Tanzania - Export Rehabilitation Program Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 7347 PROGRAM PERFORMANCE AUDIT REPORT TANZANIA EXPORT REHABILITATION CREDIT (CREDIT 1133-TA) June 30, 1988 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. LIST OF ABBREVIATIONS AND ACRONYMS AIAAC - Agricultural Imports Allocation Advisory Committee BET - Board of External Trade BOT - Bank of Tanzania EAPSA - Agricultural Projects Department (World Bank) EAIDA - Programs Department (World Bank) ERC - Export Rehabilitation Credit LC - Letter of Credit MDB - Marketing Development Board NBC - National Bank of Commerce NMC - National Milling Corporation OED - Operations Evaluation Department (World Bank) PCR - Program Completion Report PPAR - Program Performance Audit Memorandum SAA - Special Agricultural Account SAC - Structural Adjustment Credit SAP - Structural Adjustment Program FOR OFFICIAL USE OLY THE WORLD SANK Washington. D.C. 20433 U.S.A. Ofthe alf Duecto-Ce1111r4l Operatfs Evaktion June 30, 1988 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Program Performance Audit Report - Tanzania Export Rehabilitation Credit (Credit 1133-TA) Attached, for information, is a copy of a report entitled "Program Performance Audit Report - Tanzania Export Rehabilitation Credit (Credit 1133-TA)" prepared by the Operations Evaluation Department. Yves Rovani by Ram K. Chopra Attachment This document has a restricted distribution and may be used by recipients only in the performance of their oficial duties. Its contents may not otherwise be disclosed without World Bank authorization. PROGRAM PERFORMANCE AUDIT REPORT TANZANIA EXPORT REHABILITATION PROGRAM (CREDIT 1133-TA) TABLE OF CONTENTS Pate No. Preface ............................. . . . . . . . . . . . . . . . Basic Data Sheet ............................................... ii Evaluation Suimary .......... ....*............................. vi PROGRAM PERFORMANCE AUDIT MEMORANDUM i. Background: Economic Decline ............................. 1 ii. Program Design .....os................4............0............ 2 ,.iii. Program Execution .........9......... ........ ............4.. 4 IV. Program Impact ..................*............ .. ..... 5 V. Nain Issues ...................................4449..... 6 A. Program Concept .................................. 6 B. Program Design .................. .................. 10 C. BenefSts of the Program .......................... 11 PROGRAM COMPLETION REPORT i. Background ........................9*............... 15 II. Bank Response .................... ............ 16 III. The Export Rehabilitation Credit ............. 21 Objectives and Design ......... .... .... .......... 21 Implementation ................. ........................ 23 IV. The ERC in Retrospect ...................................... 25 Economic and Policy Effects .............................. 25 Role of the Bank ......................... 28 Design Weaknesses .................................. .... 30 V. Summary and Conclusions ...... ..... ... .. ..... ... 31 Annex .......... .. . .. .. . .. .. . ................. 33 Map - IBRD 19829 This document has a restricted distribution and may be used by reciplents only in the performance of their official duties. Its contents may not otberwise be disc~osed without World Bank authorization. PROGRAM PERFORMANCE AUDIT REPORT TANZANIA EXPORT REHABILITATION PROGRAM (CREDIT 1133-TA) PREFACE This is a performance audit of the Export Rehabilitation Program in Tanzania, for which Credit 1133-TA in the sum of US$50 million was lent to Government in the expectation that the decline in volume of traditional agricultural exports would be reversed and the rate of growth of non- -traditional exports accelerated. The program was approved on April 24, 1981, became effective May 12, 1981 and closed March 31, 1983. Final disbursement took place on March 4, 1985. The PPAR consiits of a Program Performance Audit Memorandum (PPAM) prepared by the Operations Evaluation Department (OED) and of a Program Completion Report (PCR) dated December 15, 1986. The PCR was prepared by the Africa Regional Office. The PPAM is based upon a review of the President's Report (No. P-3009-TA) dated March 31, 1981, the Credit Agreement and Memorandum of Understanding, both dated April 24, 1981 and the PCR; correspondence with the Borrower and internal Bank memoranda on program issues as contained in relevant Bank files have been consulted and Bank staff associated with the program have been interviewed. An OED mission visited Tanzania in February-March 1988. Discussions were held with officials of the Ministry of Finance, Ministry of Agriculture and Livestock, and the Bank of Tanzania. Field visits to regions, important for the production of major export crops, were undertaken and the impact of the program discussed with regional officials, cooperatives and farmers. The valuable assistance provided by the Government of Tanzania and staff met during the preparation of this report is gratefully acknowledged. The PCR well covers the experience of this essentially failed program and pays particular attention to fts prolonged evolution, implementation and the shared disappointment of both the Bank and the Borrower over the outcome of this Credit. The PPAM agrees with the comprehensive review provided by the PCR and expands on the weaknesses in both program concept and design and the benefits of Credit 1133-TA. As is customary in the preparation of audit reports, copies of the draft audit report were sent to the Borrower for comment on April 27, 1988. No comments were received. - 11 - PROGRAM PERFORMANCE AUDIT REPORT TANZANIA EXPORT REHABILITATION PROGRAM (CREDIT 1133-TA) BASIC DATA SHEET KEY PROGRAM DATA Appraisal Actual or Actual as X Estimate Current Estimate of Apprilsal Program Cost (USS million) 168.00 Not known /a . Credit Amount (SDR million) 56.0 56.0 1o Agreement Date 04/24/81 Closing Date 08/31/82 63/31/83 SUMMARY OF COMPLIANCE WITH AGREED CONDITIONS Reference Lb Summary Action DCA 3.05 Debt Reporting Unit to be No major Improvement at the Bank of established in the Bank of Tanzania Tanzania. by Sept. 30, 1981. DCA 8.66 Export Credit Guarantee Program to Decision to introduce the scheme was be implemented by Dec. 81, 1981. taken in late 1981, but implementation was delayed. DCA 3.67 Prompt staffing of Bureau of Recruitment procedures started, with External Trade. approval of Third Technical Assistance Project in February 1982. DCA 3.68 Plan for strengthening agricultural Initial plan submitted in December 1981 support services by Sept. 80, 1981. proved unsatisfactory, but followed by an exhaustive review of agricultural policy published in 1982. DCA 8.69 Establish a Special Agricultural Account established by April 1, 1981. Account in the Bank of Tanzania and: 1) contribute USISO million of the Government provided at most US821 million Government's own resources to Vn* but total non-Bank resources (i.e., account (in addition to proceeds of including supplier credit and bilateral the credit), in the period Aprl' 1, assistance) was USS6 million. 1981 through March 81, 1982 I - III - 1i) annually thereafter contribute No contribution as a function of export 15% of the export proceeds of earnings. (Though ad hoc agreements coffee, tea, alsal, cotton, tobacco reached with individual exporters over and cashownuts to the account. the next 6 years to allow retention of a percentage of export earnings). DCA 8.16 Study of producer prices by August Consultant study completed, but 81, 1981, and no prices annually Government delayed decisions on full thereafter vith adequate incentives. range of recommendbtions. MOU 2 Study and Understanding with IMF on 1) exchange rate policy Understanding with IMF did not take II) interest rates place. ill) Improved efficiency for parastatels. MOU 4 Review of public Investment program by June 30, 1981 to 1) meet the resource requirements of Effect no' clear. the agricultural sector ii) defer some projec6 preparation Substantial cuts were made in 1981/82 and and completion 1982/88 development budget; however, the project selection procedure was still unsystematic. il) increase allocations for Effect not clear. operations and maintenance Iv) criteria for future investment Criteria gving priority to almost all projects (to emphasize food and possible projects Introduced as part of export commodities). the Structural Adjustment Program in 1982. MOU 5 Improved budgetary control by June B8, 1981, including: 1) rolling 2-year budget Reintroduced in 1981/82, but value of second year figures not clear. ii) estimates of recurrent cost Effect not clear. Implications of Investment proposals ill) Increase current capacity Effect not clear. utilization by provision of deverred malntenance. MOU 6 Improved foreign exchange allocations by: 1) quarterly projection of foreign Six monthly allocation procedure exchange earnings Instituted; and communicated to Bank 6 to 6 months late. ii) quarterly budget of allocations Not done. for next 18 months iii) allocate foreign exchange to Not clear what priority was given to meet the needs of major producers in agriculture. Allocations were 4 to 8 the agricultural sector. months late. - iv - MOU 7 Improve agricultural policy by: 1) *Ilow crop r-ocurement and Input Not done. delivery by non-parestatal organizations 11) Improve emallholder incentives Inadequate action. ill) Improve and diversity transport Little progress. services IV) recruit technical assistance for Thrd Technical Asstanco Credit Ministry of Agriculture. approved in February 1982. Considerable subsequent progress in filling key parastatal posts. MOW 9 Improve parastatal operations by: 1) Identify and overcome the Studied by Parliamentary Task Force. weaknesses of the prasataes il) complete task force reports on Reports completed. Cooperatives National Milling Corporation by May, reintroduced as substitutes to 1981 and on Cooperatives by Dec. parastatals. Subsequently major 1981 Improvements In management of the National Milli1"g Corporation. it) critical vacancies to be filled Action delayed by a year, but undertaken by Dec. 81, 1981 with assistance from Third Technical Assistance Project. Iv) Improve financial control and Parliamentary Task Force establith*. and audit. timetable for updating accounts adopted in mid-191.: Som progress In catching up on accounts. Mau 9 Legislation for payment by results Some progress. by Sept. S u, 1981 MOW 0 Reduce scope of price management Some progress. procedures. MOW 1o Budget to bring revenue growth In Some progress. line with growth In GOP, and reappraisal of subsidiary programs. MOa 11 Hold Consultative Group meetings. Held eventually. MOU 12 Appraisal Of a follow-up ank did not agre to appraise. Rehabilitation Credit In pt. 1981. CUMULATXVE ESTIMATED AND ACTUAL DISURSEMENTS (USS Million) FY81 FY82 FY88 FY84 FY85 Estimated at Appraisal -- S9.0 -- -- -- Actual 5.00 /c 24.40 11.29 0.560 4.!d Actual as N of Retroactive Financing -- 568.80 81.49 62.40 90.96 Date of Final Dilburseament: March 4, 1985. MISSION DATA Month/ No. of Weeks Specializations Performance Rating Type of Year Persons in Field Represented (d Status (e Trend /f Problems /g Preparatlon 87/79 8 2.8 - Appraisal 11/6 16 2 - - - - Supervision I s8-84/81 8 8 AE(2), TE - - - Supervision II 65/81 2 2 E,D - - - Supervision III 97/81 1 1.5 TE - - - Supervision IV 09-12/81 2 1 E(2) 2 - - Supervision V 07/82 1 4 TE 8 8 M,P /a Actual Government cont-Ibution not known. A DCA = Development Credit Agreement, April 24, 1981 MOU - Memoranduw of Understanding, April 24, 1981 Le Retroactive financing. L AEcAgricultural Econweist/TExTransportation Economist/E=Economiet/D=Disburseent Officer. A 1=problem free or minor problems/2=moderate problems/8amajor problems. If =lamproving/2=.s6tionary/8=deteriorating. Lg P=Political/TTechnical/M=Managerial/FaFinancial/aOther -vi - PROGRAM PERFORMANCE AUDIT REPORT TANZANIA EXPORT REHABILITATION PROGRAM (CREDIT 1133-TA) EVALUATION SUMMARY Introduction The Export Rehabilitation Credit (ERC) was a response to government requeats for a third program credit at a time when performance in agriculture, tbe dominant sector of the economy in terms of outrnt, employment and exports had deteriorated and export crop production was falling. Initially discussions between the Bank and government centered on IDA support for a five-year Structural Adjustment program, but the policy changes which government was prepared to undertake were deemed insufficient and a limited first operation concentrating on agricultural export rehabilitation was agreed upon. Objectives The program expected to reverse the decline in the volume of traditional agricultural exports and to increase the rate of growth of manufactured and processed exports. This was to be achieved by: (i) improving producer returns to farmers from major export crops; (ii) ensuring foreign exchange financing .f high priority imports for the operation and maintenance of the existing productive capacity of the agricultural sector; (iii) improving the efficiency of agricultural parastatals; and (iv) introducing higher incentives for the production of non-traditional exports. Supplementary and related measures includeds (v) improvements in the foreign exchange budgetary and import allocation system; (vi) a review of all on-going and proposed investment projects; (vii) strengthening the planning, programming and budgeting functions of government; and (viii) the preparation of