Uzbekistan - Rehabilitation Loan Project

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 16566 IMPLEMENTATION COMPLETION REPORT REPUBLIC OF tJZBEKISiTAN REHABILITATION LOAN (Loain Number 386 1-lL;) May 12, 1997 Country Operations I Country Department III ELur-ope and Central Asia This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization.  CURRENCY EQUIVALENTS Currency Unit = SotLm (adopted July 1994) US$1 = 58 Soumn (as of end-March 1997) Soum per US$ Period Period Average End of Period 1995 30.2 35.5 1996 41.1 54.7 WEIGHTS AND MEASl RES Metric Systemn FISCAL YEAR OF BORROWER January I - December 31 ABBREVIATIONS AND ACRONYMS CAS Countrv Assistancc Stra)tcgy C'EM Countrx Economilc MNciiorandum ('IF Cost. Insuranicc and Fri-ight (for a traded good) FSU Former Sov ict Unlion FY Fiscal Year GDP Gross Domcstic Produkct IBRD International Bank for Rcconstruction and Developmncit IMF International Monctarx Fulid MOF Ministrv of F`in.nc. Pltl Projcct Implementat ion Unit PIA Projcct Implementation Assessment SOE Statement of Expc idittircs US$ UJiitcd Statcs Dollar Vice President: Johannes F. Linn, ECA Director: Yukon I-luanig, EC3 Division Chief Kadir T. Yurukoglu, EC3CI Responsible Staff: Erika Jorgensen, EC3C I  FOR OFFICIAL USE ONLY IMPLEMENTATION COMPLETION REPORT REPUBLIC OF UZBEKISTAN REHABILITATION LOAN (Loan Number 3861-UZ) Contents Preface .................................................. i Evaluation Summary ...................................................i Part I. Project Implementation Assessment .................... .............................. 1 A. Statement and Evaluation of Project Objectives ......................1............................I B. Achievement of Project Objectives ................................................... 3 C. Implementation Record and Major Factors Affecting the Project ........................................7 D. Project Sustainability ...................................................8 E. Bank Performance ........................................... ....... 8 F. Borrower Performance ..................................................9 G. Assessment of Outcome ..................................................9 H. Future Operations .................................................. 10 I. Key Lessons Learned .................................................. 10 Part II. Statistical Tables .................................................. 13 Appendices: A. Policy Matrix B. Form 590 (ICR Mission) C. Borrower Contribution to the ICR D. Map (IBRD 28006R) 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.   IMPLEMENTATION COMPLETION REPORT REPUBLIC OF UZBEKISTAN REHABILITATION LOAN (Loan Number 3861-UZ) Preface This is the Implementation Completion Report (ICR) for the Rehabilitation Loan in the Republic of Uzbekistan, for which loan 3861-UZ in the amount of US$160 million equivalent was approved on March 28, 1995 and made effective on April 12, 1995. The loan was closed on June 30, 1996, the original closing date. Since this was a rehabilitation loan rather than a structural adjustment loah, the disbursements were not tranched. Final disbursement took place on December 27, 1995. The ICR was prepared in the Country Operations Division I of Department III of the Europe and Central Asia Region. It was prepared by Erika Jorgensen and reviewed by Kadir T. Yurukoglu, Division Chief of EC3C1, and Marc Blanc, Project Advisor, Country Department III. The borrower provided comments that are included as an appendix to the ICR. Preparation of this ICR was begun during the Bank's final supervision and completion missions in October 1995 and October 1996. It is based on material in the project file. The borrower contributed to preparation of the ICR by preparing its own evaluation of the program's preparation and execution and by commenting on the draft ICR.   IMPLEMENTATION COMPLETION REPORT REPUBLIC OF UZBEKISTAN REHABILITATION LOAN (Loan Number 3861-UZ) Evaluation Summary Introduction I. The Bank's involvement in Uzbekistan prior to this loan had been focused on technical assistance and policy dialogue supported by economic and sector work. The first loan to Uzbekistan -- Institution Building Technical Assistance Loan (No. 3650-UZ for US$21 million, approved October 1993) -- was to support reforms in the short term to build institutions and capacity over the medium term. (See Project Implementation Assessment, PIA, para. 5.) Project Objectives 2. The Rehabilitation Loan was designed to take account of experience with other Bank rehabilitation loans to FSU countries, and so it concentrated on key issues with significant up front actions. The program objectives were to: (a) support the Government's structural reform program; (b) minimize negative impacts of transition on economic activity; and (c) improve private sector access to foreign exchange. The policy conditions to achieve those objectives covered actions related to reform of the incentive structurc, enterprise reform, and social safety net reform. (PIA, para. 4.) 3. The core policy actions on privatization of medium- and large-scale enterprises through investment funds and phasing out of state orders were formalized as dated covenants in the Loan Agreement: The deadline for one of the covenants on the privatization program was extended from July 31, 1995 to October 31, 1995 in an amendment to the Loan Agreement in August 1995. (PIA, para. 4.) 4. The Rehabilitation Loan was a one-tranche operation, but the policy conditions included timebound performance benchmarks to serve as a medium-term program that would be supported by further lending operations. An innovative feature was the phasing of disbursement of the tranche to provide the Bank with the possibility of suspending disbursements if the medium-term program went seriously off-track. (PIA, para. 6.) At the same time, the loan design created the largest risk to the program's full implementation -- that momentum would falter and delay following operations. (PIA, para. 8.)  -11- Implementation Experience and Results 5. Achievement of the broad project objectives of the Rehabilitation Program has been partial and unsatisfactory overall; but progress on the specific policy conditionality of the loan has been substantial, although ridden with delays and reversals. Implementation of a comprehensive structural reform program has been much slower than agreed with the Bank. In some policy areas, backsliding was followed eventually by corrective actions, but these were not always sufficient to fulfill the Government's obligations. The two most important policy areas where fulfillment of conditions was greatly delayed were the issuance of regulations governing privatization investment funds and the liberalization of cotton and grain pricing and procurement under state orders for 1996. However, the disbursement of the loan did help ease foreign exchange shortages, and imports other than foodstuffs and energy products more than doubled from 1994 to 1995. Sustained economic decline has been avoided, and GDP expanded slightly in 1996. But private sector access to the foreign exchange market, after improving through 1995 and the first half of 1996 as restrictions were removed and trading volumes grew, has been severely curtailed by new government restrictions in late 1996. (PIA, para. 10.) 