing a secure source of revenue. The reduction of export taxes and their substitution by a federal land tax represented a major change in the tax regime, but it was initiated and advocated by SAGyP without due consultation with Treasury and Public Revenues. 3.22 Although Congressional approval was required for implementation of the federal land tax, Government made no serious effort to secure its passage, nor did it guarantee a permanent reduction in export tax rates. The removal of key clauses in the draft land tax legislation submitted to Congress prior to Board presentation was a breach of faith with the Bank. IV. IMPACT 4.1 Achievements under the loan were mixed: (i) albeit the second tranche was one-and- a-half years late, intended resource transfers were effected; and (ii) the technical studies were effectively carried out and have proven useful in several respects; but (iii) the original policy package was either not implemented or reversed before the loan's Closing Date. Since structural changes are at the essence of an adjustment operation, the loan must be deemed a failure. 4.2 Resource Transfer. Loan funds played an important role in securing a steady foreign exchange flow. Argentina was short of foreign resources throughout the loan disbursement period; in July 1987 the Bank threatened to suspend disbursements for lack of repayment, and in April 1988 Argentina stopped payments to commercial banks. The second tranche was a major element in securing a steady flow of new money in 1987 (see paras 5.35-5.37); its release ensured continuity in funding and helped to shore up Argentina's external accounts until the crisis was overcome in 1990. Therefore, from this angle the loan was eventually successful & 2 4.3 Technical Studies. USD 9m was earmarked to finance studies and institutional development. This component was successful. (i) The soils map was completed, and Argentina's §/ The collapse of the Austral Plan and then the Spring Plan (para 1.3) was for reasons other than lack of foreign exchange. 7/ Whether the availability of such resources was ultimately positive, or just helped the Government to delay urgently needed adjustments in the exchange rate, is beyond the scope of this sectoral audit, and should rather be addressed in the context of the evaluation of the Trade Adjustment loans (para. 1.13). - 13 - Soils Atlas was published, a work which has received several awards for technical excellence, including the First Prize for a Scientific Work in the 1988-91 period. (ii) All provincial rural cadastres were updated, completed and modernized, and then articulated into a federal cadastre. This required establishing working agreements with all provinces (a feat underestimated at appraisal), and it is now being utilized for enhancing provincial revenues under the Provincial Development Loan (Ln. 3280-AR, of December 1990). (iii) The agro-economic studies led to important policy initiatives and, substantially, to the lending program followed over the last three years (para 1.14). 4.4 Policy Change. (i) Export taxes on grains (retenciones) were initially reduced to the agreed levels and beyond, but the policy was then revised and before the loan's closing they were raised to levels even higher than those prevailing before loan processing. (ii) The federal land tax was never enacted. (iii) The liberalization of the importation of tractors and agricultural machinery was implemented. 4.5 One Other Impact: Learning from Experience. The poor experience in project implementation triggered some unintended benefits through early learning from experience. On the Bank's side, the design for the second adjustment operation (Trade Policy and Export Diversification, Ln. 2815 of May 1987) involved much more ex ante discussion and preparation work; a heavier front load of conditionality; and easy-to-apply measures for second tranche release (the tranche was released within 7 months of approval, with loan targets surpassed). On the Government's side, preparation of two other adjustment operations (an earlier version of the 1988 Banking loan, and a possible privatization loan) were stopped in 1987 because they were repeating the same flaws already apparent in the ASAL. 4.6 Events after Loan Closing. The loan was closed in the final month of the Alfonsfn Government. The new Menem Government is putting in place a single federal tax system based on efficient taxes (income, VAT, social security), eliminating sector-specific taxes. Thus, today the federal land tax is a non-issue. In terms of policy consistency, grain export taxes should have been eliminated as well, but they have not been. They remain on the books. Their level has been set at zero for most grains, but taxes remain in effect for oilseeds and raw hides (Figure No 1). There is no indication that this reflects a policy decision to suppress export taxes; no attempt has been made to amend the Customs Law, which grants the ME the power to set the grain export taxes without recourse to Congress. Consequently, the Government has continued to follow the taxing practice of the past decades, whereby grain export taxes are raised after a devaluation or when international prices are high, and lowered when the currency is overvalued or international prices are low. - 14 - V. ISSUES AND LESSONS The Origins of the Loan 5.1 Most of the loan problems can be attributed or related to events at identification, preparation and appraisal. In turn, these can be mostly attributed to two causal factors, one on the Bank's side and one on the Borrower's side. 5.2 The Reasons for The Loan. The wisdom of having ever made the ASAL has often been questioned. However, even in retrospect, there are compelling reasons to have supported this operation: (i) sector importance; (ii) attractiveness of the policy proposal; (iii) resumed lending to Argentina; and (iv) quality of the Borrower's economic and agricultural teams. 