Chapter IV.6 Labor Market, Social Safety Net, Education and Training

International Monetary Fund
Published Date:
December 1991
  • ShareShare
Show Summary Details


The transition to a market economy in the USSR will involve significant adjustments and social costs, and will require putting in place effective labor market and social policies to both support the transition process and sustain the political consensus in favor of reform.

In the past, the key instrument in the USSR for providing social protection was a formal job guarantee written into the Constitution. Although such a guarantee is not possible in a market economy, through a set of appropriate policies it should be possible to minimize the costs of unemployment and to guarantee minimum incomes and standards of living. Not surprisingly, recent reform proposals leading towards a market economy have included eliminating the employment guarantee and introducing various forms of income compensation.

Open unemployment will be a completely new phenomenon in the USSR and there is considerable concern about its social and economic consequences. The starting point for the analysis, therefore, must be the labor market situation. Past and possible future labor market developments are discussed in section 2. The principal conclusion is that labor shedding could become significant if and when enterprises are confronted with a hard budget constraint. Section 3 deals with incomes and living standards. The main finding is that the traditional system of administered wages and prices and the available social transfers have been inadequate to prevent poverty or to ensure access to the basic goods and services desired by the population.

The final section reviews a range of policies involving labor markets, education and training, wage formation and social protection which could help support the reform process. It is suggested that in many cases new institutions or infrastructure to deliver such policies will need to be established. This will necessarily take time and will impinge on scarce budgetary resources. However, postponing action now would further delay the normal functioning of a market economy. The key requirement for the pursuit of effective labor market policies will be the build-up of a nation-wide public employment service. In the vocational education and training area, a reasonably well-developed infrastructure is available but needs reform to respond to the emerging skill requirements of a market economy. The system of industrial relations has to be completely overhauled and requires new institutions and “rules of the game”; there is a particular urgency to develop collective bargaining structures. But this will not occur overnight. In the meantime, an effective incomes policy is needed to contain wage cost pressures and to support macroeconomic stabilization.

The social protection of weak and vulnerable groups, including the unemployed, is of particular relevance for cushioning the immediate adverse effects of the transition process. A number of options for social safety nets are also discussed in section 4. Liberalizing the prices of basic food items would result in a severe drop of living standards for those on low and fixed incomes. Measures to protect these groups could include cash transfers, the provision of specified quantities of certain essential goods at given prices, or food-stamps. The method of delivery may be in-kind, universal or targeted. The effectiveness of the cover provided and the consequent budgetary implications of these options are examined.


a. Labor force and employment

(1) Labor supply trends

The dominant trend in the Soviet labor market over the last decade has been a sharp slowdown in the growth of the labor force (Chart 1 and Table A.1, Appendix II-1); the average annual rate of growth in the second half of the 1980s was about one third that of the second half of the 1970s. While the growth rate of the Soviet labor force in the 1980s was roughly in line with that of Eastern Europe, it was less than one half the average rate achieved in the OECD countries (Table IV.6.1). Up to the late 1960s, buoyant labor force growth had been maintained by a combination of moderate population growth and rapidly rising participation rates. But participation rates stagnated after the early 1970s and labor force growth since then has been determined mainly by the growth of the working-age population. While the working-age population grew rapidly, at almost 2 percent per annum in the 1970s, a contraction in fertility rates and an increase in death rates caused a major slowdown in the 1980s.


(Average annual growth rate)

Source: Data from the 1959, 1970 and 1979 Censuses are from M. Feshbach (1983). Data from the 1989 Census were supplied by Goskomtrud.

1. Men: 16-59; Women: 16-54.

Table IV.6.1.USSR: Labor Force Growth in International Perspective, 1980-89(Annual average growth rate, in percent)
United States1.6
United Kingdom0.8
OECD Europe1.0
Total OECD1.3
Sources: Goskomtrud; OECD; United Nations, Economic Commission for Europe, Economic Survey of Europe in 1989-90 (1990).
Sources: Goskomtrud; OECD; United Nations, Economic Commission for Europe, Economic Survey of Europe in 1989-90 (1990).

Labor force participation rates in the USSR have traditionally been higher than in most Western countries. Measured as the total labor force divided by the total population aged 16-64, the standardized labor force participation rate in 1988 was higher in the USSR than in any of the G-7 countries (Chart 2).1 However, participation rates in several small Western countries, notably the Nordic countries, were significantly higher than in the USSR. High female participation rates have tended to push up the overall participation rate in the USSR, while a low—by Western standards—pension age (men: 60 and women: 55) has tended to depress it. Indeed, in contrast to most Western countries, the participation rate for working-age women in the USSR is only marginally less than for able-bodied men. The gap between female participation rates in the USSR and in the West has narrowed over the last two decades, as rising rates in the latter have begun to catch up on a stable Soviet rate. However, the number of working hours supplied by Soviet women on average is much higher than in the West. This reflects the fact that part-time work is almost non-existent while it has expanded rapidly in most Western countries in recent years. Part-time work is concentrated among female workers in the OECD countries, while about 99 percent of Soviet working women work full-time. 2


Source: Goskomtrud; OECD (1990a)

1 Total labor force divided by the population aged 15 to 64

2. 1989

The 1980s witnessed an increasing tendency for old-age pensioners in the USSR to remain in the labor market beyond retirement age. According to the 1989 census, 17 percent of all old-age pensioners continued to work after reaching pension age, 5 percentage points higher than in 1979. This increase seems to be linked to the low value of retirement pensions, lack of indexation and the strong demand for labor. During the 1980s, the value of retirement pensions fell relative to average wages and the prices of goods and services (see section 3). This, in turn, has forced many old-age pensioners to supplement their pensions with earned income.

(2) Special features of the Soviet employment structure

In line with the experience of most Western countries, there has been a large fall in the share of agriculture in total employment in USSR and the share of services has increased. Within the service sector, social and community services have recorded the largest gains. In contrast to the pattern observed in most Western countries, however, the manufacturing industry’s share in total employment has remained constant over the last two decades. There have also been relatively small changes in the importance of sub-sectors within manufacturing, with the share of heavy industry remaining very high at the expense of consumer goods industries.

International comparisons show that the employment structure in the USSR differs significantly from the structure in Western countries (Table IV.6.2). The most striking differences are the very high shares of agriculture in the USSR,3 the virtual absence of a financial sector and a very small retail and wholesale trade sector. Apart from these three sectors, the Soviet employment structure is well within the range observed in the OECD countries, notably for the two major sectors, manufacturing and community services.

Table IV.6.2.USSR: Employment by Sector in International Perspective, 1988(Shares in percent)


1. Agriculture20.
2. Mining and quarrying1.
3. Manufacturing23.618.524.232.221.623.321.417.2
4. Electricity0.
5. Construction9.
6. Wholesale and retail trade6.
7. Transportation and
8. Financing and insurance0.511.
9. Community services25.931.421.025.930.120.730.329.0
0. Not defined4.01.4
1. Primary20.
2.–5. Secondary35.226.934.141.030.337.529.625.5
6.–0. Tertiary44.670.258.053.762.952.768.170.0
Sources: ILO for the USSR. For other countries, OECD, Labour Force Statistics, 1968-88 (1990c).

Adjusted by the ILO for the purposes of international comparisons. Because of this, the employment structure for the USSR does not correspond to that shown in Table F.2 of Appendix II-1.


Sources: ILO for the USSR. For other countries, OECD, Labour Force Statistics, 1968-88 (1990c).

Adjusted by the ILO for the purposes of international comparisons. Because of this, the employment structure for the USSR does not correspond to that shown in Table F.2 of Appendix II-1.


