103706 THE WORLD BANK OPERATIONS EVALUATION DEPARTMENT Russia: Bank Assistance for Social Protection Lawrence Thompson Director-General, Operations Evaluation: Robert Picciotto Director: Gregory K. Ingram Acting Manager: Rene Vandendries Task Manager: Gianni Zanini 2002 The World Bank Washington, D.C. OPERATIONS EVALUATION DEPARTMENT ENHANCING DEVELOPMENT EFFECTIVENESS THROUGH EXCELLENCE AND INDEPENDENCE IN EVALUATION The Operations Evaluation Department (OED) is an independent unit within the World Bank; it reports directly to the Bank’s Board of Executive Directors. OED assesses what works, and what does not; how a borrower plans to run and maintain a project; and the lasting contribution of the Bank to a country’s overall development. The goals of evaluation are to learn from experience, to provide an objective basis for assessing the results of the Bank’s work, and to provide accountability in the achievement of its objectives. It also improves Bank work by identifying and disseminating the lessons learned from experience and by framing recommendations drawn from evaluation findings. OED Working Papers are an informal series to disseminate the findings of work in progress to encourage the exchange of ideas about development effectiveness through evaluation. The findings, interpretations, and conclusions expressed here are those of the author(s) and do not necessarily reflect the views of the Board of Executive Directors of the World Bank or the governments they represent. Contact: Operations Evaluation Department Partnerships & Knowledge Programs (OEDPK) email: ecampbellpage@worldbank.org email: eline@worldbank.org Telephone: 202-458-4497 Facsimile: 202-522-3125 http:/www.worldbank.org/oed Acronyms ESSP Employment Services and Social Protection FES Federal Employment Services GDP Gross Domestic Product IMF International Monetary Fund MESI Minimum Elderly Subsistence Income MSI Minimum Subsistence Income OED Operations Evaluation Department SAL Structural Adjustment Loan SECAL Sector Adjustment Loan SPAL Social Protection Adjustment Loan SPIL Social Protection Implementation Loan WBI World Bank Institute Contents Preface ...................................................................................................................................i Executive Summary...............................................................................................................iii 1. Introduction.......................................................................................................................1 2. Sector Performance and Challenges .................................................................................1 3. Evolutio n of the Bank’s Sector Assistance Strategy ........................................................8 4. Bank Products and Services Assessment..........................................................................9 5. Development Effectiveness Impact Assessment ..............................................................17 Counterfactual...............................................................................................................17 Outcome........................................................................................................................18 6. Attribution of Bank and Program Results ........................................................................22 7. Agenda for Future Action.................................................................................................24 Annexes Annex 1: Economic and Social Indicators.............................................................................27 Table 1: Economic Indicators ...............................................................................27 Table 2: Social Indicators .....................................................................................27 Annex 2: Sectoral Assistance Strategy Matrix ......................................................................28 Annex 3: List of Persons Interviewed ...................................................................................30 Annex 4: Comments of Mr. Vladimir Gimpelson (FCPF) ....................................................32 i Preface This paper is one of the background papers prepared as an input to the Russia Country Assistance Evaluation (CAE Task Manager, Gianni Zanini) by the Operations Evaluation Department (OED) of the World Bank. Findings are based on a review of project appraisal and completion reports, sector reports, and a number of other documents produced by the borrower, the Bank, and research papers in the academic literature. The author (a Senior Fellow at The Urban Institute, Washington, D.C; former Principal Deputy Commissioner, U.S. Social Security Administration from 1993 to 1996 and former Assistant Comptroller General of the United States for Human Resources Programs, U.S. General Accounting Office from 1989 to 1993) visited Russia in February 2001 and interviewed current and retired government officials and Russian experts. Bank staff were interviewed at both headquarters and in the field office. An earlier preliminary version was discussed at a small workshop in Moscow in February 2001, with the participation of central government officials, academics, members of policy research institutes, and representatives of project implementation units of Bank- supported projects. Their valuable assistance and feedback is gratefully acknowledged. The author is grateful for the comments received so far on previous drafts by the OED peer reviewers (S. Carvalho and E. Hurwitz), the CAE task manager, other contributors to the CAE background work, Bank staff and managers (M. Rutkowski, J. Schweitzer, H. Sederlof, and M. Rashid), which have been taken into account in the June 2001version. However, the views expressed in this paper remain entirely those of the author. They do not necessarily represent the views of the World Bank. An earlier draft dated June 20, 2001 was sent to the Russian Government for review. Comments were received from Mr. Vladimir Gimpelson, a consultant engaged by the Federal Centre for Project Finance (FCPF) on behalf of the Government, and have been taken into account in this paper. iii Executive Summary 1. The social protection system that Russia inherited from the Soviet Union needed major restructuring to be effective in a market economy. Existing social benefits needed to be better targeted to help those most in need of assistance, and new benefits were needed to assist the unemployed and poor households with working-age heads. Reforms were also needed in the institutions responsible for administering social protection to reduce the role previously paid by enterprises and improve the administrative capacity of both local governments and the various national social insurance institutions. 2. From the beginning of its relationship with Russia, the Bank viewed social protection reform as one of the most important elements of its country assistance strategy, largely out of concern that a breakdown in the social safety net would undermine political support for economic reform. It developed a reform strategy that included improvements in benefit targeting and enhancement of institutional capability. The strategy was implemented through one investment loan (the Employment Services and Social Protection (ESSP) loan, approved in 1992); three major sector adjustment loans (the Social Protection Adjustment Loan (SPAL) of 1997 and the Coal Sector Adjustment Loans of 1996 and 1997); an implementation loan (the Social Protection Implementation Loan (SPIL) of 1998); and other economic and sector work financed by the Bank. There were also social protection elements in the three Structural Adjustment Loans. 3. In general, the outcome of social protection efforts during the first part of the 1990s was moderately satisfactory and the outcome in the second half of the decade was moderately unsatisfactory. Much of the reform agenda remains unrealized. Originally, the ESSP was to help the employment service deal with the effects of enterprise restructuring and help the government develop plans for reforming the pension system and social safety net. After a two-year delay in implementation, however, the loan objectives were changed to focus on assistance in introducing active labor market programs into the employment service and the procurement of computers for local offices responsible for paying pensions. The ESSP more than achieved its revised objectives, although the revised objectives fell far short of the reform needs that had been identified in the Bank’s earlier sector analyses. 4. The Coal SECALs included highly relevant social protection objectives involving the assurance that wages and social protection benefits were paid, that employment services were provided to displaced miners, and that social assets were rehabilitated before being transferred from closed mines to local governments. The project was highly efficacious in achieving the first objective, substantially efficacious in achieving the second, and disappointing in meeting the third. 5. The SPAL-SPIL package addressed most of the important remaining social protection reforms, but with mixed success. It succeeded in focusing greater government attention on the social protection issues, particularly the need for more effective targeting of spending on the poor and the complexity of the pension reform challenge. It improved the financia l status of the pension, child allowance and unemployment benefit programs, iv and the targeting of many existing social benefits. It was not successful, however, in promoting systemic pension or labor law reform, at least in the time frame originally contemplated. 6. Major barriers to success were the poor fiscal conditions prevailing through most of the 1990s, and the unsatisfactory performance of both the Bank and the borrower during the latter half of the 1990s. Where there were achievements, they are likely to be sustained. 7. Lessons to be learned from the Bank’s experience in Russia in the 1990s include the need to: (a) pay more attention to the political and institutional aspects of the reform process; (b) have reasonable expectations about the pace at which reforms can be developed, adopted, and implemented (c) deal with all of the relevant units of government, and not just the federal executive; and (d) employ lending vehicles with more flexibility to adjust lending activities as the policy development and implementation process passes through its various phases. 1 1. Introduction 1.1 Before the collapse of its communist regime in 1990, the Soviet Union maintained a comprehensive system of social protection that was closely integrated with the structure of its general economy. The system was based on the principle that every individual was entitled to (and expected to have) a job and that together with nonwage supplements provided through the enterprise, the job would provide sufficient income to guarantee a minimum standard of living. Social protection benefits were needed only for those who could not be expected to work, such as the aged, disabled, and ill. Benefit levels reflected the prevailing compensation and price structure, particula rly the heavily subsidized prices for essential consumer goods and the generous nonwage forms of compensation. Many of the system’s benefits were delivered through the workplace, casting enterprises in the role of agents of the state for purposes of operating the social protection system. 1.2 Because of the close relationship between the structure of the economic system and the structure of the social protection system, fundamental economic reform inevitably required a parallel social protection reform. Economic reform began in the middle of the 1980s with the perestroika campaign; the initial round of social protection reforms followed several years later at the end of the Soviet period. 1.3 Since the early 1990s, progress in reforming the social protection system has been slow, and the reforms have been incomplete. Reform efforts have been hindered by macroeconomic and public finance problems in the Federation, by weaknesses in the governing institutions inherited from the Soviet era, and by the cons tant pressure to maintain adequate levels of protection under the preexisting programs, which a large segment of the population relied on for sustenance. It is also likely that the size and scope of the reform challenge were not fully appreciated when the reform process began. 2. Sector Performance and Challenges 2.1 Basic opportunities and constraints. A successful social protection reform requires addressing systematically three different dimensions: policy, politics, and administration. The proposed policy changes must successfully address the particular social protection problems faced by the society in a fiscally sustainable manner and with a minimum of adverse behavioral incentives. Political success requires a consensus in support of reform, which will necessarily reflect broad acceptance of the need for reform and of the general direction that the reform effort will take. The system also requires institutions capable of collecting sufficient revenues to finance desired benefits, establishing eligibility consistently and accurately, and managing resources in a reasonably efficient and effective manner. 2.2 Reform of the social protection system found in the Soviet Union in late 1980s involved overcoming serious challenges on all three of these dimensions. On the policy dimension, the system needed to cover new social risks and restructure existing programs. To maintain affordability in the tight fiscal situation, program expansions had 2 to be offset by reductions elsewhere in the system, greatly complicating the political challenge that reform faced. Major reforms were also needed on the administrative dimension involving shifting the responsibility for delivering social services, strengthening local governmental institutions, and creating new institutions for financing social protection programs. As the Soviet Union disintegrated, moreover, the old political system also disintegrated. It was replaced by a new set of political arrangements that would prove resistant to the development of the kind of consensus that reform required. 2.3 At the beginning of 1990, there were neither cash benefits nor active labor market programs available to assist the unemployed. A Federal Employment Service (FES) existed, but its activities were tailored to the needs of the old economic system. There was also no program of cash assistance for those with low incomes who did not fit into one of several specific demographic categories, such as the aged, disabled, or war veterans. 2.4 As the transition began, lo cal governments lacked the institutional capacity and the fiscal base for operating a broad-based social protection system. The lack of capacity can be traced, in part, to the fact that the state enterprises had traditionally played such an important role in delivering social services and social protection benefits. The major social protection function performed by local governments was the delivery of old-age and disability pensions, programs in which the benefit amount was determined by a legislated formula and the entire cost was covered through transfers from other government agencies. Finally, the tax and social insurance contribution collection systems were geared to an economic structure dominated by state enterprises operating through a monopoly banking system and lacked the capacity to secure substantial compliance with the newly emerging decentralized system of privatized businesses. 