Sri Lanka - Economic Restructuring and Public Manufacturing Enterprises Adjustment Credit Projects

Document of The World Bank FOR OFFICIAL USE ONLY Report No. 15820 PERFORMANCE AUDIT REPORT SRI LANKA ECONOMIC RESTRUCTURING CREDIT (CREDIT 2128-CE) and PUBLIC MANUFACTURING ENTERPRISES ADJUSTMENT CREDIT (CREDIT 2185-CE) June 26, 1996 Operations Evaluation Department This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization. Currency Equivalents Currency Unit = Rupees (Rs) (End Year Rates Rs/US$1.00) 1990 Rs 40.24 1991 Rs 42.58 1992 Rs 46.00 1993 Rs 49.56 1994 Rs 49.98 Abbreviations and Acronyms ARC Administrative Reform Commission CSE Colombo Stock Exchange CTB Central Transport Board ERC Economic Restructuring Credit GDP Gross Domestic Product ICR Implementation Completion Report IDA International Development Association IMF International Monetary Fund JSP Janasaviya Program MOI Ministry of Industries, Science and Technology OECF Overseas Economic Cooperation Fund, Japan OED Operations Evaluation Department PAR Performance Audit Report PFP Policy Framework Paper PIMB Public Investment Management Board PMEs Public Manufacturing Enterprises PMEAC Public Manufacturing Enterprise Adjustment Credit SAF Structural Adjustment Facility SOEs State-owned enterprises UNICEF United Nations Children's Fund UNP United National Party Fiscal Year Government: January I to December 31 FOR OFFICIAL USE ONLY The World Bank Washington, D.C. 20433 U.S.A. Office of the Director-General Operations Evaluation June 26, 1996 MEMORANDUM TO THE EXECUTIVE DIRECTORS AND THE PRESIDENT SUBJECT: Performance Audit Report on the Sri Lanka Economic Restructuring Credit (Credit 2128-CE) and the Public Manufacturing Enterprises Adjustment Credit (Credit 2185-CE) Attached is the Performance Audit Report (PAR) on the Sri Lanka Economic Restructuring Credit (ERC, Credit 2128-CE) for US$106.6 million, and the Public Manufacturing Enterprises Adjustment Credit (PMEAC, Credit 2185-CE) for US$120 million. The ERC was approved on May 1, 1990 and became effective on May 4, 1990. PMEAC was approved on November 27, 1990, and became effective on December 13, 1990. Final disbursement for ERC was made on May 16, 1995 and it closed on December 31, 1994. The last disbursement for PMEAC took place on April 11, 1996 and its closing will be on May 30, 1996, almost two years later than planned. Cofinancing in the amount of US$120 million was provided for the ERC by the Overseas Economic Cooperation Fund (OECF) of Japan. These related credits were designed to support the Government of Sri Lanka's program to move to a market-driven economy. The objectives of ERC were to reduce the size of the public sector and improve its efficiency, develop the private sector, and raise the consumption levels of the poor. To reach these objectives, the ERC included measures to stabilize the economy, reform the tax and customs system, downsize the civil service, encourage privatization, and better target transfer programs. The objectives of PMEAC were to provide a new institutional framework for Public Manufacturing Enterprises (PMEs), with 14 PMEs specifically targeted for privatization. The ERC partially met its objectives. The Government deficit was reduced, tariffs were lowered or removed and the tariff structure simplified. The economic growth rate increased and the external account balance improved. However, the deficit gains were largely lost by 1995 and civil service reform failed. Targeting of anti-poverty expenditures did improve, but welfare subsidies rose. PMEAC was more successful. Most importantly, all of the 14 targeted PMEs were converted into commercial companies, and most were fully or substantially privatized. On the other hand, the new Public Investment Management Board designed to move companies to private ownership and manage the Government's holdings in the privatized enterprises and the remaining PMEs was not successful. The legal status of the Board was never clarified. 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. The Operations Evaluation Department (OED) rates the outcome of the ERC as marginally satisfactory, its sustainability as uncertain, and its institutional development as negligible. These ratings differ from those in the Implementation Completion Report (ICR), which were satisfactory, likely and partial, respectively. The main reasons for the lower ratings are that several components failed (e.g., civil service reform), were not implemented (stabilization measures) or were reversed (subsidy targeting, ending domestic price controls). OED rates Bank performance as unsatisfactory (compared to satisfactory in the ICR) because the reform agenda was too ambitious, and also because it failed to take into account some of the Bank's own research on previous problems in implementing voluntary separation packages in Sri Lanka. On PMEAC, OED rates the outcome as satisfactory, institutional development as modest, sustainability as likely, and Bank performance as satisfactory; these ratings agree with those in the ICR, except for institutional development which the ICR rated as substantial. The major lessons are: for Sri Lanka to continue to experience growth with equity it must once again address the stabilization issues it has twice failed to address. The quality and governance of public expenditures played a key role in Sri Lanka's failure to stabilize. Hence, civil service reform which is essential to improve this governance remains at the center of Sri Lanka's reform agenda. Attachment FOR OFFICIAL USE ONLY Contents Preface ......... ................................................. 3 Basi c Data Sheet... ................................................. 5 Evaluation Summary...............................................11 1. Background ........................................ ....... 19 A. The First Phase of Adjustment: 1977-88............................. 19 B. Problems of Stabilization in the 1980s: The Antecedents of the Adjustment Program ................................. .....20 C. The Targets of the Adjustment Program ..........................21 2. The Adjustment Program ......................................25 A. ERC: Objectives and Policy Instruments .........................25 Objectives ....................... ...............25 Components ...................................... .....26 B. ERC: Implementation Experience and Outcome....................26 Growth and Macro Aggregates ....................... .........26 Fiscal Measures .........................................27 Civil Service Reform ......................... .........29 Poverty Alleviation ............................ .......29 Creating a More Competitive, Transparent Economic Environment...............30 ERC Summary ..........................................31 C. PMEAC: Objectives and Policy Instruments ..................31 PMEAC: Implementation Experience and Outcome......... ...........32 Changes in Tariffs, Market Privileges and Fiscal Support... ....................32 Extent of Privatization ............................. ........33 Improved Economic Efficiency .......................... .....35 Transparency of the Process and Public Reactions.........................................36 Institutional Development ...................................37 Compliance with Credit Covenants .............................37 3. Lessons and Sustainability................................39 A. Stabilization and Governance ................................39 B. Financial Sector Development and Adjustment ........ .............39 This report was prepared by Robert Buckley (Task Manager) and Patrick Grasso (Consultant) who audited the projects in January 1996. Norma Namisato provided administrative assistance. The report was issued by the Country Policy, Industry, and Finance Division (Manuel Pefialver, Chief) of the Operations Evaluation Department (Francisco Aguirre- Sacasa, Director). 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 wiLhout World Bank authorization. 2 C. The Political Economy of Hostilities and Aid Disbursement ..............41 D. Privatization for Efficiency Gains Rather Than for Budgetary Purposes ......41 Bibliography ....................................................43 Annexes I. ERC - Status of Legal Covenants ..................................47 PMEAC: Status of Legal Covenants .....................................49 II. Aggregate Costs of the Civil Service Separation Program ...................51 III. Sri Lanka: Economic Indicators ............................ .......53 Attachments I. Letter from the Ministry of Industrial Development............... ...........55 II. Letter from The Overseas Economic Cooperation Fund, Japan.... .................. 57 3 Preface This is the Performance Audit Report (PAR) on Sri Lanka Economic Restructuring Credit (ERC), involving Credit 2128-CE for US$106.6 million, and Public Manufacturing Enterprises Adjustments Credit (PMEAC), involving Credit 2185-CE for US$120 million. The ERC was approved on May 1, 1990 and became effective on May 4, 1990. PMEAC was approved on November 27, 1990, and became effective on December 13, 1990. Final disbursement for ERC was made on May 16, 1995 and it closed on December 31, 1994. For PMEAC the last disbursement took place on April 11, 1996 and its closing will be on May 30, 1996, almost two years later than planned. Cofinancing in the amount of US$120 million was provided for the ERC by the Overseas Economic Cooperation Fund (OECF) of Japan. The PAR was prepared by the Operations Evaluation Department (OED). The report is based on the Staff Appraisal Reports, the Memoranda of the President, the Implementation Completion Reports, the official files of the projects, and Bank economic and sector work. A mission visited Sri Lanka in January 1996 to discuss the relevance and effectiveness of the projects with the Government of Sri Lanka. Their cooperation in the preparation of this report is gratefully acknowledged. The performance ratings of the PAR for ERC disagree with those of the ICR prepared by the South Asia Regional Office. The PAR provides a more critical assessment of ERC's overall outcome, institutional development and Bank performance. PMEAC's institutional development rating is also lower. The main differences between this audit and earlier analyses are: (i) the emphasis given here to identifying how the reforms pursued had antecedents in an earlier liberalization effort; and (ii) showing how these reforms and their antecedents were affected by the ongoing hostilities. The draft PAR was sent to the Borrower and the cofinancing agency for comments. The comments received from the Ministry of Industrial Development and The Overseas Economic Cooperation Fund, Japan are reproduced as Attachments to the PAR.  5 Basic Data Sheet EcoNoMIC RESTRUCTURING CREDIT (CREDIT 2128-CE) Key Project Data (amounts in US$ million) Appraisal Actual or Actual as % of estimate current estimate appraisal estimate Total project costs 90.0 109.4 122 Credit amount 90.0 106.6 118 Cofinancing 120.0 Cancellation 1.26 Date physical components completed n.a. n.a. Economic rate of return n.a. n.a. Cumulative Estimated and Actual Disbursements FY90 FY91 FY92 FY93 FY94 FY95 Appraisal estimate (US$M) 42.5 85.0 87.5 90.0 90.0 90.0 Actual (US$M) 28.0 55.6 106.0 107.2 108.0 109.4 Actual as% of appraisal 66 65 121 119 120 122 Date of final disbursement: 05/16/95 Project Dates Original Actual Identification 05/03/89 05/29/89 Preparation 12/13/89 02/13/89 Appraisal 11/06/90 01/14/90 Negotiations 03/28/90 03/26/90 Board approval 05/01/90 05/01/90 Signing 05/02/90 05/02/90 Effectiveness 05/04/90 05/04/90 Project completion 06/30/94 12/31/94 Closing date 06/30/94 12/31/94 6 Staff Inputs (staff weeks) Planned Revised Actual Through appraisal 141.0 122 60.8 Appraisal-Board 88.0 101 51.0 Board-effectiveness - - - Supervision 31.0 33.3 57.8 Completion - 6.0 - Total 260.4 262.3 169.6 Mission Data Date No. of Staff days Specializations Performanc rating Types of (month/year) persons infield represented Impl. stat. Dev.object Problems Through appraisal 10/89 10 10 Econ., Ind.. 01/90 15 15 Through Board n.a. n.a. n.a. n.a. n.a. n.a. n.a. approval Through n.a. n.a. n.a. n.a. n.a. n.a. n.a. effectiveness Supervision 07/90 4 10 Economist, S S 08/90 3 11 Technical S S 12/90 2 18 assistance S S 04/91 4 17 (op/finance) S S 11/93 3 5 S S Completion 07/95 1 2 Economist S S Other Project Data Borrower/Executing Agency: Government of Sri Lanka FOLLOW-ON OPERATIONS Operation Credit no. Amount Board date (US$ million) Poverty Alleviation 2231 57.50 04/25/91 Small & Medium Industry IV 2250 45.00 05/28/91 Private Financial Development 2484 60.00 04/20/93 7 Basic Data Sheet PUBLIC MANUFACTURING ENTERPRISES ADJUSTMENTS CREDIT (CREDIT 2185-CE) Key Project Data (amounts in US$ million) Appraisal Actual or Actual as % of estimate current estimate appraisal estimate Total project costs n.a. Credit amount 120.0 Cofinancing n.a. Cancellation n.a. Date physical components completed n.a. Economic rate of return n.a. n.a. n.a. Cumulative Estimated and Actual Disbursements FY91 FY92 FY93 FY94 FY95 Appraisal estimate (US$M) 57.5 115.0 117.5 123.3 125.8 Actual (US$M) 56.6 61.7 119.4 126.2 126.6 Actual as % of appraisal 98 54 102 102 101 Date of final disbursement: (still disbursing) Project Dates Original Actual/Estimate Identification 11/17/88 11/17/88 Preparation 04/01/88 02/28/88 Appraisal 01/14/89 05/14/90 Negotiations 09/01/89 08/01/90 Board approval 11/01/89 11/27/90 Signing 11/02/89 11/30/90 Effectiveness 12/10/89 12/13/90 Project Completion 11/30/94 11/30/94 Closing date 05/30/96 8 Staff Inputs (staff weeks) Planned Revised Actual Through appraisal 192.0 Appraisal-Board 29.5 Board-effectiveness - - Supervision 45.9 46.9 97.5 Completion - 6.0 Total 45.9 52.9 319 Mission Data Date No. of Staff days infield Specializations represented (month/year) persons Through Appraisal 01/90 2 12 Industry, Finance 04/90 2 16 05/90 1 18 Industry Appraisal through Board n.a. n.a. n.a. n.a. approval Board approval through n.a. n.a. n.a. n.a. effectiveness Supervision 03/91 Ind., finance, eng. 06/91 10/91 Industry/finance 03/92 Engineering 04/92 Engineering 11/92 Technical Assistance 12/93 Technical assistance Completion 07/95 1 1 Economics 9 Other Project Data Borrower/Executing Agency: Government of Sri Lanka FOLLOW-ON OPERA TIONS Operation Credit no. Amount Board date (US$ million) Poverty Alleviation 2231 57.50 04/25/91 Small & Medium Industry IV 2250 45.00 05/28/91 Private Financial Development 2484 60.00 04/20/93  11 Evaluation Summary 1. Prior to 1977, Sri Lanka had a closed, tightly regulated, and highly centralized economy. From independence in 1948, massive social security and welfare expenditures were financed by taxes on the export sector. Policymakers, like their counterparts in many other developing countries, had hoped to reduce the heavy dependence of the economy on foreign trade. They had also hoped to shift the economy away from dependence on having a primary commodity export structure. However, the dirigiste model ruptured. Low growth, high unemployment, and rationing nurtured disaffection. A shift to an open, market-friendly policy environment occurred in 1977. Trade liberalization, higher growth, partial liberalization of financial markets, and the availability of consumer goods helped sustain the reforms, and created an environment of broad- based increases in income. The increases allowed cuts in welfare expenditures from one of the highest levels in the world. However, the ethnic crisis that escalated after 1983, and the rebellion in the South during 1987-89 were extremely disruptive. When these disruptions are combined with the macroeconomic imbalances the result was policy paralysis, and a stagnating economy. 