6. Macroeconomic outcomes improved through 1995 and the first half of 1996 but then went off-track as the Government reacted to pressures on the current account of the balance of payments induced by external and domestic shocks. (PIA, para. I1].) The authorities loosened both monetary and fiscal policy and tightened trade and exchange restrictions. (PIA, para. 12.) The reform of incentives has been partially achieved. The improved access to foreign exchange by the private sector achieved during 1994, 1995, and the first half of 1996 has now been rescinded. (PIA, para. 14.) The phase out of state orders for cotton and grain has suffered many delays and has required continuous negotiation. (PIA, para. 16.) Reductions in input subsidies and the gradual increase of farmgate prices did constitute a significant shift in incentives for agricultural production. (PIA, para. 17.) Despite some progress on trade liberalization, exports of 4 key commodities, including cotton, remain restricted through licensing requirements. (PIA, para. 20.) Enterprise reform through privatization of medium- and large-scale enterprises using investment funds has progressed steadily since regulations were issued in July 1996. The first 300 companies were selected in 1996; and the second auction of 11 companies to the funds took place in January 1997. (PIA, para. 24.) Reform of the social safety net through reduction of consumer subsidies and targeting of household transfers as well as pension reform have helped contain budget costs. (PIA, para. 26.) 7. Overall, the loan had an unsatisfactory outcome in achieving its objectives, and sustainability of the program is uncertain. Implementation of the conditionality of the loan has been satisfactory, although the waning of borrower commitment to an ongoing reform program has resulted in widespread delays and backtracking so that the broader development objectives of the program have not been met. The reform program has moved forward but not at an accelerating pace. National income has suffered only modest declines. A framework for donor support has been undermined by the Government's abandonment of the IMF program. The volume of trading in the foreign exchange market remains high, but private sector access has been severely curtailed since November 1996. The failure to follow this loan with further adjustment operations greatly undermined its expected impact as the launching point for ongoing and  -iii - accelerating reform. Nevertheless, performance by both the Bank and the Project Implementation Unit of the Borrower are judged to have been satisfactory. (PTA, paras. 30-36.) Summary of Findings, Future Operations, and Key Lessons Learned 8. The most important finding of the project implementation experience is that a one-tranche rehabilitation loan, designed as a compromise between a rehabilitation loan and an adjustment operation, creates its own risks. If the momentum of reform slackens so that the prospect of future operations recedes, the Bank is left with little abilit0 to follow through on outstanding policy conditions. (PIA, para. 39.) Nevertheless, the policv conditions of the program remain appropriate for the problems they were intended to address In that sense, the loan was well- designed. However, the final objectives of the operation were not reached because this loan should been the first in a series of adjustment operations. with overlapping and reinforcing policy conditions. (PIA, para. 38.) 9. Preparation has continued on the Bank's next adjustment operations. The Enterprise Reform Loan was negotiated, but Board presentation was postponed from November 1996 because of macroeconomic instability. Second, a Financial Sector Adjustment Loan is under preparation, building on the analysis and recommendations of a recent economic report. Third, various operations in agriculture, including fast disbursing policy based loans, are being considered. (PIA, para. 37.) 10. One key lesson learned during preparation and implementation of the Rehabilitation Loan is that continuous support and involvement from the most senior government officials in Uzbekistan is required for every significant action. (PIA, para. 40.) Another key lesson is that the slowdown of Uzbekistan's economic reform program has increased budget costs for the Bank. Large up front investments were made by the Bank in economic analysis, policy dialogue, and institutional development during the preparation of the loan If the program had moved ahead as planned, those costs would have allowed for lower cost, follow-up operations. Instead, this operation cost the Bank about two-thirds more than rehabilitation loans to other Central Asia borrowers, a premium that exemplifies the generally substantially greater cost of doing business in Uzbekistan. (PIA, para. 41.)   IMPLEMENTATION COMPLETION REPORT REPUBLIC OF UZBEKISTAN REHABILITATION LOAN (Loan Number 3861-UZ) Part I. Project Implementation Assessment A. Statement and Evaluation of Project Objectives 1. The overall cumulative decline in Uzbekistan's real GDP during 1992 to 1994 was significantly less than in other Former Soviet Union (FSU) countries -- around 14 percent compared with an FSU average of 40 percent. Enhanced reliance on Uzbekistan's natural resources (oil and gas, and agriculture) and trade diversification (especially of cotton exports) helped cushion the effects of a large drop in industrial output. However, macroeconomic imbalances worsened through 1993 as a result of the Government's policy of subsidization of consumption and production. Inflation soared, the fiscal deficit widened; and the need for balance of payments support, because of the economic disruption resulting from the breakdown of the payments system and of trade within the FS U. became clear. The Government began to consider economic reform. 2. During 1994. the Government set out reform objectives, including privatization, establishment of property rights, trade liberalization, currency convertibility, and opening of the capital account of the balance of payments. A national currency, the soum, was introduced in Julv 1994 (to replace the interim currency introduced in November 1993). The Government eliminated all direct consumer subsidies, significantly reduced indirect subsidies (except public transport and public utility rates, which are still administratively fixed), and began freeing energy prices. It also considerably tightened fiscal policy by eliminating budgetary transfers to enterprises and by further curtailing direct consumer subsidies, thereby reducing the consolidated budget deficit for 1994 to 6.1 percent of GDP. The Bank identified the Rehabilitation Loan in support of the Government's reform program in April 1994; and in January 1995, the IMF began to support the Government's stabilization program through its Systemic Transformation Facility (for SDR100 million). 