an action program to improve the agricultural services including research, extension and other support services. The program, which provided US$50 million of IDA support was the largest in the Bank's agricultural portfolio. It was approved by the Board on April 24, 1981 and was to be implemented by the Bank of Tanzania, disbursed over a period of six months. Implementation Experience Prior to credit effectiveness, government (a) abolished export taxes on coffee and sisai and the production tax on tobacco as the first part of the program to restore producer incentives; and (b) established a Special Agricultural Account (SAA) in the Bank of Tanzania which was - vii - expected to be the conduit for the US$180 million estimated by the Bank to be the minimum annual foreign exchange requir=ments for the agricultural sector. The government was to contribute an additional US$50 million with the remaining US$80 million to be contributed by bilateral donors. Following credit effectiveness there were problems and considerable delays in implementing the ERC program. There was strong initial government resistance to bilateral donors' efforts to channel funds through the SAA allocation process. By the end of the credit period the nature and amount of government's contributions and the effectiveness of the Agricultural Imports Allocation Advisory Committee had become serious issues. It has been estimated that only US$118 million of the expected US$180 million has been channelled through the Account. Foreign exchange allocations for imp;rts fell far behind schedule which also affected IDA disbursements. The ERC closing date was postponed from March 31, 1982, when only 30% of the credit had been disbursed, to March 31, 1983; final disb-:rsement was made on March 4, 1985. Results The ERC did not achieve its objectives. As the PCR points out (para. 4.01), aggregate export earnings declined from US$563 million in 1981 to US$413 million in 1982 and to US$359 million in 1983, largely due to a continuing decline in exportable volumes of m%jor crops. Producer price increases and the provision of some of the estimated total foreign exchange requirements for the agriculture sector were insufficient to address declinirg agricultural production and exports. Sustainability Program concept was too limited to address the range and magnitude of problems affecting Tanzanian agriculture; the original Bank preposal for a structural adjustment program should have been adhered to. Contrary to expectation, there was no follow-on project. As a result, the (already) waning government commitment fell away altogether. The lack of dialogue between government and the Bank which characterized this period of Bank/Tanzania relations did, however, provoke a rethinking of strategies and policies on both sides which has since paved the way for a type of operation which this program was originally expected to comprise. Lessons The main lessons of this program are: (a) the importance of dialogue with officials charged with policy making rather than assuming that technical staff would communicate the necessary policy decisions to the politicians (PPAM, para. 35); -viii - (b) the importance of a reliable, quantitative data base for (a) understanding the effects of po'icy interventions and investment and (b) as a basis for an agreed appreciation of trends in the economy (PPAM, para. 35); (c) total funding for the overall estimated direct and indirect import requirements should have been secured before the program began (PCR, para. 2.17); (d) a detailed mechanism for monitoring the day-to-day operations should have been worked out to avoid operational problems which arose from lack of commitment by the key agency, the Ministry of Agriculture (PCR, para. 4.07-4.08); (e) precise understanding should have been reached regarding the government's contribution to the SAA (PCR, para. 4.05); and (f) the project well illustrates the inherent difficulties when the Bank fails to speak with one voice (PPAM, paras. 27-29; PCR, paras. 4.10 and 4.16). - 1 - PROGRAM PERFORMANCE AUDIT REPORT TANZANIA EXPORT REHABILITATION CREDIT (CREDIT 1133-TA) SUMMARY I. BACKGROUND: ECONOMIC DECLINE 1. Government first approached the Bank for structural adjustment support in April 1978; a period when the collapse of boom tea and coffee prices coincided with a rapid decline in Tanzanian economic performance and a balance of payments crisis. Some of the factors were external, while others can be traied to economic and political decisions taken during the seventies. While many of these decisions flowed from the Arusha Declaration, and the policy of African C-%cialism adopted by government, they were not taken in isolation. The Bank, and other important donors were well aware of the decisions being taken, and on occasion, Tanzania was singled out to the development conmmunity for praise in the way its development exemplified growth with equity (para 26). It was not until the end of the decade, well after the crisis was self-evident to all, that the Bank felt constrained to advise government that some proposed actions were felt to be inimical to the prospects for rapid and equitable economic growth. 2. External factors which contributed to the deteriorating economic performance included the severe drought in 1973/74; the oil price shocks of 1973/74 and 1979; the break-up of the East African Community and the war with Uganda in 1979; and a decline in the terms of trade. The agricultural sector report 11 puts this loss at about US$1.7 billion, or the equivalent of three years' exports; or two thirds the value of foreign aid over the previous decade. 3. Domestic factors which help explain the economic decline include policies which depressed agricultural exports, at the same time that the Basic Industries Strategy was biasing industrial production in the direction of import dependence. Agricultural exports were depressed by policies such as villagization, operation Maduka (-hich closed village retail shops), abolition of the cooperative marketing system, neglect of the infrastructure for research and extension, declining real producer prices, and reorganization of local government which undermined the tax base for the maintenance of rural roads. The Basic Industries Strategy meanwhile was leading Tanzania in the direction of increased import dependence, either because as in car or bicycle assembly the industry was dependent on the import of semi-manufactures, or for light engineering due to the need to import equipment and spares, or for fertilizers the need to import technical assistance and to pay royalties. Given an exogenous l/ Tanzania: Agricultural Sector Report No. 4052-TA, 1983. - 2 - source of demand for these industrial products, some of them would have resulted in some import replacement; however, in the absence of ample foreign exchange throughput steadily declined. 4. Three factors combined to disguise the emerging balance of payments crisis. First, the poor agricultural performance could be explained by (a) seasonal factors; and (b) startup problems as the new parastatcls established themselves, and farmers adjusted to the restriction of village life. Second, the 1970s saw a rapid expansion of donor support, so that declining export revenue was compensated for by increased aid. Third, the tea and coffee boom of 1977/78 meant that export revenues for these crops expanded even in the face of declining volume of coffee exports. All of which was compounded by an IMF/Bank recommendation, coinciding with the end of the coffee boom, for liberalization of import restrictions which largely exhausted official reserves. 5. The decline when it came was disastrous with the balance of merchandise trade falling from TSh -1.91 billion in 1977 to TSh -5.8 billion in 1980; the overall balance going from a surplus of TSh 0.9 billion in 1977 to a deficit of TSh -1.9 billion in 1978.2/ Even these changes understate the real stress on the economy, since the severity of direct controls was increased as the shortage became more severe; and increased reliance was placed on suppliers' ctedit, and other "once only" sources of foreign exchange. Decreased control of imports led to a loss of effective control as market forces were increasingly replaced by short-term crisis management. II. PROGRAM DESIGN 6. Credit 1133-TA was the Bank Group's 24th lending operation to Tanzania for the agricultural sector and the third for non-project lending. 7. As approved by the Board in April 1981, the export rehabilitation credit consisted of measures to ensure: (a) priority allocation of imports to the agricultural sector through: - the establishment of a Special Agricultural Account (SAA) at the Bank of Tanzania to meet full direct and indirect agricultural import requirements, estimated to be US$180 million p.a.; - funding of the Account to be secured in the first year, by government to provide US$50 million in matching funds and to seek additional bilateral funding to cover the US$80 million shortfall; in following years, government to assign to the Account at least 15% of receipts from the major export crops; - the agricultural sector to receive priority in the allocation of fuel; 2/ Tanzania: Agricultural Sector Report No. 4052-TA, 1983, Statistical Appendix, Table 1.1 -3- - the Ministry of Agriculture to have a representative on the Import Licensing Advisory Committee of the Bank of Tanzania; (b) promotion of nor-traditional exports through the payment of incentives pending understanding being reached under the IMF standby arrangement on an appropriate exchange rate policy; (c) improved procedures for foreign exchange budgeting and the establishment of a unit to monitor all loans and grants to predict future debt service obligations more accurately and to determine systematically import allocations; (d) a review of ongoing and future development projects to enable funds to be reallocaLed for import support and/or operation and maintenance of existing operations; (e) the introduction of a two-year rollig budget; (f) an action program to be prepared by government on ways to increase agricultural productivity, including how to improve agricultural services, research, extension and organization of other support services in the sector. 8. The credit provided US$50 million to finance 100% of the foreign cost of essential imports in four categories: agricultural chemicals, agricultural machinery, agricultural packing materials and vehicle spare parts. It was to be implemented by the Bank of Tanzania in consultation with the Ministry of Finance. The credit was to close in March 1982, eleven months after Board approval. There were no cofinanciers. 9. A Memorandum of Understanding on Follow Up Measures was signed the same day as Board approval of the Credit which provided for additional action to be taken within the overall framework of the two year IMF standby arrangement which had been approved in September 1980.3/ These comprised a review of the public expenditure program; increased provision in the recurrent budget for operation and maintenance; improvements to the system of foreign exchange budgeting and import allocation; a study of agricultural policy reform; improved parastatal operations; introduction of measures to stimulate productivity and the mobilization of domestic and external resources. Successful completion of this Export Rehabilitation Credit and the attached Memorandum would result in a follow-on program to be appraised in September 1981 - five months after credit signature. 10. Principal binefits of the program were expected to be a halt in the decline of agricultural exports while also setting the stage for long term export growth. Volumes of traditional exports were expected to recover by 9% p.a. in 1981 and 196L principally in the coffee, cotton, sisal and cashew exports. The combined value of additional export earnings 3/ although at the time of Board approval, Tanzania was in breach of this standby arrangement. from both the agricultural sector and the non-traditional exports sector was expected to be US$35 million in 1981 and US$80 million in 1982 with additional benefits expected in imports savings from increased domestic food production. 11. Risks identified were that even if export performance improved, the balance of payments situation might be such that a substantial resource gap would still remain in 1981 and 1982. Other risks included the possibility that technical assistance - to be supplied under an alternative credit - would be provided too late and that measures to secure the long term recovery of the export sector would not be taken by government in a timely manner. III. PROGRAM EXECUTION 12. The PCR (paras 3.04-3.07) well describes the history of program implementation. In summary, although the institutional mechanisms to assure priority allocation of imports co the agriculture sector were put in place, the expected results did not materialise while the remaining conditions of the credit agreement were either not met or not implemented. 13. The institutional mechanisms for allocating imports to the agriculture sector, the Special Agricultural Account in the Bank of Tanzania and the Agricultural Imports Allocation Advisory Committee (AIAAC) in the Ministry of Agriculture were established, as expected, prior to credit approval; the Import Licencing Advisory Committee was strengthened. 