2. By the late 1980s, it was clear that macroeconomic policy was on an unsustainable path. Reserves were drawn down, deficits increased, and external debt quadrupled. On the fiscal front, the efficiency of public sector enterprises did not improve; heavy subsidies were needed to sustain them. In addition, there was slippage in opening up the economy as high tariffs were also maintained. Despite these problems, on the financial side, there was some progress. As a result, while significant problems still remained at the time of adjustment in FY90, financial liberalization had occurred. This liberalization helped the financial sector to become an important resource for private sector development. The Targets of the Adjustment Program 3. Fiscal issues. On the revenue side, liberalization of trade fundamentally changed the tax base. Tax revenues fell during the 1980s as the Government found it difficult to find new sources of revenue, and a Tax Commission was established to make recommendations for reform. There were more serious problems on the expenditure side. Foremost among these was the need to restructure the civil service to make it more professional, smaller, and more consistent with a market-oriented economy. As was the case with taxes, a commission produced a study of how the civil service could be revamped. Other expenditure problems were the continual pressure for more government expenditures for poverty alleviation, and the resource drain of public enterprises. Finally, increases in military expenditures also mounted as the conflicts intensified. 4. The role of the private sector. Besides the problems with the size of government, there were also difficulties about defining the Government's role in the economy. The boundary of what was public and private was unclear. Many basic services had been nationalized or subjected to extensive controls and opaque regulations for many years. Finally, the early years of adjustment, 1978-82, was a period of severe external shocks, while the subsequent period, beginning in 1983, was one of ethnic conflict and deep civil unrest. Thus, neither period provided an environment hospitable to private sector development. 12 The Adjustment Program 5. At the time of Board approval of the Economic Restructuring Credit (ERC) and the Public Manufacturing Enterprises Adjustment Credit (PMEAC) in FY90, growth had fallen- well below its long-term trend-to 2.3 percent per year in 1989. The process of economic reform agreed to with the IMF in late 1987, and supported through a Structural Adjustment Facility (SAF) and Policy Framework Paper (PFP), had been put on hold. Implementation of adjustment at the pace envisaged in the first year PFP-SAF was not achieved. Violence had escalated in the South and spread to such a degree that work stoppages reduced production, exports, and tax collections. In 1988, the deficit reached 16 percent of GDP instead of the 11 percent target of the first year PFP. In 1989, a new government sought to revive the economy and complete the transition to a market economy. To this end, ERC included measures to reform the tax and tariff system, target transfers, downsize the civil service, and develop the private sector. PMEAC focused on privatization of state-owned manufacturing. ERC: Objectives and Policy Instruments 6. ERC had the following objectives: (a) Reduce the public sector's size and improve its efficiency. (b) Develop the private sector by reducing and rationalizing tariffs, and deregulating, privatizing, and/or restructuring: plantations, transport, the Paddy Marketing Board, and air freight operations. (c) Improving the consumption levels of the poor. ERC: Implementation Experience and Outcome 7. The ERC Appraisal Report described the PFP targets as modest but realistic given Sri Lanka's political constraints. Outside of inflation, these targets were largely achieved by the end of the PFP period in 1993. Growth was considerably higher, the deficit target was only slightly higher, and the external account balance was considerably better than the targets. In addition, considering the sources of growth-increases in foreign and private sector investment-the higher growth rate seems quite robust. Over the 1990-95 period, the growth rate compares quite favorably with those of other adjusting countries; it is higher than those realized by all other adjusters except those in East Asia. Moreover, even those targets which Sri Lanka failed to fulfill-inflation and the size of the deficit-are significantly better than those realized in the strongest performers among transition economies. Extending the perspective to 1995, however, reveals that most of the gains on deficit reduction have been lost, and although progress on reducing inflation continued, it still exceeded the PFP target for 1993. Thus, stabilization has not been realized. 8. Revenues. Moderately high rates of inflation and high real borrowing costs contributed to revenue problems which in turn made the cost of capital to the private sector very high. To offset these costs, the Government offered tax incentives, thereby reducing overall taxes to 17.2 percent of GDP in 1994, a low level of tax effort for an economy with Sri Lanka's expenditure levels. Given this result, Sri Lanka's ability to successfully stabilize was problematic. Little 13 progress was made on implementing the recommendations of the Tax Commission as proposed by ERC. 9. Civil service reform. The government performance on these reforms was clearly unsatisfactory. To induce departures, the Government introduced a voluntary separation package that was not withdrawn at the end of the reform, but left available to all retiring civil servants. Moreover, the government hired more new civil servants to replace the retiring personnel. The results were that: (i) by 1992 there were almost 10 percent more civil servants than there were in 1989; (ii) nothing was done to implement an objective, professional selection system; and (iii) the number of ministries did not decline. Indeed, they have almost doubled in number. It is hard to imagine a less satisfactory outcome. Indeed, the budgetary implications of these actions by themselves account for more than the amount by which the PFP deficit target was exceeded. There were no savings for salaries foregone and pensions increased three-fold. Moreover, in 1989, the Civil Service Reform Commission had warned that the kind of piecemeal implementation of reform that was enacted would be chaotic and set the system back, as it did. 10. Poverty alleviation. The credit's covenants called for the elimination of a number of price controls and improved targeting of welfare subsidies. The objective was to improve the targeting of assistance to those most in need at a time of price increases, and simultaneously, to keep the total level of expenditures less than 3.5 percent of GDP. Performance was unsatisfactory. The total level of expenditures ultimately exceeded 3.7 percent of GDP. Moreover, decreases in subsidy expenditures came through inflationary erosion of benefits to the poor rather than elimination from program eligibility of those who were not poor. Targeting did improve for a time, only to be replaced by a new subsidy program, with all the hallmarks of a highly politicized distribution of benefits. Regressive price subsidies were also introduced for wheat, and many of the other price controls were reintroduced after ERC's second tranche release. 11. Import tariffs and price liberalization. Export performance continued to remain strong throughout the adjustment increasing to 27 percent of GDP from 21 percent in 1988. At the same time, the tariffs on imports declined as planned as a share of GDP, and the share of imports rose. Tariffs have also been simplified as well. The deregulation of freight services has continued to result in reductions in freight charges and increases in competition. On domestic prices, a greater reliance on market forces was considerably less successful. Buses were turned over to employees, but fares and salaries continued to be controlled by the Government. The former have not been increased, and the latter increased by 30 percent. Buses, as a result, still operate at a loss. Nevertheless, given the strong performance on private and foreign investment and growth, this component is rated as satisfactory. PMEAC: Objectives and Policy Instruments 12. PMEAC had the following objectives: (a) Remove the remaining specific tariffprotections, market privileges and fiscal support for PMEs. (b) Provide a new institutionalframework for PAEs. (c) Convert PMEs into commercial companies and privatize them where possible. 14 (d) Phase out non-viable PME activities altogether. PMEAC: Implementation Experience and Outcome 13. PMEAC was more successful in meeting its more narrow objectives than was ERC. Market privileges and fiscal support were either eliminated or substantially reduced. The targeted SOEs were converted to commercial companies, and most were privatized, as planned. However, the extent to which privatization has led to improved economic efficiency is less clear. Moreover, the process has not been fully transparent, leading to perceptions of inequity, and contributing to a slowdown in privatization activities. 14. Extent ofprivatization. Overall, Sri Lanka has made dramatic progress in privatization. By 1994 at least 41 of the nearly 100 SOEs extant in the late 1980s had been divested fully or substantially, and many more were in the process of being privatized. The election of a new government in 1995 with strong union backing appears to have slowed the speed of privatization. This reflects both union resistance to privatization and a deliberate strategy of dealing with the ethnic conflict as a first priority. Nevertheless, the current Government has reiterated its commitment to the privatization program it inherited, and a new privatization commission was appointed in 1995. Significantly, there has been no renationalization of privatized companies, nor is any being contemplated. 15. Gains from privatization. Data on efficiency gains under privatization are thin. Two small-scale studies showed some indications of increased efficiency, but the picture was mixed and inconclusive. However, prior to privatization, private sector productivity growth had been much higher than public sector performance for a number of years. There also was evidence of considerable overstaffing of public enterprises. 16. The privatization process. The implementation of the privatization program has raised questions of transparency and equity. Because of the lack of a well functioning equities market in Sri Lanka, most privatizations have involved sales through tenders or negotiated private offerings. This has lead to charges that valuable companies were sold for less than market value, and the subsequent performance of shares in some companies on the stock exchange has reinforced that perception. The tension over this issue has been exacerbated by concerns over the equity of the privatization program. In addition, the stock transfer program for workers of privatized companies benefited only the relatively few workers who were employed by the specific firms (about one percent of the total national workforce). There were wide disparities in the value of the stock they were given. These equity issues appear to have reduced public support for the program. Sustaining the Adjustment 17. To sum up, in many respects, Sri Lanka has made remarkable progress on its adjustment, and it has done this in a very difficult environment. The economy has opened up, and achieved high rates of growth. It is subject to more competitive pressures, and governed by private sector investment decisions. However, stabilization was never achieved and it appears now that it is being reversed. For Sri Lanka to achieve a durable sustained adjustment, there are a number of lessons than can be drawn from this adjustment experience. Many of these lessons apply with even greater force to the Bank's support for such adjustment programs. 15 18. Stabilization and governance. Sri Lanka has not established the fundamentals of a professional public service that is free from political manipulation. Its ability to implement programs based on the effective reports of Commissions established by the Government is lacking. All too often policy appears to operate in a defensive mode, attempting to scale back and rationalize proposals for large-scale public sector programs. For Sri Lanka to continue to experience growth with equity it must once again address the stabilization problems that it has twice failed to resolve. Because control over the efficacy of government expenditures has played such an important role in Sri Lanka's failure to stabilize, a key element of a successful stabilization will be more effective management of public expenditures. This effective management requires a sound, professional civil service. Consequently, civil service reform remains at the center of Sri Lanka's ability to sustain the adjustment program. 19. The political economy of hostilities and disbursement. The failure of civil service reform is indicative of the difficulties of government policy during hostilities. When policy reversals obviate all the gains of severance pay expenditures, it is clear that policy changed quite dramatically and in response to the exigencies of the conflict. For example, the Janatha Vimukthi Peramuna (People's Liberation Front) uprising in 1987-89 generated a number of public employment and expenditure programs targeted on those who might otherwise become involved in hostilities. Rural poverty, for instance, became the target for a major poverty alleviation program in 1989 as a direct response to the insurgency in the south. Until the conflict is resolved, stabilization will almost certainly be beyond reach, and the policy horizon will remain extremely near-sighted. 20. For the Bank, perhaps the major lesson is that the adjustment program disbursed too much money over too short a time period. The reform targets were appropriate and, for the most part, the strategy was thoughtful. However, in light of the imbalances generated by the ongoing hostilities and the volatile nature of program ownership over time, it was a mistake to disburse the large amounts of these credits, more than 4.5 percent of GDP, in such a short time period. A change in political circumstances, as in fact occurred, could more easily lead to a rejection of previously agreed covenants. While this conclusion is easier to make with the benefit of hindsight, particularly in light of the much longer adjustment periods for reforming socialist economies, it should not have been difficult to make at the time. The IMF's SAF program had collapsed one year prior to credit approval. In addition, earlier adjustment lending discussions between the Bank and Sri Lanka focused on disbursing smaller amounts of assistance over a much longer time period. 21. Financial sector development and adjustment. It appears to have played an important role in Sri Lanka's high post-adjustment growth. Sri Lanka's growing financial sector was better able to mobilize resources and allocate them among private demanders of credit. Once the tariff reforms and privatization measures of the adjustment program were in place this facilitated adjustment. Recent World Bank research confirms this view. It suggests that a deeper financial sector permits countries undergoing adjustment to weather the price changes much more resiliently. However, just as the improvements in the financial sector facilitated the adjustment process, so too have the remaining financial sector distortions constrained it. A major concern of World Bank financial sector work in Sri Lanka has been the inefficiencies of the two large state commercial banks. The high interest rate margins that they charge are a severe impediment to investment. Thus, there are two types of lessons on financial sector policy. First, the productivity of the studies undertaken by the financial intermediary loans made important contributions to Sri Lanka's high growth rate. Second, the failure to complete the restructuring 16 of the state commercial banks can be expected to remain a significant impediment to the functioning of the economy. Sri Lanka's saving rate is far too low and the Government's presence in the credit markets far too high for the country to be able to afford the inefficiencies of these institutions. 