The easing of restrictions on private employment in the latter part of the 1980s has resulted in a small, but rapidly growing, private sector outside agriculture. Until the mid-1980s, the only authorized “private” sector activities were in so-called private subsidiary farming which, traditionally, was an important source of agricultural output. Three percent of total employment is estimated to have been engaged in such activities in 1989. Genuine private activity outside agriculture was legalized in 1987 after the adoption of the Law on Individual Labor Activity, allowing individuals to be self-employed. In 1989, it was estimated that 300,000 persons were legally self-employed, about 0.2 percent of total employment. Legislation permitting the leasing of land has opened up the possibility of private farming, and about 100,000 persons were engaged in such activity in the first three quarters of 1990.

The Law on Cooperatives, adopted in 1988, allows a group of people to engage in any legal activity and also to employ people. This form of enterprise spreads quickly and is common in consumer services, retail distribution, and construction. By October 1990, there were some 215,000 cooperatives, employing 5.2 million persons (more than 3.5 percent of total employment), of which somewhat fewer, however, were considered to have their main job in such activities.

Soviet enterprises tend to be much larger than enterprises in OECD countries, reflecting the central planners’ obsession with economies of scale. There are only 46,000 enterprises in industry, implying that the average enterprise employs around 830 employees. Twenty thousand of these enterprises employ less than 200 workers. This, in turn, implies that the average enterprise with more than 200 workers employs more than 1,250 employees. Although large enterprises are also the norm in Eastern European countries,4 the size distribution of Soviet enterprises is very different from the typical profile found in Western countries. For example, establishments in France, Germany and Japan with a work force of up to 100 employees account for 50-60 percent of all private sector employment.5 In these countries, establishments with more than 1,000 employees provide jobs to only 20-33 percent of the industrial work force.

(3) Regional variation in labor supply and demand

Aggregate figures on Soviet labor supply and employment conceal important regional differences. The labor market can be split into two broad regional areas according to demographic trends and employment structure: the more developed regions (Russian Republic, Baltic republics, the Ukraine, and Belorussia) and the less-developed regions (Central Asian republics, Kazakhstan, Transcaucasian republics, and Moldavia). The developed regions are characterized by sluggish population growth and stagnating or falling employment levels (Table IV.6.3). In the 1980s, the annual average population growth rate in these regions ranged from 0.4 percent (Ukraine) to 0.8 percent (Lithuania). The ageing of the population resulted in much slower growth in the working-age group than in the overall population. The Ukrainian working-age population even fell in the 10-year period ending in 1989. The dominant feature of demographic developments in the less developed regions is rapid growth of both the overall population and the working-age population. The highest annual average population growth rates in the 1980s were recorded in the Central Asian republics (ranging from 1.9 percent annual growth in Kirgizia to 3 percent growth in Tadzhikistan). Population growth was more modest in the other developing regions, especially in Moldavia, Georgia, and Armenia.

Table IV.6.3.USSR: Growth in Total and Working-Age Population by Republic, 1979-89(Average annual growth rate, in percent)



Developed regions
Russian Republic0.70.1
Less developed regions
Source: Census data from Goskomtrud.
Source: Census data from Goskomtrud.

In the developed areas, manufacturing accounts for around 30 percent of all employment, while in the less developed regions its share is generally less than 20 percent (except in Armenia) (Table IV.6.4). By the same token, the developed areas have relatively fewer workers in agriculture, ranging from 14 percent in the Russian Republic to 22 percent in the Ukraine. In the less developed regions, the employment share of agriculture exceeds 25 percent, and is especially high in the Central Asian republics, partly due to the importance of private subsidiary farming in these regions. A surprising feature of the employment structure across the USSR is that service sectors have relatively high shares in the less developed regions. For example, community, social and personal services account for a higher percentage of total employment in the Central Asian republics than in the developed regions.

Table IV.6.4.USSR: Employment Shares by Republic, 1989










Cooperatives 2Self-EmploymentTotal
Sources: For all sectors except collective farms, Ministry of Defence, private subsidiary farms, cooperatives and self-employment: Heleniak (1990a). For all other sectors: Heleniak (1990b).

Employment in the Ministry of Defence, Ministry of Internal Affairs, Central Committee of the Communist Party, and administration of Komsomol.

The figure for cooperatives includes self-employment and private agriculture.

The figure for cooperatives includes self-employment.

Sources: For all sectors except collective farms, Ministry of Defence, private subsidiary farms, cooperatives and self-employment: Heleniak (1990a). For all other sectors: Heleniak (1990b).

Employment in the Ministry of Defence, Ministry of Internal Affairs, Central Committee of the Communist Party, and administration of Komsomol.

The figure for cooperatives includes self-employment and private agriculture.

The figure for cooperatives includes self-employment.

b. Labor market: pressures and imbalances

The crucial difference between the Soviet labor market and those in the West relates to the demand for labor. While every able-bodied Soviet citizen has a constitutional obligation to work, there is in fact considerable freedom to choose where to work. The choice is the result of economic incentives, such as wages, fringe benefits and working conditions, as well as administrative regulations governing residence permits (propiski) and the availability of housing. But on the demand side Soviet managers, unlike their Western counterparts, have not been subject to strict financial discipline in determining labor requirements; instead they have been faced with a soft budget constraint and the imperative of fulfilling the production targets laid down by the central planners. Since wages have been largely determined by a very complex and cumbersome system of centrally-determined wage tariffs, wages have played little role in allocating labor to its most productive uses.

(1) Overmanning in enterprises

There is general agreement that Soviet enterprises tend to employ more workers than are needed.6 Soviet estimates of overmanning in the first half of the 1980s, based on surveys of managers, suggested that many enterprises retained between 15 to 20 percent more labor than was necessary to fulfill their production plans, and that for some branches the figure was as high as 25 percent. Soviet figures may even underestimate the extent of overmanning in industry, since decades of excess demand for labor may have fundamentally affected the perceptions of what are necessary labor requirements. Thus, comparisons of manning levels in Western-built plants in the USSR with matched plants in the West indicate that the Soviet plants employed 50 to 150 percent more workers.

While a soft budget constraint has favored labor hoarding, overmanning has also been a rational response by Soviet managers given the constraints imposed on them by the central planning system. Their overriding objective is to fulfill the plan’s targets for outputs but, at the same time, they are continually faced with potential disruption to production schedules. Overmanning has also probably been stimulated by an overheated labor market. Lack of labor discipline (“shirking”) is a prevalent feature of the organization of labor resources in the USSR. Many official campaigns have been launched to improve the situation, but to little avail. Workers spend much of their working time arranging their personal affairs, moonlighting, and queueing for scarce goods. There is some evidence that managers have competed for workers in the tight labor market by offering less-intensive working schedules. Given that their ability to offer higher wages has been to some extent circumscribed by central guidelines, this paid “leisure” was a natural alternative way of attracting workers. Labor market pressures do not seem to have eased in the 1980s, and large numbers of vacant job places continued to be recorded at the end of the decade. However, increased freedom for enterprises to set wages in recent years has meant that demand pressures may be increasingly reflected in higher wage increases rather than in less labor discipline.

The authorities have made several attempts to encourage enterprises to release excess workers over the last 20 years, but only since the mid-1980s have they enjoyed some success. A common aim of these reforms was to give enterprises and their work forces an incentive to reduce overmanning: allocating fixed wage funds to enterprises, or to a given task, so that average wages would rise with a smaller work force; or attempting to relate bonus payments to planned labor productivity growth. The failure of these reforms was due to planners revising the wage norms as soon as productivity increases occurred, thus pre-empting any rewards workers may have gained from greater work effort.

The 1986 wage reform led to an increase in tariff wages, i.e., in the basic wage scales for state employees in individual occupations and at various skill levels, but required that these higher wages be financed by enterprises out of their own resources. The 1987 Law on State Enterprises also encouraged the linking of productivity and wage levels. These measures, notably the 1986 wage reform, are thought to have reduced overmanning to some extent. By January 1990, when the wage reform had been implemented for more than 80 percent of the 72 million potentially affected workers, about 3.3 million workers (5 percent) had been displaced. Around a third of the displaced workers found a new job within the same enterprise; the remaining workers were laid-off and 30 percent of those went into retirement. This degree of labor shedding was minor, however, compared with the scale of labor hoarding in the economy.

