Report No. 11 748-RU Russia Social Protection During Transition and Beyond (in two volumes) Volume l: Main Report February 2, 1994 Human Resources Division Country Departments IlIl Europe and Central Asia Region FOR OFFICIAL USE ONLY tMICROGRAPHICS Report No: 11748 RU Type: SEC Document of the World Bank This document has a restricted distribution and may be used by recipients only in the performance of their official duties. Its contents may not otherwise be disclosed without World Bank authorization FOR OMCIUL USE ONLY Ackeowledgements This report is based on work which began under the Technical Cooperation Agreement and the findings of a mission in October/November 1992 lid by Timothy King, and the contributions of Jeni Klugman, Nicholas Barr, Lillian Liu, Keith Palmer (social protection); Simon Commander (labor markets and macro-economic policy); Sanjeev Gupta (fiscal analysis); Nonica Fong (women); Richard Jackman (employment policy); Ruslan Yemstov and Leonid Lieberman and Cecilia Ugaz (labor markets); Hannah Matras (housing) and Tanya Proskuryakova (d;a;). The principal author is Jeni Klugman. A draft of this report was given to the Government in Nay, 1993 and was the subject of fiscussions in October 1993. It was subsequently updated, as far as possible, in November 1993. The report has benefited from the constr .tive criticisms and suggestions of the peer reviewers, Emmanuel Jimmenez, Elizabeth King and Nichael Valton. A number of other people have provided valuable background Information and analysis and/or comments, including Victor Gabor, Isabel Guerrero, John Holsen, Robert Hagemann, Yukon Huang, Billy Jack, Kathie Krumm, Peter Lanjouw, Branco Milanovic, Tom Mroz, Barry Popkin, Christine Vallich and Lynette Vardle. Phillip Goldman and Mark Foley assisted in the final stages of revising the report. Rajakumari Stephen and Annie Minofu have provided tireless secretarial support throughout the process. Robert Liebenthal and Russ Cheetbam are the Division Chief and Department Director, respectively. This document ht a retddistribution amdm be ued by rcipients only in the performance of their offcia dutlL Its contents may not othwse be dicosed without World Bank authodzation. Volume I EXECUTIVE SUMNARY . . . . . . . . . . . . . . . . . . . . . . . . . . .i-xvii CHAPTER I: THE CONTEXT . . . . . . . . . . . . . . . . . . . . . . . . . . 1 Macroeconomic Developments . . . . . . . . . . . . . . . . . . . . . . . . 1 Anchors to the Stabilization . . . . . . . . . . . . . . . . . . . . 1 Output and Inventory Changes . . . . . . . . . . . . . . . . . . . . 4 Repercussions. . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 Human Development. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5 Vulnerable Groups. . . . 7 . . . . . . . . . . . . . . . . . . . . . . . . 7 Definitions and Measures of Poverty. . . . . . . . . . . . . . . . . 7 Development of the Minimum Subsistence Income. . . . . . . . . . . . 9 Findings of the Longitudinal Monitoring Survey . . . . . . . . . . . 10 CIHPTER II: CHANGING ROLES IN SOCIAL PROTECTION . . . . . . . . . . . . . 19 The Soviet Legacy. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19 The System of Cash Benefits in 1992. . . . . . . . . . . . . . . . . . . . 23 Pension Program. . . . . . . . . . . . . . . . . . . . . . . . . . . 23 Unemployment benefits. . . . . . . . . . . . . . . . . . . . . . . . 28 Sickness Benefits. . . . . . . . . . . . . . . . . . . . . . . . . . 31 Maternity Benefits . . . . . . . . . . . . . . . . . . . . . . . . . 34 Family Allowances. . . . . . . . . . . . . . . . . . . . . . . . . . 35 social Assistance Programs . . . . . . . . . . . . . . . . . . . . . 36 Nongovernmental Initiatives in Social Protection . . . . . . . . . . 41 conclusions. . . . . . . . . . . . . . . . . . . . . . .. . . 44 CHAPTER III: POLICY ISSUES AND OPTIONS IN SOCIAL PROTECTION . . . . . . . 51 Strengthening the Safety Net . . . . . . . . . . . . . . . . . . . 51 Guaranteeing Social Protection for Everybody . . . . . . . . . . . . 53 Ensuring that Social Protection Programs Protect . . . . . . . . . . 57 Guaranteeing the Minimum Level of Unemployment Compensation. . . . . 59 Financing the System of Social Protection. . . . . . . . . . . . . . . . . 60 Reform of Financing of the Employment Service. . . . . . . . . . . . 60 The Continuing Role of Eztrabudgetary Funds and Payroll Taxation . . 60 The Nature and Uses of Extrabudgetary Funds. . . . . . . . . . . . . 61 Options for Change: Intergovernmental Fiscal Relations . . . . . . . 62 The Role of Central Transfers. . . . . . . . . . . . . . . . . . . . 63 Seeking Cost Savings Wherever Pos sbble . . . . . . . . . . . . . . . . . 66 Reverting Temporarily to Uniform Adjustments to Pensions . . . . . . 66 Reducing Eligibility of Working Pensioners . . . . . . . . . . . . . 67 Reforming Sickness Benefits . . . . . . . . . . . . . . . . 67 Reforming Family Benefits. . . . . . . . . . . . . . . . . . . . . . 68 Assessing Fiscal Implications of Reform Options. . . . . . . . . . . . . . 69 Pension reform ............... .......... 69 Social Insurance Fund Reform .. . ................ . 70 Employment Fund reform . . . . . . . . . . . . . . 70 Introduction of a Minimum Cash Benef't . . . . . . . . . . . . . . . 70 Anticipating Fiscal Developments Beyond 1992 . . . . . . . . . . . . 71 Shaping for the Long-Term Social Protection. . . . . . . . . . . . . . . . 72 The Economic Issue: Limit the Aggregate Cost of the Pension System. 73 The Political Economy of Entitlements: Firmer Assurances for Individuals . . . . . . . . . . . . . . . . . . . . . . . . . . . 74 The Role of the Penslon Trust Fund . .... . . . . . . . . . . . . 76 The Role of the Private Sector in Social Protection. . . . . . . . . 79 The Role of Priva:e Pensions . . . . . . . . . . . . . . . . . . . . 80 Reforming Sick Pay . . . . . . . . . . . . . . . . . . . . . . . . . 83 Reforming the Administration of Cash Benefits. . . . . . . . . . . . . . . 87 Revenue Collection . . . . . . . . . . . . . . . . . . . . . . . . . 87 Integration of Collection and Payment Activities for Pensions. . . . 88 Strengthening Administrative Capacity. . . . . . . . . . . . . . . . 91 CHAPTER IV: AN ANALYSIS OF RECENT LABOR MARKET DERELOPMETS . . . . . . . 94 Labor Compensation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 96 The Importance of Non-wage Benefits. . . . . . . . . . . . . . . . . 97 Structure of Labor Income and Recent Trends. . . . . . . . . . . . . 100 Employment ......................... 104 The Pre-Reform Context ...................... . 104 Recent Changes in Employment in 1991 and 1992 . . . . . . . . . . . 106 Unemployment . . . . . .......................... 110 Conclusions. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113 Policy Implications: The Potential Role for an Incomes Policy in Russia. . 115 Arguments for a Wage Policy In Russia. . . . . . . . . . . . . . . . 115 Drawbacks and Problems with a Wage Policy. . . . . . . . . . . . . . 116 Experience . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117 Recommendations. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 118 CHAPTER V: IMPROVING THE EFFICIENCY OF THE LABOR NARKET . . . . . . . . . 121 The Role of the Employment Services. . . . ... . . . . . . . . . . . . . . 121 Service Delivery and Staffing. . . . . . . . . . . . . . . . . . . . 123 Training . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123 Kass redundancy. . . . . . . . . . . . . . . . . . . . . . . . . . . 125 Temporary Employment Programs. . . . . . . . . . . . . . . . . . . . 125 Employment Policy and Industrial Closures. . . . . . . . . . . . . . 127 Housing and Labor Xarket Distortions . . . . . . . . . . . . . . . . . . . 131 The Housing Sector--Past and Present ............... . 132 Now Do Real Estate Markets Work? . . . . . . . . . . . . . . . . . . 133 Linkages Between Housing and Labor Narkets . . . . . . . . . . . . . . . . 136 Labor Turnover Rates: TooLow? ........... . 136 Internal Migration . . . . . . . . . . . . . . . . . . . . . . . . . 136 Have Migration Rates Been Depressed? ... ....... . . . . . . 138 Enterprise Housing Benefits and Labor Nobility . . . . . . . . . . . 139 Enterprise Housing in the Housing Reform Debate .. . . . . . . . . . 14 Policy Recommendations ....... . .......... .. .. . 142 Volume I2 Annex 1 Tables and Graphs Annex 2 Wage and Employment Decisions in the Russian Economy: An Analysis of Developments In 1992. Prepared by Simon Comander, Leonid Liberman, and Ruslan Yeatsov Annex 3 Russia: The Role of Yomen In Rebuilding the Economy Prepared by Monica S. Fong Annex 4 Housing and Labor Market Distortions in Russia Prepared by Henna Natras LIST OF TEXT TABLES, FIGURES, AND BOXES Text Tables Table 1.1 Comparison of Selected Hnman Development Indicators. . . 6 Table 1.2 Changes in the Subsistence Minimum Income, Russia. . . . 10 Table 2.1 Federal Cash Benefits, Russian Federation (1992) . . . . 26 Table 2.2 Minimum Pensions Relative to Poverty Line (1992) . . . . 27 Table 3.1 Indicative Savings from Different Reform Options . . . . 71 Table 4.1 Type and Availability of Benefits by Firm Size . . . . 99 Table 4.2 Distribution of the Unemployed by Duration . . . . . . . 114 Table 4.3 Wage Policies in Eastern Europe. . . . . . . . . . . . . 117 Table 5.1 Ratio of Staff to Unemployed in Selected Countries . . . 123 Table 5.2 Structural Changes in the Financing of New Housing . . . 132 Figures Figure 1.1 Consumer Price Trends, Quarterly Figures (1991-1993) . 3 Figure 1.2 Incidence of Poverty in Russia Among Selected Vulnerable Groups (September 1992) . . . . . . . . . . . 12 Figure 1.3 Composition of Poverty in Russia by Household (September 1992). . . . . . . . . . . . . . . . . . . . . 13 Figure 2.1 Public Expenditure on Social Protection (1990) . . . . . 24 Figure 2.2 Employment Fund Expenditures by Category - First 6 Nonths (1993). . . . . . . . . . . . . . . . . . . . . 31 Figure 2.3 Social Insurance Fund :Expenditures by Category : First 6 Months, 1993. . . . . . . . . . . . . 33 Figure 2.4 Fund for Social Support of the Population Ezpenditures (1992). . . . . . . . . . . . . . . . . . . 41 Figure 3.1 Pension Reform ..... . 77 Figure 4.1 Russia: Real State & Industrial Wages. 103 Figure 4.2 Russia: Vage Bill and Wage Norm (January-August 1992). . . . . . . . . . . . . . . . . . 104 Figure 4.3 Russia: Employment by Sector . . . . . . . . . . . . . . 108 Figure 4.4 Russia: Unemployment and Vacancies (December 1991-June 1993). . . . . . . . . . . . . . . . 114 Figure 5.1 Structure of Housing Ownership. 1991 . . . . . . . . . . 135 Figure 5.2 Estimated Percent of Population Changing Usual Residence in One Year (1981) . . . . . . . . . . . 140 Boxes Box 1.1 The Impact of the Crisis and Individual Responses. . . . 5 Box 1.2 The Soviet System--Earnings versus Income. . . . . . . . 8 Box 1.3 The Methodology of the 1MS. . . . . . . . . . . . . . . 11 Bo 1.4 Well-being of the Elderly. . . . . . . . . . . . . . . . 17 Box 2.1 Eligibility for Unemployment Benefits. . . . . . . . . . 29 Box 2.2 Sickness Benefits. . . . . . . . . . . . . . . . . . . . 32 Box 2.3 Social Assistance in Chelyabinsk . . . . . . . . . . . . 39 Box 2.4 Identifying the Poor in Practice . . . . . . . . . . . . 39 Box 2.5 Selected Nongovernmental Initiatives in the Russian Federation .... . . . . . . . . . . . . . . . 43 Box 2.6 East Asian Experience with Social Protection ... . . . 46 Box 2.7 The Changing Nature of Enterprise Benefits Provided by Russian Enterprses .48 Box 3.1 Social Assistance for Vulnerable Groups. . . . . . . . . 53 Box 3.2 Incentive Problems and Federal Transfers: The U.S. Experience.. . . . . .66 Box 3.3 International Experience - Sick Pay Practices and Rates of Absenteeism . . . . . . . . . . . . . . . . . . 86 Box 3.4 Pilot Pensions Scheme in Moscow Oblast . . . . . . . . . 90 Box 4.1 Benefits and the Large Firm: The Case of TsAGI . . . . . 100 Box 4.2 Unemployment Measurement and Institutional Issues . . . Ill Box 4.3 The Likely Pattern of Unemployment and Policy Implications. . . . . . . . . . . . . . . . . . . 112 Box 5.1 Innovative Job Creation Efforts Undertaken by Local Employment Offices In late 1992. . . . . . . . . . 126 GUTIVE SWIRY 1. Social protection is a crucial element in Russia's transition from a command economy. Stabilization and economic restructur.ng are critical for long- term economic growth but they impose serious social costs in the short t- rm. Immediate falls in output and employment will be followed by changes in the structure of the economy across sectors and regions. This will have an impact whose effects will vary among social groups and individuals, many of whom were not previously poor. At the same time, the system of social protection must operate within the fiscal constraints imposed by the collapse of economic activity and the need to reduce the budget deficit. Unless the Russian economy stabilizes significantly and growth resumes, its people will be unable to benefit from improved access to income earning opportunities, which must be the main guarantee of social protection. Furthermore, social protection policies must support the development of an efficient labor market in Russia. 2. Strategies for social protection will play a critical roie during the transition in Russia and beyond. Current arrangements require reshaping to meet the needs of a market econowy. This report recommends changes in the design of social assistance, pensions, unemployment benefits, and other allowances and measures to improve the efficiency of the labor market. Although tkwe immediate concern is largely for those adversely affected by the economic crisis and transition, it is important that short-term measures do not undermine longer-term strategies aimed at shifting to a market-oriented system where individuals and households take primary responsibility for their own welfare. Background 3. The Soviet system of social protection differ in several important respects from that found in typical industrial market economies. The most obvious distinction was ths use of enterprises as direct instruments of social policy. The social and economic functions of the state and the enterprises overlapped considerably. The state owned and was purported to plan and direct virtually the entire process of production and distribution. Economic growth came primarily from more extensive use of inputs, and to a lesser extent from the more intensive use of inputs. The distinction between rewards to labor (wages) and transfers for welfare purposes, which is essential to a market economy, was blurred. 4. From the post-war period until the 1980s, the Soviet regime was relatively successful in several respects. The individual was provided with security through the guarantee of employment and related benefits in state enterprises and the extensive provision of free or highly subsidized goods and services. The distribution of income was relatively equal. Hunger was uncommon in the Soviet Union. Human development indicators showed a steady improvement during the Soviet period, especially in more isolated parts of the country. Literacy reached virtually 100 percent for both sexes, and the proportion of the population completing higher education was at least as high as in several industrial countries. 5. Poverty and relative deprivation, however, did exist. Groups at risk of poverty under the former system included the elderly, large families, female headed households, and the disabled. Pensioners saw their relative position gradually erode as the level cf pensions relative to the average wage fell from more than 60 percent in 1956 below 33 percent in 1990. lhere were also regional disparities in income and pockets of serious poverty, for example, where fertility rates were high. Official denial of the existence of poverty and unemployment (which was not legally recognized) meant that social assistance was at best rudimentary. Even in the late 1960s, a significant share of the population fell below the official 'social minimum." 6. In the 1980s the weaknesses of the comund economy became increasingly evident. Labor productivity was low, and investment and economic activity stagnated. It became more difficult for the state to maintain the levels of welfare and service provision for its citizens. Under perestroika, the government embarked on a series of economic reforms to revive economic growth. At the same time, it began to introduce a series of incremental reforms to the system of social protection, partly in anticipation of increasing hardship, but also to fit the needs of a market oriented-economy. 7. Since 1991 the economy has leen subject to large macroeconomic shocks. Domestic demand and output have fallen sharply, and there have been signs of hyperinflation. The decline in industrial output has probably reached 12 percent for 1993, following an almost 19 percent fall in 1992. The general rise in prices in 1992 was on the order of fifteen-fold, while prices of basic staples such as bread and dairy products increased 100 to 150-fold. Monthly inflation rates have been high and unstable: in the range of 20 to 25 percent through October 1993. Real wages have been volatile. At the end of 1992, the average real wage stood at about two-thirds of the previous year's level, and by mid-1993 it was roughly 20 percent below the level in the first half of 1991. Despite the economic turbulence there was a surprising degree of stability in employment levels, as manyf enterprises have resorted to placing workers on leave or on short-time work rather than dismiss them--a pattern that also characterized Central and Eastern Europe during the early part of the transition. 8. The economic reforms that started in early 1992 are being undertaken in conditions of continuing political uncertainty, which has led to indecisive and contradictory policies. State enterprises have not generally been subject to hard budget constraints. Arrears and extensive subsidies continue, while wage controls have been largely abandoned. The shift in Central Bank policy in mid-1992 validated the large arrears enterprises had accumulated and sent further mixed signals about the credibility of the reform program. High nominal wage claims, the perceived failure to enforce hard budget constraints, and the high concentration of market power have combined with other factors to fuel continuing inflation. 9. These economic developments have imposed serious pressure on living standards. It has been difficult for Russian households to protect themselves against inflation. There are no real estate or stock markets, and real interest rates on savings have been highly negative. The incidence of poverty is increasing and the patterns of poverty are changing. Most obvious among the emerging "new poor" are the unemployed. Open unemployment has emerged for the first time and will probably exceed 2 million by the end of 1993. Many households affected by unemployment are falling into poverty. To date, employment adjustments have mainly affected women. - ii - Figure 1. People Below tinumum Subsistence Income In Russia by Type of Household: September, 1992 Disabled 4% On mraternity leave 2% Not In labor force 16% Others 1% Working 68% Retred 10% denIX\"~ Students 2% Note: The threshold inoome ls based on a minimum eub91stence food basket, consistent with the WHO guidellnes that represents an average 68.3% of household expenditure Source: Russian Longitudinal Monitoring Survey Preliminary Results, 1993 10. The Russian Longitudinal Monitoring Survey (RLMS) provides some broad measure of current hardship.1 In late 1992 up to one-third of the sample population was below an austere threshold income. These figures are tentative, and likely conceal under reporting of income, especially of in-kind benefits such as housing as well as informal sector income. However, while the actual magnitudes are as yet uncertain, the composition of poverty is fairly clear. Many of those below the subsistence minimum are the "working poor, n as real wages have fallen significantly, and forced part-time work is widespread (Figure 1). The majority are families with children. The next largest group is the elderly. The disabled and other disadvantaged groups are smaller. Within other groups- -households affected by unemployment, those families with an above average number of children, or those with a single head or a disabled member--a 'J Th. income threshold used was redefined with World Bank assistance and is si8nificantLy lower than that which would follow from the traditionaL Soviet approach. It is based on a minimum subsistence food basket--consistent with WH0 guidelines and Russian consumption patterns--that represents 68.3 percent of subsistence income. The data--based on preliminary results of a survey of 6,000 households in autuim 1992--are tentative and could change substantially given the tremendous instability of prices, the failure to take into account both regionaL price variations and economies of scale in consumption within a family and likely underreporting of income. At the same time, the RLMS is undoubtedly the best available data source on Russian living standards. - iii - significant proportion may be classified as being in severe poverty, with incomes at less than 50 percent of the minimum subsistence level. The Role of the State in Social Protection 11. A long-term strategic view of social protection is essential to the redesign of the existing system, even though many of the changes car. only be introduced over time. The main objective of the shift to a market economy is to raise economic welfare by increasing productiv_.ty and choices. Improved living standards in Russia will be based on the achievement of a stable economy, competition among a large number of profit seeking enterprises and individuals, effectively functioning capital markets, and labor markets where wages and employment are largely determined by supply and demand. The generation of good income earning opportunities is the main route to social protection and the alleviation of long term-poverty. 12. In many industrial countries responsibility for social protection has shifted in the direction of the individual and of families. The precise boundaries between state and private actions are dynamic ar.d likely to remain controversial. Nonetheless, appropriate long-term objectives for the Russian government aiming for a market-oriented economy include the following: = The promotion of good earning opportunities. Economic growth and expanding employment are the primary guarantee of social protection. This requires the correction of major macro-imbalances, and the creation of an environment conducive to private sector activity. Alone, however, it may be insufficient. 3 Poverty relief. The government may decide that no one should fall below a minimum acceptable level of subsistence. This implies a need for income redistribution (transfers) financed through taxation. What is regarded as a minimum subsistence level is a function of economic conditions and social welfare preferences. Wages and employment should be largely market determined in the interest of efficiency. * Insurance. There is potentially a significant role for private financing in insurance. Individuals should be expected to take an increasingly larger role in smoothing their own consumption over time. The government may help fill gaps in private insurance markets in the interest of efficiency and equity. For example, unemployment insurance helps to avoid the burden of adjustment falling disproportionately on the unemployed. * Special efforts to protect the most vulnerable groups, particularly children. The incidence of poverty is significantly higher among families with a larger number of children. Hence, programs to protect such children such as vaccination campaigns, primary health care and school feeding programs are important. * The facilitation of private activity. Both the profit and non-profit sectors can make valuable contributions to social protection. For example, the state may consider contracting out social service delivery to nongovernmental organizations. - iv - 13. There are some trade offs between the pursuit of economic growth and equity. For example, t.he government may seek explicitly to redistribute income through transfers such as unemployment benefits or family allowances. Alternatively, public resources could be directed to other public good functions best undertaken by the government (for example, basic education or infrastructure). Or general government spending could be cut in the interest of reducing overall public sector involvement in the economy and increasing private savings, investment, and growth. Several East Asian economies reflect successful cases where the second and third routes were emphasized rather than the first. For example, the Republic of Korea's commitment to education meant that the bulk of social expenditures was directed toward education, complementing its export-led develupment strategy. 14. Social protection in Russia is still heavily dominated by the Soviet inheritance and has certain uharacteristics inappropriate to a market-oriented economy. In the future, social protection will require shifting organizations' and individuals' roleb so that enterprises are transformed from transfer-guarantee social service agencies to agencies providing productive employment, and so that individuals become active participants in a market economy. State transfers should be designed consistently with individual incentives to search for employment, save against the risk of future loss of income, and provide support for relatives and friends. 15. During the transition, however, special measures may be necessary to avoid hardship and preserve certain achievements of the Soviet era while establishing a system that will be viable and appropriate over the long term. In the short term, poverty alleviation policies will not only play a redistributive role, but they may also increase the efficiency and speed with which the Russian economy is transformed into a modern market system. It is especially important to target vulnerable groups to ensure continued access to basic services, and support through such means as school feeding programs. The Existing System of Social Protection 16. A striking feature of the Russian system of social protection is the fragmentation of funding sources and responsibilities. Four federal extrabudgetary funds, several ministries, and lower levels of government each play some role. In 1992 these funds represented roughly 7 percent of GDP, substantially less than what Central and Eastern European economies in transitior spend on social protection and half what some Western European governments spend on pensions alone. 17. The Pension Fund was financed in 1992 by a payroll tax of 32.6 percent.2 For the time being, it is financing a variety of benefits in addition to labor pensions, including child allowances. The Pension Fund has proved to be an efficient agent for revenue collection, securing higher compliance rates from enterprises than does the State Taxation Service. In fact, the Pension Fund ran a significant surplus in 1992 (2.7 percent of GDP). It has failed, however, to provide a minimum subsistence income for significant numbers of the elderly. At 2/ Notioneally 1 percent was on ewVploy. contribution. -V- the same time some pensions remain relatively generous. There is a clear need to mairntzin the real value of pension benefits in inflationary circumstances. However, because budgetary constraints preclude full indexation for all recipients, this effort should concentrate on those receiving the minimum benefit. Amendments to indexation rules in late 1993 recognized the latter principle. In 1993 the payroll tax to the Pension Fund was reduced to 29 percent- -the other 3.6 percent was earmarked for health insurance. 18. Unemployment benefits are paid out of the Employment Fund, which is financed by a payroll tax of 2 percent. Subnational offices of the Federal Employment Service retain 90 percent of the revenue. The Employment Fund was in surplus in 1992 and 1993 because registered unemployment was low and unemployment benefits (based on historical earnings) were meager. The present system of calculations is complex and time-consuming, and is also unnecessary in conditions of high inflation because about half the recipients en' up with only the minimum benefit. The most significant problem is that the Employment Fund will not be able to finance benefits in the face of predicted high levels of unemployment. Bank projections suggest that many oblast funds will be running into deficit as soon as unemployment rises significantly. The structure of the Employment Fund does not allow surplus revenues from better off oblasts to be reallocated to those with high unemployment and correspondingly less revenue. 19. Sick leave and maternity benefits are Fpid out of the Social Insurance Fund (SIP), which was established in 1991 as, in principle, an independent financial institution, but was in reality responsible to the successor to the former official trade inion. Its revenues derive fr'm a payroll contribution of 5.4 percent. Three-quarters of this amount is retained at the enterprise level to be spent on employees' benefits. The remainder is used to cross-subsidize enterprises and branches that have unusually high expenditure needs. In 1992 and 1993 the SIF faced falling compliance rates and charges of widespread abuse by trade union officials. The inherited system of management and control was problematic in terms of democratic accountability and the right to free association. The government recognized this, and the management structure was being overhauled in late 1993. 20. The Fund for Social Support of the Population supplements a variety of subnational programs of social assistance that are mostly in kind, using rather crude methods of targeting. It is financed through allocations from the Ministry of Social Protection. Moreover, the practical criteria for allocating resources among oblasts are unclear and apparently do not have a redistributive effect. It is limited in scope, however, representing less than 0.1 percent of GDP in 1992. 21. Subnational programs are an important aspect of current social assistance efforts.3 The central government has largely transferred responsibility for social assistance, health, and education to subnational levels of government. However, the assignment of social responsibilities was not matched by an assignment of adequate revenue. The changes in intergovernmental fiscal 3/ The Aussian Federation comprises three tiers of government. As of late 1993 there were eighty-nine subjects of the Federation includirg oblasts, Krais, and autonomous republic& (referred to in the report as oblestc). Below the kblat level are approximately 2,200 raiogs and municipalities. - vi - relations introduced in 1992 and compounded by the economic crisis resulted in a significant widening of fiscal disparities among regions. Consequently, subnational authorities are under serious--and worsening--financial pressure. Growing reliance on extrabudgetary sources can mean increased uncertainty, and the processes of central grant allocation are not transparent. This makes it difficult for subnational governments to plan recurrent expenses. As a result, virtually none of the forms of local social assistance is guaranteed over time. 22. Considerable variation across the country leads to large horizontal inequities. Prior to the breakup of the Soviet Union, oblasts in Russia already differed substantially in the amounts spent on the social sectors, with some spending as much as four times more per person than others. Such differentials are widening because the adjustment process has a much greater impact on incomes and taxable capacity in some regions than in others. Exis-ing mechanisms to redistribute revenues among regions in Russia are limited. 23. The oblasts receive revenues from a combination of shared taxes and transfers. The total allocation of transfers from the federal budget to localities amounted to less than 2 percent of GDP in 1992. The old norms for expenditure allocation (for example the number of hospital beds) were progressively abandoned between 1985 and 1991. Since then, the amount and allocation of the transfers and revenue shares seems to have resulted essentially from political negotiations between the center and the oblasts, rather than from any objective criteria reflecting tax effort or expenditure needs (such as population or per capita income). The Impact of Enterprise Restructuring 24. The economic reforms and restructuring will probably cause further falls in real incomes and large-scale unemployment. Revenue collection and expenditure demand for social protection are likely to diverge. In 1992 and 1993, despite falling output and lowered real incomes, nominal revenues for social protection remained reasonably buoyant because the revenue base--payroll taxes--continued to be levied on a work force that had not been subject to large-scale layoffs. This situation will change. Once stabilization and enterprise restructuring begins, public revenues will fall. Receipts from the inflation tax, the Drofi'; tax, and the value added tax (VAT) will fall as inflation slows and output declines further, as is expected. The extent and size of these effects depends on the speed of stabilization and structural adjustment. 25. Enterprise restructuring will have significant implications for the system of social protection. In the absence of hard budget constraints and in the light of continuing uncertainty about policy and market prospects, enterprises have preferred to hoard labor. Given the initial scale of overstaffing and the collapse of capacity utilization, unemployment is expected to rise to high levels. The private sector will become an increasingly important source of income opportunities, although in the short term its impact on total employment is likely to be small. 26. Overall, employment adjustment will place demands on the system from two directions: reducing the tax base and increasing the financing needs for unemployment compensation. Yet the requirement for unemployment benefits need not - vii - be large. On the assumption that all recipients receive a flat rate benefit set just above minimum subsistence, as recommended in this report, it is estimated that every additional 1 percent of unemployment would increase expenditure on unemployment benefits by less than 0.2 percent of GDP. Obviously, higher benefit replacement rates would require more resources. 27. The problems facing the social sectors will be exacerbated as newly market- oriented enterprises stop providing local infrastructure and services such as health and education. While withdrawal has not yet occurred on a broad scale, it is already evident in areas such as the ptivision of preschools, which Russian children attend to age seven. Divestiture of housing, which is highly subsidized and poorly maintained, is already causing considerable difficulties because many local authorities are refusing to accept the transfer of responsibility from enterprises. Subnational expenditure needs will increase as authorities take over public goods previously financed by enterprises. The problems will be particularly serious in cases of one-company towns in which the company goes bankrupt, where the local economy could collapse along with the community' s social infrastructure. Mechanisms need to be developed to allow the transfer of such responsibilities to subnational governments or private sector agents, so that critical social services continue to be available to the entire population. Under present fiscal arrangements, it will be extremely difficult for many localities to cope with sharply rising unempl.yment and falling revenue. Strengthening the Safety Net 28. Current social assistance efforts have two broad shortcomings. In the first place, existing programs become inadequate because of high inflation. Second, poverty rates among the unemployed and the working poor remain high, as these groups are not adequately covered by existing programs. Income support for the recipients of social insurance benefits (pensioners and the unemployed) has fallen to meager levels. The focus of the federal agencies, in particular the Ministry of Social Protection, tends to be on conventional categories of need, such as disability and old age, neglecting the new poor. 29. The concept of a subsistence minimum income was recently redefined, with the assistance of the World Barn, and adopted by the Russian government. A major objective of this effort was to devise a poverty line that would help policymakers seeking to target assistance in a rapidly changing environment. A food basket was developed based on a diet that is nutritionally adequate and consistent with Russian food habits, as well as with the shifts in consumption following the liberalization of prices. Clearly the realistic level of minimum subsistence depends partly on economic conditions. The official poverty line assumes that, on average, people spend 68.3 percent of their income on food. In the currently constrained context, this report recommends a minimum benefit based on food at 80 percent of expenditure (that is, a lower poverty line for the purpose of targeting scarce public resources). This, however, is only plausible in the short term while price controls remain on potentially important household expenditures such as housing. 30. If the Russian government determines that a comprehensive social safety net is an immediate priority, subnational authorities should implement a minimum cash - viii - benefit of last resort. It would embrace the principles detailed in the following paragraphs: * Eligibility should be determined solely by means testing carried out by local offices of social protection. Qualifying conditions should be kept as simple and explicit as possible so that there are no doubts about eligibility on the part of claimant or administrator. Consideration should be given to whether any element of income from work, assets, or other cash benefits should be disregarded to avoid adverse incentive effects. Certain allowances, such as disability benefits, may merit special treatment to reflect the additional necessary expenditures incurred by disadvantaged groups. c The amount awarded should be based or a subsistence minimum basket set at oblast prices and adjusted appropriately for the type of household unit (for example, the number of dependent children). * Financing the minimum cash benefit in a federal system should take the form of matching grants from the federal government to support local efforts so as to meet the goals of efficiency and equity. 31. In proposing a scheme of social assistance of last resort, strategic choices have to be made. The most important include the form that the benefit should take, how it ought to be targeted, and the most effective means of financing and administering it. 32. Form of Benefit. Currently, social assistance in cash is the exception rather than the rule. On the other hand, quantity-denominated vouchers and actual in-kind provision have a critical advantage during high inflation, because preserving real levels of money transfers through frequent indexation is cumbersome. On the other hand, international experience shows that the administrative costs of in-kind schemes are higher than for income transfers. Cash transfers have less distorting effects in goods markets and permit greater transparency in the budget compared with food subsidies, for example. Assuming there is success in stabilization and the control of inflation, cash transfers are to be the preferred form for the new basic benefit. 33. How to Target. The rationale for a means-tested cash benefit, administered by local authorities but financed partly by central transfers, is based on the fact that despite the existing array of categorical programs for various demographic groups, most of the poor in contemporary Russia do not receive any social assistance from the government. Even doubling the real level of existing benefits would only reduce poverty by about one-fifth. A minimum basic cash benefit should be introduced and should become, during the next eighteen months, the principal instrument of the social safety net. Meanwhile, a variety of schemes and initiatives, especially at the subnational level, should be encouraged. Locally designed schemes, for example, meal programs for the poor, may be better adapted to local conditions, especially in a country as large and diverse as Russia. In very poor areas the federal transfer could represent a large share of program cost. The federal government should make matching grants available to the oblast authorities to support subnational programs of social assistance. Matching grants require subnational governments to supplement central - ix - government transfers with their own money, thereby economizing on scarce central government resources and encouraging local expenditures. Hatching grants would usually be conditional on other measures being taken, for example, withdrawing general food subsidies. Such a financing system could be established relatively quickly. Once the minimum benefit is established, there are still likely to be local enhancements reflecting differences in local resources and preferences. 34. Administration. Delivery of social assistance necessarily relies on local offices of social protection. Existing administrative capacity is weak and requires strengthening. Administrative arrangements need to be developed that are capable of processing and delivering claims inexpensively and efficiently. Computerization will substantially strengthen the subnational administrative capacity of social protection offices. Staff training will also play an important role. Principles for Setting Benefits 35. A number of principles should govern benefit levels in the short term. The subsistence minimum, set at very austere levels and on the basis that 80 percent of income is spent on food, should be regarded as the system's anchor. The subsistence minimum basket should be valued in oblast prices. This is the minimum income that the federal government will seek to extend to all Russians who would otherwise fall below the threshold. Social insurance benefits should be set above the subsistence minimum to accord recognition to past contributions. 36. Pensions. The min.-umm labor pension, to which Russians with a twenty-year work history are entitled, should be set above the minimum subsistence, as should the minimum unemployment benefit. Beginning in the third quarter of 1993, only the minimum benefit becomes subject to full indexation. Benefits above the minimum increase by the same nominal amount as the minimum, thereby containing costs. This is appropriate during the transition, but the relationship to contributions would need to be reestablished in the long run. The minimum labor pension should be set slightly (say 15 percent) above the subsistence minimum and be automatically adjusted whenever the nominal cost of the subsistence minimum basket has risen by at least a certain minimum (say 10 percent), though not more often than monthly. The real minimum value of other cash benefits (social pensions and disability allowances) should be maintained at levels at least as high as minimum subsistence. 37. Unemployment Compensatlon. Unemployment has developed at a much slower rate than expected, but high levels are anticipated as the economy adjusts and restructures. The most urgent need is to develop the capacity to register claimants and to pay adequate benefits. This is a high priority of the foreign assistance program, which includes a program of computerization under a World Bank loan Russia Employment Services and Social Protection Project. But a revision of the principles on which unemployment compensation is set, as well as financing for the system, is required. Funding for proactive employment policies and income support should be separated, with proactive policies funded from the federal budget. Otherwise, income support is likely to be squeezed when unemployment rises significantly. Budgetary support will be required to pay for - x - benefits in high unemployment oblasts that have exhausted their Employment Fund revenues. 38. Minimum income support for the unemployed should be kept above the poverty threshold, as defined by the subsistence minimum, rather than being tied to the minimum wage. The unemployment benefit should not be means tested and, for at least the period of eligibility (twelve months), should be slightly higher than the subsistence minimum. 39. The present method of calculating individual unemployment benefits is administratively cumbersome and of little value during high inflation, when many people end up receiving virtually the same benefit. As the Federal Employment Service has suggested, flat rate unemployment benefits should be adopted temporarily. 40. For unemployment extending beyond twelve months, as well as for those quitting their jobs voluntarily and for first time job seekers who are registered as unemployed, individuals should be entitled to social assistance. The benefit level should be maintained at least at the level of the subsistence minimum. Recipients of this benefit may enroll in retraining programs or public works programs (although normally public works programs pay more than the subsistence minimum). Fiscal Implications of the Proposed Reforms 41. The central objective of this report is to indicate how Russia might move toward a universal and adequate system of social protection. The major constraint lies in the lack of fiscal resources. An improved safety net could be achieved largely through better prioritization of existing social expenditures and some reallocation of public expenditures toward the social sectors. At present, the government is directing less than 7 percent of the national income toward social protection (although this excludes social protection provided through state enterprises). Recent World Bank and International Monetary Fund (IMF) estimates suggest that around 22 percent of GDP is being spent on direct subsidies to enterprises, and directed credits amount to a further 21 percent of GDP. At the same time, the government is beginning to recognize the scope for reallocation of expenditure. Discussions with the Ministry of Social Protection highlighted the possibilities for shifting public resources away from enterprise subsidies and agriculture and toward targeted social priorities. 42. A number of cost savings can be achieved. Important savings can be achieved in pensions by gradually raising the retirement age and introducing a work test, and in social insurance by introducing a waiting period (say two days) before absent workers are entitled to sick pay. If only the minimum pension is fully defended against inflation, large cost savings will follow. This measure alone would have saved 2 percent of GDP in the fourth quarter of 1992. Certain reforms, such as increasing the qualifying age, will result in increased savings over time, thereby reducing pension expenditures by 35 percent by the end of the decade. 43. The largest additional cost implied by these recommendations arises from the introduction of a minimum basic benefit. Yet, even allowing for - xi - administrative costs and leakage, its implementation would probably not require more than 3 percent of GDP, based on preliminary results of the RLMS, using a poverty line below that currently applied by the government (an average of 80 percent of expenditure would be for food). About 15 percent of the Russian population would have been eligible for the proposed benefit in late 1992. 44. Indicative estimates of the overall package of these recommendations show the net cost of the reforms to be about 2 percent of GDP. This would mean that social protection expenditures in Russia would still amount to less than 9 percent of GDP. This would continue to be in line with countries with comparable levels of per capita income. The Response to Future Nacroeconomic Crises 45. A critical aspect of the system of public finance and expenditure is its ability to cope with macroeconomic downturns. In economies that are much closer to macroeconomic equilibrium than contemporary Russia, tax-financed income transfers can be counter-cyclical, that is, they can operate to stabilize the economy automatically. By contrast, there is the risk that the level and structure of transfers (entitlements) can sustain, or even cause, a fiscal crisis. The latter issue looms large in the United States today. For example, automatic indexation of social security payments constrains the government's choice of other public expenditures if taxes are not to be increased significantly. To ensure financial sustainability, the benefits structure in this report is based on an austere level of minimum subsistence. The Need for Federal Transfers 46. It has become increasingly evident that existing local-central relations are a serious problem, especially the lack of consistency between local social responsibilities and tax bases. Therefore, criteria to govern federal transfers must be defined. It is important that effective mechanisms be developed to redistribute public resources to oblasts and ralons facing economic collapse. The Bank has recommended a new approach to intergovernmental fiscal relations, that would include, the creation of a system based on shared taxes and a formula-based distribution system (Report No. 11302-RUS). With respect to social expenditures, matching grants, which are conditional on local expenditures on the particular program, would best meet the goals of efficiency and equity. Such a system avoids the leakages and increased costs that are associated with schemes based on full cost reimbursement or block grants. Local taxpayers who are sharing the burden may better hold local authorities accountable. Open-ended cost reimbursement is problematic, because subnational authorities lack incentives to contain expenditures. Natching grants encourage and reward increased local efforts, and have succeeded elsewhere (for example, the Canadian Social Assistance Plan). Programs that are appropriate targets for matching grants would be in the areas of social assistance (in particular the minimum basic benefit), proactive employment activities, health, and education. 47. The main disadvantage of matching grants--potentially adverse equity effects--can be alleviated through tax sharing and transfer formulas that take some measure of the subnational tax base into account. One readily available crude indicator would be oblast per capita income. A more sophisticated approach - xii - would also include indications of expenditure need, such as estimates of the oblast poverty gap. Special consideration might be given to localities that face economic collapse. Labor Markets and Employment Policies 48. The analysis of recent labor market developments in the Russian Federation reveals a surprising degree of stability in employment levels to the end of 1993. Open unemployment has been allowed to emerge, but amounted to only about 1 percent of the labor force by the end of the year. 49. Changes in ownership and the parallel expansion of a private sector - as yet largely concentrated in services - had important implications for the distribution of employment across public and private branches. Private enterprises respond to economic developments faster than state enterprises, and as they search out new markets, private sector labor demand will increase. The effect on the labor market, however, is not significant because, though growing fast, the private sector is still small. The private sector cannot, at this stage, absorb a significant share of workers made redundant in the course of restructuring. 50. State enterprises have exhibited considerable inertia with respect to employment. The few who have been laid off tend to have been women and white collar workers. Enterprises have sometimes placed workers on minimum wages with little or no work requirement, which effectively amounts to providing de facto unemployment compensation. Another common route is to place workers on forced leave or short-time work. This affects possibly up to one in four industrial employees. The rate and share of laid off workers in inflows to unemployment increased in late 1992, but high levels of voluntary quitting continue at the base of the skill structure. Enterprises have preferred to hoard labor in light of the uncertainty about policy and market prospects. 51. Wages have been more volatile. The price shock of January 1992 cut real wages by an estimated 40 percent. The subsequent rebound appears through several channels. First, the institutional-cum-mechanical constraints on liquidity were lifted in the middle of the year, with a major injection of credit to enterprises and the economy as a whole. Second, the excess wage tax - in the form of the minimum-wage-times-four rule - proved ineffective as the minimum wage fell to very low levels and enterprises elected to pay additional profit tax on wages paid above the cap to increase bonuses or benefits, or to do all three. This enterprise behavior can be attributed, in part, to worker control over decisionmaking and perceived high probability of continuing bailouts. The sharp acceleration in producer prices for large enterprises points to, at best, a stable pricing rule facilitating nominal claims or, at worst, an increase in the markup. Some increased differentiation in wages across skills and across regions can be observed, alongside considerable inertia in the system constraining departure from the prior tariff structure of wages. In mid-1993, the average real wage stood about 20 percent below the level in early 1991. - xiii - Figure 2. Movements in Unemployment/Vacancies March 1992 - March 1993 (Million people) 1.2 t* - . _ . ._ . . 0.8 . - .............. _ . 0.6 ......... 0.4 ........ ................ ......... ......... .. ...... 04 1192 V92 8/62 4/92 6/2 /192 7/2 8/02 0/6210/92W/212 V2ie98 2198 S/O Month -Brod -h INarrow -*-RecelvBenotit Vacnoba 52. It has long been realized that non-wage benefits provided through the enterprise were an important component of labor income. Yet it was far from clear what the value of such benefits was. The results of a small Bank survey of enterprises show that discretionary benefits are varied and pervasive across all enterprise classes. There is a positive association between enterprise size, measured by employment and the range of benefits. A significant proportion of the labor force continues to have access to enterprises providing housing, health care, and paid vacations, for example. Indeed, where single local employer is dominant, the functions of subnational authorities commonly appear to be overtaken by the enterprise. Among smaller enterprises with less dominance over the local labor market, the evidence suggests that enterprise-provided benefits are unlikely to exceed those covered by subnational governments. For housing, child care, and health facilities, enterprise expenditures at the local level were comparable to or in excess of government or non-enterprise expenditures in only 10 to 30 percent of cases, depending on the function. 53. At current prices, workers' benefits equal approximately 35 percent of income. The overall sample reveals considerable stability in the cost structure, suggesting that benefits have not been a widespread mechanism for significantly raising or lowering aggregate labor costs. Only a quarter of the enterprises had suspended any benefits. In half of those cases, housing programs had been discontinued. -xiv - The Role of the Federal Employment Service 54. Income support for the unemployed creates a foundation that allows enterprise restructuring to proceed without inflicting intolerable hardship. The next step is to implement complementary policies of a more proactive nature that can make an important contribution to the labor market transition. Modern employment services do not substitute for the economic restructuring and growth that are ultimately required to provide productive jobs for those unemployed or underemployed in jobs with poor prospects, but they do help labor mobility and redeployment. Most job growth is likely to occur in the new private sector and in small scale activities where the role of government employment services would be marginal. 55. The Federal Employment Service (FES) is engaged in activities variously described as job creation, wage subsidies, and loans to enterprises that represented about one-fifth of Employment Fund expenditures in 1992. This is in addition to the estimated 43 percent of GDP spent on subsidies and directed credits to enterprises. Large-scale indiscriminate support to enterprises is posing enormous obstacles to the reform process. At the same time, however, in response to a sharp rise in unemployment following stabilization, attempting to reduce unemployment where it is unusually high would be an appropriate policy. The experience of Eastern Europe suggests that unemployment is likely to involve regional shocks. High rates of long-term unemployment can destroy human ^apital and local communities. Labor mobility is not an answer to high regional unemployment in the short term, given housing constraints, the generally depressed labor demand, and workers who will be reluctant to move. Jobs can be created or preserved in the local economy by a variety of means, including public works and financial support to enterprises. The main objection is that each, in different ways, appears to run contrary to the commitment to an economy run on market principles. Yet financially, these schemes could pay for themselves in savings on unemployment benefits. The cheapest option could be maintaining jobs in enterprises, depending on the extent to which complementary inputs are already in place, the presence of inexpensive or even free capital, and the size of the cash flow generated by the sale of output. 56. There may be a case, in terms of economic efficiency, for continuing to support loss-making enterprises in the short run, provided that the subsidy is strictly temporary, fixed in size, and in the form of budget transfers, not directed credits. The principle of maximizing employment at minimum cost implies some cut-off cost per job figure. Financing from the federal budget would be allocated to oblast FES offices where unemployment was unusually high, perhaps 150 percent of the national average. The local employment office would be accountable for the funds spent, according to clearly established federal criteria. The proposed scheme is designed to replace, rather than augment, the existing opaque and economically costly schemes by which enterprises are subsidized. 57. Geographical mobility of labor is hampered by the scarcity of housing and the absence of a normal housing market, and also by a system of residence permits that has deterred internal migration. Although labor mobility is likely to remain limited for the foreseeable future given depressed labor demand, it is important that the obstacles to mobility be removed as much as possible. This will - xv - facilitate restructuring when the labor market becomes more buoyant. The permit system, the propiska, requires that permission be obtained from administrative authorities before relocating. The propLska is a prerequisite for access to many locally provided services. 58. This report suggests that the propiska system be eliminated and housing policies be reformed. These include more rational privatization and rent policies, as well as better management of repossessions and reallocations. All remaining legal and administrative restrictions on sliblets and exchanges in municipal housing, such as prescribed floor space norms and bans on cash payments between lessee and lessor, should be revoked or redefined so that only reasonable safety and health codes apply. Enterprises should be granted more control over their housing stock and be encouraged to shift to indirect assistance. An Agenda for Future Analysis 59. The foregoing recommendations focus upon the most urgent needs for social protection that emerged at a fairly early stage in Russia's transition from a command economy. This is by no means an exhaustive account, and further analysis will benefit from the increasing availability of information and evidence about the impact of the transition. It is useful, therefore, to set out a brief agenda for further analysis anid action in the sphere of social protection for the government and the World Bank in terms of six main points: 1. Intergovernmental fiscal relations--the lack of consistency between local social responsibilities and tax bases, and therefore the need to define criteria to govern federal transfers. 2. Scope for reallocation--highlighting and, as far as possible, quantifying, the possibilities for shifting public resources away from enterprise subsidies and agriculture and toward targeted social priorities. 3. Reform of the pension system--the creation of an effective independent agency with management and control of revenues and expenditures. 4. Constructlng a dynamic profile of poverty--highlighting changes in the position of individuals and households, as well as of regions, during the transition through continuing support for the Russian Longitudinal Monitoring Survey and providing useful and relevant results for policymakers. 5. The interrelatlon of wages and the structure of benefits--especially the implications of low wages and low productivity for the latter. 6. Social impact of enterprise reform and privatization- -especially the emergence of a new middle class. - xvi - Conclusion 60. Social protection in Russia is a prerequisite to financial stabilization and economic restructuring. The absence of assurances of minimum living standards delays thorough-going reform. At the same time, the economic crisis- -the collapse of output and the rapid inflation associated with inconsistent and unsustainable macroeconomic policies--has imposed serious hardship on millions of Russians. This has occurred even before the expected rise in open unemployment. Devolution of responsibility for basic aspects of social protection to lower levels of government has come at the cost of widening the existing disparities among regions and population groups, and is likely to worsen with the differential impact of adjustment. The existing system of social protection is extensive and complex, yet the amount of assistance given is small, and most of the poor do not receive any assistance. Evidence shows that families with chil4ren are particularly vulnerable. 61. The government faces important trade-offs among different objectives--the pursuit of economic growth, the provision of public goods, and the provision of universal welfare--over the short and longer term. Increased targeting and reductions in public expenditure on social protection help to minimize the tensions between growth and equity. Indexation should be based on defense of the minimum, replacing current ad hoc arrangements and reducing overall spending on pensions and unemployment benefits. Rationalization of the existing array of benefits and strengthening of administrative mechanisms are also required. 