22. Privatization for efficiency gains rather than for budgetary purposes. The sustainability of the specific privatizations undertaken through PMEAC is likely. However, the future of Sri Lanka's relatively successful privatization program is threatened by an emphasis on the effects that divestiture has on the government budget rather than its effect on the efficiency with which resources are used. The poor growth performance of economies with public control over production is well know. Sri Lanka is no exception to this trend. Its public sector industries have been unable to respond to market incentives. The productivity of the private sector in Sri Lanka has been, and continues to be, considerably higher than that of the public sector. Thus, the motivation for privatization is to shift resources to a more effective governance structure. The budgetary gains, or losses, associated with privatization are of secondary importance. 23. It is also important to improve the perceived fairness of the privatization process. Providing some workers shares in essentially worthless enterprises while others reap windfall profits, as has been done, is inequitable. It is also increasingly unsustainable as most profitable companies have already been privatized. 17 OED Ratings of the ERC (C2128-CE) Rating Type Rating Assessment Outcome Marginally The project's objectives were highly relevant to the country at Satisfactory approval. The adjustment program responded to the need to stabilize the economy and continue the liberalization process. The credit contributed to restoring growth and opening up the economy. However, almost all of the stabilization measures addressed by ERC were not implemented. Civil service reform was a failure; subsidy targeting was only briefly improved; and domestic price controls were reintroduced. On a positive note, tariffs and licensing restrictions were reduced and private and foreign investment increased. Institutional Negligible This adjustment was Sri Lanka's second attempt to stabilize and Development reorient the structure of the economy. While considerable reorientation has taken place, both efforts at stabilization have failed. The external environment has not been hospitable to stabilization. Nevertheless, a major weakness has been the lack of an effective, professional civil service. ERC not only failed to achieve these reforms, it made the situation worse. Before civil service reform can be taken up again, the generous retirement package to induce separation put in place by ERC must be removed. Similarly, the new poverty alleviation program and the restored price controls have all the markings of political programs rather than programs designed to help the poor. Sustainability Uncertain Sri Lanka's performance since the adjustment has been very promising. Growth has been restored, the private sector has flourished, the soft budget constraint on public firms has been tightened, and international price distortions have been reduced. On the other hand, growth impeding distortions remained and, for growth to be sustained, a third attempt at stabilization will be necessary, and renewed hostilities make it more difficult to achieve. Borrower Unsatisfactory During the course of the adjustment, the Government reversed its Performance policy stance on 10 of 16 broad categories of ERC's covenants. The civil service reform it enacted had a negative impact. Renewed hostilities created a crisis management environment that called for policy redirections, but some of the policy reversals appear excessive, and warnings of Government commissions on the likely outcome of piecemeal implementation of some reforms were ignored. Bank Unsatisfactory In hindsight, the Bank undertook a far too ambitious reform program Performance in a country that was undertaking both basic systemic change as well as stabilization in a very difficult environment. The lessons that have been learned from Bank support to other transition economies since this credit auger for fewer, more focused conditionalities, and a longer period of adjustment. In addition, the Bank can be faulted for supporting the use of a retirement package that according to Bank research, had been ineffective in Sri Lanka. Covenant Weak Covenant compliance was weak; only 6 out of 16 covenants were Compliance still in effect in 1995. 18 OED Ratings of the PMEAC (C2185-CE) Rating Type Rating Assessment Outcome Satisfactory The project's objectives were highly relevant to the country at approval. It addressed the need to reduce the number and scope of state owned enterprises (SOEs) as part of the broader liberalization of the Sri Lankan economy. Most of the SOEs targeted for privatization were in fact privatized, and there has been no re-nationalization. Institutional Modest' The major institutional change supported by the credit Development was the creation of the Public Investment Management Board, intended to supervise the commercialization of the SOEs and to manage the state's shares in enterprises. This was not particularly successful because the position of the Board, which was attached to the Ministry of Finance, vis-a-vis the line ministries that oversaw the SOEs was never clarified. Sustainability Likely There seems little danger that the privatizations that took place under the credit are not permanent. The Government has reiterated its support for the privatization program that it inherited. However, further progress is threatened because of questions surrounding the transparency and equity of the privatization program. Borrower Satisfactory The Government's performance both in preparation Performance and implementation of the credit was satisfactory. There was a one-year delay in meeting conditionality for implementation of the credit, but thereafter the Government took the necessary steps to achieve a satisfactory outcome. Bank Satisfactory The Bank satisfactorily helped design a program that Performance addressed an important need of the Borrower, had Government commitment, and was achievable. Covenant Substantial All the covenants were complied with, although with Compliance delays in a few cases, except for one which the Bank agreed had been overtaken by events and was no longer applicable. The Region agrees that the achievements were less than "substantial" but feels that the rating of "modest" is an understatement of the achievements in this area. 19 1. Background 1.1 Sri Lanka's social welfare accomplishments are well known. Its growth performance is both less documented, and less sterling. It has grown considerably less rapidly and experienced much wider oscillations than many comparators. In early 1989, after six years of civil unrest, and in immediate response to a financial crisis, a new government with a new president undertook a fundamental reorientation of the economy. It sought to complete the liberalization program begun in 1977, by rationalizing the public sector and making the private sector the engine of growth. At the same time, the Government wanted to preserve Sri Lanka's edge in human development. It sought to ensure that the benefits of growth would be widely shared, especially by the poor. In many respects, then, the adjustment program of the late 1980s had its roots in the reforms that began slightly more than a decade earlier. A. The First Phase of Adjustment: 1977-88 1.2 Prior to 1977, Sri Lanka had a closed, tightly regulated, and highly centralized economy. The government had direct control over trade, and public corporations existed across the economy. From independence in 1948, massive social security and welfare expenditures-food subsidies, for example, accounted for 15 percent of government expenditures in the early 1960s. Sri Lanka became one of the most inward-looking welfare states in the world, supported by taxes on the export sector.2 Policymakers, like their counterparts in many other developing countries, had hoped to reduce the heavy dependence of the economy on foreign trade. They had also hoped to shift the economy away from dependence on having a primary commodity export structure. 1.3 However, in the early 1970s the dirigiste model ruptured. Deterioration in the terms of trade, nationalization of plantations, drought and a significant fall in the output of plantation crops created balance of payments problems and large budget deficits. The old model of growth was no longer sustainable. Low growth, high unemployment, reaching to 24 percent, deteriorating social services, and rationing and black marketeering nurtured disaffection. The result was a landslide victory for a market-oriented party, the UNP, in the 1977 elections. 1.4 The new government had a strong mandate for reform, and the absence of effective opposition in the initial years gave its policies political momentum. There were also clear indications that with economic liberalization substantial foreign assistance would be forthcoming. Trade liberalization, higher growth, partial liberalization of financial markets, and the availability of consumer goods provided tangible benefits to consumers, and helped sustain 1 See Ramey and Ramey (1995) for a discussion of income volatility across countries. Over the 1960-85 period Sri Lanka was the 54th most volatile of 70 non-African countries. Its per capita income growth rate over this period was slightly more than half the average of that realized by Thailand, Indonesia, Malaysia, and Korea over the 1960-92 period. See Annex 3 for macro indicators for Sri Lanka. 2 De Long and Summers (1993) show that due to import restrictions, the price of equipment in Sri Lanka was the highest in the world in terms of output per worker. Besides Bank documents this section relies on Athukorala and Jayasuriya (1994), Dunham and Kelegama (1994), and Dunham and Kelegama (1995). 20 the reforms. They also helped create an environment of broad-based increases in income that allowed cuts in welfare expenditures from one of the highest levels in the world to levels approaching those of other developing countries.4 However, the pace of reform slowed as the economy experienced both internal and external shocks. B. Problems of Stabilization in the 1980s: The Antecedents of the Adjustment Program 1.5 External conditions. The downturn in the terms of trade after the second oil price hike in 1979 had an important adverse effect on the economy. Tea prices collapsed in 1978 and prices of imports such as, oil, fertilizers, sugar and investment goods rose in the following year. Between 1978 and 1982 the terms of trade deteriorated by over 40 percent, placing enormous pressure on the adjustment process and on the economy in general. It has been estimated that Sri Lanka lost the equivalent of 25 percent of its national income from terms of trade effects in the late 1970s and early 1980s.5 Then, as this shock ended another one took hold. Ethnic violence began in 1983. 1.6 Internal Conditions. As violence grew and growth receded (per capita income growth in the first 6 years after liberalization was almost twice as rapid as that in the second 6 year period), the country's deeply ingrained welfare tradition made successive governments sensitive to the situation of the poor and vulnerable sections of the community. By the mid-1980s, there was evidence that growth was accompanied by high unemployment and deteriorating standards of living. Government sensitivity to this problem was compounded by the violence, and its ability to act on these concerns was increased by the greater centralization of power that came with the establishment of an Executive Presidency in 1978. There was a sharp escalation in military costs to meet the Tamil secessionist movement in the North and East of the country after 1983. Defense costs made increasing demands on the government budget, rising from 1 percent of GDP in 1982 to 8 percent in 1986, one of the largest increases in the world, making stabilization considerably more difficult.6 1.7 To sum up, it is difficult to exaggerate how adverse Sri Lanka's external and internal environments were for stabilizing the economy and inducing the private sector to play a greater role. The 25 percent loss of national income due to terms of trade effects was almost certainly more than matched by the effects that military expenditures and violence had on economic activity, not to mention the extraordinary human costs of the conflicts.7 To get a sense of how costly these "events" were in terms of stabilization it is helpful to consider a naive counterfactual situation: had there been no violence or terms of trade effects, and correspondingly, no need to 4 Even in 1981, the earliest year for extensive comparable data, Sri Lanka had the highest expenditures on welfare as a share of government expenditures of all low- and middle-income countries. See Athukorala and Jayasuriya (1994) who cite the estimates of Balassa and McCarthy. 6 See the World Development Report 1988. Sri Lanka had the seventh largest increase in the share of GDP on defense expenditures of 54 countries for which data is available for the period. 7 A recent study by Knight, Loayza, and Villanueva (1996) shows the effects that military expenditures can have on growth and investment. From their analysis, the magnitude of the increase in military expenditure in Sri Lanka is likely to produce negative growth effects on the order of 1.5 percent per year. Over 13 years of violence, an aggregate figure of about 20 percent can be derived. While such estimates must be seen as a very rough reference point, it is clear that the violence has been very costly. 21 increase government expenditures on defense, rather than a 16 percent of GDP deficit in 1988, the deficit would have been reduced by 6 percent of GDP due to the elimination of the increase in defense expenditures. Without the violence and terms of trade effects, GDP would have been much higher. As a result, the deficit would have been on the order of 6 or 7 percent of GDP, and the deficit reductions realized in 1989 would have stabilized the economy. 1.8 Nevertheless, the experience of the 1980s was certainly not without policy failures, as noted by Lal and Rajapatirana (1989). In particular, the extent of macroeconomic instability can be gauged by the budget deficits-as high as 27 percent of GDP-and the high level of inflation-26 percent-towards the end of this period. In this environment, crisis management took precedence over stabilization and further liberalization. In many respects, these failures became the targets of the adjustment program in the late 1980s. Five antecedents of the adjustment program are identified. The first three became targets of ERC, the last two were addressed by both ERC and PMEAC. C. The Targets of the Adjustment Program 1.9 Revenues. Liberalization fundamentally changed the tax base. In the initial post- liberalization period, tax revenues increased. The increase occurred because the economic rents associated with quotas of the old regime rarely went to the Government. Whereas, the shift from quotas to tariffs at the same level of protection augmented revenue. Lowering tariffs also initially boosted revenue due to the large influx of imports. However, lower tariffs on trade and cuts in income tax eventually ate into government revenue. Taxes from previously profitable firms in the import-competing sector declined with trade liberalization, and they were not immediately replaced by taxes from newly profitable exporting firms. The Government found it difficult to find new sources of revenue, and in 1989 established a Tax Commission to study these issues and make recommendations for reform. Policy Target 1: Furthering the development of the work on tax reform and integrating it with further tariff reductions became a focus of the adjustment program. 