(2) High labor turnover

Official labor turnover rates in the USSR are not markedly out of line with those recorded in many Western countries (Chart 3). In industry, 6.5 million workers left their jobs in 1989 out of a total industrial work force of 37 million; in construction, there were 4 million job-leavers out of a total work force of 12.5 million. Measured as the total separation rate per 100 employees in industry, Soviet turnover rates in the 1980s exceeded those of Japan, France, Italy, and Sweden. They fell well short of those in the United States, however, and were slightly lower than separation rates in the United Kingdom.


Source: Goskomtrud; OECD (1986); J.C. van Ouvs (1990).

1 Number of seperations per 100 employees.

There are reasons to believe, however, that official turnover data under-record actual turnover in the USSR. Official statistics are based on surveys of large enterprises, i.e., enterprises which have independent accounting systems. In 1989, employment in these enterprises accounted for two thirds of total employment in the state sector. Experience in the West shows that labor turnover tends to be inversely related to the size of the enterprise, and scattered evidence from the USSR suggests a similar pattern. As a result, official turnover statistics probably have a downward bias of unknown magnitude.

Voluntary quits are the single most important reason for leaving jobs in both industry and construction, accounting for between 30-45 percent of all separations in recent years (Table IV.6.5). In industry, the second most important motive for job separation is to leave the labor market, e.g., for retirement, military service, return to education, or maternity leave. This is followed by termination of work assignments, disciplinary dismissals, and transfer to other enterprises. In construction, termination of work assignments, transfer to other enterprises, and exit from the labor force each account for about a fifth of all separations.

Table IV.6.5.USSR: Job Separations According to Cause, 1988-89(In percent of total)
Transfer to other enterprise6.47419.520.3
Termination of work13.212.523.319.9
Leaving the labor force29.827.620.018.6
Voluntary quits43.144.732.035.5
Disciplinary dismissals7.
Source: Goskomtrud.
Source: Goskomtrud.

The most important reason for voluntary quits—cited by a quarter of all respondents in 1987 in a Goskomstat sample survey—was the need to change geographic location. Unsatisfactory living conditions, such as bad housing and insufficient child care, dissatisfaction with working conditions, and low wages, were also cited as major factors behind the decision to quit.

The trend of recorded voluntary quits—and presumably total separations—was marked by a sharp decline in the period from 1970 to 1985, after which quit rates started to edge upwards again. Three factors seem to have been instrumental in bringing about this secular decline in quit rates. First, the ageing of the Soviet population should have reduced quit rates. In the USSR, as in the West, young workers have much higher quit rates than older workers. Second, repeated pressure from Gosplan and the branch ministries on enterprise managers to reduce unplanned turnover may have prompted them to transfer some activities to subsidiaries. Third, several administrative measures were taken in the 1980s to curtail voluntary quits. These included lengthening the period of advance notification for quits, strengthening the links between length of job tenure and entitlements to social security benefits and holidays, and temporary restrictions on the right to quit.

Given the incentive structure in the Soviet labor remuneration system, job changes have frequently not contributed to improved allocation of labor resources. One reason is the movement of skilled personnel from jobs where their skills were required to jobs which required no particular skills. Viewed from the perspective of the individual worker, such behavior may be perfectly rational; it could be a natural response to the sharp compression of wage differentials between skilled and unskilled workers (section 3). Another unwanted side-effect on labor allocation was the movement of workers from areas with tight labor markets to areas where there was no shortage of labor. Such migration flows were common up to the 1970s. Thus, labor-shortage areas, such as Siberia, had difficulties in attracting workers, whereas labor-surplus areas, such as the Central Asian republics, experienced net inflows. This could have been linked to inadequate regional variation in wages and salaries.

(3) Low unemployment despite high turnover

Official estimates of unemployment in the USSR put the number of jobless persons in 1990 at around 2 million.7 The incidence of unemployment differs markedly across republics, with the bulk of unemployment being concentrated in Kazakhstan, Central Asia, and the Transcaucasian republics.

There is little information on how either the official or unofficial unemployment estimates are derived, which makes it difficult to evaluate their reliability. One possible reason for the wide range of estimates relates to the category of “discouraged workers”, i.e., people who want a job but are not actively looking for one because they believe such search would be futile. (These individuals would be excluded from the ILO definition which requires not only that a job seeker be available for work but that he/she has also taken active steps to find work within a recent period.) This is a particularly important category for Kazakhstan and the Central Asian republics. Once an unemployment insurance system is put in place in the USSR—which is under active discussion—many of these discouraged workers are likely to register as unemployed at job placement agencies and claim benefits.

There are grounds for believing that the lower official estimate of 2 million unemployed understates the true unemployment total as it would be measured on the ILO definition. As mentioned earlier, significant labor turnover takes place in the USSR, implying that there is some search (or frictional) unemployment as part of the process of job changing. Western analysts have attempted to estimate the magnitude of search unemployment in the USSR using data on turnover together with very sketchy information on the average duration of unemployment spells. These estimates suggest an unemployment rate in the range of 1 to 3 percent for the state sector (i.e., excluding collective farms) in the mid-1980s, or between 1¼ and 3½ million persons unemployed.8 Total unemployment on the ILO definition should be higher than these estimates for two reasons. First, the estimated rates of frictional unemployment themselves are probably biased downwards because of the bias in labor turnover discussed above. Second, the estimates do not include first-time job-seekers. Some Soviet evidence suggests that this group has much longer unemployment spells than the average.

Assuming that the data on labor turnover are reasonably reliable, an unemployment stock between, say, 2 and 6 million would imply that the average duration of unemployment spells is somewhere between 1 and 3 months. Compared with the experience of most Western countries, an average spell length of not more than a quarter is relatively short. Most European OECD countries, for example, have experienced average durations of unemployment spells well in excess of 9 months.

c. Transition to a market economy: estimates of job displacement and unemployment

(1) Sources of displacement

The transition to a market economy will inevitably involve large-scale displacement of the Soviet work force. The size of this displacement will partly depend on the evolution of the macroeconomic situation but will also be determined by the following structural developments:

Reduced overmanning. As discussed above, current levels of overmanning could be around 20 percent or more; hence large-scale lay-offs should occur if and when enterprises are confronted with hard budget constraints. The extent to which they occur will depend on the cost of making workers redundant and on who bears this cost. The propensity to lay-off excess labor will also be influenced by the introduction of an unemployment insurance system: employers will be less reluctant to fire people once this system is in place.

Closure of uncompetitive firms. Moving towards a market economy will require sharp cuts in budgetary support to ailing enterprises, and increases in interest rates. Goskomtrud estimated that in 1990, 3 million workers were employed in loss-making enterprises: 2 million in 1,825 industrial firms, 0.3 million in 1,253 construction enterprises, and 0.7 million on 1,738 farms. Unless price reforms significantly improve the relative price structure faced by these enterprises, cutting off subsidies will inevitably result in plant closures and large job losses.

Structural changes. A market-determined structure is likely to create job losses in the capital goods industries. It is now government policy to transfer resources to industries producing consumer goods, and capital-intensive projects are also prime candidates for reducing the budget deficit. With capital goods industries accounting for about a fifth of total employment, abrupt reduction in capital spending would result in huge displacement.

Demilitarization. Implementation of recent disarmament agreements would imply that the Soviet armed forces would likely release manpower in the near future. The Soviet army currently consists of 4 million conscripts and professional soldiers. Moreover, the Ministry of Defence employs large numbers in unspecified activities within the USSR.