5.3 First, agriculture is a most important sector in Argentina, particularly in terms of current exports and potential for growth. At the time, it was the sector most unprotected and facing the most distorted market signals. Second, the Government had produced a program to address these problems (PRONAGRO, para 2.5) which advocated the same tax changes proposed in the chapter on agriculture of the just-finalized 1984 CEM. 5.4 Third, lending to Argentina had been virtually suspended in the last years of the military governments. With a democratic Government taking office in December 1983, it was the perfect time to re-start lending in a substantial way. Fourth, the Bank was impressed by the high quality of the Borrower's economic team and by the technical competence of the agriculture team. 5.5 In addition, adjustment lending was becoming one of the hottest lending instruments in the Bank. Adjustment operations were seen as the only way to expand disbursements in Argentina fast enough, and agriculture was the only sector in which years of academic work had generated a development strategy and a policy prescription. 5.6 The Issue of Sequencing. Was it proper to start adjusting such a distorted economy through a "back door"? Should not adjustment have rather started with a structural adjustment loan (SAL), or with a "sub-SAL" on a macroeconomic issue such as trade policy or financial sector reform? At the time, the question never arose. Adjustment was a rather new type of lending instrument, and the principle of having a macroeconomic program in place before moving into SECALs evolved only later. Nor was there a well established sequence Bank-wide to guide Regional management. Up to the time of loan identification Y, adjustment operations had been made in 26 countries. Of these, 15 had not yet had a sequel. Of the 11 which had had one or more follow-on loans, 6 countries had had only SALs; 4 had had "sub-SAL" (or broad sector) 8/ August 1984; data taken up to end of FY84. - 15 - operations followed by a repeat one (2) or by a SAL (2); and 1 country had had a SECAL followed by another (both in agriculture). LAC experience was five countries with a single SAL (3) or sub-SAL operation (2); one with the sequence sub-SAL/sub-SAL; one with the sequence sub-SAL/SAL; and one with a single (agricultural) SECAL. Up to that time, no country worldwide had had a SECAL/SAL or SECAL/sub-SAL sequence. ' 5.7 Besides, a SAL would not have been feasible at the time. On the Bank's side, analysis had focused on resumption of growth, and the Bank did not have as yet a strategy for adjustment in Argentina. The 1984 CEM had recommended measures to improve fiscal management, provide export incentives to agriculture and industry, and raise the efficiency of the financial sector, but had not offered a proposal for structural adjustment. Only by mid-1987 was the Bank to have good analytical command and policy proposals for stabilization and adjustment. On the Government's side, Argentina did not have the capacity or the political support for a full- fledged SAL. So, addressing obvious distortions in a sector with a large potential for generating additional foreign exchange seemed (and still seems in retrospect) the proper approach to starting adjustment in Argentina. 5.8 The real issue is whether macroeconomic variables had been stabilized sufficently to allow effective SECAL design and execution. Sector adjustment cannot proceed satisfactorily unless there is a stabilization program in place, including a macroeconomic framework and a detailed strategy to consolidate and widen the adjusting effort. The sequence SAL/SECAL is, thus, the ideal one, addressing macro stabilization first, followed by sectoral distortions. But if a SAL at the outset is not deemed feasible, starting adjustment through obvious changes in particular sectors may be the best way to start moving towards a wider adjustment. However, such a SECAL operation should include stabilization measures as part of its policy package. This is a lesson the Bank has learnt: most SECALs now are either embedded in a macroeconomic framework or incorporate the macro conditionality required to ensure proper implementation. 5.9 A macroeconomic framework was not in place at the time. In December 1984, the IMF approved a Standby program focused on short-term stabilization issues, without structural changes to consolidate progress over the longer term. The most serious problem the Fund saw in Argentina was the fiscal deficit, and increasing grain export taxes was the easiest and most reliable source of incremental revenue. Thus, eliminating the export taxes and replacing them by an untested land tax ran contrary to the Fund requirements LY. 9/ Usually, adjustment operations have been classified into either SALA or SECALA. There is a wide group of operations among the latter, however, which include economywide policy packages and thus are actually an intermediate form, or class, of adjustment operations between SALA and SECALS. These "Sub-SALA" or broad sector operations include loans and credits labelled Export Development or Rehabilitation; Trade Policy or Development; and Import Sector. Following the example of the RAL series, Fertilizer Loans were not included in the analysis in view of the difficulty in classifying them. 10/ Five years of adjustment lending experience was all Regional management had at the time as guidance, too short a time for sequels to have been produced. RAL I provides a longer 9-year perspective: up to the end of FY88, in 12 countries adjust- ment was initiated by a SAL to be followed by a SECAL, while in 8 countries the opposite sequence prevailed (and in 28 countries SALA were not followed by a SECAL, and in 7 countries, SECALA were not followed by a SAL). 