1.10 Expenditures. There were also policy conflicts on the expenditure side on several fronts. First, there were pressures to expand public employment as a way to address persistent, growing and sometimes violence-sensitive unemployment problems. Indeed, the role of the state remained high in terms of investment and employment. In fact, between 1977 and 1987 the increase in the share of resources under the control of central government increased by more than 9 percent of GDP. This is the second largest increase of the 18 countries examined by a World Bank study of countries which experienced macroeconomic crises. During this period the government share of GDP reached as high as 49 percent. As a result, when Sri Lanka began its adjustment program in the late 1980s, it did so from one of the highest levels of government expenditure of all adjusting economies. One way this pressure was addressed was by establishing a Presidential Commission in 1986-the Administrative Reform Commission (ARC). ARC produced a 10 report study on how the civil service could be revamped, and made into the kind of professional cadre needed to carry out public policy. See Little et al. (1993). It should be noted, however, that 50 percent of this increase was due to increases in military expenditure. 22 Policy Target 2: Civil service reform became a key component of the adjustment program. 1.11 A second aspect of the expenditure problem was Sri Lanka's legacy of concern with social welfare placed continual pressure on government expenditures, resulting ultimately in a new poverty alleviation program, the Janasaviya Program (JSP) in 1989. These pressures were exacerbated by the violence. JSP, for example, was implemented in direct response to the southern insurgency. It called for giving substantial consumption support to half the population for two years. After its enactment, however, it was almost immediately apparent that the budgetary costs of such a program, estimated at 25 percent of GDP, were well beyond the state's capacity. Policy Target 3: It was within the policy context of an emerging, unsustainable new poverty alleviation program that the adjustment program focused on both improving the targeting of poverty alleviation efforts and keeping such expenditures within sustainable limits. 1.12 The third expenditure concern was the resource drain of public enterprises. In the 1980s transfers to public manufacturing enterprises averaged 10 percent of GDP, at the same time that the productivity of these firms was growing at a fraction of that of private enterprises. Policy Target 4: The adjustment program focused considerable attention on privatizing these firms and reducing government obligations associated with supporting them. 1.13 On all of these scores, during the late 1980s policy had become one of crisis management in an unstable macroeconomic and political environment. In such an environment, the stabilization effort was bound to falter. Government expenditures accounted for a very large share of GDP. These expenditures were inefficiently allocated to public enterprises, and to what were often massive, unsustainable programs, such as the Accelerated Mahaweli, the Million Houses Program, and the JSP.9 Further, these expenditures were monitored by a civil service that was subject to political manipulation, and which operated in more than 45 ministries, many of which served no useful purpose in a market-oriented economy. 1.14 But, besides the problems with the size of government, there were also difficulties as to what was the Government's role in the economy. The boundary of what was public and private was in very basic ways unclear. Many basic services had been nationalized or subjected to extensive controls and opaque regulations. Buses, the means of motorized transport for over 90 percent of the population, were nationalized in 1959;1o land in 1972; plantations in 1976; 26 large business enterprises over the 1971-77 period. Ceilings had been placed on incomes, asset ownership, and even ownership of dwellings; and universal food subsidies; and general price, licensing and quantity controls had operated for many years. Finally, the ability to terminate employment, due to the Termination of Employment of Workmen Act, was unclear. Thus, even The first two of these programs were initiated by the Government during the 1977 liberalization. See Athukorala and Jayasuriya (1994) for an analysis of their problems. 10 Terrell (1991), p. 19. Athurokala and Jayasuriya (1994). 23 though the output of public enterprises declined during the 1970s, the public role was . 12 pervasive. Policy Target 5: The adjustment program focused on moving to greater reliance on market forces rather than government regulations to determine the prices of many goods. 1.15 To sum up, Sri Lanka's adjustment program in the late 1980s was a continuation of the reforms of the late 1970s. However, it was also an attempt to throw off the vestiges of a socialist economy. Just as the early liberalization was one of the world's first attempts at structural adjustment, the adjustment program of the late 1980s was one of the first attempts at reforming a socialist economy. As was the case with the earlier program, the later reform took place on largely uncharted grounds. At the time of the adjustment the World Bank had only selectively supported such transitions. As is shown below, when Sri Lanka's macroeconomic performance in the 1990s is viewed against the Bank's experience with reforming socialist economies two results stand out: (i) the ambitiousness of the reform agenda; and (ii) Sri Lanka's strong growth performance even with the slippages. 12 Kelegama (1992) shows that the state sector share of industrial production had declined to 26 percent of total value of industrial production, falling from 64 percent in 1980. However, the plantations and a substantial portion of the financial sector were still in public hands at the time of adjustment.  25 2. The Adjustment Program 2.1 At the time of Board approval of the Economic Restructuring Credit (ERC) and the Public Manufacturing Enterprises Adjustment Credit (PMEAC) in 1990, Sri Lanka was engulfed in a serious political crisis, waging one war in the North and East of the country, and another in the South. Growth had fallen to 2.3 percent per year in 1989. The process of economic reform, which was at best incomplete, had been put on hold. The Government was re-elected to power in 1989; it crushed the rebellion in the South; it sought to revive the economy and complete the transition from socialism. To this end, ERC included measures to reform the tax and customs system, target transfers, downsize the civil service, and develop the private sector. PMEAC focused on privatization of state-owned manufacturing. The program was prepared with IDA assistance, supported through a Structural Adjustment Facility with the IMF, and by Japan, which provided cofinancing for ERC. 2.2 The basic strategy focused on two elements: stabilization of the economy and further liberalization to make the private sector the "engine of growth." In many areas, key reforms had still to be implemented. Besides the Presidential Administrative Commission (1986) and the Tax Commission (1989), there was also a Privatization Commission, renamed the Commission on Peoplization in 1989, and a range of specialist task forces appointed to formulate policies. While the recommendations of these commissions had been accepted by the Government, they remained to be implemented. 2.3 At the beginning of adjustment, the level of private sector investment was relatively low, despite the existence of substantial policy incentives to make it more attractive. But the low level of investment was also aggravated by the lack of a vibrant equity market. For various reasons, the Colombo Stock Exchange, established in 1985, had not become an important means of mobilizing capital. Sri Lankan investors did not rely on equity finance. The weakness of this market, as is discussed below, would have implications for the approach taken to the privatization program and the public perceptions of it. A. ERC: Objectives and Policy Instruments 2.4 Besides helping to stabilize and open the economy, and target resources on the poor, ERC also began a fundamental reorientation and rationalization of the public sector. In particular, ERC had the following objectives and components:13 Objectives (a) Reduce the public sector's size and improve its efficiency. (b) Develop the private sector. (c) Improve the consumption levels of the poor. 13 See Annex I for a list of the covenants of ERC and their outcome. 26 Components (a) Reduce subsidies and rationalize revenue extraction. This measure focused on eliminating subsidies for rice, flour, fertilizer, and transport. It also worked to begin implementing the reform proposals of the Tax Commission. (b) Civil service reform. Implement the recommendations of a Presidential Administrative Reforms Committee (ARC) which called for changing the civil service system from one designed to serve the needs of a centralized, administered economy into a smaller (70,000 fewer employees), more professional structure that could serve the needs of a market economy. (c) Accelerate poverty alleviation. This measure called for maintaining expenditures on three poverty programs within agreed limits, and for the restructuring of two of the three programs to target resources more effectively. (d) Develop the private sector by reducing and rationalizing tariffs, and deregulating, privatizing, and/or restructuring: plantations, transport, the Paddy Marketing Board, and air freight operations. 2.5 In short, ERC aimed to encourage export-oriented growth of the private sector by freeing public resources. It also sought to target the remaining public expenditures in a more transparent, better focused way. Finally, it attempted to deregulate and reduce the Government presence in areas in which it had no comparative advantage. Cumulatively, these measures aspired to affect the economy in two ways: (i) by creating a more competitive, resource-efficient environment; and (ii) by the scale effects associated with reducing the large government role in the economy. This reduction in the public role would facilitate the private sector's ability to compete for and allocate resources. B. ERC: Implementation Experience and Outcome Growth and Macro Aggregates 2.6 In the year prior to ERC's approval, violence escalated, and there was a deteriorating macroeconomic situation. Tourism and foreign investment fell; taxes could not be collected due to security problems, and defense-related expenditures caused the 1988 deficit to reach 16 percent of GDP instead of 11 percent as targeted in the first-year Policy Framework Paper (PFP). Gross official reserves fell to the equivalent of three weeks of imports. The result was the program outlined in the second-year PFP, and agreed to with the IMF. Table 2.1 lays out the Macroeconomic Targets of the PFP, 1989-1993. 27 Table 2.1: Macroeconomic Targets of the PFP, 1989-1993 and Results 1989-95 Indicator Target Target Realized Realized by end 1989 (%) by end 1993 (%16) 1993 (%) 1995 (%) Real GDP growth 2-3 Around 5 6.9 5.2 Budget deficit 12.5 of GDP 8 of GDP 8.3 10.0 Inflation 12 6 11.7 8.0 External current Around 10 of GDP Around 5 of GDP 3.8 7.5 account deficit Source: Durham and Kelegama (1993), Institute of Policy Studies (1995), Econsult Annual Report 1995. 2.7 As the table shows, outside of inflation, in the short term, the targets were largely realized. Growth was considerably higher, the deficit target was slightly above the target, and the external account balance was considerably better than the target. In addition, if one considers the sources of growth-increases in foreign and private sector investment (the former increased seven-fold over the 1990-93 period to 2 percent of GDP, and the latter increased from 11.7 percent of GDP in 1989 to almost 24 percent in 1994)--the higher growth rate seems quite robust. 2.8 Over the 1990-95 period, the growth rate was more than 30 percent higher than Sri Lanka's long-term trend, and only slightly lower than that realized for a comparable time period after the 1977 liberalization (see Annex III for details). Furthermore, it is a rate that compares quite favorably with those of other adjusting countries. Sri Lanka's longer- and medium-term growth rates are higher than those realized by other South Asian adjusters, and higher than those realized by all except East Asian adjusters. Its performance relative to transition economies is remarkable. It is the only such economy not to experience negative growth in even one year, and its inflation and deficit experiences are significantly stronger than those realized in the strongest performers among the transition economies. 2.9 Extending the perspective to 1995, however, reveals that most of the gains on deficit reduction have been lost. In May 1993 the country was shocked by the assassinations of a number of prominent figures, including President Premadasa. Elections of a new president in 1994 and a new parliament in 1995 also diverted attention from economic management, and reduced Sri Lankan ownership of the program. The result was that the budget deficit returned to the levels at the beginning of the adjustment. Fiscal Measures 2.10 Revenues. Moderately high rates of inflation persisted and high real borrowing costs contributed to revenue problems. As a result of these factors, the cost of capital to the private sector was very high. Not surprisingly in an economy where the export sector had been given priority, and the private sector was designated as the "engine of growth," these kind of costs created pressures for concessions. In the event, the Government offered tax incentives in the form of tax holidays, thereby reducing overall taxes from 19 percent of GDP in 1990 to 17.2 14 Chandra and Anees (1996), OED, Table 16. 28 percent in 1994.15 This reduction contrasts with the slight increase in revenues of .2 percent of GDP by other fiscally adjusting economies. Moreover, those economies started adjustment with revenues of 22.8 percent of GDP, 17 percent higher than Sri Lanka's.16 2.11 These ad hoc tax measures distorted the overall incentive framework, by encouraging rent-seeking activities. They also had adverse effects on aggregate balances. Given the size of the budget deficit, and the reductions in revenues from import and export duties (together these duties declined from 31 percent of revenue in 1989 to 21 percent in 1993), revenue losses had to be made up on other fronts. Stop-gap measures were introduced-principally through increased excise duties, an income tax surcharge, a business turnover tax, and a defense levy. However, the stop-gaps were not sufficient to offset an overall decline in revenues as a share of GDP. Given this result, Sri Lanka's ability to successfully stabilize was problematic. None of the successful stabilizers considered by Chandra and Anees (1996) or Gupta (1996) did so with a reduction in government revenues. 2.12 Little progress was made on implementing the recommendations of the Tax Commission as proposed by ERC, even though the capital gains tax and the stamp duty were eliminated. Nor was progress made on addressing the main source of the high cost of capital-the high margins that are charged on financial intermediation-6 percent in 1994-due to the dominance of the two large and highly inefficient state commercial banks. In many respects, these high costs of intermediation represent an imposition of two taxes on the private sector through the financial sector. First, the costs of the inefficient bank operations are imposed on savers, through a less competitive and less innovative mobilization system. Second, borrowers pay for these inefficiencies through higher rates. Thus, these implicit taxes, through their effects on saving and investment decisions, are likely to have significant macro consequences. 