62. During the transition, a strengthened safety net may facilitate the process of economic restructuring that Russian authorities have been reluctant to undertake. Should the government decide that establishing a comprehensive system of social protection is an important priority, the appropriate strategy would be to base it on an austere but realistic poverty line. This route would require that existing public expenditures be reallocated and supplemented to ensure that people no longer fall below the minimum subsistence level. A primary vehicle would be the proposed minimum basic benefit, administered by local authorities and financed partly through federal matching grants. The recommendations in this report aim to enhance the capacity of the system of social protection to cope with increased strains and new challenges during this difficult period, to set out the strategic choices facing the government, and to lay foundations appropriate to the longer-term needs of a modern industrial market economy. - xvii - CHAPTER I: TME CONTEXT Nacroeconomic Developments 1.1 Russia's attempt at stabilization in 1992 proved short lived and was beset by underlying problems in the economy and by difficulties in achieving political consensus. Nonetheless, progress has been made. Structural changes have begun to work their way through the system, in particular, the transfer of ownership and the mergence of private property rights. The pervasive controls that characterized the Soviet system have weakened. But against these gains lies the reality of high inflation fueled by large fiscal imbalances and reliance on monetary financing. Moreover, the evidence suggests that the inertia of enterprises, particularly large ones, is a primary obstacle to stabilization and adjustment. 1.2 The incomes of many Russians have been radically eroded by inflation, and many of them now fall below the poverty line. Output continues to plummet: gross domestic product (GDP), which had declined by around 12.8 percent in 1991, fell about 19 percent in 1992 and is projected to fall by 12 percent in 1993. The key component of structural change--a positive supply-side resp)nse--is virtually absent. Almost all adjustment in the labor market has been on the wage side. Open unemployment has been largely postponed. The den mda on the system of social protection have grown and are likely to accelerate swiftly in t4e short term. 1.3 Beyond the hardship imposed by macrodevelopments, there are fiscal problems to be addressed. As in other economies in transition, and depending on the pace of reforms, Russia is likely to experience decreasing revenues and corresponding pressures to increase spending. Such diverging trends, combined with the urgent need to contain inflation, will place an even higher premium on the efficient use of scarce resources. This section sets the stage for the discussion of recent trends in human development and poverty, the analysis of the system of social protection as it stood in mid-1993, policy issues and options, and recent labor market developments. Anchors to Stabilization 1.4 Price liberalization was the primary component of the attempted stabilization of 1992. For Russian consumers, price liberalization initial.y meant convergence of official and nonofficial prices and some lessening of queuing in state stores. Yet the enormous price jumps of January 1992--producer prices rose by 382 percent over the previous month and consumer prices by 245 percent--were followed by continuing high inflation. This can be traced to several factors: (1) the extensive directed credits to enterprises (which accounted for around two-thirds of base money expansion in 1992), (2) expectations of future inflation, (3) the behavior of price setters in the economy and the noncompetitive structure of production, and (4) the fall in the exchange rate. 1.5 The initial stabilization effort was, in principle, to be pinned by the money supply and, to a lesser extent, by wages. There was no attempt to fix the exchange rate as the primary anchor--there were no reserves that could have I defended the rate. Unl5.ke in the Central European economies, trade liberalization was selective and partial. Thus, the external impetus to enterprises--let alone the impact of a new relative.price structure--was weak. Moreover, liberalization took place at a time when the exchange rate was clearly too low (in purchasing power parity terms), and nominal wages expressed in foreign exchange were only a fraction of what would be needed for minimal subsistence if international prices for tradable goods were to p:evail. As expected, the ruble did appreciate substantially in real terms in tbe twelve months to October 1993, but not enough to afftet the expectation that foreign exchange was the only asset that could serve as a hedge against inflation. 1.6 Several phases of the putative stabilization can be identified. In the first half of 1992, an attempt was made to keep monetary and fiscal policy in line with the announced reduction of enterprise subsidies, tax reforms, are; cuts in government expenditures. Until mid-1992, the fiscal deficit was held to within the range of 5 to 6 percent of GDP. The Federal government also sought to push the deficit down by transferring significant expenditure responsibility, particularly in the social sectors, to lower levels of government. Though intended to help balance the federal budget, decentralization in fact led to further instability, as local officials lobbied Moscow for additional funds in a negotiated system of intergovernmental fiscal relations without any clear rules. 1.7 A tight monetary policy was announced. However, real interest rates remained highly negative and credit subsidies to enterprises increased. Financial support to enterprises has flowed through a variety of channels. Enterprises have received directed credit from both the Central Bank and the Ministry of Finance, and implicit subsidies and taxes have added to the measured flow of direct subsidies. World Bank-IMF estimates suggest that implicit subsidies amounted to about 22 percent of GDP in 1992, and directed credits to a further 21 percent of GDP. Budget subsidies and off-budget import subsidies accounted, respectively, for 7 percent and 13 percent to 15 percent of GDP in 1992. 1.8 Consistency of the monetary and fiscal stance was difficult to sustain, given the position of the Central Bank and the structure of the banking system. The response of enterprises to the prevailing uncertainty was, in common with the experience in Central and Eastern Europe, to rapidly expand inter-enterprise arrears, which might have been a deliberate bargaining strategy with respect to the government. In the third quarter of 1992 there was a clear shift in government and Central Bank policy to accommodate enterprises' financing demands by netting out claims and increasing credit to the economy. Arrears are thus unlikely to be a simple measure of financial distress or an efficient means of sorting out losers and winners. Indeed, 1992 and 1993 were primarily characterized by buoyant nominal profits, given official financial support and the pricing behavior of enterprises. 1.9 Since mid-1992, the distinction between fiscal and monetary policy has become increasingly blurred. An expanding fiscal and quasi-fiscal deficit was financed mainly domestically through monetary expansion. The budget became largely irrelevant as a guide to the fiscal stance and macroeconomic developments in Russia. The formal budget deficit was estimated at 4 percent of GDP in 1992 2 and 5 percent in the first quarter of 1993, and budget expenditures were cut from around 50 percent of GDP in 1985-91 to an estimated 34 percent in 1992. Yet these figures are more a reflection of the increasing fragmentation of public finances than they are of fiscal restraint, as extrabudgetary funds have proliferated at all levels of government. Moreover, overlapping and spillover effects have emerged between the budget, off-budget import subsidies, and the government's monetary and credit policy. Preliminary IMF data suggest that off- budget government expenditures are of the same order as budget spending, and that for 1992 the consolidated (federal and local) deficit was around 24 percent (once the Central Bank's quasi-fiscal operations have been factored in). The source of these deficits has largely been transfers to enterprises, which have been and continue to be enormous. The result is deterioration in the fiscal accounts, with manifest inflationary consequences. At the same time, these figures indicate substantial scope for reallocation of public expenditures toward social priority areas. 1.10 Wages, the second anchor stabilization, proved equally ineffective through 1992. A tax was imposed on wages that exceeded an aggregate norm, that is, four times the minimum wage. However, although monthly changes in nominal wages have increasingly converged toward consumer price changes, the large cut in real wages caused by the price jump of January 1992 has not been fully offset. Average industrial wages have reconverged to the levels of mid-1991 but remain strikingly low. Furthermore, microeconomic evidence on the frequency of wage adjustments indicates lags of at least two months, which, in light of current inflation, implies significant movements in the real wage. A "sawtooth" real wage pattern persists, despite some shortening of wage contracts. In 1993, full wage indexation was not widespread, though there are signs that wage setters target consumer price changes and, in parts of the private sector, set dollarized wages. The decision to index pensions provides an important and more general precedent for wage bargainers. Figure 1.1 Consumer Price Trends Quarterly Figures 1991-93 2500 .9DatO0. 1500 ... .. 1000................... 500 ............................. ... .........,/ 0 QV4 92/1 92/2 92/8 92/4 903/1 903/2 98/3 93/4 Quarter 3 Outout and Inventory Changes 1.11 Output fell sharply over 1992 and 1993. This can be traced to a combination of supply and demand shocks resulting from sources as varied as the disruption of inter-republican trade payments, supply constraints, and reductions and shifts in public and household expenditures. The output decline was fairly general across sectors, without a precipitous fall concentrated in any one quarter, as happened in Poland in early 1990. The decline has been cumulative since 1990; it accelerated in 1992 and has run alongside persistent inflation. 1.12 The fall in household incomes and wealth was important on the demand side, together with shifts in government spending, especially for defense.1 Cuts in investment spending have been significant, particularly given the weight of capital and intermediate goods in aggregate industrial output. As noted, real wages fell dramatically in January 1992, although considering ex ante shortages, the extent of the fall in household purchasing power is difficult to assess. Retail sales fell in early 1992, but later rebounded. Official retail data understate important innovations in trade, including the proliferation of private shops. Anticipating high inflation, households drew down money balances, provoking relatively buoyant demand for goods and assets over the second half of 1992. 1.13 Trends in inventory and its composition reflect the negative demand shocks described above. For industry, there was a sharp increase in both inputs and finished goods throughout 1991 and 1992. Inputs increased most rapidly in anticipation of supply interruptions and major price adjustments. Many enterprises have responded to negative shocks not only through cutting production but also by accumulating stocks. Thus, a credible stabilization effort might imply even higher future output losses. Regercussions 1.14 To summarize, in late 1993, there were no effective anchors to prices. Inflationary expectations have increasingly affected price- and wage-setting behavior, partly sanctioned by recourse to selective indexation and further reflected in massive dollarization. Anticipation of further depreciation of the ruble and the consequent demand for hard currency have promoted exports and massive capital flight. Among other consequences, this has worsened shortages in domestic goods markets and increased upward pressures on prices. 1.15 For the Russian people, the implications of recent macroeconomic developments have been devastating. High inflation has caused serious widespread hardship. It has been difficult for Russian households to protect themselves against inflation. There are no real estate or stock markets, and interest rates on savings have been highly negative. The demand for foreign exchange as a hedge against inflation has therefore continuously pushed exchange rates down to levels that are absurd in purchasing power terms. Effective stabilization is the key }/ The share of military sector output in total output declined by more than 11 in the first nine months of 1992. Military output fell by 401 from the 1992 level, following a 26X decline in 1991. 4 to adequate social protection. While improved monetary and fiscal policies are critical, the creation of inflation-resistant instruments, including indexed financial assets, would also help significantly. 1.16 By the end of 1992, real wages had fallen to about two-thirds of the previous year's level, and by mid-1993 real wages stood about 20 percent below the level in early 1991. Since 1990, consumption has fallen by more than 20 percent. This is reflected in the rising higher share of food in total expenditure since 1989. Indeed, all households, and particularly the poor, have been dissaving. Mass open unemployment, which has not yet emerged, will impose additional strains. The disruption and uncertainty created by the economic crisis and transition has taken place in a context where existing measures of social protection have proved to be very inadequate. Human Development 1.17 The Russian Federation has made impressive achievements in human development. Literacy rates stand at virtually 100 percent for both sexes, and the proportion of the population that has completed higher education is at least as high as in a number of industrial, market-oriented countries. Russia is regarded as a middle-income country: per capita income was estimated at US$3,200 in 1991. Under the Soviet system adequate food intake was assured for all citizens. Studies suggest that the diet of even the poorest citizens was sufficient in terms of calories. But in some other respects, such as maternal mortality, the record is relatively poor. (See Table 1.1 for a comparison of basic indicators of human development.) .S=. Impact of. the Crisis aeAd ndividual Rieapones Il receat times the ecoomic situation of the Ussian people has been in ctinual fus, iven without i. .thCe emeienOe of mass mpi.ot. >n fact,: lv xdasvh been-d setio presue for sgm tin,, althbush the desUre han acelerate Since -l901.. he e4owcio ens. -ad ttdnsition bave bean. and will contin r m ti t 'X, aot wit tvn'falls output, and i"Vjtab,y 4cut in rl jiceseed 40 TOAVe fisca. ouetiaints %4ss by the collas of e.cninlo acti.ity .n te neid. to reduc the budet deficit could affect th wrovision of soctaL rotectio andthe of bs so'ia, sere, the s,ocial':. -epcusn.' ' . ill. itebly very ?:roac £OdaMe ulose- ln fileei4i -obto the eonmic reform, adin patXua to enerpie banutc and falins semlynt.*my be thoe prsnl classed as in -eed-oer wrds, -e #*0were f,ZtieXiy badly .off under, heeQ eovit syste wSil. et neot.snily orbe thef ue*t affected. levitt.. s*ulb theS relatively- '. off -siht htwe used -to cushioA ny adve bw have -b wipe ut by 4 infatio,. is s not to argue tat those csvsnt.ioeeUy rosr*d as post eogubt to be it;xed t:-o- tie.r tha the focjs 4h*oujd b b.roadstd to take $=to: Account Ue diftetimpacit otheiss and u and to daraw a distincti.n betee the old` ad tbhe nwpoo." . hiin in ii wae tot lo at vnerab.iUi Sn t -ou {a. to Uiew sb A tes -f thaeir dee .of inolvmn in a rag of ecnee tat eon .:e termd ::fci (state-rovded-. employment or t*oceme suppoztX :°socia" (no uw1stiee4 prod~etiO~ or Ob)S" end "uouaicial" or infrml (se Is9Z).. A _ori to thi enyisan the less diss th: hoshods otf.Ulo of - ecnoes, and the gater the rsec en. the o.fficial. the mr vbeable it will be durin the tnslt4*0. Vebil_ty auntoiW -* X: lsatoa, VW 35 of bouseb:lde *opined that onl te ficial ecSwas fimportant to thn(oprd with 12X in Buga and 131 Xn schosjvmka). by definiti-, .lneale hoseholds ae es lkely to haltrative Murcs o inese, and onb ly onbe..-thir hv.aces told plots Vuerble mtW ed no haecoincided with poverty under the former :Sya.-- 5 1.18 As of January 1992, the population of the Russian Federation was 148.7 million. Between 1980 and 1988, it had expanded at an annual rate of about 0.6 percent, but by 1990 it had declined to 0.3 percent. This trend is the rezult of falling birth rates, a changing age structure, and an increase in mortality. The accelerating decline in the birth rate is largely attributable to the birth of fewer children per woman--the result of an economic situation that makes large families untenable for many. Further evidence of decisions to postpone or forego childbearing is reflected in increasing abortion rates. 1.19 Relatively low life expectancy in Russia--below such other Eastern European countries as Poland and Hungary- -is attributable to high rates of chronic disease and continuing high rates of infant mortality. Low fertility rates have led to an aging population. The female-to-male ratio reflects the lasting impact of World War II and higher male mortality rates. Overall, the health status of the Russian people has deteriorated over the past fifteen years because of the significant rise in cardiovascular disease and the continuing high rates of infant and maternal mortality. A separate sector report explores the elements of health care reform needed to meet Russian needs, including health resources, financing, and structural change. Table 1.1. A Cross-Country Comparison of Selected Human Development Indicators Indicator Russia Brazil Czechoslovakia France Japan GNP per capita (USS) 3.200 2.680 3.140 19.490 25.430 Life expectancy (years) 69 66 72 77 79 Infant mortality rate (per 1.000 live births) (18-25) 57 12 7 5 Maternal mortality rates (per 100.000 live births) 52 230 14 13 16 Population per physician (1984) 222 1.080 280 320 660 Primary edt'cation (pupil-teacher ratio) 9 23 20 16 21 Daily calories (per capita) 3.386 2.751 3.632 3.465 2.956 Source: Goskomstat of Russian Federation: World Bank, World DeveloDment ReDort: UNOP. Human Develooment Report. 192. 6 Vulnerable Groups 1.20 The collapse of the totalitarian regime, which had dominated the lives of the Russian people for more than two-thirds of the century, marked the demise of a system in which income security was provided mainly through the state's guarantee of employment. While poverty was not absent under the command economy, it is likely to worsen in the course of transition (see Box 1.1). Large-scale unemployment is also expected. In the absence of stabilization, high inflation has caused widespread hardship. The following section explores what is known about poverty in the Russian Federation. The aim is to identify and analyze the main vulnerable groups, which are likely to be changing in size and nature. The finding that most of the poor do not currently receive any income transfers from the government has significant ?olicy implications. Estimates of the poverty gap (suggesting the fiscal costs of bringing all the poor up to a minimum subsistence level) are presented in Chapter III. Definitions and Measures of Poverty 1.21 Perhaps surprisingly, considering the long tradition of work on minimum subsistence requirements and family budget surveys in the Soviet Union, reliable information on vulnerable groups has never been collected. Official survey systems presented misleading profiles of need. The Family Budget Survey, which was enterprise-based (and biased toward established and larger enterprises) and confined to full-time workers plus a smaller group of pensioners, was not representative and tended to exclude groups with a marginal attachment to the work force. The resulting picture of the population was misleading, and probably underestimated the extent of need. On the other hand, the so-called poverty line was based on an unhealthful diet high in calories and animal fat. After the rapid rise in meat prices in early 1992, 90 percent of the Russian population were shown to be in poverty. It became clear that an alternative approach to identifying the poor was needed. Until recently, data availability seriously constrained the formulation of poverty reduction strategies. Poverty measurement is very controversial because the approach adopted may affect conclusions about the economic reforms. National agencies have demonstrated considerable commitment to developing a household survey to monitor systematically the effects of economic reform. 1.22 The absence of a price system suitable for assessing household welfare solely in money terms is also problematic and is worsened by high and variable rates of inflation. Moreover, much activity takes place outside the official economy that is used to monitor change. Food and housing, the most important items in most household budgets, are affected by enterprise access and, in the case of food, subsistence-type production, and their allocation still has a large non-monetary element (see Box 1.2). Alternative indicators of well-being could be useful. The share of food in total household expenditures, for example, has risen significantly since the late 1980s, when it was recorded at only 34 percent, to an estimated 65 percent in late 1992. 7 amS tS« 2 The Soviet system: emin8s wese Incee 'The *re altemative vion vegot4io the prevalsauo of ineqUaty under thb ftoer coanist regmes to Easter Erop nd the Soviet Uniao. CA the one bhnd, the main pillar of the system-in particulaer, the state amership of productive capital, guaratee of .wt ,Owsst. eentralld "at deterainetion, and ext=sive provision of public goods--suest e6gtarilAs. (in the other hand. one could emphasis* such Important features as the holding doon of mney weges, the importanace of incentives in wage sett,in, aWopriatifn by the political elite, ed the distinction between earnings and income in a xe8im at scarcity. Atkianeo and S*ck14right (*9S*) found that the disparities between the Soviet republics were in faet comparable to those in Britain (depending co the dates chosen. Although within the individual republis, including tho Russian Federation in 1959, income inequality wes significantly lower than for tho USSR as a whole. TbXs box briefly explorea the problem of interpreting money ikCOmes. The key concept of equsality in the Soviet Union was that nininaL earnings were not allowed to diverge significantly froam the average and prices were fixed at levels low enogh to provide even the poorest with minimum subsistence. The strAotu*e of relative "egas releetted low return to skill, or trade. In practice, however, the stated principle or egaLitarian distribution was not fulfilled since vast goods were invariably in scareq supply. in 199O, an estimated 68X of clothing purchases were made on the black market Braithwsbe. 1M91). Aetual well-being became vary m-ch. a mattar of access rather then nominal earnings. Individuals who lacked such acces wer forced to pay much higher free market price..: in coQtrekt. Pasty officals received salaries tht Ore ually quite modest, but their practically umnmited acceass to vesy heap or free -goods ed service (including pecial cUnics and spas) placed them at the top th t socil ladder and e r the Part attractive. :Ases ito asic gooasud svie.sin s.aeorn vouchers, end many othe 0oomotU"e ildfing food, was *Ae' Ovided thoug tah w*ok alac. Ther i a h positive correlaticn estween enterprise sioe (ss mesure *by mloyepnt) -an the prevision of boneffitis.s I onduistrieS. catrm ee contq6d to he reltively p4tiviad comparod with non- .industrial *ater etor. yeea . 'deed, a sar e. aae nociil fund e#Ct in 199Z by tbo larget eterprsesaa1 hele inaitai te -real oenspia of :therwrese Chapter I).: -~~Une - thVer fiomr symew therefore,w reXithe.ciah incm or eniture wold he accurate indicators oaetal welein. these iniciators haeroba eA "A"r einlm £ollcin6 the price Uhetalisatton, although access to s dibsiised . e d services tbrh t Ie ben remarnt 1.23 Poverty assessment generally assumes a predetermined threshold of consumption or income (poverty line). Broadly speaking, the measurement of poverty can proceed on the basis of either relative or absolute concepts.2 The most common approach to an absolute concept of poverty is to estimate the cost of a bundle of goods that will assure that basic consumption needs are met. This has characterized the development literature on the subject. In industrial countries, however, the concept of relative poverty has been more important, including in Russia, where the social minimum income was set r-t about 50 percent of the national mean in 1975 and 1985. Russian policymakers should move toward absolute measures of poverty, at least during the transition period, so that viable targeting efforts can be developed. The concept of minimum subsistence income can be utilized more effectively to direct assistance to the increasing proportion of the population likely to be classed as in need. a/ See Ravallion (1992). Another possibility is to recognize the inherent subjectivity in the notion of poverty, end simply ask individuas what income level they regard as absolutely minimal. The answzr tends to be an increasing function of income. 8 Develogment of the Minimum Subsistence Basket 1.24 In the past, Russian researchers have focused on the concept of a socially acceptable standard of living for all citizens. The social minimum was introduced in 1975 at R 50 per person per month, and then was raised in 1985 to R 70. In both cases it was equal to half the average per capita income. It was supposed to serve as a basis for the minimum pension and minimum wage, although in practice neither was adjusted upward quickly enough to maintain its real value. McAuley (1979) concluded (albeit cautiously given data constraints) that some 35 to 40 percent of the Soviet population had per capita incomes below the official poverty line in the late 1960s. By 1992, such a large proportion of the population fell below the official minimum that the concept had become irrelevant for policymakers. In Altaiski Krai, for example, the local basket was not published by government authorities in 1992 for fear that expectations of social assistance would be unduly raised. 1.25 The social minimum income has been widely criticized, not only because it does not enable policymakers to target assistance to the poor, but also because it is based on a high-protein/high-fat basket of food (selected several decades ago) that exceeds the World Health Organization/Food and Agriculture Organization of the United Nations (WHO/FAO) recommendations for daily calories and fat intake and is associated with cardiovascular and other chronic diseases. Moreover, economies of scale in household consumption were excluded. Although estimates vary, it is generally agreed that the maintenance of a given standard of living for a family of four requires less than four times the income of a single individual. Table 1.2. Canges in 3flli Subsistence 1anm., Suaim Vederatin, 1992 and 1993 Year Month Average per capita Able-bodied population Pensioners Childrn 1992 January 635 718 438 62 Februay 744 875 535 78 March 1,031 1,165 711 1,010 April 1,226 1,385 864 1,202 May 1,405 1,588 970 1,377 June 1,639 1.852 1,131 1,605 July 1,779 1,898 1,220 1,985 August 1,939 2,204 1,358 1.836 September 2.163 2,448 1,526 2,056 October 2,584 2.929 1,779 2,5i2 November 3,285 3.714 2,264 3.179 December 4,282 4,832 2,934 4,177 1993 January 5,547 6,263 3,809 5,409 February 6,756 7,633 4,683 6,550 March 8,069 9,030 5,543 8.034 April 9,683 10,836 6,652 9,641 May 12,897 14,268 8,891 13,169 June 16,527 18,130 11,305 17,303 July 21,206 23,376 14,870 21.659 August 24,764 iu,815 17,382 24,920 September 28,163 31,809 19,848 28,1S5 October 32,400 36,580 22,825 32,390 November 37,908 42,799 26,705 37,896 December (proJection) 44,352 50,075 31,245 44,338 So For calculation or the minifum subsistence level In Russia see par-graps 1.Za-6. This table snows iacone levels with an average of 68 percent spent on food, 75 percent for children ages 0-6, 73 percent for children ages 7-17, 62 percent for adults, and 83 percent for older adults. Sour¢e: See Pookin. Mohsina. and Baturin (1992). 9 1.26 An alternative basket has been developed, based on a diet that is both nutritionally acceptable and consistent with Russian food habits and traditions (Popkin, Mohzina, and Baturin 1992). It proceeds from an assessment of consumption levels of the poor (less than R 150 using the Family Budget Survey from the second quarter of 1991). The changes in food prices, in particular the elimination of subsidies for most meat and dairy products, have shifted consumption considerably. This is reflected in the revised food basket, which contains an increased proportion of bread and potatoes. It represents a much healthier diet that more than meets the WHO/FAO recommended daily nutrient requirements. 1.27 The next step in developing a poverty line is to derive an income level associated with the subsistence diet. An assumption that 80 percent of income is spent on food might be too high, especially over the longer term. The replacement of consumer durables and clothing is certain to affect income expenditure, as is the removal of remaining price controls. On the other hand, while housing and basic utilities remain subsidized, a threshold set at twice the cost of the food basket may be too generous. Setting income expenditure for food at 68.3 percent, the subsistence minimum was revised and adopted by the Ministry of Labor. This forms the basis for the results presented below. Table 1.2 presents a summary of the costs of the food basket, and the corresponding subsistence income, for 1992 and 1993. It reflects, among other things, the difficulties in dealing with nominal values. Unfortunately, regional price variations are ignored in the preliminary Russian Longitudinal Monitoring Survey (RIMS) results. Later work should take into account information about local prices, which is likely to lead to different poverty lines being drawn in different regions, and perhaps to different poverty profiles. 1.28 The World Bank has been providing technical and financial assistance to the Russian Goskomstat for a RLMS, to study the social and economic needs of the population and to measure the impact of the crisis and the transition from a command economy. The survey of 6,500 households will provide relevant and timely information on different population groups and provide a basis for the forthcoming Russian poverty assessment. Preliminary results of the initial rounds of the survey inform the analysis below. Box 1.3 highlights several important qualifications to be borne in mind when using the RIMS at this stage. It is important to monitor a range of variables relevant to poverty and behavior, starting with the base provided by the RLMS and complementing assessments with alternative measures of welfare. Findings of the Longitudinal Monitoring Survey 1.29 Latest data suggest that up to a third of Russian people live below a minimum threshold income. There are, however, serious difficulties involved in measuring income in the Russian Federation given the importance of such in-kind benefits as housing (see Boxes 1.2 and 1.3) as well as informal sector activities. If, for example incomes were under-reported by 20 percent, then the proportion of the population below the minimum threshold would fall from about 33 percent to about 23 percent. Further analysis being undertaken through a World Bank poverty assessment will explore consumption and non-income welfare indicators. 10 1.30 While the actual magnitudes are as yet uncertain, the composition of poverty and its relative incidence is fairly clear. Preliminary results from the RLI4S suggest that the incidence is most serious among households affected by unemployment, households with two or more children (especially under age six), and households with disabled members. As yet, there are relatively few openly unemployed, although this is expected to change markedly over the coming months, further altering the composition of vulnerability. There are substantial differences in the incidence of poverty among different groups (Figure 1.2). The majority of the poor come from working households, and the next most significant poverty group is the elderly (see Figure 1.3). There are relatively fewer people in the other groups that are seriously affected, such as the disabled and students. Box 1.3. Methodology of the RLM: Csaigts and Limitations The initial rounds of the BLUB were organized and undertaken in difficult circumstances in September and October 1992. The results provide a useful basis to understand the impact of the economic crisis and transition. What is presented in this report, however, represents preltminary results in a rapidly changing situation. Household aurvoys all over the world constantly face the contradictions .posed by levels of recorded consumption well An excess of reported income, There are important additional qualifcations to be kept in'mind in itterpreting the conclusions of the RUL to date, which cen be briefly sumuarised as follows: * .LVO consumtion ba*.' One poverty threshold *we used for all regions, based on a single consumption basket priced using an aveza8e of national.prices. This inotres the significant variations in prices 4nd possibly consumption patterns aroas' Russia. e Ronivalence scOaes. Clealy may scales used reflect a particular set of assumptions about the effect of economies of sca2e within the bousehold. Nonethleiss, it is unlikely that the approach adopted here, which completely g s th econmies, is satisfactoty. There is a risk that, the high rates of poverty for families *ithbchildArn are simply "by construction," rather than a reflection of reality.-. * SonZwas benefits and serVices. These were only partially included (housing was not), and the values imputed appear to be underestimates. Obviously, certain aspects will be difficult to quantify, though they are important. For example. the extent of access to subsidiaed food stores significantly affects household purchasing power. 1.31 Spatial Distribution of Poverty. Distribution patterns for money income varied considerably during the Soviet period. Some pockets of extreme poverty in money terms were revealed by the Family Budget Survey (Popkin 1992b). The regional impact of the crisis and transition is likely to increase these differentials. Recent evidence suggests that regional income differentials have increased markedly, although variations in prices make it difficult to evaluate the corresponding impact on welfare. Preliminary RLMS results reveal large variations among the twenty-one oblasts surveyed. Regional unemployment rates will vary according to, among other things, industrial concentration and the pace of reform. Industrial regions are experiencing the most severe output contraction. In 1992 unemployment rates in Yaroslavskaya oblast and North Ossetinskaya, for example, were about five times the national average. Such variations are particularly troubling because of the decentralized approach to social protection and the absence of systematic needs-based federal transfers that could direct financial support to poorer regions unable to mobilize 11 sufficient resources for basic social assistance.3 Several oblasts hit severely by unemployment are already close to exhausting their Employment Fund revenues. 1.32 Large Famllles. It appears that the majority of poor households are families with children. Up to 17 million children in Russia are estimated to live in households reporting incomes that fall below the subsistence minimum threshold. The incidence of poverty is significantly hIigher among families with an above average number of children. In European Russia this means two or more children.4 In the late 1980s, there were 823,400 families with more than three children, of which almost 100,000 had five or more children. Government assistance to families has been quite limited, and has recently been diminishing in value (see Chapter II). Figure 1.2. Incidence of Vulnerability in Russia Among Selected Groups, September 1992 9.1*w 0e subel*. miliftm IA mm tL than 60% or u,Abbnso mtnimulm with elld, . wItbebIsauuE. , stF -_w. * RO W M ..... .. ..... 1ine : ; .so 0 SO S~as5 40 a0 so105 Nogt: Based on a miniumn subsistence food basket, consistent with WHO guideUnes, which represents an average 68.3X of household expenditure. Source: Russien Lo situdinal Monitoring Survey / This issue is explored in detail in Chapter 1II, in Financina Social Protection. 1/ Various government policies designed to increase fertility were largely unsuccessful, and were abandoned under restmika. Tn St. Petersburg, for exampla, almost three-quarters of families have only one child. 12 Figure 1.3. People Below Minimu Subsistence Income Level in Russia by Type of Household. September 1992 UMout IX ao o, 8 Note: Threshold Income Is boased on a minimum aubaleteno food basket consistent with WHO guidelines, that represents an average 686.% of household expenditure. Souroe: Russian Longitudinal Monitoring Survey Prellminary Results 1.33 Preliminary results from the RIMS suggest that almost one third of families with children lived below the minimum threshold income in the latter part of 1992; among families with three or more children, the relative incidence of poverty is much higher. Many of the latter group reported incomes at less than half the poverty line. Looking at the figures slightly differently, the incidence of poverty among households with young chlldren is also striking. According to the RLMS, in September 1992 rates of poverty among families with children under age six were significantly higher than for households with no children under six. It is important to note, however, that household economies of scale in such areas as food purchases and preparation are excluded. 1.34 A profile of large families gathered from surveys of Taganrog (an industrial city considered representative of European Russia) shows that the parents' educational background is relatively low (Prokofieva 1992). Among mothers of large families, only 4 percent have higher education, and 42 percent have less than secondary education, compared with national averages of 11 percent and 21 percent, respectively. 1.35 Poverty in large families can be partly attributed to the relatively fewer active (that is, paid) members in the household. Prokofieva (1992) reports that in 40 percent of the cases, mothers were either not working or were poorly paid, and that their employment status and career prospects were often deteriorating. This situation is likely to have worsened with rising unemployment, a category in which women are over-represented. 1.36 Incomplete Families. A significant number of families with children under eighteen are incomplete, that is, without both parents. The majority are headed 13 by women: 2.2 million families (11 percent of all households) consist of mother and children only; households with father and children constitute only about 1 percent of all households. Despite special government allowances,5 the incidence of poverty among female- headed households with young children is even worse than for families generally. The initial RLMS results suggest that up to 55 percent of female-headed households with children under six and 48 percent of those with children under sixteen reported incomes below the subsistence minimum. 1.37 Families with one breadwinner often fall below the subsistence minimum after the birth of a child: up to 30 percent in the case of the first child, and 50 percent for subsequent births (see also Gazes and Le Cacheux 1992). Households rely.ag on a single breadwinner with low earnings are currently at high risk because of the erosion of family allowances and because widespread female unemployment and partial employment have had serious repercussions for this vulnerable group. 1.38 The Unemployed. One group that did not feature prominently in traditional surveys of poverty in the Soviet Union was the unemployed. The economic crisis and transition has been associated with increasing levels of joblessness, although at slower rates than expected. According to the Federal Employment Service, as of December 1992 about 1.5 million Russians were out of work, one- third of whom were registered as unemployed (see Box 4.2 in Chapter IV on measurement of unemployment). The shape and dynamics of the transition in the Russian labor market are analyzed in Chapter IV. The focus here is on the Lmplications of unemployment from a poverty perspective. 1.39 The replacement rate of an individual's unemployment benefit is based on work history and past earnings (see Box 2.1 in Chapter II). With high inflation, the real value of benefits based on past earnings erodes rapidly. In early 1992, when benefits were calculated on the basis of the previous year's average earnings, the majority ended up with the minimum benefit. Although a change in the rules for calculation reduced this phenomenon, it is estimated that more than half of all recipients receive only the minimum, with as many as 80 percent of recipients in some regions receiving the minimum. The average unemployment benefit paid in December 1992 was R 1,064, which was far below the subsistence minimum (R 4,282). The minimum benefit stood at R 900 (equivalent to the minimum wage). 1.40 The duration of unemployment is an important dimension of household well- being, since it is easier to cushion current consumption levels during temporary spells. Yet in areas of relatively high unemployment, such as the Moscow, Kaliningrad, and northern regions, durations are already more protracted. In Moscow, nearly 20 percent of the registered unemployed reported durations exceeding eight months in July 1992. Women, who are already dtsproportionately affected by unemployment, also experience longer spells of unemployment (see Table 4.2 in Chapter IV). The maximum duration of unemployment benefits is twelve months. 1.41 Many of the unemployed do not register themselves as such, perhaps because many vacancies are not known to the employment service, and unemployment benefits S Transfers are made only to families living apart from each other, not to the many that are incomplete *nd do not pool resources but that continue to llve together (that is, because of housing shortages). 14 are low. Nonetheless, there was a rapid expansion in the proportion of those eligible for benefits: from 19 percent of the narrow unemployment measure in December 1991 to 60 percent in September 1992. This can be explained by the delayed termination of benefits to individuals whose severance pay has expired and the rising share of eligible categories. The Federal Employment Service estimated that 56 percent of the unemployed were receiving benefits between January and July 1992, ranging from only 13 percent in Kaliningrad oblast to 98 percent in Aginsky-Buryatsky autonomous oblast. It is not clear how those without any state income support are coping; some may be working in informal markets, others are turning to household production. Considering the low level of benefits and the high rate of nonregistration, it is not surprising that preliminary results from the RLMS show the incidence of poverty among families affected by unemployment to be relatively high, of whom 14 percent reported less than half the subsistence minimum income. 1.42 Who are the unemployed in Russia? At the end of 1993, they were overwhelmingly (74 percent) women. This average masks regional and local concentrations of unemployed women that are as high as 89 percent in the Nenets region. And as mentioned above, unemployed women looking for work have had greater difficulty than have unemployed men. The overall placement rates of the registered unemployed were 50 percent for men, but only 23 percent for women. 1.43 The Elderly. The number of old age pensioners is large (partly reflecting low entitlement ages) and is increasing in both relative and absolute terms. Almost 19.5 percent of the population fell into this group in 1990, as compared with 17.3 percent in 1980. For the reasons mentioned above, women greatly outnumber men among pensioners. 1.44 The elderly represent an important dimension of poverty. Preliminary results from the RIMS suggest that, overall, about 19 percent of poor households are headed by an elderly person. Surveys in Taganrog in 1990 showed that 30 percent of those in poverty were lone pensioners, and 32 percent were pensioners in working households. 1.45 Not all pensioners are poor. In fact, the incidence of poverty among households with elderly members, as revealed by the RIMS for September 1992, was lower than for the general population.'6 Note, however, that the RIMS results are based on households, and almost half the pensioners in the sample lived in households with non-elderly members. In about one-fifth of cases, the pension is r,ot the sole source of income for the recipient: there is no prohibition against working attached to receipt of pensions, and the entitlement age is relatively early. A recent ILO survey found that older age workers constitute a fairly significant share of total employment (15 percent), and that their sectoral pattern of employment reflected that of the labor force generally. The survey revealed little evidence that older workers were marginalized in the economic downturn of 1990-91 (Standing 1992). 1.46 Beginning in the 1950s, there was a gradual deterioration in pensions, with the average replacement rate falling from more than 60 percent in 1956 to less jV This may be due, Xi part, to the ti4in of the survey; a penion increase ma have teporarily appeared to improve household well-being. 15 than 33 percent in 1990.7 By early 1992, this had changed, with average pensions reaching relatively high levels in historical terms (at 82 percent of the average wage). By September 1992 this level had fallen to only 20 percent. Box 1.4 presents some alarming evidence about the welfare of the elderly. 1.46 About half of the elderly received only the minimum pension in 1992; minimum pensions fell below the minimum subsistence for the aged throughout the year (see Table 2.2) in Chapter II. Automatic quarterly indexation of pensions was introduced in early 1993. Local authorities may raise additional resources to assist the elderly, depending partly on the f )nomic base and the age structure of the population. For example, in Sumara, which has a relatively aged population, extensive social assistance programs could be maintained because of the proximity to industry, and in Tuman, a young region engaged in oil extraction where retirement tends to be early, there were financial resources to pay higher pensions. This can be contrasted to Novosibirsk, where in September 1992 local supplements benefited less than one-third of pensioners, and averaged only R 316 per person. 1.47 The Working Poor. Focusing on demographic categories such as the elderly and single-headed households runs the risk of neglecting an important dimension of vulnerability among the so-called working poor. 8 There is a significant group of working poor people in the prime age group (ages 20 to 50) in the Russian Federation. Preliminary RIMS results suggest that almost 66 percent of poor households are headed by someone who works in an enterprise, a cooperative, or a state or collective farm. 1.48 The presence of the working poor under the Soviet system is explained, in part, in terms of low minimum wages. The minimum wage, which was raised from R 60 to R 70 in 1977, remained at the same level until 1991. In November 1992 it stood at R 900, which was below the minimum pension and only 27 percent of the minimum subsistence levels. In 1993 it was indexed periodically to reach almost R 8,000 by July. Survey data suggest that some enterprises, particularly in engineering and light industry, have begun to place large parts of their labor force at (or near) the minimum wage (see for example, Box 4.1 in Chapter IV). j/ Average replacement rate is defined here as the ratio between the average pension and the current average wage (as opposed to the relationship between the individual's pension and past earnings). O There is some circular reasoning being employed here. Of course the working poor are poor, by definition. However, these tautological terms are in general use. 16 -,45. Well-Boins of th* Eledrly Alamins insight. into the UviAn conditions of the eldesly emered from surveys that weo carried out in Moscow, Yaeaterlbug., ad YV'!svaM im the tirat half of 1992. The U.S. charitable asency, CAMS, carried out surveys betwea April .ad December 1992, using a random sample dawn from a List of all pensioners in each siteo Tbe initial results contain evidence of possibly severe danger to life and health-in particular. reporta of lets. weight less--although it muwt be noted that the date axe not dernved from actual measureomnt but ftom self-reporting. Sixty-cne percent of pensioners in 8,sccw and 401 in Yiketerinburg said they lost five kilostaae or more in the past six months, and when responding to a modified qy1stian, 501 of respondents i Ylerevan reported losaig five kilograms or moxe. Subsequent surveys An reports of Gadaria, St. Petersburg, and Itkurtak delivered strikingly aimilar resuLts on weigbt less, with a range of 482 to 521 reporting a lose of more than five kilograms ever the previous aix months. The picture was correspondingly grim in terms of income security. ln April, median peasions in Moacow (R 410) and Yetatriniug, (R 450) were well below the K 780 estimated by the World Benk to be the minimum subsistence level at that time. In general, the results were internally consistent--those at greatest risk (from health problems, as well as socioeconomic stress) had the most frequont reports of weight loss. The most striking similarities lay in the very limited amounts of savings held end in the minimal suppot. being provided by telatives, neighbors, and friends. The mail differente arose batween rural and urban are"a. The former had wuch greater access to household end land plots, end therefore to home-grown vesetables. Although these surveys are currently incomplete, fragmentary, and not wholly consistent with alternative data sources, the inital .resuLts point to an extremely erious situation. Source: MAUE, soaow 193 1.49 Despite higher education achievements than men, women outnumber men in low- skilled labor two to one.9 This is translated into differences in average wages, which for women are about two-thirds those of men.10 In Taganrog, 95 percent of industrial workers earning the minimum wage were found to be women. This pattern is especially evident in feminized occupations and industries: overall, the higher the proportion of women, the lower the average wage, a tendency that has increased in recent years and is fairly robust across sectors and occupations. 1.50 Recent economic developments have heightened the need to focus on the working poor. This group has been badly hit by sharp declines in their real (cash)"1 wages. Increasing numbers of enterprises have placed workers on minimum wages, and have resorted to short-time work and/or temporary closures in the face of demand- and supply-side shocks. Rules and practices in cases of short-time work vary, although generally there is either reduced pay or no pay at all. This phenomenon has forced millions of workers to accept drastic wage cuts in return for the possibility of retaining their jobs. A Goskomstat survey of industrial enterprises in August 1992 found that about one in four workers was either on forced leave or engaged in short time work. The incidence of partial employment varies across branches of industry and across regions. Certain 2/ The mismatch between women's lower skill levels and their higher educational qualifications is far greater than for men: 621 of all workers with higher or specialized education are women, and 382 are men (see Annex 2). a/ Women are reported to accept joba at Loer levels of skill and remuneration in exchange for such non- monetary compensation as aecess to childeare. Their responsibility for housework and child care mar assume priority over career considerations, because of the inadequacy of other support mechanisms and the traditional unwillingness of Russian men to take on domestic tasks. )Al The distinction between cash (monetary) and non-monetary compensation is important, since the value of the latter has tended to be maintained in real terms. 17 branches, such as engineering and light industry, have resorted to reduced shifts. In Altai Krai, local authorities have reported that virtually all enterprises were operating on a short-time basis. In one case, for example, a textile factory had been closed for six weeks, and workers were being paid an average of R 3,000 a month instead of the R 8,000 stipulated in their contracts. Women's concentration in textiles suggests that they are more vulnerable to partial employment, over and above their higher unemployment levels. 1.51 The repercussions of forced partial employment could be very serious, especially in households with one wage earner or where more than one wage earner can be put in this position. Evidently, workers do not feel that they have any alternative but to accept reduced compensation; few respond by leaving the enterprise. Workers forced into partial employment are probably also those most likely to become unemployed, making them particularly vulnerable to poverty as household savings and reserves will already be diminished. (Some may be turning to the informal sector for income-generating activities, although further analysis is required.) 18 CHAPTER II: CHANGING ROIES IN SOCIAL PROTECTION 2.1 Social protection systems involve three main sets of actors: individuals, enterprises, and the state. Their respective roles vary across economic systems and over time. The current situation in Russia, shaped under the Soviet system, exhibits characteristics that are inappropriate in a market-oriented economy. The most obvious is the direct utilization of enterprises as a welfare instrument of the state. Other shortcomings include large gaps in coverage and the inadequacy of benefits. Developing an effective social protection system would involve changing roles and relationships, so that the enterprises are transformed from transfer guarantee agencies to providers of productive employment, and individuals become active participants in the market economy. In the transition period, however, special measures will be necessary to avoid unnecessary hardship and preserve certain achievements of the Soviet era, while establishing a system that will be viable over the long term. This chapter explores these issues, with emphasis on the existing state system, thereby setting the stage for an analysis of the options facing Russian policymakers. The Soviet Legacy 2.2 In the Soviet era, the Indlvidual enjoyed security in state employment and received various social benefits through the enterprise. Job dismissal was a rare occurrence even in cases of redundancy or incompetence. On the other side of the coin, there was a positive obligation to work, mandated by the Constitution. Until recently, this meant working for the state or state-owned enterprises, since private sector and individual initiative and economic activity were severely restricted. Moreover, there was no freedom of movement, as the state imposed constraints through a system of residential permits (propiskl), and housing shortages were endemic in urban areas. Labor morale and effort tended to be low. 2.3 In return, the Soviet state guaranteed citizens employment in a centrally planner' economy based on state-owned enterprises. Basic social services were avail- ,le without charge, although access and quality varied, partly because of the importance of and variations in enterprise-provided services. There was assorted, although limited, assistance for those who did not work. In principle, price controls on basic consumer goods ensured affordability, although not availability, since shortages were pervasive. Moreover, hidden inflation was substantial. In the absence of regular indexation, recipients of income transfers experienced gradual erosion of the real value of their benefits. 2.4 The social protection system under state socialism comprised separate pension programs for the non-agricultural sector (although including those working on state farms) and for collective farms, cash benefits for sickness and maternity, allowances for families with children, and limited social services for the single elderly and disabled pensioners, and for orphans. While pension eligibility depended on relatively low age and tenure requirements, the system was similar to those of Western countries in several respects. Benefits were wage-related, with separate formulae for the urban and agricultural labor forces 19 (as in France and Germany).' Pension financing was based on the "pay-as-you-go" method, whereby current obligations to beneficiaries were funded entirely by the current generation of workers. The Soviet social security system was financed primarily from general revenues. 2.5 The official social insurance system was the only income security program available for the average citizen. There was no private insurance, and only limited social assistance from subnational governments was available to those suffering economic hardship. The existence of the poor was not officially acknowledged, since it was thought that all adults not working full time fell into clearly defined groups (the disabled, for example) whose needs were supposed to be met through categorical entitlement programs. Indeed, the Soviet system was notable for its limited public assistance to those who never worked and to those temporarily unable to work, (say, because of a need to stay home with young children). Assistance for the elderly and the disabled who did not qualify for pensions was meager. Thus, Soviet welfare programs had significant gaps- -a state of affairs that had contributed much to the poverty of the disadvantaged groups described in Chapter I. 2.6 From the mid-1980s, partly in anticipation of growing hardship during economic restructuring and the state's changing role in social protection, a series of incremental and systemic changes were adopted. By 1991 the Soviet regime had overhauled the state pension program; initiated voluntary supplementary pension plans; established an employment service to provide training, job referral, and cash benefits to the unemployed; enhanced cash allowances paid to families with children; extended in-home services (such as help with shopping); and set up territorial centers for short- and long-term care of the elderly living alone and of the disabled (in addition to the established boarding homes). Moreover, subnational governments and community organizations were permitted to engage in charitable activities. Each of these aspects is examined later in the chapter, in the section on the system of cash benefits. 2.7 Now, focus of the federal agencies--in particular the Ministry of Social Protection (MSP)--remains on conventional categories of need, such as the disabled. There is a real risk of bypassing the large and growing sections of the population who are being hit severely by the economic crisis and adjustment, but who are not eligible for any federal social assistance. What is 'learly lacking is some measure of last resort to help prevent people from falling into poverty. 2.8 Current social assistance programs have a number of defects. First, coverage is not comprehensive, and it is especially deficient with respect to emerging vulnerable groups: women, who are disproportionately represented among the unemployed and the poorly paid; those who are unemployed but not receiving benefits; and the working poor (see Chapter I). The MSP continues to focus almost exclusively on the elderly and the disabled as the major client groups of social assistance (see Annex 1, Table A1.6). Second, excessive local variation I/ Unless otherwise specified, for rferences to non-Soviet social security programs, see Social Securitv Proimams throushout th< World, 1989, Washington, D.C., .S. o,ernment Printing Office, 1990. 20 leads to horizontal inequities among individuals.2 This raises the question of the respective roles of central and subnational authorities, especially with respect to financing.3 Although it is appropriate for subnational authorities to determine which individuals satisfy the criteria for social assistance, it is not their province to determine the criteria themselves. Limited available evidence suggests that subnational authorities are capable of responding to local needs. But the difficulties of designing a comprehensive social assistance scheme--one that takes full account of interactions between the various cash and in-kind benefits--must not be underestimated. It is not a job for those consumed with trying to avert widespread disaster. Central guidelines on the circumstances under which help should be given, and what that assistance should be, are urgently needed. 2.9 Third, subnational social assistance is typically provided in kind, such as food vouchers. Any cash payments made are typically for specific items, for example, to help with heating bills. In a time of rapid inflation, as a means to guarantee the maintenance of minimal consumption levels and where social assistance is regarded as a temporary expedient, assistance in kind rather than cash may well be appropriate, and overriding concerns, such as for nutritional needs, may justify such measures as providing milk at school. But international experience shows that the administrative costs of such schemes are higher than the cost of income transfers. Cash transfers allow greater transparency in the budget and have less distorting effects in goods markets. 2.10 Fourth, there is no guaranteed source of funding for the subnational measures that do exist. There is growing reliance on extrabudgetary sources, but funding for these is typically unstable and the processes of allocation are not transparent. This makes it difficult for subnational governments to plan recurrent expenses. It has also prompted many oblast and raion governments to invest their initial capital in what they hope will be profit-making endeavors in order to expand their resources (or at least to maintain the real value of the unused funds) (see para. 2.58). In consequence, virtually none of the forms of subnational social assistance is guaranteed to individuals over time. Assistance that is awarded in January may not be available again in February. The result is that social assistance to the poor varies over time and across regions according to available resources. The implications for poor, vulnerable groups are potentially devastating. S/ For example, a retiree in Chelyabinsk Uving on the normal (November 1992) minimum pension of R 1,320-- when the poverty threshold was estimated at R 2,464--could expect to receive a bread donation (R 250) and free use of public transport (a value of R 60 per month). Other forms of assistance might have been provided as available, although they could not be relied upon on a regular basis. Sone pensioners could receive a book of vouchers entitling them to a meal every other day in a local canteen (a monthly value of R 450) and, depending on the availability of resources and whether they had received similar help recently, a pair of shoes (a value of R 600). Programs vary between oblasts. For example, Altai Xrai ran no comparable bread scheme for retirees. Indeed, evn within Chelyabinks, the criteria employed by the raions for this schema varied, depending on local resources and the assessment of local need. OThe Russian Federation is a three-tiered federal state. The central government has direct relations with the ninety-one oblast level governments. Below the oblasts have approximately 2,200 raions and municipalities subordinate to them. This report refers extensively to "subnational governmuents' to inclnda oblasts end raions, which have extensive social sector responsibilities. 