3. The Bank's Country Assistance Strategy (CAS) for Uzbekistan (March 3, 1995, Report No. 14019-UZ), presented simultaneously with the Rehabilitation Loan to the Bank's Board, laid out a strategy of graduated response in which the level and composition of Bank lending would be adjusted according to the pace at which the Government executed its economic reform program. In the high-case scenario, with strong progress in the reform program, it was expected that the Rehabilitation Loan would be the first in a series of policy based operations in support of the Government's reform program, focusing especially on enterprise restructuring and financial sector reform. (CAS. para. 60.) 4. The Rehabilitation Loan was designed to take account of experience with other Bank rehabilitation loans to FSU countries, and so it concentrated on key issues with significant up front actions (President's Report, Rehabilitation Loan to Uzbekistan, March 3, 1995, Report No. P-6525-UZ, para. 50). The program objectives were to: (a) support the Government in the design and implementation of its structural reform program,  Republic of IJzbekistan: Implementation Completion Report 2 (b) ease the shortage of foreign exchange needed to finance critical imports and minimize the negative impact on the level of economic activity stemming from delay of reform and transition disruption; and (c) deepen the existing foreign exchange market to improN e private sector access. The policy conditions to achieve those objectives, as specified in the policy matrix in the President's Report (Annex 2) and as shown in Part HI, Table 5, and in Appendix A, covered actions related to reform of the incentive structure (phase out of state orders, trade and foreign exchange liberalization, elimination of subsidies, and strengthening of antimonopoly policy), enterprise reform (privatization with broad participation, and improved governance of state-owvned enterprises) and social safety net reform (pension reform). The core policy actions, formalized as dated covenants in the Loan Agreement (April 5, 1995), were: (i) designing and implementing a privatization program for medium- and large-scale enterprises using investment funds. and (ii) phase out of state orders for the 1995 and 1996 crops and increasing procurement prices for the 1995 crop. The deadline for one of the covenants on the privatization program was extended from July 3 1, 1995 to October 3 1, 1995 in an amendment to the Loan Agreement in August 1995. 5. The Bank's involvement in Uzbekistan prior to this loan had been focused on technical assistance and policy dialogue supported by economic and sector work The first loan to Uzbekistan -- Institution Building Technical Assistance Loan (No. 3650-UZ for US$21 million, approved October 1993) -- was to support the Government in formulating and implementing reforms in the short-term, and in building institutions and capacity to implement and sustain reforms over the medium-term. The first Country Economic Memorandum (1993) and two economic reports. 'Subsidies and Transfers" (June 1994) and ' Adjusting Social Protection" (December 1994) had provided the basis for advice to the Government. 6. The Rehabilitation Loan was a one-tranche operation, but the policy conditions included time bound performance benchmarks that extended beyond the expected disbursement period. These benchmarks were expected to serve as a continuous evaluation instrument to be used in macroeconomic and sectoral dialogue and as a medium-term program that wnould be supported by the Bank's further lending operations. (President's Report, para. 51.) An innovatixe feature of this loan was the phasing of disbursement of the tranche by setting a monthly maximum for withdrawal applications. The extended disbursement period was meant to provide the Bank with the possibility of suspending disbursements if the medium-term program went seriously off-track. 7. Successful implementation of the Rehabilitation Program was important to lay the foundation for the Bank's strategy in Uzbekistan for FY96 to FY98, as described in the accompanying CAS document. Specifically, the Rehabilitation Loan was designed to help the Government prepare and begin to implement its structural reform program, which could then be supported by follow-on adjustment operations. The objectives of the program and the policy conditions as described in the President's Report were specific, the core actions were emphasized through the use of dated loan covenants, and clarification was provided in the Minutes of Negotiations. 8. The most risky aspects of the program design were those areas of action that were meant to be medium-term monitorable indicators, and which therefore would likely require the support of other operations, as would the two areas that involved the most complexity -- phase out of state orders and a new privatization program. These were also the two areas of action chosen for loan covenants. But while the program considered as a whole was demanding, it was designed to ensure forward momentum along the path of transition and reform. At the same time, this design also created the largest risk to the program's  Republic of Uzbekistan: Implementation Completion Report 3 full implementation -- that momentum would falter and delay following operations. From minutes of meetings discussing the design of the loan, it appears that this was a recognized risk when the loan was prepared, although it was not stated explicitly in the President's Report. 9. At the time of loan approval, the policy conditions were ambitious but realistic, with the one- tranche loan intended to provide support mainly for the policv reforms taken before effectiveness, and medium-term actions included in the policy matrix to generate continuity to the next adjustment operation. At the time of approval of the loan, the IMF was supporting the Government's stabilization program through its Systemic Transformation Facility, and discussions had begun on a possible stand-by arrangement. The Government's Letter of Development Policv in the President's Report included the policy actions supported by the Rehabilitation Loan within a broader and longer program of reform. Moreover, the Government had taken many significant policy steps before Board presentation of the loan, which the Bank considered to be supported by the loan, and almost half of the actions supported by dated covenants in the Loan Agreement had been completed bv the time of negotiations in February 1995: (a) Reform of the incentive structure. By end-1994. most industrial state orders and state orders for all agricultural