14. Import allocations were, however, consistently behind schedule initially due to procedural delays within the Bank of Tanzania, latterly to (i) suspension, in October 1981, of Bank disbursements as well as to (ii) dwindling government support for both the SAA and AIAAC once it became evident that the Bank was not going to appraise a follow-on project. By the original credit closing of March 1982 only 30% had been disbursed. It was a further two and a half years before final disbursement was made. 15. Actual government contribution to the Special Agricultural Account remains in doubt; reflecting in part government resistance to the idea of a fixed contribution (PCR para 4.09). The PCR cites four different figures ranging from US$62.4 million to US$13.2 million. No data is available from the Bank of Tanzania. The PCR estimates that the Account served as a conduit for US$118 million (comprising IDA support of US$50 million, maximum government contribution of US$21 million, the remainder from bilateral sources) rather than the US$180 million estimated at appraisal. The azechanism to guarantee future funding of the Account for agricultural imports was not put in place. 16. Actual allocations to agriculture within the four categories varied substantially from that expected at appraisal with a much higher than expected contribution being made for agricultural chemicals, principally coffee fungicides (+465%), offset by lower than expected allocations for agricultural equipment (-20%) and vehicle replacement parts (-45%). -5 - 17. Measures to promote non-traditional exports were not successfult the incentive scheme was not introduced as expected while the simplification and centralization of export clearances were similarly not implemented. Progress in the remaining conditions of the credit agreement was minimal, largely due to weak government commitment: - the necessary instruments for foreign exchange management were not put in place by the Bank of Tanzania; - the draft export credit guarantee program was deLayed in implementation; - the pricing study was completed but decisions on the recommendations was delayed; - the plan for strengthening the agricultural support services was delayed and then proved unsatisfactory; - little progress was noted in respect of the Memorandum of Understanding (see Basic Data Sheet). 18. A serious review of the agricultural project portfolio was undertaken which proved beneficial (para 43) but in declining to appraise the expected follow on program the Bank concluded that overall the program had failed to reach its stated objectives. Credit closing was postponed to March 31, 1983 and then delayed informally three times, to allow commitments to be disbursed, final disbursement took place on March 4, 1985. IV. PROGRAM IMPACT 19. The stated objectives of the credit were not reached. Aggregate export earnings continued to declines from US$563 million in 1981 to US$413 million in 1982 and to US$359 million in 1983. This was due largely to a drop in the exportable volumes of major crops. Foreign exchange earnings from each of the major export crops were lower in 1983 than in 1981. Few of the expected changes deriving from policy reform were effected since most were not implemented. Reasons for the program not reaching its objectives can be traced to weak government commitment, itself an integral part of the flawed program concept and design (paras 22-40). 20. Sustainability of the program is unlikely. As the PCR points out, government could see no useful purpose in continuing to support the institutional mechanisms which were established under the auspices of the credit - the SAA and the AIAAC; Bank support similarly fell away when it became evident that the government had no intention of introducing any of the policy reforms which were expected to be accomplished through the provision of the credit. The program did, however, provoke a rethinking of policy on both the part of the Bank and the borrower which has since resulted in a broad based program intended to address the deep rooted problems of the economy in general and the agricultural sector in particular. -6- V. MAIN ISSUES 21. This credit marked a turning point in Bank lending for agriculture in Tanzania. Prior to this operation, the project pipeline had increased rapidly, peaking between 1978-81 when this credit was being negotiated: seven projects were approved for total support of US$127.5 million. Thereafter, Bank lending to agriculture stopped altogether for four years and when resumed took the form of a multi-sector rehabilitation program. Program experience, therefore, has to be viewed within the context of a broader time perspective than that which constitutes immediate program history. A. Program Concept 22. Negotiations for Credit 1133-TA extended over three years. Although compromise was eventually reached between the Bank and the borrower both parties were left unsatisfied; the credit as negotiated did not fit well with prevailing economic conditions in Tanzania. From the outset, and for the following two and half years, the Bank correctly identified the need for a broad-based reform program but the credit which finally emerged in April 1981 was limited to the agricultural sector. This change was a fundamental flaw in program concept given the economic decline of this period (paras 1-5). Government also was anxious to have a program which addressed both short and long run problems of the economy seeing the operation as providing necessary support to adhere to IMF conditionality. 23. This lack of a Structural Adjustment program, however, was due to what has been described as "a fairly intensive dialogue of the deaf" going on between government and the Bank during these years. Essentially government's proposals were broad statements of intent, not action oriented programs based upon conditionality which, for the Bank, were central to any new loan. The Bank meanwhile was sending conflicting signals to government, different mission leaders headed each new mission, and in groping for the "right" policy package differing prescriptions were put forward by the programs and projects divisions. 24. From the beginning of the negotiations with the Bank, the Tanzanians considered that the cause and extent of economic difficulties were principally due to exogenous variables and not to domestic policies. This in part because of the assessment of the prevxous balance of payments crisis and its apparent resolution with the boom years of 1977-78 and second because of the Bank's uncritical support during this period for precisely those policies which were now regarded as contributing to declining agricultural performance. Government, therefore, did not see the need for either policy reform or conditionality seeking instead a speedy injection of external resources which was also regarded as necessary to adhere to IMF conditionality. 25. After almost two years of negotiation, government was, therefore, taken aback with the Bank's request for extensive and additional government action (over 40 separate issues were cited) when it presented its Economic Recovery Program in May 1980. Its subsequent, revised program in September for a five-year program of assistance of US$500 million, was similarly -7 - deemed insufficient by the Bank as the basis for an agreed program. The extent of conditionality expected by the Bank came as an understandable shock to the Tanzanians. 26. Throughout the seventies, Bank country strategy to Tanzania was one of consistent and positive support for government policies. The Bank described government development priorities as being well conceived;A/ in later years, the Bank concluded that there were no fundamental differences between the Bank and Tanzania on development objectives nor any differences of opinion on important policy measures.5/ While the Bank also noted that economic growth6l was not as good as expected the conclusion was that Tanzania nevertheless deserved maximum Bank group effort. In the late seventies, when the Bank noted that there was need for a better balance between allocation of resources and priority investment, cautioning against further capital intensive, low priority projects, the Tanzanians replied that almost all such projects in fact had been supported by the Bank.7/ Even as late as January 1980, the Bank reiterated that Tanzania's development objectives were sound and in line with the country's needs.8/ Such statements were supported by a substantial increase in the project pipeline throughout the period.9/ 27. The Bank, for its part, expected to finance a structural adjustm.-nt operation; several differing policy prescriptions were proposed over the years. When agreement on conditionality could not be reached, the agricultural projects division continued to argue the case for policy reform in agriculture. 'his was then added to the credit at a late stage and lacked any real government commitment since agreement in principle on a more restricted program had already been reached between the Bank's programs division and government. 28. Preliminary discussions between the Bank and government for the proposed SAL began in April 1978 and were intermittent until the ongoing negotiations with the IMF reached an impasse one year later. In early 1979, with government increasingly concerned about the deteriorating balance of payments situation, the Bank took the initiative and sent a preparation mission. Specific proposals for a framework of program lending covered public investment and its financing, trade policy reform and 4/ Country Program Paper 1973. 5/ Country Program Paper 1975. 61 Estimates of which were consistently inflated by the Bank in assuming a 9% growth of subsistence agriculture. 7/ McNamara's meeting with the Tanzanian delegation, October 25, 1978. 8/ Country Program Paper, January 1980. 9/ Twenty two investment projects covering export crops: tobacco, tea, cotton, cashew, sugar, coconut; food crops; dairy and livestock activities; fisheries and forestry were approved between 1965-81. - 8 - parastatal efficiency.10/ A second and new Bank preparation mission (under a different mission leader) followed in March, which concluded that the magnitude of Tanzania's balance of payments difficulties were such that any adjustment credit must cover a wide spectrum: (i) fiscal reform within, and additional to, the overall framework of IMF discussions; (ii) balance of payments, with particular emphasis on export promotion and the broader aspects of trade policy including the rate of exchange and related incentive policies; (iii) strengthening of agricultural administration, supporting services and improving production price incentives; (iv) improved regional planning and transportation; and (v) enhanced parastatal performance. The Bank continued to expect that such adjustments would take place in parallel with the Fund. The program was estimated to require US$50 million annually over several years; a Consultative Group meeting was also to be convened to assist with the remainder of the resource gap; and the credit was entered in the lending program with the expectation that a second Structural Adjustment Credit would follow.11/ 29. Meanwhile the Bank office ir Tanzania was calling for an even broader approach; in addition to addressing macro economic constraints and poor productivity in agriculture it proposed support for increasing productivity in the industrial sector.12/ 30. When it was concluded at the Annual Meetings in 1980 that, in the absence of government agreement, a SAC would not be possible, the Bank, despite its earlier assessment of the need for broad based structural support, compromised with a limited program. As appraised in November 1980, the credit was expected to focus on export rehabilitation. While there were supplementary measures the Bank dropped the recommendation for strengthening the agricultural supporting services.13/ Government meanwhile was still very anxious for support to help meet IMF obligations due by the end of 1980. 31. The projects division continued to believe that the agriculture sector in Tanzania had to be fundamentally restructured. It did not consider that it would be enough merely to plan and implement bet .er the 10/ Aide memoire for a Framework of Program Lending in Tanzania dated June 8, 1979. ll/ Structural Adjustment Initiating Memorandum, May 9, 1980. 12/ Letter dated September 20, 1980. 13/ Appraisal Aide Memoire, November 1980. - 9 - existing government strategies which essentially was what government proposed to do. Instead, the projects division believed that priority had to be given to strengthening agricultural research and extension, manpower training, improving agricultural planning, clarifying the respective roles of state versus peasant production, the use of tractorization and targeting investment in regions with high, not low, production potential. 