21 2.11 State enterprises traditionally played a hybrid role, both as a purchaser of labor services and as a financier and provider of social services. Money wages were low, although supplemented by access to non-wage benefits (see below). The basic function of managers was to follow plans that covered most aspects of technical, financial, and industrial activity. Enterprises thus tended to behave as output maximizers, hoarding labor and receiving extensive subsidies from the state. Excess employment was widespread and substantial--estimated at around 20 percent in 1990. It appears to have arisen under conditions of soft budget constraints, prompted by a combination of output objectives, the unreliability of other input supplies, and the needs of production technology. 2.12 At the same time, the state system of social protection imposed significant administrative burdens on employers, including the payment of sick and maternity benefits and family allowances and preclaims activity on pensions (including copying out the employee's workbook in full). The Soviet state mandated a series of leaves and concessions to facilitate the high labor force participation of women, who were also expected to undertake housework and child care in the home. The burdens implicit in these practices were shouldered by enterprises in the interest of social responsibility, as was the provision of other social services. 2.13 Under the Soviet system, in common with other socialist economies, workers' compensation packages often included housing and such services as day care and kindergartens (see table 4.1 in Chapter IV). Public enterprises often undertook capital investments to benefit the local communities--for example, the construction of schools, which would then be transferred to subnational authorities. Nonetheless, the relative significance of enterprise provision of social services should not be exaggerated. Where an enterprise acted as a locally dominant employer, it would often fulfill the service functions that would otherwise be performed by a subnational authority. On the other hand, among smaller enterprises with less local labor market dominance, survey evidence suggests that enterprise-provided social benefits may be much less important than those provided by the state. In the Bank sample, enterprise expenditure at the local level for housing, child care, and health facilities was comparable to, or exceeded, subnational authority expenditures in only 10 to 30 percent of cases, depending on the function. 2.14 Housing is the most significant benefit provided by enterprises, accounting for an estimated 14 percent of total labor compensation. Enterprises, especially in areas with persistent labor shortages, relied increasingly on the provision of housing to attract labor and reduce turnover. In 1992, enterprises managed about 57 percent of the total state-owned stock, although many enterprises were seeking to transfer responsibility to local auithorities. Regional differences show that the share of enterprise housing is highest where heavy and defense industries predominate, which reflects previous priorities in industrial policy. Beyond the direct provision of housing, there was an array of in-kind and cash forms of assistance; for example, until recently, housing construction loans were channeled through enterprises. 2.15 The foregoing picture should not be interpreted as a denial of private initiative under the former system. In Russia, as in other former socialist economies, private sector output and employment tended to be severely underreported. Many individuals engaged in private enterprise in addition to 22 their official job. Even before the reforms began, many households resorted to ways to produce and exchange goods and services outside the official planned economy. 2.16 The impact of subnational nongovernmental organizations (NGOs) has been shaped, and limited by, distinct features of the Soviet context. Foremost, the government's extensive involvement in economic and social life inevitably limited the scope, and arguably the need, for private welfare-oriented organizations. The Soviet regime mistrusted social activity not specifically authorized by the state. Beyond a few high-profile, state-approved, and highly centralized organizations, such as the Children's Fund, charitable activity was very limited during the Soviet period. Historically, independent organizations tended to be part of the political dissidence movement, rather than groups formed to meet direct welfare objectives. Only in the past five years or so have the latter type of organizations begun to evolve, generally from the grass-roots and, as elsewhere in the world, through the commitment and energy of dedicated individuals. In 1991 a Law on Public Organizations granted citizens the right to form charities. The System of Cash Benefits in 1992 2.17 Federal cash benefits in Russia are funded from the budget and four extrabudgetary funds: the Pension Fund (PF), the Employment Fund (EF), the Social Insurance Fund (SIF), and the Fund for Social Support (PSS) of the Ministry of Social Protection. Table 2.1 summarizes the sources of their financing together with benefits provided. In 1992, the Russian Federation spent roughly 7.3 percent of gross domestic product (GDP) on providing cash benefits from these four extrabudgetary funds and the federal budget, and for the first nine months of 1993 it spent 7.8 percent.4 For comparison, in 1990 several countries were devoting 12 to 15 percent or more of GDP to public pension expenditures alone (see Figure 2.1). The Pension Program 2.18 The Russian pension program, which took effect in March 1991, covers old age, disability (general and work-related), and survivors' benefits. There are two broad types of pensions: labor pensions, which are paid on the basis of a contribution record, and social pensions, which are paid to individuals with fewer than the five qualifying years of employment. About 75 percent of all pensioners (34.5 million people) receive the former. The Pension Law of 1990 was amended on several occasions in 1991 and 1992 to adjust the minimum pension and computation of benefits. Qualifying conditions and administrative agencies have remained unchanged since 1990. The calculation of pension benefits is complicated by several factcrs, including bonus for extra years of service and regional coefficients. (Eligibility and levels of benefits are summarized in Annex 1, Box Al.l, and numbers of beneficiaries by category are given in Table 1.1, Annex 1.) k/ This estimate also includes the cost of administering some of the cash benefits. In circumstances of rapid inflation, this estimate can only be an approximation. The overall level is likely overestimated beeause of some "double counting" of PF and federal budget expenditure. 23 Figure 2.1. Public Expenditure on Social Protection as a Percentage of CDP, 1990 Percentage of GDP 26- 20 - 19.3 16.4 16 12.6 iti 10 6 7.2 0-4 Paraguay Mexico Kyrgyzatan Russia Czeohoslov. Turkey Chlie OECD Low Ino. Brazli Hungary Eg=: Includes cash benefits (pensions, unemployment, and social assistance), but may exclude some subnational assistance. Subsidies are not included. Figures for Russia and Kyrgysatan are for 1992. 2.19 Given the relatively large number of pensioners in Russia and rapid inflation, the issue of the adequacy of pension benefits is controversial. According to the 1990 law, pension benefits are to be revalued annually by the Parliament in line with the cost of living. A series of amendments have sought to partially compensate for price inflation. In addition, the provisions for the benefit formula and the basis of benefit adjustments have changed under the conflicting pressures to reduce public expenditure on the one hand and to protect minimum living standards for pensioners (whose savings have been virtually wiped out by inflation) on the other. In general, fiscal austerity prevailed in 1992. 2.20 During 1992 the parliament legislated revaluations of past wages that form the basis for calculating pension benefits. First, a series of coefficients were introduced for adjusting past earnings for the recalculation of benefits.5 Second, an amendment in April 1992 (effective January 1993) allowed pensioners to request that their benefits be calculated from wages over the preceding year. This is a temporary measure, however, and applies only to earnings during 1992- ,/ This revaluation upgraded past earnings by a coefficient ranging from 11.2 for pre-1971 earnings, to 2.9 for earnings in 1991. An October amendment furtber upgraded the 1991 coefficient to 5.5. 24 93. The parliament, concerned with overall pension costs, vacillated over lowering the pension ceiling, then subsequently raised the maximum benefit. 2.21 The basis for determining the minimum pension has been altered at various times. In May 1992, it was linked to the minimum wage, thus superseding the 1990 provision for adjustments according to cost of living. The link to the minimum wage was ignored in October 1992 when the minimum pension, but not the minimum wage, was upgraded by 2.5 times. Nevertheless, throughout 1992 and 1993 the minimum pension remained below the poverty line (see Table 2.2). An amendment in early 1993 promised to index the minimum pension quarterly, and the minimum pension was increased to R 4,275. On May 1, 1993, the minimum pension was increased to R 8,122, and the maximum pension to more than R 20,000. The replacement rate--the relation of the average pension to the average wage--has been fluctuating considerably over time. In February 1993 the replacement rate (relative to the average wage) was 31 percent, up from 18 percent in September 1992. But by July 1993 it had fallen back to 26 percent. From November 1, 1993, all pensions are to be indexed at a flat rate. 2.22 Policy, Financial, and Administrative Responsibilities. Through 1992 and most of 1993 the Ministry of Social Protection (MSP), the Pension Fund, and the Commission on Social Policy of the Supreme Soviet all played a part in arrangements that did not facilitate coordinatior between the financial and administrative components of the pension program The Commission determined policy and has the power of legislative initiat' e. The NSP was charged with coordinating the policy discussions. It is respoInsible for implementing the law as it relates to the structure of benefits through oblast and raion social security offices, and for calculating entitlements. The PF collects payroll contributions and manages the extrabudgetary fund independently from the Ministry of Finance; it reported to the Parliament through the Commission on Social Policy. In October 1993 the role of the Parliament was ended, and the PF was headed by government appointees. 2.23 The PF is an independent agency employing more than 8,000 people and is managed by a permanent executive directorate. In January 1993, the PF was responsible for the payment of more than 46 million pensions, of which more than 34 million were work-related pensions financed from the Pension Fund's own resources (see Table 4.2 Chapter IV). The bulk of PF revenues originates from employer contributions. In 1992 non-agricultural enterprises contributed 31.6 percent of the wage bill, farms paid 20.6 percent, and individual employees contributed 1 percent of their earnings. The self-employed contributed 5 percent of their pay. In early 1993 the rate for enterprises' contributions was reduced to 28 percent (see below). 2.24 The PF greatly reduced funding responsibility for non-insured grouDs over 1992, although this trend was reversed in early 1993. Family benefits for children from eighteen months to six years old are routed through the PF but financed from the state budget.6 In 1992 the budget reimbursed the Pension Fund for expenditures on social pensions and Chernobyl victims (the Employment Fund 1J Family allowances paid to children less than eighteen months olA were shifted to the Social Insurance Fund starting April 1, 1992. 25 (see paragraph 2.321 assumes the cost of benefits paid to older unemployed workers who are allowed to retire early). There has been continuing debate over where financing responsibility for certain benefits should lie: the budget or the PF. For most of 1993, the PF was not being reimbursed by the budget. It could be argued that PF responsibility for non-insurance benefits clouds the link between benefits and contributions. There are no individual contribution accounts at present, although proposals are being considered for the early introduction of such a system. Table 21: Sumeary of Federal Cash Benefits in the Russian Federation, 1992 & 1993 (First Quarter) Expenditure as a Sources of finance percentage of GNP payroll tax (Z) Fund type Expenditures transfer budget - 1993 1992 (first quarter) Pension fund Old age, survivors, disability, Employer: 20.6 Yes 5.2 6.1 and selected military pensions; to 31.6 payments to those affected by Chernobyl; selected family Employee: 1 allowances. Employment fund Unemployment benefits; funding Employer: 2 Yes 0.1 0.6 of public works programs; (11 in 1992) administration of training programs; funding the establishment of an information system. Social insurance Sickness, maternity, and Employer: 5.4 No 0.8 0.8 Fund funeral benefits; child allowances; subsidization of sanatoriums. Fund for social Assistance to those in Budget transfer Yes 0.1 0.02 support financial distress, such as orphans, lone pensioners, the handicapped, etc. Source: Expenditure figures are based on the annual reports for 1992 and first quarter actual reports for 1993 for the PF, EF, SIP, and FSS. 26 2.25 In 1991, the first year of its operation, the PF generated a surplus of R 12 billion (Annex 1, Table A1.2).' About 90X of its revenue originated from payroll contributions, and most of the remainder from budgetary transfers to cover the cost of social and military pensions and certain child allowances administered by the PF (Table A1.2). Between 1991 and 1992, payroll contributions as a percentage of total revenues rose from 90 to 94 percent, and expenditures on labor pensions rose from more than 77 percent to about 82 percent of overall expenditures. 2.26 Despite nominal increases in labor pensions at various times in 1992, the average pension as a multiple of the dependency ratio has been substantially lower than the level of the average wage to which insurance contributions are linked. As a result, the cumulative PF surplus amounted to more than R 420 billion, or 2.7 percent of GDP. The financial outcome for the PF is the result of a number of factors, including the level of payroll compliance, the evolution of wages, the number of pensioners in the system, and the average pension benefit. The Fund also benefited from the shifting of responsibility for payment of allowances for children under eighteen months old from the PF to the SIF beginning in 1992. For the first half of 1993, the cumulative PF surplus decreased to 1.8 percent of GDP. Table 2.2.: Minimum Pensions Relative to the Poverty Line, 1992 and 1993 month minimum pension Subsistence minimum income February 1992 342 468 August 1992 1,320 1,361 November 1992 2,250 2,464 February 1993 4,275 4,683 May 1993 8.122 8,891 August 1993 14,620 17,696 (estimate) Column 2 is based on government estimates of the subsistence minimum income for women over 55 years and men over 60 years, using 83 percent food expenditure. For discussion, see Cbapter I. Z/ This can be seen where surplus (s) is the difference between revenues (R) and expenditures (E) s - R - E w (tw - er)L or, to reformulate where w is the average wage, t is the payroll contribution rate, e is the average pension, r is the ratio of pensioners to workers (the age dependency ratio), and L is the number of workers. t approximately equals .30 and r equaL .46. Together with the information on average wages and pensions, tw > er, and a positive surplus exists. 27 2.27 Subnational Pension Administration. The Pension Fund has established branches in the oblasts and in more than 2,000 raions. Raion representatives collect contributions and forward them to oblast departments of social protection for consolidation; benefit expenditures are also transferred to the oblasts for payments. Surpluses are then forwarded to the PF in Moscow; deficits are covered by transfers from the central office of the PF. Generally, northern regions are surplus oblasts, because pensioners prefer to reside in areas with a moderate climate. In the third quarter of 1992, there were forty-one oblasts running deficits on their PF. 2.28 Oblast PFs are permitted to retain penalties levied on delinquent enterprises. These amounts are then partly used to help finance subnational social assistance programs. The oblast PF decides what programs to support and the level of supervision required for the use of such funds. For example, in Altai Krai the PF insists on distributing the monies to raion governments and on monitoring their use directly, thus bypassing the Krai authorities. In Chelyabinsk, however, the oblast PF allows the Oblast Department of Social Protection to dispense PF allocations for social programs as it sees fit. 2.29 The Ministry of Social Protection (MSP) is responsible for the implementation of legislative programs for pensions, and oversees oblast and raion operations. Raion offices register claims and calculate entitlements, forward papers to the oblast offices (where the calculations are checked and entered onto a mainframe computer system), and arrange payment. Day-to-day control is vested in the local administration (Social Protection Departments), although guidance on methodology is received from the MSP. The Social Protection Departments also administer the social assistance programs at the oblast and raion levels (see Annex 1, Box Al.l). 2.30 In 1992 and 1993 problems of pension adequacy were compounded in some areas by delays in payment. Periodic shortages of liquidity and the time-consuming (manual) process of frequent recalculation of individual pensions no doubt played some role. The PF and the MSP each tended to blame the other for the problems, leading to proposals that the two agencies' pension responsibilities be combined. This issue is discussed in paragraphs 3.85 through 3.93. UnemRloyment Benefits 2.31 Throughout most of the Soviet period, it was illegal for individuals to be simply unemployed. With perestroika, however, it began to be recognized that if the labor market was to function more efficiently there had to be at least fractional unemployment. In response to emerging realities, a system of providing for the unemployed began to evolve. In 1991 a new (Soviet) Employment Law made it possible to register as unemployed and to receive unemployment benefits of limited duration (see Box 2.1). 2.32 Unemployment and related benefits are paid from the Employment Fund, which began operation in July 1991 under the provisions of the Employment Law. The EF was established to pay not only unemployment benefits but also all other costs of operating an employment service. The EF, which operates as an extrabudgetary fund and is used to finance (and is managed by) the Federal Employment Service (FES), has no separate existence of its own. The FES replaced the Employment 28 Committee by Presidential decree in June 1992, when the link with the Ministry of Labor was also broken. The FES is responsible for providing overall leadership, policy direction, and management to the national system of employment offices established at the oblast and raion levels. Box 2. Eligibility for Unemployment Benefits Under the Employment Law 199l eligibility for buloyment benefits depends on satisfying two requirements: (1) at least twelve weeks work in the previous year, and (2) an aetively seeking work test. In practice, benefits are paid upon satisfaction of (2) alone, and new entrants, for example, do receive benaefits. The legislation is unclear with respect to the poition or individuals who quit voluntarily or arw dimissed for disciplinasr reasens. Before being classified as unemployed, reduadant workers are entitled to receive their avera8e wage for a period of job search not exceeding three months. Benefits are then paid for a limited duration only: twelve manths. plus an additional week of benefits for every year worked over twenty years (twnty-five for men), up to a maxirrm of twenty-four months. Benefits are withdrawn if the recipient, refuses two "appropriate" 4ob offers, although he or she a"ey till use the Uploywent Service as a soutro of Job infomnation. Benefits are genorally paid on an earnings-related basis, and require extremely camplex methods of calculation. The rates vary over the year as follows: (1) during mouths one to three the benefit is 75X of the individual's average wag over tho previous two months; (2) 602 of the average wage between mcwths three to seven. end (3) 45X for moaths seven to twelve. Supplemet are paid to unemployment benefits where the recipient has dependents: each dependet taises an individual's benefit by DIO, up to 100? of their previos vage. Tratrin supplments (752 of the individual's previous wage, or equal to the minimum wage) are pa!4 for the duration of the training period for -those with at least one year's work history. Applicants seeking work who ae not, eligible for unemploymet benefits may recetive _smto of the minimum _wge. 2.33 The day-to-day operations of the Federal Employment Service (for example, registration for work and benefits) occur in the li.jon employment offices. Every applicant has a workbook in which the enterprise manually records his or her work record. The workbook record is the basis upon which the employment office calculates maximum duration of benefit down to the last day. Processing the claim is time-consuming, because, among other issues, under the law the maximum duration is increased by one week for every year of service after twenty years (twenty-five for men). With subnational administrative services in a relatively embryonic form, this could become difficult if there is a rapid expansion of unemployment. Proposed amendments that would allow the temporary use of flat- rate benefits at two levels--a higher level for applicants who have lost their jobs and a lower level for first-time or returning job seekers (and only for six months) - -have not yet been accepted by the Parliament. Unemployment benefits are normally paid through the Savings Bank. 2.34 The EF is financed by payroll contributions from all enterprises. In January 1993 contributions increased from 1 to 2 percent. In the event of deficits, budgetary transfers have been authorized. Contributions are collected at the raion level. In theory, of each R 100 collected, the raion is to keep R 45 and transfer R 55 upward; the oblast keeps R 45 and transfers the remaining R 10 to the center, or the headquarters of the FES. But in practice, neither the collection of contributions nor their transmission upward is wholly reliable. Tax compliance was estimated at only 60 percent during the first six months of 1992--substantially lower than the compliance rate for the PF. And financial 29 arrangements differ among oblasts - -in particular, the proportion of contributions that is retained at the subnational levels. In recent quarters, several oblasts have failed to transfer the required 10 percent to the center. 2.35 Largely because relatively few individuals sought unemployment benefits, the EF generated a small surplus (about R 2 billion) in the second half of 1991 (Annex 1, Table A1.4). These trends continued into 1992. Revenues exceeded expenditures by R 28.6 billion in 1992, principally because the nominal wages rose, whereas the number of registered unemployed was substantially less than expected.8 2.36 While in 1992 nearly half of Employment Fund expenditures went to cover the cost of administration, the relative shate declined in 1993 to 35 percent. The "Job creation" function of extending loans to enterprises at below-market interest rates (between 30 and 50 percent annually) accounted for 19 percent of total expenditure in the first half of 1993, down from 26 percent in 1992. The EF has also used a part of its funds to remodel offices and advance funds to municipalities to construct housing for fund employees. So far, few resources have been expended on either labor training or public works (see Figure 2.2). 2.37 The financial position of the Employment Fund is conditional on several factors, including the average duration of registered unemployment, the level of registered and unregistered unemployment, and expenditures incurred for job creation activities. Over 1992 the Employment Fund generated a larger-than- projected surplus of more than R 25 billion and R 93 billion for the first half of 1993. 2.38 There are a number of problems with the present system of social protection for the unemployed. There is real concern that the EF is too small and will not be able to finance benefits even at meager levels of income support in the face of rapidly rising levels of unemployment (see Chapter I), the recent increase in the payroll contribution notwithstanding. Already several oblasts have placed all recipients on minimum benefits. According to current projections, many oblast funds will soon run into deficit. Quite apart from its size, the present structure of the EF does not allow it to shift public resources from oblasts with low unemployment (higher revenues and expenditure demands) to oblasts with high unemployment. And sadly, experience in Eastern Europe suggests that proactive programs, such as retraining, which are financed from the same source, are likely to suffer cutbacks just as the need for such activities increases. Possible solutions to these problems are discussed in Chapter III. 1/ Only about half of those registered claim benefits mainly because redundant workers are entitled to severance pay, that is, full wages ftrn their former employers during the first three months of unemployment (see Box 4.2 in Chapter IV). 30 Figure 2.2 Employment Fund Expenditures by Category Flrst Six MOAths of 1993 Muhft0heVn 411 W.IItuS St Souce: Federal Employment Service Sickness Benefits 2.39 In 1933 sickness and disability benefits were placed under the control of the official trade union, although it was financed from the state budget through a payroll tax rather than union dues. There was considerable scope for discretion in the decisionmaking of factory-level social insurance committees. There were, for example, eighteen categories of disability allowances, with availability of the allowances depending on such specific factors as the role played by alcohol. Evidence suggests that it became common practice for union officials to use these benefits as instruments of labor discipline, as well as means of bribery and favoritism. 2.40 In 1991 the Social Insurance Fund (SIF) was established, in principle, as an independent financial institution with a board of directors reporting to the Independent Federation of Trade Unions (IFTU), the successor to the official trade union. In late 1993 the government decided that control was to be shifted from the IFTU, although the proposed management structure was not yet clear. SIF revenues are derived from a payroll contribution of 5.4 percent (5 percent in the case of collective farms). Of the 5.4 percent deducted from the enterprise account, 74 percent is spent at the same enterprise on benefits for its own workers. Revenue collection agents, formally employed by the SIF (but in practice working for both the SIF and IFTU), operate at the enterprise, raion, oblast, and central levels. All benefits paid out of the SIF are guaranteed by the state. If the SIF is unable to pay for them, the obligations have to be met through budgetary transfers. The SIF pays benefits only to those who are employed (see Box 2.2). 31 A persm is eligible ft sick pt tho moment a cOntract of employment is signed. There are no ontribution or previous-work ciditLon fto the LUnes or Wxury, whether or not it is work-related. The legislation applies to all emplyees Isn the publie and private sectors. Duration is generally until recoVery or until eligibility for tiasbiity pension is established. Benefit lloae difter, depenn on whether or not t.he ilxness or Wury is work-related. In practice, oxrk-related ts defined an accidents at work, plus a rather broad list of other illnesses. The bemefit in sob eases is pid entirely by the employes at 100Z of the individual's previous wage. There it a mintimm beetitt equal to 902 of the mnimm wage. Sick pay fox non ork-rel.ated o4uses to paid from the Sooial Ineurance lumd at the following percentages of the individual', previous wage: 1001 for those with at least eight years of service and for parants with three or more ohldren under eighteen; 80S for individuals with five to eisht years of light servioe; and 60X othewise. There is a minimum benefit of 90X of the minimum wage and a maximum benefit equal to twice the relevant tariff wage. RPlated benefits include payMnts to those who must stay at bhte to care for a sick family metbet. benefit levels are the same an for nen-work-related sick pay, although the duration rules are more complex (the normal axlium is three days benefit to care fot an adult and fourteen days to care for a chiA4). 2.41 The remaining 26 percent of the SIF contribution is transferred upward. Allocation is determined by government decree: only 6 percent of the 26 percent goes to the central level, with the remaining 20 percent divided between the Oblast Industrial Committee (for the particular branch of industry) and the Oblast Committee of the IFTU. There is some flexibility and possibility of redistribution, especially at the oblast level. Expenditures include construction of sanatoriums, enforcement of labor legislation (300,000 inspectors are engaged nationally), and provision of additional benefits (for example, for victims of Chernobyl). Since eligibility is established at work, the administrative costs of the central authorities are low (less than 1 percent of total benefit expenditure). 2.42 Wages rose faster than benefit expenditures in 1991, leading to an SIF surplus of R 5.2 billion. More than one-half of the expenditure was on sick pay, followed by purchases of sanatorium vouchers. And in 1992, despite paying certain child allowances as of April, the SIF generated an accumulated surplus of R 44.1 billion. During the first half of 1993, the accumulated surplus dropped to R 31.7 billion. More than one-half of its expenditure was on sick pay, and 15 percent on purchasing sarnatorium vouchers( see Figure 2.3). 2.43 Declining compliance rates and delayed payment have become apparent, especially during the liquidity crisis in mid-1992. A Presidential decree has established a list of priority payments in which the Pension Fund, federal taxes, and wages (in that order) rank above contributions to the SIF. This factor, together with the uncertainty associated with other SIF expenditures, makes it difficult to estimate the overall financial position of the SIP. Nevertheless, it is clear that in 1992 and 1993 the SIF was running at a considerable surplus. 2.44 Official estimates show that 65 million workers are within the ambit of SIF benefits, although the actual number of beneficiaries may be smaller. Since 1990, benefits and services are, in principle, available for non-union members; that is, all employees are automatically members of the SIF, whereas the official 32 trade union claims only 50 million members out of a total work force of 72.5 million. A major difficulty with the system that prevailed to October 1993, however, was the failure of trade union officials to keep SIF contributions (which legally belong to the state) separate from the union's own funds. A preliminary investigation undertaken by a Supreme Soviet commission in 1992 found that the SIF was being used to finance, among other things, salaries of union staff and the establishment of banking structures--an apparent diversion of R 1 billion. Figure 2.3. Social Insurance Fund: Expenditure by Category : First Six Nouths, 1993 Sourc: Socil lusrance Fund 2.45 There was also evidence of widespread abuse by officials of the former official trade union; allegations have been made that members of independent unions do not receive, or are threatened with denial of, benefits. Both this and the problems described above suggest fundamental difficulties in the system of management and control that prevailed up to October 1993, both for democratic accountability and the right of the individual to free association. 2.46 The official trade union, which has a presence in virtually every work place, was the only communist-run organization to survive the collapse of the Soviet state with their leadership, property, and membership largely intact. Although diverse coverage remains extensive, the situation is in a state of flux; diverse systems are emerging outside the SIF, including (1) independent trade unions for small- and medium-scale enterprises in the private sector (collecting a 13 percent contribution), and (2) the Independent Federation of Coal Miners, an individualized system of mandatory savings that is operated on an annual basis and involves no real risk pooling. There are about 40 trade union associations outside the IFTU (for example, railroad workers) --covering an estimated one-fifth of the workforce according to an official of the Ministry of Labor. Many of the new unions are small, however, often with fewer than 3,000 workers and restricted to a few industries. 2.47 Although effectively in existence since 1933, the Social Insurance Fund lacks legislative basis, and is governed by various decrees and resolutions. Current reform proposals include shifting the SIF from trade union control to control by the Ministry of Labor or an agency managed by a board representing various interested groups. Several draft laws were circulated in 1993 and 33 engendered intense political battles. In October 1993 it became clear that the trade union will lose its monopoly position with respect to sick pay, and that responsibility would be transferred to an agency of the government. (The issue of sick pay, including an analysis of the ineentive and economic effects of alternative schemes, is discussed in Chapter II, paragraphs 3.114 through 3.126.) Maternity Benefits 2.48 Near universal labor force participation under the Soviet system was made possible by a system of concessions to women that assisted them in combining full-time work with part-time child rearing. Maternity leaves provided in Russia are more extensive than in many Western countries (see Annex 1, Table Al.7), but since benefits are tied to the minimum wage rather than to the previous wage, they provide relatively less compensation. Although the benefits have been adjusted from time to time, it is widely felt that they are not adequate to meet the increased cost of living for families, particularly families headed by women. Maternity benefits are administered by the employer under the SIF scheme described in the foregoing section. There are also various allowances for families with children and working mothers (discussed in the next section). 2.49 There is a maternity grant (that is, a one-time payment set at three times the minimum wage) paid to all mothers. The eligibility requirement for the maternity allowance (a continuing benefit) is the same as for the sickness benefit: it is necessary to have signed a work contract. The maternity allowance is 100 percent of the mother's wage, regardless of length of service. Allowance payments begin after thirty weeks of pregnancy, and normally continue for up to 126 days. There is a complex set of rules about duration of payments for pregnancies with complications and multiple births, for pregnancies in the Chernobyl region, and for those in rural areas. The maximum benefit period is 180 days. 2.50 Maternity and family benefits from different sources have been administered at the enterprise level. Although this may have worked well under central planning, the system is becoming less appropriate. Benefits administration is a complex and time-consuming task for the enterprise. Although official data about enforcement are lacking, there is evidence of non-payment to many women who have wound up seeking assistance from women's organizations to obtain their rightful allowances. 2.51 Streamlining the system to permit greater transparency and administrative ease will be necessary so as to safeguard the achievements of the family benefit system in Russia during the transition and beyond. Financing allowances through several different agencies (including the enterprise) and without a clear accounting of costs was viable in a centrally planned economy in which social costs were spread out among different entities. In profit-oriented enterprises, however, management may have incentives to hire persons for whom concessions do not apply. Employers may see all women as potential caretakers of children, the sick, and the elderly, in view of the current division of labor in the household. Women may also be offered lower salaries for the same work to compensate for perceived increases in the cost of female labor (although this is a dubious distinction--the social insurance deduction is levied on all employees). In this 34 way, the concessions to women are transformed into obstacles to women's employment. Family Allowances 2.52 No unified legislation currently governs family allowances in Russia. Through various decrees, there are twelve types of benefits now payable to families with children (see Annex 1, Table A1.5). The tendency since the mid- 1980s has been to remove means testing for family benefits and to expand benefits to all families with children. Some disadvantaged families receive additional grants. Families may receive all benefits for which they are eligible, with no offsets. Funding is complex: federal budget allocations, the Social Insurance Fund, the Pension Fund, and subnational government budgets all play a role. 2.53 The system of family benefits is complex but can be grouped into three broad categories: those payable to all families with children, regardless of income or other qualifying conditions; those payable to working mothers; and those payable to disadvantaged families. Table Al.5 in Annex 1 sets out each of the benefits and their levels. All families receive (1) a one-time grant for the birth of each child, (2) monthly payments to mothers caring for children up to eighteen months old, (3) monthly allowances to children ages eighteen months to six years old, and (4) quarterly cost-of-living compensations for all children up to eighteen years of age. Working mothers are further granted (1) paid maternity leave to care for a child under eighteen months, and (2) compensation for unpaid leave to care for a child under three years. Allowances to disadvantaged families include those payable to (1) families with single parents and unmarried mothers (children up to sixteen years old, or up to eighteen if the child is a student and not receiving stipends); (2) children whose parents are evading support (to age sixteen, or eighteen as above); (3) children not receiving other payments (to age sixteen, or eighteen as above); and (4) foster children (to age sixteen, or eighteen as above); (5) non-working individuals caring for a disabled child under the age of sixteen; and (6) children infected with HIV or having AIDS (to age eighteen).' 2.54 Since 1990, family benefits have been tied to the minimum wage, and have been adjusted periodically (although not always in tandem with the minimum wage). In 1992, these benefits were adjusted five times, yet the real value has been greatly eroded. The connection to the minimum wage seems to be diminishing, although no official decision has been made to this effect. In December 1992 all the benefits were doubled; the most recent upgrading, undertaken by Presidential decree on February 1, contained a long list of new levels of specific benefits rather than a uniform coefficient. 2.55 The only child allowance tiat is currently means tested is the annual compensation for the rise in prices for children's clothing. It was introduced in 1991, and was originally paid to all school children. In August 1992 the benefit was reformulated by the Ministry of Labor: beginning in early 1993 the allowance was R 480 for every child under thirteen and R 630 for every student over thirteen years of age (multiplied by the relevant regional coefficient), v Benefits to children of miUtazy perosmel axe not included here. 35 provided that the family per capita income did not exceed four tilnes the minimum wage. Income for the purpose of this al!kowance includes only wages and pensions received by family members; all other social benefits are excluded. Social Assistance Programs 2.56 Social Assistance at the Federal Level. It is "a common characteristic of most intergovernmental fiscal systems that there is considerable confusion and obscurity as to who exactly is responsible for what and how precisely various public services are delivered and paid for" (Bird 1990). Russia is no exception. The structure of responsibility with respect to social assistance is rather complex, shared among several levels of government and various extrabudgetary funds. Measures are being adopted by oblast and raion governments across the country (for example, to provide for the homeless or to finance food subsidies for basic commodities), and are often financed through the creation of extrabudgetary funds, in emulation of a common central government practice. This strategy has been described as pushing the deficit down. 2.57 The lack of taxation bases adequate to finance perceived local social needs has had at least three major repercussions. First, it has led subnational officials to lobby higher levels of government for more funds on an almost continuous basis. Projected revenue shortfalls are dealt with, among others, by incurring arrears and by deferring expenditures. There are also monthly negotiations between the oblast administration and the Ministry of Finance, while the chairman of the oblast Soviet lobbies the Supreme Soviet, and so forth. This manifestation of the continued soft budget constraint in subnational governments partly vitiates the presumed benefits of decentralization, and has led to a substantial lack of clarity in the budget process. And current practices certainly have not contributed to the stability, security, or objectivity of federal grants. 2.58 Second, subnational authorities are motivated to pursue alternative sources of finance. In particular, many subnational officials have become quite entrepreneurial. The St. Petersburg Committee for Social Protection, for example, has since December 1991 been active in newspaper and book publishing. Profits, which amounted to R 10 million in 1991, are being directed to charitable measures. This practice, which also emerged in Eastern Europe (Poland, for example), appears to be fundamentally inconsistent with the thrust of the reform program. A third and related result of the lack of adequate tax revenues is that oblast governments rely increasingly on enterprises to finance local investments and thus contribute to political pressure to extend credit to these enterprises (which enjoy easier access to credit than do subnational governments). Damaging repercussions follow for both macroeconomic stability and resource allocation. 2.59 Perhaps the most worrying aspect is the scope for worsening disparities in the provision of social assistance, largely because of the absence of effective redistributive policies among oblasts. In Russia, people are relatively immobile, and individuals' locations are largely determined by historical place of residence. Where the poor constitute a relatively large part of an oblast's 36 population, they will tend to receive relatively smaller transfers.10 And there is already evidence that these inter-oblast disparities are likely to be exacerbated under the current scheme in Russia, because poorer regions have correspondingly lower fiscal potential. Decentralization has increased the already significant fiscal disparities among regions. An index of per capita public expenditure (budget plus enterprise social expenditure) varied from 1 to 18 between regions in 1992. The differences in social expenditure broadly reflect variations in tax revenue, indicating an absence of equalization among oblasts. 2.60 The responsibilities of subnational authorities will tend to grow during the course of transition from a command economy. Not only will the importance of assistance for the poor loom larger, but subnational authorities will have to step forward to assume responsibility for social welfare as enterprises retreat. In the next section we examine centrally initiated programs of social assistance, before going on to discuss the issues that are arising with respect to social assistance at the subnational level. The relevant extrabudgetary funds are also discussed. In Chapter III, series of recommendations to improve current financing arrangements are made, including the need for federal transfers. 2.61 Centrally Initiated Social Assistance. In the absence of legislation governing social assistance, the Russian government relies primarily on Soviet practices and regulations. The elderly and disabled living alone who have no other means of support are the main vulnerable groups for whom three basic types of social services are provided: boarding homes for permanent residence; in-home services for those requiring assistance for shopping and other chores; and territorial centers for day care and/or as long-term nursing home'.. In addition, a 1991 Soviet law laid the foundation of providing material assistance to the disabled of all ages1l; and a 1992 federal decree outlined emergency relief services, extending the scope of assistance beyond the two primary groups to include any individual in dire need. 2.62 Oblast social protection departments develop more detailed rules for implementation for these services. Eligibility, user fees, and other aspects may vary across regions. In-home services are typically free. Beyond the advantage that individuals can remain in the community, the per capita cost of day care is about one-seventh the cost for boarding homes. These programs, which are fairly well established under the guidance of the Ministry of Social Protection, would until recently have been regarded as the full extent of social assistance in Russia. Ij/ Insofar as labor and capital mobility increase during the transition, a relatively decentralized fiscal system may constrain actual assistance to the poor in a number of ways. Competition among local authorities to attract investment may severely limit levels of taxation and therefore the amount of redistribution that is likely to take place within Jurisdictions. Areas with low taxation may attract individual wealthy people, and there may be a fear that relatively generous local social support will attract the mobile poor. U1S. empirical evidence provides som support for the view thaL benefit, differentials exert a significant influence on the location decisions of the poor, and that this, in turn, depresses the level of social assistance (Brown and Oates 1985). "I For example, within the 1991 law are "Provisions on Territorial Centers for Social Services to Pensioners," and "Provisions for In-Home Social Assistance to Lone Old-Age and Disabled Pensioners" of 1987; and "On the Foundation for Social Protection of the Disabled in the U.S.S.R." of December 1990. 37 2.63 Subnational Social Assistance. Of the aforementioned services, only boarding homes are fully supported by the central budget, although it is difficult to discriminate between the proportion of central and subnational expenditures in social assistance. Since late 1991 subnational government aid to the poor has greatly expanded to meet rapidly rising demands. While there are no aggregate data to assess the impact of such efforts nationwide, available information suggests that these programs are extensive, and could play an important role in alleviating poverty. 2.64 Subnational programs may or may not be explicitly targeted. Some benefit all local residents, such as producer or retail price subsidies of staples (for example, bread up to late 1993) or free garden plots for growing vegetables. Others are directed to broad categories of vulnerable groups (access to discount stores, free transportation for pensioners), to the poorest groups (access to soup kitchens and meal coupons), or to individuals (subsidized telephones for housebound single elderly or disabled, discount energy for poor large families). To meet rapidly growing demands for assistance, many subnational governments are using a combination of criteria- -based on both category (for example, pensioners) and income (for example, minimum pension or less)--to target the broad categories of vulnerable groups and thus effectively utilize limited resources. In Chelyabinsk oblast, for example, pensioner eligibility for social assistance is limited to those receiving less than the minimum pension; in Novosibirsk, less than a locally established subsistence minimum; and in Barnaul, less than twice the minimum wage. 2.65 Typically, these subnational programs have three things in common: (1) introducing an income test to identify the poorest in the generally vulnerable groups; (2) expanding in-kind transfers rather than cash benefits; and (3) financing assistance through noncentral resources, with some central government subsidies. Organizational and administrative arrangements vary considerably, even among raions in the same oblast. (Boxes 2.3 and 2.4 present brief case studies.) 2.66 Benefits in kind are substantially more extensive than cash payments and are generally administered by social workers at the raion level. In Chelyabinsk, for example, of about R 27 million budgeted for benefits to the needy in 1993, only 3 percent will be in cash allowances, while 27 percent will be for in-kind benefits, and 70 percent will be for subsidies, mainly for bread. The need for material social assistance appears to be administratively determined by whether applicants fell into one of several categories of vulnerable groups identified in Presidential decrees (that is, elderly and disabled people, large families, orphans, and single parents). The oblasts sometimes produce more detailed lists. For example, Chelyabinsk oblast has identified fourteen vulnerable categories vsee Box 2.4). Crude income tests may be applied, normally extending no further than establishing whether individual income exceeds the minimum pension. Such tests, however, make it possible for social workers to exercise discretion, for example, in cases where there was ill health or indebtedness. 38 ox 2.3. Social Asaistance in Chelyobinak Since mid-1991. social assistance activities in Chelyabinak bave fallen into three main areass 1. In-home services, provided to elderly end disabled people, to fsmilies with three ot more children, and to families headed by a single parent. Social workers visit their clients three times a week to help with all evexrday activities apart from cooking and cleaning. 2. Day cent (located in thirteen raioea), providing, for example, occupational therapy and other facilities for disabled people. 3. Matereil assistance, covering various programs such as meal vouchers, clothes, humanitarian aid, medical aid, and one-timo cash benefits. Personal and finaneial details are reconfirmed each time an application for assistance is made. Social assistance centers are located in every raion to deliver in-hams services and material assistance. It is planned that, eventually, the register of people in need will be recorded on computer systems at the rtion level, which will provide summary information to an oblast "information bank." So far only fourteen of seventy ralona have computer support, Few cash benefits are provided as the policy is to maximize benefits in kind. One novel feature is a system of vouchers that can be redeemed against the price of meals in cafeterias contracted with the social assistance centers. Books of thirty vouchers--each with a face value of R 30-- are awarded by social workers on the basis of need and are designed to last two months. This program is severely limited, bowever, in one raicn serving a needy population of 24,000, only 725 books were available for iasus iu a month. I EoxLZ.4. Identifying the Poor in Practice The ablest adfniatrtatif in Chelyabsink ha. identified the following groups as those most in need of assistance: 1. Non-working psiaonets with iunoaes et or below the Minima pension. 2. Disabled people in groups I or It with incomes at or below the minimum pensian. 3. Non-wotking disabled peple in Group III with Incoeos at or beloW the Oun4MW, pension. 4. Familie, with children less then three years old in which the average individual income of eacb fAmily meeber is at or below the minimum wage. 5. Families With a dis_ t ild lesthan aten years of age. 6. Cbildren iUviS with suara.-. 7. Children with daetes. 8, ^*XAdtt with dibee*s Type I (ijnsulln dependency). 9. OCqphane" students. 1O. Stuxdnt failihes with O)4)AWI. 11. Famlies with three ax morea hildzan under the age of eighteen. 12. Women more than twenty weeks into their pregnancy. 13. Sifg -parent tamilies witb illagitiU ate children, 14. Single-parent fmailUe with. on. prtO*e missing (fore.xmpwle, due tO impriwonrent or death), 5he ai3 is to comle ai rgist *f every Individual in need in the oblaot. This includes detnal of depnnsnd socs port, income and expenditures, living conditions, and hbalth statusW ZndivlAenl ar identified on tbeir -on initlative, by third parties (for example, aoir bobaehalf, or as a &agult of pxoactive investigatiolns f t fcia rOd. o krs obain the Anformation about iividual cirsmatena which i them use0d be aoia what level of a lsiao abud be provided. It is i tbat mor thasn J0,:O:i tire lst yventually be interviewed as part of this progrm. j 2.67 The present arrangements are administratively demanding. The emphasis is on individual interviews to verify details relating to income and expenditure. However commendable, this proactive approach toward identifying people in need 39 adds to the administrative burden. Moreover, because assistance is not awarded on an ongoing basis but is determined on each application, social workers have to deal with the same case on numerous occasions. In one raion, an estimated 6,000 people in need had generated more than 25,000 applications for help in early 1992. 2.68 Extrabudgetary Funds for Social Asslstance. A striking feature of social protection, as elsewhere in the Russian Federation, is the widespread practice of shifting items off-budget. There are at least four types of extrabudgetary funds earmarked for social assistance: the Supreme Soviet Fund for Social Protection (SP Fund), the NSP Fund for Social Support (FSS), and various oblast and raion funds for social protection. Although there is some realization that this practice is not sound financial management, it is nonetheless pursued. This is partly due to the desire to appear more resource-poor than other oblasts and raions that are also trying to attract additional subventions. However, these funds generally do not have a regular source of revenue (beyond, in some cases, a share of privatization proceeds). 2.69 The sources, financial status, and criteria for allocation of the SP Fund are not very clear, and no annual report is published. This fund is distributed to and disbursed by oblast and raion soviets either independently or through the subnational social protection departments (depending on the working relationship between local soviets and governments). The budgetary transfer in 1992 totaled R 8 billion. A total of R 15 billion was budgeted for 1993, of which R 10 billion were already distributed in the first quarter. There was no surplus in 1992. Social Protection Fund-supported activities included financing cash transfers to vulnerable groups and rehabilitation of the handicapped. In late 1993, the future of this fund was unclear. 2.70 The Fund for Social Support has been a funding source for subnationally initiated social assistance programs, although it amounted to less than 0.1 percent of GDP in 1992. The MSP claims that distribution of the FSS Fund is directed to the so-called 'emergency raions,"--areas where the decline of the military-industrial complex is creating hardships, zones suffering from ecological disasters or ethnic conflicts, and the like. However, evidence suggests that actual distribution does not enhance inter-oblast equity. In the first quarter of 1992 just twenty-three oblasts received allocations that totaled R 1.3 billion; many of the poorest oblasts received nothing (see Report No. 11302 RUS). 2.71 The FSS began operation in early 1992 with an endowment of R 10 billion, given a one-time transfer of Party assets and the taxation of revalued inventories. And it receives a part of the receipts from the revaluation of commodities in state stores and ruble receipts from the sale of food aid. But the FSS has not been receiving other revenues designated to it. For example, the FSS should have received 10 percent of the revenue from the privatization of federal property (totaling R 1.8 billion, none of which was received). There have also been problems on the expenditure side: R 1.22 billion was sunk into unprofitable ventures. Nonetheless, at the end of 1992, the FSS had a surplus of R 3.3 billion. 40 2.72 A large part of FSS expenditures and the expenditures undertaken by the NSP (R 330 million through the end of May 1992) appear to be reasonably well targeted. Programs include assistance to orphans and to lone pensioners, the establishment of free kitchens, aid to children affected by Chernobyl, funding of institutes and clinics doing research and fitting artificial limbs, purchases of wheelchairs for the handicapped, and the provision of emergency assistance (see Figure 2.4). However, 24 percent of the total federal expenditure of the FSS has been invested in unprofitable enterprises, usually those producing consumer goods. The rationale for making such investments is questionable. An FSS Council of Experts established in early 1993 with a permanent director and rotating membership, is to evaluate investment proposals, conduct feasibility studies, and make recommendations to its board. One task of the council was to retrieve the money lost in the aforementioned investments. Figure 2.4 Fund for Social Support of Population Expenditures, 1992 (X of total) B=k_ AW Wm m~ a 2.73 Oblasts are estimated to have spent Rt 4.5 billion of their own budgetary revenue on social protection through September 1992; this represents less than 0.1 percent of GDP. Unfortunately, detailed data on this are not available. Subnational Social Protection Funds may provide significantly larger amounts for social assistance. Overall, however, inadequate financing appears to frequently constrain social protection effort. The FS spends some resources (some R 2 billion) supporting subnational Social Protection Funds, but these funds are nominally independent from the center. Evidence suggests that there has been some inappropriate use of revenues on the part of subuational FSS offices. Nong-overnmental Initiatives in Social Protection 2.74 Any assessment of social protection in the Russian Federation would be incomplete if the role and efforts of the private, non-profit sector were 41 neglected. Numerous initiatives have emerged in recent times, representing collective responses to both the liberalization of political and legal constraints that had inhibited independent action and the growing hardships suffered by vulnerable groups. Yet beyond the impact of the Soviet legacy, these organizations have been constrained by the difficult economic conditions from which they have emerged. Indeed, the past year has seen the closure of some charitable organizations. Anecdotal evidence suggests very limited financial resources, except for cases in which the charity has a corporate sponsor--for example, a specific factory. 2.75 Lack of a clear legal framework and status exacerbates the situation: current Russian law neither defines what constitutes a charicy or its activities nor provides for the legal status of charities. Lack of legal status means that charities are not tax exempt,22 that there are no tax deductions on donations, that problems can arise at customs, and so on. These problems could be addressed by draft legislation that provides tax exemptions at the rate of 50 percent for non-profit organizations, and 100 percent for charitable organizations. Since commercial activity has necessarily become an important source of revenue, the distinguishing factor in determining status should be how an organization spends its income, not its sources. 2.76 Relationships with government agencies and the degree of government intervention can have a critical bearing on the activities of nongovernmental organizations (NGOs). Oblast and raion authorities often play a more significant role than the federal government. Not surprisingly, the relationships vary significantly, from constructive consultation to mutual animosity. In some cases, organizations have agreements signed with mayors of towns in the local region. Some charities have begun to explore contracting out services from state services. There have been reports that some efforts at cooperation with subnational government authorities have broken down, due, for example, to financial discrepancies. Some NGOs are actively engaged in lobbying on behalf of their clients and have links with members of the Supreme Soviet and subnational authorities. 2.77 A wide spectrum of organizations has emerged in the non-profit sector, with a variety of objectives, sizes, and geographical coverage. This is reflected in differences in their ability to reach the poor, to operate cost-effectively, and to innovate and adapt. Informed sources (United Way, Moscow) have estimated that some 5,000 non-profit organizations and 550 formal charities are currently working in Russia. In Moscow alone, these organizations employ more than 10,000 volunteers. (Box 2.5 describes some of the diversity of current operations.) a/ Exceptions are certain designated organizations, such as the Lenin Children Fund, and those for veterans and invaUds. 