commodities except cotton and grain were eliminated. The Government had reduced state orders for the 1994 cotton crop and for the 1995 cotton and grain crops and increased procurement prices. Progress had been made in liberalizing the external trade regime, by removing the tax on foreign exchange proceeds and reducing export licensing to cover only 11 commodities. A weeklv auction mechanism for foreign exchange was established in 1994, and the exchange rate was unified. Direct food subsidies were removed, and full pass through on industrial tariffs for gas, electricity, and heat was implemented. (b) Enterprise reform. Privatization of small-scale enterprises, retail shops. trade and service establishments, and housing was mostly completed by end-1994, although mainly through direct transfer or sale to staff. Governance of state-owned enterprises was improved: (i) ownership, property, and management rights were clarified, with the transformation of several hundred medium and large state enterprises into open joint stock companies; and (ii) greater financial discipline was imposed through improved reporting of arrears and the identification of the 30 largest loss makers in 1994-95. (c) Social safety net reform. The Government began shifting towards a more targeted system of social protection in 1994, especially for pensions. A draft resolution set out a revised system for adjusting pension benefits. B. Achievement of Project Objectives 10. Achievement of the broad project objectives of the Rehabilitation Program (para. 4 above) has been partial and unsatisfactory overall; but progress on the specific policy conditionality of the loan has been substantial, although ridden with delays and reversals. Implementation of a comprehensive structural reform program has been much slower than agreed with the Bank. In some policy areas, backsliding was followed eventually by corrective actions, but these were not always sufficient to fulfill the Government's obligations under the Rehabilitation Program. The two most important policy areas where fulfillment of conditions was greatly delayed were the issuance of regulations governing privatization investment funds and the liberalization of cotton and grain pricing and procurement under state orders for 1996.' However, Of the 16 conditions set out in the policy matrix of the President's Report as medium-term actions (but not tranche release conditions, since this was a one-tranche operation), 11 are almost completed, although some policy objectives have been  Republic ol Uzbekistan: Implementation Completion Report 4 the disbursement of the loan did help ease foreign exchange shortages, and imports other than foodstuffs and energy products more than doubled from 1994 to 1995. Sustained economic decline has been avoided, and GDP expanded slightly in 1996. But private sector access to the foreign exchange market, after improving through 1995 and the first half of 1996 as restrictions were removed and trading volumes grew, has been severely curtailed by new government restrictions in late 1996. Table I presents the available quantitative indicators of the program's impact so far. 11. Macroeconomic Outcomes. Macroeconomic performance improved through 1995 and the first half of 1996 but then went off-track. The IMF approved a fifteen-month stand-by arrangement of SDR124.7 million in December 1995, a program designed to support stabilization policies to keep inflation to 34 percent or less in 1996 and confine the decline in GDP to around I percent. The Central Bank was to maintain a positive real refinance rate and increase the proportion of interbank credit intermediated through the credit auction mechanism; and the Government was to keep the fiscal deficit to about 4 percent of GDP. Structural benchmarks were designed to reinforce the Rehabilitation Loan and bridge to the next planned Bank adjustment loan on enterprise reformn. The first review of the stand-by was completed in June 1996 and, until the third quarter of 1996, Uzbekistan largely met the stabilization objectives of the program. 12. GDP declined 0.9 percent in 1995, a better performrance than had been expected. Consumer price increases moderated from their extreme levels in 1994 to average 305 percent for 1995, slowed by fiscal and monetary tightening in the second half of the year. The fiscal deficit in 1995 dropped to 4.1 percent. However, the moderately successful stabilization of 1995 began to dissipate halfway through 1996 as the Government reacted to pressures on the current account of the balance of payments induced by a drop in world cotton prices, a lower than expected wheat harvest, and declining forecasts of the cotton crop. The authorities loosened both monetary' and fiscal poliev and tightened trade and exchange restrictions. GDP still expanded by 1.6 percent for the year, but annual inflation for 1996 registered at 54 percent, above the IMF target, because of large injections of credit into the economy in the fourth quarter. The small surplus in the budget that had emerged in the first quarter of 1996 was overwhelmed by unbudgeted expenditures in the last quarter for onlending to agriculture and clearance of pension arrears. For the full year of 1996, the budget deficit was about 7 percent of GDP, well above the IMF program target. (See Table 2.) 13. The auction determined official exchange rate depreciated by 40 percent in 1995, pushing the real exchange rate to about 85 percent above its September 1994 level. During 1996, the nominal exchange rate remained relatively stable through the first three quarters of 1996 and then depreciated by 38 percent in the fourth quarter, leaving the year end real exchange rate at approximately the level at which it started the year. The external current account, close to balanced in 1995, loomed to a deficit of 7.9 percent of GDP in 1996, responding to the deterioration in terms of trade but also the temporary liberalization of the foreign currency market that allowed significantly higher imports in 1995 and the first three quarters of 1996. Gross international reserves exceeded six months of imports in 1995 and continued to cover almost five months of imports at the close of 1996, a 12 percent increase from the end of the third quarter of 1996 which was achieved through the rationing of foreign exchange and use of preferential exchange rates. As a result of these policy shifts, the IMF stand-by is expected to lapse in March 1997 without the last two disbursements having become available, and the Bank's next planned adjustment operation, the Enterprise Reform Loan, cannot be presented to the Board. 14. Reform of Incentives. The overriding goal of the Rehabilitation Program was to strengthen the development of Uzbekistan's incipient foreign exchange market and improve access to foreign exchange pursued through slightly different actions than originally envisioned. Another 4 conditions (on procurement; 1 997 state orders, export licensing; and the trucking sector) are on their way to being met. On only one medium-term policy condition -- the reform of agricultural taxation -- is there little sign of action.  