2.5 Sector performance. The government structure and the political situation during the decade of the 1990s were a further barrier to reform. The Russian Federation withdrew from the Soviet Union in 1991 after a two-year power struggle between the leaders of the two political units. In the fall of 1993, another power struggle erupted between the president of the Russian Federation and the Russian Supreme Soviet, leading eventually to the adoption of a new constitution and elections to select a new State Duma. The first Duma was elected to a two-year term in December 1993 and a second Duma was elected to a four- year term in December 1995. No single party had a majority in either. The former communists constituted the largest single block in the second Duma and were generally hostile to the president’s reform efforts. President Yeltsin overcame generally low popularity ratings to be reelected in the summer of 1996, but was never able to generate strong political support for his reform program and resigned in December 1999. 2.6 Early in the reform process, local governments were given responsibility for delivering health and education services without being given a corresponding source of revenue, leading to deterioration in the quality of both the services being delivered and the physical facilities in which they were delivered. Although there were some 3 intergovernmental assistance in the form of budget transfers from Moscow, the amounts were inadequate and the allocation process was opaque and unsystematic. 2.7 The decentralization of power further complicated the reform effort. Much of the authority for the social protection system rested with local governments with no clear mechanism for the central government to exercise leadership or control. At the center, moreover, responsibility for social protection matters within the executive branch was spread across several ministries, and responsibility for policy development was divided between the executive and a frequently hostile parliament. Further barriers to effective policy development and coordination were the frequent change in governments over the period and continual reorganizations (or debates about reorganization) of the social protection institutions. 2.8 The combination of macroeconomic weakness and microeconomic reforms presented additional challenges for the social protection system. As officially measured, Russia’s gross domestic product (GDP) declined continuously from 1991 through 1996, falling by over 40 percent during these six years. 1 The economy grew slowly in 1997, but shrank again in 1998 as a result of the fiscal crisis. 2.9 The unemployment rate as measured in household surveys rose continuously over the whole period. For a variety of reasons, however, the number that actually registered as unemployed remained small. The reform of the price systems in the early 1990s was followed by extremely rapid inflation over the period 1992 through 1995. Inflation was brought under temporary control in 1996 and 1997, but erupted again in 1998 and 1999 as a result of the fiscal crisis. 2.10 Meanwhile, real wages declined steadily over the four years from 1992 through 1995, from 3.2 times the minimum subsistence income (MSI) at the beginning of this period to 1.8 times the MSI in 1995. Real wages recovered in 1996 and 1997, but fell again as a result of the financial crisis. By 1999, the average wage had fallen further, to 1.7 times the MSI. 2 2.11 Because of these economic problems, Russia struggled over most of the decade to maintain adequate benefit levels in its major social protection programs and to develop an effective response to rising numbers of poor people. Over most of the 1990s, the average pension fluctuated just above the level of the minimum elderly subsistence income (MESI), but it fell to just 70 percent of MESI as a result of the financial crisis of 1998. Meanwhile, the minimum pension declined steadily relative to both MESI and to the average pension over the 1992-1995 period, dropping to under 50 percent of MESI in 1995. The minimum was increased to 76 percent of MESI in 1997, but fell to 40 percent of MESI [in 1998?] as a result of the financial crisis. 1 See Annex 1, Table 1. 2 See Annex 1, Table 2. 4 2.12 The decline in real benefit levels, particularly the minimum benefit, was only part of the problem, however. Macroeconomic weaknesses and the slow pace of economic reform lead to arrears in enterprise wage and tax payments, which undermined the revenue base of both the general budget and the social funds. The result was repeated instances of delays in making social protection payments. Serious problems with payment arrears emerged in 1994 and 1995 in the pension, unemployment, and child allowance programs. Pension arrears were cleared in 1996, emerged again during the fiscal crisis of 1998, and were cleared again in 1999. Legislation adopted in 1997 and 1998 allowed unemployment benefit arrears to be cleared and full child allowance payments to be made on a prospective basis. Some of the child allowance payments that should have been made in the mid-1990s have never been paid, however. 2.13 Poverty rates increased sharply as a result of price decontrol in 1992, and appear to have drifted upward more or less continuously since then. According to World Bank estimates, the 1998 fiscal crisis left almost half of the population living in poverty. Beginning in 1994, the addition to estimated household expend itures of an imputed value for the use of consumer durables caused one-time downward shift in that measure of the poverty rate. Apparently, this methodological change is the major reason that the official figures show lower poverty rates in the second half of the 1990s than were observed in the first half. 2.14 The first round of reforms (1990-1992). The Soviet leadership introduced a series of reforms in 1990 and 1991 that were designed to adjust the social protection system to a market economy. They included: • A new system of active employment programs and unemployment benefits, which lay the groundwork for dealing with the labor market consequences of enterprise restructuring. Responsibility for these new programs was given to the FES, an existing institution with no experience in operating either type of program. • The shifting of the responsibility for financing major elements of the social protection system from the general government budget to a new set of institutions and revenue sources, which provided the basis for developing a more rational system for financing the social protection system. Pension contributions were to be collected by a new Russian Pension Fund; contributions to finance, sickness and maternity benefits were to be collected by a new Social Insurance Fund; and a new tax was to be levied to finance the employment programs and be collected by a new Employment Fund, which was to be managed by the Federal Employment Service. • A new system of family allowances, which gave small additional stipends to households with children. The system relied primarily on extra payments from enterprises. 2.15 The first round of reforms filled the unemployment protection gap, a change that was considered at the time to be a priority due to the concern that enterprise restructuring 5 would soon generate serious unemployment, undermining the first principle of the traditional Soviet social protection system. In fact, registered unemployment never became very significant, despite an increase in joblessness as measured by the International Labor Organization concepts. The low level of registered unemployment appears to have been the result of two factors, the tendency for workers to remain technically affiliated with their former enterprises, even if no wages were being paid, and the low level of unemployment benefits available to those who did register. 2.16 Several aspects of the new unemployment benefit would generate further reform efforts over the decade, however, involving both policy issues and implementation issues. Among the policy concerns were overly generous benefit entitlement rules (for example, making new labor force entrants eligible) and the authority to use employment fund monies to subsidize employment in existing enterprises. The latter would prove to be a barrier to the restructuring effort by prolonging the life of failing enterprises. Implementation concerns included a general lack of confidence in the effectiveness of the collection mechanisms and in the integrity of the financial management systems at the Employment Fund and limitations on the ability of the Employment Fund to transfer resources from low-unemployment regions to high-unemployment regions. Finally, inflation would seriously erode real benefit levels soon after the program was established; benefit levels remained inadequate for the rest of the decade. 2.17 Pension adjustments introduced in this round of reform left the basic structure of pension benefits unchanged. The subsequent deterioration in the economic situation caused these benefits to become inadequate (when they were paid at all), even though the pension contribution rate was 28 percent of wages, a level that many considered a hindrance to economic growth. The combination of inadequate benefits and high contribution rates signaled the need for a more fundamental reform of the pension program that would adjust entitlement and benefit calculation rules. The case for immediate attention to pension reform was strengthened further by a projected deterioration in the demographic situation shortly after the turn of the century. 2.18 Although the new Social Insurance Fund collected contributions for sickness, maternity, and related benefits, many of the benefits continued to be paid through the enterprises, which remitted their contributions on a net basis. This arrangement contributed to later concerns about the financial management of the Social Insurance Fund. In addition, many of the benefits provided through this fund for somewhat vague “rest and recreation” programs involved stays at spas and similar activities, raising questions about their priority. Also, the idea that social insurance premiums would be collected by three different funds operating essentially independently of each other as well as independently of the State Tax Service (now the Ministry of Taxation) struck many as, at best, a needless duplication of administrative arrangements. Debates about consolidating contribution collection responsibilities at the tax service continued through the rest of the decade, further destabilizing the social protection institutions. 2.19 In principle, the family allowances should have helped the working poor. In practice, however, they appear to have had only a limited impact on overall poverty 6 trends. In part, they were neither large enough nor sufficiently well targeted on those with the least adequate income to cause a significant reduction in poverty. In addition, since they were a liability of the enterprises, they often were not paid at all. 2.20 In mid-1992, the post-Soviet Russian government endorsed rapid implementation of further social protection reforms, including increases in minimum unemployment and pension benefits, creation of a new means-tested cash benefit, and improved program management. 3 By mid-1993, however, government interest in the reform effort faded, and, at least until 1995, social protection issues received comparatively little attention, despite rising poverty rates and declining average benefits. The implicit assumption was that economic growth was the most effective strategy for solving the emerging social protection problems. 2.21 The second round of reforms (1995-1997). A second round of reform began in the 1995-96 period, when Russia was forced to borrow from the International Monetary Fund (IMF) to cover its government budget deficit. One of the major uses of IMF funds was the clearing of the pension arrears, highlighting the fiscal imperative of policy and institutional reform in the social protection area. 2.22 During these years, both the IMF and the World Bank urged the Russian government to pay greater attention to the need to improve the social protection system, particularly for the poorest. There was concern that continuing deterioration in the social environment would undermine political support for further economic reforms. At the same time, tight budget constraints required that reforms not produce a net increase in spending; the cost of improvements for the poor had to be covered by reductions in payments to the non-poor. 2.23 The major objectives of the second round of reforms were to improve targeting of existing social protection programs (ensure that a higher fraction of total benefits went to the poor), strengthen the institutio ns that financed and delivered them to ensure more adequate financing, and reform the pension system to simultaneously produce more adequate benefit levels and greater fiscal sustainability. At least in the short run, much of the pension reform effort failed, but other reforms succeeded, including: • increased minimum benefits in the pension and unemployment programs to provide a more adequate floor of protection; • improved targeting of children’s allowances and providing a more stable source for their financing; • tightened eligibility for unemployment benefits and improving their financing by shifting more resources from surplus regions to deficit regions; and 3 “Memorandum on Economic Reform Policies,” Annex 1, Report of the President on a Proposed Rehabilitation Loan, July 7, 1992, P-5834-RU. 7 • increased emphasis on targeting of the rest of the package of social benefits and privileges and trimming the cost of maternity and sickness benefits. 2.24 The unfinished reform agenda. Although targeting was improved in many programs, further improvements are needed, particularly in housing and utility subsidies. In addition, the pension system still must be reformed. 2.25 Throughout the decade of the 1990s, outside experts urged Russia to create a new, antipoverty benefit in the form of a locally managed system of cash assistance. So far, however, the Russian government has not been willing to adopt this as a part of its own agenda, although it has authorized localities to provide such benefits from their own resources if they wish. The absence of a general antipoverty benefit continues to be part of the unfinished agenda of social policy experts, however, even if it is not on the government’s agenda. 2.26 In many ways, the reform agenda established in Bank analytical work in the early 1990s continues to provide a valuable blueprint. In July 2000, the newly elected government of the Russian Federation issued an Action Plan for Social Policy and Economic Modernization. Consistent with earlier Bank analyses, the plan calls for greater targeting of social assistance benefits, reductions in producer subsidies and in “socially unjustified” privileges, and rationalization of the system of intergovernmental transfers that would support more targeted assistance. The government and the Duma appear to have reached agreement in mid-2001 on a pension reform strategy that combines pay-as-you- go and funded approaches to financing. As of October, however, the exact benefit structure and the financial details of the plan were not clear. 2.27 Russia adopted two major structural changes as part of a new tax code in 2000. The Employment Fund was abolished and its operations merged into the general budget. This breaks the linkage between employment fund revenues and expenditures both regionally and nationally. In addition, responsibility for collecting pension contributions was shifted from the Pension Fund to the Ministry of Taxation. The Pension Fund retains responsibility for collecting and storing earnings records and is in the process of assuming from the local social protection offices the responsibility for paying pensions in those regions where it had not previously had that responsibility. The new institutional arrangement can produce a more efficient pension administration, but only if the Pension Fund and Ministry of Taxation can develop effective procedures for coordinating information on enterprise contribution payments. 