32. When it then became evident that these policy changes were not to form the core of the proposed Bank credit, the projects division then argued that, at the least, the ERC should be used as a vehicle through which such dialogue could begin with government to secure these key policy changes; if not then the credit would be "pouring money down a fast drain".14/ At the last minute, therefore, a covenant was included requiring government to prepare an agricultural policy plan. 33. The final emergence of the ERC in March 1981, therefore, left all parties unsatisfied. The Bank had tried and failed to secure a SAC while the government felt that the Bank kept coming up with new and different policy packages while also unduly procrastinating at a time when foreign exchange shortages were acute and thus further jeopardize government's ability to meet the ceilings set by the IMF. 34. This then raises the question as to what alternative was open to the Bank. At that time, structural adjustment lending was a new Bank instrument and thus one for which the Bank was concerned to establish its credibility; given that the necessary conditionality could not be agreed, one alternative would have been to follow the policy which was subsequently adopted: halt all lending. This already had constituency within the Bank. The debate over whether the Bank should, in the face of little acceptance of the necessary conditionality, continue to lend to avoid further deterioration or to halt all lending until policy reform is agreed upon, is not new. 35. The lesson which is concluded by the audit, however, concerns the more fundamental question as to why the quality of the dialogue between government and the Bank was so poor as to preclude what prevailing economic conditions clearly called for: a structural adjustment program. In this respect, the Bank would have considerably benefitted if there had been reliable quantitative data on key performance indicators of the ecovnmy. Little such informetion existed. Data on public expenditure including actual investment trends was weak;15/ foreign exchange requirements for the 14/ Memo dated February 4, 1981. 15/ Early Bank missions were conducting extensive reviews in order to "provide estimates of the total cost of the existing and proposed public investment program so that comparisons could be made with available resources and a framework established for reconc.ling the two" Aide Memoire June 8, 1979. While in a memo dated March 5, 1981, it was noted that the Bank did not know what the actual allocations to agriculture and the supporting services had been in recent years. - 10 - agricultural sector were unknown;6/ even less information was available concerning the performance, liabilities and expenditures of parastatals 17/ which had a commanding presence in the agriculture sector. If ongoing Bank sector work had been directed towards such analysis, there is no doubt that this would have enabled predictive models of the Tanzeniai. economy to be constructed, which, at the time of credit negotiations, tould then have provide the basis for some agreei appreciation with government of trends in the economy as well as providing a rough, quantitative understanding of the effects of proposed policy interventions and investment. Finally it is also apparent that the Bank confined its discussions to government officials, not to the party despite its preeminance in policy making.18/ B. Program Design 36. The program as designed had a number of flaws. First, although a key objective was to halt the decline in agricultural export earnings by ensuring that vital foreign exchange financing of total import requirements was available, there was no provision in the credit agreement which would guarantee this. It is unclear why the Bank concentrated on ensuring that only US$100 million would be available if it considered US$180 million to be the minimum necessary for the sector as a whole. 37. Second, the fact that there is considerable uncertainty regarding government's actual contribution to the Special Agricultural Account and the fact that the final supervision mission raised the question as to the extent to which Bank support had substituted for government funding to agriculture, illustrate the weaknesses in (a) not ensuring payments made in parallel and (b) the monitoring of government allocation requirement. 38. Third, given that the Bank was also anxious to see a system put in place which would continue to secure the necessary import requirements to the sector after the closure of the credit one year later, the proposal put forward by the Tanzanians, but overruled by the Bank, to retain, from the outset, 10-15% of export earnings may well have been more sustainable than the once and for all US$50 million which the government was all along reluctant to agree to, and which as it turns out was not committed. And, as the PCR points out, this retention of export earnings would have, in any event, comprised almost US$50 million. 16/ The purpose of the MDB's study on "Estimates of 1981182 Import Requirements for the Production, Processing and Marketing of Major Crops in Mainland Tanzania", August 1981 was precisely to find out the size of total foreign exchange needs of the agriculture sector. 171 Aide Memoire, September 25, 1979 18/ When the government presented its second and revised economic recovery program to the Bank, the analysis to President McNamera stated: "We assume (my italics) that he (Minister of Finance) has convinced the political leadership of Tanzania that important changes are necessary but we are not sure of the extent of the commitment that the proposals enjoy" memo dated September 26, 1980. - 11 - 39. Fourth, the program was unduly optimistic as to the likely achievement of the longer-term policy actions when in essence the operati.on was to be a fast disbursing credit (six months). This then became even more unlikely when the Bank then failed to go ahead with the second program of support as had been expected. 40. Program experience also illustrates that, in the Tanzania scenario, while prices may be one factor accounting for production decline, reversing this trend takes more than "getting the prices right". Export volumes continued to decline over the period despite the government having made the necessary price adjustments prior to the credit agreement. C. Benefits of the Program 41. The very fact that there was little meeting of the minds between various divisions of the Bank and the government throughout the three years of negotiations, provided th2 impetus for all to reevaluate critically their respective strategies, which in turn formed the basis in 1986 for a broad-based structural adjustmeat along the lines which had been expected at the outset of this project. For the Bank's part, a serious effort was begun to address the gaps in the Bank's knowledge regarding the agricultural sector in Tanzania. At the beginning of 1981, a major agriculture sector review was initiated and prepared over a period of nine months and several missions. Meanwhile goverranent instigated in 1982 its own high level review of the agriculture sector which in turn led to substantial reformulation of government policy. 42. At the same time, and at the suggestion of the Bank, an Advisory Group was established in 1981. This was funded under the third Technical Assistance credit but originally proposed within the context of the ERC precisely in an effort to endeavor to restore some degree of dialogue between the Bank and government at a time when negotiations had once again broken down with the Fund. The establishment of this Group of three independent experts and a secretariat, largely comprising Tanzanian officials, proved successful; by interposing a third party--at sufficiently high level--it provided a more neutral channel for the transmission of ideas and also demonstrated to government that the Bank was willing to take advice as well as to give it. From this emerged the Tanzanian Structural Adjustment program which was approved by Parliament in June 1982. 43. This program loan also had a subsidiary benefit at the project level in continuing a serious review between government and the Bank on the performance of the agricultural portfolio first initiated in September/October 1980. As a result, it was decided against going ahead with the processing of some projects while the poorly performing projects were either cut back and/or subsequently closed down. 44. In conclusion, therefore, it can be said that the long drawn out negotiations which characterized this essentially failed credit did result in the Bank (i) reaching a decisive turning -point in its agricultural lending program to Tanzania; and (ii) substantiating and quattifying the issues which had to be addressed if the declining trends in the economy - 12 - were to be reversed. It is evident that this should have been done on a much more timely basis to have paved the way for an operation more appropriate to tackling the deep-rooted structural problems in the sector. - 13 - PROGRAM CCHPLETION REPORT TANZANIA EXPORT REHABILITATION PROGRAM CREDIT 1133-TA December 15, 1986 - 15 - TANZANIA EXPORT RERABILITATION PROGRAM (IDA Credit 1133-TA) PROGRAM COMPLETION REPORT I. BACKGROUND The Economy 1.01 Both internal and external factors contributed to a severe deterioration in the Tanzanian economy during the late 1970s. By 1980, growth in the monetized and productive sectors had come to a standstill, while export volumes were shrinking. Deficiencies in the policy environment and the chronic lack of foreign exchange for maintenance and repairs were causing extensive decapitalization, particularly in key agricultural and transport sectors. A series of adverse external events, including oil price increases, the collapse of the East African Community in 1977, the war with Uganda in 1978/79 and another year of poor rains in 1980, further aggravated the existing problems and made more urgent the need for restructuring the economy. 1.02 The annual growth r, te of real GDP, which had averaged 5 percent during 1973-1978, declined to 3.5 percent in 1979 and 2.6 percent in 1980. By then, GDP per head was falling. Meanwhile, the structure of the economy was changing, away from the productive monetary sector, towards subsistence and service activities. Excluding public administration, commereq and other services, the monetized sector grew by only 2.3 percent a year, well below the population growth rate. The pervasive shift to subsistence production, and the emergence of a parallel economy, including an active smuggling trade across the borders, seriously inhibited Government .efforts to mobilize domestic resources. Despite new tax measures, growth in domestic revenue fell behind the growth in current spending between 1977 and 1980. Expenditure on maintenance and upkeep, however, remained static at about one-third of the recommended requirements for maintaining the public sector's capital stock, including parastatal assets. 1.03 The poor performance of agriculture,-the dominant economic sector in terms of output, employment and exports, resulted in the 1978-1980 economic crisis. Although food production probably kept up with the increase in total population, a growing proportion of rural food surpluses was being smuggled out of the country. Consequently food supply was falling behind the rate of urban population growth. Export cvip pioduction declined. -The general direction of the Government's sector strategy emphasized the transformation of the institutional structure of rural development over measures designed to improve agricultural production directly. The proliferation of parastatals and other public sector involvement in the production of major crops, their marketing and the supply of physical and technological inputs, overstretched limited available management and skilled manpower resources. - 16 - 1.04 The prevailing price structures did not provide farmers with adequate incentives for increased productivity. The price terms of trade for marketed smallholder production of food and export crops had fallen 20 percent during the 1970s. They fell 30 percent for export crops alone. Declining returns on production for export were not related to trends in international trade prices bit mainly reflected rising marketing and administrative costs of the individual crop parastatals, and high export duties. These two elements reduced the farmers' share of export earnings from coffee, cotton, tobacco and cashews to below 50 percent. These price developments led to a diversion of supplies into unofficial channels (e.gs, coffee), neglect of tree and bush crops (e.g., coffee, sisal and cashews), and s-ibstitution of subsistence production for annual export crops (e.g., cotton and tobacco). 1.05 The supply of imports to cover agriculture's basic requirements was inadequate and mainly responsible for the erosion of the sector's capital base. The support processing industries and transport infrastruc- ture were equally import-starved. Overall, the sector's share of foreign exchange allocations did not reflect its importance to the economy and was disproportionately small compared with its contribution to foreign ex%.Aange earnings. 