Republic of Uzbekistan: Implementation Completion Report 5 by the private sector. The substantial progress towards this goal achieved during 1994, 1995, and the first half of 1996 has now been rescinded. After a new national currency, the soum, was introduced in July 1994, the Central Bank restricted access to foreign exchange through licensing and new currency restrictions, leading to a wide disparity between official and unofficial exchange rates; but by October 1994, a single unified exchange rate system was in place. Enterprise access to foreign exchange through a system of twice weekly interbank foreign exchange auctions was introduced in the beginning of 1995. These auctions, which were initially heavily controlled through a "patent" or license system, were liberalized. Restrictions on access by individuals to buy foreign exchange were abolished in July 1995; and the surrender requirement for export earnings and the license system for access to the foreign exchange auctions by enterprises were liberalized in October 1995. In response, the weekly volume of foreign exchange traded increased significantly, from about $1 million per week when the auction was introduced in April 1994, to $10 million per week by January 1995, $30 million by June 1995, and almost $70 million each week by November 1996. (See Table 1.) 15. The recent restrictions on the foreign exchange market have severely hampered access but have not dampened demand for foreign exchange at the official rate, and so sales of dollars at the auctions have remained robust in early 1997. However, the interbank auction has been limited to only once a week (rather than thrice weekly as of October 1996 and daily as of December 1996 to which the Government committed under the IMF stand-by). In addition, the spread to the rate charged by commercial banks for cash at exchange bureaus widened in October but then narrowed again. The premium to the curb market rate, however, rose sharply from October through November, and has remained at over 100 percent, indicating the severe disequilibrium induced by the recent restrictions on foreign exchange. (See Table 1.) 16. The phase out of state orders for cotton and grain, a policy area where the Bank acquiesced to the Government's desire for gradualism, nevertheless has suffered many delays and has required continuous negotiation. The required reductions in state orders and increased prices for 1995 cotton and grain were announced by Presidential Decree in January and February 1995, before Board presentation in March 1995. State orders covered 60 percent of the 1995 cotton crop at 50 percent of world prices, and 50 percent of the 1995 grain crop at 50 percent of world prices. (See Table 1.) In September 1995, the Bank discovered that the grain pnrce had been set at only 40 percent of the benchmark world price agreed at negotiations (i.e., the US Gulf port price for wheat plus transportation costs from southern Kazakstan), but the Government argued that Uzbekistan's pattern of grain imports had shifted (and so the lower CIF price for imports from Kazakstan was the appropriate import parity price). Thus, it only becarne clear ex post that the Government did not meet its commitment of a 1995 gram price of 50 percent of the world price (although it was in compliance at the time of Board presentation). 17. In 1995, wheat procurement prices were kept low by the Governnent to try to contain inflationary pressures, despite the conflicting aim of achieving self sufficiency. Rice and maize were removed from state orders in 1995, and their liberalized wholesale prices increased to match approximately border price equivalents. Subsidies of agricultural inputs were reduced so that both fertilizers and agricultural chemicals were priced at world equivalents (and the removal of these large subsidies occurred with admirable speed). Farmers also pay world prices for equipment and machinery, imported mainly from Russia. Although the agricultural sector does continue to receive some implicit subsidies in the form of tax exemptions and low water charges, these developments in agricultural policy together constituted a significant shift in incentives for production, away from input subsidies and towards higher output prices at the farmgate. 18. An agreement on quotas for 1996 cotton and grain was finally reached in March 1996, after extended discussions. (The Government had issued orders and prices inconsistent with its agreements with  Republic of Uzbekistan: Implementation Completion Report 6 the Bank and the IMF in January.) The procurement levels were set at 40 and 25 percent of the cotton and grain crops, respectively. In April 1996, the Bank agreed to the Government's proposal of $130 per ton as a minimum purchase price for grain for the 1996 crop, which by alternative calculations could be considered to be 75 percent of world price. The cotton price was set in local currency terms in August 1996; and at August exchange rates, it converted to over 60 percent of the world price. However, the subsequent 50 percent depreciation of the soum to the end of 1996, when cotton farmers could expect to be paid, reduced the cotton price to about 40 percent of world price. State orders for 1997 have not yet been announced, and there are signals that the Government will not honor its commitment to eliminate state procurement of grains and reduce further state procurement of cotton. 19. For goods other than cotton and wheat, work is ongoing to replace the former system of state orders with a system of state procurement by competitive tender. The Bank has been providing advice, and an Institutional Development Fund grant of $160,000 to assist in the development of public procurement legislation and training of personnel in procurement procedures was approved in October 1996. 20. Despite some progress on trade liberalization, exports of 4 key commodities, including cotton, remain restricted through licensing requirements. The Government took several steps in 1995 and 1996 to liberalize trade. It reduced the number of items subject to export licensing from 11 to 4 (cotton, oil, ferrous metals, nonferrous metals), and pledged to eliminate licensing for all goods except cotton, by January 1, 1997 (instead of initiating auctions of export licenses for oil and ferrous and nonferrous metals, as committed under the program). Export licensing for cotton had been placed under simplified procedures by mid-I1995 but not sufficiently to render them automatic as specified in the policy matrix of the President's Report. New import and export tariffs set in October 1995 created high effective protection for some commodities, but the Government began rationalizing the tariff system with a consolidation and reduction of rates on April 1, 1996. The maximum export tax rate and the number of products subject to export duties were also reduced. Rough calculations of average tariffs show reasonable declines over the period. (See Table 1.) 