8 3. Evolution of the Bank’s Sector Assistance Strategy 3.1 Bank assistance during the period included analytical studies, conferences and workshops, and both investment and adjustment loans. Bank staff participated in the major study of the Soviet Economy conducted in cooperation with the IMF, the European Bank for Reconstruction and Development and the Organization for Economic Cooperation and Development and released early in 1991. The Bank cosponsored the initial rounds of a household survey to develop reliable estimates of the distribution of income and the impact of social protection programs. That survey provided the basic material for a series of reports focusing on patterns of poverty and the effectiveness of the social protection system, beginning in 1994 and continuing through the rest of the decade. 3.2 Over the decade, the Bank also produced several reports analyzing the structure of the Russian social protection system and recommending reform strategies. Initially these looked broadly at the entire protection system, and in particular the pension system, which was the largest component. In later years, reports focused on more specific social protection issues. A 1995 study focused on the social impact of the restructuring then going on in the coal industry, while work done somewhat later in the 1990s focused on techniques for improved targeting of social assistance benefits. Bank staff worked informally with Russian officials in 2000 to help develop the government’s current social policy reform agenda and output measures appropriate for measuring its impact. 3.3 The World Bank Institute (WBI) and regional staff organized workshops and seminars designed to inform journalists, policy makers, and administrators about major social protection reform issues. Various sessions covered issues in benefit targeting, pension reform, and labor law reform, among others. These activities occurred primarily in 1997, 1998, and 1999. 3.4 Bank loans directly specifically at social protection issues during the 1990s included the Employment Services and Social Protection Project (ESSP, effective in 1994 for $70 million), the two Coal Sector Adjustment Loans (Coal SECAL I and II, effective in 1996 and 1997 for $500 million and $800 million, respectively), the Social Protection Adjustment Loan (SPAL, effective in 1997 for $800 million), and the Social Protection Implementation Loan (SPIL, effective in 1998 for $28.6 million). The three Structural Adjustment Loans (SAL I and II, effective in 1997 for $600 million and $800 million respectively and SAL III, effective in 1998 for $1.5 billion) also included some social protection conditions. 4 3.5 The ESSP was the only direct investment loan. It was originally designed primarily to help develop the capacity of the Employment Service to handle an expected increase in unemployment claims associated with industrial restructuring. It 4 The ESSP was approved in November 1992, but became effective only in September 1994 for $70 million, of which $10 million were eventually canceled. A Coal Sector Restructuring Implementation Assistance Loan of $25 million was processed in conjunction with the first of the Coal SECALs. 9 subsequently became primarily a vehicle for upgrading the information processing capacity of the pension payment function within local social protection offices. 3.6 Several of the conditions accompanying the Coal SECALs focused on the social issues associated with coal mine closures. Major concerns included the payment of past wages, severance payments, and disability benefits to miners affected by the closures, the financing of active labor market policies to mitigate the unemployment problems in coal communities, and the assurance of adequate funding for rehabilitation and maintenance of the housing and social assets (schools, roads, utilities, and so forth) previously owned by the coal companies. 3.7 Conditions accompanying the SPAL focused on reform of specific social protection institutions, including the pension system, labor law, social assistance, the unemployment insurance system, and the system for sickness and maternity pay. A general theme was to encourage improvements in benefits for the poorest to be financed by reducing benefits to the better off. 3.8 The only social protection issue addressed in SAL I and SAL II involved clearing pension arrears. SAL III required further steps to clear pension arrears, imposed several other social protection conditions that were consistent with those contained in the SPAL, and imposed one that was arguably inconsistent with the general thrust of the SPAL. 5 An IMF Extended Fund Facility effective in 1996 also involved commitments by the Russian government to clear pension arrears and improve collections in its various social funds. 6 4. Bank Products and Services Assessment 4.1 Diagnosis and proposed solutions. From the beginning, the Bank’s country assistance strategies treated social protection reform as primarily a means to an end. The social safety net needed to be strengthened in order to facilitate industrial and labor market restructuring and to protect the disadvantaged during the reform process; the link was particularly clear in the Coal SECALs. When social protection reforms stalled in the mid-1990s, Bank officials expressed concern that the whole economic reform program would be jeopardized. It was 1999 before the Bank’s assistance strategy targeted poverty reduction as an end in itself. 4.2 Despite the more limited strategic vision, the social protection reform agenda articulated in the Bank’s analytical work was much broader and remained fairly consistent throughout the period. Its major elements included: 5 The potentially inconsistent condition called for shifting the collection function of the Employment Fund to the Tax Ministry, at a time when the SPIL was financing financial management improvements at the Employment Fund. 6 The Bank also loaned Russia $200 million in 1996 for a pilot program to rehabilitate community social facilities, primarily schools and health facilities. A proposed $100 million investment loan to strengthen the administration of the Russian Pension Fund never became effective. An $80 million loan to finance demonstration projects involving restructuring of northern communities was being processed in early 2001. 10 • Increasing minimum-benefit levels in the pension and unemployment insurance programs to minimum-subsistence levels. As a practical matter, this primarily meant increases in minimum pension benefits. Although unemployment insurance benefits were even less adequate than pension benefits, they were also far less important quantitatively. • Creating a new program of targeted cash assistance to provide an antipoverty benefit, particularly to the working poor. • Financing the benefit expansions through reductions in other elements of the social protection system. Suggestions for reductions included the following: (a) reducing subsidies to inefficient state enterprises; (b) reducing housing, utility, transit, and other subsidies to higher- income households; and (c) reducing pension outlays by increasing the general retirement age, phasing out the early-retirement preferences, and limiting the rate of growth of benefits above the minimum level. 4.3 The Bank’s analytical work was particularly effective in documenting the need for better targeting of social protection spending. It noted, for example, that poverty was as significant a problem among working-age families as among pension recipients and that the distribution of the existing social protection payments was, if anything unfair to the poor. The analytical work also focused specifically on patterns of economic disadvantage among women, including the poorer employment prospects for women workers and the need for more effective support for aged widows. With one notable exception, however, the analytical work largely ignored the practical political and institutional issues associated with achieving the proposed social protection reforms. The exception involved the analysis of procedures that could allow benefits to be targeted even if household income was not well documented. 4.4 The social protection agenda developed by the Bank is generally regarded as being analytically correct but, at least in the context of the 1990s, overly ambitious and politically unrealistic. The analysis probably also underestimated the time that would be required to develop a consensus about particular reform solutions and to enact and implement legislation. 4.5 After its brief initial enthusiasm in 1992, the Russian government found that the benefit reductions were not feasible, at least in the political environment of the time, and lost interest in most of the key elements of the reform agenda. Through the 1990s, the government resisted increasing unemployment benefits by anything approaching the amounts recommended by Bank analysts, although it has recently acknowledged the desirability of implementing this change eventually. It also has been unwilling to advance major reductions in pension entitlements such as a general increase in the retirement age or a reduction in early retirement privileges, although it has been willing to look for alternative mechanisms for financing the latter. It has generally supported greater targeting of social protection benefits, and some progress has been made in achieving this element of the reform package. However, the targeting improvements 11 have not been sufficient to relieve the general fiscal pressure on the social protection system. For this reason, the national government has not been willing to assume fiscal responsibility for a new program of cash social assistance, although it has allowed local governments that are willing and able to finance such assistance on their own to adopt such programs. 4.6 Bank-sponsored seminars and workshops appear to have been fairly well targeted. Workshops were held on pension reform, assistance targeting, and labor la w reform. Each was an important policy area where knowledge transfer activities had great potential value. These activities are generally credited with elevating the debate about pension and labor law reform options, which is probably a prerequisite for an eventual agreement on a reform package. They are also credited with increasing the acceptance of targeting of social benefits among both policymakers and local program administrators. 4.7 The record with respect to administrative reforms is mixed. The need to strengthen social protection administration was recognized from the beginning of the Bank’s involvement with Russia, and the first social protection loan was aimed at strengthening the Federal Employment Service (FES). Although some of the intended improvements were ultimately financed through other means, the project did succeed in building the FES capacity to conduct active labor market policies and improve the capacity of the agencies responsible for paying pensions. An additional loan to strengthen the Russian Pension Fund was prepared later in the decade, but was dropped before approval. On the other hand, despite early recognition of the fiscal and administrative challenges involved in implementing a general program of cash assistance, the Bank appears to have made little effort to encourage improvements in this area. 7 4.8 In the meantime, the Bank’s analysis of the social protection issues associated with mine closures appears to have been both accurate and complete. The social protection aspects of the Coal SECALs were designed to address the problems identified in the earlier Bank studies. The design of these projects was coordinated with the Trade Union Federation, which seems to be generally supportive of their original objectives and approach. The major complaint from both local government officials and the trade unions is that the government failed (and the Bank failed to require it) to meet the financial commitments it had made, undermining the value of the social protection component of the coal loans. 4.9 ESSP loan. The first social protection loan was the Employment Service and Social Protection (ESSP) loan, approved by the Bank in November 1992. At the time, it was seen as the first step in the social protection reform process. Its goals reflected the reform prescriptions contained in the earliest analytical reviews and endorsed by the government earlier in 1992. They were (a) to develop the capacity at the Employment Service to process the expected increased volume of unemployment claims and to implement active labor market policies and (b) to assist the Ministry of Social Protection 7 Bank staff are now preparing a Regional Fiscal Reform Loan that is scheduled for Board presentation later in 2001 and will begin to deal with fiscal reforms at the subnational level on a demonstration basis. 12 in reforming the social safety net and the pension systems. The first goal was to be achieved through assistance in computerization of FES offices and in the design and implementation of a variety of active labor market policies. The second was to be achieved through assistance in developing program reform plans, developing computer systems for processing pensions in three pilot regions, and developing a master plan for a computerized social security system. The project was to be implemented over the period February 1993 through October 1994. Ninety percent of the Bank’s funding was budgeted for the employment service component. 4.10 Loan processing and procurement difficulties caused major implementation delays. The loan did not become effective until October 1994 and it disbursed far more slowly than originally planned. In part, the delays can be traced to the government’s lack of familiarity with Bank processes. In part, they reflect the Bank’s own inability to process smoothly a large procurement of personal computers. By the time the loan became effective, the FES had used its own resources to acquire the computers and staff training, which had been the largest single component of the original loan, and had used a loan advance and other donor funds to begin developing active labor market programs. Also, by then, the Ministry of Social Protection had completed its design and pilot testing of the new pension-processing computer systems and wanted to begin implementing the systems throughout the country. The loan was therefore restructured to focus more on the introduction of active labor market policies in the FES and the comp uterization of the local social protection offices. By this time, also, the government had lost interest in pursuing the large reform agenda contemplated earlier, and the components that were focused on assistance in designing and implementing reforms in the pension and social assistance areas and in developing a master plan for computerizing the entire social security system were dropped. 4.11 Additional procurement delays lead to a second reallocation to further increase support for computerization of social protection offices and several postponements in the original loan closing date. By the time the loan was finally closed in April 2000, three- quarters of the loan had been used for computerization of the social protection offices and one-quarter for the development of active labor market policies. The latter activity had been supplemented by $1.2 million in assistance from eight other donors. 4.12 Coal SECALs. The first of the Bank’s two Coal SECAL loans was negotiated in mid-1996, and the second was negotiated in late 1997. Though technically for budget support, disbursement of each portion of each of the loans was conditioned on the borrower’s meeting of specified conditions, including several pertaining to social protection issues. 4.13 Social protection conditions attached to the first coal loan sought to address each of the problems highlighted in the Bank’s earlier analysis. In particular, the government committed to a specified minimum level of financial support to ensure that individuals received the wage and benefit payments to which they were entitled, that communities had funds for job training and community development, and funds were available for the rehabilitation of the social assets being transferred from the mines. It agreed to establish 13 new mechanisms to assure that these subsidies were put to their intended uses. It also agreed to create an infrastructure for delivering employment services in affected communities and to monitor (and report on) five different indicators of the social impact of mine closing through an ongoing program of surveys of the affected population. 