11/ The IMF eventually agreed to the policy changes to be supported by ASAL when compensatory tax provisions were incorporated to make the policy changes revenue-neutral. The issue of Bank/Fund relations is beyond the scope of this sectoral audit, and should rather be addressed in the context of the evaluation of the Trade Adjustment Loans (para. 1.13). - 16 - 5.10 In June 1985, in the middle of the ASAL appraisal mission, the Government announced the Plan Austral. This Plan took the Bank by surprise. The Bank had not been consulted about it; the loan had not been intended or designed to support it, nor was it modified afterwards to this effect. 5.11 To a certain extent, the failure of the ASAL must be placed in the perspective of the collapse of the whole stabilization program. In this respect, the appraisal mission's Aide Memoire contained a note which in retrospect turned prophetic ("the mission is extremely concerned with the current macroeconomic environment"), followed by the naive expectation that "the mission assumes that the economy will have been stabilized by January 1986". 5.12 Fast Preparation. In spite of it being the first adjustment operation in Argentina, and of the limited knowledge available at the time on both the economy at large and the agricultural sector, the loan was processed extremely quickly. The whole process took barely one year, from identification in July 1984, Government request in September, through a single preparation mission (November 1984), to appraisal in May/June 1985. Remaining Bank doubts on the federal land tax proposals required a post-appraisal mission in August, 1985 (para 2.17). One wonders whether a more protracted process would have allowed the Bank to weigh more properly the multiple indications that the project might not achieve its objectives (paras 2.11-2.14 and 2.18). 5.13 Poor Preparation. Speed is not an issue if a loan is properly prepared. Unfortunate- ly, this loan was not. The overall tax system and the role played in it by the grain export taxes was not properly researched. Neither was the loan's institutional, political, accounting, and disbursement aspects. The preparation team did not analyze important loan elements, and only identified (but did not quantify and follow through) their economic and fiscal effects. 5.14 (i) Overall Tax Regime. Grain export taxes were one of the largest individual sources of fiscal revenue, in some years generating more than 20% of total tax revenue. Their replacement by a federal land tax would introduce major changes in the structure of the tax system. But the overall tax system was not analyzed, nor was the role of the two taxes in it. There seemed to be no awareness of the adjustments being introduced by the Government at the time into the tax system - moving towards a system centered on VAT and income taxes - or of the fact that a land tax would run counter to them. The Loan Agreement stipulated that compensatory tax revenue had to come from the agricultural sector, but the other sectors' tax burdens had not been analyzed, and the fact that agriculture was already relatively overtaxed was ignored. 5.15 (ii) Sectoral Taxation. The proposed tax swap implied the substitution of a direct tax for an indirect one, and a change in the subjects of taxation (and tax burden) from grain exporters and producers to landowners; from small landowners (typically grain producers) to large landowners (typically beef producers); and from non-owner producers (mainly contractors) to owner-producers. It would subject land to three different land taxes - federal, provincial, and municipal (plus its inclusion in the totals for computing the net worth and capital taxes). None of those aspects was analyzed, nor was the deeply-seated opposition - and always very effective in the past - of the two largest and most powerful landowner associations to a federal land tax. (The current tax rate per province is shown in Figure No 2.) - 17 - 5.16 (iii) Negotiation of the Tax Substitution. The proposed tax swap was conceived as a change in tax instruments while keeping the overall sector tax burden constant. Such a tax swap would have required a negotiated agreement among the three parties involved, namely, the Secretariat of Agriculture (which should have obtained assurances of a more efficient incentives system), the Secretariat of Finance (which needed assurances of getting a similar level of reve- nue), and the four agricultural associations (which should have received assurances that the new tax would not be an additional burden, and the old tax would not be reinstated) 2. Staff at the Secretariat of Agriculture claim that such negotiations were held in the context of the discussion of PRONAGRO, while staff at Treasury claim to have known nothing about them. The agricultural associations were divided on the issue, with the two largest and more powerful ones rejecting any land tax and the two others favoring the swap. Two facts stand out. First, the discussion was never set in the terms implied by the policy package, i.e., a negotiated agreement between the Government and the agricultural sector had not been reached. Farmers and landowners were never confronted with the land tax and swap proposal as a real, immediate policy issue L. Nor were they asked to commit themselves to it, or offered a guarantee that the new land tax would in fact replace the export taxes and the latter would not be reinstated at a later day (as it indeed happened, see paras 4.4 (i) and 4.6, above). And second, Bank missions did not check and confirm whether a firm agreement on, and commitment to, the tax swap had actually been reached. 