42 Dow: 2.S. Selected Nongovernmental Initiatives in the Rpasian Federation An interesting example of a secular otaeniration is the Russian Journalists Charity Foundation, which has aided over 89,000 people since it emerged in 1990 as a joint initiative of professional colleagues. It compiles lists of the naedy based on requests, its aon assessments, and other NGOs. The lists are sent to companies and associations abroad. The foumdation does not rely on government information and categorimation, which are regarded aa not sufficiently accurate or dynamic in identifying the needy, Interestingly, the foundation has sought to shift its focus from food and clothing to what is regarded as longer-term constructive support in facilitating income-earning opportunities. These include sewing machines and tools provided to orphanages and miorobakeries for the handicapped. This equipment is given free of charge, on the condition that it not be sold. The Foundation also operates a free legal advice center. The aeventy-six institutions assisted include a soup kitchen, a kindergarten, and societies of large families, invalids, and foster families, as well as hospitals, orphanages, and boarding schools. Geographically, activities extend beyond Moscow, to Zelenograd. Xoloma, Zaraisk, Halashaike, Yaroslavl, Kostrema, Oah, and the Tver region. Leningrad bas been regarded as the region that took the lead in the formation and proliferation of local support groups. Established in 1986 in St. Petersburg, Angel of the Neva is a nongovernmental, nonreligious, oharitable organiaation. It is aetive in several areas, including training (for example, survival courses for the elderly, and cormunity leadership and mobilization for those working in the area of social assistance); the storage and distribution of humanitarian aid (cases for assistance--which aumbered 250,000 between January and June 1992--are selected with volunteer help); provision of a soup kitchen; assistance to specific groups, such as children's polyclinical health care; and assistance as a liaison with more than 100 other governmental and nongovernmental agencies involved in practical social assistance. A computerized data bank of clients was set up in September 1992. However Angel suffers the financiag difficulties alluded to in the text, despite relatively low operating costs. Its fends derive entirely from donations and international bummaitarian assistance, and the former fell dramatically in 1992, forcing, among other things, closure of the soup kiethen. In even more remote regions, Memorial is ac wel-known orgenization associated with the dissidence movement of the Soviet ora. It is involved in several areas of activity: human rights, research and education, and charitable assistance. The origins of the latter lay in the poor living conditions of poLitical dissidents, ex-prisoners, and their families. Many people suffered from poLitical oppression during the Soviet period. Memorial seeks to assist survivors through counseling and material assistance. The resources available for these activities are very limited. Pinancing comes mainly from foreign sources--a recent appeal by the Moscow branch to over fifty Russian commercial enterprises raised only R 15,000 from two small enterprises. In some parts of the country a sense of increasing political apothy and financial constraints have combined to diminish, ond even- close, several branches of the organization. Nonetheless, ocoaional on-off foreign donations vaiblo- irregular grants to those on the list, as In Chelyabinsk. Tn August 1992 the bank balance of Hemorial Chelyabinsk had dwindled to R 4,000, when it was finally boosted by a modest local government grant of R 100,000. 2.78 Distribution of international humanitarian assistance in Russia has been done through nongovernmental organizations and through municipal committees at the raion level. Local charities, such as Moscow Charity House, target needy households and deliver humanitarian assistance. These local charities may deliver assistance door-to-door in cases where simply delivering the assistance in bulk to the neighborhood is likely to fail to benefit the most needy. Involvement in aid distribution can help to strengthen the institutional capacity of local nongovernmental organizations. Other more formally organized and relatively large-scale international nongovernmental organizations have been involved in emergency social assistance in different parts of the Russian Federation over the past year. CARE, for example, was involved in a food monitoring project, assisting in the distribution and monitoring of supplies from the United States, and has undertaken surveys of vulnerable groups (see Chapter I). 2.79 Women's groups represent a particularly important sphere of NGO activity. A number of independent women' s groups has been emerging in addition to the 43 officially sponsored Soviet Union of Women, which previously dominated the scene. The focus of the women's groups includes a wide range of professional, humanitarian, entrepreneurial, political, and self-help objectives, although their level of activity is still modest. These organizations have a key role to play, not only in the changing social and political life of Russian women, but also in assisting the economic transformation. A First Independent Women's Forum of such groups was held in 1991, and a second is planned. Approximately 700 women's organizations are registered with the Ministry of Justice. (A selected list of organizations is given in Annex 2.) Conclusions 2.80 The transition from a command economy has been marked by the accumulation and worsening of large macro imbalances. Enhanced enterprise autonomy has been accompanied by continuing subsidies and transfers. Nevertheless, the picture of the interrelationships between individuals, enterprises, and the state remains iargely unchanged in essential respects. 13 It is necessary that these roles ch..ge fundamentally during the course of successful transition. Economic theory and practice in industrial countries suggest appropriate future directions. State intervention in markets is justified on efficiency grounds when the assumptions of perfect competition, no externalities, and perfect information do not hold. The capacity of the family and of individuals to fulfill the role envisaged by market economics depends on having the information to make informed choices and the personal mobility and resources to act on these choices. In addition, the equity objective of ensuring minimum living standards for the population may imply specific public interventions. Families may lack the resources to provide for their members adequately. 2.81 In many industrial countries there has been a shifting of responsibility in the direction of the individual and families. The precise boundaries between state and private action are dynamic and likely to remain controversial. Nonetheless, it is possible to identify the following as among the appropriate objectives for the Russian government as it aims for a market-oriented economy: * Promotion of earning opportunIties. Economic growth and expanding employment are the primary guarantee of social protection. This requires the correction of major macro imbalances and the creation of an environment conducive to private sector activity. e Poverty relief. No one should fall below a minimum acceptable level of subsistence. This implies the need for income redistribution (transfers) financed through taxation. Wages and employment should be largely market determined in the interest of efficiency. * Insurance. The government may have a role in filling gaps in private insurance markets in the interests of efficiency and equity. For example, unemployment insurance helps to avoid the burden of adjustment falling disproportionately on certain sections of society. Disability and survivors' pensions can be regarded as insurance benefits. There is also a potentially significant role J.Wlote that the distinction between levels of goverment is ignored at this stage. 44 because of tax incentives and partly because public pensions tend to be fairly meager. * Facilitation of private activity. Both the profit and non-profit sectors can make valuable contributions to social protection. For example, the state may consider contracting out as a more efficient way to deliver social services. A series of specific recommendations with respect to nongovernmental organizations are made below. * Special efforts to protect the most vulnerable groups, particularly children. The incidence of poverty is significantly higher among families with a larger number of children. Hence, programs to protect such children such as vaccination campaigns, primary health care and school feeding programs are important. 2.82 These objectives may potentially be at odds with one another. Russia faces strategic choices in the design of social protection, including the choice between the immediate provision of a high degree of social protection versus relatively low levels of social protection, alongside cheaper labor costs, more rapid employment growth, and (potentially) higher public spending on human and physical capital formation. In several East Asian countries extraordinary increases in the availability and quality of education have been associated with high rates of savings and investment, limited public intervention in labor markets, and the rapid growth of output, labor demand and real wages (Birdsall and Sabot 1993). (Box 2.6 discusses the Korean and Japanese experiences.) 2.83 What do the above objectives and trade-offs imply for the respective roles of the individual, the enterprise and the stat3? For the state, at this stage, fiscal and monetary restraint is the paramount task. The general problem of scarce public resources is exaggerated in the current context, given the need for stabilization and the problems of restructuring. All of the transition economies of Eastern and Central Europe have faced a fiscal crisis, as output and hence, tax, revenues fell while the needs of the social safety net rose. Stabilization and adjustment clearly require a shift in state-enterprise relations, toward an arrangement in which budget constraints on enterprises are hard and property rights are clearly defined. Encouragement of private sector activity requires a series of reforms, chief among them, reform of the legal-regulatory environment. This would include legislation to protect private property, to govern contracts and corporate activity, and to ensure prudential regulation of the financial system. Policy measures to encourage micro enterprise development would include market deregulation and licensing and possibly tax incentives. 2.84 With respect to the system of social protection, a better balance is required between encouraging employer participation in order to reduce unnecessary government expenditure and overburdening enterprises with administrative overheads that inhibit competitiveness. In view of increasing labor mobility and the growing profit orientation of enterprises, the existing system of benefit administration will become untenable. Consideration needs to be given to shifting away from employer responsibility for the administration of family allowances and benefit claims. The extensive labor market concessions granted to women should not be allowed to become barriers to their relative 45 family allowances and benefit claims. The extensive labor market concessions granted to women should not be allowed to become barriers to their relative attractiveness as employees. The gradual elimination of child care and preschools at the work place should be accompanied by increased state involvement in these spheres, and the adoption of measures such as contracting out to encourage private sector activity. Box 2.6 East Asian Experience with Social Protection Several East Asian countries chose to emphseise investment in human capital and growth rather than social protection, especially ii the initial stages of developlment. Japan and Korea both succeeded in achieving rapid growth and reductions in absolute poverty. Economically and physically devastated, post-war Japan concentrated on economic growth and reconstruction, giving secondary attention to the establishmet of a comprehensive social protection system. Its people achieved a remarkable economic growth record to become one of the most competitive and affluent countries in the world During the period of economic reconstruction, social policy tar8et the poor. As Japan's income grew, the majority of the Japanese were able to afford contributims to the social insurance systen, and the primary objective of social protection was transformed from poverty relief to poverty prevention. Whereas, in 1951 Japan spent 3.52 of its national income on social protection, the figure rose to 13.7S by 1990. Pufblic assistance. funded from genedral revwaes at alL Levels of government, is the lest-resort guarantee of a minimum standard of living. It is strictly ueans-t-sted with recipients being required to seek assistance from relatives first. Child allowances were not introduced until 1971 end were initially modest, limited to families with more than three young children, but became more generous over time as economic prosperity allowed. In 1991M, chil allowances are received from the first child under 3 and the income test is loberal enough so that three-fourth of the children receive a benefit. Japan's pension system has- alSo xpandad aimt. 1973; benafits aro now indexed and past salaries updated. In 199 the typical retires and his dependent.apouse bave 662 of their income replaced througb pensions. In the 1980s, tte Korean "ystem proved able to adjust to a competitive market with rare cases of bankruptcy and mas aoffs, es putsued on evport-ld. development strategy that focused on labor- intensive induStries that Stim%ulated labor demand and curtailed the maber of unemployed and underemployed. The consequent increase in the incoms of Low-skilled workers improved the equality of income distribution, despite the low priority of socal poolicy and laak o£ specific attention to equity [see empirical studies by Choo (1980), Adelman end Robinson (197$), Oad Renaud (19?6)i. Education comprised the bulk of Korea's welfare spending An 1984, compared to other nations, for example Argentina, Chile, Sweden. end the United States, which conoentrated on social security. This strong emphasis on education manifested itself in the expanded constructicn of schools and ln an improved quality of instruction. Government expenditures alo rtose in the Areas oif medical ssistance and primary bealth care. Korea's Medical Care Assistance Program, establithed in 197J, provides medical services for those unable io pay for thbm. In 1946, over 102 d te p lton ws covered under this proam. This belps to combat a ajr o f onic p t in rfs hi. tto edical o. Korea's primary health eare program established aesounity health asX oeet.ersf In order to provide advice and preventative services to the rural popwlation and low-iome urben residents. 2.85 In the relationships between the individual and the state, there should not be serious adverse incentives affecting labor supply decisions, inter-household transfers, and inter-temporal decision-making (see paragraph 3.12 in Chapter III). Such effects could -eatly alter the incidence of income transfers. There is a real risk that low .ncome individuals experience far higher implicit marginal tax rates than do individuals higher up the income scale. This obstacle to the individual's transition from welfare to paid work is evident in many countries, including Canada, Britain, and Israel. Institutional impediments to mobility in Russia should be removed (see paragraphs 5.34 ff in Chapter V). At the same time, it is appropriate that the state levy taxes in order to provide a minimum subsistence for those citizens who otherwise would fall below that poverty level. Welfare objectives should not be pursued largely through influencing enterprise decisions about wages and employment. 46 2.86 The basic role of enterprises will be to produce goods and services according to domestic and international demands. Indiscriminate and large-scale support from the government will cease. Enforcement of such constraints will lead enterprises to adjust their behavior to the new economic environment. Many enterprises will downsize, some may liquidate, and many will alter their product mix and their technology. Enterprises will be expected to respond to market signals and will, to a significant extent, be liberated from the distributional and welfare objectives of the government. New private sector activities will become increasingly important. The contribution of enterprises to social protection will take the form of contracts for labor services in return for a package of compensation (in cash and/or in kind). In fact, it is expected that the gross wage share of GDP will rise from the currently low levels as individuals are remunerated more in cash rather than with subsidized goods and services and transfer payments. 2.87 During the transition, however, there is a case for continued state intervention in the enterprise sector in the following areas: * Employment subsidies in regions of unusually high unemployment. Strictly temporary subsidies, fixed in size, transparent, and limited to regions suffering the worst shocks, can assist the objectives of enterprise restructuring and the development of well- functioning labor markets (see Chapter V). * Transitional support to allow the continued provision of social services, in situations in which subnational authorities are financially or administratively unable to meet immediate needs, and in cases where it is inappropriate to rely on private provision. 2.88 As the challenge in Russia is to avoid massive disruption to existing basic services, it is not critical that enterprises immediately divest themselves of social functions. Immediate withdraw could be immensely costly, because the state would not be capable of taking up the functions that are properly regarded as public goods, such as the provision of basic health care and education. Orderly devolution is an important objective during the transition and will help to preserve the impressive human capital achievements of the Soviet era. State support may be extended to enterprises to maintain service provision through, for example, forms of contracting out. State regulations that inhibit service provision, including strict rent controls, should be liberalized. As far as possible, support that is extended should be in the form of explicit budget subsidies, as opposed to, say, tax expenditures (incentives). Current fiscal arrangements, for example, encourage the provision of housing. 2.89 The basic point is that enterprises should become market oriented and less of a fiscal burden on the state. This may or may not involve a curtailed role in the provision of social services in the short term. Employers and individuals in Russia, as in many other industrial countries, may well prefer that compensation for labor services partly take a non-monetary form. The ratio of non-wage to total labor costs averaged roughly 30 percent in the OECD in the period 1966-81, with a range from 25 to 40 percent. Tax treatment of benefits clearly have an important influence. The total cost of all social services currently being provided by the typical Russian enterprise (between 35 and 40 47 percent of the wage bill) is therefore broadly comparable to the patterns observed in many industrial countries, in which non-wage benefits include group life, health, and disability insurance, subsidized child care, and so on. In the U.S. private sector in 1990, the cost of voluntary non-wage benefits amounted to 25 to 30 percent of business payrolls. 2.90 Reflecting the significant inertia on the part of enterprises in general in Russia and continuing widespread subsidization, adjustments with respect to social provision have been limited (see Box 2.7). The most widespread changes appear to be taking place with respect to the financing and provision of preschools. In addition, in 1992 and 1993 many enterprises were seeking to transfer responsibility for housing (and the large subsidies due to rent controls) to local authorities. The latter were often reluctant to accept this burden. Overall, it seems unlikely that many enterprises will suspend housing programs in the short run. The World Bank survey found that more than one-third of the sample was continuing to construct housing; the same proportion declared that housing would be the last benefit--of all benefits provided--that either management or workers would be prepared to relinquish. EBx 2 7. Tho Cheeging Nature of Entsrpris Eamef ita Provided by Russian Enterprises A significant share of Russian workers hae entitlements to housing. child care, and paid vacations, largely irreapoetive of their status within the ntetprise, Housing assistance may take the form of actual provision, or various forms of Gash or in-kind assistance (see pargraph S.62 in Chapter V). Rnts are very Low. Occupants obtain an implicit ownership right, beyond the termination of euployient. Currently, ther are able to have th legaL title of the apartment transferred lato their anm names without paying ny purchase price. Mad.um and large enterprises normally operate a health facilty such as a hospital or polyelknie, sall enterprises may have an in-house faciUt.y (fot example. a resi4ent doctor), as well as cotracts with outside heloth care providets far their emplyeers. Health care provided tbrough the enterprise is free and often of muc higher qpality then that available in public facilities. Entetprises ate icrseasingly tending to contract outside facilities rather then eongae in direct provision. The most important educatlon sevice financed end operated by enterprises for their employees has been kindergartens for children between tree and seven years old. Fsee provision has facilitated very high rates of fomale Labor force pasticipation. Enterprises appear to have been rapidly divesting thoeselves of thi fumaction, and the authorities ase finding it very diff4cult to fill the void. In Chelyabinsk, for example, about two-thirds of the S00 kindergarten. in the city were operated by enterprises. that, as they be4#.. Joint stook companies, sre shifting away f this activity, especially the.sSmaller entepises.T oity administration pretosed funding these facilities tbrough the imposition of an addtional tax. In Psvlovskai raion, of eighty-nine kindergartana, only eight had beloned to govarnment authorities, and an additional toenty-three were being shifted cut of enterprises in late 1992. The local Soviet has resolved to accept all kinAergarten offered. The same yatem gad treand are evident with respect to day care. State enterprises often provide financial assistancs to local schOols in return fos preferential aceass for employees' children. This may take the for, of wage supplemnots to ettract better teachers. Access tc consumption goods has traditionally boon an mportant aspoct of worker ccmpaeation, The enterprises may obtain th goods--ransing Sn foodatuffs to durebles--through direct purchase, barter, or eve frm their on fams. hese re made available to employees at substantially lees tha market priCes. Evidenes suggests that many enterprises ontinue to provide this beefit. 2.91 It is as yet unclear how swiftly enterprise behavior will change in response to changing conditions in Russia, for example, the withdrawal of subsidies. Transition economies in Europe may provide some indication as to future trends. In Poland, for example, the social assets of the enterprises included kindergartens, cafeterias, and housing, but cost relatively little--in 48 most cases less than 2 percent and very rarely more than 5 percent of total costs (Pinto 1993). Recent survey evidence suggests that the rapid deterioration of the financial condition of Polish enterprises in 1991 led to some adjustments. Generally, the assets were transferred to subnational authorities, often free of charge, or to commercial ventures established by former employees. However, the giveaway process was slowed by the costs to the potential donor (gift tax and obligation to show title) and the scarcity of potential benefactors (given high maintenance costs and the glut of such assets). Alternative solutions can involve cost recovery, for example, vacation resorts leased or operated on a commercial basis. Worker resistance to enterprise divestment has apparently been minimal, which is attributed to the fact that these enterprises lacked adequate resources to operate the facilities satisfactorily in any case. 2.92 For individuals in Russia, increased choices and greater scope for the expression of self-interest also mean that people are expected to take greater responsibility for their own livelihoods. This would include, for example, individual savings to help cushion future loss of income. Increasing efficiency and living standards often involve reduced job security, as many workers are in jobs in which their productivity is very low and will have to find new jobs. The returns to education and skills are likely to rise, and there are likely to be increasing differentials in living standards among individuals. 2.93 In the near term, many Russians are likely to focus on survival strategies, and some may be unwilling to actively take advantage of emerging opportunities. There may be modest responses, in the forms of home production of foodstuffs and hoarding; or responses might extend further to taking on additional paid work or selling in street bazaars. There are pervasive social networks for exchanges of goods and services, and for private transfers (in cash or kind), which may be redistributive in impact. Available evidence suggests that many initiatives could be described as self-help and resource pooling. For example, several large families may combine their resources to purchase a new refrigerator. Individuals or groups may join forces to help with house building. And there are associations of invalids and their parents. Remittances from abroad are unlikely to be significant- -only 8 percent of Russians say that they have even one friend or relative living in a market-based industrial country. Sample surveys are helping to identify behavioral changes. As might be expected, Russian youth appear markedly more oriented toward entrepreneurship than the rest of the population, while the majority, including the elderly, seek government protection, accept falling living standards, and shift to cheaper forms of consumption. Another important factor affecting behavioral change is education, although this overlaps preferences based on age. 2.94 Evidence reflects a growing interest in starting small businesses, but business know-how, marketing skills, and capital are lacking. In order to channel these initiatives into commercially viable activities, training in small- scale entrepreneurship as well as in technical skills and in marketing are needed. Such courses have been offered by the Women's Union of Russia on a pilot basis over the past year in Moscow and have enjoyed considerable success. Women's nongovernmental organizations also provide the moral and psychological support required to start a small business. Their outreach capability facilitates access for women and provides needed feedback on training. In this context, nongovernmental organizations are especially well placed to mount pilot 49 training programs that could subsequently be translated into larger-scale programs under the auspices of the Federal Employment Service. At present, there remains substantial unmet demand for training and advice: applications far outstrip the number of available places in Moscow and in provincial chapters of the Women's Union that have initiated similar programs. 2.95 In response to the pressures imposed by the transition, households may seek to rely increasingly on social relations and subsistence production, rather than on cash exchanges. The characteristics of those adopting such a defensive strategy resemble the characteristics of the general population, in terms of age, income, and education (Rose 1992). A defensive strategy, which involves retaining jobs in the formal economy as long as possible, complemented by activities that can provide some sort of insurance, can be seen as a rational household response to declining incomes and widespread uncertainty, not least about future employment prospects. Such a response is encouraged to some extent by government policy. In Altai Krai, for example, in 1992 one way that school teachers are being compensated is in the form of land grants. Such compensation, however, represents a retreat--a step back from the specialization of work that characterizes modern industrial economies- -and is an inefficient use of valuable human capital. An engineer tilling a vegetable plot for his or her family's subsistence consumption simply reduces market demand for the product and does nothing to promote transition to the market or economic growth. 50 CHAPTER III: POLICY ISSUES AND OPTIONS IN SOCIAL PROTECTION 3.1 In a time of macroeconomic crisis with hyperinflation, the paramount question is the adequacy of the social safety net. It is also imperative to consider whether some immediate reforms would save scarce public resources. But actions taken in the short run have also to consider the longer-term need to develop a social security system appropriate for the competitive market economy emerging in Russia. Over the long term, a stable framework should be established within which individuals can make important lifetime decisions. Economic security is not simply a matter of the adequacy of today's consumption, but also peace of mind about the future--a peace of mind that for many has been shattered by the events of the last two years. Frequent changes to the system of social protection and mixed signals to individuals have reduced the sense of security such a system should provide. 3.2 A key set of questions concerns the roles of different levels of government in providing a social safety net. The Russian Federation has decentralized financial responsibility for social programs to oblast and raion authorities. But the economic fortunes of different localities are likely to differ greatly, and some may require federal assistance if their safety nets are to be adequate. The next section explores the relevant issues and goes on to suggest immediate policy actions. It incluies a discussion of the role of extrabudgetary funds and what should be done with the savings of present surpluses of the Pension Fund. The chapter then addresses the policy choices in the context of fiscal decentralization and presents appropriate mechanisms for central transfers. Strengthening the Safety Net 3.3 Strengthening the safety net in Russia will require three main steps: extending coverage to assist all those in need, ensuring that the programs provide enough protection, and guaranteeing support for the unemployed. These are discussed in turn. Guaranteeing Social Protection for Everybody 3.4 It is widely regarded as a critical function of the state, even in market- oriented economies, to provide a guarantee of minimal living standards to its citizens--at least in countries that can afford to do so. The Soviet regime claimed to do this, and now the onus is on the Russian government. This guarantee is provided chiefly through two means. First, citizens may be entitled to income transfers because of prior contributions to a social insurance scheme. Second, some programs give social assistance to those who, for whatever reason, require additional income support. These include the working poor, homeless people, and newly arrived refugees. Social insurance can be regarded as earned- - these schemes have well-publicized rules and, in most countries, are reasonably well-established. Benefits are usually paid independently of the economic circumstances of the recipient, as are old age pensions in Russia, although--as already observed, recent years have been difficult for pensioners (see Table 2.2 in Chapter II). 51 3.5 Social assistance schemes will become increasingly critical as the market economy develops, entailing higher levels of unemployment. In order to ensure that only those truly in need receive benefits, social assistance schemes are typically targeted through restrictive rules. In introducing a benefit of last resort to prevent Russians from falling into poverty, the following must be addressed: the form of the basic minimum, the targeting of the benefit, and financing and administration. 3.6 Cash or In-Kind? Economists prefer cash transfers in order to allow individual choice in consumption decisions. Cash transfers have less distorting effects in goods markets, compared with in-kind benefits. Because the various groups at risk of poverty in Russia are likely to differ in terms of their resources, living arrangements, and other relevant parameters, income transfers have the advantage of avoiding the predetermination of individual needs. Cash transfers also permit greater transparency in the budget compared to, for example, food subsidies, and are less expensive to administer (see Box 3.1). 3.7 In Russia, however, social assistance in cash is the exception rather than the rule'. While this is largely attributable to the tendency of government authorities to respond to the crisis by extending in-kind assistance, it has been reinforced by the arrival of humanitarian aid in non-cash form. Quantity- denominated vouchers and in-kind provision can be advantageous during high inflation as a means of protecting minimum levels of consumption. Preserving real levels of income transfers through frequent indexation is administratively cumbersome. 3.8 On balance, cash transfers are the recommended basis for the new minimum basic benefit. Existing schemes of in-kind provision, however, are playing a role during the economic crisis and the continuation of subnational forms of assistance in kind should continue, at least in the short term. The proposed minimum cash benefit will take time to develop and implement, and will be limited in usefulness until the economy stabilizes. Once a minimum cash benefit is established, some existing programs may be phased out, while others may be retained as components of the safety net (as is the case in several industrial countries). 3.9 Means Tested or Categorlcal? In the Russian context, which is characterized by severe fiscal constraints and where most of the population is not poor, efforts must be concentrated on those who fall below a defined poverty line. At the same time, however, there are problems associated with targeting, including administrative costs and possible adverse incentive effects. There is always the risk that some individuals truly in need will not receive benefits. While there is a tendency to focus purely on the overall fiscal savings and exclusion of the non-poor (see, for example, Grosh 1992), stricter targeting can reduce the coverage of the poor. In general, the higher the priority given to raising the welfare of the poor, the more important it is to eliminate errors of exclusion. Here we explore the case for means-tested targeting, compared with categorical targeting (that is based on targeting some observable proxy for I/The temporaly means-tested compensation for bread price increases introduced in September 193 is an exception. 52 income, such as number of children). Paragraphs 3.14 and 3.15 discuss the possibility of introducing geographic targeting and self-selection. go3: Social asslstance for Vulnerable Grooma There are alternative approaches to protecting the poor, all of which involve trade- offs. This box highlights the major features and drawbacks of each. 1. Continued subside of very basic foods. To the extent possible, the items selected would be those that mOast people do not consume or consume only in small quantities. The problems with this approach are two-fold. first, given constraints in food supply in Russia over the past decade, food consumption patterns do not vary much across income classes: it is very difficult to identify products that are a significant part of the consumption bundle of the poor and are only a amall part of the mon-poor's oonsumption. All income groups, for example, consume similar quantitios of bread end potatoesa. Second, price controls alone do not solve anything and can le-ad to shortages, especially if reflected in low prices to producers. 2. Food entitlements through the nrovision of counons to poor households. These provide better social protection than cash in a high inflation context, but carry risks from the perspective of the macroeconmiec and stabilisation policy. since coupons maintain their real value through increased expenditure commitents, which cannot be known in advance. Such a scheme may also be relatively ecmplex and expensive to police and administer. Nonetheless, coupons are less xpens*ve gad more compatible with market reforms than unlimited universal access to subsidised foods, 3. Direct srovisivn of food. While direct distribution of domestic food supplies is possible in principle, in practice such schemes have utilized international food aid. To provide food directly to thousands of distribution centers through a set of government orders is logistically and administratively complex. Al. food aid, however distributed, is likely to reduce demand for domestic food. and hen4e Ca= Incmese, which in turn may lead to procurement subsidies that worsen the fiscal deficit. -Food aid also gives the least opportunity for choice by recipients and reflects the coemand econminc system that Russia is trying to move away from. Food aid mty be boet used as a ooo4dity to be sold, with the proceeds used to protect the vulnerable. 4. CXs;,se:nst&n t ke on. Allowing market prices to prevail for all products and providing a degree of cash compensation to the poorest is, in principle, the most economically efficlent sppr*a6h. The major problem arises in high inflation, which requires full and frequent indexation. Benefit adjustments should be based on price movements of a besket of baic 3.10 Fully efficient means testing allows the benefit to be concentrated on the poor, reducing the cost of a given poverty reduction or achieving a greater poverty reduction for a given budget (see Atkinson 1992). Benefits can be tailored to current individual circumstances, confined to those who need it most, and operated as an effective mechanism for redistribution of income. Such programs are widely used in industrial countries, for example, Australia, where the role of income testing has been expanded in recent years. Countries that rely on a means-tested approach to pensions, such as Australia and Canada, have higher benefit guarantees than other industrial countries. 3.11 Administrative costs, although difficult to quantify, are frequently mentioned as a disadvantage of means-tested programs. Information needs to be updated regularly. The evidence, however, is mixed. A recent World Bank study found that there was no apparent correlation between the incidence of program expenditures and administrative costs--which was attributed to the fact that screening costs are typically a small component of total administrative costs. Where the existing physical and information infrastructure is adequate, additional administrative costs are likely to be less important (Grosh 1992). 53 (Universal programs are clearly cheaper to administer than means-tested programs, although that is not the relevant comparison here.) 3.12 Both income testing and categorical restrictions raise potentially adverse incentive effects. Economists' traditional concerns are based on a well-known static model of labor supply, which implies unambiguously that transfer programs will reduce hours of work in the eligible low income population. This is associated with high rates of benefit withdrawal. In general, evidence from the United States suggests that work effort would be higher in the absence of most cash benefit programs, although there is considerable uncertainty about the magnitude of the work response.2 Concern about the poverty trap has led governments to taper the withdrawal of benefits, with the reduction being less than one for one. The definition of the recipient unit is also important: if the total income of the household is the relevant basis, all members face a common marginal tax rate. Another approach is to make the transfer conditional on a work test (the U.S. Earned Income Tax Credit is an example). This avoids full dependency, although work effort may still be reduced to the minimum required.3 Insofar as the proposed minimum benefit will be set at a very austere level of subsistence, the likelihood of adverse incentive effects is correspondingly small.4 3.13 Approaching poverty reduction on the basis of categorical criteria clearly requires some country-specific empirical basis. In Russia, it is known that children in large families are at a high risk of poverty (see Chapter I). However, the fundamental weakness of relying exclusively on categorical benefits lies in the likelihood of poor individuals continuing to fall through the net. Russia has already established a relatively extensive system of benefits for groups designated as at risk of poverty, such as children in large families. Yet most of the poor--54 percent--currently to not receive any transfers from the government. The categorical approach has manifestly failed to provide a floor for increasingly large numbers of people who do not fall into conventional categories of need, the most obvious examples being the working poor and the long-term unemployed. Chapter I have highlighted the changing nature of poverty during the transition from a command economy. Even if the real level of existing benefits were doubled, simulations based on preliminary results of the Russian Longitudinal Monitoring Survey show that the incidence of poverty would only be reduced by about one-fifth, from 37 to 28 percent. Thus it has become clear that categorical benefits alone will not suffice, even if set at levels far more generous than they presently are. 3.14 Targeting benefits on a geographic basis may be a useful broad proxy for need. In Albania, for example, many rural areas are relatively poor, y See, for example, R. Moffitt t1992). t An important additionaL finding in the U.S. 1iterature is that the labor supply reduction associated with the AFDC seldom arises from initially ineligible female heads of household who lower their hours of work below the break-even point to qualify for the program (Moffitt 1992). A/ Work disincentives are far more likely when the benefits are apprnximetely equal to the aarr.ings a recipient would receive if he or she worked fuLl time at the minimum wage, as is the case for Aid for Families with Dependent Children (AFDC) in the United States. 54 particularly in the mountainous regions (Alderman 1992). In Russia, available evidence does not point to such clear distinctions between rural and urban areas, although it is clear that oblasts do differ significantly in terms of per capita income and likely extent of need. The proposed Equalization Fund is designed to help alleviate spatial disparities in Russia (see paragraph 3.45-3.52), while there will be further targeting at the subnational level. 3.15 Targeting may also rely on self-selection. This occurs when informed individuals do not make use of programs for reasons of stigma, opportunity costs, and product quality. In general, however, it is difficult to find a suitable vehicle for self-selecting programs. The various factors that underlie the notion of stigma are unlikely to be significant in the Russian context, because the state has traditionally adopted a fairly extensive role in the provision of income support, queuing is common generally and not an obvious deterrent, and product quality is often poor. On the other hand, the Bank has recommended a coupon system for lower-quality bread in the Kyrghyz republic on the grounds that product quality and the effort involved in collecting one's coupons would entail an element of self-selection (1993 Bank Report No. 11535, paragraph 4.86). There is scope for such schemes in Russia and, as emphasized above, the efforts of subnational authorities in the provision of in-kind benefits to needy groups should be encouraged. The proposed scheme relies upon self-selection in the first instance, because individuals must come forth to apply for the benefit, but there are additional restrictions. 3.16 On balance, it is recommended that a minimum basic benefit be provided using means testing. In contemporary Russia, most of the poor fall outside categorical programs for one reason or another. The evidence strongly suggests that a categorical approach cannot, at present, meet the important objective of universality. At the same time, categorical programs should continue since, like child allowances, they are well-established and reasonably well-targeted and also because it is likely to take up to eighteen months to design and implement a new means-tested system. Preparation time is required to resolve the details related to the conditions of entitlement and to train staff and provide them with the necessary technical support. After the establishment of a means-tested benefit, it would be advisable to review the levels and scope of the existing categorical programs, but not necessarily to abandon those that are reasonably well-targeted. 3.17 Subnatfonal or Central? Current social assistance arrangements rely almost entirely on efforts developed at the subnational level. Only the most basic policy advice is provided by the c. entral government. Oblast and raion administrations have to determine in detail what programs they wish to provide (and can afford), and who should benefit from them. The available evidence suggests that subnational authorities are capable of performing this task, on an increasing scale, given sufficient support and funds. However, a basic ability to extend short-term crisis measures, while commendable, does not equate with the development of a comprehensive and cogent system of social protection. The latter task requires considerable resources and must take into account the relationships among the various benefits, suggesting that it should be undertaken at the federal level. There is an urgent need for the Ministry of Social Protection to adopt a proactive role in this area. 55 3.18 At the same time, subnational schemes may better meet local needs. This argument clearly carries weight in a country as large and diverse as Russia, although the differences in programs examined by the mission appeared to have more to do with availability of resources than specific local living conditions. The advantage of regional variations in programs could be exploited through subnational enhancements to a central scheme, over and above a federally supported minimum. This concept is supported by the Ministry of Social Protection. The critical issue of financing the minimum basic benefit is examined below. 3.19 Administration. The Ministry of Social Protection has the primary responsibility to implement programs for social assistance, and its subnational offices implement centrally mandated schemes. The program of minimum basic benefits will cover up to about 15 percent of the Russian population, of whom some 30 percent are children. Currently, the total staff of the social protection system is 244,000, including some 84,900 social workers. 3.20 In introducing a new minimum benefit, a public relations campaign should be developed so individuals can determine on their own whether they qualify for benefits, thereby alleviating strains on subnational offices. Some verification will occur through documentation. Children will be checked against school and medical records; adults against employment, medical, and other records; and pensioners against pension receipts. Most types of income would be counted, except in cases where the income component is determined to be either too small to merit consideration or counter to the thrust of the reform program. The main goal should be to provide a rough assessment of the needs of families for social assistance. Some degree of leakage can be tolerated, at least over the short term. The proposed system of matching grants will help to minimize the risk of misuse and diversion of funds at the subnational level. 3.21 Means-tested approaches are being considered on a much larger scale with respect to the housing sector.5 Pilot schemes are being implemented in Moscow and other regions. Income testing for housing allowances in Moscow is to occur every six months, and more frequently when families experience an unanticipated drop in income. The process involves self-selection (facilitated by a public information campaign outlining qualification levels and necessary documentation) leading to an interview with a social worker. Those who do not reapply after six months become ineligible. 3.22 The introduction of a minimum cash benefit of last resort, to be implemented by subnational authorities, is an immediate priority. It needs to embrace the following principles: * Eligibility should be based on means testing by subnational offices of social protection. Qualifying conditions should be kept as simple as possible so that there are no doubts about eligibility on the part of either claimant or administrator. Consideration should be given to whether any element of income from work or assets, or 5/Fot an overview of these proposals and the issue of housing administration, see Jill lhadduri and Raymond Struyk, 'Options for Admi-!sterlng Houing Allowances in Russia," The Urban Institute, March 1993. 56 other cash benefits, should be disregarded in order to avoid adverse incentive effects.6 Certain allowances, such as disability benefits, may merit special treatment in order to reflect extra necessary expenditure incurred by disadvantaged groups. * The amount awarded should be based on the oblast subsistence minimum basket, adjusted appropriately for the type of household unit, for example, the number of dependent children (as already defined by the Ministry of Labor and described in Chapter I). This would allow each member of the family unit to obtain essential living expenses such as food, fuel, and clothing. Above the minimum standards there are likely to be subnational variants, reflecting differences in local resources and political preferences. * The basis of claims should be self-selection and self-assessment in order to reduce administrative overheads. Claimants should fill out forms providing the details needed for administrators to carry out the calculation outlined above. Verification of details should be restricted to cases of doubt. There should be no reliance on household visits. * The provisions for the benefit should be enshrined in national legislation, with clear rights of redress to an appeal body for claimants dissatisfied with the decision on their claims. * The question of financing minimum cash benefits in a federal system must be addressed. Although the overall cost of such a scheme does appear to be affordable (less than 3 percent of national income) poorer oblasts are unlikely to be able to afford a minimum basic benefit. The urgent question of whether this should be achieved by reimbursement for the particular program or by a more general equalization fund is discussed below (beginning with paragraph 3.45). Matching grants are recommended, conditional on supplementary local expenditures, in order to meet the goals of efficiency and equity. 3.23 Finally, it is emphasized that several entry mechanisms into the system of social assistance should continue. While the minimum basic cash benefit will form a critical basis of the social safety net, other schemes and initiatives, especially at the subnational level, should be encouraged. In the interim period, prior to the establishment of the minimum basic benefit, federal support should be extended to other types of subnational social assistance efforts in the form of matching grants (see below). Ensuring that Social Protection Programs Protect 3.24 High and unstable rates of inflation in the past year have seriously undermined the effectiveness of existing programs. Clearly, stabilization is an extremely important goal, especially for those who depend on income transfers. i/ This is a difficult question, which may be an appropriate subject for technical assistance. 57 In the interim, the need for adequate social protection raises the controversial topic of indexation. In early 1993 the Parliament began to adjust pensions quarterly in line with changes in the cost of living. In pursuit of that policy, payments were increased by 90 percent on February 1, 1993, and automatic adjustments were legislatively provided for. On the other hand, the government has been understandably reluctant to introduce automatic indexation because of potential inflationary and fiscal implications. Nevertheless, the fact that ad hoc adjustments have left many beneficiaries below the poverty line suggests that there is little alternative. Otherwise, average entitlements have to be set much higher in real terms to prevent severe hardship during the periods between adjustments. Moreover, at current inflation rates, quarterly adjustments are too infrequent. 3.25 The principles that should govern benefit levels in the short term can be stated as follows. The subsistence minimum, set at very austere levels and on the basis that 80 percent of income is spent on food, should be regarded as the anchor of the system. The minimum basic benefit should be set at the subsistence level, with the basket valued in oblast prices. This is the minimum income that the federal government will seek to extend to all Russians who would otherwise fall below the threshold. Social insurance benefits should be set above the subsistence minimum, to accord recognition to past contributions. The minimum labor pension, to which Russians with a twenty-year work history are entitled, should be set above the minimum subsistence. The minimum unemployment benefit should be set at a level that exceeds the minimum pension. Only the minimum benefit should be subject to full indexation. In late 1993 the government adopted the latter as a temporary measure so that benefits above the minimum should increase by the same nominal amount as the minimum, thereby significantly reducing the fiscal implications of full price adjustments throughout the benefit structure. 3.26 The minimum labor pension should be set slightly (say, 15 percent) above the subsistence minimum, and be automatically adjusted whenever the nominal cost of the subsistence minimum basket has risen by a certain minimum of at least, say, 10 percent (though not more frequently than monthly). 3.27 Formally, the level of minimum cash benefits has been linked to the minimum wage. In recent practice, however, this link was not observed; the minimum wage was allowed to erode very significantly in real terms, while minimum pensions at best were subject to irregular adjustments. In early 1993, the minimum wage was increased to R 4,275, once again equivalent to the minimum pension. A distinction has to be made between the minimum wage, which is apt to be used as a numeraire for many other wages in the system, and the wages that are paid. It is important that the level of minimum benefit, including the minimum unemployment benefit, remains significantly below the minimum wage paid for individuals who are working. ( A number of enterprises have been paying idle workers the minimum wage rather than laying them off.) The government needs to make as little use of the concept of the minimum wage as politically feasible in its policy actions. For example, it is unfortunate that senior official salaries and the share of wages that can be treated as costs for tax purposes are set as multiples of the minimum wage. 58 Guaranteeing the Minimum Level of Unemployment Co9Mesati oa 3.28 In Chapters I and IV, it is shown that unemployment has developed at a much slower rate than was widely expected, but that high levels are inevitable if economic reforms lead to a restructuring of the economy. Retaining unemployed people in their work places at a reduced wage (such as through short-time working or periodic layoffs) can be justified only in exceptional circumstances (see Chapter V). Unemployment must therefore be expected to increase very substantially in the future. 3.29 Faced with this unprecedented problem, the Federal Employment Service (FES), assisted by consultants with relevant international experience, is developing a national network of employment offices. This includes proactive programs to provide emergency public works and job counseling and promote retraining, which are not simply measures of social protection but ways to improve the efficiency of a market economy. Employment services cannot, of course, substitute for the economic restructuring and growth that are ultimately needed to provide productive jobs for those unemployed or underemployed in activities with poor prospects. In the short run the most urgent need is to develop the capacity to register claimants and to pay adequate benefits. This is a high priority of the program of computerization under a World Bank loan. But a revision of the principles on which unemployment compensation is set is also required. 3.30 Minimum income support for the unemployed should be kept above the poverty threshold, as defined by the subsistence minimum. The unemployment benefit should not be means tested and, for at least the period of eligibility (twelve months), this should be slightly higher than the subsistence minimum. 3.31 For those whose unemployment extends beyond twelve months, for those who quit voluntarily, and for first time job seekers who are registered as unemployed, the benefit level should be at least as high as the subsistence minimum. This will require legislative changes, because currently no distinction is made between those were involuntarily laid off and those who quit their jobs. Recipients of this benefit may enroll in retraining programs or public works programs (although normally, public works programs will pay more than this benefit). 3.32 The present method of calculating individual benefits is time-consuming and pointless during high inflation, as most people end up receiving virtually the same benefit. The method can also prove administratively cumbersome in the face of a rapid expansion of unemployment in a still largely uncomputerized system, and thus delay t't' payment of benefits. In 1992, the FES proposed a legislative amendment that would allow the temporary use of flat rate benefits. The Parliament and trade unions resisted the proposal. We recommend that flat rate unemployment benefits be adopted on a temporary basis. 59 Financing the System of Social Protection Reform of Financing of the Employment Service 3.33 Sole reliance on payroll taxation to finance unemployment benefits may seem absurd. The yield of the tax falls as the need for expenditure rises. This is, however, the way that such benefits are financed in most industrial countries, presumably in order to emphasize their insurance character. Even though the tax rate doubled at the beginning of 1993, it remains low in relation to probable future needs. Moreover, with 90 percent of the revenue retained at the oblast and raion levels, public resources cannot be shifted from oblasts with low unemployment (higher revenues and lower expenditure demands) to those oblasts with high unemployment experiencing collapse of their tax base. 3.34 At present, the Finance Ministry is empowered to replenish the Employment Fund (EF) from general revenues as needed to pay unemployment benefits. This is essential. The conditions under which budgetary support will be given are not, however, spelled out, and too narrow an interpretation could have adverse consequences. For example, if the requirement is that the Employment Fund on a national level be exhausted, this would ignore (1) the lack of capacity to shift unused balances from one oblast to another, and (2) the likelihood that proactive programs, which are financed from the same Fund, will suffer cutbacks just as the need for such activities increases. 3.35 Suggested solutions have included centralizing a much larger proportion of employer contributions and earmarking Pension Fund surpluses. The former has been at least temporarily rejected because it might reduce the incentive for raion offices to chase delinquent enterprises. In the present cIrcumstances, in which employment offices have limited resources for enforcement, this is a realistic concern. The issues of revenue collection and the possible use of Pension Fund surpluses are discussed below. 3.36 An appropriate option would be to finance proactive policies from the budget, at least in oblasts with unemployment over a certain level.7 The first step is to finance proactive policies directly from budgetary resources rather than from the Employment Fund. The EF would continue to be largely retained at the subnational level, solely for paying unemployment benefits. In the currently constrained fiscal context, it is unwise to recommend the reduction of any tax bases or rates. Thus, the payroll tax for the Employment Fund should be retained at its present 2 percent level. Given the generally low average replacement rates, this is likely to be adequate to finance benefits for local unemployment of up to 5 percent, depending on compliance rates. The Continuing Role of Extrabudfetary Funds and Payroll Taxation 3.37 The extrabudgetary funds that have proliferated at all levels of government play a critical role in financing social protection. On the one hand, this has y/ In an ecoromy much closer to macroeconomic equilibrium than contemporary Russia, this would have an additional advantage--that running an unplanned budget deficit to finance unexpected unemployment is a useful stabiliser. 60 obscured the fiscal position of consolidated government to a considerable extent and could seriously limit the efficacy of government attempts at stabilization. On the other hand, there is a widespread belief that extrabudgetary funds have enabled the mobilization of additional resources for social protection. 3.38 With respect to the insurance-based benefits, it is feasible to have a social security system based on individually identified contributions and individually calculated benefits that comingles payroll tax revenues with other taxes, so that social insurance benefits become a general budgetary obligation. It is the contribution that determines the earned right, rather than the existence of a separate fund. In practice, however, the concept of earmarked individual contributions tends to entail the establishment of a notional fund. The few countries that pay pensions out of general revenues have flat-rate or means-tested pensions. 3.39 In principle, earmarking funds for social putposes need not be based on a payroll tax. A proportion of income taxes could be earmarked as contributions to a national pension scheme. In practice, however, the notion of the fund seems invariably implies a payroll tax. In Russia, at least in the near future, fiscal constraints, the undeveloped personal income tax system, and the objective of individualizing entitlement programs (see paragraphs 3.86-3.88) argue for continuing with wage bill taxation. For the time belng, the Pension Fund should continue to be based on a payroll tax of 28 percent and should maintain a separate existence from the government budget. Further institutional reforms to the Pension Fund are recommended below (see paragraphs 3.91-3.98). The Nature and Uses of Extrabudgetary Funds 3.40 If there are to be funds, tne next question concerns the principles on which contributions should be set. Should the funds aim to be self-financing in the short run (known in the pension literature as pay-as-you-go), in the long run (known as fully or partly funded), or in some other form? Should the funds be centralized or should funds be organized on a regional basis? A third set of questions concerns the purposes, if any, other than the payment of benefits for which the funds might be used and who should manage them. These are not, of course, wholly independent questions. 3.41 The issues are urgent. In 1992 and 1993 the Pension Fund ran substantial surpluses. Contributions to the Fund are deposited in the Central Bank and earn annual interest of 3 percent (the Pension Fund management, however, has independent control over penalty payments received, which are substantial). It is possible that the Pension Fund might become an independent financial institution. In Moscow oblast, it was proposed to open a commercial bank to take over pension payments from the post office. The establishment of a hybrid investment/commercial banking network with the Pension Fund owning 51 percent of the stock has been envisaged. While this may reduce the high charges for postal delivery of pensions, in other respects the proposal is clearly undesirable.8 j/ The broadly desirabl.e goal of nore competitive retail banking could be achieved by breaking up the Savings Bank on some lines other than a purely geographic basis and by encouraging existing commercial banks to expand to serve households as well as enterprises. 