De-bureaucratization. With the central government administration and especially the branch ministries losing much of their functions, job losses are to be expected in the state bureaucracy as well.9

(2) Destinations of displaced workers

Displaced workers will face four alternatives:

Re-employment. There are several sectors in which there is a large potential to expand employment. Consumer goods industries, wholesale and retail trade, and the financial sector account for a very small proportion of total employment. Although private employment has expanded rapidly in these areas in the past two years, it would require massive growth in new small enterprises to absorb a significant share of the large-scale lay-offs from the state sector. Creating the right climate for entrepreneurs and ensuring that they have adequate access to finance and business advice will be crucial factors in this regard (see Chapters IV.2 and IV.5).

Exit from the labor force. Some displaced workers will leave the labor market. This option is likely to be availed of by older workers who are close to retirement. The degree to which this is a viable option for them will depend on the generosity of retirement pensions and severance payments, and perhaps also on the extent to which part-time work becomes more widespread as the service sector expands.

Emigration. Once a new emigration law has gone into effect giving Soviet citizens the right to leave the USSR, there could be a sharp rise in emigration, especially of highly-skilled workers. The Government has indicated that several million Soviet citizens would like to emigrate to the West in the coming years. The extent to which large-scale emigration materializes will partly depend on the response of Western governments to increased immigration pressures.

Unemployment. The degree to which job displacement gets translated into higher unemployment is likely to depend importantly on the personal characteristics of the redundant workers, as well as the state of the labor market and employment policies. Experience from the industrialized countries shows that while prime-age men are most likely to be made redundant, they will generally have short unemployment spells. Older workers, women, and unskilled workers tend to have much longer unemployment spells if they lose their job. The proportion of these groups in overall lay-offs can have a decisive influence on the stock of unemployed workers. Soviet officials expect that these groups will be strongly affected by redundancies.

The destination of displaced workers will also depend importantly on the evolution of wages. If wages rise relative to producer prices or other input prices, employers will be more reluctant to expand production or they will use other inputs more intensively. In principle, growing unemployment in a market economy should prevent real wages from deviating too far from their full-employment equilibrium value. However, experience from many market economies shows that real wages tend to respond only gradually to high unemployment.10

(3) Soviet estimates of displacement and unemployment

Goskomtrud has projected the labor market implications of a move towards a market economy under two different scenarios: (1) a transition period of 8-10 years; and (2) a transition period of 2 years (Table IV.6.6). The Shatalin plan also gave an estimate of the unemployment consequences in 1991 of the “500 days” transition program.

Table IV.6.6.USSR: Estimates of Job Displacement and Unemployment in the Transition to a Market Economy, 1991



Inflows into unemployment21.225.0
Of which:
New entrants0.70.7
Increase in unemployment0.75.3
Memorandum item:
Unemployment, end-1990 (millions)
Source: Goskomtrud.
Source: Goskomtrud.

According to Goskomtrud’s “gradual” scenario, unemployment would increase only moderately despite a sizable increase in the number of redundancies. It was projected that employment in the state sector would be reduced by a third (40 million) over the period of 8-10 years, with a fall of 9.5 million expected in 1991. About two-thirds of the contraction in 1991 was projected to come about through workers leaving the state sector on their own accord. Around 500,000 workers were expected to be made redundant due to ownership changes, presumably due to reduced overmanning. The state sector was also assumed to lay off 3 million workers: 1½ million as a result of loss-making enterprises closing down, and 1½ million due to structural changes in industry.

As well as 3 million lay-offs, Goskomtrud assumed that voluntary quits would remain at their 1990 level. This latter assumption seems most unrealistic as evidence from market economies suggests that quits should decline significantly in an environment of slack labor markets and rising unemployment. A salient feature of these projections is that total employment in the Soviet economy is expected to remain at its 1990 level. This in turn implies that the number of unemployed persons would increase solely in line with the projected increase in the labor force of some 700,000 persons. Indeed, the gradual increase in unemployment in the first half of the 1990s under these projections, up to 4½ million (an unemployment rate of 3¼ percent) in 1994, is entirely accounted for by the projected increase in the labor force.

In the “shock” scenario, Goskomtrud estimated that layoffs would be 6 million in 1991. Compared with the gradual scenario, the increase in redundancies would arise from more rapid structural changes in the state sector and faster privatization. The unemployment consequences of these developments were not explicitly stated by Goskomtrud, but if the job creation capacity of the economy were not impaired by the shock, the number of unemployed persons could still rise by 3¼ million in the first year of the program. If this is added to the current official estimates of unemployment (2-2.5 million), it would imply an unemployment rate of around 4 percent in 1991.

The Shatalin program had expected unemployment to rise by 5 million in the first year of a rapid transition to a market economy. This increase was to have resulted from 25 million workers changing jobs, an estimated increase over the previous year of about 50 percent, of which almost 20 million were expected to find new jobs. The labor force (the employed plus the unemployed) was expected to increase by only 0.5 percent. The near doubling of unemployment to 11.6 million (an unemployment rate of over 8 percent) would be associated with a doubling of the average duration of an unemployment spell, from one to two months.

(4) An assessment of the Soviet estimates

The Goskomtrud estimates of displacement and unemployment look modest compared with the experience of Eastern European countries which have embarked on a rapid transition to a market economy. Since the start of the Polish program in January 1990, recorded unemployment had risen by end-1990 to about 6 percent of the total labor force. The unemployment rate in the former GDR had increased to over 7 percent during the same period. In both cases, firms reacted swiftly to tighter financial discipline and greater competition by laying off workers, and job creation in the rest of the economy failed to keep pace. Sizable numbers of unfilled vacancies prior to the transition all but disappeared in the new environment. In the first quarter of 1990, vacancies in Poland were only 7 percent of what they had been one year earlier; and unfilled vacancies in the GDR fell by 75 percent in the first five months of 1990.

Possible reasons for relatively low unemployment in the Goskomtrud projections are that they did not fully take account of overmanning in enterprises, or that even under the “shock” scenario that it was assumed that enterprises would not immediately be subjected to hard budget constraints and competition. Another reason for low unemployment in the Goskomtrud projections was that virtually no increase was assumed in the average duration of unemployment spells during the period of restructuring. Keeping the average duration of unemployment constant at a time when there is a steep increase in redundancies would appear to be highly unrealistic.

The Soviet unemployment rate could easily exceed 5 percent during the transition to a market economy on relatively modest assumptions about the current degree of overmanning and the average duration of unemployment spells. This can be illustrated by two hypothetical scenarios for 1991-92 (Table IV.6.7). In both scenarios, enterprises are assumed to eliminate overmanning in two years and the initial stock of unemployment is set at 2½ million as of the end of 1990. Lay-offs are assumed to be relatively evenly spread across quarters in the “gradual” scenario. In the “radical” scenario, they are assumed to follow an inverted U-shape path, beginning with a slow start, a peak, and dwindling towards the end of the period. Enterprises currently operating at a loss are assumed to go out of business in one year, the layoffs being evenly spread over the year. In addition to redundant workers, inflows into unemployment will include new entrants to the labor market and voluntary quits. Voluntary quits are assumed to be inversely related to lay-offs, but the growth in the labor force is the same in both scenarios. The duration of unemployment spells increases gradually in the course of the two years.

Table IV.6.7.USSR: Unemployment Scenarios 1(In thousands)


to the



of Loss-Making









Gradual Scenario
Radical Scenario

It is assumed that the overall quit rate in the Soviet economy is 12.5 percent in the steady state. The increases in the labor force of 700,000 in 1991 and 600,000 in 1992 are estimates supplied by Goskomtrud. The estimate for employment in enterprises currently operating at a loss was also supplied by Goskomtrud. Overmanning rates of 12.5 percent and 25 percent in the gradual and radical scenarios, respectively, refer to the state sector plus collective farms. The quarterly outflow from unemployment is calculated as the reciprocal of the estimated average unemployment spell measured in months multiplied by three, which in turn is multiplied by the stock of unemployment in the previous quarter. This method is based on the assumption that those who are unemployed experience only one spell of unemployment during a single quarter. It also assumes that the labor market is in a steady state, in which eventually inflows into and outflows from unemployment cancel each other out.