4.14 The initial reform agenda proved to be too ambitious for the time available and four of nine second tranche conditions had to be waived at the end of 1996. Three of the waivers involved social protection issues: (a) the promised targeted subsidies were late and the targeting mechanism was not sufficiently well developed, (b) the social impact monitoring was late getting started, and (c) the community development and job training programs were not ready for implementation. 4.15 The second loan package broadened the reach of the social protection conditions by extending the government guarantee of the payment of severance benefits and wage arrears to miners declared redundant due to downsizing and to those leaving to collect old-age or disability pensions. The government also agreed to further modifications in the mechanisms established to prevent misuse of the subsidies and to continue the social impact monitoring. Subsidies for rehabilitation of social assets were shifted from the mine closing program to another account in the budget, accompanied by assurances that the allocation would be adequate. 4.16 The government was not able to meet the conditions for disbursement of the second tranche, owing in part to the 1998 financial crisis, and the remaining portion of the second loan was subsequently rescheduled. The revision dropped the conditions related to minimum subsidy payments for social protection purposes in favor of a general requirement that subsidies for “priority” purposes (which included social protection) should constitute at least 65 percent of all subsidies. Requirements relating to subsidies for rehabilitation of social assets were also dropped. The requirements for improving the mechanisms to prevent misuse and continuing the social impact monitoring were retained. 4.17 Through the end of 2000, government payments to finance the mine closing expenses had covered around one-quarter of the amount originally agreed to in the individual mine-closing plans. The shortfalls have been particular severe with respect to the costs of restoring the environment, rehabilitating the social capital, and relocating those whose housing was made uninhabitable by mining. It can be argued that the funding shortfall is the result of the structure of the Bank’s loan. Fully funding the plans for the closed mines would have cost roughly $1 billion, less than the combined amount of the two Coal SECALs. If the loans had actually been used to cover the cost of the sector adjustment, the mine-closing process would have left far fewer social protection problems. 8 8 The Bank’s European and Central Asia region management notes that by October 2001, about two-thirds of the social assets of all mines (operating as well as closed) had been turned over to local communities, presumably after at least minimal rehabilitation. 14 4.18 SPAL and SPIL Objectives. The overall objective of the Social Protection Adjustment Loan was “…. to gradually develop an effective system of income support and poverty relief to meet the needs of a market economy.” This included pursuing poverty alleviation by “…raising and protecting minimum benefits, and through better targeting.” It also wo uld “…help establish a viable social safety net by introducing structural reforms in pensions and welfare programs. 9 ” The SPAL addressed most parts of the Bank’s social reform policy agenda through conditions attached to the disbursement of three tranches of budget support. There was no particular link between the amount of the budget support provided and any additional social protection expenditures required under the loan agreement. In fact, the package of adjustments envisioned in SPAL was supposed to be budget neutral. 4.19 Improvements sought in the pension area included the following: (a) increasing the adequacy of the minimum pension, (b) clearing the arrears, (c) strengthening financial management in general and collections in particular at the Pension Fund, (d) creating the proper legal environment for supplemental pensions, and (e) initiating a systemic reform of the pension system. The loan sought improved benefit targeting through (a) reforming the structure of unemployment benefits, (b) gaining specific authorization for means- testing of children’s allowances, (c) conducting pilots to test several indicator-based methods for implementing targeted assistance, and (d) shifting a portion of the cost-of- sickness benefits to employers. The financial security of social protection benefit promises was to be enhanced by (a)increasing the ability of the Employment Fund to transfer monies to regions with high unemployment and (b) gaining full federal financing of at least a portion of the children’s allowance. Finally, a new labor code was to be adopted that would facilitate worker mobility, particularly out-of-state enterprises. 4.20 The vision of systemic pension reform differed somewhat from earlier Bank recommendations. Earlier studies had advocated a higher minimum benefit, a flatter benefit structure, and higher retirement ages. While the SPAL also called for a higher minimum, it ignored the other two traditional elements, focusing instead on the creation of a three-pillar system containing a fully or partially funded, second pension pillar. The approach contained in the SPAL was consistent with the Bank’s approach to pension reforms elsewhere in the world. 10 It also recognized the political barriers to increasing the retirement age and narrowing benefit differentials. However, it amounted to a prescription for increased aggregate pension spending, at least in the short run, complicating the fiscal challenge of social protection reform. 4.21 SPAL documentation noted several important bene fits in the SPAL package for women. Chief among these was the increase in minimum pension and unemployment benefits, each of which would disproportionately benefit women. The package did not 9 Report of the President on a Proposed Social Protection Adjustment Loan, June 5, 1997, Report No. P-7148-RU, p. 14. 10 World Bank, Averting the Old Age Crisis, New York: Oxford University Press, 1994. 15 address the gender aspects of the rest of the pension reform package, however, some aspects of which could have had a negative impact on women. 11 4.22 The SPIL provided financing for the technical assistance to meet the reform conditions, including (a) building actuarial expertise in the government, (b) developing a public information program, and (c) assisting the Ministry of Labor in the development of comprehensive pension reform. The SPIL package also offered assistance to the government to improve the statistical system that tracked social welfare needs and benefit receipt, and provided much of the financing for the seminars and workshops to acquaint government officials, media representatives, and representatives of other relevant interest groups with international social protection practices and principles, particularly in the areas of pensions, labor law, and social assistance. 4.23 A proposed Pension Reform Implementation Loan would have focused on building the pension administration by financing the acquisition of information technology equipment and helping design key business process and computer systems. Loan processing stalled, however, when the Pension Fund was unable to develop a financial management system acceptable to the Bank. Eventually, the Pension Fund decided to use its own resources to purchase the information technology equipment it wanted and to drop the other tasks envisioned by the loan. 4.24 SPAL and SPIL achievements. The SPAL program was successful in improving the financing and targeting of child allowances, unemployment benefits, and sickness benefits. Minimum unemployment benefits were increased (and the maximum reduced), eligibility was restricted for those with little prior labor force experience, the duration was shortened in certain circumstances, and the employment fund was given greater flexibility to shift funds among regions. Responsibility for administering child allowances was shifted from enterprises to local governments, local governments were authorized to exclude higher-income families and the responsibility for financing benefits for the youngest children was shifted to the federal government. Legislation to limit the size of sickness benefits and shift some of the responsibility for financing them to employers was introduced and appears likely to be enacted. 4.25 SPAL objectives to improve the legal basis of voluntary pensions and eliminate pension arrears were achieved, though the elimination of arrears also required help from the three structural adjustment loans. At least initially, the minimum pension benefit was increased to 80 percent of MESI, the target provided in the SPAL, but the benefit could not be maintained at this level after the 1998 fiscal crisis. The third tranche of the SPAL was renegotiated in late 1998, and more modest conditions were established for the minimum pension benefit which by that time had fallen to less than 50 percent of MESI. 11 A pension reform that tightens the linkage between preretirement earnings and retirement benefits may have a negative impact on women due to their lower average earnings levels and the less regular work careers. Depending on the details of the reform, however, this impact may be offset by other changes being introduced at the same time. 16 4.26 Assessing the impact of the pilot programs and the financial management reviews is more difficult. The targeting pilot programs were concluded as planned, and their results appear to have encouraged greater targeting of many existing social benefits. They have not lead to the adoption of a new program of targeted cash assistance for the working-age poor, however. Barriers to the adoption of such programs appear to include some disagreement about how effective the pilots were and, more importantly, continued reluctance on the part of the federal government to assume the fiscal liability necessary to initiate a new federal-regional program of social assistance. There has also been little effort to enhance the administrative capacity of the local governments that would have to deliver such a benefit, if one existed. 4.27 Financial management reforms appeared to be making some headway at the Employment Fund, but any positive impacts have probably been lost as a result of the abolishing of the Employment Fund late in 2000. It is too early to see any impact of the financial management review at the Pension Fund, since that activity proceeded more slowly than did the parallel activity at the Employment Fund. The value of the review at the Pension Fund is also undermined by the change in its role that was also legislated at the end of 2000. 4.28 The greatest disappointment in the SPAL/SPIL package was the failure of systemic pension reform. Technically, the SPAL only required that the government agree to a reform plan and submit draft legislation to the Duma, conditions with which the government complied. In fact, however, the government was not able to develop a reform proposal that was either politically viable or fiscally sustainable. The failure was due in part to philosophical differences among different parts of the government and in part to reluctance about proposing benefit retrenchments. The philosophical dispute involved the role to be played in the reform by mandatory, individual, funded accounts. The absence of any benefit retrenchments raised serious doubts about the fiscal sustainability of the proposal. Debate about the shape of a pension reform has continued, leading to a new set of proposals being considered by the Duma in 2001, which appear likely to be enacted. They resolve the basic structural debate by combining pay-as-you- go and advanced funded approaches in a benefit struc ture that includes both flat and earning-related benefit elements. There is still no adjustment in the retirement age, however, and it is not clear that the plan will be fiscally sustainable. The Duma also gave preliminary approval to a labor law reform in June of 2001. 4.29 As of October, 2001 much of the anticipated institutional capacity building in the pension area had also failed to materialize. 12 Some was lost when the Pension Reform Implementation Loan died. Other efforts covered by the SPIL either were not successful or were not attempted. These include creation of an actuarial capability in the government, initiation of a public information campaign about the need for pension reform, and general support for policy development in the pension area. 12 It must be recognized that this assessment of the impact of the SPIL occurred prior to the closing of the loan, which was extended from the end of 2001 to the end of 2002. An assessment after the completion of all projects still planned or underway may produce a different conclusion. 17 4.30 SAL I, II and III. All three SALs included conditions, which were indeed met, relating to the clearing of pension arrears. SAL III also included conditions requiring certain reforms in the operation of extra budgetary funds, including regular audits of several of the social protection funds, improved targeting of expenditures of the Social Insurance Fund, and changes designed to balance the finances of the pension fund, all of which were requirements that were consistent with the SPAL. A second tranche condition of SAL III would have required that the collection function of the Employment Fund be transferred to the State Tax Service, a reform ultimately adopted in 2000 and one not entirely consistent with the SPAL’s emphasis on improvement of the financial management of that fund. 4.31 Coordination. The ESSP project benefited from the participation of eight other bilateral donors who assisted in the development of an active labor market capacity in the employment service through donation of technical assistance and materials and financing of study tours. Much of the SPAL agenda was developed in late 1995 and early 1996 when the Bank was working with the IMF to develop conditions to be associated with an Extended Finance Facility. For a variety of reasons, most of the social protection elements eventually were spun off into a separate World Bank loan. The European Union’s program financed technical assistance to the Ministry of Labor for pension reform at the same time as the SPAL was being disbursed, though their efforts were largely independent of the Bank’s, in part due to philosophical differences about the substance of the reform. 5. Development Effectiveness Impact Assessment Counterfactual 5.1 It is quite likely that the socia l conditions in communities affected by coal mine closures would have been substantially worse without the Bank’s intervention, even if the government had not followed through on all of its commitments. Bank intervention also probably substantially improved the economic position of the poorest families by (a) accelerating the trend to targeting of social benefits, (b) increasing minimum pension and unemployment benefits, and (c) improving the financing of children’s allowances and unemployment benefits. Bank intervention through the ESSP produced major improvements in the administration of the pension system, which may well not have occurred if the Russian government had to finance the computer purchases itself. 13 It is less clear whether the capacity build ing activities at the employment service would have proceeded in the absence of the ESSP project, since other donors also supported these activities. 