2.13 Expenditures. With continued war in the north and east, defense expenditure rose from 9 percent of expenditures in 1990 to 15 percent in 1994, and together with the cost of rehabilitating refugees, made significant demands on the government budget. Besides higher defense expenditures, an increase in transfers to public corporations between 1990 and 1994 also placed upward pressure on expenditures. Finally, reductions in civil service salaries were not realized, price subsidies-eliminated as part of the credit conditionality on rice, wheat, and fertilizer-were reintroduced, and subsidy expenditures increased. Stabilization under such circumstances was problematic. Barbone and Polackova (1996) empirically examine public finance revenue and expenditure patterns for 47 economies for 1992. Standardizing for per capita income levels they estimate the tax effort that corresponds to the country's debt level, a measure of educational level, and an economy's openness. If Sri Lanka is analyzed with the coefficients of their model, its government revenue efforts are low. They are less than the tax effort, 18.3 percent of GDP, predicted by the model. and while the level of effort 17.2/18.3= 94 percent, as well as the level itself is relatively high. it is lower than 19 of the 26 transition countries they considered. In addition, of the 27 fiscally-adjusting countries considered by Gupta (1996), only 2 reduced revenues to lower levels than Sri Lanka. In neither case was growth sustained after adjustment. 16 Gupta (1996) 29 Civil Service Reform 2.14 In Sri Lanka, reform of the civil service lies at the heart of building an effective governance structure. The reforms proposed by the ARC commission, and covenanted by ERC, sought to make the civil service: (i) a professional organization insulated from narrow political concerns; (ii) able to support both rapid private market development and intervene efficiently where markets fail to function effectively; and (iii) considerably smaller. Both the Government and Bank performance on these reforms, which were introduced with enthusiasm by the Presidential Commission, must be considered unsatisfactory. 2.15 To launch the program, the government introduced a generous voluntary separation package with a target of 60,000 departures over three years. The separation package was not withdrawn at the end of the reform, but left available to all retiring civil servants. Moreover, the government hired more new civil servants to replace the retiring personnel and increased their wages by 30 percent. The results were: (i) that by 1992 there were almost 10 percent more civil servants than there were in 1989; (ii) many of those who would have retired anyhow, also received separation packages; (iii) nothing was done to implement an objective, professional selection system; and (iv) the number of ministries did not decline, as recommended by the ARC. Indeed, they have expanded to 89 ministries from 46 in 1989.17 2.16 It is hard to imagine a less satisfactory outcome. There were no savings for salaries foregone and pensions increased three-fold. The results are even worse when two factors are noted. First, the same type of separation and rehiring of those separated occurred repeatedly in the transport sector in Sri Lanka in the 1980s, and the World Bank was researching this issue while the credit was under design. Second, perhaps the foremost recommendation of the ARC was that "the implementation of bits and pieces of the suggested agenda could be dangerous in that it could lead to incoherent and chaotic results which would put the administrative system back rather than propel it forward." The ARC prediction could hardly have been more prescient. Reform was implemented on a piecemeal basis, with results so adverse that the direct budgetary costs for the separation payments alone cause the deficit to exceed PFP targets.'9 Poverty Alleviation 2.17 In the wake of the 1987-89 insurgency in the south, rural poverty became the motivating force for a major new poverty alleviation program, the Janasaviya Program, established in 1989. Other measures to alleviate poverty-such as the mid-day meal program, free school uniform program, and various general commodity subsidies for fertilizer and flour and bread, and the food stamp program were also designed to subsidize the costs of many essential goods. Moreover, because of growing regional disparities in industrial development, a program to 17 The results are from Wanasinghe (1994). 18 See Report No I of the Administrative Reforms Committee, cited in Wanasinghe (1994). 19 See the appendix for the assumptions and method of calculating the costs of the separation expenditures. It modifies an approach used by Terrell (1991). 30 establish 200 garment factories in rural areas began in late 1991.20 As noted earlier, the new poverty program was extremely overambitious and was implemented in "waves" to make its costs more manageable. 2.18 Welfare subsidies rose. Having fallen from 21 percent of government current expenditure in 1979 to 4.5 percent in 1988, they surged up again to 12 percent in 1990 and are at that level, which is equal to 3.7 percent of GDP in 1995. ERC's approach to poverty alleviation stressed both the sustainability of the level of expenditures, and concern with the effects that price liberalizations would have on the position of low income households. The credit's covenants called for the elimination of price controls on wheat, fertilizer, rice, and privatization of bus transportation, which implied an increase in transport costs. The objective was to improve the targeting of assistance to those most in need at a time of price increases and simultaneously to keep the total level of expenditures less than 3.5 percent of GDP. 2.19 Government performance was unsatisfactory. Decreases in subsidy expenditures came through inflationary erosion of benefits to the poor rather than elimination from program eligibility of those who were not poor. Targeting of JSP did improve, but was reversed after the second tranche release. Moreover, the design of the program was never very well thought out. Eventually, in 1994, it was replaced by a new subsidy program, one that has all the hallmarks of a highly politicized distribution of benefits.21 Regressively allocated price subsidies were also introduced for wheat, and many of the other price controls were also reintroduced after the second tranche release. Creating a More Competitive, Transparent Economic Environment 2.20 Tariffs and prices. Strong export performance continued throughout the adjustment period. In 1994 exports accounted for 27 percent of GDP, up from 21 percent in 1988. Similarly, the tariffs on imports declined as a share of GDP even though the share of imports, particularly of investment goods, rose. As a result, effective import duties were cut in half falling from an average implicit tax rate of 17 percent to 9.5 percent. The tariffs have not only been lowered and removed, they have also been simplified as well. Nevertheless, in some instances, specific duties have been maintained which are equivalent to very high rates, and quantitative restrictions for some key agricultural goods remain, even if they are periodically lifted. The deregulation of freight services has continued to result in reductions in freight charges and increases in competition. 2.21 On domestic prices, a greater reliance on market forces was considerably less successful. For instance, public buses were "peoplized," in a privatization-like program that turned the buses and half the equity of the companies over to employees. This process was to lead to complete divestiture by 1991. However, in the event divestiture did not occur: fares continued to be controlled by the Government and have not been increased, and salaries were raised by the Government by 30 percent. The buses, as a result, still operate at a loss. Similarly, rail transport subsidies, which were also targeted by ERC, remained in 1994. Fertilizer and rice subsidies were also restored. In general. then, although restrictions on international prices have been 20 The program entitled companies to special export concessions and reflected little concern with the economic viability of specific locations. 21 See the discussion of the new subsidy program. the Samurdhi Program, in Institute for Policy Studies (1995). 31 greatly reduced, with highly beneficial effects on the economy, the same is not true about some important domestic prices. ERC Summary 2.22 To sum up, while Sri Lanka has made remarkable progress on its adjustment and reform agenda, it has neither stabilized nor has it established the fundamentals of a professional public service that is free from political manipulation. It appears to be unable to implement programs based on the effective reports of Commissions. All too often public policy appears to operate in a defensive mode, attempting to scale back and rationalize proposals of political parties for large-scale public sector programs, the most recent of which is the Samurdhi poverty program. These proposals go forward while commission reports are ignored. For Sri Lanka to continue to experience growth with equity it must once again address the stabilization problems that it has twice failed to resolve, most recently with ERC. Because control over the efficacy of government expenditures has played such an important role in Sri Lanka's failure to stabilize, a key element of a successful stabilization will necessarily be more effective management of public expenditures. This effective management can only be realized by a professional civil service. Therefore, civil service reform remains at the center of Sri Lanka's stabilization agenda. C. PMEAC: Objectives and Policy Instruments 2.23 PMEAC represented a dramatic reversal of long-standing Sri Lankan policy on government ownership and management of manufacturing and other enterprises. It was designed shortly after (and as a complement to) ERC, to support the implementation of a program of comprehensive reform of public manufacturing enterprises. It was based on the recommendations contained in the Public Manufacturing Enterprise Sensitivity and Implementation Framework Studies carried out with IDA financing by the Presidential Commission for Privatization. This program called for making changes in the regulatory and institutional arrangements within which PMEs operate, and taking steps to adjust PMEs to a more commercial and competitive environment. For PMEAC, this meant the following specific objectives: (a) Remove the remaining specific tariff protections, market privileges and fiscal support for PMEs. The effect of these changes would be to increase the competitive pressures on PMEs. (b) Provide a new institutional framework for PMEs. For those PMEs converted to company status, but in which the Government would continue to hold a majority interest, this framework was expected to induce adoption of an autonomous commercial operating mode, increase accountability for performance, and promote flexible responses to increased competition. These changes were designed to maximize returns on the Government's investment in the PMEs. (c) Convert PMEs into commercial companies and privatize them where possible. This would be done through changing the legal status of PMEs, reconfiguring their financial structure, privatizing PMEs where possible, and liquidating others through joint ventures with foreign owners and domestic sales of stock. 32 These changes would remove government regulations over labor, wages, sourcing of inputs, and pricing of outputs. (d) Phase out non-viable PME activities altogether. 2.24 Thus, PMEAC was designed to support the ERC's strategy of economic growth, budget control and reduced Government presence in the Sri Lankan economy. It sought to do this by (i) moving selected SOEs from public to private ownership, and (ii) subjecting the remaining SOEs a more competitive and commercial operating environment. Specifically, it targeted 10 SOEs controlled by the Ministry of Industries, Science and Technology (MOI) and four textile mills under the Ministry of Textiles for privatization, and supported regulatory, incentive and trade law reforms designed to make all PMEs more competitive, commercial entities. It also financed strategic studies of several industries, and the establishment of the Public Investment Management Board (PIMB), under the Ministry of Finance, to manage the Government's remaining investments in converted and privatized companies and to assist in the restructuring of companies not immediately privatized. PMEAC: Implementation Experience and Outcome 2.25 PMEAC was more successful in meeting its narrower objectives than was ERC. Market privileges and fiscal support were either eliminated or substantially reduced. The targeted SOEs were converted to commercial companies, and most were privatized, as planned. However, the extent to which privatization has led to improved economic efficiency is less clear, especially in light of Sri Lanka's failure to fully implement structural reforms under the ERC, as discussed above. Moreover, the process has not been fully transparent, leading to perceptions of inequity, and contributing to a slowdown in privatization activities. Finally, the PIMB did not work well. As a company registered under the Companies Ordinance-to protect it from political pressures-but functioning directly under the Ministry of Finance, its relationship with the line ministries that actually controlled the PMEs was unclear. Changes in Tariffs, Market Privileges and Fiscal Support 2.26 A major goal of the PMEAC was to force SOEs to behave more like commercial firms by removing specific tariff protections, market privileges and fiscal support for the companies. To a large extent, this was accomplished through ERC. The tariff structure had largely been rationalized into three broad bands by 1995, with a maximum rate of 35 percent, and most specific duties and quantitative restrictions on imports had been removed (with some exceptions, mostly in agriculture). 2.27 On the budget front, PMEAC was less successful. Shortly before implementation of the project, total government transfers to public corporations fell substantially, as shown in the following table. This reflected a large decline in capital expenditures. However, these transfers generally rose throughout the period of the project, and were expected to nearly double between 1994 and 1995, exceeding the level of 1988, despite the substantial privatization of public corporations. 33 Table 2.2: Transfers to Public Corporations, 1988-95 Transfers 1988 1989 1990 1991 1992 1993 1994 1995 (prov.) (est.) Current 1,725 N.A. 2,639 1,743 1,428 1,571 2,634 7,757 Capital 9,683 N.A. 1,249 2,993 4,610 4,311 4,246 5,230 Total 11,408 N.A. 3,888 4,736 6,038 5,882 6,880 12,987 Sources: For 1988, Athukorala and Jayasuriya (1994), p. 144; other years from the Central Bank of Sri Lanka Annual Report 1994, p. 131. Extent of Privatization 2.28 The targeted privatization actions supported by PMEAC grew out of the Government of Sri Lanka's broader efforts to convert large numbers of SOEs to a commercial mode of operation or to outright private ownership. ERC, for example, was designed in part to support (i) improved productivity in the tree crop industry, in part through private management, and (ii) partial privatization of the bus transportation industry. Moreover, other privatizations occurred in addition to those targeted in these two credits. Thus, privatization under PMEAC must be considered in this broader context. 