5.17 (iv) Price Structure and Income Distribution. Loan processing documents mention the main economic consequences of the tax changes, but their micro and macro-economic implications were not analyzed or quantified. First, grain export taxes kept domestic grain prices low, thus increasing real incomes - particularly of the poor - and favoring urbanization and industrial development. Second, they protected in particular the agro-industrial sector. And third, the removal of export taxes would lead to a major income transfer from consumers to grain producers, estimated at 1.5% of GDP in the IMLY . Besides ignoring the all-powerful labor unions and industrial lobby, the Bank did not analyze the significance and impact of such a change in income distribution . 5.18 (v) Institutional Arrangements. Throughout the loan's history, the Bank worked basically with the SAGyP. While the SAGyP had responsibility for issues bearing directly on the sector, the overall leadership of economic policy - including the critical exchange rate, taxation, fiscal and monetary policies - lay with the parent ME, while tax administration lay with the Secretariat of Finance. SAGyP was technically strong but politically weak, without a support group or significant constituency. As mentioned above, ME did not even receive the preparation 12/ There are four such associations in Argentina: the Rural Society, grouping the largest and most powerful landowners and cattle raisers, basically in the Pampa; the Rural Confederations, which is a third-tier organization grouping some 300 rural societies nationwide, which in turn group large and medium-sized landowners; the Agricultural Federation, which groups small-scale farmers, many of which are not landed; and the Agricultural Inter-Cooperative Confederation, another third-tier organization which groups 13 federations and 1,300 individual cooperatives, basically devoted to processing and marketing of agricultural products. I/ One of the interviewees called them "meetings at the wishful thinking level". 14/ Incidentally, the figure of 1.5% of GDP mentioned in the IM was wrong; 1.5% GDP was just the magnitude of the tax to be removed, which applied only to the portion of the production which was exported. Price increases resulting from the tax removal would affect the whole volume of production, and the ensuing income transfer would be about twice as large. - 18 - mission; contacts thereafter were rather superficial; the question of whether the Ministry actually supported the policy package was not discussed - it was assumed on the basis of the Minister's signing the Letter of Development Policy, where the package was described. The Secretariat of Finance was not officially approached to assess its support of the proposed tax swap - it actually opposed it, and in due course, fought internally against it. 5.19 (vi) Disbursement Procedures. These were ignored during preparation and appraisal. The Central Bank, the designated authority for SOE preparation, was not visited by either mission and was unaware of the steps and responsibilities involved. This led to major delays in loan disbursements. Current rules for processing lending operations, which call for explicit clearance of the yellow cover package by the respective Disbursements Division, should help to prevent such oversights. 5.20 The Causal Factors. Most of the problems detailed in the preceding paragraphs can be attributed to deficiencies in the way both the Bank and the Government handled the loan. 5.21 (i) The Bank's side. The onus of handling the very first adjustment operation ever in Argentina was put on the Projects Department's Agriculture Division. This was a poor decision. The Division developed the policy package in terms of the sector's potential and incen- tives, but lacked the global macroeconomic support that only the Programs Department could provide. The failure of the latter to provide such support during identification and preparation was due to the focus of its recent economic work and its temporary shortage of specialized staff. On the former, Programs had focused on the medium-term resumption of growth rather than on stabilization and structural change (para 5.8). On the latter, the two macro-economists working on Argentina had left shortly after completing the CEM. Another economist, then working on Bolivia, was hurried in to join the appraisal mission V, but he then also left the Division. Therefore, Programs was not in a position to provide the required macro-economic input into project preparation and appraisal L 0. Most of these institutional and management issues were addressed at the 1987 reorganization, when the program and sector divisions serving the same countries were put under the same Department. Currently, country operations divisions carry the main responsibility for macro-economic aspects of most SECALs. 5.22 (ii) The Government's side. No consensus existed within Government on the tax swap issue. ME did not get into the details of the work being done by SAGyP. Although weekly briefed by the Secretary, senior staff in ME did not alert the Secretariat of Finance of the tax swap proposed under the loan, nor did they alert the Secretary of Agriculture of the diverging objectives of the proposed land tax and the on-going tax reform. Even if he knew the new proposals were being prepared, the Secretary of Finance may have chosen not to oppose what 15/ When the project officer fell ill on the eve of departure, the Programs' economist assumed mission leadership. J/ The role of Programs changed substantially after appraisal, when the project officer in charge rotated out of the Region; a new country economist husbanded the loan through Negotiations and Board presentation. L7/ In the audit's view, the fact that the Agriculture Division limited its contacts basically to the Secretariat of Agriculture (paras 3.19 and 5.23) is also related to Programs' lack of leadership on what was a major change in the country's tax, fiscal, real wage, and industrial protection regimes. - 19 - looked at the time to be preliminary plans until a firm proposal was sent to his desk - which actually happened only after the Minister had signed the Letter of Development Policy. Weaknesses of the Loan's Policy Package 5.23 The Reduction in the Grain Export Taxes. Export taxes are a flexible, easily collect- able, reliable form of taxation, used historically in Argentina to extract revenue from the agricultural sector and particularly to mop up windfall benefits following devaluations. Converse- ly, they usually are reduced in case of revaluation and/or a substantial drop in international prices. 