It is assumed that the overall quit rate in the Soviet economy is 12.5 percent in the steady state. The increases in the labor force of 700,000 in 1991 and 600,000 in 1992 are estimates supplied by Goskomtrud. The estimate for employment in enterprises currently operating at a loss was also supplied by Goskomtrud. Overmanning rates of 12.5 percent and 25 percent in the gradual and radical scenarios, respectively, refer to the state sector plus collective farms. The quarterly outflow from unemployment is calculated as the reciprocal of the estimated average unemployment spell measured in months multiplied by three, which in turn is multiplied by the stock of unemployment in the previous quarter. This method is based on the assumption that those who are unemployed experience only one spell of unemployment during a single quarter. It also assumes that the labor market is in a steady state, in which eventually inflows into and outflows from unemployment cancel each other out.

With these assumptions, an unemployment trajectory can be computed for different starting values for overmanning and different lengths of the average duration of unemployment spells. If current overmanning is assumed to be around 12½ percent and the average duration of unemployment is assumed to increase to 7½ months (i.e., the gradual scenario), the unemployment rate could climb to around 9 percent before descending to its steady-state value of around 7½ percent.11 If overmanning is assumed to be around 25 percent and the average duration of unemployment increases to 10 months (i.e., the radical scenario), the unemployment rate would rise to around 12 percent in the first year of the transition period. The steady-state value would be around 10 percent. Judging from the experience of many West European countries, an average duration of unemployment of 10 months would still be relatively low. While care should be taken not to read too much into such mechanical calculations, they do show that the unemployment consequences of a transition to a market economy could be much sharper than recent Goskomtrud projections.


Living standards in the USSR have, in principle, been determined by an employment guarantee, in conjunction with administratively set wage rates and transfers. Wages and transfers, such as pensions, have been set at levels reflecting the system of controlled prices for consumer goods. In practice, however, the ability to mould living standards has been severely limited in recent years. The reforms of the late 1980s conferred a degree of decentralized decision making to enterprises in the setting of wages, bonuses, and in-kind transfers to employees. At the same time, the state procurement and distribution system has been under increasing pressure and has not been able to cope with increased demand. This has resulted in shortages and queues, as well as higher parallel market prices for some goods in particular areas. High cash wages have therefore not been adequate to ensure significantly increased access over goods and services. Moreover, groups depending on fixed transfers, such as pensions, have experienced declining standards relative to both wages and prices. It is not surprising, therefore, that there have been widening differentials in both living standards and poverty across various regions of the USSR.

a. Wages and incomes

A system of tariff wages has been in operation on a national scale in the USSR since the reforms of wage policy and administration in the mid-1950s. In addition to tariff wages, most employees earn bonuses, which are determined and allocated at the enterprise level. The social consumption fund, financed by contributions from enterprises as well as the union budget, provides for additional cash and in-kind payments to employees. While its distribution is less well documented,12 the system of in-kind payments is particularly important, and may vary considerably within as well as across enterprises. Even more limited is the information base regarding incomes from private activities.

(1) Tariff wages and bonuses

A major issue in relation to the administrative determination of the structure of wages has been the tension between the desire to control the distribution of income and the need to provide appropriate work incentives. As a result, an elaborate system of tariff wages has been devised for workers in the state-owned productive sector. There are (1) basic rates, specifying wages for the least skilled for each branch of the economy; (2) skill scales reflecting differentials relative to the least skilled; (3) additional coefficients for arduous work; and (4) regional coefficients to attract workers to inhospitable parts of the country. Wages in the so called non-productive sector (e.g., services, health care and administration) are determined in relation to similar jobs in the productive sector. While collective farmers are not covered by the tariff system, they are nonetheless guaranteed minimum wages.13

Bonuses have become a significant element of the monthly average wage. The importance of bonuses varies considerably, however, and they also have tended to vary by skill levels.

Despite the elaborate paraphernalia of the tariff wage structure, there was a compression of wage relativities over the 1956-66 period, although a widening of differentials since 1986. At the beginning of the wage reform of 1956, the ratio of earnings in the ninth decile relative to those of the first decile was reported to be 4.4.14 This ratio declined, however, to 2.8 in 1968, as a result of a cap on maximum wages, as well as the 50 percent increase in the minimum wage to rub 60 with effect from January 1968.15 The ratio stabilized between 3.0 and 3.3 until wage reforms began in 1986; by 1989 the ratio between the ninth and first deciles had risen to around 4.

A consequence of the earlier compression of differentials was the reduction in the salaries of professionals and technical staff relative to workers’ wages, although this may also have been due to an excess supply of technical and professional graduates. Salaries relative to wages in the construction sector, for example, fell below parity during the 1980s—they had been twice as high in 1955 (Table IV.6.8). Despite differential access to superior and scarce goods and services, particularly for managerial staff and administrators, by the mid-1980s it was thought that better pay was attracting technical staff into blue collar occupations. The 1986 wage reform was partly designed to rectify this structural problem, and there have been larger increases in salaries than wages subsequently.

Table IV.6.8.USSR: Relation Between Earnings of Salaried Workers and Wage Earners, 1955-88(Earnings of wage earners = 100)
Sources: Trud v SSSR (1988), pp. 189,209; Narkhoz 1988 (1989), p. 77, and Chapman (1990).
Sources: Trud v SSSR (1988), pp. 189,209; Narkhoz 1988 (1989), p. 77, and Chapman (1990).

(2) Sectoral and regional variations

The sectoral pattern of wages has remained fairly stable despite the reforms to date, with a few exceptions. Earnings in industry and transport have remained above average, and those in agriculture, housing, health, education, culture and art have been well below average throughout the 1980s (Table IV.6.9). Despite above-average growth in agricultural wages in the first half of the decade, they were no higher, relative to the overall average wage, at the end of the decade than in 1980. The improvement in salary levels relative to wages has had an effect in changing sectoral rankings only in financial services and administration, and that only in 1990. In other sectors, while the position of technicians and administrators relative to wage earners may have improved, overall sectoral rankings appear to have been unaffected. Within the industrial sector, there were declines in wage levels relative to the national average in the fuel, metallurgy, chemical and petrochemical industries, but increases in light and food industries, and in timber and wood processing.16 This reflects the recent shift of emphasis from heavy to consumer goods industries.

Table IV.6.9.USSR: Monthly Wages by Industry, First Half 1990(USSR average wage = 1.00)
Information services0.760.920.96
Housing & services0.790.780.78
Health & social services0.750.690.69
Financial services0.960.941.22
Memorandum item:
Average wage in the national economy
(rubles per month)168.9235.8256.0
Source: Calculated from data provided by Goskomtrud.
Source: Calculated from data provided by Goskomtrud.

Wage differences remain marked across republics. To some extent this is due to the location of industries, with the high wage sectors generally absent from the poorer regions. Agricultural activities normally attract lower wages. There is, moreover, variation across republics for the same sectors. Comparisons for a relatively less developed republic (Uzbekistan) with one of the more developed republics (the RSFSR) suggest that in 1987 average wages were lower in each sector in the poorer republic (Table IV.6.10).

Table IV.6.10.USSR: Monthly Wages by Republic and Branch, 1987(USSR average wage = 1.00)

Republican economy1.070.84
Information services0.870.72
Housing & services0.800.62
Health & social services0.750.67
Financial services1.050.83
Memorandum item:
Average monthly wage (rubles)216.1169.7
Source: Trud v SSSR (1988).
Source: Trud v SSSR (1988).