21. The other area of action related to trade was the development of an agricultural taxation system to replace lost revenues as state orders and export taxes were reduced, particularly for cotton. No actions have yet been taken by the Government. The Bank has provided policy advice, with tvs o missions and two policy notes to the Government (in late 1995 and early 1996). 22. The elimination of subsidies is back on-track after delays. The price differential between household and industrial energy prices was reduced in April 1996 and further in December 1996 to meet the levels agreed under the Rehabilitation Program, a 12-month delay.2 However, the Government continued to cover 80 percent of the costs of heating and hot water in 1995, accounting for most of the 4 percent of GDP spent on subsidies from the budget. and prices for heating, hot water, and housing are not adjusted on the basis on full cost recovery. 23. The strengthening of antimonopoly policy was expected to be supported by the Enterprise Reform Loan. A new antimonopoly law, which redefines monopolies as only those companies with a market share greater than 65 percent, was passed in December 1996, 18 months late; and an independent antimonopoly 2 In April 1996, the differential for electricity prices was reduced by 33 percent and that for gas by 35 percent. In December 1996, the differential for electricity was reduced by an additional 85 percent and for gas by 41 percent.  Republic of Uzbekistan: Implementation Completion Report 7 authority has been set up. Demonopolization has progressed in the wholesale trade sector;3 but a competitive commercial trucking sector, crucial for private sector development, is not yet a reality. The Government has promised to sell its shares in trucking companies through the new mass privatization program, auction 20 percent of the cargo truck fleet, and remove restrictions on privatized gas stations. 24. Enterprise Reform. Privatization is an area of slow but steady success. About 95 percent of the 10,400 small-scale enterprises in the Republican level privatization prograrn had been privatized, and 96 percent of housing units have been privatized. Almost all medium and large state enterprises have been transformed into open joint stock companies; and of the 1,600 medium- and large-scale enterprises slated for privatization, about 1,000 have had some shares sold to managers and employees. The other 600 are to be privatized through sales of shares to the public through the Privatization Investment Funds, a mass privatization program intended for support by the Enterprise Reform Loan. Regulations for this mass privatization scheme were to have been issued by July 31, 1995. a deadline extended by the Bank to October 31, 1995. The regulations were issued in July 1996, 12 months late; but progress since then has been steady. The first 300 companies were selected in 1996. and the second auction of 11 companies to the funds took place in January 1997. 25. The governance of SOEs has improved, especiallv the strengthening of financial discipline. After some initial difficulties the Credit Bureau, set up in the Central Bank to collect data on enterprise arrears, has made progress. Since November 1995, 30 of the SOEs with the highest arrears have been monitored closely, and the Government has announced its intention to either privatize or liquidate them. The further accumulation of enterprise arrears has been discouraged through two Presidential Decrees in May 1995 and January 1996. 26. Reform of the Social Safety Net. The curtailing and targeting of consumer subsidies and the establishment of unemplovment benefits and family allowances greatly improved targeting and further reduced the fiscal burden of the social safety net, as has pension reform. Although the Parliament has not vet enacted the revised scheme of pension adjustment into law, costs have been contained, which is an important achievement, given how large Uzbekistan's pension expenditures have been relative to its demographic structure and level of GDP. Only about 6 percent of the population is over the age of 60, yet pension benefits had been near 10 percent of GDP, partly because another 6 percent of the population below age 60 were also collecting pensions and because the average pension was three times the minimum wage. The structure of pensions was flattened by a Presidential Decree in August 1995 (which increased the minimum pension by more than the average). As a result, pension payments fell dramatically, from over 11 percent of GDP in 1993 to 6 percent in 1995. (See Table I.) C. Implementation Record and Major Factors Affecting the Project 27. Implementation of the Rehabilitation Program has encountered moderate problems. Generally acceptable compliance with the specific policy conditions has been marred by delays and faltering commitment. The overall timetable for this loan, as laid out in December 1994 in the Initiating Memorandum, was followed successfully. Disbursement was distributed over a number of months (as described in para. 6); with US$32 million allowed in retroactiRe financing for policy actions completed prior to loan effectiveness, then US$16 million allowed for each of April, May, and June 1995; and then 3The wholesale trade of consumption goods was demonopolized in 1995: the rural wholesale trade is now wholly private, and the Government has promised to sell its remaining shares in urban wholesale companies. But wholesale trade in raw materials and intermediate inputs remains dominated by a single association.  Republic of Uzbekistan: linplementation Completion Report 8 US$20 million allowed per month for the remaining funds. The loan disbursed on schedule, with the final money released in December 1995. 28. A series of economic shocks undermined the Government's commitment to its reform program. World cotton prices fell by 15 percent in 1996, pushing down the country's terms of trade by 6 percent. The wheat harvest in May 1996 came in below target. Sales of reserves through the official currency auctions were growing as imports increased. Together, these events fed the Government's anxiety over the level of foreign reserves, as well as anxiety over letting markets work. The Government had always expressed concern over exposing the key commodities of cotton and wheat to world commodity price fluctuations; and these developments confirmed the fears of instability induced by factors outside government control. 