5.2 The Bank could have further improved social conditions in coal mining communities had it tied disbursements under the Coal SECAL more closely to the 13 Unlike the FES and the Russian Pension Fund, each of which used their own resources to purchase computers, the local social protection offices would have had to rely on the resources of their regional governments in the absence of the ESSP loan. 18 government’s financing of its social, environmental, and technical commitments. If even half of the money lent to the government under these two loans had been used to discharge the government’s commitments, local governments would have received most of the financing they had been promised at the time the mine closing plans were approved. 5.3 It is unlikely that any different strategy would have produced success in the reform of either pensions or labor law during the 1990s. In each case, there was insufficient political consensus about the need for reform or the direction that reform should take. Given the political rivalry between the Duma and the Yeltsin Administration, moreover, it is unlikely that any actions taken by the Bank could have produced such a consensus. Outcome 5.4 Relevance. Overall, the objectives of the Bank’s social protection activities in the 1990s were substantially relevant to the development challenges facing Russia. In general, the design of the activities followed closely the results of the policy studies, which generally produced accurate assessments of the problems and challenges facing Russia, although they largely ignored the institutional and political aspects of reform. 5.5 As originally designed, the ESSP project was a highly relevant response to the perceived weaknesses in the employment service in the early 1990s and the need to redesign the social safety net and pension systems. The redesign of the project in 1994 caused the objectives to become less relevant, however, by dropping the elements related to redesign of the system and focusing the majority of project resources on computerizing existing processes. 14 5.6 The social protection objectives of the Coal SECAL were highly relevant responses to the Bank’s analysis of the social protection consequences of earlier mine closings. The SALs were highly relevant in ensuring , among other things, that promised pension benefits were actually paid. 5.7 Most of the elements of the SPAL and SPIL were also highly relevant to the social protection issues developed in the Bank’s analytical work and perceived by government officials at the time. In particular, the loan focused attention on improved targeting of existing bene fits, developing of approaches that would allow the creation of a new program of social assistance, and strengthening of the existing social protection program administration. 14 In retrospect, one can question the priority of building capacity to handle unemployment payments and operate active labor market policies during the 1990s in Russia, as the economy in general and the labor market in particular did not develop as originally expected. The designers of the ESSP could hardly be faulted for this, however, as it was not foreseen by any of the policymakers and policy analysts at the time. 19 5.8 Efficacy. As of October 2001, the efficacy of the Bank’s social protection efforts varied widely. The ESSP was only modestly efficacious in meeting its original (highly relevant) objectives, but was substantially efficacious in achieving its revised (but less relevant) objectives. Two of the three major social protection objectives of the Coal SECAL were achieved, but the efficacy of the third, the rehabilitation of social assets, was modest at best. The SPAL achieved objectives related to better financing and targeting of certain existing benefits, but it failed in its objective to initiate major reforms of the pension system or the labor law and many of the institutional improvements envisioned under the SPIL have yet to materialize. The combination of the SPAL and the SALs succeeded in clearing pension benefit arrears. 5.9 Working with other donors, the ESSP project appears to have been substantially effective in building capacity in the employment services and increasing the efficiency of the pension payment functions of the local social protection offices. In part because of the way the project evolved from its original design, however, there was never a clear plan for maintaining the software and updating the hardware as program provisions changed and the machines became obsolete, nor was there a plan to use the installed computers as a mechanism for improving program management or policy analysis capabilities. 5.10 In some respects, the social protection elements of the Coal SECAL were exemplary. It included innovative institutional reforms designed to ensure that wages, severance pay, and disability benefit payments reached their intended beneficiaries and an innovative system for monitoring the impact of the whole set of reforms included in the Coal SECAL package. The approach was highly effective in ensuring the proper implementation of the payment reforms. On the other hand, the project was at best only modestly effective at ensuring that adequate provision was made for updating and maintaining the social assets being transferred from the mining companies. Moreover, although the job creation activities are popular, the early data suggest that the cost per job created is quite high. All in all, therefore, the Coal SECAL should be viewed as moderately efficacious. 5.11 Some of the elements of the SPAL were highly effective, particularly the reforms of child allowances and sickness benefits, and the clearing of pension arrears. The targeting pilots were effective in encouraging better targeting of existing social benefits, but have not been effective in promoting the development of a universal, targeted cash assistance benefit, a result that can be judged as having modest efficacy. Neither the pension nor the labor law reform efforts led to the kind of systemic reform originally envisioned within the time frame originally envisioned. Also, the financial crisis of 1998 showed that, along with other contemporaneous adjustment operations, the SPAL was not successful in ensuring a stable budget and macroeconomic environment. 5.12 Together with the seminars and workshops funded through SPIL, WBI, and Policy and Human Resources Development Fund grants, the SPAL was effective in elevating social protection concerns on the government’s policy agenda. They also educated the elites in the government about the issues that future reforms will have to 20 address and the general options available to them, particularly with respect to pension and labor law reform. 5.13 SPIL resources have been important in achieving acceptance of targeting and in advancing knowledge about pension reform. In other respects, however, the SPIL has been a disappointment. Several of the more attractive projects envisioned in the SPIL have yet to be implemented, including assistance to construct a new household survey, public information about the need for pension reform, and creation of an actuarial capacity to analyze the cost of various social protection reform options. The SPIL is ongoing, but to date its efficacy has been modest, at best. 5.14 Efficiency. The ESSP appears to have been substantially efficient. Bank resources were leveraged effectively through the cooperation of other donors in the employment services component. For a relatively modest investment, the computerization component had a significant impact on the quality of service as measured by the time required to process initial applications or implement benefit adjustments and by the reduction in computation errors. 5.15 Efficiency does not appear a relevant concept for the various adjustment loans with social protection elements, as the size of these loans was not linked in any systematic way to the activities to be undertaken as a condition for receiving the loan. 5.16 Outcome. Comparing the relevance, efficacy, and efficiency of the different social protection operations, it appears that the coal SECAL could be judged satisfactory, the ESSP judged moderately satisfactory, and the SPAL judged moderately unsatisfactory. At the time of the latest Bank review of the SPIL earlier in 2001, it was also moderately unsatisfactory. 15 Although the Bank devoted significant additional resources to producing competent policy analysis work, seminars, and workshops, they did not lead to any major changes in government policy separate from the reforms addressed in the individual loan packages, and the social protection components of the three SALs were not substantial enough to affect the overall rating. The social protection effort over the first half of the decade, therefore, should be judged to be moderately satisfactory, while the effort over the second half of the decade should be judged moderately unsatisfactory. 5.17 Institutional Development Impact . The Bank's efforts in the first half of the 1990s had a substantial impact on institutional development, but its impact in the second half was only modest. The Bank had a substantial institutional impact in the coal SECAL, parts of the ESSP projects, and parts of the SPAL, but to date has had only a modest impact in other aspects, particularly in parts of the SPAL-SPIL package. 15 The Federal Center for Project Finance of the Russian Federation asked Vladimir Gimpelson, a consultant, to comment on an earlier version of this paper. Mr. Gimpelson’s overall assessment of the ESSP was somewhat less favorable and his overall assessment of the SPAL and SPIL somewhat more favorable than this assessment. Mr. Gimpelson’s report is at Annex 4. 21 5.18 The social protection elements in the coal project were critical to the closing of a substantial number of unproductive mines, allowing a significant reduction in budget subsidies and making a significant contribution to the efficient allocation of the country’s financial resources. The ESSP created new programs to improve the operations of labor markets and increased the efficiency of the local social protection offices, not only in delivering pension benefits, but also in handling other workloads that could be programmed onto the pension computers. 16 5.19 The institutional development impact of the various elements of the SPAL-SPIL package varied from high (reform of the children’s allowances and unemployment benefits) to modest (pension reform and labor law reform). The children’s allowance reform produced a more effectively targeted benefit and a new institutional arrangement to ensure that benefits, financed by direct federal transfers, were paid regularly. The unemployment bene fit reform also produced a more effectively targeted benefit with more sustainable financing. The increase in the minimum pension also produced a modest improvement in the targeting of pension benefits, while the pilot programs appear to have lead to increased targeting of many current social protection benefits. The SPAL can also be credited with advancing the dialogue about pension reform and the understanding of the reform options among the key policy makers in Russia, even if it did not produced a reform consistent with the original timetable. 5.20 Sustainability. Overall, the achievements of the Bank’s social protection efforts are likely to be sustainable. As executed under the ESSP, the sustainability of the computerization effort under the ESSP was open to some question owing to the lack of a plan for financing the replacement of equipment as it aged. The recent reorganization that transfers the pension payment responsibility from local governments to the Russian Pension Fund will probably solve this problem, however. 5.21 The gains from the Coal SECAL are highly likely to be sustained, as the mines are unlikely to be reopened. Similarly, the improvements in children’s allowances and unemployment benefits are highly likely to be sustained. The acceptance of improved targeting for other social benefits is also likely to grow, making the gains from the targeting pilots also substantially sustainable, even if the pilots did not lead to a new program of targeted cash assistance. Prospects for sustaining the systemic pension and labor law reform achievements are unclear at this time, although recent progress has been made on both fronts. 16 Mr. Gimpelson has a slightly more favorable assessment of the impact of the SPAL and SPIL and less favorable impact of the impact of the ESSP. See Annex 4. 22 6. Attribution of Bank and Program Results 6.1 External factors. A major barrier to social protection reform was the poor fiscal condition of the government, in part a product of the poor macroeconomic environment. The 1998 financial crisis had a particularly negative impact. While that crisis was largely the result of poor government policy choices, it was also partially attributable to external factors, particularly the low prices for Russian exports (especially oil) and the financial crisis in Asia. Some have argued that the Bank and the IMF are also partially responsible by continuing to lend to Russia despite the lack of reform of its financial sector. 6.2 Bank performance. The Bank’s performance over the first half of the period can be rated satisfactory, but its performance over the second half was unsatisfactory. The Bank’s non- lending activities made a substantial and positive impact throughout the period. In the early 1990s, Bank analysts developed a policy agenda that guided much of the subsequent reform efforts, and the Bank helped to create new data sources for analyzing social protectio n issues. Later, the bank supplied important analyses of trends in poverty incidence and the target efficiency of the current social protection system, and helped educate Russian journalists as well as policy and political elites about various social protection issues. 6.3 The Bank responded quickly to the economic crisis of the early 1990s with the ESSP loan package. That package was well designed to deal with what then appeared to be the most pressing immediate problem, the expected upsurge in unemployment. It also promised resources to help the Russian government develop its own social protection reform plans, even if the government’s 1992 reform scheduled (and by implication, the ESSP agenda) was overly ambitious. Bank staff were flexible in adjusting the loan package when the delay in implementation rendered the original design moot, producing a positive outcome for the project. The staff can be faulted, however, for possibly being too flexible in allowing the Russian government to drop all elements related to planning for policy and administrative reforms, thereby losing the opportunity to lay a better foundation for subsequent reform efforts. 6.4 On balance, Bank performance in the second half of the 1990s was less successful. Bank performance on the social protection aspects of the Coal SECALs included highly satisfactory design and follow-through in the areas of social monitoring and ensurance of wage and pension payments. Less satisfactory aspects were the unrealistic implementation schedule established in the first loan and the loosening of the commitments under the second loan to rehabilitate social assets. 6.5 The failures of the SPAL can be linked in large part to shortcomings in its design, the major contributor to the generally unsatisfactory assessment of Bank performance in the late 1990s. First, the SPAL time frame was probably too short to allow development of meaningful pension and labor law reform. Each required sufficient time for analysis of the issues, education of the interested parties (including members of the Duma), and development of a consensus about the exact nature of the reform to be undertaken, yet the loan required the government to reach a final decision in each area as a condition for the 23 second tranche, expected to occur six months after loan effectiveness. 