61 State establishment of a national network of financial offices with a captive clientele of pension recipients is not compatible with market-induced competition. Indeed, by its monopoly of such a large part of the market, such a network may discourage other entrants. The Pension Fund should not establish or become an independent financial institution. 3.42 Purely commercial ventures by the other extrabudgetary funds should also be strongly discouraged. Faced with inflation rates of up to 30 percent a month, it is understandable that in some places the Employment Fund and the Social Support Fund are being used in various innovative activities. The EF is advancing loans at below market interest rates to new and old enterprises with the notional objective oE creating jobs. This activity is taking place without sufficiently clear or transparent guidelines. Similarly, the Fund for Social Support (FSS) is investing in enterprises that are producing consumer goods. There is a case for enterprise support to avoid very high levels of unemployment but on the basis of clearly defined criteria (see paragraphs 5.22-5.33 in Chapter V). Pptions for Change: Intergovernmental Fiscal Relations 3.43 As a recently formed independent federation. Russia is still eearch5.ng for a workable division of policy responsibilities and resource allocation among central and subnational governments. This is a problem for any government: there is an inherent tension between the virtues of decentralization and the need to prevent excessive economic and social disparities among political units. On the one hand, transferring revenue and expenditure powers to those levels of government closer to the people (taxpayers! should lead to greater willingness of beneficiaries to share the burden and actively participate in the design and management of programs, and thus better hold decisionmakers to account. On the other hand, important considerations call for a well-financed central government that can provide a safety net both for individuals and for the provision of basic services in areas of economic distress. 3.44 In Russia, the central government has assigned major social sector expenditures to oblast authorities, which have in turn transferred some of these responsibilities and taxes to raions. This has caused problems. There does not appear to be a correspondence between the taxes assigned to the subnational levels of government and their present expenditure responsibilities, and recent legislation9 does not take into account the need for equalization. Nor does the allocation of taxation bases accord well with what are generally thought to be rational principles of assignment. These issues were discussed in a recent World Bank report, Intergovernmental Fiscal Relations for the Russian Federation (Report No. 11302-RUS), and need not be repeated here. It was concluded that "subnational governments may be caught in a vice of declining tax revenues and increasing expenditures" (paragraph 2.11). Decentralization has increased the already wide fiscal disparities among the regions of Russia, reflecting the marginal equalizing (and in some cases counter-equalizing) role of federal 2/ Specifically, the Law an Rishts of Local Self-Government and the Law on Principles of Taxation (which have not yet been implemented). 62 grants. The emerging difficulties that have become evident in the context of social assistance were highlighted above (paragraphs 2.67-2.71 in Chapter II). The Role of Central Transfers 3.45 On equity grounds, there is a need for central authorities to redistribute tax receipts among regions. Even federal countries, like Australia and Germany, normally regard the maintenance of minimal living standards as too important to be left entirely to subnational authority discretion. This need not imply fully centralized financing, but it does suggest that at least some central support needs to be available to oblasts that have an unusually high incidence of poverty. Moreover, given the magnitude of economic restructuring that is required and the likelihood that some local jurisdictions will be faced with virtual collapse of their economic bases, the central government will not be able to abdicate responsibility for many areas of social expenditure. 3.46 The recent World Bank report recommends a new approach to intergovernmental fiscal relations in the Russian Federation (1993 Report No. 11302-RUS). It would represent a shift back toward tax pooling and sharing, rather than tax assignment. To summarize, the framework has four components: (1) a Common Pool of revenue that is notionally divided between federal and oblast goverrments, based on their respective expenditure responsibilities; (2) allocation of a portion of the pool on a derivation basis (that is, to the oblasts where the taxes are actually collected); (3) allocation of the remaining part through an equalization formula; and (4) subnational taxes and surcharges. The center would also reserve certain taxes for itself (such as those on trade), and all taxes would continue to be collected by the State Taxation Service. In addition, because the protection of vulnerable groups during the transition is a national priority, the World Bank has suggested that the federal budget resume financing if subnational revenues do not suffice (Russian Economic Reform, September 1992, paragraph 9.29). This raises several issues, such as the relative merits of block versus tied allocations, the role of matching grants, and closed versus open-ended reimbursement. It proceeds on the assumption that the central government wishes to ensure minimum levels of income security across the country. 3.47 In principle, central grants (or taxes shared on the basis of an equalizing formula) can have several objectives. These include reducing fiscal imbalances created by the system of tax and expenditure assignments or by different resource endowments and economic and demographic structures. The objectives also include ensuring a minimum pattern in the provision of services in each locality, while creating incentives for least-cost provision and internalizing the externalities.1" The degree to which central transfers (or taxes shared on the basis of an equalizing formula) are able to reduce disparities depends largely on their size relative to the derivation part of the Common Pool, which will reflect the distributional objectives of the government, its assessment of incentive effects and political considerations, such as the risk of promoting separatist tendencies among wealthier oblasts. IV For exawple, the benefits of public health and education are not confined to individual localities. 63 3.48 Block grants (lump sum transfers) are efficient in the sense that oblast authorities may decide how best to spend the money in light of local democratically expressed preferences. In practice, however, this may not ensure adequate social programs, and central guidelines will be needed to ensure that effective social programs, including a safety net, are maintained. Legislation and exhortation may not offer adequate assurance that the desired minimum standards will be implemented. 3.49 The allocation of categorical grants can be based on such formulas as indicators of need (for example, the number of residents below the poverty line) and/or a case-by-case review of proposals. These may have the advantage that the center could direct funds to areas of priority for social assistance (assuming, of course, that the center would--and that subnational governments would not--do so). Funds thereby directed to priority areas may be additional, or alternatively, may simply substitute for what would have been spent in any case, since central grants are fungible. Concern has often been expressed that local governments come to view grants as substitutes for local taxes and user charges (Bird 1990; World Bank 1988, pp. 165-66). 3.50 Alternatively, matching grants--which are usually conditional--require subnationai governmentLs to supplement any transfers with their own money.1' Matching grants may induce a larger degree of subnational involvement for particular expenditure programs. It remains in the locality's interest to continue raising revenues. The Canada Assistance Plan is a case in point; it commits the federal government to sharing with the provinces, on a fifty-fifty basis, allowable costs in social assistance. Through this approach the center hoped--and to a large extent succeeded--to draw provincial activity in desired directions. (Block grant formulas can also specifically include subnational tax efforts.) 3.51 Oblasts differ from one another in their capacity to raise their own revenues and in their needs for expenditure. A central equalization mechanism could seek to capture both in one formula. For example, tax capacity might be measured by an indicator of average real income and need, by some measure of poverty. If the equalization formula was based on some weighted combination of these, would the formula be calculable? Some of the data, such as population, age structure, and price indices, are already available by oblast. Data on oblast poverty incidence could be derived from the results of the RIMS only if the survey were to be greatly expanded, which is under consideration. A measure of the oblast poverty gap would capture both the numbers of people below the local poverty threshold and the seriousness of income shortfalls. 3.52 Assuming that it is possible to rank every oblast on the basis of (1) average per capita real income and (2) oblast poverty gap, several prior questions must be addressed in formulating an equalization mechanism. It is necessary to decide whether to use one, the other, or both indicators. Giving high weight to average incomes may underestimate the poverty problem; conversely, giving too much weight to the poverty gap may penalize oblasts that have effectively reduced intra-oblast disparities. If there are no matching J/ See Inteeaovernmeintal Relatiow in the Russian Federation 1993 (Report No. 11302-RU), Annex 2. 64 grants for social assistance (see below), the argument for increasing the weight of some measure of the poverty gap in the equalization formula is stronger. On the other hand, if oblast authorities have a generally satisfactory approach to alleviating intra-oblast disparities, the index of per capita income should be the dominant indicator. 3.53 The proposal for matching grants represents a type of targeted intergovernmental grant that is justified in light of the importance of the expenditure. A priority for federal financing should be the minimum basic cash benefit (begin with paragraph 3.19). Before implementation of the minimum basic benefit, matching grants should be made available for other subnational targeted forms of social assistance, which may include in-kind benefits. For administrative reasons, entitlement to such assistance is best determined at the local level, so that a mechanism for central financial transfers i' needed. 3.54 Definitive guidelines are not available to assist the Russian government in setting the precise matching rate for particular expenditure programs, or in determining how those rates should be varied in accordance with the characteristics of different subnational governments.12 The matching rate for each program may be seen as comprising two components. First, the rate should reftlect the degree of interest of the central government in tbe provision of the particular program, that is, the externalities associated with the program. Second, the rate applicable to any locality should be inversely related to the need (income level or capacity) of the subnational government. Basic social programs would rank highly with respect to the first component. However, the most appropriate formula for any service can probably be best determined by trial and error. 3.55 Would it be possible to structure matching grants for social assistance so as to meet the goal of equity, the desire to encourage subnational efforts at social assistance, and the provision of appropriate incentives to subnational authorities? More fundamentally, what are the advantages and disadvantages of matching grant schemes operating to supplement equalization transfers? Oblast price differentials are particularly relevant to the minimum basic benefit, which is based on a minimum consumption basket costed at local prices that favors reimbursement for actual expenditures, rather than uniform transfers. On the other hand, central transfers may have adverse incentive repercussions. Subnational authorities may feel able to reduce their programs of social assistance in the knowledge that the federal government is prepared to meet the resulting gaps, which would be problematic insofar as evidence shows initiatives are being undertaken by oblast and raion governments, on a widespread scale, to supplement the incomes of the poor within their jurisdictions (see Chapter II). 3.56 An added incentive problem is likely to arise with open-ended cost reimbursement grants: subnational authorities lack any incentive to limit the entitlement of local citizens to federal funds. This is particularly probable in a scheme where information requirements rely entirely upon subnational authorities. Generally, one would expect full cost reimbursement to be very &,/ Se. Interaovernmental Relation. in the Rusmian Federation LRenort No. 11302-RU). Amnex 2. for a detailed discussion. 65 inefficient because of these incentive effects, beyond any potential problems relating to fraud (see Box 3.2). Open-ended matching grants may make it difficult for the federal government to achieve overall budgetary control. As in the Canadian social assistance scheme, where a cap was imposed on federal payments to the wealthier provinces. Ceilings related to oblast per capita income or incidence of poverty should be considered in Russia. I oJetJensfe Rs 3L2. Incentive Problms and Pederal Transfers: The U.S. Experience Russia can learn from the expetiences and mistakes of other federal countries in the design of transfers. The experience of the federally funded Medicare system in the United States ia a relevant case for comparison. Originally operating as a cost-reimbursie3t seheme, hospitals reported their actual costs to the center and were relibursed the full amount. The evident result was escalating expenditures since hospitals had no incentives to contain costs. The sche"e was amended to take the form of lump-sum payments (for type of illness), so that medical facilities would not have incentives to perforzm unnecessary teasts and procedures. Under Medicaid, the federal goverent partially reimburses states for their payments to healtb providers of the needy. A legislative loophole enabled states to increase their payments to bospitals in order to generate more federal matching funds, then receive a kickback from the hospitals, In soe states, only a smell proportion of the windfall went to expanding health services for welfare recipients. Fetderal Medicaid money was thus diverted to sucb other purposes as highway construction. 3.57 It is recommended that conditional matching grants be developed for social assistance, with the exact formula (matching rates) based on estimates of oblast poverty rates. Once the minimum basic benefit is established, matching grants should be directed only to expenditures on this benefit. Seeking Cost Savlngs Wherever Possible 3.58 Because the foregoing recommendations would lead to an increase in real public expenditure at a time of fiscal crisis, it is essential to try to find ways to reduce costs elsewhere. Revertine TemRorarily to Uniform Adiustments to Pensions 3.59 By far the largest social protection expenditure is the program of labor pensions. In report discussions, there was general government agreement about the need to restructure the system of pensions. Here the most immediate possibility is to restrict the degree to which pensions are earnings related. If there had been a long history of differentiated individual contributions into an earmarked fund, it might be suggested that those who have contributed more are entitled to higher pensions. In fact, early in 1992, many pensions fell to minimum levels, and existing differentials were effectively curtailed by the effect of rapid inflation on unindexed pensions. In May, the Supreme Soviet partially restored earnings-related pensions, setting the maximum at twice the minimum. In October, the maximum rose to three times the minimum, as in the original Pension Law. The ceiling has a critical impact on total pension expenditures. 3.60 In the third quarter of 1993, the government determined that inflation adjustments to all pensioners take the form of uniform amounts, calculated on a basis that would maintain the real value of the minimum pension. This should be 66 a temporary measure to permit adequate minimum pensions and which will be a priority for reversal as soon as the fiscal situation becomes reasonably stable. Reducing the Eligibility of Working Pensioners 3.61 The cost of any benefit scheme is determined by the size of the benefit, the number of eligible recipients, and the administration costs (administrative questions are addressed below). In terms of reducing eligibility, old age pensions are again the most obvious candidate, because of both overall cost and relatively wide access. Reducing eligibility is more difficult than allowing pension differentials to erode with inflation. Existing recipients have to be "grandfathered": the change in eligibility can apply only to potential new recipients, and the cost savings will occur only slowly. The following possibilities should also be considered: * A possible exception to the grandfather clause, which could be implemented immediately, is to prevent pensioners who work from accumulating additional years of service that can be periodically used to increase pension benefits. Not only does the existing practice raise average pensions, it also incurs significant administrative costs because pensions have to be frequently recalculated, typically by hand. _ A more important measure is to reduce the extent to which old age pensions are paid irrespective of whether the individual continues to work. It seems inequitable to pay pensions to those who continue in their regular occupations, at the price of having to restrict pensions to a level that leaves those solely dependent on pensions in poverty. At the same time, it also seems unfair to restrict the economic activity of pensioners--to deter them from starting small businesses, for example. It should not be illegal for pensioners to work. An appropriate policy that would not prevent pensioners from working would involve enforcement of existing legislation that requires individuals to leave their current employment in order to be eligible for a pension. In due course, as the economy recovers, as the retirement age increases, and as a new contributory pension system develops, this measure might be reversed. One may argue that the more an individual can be said to have earned an entitlement to a pension, the stronger his or her claim to receive it whether or not he or she is working. 3.62 Both possibilites listed above could have negative effects on women, who are on average less well paid, and in the past have been able to use the combination of earlier pension age and the continuation of employment to try to increase lifetime earnings. A floor could be built in to safeguard against this differential impact. Reforming Sickness Benefits 3.63 The following paragraphs show that there is a robust efficiency argument for public intervention in the provision of sick pay (paragraphs 3.113-3.125). 67 While employer mandates have important economic and practical advantages over publicly financed provisions, the main risk is that groups who are at high risk of illness will be discriminated against in hiring decisions. Employer bias against those who are perceived to be expensive to insure may lead to labor market segmentation, to the detriment of individuals who are likely to be already disadvantaged. Whatever approach is adopted, it is difficult to design a scheme such that the sick would stay home and the fit go to work. It may well be appropriate to incorporate features, such as waiting periods of up to two days for the employee, to discourage cheating. In the short term, certain reforms of the current Russian system are appropriate. 3.64 In late 1993, the future of the Social Insurance Fund was unclear. The World Bank recommends that in the short term, the functions and revenue base of the Social Insurance Fund, which is used to pay sick leave, should be transferred to oblast authorities, and the payroll contribution should be collected by the Pension Fund. Cost savings would be achieved by the recent government decision to transfer the work of the Social Insurance Fund out of the control of the former official trade union by greatly reducing the proportion of the fund that is devoted to things other than its primary purpose. Administering cash benefits is not an appropriate role for a monopoly trade union in a market economy and, as competing unions increasingly emerge, the charges that the Federation of Independent Trade Unions is profiting from this and/or favoring its own members will grow. 3.65 Further budgetary savings could come from transferring to employers and/or employees the first part of sick leave. It is common practice in OECD countries for the employer to directly bear the wage costs of the sick employee during the initial period of illness. This may be mandated by the government. For Russia, the International Monetary Fund has suggested one month. Given potentially adverse equity effects (see paragraph 3.120), no recommendation is made at this point. A number of countries(for examplc, Japan, Greece, and Italy) impose waiting periods on sick individuals (three days). It is recommended that the employer have discretion whether to impose a two-day waiting period before employees become eligible for sick pay and that consideration be given to the introduction of employer mandates on sick pay. Reforming Familv Benefits 3.66 Further analytical work to determine different levels and types of financial support needed by families and how it can best be provided is required, with a view to rationalizing the complex system of entitlement to child allowances and improving targeting and administration. Although now eroded by inflation, child allowances have been generous by OECD standards. 3.67 It is not recommended that child allowances be income-tested. Family allowances are inherently well-targeted, since the presence of children, particularly three or more, sharply increases the likelihood that a household will be in poverty (see Chapter I). Thc presence of young children usually coincides with periods in which family incomes are at a relatively low point of the life cycle, given young wage earners and diminished second-earner income. Family allowances have low administrative costs, given basic information requirements (initially a birth certificate and proof for student status after 68 age sixteen). If a system of income testing is to be developed, as is recommended for the minimum basis benefit, household income and needs have to be considered as a whole, with the presence of children only one consideration. 3.68 The current system of assigning both administration and responsibility for some elements of family policy to the enterprise, through leaves and allowances to mothers of small children, also needs review and adjustment in light of the changing roles of enterprises and government agencies. The need to establish a high-level body at the ministerial level for the implementation and coordination of family policy and the provision of services of social support should be considered. Assessing Fiscal Implications of Re!orm Options 3.69 This section presents estimates of the overall impact of the reform package, based on the additional costs and financial savings that would have occurred if the proposed reforms had been implemented in 1992 (See Table 3.1). The list is not comprehensive because the fiscal impact of some proposals cannot be computed using available information. Calculations are based on expenditures and revenues for 1992. Some calculations are only illustrative--comparing actual expenditures to what would have been spent had the reform been implemented in a certain quarter. The estimates of fiscal savings and incremental costs must be treated cautiously. The present size, let alone, future projections, of the deficit or quasi-fiscal deficit cannot be determined with any confidence. At this stage, it is very difficult to go beyond the general trends of revenues and expenditure. Pension Reform 3.70 The following are issues to consider in addressing pension reform: * Introduction of a retirement test. In this case, a pension is paid only if the individual retires. It is estimated that around 25 percent of pensioners are still working. If they had continued to do so even after the retirement test was introduced, the PF would have saved R 160 billion (or R 13.4 billion on a monthly basis) in 1992.13 If some pensioners had decided to retire instead, the expenditure savitgs would have been correspondingly less, and the revenues of the PF would have been affected. * A pension paid only If the individual leaves his or her old Job. This simply involves enforcing existing rules. Even if there are no financial savings to the Pension Fund, this would help facilitate restructuring of the economy. * A minimum penslon to be paid to those who continue to work until they retire. If all working pensioners had continued to work and were paid only the minimum pension during 1992, the financial savings for the full year would have amounted to R 95 billion. IV This assunes that the average pension for working pensioners is the same as that of other pensioners. 69 * Unlform adjustments, rather than adjustments based on earnings-- which would entail considerable savings. In 1992, a fourth quarter uniform adjustment would have reduced pension expenditures by about R 200 billion--or 2 percent of fourth quarter GDP. Social Insurance Fund Reform 3.71 Levylng User Fees that Cover Costs of Sanatoria. The subsidy to sanatorium users for 1992 was estimated to range between R 15 and 18 billion. Assuming that demand is price-inelastic, full cost recovery would have saved this amount. E&mlovment Fund Reform 3.72 Financing proactive employment programs from general revenues would have had an overall revenue-neutral effect in 1992. In the future, when proactive employment expenditures would otherwise have been squeezed by the payment of unemployment benefits, additional budgetary expenditures will be incurred. At this stage, it is not possible to estimate the amounts involved. Introduction of a Minimum Cash Benefit 3.73 Given reliable estimates of the size of the poverty gap in Russia today, we are able to derive a sense of the amounts of money required to lift everyone to the level of minimum subsistence. The calculations are based on preliminary results of the RLMS, which recently became available. According to those rough estimates, it appears that the poverty gap in 1992 amounted to just over R 264 billion, or 1.7 percent of GDP. For about half of those in poverty, a minimum cash benefit is recommended. Assuming that administrative costs equal 10 percent of benefit expenditures and that there is a leakage rate of 30 percent implies that raising minimum benefits up to the poverty line would cost about 3.0 percent of GDP. The estimates of administrative costs and leakage rates are based on the Latin American experience with individual assessment programs.14 In fact, the administrative costs and leakages are likely to be lower, since about half of the incremental costs would represent increases in established benefits (pensions and unemployment benefits), which are not means tested. This should be put in the perspective of the existing structure of public expenditure, where budgetary subsidies to enterprises amount to over 22 percent of GDP and an additional 21 percent of GDP is being directed to enterprises and special programs through directed credits from the Central Bank and the Ministry of Finance with a substantial subsidy component. There is clearly scope for reallocation of resources so as to benefit the needy of Russia. 3.74 The estimates used are based on a very low poverty threshold, which can be justified as a way to focus attention on the worst off in the short run. Once l 0Taking the worst case scenario of Latin American individual asessment programs (Grosh 1992) hardly alters the picture. In this case, administrative costs were 28.5 of program costs. Even making the cnmervative assumption that a Russian program with such high administrative expenses would not result in better targetin8 (with leakage rates remaining at about 302), we find that the overall cost is still approximately 32 of GOP. We can also take a more optimistic case, and aesume that the program achieves the efficiency of the U.S. Supplemental Security inous program, with administrative costs of 8.72 end a leakage rate of 3.32 (Mackintosh 1985). This results in a figure of about 2X of GDP. 70 the economy recovers and growth resumes, the proportion of food in the minimum subsistence income calculation ought to be reduced from an average of 80 percent to 68.3 percent. Following the liberalization of prices for utilities and housing, non-food expenditures will become a much more important component of household expenditures. The higher poverty threshold i'. that which was agreed upon by the Russian government following extensive work and discussions. It is still very austere compared with the United States, for example, where food comprises about one-third of subsistence income. AnticiRating Future Fiscal Developments 3.75 In 1992, despite falling output and falling real incomes, the revenue sources of social protection remained reasonably buoyant. This was largely because the primary revenue base- -payroll taxes- -continued to be levied on a work force that had not yet been subject to large-scale retrenchment. Real wages, after the sharp fall at the beginning of 1992, subsequently rebounded. As a result, each major extrabudgetary fund was running a surplus through to the end of 1993, while benefit levels and unemployment remained relatively low. 3.76 This situation will change. The general trend is clear: once stabilization and enterprise restructuring begins, public revenue will fall. Receipts from the inflation tax, the profit tax, and the value added tax will fall as inflation slows and output declines further. The payroll tax collections will fall as unemployment rises. The extent and size of these effects depends on the speed of stabilization and structural adjustment. It is virtually impossible to quantify general projections. Not only are basic data lacking, but the macroeconomic policy framework has been in total flux. Table 3.1. Indicative Savings from Different Reform Options Estimated savings, first 6 Percentage of Reform Measure months of 1993 GDP Uniform pension adJustments* 527 1.3 Retirement test for pensioners 196.3 0.5 Minimm pension for those who work 157.3 0.4 Full cost recovery--sanatoria use 18 0.04 Note: * calculated on the basis of fourth-quarter savings. Source staff estimates. 71 3.77 Employment adjustment will place demands on the system from both directions: reducing the tax base generally and increasing the financing needs for unemployment compensation. Yet the magnitude of the financial needs is not as large as might be feared. On the assumption that all recipients receive a flat rate benefit set above minimum subsistence, as recommended in this report, it is estimated that every additional 1 percent of unemployment would increase expenditures on unemployment benefits by less than 0.2 percent of GDP. 3.78 The overall impact of unemployment on the quasi-fiscal deficit is difficult to determine. The first round effects in a relatively simple case would show budget savings equal to the size of labor costs (wage bill plus payroll taxes), incremental costs (which include unemployment benefits and additional social assistance to keep families affected by unemployment above the poverty line), the loss of payroll tax revenues, and reduced value added tax revenues. Assuming that the share of wages in GDP is less than 40 percent, and ignoring any accompanying falls in output, the conclusions that emerge are that the incremental costs from a 1 percent increase in unemployment are quite small (less than .3 percent of GDP); and furthermore that the net benefit to the quasi-fiscal deficit of withdrawing enterprise subsidies and forcing people to become openly unemployed and supported by the budget is quite small. 3.79 More generally, the formal budget deficit is likely to widen following a sharper separation than exists now between monetary and fiscal policies. The Bank has recommended that existing subsidies be replaced, as much as possible, with compensating income transfers, and that all producer subsidies be reduced and made explicit. This would simultaneously reduce the real deficit and, by increasing the transparency of macroeconomic policy, increase the formal budget deficit. Shaping the Long-Term System of Social Protection 3. q0 Any social security system has to be considered with respect to its finance and its administration. In Russia, added complexity derives from the federal nature of the government and the need to determine the role of each level of government, both financially and administratively. At the center, the well-known confusion of roles and poor relations between the executive and legislative branches is evident in the absence of any consensual process of benefit determination and in competing proposals to take over the administration of pensions. 3.81 A critical aspect of the system of public finance and expenditure is its ability to cope with macroeconomic downturns. It has long been recognized that in economies that are much closer to macroeconomic equilibrium than contemporary Russia, tax-financed income transfers can be counter-cyclical--that is, the transfers operate to automatically stabilize the economy. On the other hand, there is the risk that the level and structure of transfers (entitlements) could in fact sustain, or even cause, a fiscal crisis. The latter issue looms large in the United States today. For example, automatic indexation of social security payments constrains the government's choice of other public expenditures if taxes are not to be increased significantly. Certainly, entitlement spending is a policy c'aoice, but if it is politically difficult to alter once in place, future fiscal reforms can be jeopardized. The proposals made in this report are not 72 meant to be cast in stone; moreover, in order to minimize this risk to future fiscal reforms, the benefits structure is anchored to a level of minimum subsistence. 3.82 Indeed, at present there appears to be the worst of all worlds. There is no clear pension policy, no link between cash benefits and other fiscal issues, no consensus on what should be done and maximum ill will between the executive and the legislature. The branches have overlapping roles. This not only prevents an adequate evaluation of alternative policies and competing claims on resources, but could lead to a situation where one side can seek to make political capital out of appearing to be more generous than the other. While an urgent priority is to prevent such disagreements from seriously damaging the well-being of vulnerable groups or worsening the fiscal situation, the intensely political nature of these issues limits what outside recommendations can contribute to the long-run solution of the underlying problems. Nevertheless, a focus on the key economic, administrative, and financial issues identifies some critical measures that do not depend on resolving issues of governance, and also help to illuminate the issues of governance. Our focus is on policy actions that should be taken in the next twelve to eighteen months; in some cases, our recommendations are to proceed more gradually than some current proposals suggest. The Economic Issue: -Limit the Ataregate Coqt- nf the Pen-sion Sy2te 3.83 Everybody's consumption comes from current GMM. Claims that can be made on national income depend on (1) direct income from selling labor services, (2) entitlement due from legal possession of tgrets, (3) entitlement based on politically determined rules, such as pension rights, or (4) ex gratia assistance.15 The transfer of incomes is from those directly involved in producing it to those who impose a burden on the working generation. Since Russia, like other industrial countries, faces an aging population and the emergence of long-term unemployment, this burden will increase. The social security system affects the size of the burden in two ways. The first is through its effects on individual behavior--the incentives to work and to save, which affect the size of the national income out of which benefits must be paid. These effects are complex even in countries that have experienced much longer periods of relative stability than has Russia, and there is no consensus over their magnitudes. Since many economists believe that an adequate public social security system is likely to be associated with lower private saving, an important question is the impact of the scheme on public saving. Although this report has a limited discussion of the uses of extrabudgetary funds, a full examination of incentive effects is beyond its scope. The second determinant of the burden concerns eligibility rules. The key question here is the age at which pensions are paid. 3.84 Mosr industrial countries have fixed the retirement age at around sixty to sixty-five. Although there has been some tendency for this to decline with 0/ Both (2) and (3) depend crucially on legal and political recognition of what constitutes a right to a etream of income that is not a payment for direct labor service. Altbough pension entitlements are not marketable, they do have much in coimon with property income. 73 increasing prosperity and prolonged economic recession, a rising life expectancy and the general aging of the population are offsetting forces. In the United States, the age of eligibility for retirement on full social security will rise slowly, from sixty-five to sixty-seven between 2003 and 2027. Retirement ages in Russia are low and should gradually increase. An idea that was proposed in the Parliament has been to raise the retirement age by one month per year until it has increased by five years--that is, over sixty years. While this is the same rate of increase as proposed for the United States, the different starting point makes it too slow to achieve useful cost savings. A ten-year time horizon- -increasing the retirement age by six months per year would be more effective. The increase in retirement age should extend for ten additional years for women so as to enable equality with men. 3.85 Although the impact of this reform would ultimately be considerable, it will take time to come into effect (see Figure 3.1). It is fairer to announce the schedule for retirement-age increases several years in advance to help older workers plan their post-retirement life, and it may be inappropriate to raise the age during a period of rapidly rising unemployment such as that expected in 1993- 94. The government indicated in report discussions that such measures are under consideration. 3.86 We recommend raising the retirement age by six months each year beginning in January 1996 until pensionable ages reach age sixty-five for men (in 2005) and for women (in 2015). The decision should be made and announced in 1993. The Political Economy of Entitlements: Assurances for Individuals 3.87 A system of social protection provides security both to those who receive benefits and to those who feel that the benefits are available, if and when required. For this reason, such schemes are often regarded as social insurance. A right to a pension is like a right to income from property and, for those who expect to cease working, such a right is enormously important. A desire for an implicit guarantee of future benefits is likely to become increasingly important as the population ages and as the proportion of the national income that needs to be devoted to supporting the non-working population increases. It is in the interest of beneficiaries (who remain a minority of voters) to establish claims to GNP that are as independent of the political process as possible. An important distinction in most market economies is between entitlement programs and assistance programs. As noted, this already exists to some extent in Russia. The existence of an earmarked tax on prospective beneficiaries--which for practical purposes means a payroll tax--is an important step toward establishing a particular benefit as an entitlement. 3.88 The difference between insurance and assistance is more a difference of political perception than of economic or legal logic. There may be no difference in economic impact between a payment to an individual that is regarded as a charity by beneficent taxpayers and one that is considered a just return on prior contributions, but these can be perceived quite differently from a political standpoint. It may be argued that the concept of a pension as being earned depends on the degree to which pensions represent appropriate amnuities based on the value of prior contributions. Even where, as with the Russian Pension Fund, there has been no individualization of contributions, the very earmarking of the 74 payroll tax provides ammunition for proponents of pensioners. Accepting that the Pension Fund may indeed be larger than needed, some have argued that it should be reduced in size rather than used for other purposes such as health. In other countries, the concept of a pension as having been earned can be strong, even if pension payments far exceed any realistic return on prior contributions--as in the United States. Similarly, prior contributions to an employment fund make unemployment benefits an entitlement, irrespective of the actuarial correctness of the payment. 3.89 This leads to two recommendations: * Increase employee participation. The real onus of social insurance contributions is on the employee3, irrespective of whether the statutory liability falls on individual or employer. 16 However, there may be an important political distinction. The Russian system currently attributes only 1 percent from a total payroll tax of 39 percent to an employee contribution, and no records are kept that can identify the amount contributed on behalf of any individual. Attributing part of this tax to an identifiable employee would strengthen the notion of pensions as having been earned, which shculd increase the sense of security of those who depend on them. The knowledge that pensions are paid for by individual contributors, rather than solely by employers, may moderate trade union and political demands for increased benefits. With respect to pensions, without increasing payroll taxes for the purpose, one-third to one- half of the pension contribution should be identified as a participant contribution, and this should be reflected in individual pay slips. Similarly, half of the recent (1 percent) increase in the payroll tax contributions to the Employment Fund should be designated as a worker contribution. 3 Identify individual contributions. Designating certain contributions as nominally made by participants is only the first step needed to establish the concept of individually earned rights. In the market economy emerging in Russia, contributory entitlement programs will have to be compatible with the needs of an efficient labor market and a wide range of life histories and personal preferences. Many individuals will change employers several times; others will stop working temporarily, some will become periodically self-employed, many will change their oblast of residence, rnd some will retire to a place in which they have never worked. While some of this information can be reconstructed from personal workbooks, these are vulnerable to loss and forgery. The current system utilizes only relatively recent information on earnings, although work histories are reviewed. However, using recent earnings as a determinant of pensions could have a distorting effect on pension wj/ The supply of labor as a whole is assumed to be relatively inelastic (and in some cases even backward bendlng) to wage levels, especially in economies that offer little opportunity for part-time work. The amount of labor demanded by profit-minded employers is likely to be affected by the cost of labor. It is therefore reasonable to assume that the cost of a payroll tax is very largely borne by the labor force, whether or not it is explicitly attributed to workers. 75 levels, since it is in the interest of the individual to maximize these earnings even at the expense of earnings in earlier years. The dangers are worse where managerial interest in either profits or the sizi of pension payments is small. Full lifetime work and earning histories will be important in any contributory scheme. 3.90 The future Russian social security system will therefore require individual lifetime identification nationwide, at least for determining pension entitlement. This will take years and extensive computerization to be fully implemented. Compared with other countries that earlier established the basic design of their social security systems, Russia has the advantage of access to computer capabilities from the outset. Data need to be centralized independently of financial arrangements or other administrative considerations. It is important that the new computer system incorporate a centrally designed software package that will be utilized by all Ministry offices. Although the information for cash benefits other than pensAons is less demanding of an individual's work history, wherever there is a requirement for a contribution record there needs to be a national data base, and there are clear economies of scale if all entitlement schemes can draw on it. In contrast, social assistance schemes require only current information that can be obtained locally. 3.91 We recommend that the basic design of a system of individual lifetime records of contributions and work histories be developed as soon as possible so that intermediate decisiors about computerization will facilitate rather than impede its attainment. The Role of the Pension Trust Fund 3.92 There should contlnue to be an extrabudgetary fund for pensions. To distinguish this from the present cone.ept of the Pension Fund, let us call this the Pension Trust Fund (PTF). The principles under which the PTF should be managed are discussed in this section. 3.93 An important question lies in the long-term investments of Pension Fund surpluses. Whether the Pension Trust Fund should acquire assets, such as long- term government bonds, or even become an investing institution is a complex issue. Funding would add strength to the sense of entitlement and would also remove the determination of benefit rates from the political arena. Some models of funded schemes envisage these as mandatory savings schemes, with individuals choosing from a competitive set of privately managed funds. The proportion of the GNP required to support retired non-workers will increase as the population ages. It is in the interest of each generation that pensions be protected by some kind of fund, so that their pension rights are better entrenched. 76 Figure 3.1 Pens i on Reform Change In the Number of Pensioners 50 . so~~~~~~~~er 40 30 0. - 20 10 195 97 1999 200i1200312005 2007 2092120305 20'17219 199 1998 2000 2002 2004 2008 2008 201 2022142 2018 202 years 3.94 In many ways, long-term investments could be portrayed as an attractive option. The Pension Fund is running a surplus, and the development of institutional investors could help the process of capital market development. Against this, however, are the extraordinary risks connected with investing in contemporary Russia and the generally poor record of state-run pension funds as investors.17 It is sobering to observe that most pay-as-you-go schemes started as funded programs which that accumulated investment funds. The absence of experienced investment managers makes it uncertain that institutional investors with long-term objectives would emerge from active funding. Moreover, in the present fiscal situation the additional expenditures entailed by investing pension fund surpluses will add to inflation. Nevertheless, the possibility of using privatizable assets to build up actively managed pension funds warrants ,U The poor performance of pension investment is attributed to two factors: high inflation rates and poor investment policies. An international survey concluded that state-run funded schames 'offer powerful evidence that this option may only invite squandering capital funds in wasteful, low-yield irvestments ishichl should give pause to anyone proposing similar accunulations elsewhere" ( Jean-Jacques Rosa, 1982, p. 2,121.). A study of pension finance in Latin America found that, of eight representative countries, between 1980 and 1987 only three had positive average annual real yields on investment. "The rest had negative yields as low as -21 percent annually, hence decapitalizing the reserves" (Carmelo Masa-Lago 1990). 77 additional study. Even if such an approach is adopted, actual pensions should not be made entirely dependent on fund performance, at least until the economy stabilizes and the profitability of the funds is established. 3.95 As an identifiable fund, even if its only assets are short-term government debt, the PTF requires management. Handing it entirely over to the Ministry of Finance would vitiate its fund nature. This implies that there should be some sort of joint control. We suggest that the PTF be managed by trustees representing the executive and the legislative branches of government. The trustees of the PTF would make recommendations about levels of benefits. The Ministries of Finance and of Social Protection and the Parliament should all be represented as well as perhaps others such as trade unions. This could begin to ease the problems of governance that characterized 1992 and most of 1993. 3.96 In the long run, the PTF might become at least partially funded, but it is clear that for the time being an actuarially based PTF operating on a pay-as-you- go basis would require a smaller payroll tax than the Pension Fund now receives. What should be done with the surplus? Several proposals have been made, although neither the Parliament nor the government ministries appear to have produced actuarial-based projections of pension payments in the forthcoming months. Arguing that taxes are too high and that money taxed to finance pensions should be returned to the taxpayer if not needed, it was suggested that the payroll tax be reduced to about 25 percent. In 1992 Finance Ministry officials suggested that the reduction could be larger than 8 percent ot payroll (and proposed reallocating the proceeds of the payroll tax, rather than reducing it). Although it would in principle be desirable to reduce the payroll tax and rely more on the profits tax in the face of growing unemployment, existing fiscal constraints weigh against reducing the payroll tax. With any improvement of the fiscal situation, an early priority would be to reduce the payroll tax. 3.97 Within Russia, there have been discussion of proposals to decentralize the Pension Fund system. This would help avert a revolt by "surplus" regions that are increasingly resentful of high contribution rates. In its final months the U.S.S.R. Pund did not receive transfers from several republics, including the Russian Federation. Under an MSP proposal, the Pension Pund would serve as a coordinator of oblast funds, setting a national contribution rate to allow a reserve not exceeding three-months' benefit payments. It would continue to redistribute funds from surplus to deficit oblasts. Oblast pension funds, however, would have the authority to levy contributions above the national rate- - to be spent locally on social protection or to be invested in the capital market. It has been proposed that oblasts retain a proportion of any surplus collected, rather than transfer the total surplus to Moscow, as at present. Some oblast pension funds (Altai and Chelyabinsk, for example) are already engaged in subnational social assistance programs, using the fines and penalties collected. Administratively, it might be unwise to expand the Pension Fund's authority through financing of social assistance or by developing Pension Fund expertise to monitor local programs. Such expanded authority could develop into efforts that run parallel to oblast and raion offices of social protection and thus cause administrative confusion and waste. 3.98 We strongly recommend that the Pension Trust Fund remain centralized. Although Russia has become a truly federal country and is increasingly devolvint 78 responsibilities to subnational government, the case for continued centralization of pension administration is overwhelming. Even with limited mobility in the Soviet period, many individuals found ways to arrange their retirement near the Black Sea and the Baltic Coasts, which contributed to relatively elderly age structures in Ukraine and the Baltics. Further demographic imbalances among regions are likely as Russia becomes a market economy with a housing market that does not impede mobility. 3.99 It has been suggested that the surplus might be directed toward other social expenditures, for example by supplementing the Emplcyment Fund or being earmarked for health expenditures. In 1993 a decision was made to direct 3.6 percent of the payroll tax to Health Insurarnce Funds. There is not, however, any need to earmark the surplus funds. Money held in an inactive account at the Central Bank (or as government debt that would in any case have to be created) has the same macroeconomic effect as if the money had been paid directly to the Ministry of Finance. The most logical option would be for the PTF to continue to run surpluses and to finance some of the budget deficit by holding long-term government bonds. If government bonds are safeguarded at least moderately against inflation, they could serve as the basis for eventual partial funding. For the present, however, the bonds would not be marketable, and the PTF would not be an active participant in capital markets. We recommend that the PTF be allowed to continue to run surpluses, whizh could be invested in government bonds. The Role of the Private Sector in Social Protection 3.100 As the economy shifts from a centralized command system, social protection is likely to become increasingly pluralistic. One dimension lies in the delivery of social welfare programs. Another is the development of private pension schemes. Governments in the OECD have increasingly begun to rely on contracting out social services to private agencies.'8 The monopoly provision of state services can undermine the efforts of small private organizations to provide services. In Poland, private organizations functioned best in small communities suffering from a lack of state-provided services. While these arrangements can lead to efficiency gains in the provision of services, the experience of Central and Eastern Europe suggests that the idea of utilizing private sector services over the short term should be approached cautiously in Russia. 3.101 In this context, certain factors must be considered in developing the private delivery of social services. First, the financing and provision of such services must be coordinated among the raion, oblast, and federal levels. In Hungary, the central government was reluctant to bypass local agencies by contracting to private providers, especially since central payments to subnational and private agencies are determined on a uniform per capita basis. Local authorities may demand preferential access to funding. Allowing the local public authorities to play a major role in contracting out can help to alleviate such problems and is the course generally adopted by the OECD countries. Second, 11/ See "Social Welfare Services Delivered by the Private Sectors." Innovation and Dloament, November 1992 (Study Ser'.as: Occasional Papers, no. 10). 79 there must be mechanisms in place to ensure that minimum national standards are being followed by private agencies. Monitoring provisions in the contract should include cost accounting, standards of care, and consumer satisfaction. Breaches should be subject to strict financial sanctions. 3.102 Competitive pressures can provide incentives for private agencies to deliver benefits at relatively low cost. But such gains depend on a functioning market, where more than one provider must submit bids for a particular service. A legal system capable of allowing agencies to enforce contractual agreements is also necessary. With little private experience in the provision of social services, Russia cannot expect a privatization of service delivery to occur overnight. On the other hand, there are a wide range of options that countries pursue when employing the private sector, ranging from commercial agencies to non-profit or charitable organizations (such as the church>. 3.103 Nongovernmental organizations (NGOs) could play an important complementary role in Russia's efforts to develop a comprehensive system of social assistance. The energy and ingenuity of these organizations has already beea evidenced, even in the very difficult circumstances of recent years. Beyond recognizing the major contributions already being made by NGOs and their potential enhanced role, the government could help to create an environment :hat is more conducive to NGO activity by adopting the following measures: 3 Legislation should be adopted that clarifies the legal status of charity in Russia and accords tax advantages that appropriately recognize the humanitarian nature of their activities. * The legal and procedural barriers facing NGOs should be minimized and the registration process made as simple as possible. - At all levels of government, but particularly at the raion and oblast levels, authorities should actively consider the delegation (contracting out) to NGOs of some aspects of social assistance and the delivery of services to the needy. NGO involvement should not be confined to the implementation of projects and programs, but should extend to participation in the design activities. Women's NGOs, which have already begun to facilitate commercial activity through training and advice, are well placed to mount pilot training in fields in which women have expressed interest. * Public works for the unemployed could be operated or sponsored by NGOs. The Role of Private Pensions 3.104 It is likely that a system of private pensions will eventually emerge in Russia. Privately funded schemes complement public pension schemes in all OECD countries. These are supplemental, in that they replace only part of the state.pension, and are often optional, in that the individual can choose to remain entirely within the state scheme. Private pensions generally operate as low risk and lower tax bearing forms of saving, although the returns are not guaranteed. 80 3.105 Government interest in relying more upon private arrangements and less on public schemes is evident across most industrial countries for fiscal, demographic, and other reasons. In some high income countries, private pension schemes, encouraged by favorable tax treatment, have assumed a major role in overall old age security support. Examples include the Netherlands, where coverage reaches two-thirds of the private sector labor force, and Switzerland, where employer provided private pensions as mandatory coverage rose to 92 percent by 1987. However, in no country except France does more than one-third of the elderly receive private pensions. 3.106 In Russia, private pensions emerged on the reform agenda in December 1991, with the issue of a Presidential Decree on Non-State Pension Funds. This attempted to establish the framework for private pensions by allowing non-state entities to set up pension funds for individuals. Further draft decrees were circulated in late 1992 and early 1993 on occupational pensions, on regulation and inspection of non-state pension funds, and on taxation arrangements. 3.107 In recent months, private pension funds have come into existence in Russia through commercial banks and other entities, as a financial infrastructure has begun to develop. Several insurance companies are selling individual annuities and group plans. The Ministry of Labor has supported the Russian Fund of Pension Insurance, which is seeking enterprise and trade union interest. Some consulting companies, acting as investment enterprises, have set up pension funds. One estimate is that there are at least several hundred thousand employees already enrolled in various private plans. The premiums are paid by employers at rates of around 5 to 10 percent of the individual's wage. 3.108 The potential advantages of increasing the role of private pensions as one of several possible sources of retirement income can be summarized as follows: * Private schemes reduce the fiscal costs of supporting the aged, as costs are shifted to private arrangements; although there are additional budgetary implications in regulations and tax expenditures. * Funded pension schemes are believed to increase savings. Much more controversial is whether public pensions have any dampening effects on total (public plus private) savings. * Public schemes are seldom generous; private pensions could improve the adequacy of retirement income, and in particular, the replacement rates for high-earning workers. However, international experience suggests that private pension fund performance has not been good and that full protection against inflation is rare. 3.109 A variety of schemes fall under the heading of private pensions. Occupational pensions are generally mandatory for eligible employees, typically cover an entire industry, and are wholly or partly supported by employer contributions. When there is only one employer, there may be problems in the transfer and portability of rights. Personal pension plans may be offered through financial institutions, including life insurance companies. Because of 81 high marketing and administrative overhead costs, personal pension plans flourish primarily where favored by income tax provisions. Other savings instruments that deserve mention include postal savings and mutual funds and goveriment securities, all accessed through brokers or similar financial servicis. Defined- benefit schemes predominate among private pensions, typically based on the final salary and years of service. The supplementary nature of private pensions is evident in the modest average size of the benefits. 3.110 The most important conclusions that emerge from international experience with private pension financing concern regulatior. and the problem of inflation. Regulation of the pension market is required w.n at least two ways. The conduct of private pension companies requires regulation to protect consumer interests in an area too complex for individuals to protect themselves. For the same reason, consumer choice should be restricted to approved pension funds. Private sector assets carry a large inflation risk. Although internationally diversified pension funds can resist purely domestic inflation, the greater the common inflationary shock, the less able are funded pensions to resist it. In private pension plans, benefits often become eroded by inflation, although this often characterizes public schemes as well, where lags in benefit adjustments may (deliberately) be used to make fiscal savings. 