29. From the Bank's perspective, however, while these factors external to government control had a partial negative impact on the program, it was the factors under the control of the Government, i.e., its commitment to carry through on the reform program that limited the significant achievements of the loan to the set of policy reforms completed before Board presentation (and the smaller set completed afterwards), rather than allowing it to become a launching pad for an ongoing reform program. In particular, the Government's desire to deviate from the detailed interpretation of conditions laid out in the Minutes of Negotiations regarding minimum prices to be set for state orders created repeated difficulties during supervision. D. Project Sustainability 30. The program is likely to maintain a number of its achievements, especially those that are difficult to reverse, but overall sustainability is uncertain. Those areas where achievements are least likely to be undenrined are price liberalization and the elimination of consumer and enterprise subsidies; corporatization and privatization of enterprises; improved financial discipline of state enterprises; elimination of state orders on all commodities except cotton and grain, and reduction of input subsidies in agriculture. These are areas xvhere the Government has relinquished control and where the reestablishment of government intervention would likely have a high budgetary cost. 31. The Bank remains confident that despite this period of hesitation, the Government will return eventually to a faster paced, market-oriented reform path. But in the absence of movement towards further reforms in the short-run, which the Bank had intended to support with follow on projects, the benefits of transition will not begin to be felt. Also, extended delays of privatization and liberalization of the economy will damage nascent markets. Expectations may be permanently affected, rendering the next bout of stabilization more difficult. especially if financial disintermediation and currency substitution by the populace rise significantly. Higher real interest rates and greater depreciation would then become necessary (and so a larger contraction of real incomes) before macroeconomic equilibrium can be reestablished. E. Bank Performance 32. The preparation of the Rehabilitation Program involved lengthy technical collaboration. Bank staff and consultants spent 150 weeks to bring the loan to appraisal, and missions to Uzbekistan included specialists in economics, privatization, enterprise reform, the financial sector, the energy sector, trade, and the legal framework . As mentioned in para. 5 above, the formulation of the program was preceded by economic and sector work in the first, comprehensive CEM and economic reports on subsidies and on social protection. The Bank was closely coordinated with the IMF's activities in the country.  Republic ot Uzbekistan: Implementation Completion Report 9 33. The Bank was highly responsive to borrower requests for additional technical assistance and analysis of issues related to the program. For example, in August 1995, the Bank provided key technical assistance to the Government to address concerns about revenue losses and uncertain impact on producer incentives of phasing out state orders. A mission initiated the design of an alternative scheme of agricultural taxation, leaving a report with the Government in October 1995. Another mission visited in early 1996 to provide advice on trade policy. Reports were sent to the Government on both trade policy and on reform of cotton taxation in early 1996. The Bank is also providing significant assistance with the development of standard state procurement legislation and procedures through an Institutional Development Grant approved in October 1996. 34. Appraisal of the loan was delayed from mid-1994 until January 1995 because of the considerable slippage by the Government in nearly every area of its reform program (which wvas also to be supported by the IMF's Structural Transformation Facility) and the need for more agreement on the details of the structural reforms to be supported by the loan.4 Later, supervision was made difficult bv the loan's single tranche structure, and delays of policy implementation were frequent. Missions supervised the program in May. August, and September 1995. and in March 1996; and 59 staff weeks xvere used to trv to sustain momentum in the program. The Bank continues to monitor the measures agreed under this program, but in the absence of following operations and especially with the unvwinding of both stabilization and structural reforms in late 1996. policy dialogue lacks the solid foundation it needs to be effective. F. Borrower Performance 35. The overall implementation of the program wvas the responsibility of the Project Implementation Unit established as part of the Department for the Coordtnation of External Economic Activitv under the Institution Building/Technical Assistance Loan. This unit was well-staffed and operated effectively. The delays in meeting policy conditionality arose from wavering commitment on the part of the most senior government officials, not from a lack of coordination or institutional weaknesses. G. Assessment of Outcome 36. Overall, the loan had an unsatisfactory outcome in achieving its objectives. Implementation of the conditionality of the loan has been satisfactory. although the waning of borrower commitment to an ongoing reform program has resulted in widespread delays and backtracking so that the broader development objectives of the program have not been met. The development impact of the program. as described in the discussion of the loan's benefits in the President's Report (para. 72), was expected to be the acceleration of implementation of a comprehensive structural reform program, avoidance of sustained economic decline; provision of a framework for donor support; deepening of the foreign exchange market; and increased access to the foreign exchange market for the private sector, Briefly, the reform program has moved forvard but not at an accelerating pace. National income has suffered only modest declines. A framework for donor support has been undermined by the Gov ernument's abandonment of the IMF program. The volume of trading in the foreign exchange market remains high, but private sector access has been severely curtailed since November 1996. The failure to follow this loan with further adjustment operations greatly undermined its expected impact as the launching point for ongoing and accelerating reform. 