17 Second, the terms of the loan failed to guarantee that either pension reform or labor law reform was actually adopted, since in these areas the conditions for tranche release required only that legislation be submitted to the Duma. The government was able to comply with the condition without having developed a politically acceptable (or, in the case of pensions, financially sustainable) reform package. In contrast, the loan required that reforms in the structure of child allowances and the financing of unemployment benefits be enacted as second and third tranche conditions. Third, there appears to have been no provision for implementing any assistance program that might have emerged from the pilots, either as a part of SPAL or as part of a subsequent investment loan. 18 In contrast, SPAL documents do note plans for follow-on lending to support pension reform. 6.6 Supervision under the SPAL-SPIL was satisfactory. Bank staff arranged for substantial amounts of technical assistance to facilitate reform and kept well abreast of the status of implementation. Third-tranche loan conditions were renegotiated when the 1998 crisis made attainment of some of the original objectives impossible, and several other worthwhile goals were added to the package at that time. On the other hand, it is doubtful that the pension reform plan accepted by the Bank was actually fiscally viable, and a more complete analysis of the gender impact of the proposed reforms, particularly the pension reform, would have been useful. 6.7 Borrower performance. The performance of the borrower was generally satisfactory over the first half of the period but unsatisfactory over the second half. The major objective of the original ESSP package was to strengthen the employment service, an objective that was fully met due in large part to the initiative of the borrower. The borrower also helped restructure the loan to gain substantial improvements in the quality of the service delivered by local pension offices. Not all aspects were positive, however. One negative was the dropping of the various planning elements of the ESSP. A second was the implementation of an abrupt change in policy on customs’ duties that slowed the implementation of the ESSP computer installations. 6.8 Unsatisfactory government performance was a contributing cause to most of the major social protection reform failures of the last half of the 1990s. The government failed to maintain a satisfactory macroeconomic environment, leading to the 1998 crisis that undermined a number of reform efforts. It did not follow through on its commitment to fully fund the coal restructuring program, and it funded the assistance targeting pilots only after being threatened with cancellation of the second tranche of the SPAL. The government was not able to develop an acceptable financial management system to allow a pension reform implementation loan under preparation to go forward. Finally, it was not able to organize a policy development process that could produce reasonable pension 17 Since the tranches did not have specific release dates, in principle, second tranche release could have been delayed to give more time for policy development. The possibilities for delay were limited, however, as the loan was to close at the end of 1998. 18 The need for a social assistance implementation project was acknowledged subsequently in the Bank’s report, “Russia’s Social Protection Malaise,” issued in March 1999. 24 reform and labor law reform plans in the time period to which it had committed. In these latter two cases, the government complied with the letter of the loan agreement without complying with its spirit. 7. Agenda for Future Action 7.1 One major lesson to be learned from the Bank’s Russia experience is the need to pay more attention to the political and institutional aspects of the reform process. The Bank’s analytical work provided quite competent and insightful assessments of the social protection needs of the population, the patterns of poverty, and the broad program policy reform options, but it largely ignored institutional and political aspects of reform. The limited progress in actually implementing reform came from the difficulty in building institutions and developing political consensus, not an insufficiency in information about the need for reform or the policy directions that reform could take. 7.2 Bank officials spoke frequently about the need to strengthen the social protection institutions as a part of the reform process, but the Bank never developed a consistent plan for achieving administrative reforms, particularly in the areas of means-tested social assistance and contributio n collections. The ESSP loan and the proposed Pension Reform Implementation Loan did focus on administrative improvements, but the pension implementation loan was never implemented and the ESSP was primarily an ad hoc response to a potentially pressing capacity problem. Moreover, the ESSP was not implemented according to the initial schedule, and the activities it actually financed allowed the computerization of one activity in one set of local government offices, without a plan for how that would fit with the administrative aspects of the rest of the reform agenda. 7.3 Institutional considerations may also justify adoption of second-best policy responses. For example, little progress has been made over the last decade in getting a universal cash assistance benefit implemented, but there has been significant progress in improving the targeting of many existing benefits and increasing minimum pensions and unemployment benefits. The Bank should consider whether an antipoverty strategy is focused on making further improvements in the administrative capacity that would strengthen targeting of the existing social protection programs, would have a better chance of success than continuing the effort to create a new assistance program. 7.4 A second lesson is the need to have reasonable expectations about the pace at which a society can build new public institutions and adopt new social attitudes. The social protection aspects of the coal SECALs, particularly those involving community development and refurbishing of social assets, probably could never have been fully implemented within the time frame originally contemplated for that loan. Similarly, developing the political consensus required for a major pension reform for a new program of social assistance requires more time than was available between the tranches of a SPAL. 25 7.5 A third lesson is the need to deal with all of the relevant governmental units in developing and executing a policy agenda, particularly in a country with an independent legislature and a decentralized government. During this period of time, the Bank dealt primarily with the executive branch of the federal government. There was little dialogue with regional or local governments or with the Duma. At least in the social protection area, there was also little dialogue with interest groups or outside opinion leaders, in part because these institutions of civil society were in the process of forming themselves. The near exclusive attention paid to the government ministries left the Bank in a position where it was unable to assess the political realism of the proposals it was offering, assist the national government in developing a political consensus behind proposals jointly endorsed by the Bank and the government, or protect local governments from inadequate follow-through on national government commitments. 7.6 The Russian experience also raises questions about the conditions under which budget adjustment loans should be used as a vehicle for social protection reforms. The approach appears to have been successful when used to assist in the implementation of reforms that had already been largely designed and broadly agreed to, such as in the Coal SECAL, or that enjoyed fairly broad political acceptance at the time the loan was negotiated, such as in the area of child allowances. The approach was less successful in areas where there was little consensus at the time the loan was negotiated, such as the reforms of pensions, labor law, or cash assistance. It was also not successful where important conditions were subsequently waived, as with the social asset spending under the Coal SECAL or the minimum pension under the SPAL. 7.7 Some reforms require short-range expenditures in order to achieve long-term efficiencies. The coal sector reforms are one example and implementation of pension reforms represents potentially another example. In these situations, a loan to help cover the one-time adjustment cost seems entirely appropriate. The Bank needs to disburse on a schedule that allows it to make sure that the government follows through on its commitments, however. 7.8 More care is needed in structuring an adjustment loan that attaches conditions about social protection reform to money that will be used for entirely different purposes, however. Such a loan has several potential positive impacts. Its existence gives the government an incentive to proceed with the reform effort and gains the Bank a seat at the table and some influence over the reform agenda. In the past, these loans have tended to run into trouble, however, when subsequent developments make the original set of conditions or the original time frame inappropriate, because they do not have sufficient flexibility to adjust to changing circumstances. 7.9 The Bank should look for strategies that build greater flexibility into its lending operations and make sure that it does not pay too high a price for the benefits obtained. One promising approach that provides greater flexibility is the division of the loan into smaller, floating tranches, as was done in the restructured Coal SECAL II. In this way, failure to meet one of the major conditions (such as the failure to maintain the minimum pension in 1998) does not block disbursement of the entire tranche. Each of the major 26 elements of the reform program can proceed at its own pace when unforeseen events occur. Another promising approach is program lending, under which the Bank would commit to one or more follow-on loans of appropriate size and structure, but not impose an ex-ante schedule of specific actions at the time the first loan is negotiated. 7.10 Although an existing loan can buy a seat at the table, the Bank should be sure that it does not overpay for that privilege. The Bank has many resources to offer countries contemplating reform, including superb analytical talent to assist in diagnosing problems, developing policy options, and developing systems for monitoring policy implementation; resources to educate policy elites about policy issues and the consequences of choosing difference options; potential future investment loans to strengthen social protection institutions; and future adjustment loans to deal with modest adjustment costs. If these are not sufficiently attractive by themselves to gain the Bank a seat at the table, the client is likely not ready to undertake a serious reform effort. 7.11 Where there are less controversial elements of the reform package, such as the SPAL reforms in unemployment financing or child assistance targeting, an adjustment loan with conditions may be an effective instrument to advance reform. Where reform will require a more extended process of development and discussion, a more flexible approach is likely to be more effective. 27 Annex 1 Economic and Social Indicators, 1991-1998 Table 1. Economic Indicators Unemployment Rate (%) GDP Growth CPI Growth Year (% per year) (% per Total1 Registered2 year) 1991 - 5.0 93 1992 -14.5 2509 4.7 0.8 1993 - 8.7 874 5.5 1.1 1994 -12.7 308 7.4 2.3 1995 - 4.1 198 8.5 3.3 1996 - 3.4 48 9.6 3.6 1997 + 0.9 15 10.8 1.9 1998 -4.9 28 11.9 2.9 +3.2 86 12.7 1.8 1. ILO concept, from household surveys; 2. Registered with the employment service. Source: IMF Russia Country Report, 2000 (GDP and CPI, 1993-99): Goskomstat, Russian Statistical Yearbook, 1994 and 2000 (unemployment rates, GDP and CPI 1991-2) Table 2. Social Indicators, 1991-1999 Ratio, Household Poverty Year Ratio, Average Ratio, Ratio, Rate (%) Average Pension to Average Minimum RLMS Wage to Average Pension to Pension to Goskom- (World MSI1 Wage MESI2 MESI2 stat Bank) 1991 na 0.36 na Na 11.7 1992 3.2 0.27 1.2 0.85 33.5 26.8 1993 2.8 0.34 1.4 0.78 31.5 36.9 1994 2.5 0.36 1.3 0.66 22.4 37.6 1995 1.8 0.40 1.0 0.48 24.7 41.1 1996 2.1 0.38 1.2 0.73 22.1 43.2 1997 2.3 0.35 1.1 0.76 20.8 1998 2.1 0.38 1.1 0.67 23.8 49.0 1999 1.7 0.28 0.7 0.40 29.9 Na = not available; 1. Per Capita Minimum Subsistence Income; 2. Per Capita Minimum Elderly Subsistence Income. Sources: Goskomstat, Russian Statistical Yearbook, 2000 and Social Conditions and Living Standards of the Population, 1999; World Bank, “Poverty in Russia” and “Targeting the Long-term Poor.” 28 Annex 2 Sectoral Assistance Strategy Matrix, 1992-2000 Strategic Objectives Strategic Actions Bank Assistance Process Indicators Evaluation of Assistance FY 92-94: FY 92-94: FY 92-94: FY 92-94: FY 92-94: In its first County Assistance Redesign social safety net to Lending: Employment Elements involving the Broader objectives involving Strategy (CAS), the Bank’s provide more adequate Service and Social Protection Employment Service redesign of the social safety primary focus was the coverage for the poor, more Loan computerization and redesign net: Unsatisfactory establishment of a market effective targeting of all of the safety net were economy. In that context, benefits, and a sustainable Economic Sector Work dropped. one of four medium-term pension system. (ESW): Analyses of social Technical assistance to objectives was strengthening protection system; 39 model offices; training of Employment Service: the social safety net to protect Strengthen institutions that development of more 242 vocational education Satisfactory those harmed by the deliver social benefits, adequate measurement of trainers, 200 job club leaders, economic reform process. particularly the employment poverty rate. 60 trainers of entrepreneurial Subsequent CASs maintained service. skills. the emphasis on an adequate Computerization of local social protection system in Provision of 14,000 social protection offices: order to preserve support for workstations to computerize Satisfactory economic reform. local social protection offices in 2,266 rayons. FY 95-96: FY 95-96: FY 95-96: FY 95-96: FY 95-96: 1. Moderate impact of 1. Institution building [Community Social [Rehabilitation of education, transition on vulnerable (Primarily FES) Infrastructure Project] health, water supply and groups 2. Improve targeting of sanitation facilities in two 2. Maintain social service social welfare expenditures [Health Services Project] Oblasts] infrastructure 3. Improve Federal/local 3. Promote more flexible fiscal relationships; establish ESW labor markets minimum standards for local poverty alleviation efforts 29 Annex 2 FY97-00: FY97-00: FY97-00: FY97-00: FY97-00: 1. Fundamental reform of the 1. Eliminate pension and 1. Social Protection 1. Pension and 1. Eliminate wage and social safety net to facilitate wage arrears Adjustment Loan unemployment benefit arrears benefit arrears: Satisfactory labor market transitions 2. Maintain adequate 2. Social Protection eliminated 2. Better targeting to minimum pensions and Implementation Loan 2. Minimum pension equal to 2. Improve targeting of maximize social protection unemployment benefits 3. Coal Sector Adjustment 40% of MESI in 1999; current benefits: Satisfactory impact within current 3. Protect communities for Loans minimum unemployment resource constraints the impact of coal mine 4. Structural Adjustment benefit raised to 20% of 3. Protect communities from closings Loan III regional MSI. the impact of coal mine 4. Rewrite labor code 5. Pension Reform 3. Full wage and severance closings: Moderately 5. Build more effective Implementation Loan (never payments to over 90% of Satisfactory pension institution effective) redundant coal miners; 6. Improve targeting of 6. Northern Restructuring significant shortfalls in 4. Reform pensions: current assistance programs Loan (in preparation) allocations for rehabilitating Unsatisfactory and create new program of and maintaining social assets. cash assistance 4. Labor code not revised. 