3.111 Given uninsurable inflation risk, the state may give private pensions at least a partial guarantee.19 One possibility is to issue indexed government bonds (that is, bonds with a guaranteed positive real rate of return). Note, however, that although the underlying pension is organized on a funded basis, the inflation guarantee is pay-as-you-go. The inflation risk is even greater during times of fundamental structural change and in economies with no experience of financial markets. Extreme uncertainty about the prospects of pension portfolios in countries like Russia will persist until prices have been liberalized and stabilization has been effective. Thus, there is a strong case for providing at least a partial guarantee from the state. Otherwise, pensioners would face substantially more inflation risk than wage earners. In addition, the collapse of private pension schemes during the infancy of a market economy might damage the political viability of further reforms. 3.112 The management of private pension funds is difficult, as is the more general operation of financial markets. Both sets of activities require operational expertise and experience, as well as the capacity to construct an appropriate regulatory regime. Private pensions require planning and operational skills that do not yet exist on any scale in Russia. A current example of potential problems can be found in the United States. The Pension Benefit Guarantee Corporation insures the benefits of participants of private defined- benefit pension plans, financed by premiums paid by the plans.20 D1J In the United Ctuadom. for instance, approved private schemes are indeimified against inflation of more then 5S per year. ani The financial condition of the Pension Benefit Guarantee Corporation has recently been questioned in liSht of pension fund failures and has been compared with the Savings and Loan fiasco. However, there are several reasons to expect that the similarities are sonewhat superficial. 82 3.113 This list o. cautions should not be interpreted as arguing against private pensions, merelI against expecting them to do what they are not capable of doing. Private pensions have a useful medium- and long-term role. They appeal to individual responsibility, improve consumer choice, assist the development of capital markets, and may eventually reduce public welfare expenditure. Subject to a suitable regulatory regime, the phased introduction of private pensions is therefore a useful complement to the state scheme. In the short term, however, the need for state guarantees, which could severely reduce the potential budgetary savings of sw'tching to privato pensions, combined with the need for strict regulation, suggests that private pensions should be introduced gradually. Reforming Sick Pay 3.114 The discussion in Chapter II highlighted the problems of the present system of sick pay in the Russian Federation. These primarily revolve around the difficulties in the system of management and control and the inappropriateness of the monopoly role enjoyed by the former official trade union. It is clear that basic changes to the system are warranted. An important distinction should be drawn between sickness and maternity benefits, which are both currently managed by the Social Insurance Fund (SIF). It is recommended that maternity benefits continue to be available on the same terms to working women, to be financed from the budget and administered through the oblast Department of Social Protection. A distinction also should be drawn between sick pay and disability pensions, which may continue workers' income over more extended periods. Further absence due to accidents at work are excluded. This section focuses on sick pay, which represents the majority of expenditures made by the SIF. It presents the reform options available to the Russian government, in light of international experience, and an analysis of the incentive and equity effects of alternative approaches. 3.115 Payments to employees for days on which they do not work can be the responsibility of the government, the employee, or soae third party. Each route could imply different costs and different incentive effects. The main function is one of insurance, ensuring a stable receipt of income during the year. Broadly speaking, there are three issues to be addressed here: (1) the desirability of any government intervention, (2) the form of government intervention (for example, direct public provision versus mandated private provision) if it is deemed desirable, and (3) the design of sick pay schedules. Box 3.3 presents a summary of international experience, with respect to patterns of sick pay and absenteeism. 3.116 The Desirablllty of Government Intervention. As in the area of health insurance, market failure may require government intervention in sick pay. In addition, the (re)distributive goals of the government ma- be furthered by sick leave policy if other, more direct transfer instruments are not available, especially insofar as there is a correlation between poverty and sickness. 3.117 The presence of externalities to sick leave suggests that it will be underprovided in a free market. Society may be unwilling to allow individuals to fall into abject poverty if they become sick. Underprovision of sick leave entitlements partly derives from information constraints and adverse selection. 83 If different employees have different probabilities of needing time off, and if these probabilities are unobservable to the provider of the benefit, only high risk individuals will be willing to pay the initial cost-covering premium. The externality comes from the fact that if everyone pays the same premium--wnich they must if the insurer cannot distinguish between high and low risks--the low risks provide a subsidy to the high risks. However, this will lead to higher than expected costs, premium increases, and further dilution of the insurance pool. As a result, there may be no insurance provided (see Rothschild and Stiglitz 1976). 3.118 There is some evidence to suggest that enterprises with higher risk employees are less likely to offer sick leave benefits (see Box 3.3). Clearly, a higher average risk of having to take days off implies a higher premium for any given level of insurance, but it is not immediately clear why less insurance should be observed. One possibility is that enterprises with higher average risks also have a higher variance of risks, so the adverse selection problem noted above would be correlated with the size of expected losses. Another reason could be that high risk occupations yield compensation close to the minimum wage, in which case the wage cannot be reduced to pay the premium. However, this would only explain why insurance is not provided by the enterprise through a reduction in wages; the enterprise could still offer group insurance to its workers, or it could be bought through a third party (which would involve further adverse selection problems). Lack of insurance may be due to lack of foresight on the part of individuals (like in the merit good argument above). 3.119 The Form of Government Intervention: Mandate versus Public Provision. The government can intervene by mandating that employers provide compensation for workers during periods of absence or by providing sick pay publicly. The main argument in favor of mandates is that they have less distortionary costs associated with them than publicly provided benefits financed out of general revenue (see Summers 1988). A further strong practical argument for employers' directly bearing the cost of sick pay in the present economically depressed context is that it alleviates the risk of employer-employee collusion in resorting tO publicly provided sick pay to minimize enterprise costs, especially since termination involves additional costs (such as severance pay). 3.120 There are some problems with employer mandates for sick pay, however. As mentioned above, real wage rigidities could lead to unemployment rather than wage adjustment in the face of mandated benefits.21 This points to public provision only if the relevant taxes are not borne by minimum or low wage workers, which is unlikely to be the case where public provision is financed through a payroll tax (as in Russia at present). The existences of minimum wages in the first place suggests that the government's ability to target tax and expenditure policies is limited, such that financing mechanisms which exempt low wage earners would not be available. In practice, this problem probably applies equally to public provision and mandates. 3/ To the extent that high inflation makes reductions in the value of real wages easier, this problem may be less severe in Russia than elsewhere, at least during the transition. 84 3.121- Perhaps the most significant drawback of a mandate scheme is that different types of workers may impose varying additional costs on enterprises.U In particular, if an employer is required to pay sick leave benefits, he or she is likely to prefer applicants perceived to have a lower probability of becoming ill than others. To the extent that such information is observable, employment decisions will be biased. Examples include women of child-bearing age, individuals with known histories of illness, older workers, and so on. Public provision does not induce such a distortion in the employment dec Uion, because the employer does not bear a direct cost when sick leave is taken. In this sense, mandating benefits can hurt those who are supposed to be helred. 3.122 Finally, mandates generally give the government less redistributive scope than public provision, given the link between mandates and public provision. However, this is not a strong reason to avoid mandates in the case of sick pay. First, if the reason for government intervention is to provide a merit good to individuals, the underlying assumption is that the good is worth more to them than the price paid. A similar argument holds if the intervention is due to efficiency reasoDs (externalities). 3.123 Design of Slck-Pay Schedules. Once the form of government intervention has been decided, it is necessary to establish the details of the insurance mechanism. There are a number of incentive problems that could arise. Employees may take too much time off (adverse selection at the individual level), employers could claim more sick days from the government than were taken (this would happen if the enterprise was responsible for paying sick leave benefits and was reimbursed by the government, say, at the end of the year--which is adverse selection at the enterprise level), and employees may not take precautions against requiring sick leave (individual moral hazard). 3.124 The extent to which these incentive problems are important will determine the shape of the payment schedule--for example, whether there should be a waiting period of a few days, whether the replacement rate should be equal to or less than the wage, and whether there should be a limit to the number of sick days available. A formal analysis of this problem requires a model of individuals' preferences and needs for days off. Individuals differ in the responsiveness of their labor supply decisions with respect to payment incentives. Because individuals derive utility from leisure and disutility from work, in general we would expect that any compensation for absence would distort incentives. Individuals would take more days off than they needed, but it is then necessary to model the need aspect. It might be iL. terms of disutility of work, productivity on the job, or something else. R1/ A distinction can be dram betwn employers who expose their workers to significant health risks (and thereby derive a benefit), s ad those engeing individuals Who have non-employment-related risks of sickness (observable or not). The problem is to void discrimination against the latter. 85 DU 3.3. latetnational Expiertoe: Sick kay Practices and Ratet of Absenteeism Practices with respeet to siek pay vary across industrial countries. Thero may be distinctions In treatment among variou sroups of employees (private versus public, large versus small enterprisea, Low/minimum wage versus better paid full-time versua part-time. and so on). It ts Important to note how the schemes are fnanced (whether by the employer, or through the govermwnt budget) an well as the design of the scho" (whether there is full earnings replacemont, waiting period. and no on.) In the United Statos, for exaMple, abort-tern protection against earnings loss due to illness can take the form of stak leave, sickness and accident insurar'ee, or both. Sick leave normally providea 100l of the individuol's pay, whereas uickness and accident iusurance only replaces 50 to 671 percent (either sernings-based or flat amount), and involves a waiting period (one to seven days). Where beth schemes are available, coordination can take two forms: atarting insurance banefits after tick pay has been oxhausted, or paying benefits ooncurr4atly but reducing sick pay so that the total amount paid does not exceed normal, salary. Paid sick leave is more prevalent among white collar ooua,ations, whereas sickness and accident insurance is more common for blue collar wor*ers. In most cases (M)X) insurance contributions for full-tine employees are paid for by the employer. In general, full payment of salary up to a certain number of days per year in the event of sicknoss characterises the public sector, as in, for example Australia. In the United States in 1990 IMX o*- al employees of at-te and locaL 'gvarnment were covered by paid sick leave plans, which were generally provided without any waiting period. In the private sector, the market generates an uneven pattern of bentfit provision. Coverage varies aeoosa occupations and enterprise size. In the United States, in 1)91, 672 of all fall-time employees in medium and large establishments had sick leave^- ranging from 871 for professional and administrative workers, down to less than 502 for production and service occupations. In small enterprises, three-fifths of all workers were covered--the range Was froc three-quarters of professional employees down to leos than half of production workers. Part-time workers are unlikely to be covered in the absence of mandates or public provision--only 301 are entitled to saik pay In the United Sttats. 1 Governments in a number of industrial countries have enacted legislation that provides for sickness insurance for all employees, often from the firat day of sickness. In Denmark, under the reforms introd%ced in 1973, employers wore t. finance benefits during the first tive weeks. Tn some ':ountries, including Ireland and Spain, certain minmlau prior insurance coatributions are required. :All Swedieb rasidents are coveredly ompulsory health -nsurance and, after a waiting period, qualify fot ick:benetits. P Urivate {legay enforceable) arrngeents for sick leave, which are the subject of collectivsn aretements, may eo-ezist beside publica shemes that wuld come 1nto play in the absence of the former. Ya~ Australia, sick leave is-the sub4eot, of industrial awards, while workers outaide such agreements may apply to the government--for assistanoe in the ev2nt of sickness. The publicly provided benefits are paid on the. umo condAtions a the benefit (that is, flat rate and means-tested). Zn several countries, reforms over the ist decade have sought to reduce the public provision of sick ' auy. 1n th* iailty l98's, the tritieh government .u.tified its withdrawal on the basis of evidence of i- it;snsiwv vpi'vte psovision of eaIrnigs xeplacement during sickness, although it continues to prti ally finance sick leave through deductions o enterprises' tax liability. In the opposite direction, the Ge*fn gsvemlment aa relve d entetpri$eA of paying wges eor days of sick leawe by I assiDng the cost onto the Stae. i--here are srignif nt variatioas in rates of absenteeism across countries, sectors, - terprises*.and individuals. The anofal average nwher of day's taken ranges fron around twenty-one (Prence) end twenty-three (Sweden) to currently reported rates if eight days in Russia. While the doteramiants a' n" t exactlk clear,-certai# risk factars haem bean identified as increasing the lSkelihood that. an individual. wdll be bhent from work. )orkers who are older--especially over fifty yearse-nd wumen tend to be away meor often. Hesith status obviously plays a role. There are varetiona aliross occupatons; workers in industvy and construction are absent more often, deseite the fact tbat thert arc lees likey to enjoy full earnings replacement end tend to be more strictly aonitored. thete appears to be an inverse%correlation between level of skill and pay, ari absenteeism. A 4ietinctl.on abould be drsm- between Prolonged ebsences (which constitute the overwhelming bulk of ,.he total' ind frequent or repeated:absences (which are also important, but less so than the first). Whet is theh ipact of the prevision of earnings replac_emet on labor supply? U.S. research suggests hat *ligibility tot sick pay increases wotkers' absenteeism rates. Similar findings emerged 'in Norway, for eq le, when legislation in 197B.providing for fulL pay from the first day of sickness appeared to lead to a substantial incxese in sabenteee$m. This is consistent with what economic theory ind intuition would iead us to expect: any earnings replacement would distort incentives. -thoer is nonetheless a trbust efficiency argument that ftaors pubLic intervention (see pfragrapbh P. 1t5-8.117). 86 3.125 Without going into a formal analysis of the problem, it is evident that certain measures are available to combat different adverse effects. First, the need for costly supervision of short periods of sickness can be avoided by the imposition of a waiting period. Second, the incentive for malingering is curtailed via a cut-off point at which sick pay entitlements cease. Empirical evidence suggests that the bulk of total absences from work are accounted for by a relatively small number of employees who take long periods off work. 3.126 To summarize, it has been shown that there is a robust efficiency argument for public intervention in the provision of sick pay. While employer mandates have tmportant economic and practical advantages over public provision, the main risk is tnat groups who are at high risk of illness will be discriminated against in hiring decisions. Bias in hiring decisions against those who are perceived to be expensive to Insure may lead to labor market segmentation to the detriment of individuals who are likely to be already disadvantaged. Whatever approach is adopted, it is difficult to design a scheme such that the sick would stay home and the fit go to work. It may well be appropriate to incorporate features, such as waiting periods of up to two days for the employee, to discourage cheating. Reforming the Administration of Cash Benefits 3.127 The centralization of an information base and/or a common pool of financing does not necessarily entail a single centralized administrative structure. But the number of different agencies involved in the administration of various parts of the cash benefit ..ystem having both distinct and overlapping responsibilities has already given rise to the need to design the system more efficiently and to present alternative proposals for reform. There has been some tendency to assume that the solution to a lack of coordination is vertical integration, but economies of scale may well be better achieved if several cash benefits share a common administrative mechanism rather than having all administrative functions for any one benefit being performed by one organization. Revenue Collection 3.128 At present, revenues for the Pension Fund, the Employment Fund and, the Social Insurance Fund are separately collected and monitored, in complete independence from the activities of the State Tax Service. Each of the agencies has established its own mechanisms for identifying prospective contributors, the recording of receipts and the checking of these against anticipated resources, the investigation and detection of non-compliance, and the pursuance of payment. This has inevitably resulted in a duplication of resources, often to very little gain. The Pension Fund is the most active in enforcing compliance by imposing penalties for late payments and understated payrolls. 3.129 However, if Pension Fund inspectors discover that an enterprise has been understating its wage bill, this information is not automatically passed on to the Employment Fund. In addition, employers face an additional administrative burden of making several different payments to the various agencies, a burden that will become more significant if and #hen a system of individualized contributions is introduced. The problem of enforcing compliance is likely to 87 worsen as the role of large and state-owned enterprises becomes relatively less significant. In addition, as enterprises become more profit-oriented and as new businesses (many of them small) are created, there will be pressure to rationalize and reduce the administrative burdens placed on them. 3.130 The introduction of formal mechanisms to require the collection agency of each Fund to report suspected non-compliance to the others would be an improvement, but not an adequate solution to the problem. Such mechanisms would reduce, but not eliminate, the duplication of activities among agencies. A possible additional step would be to nominate, on a case-by-case basis, one of the existing collect'-on agencies to pursue non-compliance. In practice, this is likely to imply the use of the Pension Fund, which has proved most efficient in these matters. Another possibility would be to create one collection agency to service all the funds, with responsibility for detecting non-compliance and pursuing payment on behalf of the different funds. This could significantly reduce the current level of duplication of resources. The funds would then concentrate on accounting for--and distributing the monies collected among--the payment agencies. 3.131 Rather than create another agency, the responsibility for the collection of contributions could be given to the State Taxation Service. Such a system could only work with adequate computer support to enable detailed exchanges of information about contributions. A practical point against such a move is that the PF is reportedly more efficient than the STS in collecting revenues. The mission was not able to explore this option. Each of the aforementioned options would strengthen, to some extent, the institutional capacity for dealing with revenue collection in a more efficient way and in the long run might enable a reduction in payroll charges. The Integration of Collection and Payment Activities for Pensions 3.132 A more radical approach concerns whether, in the case of pensions, the organization of collection and payment authorities should be drawn together. At present the Pension Fund offices at oblast and raion levels have no institutional links with their Social Protection counterparts. There are moves, however, to alter these arrangements so that the collection, calculation, and payment functions would be integrated into one agency. The most tangible step to date is a scheme running in Moscow oblast, at the direction of the Parliament, that places the registration, calculation, and payment of pensions in the oblast under the direct control of the director of the Pension Fund (see Box 3.4). In 1992 it appeared that this arrangement was regarded by the Pension Fund and the Parliament much more as a pilot (for eventual national replication) than as an experiment. 3.133 In reviewing these proposals, two questions must be considered: (1) are the functions of the Pension Fund and Social Protection offices currently, or likely to become, so inter-linked that coordination under the control of one management body is required; and (2) are there substantial cost savings to be made by integration? 3.134 Until October 1993, interaction between the Pension Fund and the Social Protection agencies at subnational levels was limited to the periodic 88 billing of the Fund for the monies to pay benefit recipients. The introduction of individualized contributions would, however, have a major impact on working relationships between the agencies, because those calculating entitlement must be able (either through computerized links with the collection agency or other means) to establish whether the applicant had an adequate contribution record. The process of establishing contribution records on an individual basis will take several years, but even when this occurs, the case for integration would still be open. In Great Britain, for example, a conscious decision has been made to separate the responsibility for collection and payment functions between two agencies, albeit under the control of the same government department. Such a delineation in organizational terms is thought to provide a clearer management focus and to ensure that one activity does not gain undue precedence over the other. 3.135 However, Russia's sheer size and the relatively small numbers of staff located within each of the local agencies point toward integration. In Altai Krai and Chelyabinsk, for example, the average number of Pension Fund staff in each raion typically amounted to only two or three. Even in Moscow raions, the average was only eleven. Although a larger number of Social Protection staff work on pensions in raions, it is still common for these to be in single figures also. Such small units, sometimes several hundred kilometers from the oblast office and with limited transport and communications links, clearly raise management issues relating to control, support, and flexibility. Maximum efficiency in the use of administrative resources very often comes with scale. Greater flexibility can increase the ability to respond quickly to crises and the longer-term interchangeability of staff between functions (thus improving, among other things, career development prospects and staff motivation). 3.136 The present arrangements, which separate pension claim activities between raion and oblast offices, are ineflicient and result in average clearance times of about six weeks, even after all the information has been received from the applicant and/or enterprise. These problems are well recognized and further decentralization is planned once the computer support is available to enable raions to deal with mass recalculations and payments. As a result, it is anticipated that oblast offices will be downsized, with staff from these offices being redeployed in the raions. 3.137 Some caution is required in this respect: there is a tendency to assume that all work must be carried out locally. Decentralization should not be seen as simply dividing oblast posts proportionately between raions. Only work that requires face-to-face contact with claimants and employers needs to be location-specific. Processing can be carried out anywhere, and it may be more cost-effective to carry out all the processing work on applications and recalculations in remote processing centers, which could enjoy economies of scale. The computer systems will require special security arrangements to prevent abuse. In addition, support is needed to cope with the general maintenance of information technology and also with specific problems. Such arrangements are more difficult to establish in small, remote management units. 3.138 With respect to payment mechanisms, as far as possible, forms of payment other than cash should be used, including more extensive promotion of payment via savings bank accounts and/or some form of money order that is 89 b&L.~ ilot ensios Sohiteo HIMow Oblant !n Julyr 199Z the Sup* Soviet, *Missioned e piLot oSawola in X osewalasti dxswit tosether the operational esponsJbilitias for peios under one enaent fninu. As a rS ult, al staff-'i&ebher Collecting, atributina to th tensionk Fod or caeulat;ig ed pyin berefits to the £.6 lopsioners in the oblast-n*e tVort t4 the local 4$Ztor of the Pension FArd. Ihe new Consolidated pension Organization continues to calculate and a Other ash bnmefits, a% ob as ftmily allowanges and tbo social pension. tesponalbility for seoeal assistance programs has bem retained by the residual sociaL proteotion deprtment of the Kisttry of Soolal Protection, New aecouacdation hs been secuted in each of the sisty-five rains sO that the 3,000-plus pensions staff van be cm-located in new pension departaiet offices. While clear do _0ation lines have bee retained between the staff of the previously separate org,aisaticna, including separate omnagsemet chains below ditector level, the Lonrtezn plan is to encOurXge iuterohangebility. As a result of the availibiUty of Pension Fund resources. beneftt admixistration staffing has been incressed, more computer assistance introduced, and standards of aooamedtion improved. A.number of other iitiatives are being tried s part of the pilot sexercise lhese include the oreation of a -bat, ocatrotled by the Pention Fnd, with branchs ign ech of the maw penions departmts. The Pensio F1und intends to depoalt the penalties obtained znD nou-cnmplying enterprises Into this bak, wih was due to etaxt operatig in Jauwary 1993, and then to make loans OQ oa;Xanie" at competit i Xates. Tha ovtebon of this now financil Institution was apprved by the Central Bank, despito oWosition frm othsr banks fearful of the effect of a removal of an estimated R S billio fran coircUslaon. Theba fd is also setting. up a oepayent delv ei, aatefrm the post office. eogotations with the nterrises that would undertake *te task are'likely to lead to charges of 32 or leis, thus undercutting the rats levied bI the post office.. encashable at the post office. Home delivery should be phased out, except possibly for a minority of housebound clients. Even here, however, it should be possible to view cash deliveries as the exception rather than the rule, if proper arrangements are established to enable friends and/or relatives to collect payments. While the reduction of revenue from pension delivery may have a marked adverse financial impact on the postal service, this is not a reason for resisting the change. 3.139 Potential gains in administrative efficiency present the most compelling argument for integration. A properly conducted pilot is required in representative regions, followed by very careful evaluation. The degree to which integration appears to (1) reduce costs, (2) improve flexibility, (3) streamline management functions, and (4) generally improve service levels and overall performance should be set against the risk that either the collection or payment functions are being afforded lower priority. Key measures of success must include shorter clearance times for pension applications and recalculations, improved revenue compliance rates, and reduced overall staffing levels pre- and post-integration and the relative numbers of staff at each grade. The results of such an exercise should provide a good basis for evaluating integration proposals. In particular, this should inform decisions about the nature of integration (that is, whether collection and calculation functions should be simply different duties performed by staff in one integrated office, or whether there is a need to maintain separate chains of command and career structures within each function). The findings should also help minimize the amount of disruption faced by the Pension Fund and subnational departments of social protection during any transitional period. 90 3.140 If the funds are to be collected by one agency or by the State Tax Service, and if management at the center is by a set of trustees, and if collection and payment activities at a local level are to be linked, the issue of whether the former Pension Fund takes over the work of the Department of Social Protection or the other way around seems unimportant. One possibility is to establish an independent Federal Pension Agency (FPA) to oversee the newly integrated collection and payment agencies operating at a subnational level. The possibility of subcontracting its collection responsibilities to the State Taxation Service requires further study, but in any event responsibility for collecting and enforcing payroll taxes needs to be unified. If payroll contributions are collected separately from other enterprise taxes, they would be paid into a consolidated Insurance Fund from which all entitlements would be paid. 3.141 The FPA could be set up under one chief executive, as is the FES, or under a board of trustees. In either case, appointments would be made with the agreement of, inter alia, the Ministry of Social Protection and the Parliament, so that the Agency has the confidence and support needed to develop a more streamlined and effective administrative system. Since it is unrealistic to expect agreement on the appointment of one individual, a board of representative trustees would be more feasible. This approach has the added advantage of direct client representation through board membership. 3.142 The links between entitlement to the unemployment benefit and the actively seeking work test suggest that unemployment benefit payment functions remain separate from the FPA.23 Recipients of unemployment benefits should have an incentive j make regular visits to the Employment Office, which, in turn, could ensure that they are still actively seeking work (and of course have not found a job). The funding for the benefit could be provided a consolidated Federal Insurance Fund, while the administration of the unemployment benefit could remain with the Federal Employment Service. Strenethening Administrative Capacity 3.143 The administrative system of social protection is burdened by a heavy workload that is expected to grow over time. While short-term measures to address this include maintaining or raising staffing levels, the Ministry of Social Protection will have to work toward developing administrative arrangements capable of processing and distributing claims inexpensively and efficiently. 3.144 Specialists play a key role in effective institutional development. Designers of information systems can assist with data base design and management. Trained fieldworkers and household surveys can help monitor the effects of benefit disbursements on targeted populations. Given the expected number of cases that will have to be administered and the current reliance on non- computerized assessment records such as workbooks, the objective should be to design a relatively simple administrative system with a minimum of paperwork. 23/ This is the situation in most countries. In Ireland, however, the registration and payment of unemployment benefits are handled through one public social welfare agency in cooperation with the post office, quite separate from the employment offices run by the Irish Training and mployment Authority. 91 3.145 Administrative capacity is not solely a function of designing a rational institutional framework for the Ministry of Social Protection, though this is important. It is also necessary to establish a framework of legislative measures that allows the institutions to quickly respond to changes in regulations and budgetary allocations. It will be important to implement measures that are enforceable, but not overly detailed, so that administrative staff have some degree of flexibility in delivering benefits (and minimizing paperwork). 3.146 Strengthening institutions requires the establishment, of clear lines of authority, within and between agencies. Within agencies, personnel policies should be drafted and adhered to. No country has been able to eliminate the problem of institutional corruption, and the Russian government has acknowledged that this is a widespread problem. In general, it can best be minimized through the carrot and stick approach- -a reasonable promotion and renumeration system and the tangible threat of sanctions if corrupt practices are uncovered. Regular and detailed audits should also be implemented. Information about departmental activities should be made accessible to the public, and annual reports should allow for outside review. A system of regular legislative oversight should also be adopted. Between agencies, there should be clearly delineated lines of budgetary and administrative authority. Central authorities should be empowered to enforce directives within their competence through incentive mechanisms such as matching grants and appointments. 3.147 Computerization will substantially strengthen the subnational administrative capacity of the Department of Social Protection over the medium to long term. A pilot program covering 11 percent of all Russian pensioners proposed under the Employment Services and Social Protection project should establish the framework for a centralized computer network linking the MSP with the oblasts and raions. The system should eventually allow different departments to exchange essential information about beneficiaries. Such measures would greatly facilitate means assessment by the MSP. Implementing the system, given the lack of trained network administrators and the outmoded telecommunications network, will take some time. It is recommended in the interim that local pension offices resist the temptation to purchase computers until they are linked to the central network, but they should ensure that they have the necessary number of workers to process applications (see paragraph 3.137). According to one estimate, raising administrative costs by 1 percent will buy twice as many staff workers in Russia as in the United Kingdom (Marshall and Thompson 1993). 3.148 Staff training is an important component of overall administrative reform. In this case, Roskadry (the Governmental Main Directorate of Staff Training for Public Service) can play a key role in helping to improve the staffing of the branch ministries, which should provide their own focused in- house training to supplement the work of Roskadry.2' Such training can focus on the full spectrum of services that a modern social service administration should provide, including procedures used to administer benefits, computer systems used for means testing and record keeping, and trouble-shooting when 4/Ses World Banlk, "Maket-Oriented Trainin8: Project Identification Report," Human Resources Division, Europe and Central Asia Department III, 1993. 92 things go awry. Higher level administrators should be provided with opportunities to obtain training in financial management and accounting, for example. Such training can be enhanced by attempts to instill a sense of administrative professionalism among workers, and should be complemented by a system that encourages merit-based bonuses and promotions. The training requirements would imply a relatively low level of investment. 93 CHAPTER IV: Al ANALYSIS OF RECENT LABOR MARKET DEVELOPMENTS 4.1 For any country, access to decent income earning opportunities is the best guarantee of social protection. For Russia, the movement toward a more efficient labor market is, in addition, a critical aspect of successful transition to a market-based economy. The transition will be a difficult one, even though labor allocation was the closest approximation under the Soviet system to a market process. Beyond the liberalization of wage and employment decisions, an effectively functioning labor market requires an institutional structure - including competitive markets, private property, and labor mobility- -that depends on longer term structural reforms. In the short term, the greatest challenge of labor policy will be to handle the likely scale and rate of increase .n unemployment. 4.2 Under the Soviet system, money wages were low, wage dispersion was slight and bore little relation to productivity, and the returns to human capital investment were relatively low. Overstaffing was extensive and widespread (estimated to be on the order of 20 percent prior to the reforms), and labor productivity was generally low. Ir addition, the role of enterprises is rather complex, often playing an important part in the provision of social services as well as the purchase of labor services (see Box 2.7 in Chapter II). Each of these aspects has important policy implications. 4.3 The relaxation of the tight Soviet era central controls on enterprises yielded stronger nominal wage claims and considerable labor turnover. It did not result in substantial employment losses, despite significant drops in output. After a sharp acceleration in unemployment in early 1992, the rate of increase decelerated. In the second and third quarters of 1993, total registered unemployment actually fell. Over 1991-92, GDP fell by some 30 percent and total state sector employment by 3 percent each year. The limited employment adjustment can be attributed not only to the implicit moral economy of the Socialist era but also to significant adjustments in the level of real wages. The compression imposed by the previous wage structure may be beginning to come apart, although at the end of 1992 the process was still somewhat muted. There is evidence of employment adjustment through extensive use of short-time work weeks and forced vacations, which amounts to marginal employment status without formal separation. 4.4 Overall, the structure of real wages and employment levels in the state sector of the economy continued to exhibit surprising stability, even into 1993. There is immense inertia in the system. Liberalization of wage and employment decisions has not produced an effectively functioning labor market, and labor productivity can be assumed to have fallen further. Various palliative measures to alleviate labor market strains, such as reduced working time, nonetheless have a useful role to play, particularly since the current profitability of enterprises is not a reliable indicator of future viability. 4.5 The implications for policy are reasonably clear. The employment overhang remains very large. Unemployment will rise significantly once the widespread subsidization of enterprises ends and a degree of stabilization is attained. 94 Very high levels of unemployment, such as those experienced in regions of Central and Eastern Europe, can create economic waste, social distress, and severe personal hardship. The policy challenge is to minimize major labor market disruptions as far as possible in line with macroeconomic goals. Several short- term measures are important. First is the provision of income support for the increasing numbers of unemployed, to avoid a rapid surge in poverty (see Chapter II). Second, a wage policy could complement tight monetary and fiscal policies to enable a downward revision of nominal wage claims and influence the unemployment-inflation trade-off (this is addressed in paragraphs 4.66ff). Third, to slow down the rise in unemployment and to help facilitate an orderly transition, employment subsidies, explicit and fixed in size and duration, could be extended to enterprises in regions where unemployment significantly exceeds the national average. Finally, the Federal Employment Service must strengthen its capacity to actively support the unemployed through counseling, retraining, and public works schemes. 4.6 As enterprises lose access to extensive subsidies and transfers and become more profit oriented (or cease operations), there will be a potentially serious disruption in the social services they formerly financed or delivered. Enterprises finance a broad range of social services, whose importance to the local community rises with the enterprises' dominance as a local employer. There are over one hundred non-military enterprises employing over 10,000 workers in Russia. There are likely many more enterprises that are the locally dominant employer (although the actual number of one-company towns is unknown). The withdrawal of these enterprises could result in the collapse of local community social infrastructure. As yet, there is little evidence to suggest that this has occurred on any wide scale, but substantial changes are likely. (The changing role of the Russian enterprise in the sphere of social protection is specifically addressed in Chapter II.) 4.7 In the longer term, structural reforms will make an important contribution to the effective functioning of labor markets. In a market-oriented economy, enterprises have incentives to minimize costs and to use labor efficiently. The supply response of enterprises is a key aspect of structural adjustment and will create new demands for labor. Market-determined wages should give workers appropriate incentives to work and acquire skills and to move between enterprises and localities. More generally, the economy should provide the incentives and inspire the confidence to establish new businesses and to take risks. What is needed to achieve this is the emergence of better functioning capital markets and a series of reforms to enhance labor mobility, such as the abolition of the proplska (resident permits) and the alleviation of constraints imposed by housing shortages (see Chapter V). 4.8 This chapter discusses the system of labor compersation, including the role of non-monetary benefits, and explores the evolution and structure of unemployment. (Annex 2 presents a fuller discussion of wage and employment developments and enterprise behavior and provides a simple bargaining model that captures the main features.) Most of the discussion concerns the state sector. Recent and increasingly significant changes in control and ownership are largely 95 ignored.I Information on the private sector is difficult to secure, and it is as yet unclear that title changes among state enterprises correspond to any meaningful change in behavior. 4.9 The chapter draws on three sources of data: the official data provided by the Russian Goskomstat, an International Labor Organization (IW) establishment survey, and a new survey of forty-one enterprises in the Moscow and Volga regions (referred to here as the Bank survey). The sample covered randomly selected enterprises in ten branches of the economy, including trade and services. Interviews were held over a period of two weeks in mid-November 1992. While the sample size is clearly small and by no means representative, it is a useful check on more aggregated data and provides a range of information unavailable elsewhere. Labor Compensation 4.10 Wages in the Soviet system were intended to square a circle: to fulfill basic efficiency criteria (in other words, to motivate workers) while avoiding the distributional effects likely to accompany an efficiency-based scheme of payments. The result was a centralized tariff system that tolerated only limited wage dispersion by skill, seniority, and other attributes. A set of regional coefficients was applied to all ranges of the payments structure to reflect differences in local prices and working conditions, and to channel labor to priority sectors.2 Roughly three-quarters of enterprises paid monetary supplements, which constituted on average over 40 percent of total earnings. 4.11 There was also an increasing use of bonuses. The size of the bonus fund was determined by central agencies, but bonuses were distributed at the discretion of the enterprise. By the mid-1980s, bonuses constituted 17 percent of the average wage and just under 21 percent in industry. The performance of work collectives was the dominant base for bonus calculations, with a further 15 to 20 percent of cases relating bonus payments more mechanically to the enterprise's gross revenues. Thus, while the Soviet wage system remained centrally regulated, payments practices became more explicitly linked to individual or enterprise-level performance. 4.12 Several features distinguished the structure of relative wages. One is that returns to skills or grade were relatively low. A 1991 survey of 500 enterprises in the Moscow and St.Petersburg regions found that managerial wages rarely exceeded 2.5 times the wages of unskilled workers (Standing 1992). The World Bank survey found that management wages in the third quarter of 1991 averaged barely 35 percent higher than those of professionals and skilled workers. A compressed structure of wage differentials over job functions continues to hold even after adjustments are made for non-wage monetary payments. The within-enterprise distribution for joint-stock and cooperative enterprises was similar to that for the dominant state sector. k/ Already, there is ewidence of widespread small-scale privatization in the retail sector (particularly in Moscow) and spontaneous privatization in state enterprises. Indeed. Aslund (1992) suggests that private sector employment in Russia may mow approach 20X of total employment. I/ A full description of the system is provided in Oxenatiemna (1990). See also Adam (1980). 96 4.13 Another feature of the relative wage structure was the privileged position of industrial workers. Priority sectors such as energy and, to some extent, the military had wage levels at the top of the ladder. Factoring in bonus payments increases this differentiation. Nonetheless, the spreads across branches were rather low. Indeed, variation in the average state sector wage across branches was no more than 25 percent in 1985 (Oxenstierna 1990). 4.14 Overall, greater decentralization of wage decisions and increasing private sector activity was associated with greater dispersion in wages across skill and other categories. This change resulted from the explicit objective of inducing greater worker effort and more active managerial behavior by associating earnings more closely to enterprise performance. While the degree of differentiation should not be exaggerated, there is some evidence to indicate that these shifts have begun to reverse the earlier (pre-1986) tendency to less wage dispersion. 4.15 The Enterprise Law of 1988 increased enterprise discretion in disposing of bonus payments and abolishing maximum wage ceilings. The law explicitly linked wages and bonuses to financial results--the amounts available being directly associated with gross revenues in the case of wages and with residual income in the case of bonuses. Enterprises were intended to be self-financing (samofinansirovanie), with payments to the budget predetermined and fixed over five-year periods, and were allowed to determine the size and structure of employment as well as wages. The law gave workers an active say in management, including the election of the head of the enterprise. The Imgortance of Non-Wage Benefits 4.16 It has long been realized that benefits provided through the enterprise were an important component of labor income, yet the value of these benefits was far from clear. The matter acquires particular importance in the light of suggestions that local authorities and other institutions should take over the social programs previously run by enterprises. Accordingly, it is useful to list and quantify the benefits provided, in part as a prelude to estimating the likely implications of enterprise divestment. 4.17 Aggregate Data. Aggregate data show that cash benefits paid by the enterprise account for at least 6.3 percent of total employee income.3 Regional data reveal fairly low dispersion rates for cash benefits as a share of income. When exogenously determined regional coefficients and supplements are taken into account, regional variation is low in both per capita wages and benefits. In other words, regional variations in total per capita wages are largely explained by wage coefficients. 4.18 There is no positive correlation between levels of wages and levels of benefits. Thus for industry, the share of benefits is at the national average despite markedly higher than average wages. If anything, there is a weak negative correlation between wages and benefits. Disaggregate data suggest that inter-branch variation in the share of benefits increased sharply between 1991 S/ Total income comprises wages alnd bonuses, plus cash assistance, dividends, enterprise-provided cash benefits, and cash benefits provided by the local authority. 97 and 1992. There is a general increase in the share of benefits in total per capita income, but no evident association between the relative wage level and the benefits share. 4.19 Survey Evidence. A list of the types of benefits provided by enterprises is given in Table 4.1.' A significant proportion of the labor force continues to have entitlements to child care, paid vacations, housing, and vacation homes. The focus is on discretionary benefits, rather than mandatory benefits--such as maternity allowances.5 It is clear that discretionary benefits are both varied and pervasive across enterprises of all size classes. Further, there is a positive association between enterprise size, as measured by employment, and range of benefits. The absence of housing programs among several of the largest enterprises may be unrepresentative. Bigger samples, such as the ILD survey of 500 enterprises, point to housing programs as an almost defining feature of larger enterprises (see Standing 1992). This is more likely to be the case for enterprises that are locally dominant employers and that have taken on the functions of local authorities. By contrast, the evidence suggests that the benefits provided by relatively smaller enterprises may not exceed those provided by local authorities. Indeed, in the Bank sample, enterprise expenditures at the local level for housing, child care, and health facilities were comparable to those of local authorities in only 10 to 30 percent of cases, depending on the function. There is thus a need to avoid a simplistic view of the relative significance of enterprise-level functions. 4.20 Such extensive provisions of benefits has ensured that industrial workers have continued to be relatively privileged. There has been little distinction in terms of access to facilities by category of industrial employee. There has been fair convergence in the range of benefits offered by industrial enterprises, except that temporary workers rarely have any entitlement to benefits.6 Finally, the surveys indicate that in both 1991 and 1992, the wage and social development funds typically increased relative to the production fund. 4.21 Cost of Benefits. There has been little information on the shares of benefits costs in enterprises' labor payments. The Bank survey is one of the first to provide a reasonably detailed picture on this score. The enterprises' wage bills (inclusive of the excess wage tax) account for roughly half of total labor costs. Adding bonuses raises the share to just over 60 perc--t. The remaining labor costs are distributed fairly evenly between the Pension, Social Insurance, and Employment funds (which are mandatory) and, most significantly, the social fund. Cash benefits account for about one-third of Social Fund allocations (6.3 percent of 20 percent). While there is considerable variance over enterprise size classes and branches in the respective shares and levels, in general there has been a surprising degree of stability. J/ Annex 2 provides a more detailed analysis of the survey results and benefits (Commander, Lieberman, and Yeintsov 1993). V MFndatory benefits are detailed in Chapter I. I/ See Standing (1992) for more details. 98 Table 4.1: Type and Availability of Benefits by Enterprise Size, November 1992 Enterprise Size by number of employees 80-350 351-700 701-900 901-1,500 >1,501 Sample Size 10 11 10 7 3 Housing (permanent) 2 5 8 5 0 Housing (temp) 1 3 4 3 0 Kindergarten 1 5 8 6 2 Land for dachas 3 7 7 6 3 Canteen (subsid) 4 6 8 7 2 Polyclinic access 2 2 5 5 2 Ccmcnity house 0 1 3 3 0 Fitness facilities 0 0 1 2 0 Sanatorium 0 6 4 4 1 Food store witb subadized prices 8 9 7 6 2 Housing rents assistance 0 0 0 1 0 Other form of housing help 0 6 4 4 0 Transport allowance 3 3 1 3 2 Child care allowance 10 11 9 7 3 Paid vacation 4 8 7 3 0 Pre-dismissal allowance 8 10 8 4 3 Sanatorium vouchers 8 10 9 6 3 Source World Bank survey 4.22 There has been relatively limited adjustment in benefits (see Box 4.1). Barely a quarter of the enterprises in the World Bank survey had suspended any benefits; half of those cases were housing-related programs. Sixty percent of enterprises with housing programs were intending to transfer their housing stock to local authorities.7 Questions about the relative priority of different benefits yielded mixed results. Housing benefits figured prominently, with an equal number of enterprises prepared to shed such benefits first or last. Employees were unambiguous in giving priority to retaining housing benefits. Housing benefits generally constitute the largest direct and implicit cost to enterprises (given rent controls) and the largest current and anticipated income component for households, given housing shortages and constraints on mobility. (The issue of enterprise housing and further survey results are explored in greater depth in Chapter V.) V The legal status of enterprise housing is quite unclear (see Chapter V). 99 - 44. * -efits, ad the Lage fntarprste: The Case of tsAtT SaI,4 ar the Centre). Aete end Uydrodynamioe lbwtitte. is a r veoUno eilo of a Uarge Soviet- ty"etarpriwet a loaolty dominant -aployer witb an extemsive benefit progsa tmd aonsiderable reliance on budgetary income. Situated in 2hoahy, 40 kilometer outaida Moscow, the entrprise concentrates an resarh and d&v elort in the field of aerodyamics. All W.5'uiu aircaft are tested and in part des igeA a tslM. -the oompaW's work toe. e eeds 10,000 in a city with a total population of under 90,00O d ew Otor 5ouroees of omplt.msnt 2he high degree of vertical and horisental integratiom nwoeseartly Implies a fairli heterogeneous workforce with, however, a cace of highly skilled aeienttfil sta£f and a biatory of relatively high wage payments. The lattes is no longer true ea Problems with input supplies tom other parts of the CiS, an evergy price shook (given the high energy intoosltw of Its aati#ity),-and initial doubt over the viability of the Russian aixoraft Industxy have undemined the entepriaoQ' ability to pay workers and taintain outstanding obligation, the repnso to ne8ative ehoe'jg to vurrent output in 1992 w e Limited labor retrenchment-- some 5X of the work forcea swe as ed--alo i use o.f inentives to induce further w. lUntary separationst A furthe 25% of the ofoe--ler unilled or ami-akiled staff--have been placed on. ina wee with tt CA0" reqirumsnt to attend work. Almot allU paced on nimnm w8e.r. oetis4 is- attributable both to disovilWnatlio and to the low status of womn's work. De featoj- the-nterprseprovides umploment betfits to around s quarter of the labor force. At the same tim, eleat wae differtiation h# ocaurod as indtvidual imite within the enterprise attempt to -aproprite orent itmooms StUOM partiuaxly wth respect to unit-specitic outside work contrats.W wIie it Is too eary .to conclude, it appears tihat manaers of viable units are attempting to induce 8os-se:frM > - PriVLtiatrion. Amon- other coneAquesc. ths esul-ts !i an OM wllingness to- support enterprSe ol-- -:n.it.:- - nterrise-provided onuefits are vry substantil, prin over as of toal coits in 199l. Tbey inOluie pvsion f hU enfs-a£ril.1sa (e-len rtes), 555aprten hose (an a- furtbh woder t . t * ureenad i flooramn s cqpleted in 192. land f g ctrt , 4t W inae, meia=et Qi? es, and ti4ewth. -ubsidtad fod ell e baef*it: the o* f" 11 i- l apvided Wr0ers withd- t food d }ia*or-t entriblic -gd.- oi ret I - , c . -,. x s, -';_1. X - ''' tenants may nt, ps et i e of th ong lund ra depetq an *tor#*bta roeabIiein s- -on _omeo e. O 4 es hav e Oe otat e benefits. schs p*:deiaa alon. supne. hut. the enepa lha greal a icoelera.ited.. housing oonstrui* ed- ivs .a priIty to housin, the pamIs largely *fuei.y .a -- a_iwlo 1,emau o `f was-- ..p enterprIse's tools). Gwel t e,. ,. ,i, r f t , f ,~.tai.. :Y : : . , :' Y - apat2 nt thA mo ie coprehensiv measide iwtof. larorie incomen fianctsing botnuhs tamets psocia, feund lhIo paymen ts, and srzuoial e inuapnce, othelays,eue to wa ro,vafer taes. e Tloeres ith fair dtn varibilto finethe distribuetionstrotf lab4oru incomeptyae o fe th principa aparonents oover time The nuban eto e wagetna s oareno diplayns. considgemrabet qurotsrl vadriatso ovn aseroi) a l entepisg~ ewn si elo thases tEtlhouigh itoh isnoreil sarleas hen mesrcured ofe thabyer Som ciald fucnt expenditrsaesmlrl akdb ut istrfagintaqatr variabltinthdistributrmion oflborinoe over the prr.Soincipald payments constitute over a quarter of labor income across the full Bank survey sample; that share is generally positively correlated with enterprise size. These numbers understate the value of non-cash benefits, since they are based only on current financial flows and ignore capital and depreciation costs. 100 Furthermore, there was a very sharp increase in social fund allocations by the largest enterprises in 1992. The social fund constituted over half of labor income for these enterprises in the third quarter of 1992. This appears to confirm the perception that some enterprises have avoided wage controls in part by increasing labor compensation through benefits. Indeed, in response to high monthly price changes, enterprises have sought to compensate workers not only by explicitly associating wages with price changes but also by increasing benefits that help maintain real levels of consumption.8 4.24 On the income side, there is considerable stability in the levels and shares of the different components. Wages accounted for roughly 50 percent of total labor income, up to 65 percent including bonus payments. Thus it seems that benefits, dissociated from more conventional cash-wage or effort-related incentives, comprise approximately 35 percent of labor irncome at current prices. 4.25 Discretionary non-wage benefits amount to about 20 percent of enterprises' labor costs, wages and bonuses for 60 percent, and mandatory payroll taxes for 20 percent. Overall, these relative shares have been fairly stable over time, although there is evidence that some enterprises have sought to avoid wage controls by shifting to higher benefits payments. Clearly, a more satisfactory measure would be achieved by attributing shadow prices to the benefits. Given the volatility of prices and the uncertainty about the levels at which key prices are likely to settle, this task is not attempted at this point. 4.26 These activities of enterprises should be put in perspective, however. At 35 to 40 percent of the wage bill, the total _.ost of all social services provided by the typical Russian enterprise is broadly comparable to the patterns observed in many industrial countries, where employer-provided benefits may include group life, health, and disability insurance, pension and retirement packages, subsidized child care, and so on. The ratio of non-wage to total labor costs averaged roughly 30 percent in OECD countries in the period 1966-81, with a range from 25 to 40 percent. In the United States in 1990, the cost of voluntary benefits amounted to 25 to 30 percent of private sector payrolls, while those mandated by law added approximately 10 percent. 