1.06 The dwindling export crop production was reflected in the worsen- ing balance of payments. By 1978-1980 the volume of exports was one-third below the mid-1960s levels, and appeared to be on a continuing downward trend. Of the six main export crops only tea shipments had risen consis- tently since the mid-1960s; sisal and cotton steadily declined; cashews and tobacco fell from their mid-1970s peaks and coffee supplies were static. Large foodgrain imports we:e required in 1980 to compensate for shortfalls in domestic supply. Export prices had fallen from the coffee boom peak of 1977 while import prices continued to rise, pushed by soaring oil prices in 1978/79, and the terms of trade worsened. The Government severely cut back on the volume of imports, including essential production inputs and spare parts, which compounded the existing problems-,of low capacity utilization. The current account deficit grew from US$183 mirlion in 1977 to US$637 million in 1980. Although the flow of aid increased, it was insufficient to cover this widening gap. The Government had to resort to short-term borrowing and accumulated major import payment arrears for the first time since Independence. After 1978, the net level of external reserves (including external payments arrears), was negative. It deteriorated from a negative US$209 million at end-1979 to negative US$501 million at the end of 1983. II. BANK RESPONSE 2.01 The Export Rehabt.itation Credit was a response to the Government's requests for non-project assistance during the period 1978-1980. (Two previous operations of this kind had been approved in December 1974 and March 1977.) - 17 - 2.02 A program credit was first discussed, and formally applied for, in April-May 1978. The Government withdrew the application after a joint review with IDA staff of the FY79 lending program. The possibility of a program loan was raised again at a meeting between the Bank's President and Tanzania's Principal Secretary for Finance in September 1978. The Bank's position then was that although a program loan was not contingent upon meeting the Fund's requirements in every detail, it was essential that a firm program dealing with both long-term and short-term problems be worked out. Such a program would have to address the need to promote exports and to balance available resources and prio ity needs. 2.03 Early in 1979, discussions between Tan2ania and the Fund had reached an impasse. The Bank did not want to undercut Tanzania or the Fund, but was willing to play an honest broker role. In June 1979, the Bank presented an aide-memoire to the Government in which it set out a general framework for possible future program lending. A program loan preparation mission went to Tanzania in August 1979. Its aide-memoire to the Gvvernment suggested a loan package which would require the Government to introduce: (i) a program of export incentives; (ii) a three-year development expenditure program; (iii) fiscal improvements; and (iv) increased interest rates to mobilize domestic savings (this proposal was already a sticking point in the negotiations with the Fund). The mission explained that for a program loan to be processed expeditiously, the Government would need to reach agreement with the IMF and would have to prepare a policy package as outlined in the aide-memoire. 2.04 During the October 1979 Annual Meeting discussions with the Tanzanian Delegation, the Bank reiterated the need for a program that would provide for a consistent development of exports, including a review of export crop prices and measures to improve the efficiency of crop _parastatals. The Bank would also look for tighter budget controls and a review of the investment program. A Bank mission to Tanzania in January 1980 stressed the importance of budgeting improvements, parastatal reform and export crop price increases. It also made*more precise agriculture- related policy suggestions dealing with the supply of inputs and output marketing, extension and research services and strengthening the Ministries of Planning and Agriculture. The Bank committed itself to respond as quickly as possible to Tanzania's needs in the framework of the structural adjustment program, and aimed at September 1980 as a target date for a program support credit, including the possibility of a consultative group meeting along the way. 2.05 A Structural Adjustment Credit (SAC) of US$50 million to Tanzania was included in the Eastern Africa regional lehding program for FY81 with a tentative timetable leading to Board presentation in September 1980. The Initiating Memorandum for the SAC proposed that it be used for increased capacity utilization especially in agriculture and transport, probably linked to a broadly defined list of inports. An understanding with the Government was to be sought on adjustment measures and reforms in: (i) budgeting and fiscal areas; (ii) import support for export-oriented - 18 - agriculture and related transport needs; (iii) improvements in the institutional framework for agriculture; (iv) parastatal management; and (v) regional planning. The memorandum proposed a series of structural adjustment credits to be granted as the program phases were successfully implemented. 2.06 In May 1980, the Government prepared an Economic Report with its own assessment of the economic situation and a broad identification of policy adjustments required. Three particular objectives set out in the report, which elicited positive Bank support, dealt with: (i) fiscal and monetary affairs (specifically, reduction of the budget deficit); (ii) the balance of payments (liberalization of imports and incentives to exports); and, (iii) improvements in administrative efficiency, particularly in the parastatals. The report was not, however, sufficiently explicit on how the Government proposed to achieve the proposed structural adjustment. 2.07? An appraisal mission for the SAC went to Tanzania in mid-June of 1980. Hovever, given the multiplicity of issues raised by the Bank, the manpower censtraints, the fact that key people were in the thick of budget pregarations and the uncertainties related to the impending elections, the Government wes not able to adequately respond to the Bank's queries at such short notice. Consequently, the appraisal was not carried out. 2.08 Tanzania finalized a Structural Adjustment Action Program and presented it to the Bank in mid-September 1980, with a formal request for a 5-year IDA Structural Adjustment Credit Program as follows: 1980/81: US$75 million; 1981/82: US$125 million; 1982/83: US$125 million; 1983/84: US$100 million; 1984/85: US$75 million. The request was discussed during the Annual Meeting when the Bank told the Tanzanian delegation that the latest policy package still fell short of what was needed to sustain economic growth in the long term. The Tanzanian delegation argued that a detailed response to the Bank's concerns would need substantial additional effort and time. This, in turn, would mean that the balance of payments support would be further delayed, thus changing the assumptions underlying the rehabilitation program. It was decided that the Bank would prepare a detailed questionnaire for the Government, the answers to which would be discussed and agreed during the (second) Bank appraisal mission scheduled for end-October. Meanwhile it was agreed to go for a more limited first operation, concentrated on export rehabilitation. Compared with the multipurpose SAC, the proposed ERC was to have a single theme of agricul- ture export revival by guaranteeing the sector's imported input require- ments on a self-sustaining basis. 2.09 The appraisal mission for the Export Rehabilitation Credit visited Tanzania in November 1980. It prepared an aide-memoire for the Government with detailed recommendations for policy actions. The recommendations concerned: (i) increases in agricultural producer prices; (ii) a review of on-going development projects (including World Bank - 19 - projects) to allocate or reallocate funds to priority operating and main- tenance needs; (ii) an understanding that the ERC would lead to a series of credits, if the Government were able to formulate and implement a comprehensive program of structural adjustment (to include tightening current and development budget management, and the establishment of a multi-year budgeting system); and (iii) adherence to the IMF program. 2.10 Strong ministerial reservations were expressed on some of the aide-memoire recommendations during and after the meeting which concluded the ERC appraisal visit. The Government's reluctance to introduce reforms had an adverse impact on the implementation of the ERC. By January 3, 1981, however, a Government position paper had been prepared which agreed to most of the Bank's recommendations (regional pricing to replace panter- ritorial pricing, auction systems for tobacco and sisal and development project reviews). In addition, the Government asked for technical assist- ance for restructuring the National Milling Corporation (NMC). The proposal for agricultural export retentions was agreed to, at a 10 percent rate. The only real difference concerned the Government!s express wish to separate the IMF program from the Bank's program loan. The suggested time-frame was also broadly agreed to, but the Government asked the Bank to . speed up the credit processin so that it could become effective in February 1981, instead of March, as scheduled. The Bank considered the Tanzanian response substantive and positive ,enough to justify going ahead with the credit. On January 31st, the Government abolished export taxes on coffee and sisal and the production tax on tobacco, as the first move on the Bank's program. 2.11 Within the Bank, however, some staff believed that the Govern- ment's response was insufficient. They proposed that an action program for agriculture concerning long-term policy matters (including research, extension and rganization of support services) should be added to the understandings already reached with the Government. They strongly urged that the ERC should be used as a vehicle for a constructive dialogue between the Bank and the Government with the view of limiting Government intervention to those areas in which it had the planning, management and implementing capacity to be effective. A draft memorandum of understanding incorporating the proposed agricultural policy program was circulated on February 12. Its main provisions were eventually incorporated in the ERC submission to the Bank's Loan Committee as part of a memorandum of understanding to be discussed with the Tanzanians during the ERC negotiations. Negotiations took place in Washington early in March, followed by further discussions in Tanzania between the Government and the Bank's Regional Vice President. 2.12 Negotiations with the Tanzanians centered on: (i) producer price changes which would provide sufficient incantives for increased production of export and food crops in the future; (ii) improvements in the foreign exchange allocation system including establishment and maintenance of a Special Account at the Bank of Tanzania which would be used to finance imports of production inputs for the agricultural sector; (iii) measures to - 20 - promote non-traditional exports, including the establishment of an export credit insurance scheme and provision of additional staffing for the Board of External Trade (BET) by December 31, 1981; and, (iv) preparation of an agricultural services action program by September 30, 1981 incorporating the Government's plans for research, extension, input delivery and crop collection in rural areas (a draft program to be available for IDA review by end-August 1981). 2.13 The Government was expected to have moved on producer price adjustmeats and the establishment of the Special Account by the time negotiations started. Other items on which the Government was expected to have acted were the preparation of timetables for updating parastatal accounts, the introduction of non-traditional export incentives (except (iii) above) and the appointment of representatives of the Ministry of Agriculture and of the BET to the Import Licensing Advisory Committee of the Bank of Tanzania. 2.14 The Bank advised the Government that it considered the export rehabilitation measures included in the ERC agreement as a first step in the necessary reform of the economy. To strengthen the impact of this operation and relate it to future more broad-based support, the Bank hoped for a parallel agreement on a Memorandum of Understanding outlining the Government's intentions in the following areas: (i) review of the public investment program, in the context of which the Bank Group's lending program would be discussed to determine the possibility of reallocating project resources to general economic adjustment, to be completed by end-June 1981; (ii) provisions for operations and maintenance requirements and preparation of a rolling two-year forward budget which would integrate development and recurrent expenditures also by end-June 1981; (iii) foreign exchange budgeting improvements; (iv) adjustments in agricultural policies; (v) strengthening policies dealing with wages, prices and incentives; (vi) improved domestic resource mobilization and cooperation with the Bank in organizing a consultative group meeting. 2.15 The Tanzanian delegation was reluctant to agree to the Govern- ment's share in funding the Special Account designed to finance an absolute amount of agricultural sector imports. It was to be part of an agricul- tural support fund which would pay for imports of at least US$100 million equivalent during its first year of operations. US$50 million would have - 21 - to come from the Government and the balance from IDA. The Bank envisaged an ultimate financing target for the fund of US$180 million, the then current estimate of direc: and indirect import requirements of the agricul- tural sector and supporting activities (including transport, marketing and processing, but excluding fuel). 