5. Create new program of 7. Improve financial 5. Individual pension cash assistance: management and public accounts initiative completely Unsatisfactory sector administration implemented 8. Cost-effective assistance 6. Targeting pilots to the north completed; increased targeting of current benefits; few new cash assistance programs; children’s allowances targeted on low- income households and payments are current 7. Financial management reviews completed. 30 Annex 3 List of Persons Interviewed Russian Federal Government Mr. Valery Janvariov, Deputy Minister of Labor and Social Protection Mr. Alexander Zhukov, Chairman, Budget Committee, State Duma Mr. Boris A Dudenkov, First Deputy President, Social Insurance Fund Mr. Grigory Y. Glazkov, Head of Department, Ministry of Finance Mr. Mikhail Dmitriyev, Deputy Minister of Economy Dr. Evgueni Gontmacher, Head of Department of Social Development, Aparat Mr. Mikhail Lopatin, Advisor on Social Affairs to the First Deputy Prime Minister Mr. Vladimir Zinin, Chief of Department, Ministry of Labor and Social Protection Ms. Ludmila Rau, Chief of Department, Social Insurance Fund Ms. Loubov Eltsova, Deputy Head, Federal Employment Services Mr. Anatoly Kolesnik Deputy Chairman, Russian Pension Fund Ms. Natalia Petrova, Head of Department of Accounting, Russian Pension Fund Mr. Vladimir Dubrovski, Deputy Chairman, Social Insurance Fund Tula Oblast Government Mr. Yury Pavlovich Semyonov, Director of the Tula Regional Social Insurance Fund Mrs. Natalia Vyacheslavovna Nikolayeva, Deputy Director of the Tula Social Insurance Fund Mr. Alexander Petrovich Rybalchenko, Deputy Director of the Department for Social Protection of Population, Tula Oblast Administration Mrs. Tatiana Nikolayevna Bukolova, Head of the Department for Automatic Communication Systems, Department for Social Protection of Population. Former Government Officials Mr. Dmitry Vasiliev, Former Deputy Minister for Privatization Mr. Yegor Guidar, Former Prime Minister Mr. Sergei Kalashnikov, Former Minister of Labor and Social Protection Mr. Oleg N. Sysuev, Former Minister of Labor and Social Protection and Deputy Prime Minister Russian Private Sector Mr. Igor V. Kolosnitsin, Institute for the Economy in Transition Mr. Mikhail Shmakov, President, Federation of Independent Trade Unions of Russia Mr. Igor Shanin , Secretary of Federation of Independent Trade Unions. 31 Annex 3 World Bank Mr. Andrei Darusenkov, Moscow Office Mr.Vadim Voronin, Moscow Office Mr. Andrei Markov, Moscow Office Mr. Joseph Procak, Russian Anchor Unit Mr. Tim King, retired Europe Central Asia Human Development Sector Unit (ECSHD) Ms. Anastossia Alexandrova, Moscow Office Ms. Elena Zotova, Moscow Office Mr. Andre Markov, Former Head of Social Protection, Moscow Office Mr. Michael Carter, Former Head of Moscow Office Ms. Kathryn Dahlmeier, Task Team Leader, ECSHD Mr. Hjalte Sederlof, Former Project Director, ECSHD Ms. Mansoora Rashid, Protect Director, ECSHD Ms. Jeanine Braithwaite, Senior Economist, Poverty Reduction and Economic Management Sector Unit (ECSPE) Ms. Donna Edgerton, Information Technology Consultant, ECSHD 32 Annex 4 Comments on Lawrence Thompson's Evaluation Report By Vladimir Gimpelson Consultant FCPF Summary The Federal Center for Project Finance of the Russian Federation asked Vladimir Gimpelson, a consultant, to comment on an earlier version of this paper. Mr. Gimpelson agreed with much of the analysis of the report, including the importance of the political and economic environment in influencing the Bank's social protection projects. Mr. Gimpelson's differed somewhat with the assessment presented here of the Bank's two major social protection loans during this period. His assessment of the Employment Services and Social Protection Loan was less favorable than that presented in this report, while his assessment of the Social Protection Adjustment Loan and the associated Social Protection Implementation Loan was more favorable. With respect to the ESSP loan, Mr. Gimpelson notes that both the Bank and the borrower should have recognized that the emphasis on active labor market policies was misplaced prior to the effectuation of the loan in 1994, since it was clear by then that the Russian transition was not going to lead to rapid and significant increases in unemployment in a manner similar to the experience in Eastern Europe. He also argues that the decision to drop the pension reform components from the loan robbed it of most of its potential for institution building. With respect to the SPAL, Mr. Gimpelson faults the Bank for establishing what it should have known was an unrealistic timeframe for the pension and labor law reforms. He faults the government for agreeing to undertake these reforms when, prior to the 1998 crisis, development of a consensus around either reform was not possible. He notes, however, that the SPAL process has had a positive longer-term impact, leading to action in the Duma in 2001 that is likely to produce reforms in both areas. His more favorable assessment of the SPAL-SPIL effort is primarily due to this longer term impact. All of Mr. Gimpelson's points are valid, and his assessment presents a reasonable, if modest, alternative to the view presented in this report. Each of these projects had multiple objectives, some of which were achieved fully, some of which were achieved partially, and some of which were not achieved at all. Any assessment of the overall record necessarily involves implicitly weighting the different elements, and slight changes in one's assessment of the relative importance of the various components can lead to a modest alternation in the resulting overall assessment. This accounts for much of the difference between Mr. Gimpelson's views and the view represented in this report. In addition, this report is being completed at the end of October, 2001, at which time it is still too early to know how successful the current labor law and pension reform efforts will prove to be and therefore how much credit the SPAL-SPIL package should be given for advancing the debate around each. Comments on Lawrence Thompson's Evaluation Report The main purpose of these comments is to contribute to a thorough discussion between the World Bank and the Russian Federation (RF) Government over Russia's social policy in the 1990s. This text contains an analysis of the evaluation report prepared within the Bank's Country Assistance Evaluation by World Bank Consultant L. Thompson. 33 Annex 4 Mr.Thompson bases his findings on a review of project appraisal and completion reports, sector reports, research papers in the academic literature, and a number of other documents produced by the Bank and the Borrower. Overall, Mr Thompson has presented a thorough, professional analysis covering all major aspects of co-operation between the Russian Federation and the Bank in the social protection area. The author considers in sufficient detail the design of specific loans, their objectives, tasks, implementation conditions and constraints. Successes and failures in projects implementation are discussed in connection with the general political and macroeconomic situation in Russia in the 1990s. Mr. Thompson's Report focuses on the following projects: the Employment Services and Social Protection Loan (ESSP), the Social Protection Adjustment Loan (SPAL), and the Social Protection Implementation Loan (SPIL). Relatively less attention is paid to the Coal Sector Restructuring Implementation Assistance Project. The Regional Social Infrastructure Support Project was left beyond the scope of the discussion in Mr. Thompson’s Report. In project evaluation the author emphasizes standard OED criteria as follows: • Relevance of objectives • Efficacy • Efficiency • Sustainability • Institutional development impact (IDI) • Overall evaluation of the projects. Most of the specific conclusions made in the Report seem to be correct; but, in my view, some need more detailed comments. I believe, however, that a few more general conclusions in Mr. Thompson's text do not seem to be unambiguously derived from the project-specific evaluations. In this case, the logic linking specific and general conclusions could be disputed. The following part of this material discusses the logic and findings of Mr Thompson’s Report. Overview In 1992, the Russian Federation commenced reforms to transition to a market economy. A properly functioning market economy assumes a social protection system that would differ radically from those in socialist planned economies. The social safety net in a market economy is based on a different philosophy, other principles of financing, and another type of distribution of responsibilities between the government and civil society. As Mr. Thompson notes, from the beginning of its co-operation with Russia, the World Bank viewed social protection reforms as an instrumental part of the economic reform program. The underlying logic was simple: radical reforms lead to rising poverty due to emergence of multiple losers. This group of losers includes mostly aged, low skilled, unemployed, disabled, and poorly educated. Further expansion of this group as a result of economic restructuring would undermine political support for economic reforms. Against this background, compensatory measures are needed to protect and support these groups until resumed economic growth improves their living standards. Thus, in the early years, social protection measures were considered mainly a compensatory mechanism. Only later – in the second half of the 1990s – did the Bank’s attitude to social protection reform shift to viewing social protection goals as relatively independent. But even in that period, restructuring of the social safety net was considered an instrument for increasing efficiency of this system, while containing its fiscal 34 Annex 4 pressure. The emphasis was on redistributive mechanisms (e.g. better targeting in social assistance). What is needed to implement a rapid structural adjustment of the entire social security system? It is "only" the combination of three basic conditions: 1. "Technical" conditions, i.e. the presence of a clearly defined comprehensive blue-print describing what must be created and a methodology for achieving this objective. 2. Political conditions for reforms: • A consensus between the major political and social actors in regard of the necessity and content of reforms; • The Government's political will to push ahead implement (often unpopular) reforms; • A political system with a limited number of veto gates, enabling fast adoption and implementation of the social sector reform; and • Efficient social sector reform management (the efficient bureaucracy at all levels of the Government). 3. Availability of financial resources as such reforms are expensive. Given these conditions, it is clear that reforming Russia's social safety net could be neither easy nor fast in the mid-1990s. As far as I understand the substance of co-operation between the Bank and the Russian Government, the Bank sought to facilitate the creation of most of the conditions listed above by providing relevant support. Naturally, the Bank cannot influence most of the political conditions (e.g. the Government’s political will or the structure of political institutions). However, it is necessary for the Bank to consider how these circumstances may affect project implementation. Evaluation of the political conditions for reform should be an integral part of the design of a relevant program and its implementation policy. There were multiple obstacles to implementing the structural adjustment of Russia's social protection system in the 1990s. Among them were: technical difficulties in building a new social protection system (including the lack of the experts' consensus as to what it should look like); the unfavourable macroeconomic environment; redistributive consequences of structural reforms in this sector,19 politico-economic constraints; and the financial costs of the reforms. These constraints illustrate the challenges that have to be taken into account in the process of initiating and implementing joint co-operation projects by the Bank and the Borrower. A thorough and realistic evaluation of these challenges is a prerequisite for successful completion of the projects. Whether and to what the extent these challenges were taken into account by the bank and the borrower should be evaluated in order to facilitate future co-operation. Evaluation of Individual Projects in L. Thompson's Report 1. The Employment Services and Social Protection Project (ESSP) 19 This results in new additional political restraints 35 Annex 4 The loan to finance the Employment Service and Social Protection Project was the only direct investment loan. It was originally designed to provide technical assistance in developing the capacity of the Employment Service. This should ensure a proper response to an anticipated rise in unemployment claims as a result of industrial restructuring, Later, the major objective was changed to focus mostly on computerisation and upgrading the data processing capacity of local social protection offices dealing with pension payments. The outcome of this project has been rated by Mr. Thompson as modestly satisfactory. There are, however, arguments for a more critical assessment. Moreover, the discussion of the results of this project in Mr. Thompson's Report gives little support for this satisfactory rating. Based on Mr. Thompson's description of the project, I would rate it as unsatisfactory rather than satisfactory. Possible criticisms relate both to the principal elements of the project’s initial design and content and to its actual outcome. The loan agreement was signed in April 1994, with a loan closing date of April 1996. The project completion date was extended more than once. As a result, it was actually closed only in 2000. Also, the objectives and tasks of the project were changed considerably, part of the loan funds was cancelled, and part of all expenditures was financed from other sources. The project was initiated in 1992 with little reliable labour market data on hand. The Staff Appraisal Report was dated October 1992, the same month in which the first labour force survey based on an internationally compatible methodology was conducted by the RF Goskomstat. Survey results became available late in 1993. The project was therefore designed mostly on the basis of rather general assumptions and extrapolation of developments seen by that time in Central and Eastern European countries. A transformational recession and potential structural reforms were expected to increase unemployment. This could bring massive poverty and, as a consequence, the loss of political support for reforms. In order to mitigate these consequences it was necessary to provide unemployed with temporary cash benefits and to set up retraining facilities. Aged and unemployed people were considered as the group most vulnerable to poverty. By the end of 1992, however, it was already quite clear that the initial unemployment expectations were incorrect, and that the major developments in Russia would differ from those observed in Poland or Bulgaria. The “Russian way of labour market adjustment” did not lead to a rapid rise in joblessness (at least, in the number of actual benefits claimants). A significant portion of the poor would be those who were still employed. Predictions of unemployment growth made in the first half of the 1990s proved to be absolutely incorrect. Though the OECD/ILO type of the unemployment rate grew steadily, that of registered unemployment remained extremely low throughout the 1990s. The reasons for these divergent trends were numerous and their analysis is out of the scope of this paper. However, it was the registered or claimant unemployment was the main concern of the employment service. The central elements in the Employment Service Component of the ESSP project were designed to improve the structure of labour supply (with emphasis on introducing active labour market programmes into the employment service), develop job brokerage