4.27 The Wage Path since 1991. It is difficult to assess the recent path of real wages. The obvious problem of measurement under conditions of cash constraints and goods shortages is further complicated by cash shortages and liquidity constraints in 1992. Both drive a wedge between notional and actual claims, making identification of the true real wage almost impossible. In addition, forcing employees into increased short-time work and unpaid vacations would reduce actual nominal wage payments. These developments, paralleling the sharp fall in capacity utilization, obviously reduce effective wage payments. on the other hand, secondary work was omitted from aggregate wage income. The share of private and semi-private employment has apparently risen dramatically. A survey administered in mid-1993 with wide regional coverage indicates that around 22 percent worked in a semi-private entity as their principal job. 1/ When enterprises in the World Bank Lurvey in November 1992 were asked how they protected workers against price increases, 602 of respondents indit,ted an explicit attempt at wage indexation and 402 at indexation and enhanced social or benefits progrms. 101 4.28 At first glance, official wage data show sectoral real claims accelerating in 1991. If money illusion and goods shortages are ignored at the peak in December 1991, real wages were roughly 50 percent higher than in January 1991 and double the level of January 1988 when the Enterprise Law came into effect. However, cash constraints and goods shortages make comparison problematic over time. 4.29 The climb in real wages was followed by a precipitous fall over the first quarter of 1992, of more than 55 percent since late 1991. The decline is larger yet--around 75 percent--when measuring average industry sector product wages. There are some signs of slight upward drift again (see Figure 4.1). Even so, by May 1993 real wages for the state and industrial sector were about 20 percent below the level in the first half of 1991. 4.30 There is a high variation in nominal wages across regions. Even after wages are deflated by regional retail prices, there is still considerable dispersion and instability. While the regional wage rankings remain largely unchanged, the coefficient of variation remained high- -around 0.5 throughout 1992 and 1993. Differences in the path of regional real wages seem to be driven more by divergent nominal wage claims than by variations in regional inflation. Exploring the evolution of regional wages allows non-trivial differences in regional price levels and the system of regional wage coefficients applied to base wages to be accounted for. For institutional reasons, the regional data also appear to be more reliable. 4.31 The broad picture is that the contraction in real wages associated with the price shock of January 1992 has subsequently been largely reversed. This conclusion holds across most sectors, though current statistics do not adequately capture the private sector. The ILO survey reports real wage contractions in the private sector slightly lower than in the state sector. Official data for the private sector, cooperatives, and joint ventures report a wage fall over early 1991, exceeding that for state enterprises, and continuing until mid-1992. The situation is muddied, particularly given the rapid and ambiguous transfers of enterprise titles occurring over this period. Sectors with above average wage expansion in 1991 maintained that trend, while smaller wage increments characterized traditionally low wage sectors, particularly in the non-material branches. 4.32 Trends In Relative Wages by Industrial Sector. Relative sectoral wages to late 1993 reflect very little change since 1991, indicating the continuing significance of institutional features in the wage setting system. The Bank survey indicates that in more than 80 percent of cases wages remained administratively set rather than achieved through bargaining--an indication of a continuing de facto role for the tariff wage structure. 4.33 There is scattered evidence that wage differentials have widened in favor of groups of organized- -largely but not exclusively blue-collar- -workers and that this process has reflected more the exercise of respective bargaining powers rather than any explicit association of wages to output changes (Commander and Yemtsov 1992). Workers in the mining and energy sectors expanded their wage differentials both with respect to industry in general and also with respect to skilled and professional employees. Yet while certain groups of organized labor 102 have consolidated their role as wage leaders, this process is less clear elsewhere. The Bank survey reveals ambiguous evidence of a change in wage relativities. While over two-fifths of sample ent6 prises reported an increase in wage differentials, this proportion was exactly matched by enterprises reporting no change. The remaining 17 percent reported decreasing inequality in wages over 1992. For enterprises that did report an increase in wage differentials, most attributed the increase to market features rather than, say, to any explicit association to private sector wages or linkage to prices. Figure 4.1 Russia: Real State & Industrial Wages Jan 91 - Feb 93 Cdec 90 prfces) 480 460- 440- 420- 400 3so~~~ I dsr ae+SaeWg 340- 0 320- 300 280- 260- 240- 220 200- 140 thJal¾91d qrs of"1y I and I seop I n cls cn21 marge JtcI esp in rt ofi93s feb apr Jun aug oct dec91 feb apr jun aug oct dec92 feb C Industry Wage + state wage 4.34 For enterprises in the Bank survey, the actual evolution of wages between the third quarters of 1991 and 1992 shows close convergence in rates of increase across the main grade categories. The only apparent losers in relative terms were professional staff, such as engineers, but the shift in relativities is not large. However, it is clear that wage changes at the top of the grade structure (for the best paid managers) have been consistently higher, especially in the largest enterprises. Differentials within functional categories have also widened, again among top managers. Overall, this suggests that the compression imposed by the previous tariff wage structure is beginning to come apart, even if the process remained (as of November 1992) somewhat muted. 103 4.35 Wage Controls. By 1992, most wage controls had been jettisoned, both explicitly and implicitly, through reductions in the share of enterprise income transferred to the state budget. Nevertheless, departures from the earlier structure of wages and their relativities were reasonably contained. The nominal wage explosion was concentrated in the second half of 1991. In 1992, wages might have been intended as an anchor in the projected stabilization, since wage payments greater than four times the minimum wage were taxed at the same rate as profits (32 percent). For 1993, the tax rate was increased to 50 percent, and the enterprise income tax was reduced. Figure 4.2 presents estimates of warranted (that is, four times the minimum) and actual wage payments for each month from January to August 1992. The obvious conclusion is that the wage rule has been weak and unable to contain wage claims. In all months except January the actual economywide wage bill has exceeded the norm-given wage bill by a significant margin. The ratio of the actual to the norm averaged 1.6 over the first eight months of 1992. Figure 4.2 Russ i a Wage Bl I I aLncd Wage Norm January - August 1992 360 340- 320- 300- 280 260- 240I C A 220- Empl 200n 18 1200 140- 120 100 80~~~~~~~~~~~~~~~~~~~~~~~~~~~~~_ Jaen 92 fob nur apr- may Jukne July aug 92 1992 0 Actual + Norm Employmenmt The Pre-Reform Context 4.36 Labor force participation rates were high in all socialist regimes, in part because of legal restraints on non-working. This includes strikingly high levels of participation by women, possibly around 85 percent in 1987. The high participation of pensioners in the labor force--estimated at 30 to 35 percent in the early 1980s--is encouraged by the ease with which pension rights can be maintained and accumulated alongside full- or part-time work (see Chapter II). 104 4.37 The labor force is highly skilled. The Russian people can boast of high educational achievements, with mean years of schooling of 11.1, compared to 10.4 in Japan and 9.6 in Norway. Secondary completion rates are higher than those in the United States. Russia has one of the most extensive systems of education and training in the world, including over one hundred post-university training institutes. 4.38 The female labor force is highly educated: 47 percent of women have completed higher and secondary specialized or technical education, compared with 34 percent of men (Annex 1, Table A1.8). Professional women are concentrated in certain fields, including economics, law, education, and medicine, all traditionally considered low status occupations. Certain factors encourage women to seek white collar jobs. Women have higher qualifications for the same pay. Higher education is necessary for women to reach levels of pay comparable to those of less skilled male workers. Women have cited working hours that are compatible with family life and a pleasant working environment as important reasons for job preferences. 4.39 The occupational structure of industry was, and remains, rather particular. A heavy bias toward unskilled and skilled manual labor reflects the level of production technology and the extensive growth strategy adopted in earlier periods. The high degree of industrial concentration yielded large enterprises with high average employment. An emphasis on heavy industry and military production generated a pattern of labor allocation significantly different from that in market industrial economies. The services sector was rudimentarily developed; for example, the trade and communication branches accounted for less than 9 percent of total employment in 1985. Open unemployment was kept very low (although estimates before 1990 range widely). Frictional unemployment was variously put at between 0.6 percent and 2.7 percent of the Soviet labor force in the 1980s. The consistently higher number of vacancies relative to job seekers (pre-1991) and the apparent downward mobility of skilled workers point to major problems of mismatch, related largely to the educational and training system and the rules for compulsory placement of professionals.9 4.40 Turnover rates were reasonably high, averaging around 14 percent in industry between 1988 and 1991. The persistence of apparent shortages for skilled and production workers has promoted job turnover, as these workers chase relative wage adjustments. Buoyant demand for production workers can be traced to fixed factors or technology. It is striking that turnover in 1991 remained at high levels: in the Moscow and St.Petersburg regions over 15 percent in industry. Explicit measures to reduce turnover, such as provision of enterprise housing, contributed to rather localized patterns of turnover. Because enterprise-supplied housing granted implicit ownership rights to the occupants, even if they left the enterprise, workers could move easily between enterprises, but not between localities. ./ An extended discussion is provided in Omnstierna (1990. pp. 215ff). 105 Recent Changes in Employment 4.41 Downward employment adjustments can be observed since 1986-87, before any attempts at systemwide reform. Figure 4.3 shows that aggregate industrial employment in 1991 was nearly 8 percent below the peak in 1986, a trend that holds for the major branches except construction. This drop is generally related to the Enterprise Law and the wage-setting regime allowed under reform. Greater decentralization of wage bargaining, combined with an explicit tax on wages, created incentives for employment cuts and enhanced labor flows conditioned by relative wage effects. Involuntary labor shedding has been concentrated on non- production workers and, in particular, on women. Overall, the bulk of job losses have been concentrated in the state budget (non-material) sector, not in industry. This also explains the high share of female and clerical workers in total unemployment (see paragraphs 4.57, and 4.59). 4.42 Since 1991, however, the apparently limited loss in employment is striking, given the size of shocks to output. This holds for both the state sector as a whole and for industry. Total state sector employment fell by less than 3.5 percent between January 1991 and January 1992 and by a further 1.3 percent between January and August 1992. In the same period, GDP is estimated to have fallen by about 30 percent. The decline in employment in industry is even smaller: less than 3 percent and falling a further 0.5 percent between January and June 1992. Part-time work also appears to have declined over the course of 1992, possibly by around 5 to 6 percent in industry. 4.43 Enterprises may adjust to falling output demand in ways other than outright layoffs. A significant development--particularly since June 1992--has been an increase in the numbers of workers on short-time work and forced holidays. A Goskomstat survey of 3,000 industrial enterprises in August 1992 indicated that many enterprises have placed some of their labor force on unpaid leave or short- time work, and that this share had doubled since June 1992. About 13 percent of workers were on forced leave (typically for three months without pay) and 16 percent were on short work weeks. Almost one in four employees was subject to one or the other tactic. Other evidence suggests that in a significant number of state enterprises, management has effectively lost control over labor discipline,'0 a further negative shock to labor productivity. 4.44 The enterprise strategy of choosing unpaid leave or short-time work is consistent with a general reluctance to retrench. This strategy induces active job searching or secondary employment by effectively reducing wages and time spent in primary employment. Still, the Bank survey did not find that this was a wholly generalized process. There was evidence that use of involuntary leave and short-time work had increased over 1992, but that by the third quarter of 1992 under 35 percent of enterprises reported using forced leave and 17 percent short-time work. In summary, under 5 percent of the total sample labor force was on prolonged involuntary leave, nearly 90 percent of whom were men concentrated in machine building and light industry. A further 5 percent of workers were reported on short-time work, although 70 percent of these were from one ,IV Although this does not show up in additional claims for sick pay, a time hbno?ed method of shairking. 106 enterprise. Unfortunately, the data on short-time work and unpaid leave remain fragmentary. 4.45 Regional employment data show increasing variance in 1992. In the Northwest, including St. Petersburg and the military-dominated enclave of Kaliningrad, experienced employment contraction of about 5 percent. Decline in military production is likely to have been responsible for a significant share of employment loss in the Northwest. But in general, this association cannot be assumed. There have been no major job losses in other military-dominated regions, such as the Urals, despite anecdotal reports of huge drops in output. In Novosibirsk and Saratov, two oblasts where defense represented about 47 percent of total employment prior to 1991, no unusually strong downward pressure on employment has been observed. 4.46 While the story on the state side is one of gradual job loss and high rates of worker movements between enterprises, the picture for the private and cooperative sectors, where some job creation might be expected, is less clear. The widespread reclassification of titles that occurred in the last two years is responsible for some of the lack of clarity.'1 Official data for cooperatives show a 40 percent decline in employment between January and July 1992, while employment in private enterprise increased by about 10 percent. Even so, combined employment in these categories in mid-1992 probably constituted no more than 3 percent of total employment. And the relatively high share of part-time workers--16 percent compared with an economywide average of 4 percent--suggests that many of these enterprises are small. Nonetheless, it is likely that official data grossly underestimate private sector employment. 4.47 Employment Changes: Enterprise-Level Data. Enterprise-level data are a useful countercheck to official data on the state sector, which are subject to measurement error, in part because of the breakdown in traditional reporting and coverage. Employment decisions can be explored using the World Bank survey, as well as information from the ILO data set. The Bank survey data yield striking results: (1) high rates of turnover; (2) low levels of involuntary retrenchment across all enterprise sizes and branches; (3) considerable new hiring by enterprises, largely to replace workers who have left; and (4) consequently, a generally low level of job losses through 1992. 4.48 The acceleration in job separations in 1991 and 1992 does not reflect a major change in enterprise behavior. The lack of a robust association between employment changes and sales changes (and by implication, to output changes--see below) points to a generalized but weak process of labor releases and gross flows induced by a change in employment rules and wage setting practice. It is revealing that the continuing high rates of turnover remain dominated by voluntary separations. Involuntary retrenchments amounted to less than a quarter of total separations. A similar pattern emerges in the ILO data. The principal reason for involuntary separation remains employee conduct rather than current or predicted patterns of output or demand. Falling demand for enterprise output 11/ The IL survey in June reports 27Z of establishments classed as leasehold (552 state and 182 private), but the distinction apparently is not very meaningful in terms of economic behavior. The state sector may be shrinking rapidly (down 8S over their sample between September 1991 and June 1992), but primarily by means of title changes. 107 was cited in less than 3 percent of cases as a factor motivating employment cutbacks. Survey evidence thuw shows that voluntary separations have dominated, motivated by changes in relative wages. Evidence suggests that forced separations have been concentrated among clerical staff, unskilled manual workers, and, to a lesser extent, among temporary labor.U Figure 4.3 Russia: Employment by Sector 1980 - 1991 26 24 - 22- 20 19 2 14- E 12- 4- 2 1 1 I , _ I - I I I - I I I I 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990 1991 tI Industry + saric o constr A transport X trade 4.49 Although a clear majority of enterprises report excess employment overall (see paragraph 4.3), enterprises still complain of labor shortages among skilled manual workers. The dominance of voluntary separations is thus associated with the continued posting of vacancies as well as reported overstaffing. This pattern reflects the stability of enterprise objectives and reinforces the view that gross flows have been largely driven by institutional shocks rather than by a process of enterprise-level restructuring. Consequently, involuntary separations have been limited, considering the size of negative shocks to output. 4.50 One would expect an association between output and employment changes and domestic relative prices. A simple equilibrium framework would predict a I llote that tanporay labor unte" to less than 11 of total lUdutrial mmploymt in 1991. 108 positive association between the former, and relative prices as indicative of a shift in the demand curve. A negative association would suggest a shift in the supply curve. But preliminary regressions relating output and employment as well as changes in output and employment to relative prices at branch and regional levels provide ambiguous results. The output to employment link is largely absent, although there is a weak correlation in the changes when regional data are used. More promising is the association of branch level output changes to relative prices. Relating the change in branch output relative to industrial- sector output with the change in branch relative prices shows a clear and robust association over most branches. The association is particularly evident for light industry and the machine-building industry, which are experiencing the strongest relative decline in output and prices. This provides limited support for some structural dimension to output losses but, equally, provides little evidence that this has translated systematically into employment changes. Once again, there is evidence of considerable rigidity with respect to employment. 4.51 A correlation between employment losses and declines in relative wages can be seen in the analysis of regional wage movements and relativities. Thus, while wages in the extremities (for example, in the Far North) remain adjusted upward uniformly by coefficients of 1.5 to 2.0, wages in the Northwestern and Central regions--the areas with the largest net job losses--declined relative to the national average in early and mid-1992. This is supported by the ILO survey of June 1992 that found wages to be systematically lower in enterprises with the largest contraction in employment (see Standing 1992). The main loser in terms of relative wages is engineering, which also accounts for the bulk of job losses in industry. This may indicate the emergence of a weak but systematic association between wage behavior and levels of unemployment in regional labor markets. 4.52 The overriding impression that emerges for employment is one of significant inertia. The analysis has pointed to an absence of a robust association between sales and employment changes, between output and employment changes, and between output and employment and relative prices. It suggests that the employment picture is dominated by gross labor flows conditioned on relative wage changes resulting from the relaxation of central controls. While the data sets used do not permit identification of the destination of those who quit their jobs, information on the characteristics of those who quit through 1991 suggests that leaving the labor force was not the dominant motivation. 4.53 Labor Hoarding. The relatively gentle decline in employment relative to output in industry is particularly striking, given extensive and continuing labor hoarding. Nearly two-thirds of sampled enterprises reported excess employment levels in the third quarter of 1992. This excess employment was fairly equally distributed across enterprise sizes and branches, at between 8 percent and 14 percent of current employment, except for the largest enterprises where the estimate was below 1 percent. 4.54 Institutional factors governing dismissals or the presence of a union might be expected to impede involuntary separations, but this is not generally the 109 case.13 Unions are present in most work forces, but they wield negligible bargaining power. In only 10 percent of cases where excess employment was present were dismissal rules and worker protest cited as factors of any significance in governing employment decisions. 4.55 Labor hoarding can be traced to a combination of factors, including production technology and an apparent willingness to provide fall-back wages (the minimum wage) within the enterprise. In nearly two-thirds of cases, the enterprise motive for hoarding labor was the belief that output would shortly expand. In 25 percent of cases, enterprises argued that such workers were not a significant financial burden to the enterprise. Evidence from the survey suggests that some enterprises, particularly in engineering and light industry, have begun to place parts of their labor force on or around minimum wages with minimal work requirements. Overall, minimum wages were reported for nearly 4.5 percent of the sample work force (13 percent for engineering enterprises and 9 percent for light industry). This may be seen as a rational response on the part of enterprises: taking benefits (primarily housing) as a short-run fixed cost, wage reductions to the minimum level would be equivalent to outright severance costs. In the case of an aerodynamics design enterprise, around a quarter of the work force of 10,000 (primarily the unskilled) had been placed on minimum wages (see Box 4.1). This amounts to de facto provision of unemployment benefits within the enterprise with, of course, the difference that employees maintain access to enterprise-provided benefits. Assuming the rough distribution of wages in total labor costs from the enterprise side and constant access to non-wage benefits, resorting to minimum wages would have allowed enterprises to make per capita cost savings of at least 45 percent over 1992. Unemployment 4.56 The fear of mass unemployment in Russia haunts current policy discussions. This anxiety is promoted by the general collapse of output and the recognition that structural changes are likely to induce significant job destruction that is only partially offset by creation in growing sectors. As yet, although the changes to measured unemployment have been significant in absolute numbers, the base has been very low. The apparently dominant strategy of enterprises since 1991 has not been to induce large-scale retrenchment. Further, turnover has remained high and separations in 1991 remained dominated by quits. What is striking is that this broad picture continued at least to the end of 1993. Such adjustments have largely been borne by wages (and possibly by benefits), with little change on the quantities side. 4.57 The high educational level of the unemployed has been a striking feature to date. Unemployment has disproportionately affected skilled labor, and more than half of the registered unemployed have secondary, specialized, or higher education. For women, this share rises to nearly two-thirds. Thirty percent of the registered unemployed in January 1992 had a higher education, compared with 15 percent in the labor force as a whole. White collar workers such as engineers flf Firing d.oisions were taken by the administration alone in 60S of oases, by the administration in association with the trade union in 15X of cases, and by the factory council in the remainng 15. Opposition to dismissals arose from the trade uniona in half the sample, but significantly, no opposition was reported in a third of the cases. 110 and scientists figure prominently among the unemployed. Vacancies for new jobs, on the other hand, tend to be concentrated in areas of blue collar work, where males predominate (for example, machine fitters and builders) or where preference is now given to males. When unemployment becomes driven by plant closures, it is expected that heavy industry and defense will be strongly affected and that the structure of unemployment will shift toward blte collar workers. Nonetheless, it is likely that the educational background oaf the unemployed will remain high compared to that in other countries. In particular, engineers are likely to be in surplus. B oj2. uummlnt: Ibaauaeut ad 1atnit4tonal Issues Russian statistics neeurse unenpLoywwt in a Variety of was. The broad category of job soekere includes people laedng fu- or put-tun a pluent who wre lokfor work and hbe relstued in a raicn offiee of the Fedral iup1yawt service. Us"uld inlude datmbllzed £tltary zetrmenhed s f-I*Audf tho*4 ftsiOU94 by eAs" l£Ay*-i:e e.Aren"t. PeO4 weturning to the 1abor forca as vei 'tbose v*1=tarA1y 'witU pviosU eoUgWat. Workers diamsued bY enterprises have X e tibt t-o* U0OM OC tffe iasntt 500#o#, pay mid UtaUaJ 8 t 4eSsif Led ia this poel, r .aer thn .eg:- t d . t ab v. A swhoaq aeto withthe status of ns 3: th naroaw measre. These incude, in . iiaoiple, all nstM'e :Jch teerabere wi sot. be matched Z4tb tpo~nt. lby th* MRS. At the cud of ~139hi,s _1q mntd. a3o;3S 1) o$ total. Job seekers, bat basen tWubgh 1992 to ner 401,Ti pat ba bee FmI_ gerne by th fleaio obSles heemnh aft severance.. The4 ?har# .le sxpeeted to :*. tuxthe-~ Thea sMW*t.i** ara Ukety to be bia*sd dowward ; for several -eesove.. Thtosbe who do not regist at an anpiopet ofie ibecauSe unuuploment benetits Ar w a e ad siDd o a *e-o mn rap4id to i4 ve s,theo :a0% to re5 e e*aekigure.4 plo, th eptoih of hidsen im yt emergat ms ures - =.(Box 4.2 expaid th a frant astoes).±Te broad measure . :. a doubling ot reitered job95 thsisearchers andbtA the unmloyenter,ate b eptween ece1 ber t 1991 and oetevebene99f in en thdv*4om ae indusr ealstrucur var'aies hidtory a acrostihey aountry oregwona. voardaiation inr ui4$nemplymentie rat es would beti expctd.rfslo Dtwon h 4.58 Figure 4.4 plots the path of unemployment over several measures for 1992 (Box 4.2 explains the different measures). The broad measure shows a doubling of registered job searchers and the unemployment rate between December 1991 and October 1992. Since the industrial structure varies widely across the country, regional variations in unemployment rates would be expected. Data on the regional distribution of unemployment reveal that while a significant share of total unemployment (22 percent) remains concentrated in the central region, which includes Moscow, there is relatively little regional dispersion. According to the broad measure of job searchers, the Northern region consistently has the highest unemployment rate, although it is less than a percentage point above the Russian mean. Variation across regions has increased slightly since June 1992. There exist isolated pockets of relatively high unemployment--Yaroslav oblast in the central region (where unemployment in February 1993 exceeded 4 percent) and parts of Northern Caucasus, for example. It is likely that regional variations will become more pronounced as enterprises restructure and energy prices rise to world levels, and relatively high rates of unemployment will emerge in certain regions (see Box 4.3). 111 Bn A43. Tho Likely Pattern of UnmtpJ.oynwt and Policy mplicationas Economic policies whih bhave teaulted in hyperitnlation and generaL chaos are unsustainable. the economic costs of stabilization end reatructuring have been postpored but not avoided. All the Eastern European countries that introduced stbilisetion programs axserienoed a seavre *hock resulting In a shaVp output decline across the econoy as a whole. Uneployment rzates by mid- 1992 had risen to around 17S in Bulgaria. 13S in Hungary, 142 in Poland, 112 in Blovakia, end 41 and rising repidly in Rcmnmia. Only in the Czech Republic has large-scale unemployment been avoided, Even so, aggregate shocks have seotoral effects. These depend on the mis and specifics of fiscal and monetary policy, the exchange rate regime, trade poll&y end the use of tariffs, and so on. Eastern European experience indicates the extett to uhich the impact of transition is uneven across industries, regions, end social groups. In Poland, in the first year of reform, the decline in output ranged from 10X (power) to over 402 (textiles), While waploynant changes varied mAch less (from an increase of 2X in pwer to a fall of 11X in textiles), the implications for regional unemployment rates w. o substantial, with a rause fOm 22 (Warsaw) to 121 (Suwaiskie) by December 1990. Likewise, in the first year of reforms in Bulgarie, changes in output ranged from an increase of 52 (non-ferrous metallurgy) to. fall of al)nat Sl i-n electronico. Again, employment chon8es showd a much Maller variation (which in Bulgaria appears to have no correlation with the changes in output), ranging from an increase of 4X (gas and oil production) to a felL of 1S5 (other ikndustries). In 1992, regional unemployment rates in Bulgaria ranged from 121 (SofIa) up to 27X Ofihailovgred). Unemploymeat rates tend to be highest in regions dominated by old manufacturing industries and lowest in tradiag centers, most often in the capital caty. Avoiding very high rates of regional unemployment Is important partly bacause its ill-effects are both cwuulativo and Interective. Uigh rates imply long spells, and long-tenm unemploymnt can have particularly harmful effects an peoples High rates of unemployment, in a locality, can lead to disintegration of the-local econoqm and the social system btJed on ,it Unacceptably high levels of unemployMent could well destroy political support for tbe refor= process. In the short term, then. it is Important, witbout undermining the objectltea of the reform package, to attack tha potential emergence of Very higb rates of pne"#l4ywent In p*ti,oular localties. Seh possible shpe of such a policy is addresed in Chapter V. 4.59 Evidence on the composition and duration of unemployment is presented in Table 4.2. For the newly unemployed, the minimum duration of unemployment rose from less than three months in January 1992 to over five months in 1993.14 By early 1993, over one in ten currently unemployed had been without work for over one year. In areas of relatively high unemployment, such as the northern and Moscow regions, unemployment durations are already protracted. Around Moscow, nearly 20 percent of registered unemployed reported durations of eight to twelve months by July 1992. Another striking feature is the high proportion of women and their continuous distribution over respective durations of unemployment. On January 1, 1992 over 72 percent of the registered unemployed were women. Their large share in unemployment can be attributed to discrimination against clerical and administrative staff, positions held primarily by women, as enterprises seek to maintain their employment of production workers. Young people, who are likely to be new entrants to the labor force, also constitute a large share of the unemployed. In early 1993, nearly 19 percent of the registered unemployed were under age twenty-two, and a further 20 percent were between twenty-two and thirty. Additional information on the skill and other attributes of the unemployed indicates a strong bias toward white collar workers. In other words, unemployment has yet to severely hit the male, blue coilar work force. )/ When unemployment is rising, minimum duration can be estimated by dividing the stock of unemployment of the monthly inflow into unemployment. 112 4.60 Mass layoffs are a crude proxy for measuring any apparent shake-out in inflows to unemployment. In 1992, mass layoffs accounted for only 8 percent of separations. With the exception of the central region, where the level of layoffs is disproportionately high, regional data show such separations to be fairly uniformly distributed. This again points to a common response function and path across regions. 4.61 Outflows from unemployment show a clear drop in the ratio of job finds to the stock of unemployed. The absolute numbers of unemployed who found jobs during 1992 changed a little. This accords with economywide data that indicate a continuing high volume of hiring in the economy through 1992. The fact that a consistent 40 percent of the outflows from unemployment are reported as finding jobs is remarkable in a context of such large declines in output. According to national FES data, private entities account for only 15 percent of total outflows to jobs. Data on the efficiency of the job search shows that in September and October 1992, over 30 percent of those who found a job did so within ten days of registering as unemployed (after the exhaustion of severance pay). 4.62 The pattern of vacancies shows a clear and sharp rise until September 1991 and a significant fall thereafter. Posted labor demand by enterprises at the employment service fell by around 50 percent between January and October 1992. Most (more than 85 percent) of these posted vacancies are for manual workers, reflecting the historical bias in the nature of labor demand of Russian enterprises. Conclusions 4.63 While wages initially bore almost all the adjustment, there were far stronger signs of a wage push in late 1992. Employment reductions began to accelerate in the latter part of 1992, but from a low base. The deceleration in the growth of unemployment since the end of 1992 can primarily be explained by a fall in the inflow rate. Job losses are largely offset by hires, and much of the churning throughout the labor market appears to be through voluntary separations. This may in part be linked to the emerging process of wage differentiation and to the relatively buoyant demand for labor posted by enterprises. The latter phenomenon can be traced not simply to institutional inertia but also to short-run constraints on production exercised by technology and the associated level and structure of labor demand. Accelerating nominal wage claims have been fueled primarily by the reversal of earlier announced reforms (particularly with regard to the monetary stance) and workers' evident intentions to link wage claims to price changes. The spiral is not a function of trade union power in the conventional sense, but it does reflect the control structure and decisionmaking rules that characterize the bulk of Russian enterprises. Lax monetary policy and decentralized insider power, giving rise to relative employment stability and real wage rigidity, are powerful ingredients for hyperinflation. 113 Figure 4.4 Russia: Unemployment and Vacancies Oecenber 1991 - June 1993 1 2 1.1 I e' 0.L7 ew% 0. o 0. 0.5 0.4 0.3 0.2 0.1 0 dec91 fe'b ar Ju'n aug oct dec'92 feb alpr June Jan92 maor May Jul SeOP nov jan93 mar may o Unemfp Wa-od + Unenvp Narrow 0 Rag Ist UnGMi A, Vacancies Table 4.2. Distribution of the Unemployed by Duration 1992 % of unemployed-to total stock Total Duration All Unemployed Youth (months) Women 16-29 <1 27 31 70 1-4 49 34 72 4-8 18 32 72 8-12 6 20 65 Total 100 32 71 Note: Youth and women are not mutually exclusive. 114 Policy Implications: The Potential Role for an Incomes Policy in Russia 4.64 Of foremost importance in the present context is high and possibly accelerating inflation, which can be largely attributed to lax monetary and credit policies. A decentralization of enterprise decisionmaking and a concentration of market power have also been observed. Inflationary expectations are high, although there has not been any widespread shift to full indexation in the economy. Survey evidence suggests that wages have not diverged significantly from the old centralized tariff system and that collective bargaining is not widespread. In fact, real wages have fallen significantly since 1991, although with some rebound in early 1992. 4.65 Incomes policies attempt to influence the determination of wages and prices in situations where the participants, either enterprises or labor, have some degree of monopolistic power (Bosworth 1991). Here, the focus is on wages. The objective of a wage policy is to control inflation while attempting to avoid mass unemployment; in other words, to influence the unemployment-inflation trade-off. Arguments for a Wage Policy in Russia 4.66 The primary sources of inflationary pressure in contemporary Russia are evidently not cost-push in nature. Most significant has been the laxity of monetary and fiscal policies. There is no real distinction between monetary and fiscal policies, since any budget deficit is covered by Central Bank credits. 4.67 In such a context, an incomes policy cannot be a leading instrument for tackling inflation. It can only complemrnt policies of monetary and fiscal restraint. In Western economies, incomes policies have often failed because they were used as a substitute for the adoption of sound macroeconomic policies that were seen as painful and politically difficult. Stabilization will require the adoption of tight monetary and fiscal policies, including a rationalization and reduction in the volume of subsidies and transfers to enterprises. A wage policy may well assume importance under stabilization, given the need to keep costs in line with the downturn in aggregate demand. Inflationary expectations are likely to be high and volatile, and, relatedly, the upward momentum in prices would be such that indexation coefficients would be effectively too high. At the same time, where there is extensive state ownership of enterprises and worker involvement in enterprise decisions, "capital" lacks an effective advocate. Wage policy may assist in a downward revision of these patterns and practices by substituting for effective enterprise governance and management and by breaking inflationary expectations. 4.68 Nominal inertia, the typical argument for a wage policy, is not yet evident in Russia. In the course of stabilization, however, there is a risk that nominal wage increases will exceed those of prices and that the resulting increases in real wages would feed through rapidly to unemployment. It is likely that the impact of unemployment on wage behavior will suffer lags, given the history and continuing levels of unemployment, which are significantly below equilibrium levels. A wage policy could help to avoid this scenario. 4.69 A further argument is that an incomes policy would help to alleviate the short-run inflationary implications of weak competitive pressures. In Russia, as 115 in many socialist economies, the pursuit of scale advantages led to highly concentrated production structures. Decontrol in the absence of competitive markets can result in the cartelization of production, weak managerial resistance to wage claims, and large price increases (see Chapter II). Drawbacks and Problems with a Wage Policy 4.70 Wage policies are subject to a range of criticisms, from the potentially serious efficiency losses to the feasibility of implementation. These are addressed in turn. 4.71 The Need for Changes In the Structure of Relative Wages. The Soviet system of wages was based on a centralized tariff, with low returns to skill or grades. There was inadequate variation in wages both by skill and by region (Jaclkman and Rutkowski 1993), and differences that did exist were unrelated to individual productivity. In effectively functioning labor markets, wage differentials operate as a guide to resource allocation decisions. Thus, it would seem undesirable to entrench an previous structure of wages that is wrong. But in situations where the basic assumptions underlying allocative efficiency (competition, labor mobility, and so on) are absent, the forces that are supposed to guide efficient allocation decisions do not exist. Even if wages are left to sort themselves out, shifts in accordance with allocative efficiency need not follow. Indeed, the decentralization of wage decisions has not greatly increased the extent of sectoral differentiation in Russia (see paragraph 4.15). Therefore, the case for presuming that adverse efficiency effects necessarily flow from centralized wage restraints is not strong. 4.72 There is a danger that incomes policies can become a mechanism for the reassertion of centralized control. Historically, wage setting was an administrative and highly centralized exercise. In adopting a wage policy during the transition, the Russian government should try to limit wage movements without defeating the underlying shift toward more open and competitive markets. 4.73 The Weakness of Existing Trade Tnions. The official Federation of Independent Trade Unions (IFTU) was the only element of the party apparatus to survive the transition from totalitarianism largely intact in terms of its assets, privileges, and membership. Truly independent trade unions, not associated with the old regime, are limited in size and coverage. The IFTU has been able to preserve its position partly by its continuing control over important employee benefits (see Chapter II). Yet, despite a presence in virtually every Russian work place, the IFTU apparently does not enjoy the trust and confidence of labor. There is no clear separation of worker and management interests. Indeed, the IFTU had continued the Soviet practice of supporting the bloc that represents state enterprise directors, undermining attempts to establish a tripartite bargaining mechanism during 1992. 4.74 Union coverage is likely to decrease as private sector activity expands and the average size of enterprises declines. The transfer of control away from the Social Insurance Fund would also likely induce increasing numbers of workers to stop paying their union dues. The weakness of trade unions detracts from the viability of a consensus-based approach. 116 EAR.lre=6 4.75 Recent experience casts some light on the choice between a wage policy based on centralized norms (tripartite) and a policy based on tax penalties. In Bulgaria, Hungary, Poland, Romania, and the former Czechoslovakia, some form of wage controls have been in continued and widespread use. All are tax-based, most have an explicit link to enterprise profitability, and most exempt the private sector (see Table 4.3). Coordinated systems for centralized wage determination, which have operated in Sweden, Austria, and some Central and Eastern European economies, also reveal some relevant lessons. Table 4.,. Wage Policies in Eastern Europe Real wges, tegeac 1991 Coutay Meam Period (la 1990) Prf-link Tax-bsd Tax ates Private Sor Bula Equal absoblt 6 morths -39% 1-55% Q 80 80 - Exempt Increase (adJ. QZ) -50% aU 1991 (none/ +1%) CSFPRb Celing on Quarty -10%/-20%(Q2) Yes Yes 200% if > 3% Exemp wage bl (n1l./ -3%) 750% if > 5% Huny Cein on Yeadly -10%1-7%(Q2) Yes Yes 50% BUMP wage bill (-3%1-3%) Poland hIdexaion Monthly +3%/-1S%(Q3) Yes Yes 100% if <3% Bxemp of avage wage (-30%/-30%) 200% if > 3% In 1991 500% if > 5% (wae bll in 1990) a. Ford pvmu- mnobya. wosav ew teo dimiaely Far wma amaod eaooeadulvm w weacn us k wit ocw via a ceih arn do NW wag bit. Ilb nit wm ovbba =rt for we bureus abne do asaiin war. wd b. T1. wagebMitt ui. barn... A wvge4utff aynmsNail jas. C. A Tlawwagewa w, rmhamy 9. am fin.0 urWpdo(eeI wa atefr tor u oveal wpt bml c g_ts t -in erf_ . Fr t990 k e a w e s itt abowe wes dite w exed s a uxteas. For 1991. atoiwe coIarausoib 1£8. D a- l8i4 b,* 285. do pu of I rm, abov 185 mu wed aft U, a Ipaeoxu . Above23%. U,, slut hams is med. 1Pbriaulv 1w" labor dasftn is give. bayI wageau upnus ~r55 wb aupioyum3 istU.d d. Caumaeistied fimne added miie so atts eingia Mantaty wspe seat were dknbutd In immsy 3989. 4.76 Social compacts are diffl : ult to achieve and sustain, partly because it is difficult for democratic trade unions to negotiate and preside over declining living standards. There may be a general acceptance that real wages have to fall when a conscious tripartite realization facilitates a substantial fall virtually overnight. Bulgaria's wage policy was associated with a significant fall in real wages (about 35 percent) as a tight incomes policy based on a social compact was supported by fiscal and monetary restraint. At the same time, however, unemployment rose rapidly to high levels (about 16 percent in late 1992), so the wage policy cannot be regarded as generally successful. In Czechoslovakia in 1991 a tripartite agreement complemented a tax-based wage policy (see Table 4.3). A fall in the real wage was specified, with some dispute about whether this was a target or a minimum, since the target was actually exceeded. In the second 117 half of 1992, in the absence of consensus, the Czech government set guidelines for wage increases and taxed the excess. 4.77 Politics, institutions, and social consensus are all important determinants of wage policy, from the political strength and credibility of the trade union movement to the central-local balance of trade union power (Ahmad, Corricelli, and Revenga 1992). Compare the experiences of Bulgaria, where tripartitism worked, and Romania where an absence of coordination resulted in a wage push in 1990 and 1991 and ineffectual implementation of the excess wage tax. A relatively high degree of coordination and tripartite agreements, whether industry level or economywide, is preferable to tax-based wage policies (Commander 1992) to avoid the distortionary effects of taxation. However, where consensus is absent or not feasible, a more punitive approach could be based on an excess wage tax. In Poland, for example, where there was no tripartite agreement and local trade unions are relatively strong, this route was adopted. 4.78 Russia faces difficulties in implementing a consensus-based approach. Consensus requires a degree of stability and commitment. This is very difficult to obtain and sustain under conditions of pervasive uncertainty and the risk of decapitalization by insiders. A corporate-type approach requires confederations of these representatives of labor and capital who are competent to make decisions binding on their members. Currently in Russia, effective worker representation is weak and seems to be weakening. Moreover, centralized bargaining is more difficult in a large, increasingly decentralized federation such as Russia compared to smaller countries. Finally, an appropriate framework of (quasi- legal) tribunals is required to adjudicate disputes that arise. 4.79 These conditions lead to the recommendation that a wage policy in Russia be based on taxation, rather than on institutions and commitments that likely do not exist. The administrative apparatus for tax-based policy is already in place, thus facilitating its implementation. This means that punitive tax penalties on enterprises exceeding some defined wage norm are basically a continuation of regulatory policies adopted in the Communist era. Recommendations 4.80 An effective wage policy could make an important contribution after stabilization, if nominal inertia in wage setting worsens unemployment trends. It could be used to affect the iniflation-unemployment trade-off. However, it could not rely on the present wage tax, which is based on arbitrary norms and low tax rates. Moreover, any wage policy will be ineffectual in the absence of certain preconditions. First, there is the need for monetary and fiscal restraint. With this, hard budget constraints must be imposed on enterprises. This would include enterprise exit procedures (that is, operative rules with respect to bankruptcy and liquidation). Otherwise, there will be a continuation of the current situation. Russian enterprises have opted to pay the penalties without much impact on their wage decisions. Indeed, in Poland in 1990, many enterprises facing soft budget constraints were prepared to incur tax penalties on excess wage increases at much higher rates (200 to 500 percent) than those that are currently being imposed in Russia. More specifically, a wage policy should address the following points. 118 4.81 Wages should be defined to Include non-cash benefits. A significant proportion of labor income in Russia, as in many other industrial countries, comes in a non-wage form. This must be taken into account in an attempt to restrain labor costs. There is already evidence that Russian enterprises have sought to avoid wage controls by increasingly resorting to compensation from their social funds. 4.82 The scheme should be as slmple to administer as possible. Initially, a standard should be set for increases in the overall wage bill. On the one hand, a wage-bill tax is simpler to administer than alternative schemes. Employment reductions would allow greater flexibility and relative wage adjustments. On the other hand, a wage-rate tax has the advantage that it permits payrall and employment to be increased in expanding enterprises, and it avoids encouraging unwelcome adjustments to the skill structure within enterprises, which a tax based on the wage bill might. However, a wage-rate tax is more difficult to implement: Romania faced serious difficulties in implementing a wage rate-tax. Both forms of taxation could lead to bias in favor of low-skilled labor. 4.83 There should be a uniform standard for wage changes across the economy or major sectors. The standard should be preannounced and not subject to exemptions. The Western experience with productivity bargaining and other enterprise-specific standards has been problematic. Moreover, enterprise profits during the transition have more to do with monopoly rents than with worker performance. The Polish tax rule for 1991 encouraged enterprises with market power to pass wage increases through to prices. 4.84 Indexation coefficients and the frequency of adjustments have to be carefully considered. The longer the interval for indexation adjustments, the smaller the transmission of temporary price jumps to sustained inflation. Conversely, very high rates of monthly inflation imply correspondingly large drops in real wages if nominal wages are not adjusted. It would be difficult to sustain a single policy over any length of time. Initially, wages should be indexed not more frequently than monthly; then, with successful stabilization, not more often than quarterly. Indexation should be linked to the expected or announced inflation path rather than to past inflation. But when inflation is accelerating and very unstable, backward-looking indexation is preferable--hence the importance again of stabilization. Forward-looking indexation permits compensation for quarters when inflation is higher than anticipated. Indexation should be partial, in that price increases due to a deterioration in the terms of trade should not be included in the price index, since the loss of real income that would imply should be partly borne by labor. 4.85 Wage policy rules should be temporary. Unless the wage policies are limited in duration, the risks of relative price distortions and rigidities and adverse effects on labor productivity and effort--and of subsequent wage explosions--are magnified. This can be seen in the experiences of several Latin American countries with heterodox adjustment programs. 4.86 To summarize, the priority of the Russian government should be to stabilize the economy, which will require monetary and fiscal restraint and an end to large-scale, indiscriminate subsidization of the enterprise sector. With stabilization, there is a risk that wage claims could lead to a reacceleration 119 of inflation, given time lags in the adjustment of wage-setting behavior. In order to avoid this and consequent repercussions for unemployment (when the overall rate is already likely to be high), it could be appropriate to institute a temporary wage policy. While a cooperative approach to wage setting is preferable, it is unlikely that a sufficient degree of consensus could be attained in contemporary Russia. Therefore, a wage policy would follow a more decentralized, punitive approach and rely on tax instruments. Initially, a tax would be imposed on the wage bill at a high rate, with a low coefficient of indexation. Subsequently, the tax could shift to the wage rate at a lower tax rate, and with a higher indexation coefficient and less frequent periods of adjustment. The tax-based wage policy would then be phased out. 120 CHAPTER V: IMPROVING THE EFFICIENCY OF THE LABOR MARKET 5.1 In the post-Stalin period labor allocation was in fact achieved more through market-type mechanisms than through compulsion or a plan. Most individuals were free to seek employment in response to alternative opportunities (subject to mobility constraints; see the section discussing housing later in this chapter, and paragraphs 5.24-5.25), although until recently unemployment was in principle illegal. However the centralized wage tariff system, even combined with the widespread use of bonuses, resulted in an overall structure of relative wages characterized by low returns to skill or grade. Incentives to enterprises were such that there was a substantial degree of overstaffing, particularly in industrial enterprises. This has increased dramatically given the large drops in output. 5.2 Chapter V focuses on measures that would facilitate the movement of workers to more productive uses, on the assumption that enterprises face incentives to utilize labor in a more market-oriented way: First we explore the role of the Federal Employment Service in the light of international experience and emerging needs in the specific Russian context. Particular attention is accorded to the need to develop proactive programs (the issue of income support was covered in detail in Chapter II). This chapter is concerned with the objectives and design of employment policy in Russia after stabilization, and in particular with the approach to enterprises that face closure or mass layoffs. It builds on the extensive documentation underpinning the Employment Services and Social Protection Project (Staff Appraisal Report No. 11209-RUS), which is providing assistance in these areas. 5.3 Given the aforementioned project, this report is able to turn to related issues that remain to be addressed in formulating programs to deal with mass unemployment and the linkages between housing distortions and the labor market transition. The geographical mobility of labor is hampered by the scarcity of housing, by the absence of anything that resembles a normal housing market, and by a system of residence permits that has deterred internal migration. While labor mobility is likely to remain limited for the foreseeable future given depressed labor demand, it is important that the obstacles to mobility be removed as far as possible. The discussion of housing reforms highlights policy options that will facilitate restructuring and flexibility when the labor market does become more buoyant. The Role of the Employment Services 5.4 During perestroika, it began to be recognized that if the labor market was to function more efficiently there had to be at least frictional unemployment. In 1991, a new (Soviet) Employment Law made it possible to register as unemployed and to receive unemployment benefits for a limited duration, and an Employment Fund was established. These arrangements became part of the law of the Russian Federation. The employment services became the responsibility of the Employment Committee, which was part of the Ministry of Labor. In June 1992, the link with the Ministry of Labor was severed, and an independent Federal Employment Service (FES) was set up. 121 5.5 The FES is responsible for providing overall leadership, policy direction, and management of the nationwide system of employment offices and programs, primarily through oblast employment offices. Its functions include planning and budgeting the nationwide programs, managing the Employment Fund, establishing objectives and performance standards with regional offices, developing program designs and procedures, and providing oversight for some of the oblast FES offices and the 2,121 raion employment centers and bureaus. The system is fairly decentralized, in that oblasts and raions enjoy a certain level of financial and decisionmaking autonomy. The oblast employment offices are responsible for managing employment programs, and the raion offices within their administrative jurisdiction, including planning and budgeting, provide guidance and technical assistance and oversight of their operations. The raion employment offices deliver the programs authorized by the Employment Law directly to clients. Raion offices are the primary interface of the FES with the general public, where the day-to-day operations of the employment service occur (for example, registration for job placement and unemployment benefits). Payment of unemployment benefits is made through the Savings Bank, an organization that has oblast-level and raion-level offices and some 41,000 branch offices. 5.6 The FES has undergone two reorganizations in the past twelve months. At the center, its capacity is clearly overstretched, with about 150 staff on board out of a total of 210 authorized positions. Building on existing procedures, new management systems to streamline communications between headquarters and the oblasts began to be developed in 1992, including a revised budget and planning system, a formalized directives system, and an initial Management Information System (HIS). These initiatives are being developed with Bank assistance, first through funding under the Technical Cooperation Program (TCP), and later under a Project Preparation Facility, followed by a loan (approved by the Board in November 1992) for the Employment Services and Social Protection Project. The problems of the system of unemployment compensation--complicated calculations, inadequate benefits, and inappropriate financing arrangements--were addressed in Chapter IV. The following sections highlight proactive activities being undertaken or planned by the FES. 5.7 All the market industrial countries reflect a belief that there is an important role for government in the provision of employment services. The various forms of market failure here range from information asymmetries as to job availability, to the absence of capital markets to facilitate training, to investments in human capital. The mass of empirical studies tends to support what economic theory would lead us to expect. Activities that seek to allow a more rapid matching of job vacancies with job seekers through the provision of an information system and assessments of clients are both useful and cost- effective. Counseling the unemployed, including aptitude tests and training in job search techniques, also reflects positive results, particularly for the long- term unemployed. Another lesson that has clearly emerged from the experience of Eastern Europe is that a legal and budgetary distinction must be made between the source of financing for income support and that for proactive employment programs (see Chapter III). 122 Service Delivery and Staffing 5.8 Local capacity may be inadequate to deal with large scale unemployment. In the Soviet period, employment offices (Job Placement Bureaus) helped to put job seekers and employers in touch with one another in an era when open unemployment was unknown. Despite their enlarged importance, employment offices remain thinly staffed and typically short of office space and computing equipment. The Bank project has a large computerization component. Given the present staff of 21,500, if unemployment rose, as projected, to an enormous 5 million, there would be one staff for every 232 unemployed, compared to Sweden, where there are only nine unemployed for each staff member. 5.9 Staff are a critical input to employment services, particularly with the emphasis on proactive policies as distinct from benefit delivery. Available evidence reveals wide inter-country variations in levels of staffing, although it is difficult to compare client-staff ratios without knowing the exact composition of the services delivered (see Table 5.1). It is, however, fairly clear that the increasing demands placed on staff (with rising unemployment and the emphasis on proactive policies) and limited funds have created serious problems with staff recruitment and training, even in countries with well- established employment services. Increased staffing may reduce total costs by reducing the number of clients but not cests per client. In Russia, existing staff could be better utilized with a simplified structure of unemployment benefits and technical support to repl _e the existing manual time-consuming methods. Table 5.1: Ratio of Staff to Unemployed in Employment Services in Selected Countries, 1988 Countrv Ratio Austria 1:33 Germany 1:37 Poland 1:225 Portugal 1:120 Sweden 1:9 Russia (Dec. 1992) 1:3 (Dec. 1994) 1:232 United Kingdom 1:98 NM: Staff estimates for Russia were based on 518,000 registered unemployed and 21,500 field staff, as of December 1992. December 1994 estimates are based on a projection of 5 million unemployed. Source: Fretwell (1992) Training 5.10 Beyond placement, employment service activities should extend to counseling on educational and training opportunities. When mass unemployment emerges, the 123 unemployed are likely to come from sectors possessing low or obsolete skills and having neither an employer nor private resources to support their retraining. Hence, there is a strong prima facie case for government support for training the unemployed. However, it should be noted that evaluations of such efforts in industrial countries have shown mixed results. While employment service offices with job training programs have higher wage/replacement ratios, a key problem identified in many industrial countries is that neither employers nor unemployed workers have much faith in government training courses for the unemployed. Linked to this is a conflict between efficiency and equity- -whether to target the best and brightest whose employment prospects are better in any case, or to concentrate efforts on the less able, who may have difficulties with the course and also in finding work afterward but for whom training may have a greater incremental benefit. In practice, certain groups tend not to enroll (older, less educated, and female), sometimes because they are actively discouraged (as are women in Poland). Programs tend to upgrade the skills of those workers who already had the most likelihood of reemployment. 