2.16 The Tanzanian delegation was worried lest the Government's capa- city to meet its share would be constrained by emergency events (including falling export prices). After paying for fuel and food imports, the Government expected to have only US140 million at its disposal in 1981 for the remaining foreign exchange needs of the economy (including US$35 million in debt service obligations to the Bank Group). Consequently, a firm commitment to allocate US$50 million to agricultural sector imports might prove difficult to honor. The Tanzanians preferred alternative was to allocate to the fund 15 percent of export earnings from six major export crops in 1981/82 which would be a starting point for Government allocations in subsequent years of the adjustment program. The Bank remained firm and the Tanzanian Government eventually agreed to the US$50 million. The Bank assured the Government that it would review the matter if for reasons beyond its control, Tanzania failed to fulfill this commitment. 2.17 The Government's contribution of US$50 million, plus US$50 million from IDA still fell US$80 million short of what was needed, unless this 45 percent shortfall could be funded from new bilateral resources. With hindsight, the Government's preference for a 15% export earnings should have been accepted. The 15% would have yielded US$45 million on the basis of US$300 million worth of main crop export earnings that were virtually certain to be realized. The mechanism for capturing the 15% proportional contribution would have been put in place during the credit period, and might have had a better chance of survival as part of a longer term agricultural export rehabilitation program than a fixed sum contribution. III. THE EXPORT REHABILITATION CREDIT Objectives and Design 3.01 The ERC had two broad objectives: - To reverse the decline in the volume of traditional agricul- tural exports, and - To increase the rate of growth of manufactured and processed exports. 3.02 Central to the execution of the ERC was the creation of the Special Agricultural Account (SAA) at the Bank of Tanzania. During its first year of operation (April 1, 1981 - March 31, 1982), the Government's cumulative commitment of US$50 was to increase on a quarterly basis from - 22 - US$10 million equivalent at end June 1981, to US$20 million at end- September 1981, to US$30 million at end-December 1981 and US$50 million by March 31, 1982. The Government and IDA's contributions each of US$50 million, would create an agricultural import support fund of at least US$100 million. In addition, the Government was to make every effort to raise additional resources from bilateral and multilateral donors. In this regard the Netherlands provided the equivalent of US$18 million in addi- tional import finance for agro-industries in 1981/82 and a further US$8 million equivalent in 1984, administered as part of the SAA. Swedish, Norwegian and Danish funds were also channelled through the SAA. 3.03 To achieve maximum effectiveness in allocating the SAA funds, and to minimize leakages into non-export sectors, imports under the ERC were restricted, both in terms of eligible goods and of eligible importers. The credit financed 100 percent of the foreign exchange costs (cif) of eligible imports. It was administered by the Bank of Tanzania (BOT) in consultation with the Ministry of Finance and the BOT Import Licensing Advisory Committee, which was expanded to include representatives of the Board of External Trade and the Ministry of Agriculture (Kilimo). To maximize Kilimo's input to the import licensing process, an Agricultural Imports Allocation Advisory Committee (AIAAC) was set up within Kilimo to prepare recommendations for the BOT licensing committee. 3.04 The Tanzania shilling equivalent generated by the sale of foreign exchange to importers under the program was credited to a special account at the Bank of Tanzania to be used for rural development purposes. Importers had to follow the normal procedures for obtaining licenses and opening letters of credit. Broad international tendering among suppliers in Bank member countries, Taiwan and Switzerland was required for contracts over US$5 million equivalent. The minimum 'mport order under the ERC was fixed as US$10,000 equivalent. The list of eligible importers consisted of about 90 enterprises (mainly parastatal) entitled to import licenses financed under the ERC arrangements. 3.05 To speed up disbursements, imports financed after November 15, 1980, would be paid for retroactively under the ERC, as soon as it became operative. IDA disbursements were made against full documentation, including original invoices returned to the BOT under existing foreign exchange regulations. Except for the retroactive financing, which was arranged under IDA Disbursement Procedure I (Application for Reimbursement), most of the IDA financing was subject to Disbursement Procedure VI (Application for a Qualified Agreement to Reimburse) under which the National Bank of Commerce (NBC) (representing the borrower) requested IDA to issue to a commercial bank in the supplier's country its qualified agreement to reimburse the commercial bank for payments made under a letter of credit. - 23 - Implemeutation 3.06 The Government made an early start on ERC implementation with institutional arrangements for, and its first quarterly contributions to, the SAA. However, the agricultural import program fell behind schedule because of delays in BOT foreign exchange allocations. These also slowed down IDA disbursements. Bank supervision missions began to see import hold-ups as a major problem halfway through the original ERC period. Subsequently, the effectiveness of the AIAAC in setting import priorities and the nature and amount of the Government's contributions to the SAA became serious issues. Charting the implementation through the Bank's monitoring and supervision work, the following story emerged: - May 1981. A supervision mission reported substantial progress in establishing the reporting and monitoring arrangements for the SAA. The AIAAC had been set up and two expatriate staff assigned to it. A general sense of cooperation and commitment was noted among the three agencies most concerned with the ERC: The BOT, NBC and the Ministry of Agriculture (Kilimo). - June 1981. A mission reported a deepening foreign exchange crisis, and little hope of an early agreement with the IMF. The Government reluctantly agreed to Dutch import support of US$18 million equivalent to be channelled through the SAA. Disburse- ment of the US$5 million ERC retroactive financing was completed. - July 1981. A supervision mission found that the BOT had not met its July 1 deadline for announcing foreign exchange allocations for the import period July-December 1981. The Government's first quarterly contribution to the SAA (April 1 - June 30, 1981) estimated at US$18.2 million in terms of opened letters of credit, far exceeded the expected total of US$10 million. - September 1981. A supervision mission found import allocations six weeks behind schedule. No further IDA disbursements had been made. The NBC reported that Government commitments to the SAA had risen to US$37 million equivalent. The Tanzanians regretted that the supervision mission was not there to appraise a repeat of the program credit. Halfway into the credit period there was little progress on the other provisions of the ERC Agreement and the attached Memorandum. This e2lay was discussed with the Tanzanian delegation at the Aunual Meeting and a follow-up letter from the Bank to the Principal Secretary in the Ministry of Finan4.e stressed the relationship between further Bank lending and the veed for an effective recovery program. - October 1981. Due to overdue debt service payments, disburse- ments to Tanzania were temporarily suspended. This was unfortu- nate as requests to IDA to reimburse under the ERC had only just started to flow. - 24 - - November 1981. An agricultural sector mission found that BOT import allocation procedures were not working satisfactorily and that the role of the AIAAC was diminishing. A letter was sent to the Principal Secretary (Agriculture) suggesting ways of support- ing the Committee's work. Kilimo officials expressed concern about the uncertainty of continued IDA support for the SAA, and reservations about the future of the AIAAC in the absence of further program aid. - March 1982. The ERC closing date was postponed to March 1, 1983. The undisbursed IDA balance, as of March 31, stood at US$34.7 million, most of which had been committed. - June 1982. The last official ERC supervision mission dealt mainly with the agricultural aspects of the program credit. It found very little progress made with respect to achieving the program's agricultural development objectives. In addition, the mission found that Government allocations to the SAA over the ERC period had amounted to only US$29.7 million. - July 1982. The Government published a Structural Adjustment Program for the economy. - March 1983. The credit appeared to be fully committed and IDA stopped approving new commitments under the ERC although there was an undisbursed balance of US$8.8 million equivalent at the time. The credit closing date was informally postponed to September 30, 1983. However, since ov-r 60 agreements to rei!- burse had been issued to various correspondent banks, the Loan Department needed additionr -ime to infcim these banks that the expiry date of the letters of credit (LCs) had lapsed, the credit would be closed, and in the absence of a a reply, the IDA agree- ments to reimburse would be cancelled. It was necessary to give the correspondent banks 30 days within which to respond to IDA, so the deadline was postponed to December 31, 1983. - December 1983. The majority of banks contacted by IDA confirmed that they had cancelled their respective agreements to reimburse because suppliers did not present LCs. The Loan Department subsequently cancelled all the outstanding agreements, including those with banks which failed to respond. It was then found that there was an undisbursed balance of US$4.1 million in the ERC account. - April 1984. It was agreed with the Government that the undis- bursed ERC balance would be used for direct reimbursement for foreign exchange expendtures made by the Government under the project before March 31, 1983 (mainly those affected by the IDA decision rot to approve new commitments in March 1983). - 25 - - September 1984. Disbursements were completed. - October 1984. The Netherlands Government allocated SDR 7.7 million of fast-disbursing (by December 31, 1984) import support to be channelled through the SAA. - April 1985. The Bank advised the Government that the credit of SDR 40.9 million had been fully disbursed and that the credit account had been closed. 3.07 The initial allocations and actual expenditures under the ERC are shown below. Variations were covered by Clause 3 of Schedule I attached to the ERC agreement. Initial Actual Allocation Allocation % Variation Category ----(SDR million)- 1) Agricultural equipment, machinery, replacement parts & spares 14.72 11.88 -19.3 2) Unprocessedrubber& agricultural packing mat6rials, including tin plates & jute bags 12.26 12.84 +4.7 3) Agricultural chemicals, including coffee fungicide 1.66 9.42 +467.5 4) Vehicle replacement parts 12.26 9.76 -44.9 TOTAL 40.90 40.90 - IV. THE ERC IN RETROSPECT Economic and Policy Effects 4.01 While the credit did not achieve most of its stated objectives, it is likely that without the rehabilitation program, Tanzania's export capacity would have been even lower than it was during the four-year period of ERC disbursements. Aggregate export earnings declined from USS563 million in 1981 to US$413 million in 1982 and US$359 million in 1983. This was due mainly to the drop in exportable volumes of major crops. Persistent bottlenecks in domestic production and marketing caused foreign exchange earnings from each one of the major export crops to be lower in 1983 than in 1981. Only tea and cashewnut exports were showing signs of improvement by 1984. In these circumstances it is difficult to assess conclusively the impact of the producer price increases or other elements of the rehabilitation program on production trends. - 26 - 4.02 The ERC minimized the damage caused to agriculture by the economy's gradual import starvation, but on its own could not reverse the trend. Total imports declined from US$1160 million to US$1090 million between 1981 and 1982, then fell more sharply, to US$819 million in 1983. Shortages of imported inputs continued to disrupt agricultural and manufac- turing production, accelerated infrastructural decay, particularly in transport, and caused widespread underutilization of capacity. 4.03 The Bank and the Government were fully aware of the risks to the ERC at the time that it was agreed. The President's Report drew particular attention to the possibility that additional concessional financing would not be made availaule from other sources, resulting in the need tc cut imports even further and hence reduce the benefitLi of the proposed credit. A slow but continuing movement Uowards various agricultural policy reforms may have been one of rehabilitation effects of the ERC. Thus some key elements in the program for agricultural development included in the 1984/85 Budget and Plan were pioposals made at the time of the ERC prepara- tion and implementation. These included the Government's commitment to: (i) a priority shift in favor of agriculture by increasing the sector's share of both current and capital spending; (ii) concrete measures to ensure that recurrent cost imports for agriculture would be reasonably met; (iii) better producer incentives, including increases in producer prices, and a 40 percent devaluation; and, (iv) crop marketing and distribution improvements. 