capacity, and ensure unemployment benefits administration. The authors of the project probably believed that improving (through retraining) the structure of labour supply and enhancing job brokerage capacity of local employment centres would reduce the level of unemployment or maintain it at a relatively low level. As a matter of fact, the main problems associated with unemployment in 36 Annex 4 Russia were (and still are) not in the supply of labour, , which proved to be rather flexible 20 , but in labour demand, which remained very low throughout the ‘90s. Russia's economy, experiencing prolonged and deep recession, created few new formal jobs, but destroyed old ones at a rather high rate. Besides, it was hardly possible to rapidly organize large-scale retraining facilities as the demand structure was unclear, and both trainers and jobs were in short supply. In addition to the extremely sluggish job creation throughout the period under review, the employment service had a very low penetration rate, controlling one tenth of all vacancies. Thus, the active labour market programmes were doomed to be inefficient. It is noteworthy that so far there has not been any ALMP evaluation conducted. The issue of increasing labour demand was never discussed within the framework of World Bank-funded social sector support initiatives. A significant positive result of the project was the enhanced institutional capability of the Employment Service units. They implemented modern technologies and IT solutions, built a network of model offices, started to develop active labour market programmes, set up analytical departments within their organizational structure, created data collection facilities, etc. However, evaluating the efficacy and sustainability of this project is not easy. 21 First, actual trends in registered unemployment had little to do with any forecasts. Second, the unprecedented gap between the levels of registered and surveyed (ILO type) unemployment indicated crucial weaknesses of the registration system throughout the period. Third, the difficulties (and arrears) in paying unemployment benefits emerged in the mid-'90s put into question, along with some other factors, the institutional capacity of the employment service to provide income support. Of course, all this was caused by a variety of factors, and the employment service's capability could have been much weaker if this project had not been implemented. The ESSP pension system component was aimed primarily at upgrading the data processing facilities at the local offices responsible for paying pensions. To achieve this, the component included provision of assistance in implementing automated data processing systems in three pilot regions and preparing a comprehensive plan for developing the automated pension payment system. Initially, it was intended to draw up proposals on reforming the pension policy and to work out social assistance measures. These strategic tasks, however, were later cancelled, and emphasis was placed on the computerization of pension data processing. Within the limited technical framework (procurement of computers for the local offices in the pilot regions), it should be admitted that its tasks were implemented. However, it is difficult to say to what extent this has influenced the institutional development of the country’s pension system. The institution-building tasks were excluded from the project, and were discussed later within the framework of the SPIL and SPAL projects. Furthermore, the paradigm of pension reforms implied in the early ‘90s changed significantly by the second half of the ‘90s. Due to various circumstances (some of which are discussed in Mr Thompson's Report), the tasks and objectives of this loan underwent multiple revisions. As a result, the gap between the project's initial strategic objectives and its actual technical content expanded. The sum of the 20 Very intensive labour turnover, including flows in and out of labor force, flows to education and training, occupational changes, etc. 21 It should be also said that the creation of local employment centres was financed by bilateral donors as well, but some leverage effect due to the WB involvement did exist. 37 Annex 4 revised individual objectives does not make it possible to reconstruct the project's initial strategic priorities. The schedule of the project was also broken. Its start-up was scheduled for 1993 and completion for 1994, but it was actually launched only in 1994 and completed in 2000. The objectives and tasks were changed; of the total amount of loan proceeds of US$ 70 million an amount of US$ 10 million was cancelled; and a significant part of procured items was financed from other sources. In my opinion, the responsibility for the above listed weaknesses of the project should be shared by the Bank and the Borrower. They relate both to the project’s design and its administration. Indeed, the initially ambitious objectives finally shrank to purely technical (though important!) tasks, and a series of failures that led to loan completion date extensions, the inability to meet the initial deadlines, etc. It should be noted also that at that time researchers (both foreign and domestic) had neither adequate knowledge of how the Russian labour market was working nor reliable and representative data. This project could have initiated more active research on social protection issues, but in its initial phase the Borrower was not ready to engage in detailed discussions of the existing problems and challenges. Adequate information on Russia’s labour market and social protection issues as well as related analytical studies was provided later. The last point might be an argument in favour of the project's relevance, but not of its potential effectiveness. The foregoing makes me to be more cautious (in comparison with Mr Thompson's evaluation) in assessing this project. 2. Social Protection Components in the Coal Sector Adjustment Loans (Coal SECALs) Mr. Thompson’s Report pays comparatively less attention to analyzing and evaluating the Coal Sector Adjustment Project. Though the related loans were earmarked for replacement of budgetary revenues, their disbursement was conditioned on the Borrower’s meeting of specific requirements, including the element of social protection in the coal mining regions. In particular, the government assumed an obligation to provide financial support to ensure that that wages and social protection benefits were paid to displaced miners, that communities had funds for job training and community development, and that social assets were rehabilitated before being transferred from closed mines to local municipalities. Among the conditions accompanying the loans was the provision of assistance in delivering employment services in the communities affected by mine closures and monitoring of five different indicators of the social impact of mine closing. Overall, Mr. Thompson positively evaluates the results of the social protection programme linked to the Coal SECALs. Basing on project design and implementation documents, I share this assessment. However, on the whole, the institutional impact of this sector project in the social protection area was limited. 3. The Social Protection Adjustment Loan (SPAL) and the Social Protection Implementation Loan (SPIL) These two loans can be reviewed simultaneously as SPIL was intended to finance measures in order to meet the conditions of the SPAL. Mr Thompson rates the SPAL-SPIL project implementation as moderately unsatisfactory rather than satisfactory. There is a strong basis for this negative assessment as, from the formal side, a number of the most significant conditions of the SPAL (submission of draft pension reform legislation and a new labour code to the Duma) were not met as scheduled. The question is, however, whether these conditions could 38 Annex 4 have been met as planned, and whether they were fulfilled later. Two more questions are whether the Bank and the Borrower could have foreseen this failure and why they agreed to the obviously unrealistic conditions. Tasks and objectives of the SPAL and SPIL. The SPAL was generally designed to “gradually develop an effective system of income support and poverty relief to meet the needs of a market economy.” This included pursuing poverty alleviation by “…raising and protecting minimum benefits, and through better targeting.” It also would “…help establish a viable social safety net by introducing structural reforms in pensions and welfare programmes.” The SPAL addressed a whole set of components of the Bank’s social reform policy agenda through conditions attached to the disbursement of three tranches of the loan to finance budget deficit. The activities envisioned in the SPAL were to cover virtually all areas on the social reform policy agenda.22 This social protection agenda was too far-reaching and overly ambitious, more so as it was to be implemented within a very short period, from June 1997 through December 1998, under the original agreement. Project Outcomes and Their Assessment As Mr. Thompson states, the major part of conditions scheduled under the SPIL-SPAL loans was fulfilled. However, the most complicated elements of the reforms envisioned therein – pension reform (transition from PAYG to a three-pillar system with a funded pension component) and liberalization of labour law – were not implemented within the time frame originally contemplated. These failures explain the negative assessment of the SPIL-SPAL project in Mr Thompson’s Report. One cannot dismiss this criticism related to 1997-98 time period. The Government was not able to reach agreement on these reforms and present new agreed legislation in a timely manner. However, at the present moment, these conditions seem now to have been met. Looking back from 2001, we could raise serious doubts that social reforms as politically and technically complex as these had any chance to be implemented under the SPAL-SPIL schedule as planned. Having world-wide experience in social sector reform, the Bank could, and probably should, have predicted most of the barriers to project implementation. 23 Consequently, the Bank should fully share with the Borrower the fault for the unjustified high expectations. It certainly does not mean that the RF Government, with its priority on covering the federal budget deficit in 1997 and early 1998 should not share the blame. At that time, the Government was preoccupied much more with financing budget expenditures, clearing pension and wage arrears and paying benefits and allowances. Deep restructuring of the social protection system was clearly out of its short-term agenda. The Bank’s underestimated the time factor for this extremely ambitious reform project and Mr Thompson fully admits this fact. However, he clears the Bank from this fault putting most of the blame on the Borrower.24 Analysing today (in early July 2001) the history of social protection reforms in Russia in the mid-1990s, one should admit that radical reshuffling of the pension system and labour law were premature before the 1998 crisis. Nether neither the Government nor the general public was 22 See the detailed list of conditions in the Loan Agreement. 23 There is a vast literature discussing the political economy of the WB initiated social sector reforms in different countries. 24 Note: This comment reflects the draft upon which Mr. Gimpelson was commenting. The assessment has since been modified. 39 Annex 4 ready to accept them. Assume, moreover, that by an accident, the Government had completed their implementation before the 1998 collapse, say, in late 1997 or in early 1998). Then, the post- crisis after-effects could have been much more adverse, including drastic devaluation or loss of a part of the assets of newly created pension system or a dramatic increase in the level of unemployment. 25 In that environment, new social protection institutions might have proven not to be viable and efficient enough. As a result, the reforms that had just been implemented in a hurry might have turned out to be unsustainable and rejected for years to come. Ultimately, the 1998 crisis contributed to Russia’s macroeconomic stability and partial recovery in output, thus creating more favourable conditions for the social reforms under review. Failing in their crucial components in the short run, SPAL and SPIL have had longer- term positive impact. They gave strong impetus for preparing the pension and labour law reform in Russia. SPAL-SPIL projects initiated more domestic research, more expert and public discussions, and finally contributed to turnaround in the public opinion towards reforming the social protection institutions. The macroeconomic and political stability in Russia enjoyed in 2000-2001 opened a new window of opportunity for more radical solutions. Ultimately, all this resulted in the agreed drafts of pension and labour laws submitted to the State Duma in 2001. In this sense, the 2-3 years delay in the pension and labour law reforms has turned out to the better. Political Economy of Russia’s Social Protection Reform Mr. Thompson’s Report pays much attention to the political economy of the social protection reform. This is a strong point of the report. The Report concludes that ignoring the political economy reduces chances for success. This relates to the issues like the pace of reforms, relationships between agencies within the government and between branches of power, co- ordination and collective action problems, configuration of veto gates, redistributive conflicts, electoral politics, and so on. Even the most perfect design of the reform can be seriously undermined by various institutional and political factors. Reminding this is important because it is mainly political economy constraints that stalled or reversed Russian reforms in the 1990s. The point made in the Report that the Bank has largely ignored these issues seems very plausible. Overly ambitious programmes and unrealistic time schedules resulted in frequent rescheduling and cancellation of initial agreements. In my opinion, the necessity of paying more attention to these issues must be one of the most significant lessons for the future learned from the past experience in the co-operation between the Bank and the Russian Federation. Conclusion I believe that Mr. Thompson has offered an objective assessment of the most important social protection projects. Still, it seems to me that Mr Thompson’s overall evaluation of the ESSP and SPAL-SPIL projects could be disputed as not always being derived from his project- specific estimates. The overall ratings seem implicitly to reflect subjective and emotional perceptions of the Russian transition among international experts and observers. Thus, the 25 Certainly, one could assume that the timely pension or labour law reform would have prevented the country from the 1998 crisis. Still, it is very unlikely as the financial collapse was caused mostly by macroeconomic mismanagement and fiscal problems and the social protection reform was supposed to be budget neutral. 40 Annex 4 unreasonably high expectations from the Bank at the early stage of the reform process gave way to the equally unjustified disappointment by the end of the 1990s. One can argue that the more technical and specific the project is and the less the institution-building component is, the more likely the project will be implemented smoothly. Assessing the projects’ institutional development impact beyond the narrow time frames of the initial schedules, one could that the actual institutional effect the SPAL-SPIL loans was rather significant, while that of the ESSP was quite limited. If there were no SPIL-SPAL package that initiated intensive preparatory work for the pension and labour law reforms in the late 1990s, the Russian Federation would have been hardly able to have key pension laws and the new labour code approved by the Duma in the middle of 2001. wb157917 O:\PUBs\Working Papers\OED\Country\Russia\Final\SP Thompson v4.doc September 5, 2002 10:30 AM