5.24 Four issues stem from the above. First, the elimination of the discretionary power of ME to set export taxes would have addressed only part of an important issue. Both the exchange rate and the export tax regimes were closely related to each other, and had to be confronted simultaneously. But exchange rate policy was the Fund's domain, leaving the Bank with no mandate to deal with that other part of the issue. The Standby Arrangement approved by the Fund in December 1984 (between loan preparation and appraisal) committed Argentina to keep a realistic exchange rate policy. The tasks remaining for the Bank were to ensure Government's compliance, and to extend it over the loan period, which was to be longer that the Standby. The draft Initiating Memorandum thus included two proposals: (i) to have cross- conditionality with the Fund Standby Arrangement; and (ii) to maintain a joint Bank/Fund dialogue on exchange rate policy. The former was rejected by management on institutional grounds; the second was rejected by the Fund on jurisdictional grounds ff. Therefore, the loan's policy package included only half-a-solution, for nothing in the Loan Agreement required the Government to maintain for the duration of the loan an exchange rate consistent with the removal of the export taxes. 5.25 Second, there were diverging views on the grain export taxes between the Bank and the Fund, and between the Secretariats of Agriculture and of Finance. SAGyP and the Bank viewed these taxes negatively, on efficiency grounds, and proposed their outright elimination - later tempered to a partial reduction. Finance and the Fund viewed them positively, as expedient sources of revenue in view of the catastrophic fiscal deficit, and proposed increasing them. Neither conflict was resolved during preparation, appraisal or negotiation L. In the event, Finance and the Fund prevailed, and export taxes have continued as before. Thus, they were reduced before preparation (a signal the Bank misread as an expression of policy acceptance), raised during both the preparation and the appraisal missions, lowered during the implementation period, and raised before loan closing, each time without the Bank being consulted or even informed beforehand. 5.26 Third, Agriculture and the Bank viewed the initial reductions on the export taxes as a sort of down payment on the future introduction of the compensatory land tax. They thus expected landowners to appreciate such reductions from this perspective. But the landowners 18/ "The proposal [in the draft IM, to maintain a joint Bank/Fund dialogue on exchange rate policy] ... would not seem to conform to the agreed principles of division of responsibilities between our two institutions and may raise a jurisdictional issue." Memo of May Ist, 1985 from Deputy Division Chief, IMF, to Chief, River Plate Division, World Bank. 19/ As mentioned above, the Fund agreed to the loan's policy package only after adequate revenue substitutes had been included in the draft Loan Agreement (footnote 6). - 20 - interpreted the reductions as just a new turn in the well-known historical practice of using export taxes to compensate for changes in exchange rate or international prices, and not as a new sector benefit to be "repaid" by accepting a land tax in its stead. Further, they were never offered guarantees that the discretionary power to set them would be removed from ME. Subsequent events proved them right. 5.27 And fourth, the issue of 'optimal taxation' that arises in the context of Argentina's substantial share in the international grain markets was raised during loan processing by the Bank's technical review departments (i.e., export taxes should not be reduced to zero but to their "optimal" level). Curiously, the proposal was rejected as complicated and difficult to compute. 5.28 The Introduction of a Federal Land Tax. Four times since the 1950's Governments had unsuccessfully tried to introduce a land tax. Repeated warnings against trying again were received during loan processing (para 2.18). Two obvious questions are: (i) why did the Alfonsin Government attempt it again, and (ii) why did the Bank support this attempt. 5.29 The Government's Story. In 1983, Alfonsin won the presidential election on a platform which included the federal land tax. The losing Peronist candidate's plank also included it. Thus, the proposal would have been expected to carry a solid majority of the Congress votes. Second, being the first elected government after eight years of military rule and the South Atlantic war defeat gave the Alfonsin Government a strong sense of mandate and empowerment, together with a conviction that through either legislation or mass mobilization it could change Argentine society ("voluntarismo politico"). And third, the land tax proposal was considerably milder than its predecessors, and did not include any of the politically sensitive issues which had doomed them. Therefore, the land tax seemed acceptable and affordable. The unresolved part of the story is the inability of the Government to settle its internal conflict (paras 5.14; 5.18 and 5.25); eventually the draft Land Tax Act was sent to Congress without the signature of the Secretary of Finance, the Government's tax man. 5.30 The two most powerful landowner associations mounted a massive attack on the draft Act (and on PRONAGRO, which included the land tax proposal), including parliamentary obstruction and back door access to the President. Meanwhile, the Government's strength to push for its approval had been sapped by the decline in international prices (which made lobbying Congress for additional sector taxes a non-viable proposition) and by the defeat at the September 1987 elections. 