2.29 Privatization under PMEAC. For the most part, PMEAC achieved its privatization objectives, which were the major focus of the project. The four textile mills were fully divested. Each of the ten MOI corporations and two Government-Owned Business Units (GOBUs) targeted by the credit was converted to a company by 1992. Subsequently, eight of these companies were privatized, in whole or substantial part, and a ninth was closed down and its assets sold. Concerning the remaining three companies, security considerations have been cited for the failure to divest two, and the third is in receivership, with the Treasury paying employee's salaries.22 2.30 Of the eight completed MOI privatizations, five involved 100 percent divestiture of the Government's ownership position (in leather products, tires and hardware, oxygen and motor car trading). In a sixth instance (ceramics), 73 percent of the company was divested. The remaining two companies were restructured with the hope of eventually privatizing them. One, the National Salt Company, was reorganized into a holding company (Lanka Salt, Ltd.) and a subsidiary (Puttalam Salt, Ltd.); while the Government retains ownership of the holding company, 60 percent of Puttalam Salt was to be sold through a public share offering, which has not yet materialized. The other, the State Mining and Mineral Development Corp., was broken up into three companies. The smallest, Kahatagaha Graphite Lanka, Ltd., (representing Rs. 13 million in capitalization) was entirely privatized, but the Government retained ownership of 50 percent of the shares in the much larger Bogala Graphite Lanka, Ltd. (capitalized at Rs. 280 million). Negotiations for the sale of the third component, Lanka Phosphate, Ltd. (with over Rs. 72 million in equity capital) have been taking place for a long time with a foreign investor, but had not been finalized at the time of this audit. 22 Hulugalle (1994), Annexes I and II. The author provided data updated to early 1996. 34 2.31 Privatization under the ERC. ERC's privatization component was somewhat less successful. Some privatization did take place in both bus transport and tree crops, but the results have not been as beneficial as anticipated. 2.32 The bus systems operated by the Central Transport Board (CTB) were broken up into 104 depots and 13 workshops by 1991. As part of the privatization drive, 95 bus companies and a tire rebuilding factory were established by 1993, with employees receiving 50 percent of the stock in each, and the remainder held in trust by commercial banks on behalf of the Government for 3 to 5 years. During the same period, 5 of the 13 workshops also were privatized, with 60 percent of the stock in each sold on the market and the remainder provided to employees. However, the Government has retained a major role in the transport industry. The CTB turned over ownership of bus depot land and buildings to local transport boards, which then leased them to the operating companies. In addition, the Government has continued to intervene in wage- and fare-setting, making it difficult for the bus companies to generate sufficient income to ensure adequate maintenance of buses. An evaluation of the ERC and PMEAC by the cofinancier-the Overseas Economic Cooperation Fund (OECF) of Japan-found some improvements, including an increase in the number and reliability of buses and higher sales volume (per mile and per employee). However, it also found continuing problems, notably continuing operating losses (partly attributable to Government imposed 30 percent wage increase) and a reduction in some critical services (such as late-night and early-morning service in the Colombo suburbs). 2.33 The tree crop industries-tea, rubber and coconut-also took the first steps toward privatization. In 1992, over 400 state-owned estates were consolidated into 22 regional plantation companies with 99-year leases on the land. Management of these companies was contracted out to private firms on a profit-sharing basis for four and one-half years. The OECF evaluation found some improvements in management and productivity. But this arrangement has not been fully successful because of two major problems. First, the management contracts are for only a few years, while the SOEs continue to own the land. Second, the Government continues to control wages and terms of employment. Thus, the management companies are discouraged from taking steps to improve productivity through long-term investments or changes 23 in labor pricing or practices. 2.34 Privatization in the broader context. Despite the problems described above, Sri Lanka has taken dramatic steps toward privatization. In the late 1980s there were nearly 100 SOEs, but by 1994 at least 41 of these had been divested fully or substantially, and many more were in the process of being privatized. Although, some of the largest SOEs were not yet slated for privatization, especially in the areas of finance (e.g., Bank of Ceylon), transportation (e.g., Air Lanka), and energy (e.g., Ceylon Petroleum), the experience in other privatizing countries shows it is often a sensible strategy to privatize small and medium sized enterprises first to gain 24 experience and establish procedures before tackling the larger, more complicated enterprises. Moreover, in the context of worldwide experience, Sri Lanka's privatization efforts have been 23 World Bank (1994a). 24 Kikeri et al. (1992), p. 54. 35 notable. For example, for the entire period 1980-91, only 122 SOEs had been privatized in all of Asia.25 2.35 The election of a new government in 1995 with strong union backing appears to have slowed the speed of privatization. This reflects both union resistance to privatization and a deliberate strategy of dealing with the ethnic conflict as a first priority. Moreover, the Government has tried to prevent privatization from reducing overall employment in the affected companies, not only as part of the process of preparing them for sale, but also after the new owners have taken control. Nevertheless, the current Government has reiterated its commitment to the privatization program it inherited, and a new privatization commission was appointed in 1995. Significantly, there has been no re-nationalization of privatized companies, nor is any being contemplated. Improved Economic Efficiency 2.36 One concern about Sri Lanka's privatization program has been that it has focused less on increasing economic efficiency than on addressing government budget problems. The previous government initiated the program in large measure as a way of reducing government expenditures to support SOEs. The new government has continued to pursue it-if more slowly-giving more emphasis to raising revenue through sales of government assets to help balance the budget, a strategy pursued by a number of developed countries, such as the United States, in recent years. Nonetheless, the primary benefits of privatization should accrue through improved economic performance, not better fiscal aggregates. 2.37 Data on the economic efficiency of Sri Lanka's privatization efforts are thin. One study was limited to the consideration only of the sales and profitability of privatized firms, not broader indices of economic efficiency.26 Moreover, only 4 of the II firms it examined were targeted under PMEAC. The authors compared the performance of firms before and after privatization and found mixed results for changes in profits and sales. Among the five firms for which sales per employee could be computed, 3 showed substantial gains, and none exhibited declines. 2.38 The most recent data on economic performance are from a survey of 15 companies that had been privatized for between 12 and 30 months by 1992, again including only four of the firms that were part of PMEAC. Both for the full sample and for four PMEAC firms there was a notable increase, on average, in the value of production (25.4 percent) and sales (48.1 percent) after privatization. And while these varied considerably among firms, only one reported a decline (of 5 percent) in production value, and none in the value of sales. By contrast, the value of industrial output among all SOEs was virtually unchanged between 1990 and 1992, 25 Kikeri et al. (1992), pp. 22-23. Of course, Sri Lanka's efforts are dwarfed by those in the former German Democratic Republic, where some 4500 firms were privatized within 18 months in the early 1990s. However, the latter privatization program was undertaken within the context of the highly developed German economy, a luxury unavailable to Sri Lanka. 26 White and Kelegama (1994). 27 Hulugalle (1994), Annex VIII. 36 essentially the period covered by this study.28 Employment declines in the privatized firms were generally modest, with major losses in some companies offset by corresponding gains in others. Overall, employment in these 15 firms fell by a total of 580, or an average of 1.1 percent for each firm.29 Transparency of the Process and Public Reactions 2.39 The implementation of the privatization program has raised questions of transparency and equity that have contributed to a slowdown in the process. This problem arose in part because of the lack of a well functioning equities market in Sri Lanka; the Colombo Stock Exchange (CSE), opened in 1985, listed only 200 companies at the end of 1993, and equity and debt financing combined accounted for only 20 percent of estimated private investment expenditure in 1992.30 As a result, most privatizations have involved sales through tenders or negotiated private offerings rather than unrestricted public offerings in the equities market.31 The explicit tradeoff here has been between moving rapidly before opponents of privatization could mobilize and accepting some reduction in transparency on the one hand, or waiting for the development of stronger capital markets and risking a loss of momentum, on the other. In effect, the Government chose the former option. 2.40 Not surprisingly, then, privatization through tenders and negotiated sales has lead to widespread charges that valuable companies were sold for less than market value. In some cases, the subsequent performance of shares on the stock exchange has reinforced the perception that companies were undervalued at sale. For example, Ceylon Oxygen was sold at Rs. 15 per share, but opened on the CSE at Rs. 87.50, and United Motors was sold for Rs. 10 per share, but 32 within 18 months was listed at Rs. 29 on the CSE. The problems with the valuation of SOEs for sale in part reflects the weakness of the valuation system. The Valuation Department, traditionally involved in government procurement, tended to consider only tangible assets such as land and structures when advising the Government on the value SOEs, not their economic potential. Other developing countries pursuing privatization programs-including Guinea, Mexico and Pakistan-also have experienced criticisms of the transparency of the process, which reduced public support for privatization and led to a slowdown of their efforts.33 2.41 The tensions raised by perceptions of undervaluation have been exacerbated by concerns over the equity of the privatization program. In part to counter the concern that selling of national assets to private (especially foreign) owners was tantamount to "giving away" national assets at bargain-basement prices, the Government adopted a policy of providing shares to employees of privatized SOEs free of charge. In most cases, this amounted to 10 percent of total 28 World Bank (1995). 29 Hulugalle (1994), Annex VIII. 30 World Bank (1995). 31 Hulugalle (1994), Annex I. 32 Kelegama (1994), pp. 145-148. Kikeri et al. (1992), pp. 70-71. 37 equity, but in bus transport it was 50 percent. To some extent, this policy of distributing stock to employees succeeded in tempering union opposition to privatization. 2.42 However, it also created a number of equity problems. First, relatively few workers benefited from the transfer of stock. By mid-1993 a total of 74,000 persons in privatized companies were covered by these stock transfers. 34 But they represented less than 20 percent of all workers in SOEs at the start of the privatization process in 1989, and only 1 percent of the total work force. Second, even among the covered workers there were wide disparities in the value of the stock they were given; those in highly profitable companies, such as the Distilleries Company, found themselves beneficiaries of a relative windfall, while those in companies with heavy losses, such as the bus depots, were holders of relatively valueless stock. These equity issues appear to have reduced public support for the program. Institutional Development 2.43 The major institutional component financed by PMEAC was the strengthening of the PIMB. It had been established in 1989 to manage the Government's remaining holdings in converted or privatized SOEs, and was expected to provide the technical expertise in financial restructuring and privatization that the line ministries lacked. As an incorporated company, accountable to the Ministry of Finance, it was in a position to hold shares directly and enter into legal agreements; this was expected to allow PIMB to take a leading role in the privatization of PMEs and to apply commercial standards to the operation of those remaining under Government ownership. 2.44 The PIMB was not as successful as other components supported by PMEAC. It had a legally ambiguous role that limited its effectiveness. The line ministries retained control over the privatization process and continued to operate the remaining SOEs. Moreover, the PIMB's expected private-sector orientation was weakened by the appointment of senior staff from the line ministries as Board members. By 1994 the PIMB had been disbanded.35 Compliance with Credit Covenants 2.45 PMEAC included 10 specific covenants, 8 of which specified actions to be taken before release of the second tranche. All of the covenants were complied with, except one. That called for adoption of a strategy to restructure and modernize the cement industry. However, two of the three cement plants were privatized by the end of 1992, and the remaining one was in the conflict zone. Under the circumstances, this covenant became superfluous and IDA agreed it need not be carried out. 2.46 Four covenants were complied with only after delays ranging up to 13 months. These were major covenants, including conversion of MOI enterprises into companies, divestiture of at least half the asset value of Treasury-held shares in the mining and mineral sector and in other PMEs, and adoption of a strategy to restructure and privatize salt operations. However, despite these delays, the project was completed on time. Kelegama (1994), pp. 149-52. Hulugalle (1994), pp. 18-19.  39 3. Lessons and Sustainability 3.1 Sri Lanka has made remarkable progress on its adjustment, and it has done this in a very difficult environment. The economy has opened up, and achieved high rates of growth. It is subject to more competitive pressures, and governed by private sector investment decisions. However, stabilization was never achieved and it appears now that it is being reversed. For Sri Lanka to achieve a durable sustained adjustment, there are a number of lessons than can be drawn from this adjustment experience. Many of these lessons apply with even greater force to the Bank's support for such adjustment programs. A. Stabilization and Governance 3.2 Sri Lanka has not established the fundamentals of a professional public service that is free from political manipulation. Its ability to implement programs based on the effective reports of Commissions established by the Government is lacking. All too often policy appears to operate in a defensive mode, attempting to scale back and rationalize proposals for large-scale public sector programs. For Sri Lanka to continue to experience growth with equity it must once again address the stabilization problems that it has twice failed to resolve. Because control over the efficacy of government expenditures has played such an important role in Sri Lanka's failure to stabilize, a key element of a successful stabilization will be more effective management of public expenditures. This effective management requires a sound, professional civil service. Consequently, civil service reform remains at the center of Sri Lanka's ability to sustain the adjustment program. 