5.11 The high level of education in Russia is likely to provide the foundation for its economic recovery and eventual rise to prosperity. In 1990, about 12 percent of Soviet adults had completed higher education, which exceeded the levels achieved by several Western European countries. In addition to its general education, Russia has a large and complex training system with a tradition dating back over more than a century. In recent decades, preemployment training has been provided to about two-thirds of each age cohort graduating from the general education system. Those who leave school after only nine years of general education entered either a vocational school of the Ministry of Education or a technical secondary school usually operated by sectoral ministries. Once employed, many workers at all levels, from operative jobs to positions requiring postgraduate qualifications, have been provided with in-service training. 5.12 Both higher education and the training system have suffered from excessive specialization oriented to industrial needs, as most vocational and technical schools were very closely linked to specific enterprises. In the near future, however, requiring the identification of specific impending vacancies before retraining an unemployed person is unrealistic. The demand for labor will not be at all buoyant. Consequently, emphasis should be placed on training of more generic skills, such as electronics, quality control, customer service, commercial skills, and computer skills. For some skills, such as accounting, there is already a very clear unmet demand. At the same time, allowance must be made for the unskilled and relatively less educated people who have become the core of the unemployed in market industrial economies. Computer literacy is unlikely to be an appropriate goal for the bulk of unemployed once joblessness emerges on a mass scale. 5.13 Training does not, however, create jobs in a stagnant economy. In any case, Russia suffers from a serious shortage of qualified trainers in the subject areas where job prospects are likely to be brightest. In general, resource constraints will severely limit the possibility of using training as a major policy instrument to deal with mass unemployment. Nonetheless, by sponsoring training on behalf of the unemployed through contracts with existing institutions, the FES can stimulate continuing training activities so as to enable workers to adjust to changing demands. Under the Employment Service and 124 Social Protection Project, new training strategies and policies will be developed, and materials provided to deliver flexible adult training programs in response to the needs of the emerging market economy. Indeed, FES forecasts predict increasing training efforts. In addition, NGOs, especially those that focus on the needs of women, have begun to undertake pilot training programs. Mass Redundancy 5.14 Many countries require that workers' representatives be given notice (say ninety days) and proper consultation prior to large-scale retrenchment. Furthermore, a large group of European countries, including France, Spain, Greece, Germany, and the Netherlands, require official approval of the social plan for workers who are to lose their jobs through mass redundancies. While one could argue that when shedding jobs becomes complex and costly, disincentives are created to capital investment and to domestic restructuring --the evidence points to the value of on-site employment services set up as early as possible before the actual layoff date, and to giving advance notice. 5.15 The most effective programs appear to be those where both employers and workers are directly involved in the design and delivery of services. This approach has been followed in Canada and the United States and is being adapted and used in Hungary and Poland. In Canada, between 1971 and 1981, labor- management committees, set up with the assistance of the Industrial Adjustment Service, found sixty-six jobs for every one-hundred workers displaced (usually within a year). On average, IAS assistance reduced the jobless spell by two weeks (thereby virtually paying for itself). The scale of adjustment in Canada was obviously much smaller than what is anticipated to occur in Russia. Nonetheless, efforts have begun to test the Canadian model in Vladimir, an industrial center about 200 kilometers from Moscow. After some initial misunderstandings of the purpose of the program, industry and the community have responded with enthusiasm. Tegmorary EmRloment Programs 5.16 While the foregoing activities have an important role in facilitating labor mobility, they cannot be the spearhead for tackling mass unemployment, because, for obvious reasons, areas of high unemployment are unlikely to have jobs available in any case, and any vacancies that arise are likely to be quickly filled without the assistance of the employment service. A statutory provision enables the FES to use money from the Employment Fund to finance public works programs run by local authorities. Temporary employment schemes are a useful strategy that could be utilized in Russia in response to rising unemployment and to the large numbers of individuals exhausting their entitlements to unemployment benefits. To date, however, public works have been rather limited in scale and sometimes questionable in nature, as explained below. Support to strengthen the FES's capacity to develop substantial labor-intensive public works and/or temporary employment programs has been provided under the Employment Services and Social Protection Project. 5.17 Available evidence suggests that efforts to date have been undertaken on an ad hoc basis at the local level. There often tends to be some confusion between the categories of "job creation," "public works," and "wage (and 125 enterprise) subsidies,|' which is reflected in the accounts of the Employment Fund. For example, funds under each of these headings have been used to promote new local enterprises and effectively subsidize existing jobs. The Russian approach to public works is also striking because participants are paid both unemployment benefits as well as a wage enumeration for participation in the scheme. The costs of public works may be partly, or largely, provided by local authorities rather than the Employment Fund, which makes it difficult to clarify the full extent of activities. Box 5.1 presents some examples of job creation- type schemes. Box Innovattive Job Creation Efforts Undertaken by Local Employment Offioes in Late 1992 In Novosibirsk, the local MES regarded public works as a way to maintain skill levela. Individual incentives to participate lie in the payment of both the unnploymont benefit and the wage rate (although generally it was for piece-type work) end in the fact that participation was regarded as "work bistory." The activities were Jointly financed by the enterprise itself, the local budget, and by the Employment. Fund for such expenses a, publicity. -Public works in Altai Krai were not used extensively, due to the perceived high costs involved. In the first-three quarters of 1992, about R 234,000 was spent on pAblic works, out of a total revenue of R 283 million. Job subsidies were boing given on a limited scale: enterprises, often in budget sphere, were contracted to hire unemployed workers. The cost arrangenakta varied. In Pavloski raion -in Altai Xrai, Employmnt Fund resources were being used for such public works as tree planting, clean-ups, and minoz civil works--engaging up to seventy people. bwever, moat of the workers were .not.unemployedl Th-e local employment service classed as 'Job creation" loans to enterprises at for below market rates (10%). In St. Petersburg, the FES has engaged in pilot programs that advance credit at low irterest rates to anterprises-on the -expectation that the enterprise will become competitive within two to thbree years. The schemes, which ware rationalized as cheaper than paying unemployment benetits, were ; .id to he Wmore advanced there thn anywhere else in Russia. In this region, the oblast FES office reported th.at y reque8is for Job subsidies were received from factory managers. but these were often too vague and iffitiult to supervise. Low interest loans of R I million and R 2 million had bee adveaced to two factories. In addition, public works schemes were rum for seasonal workers. Vocisions to tinace such aotivities are nude by the oblest committee. 5.18 Public works cannot provide more than a partial answer to high and rising unemployment. There are several problems. First, in industrial countries, for example, France, Belgium, and Ireland, schemes serving the long-term unemployed experience successful transition rates (to regular jobs) of only 20 to 30 percent. Second, neither the FES, nor the Russian government, have the borrowing capacity or surplus management resources to run large-scale programs (unlike the United States in the 1930s, for example). The opportunity that public works schemes provide for the diversion of resources intended for the unemployed, most likely to contractors and input suppliers, and the concurrent potential for corruption, are great. Third, given existing budgets for infrastructure expenditure, there is a real risk that federal grants for public works will be used to substitute what would have been financed from the local budget in any case, thereby displacing existing local employees. Fourth, there is inevitably tension between the multiple objectives expected of public works schemes, in particular between that of producing useful output (which requires expensive complementary inputs) and makework projects that absorb large numbers of workers. Finally, it is clear that the generation of adequate income-earning opportunities in Russia requires the relative decline of the state sector and the expansion of the private sector. Indeed, in terms of income generation, it seems unlikely 126 that new public works can do as well as new private enterprises (which they will tend to displace) or existing enterprises that already produce output that is in some sense marketable (see paragraphs 5.27ff). The conclusion is that even in regions of high unemployment a public works program should only be one element of a local employment strategy incorporati'w training, job counseling, small business promotion, and other measures. lf; 5.19 Nevertheless, it is clearly useful to bvoaden the options available to the unemployed, to provide worthwhile social services or infrastructure, and to facilitate the development of local nongovernmental organizations and other contracting agencies. The program wage should be set moderately above the level of the unemployment benefit (say 150 percent), which is unlikely to attract those who could otherwise obtain remunerative employment. The duration would be strictly temporary--limited to twelve months, and possibly part-time, to allow the job search to continue. Given the likely composition of the Russian unemployed, the range of jobs under consideration for public works should extend beyond unskilled and semi-skilled to, for example, local facilities like libraries and hospitals, domiciliary care for the elderly, and community-based associations. At the same time, temporary jobs should be labor-intensive and require minimum capital outlays. For many participants, there will be a need for nursery facilities and projects that provide child care. 5.20 Public works represent a proactive employment policy that should be funded from general budget revenue, not the Employment Fund. Financing for public works should be allocated among oblasts based on the relative severity of unemployment and the program plans submitted by local authorities (that is, the availability of suitable projects). The federal government may impose matching requirements on grants for public works in order to pi vide greater local commitment to successful implementation. Local authoriti s in turn, should be encouraged to solicit applications for the provision of community services from local nongovernmental organizations. Enterprise m*sures that share the objective with public works of increasing the number of jobs in the local economy are explored in detail in the next section. Eployment Policy and Industrial Closures 5.21 This section is concerned with the context, objectives, and design of the employment policy in Russia in conditions of high unemployment, and in particular, with the approach toward enterprises that face closure or mass layoffs. As the previous chapter has shown, while enterprises have been adjusting in various ways to partial price and trade liberalization, including in some cases through reduction in employment, the changes have so far been relatively modest and unemployment has grown rather slowly. The introduction of policies to restore price stability.in Russia will lead to a much sharper rise in unemployment, and unemployment rates in aggregate can be expected to be at least as high (perhaps 10-15 percent), and as varied across sectors, as in the countries of Central and Eastern Europe. It follows that, in addition to an economywide problem of high unemployment, there will be much higher unemployment rates in particular sectors. Most seriously, there are likely to be regions, oblasts, or towns where the local economy faces collapse and where unemployment rates may rise far above the national average (see Box 4.3). Despite an initial preponderance of women axo white collar workers among the 127 - unemployed, as unemployment continues to grow it can be expected that the unskilled and young people will experience above average unemployment rates. 5.22 An important objective of labor market policy in response to a sharp rise in unemployment following stabilization should be to reduce unemployment in localities where it is particularly high, rather than throughout the economy as a whole. The overall levels of employment and unemployment in the economy in the wake of stabilization will be primarily determined by the stance of macroeconomic policy, and by the success of the stabilization program in restraining inflation and achieving external competitiveness. 5.23 Focusing first on the regional problem, a solution may be sought in principle, either by encouraging workers to move, or by creating or preserving jobs in the local economy. As a general principle, institutional constraints to mobility should be removed. However, evidence points to the conclusion that people are very reluctant to move in times of general depression. It is also difficult to persuade local communities to accept migrants when many of their own workers are out of work. Hence, policies aimed at improving labor mobility are not the most promising approach in conditions of high unemployment. Somewhat similar arguments apply to retraining. This is particularly the case in Russia today, given the impediments to migration and the lack of appropriately trained people and other resources for training. 5.24 The development of proactive programs by the FES (job matching, training schemes, and other measures to promote labor mobility and encourage labor market flexibility) is of vital importance for enhancing productivity and reducing unemployment due to structural imbalances in the labor market in the medium term. As a market economy develops, the location of industries, enterprises' technologies, and their skill mix should be determined by considerations of efficiency and competitiveness with workers having to migrate and acquire new skills in response. But it is unrealistic to assume that there can be sufficient mobility in the short run to tackle the problem of high sectoral unemployment rates, when aggregate unemployment is in double figures. Labor market flexibility and inter-sectoral mobility are of secondary importance if labor demand is collapsing everywhere. 5.25 Jobs can be created or preserved in the local economy through a variety of means: public works, temporary employment schemes in local government or public administration, employment subsidies to state enterprises or to private enterprises, or financial support to loss-making enterprises. The choice between these options will be influenced by financial costs, by the consistency of the policy with overall economic objectives for restructuring the economy, and by the effects (and side-effects) of any scheme on economic efficiency. 5.26 From a financial point of view, the critical question is whether a scheme pays for itself in terms of savings in unemployment benefit payments. A cost-neutral scheme would be one where the only outlays were wages; all those employed would otherwise have been unemployed, and the wage paid is the minimum wage (which is equal to the unemployment benefit). Public works and temporary employment schemes tend to be more expensive, typically requiring expenditures on complementary inputs, while workers may have to be paid above 128 the minimum wage. The output, even if socially productive, is typically non- marketable, and hence provides no income to defray the costs. Employment subsidies can be set at a rate s_ngrificantly lower than the minimum wage, but unless they are very carefully targeted, many of the jobs subsidized will not be additional. Continuing support for loss-making enterprises may be the cheapest of the options, depending on the extent to which (1) complementary inputs are already in place, and thus inexpensive (for example, management) or even free (for example, fixed capital); and (2) the enterprise produces positive value added goods (with inputs and outputs measured at world prices), thus generating a net cash flow. 5.27 The main objection to the principle underlying any schemes for subsidizing employment is that each, in different ways, appears to undermine the commitment to an economy run on market principles. Public works and temporary employment schemes seem to revive the role of government in the economy, though arguably infrastructural improvements can be presented as necessary accompaniments of private sector development. Employment subsidies to the private sector, though in a sense encouraging its growth, do so at the expense of obscuring the need for enterprises to compete in the market rather than to rely on the government. Continuing support for loss-making enterprises appears contrary to the whole idea of economic restructuring. The perception of such enterprises as industrial dinosaurs that produce inferior, unmarketable output, using obsolete technology and antiquated capital, and whose existence should be terminated as quickly as possible, is too simplistic. It is misleading to make crude comparisons with enterprises in market industrial economies taking no account of differences in endowments or relative factor prices between Russia and the West. Even as a market economy fully exposed to world prices, Russia will probably continue producing in a wide range of industries. For many enterprises, losses may be more a temporary consequence of the recession following stabilization and the weakness of the banking system, which has enhanced the risk of financial market failures, than they are an indicator of longer-term unprofitability. 5.28 It follows that there may be a case in terms of economic efficiency for continuing to support enterprises that are loss-making in the short run, particularly where such enterprises have positive value added and some prospect of commercial viability in the medium term. The aim of policies to support enterprises should be to maximize employment effects at minimum cost, while not creating inefficiencies and perverse incentives. These considerations affect the design of any subsidies, the criteria for allocating finance, administrative arrangements for determining which enterprises to support, and transitional arrangements. This is not an argument for maintaining the multiplicity of current schemes that includes directed credits that subsidize enterprises and involve substantial waste and economic costs and is opaque and resistant to monitoring and accountability. 5.29 In terms of design, major distortionary effects can be avoided by establishing the following principles for enterprise support: * Fixed duration. The subsidy must be temporary, for a fixed period specified in advance. This is primarily because the case for enterprise support is basically a short-term one. The government must avoid any 129 commitment to support what may turn out to be hopeless cases indefinitely, and should encourage enterprises to restructure themselves and adjust to the new market environment. * Fixed In size. The subsidy should be restricted to a fixed amount, rather than related to the efiterprise's wage bill or to its losses (either of which would enable enterprises to increase its subsidy by raising its wages). Given the instability of inflation, however, it would need to be linked to some index, such as the minimum wage or the unemployment benefit. * Conditional on employment. 7t is obviously necessary to ensure that the enterprises actually do maintain the employment levels for which they receive a subsidy. There are arguments for a predetermined level of employment, but in Russian circumstances probably a stronger case exists for allowing enterprises to 6lter their employment levels ex post, with the subsidy payment related I& the number actually employed up to a given level. For example, a% enterprise might receive a subsidy of x percent of the minimum wage gbr each full-time employee up to the number employed at the time the subsidy was agreed. * Take the form of budget grants, not directed credits. The financial system, especially the banking sector, should not be forced to assume quasi-fiscal responsibilities of the government. Enterprise support is best provided in the form of direct budget transfers. 5.30 Finances should be allocated to FES offices in areas of (actual or potentially) high unemployment, which might be defined as those where the unemployment rate is at least 150 percent of the national average, or would become so because a particular enterprise proposed to close. An area should be the size of a local labor market (that is, based on the distances people can be expected to travel to work), which might correspond to oblasts or municipalities in urban areas and to raions in rural areas. The purpose of the policy is to be selective between localities--and those with the greatest economic difficulties will be least able to finance support schemes from local resources. Hence, finance would generally be provided from the central budget, although there might be ci.cumstances in which subnational authorities could reallocate resources interna iy (for example, an oblast with a limited area of severe economic depression=might support enterprises there from its own rather than from central resoui. -es). In this event, the above principles regarding duration, conditionalityvand so forth should still apply. 5.31 The administration of the schemes would be in the hands of the local employment office, whose job it would be to allocate the funds in the most effective way. It would need to institute criteria for determining which enterprises to support. These conditions should be objective, although it will be difficult to lay down fixed operational rules in a time of rapid change. To some extent procedures will have to be discretionary. Enterprises could be supported only if they were on the brink of closure or had announced, though not yet implemented, proposals for mass layoffs. The basic principle of maximizing employment at mirnimum cost implies some cut-off cost per job figure. The employment office would either negotiate an arrangement with 130 individual enterprises, or it might, in the manner of the Treuhandanstalt in East Germany, invite bids from enterprises prepared to preserve jobs in return for financial support. The local employment office would be accountable to the central government for the use of the funds, which could only be spent on schemes meeting the cstablished criteria. The proposed scheme is designed to replace, rather than to augment, current activities of the FES that fall under the heading of job creation. Augmentation which does not yet reflect any clear and consistent rationale, represented about one-fifth of Employment Fund expenditures in 1992 (see Box 5.1). 5.32 The current situation of widespread and indiscriminate support for enterprises creates peculiar difficulties for the transition. The government, faced with massive insolvency across the state sector, will need to make immediate decisions on industrial policy. Clearly, a vital element of a stabilization program is that any support be selective and transparent. The role of the employment offices is to come in, as a second string, so as to cushion the worst of the regional unemployment repercussions of industrial restructuring. Enterprises from which central subsidies have been withdrawn will in the first instance be expected to face the verdict of the market, but those that cannot survive unaided and operate in an area of unusually high unemployment might be considered for support by the FES. In the initial period following stabilization, it may not be clear which areas will be worst affected by unemployment. A policy for contingency funding could be devised to enable employment offices to support enterprises in the event of an economic emergency. This contingency fund should be accessed only if and when the local unemployment rate exceeds 10 to 15 percent. To the extent that the Eastern European experience is relevant, a rise of this magnitude over a period of about a year could be expected showing sufficient regional variation to enable problem areas to be identified (see Box 4.4 in Chapter IV). Housing and Labor Market Distortions 5.33 Housing in the Soviet Union was provided practically free of charge and complemented the system of low cash wages. However, state provision has left a legacy of chronic shortages reflected in long queues and quality problems made worse by poor maintenance. At present, urban housing is still built, allocated, and managed by the state and state enterprises. It is argued that persistent distortions in the housing markets of the former socialist economies of Russia and Eastern Europe impede adjustment toward efficient labor markets by reducing labor mobility. 5.34 This section shows how housing shortages and the housing allocation system (waiting lists, the implicit ownership rights of state stock renters, and the provision of housing by enterprises) have kept inter-regional household mobility at very low levels. A brief overview of recent trends in the housing sector precedes an exploration of the linkages between the housing and labor markets, and in particular, the role of enterprise housing. A series of recommendations are presented aiming, in particular, at increasing mobility in the existing stock of housing. 131 The Housing Sector: Past and Present 5.35 The most pronounced sectoral trend since 1990 has been a sharp decline of housing output. A 20 percent drop was registered in 1991, with a further 34 percent decline during the first half of 1992 over the same period of 1991. This has affected all the major housing subsectors--state, enterprise, cooperative, and individual construction--and is attributed to high and rising costs in the construction sector, major cuts in public outlays, and the collapse of the state supply system. There has been a shift within state housing investment from direct budget finances to enterprise funds, and (to a lesser degree) to non-state investors. The contribution of direct budget financing to the total new stock fell from 81 percent in 1986 to 13 percent in the first half of 1992. At the same time, the share of enterprises increased from 7 percent to 46 percent (see Table 5.2). Table 5.2. Structural Changes in Financing of New Housing (percentage of floor space) 1992 1995 Source of financing 1986 1989 1991 preliminary variation I/II Total new stock 100 100 100 100 100 State-financed, from 80.5 79.5 73.9 70 56 / 51 Central budget 80.5 39.2 28.4 13 9 / 13 Other state - 40.3 45.5 56 47 f 38 Enterprises - 40.3 45.5 46 37 / 26 Local budgets - - - 10 10 12 Housing cooperatives 5.4 5.3 5.0 4 4/ 8 Individual 5.9 8.4 9.7 10 14 I 23 | Other l 8.2 6.8 11.4 16 26 I 18 a/b For 1992, August projections (with 10 percent fall of housing output). Forecasts for 1995 are based on optimistic (twice the 1991 output level) and pessimistic scenarios (approximately 90 percent of the 1991 Levl). Source: Ministries of Construction and Economy. 5.36 Since 1991, on- and off-budget housing subsidies have grown rapidly, largely due to rising maintenance costs. In 1992, housing is expected to represent at least 6 percent of all federal (consolidated) budget expenditures, and housing investment about 25 percent of all budget-financed investments. In 1992, financial responsibility for the operation of the existing stock was passed to municipal governments.1 Since price liberalization excluded housing rents (which remain frozen at their 1928 nominal levels), the coverage of maintenance costs by rental revenues has deteriorated sharply, from nearly half of all maintenance in 1989 to less than a third in 1991. At present, the coverage is only 3 to 5X. T The terms "mnicipal" or "local" are used throughout this section to dote the Ioest level of governmnt, that is, the level of city and raion. 132 5.37 In mid-1992, rents constituted barely 0.5 percent 'f expenditures of urban working households. Yet under current economic conditions, the transition to a situation where most households pay full market price for their housing will be extremely difficult. Recent estimates indicate that if rents in Moscow were raised to cover even half the cost of maintenance provision, severe hardship would be imposed on poor households.2 The difficult questions surrounding rent reform and housing allowances are not addressed in this chapter. 5.38 Housing shortages in Russia are mostly an urban phenomenon, while in rural areas the problem is more of low quality. Crowding indicators, at an average of 16.4 metersz per capita floor space, are the worst (after Rumania) among the former socialist countries of Europe. Another indicator of scarcity is the housing waiting lists.3 Even though one must live in conditions of serious overcrowding before signing up, the number of households officially "waiting" expanded from 12.7 million in 1986 to 14.5 million by the end of 1990 (with an additional 1.6 million on the cooperative lists, or 26 percent of all urban households). Almost a fifth have been on waiting lists for over a decade. How Do Real Estate Markets Work? 5.39 In Eastern Europe, as in Hungary and Poland, the marginal free housing market involved a range of grey (and, to a lesser extent, black) activities, and had long coexisted with various forms of state housing. Private real estate agencies were legitimate businesses, and the grey activities consisted chiefly of hard currency transactions and underreported prices. In Russia, by way of contrast, housing markets have been much less important. This is due to a combination of the practically nonexistent private sector in the Soviet economy and tighter restrictions on the ownership rights of individually owned and cooperative housing, as well as direct policies controlling population mobility (the propiska system). 5.40 Until recently, housing markets (besides markets for the vacation houses, or dachas, which have developed around large cities) were largely confined to the very limited exchange and rental markets. Such transactions were permitted by law, although they were highly restricted. While most regulations remain, enforcement has started to vary increasingly across the country. Unfortunately, data on the volume of these transactions, in particular private rentals, is very scarce. 5.41 The foundations of a housing market have been established by three major laws--the 1990 Law on 0-nership, establishing the private ownership of housing; the 1991 Law on Privatization of the RSFSR Housing Stock; and the December 1992 Law on Basic Principles of Federal Housing Policy. In 1991, the first housing auctions were organized, although few units have been auctioned ' Struyk, and others (1992). This data may include numerous instances of double counting, since members of the same household may enlist in their municipality of residence as well as in their places of work. 133 outside Moscow. Generally, auction activities have been decreasing. The buyers are mainly enterprises, and prices are extremely high. Real estate agencies have been established, and newspapers carry extensive transaction reports as well as listings, the most common source of information for those seeking to buy or rent. Better apartments in Moscow and St. Petersburg are typically sold for U.S. dollars.4 These prices exceed the ability to pay of all but the most affluent. Although financing can theoretically be obtained at an 80 percent annual interest rate (still negative in real terms), practically all of these transactions are realized in cash. 5.42 The most common way to obtain housing has been to join a queue, either at the place of residence for local government housing or at the place of work for enterprise housing. Not everyone has access to state housing. Localities have established threshold norms, expressed in square meters of living space per person, to qualify for improved living conditions, that is, the right to join a housing queue. Depending on the region, one can wait from several to over fifteen years (in large cities). In addition, migrants must wait a number of years, which vary across the country, before they are allowed to even sign up. Labor contracts witi special categories of workers, which include immediate occupancy of enterprise housing, are an exception. 5.43 State-owned units are allocated to households on a rental basis with no time limit. The implicit ownership right embodied in the right of permanent occupancy is an important feature of the system. The termination of a rental contract is virtually impossible. Until recently, eviction (which necessitates provision of alternative dwelling) was only allowed in the event of serious misconduct. The December 1992 law on federal housing policy allows eviction in cases of persistent (six months) non-payment of rent or utilities.5 5.44 The July 1991 Law on Housing Privatization provides this option to tenants. Continuing uncertainties, however, are likely to hamper the process, including the future of rent control, the ownership of urban land,6 and the anticipated introduction of property taxes. As of early 1993, about 8 percent of all state- owned rental units were privatized nationally. Moscow has privatized the highest share of its stock (13 percent), but in the majority of oblasts the share remains negligible. In December 1992, the national law was amended to become a give-away transfer, which should accelerate the Drocess. ' Now municipa. apartments in Moscow are sold at auctions, the price range in December 1992 was R 150.000 to R 220,000 per meter. Alternative dwelling still has to be provided, but the only requirement is that, instead of meeting the social norms, it must reach standards acceptable for hostel-type accommodations. * A constitutional amdment allowing private ownership of urban land was passed in December 1992, but its direct impact on housing privatization is unclear since it does not expressly include land on which privatised buildings are locted. 134 Figure 5.1. Structure of Housitg Ownership, 1991 MWt* o.m.'- Note ExUOtue hWt*te * O"dontntko,te Et'ttaee "WSud WtOtho VW Otheg COoOPW.IW tWOe at e,'tetfte.e 5.45 Renting and subletting has long been allowed, the main restriction being local per capita space norms. Even though many households occupy apartments under much worse conditions, this regulation can be used by local authorities to control relocations. Rents exceeding that paid by the. head tenant are deemed illegal income. In fact, charging much higher rents has become fairly common practice, especially in areas where labor shortages and acute housing shortages coincided. At present, free market rents are being charged for privately owned units.' 5.46 For individuals wishing to relocate for any reason- -including better job prospects- -the exchange market is often the only feasible alternative, particularly for those migrating to another city. As with private rentals, exchange activities have developed on the boundaries between legitimate activities and the grey and black markets. Housing exchanges are legal, and subject to many restrictions. Data regarding the volume of exchanges is difficult to obtain. It is estimated that some 110,000 exchanges take place in Moscow each year, involving about 4 percent of all the city's state-owned stock. A survey of housing exchanges between Novosibirsk and the rest of the country, conducted in the mid-1980s, indicated that the volume increased 1.5 times over the preceding decade. 5.47 Sublets, privatization, and to some extent exchanges each raise similar concerns in that they allow households, especially those who received In Mosoow. rents for weU.-Iooeted studio apartments vary from R 8,000 to R 10,000 per month; the same apartments rent for R 5,000 in Volgograd, R 3,000 in Saratov, and R 6,000 to R 12,000 in Y.katiuburg. in mid-1992, wben the average monthly household income was approximately R 6,800. 135 preferential treatment under the old system, to capitalize on the state subsidy. Yet placing a ban on subletting would be costly, not only to enforce, but, more importantly, it would sharply reduce the availability of rental housing and raise prices. Measures such as those recommended in paragraph 5.77 below could encourage exchanges and reduce windfalls. Linkages Between Housing and Labor Markets Labor Turnover Rates: Too Low? 5.48 While labor turnover in pre-reform Russia has been regarded as fairly high by some, there seems to be no consensus on the issue. Turnover rates, even after declining in the late 1970s, were persistently above the 7 to 9 percent limit set by planners. Unfortunately, because of methodological differences, Russian turnover rates cannot be directly compared with those of other former socialist countries with labor shortages. World Bank analysis of the labor market in Poland suggested that in view of these shortages, labor turnover rates of about 20 percent were indeed too low. Whether labor shortages enhanced turnover is, however, not clear; in fact, Soviet data for the 1967-81 period show that industrial turnover tended to decrease as labor shortages increased.8 Lower turnover rates may be accompanied by higher intrafactory downward mobility. Workers' willingness to accept lower paying or lower ranking jobs rather than to be dismissed is likely to be linked to enterprise benefits like housing, which reduce their external mobility. 5.49 In 1991 and 1992, despite the shrinking economy, voluntary quits remained, the main reported form of labor turnover (see Chapter IV). On the one hand, one would expect voluntary separation rates to be affected by housing scarcity directly, given enterprise provision of housing, and indirectly through the suppressed rate of inter-locational mobility. On the other hand, some turnover may be due to a search for better benefits (including housing) when wage differentials are flat.9 5.50 A sharp drop in the turnover rates in the early 1980s followed the reintroduction of co.mpulsory measures of labor allocation. This trend was reversed in the second half of the decade, when these measures were largely abandoned and new legislation allowed income-earning opportunities outside the state sector. The higher rates may reflect some abatement in mobility constraints, such as the absence of proplska requirements for non-state employment and a less strict compliance on the part of state enterprises. Internal Migration 5.51 Efforts to control employment flows extended to a direct monitoring mechanism, as well as an implementation tool, in the form of the propiska. The Malle 1987. 9 Standing (1992) found housing to be one of the causes of voluntary quits in 1991. Housing's share in reasons for resignation was low in state enterprises (21) end the highest in coopeative-type enterprises (8). 136 system, which essentially requires that permission be obtained from the authorities before relocating, has worked in conjunction with housing and employment laws to control migration flows. The propiska is also a prerequisite for access to many locally provided services. 5.52 The propiska is issued automatically to people born in the municipality as well as to immigrant spouses, but may be very difficult for others to obtain. Requests for propiska are screened by local authorities, since all levels of government can enact regulations affecting theii area of jurisdiction. The degree of difficulty varies across locations, the most attractive cities having the most restrictive rules. 5.53 As a rule, employers must not accept job applicants without a local propiska. However depending on their importance to the local economy, enterprises are granted quotas for bringing in outside labor when the local labor market fails to supply workers, in particular for low-paying and low-status jobs. These workers are subsequently housed in hostels with a temporary propiska. In Russian cities, over 5 percent of the population resides in hostels (as compared to only 0.3 percent in Poland). In Moscow non-resident workers on temporary propiska rose from 25 percent of all employees in 1974 to 33 percent in 1980. Eventually, hostel residents can apply for a permanent propiska. Unemployed workers can only register with the FES in the municipality of their propiska. 5.54 Since the propiska is tied to its holder's dwelling, households wishing to relocate must find housing. The problem is a difficult one, since in principle a person cannot obtain one without the other. Propiska is required not only for the actual allocation of state-owned or cooperative housing, but also as a condition to sign onto lists. Options available to immigrant households are limited given the cost of free market rental and the requirement that home buyers hold the local propiska. It is not surprising that a black propiska market has emerged in many large cities.'0 5.55 The basic premise of a bill "On the Choice of the Place to Live of the Citizens within the Territory of the Russian Federation" is the individual right to freedom of relocation, although compulsory registration would continue. Both the draft law and the parliament's resolution on its implementation would render interference in private relocation decisions illegal. However, it is not clear whether the proposed scheme would be followed by repeal of housing restrictions. The draft law provides that registration would be issu,ed (upon the presentation of the required documents) to the buyers of private housing as well as renters, which is a step forward, but the issue of local housing norms and standards remains to be addressed. Moreover, the draft implies that vaguely described "justifiable" restrictions may be imposed. These are likely to be used to limit inflows to areas where the welfare of existing residents is perceived to be threatened. Since 1991, the city of Moscow has been selling propiska to enterprises to enable them to bring in desired workers. In late 1992, the auction price of Moscow's propiska was approximately R 1.5 to 2.0 million per household. 137 Have Migration Rates Been Depressed? 5.56 Soviet migration and labor allocation policies were largely ineffective at achieving the redistribution of labor resources between surplus and deficit regions. Likewise, the growth of large cities, with their ostensibly strict migration controls, was not contained. In the decade before 1989, Moscow and St. Petersburg and surrounding regions, while accounting for only 8% of the Soviet population, absorbed 451 of all migrants. On the other hand, the growing concentration of people on city fringes over the last thirty years appears to be linked to the propiska and urban housing shortages. This has not been the western scenario of "suburban flight," but rather a second best solution for rural migrants seeking access to the vastly superior provision of goods and services in larger cities, yet who were prevented from becoming city residents because of the propiska and lack of housing opportunities, Suburban dachas have helped to meet the demand for accommodation. 5.57 Even if many migrants managed to circumvent residential restrictions, the expected outcome would still be distorted patterns of internal migration, including depressed migration rates. Establishing the magnitude of these effects is very difficult. Over the years, the collection and publication of data on migration was periodically suppressed. Officially published statistics collected on the basis of the propiska system may have significantly underestimated mobility. These problems compound the already difficult task of determining mobility rates. 5.58 Although international comparisons of internal migration are difficult, migration specialists agree that rates in the centrally planned economies tended to be much lower than for market economies. In 1971, a year for which tentative comparisons are available, Poland, Hungary, and Czechoslovakia had net internal migration rates that were approximately 2.4 times lower than those for the United States and West Germany. The official internal migration rates for the U.S.S.R. during the 1970s and 1980s were slightly lower than Poland's. The net urban migration rate per 1,000 people in the U.S.S.R. was 7.4 during the 1976-79 period and 5.1 between 1980 and 1988. The corresponding figures for Poland are 10.8 and 6.3. 5.59 A housing opportunities approach shows mobility rates in Western Europe where government intervention in housing markets is stronger, to be on average half of those in Australia, Canada, New Zealand, and the United States. The figures are based on the residential mobiliLy rate, defined as a percentage of the population moving from one dwelling to another during a given period of time. While no such data exists for the former U.S.S.R., national housing experts estimated the rate to lie between 3 and 4% (see Figure 5.2). 5.60 Mobility could be affected by unmeasured characteristics of housing markets, such as their openness and the ease of housing exchanges. This is likely to be true for Eastern Europe and Russia. Given serious housing scarcities, a relationship between shortages and mobility could also be expected. Although such an association was found by Mayo and Stein (1987) for Poland, the results of a similar study in Russia have to be treated more cautiously. Karel et al. (1988) concluded that labor is deterred from moving to regions with increasing housing shortages. During 1966-85, migration inflow areas had higher 138 housing output than the outflow regions. In their all-republic regression model of net migration per capita, residential construction was the single most important explanatory factor, accounting for 44 percent of variation in the dependent variable. However, in the migration volume (in-flow and out-flow) models, the same housing variable has revealed little explanatory power. This may be due to model misspecification, even though (according to numerous surveys) "better housing opportunities" shows consistently as an important motive for migration. Another plausible explanation lies in the significant numbers of housing exchanges; these would have a negligible effect on net migration, but would certainly affect volume. 5.61 Recent data suggest that rates of internal migration have declined further. The actual scale of anticipated massive refugee flight has so far been moderate: at the end of 1991, there were approximately 200,000 refugees in Russia (much less than anticipated). In any case, an influx of refugees would certainly exacerbate housing deficits, together with the problem of army demobilization. Enterprise Housing Benefits and Labor Mobility 5.62 Over the last two decades, enterprises--especially those located in areas with persistent labor shortages, as in the north and east--came to rely increasingly on the provision of housing to attract workers. At present, enterprises manage about 57 percent of the total state-owned stock. Regional differences in the share reflect previous priorities in industrial policy, being highest where heavy and defense industries dominate. Enterprises currently dominate new construction. 5.63 Housing Benefits. Enterprises may offer rental housing, including various accommodations with shared facilities. There is also an array of in-kind and cash forms of housing assistance for which data are largely anecdotal. These benefits do not constitute an entitlement. In the past, legislation encouraged, but did not oblige, enterprises to extend housing assistance to employees. This could involve mediation in securing land plots. Cash assistance takes the form of rental allowances, coop downpayment loans, and partial reimbursement of housing debt repayments. Until recently, the Sberbank (Savings Bank) housing loans for individual construction were channeled through enterprises. 5.64 The World Bank survey conducted in November 1992 shows that most enterprises (75 percent of respondents) offer some form of housing assistance. Enterprises that do not tend to be very small. More than 1/3 of enterprises providing housing assistance have at least three different housing programs. Help with rent payments was rarely available and this practice is reported to have been largely discontinued. Programs furnishing land plots are the most common type of housing assistance and are available in more than half of the enterprises. About one-fourth offered help in both acquisition of land plots and construction loans. 139 Figure 5.2. Estimated Percentage of Population Changing Usual Residence in One Year, 1981 Selected Industrial Countries AUSTRIA 8 FRANCE SWEDEN _ __ AUSTRALIA JI UNITED STATES CANADA I I l 18 NEW ZEALAND |_19 SOVIET UNION _ O 6 10 16 20 26 Source: L. Long, 1991; Internal World Bank Reports 5.65 The provision of housing by enterprises has arguably been one of the most important factors affecting labor mobility. This has at least two dimensions: (1) conditioning access on length of employment, and (2) reducing inter- locational mobility, as workers even once retrenched are reluctant to leave their current apartments. Enterprise housing generally depends on the length of the individual's employment in tl-e e-nterprise. The waiting period varies greatly across branches and regions, and could well exceed ten years, but would nevertheless still be shorter than the alternatives. Workers whose housing conditions would fail to meet the criteria for access to municipal housing may qualify for enterprise lists. An employee's qualifications and contribution to the company may entitle the family to an apartment of better quality than municipal norms. 5.66 As mentioned above, enterprises exercise vary limited control over their existing rental stock. A permanent occupancy right is acquired that in principle cannot be terminated upon termination of employment. Among erterprises reporting permanent housing in the Bank survey, all have tenants who are not their employees. Indeed, nearly 40 percent of current tenants were not current enterprise employees. The occupancy of enterprise housing by families not 140 receiving an income from the enterprise creates a serious problem for housing reform. The real value of housing services is likely to be such a large proportion of the families' real income that it will be impossible simply to raise rents without increasing money incomes substantially. Where tenants are employees, there is an option of raising rents to a market level and compensating those concerned by paying higher wages, but this is not feasible for all tenants. Enterprise Housing in the Housing Reform Debate 5.67 Recent policy statements and legislative draft proposals, in particular, the recent Law on Basic Principles , state that enterprises and public entities providing housing for their employees should be the principal players in the housing delivery system, at least during the transition (see Table 5.2 forecasts). Whether enterprises are willing and able to remain in this role is unclear. Fiscal incentives encourage employer provision of housing benefits through federal and municipal tax exemptions and rebates. Since 1991, all housing investment financed from the enterprise social funds is exempt from profit tax. 5.68 Recent legislation has left the actual ownership of the existing housing stock of enterprises unclear. It appears that the ultimate control over enterprise housing remains with the corresponding level of government. It is legally subject to privatization upon individual occupant demand. The Bank survey found that privatization of apartments had begun in more than half of the enterprises, involving between 1 and 13% of all apartments. One small company had already privatized its entire housing stock. Elsewhere, there are reports that some enterprises have sought to obstruct the privatization process. There have also been incidents of enterprises abandoning their housing, in which case the stock would be transferred to municipal authorities. Some enterprises have begun to shift from direct low interest loans and grants for employees who are engaged in construction or are members of housing cooperatives. 5.69 Enterprises, in the short run, may not tend to suspend housing programs. Beyond the impact of fiscal incentives, the proposed enterprise privatization law vests employees with substantial managerial control. Consequently, it is suggested that housing would be one of the last benefits to be discontinued, even if the scope of the programs is curtailed and their character changed. The Bank survey found that more than one-third of enterprises offering housing benefits show capital expenditures in 1991 or 1992, and the same number of companies indicate that they have no plans to reduce their construction programs. Only two enterprises declared plans to discontinue all housing programs. In fact, over one-third of enterprises listed housing assistance as the last benefit (among all others provided) that both management and workers would relinquish. If alternative options to obtain housing become less accessible, it will become an increasingly important enterprise benefit. At the same time, significant uncertainty shrouds the factors that affect housing programs, the most important being rising construction costs, legislative reforms regarding rent and the tenancy rights, the implications of regulations allowing private land ownership, privatization, and the rights of persons on housing lists. 141 Policy Recommendations 5.70 Not only is the pace of the transition to a market-based housing system in Russia more severely constrained by macroeconomic and political factors than in Eastern Europe, but it embarks from a much higher degree of state monopoly over production, distribution, and management of housing. In any case, as emphasized above, it cannot be expected that housing policy reform would alleviate problems of high regional unemployment in the short run. This section highlights policy reforms that could have a significant beneficial impact on household, and therefore labor mobility. Annex 3 looks at a wider range of sector reforms, including measures to revive new construction. 5.71 The Need te Emphasize Coordination and Implementation of Reforms. Fueled by the apparent consensus for the general market direction of housing reforms, the legislature has introduced several piecemeal measures without a coherent, long-term policy framework. A fragmented approach has contributed greatly to the prevailing confusion. Uncertainties as to the status of existing regulations and the interpretation of new laws and decrees, and the absence of clear implementation procedures, create a situation in which local authorities, enterprises, and banks, as well as individuals, find it very difficult to make informed choices. Lack of coordination among those active in housing policy- making and delays in implementation could inhibit the labor market transition and industrial restructuring. 5.72 The first attempt to develop a framework for comprehensive housing reform, the Law on the Principles of Federal Housing Policy (December 1992), reconfirm the right to private ownership of housing, and the introduction of private ownership of urban land is welcomed. Embracing a set of broad guidelines for future actions, the law should accelerate the development of specific legislation in key areas of housing policy and encourage a much needed emphasis on implementation. 5.73 Elimination of the Propiska System. The proposed legislation On the Choice of the Place to Live of Citizens should be implemented as soon as possible. Beyond direct control, the propiska system impedes private transactions by significantly raising transaction costs. All housing restrictions, beyond those required for sanitary considerations, should be abolished. Given fiscal constraints, some local governments are likely to use the clause allowing "justifiable" restriction to limit immigration. This should be avoided. 5.74 Emphasize Changes for Existing Stock. While reforms aimed at existing housing and new construction should be considered simultaneously, those targeting the existing stock have the potential to improve labor mobility relatively quickly. Attempts should be made to achieve more efficient use of state-owned housing through more rational privatization and rent policies, better management of repossessions and reallocations, together with repeal of restrictions on exchanges and sublets. 5.75 Escalating housing maintenance subsidies are unsustainable. If rents are not adjusted and revenues continue to fall rapidly in real terms, municipal governments and other owners of state housing are likely to cut furthter already low maintenance budgets, which in turn will cause rapid deterioration of the 142 stock. Rent increases will both raise revenue and encourage mobility through privatization and exchanges. Russia has the lowest cost recovery levels on state rental stock and the lowest rent burden on household incomes of all the former socialist countries of Eastern Europe. Yet strategies for rent adjustments have to be developed in the context of falling real incomes. Almost any meaningful increase of the basic tariff will have to be accompanied by housing allowance programs, so net budget savings may be very small. 5.76 Options for rent reform should combine rent adjustments and housing allowance schemes.11 Programs should not be standardized across the country. The Russian Federation, unlike most Eastern European countries, has recognized the local aspect of housing by empowering municipal governments to adjust rents. Other options to be considered concurrently with the basic rent tariff increase include: (1) implementing higher rent increases on housing space exceeding the "social norm", possibly with a downward adjustment of such norms to minimize the per unit subsidy; (2) further differentiation of tariffs to account for locational advantages; and (3) targeting the best parts of housing stock for faster rent increases, possibly covering at least current maintenance. 5.77 All remaining legal and administrative restrictions on sublets, and exchanges in municipal housing, such as prescribed floor space norms and bans on cash payments, should be revoked or redefined so that only reasonable safety and health codes apply. Despite existing constraints, apartment exchanges have been relatively frequent and their number should increase with rent adjustments. Exchanges contribute significantly to both local and interregional mobility and should be facilitated. 5.78 Privatization. In late 1992 the housing privatization program shifted to a free transfer, thereby accelerating the process. The demand for transfers would also be stimulated by implementing rent increases, property taxes (uncertainty about their levels is often reported to delay privatization decisions), and condominium law for privatized or partially privatized buildings. Since urban land can now be privately owned, owners' rights to land on which privatized buildings are located need to be clarified. 5.79 Privatization of housing will have a direct impact on mobility as increased numbers of units are placed on rental and resale markets. New construction could be stimulated, especially if owners can use their privatized dwelling as a collateral for housing loans. It may also contribute to the transition in a different way, if privatized dwellings can be used as a collateral for small business loans. 5.80 Enterprise Housing. Although enterprises should eventually be freed from housing responsibilities, immediate withdrawal would cause an acute crisis. Enterprises currently account for approximately 65 percent of financing of new housing stock as budgetary allocations continue to fall and the private sector is undeveloped. New tax legislation creates incentives for enterprises to provide housing benefits and, as noted, the enterprise privatization law gives ' Housing allowance programs are currently in use or under consideration in several Eastern European countries. The first program in Russia is being designed for Moscow. 143 substantial control to workers, who are likely to support continuance of housing programs. 5.81 During the transition, while enterprises continue to play an important role in the provision of housing, enterprises should exercise more control over their housing stock. This implies the right to determine rents, establish one's own policy with respect to rentals and exchanges, and to evict workers after the termination of employment. The occupancy of enterprise housing by families not directly engaged by the enterprise creates an obstacle to housing reform. Such changes should at least apply to new units. In addition, incentives could encourage enterprises to shift away from direct provision to other assistance programs, such as short- to medium-term downpayment loans. This could improve labor mobility by creating the potential for greater flexibility and by supporting private ownership and thus the development of housing markets. 5.82 Conversions of Dachas. Recent legislation encourages the conversion of "temporary" homes meeting normative requirements into permanent residences. This could significantly expand the existing stock of housing in areas where shortages are most acute. 5.83 Waiting Lists. Despite what the privatization law implies, it is unrealistic to expect that free housing will continue to be provided to households on waiting lists. It is also obvious that waiting periods will expand tremendously, with no ultimate guarantee of housing. Some households will simply lose hope and drop out of the queue to seek other options, as has been the case in Poland. Poland has sought to shorten its waiting lists further through a program whereby households are offered preferential access to land and a construction grant, though it is unclear how many households have taken advantage of this offer. Similar programs could be considered in Russia. 144