5.31 The Bank's Story. The 1984 CEM had included a chapter on agriculture which advocated a federal land tax. Thus, both Programs and Projects staff were delighted to learn that the Government was preparing a similar proposal. This perceived fruitful matching of technical analysis and political will, with resource transfer needs and an interest in structural adjustment (paras 5.3-5.6) gave loan processing its impetus. To ensure that essential steps were taken early, submission of the draft Land Tax Act to Congress was made a condition of Board presentation, while the technical studies (economic and cadastral) required to support it were to be carried out during implementation. This was a wrong sequence, for drafting the Act ought to have followed completion of the technical studies. Hence, a longer preparation process was required (paras 5.12-5.13). - 21 - 5.32 Without fully realizing the irony of asking for the Act first and the technical studies later, the Bank focused its analytical capacities on the latter. A post-appraisal mission was carried out solely to address technical objections raised by the Bank's Agriculture Department. The Bank, however, missed the bigger picture, for loan preparation and appraisal did not properly address the role of the land tax in the on-going overhaul of the tax regime (para 5.14) and the tax's legal aspects (para 2.18). 5.33 But the most serious flaw was not appraising the political viability of the land tax proposal. The issue was explicitly raised within the Bank, but the Loan Committee decided against it, on the grounds that the Bank had to trust the Government's assessment. The loan history would have been different - the loan itself might not have materialized - had such an appraisal been carried out. This should no longer be a problem; current appraisal procedures, which include governance and institutional assessment, require an analysis of whether policy change proposals have adequate political support. Events Surrounding the Second Tranche Release 5.34 Release of the second tranche of the loan proved a complex process. Once the draft Land Tax Act was withdrawn from consideration (para 3.2) and Government had refused to remove import surcharges on agro-chemicals (p. 3.6), two major conditions for second tranche release became unattainable. Meanwhile, the second tranche funds had become (not by Bank intent) a part of the financial package agreed between the Government and the commercial banks, with developed country government support, as Argentina came to the verge of running out of foreign exchange. Pressure from several quarters to release the tranche was intense. Management had to go to the Board to get the conditionality waivers approved, which the Board reluctantly agreed to do only after a three-hour, tense debate. 5.35 The External Financial Condition of Argentina at the Time. Up to 1986, Argentina managed to secure enough foreign liquidity to avoid major external payment problems. The December 1984 Standby Arrangement and the first tranche of ASAL (plus both tranches of the Trade Policy Loan; para 1.13) contributed to ensure this liquidity. But the drastic drop in interna- tional grain prices after the loan became effective severely reduced foreign exchange inflows, and by early 1987 Argentina had missed payments to commercial banks. By July 1987, delays in payments had prompted Bank Regional management to consider suspension of disbursements. 5.36 In this deteriorating context, commercial banks formed an Advisory Committee and a Steering Committee, both chaired by Citibank. In February and March 1987, Bank and Fund staff participated in meetings of these Committees in Washington and New York, presenting their respective base case projections and funding commitment and disbursement programs. In April 1987, the Government and the commercial banks reached an agreement on a lending program extending through the first quarter of 1988. The main issue was the amount of "new new" money- commercial banks would provide to Argentina. The agreement did not include commit- ments from either the Bank or the Fund, but both the Advisory Committee and the Government wanted to link the agreed commercial bank disbursements to specific levels of Bank and Fund 20/ "New new" money was the term used at the time for lending which was additional in net terms, i.e., net of principal and interest payments. - 22 - disbursements. Initially, Bank staff opposed any such disbursement cross-linkage, but in the event they agreed on the dual grounds that the Government wanted to pursue that avenue (actually, it needed to L) and that no risk for the Bank was involved. Four tranches of new money disbursements by commercial banks were agreed (triggered by Bank disbursement amounts shown in parenthesis, cumulative from January 1987: August 1987 (USD 150m), October (USD 300m), December (USD 450m), and first quarter of 1988 (USD 500m)); the Bank saw the agreed levels as "low risk". 