3.3 Without improved governance abilities for the public sector, failure in sustaining the framework of a market-based economy is almost certain. Importantly, while both ERC, and the Administrative Commission on which it based its recommendations, emphasized that gains were possible from reductions in staff, they also stressed the need for increased professionalism as a distinct issue. This is an important distinction because Sri Lanka's total expenditures on civil service wages and salaries, 5.1 percent of GDP, are not usually high relative to those of other adjusting economies.36 Thus, in many respects, the most pressing problem with the civil service is the quality rather than the quantity of the expenditures. The main lesson is that civil service reform should be seen, as the ARC stressed, as a comprehensive package rather than as a mechanical exercise driven by the number of separations. B. Financial Sector Development and Adjustment 3.4 One factor that appears to have played an important role in Sri Lanka's success is the liberalization and corresponding deepening of the financial sector that took place over the 1980s. No other adjusting country with such an extensive government role in the economy had the advantage of such a system. Through a series of eight financial intermediary loans, beginning in 1979, the World Bank engaged the Government in a number of studies and policy reforms that had a beneficial effect on the development of this sector as well as on trade policy. The policy reforms that accompanied the credits helped to establish positive real interest rates, an auction market for Treasury securities, the ultimate privatization of the two leading Development 36 Gupta (1995). 40 Finance Institutions, and the establishment of a credit rating agency. They also helped identify the real costs and distortions of the import tariff system. 3.5 The credits also produced studies that identified the problems associated with: (i) the functioning of the two large state commercial banks, and (ii) the trade policy regime. The former study detailed the adverse effects that the high interest rate margins of these lenders had on the cost of borrowing-real rates were 12 percent in 1993-and the efficiency of the financial sector. The latter study provided details on the perverse effects of the tariff regime, as well as ways to simplify and reduce these barriers. As a result, while at the time of adjustment there were still problems in the financial sector and tariff structure, there was also real progress. The financial sector's funding for private sector borrowers grew from less than 50 percent of the domestic credit provided by the financial sector in 1979-a sector that had been offering negative interest rates and stagnating for a number of years-to more than 70 percent of the domestic credit provided by a larger sector in 1993. In addition, the study of tariff structure helped in developing adjustment loan conditionalities. 3.6 Thus, Sri Lanka's growing financial sector was able to mobilize resources and allocate them among private demanders of credit. Once the tariff reforms and privatization measures of the adjustment program were in place this facilitated adjustment. Recent World Bank research confirms this view. It suggests that a deeper financial sector acts like a "shock absorber" and permits countries undergoing adjustment to weather the price changes more resiliently. When such a system is in place, lenders discriminate between firms with liquidity problems (due to the new price environment) and those with solvency problems. Without such an ability to discriminate insolvent firms continue to receive support long after they should. Adjusting countries with greater financial depth grew considerably more rapidly, by 1.8 percent on aper capita basis, after adjustment, than did countries with less depth, which experienced negative growth. Indeed, this single indicator alone is a good guide to the success or failure of adjustment reforms.37 3.7 However, just as the improvements in the financial sector almost certainly facilitated the adjustment process, so too do the remaining financial sector distortions constrain it. A major concern of World Bank financial sector work in Sri Lanka has focused on the costs of the inefficiencies of the two large state commercial banks. The high interest rate margins that these institutions charge are, as a recent survey of Sri Lankan entrepreneurs shows, a severe impediment to investment. The resulting high cost of borrowing also creates rationales for the Government to engage in the kind of ad hoc tax relief measures of the sort described earlier. One of the key lessons of successful adjusters is that this kind of reduction in the government's revenue base will make adjustment very difficult. 3.8 Thus, there are two types of lessons on financial sector policy. First, the productivity of the studies and reforms undertaken as part of the series of financial intermediary loans was almost certainly extremely high. These policies contributed to the increase in financial depth that occurred during the 1980s, and had important contributory roles in Sri Lanka's high post- adjustment growth rate. The second lesson is less positive. The failure to complete the restructuring of the state commercial banks can be expected to remain a significant impediment See King and Levine (1993). 38 See Coopers and Lybrand (1994). 41 to the functioning of a more effective, liberalized economy. Sri Lanka's saving rate is far too low-less than half that of comparator countries-and the Government's presence in the credit markets far too high for the country to be able to afford the costs of inefficiencies of these institutions. C. The Political Economy of Hostilities and Aid Disbursement 3.9 The failure of civil service reform is indicative of the difficulties of coherent government policy in an environment characterized by hostilities. When the expenditures on the severance payments to induce separation were almost immediately offset by new hiring, it is clear that policy directions changed quite dramatically and quickly. Almost certainly the new policies- which undermined the existing policies-were in response to other, more pressing, events. Study after study shows that many of the policy choices made were the result of the exigencies of the conflicts. Until the current conflict is resolved, the policy horizon will almost certainly remain extremely near-sighted and immediate. 3.10 For the Bank, perhaps the major lesson is that the adjustment program disbursed too much money over too short a time period. The reform targets were appropriate and, for the most part, the strategy was thoughtful. However, in light of the imbalances generated by the ongoing hostilities and the loss of program ownership as time passed, it was a mistake to disburse the large amounts of these credits, more than 4.5 percent of GDP, in such a short time period. While this conclusion is easier to make with the benefit of hindsight, particularly in light of the much longer adjustment periods for reforming socialist economies, it should not have been difficult to make at the time. The IMF's SAF program had collapsed one year prior to credit approval. In addition, earlier adjustment lending discussions between the Bank and Sri Lanka focused on disbursing smaller amounts of assistance over a much longer time period. 3.11 The main lesson, for the Bank, is that while policy-oriented lending can be very productive-as shown by the financial sector lending-the timing and targets of this support are crucial. The timing of aid disbursement in a country as aid-dependent as Sri Lanka-it receives the second largest amount of aid as a share of GDP of non-African countries-is particularly important. Better coordination among donors is essential if aid is to be used effectively. For example, perhaps the main reason that the Sri Lankans rejected the Bank's 1983-84 proposals for a slower disbursing and smaller adjustment program was the availability of other assistance. D. Privatization for Efficiency Gains Rather Than for Budgetary Purposes 3.12 The sustainability of the specific privatizations undertaken through the PMEAC is likely. However, the future of Sri Lanka's relatively successful privatization program is threatened by an emphasis on the effects that divestiture has on the government budget rather than its effect on the efficiency with which resources are used.40 The poor growth performance of economies with public control over production is well know.41 Sri Lanka is no exception to this trend. Its public sector industries have been unable to respond to market incentives. Price increases have not See Dunham and Kelegama (1995), Terrell (1991), Athukorala and Jayasuriya (1994). 40 Kikeri et al. (1992), pp. 6-7. 41 Easterly and Fischer (1995). 42 elicited corresponding supply increases, and price decreases have resulted in losses for the Government. The productivity of the private sector in Sri Lanka has been, and continues to be, considerably higher than that of the public sector. Thus, the motivation for privatization is to shift resources to a more effective governance structure. The budgetary gains, or losses, associated with privatization are of secondary importance.42 3.13 It is also important to improve the perceived fairness of the privatization process. Providing some workers shares in essentially worthless enterprises while others reap windfall profits as the worth of their stock in undervalued companies soars on equities markets is inequitable. It is also increasingly unsustainable as most profitable companies have already been privatized. A mechanism that permitted broader sharing of the gains could address this issue and increase support for privatization. 42 For discussion of criteria for privatization, see World Bank (1994b). 43 Bibliography Athukorala, Premachandra, and Sisira Jayasuriya. (1994). Macroeconomic Policies, Crises, and Growth in Sri Lanka, 1969-90. Washington, D.C.: World Bank. Barbone, Luca, and Hana Polackova. (1996). "Public Finances and Economic Transition." Policy Research Working Paper no. 1585. Washington, D.C.: World Bank. Bruton, Henry J. (1992). "Sri Lanka and Malaysia." The Political Economy ofPoverty, Equity, and Growth. A World Bank Comparative Study. New York, NY: Oxford UP. Central American Bank for Economic Integration. XX0VAnnual Report 1994/95 Fiscal Year Central Bank of Sri Lanka. (1988). Annual Report 1987. Colombo, Sri Lanka: Central Bank. ___. (1995). Annual Report 1994. Colombo, Sri Lanka: Central Bank. Chandra, Vandana, and Salman Anees. (1996). "Fiscal Adjustment under SALs and SECALs: A Macroeconomic Overview." Background Paper for OED's study 'Fiscal Management under SALs and SECALs,' March. Coopers & Lybrand. (1992). "International Financial Audit of Bank of Ceylon for the year ended 31 December 1991. Final Audit Report." . (1992). "International Financial Audit of People's Bank for the year ended 31 December 1991. Final Audit Report." Coopers & Lybrand Deloitte. (1990). "Sri Lanka. Economic Restructuring Credit. Administrative Reform Programme." Appraisal Report, London, February. De Long, J. Bradford, and Lawrence H. Summers. (1993). "How strongly do developing economies benefit from equipment investment? Journal of Monetary Economics 32(3): 395-415. Dunham, David, and Saman Kelegama. (1995). Economic Reform and Governance: the second wave of liberalisation in Sri Lanka 1989-93. Research Studies Governance Series No. 2. Colombo, Sri Lanka: Institute of Policy Studies. . (1994). Stabilization and Liberalization: a closer look at the Sri Lankan experience 1977-93. Colombo, Sri Lanka: Institute of Policy Studies. Econsult. Annual Report 1995 - Monetarism With A Human Face. Colombo, Sri Lanka: Econsult. Fiszbein, Ariel. (1992). "Labor Entrenchment and Redundancy Compensation in State Owned Enterprises: The Case of Sri Lanka. World Bank SASVP Internal Discussion Paper, Report no. IDP- 121, December. 44 Gupta, Poonam. (1996). An Analysis of Revenue Versus Expenditure Focus in Bank-supported Lending. Draft paper, February. Hulugalle, Sriyani. (1994). Privatization Program of Public Sector Commercial Enterprises: Sri Lankan Experience. Colombo, Sri Lanka: World Bank (unpublished). Institute of Policy Studies. (1995). Sri Lanka: State of the Economy, 1995. Colombo, Sri Lanka: Institute of Policy Studies. Karunaratne, S.A. (1991). A Macroeconomic Overview of Sri Lanka 1970-87. With Special Reference to Stabilisation Issues. Research Studies Macro-Economic Series No. 5. Colombo, Sri Lanka: Institute of Policy Studies. Kelegama, S. (1992). "Effects of Present Public Sector Monopoly or Protected Activities on Private Sector Development" and "Progress, Prospects, and Problems in the Privatization Process and its Impact on Private Sector." Paper prepared for the Joint IDA/GOSL Private Sector Assessment (PSA) Study, February. Kelegama, Saman. (1994). "The Impact of Privatization on Distributional Equity." In V. V. Ramanadham (ed.) Kikeri, Sunita, et al. (1992). Privatization: The Lessons ofExperience. Washington, D.C.: World Bank. Knight, Malcolm, et al. (1996). "The Peace Dividend. Military Spending Cuts and Economic Growth." Policy Research Working Paper No. 1577. Washington, D.C.: World Bank. Lal, Deepak, and Sarath Rajapatirana. (1989). "Impediments to Trade Liberalization in Sri Lanka." Trade Policy Research Centre, London: Gower. Little, I.M.D., et al. (1993). Boom, Crisis, and Adjustment. The Macroeconomic Experience of Developing Countries. New York, NY: Oxford UP. Operations Evaluation Department. (1987). Sri Lanka and the World Bank. A Review of a Relationship. Washington, D.C.: World Bank. Prywes, Menahem. (1995). "Unemployment in Sri Lanka: Sources and Solutions." World Bank South Asia Region Internal Discussion Paper, Report no. IDP-154. Ramey, Garey, and Valerie A. Ramey. (1995). "Cross-Country Evidence on the Link Between Volatility and Growth." The American Economic Review 85(5):1138-1151. Terrell, Kathering. (1991). "Labor Redundancy in the Transport Sector: The Case of Sri Lanka." Draft paper part of a series being carried out under th eWB Research Committee project no. 675-2 1. UNICEF.(1985). Sri Lanka: the social impact of economic policies during the last decade. Colombo, Sri Lanka: UNICEF. 45 Wanasinghe, Shelton. (1995). Activating the Administrative Reform Process in Sri Lanka. Research Studies Governance Series No. 1, Colombo, Sri Lanka: Institute of Policy Studies. White, Howard, and Saman Kelegama. (1993). External Shocks, Adjustment Policies and the Current Account: The Case ofSri Lanka, 1971-1991. Research Studies Macroeconomic Series No. 7, Colombo, Sri Lanka: Institute of Policy Studies. . (1994). "The Fiscal Implications of Privatisation in Developing Countries: The Sri Lankan Experience." Working Paper Series No. 179. The Hague: Institute of Social Studies. World Bank. (1995). Sri Lanka: Private Sector Assessment. Report No. 12514-CE. Washington, D.C.: World Bank. . (1994a). Sri Lanka: Tree Crops Strategy. Report No. 12356-CE. Washington, D.C.: World Bank. . (1994b). World Bank Assistance to Privatization in Developing Countries. Report No. 13273. Washington, D.C.: World Bank. . (1996). World Development Report 1996: From Plan to Market. Report No. 15441. Washington, D.C.: World Bank.  47 Annex I ERC: Status of Legal Covenants Descriptions of covenant Comments Rice, wheat-flour, fertilizer, and The price of rice is subsidized indirectly through purchases transport subsidies not reinstated. by the Paddy Marketing Board, wheat-flour is subsidized directly at a cost of about Rs 6 billion in 1995, and fertilizer is subsidized directly at a cost of Rs 1.5 billion in 1995. Bus fares regulated and rail subsidies remain. These reversals came after the release of the second tranche. Agreement on plan to restructure the Not done. sugar industry. Agreement on plan to implement The capital gains tax and stamp duty on share transfers were recommendations of the Tax eliminated in the 1991 budget but the VAT has not been Commission introduced, and other ad hoc levies were introduced Macroeconomic policies agreed in the Done. second year PFP implemented. Reduction in central and provincial Not done. Second tranche release followed by significant administration staff. new hiring. Agreement on a plan to reduce central There was an initial reduction in the civil service. This was and provincial administration staff in followed by hiring which more than reversed the initial 1990, and the plan is reflected in the reduction. budget. Cabinet approval of legislation to Not done. establish a civil service commission. Conversion of the Employees The Board waived this condition for second tranche release Provident Fund into an employee because parliament could not approve the conversion pension scheme. because it was not in session. The conversion was not approved because of design problems and union opposition. Agreement on a plan to reduce the Not done. cost of the civil service pension scheme. Deregulation of air freight operation Done. The Government allowed all airlines to register to carry airfreight. 