4.04 Some institutional and procedural improvements introduced under the ERC proved worthwhile, if short lived in the cases of the SAA and the AIAAC (para. 4.09). A Presidential Commission was appointed in May 1982 to undertake a major review of agricultural policy and a substantive policy paper was released in November 1982, embodying many of the detailed recom- mendations of the Commission. This policy review was wider in scope than the action program proposed during negotiations of the Export Rehabilita- tion Program but lacked details regarding the practical implications of some proposed changes of policy. Subsequently, within the context of Tanzania's own Structural Adjustment Program (SAP), and with valuable inputs from the Bank's 1983 Agricultural Sector Report, policies were refined and the general development emphasis was increasingly put on agriculture as the priority sector. Forty other credit conditional/, such as the measures to promote non-traditional exports, were introduced more or less on schedule. However, their impact was limited, partly by production constraints, due to the shortage of foreign exchange to finance inputs for manufacturing industries, partly by market constraints as exporters were reluctant to sell more to neighboring countries such as Uganda, Mozambique and Zambia which already owed considerable import arrears to Tanzania. Undertakings such as the setting up of a debt monitoring unit at the Bank of Tanzania and the appointment of key personnel to the Board of External Trade were not met by the agreed target dates, and were partly superseded by institutional changes introduced in the context of the SAP. 1/ See Annex for Summary of Performance. - 27 - 4.05 Initially Bank missions were well satisfied with the institu- tional arrangements made for the implementation of the ERC "core program", the establishment of the SAA at the BOT and the appointment of an AIAAC at the Ministry of Agriculture. The SAA eventually served as a conduit for US$118 million equivalent of import support financed by the IDA credit, the Government's own resources and bilateral credits from the Netherlands, Sweden, Denmark, and Norway. Controversy still surrounds the question of how much the Government contributed to the SAA. In the first 6 months of the ERC period the National Bank of 'ommerce reported Government contribu- tions to the SAA of USS37 millian, which would have been US$17 million more than what was due in the period under the ERC provisions. By the time figures became available on actual disbursements in the ERC context, the picture looked different. Depending on which figures are preferred the Government contributed US$62.4 million (allocations from own foreign exchange resources, plus suppliers' credits which were never formally excluded from the Government's contribution), US$29.7 million (own foreign exchange allocations), US$21.7 million (based on letter of credit (LC) drawings plus direct payments), or US$13.2 million (based on LC drawings plus direct payments, adjusted to exclude a proportion of imports by industries which also served non-agricultural sectors). There was no early , warning for this problem because other delays (notably in the BOT allocation information system) obscured the shortfall in own foreign exchange disbursements during the relevant import periods July-December 1981 and January-June 1982. Because of inadequate monitoring of SAA disbursements while the account was fully operational, and lack of data on foreign exchange allocations to agriculture before (or after) the ERC, there is no accurate measure for the Government's con4ribution. 4.06 Disbursements of the IDA contribution to the SAA were much slower than expected. Compared with the appraisal estimate that the entire credit would have been disbursed by end-March 1982, 70 percent remained undis- bursed at that time, although most of this balance was thought to have been committed. Several factors contributed to the disbursement delays: (i) for both of the six-month allocation periods BOT completed the exercise of informing foreign exchange applicants of their allocations 3 or 4 months late, and the Bank was informed nearer the end of each period; (ii) the delay was further compounded by the fact that after obtaining import licenses from BOT, recipients' LC applications had to be approved again by the BOT, which should not have been necessary for IDA-supported LCs because the required foreign exchange backing was known to be avilable; (iii) there were administrative delays in preparing and handling documentation for obtaining IDA qualified guarantees to reimburse commercial banks, both at the Tanzanian end and in Washington. 4.07 The performance of the AIAAC was good insofar as it was allowed to enter the foreign exchange allocation process. AIAAC prepared detailed allocation plans (based among others on the net foreign exchange generation/saving capacity of each recipient enterprise2/) for the periods July-December 1981, January-June 1982, and an indicative program for July- 2/ The basic data for these assessments was prepared by the Marketing Development Board in March 1981, updated in August 1981. - 28 - December 1982, after the ERC was formally completed. The last six-month program was made redundant by the then deepening foreign exchange crisis. The recommended programs were welcomed by the BOT advisory committee. In its actual allocations the BOT followed the main line of AIAAC recommenda- tions, subject to its own revisions, particularly in the allocation of bilateral assistance and suppliers' credits. The AIAAC was not informed of these changes nor of the final allocations. As a result the allocation process was not as thoroughly monitored as had been intended. The AIAAC also carried out a number of field visits and submitted reports on the needs and problems of individual agricultural sector importers. Lack of resources meant that these contacts were not followed up after the ERC, thus wasting what might have been an opportunity for finding out the impact of ERC imports on capacity utilization and export generation in individual enterprises. 4.08 The fact that AIAAC was located within the Ministry of Agricul- ture was its main strength. In theory, the Committee had access to all Kilimo's people and data resources and through the Principal Secretary a voice in the BOT import licensing committee. However, Kilimo located it in the Technical Services Division which had no interest in its activi- ties. It was left to fend as best it could without a'budget or any full-time Tanzanian staff. Three expatriates (two provided with Dutch aid, and one temporarily borrowed from the IDA-funded MDB) carried out most of tae work. If it had been located in the Planning Division of Kilimo or in the BOT its work might have been more effective. However, such a move would have prevented the cooperative relationship that existed at least during the crucial first six months of the ERC between AIAAC and the BOT. Role of the Bank 4.09 The Bank played a key tole in developing the export rehabilita- tion program. It was involved in most sectors of the economy (there were 44 Bank Group projects under execution at the time that the credit was signed), particularly in the agricultural sector. Bank staff were increas:'.ngly conLerned about the way in which the development potential of these projects and of the economy as a whole was being eroded by the shortage of foreign and domestic resources. The operation enabled the Bank to continue its efforts to persuade the Government of the need for major policy adjustments. Although the Government agreed to the conditions of the credit, and the Memorandum of Understanding, the agricultural provi- sions did not appear to have received Kilimo's wholehearte support, nor did other Ministries show much commitment to the ERC's success. The Government was very reluctant to allocate an absolute amount of US$50 million to the SAA, instead preferring the export earnings formula. Bilateral donors' efforts to channel funds through the SAA allocation process also met with strong initial Government resistance, despite the ERC agreement provision to the contrary. When the Bank did not send a formal appraisal mission for a second ERC, or a structural adjustment credit, despite a commitment to this effect in the Memorandum of Understanding, the momentum sustaining the successful parts of the ERC gradually disappeared. The Government saw no useful purpose in maintaining either the SAA or the AIAAC once the ERC funds wer. fully disbursed. - 29 - 4.10 Within the Bank, responsibility for the ERC was shared between the agricultural projects (EAPSA) and the programs (EA1DA) departments. Among the staff involved, there were differing views and expectations of the project. Programs staff favored an operation focussing on limited objectives pertinent to export rehabilitation. On the other hand, agricultural projects staff had become increasingly frustrated by their apparent inability to persuade the Government to make adjustmenPs to current agricultural sector policies. The existing portfolio of Bank Group appraised and on-going agricultural projects did not have the leverage which a series of fast-disbursing program credits promised to have. Hence, the ERC appeared to provide a golden opportunity for a breakthrough in persuading the Government to act. Much hard work went into the preparation of agricultural sector reform proposals in the hope that they would be used as policy conditions for the ERC (as a first step in a multi-year IDA support program). These were included as covenants in the credit agreement: (i) formulation of an action program for agricultural support (section 3.08); (ii) setting up, funding and operating the Special Agricultural Account (section 3.09); and, (iii) to commission a study of producer price setting for food and export crops, and to act on its recommendations for annual producer price adjustments (section 3.10). Others were included in the Memorandum of Understanding (see paras. 23-24). However, none of these covenants or understandings were satis- factorily met. 4.11 In light of the performance under the ERC, the Tanzanians were aware of the Bank's uncertainty on whether or not to nontinue program lending, with or without an agreement with the IMF. The letters sent to the Government after the final-ERC supervision in July-August 1982 focussed on the shortcomings in meeting the ERC foreign exchange allocation and agricultural policy conditions, which were ostensibly the main reasons for not proceeding with the appraisal of further program lending. One of the Bank's letters stated "the Government did not allocate the US$50 million ... instead, according to the records compiled by BOT and NBC as little as US$13.2 million were disbursed ... Larger amounts of foreign exchange appear to have been expended on industry and other purposes in the economy." Given the uncertainty surrounding the allocation data and the fact that there was no Government commitment to spend proportionately more cn agriculture than on the rest of the economy, this letter might have appeared unjustified. Meanwhile, the Government had published the Structural Adjustment Program largely as a response to earlier advice that further Bank lending would be related to the implementation of an effective recovery program. The Bank review of the SAP took place in November 1982. >4.12 When Bank staff realised that the ERC would not fully achieve its purpose, they had the option of trying for an improved ERC. However, the size of financial commitment required (and the concurrent - as yet indefinite - policy changes that would have been expected from the Government in 1982/83) made this an unrealistic option. In the event, the Bank opted to abstain from further program lending. The ERC experience left both sides unsatisfied. Because of inadequate performance on the. policy front under the ERC, the Bank decided against a follow-up program credit. - 30 - Design Weaknesses 4.13 The credit suffered from a number of weaknesses which contributed in varying degrees to its Limited success. The credit was not large enough to fund the program although bilateral donors committed substantial resources in addition to IDA. If the MDB study was believed to provide a valid basis for the ERC, then the $100 million plus, which the ERC was to raise could not have covered the $180 million agricultural sector import requirements estimated by the study. Some of the conditions attached to the credit were insufficiently discussed and understood before they were included in the credit agreement. First, there was the trade-ofO between a rapid disbursement credit and medium-term economic policy reforms. These dual objectives became even more incompatible when the credit was limited to one year instead of being part of a multi-year program, as it was intended to be. Second, there were conditions not fulfilled by the Government, conditions which the Government had been most unhappy about, namely the agricultural action plan and'the US$50 million flat rate contribution to the SAA. The addition of an agricultural services action program at the Decision Meecng represented a significant and demanding alteration to understandings reached with the Tanzanians during the appraisal mission. Such an addition should have been carefully prepared with greater direct involvement of Bank staff. The reluctance of the Tanzanian authorities to commit US$50 mllion to the SAA should have been discussed more comprehensively, even at'the cost of either redesigning or delaying the credit. The Government was prepared to adhere to the export earnings retention system originally proposed by the Bank, and offered to increase the retention size f-rom 10 to 15 per cent. This would have yielded an identifiable contribution requirement from the Government of US$45 million in 1981/82. Third, the monitoring of the conditions was unsatisfactory particularly for the SAA contributions. A specific measurement should h-sve been attached to the Government's allocation requirement in terms of BOT allocations, LCs opened, inclusion or not of suppliers' credits, etc. As it was, the rules were belatedly formulated. 