5.37 But then the arrangements started to unravel. The last disbursements from the commercial banks under the April 1987 Agreement, and from the Fund, were made in the last week of March 1988. The same month, Fund staff completed its review and informed its Board of Directors that the Argentina program was not fully funded, making Argentina ineligible for further support. And the Bank's Banking Adjustment Loan, approved in March 1988, could not be signed (it was eventually canceled unsigned). Suddenly, Argentina was without adequate foreign funding, and in April 1988 it stopped international paymentsL. ASAL's second tranche became the only foreseeable fund source massive enough to tide the country through. To help in filling the gap, the USA Treasury made a short-term USD 500m bridge loan towards the loan's second tranche . As the term for the bridge loan approached, pressure increased on the Bank to release the tranche. 5.38 The Process from the Dismissal of the Draft Land Tax Act to the New Condition- ality. But conditions for second tranche release had not been, and could no longer be, fulfilled (the five conditions are listed in para 2.25.c). For some six months after the dismissal of the draft Act, the Bank kept pressure on the Government to deliver on the land tax commitment. In the second half of CY 1987, the Bank eventually acknowledged the demise of the land tax and started searching for alternative release conditionality. Contrary to the original project concept and history (but not necessarily in contradiction with the literal text of the Memorandum of the President), the Region shifted its stress to the adequate compensation for revenue lost from reducing the export taxes, and to interpret the land tax as only one possible tax instrument, inter alia, to generate such compensatory revenue. In October 1987, Regional management and the Government agreed on new bases for release, accepting a recently-announced tax package as an adequate substitute source of revenue to replace foregone revenue from export taxes. 5.39 It still took the Government another six months to assign tractors and agricultural machinery to the "automatic entry" category, but import tariffs on agrochemicals were not reduced 21/ Government emphasized that a linkage was required in order to obtain the requisite amount of fresh financing. 22/ The undeclared moratorium affected all payments to commercial banks on account of old debt, renegotiated debt (1985- 1987) and "new new" money (1983; 195 and 1987). Payments on arrears totalled USD 1.95b as of December 31, 1988, and USD 3.56 as of June 30, 1989. Argentina continued making payments to international organizations (World Bank, Inter American Development Bank, and Fund), and to commercial banks on account of commercial credit lines; loans under the Trade Credit and Deposit Facility; private sector debt, and Exit Bonds and New Money Banks issued in connection with the April 1987 Agreement. Machirea J.L. and J. Sommer. El Maneio de la Deuda Externa en Condiciones de Crisis de Balanza de Pagos: la Moratoria 1988-89 [External Debt Management under Balance of Payment Crisis: the Moratorium of 1988-89], CEDES, Document No. 59, Buenos Aires, 1990. 23/ This was not the first time the USA Treasury had made such a loan to Argentina; in 1987, a similar bridge loan had been made towards the Fund's Compensatory Finance Facility, which in the event put pressure on the Fund to release such funds. - 23 - due to pressing fiscal demands and Fund conditionality. Thus, the Bank could point to compliance with three conditions, but the other two were now out of reach. Management had no option but to go to the Board for a waiver of conditionality. 5.40 The Board's Involvement on Second Tranche Release. For the first time ever, in May 1988 Bank management had to request an approval of waivers of second tranche conditionality L'. It was not the best of years. The Fund's Board discussion on Argentina (March 1988) had not been smooth, with staff being accused of being too soft in accommodating towards the country. Also in March, the Bank's Board had approved the Banking Sector loan after a long discussion centered on Argentina's economic and financial situation (described by Regional management as "the most difficult that any middle-income, highly-indebted country faces"), the role of adjustment lending to address such situations, and particularly, the nature of second tranche release conditionality and its balance with proper front-loading. Earlier in March, the Board had analyzed adjustment lending as a whole !'. Almost all possible aspects were covered in the five-hour discussion, including the role of front-loading and the legitimacy of Bank lending for adjustment - which included quotations from the September 20, 1946 Board meeting on the purpose and objectives of the Bank. And then, relations between the Bank and the Fund on Argentina were souring - eventually leading to the open September 1988 confrontation. 5.41 It was against this background that, on June 2, 1988, the Board sat to consider management's recommendation. The discussion paid little attention to the specifics of the Argentina case, focusing instead 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, compliance with which was a contract condition and without which the contract had to be cancelled. They tended to see the question on "project" grounds, asking themselves whether the individual operation should continue or be canceled. In the specific case of Argentina, the borrower had not complied with two critical conditions and thus, in their view, the loan had to be canceled. 5.42 The second group, in contrast, considered adjustment to be a longer-term process, supported by the Bank 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 being a part of such series. Thus, failure to fulfill all 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. Under either position, a secondary issue was what to do with conditionality negotiated at the onset of the project if conditions had changed during implementation. 