48 Descriptions of covenant Comments Agreement on a plan to introduce a In September 1990 a four-band tariff was introduced and the four-band tariff and to reduce the maximum rate was lowered to 50%. In the 1995 budget the maximum tariff. maximum rate was lowered to 45%. However several goods remained outside the standard tariff system, were imported under exemptions, or were subject to quotas. Increase export competitiveness and Done. other industrial reforms Limit the cost for the subsidy Done, but by 1995 exceeded 3.5%. programs to no more than 3.0-3.5% of GDP Increase economic returns to public Not done. investment projects. Improve the productivity of tree-crop Done. estates. Restructure subsidy programs for Done, but reversed after second tranche release.. better targeting. Note: There were more detailed covenants than the 16 listed here. For example, there were 16 conditionalities for the second tranche release alone. The listing here aggregates over similar targets. 49 PMEAC: Status of Legal Covenants Descriptions of covenant Comments Open and maintain a special account in US Done. dollars. Audit records and accounts and provide IDA Done. with audit reports within 6 months of the end of the fiscal year. Divestiture of at least 40 percent of the Done as a condition of second tranche release. Treasury's shares in public companies whose asset value, in aggregate, would correspond to at least half of the total asset value of MOI's mining and mineral enterprises (excluding that of the Salt Corporation). Divestiture of at least 60 percent of the Done as a condition of second tranche release. Treasury's shares in public companies whose asset value, in aggregate, would correspond to at least half of the total asset value of MOI's non-mining and non-mineral enterprises (excluding that of the Paper Corporation). Satisfactory progress in the conversion of all Done as a condition of second tranche release. MOI's enterprises into public companies. Submission of a bill to Parliament to enact an Done as a condition of second tranche release. The Act Industrial Promotion Act. was passed in 1990. Submission of a bill to Parliament to enact a Done as a condition of second tranche release. The Act Mining and Mineral Development Act. was passed in 1992. Adoption by the Government of a strategy to Done as a condition of second tranche release. restructure and privatize existing salt extracting and processing operations. Adoption by the Government of a strategy to Done as a condition of second tranche release. streamline the paper operations of the National Paper Corporation. Adoption by the Government of a strategy to Not done. IDA agreed that the restructuring and restructure and modernize the cement industry. modernization strategy was no longer needed or desirable because two of three cement plants were privatized by 1992 and the remaining plant was in the conflict zone.  51 Annex H Aggregate Costs of the Civil Service Separation Program As part of a program to reduce civil service employment, the Sri Lankan Government made generous separation packages available to induce staff to leave government employment. The full details of the program are described in Coopers and Lybrand (1991). Fiszbein (1992) provides an analysis of what the optimal separation package should be to induce a voluntary separation. Given that the objective of this part of the reform was a reduction in staff, the program suffered from a number of failings: (i) More staff were rehired than separated. As a result, civil service employment increased by almost 10 percent; (ii) Separation benefits were given to many who would have retired anyhow; (iii) Concerns have been expressed that the best staff left; and (iv) the separation package is still in place. To get a sense of the direct budgetary costs of this program, a simple present value calculation of total program costs is constructed. It relies on the analysis in the Coopers and Lybrand (1991) study of the costs of the program. That analysis made careful estimates of how many separating employees would be eligible for retirement benefits. It also estimated how the number of separatees would be allocated over job classifications. These estimates allow for an aggregate costing of the program. For 20,000 separations the estimated costs were US$30 million. However, in the event, Wanasinghe (1994) estimates that there were 30,000 separations. We assume that the additional 10,000 separations have the same cost profile as the 20,000, so aggregate costs are US$45 million. If, for simplicity, separation benefits are expended immediately, then the present value of these expenditures can be calculated according to the following equation: T C=E + Lc '+Ad-P (1) Where C = present value of program costs; E = severance expenditures; assumed to be US$45 million or .6 of GDP; LCi = savings on labor costs in period i; Pk = contribution of laid off workers elsewhere in the economy in period i; Pj= foregone contribution of laid off worker from civil service employment in period i; r = a discount rate of 10 percent Because 10 percent more workers were hired than laid off, there are no savings on labor and LC is equal to negative .1 times the 5.1 percent of GDP expended on civil service wages or -.5 percent of GDP per year. 52 Due to the lack of empirical evidence, a number of different assumptions can be made about the productivity of those employed in the civil service and outside of it, and the relationship of their productivity to their compensation. For example, among other things, one could assume: 1. Productivity in civil service is negative 25 percent of wages. In other sectors it is 100 percent of wages, but wages in these other sectors are, following Prywes' (1995) estimates, equal to 40 percent of government wages; 2. Productivity equals wages in civil service and other sectors, and they are equal to each other; or 3. Productivity in civil service equals the productivity of other sectors; other sectors are paid 40 percent of civil service wage, and productivity equals compensation. However, because less than 10 percent of civil service employees changed jobs, and LC, the lack of savings on labor costs per year, is more than four times larger than the job changing effects, the overall results are largely insensitive to assumptions about worker productivity and wage compensation. If we follow the simplest assumption, i.e., 2, then equation (1) can be restated in terms of GDP and budgetary effects as: 6 C =.6+ -4l+ ' = 4.3%ofGYDP In other words, on the order of 40 percent of the 1995 budget deficit in 1995, and more than 100 percent of the amount by which the deficit exceeds the PFP target can be attributed to the way the civil service reform was enacted. Under assumption 3, which seems more realistic the budgetary costs fall to about 3.8% of GDP, still more than the total amount by which the PFP target was exceeded. Consequently, the failed civil service reform was costly not only in its failure to improve governance, but also directly. SRI LANKA: ECONOMIC INDICATORS (Page 1) 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 18 99 90 9 Average Average GDP Real GDP Growth 5.0% 4.3% 1.5% 2.7% 2.3% 6.2% 4.8% 4.3% 6.9% 5.6% 3.2% 5.6% 4.4% Government Consumption (% of GDP) 10% 10% 10% 10% 10% 10% 10% 10% 9% 10% 10% 10% 10% Private Consumption (% of GDP) 78% 78% 77% 78% 77% 76% 77% 75% 75% 75% 78% 76% 77% Gross Domestic Capital Formation (% of GDP) 24% 24% 23% 23% 22% 22% 23% 24% 25% 27% 23% 24% 23% Gross Domestic Savings (% of GDP) 12% 12% 13% 12% 12% 14% 13% 15% 16% 15% 12% 15% 13% Exchange Rates (Rupees) -- - - _i__-_3 Market Rate: Per US$, Period Average 27.2 28.0 29.4 31.8 36.0 40.1 41.4 43.8 48.3 49.4 30.5 44.6 37.5 Market Exchange Rate Index (1990= 100) 148 143 136 126 112 100 97 91 83 81 133 90 112 Interest Rates (Percent Per Annum) Treasury Bill Rate 13% 10% 7% 14% 15% 14% 14% 16% 17% 12% 12% 12% Deposit Rate (National Savings Bank) 12% 12% 12% 12% 14% 16% 14% 14% 14% 14% 12% 14% 13% Lending Rate (DFCC) 14-21 14-18 14-19 14-19 14-19 14-19 16-22 17.5-23.5 18-24 18-22 Lending Rate (NDB) 7.0-14 7.0-14 7.0-14 8.2-15 10.3-17.5 11-18.5 11.3-18.5 13.9-20.5 13.7-20.5 13.6-22.0 Prices: Period Averages (1990=100) Inflation Rate (change in CPI) 1.5% 8.0% 7.7% 14.0% 11.6% 21.5% 12.2% 11.4% 11.7% 8.4% 8.5% 13.1% 10.8% InflationRate(changein WPI) -15.2% -3.0% 13.4% 17.8% 9.1% 22.2% 9.2% 8.7% 7.6% 5.0% 4.4% 10.5% 7.5% Change in Industrial Wage Rate Index 9.3% 5.4% 5.1% 25.0% 15.2% 18.6% 11.7% 14.9% 21.0% 2.2% 12.0% 13.7% 12.8% Trade Balance Exports as a % of GDP 22% 19% 21% 21% 22% 24% 22% 25% 28% 27% 21% 25% 23% Tea Exports as a % of Total Exports 33% 27% 26% 26% 24% 26% 22% 14% 14% 13% 27% 18% 23% Total imports (fob) as a % of GDP 33% 30% 31% 32% 32% 33% 34% 36% 39% 41% 32% 37% 34% TradeBalance(%ofGDP) -11.0% -11.4% -9.9% -10.9% -9.5% -9.7% -11.9% -10.7% -11.1% -13.4% -10.5% -11.4% -10.9% Current Account Balance Current Account Deficit (% ofGDP) -7.0% -6.7% -5.1% -5.6% -4.4% -3.2% -5.4% -4.5% -3.8% -6.5% -5.7% -4.7% -5.2% Current Account Deficit before grants (% of GDP) -9.9% 9.5% -7.8% -8.5% -7.1% -5.5% -7.6% -6.4% -5.3% -7.9% -4.8% -6.5% -5.7% Foreign Direct Investment (US$ million) 26.2 29.7 59.5 45.7 19.7 43.4 48.4 122.6 194.5 - - 36.2 81.8 59.0 Foreign Direct Investment (Growth Rate) -20% 14% 100% -23% -57% 120% 12% 154% 59% -14% 3% 66% 34% Fiscal Balance Government Deficit (% of GDP) -11.6% -11.4% -11.1% -15.1% -11.2% -9.9% -12.1% -8.0% -8.3% -10.2% -12.1% -9.7% -10.9% SRI LANKA: ECONOMIC INDICATORS (Page 2) 1985 1986 1987 1988 1989 1990 1991 1992 1993 1994 Average Average Average 1985-1989 1990-1994 1985-1994 Percent Financed from Domestic Sources 55% 50% 67% 75% 73% 54% 45% 68% 69% 66% 64% 60% 62% Percent Financed from Foreign Sources 45% 50% 33% 25% 27% 46% 55% 32% 31% 35% 36% 40% 38% Net Domestic Borrowing (% of GDP) 5.3% 5.1% 5.8% 9.5% 6.3% 4.2% 4.3% 3.7% 4.9% 6.5% 6.4% 4.7% 5.5% Net Foreign Borrowing (% of GDP) 4.4% 5.0% 2.9% 3.2% 2.4% 3.6% 5.2% 1.7% 2.0% 2.0% 3.6% 2.9% 3.2% Tax Revenue as a % of GDP 18.7% 17.4% 17.9% 16.2% 18.9% 19.0% 18.3% 18.0% 17.5% 17.2% 17.8% 18.0% 17.9% Tax Revenue as a % of Total Revenue 84% 84% 83% 86% 88% 90% 89% 89% 89% 90% 85% 89% 87% Revenue from Corporate Taxes (% of Total) 11% 9% 8% 8% 5% 6% 8% 8% 7% 8% 8% 8% Revenue from Individual Taxes (% of Total) 4% 4% 4% 4% 4% 4% 5% 5% 5% 4% 5% 4% Revenue from Export Duties (% of Total) 8% 4% 5% 4% 3% 4% 1% 1% 0% 5% 2% 3% Revenuefrom Import Duties (% of Total) 22% 25% 26% 26% 28% 25% 24% 24% 21% 25% 24% 25% 55 Attachment I MDiST= OF IDUeRAL DEVEOPMIr *-. e. 570. .a. au. "a, P. 0.O x m. 73wc. ma . =k dl, rA a. '3f.fl& . G.•Gar -. .nc~k 3. .r*--S. 5th Jine 1996. M'yNo . YeéNo.Dala Mr. Manuel Penalver, * ~ Divis ion Chief., CoUntry Policy, Industry, and Finance Operations Evaluation Department, The World Bank. Dear. Mr. Penalver, Ro: ECon~-C Restructuring Credit (Cr.2128-Cm). an Public Mnnfact:ne-ng Eter~riMes ådjustent Credit (Cr.2185-CE) - Perfozwanc Audit Report I have to thanic you for your letter dated 9th May '96 . forwarding the OED report on the above projects. I have only a few comments to make and these are res- tricted to the PMEAC. Thuy are as followw: - Para. 2.25: The last septence- reförs to the fact that the PfIB did .nt work well. The PI vas a company registered .under the Companies Ordinance and therefore its relationship with line Ministriss was. unclear. Though the Intention for -GattIng up the . PI4 was to withutand any political pressures, it failed du. to the fact that it functioned directly nder the Ministry of i'inance. Para 2.30: The position regarding Lanka Salt Ltd. iu that the operations at Puttalam which originally cme under the purviev of this company vre seperated and a subsidiary under the titl .'Puttalam Salt Ltd.' mot up to look after · ua.} 421401 8asuui IeLs 4440 'ea uas Mtc }}~ }L. 56 those operations. Puttalam Salt Ltd. was thereatter. identified for priva4tisation, but could not be proceeded vith due - the General Election that intervened. Wben the majority holding of Puttalan Salt Ltd. is divested to the private sector, it will end up as a private company. The reference to Lanka 'Phosphate Ltd.. in the last. sentence of this para needs revision. Lanka Phosphate Ltd. -has the mining rights for the unique phosphate deposit in the North Central Province. Proposals were called to set. up *A joint, venture with Lanka Phosphate Ltd. to manufacture phosphate fertiliser for export utilising this deposit. Discussions are still proceeding with one of the largest manufacturers of phosphate fertilizer on this basis. If the negotiations succeed, equity sharing will be such that it would. result in a major divestiture. Para 2.44: My comments on para 2.25 apply. As far as valuation of enterprises was concerned, the majority of them - were done by the Chief Valuer of government. . -It would have been useful, if at .the initial stages, aqoe guidelines had been set 'in respect of valuation of enterprises and divestiture, particularly when it .took. place outside the -Stock Exchange. Yours Sincerely, Vincent P ultant Attachment II 57 THE OVERSEAS ECONOMIC COOPERATION FUND, JAPAN COLOMBO OFFICE: 3rd Floor, NDB Building, 40 Navam Mawatha, Colombo 2. Tel: 422255 422260 422264 Fax: 422272 TIx: 23278 OECF CE June 04, 1996. Mr. Manuel Penalver Division Chief Country Policy, Industry, and Finance Operations Evaluation Department World Bank. Dear Mr. Penalver, Re: SRI LANKA - Economic Restructuring Credit Performance Audit Report Thank you very much for sending us the captioned report and providing us the opportunity to comment on it. We almost fully agree with your evaluation. We consider that the achievement of ERC was 'mixed'. We particularly agree that ERC was less successful in the areas of civil service reform and rationalisation of GOSL's massive and ineffective poverty alleviation programme. We also agree that the privatization should have been/should be pursued for efficiency gains rather than for budgetary purposes. This consideration should be much more important from now on because GOSL is now intending to privatize public utilities. We consider that our evaluation results on the privatization of CTB and the public plantation companies are well taken in your report. GOSL recently made a decision to fully privatize plantation companies. We consider this as a brave step and we believe that you share the same view. With best regards. Kazushi Hashimoto Director 3rd Division Operations Department II cc: OECF Washington Office OECF Colombo Office OECF RIDA (Evaluation Group)