The main form of private activity until now has been that in cooperatives. Average earnings in cooperatives were around rub 367 per month in the first half of 1990 (Table IV.6.11), compared to rub 256 in the national economy (Table IV.6.9). The variation in earnings of cooperative workers across sectors tends to reflect the relative earnings potential in state-owned enterprises, with the highest wages in reprocessing and construction. It has been argued that cooperative members do not report their full income, and that these earnings reportedly range up to rub 900 per month.17

Table IV.6.11.USSR: Cooperatives — Employment and Average Wages, January-June 1990

Average Monthly Wage

Consumer services383.4261.74
Consumer goods877.1345.36
Medical care54.3306.02
Care of sick/elderly2.4145.83
Source: Goskomstat.
Source: Goskomstat.

(3) Wage adequacy and other benefits

Wage levels have apparently been designed such that a family with children would require two wage earners to satisfy basic needs, given current prices. Goskomtrud estimates suggest that, with the appearance of a child, 30 percent of single earner families fall below a poverty threshold of rub 75 per capita per month, and with more than one child more than 50 percent of single earner households fall below the poverty line. This puts pressure on the spouse to return to full time work in order to make ends meet. It also highlights the difficulties faced by single mothers with children. In addition, women tend to work in services and light industry, which traditionally have been low-paying; in general, they also appear not to advance as far as men.18

In addition to the payments out of the wage fund, benefits provided from the social consumption fund (SCF), often in kind, are an important element in determining the real living standards of recipients. The SCF includes both the totality of social benefits paid directly by the enterprise, and those provided by the Government, including (1) free benefits in kind: mainly educational and health services; (2) money payments and allowances—for vacations, stipends, sick leave, pensions, maternity leave, and income supplements; and (3) subsidized services, including rental housing, children’s activities, and vacation homes. Direct subsidies on food items, while an essential ingredient in the determination of the standard of living, are not included in the SCF. Out of the total SCF of rub 175 billion in 1988, just under rub 75 billion was spent on pensions and allowances, and around rub 100 billion accrued in the form of benefits in kind and other cash transfers. In addition, there were overt budgetary outlays of over rub 85 billion on account of subsidies to consumers for food items (Table IV.6.12). In 1988, various in-kind benefits and price subsidies together probably were equal to almost one-third of the size of the total money incomes (including transfers) of the population received from the socialized sector (Table IV.6.12 and Table D.2, Appendix II-1).

Table IV.6.12.USSR: Expenditures from Social Consumption Fund and Direct Subsidies, 1988-89(In billions of rubles)
Total social consumption fund175.4
Health care25.4
Of which:
Direct subsidies86.397.6
Meat and meat products21.722.7
Milk and milk products15.817.5
Source: Sotsial’noe razvitie SSSR (1990).
Source: Sotsial’noe razvitie SSSR (1990).

The SCF may indeed have been a policy instrument in the equalization of the income distribution.19 However, while a significant number of families with low or no earnings may be lifted above the poverty threshold through SCF payments, low pensions and allowances led to many retired, one-parent, and large families remaining below the poverty line. A significant proportion of retired people only manage to stay above the line by continuing to work. As for free education, it has been argued that low-income families received larger benefits because they have more children and there is universal education. The situation with respect to the health care services seems to favor the better-off, who appear to have easier access than the poorer groups.

In any event, living standards appear not to be dependent either on an individual’s productivity or level of remuneration. Moreover, since various fringe benefits—such as places in pre-school, the better hospitals, dachas or dining rooms—are in short supply, privileges in this respect tend to confer greater access to such benefits than wage earnings per se.20

(4) Food subsidies

Highly subsidized meat and dairy products are more important in the budgets of households living in relatively high income republics, whereas very limited quantities are either available or consumed in the poorer Central Asian republics. Moreover, high income families tend to benefit more than low income families from food subsidies, given their higher per capita consumption of these products. Consequently, the commodity pattern of subsidies favors both richer households and regions. It should also be emphasized that many commodities, including grain and bread, bear implicit subsidies and are at well below world prices. Reforms aimed at ensuring appropriate prices to farmers will transform these implicit subsidies into explicit subsidies unless measures are taken to adjust consumer prices.21

b. Inequality, poverty and need

The foregoing discussion suggests that both the setting of the wage mechanism and the system of transfers through the social consumption funds has, at least until recently, been directed by considerations of equality. But while the need to prevent wide income differentials appears to have been a political objective, the existence of poverty has not been a matter of particular concern, since it was assumed that the administratively determined system of wages and transfers would be adequate to meet minimum living standards. An official poverty line was not established until 1989. It is more appropriate, however, to define poverty in a broader sense, to encompass not only the lack of monetary means to ensure a minimum command over material resources, but to reflect unsatisfactory social conditions, including health, sanitation and access to services.

(1) The distribution of income

The distribution of income in the USSR is not very concentrated by international standards. The contrast with China is particularly instructive. In China, reforms in rural areas were instituted in 1978 leading to a dismantling of communes and a move towards a market oriented system. While this rapidly increased average incomes, there was also an increase in inequality: both because the new income generating activities led to relatively high incomes, and because the social support system which depended on the communes to protect the vulnerable was not replaced. In urban areas, the pace of reform has been much more limited. A comparison of rural and urban areas shows a much lower incidence of inequality in the latter in 1987, using different inequality measures to reflect the different types of inequality.22 It is interesting that the pattern of inequality in the USSR in 1989 roughly corresponds to that obtained in urban China in 1987 (Table IV.6.13). Given that rural China has undergone the transformation to a market-based system (and urban areas have not), this sector provides an interesting contrast and suggests likely directions of change for the USSR. At present Soviet income is much more equally distributed than in rural China, with both relatively fewer well off and relatively fewer poor individuals. While there is likely to be some increase in inequality in the transition period, it is more important from a policy perspective to be concerned about minimum living standards of the poor rather than inequality per se.

Table IV.6.13.USSR: Measures of Income Inequality, China and USSR 1
China (1987)
USSR (1989)UrbanPeasants
Coefficient of variation0.5350.6141.010
Gini coefficient0.2670.2580.285
Atkinson index
Parameter = 20.2320.2160.284
Parameter = 5.00.5540.5650.849
Sources: Estimates for the USSR are based on Goskomstat data. Estimates for China are from Ahmad and Wang, in press.

Both sets of estimates are based on grouped data on the assumption of linear interpolation of the withingroup distributions. The parameters for the Atkinson index refer to the level of inequality aversion.

Sources: Estimates for the USSR are based on Goskomstat data. Estimates for China are from Ahmad and Wang, in press.

Both sets of estimates are based on grouped data on the assumption of linear interpolation of the withingroup distributions. The parameters for the Atkinson index refer to the level of inequality aversion.

(2) Poverty

The relatively egalitarian distribution of income in the USSR, however, does not mean that poverty does not exist. Indeed, the minimum wage, which is currently rub 70 per month, has not been adjusted since the 1970s. In order to construct minimum incomes, a representative consumption basket assessed at administered prices was chosen by Goskomtrud. The method did not allow for the non-availability of goods at administered prices, or that parallel market purchases might be at prices considerably higher than the administered levels. The disutility and time loss involved in extensive queueing was not accounted for either. Goskomtrud estimated that in 1988, the minimum income level should have been rub 100 per capita per month, and that at such a threshold 80 million persons (or over a quarter of the population) would have been classified as poor. However, the official poverty line chosen was rub 75 per capita per month, and according to this threshold there were 43 million persons, or around 15 percent of the population, living in poverty. While the numbers falling below a constant nominal poverty line have decreased since 1988 (to around 30 million in 1989—see Table IV.6.14), higher parallel market prices in conjunction with greater shortages should have led to a revision of the poverty line. Estimates by Goskomtrud and its affiliated research institutes suggest that both the numbers of the poor and the proportion of the population in poverty has increased. Given the large proportion of households concentrated between rub 75 and rub 150 per capita (45 percent of the total population in 1989—see Table IV.6.14), small adjustments to the definition of the poverty line would lead to large changes in the numbers classified as poor.