24/ A second occasion soon followed, for the Congo SAL, in November 1988. 25/ The basis for the discussion was the "Interim Report on Adjustment Lending". After the Board discussion and further elaboration, the document was completed and distributed to the Board as the "Report on Adjustment Lending" (RAL 1; Report N* R88-199), on August 8, 1988. - 24 - 5.43 The debate was heated. In addressing particular concerns expressed by several of the chairs, management committed itself to prepare a paper on administration of conditionality in adjustment lending, and to keep the land tax issue alive in lending and policy dialogue with Argentina. This audit found no evidence that management delivered on either. 1' VI. MAIN LESSONS 6.1 The loan was among the early crop of sector adjustment lending operations and yields some important issues. As indicated in the relevant paragraphs, many of them have already been internalized in the Bank and are now covered by revised procedures. The most important lessons are: (i) If a new tax is going to substitute for an existing one, conditionality barring the re- introduction of the latter should be included (paras 2.10; 3.4 and 3.20); and all affected parties must be explicitly contacted and the conditions for the tax swap negotiated and agreed (para 5.16); (ii) Ownership by Government must be explicitly confirmed (para 2.12); if there are disagreements within the Government (a frequent occurrence), Bank staff should confirm that those in favor command the high ground (paras 1.11; 3.19; 5.18 and 5.22); (iii) Warnings that the policy package being negotiated may not be legally institutionally or politically feasible should be analyzed (paras 2.11-2.14 and 2.18); the political viability of a loan or credit should be assessed as part of overall appraisal procedures (paras 2.15 and 5.33); (iv) A basic agreement with the Fund on the overall direction of adjustment and on the particular policy tools to be utilized for stabilization and for adjustment is critical (paras 2.20; 3.6; 3.18; 5.9 and 5.25); (v) Result of expected legislative action cannot be defined implicitly or explicitly as a tranche condition, for the Executive Branch of many governments cannot commit itself to such decisions (paras 3.2; 3.8 and 3.12); (vi) Sector adjustment cannot proceed satisfactorily unless there is a stabilization program in place, including a macroeconomic framework (para 5.8-5.9); such macroeconomic framework must be provided by the respective country operations division (para 5.21); 2f Both the Bank and the Government burnt themselves badly on the whole process. Not a single reference to either adjustment in the agricultural sector, the ASAL itself, or a federal land tax can be found in any Bank file or document after release of the second tranche, not even in the 1988 CEM; the 1988 PER, the green cover report on Public Finance Review (Report NO 10827, of July 21, 1992); or, except for half a line in the Executive Summary, the one on Agricultural Sector Review (Report No 7733-AR, of June 30, 1989). - 25 - (vii) The Bank should not accept "passive" cross-conditionality on its actions, decisions, commitments or disbursements, for default on the part of somebody else eventually falls on the Bank shoulders (para 5.36) U; and (viii) A clearer policy on the nature of Bank support to adjustment (including the issue of what is paramount, the individual adjustment operation or the overall adjustment process, and rules on the administration of conditionality) is necessary for smooth Bank operations (paras 5.42-5.45). L/ The issue is different with "active" cross-conditionality, i.e., when the Bank is one of the parties in a cofinancing arrangement and participates fully as such, often as "syndicate" leader. There the Bank decides on its own will to get involved, discusses the issues, and takes the ensuing risks.  - 27 - FIG URE I EXPORT TAXES ON GRAINS AS OF JUNE EACH YEAR 0.1 0.4 Jn.84 Jun.8s Jun.86 Jun.87 Jun.88 Jun.89 Jun.9o Jun.91 Jun.92 Identfi~cation Appraisal Effecbveness Release Loen of 2nd. Closing Tranche Source: DA TAFIEL, Fundación de Investigaciones Económicas Latinoamerican.  - 29 - FIG URE 2 ARGENTINA: CURRENT LAND TAX RATE PER PRO VINCE (%) 8 MMø U 62~ .4 - 1 - o z" ø % 0 -. -0 .3 C., la o..u 0-- -U 4U-.~.. ~4 *a - 4.... e .U.0w.a eo *O6 o N .4 • s gøU~ eUUC UU Ma. Z,oami:mmm am.ø Note: Some provinces have a single rate, appliedproportionally on the value of land Otherprovinces have a basic (minimum rate applied as above, plus a progressive rate on top of it, which could go up to the maximum indicated in the figure above.  - 31 - TABLE 1 ARGENTINA: STRUCTURAL ADJUSTMENT LENDING Loan No. Name Amount Approval Closing Current Status I_ IJUSD m. IDate Date A. Adjustment Loans 2675 Agricultural Sector Loan 350.00 04/03/86 06/30/89 Evaluated by OED 2815 Trade Policy and Export 500.00 05/05/87 12/31/90 Closed Diversification Loan 2923 Banking Sector Loan 400.00 03/29/88 11/30/89 Never signed 2996 Second Trade Policy Loan 300.00 10/27/88 12/31/90 Closed 3280 Provincial Development Loan 200.00 12/18/90 12/31/96 First tranche released 3291 Public Enterprise Reform 300.00 02/12/91 12/31/93 First tranche released Adjustment Loan 3394 Public Sector Reform Loan 325.00 07/30/91 12,31/93 First tranche released B. Technical Assistance Loans Related to Adjustment 2712 Sector Management Loan 18.50 06/03/86 12/31/91 Closed 3015 Tax Administration Technical 6.50 01/24/89 09/30/93 Under implementation Assistance Loan 3292 Public Enterprise Reform 23.00 02/12/91 06/30/95 Under implementation Technical Assistance Loan 3362 Public Sector Reform 23.00 06/25/91 06/30/95 Under implementation Technical Assistance Loan 3460 Second Tax Administration 20.00 04/14/92 02/28/92 Under implementation Loan  - 33 - & 1 3 3 3 - - _ _ _ _ _ _ _---------- 1[ 1 i~ z ~ 1 1 1 ci