4.14 The Bank was overly optimistic about the timing of the ERC: in May 1980 the credit was still scheduled for September 1980 which undoubtedly encouraged the Tanzanians to count on the IDA US$50 million being available by end-December 1980, to help meet ceilings set by the IMF. The credit was not available and the IMF conditions were not met. Seen from a Tanzanian perspective the Bank dragged its feet and was partly responsible for the lapse of the stand-by program. IDA disbursements of the credit would probably have been smoother if they had involved the use of a revolving fund. This was proposed within the original working party but was subsequently dropped. Monitoring arrangements for the project should have included specifications for the timing of BOT foreign exchange allocation procedures to avoid the 3 or 4 months delay incurred in each of the two six-month import periods covered by the ERC. - 31 - 4.15 The adaptation rather than radical reform of institutional arrangements for agricultural import licensing was a sensible approach, implemented emoothly with relatively little friction among the agencies concerned. Unfortunately the AIAAC lacked status and funds. More attention should also have been paid to setting up a systpm of import allocation data flows between the AIAAC, the BOT and the ERC benefi- ciaries. 4.16 The main conclusions of the ERC experience were first, that if the MDB estimate of US$180 million was a valid basis for the approved program, then secure funding for the EAC should have approached that amount. Second, given the cooperative relationship required for the successful implementation of a program credit, it was unwise to introduce new conditions (the agricultural services action plan) without comprehen- sive discussion when negotiations were in the final stages. Third, the mechanism for monitoring day-to-day operations of the credit should have been worked out in detail with the agencies involved. This might have avoided operational problems such as those arising from lack of commitment on the part of the key agency of the Ministry of Agriculture. Fourth, Bank staff were aware of the risk that the Government's US$50 million funding of the SAA might be a substitute for rather than in addition to, "normal" foreign exchange allocations to agricultural sector imports. For that purpose precise rules should have been laid down as to what was, or was .not, to be included in the Government's ERC contribution. Finally, a program credit intended to rehabilitate a specific economic sector, in this case export-oriented agriculture, would have benefited from better coordination among the concerned Bank staff at all stages, including supervision missions. V. SUMMARY AND CONCLUSIONS 5.01 The Tanzania Export Rehabilitation Credit for 1981/82 was a response to the Government's requests for non-project assistance during the period 1978-1980. At that time the performance of agriculture, the dominant economic sector in terms of output, employment and exports, had deteriorated and export crop production was falling. The general direction of the Government's sector strategy, meanwhile, emphasized the transforma- tion of the institutional structure of rural development over measures designed to improve agricultural production directly. 5.02 Initially, discussions between the Government and the Bank centered on IDA support for a 5-year Structural Adjustment Program. However, the policy changes which the Government was peepared to undertake were considered to be insufficient for a Structural Adjustment Credit (SAC) and a limited first operation, concentrated on agricultural export rehabi- litation, was agreed upon. 5.03 An aide-memoire presented to the Government at end-1980 outlined the policy actions recommended by the Bank for implementation in conjuncr- ion with the ERC. The Government had broadly agreed to these recommenda- tions by January 1981 and moved to increase crop producer prices as a first step towards ERC implementation. - 32 - 5.04 The import requirements to be met by the ERC were identified in an MDB study which formed the basis for the credit. The study estimated the minimum import needs for the rehabilitation of export crop production at US$180 million. However, firm financing was US$100 million, US$50 million each from IDA and the Government. The balance was unspecified amounts expected to be contributed by bilateral donors. The agricultural policy program which supported the ERC was introduced belatedly and without adequate discussion. The Bank also proposed a change in the mechanism for Government contributions to the SAA. Instead of retaining a proportion of export earnings, the Bank proposed that the Government make a flat contri- bution of US$50 million. The Government accepted this commitment with great reluctance. 5.05 By the time the ERC reached the negotiating stage the single agricultural export rehabilitation purpose, which made it a preferred alterdative to a SAC, had been made complex by the add-on agricultural services action plan. 5.06 There were problems and considerable delays in implementing the ERC program. Foreign exchange allocations for imports fell far behind schedule, which also affected IDA disbursements. There was strong initial Government resistance to b1lateral donors' efforts to channel funds through the SAA allocation process, despite the ERC agreement provision to the contrary. By the end of the credit period the nature and amount of the Government's contributions to the SAA and the effectiveness of the AIAAC had become serious issues. The ERC closing date was postponed from March 1982 to June 1984. The credit account was finally closed in April 1985. The SAA eventually served as a conduit for USS;8 million equivalent of 4 import support financed by IDA, the Government's own resources and bilateral credits from the Netherlands, Sweden, Denmark, and Norway. 5.07 The ERC could not achieve most of its main objectives, because most of the necessary conditions were not met in a timely fashion, and the inadequate tunding for imported inputs limited the export recovery. It is likely, however, that without the ERC, Tanzania's export capacity would have been even lower than it was during the four-year period of ERC disbursements. The initial thrust for policy reforms introduced in 1984 can be traced back to discussions and studies related to, or initiated because of, the ERC program. Apart from underfunding, the credit suffered from several operational and design weaknesses. Among the former, the lack of close collaboration within the Bank, and lack of Tanzanian support for two key conditions introduced belatedly in the process undermined the cooperative relationship required for the implementation of a program credit. On the technical side, the credit would have benefited from a real commitment on the part of Kilimo to the success of AIAAC, and detailed specifications for monitoring import procedures and the Government's contributions to the SAA. -0- (eg-3-tan-pcr) - 33 - ANNEX kRC COVENANTS AND MEMORANDUM OF UNDERSTANDING: GOVERNMENT PERFORMANCE B! END-AUGUST 1982 Specific Measures Government Action (1) Abolish export taxes on coffee Action taken in April and and sisal, as well as the July 1981. production tax on tobacco, and increase producer prices for these crops substantially. (2) Complete a study of the Consultant's study was producer price setting system completed, but Government by August 31, 1981 with a view delayed decisions on fully to improving the incentive range of recommendations. ; ffects of such prices in uture. (3) Establish a Special Action taken on April 1, Agricultural Account at the 1981. Bank of Tanzania to meet a larger share of the direct and indirect import requirements of the agricultural sector. (4) During the first year of The Government provided no operation the Special more than US$21 million to Agricultural Account (April 1, the Account,.but.total 1981 - March 31, 1981), make resources channelled thiough available US$50 million from the Account reached US$118 the Government's own resources, million, including assistance in addition to the IDA credit, from Sweden, the Netherlands to finance the importation of and Denmark. US$100 million of inputs for agriculture. (5) Assign priority to the The Government established a agricultural sector in the stringent fuel rationing allocation of fuel oil, heavy system in mid-1981, but it diesel oil and lubricants for was not clear how priority production, processing and was assigned to the transport. agricultural sector. (6) Appoint a representative of the The Principal Secretary for Ministry of Agriculture to the Agriculture was given a seat Import Licensing Advisory on the Committee. Committee chaired by the Bank of Tanzania. - 34 - ANNEX Page 2 (7) Have import requests froa-the The Agricultural Imports agricultural sector screened by Allocation Advisory Committee a specialist committee at the (AIAAC) was set up in April Ministry of Agriculture. 1981 to receive and comment on Import applicativns from the sector. (8) Establish a Task Force to A Parliamentary Task Force improve the financial was established and Treasury accounting practices of introduced new controls. The parastatals. As a first step, timetable was finalized in prepare a detailed timetable mid-1981, and there was some for*the completion and audit of progress in updating the accounts of eight key accounts. agricultural parastatals. (9) Initiate international Action on this matter was recruitment of key personnel delayed by six months, owing for thirty positions at the to the Ministry's wish to Ministry of Agriculture and consult the agencies major agricultural parastatals concerned; a Third Technical by December 31, 1982. Assistance Credit was approved by IDA in February 1982 to finance these individuals. (10) Establish a more effective Progress in this area was transport planning and limited. management systems for the major agricultural parastatals. (11) Introduce a cash bonus system The cash bonus system was for non-traditional exporters introduced in November 1981. and simplify other export There were some revisions of promotion procedures. export procedures along the lines proposed. (12) Introduce an export credit The decision to introduce guarantee program, if suitable, such a program was taken in by December 31, 1981. late 1981, but implementation was delayed. (13) Fill two key vacancies at the BET started recruitment Board of External Trade (BET), procedures. i.e., a Trade Policy Manager and Export,Promotion Manager, by December 31, 1981. - 35 - ANNEX Page (14) Establish a strong debt and Treasury claimed to have aid monitoring unit within the strengthened its own Bank of Tanzania (BOT). monitoring procedures, but no major improvement appeared at BOT. (15) Review ongoing and future The Government made development projects with a substantial cuts to its view to consolidating overall development budget in public investment program. 1981/82 and 1982/83; however, the project selection process was still unsystematic. (16) Re-introduce two year rolling Two year budgets were budgets for the development reintroduced in 1981/82; the and recurrent accounts so as practical value of second to facilitate medium term year budget figures was financial planning. doubtful. (17) Prepare an action plan by The Government requested, and September 30, 1981 outlining was granted, a postponement studies and steps to be taken of the deadline to December to improve major aspects of 31, 1981. only an initial agricultural services (e.g. draft of the action program research, extension, input was prepared, but it proved delivery, crop collection), unsatisfactory. The Government appointed a Commission in iay 1982 (under the Chairmanship of the Economic Adviser to the President) to prepare recommendations by September 1982 for revising agricultural policy. Source: Bank staff draft ERC assessment (September 1982). IBR TANZ UNDA POPULATIO ( - AND TRANS20 D A: WO RWANDA K EN Y A --- Oil Pipeline AZ- National Porks (N.P.) ond -LI'ul Game.Reserves (G-R.)8 Rivers Internotional Airports BUR UsND AME. IHA Scheduled Service Airports .m hi vA - Main Roads moz.aOao Secordory Roods RilroadsN -- Region Boundories Internotional Boundaries Population Density by Region (Inhabitants Per Square Mile) I, OM A rOver 1500 (Dar Es Salaam) 250 - 1500 (Pemba and Zanzibar Islands) r 100-250 so - 100 Lees tisan so TANZANIA IN 1967 R D ES SALAAM - oII ,os * s~oddled Ai, Ser G-R PLv - .dS-Id. e-d, _ N,ional Porks ond Game Rsere - UGANDA ANANDA G.R IND. KENYA . RE \ noto*" + -- -sa"u G.R T \T iyu Nc be- - N D I 0Nz A m B I A 7Ld * s..TWARA ZAMBtIA As v.-- MAmAWI .O"l|O." rov .'"* S-- MOZMBZAEMB...QUEe M O A MB l U Erea 3,6Aff