Table IV.6.14.USSR: Republican Income Distribution and Poverty Levels, 1989

Households by Per Capita Income Groups


rub 75





rub 200
(Percent of total population)
Source: Goskomstat.
Source: Goskomstat.

It is difficult to compare poverty in other countries with that in the USSR, given the absence of purchasing power parity estimates for the latter. However, the 1990 World Development Report, using a cut-off point on a purchasing power parity (PPP) basis of US$370 per capita per year, found 33 percent of the population in developing countries to be poor, ranging from 47 percent in Sub-Saharan Africa, to 51 percent in South Asia, to 20 percent in East Asia, and 8 percent in Poland. (In terms of local standards, however, about 20 percent of the population in Poland was considered to live below the poverty line in the late 1980s.) At the commercial exchange rate of the USSR, the US$370 poverty line would translate into about rub 50 per capita per month, although at the noncommercial rate the poverty line should have been about rub 180. On this basis, a poverty line between rub 100 and rub 150 might be considered comparable to international standards.

The regional distribution of incomes for 1989 indicates high concentrations of poverty in particular republics. Thus in Tadzhikistan and Uzbekistan, 51 percent and 43 percent, respectively, of the population fell below the poverty threshold in 1989, compared to less than 2 percent in Estonia, and 5 percent in the Russian Republic. The substantial difference in poverty levels between the RSFSR and Uzbekistan is explained not so much by variations in salary levels (although these do exist, as noted above), but by differences in family composition and size.23 Uzbekistan, for example, has quite a different demographic profile from the RSFSR with a younger population and a high under-15 dependency ratio. On the other hand, the RSFSR tends to have a smaller average family size, but with an ageing population structure. Higher income levels in the RSFSR are, however, no guarantee of an equivalently higher standard of living, given the regional variation in shortages of basic goods and services. Azerbaidzhan, for example, with relatively low wage levels, does not suffer one of the lowest living standards in the union, because of relatively abundant supplies of essentials at relatively low prices. In some cases, however, such as Uzbekistan, the relative abundance of cheap fruits and vegetables is not adequate to completely counterbalance low money income levels.

Actual consumption patterns are therefore a much more accurate indicator of living standards than wage levels. For the main items currently carrying a budgetary subsidy, there are significant differences in consumption patterns across republics (Table IV.6.15).24 For example, per capita consumption of meat and products, milk and eggs in Uzbekistan is roughly half that in the RSFSR, and consumption of potatoes a third. However, the consumption of vegetables and fruits in Uzbekistan is somewhat higher. This pattern is also reflected in the consumption of durables, with the per capita sale of television sets and refrigerators in Uzbekistan (6.3 and 16 units per 1,000 inhabitants respectively in 1984) less than half or around half that in the RSFSR (16.3 and 31.4 per 1,000 inhabitants).25

Table IV.6.15.USSR: Distribution of Annual Per Capita Consumption of Major Food Items Ranked by Households According to Monthly Per Capita Income, by Republic, 1989 1


rub 75










Meat and Meat Products (kilograms per capita)
Milk/Dairy Products (kilograms per capita)
Eggs (number per capita)
Vegetables (kilograms per capita)
Potatoes (kilograms per capita)
Source: Goskomstat.

For each republic and type of commodity, the average consumption in quantity terms and the mean consumption of those below the poverty line (rub 75 per capita per month) and other income groups.

Source: Goskomstat.

For each republic and type of commodity, the average consumption in quantity terms and the mean consumption of those below the poverty line (rub 75 per capita per month) and other income groups.

There are clearly price variations too, both across income categories and across regions (Table IV.6.16). In the RSFSR, a kilogram of meat costs rub 2.58, somewhat below the national average of rub 2.68, but in Uzbekistan the price is rub 3.49 per kilogram. And in Uzbekistan, as in other Central Asian republics, the poorest income group pays a higher price for meat per kilogram than the richer households. This might suggest that the poorer segments of the population are forced to buy in parallel, higher priced markets for some commodities in short supply, as they may have less access to supplies at lower, controlled prices. This pattern is not borne out, however, for some of the other highly subsidized products, and therefore it is difficult to generalize.

Table IV.6.16.USSR: Regional and Household Specific Price Variations(Rubles per kilogram)

Households with

Per Capita Income



rub 75


rub 200
Meat and Meat Products
Milk/Dairy Products
Source: Goskomstat.
Source: Goskomstat.

(3) Social indicators

Living standards involve not only access over goods, described above, but also the quality of services and provision and quality of public goods such as preventive health care. Two fairly general indicators are life expectancy at birth and infant mortality rates. They capture a variety of factors reflecting social conditions and standards of living. With life expectancy at birth for men and women estimated at 64.2 and 73.3 years respectively in 1987, the USSR trailed behind all Eastern European countries and industrial countries except Turkey (Table IV.6.17). The infant mortality rate in the USSR was also higher than in all other countries surveyed except Romania.

Table IV.6.17.USSR: Life Expectancy and Infant Mortality in the USSR and Selected Countries, 1982-86 1
Life Expectancy at BirthInfant Mortality
MaleFemale(Deaths per

thousand live births)
(in years)
OECD average71.678.08.42
Germany, Fed. Rep71.277.88.3
New Zealand71.277.79.8
United Kingdom71.577.49.1
United States71.578.310.0
East European average67.274.417.6
Sources: U.S. Department of Health and Human Services (1989, 1990).

Life expectancy data for most OECD countries and all data for East European countries are for a selected year between 1982 and 1986.

Excludes Turkey.


Sources: U.S. Department of Health and Human Services (1989, 1990).

Life expectancy data for most OECD countries and all data for East European countries are for a selected year between 1982 and 1986.

Excludes Turkey.


As with other indicators, both life expectancy at birth and infant mortality rates vary across republics in the USSR. The standards in the RSFSR are fairly similar to those in Eastern Europe. However, both measures indicate somewhat lower living standards in Central Asia and the Caucasus. Infant mortality in Uzbekistan, Turkmenistan and Tadzhikistan in 1988 was 43.3, 48.9 and 53.3 per thousand respectively,26 roughly twice the overall average for the USSR and reported to be increasing rapidly. In some regions of Uzbekistan, every tenth child reportedly dies at birth.27 To some extent this reflects the lower availability of doctors, hospital beds, and quality of care.

c. Pensions and family allowances

There are two broad categories of households that are particularly disadvantaged under present circumstances in the USSR: the elderly relying on pensions, and families with large numbers of children. Both groups have a combination of low per capita incomes and a greater need for health care facilities and services.

(1) Pensions

The provisions of the current old age pension system in the USSR date from 1956. At present there are around 60 million pensioners who received an estimated rub 65 billion in pensions in 1990 (Table 4, Appendix III. 1-3; also see Table IV.6.18). The average pension in 1990 was about rub 100 per month (for data through 1988, see Table IV.6.19). The pension legislation was supplemented in 1964 to extend coverage to collective farm workers (kolkhozniki), although the terms were less generous than for workers on state farms, who were treated on a par with other state employees. It was envisaged that the differences would be gradually reduced, and while this may have happened to some degree, they have not been entirely eliminated. The retirement age is relatively low in the USSR (well below the OECD average), and large groups qualify for pensions at an earlier age because of the nature of their job (workers in hazardous jobs). There is a precipitous drop in living standards on retirement, which forces elderly workers to remain employed if possible. In 1987, over 50 percent of state pensions and 90 percent of collective farmers’ pensions were below rub 80 per month (Table IV.6.20), thus fairly close to or below the official poverty line. Minimum pensions have been rub 70 per month since 1981, when they were set equal to the minimum wage. Maximum state pensions were fixed at rub 120 in 1956, at a time when the average wage was rub 73 per month. The average wage exceeded rub 250 per month in 1990 (Table IV.6.9), and the replacement rate for a retiring worker receiving the average wage has dropped from around 100 percent in 1956 to under 50 percent. Fixing pensions in absolute terms has clearly resulted in eroding standards of living for old-age pensioners.