4 The Bank determined that progress towards agreement with the Government was required on a privatization blueprinit, methods for imposing financial discipline on state enterprises, increasing the procurement price on state orders for cotton and accelerating further reductions of state orders, and reduction of pension payments. By December 1994, the Bank assented to the Govermment's view that the privatization program should not include vouchers, but agreement still had not been reached on automatic granting of cotton export licenses and a methodology for changing pension indexation to reduce the budget burden. These unsettled issues were taken up during appraisal  Republic of Uzbekistan: Implementation Completion Report 10 H. Future Operations 37. Despite the hiatus in reform and the deterioration of macroeconomic conditions since late 1996, the Bank has continued some preparation of the next adjustment operations. The Enterprise Reform Loan was negotiated, but Board presentation was postponed from November 1996 because of macroeconomic instability. However, privatization is progressing; and the Bank is providing extensive technical assistance in cooperation with other donors. Second, a Financial Sector Adjustment Loan is under preparation,5 building on the analysis and recommendations of the recent economic report, "Uzbekistan: Creating Financial Markets: A Review of the Financial Sector" (February 1997, Report No. 15430-UZ). Third, various operations in agriculture, including fast disbursing policy based loans, are being considered, and a policy framework for the sector is under discussion with the Government. I. Key Lessons Learned 38. The policy conditions of the program remain appropriate for the problems they were intended to address. In that sense, the loan was well-designed. However, the final objectives of the operation were not reached because this loan should been the first in a series of adjustment operations, with overlapping and reinforcing policy conditions. 39. The most important finding of the project implementation experience is that a one-tranche rehabilitation loan, designed as a compromise between a rehabilitation loan and an adjustment operation, creates its own risks. If the momentum of reform slackens so that the prospect of future operations recedes, the Bank is left with little ability to follow through on outstanding policy conditions. 40. During preparation and implementation of the Rehabilitation Loan, it also became apparent that continuous support and involvement from the most senior government officials in Uzbekistan is required for every significant action. 41. An important side effect of the slowdown of Uzbekistan's economic reform program has been increased budget costs for the Bank in completing this loan, because large up front investments were made by the Bank in economic analysis, policy dialogue, and institutional development. If the program had moved ahead as planned, those costs would have allowed for lower cost, follow-up operations. This operation cost the Bank over $680,000 from identification to completion report, which was about two- thirds more costly than other Bank rehabilitation loans to Central Asia borrowers. This premium exemplifies the generally substantially greater cost of doing business in Uzbekistan compared to the Bank's other Central Asian clients. 5 The Govemrnent signed a Project Preparation Facility Agreement in January 1997.  Republic of lJzbekistan: Implementation Completion Report 11 Table 1: Indicators of Progress with Respect to Loan Objectives 1992 1993 1994 1995 1996 I. Progress on Structural Reform Program Cotton under state orders (% output) 100 80 67 60 40 Grain under state orders (% output) 100 75 60 50 25 State price for cotton (% of world price) - - 14 50 41 State price for grain (% of world price) - - 28 40 75 Budgetary subsidies (% of GDP) 21.0 25.8 2.7 3.9 4.8 Pension payments (% of GDP) 9.7 11.4 - 6.0 - Privatized enterprises (% of total) - small enterprises 0 64 80 96 - medium and large enterprises 0 16 33 39 - Average unweighted tariffs (%) - imports 0 0 0 18 11 - exports 0 0 0 32 26 II. Ease Foreign Exchange Shortages and Minimize Impact on Economic Activity Non-food non-energy imports (US$ mln.) 628 1,973 1,191 2,851 3,271 Total imports (US$ min.) 1,660 3,255 2,727 3,597 4,712 Current account balance (US$ mln.) -238 -429 118 -50 -1,075 Reserves (US$ mln.) 530 1,021 1,330 1,868 1,901 GDP (% change) -11.1 -2 3 -4.2 -0.9 1.6 - industry - -4 2 -6.6 -5.9 3.4 - agriculture - 15 -3.4 2.3 -18.6 III. Deepen Existing Foreign Exchange Market to Improve Private Sector Access Frequency of auctions (per week) 0 1 2 2 Weekly sales of foreign exchange (US$ mln.) 0 5 34 61 Exchange bureau premium (% over auction) - - 20 11 Curb market premium (% over auction) 6 39 49 92 Note: Exchange bureau and curb market premia are based on end-of-period rates. Sources: Government of Uzbekistan, IMF, and staff estimates.  Republic of lJzbekistan: Implementation Completion Report 12 Table 2: Selected Economic Indicators 1992 1993 1994 1995 1996 1. Economic Activit GDP (% change) -11.1 -2.3 -4.2 -0.9 1.6 II. Prices and Money Consumer prices (% change) 645 534 1568 305 54 Broad money (% change) .. .. 680 158 99 Exchange rate (sumlUS$, e.o.p.) .. 1.3 25.0 35.5 55.0 111. Government Finances (% of GDP) Total revenues 31.3 36.0 29.2 34.6 32.2 Total expenditures 43.4 3X.8 33.3 37.6 35.6 Budget balance -12.1 -17.6 -4.1 -3.5 -8.8 IV. Balance of Payments Exports (US$ mln.) 1424 2877 2940 3805 3781 Imports (US$ mln.) 1660 3255 2727 3597 4711 Current account balance (US$ mln.) -238 -429 118 -50 -1075 Current account balance (% of GDP) -11.8 -7.8 2.1 -0.5 -7.9 External debt stock (US$ mlii.) 33 820 793 1499 1964 Debt service ratio (% of exports) .. 2.5 10.5 16.8 17.9 Reserves (months of imports) 3.8 3.8 5.7 6.2 4.8 Note: Consumer price inflation is period average. Sources: Governtnent of Uzbekistan, IMF. and staff estimates.  Part II. Statistical Tables Table 1: Summary of Assessment Table 2: Related Bank Loans/Credits Table 3: Project Timetable Table 4: Loan/Credit Disbursements: Cumulative Estimated and Actual Table 5: Key Indicators for Project Implementation Table 6: * Key Indicators for Project Operation Table 7: * Studies Included in Project Table 8A: Project Costs Table 8B: Project Financing Table 9: * Economic Costs and Benefits Table 1(0: Status of Legal Covenants Table 11: * Compliance with Operational Manual Statements Table 12: Bank Resources: Staff Inputs Table 13: Bank Resources: Missions * Standard Tables 6, 7 and 9 are not included because they are not applicable to this Rehabilitation Loan. Table 11 is not included because there have been no instances of noncompliance with Operational Manual Statements.  Republic of Uzbekistan: Inplementation Completion Report 14 Table 1: Summary of Assessments A. Achievement of Objectives Substantial Partial Negligible Not Applicable Macro Policies El / E3 El Sector Policies El / El El Financial Objectives / El l E Institutional Development E E El Physical Objectives ElE E / Poverty Reduction E E O Gender Issues EE El / Other Social Objectives / El El Environmental Objectives E El I / Public Sector Management E a 0 E Private Sector Development l El I/ El Other (specify) El E El I B. Project Sustainability Likely Unlikely Uncertain O ElO / Highly C. Bank Performance Satisfactorv Satisfactory Deficient Identification V E E Preparation Assistance I O E Appraisal El E Supervision E / E Highly D. Borrower Performance Satisfactory Satisfactory Deficient Preparation El I O Implementation El I