Table IV.6.18.USSR: Total Pension Expenditures(In thousands of rubles)
State pensions45,32350,19855,72856,78058,585
Of which:
Old age29,39534,00739,33340,50143,164
Loss of breadwinner6,9897,2247,1696,4365,723
Collective farms11,85811,13510,0739,98210,432
Of which:
Old age10,3399.7698,9268,8739,383
Loss of breadwinner697630502470419
Source: Sotsial’noe razvitie SSSR (1990).
Source: Sotsial’noe razvitie SSSR (1990).
Table IV.6.19.USSR: Average Pensions(In rubles per month)
State SectorCollective Farms
Old age62.771.687.291.793.925.
Workman’s disability49.752.757.659.760.827.834.640.645.346.8
General disability45.253.464.468.
Loss of breadwinner35.939.146.348.048.825.131.338.741.141.9
Source: Sotsial’noe razvitie SSSR (1990).
Source: Sotsial’noe razvitie SSSR (1990).
Table IV.6.20.USSR: Distribution of Pensions, 1987(In percent)
State PensionsCollective Farms
TotalOld ageTotalOld age
Less than rub 8050.441.592.992.8
rub 80-10014.917.03.63.6
rub 100-12011.313.11.81.8
More than rub 12023.428.41.71.8
Source: Sotsial’noe razvitie (1990), p. 107.
Source: Sotsial’noe razvitie (1990), p. 107.

The increase in the number of pensioners has been quite dramatic, and is set to continue. Between 1960 and 1981, while the total number of pensioners increased by 143 percent, there was a 650 percent increase in old age pensioners, predominantly from the state sector. Over the period 1979-89, persons of retirement age increased from just under 27 percent of the working population to almost 31 percent (Table IV.6.21). A similar increase is projected during the 1990s, and by the year 2000 the ratio of persons of pensionable age to the working population is expected to reach 34 percent.

Table IV.6.21.USSR: Persons Above Retirement Age in the 1979 and 1989 Censuses 1








Source: Goskomstat.

“Retirees” refer to persons above retirement age, which is 55 years for women and 60 years for men. The “working” population refers to the population of legal working age. The dependency ratio does not account for the fact that some “retirees” may be working, and that some persons of legal working age may not be in the labor market.

Source: Goskomstat.

“Retirees” refer to persons above retirement age, which is 55 years for women and 60 years for men. The “working” population refers to the population of legal working age. The dependency ratio does not account for the fact that some “retirees” may be working, and that some persons of legal working age may not be in the labor market.

Significant regional differences in demographic characteristics are reflected in regional patterns of expenditures on pensions. While in the Ukraine the old age dependency ratio in 1979 was already higher than the current national average, and approached 38 percent in 1989, the ratio in Uzbekistan actually declined from 18 percent to just over 16 percent over the same period. These differences impose a severe constraint on the degree of pooling that is feasible on account of pensions. A national pool would redistribute resources from regions with low old-age dependency ratios to the regions with concentrations of retirees, which under Soviet circumstances would imply redistribution from poorer to richer republics. This highlights a fundamental dilemma in the formulation of a union-wide pension policy: the extent of the ageing of the population dictates as wide a pooling of risks across generations as possible, but the inter-regional redistribution that is involved may not be acceptable.

The all-union pension legislation of 1956, which provided considerable extensions in the coverage of the system and a substantial increase in benefit levels,28 was designed to be financed through a payroll tax ranging between 4.4 and 9 percent, with minimal contributions from the national budget. At present, old-age pensions are funded on a pay-as-you-go basis, through payroll taxes on enterprises and contributions from general budget revenues. Enterprises in 1990 paid 12 percent on average of their wage fund (tariff wages and bonuses) to the social security budget, which finances old-age pensions and other social expenditure, but about 40 percent of the expenditures were covered by general budget revenues. The two major factors responsible for the increase in the payroll tax were the extension of coverage to include workers on collective farms—approaching universality without a commensurate increase in resources—and the substantial increase in the number of pensioners due to the ageing of the population.

There are significant difficulties with administering the pension system. According to Goskomtrud, the present archaic system of accounting on the basis of the wage fund of enterprises cannot be changed easily. Unlike in many Western countries, there is no system of social security numbers and individual contributions cannot be identified. The authorities are attempting to computerize the system, but this is difficult without citizen codes. An important measure to be introduced under the new pension reform is a contribution to be paid by individuals, for the first time, on their monthly earnings at a rate of 1 percent. This would introduce individual accounting procedures and would also be an important signal for persuading individuals to contribute to their retirement, as the market determination of wages develops.29 Another difficulty in administering the system lies in the fact that the place of work and of retirement often differs. Many workers temporarily migrate to Eastern Siberia for high wages with high pension entitlements and early retirement but seldom live there in their old age. Administering benefits to such individuals poses particular difficulties.

(2) Allowances

There are a number of allowances for particular contingencies, some of which are means-tested. Meager child allowances have long been payable to low income households. Maternity benefits and birth grants, as well as temporary disability allowances, also exist. Nonetheless, the expenditures on these items (apart from the temporary disability allowances, which were almost rub 9 billion in 1989), have been fairly limited, and only the maternity grant (rub 2.37 billion in 1989), and care allowances for children under the age of one (rub 1.51 billion in 1989), have been of any consequence.

New benefits, or adjustments to the levels necessary to make an impact on the poverty of particular groups, are likely to have major budgetary or financing consequences. Moreover, the ability to pay and the need for social assistance does not often coincide within the same area (e.g. in Central Asia), and continuation of such transfers is likely to be contingent upon resources from the union and the higher income republics. Means-testing is frequently advocated to limit budgetary outlays.

The payment of means-tested child allowances is currently one of the provisions of the social security system. Unlike in industrial countries, such as the United States or the United Kingdom, means-tested benefits in the USSR in the past have had relatively high take-up rates, as adequate income levels have been perceived as a right (although not always fulfilled). In Uzbekistan, as noted earlier, 43 percent of households qualified for assistance. However, the means-test did not include income from garden plots and home grown produce, thus indicating potential difficulties.

The experience of means-testing in developing countries is quite poor,30 and that of developed countries has not been particularly successful either. Administration is a major issue, as is the determination of the appropriate cut-off point for the means-test. It is not easy to determine either sources of income or whether households or individuals are likely to fall below a particular poverty line. This issue is particularly severe in respect of services, the informal sector, and in agriculture. Nonetheless, the administration has to be local if there is to be any chance of accurately identifying those below the cut-off point. However, with local and regional administration of funds from higher levels of government, there is usually little local incentive to limit expenditures.

Potentially the most serious issue with means-testing involves disincentive effects. This issue will become more important as benefit levels rise as proposed by the authorities (see section 4.e). As the allowances are based on per capita income, benefits for a four-child family could be substantial enough to discourage a shift into productive work. There is thus a need to design social assistance programs in such a way that disincentive effects are minimized. The difficulty with means-testing is that high marginal tax rates are generated at the point at which the benefit is withdrawn, leading to what is called a “poverty trap”. However, targeting by household, rather than on a per capita basis, may avoid this distortion to the overall tax-benefit structure.


a. Labor market policies

This section deals primarily with the role of labor market policies during the transition period. It also addresses some issues of improving economic performance through labor market policy when the period of large-scale labor displacement is over.

(1) Labor market information and matching

High turnover in the Soviet labor market suggests that many workers are accustomed to market-like behavior in regard to job search; two thirds of all hiring reportedly takes place at the enterprise gate and about 20 percent is the result of placement efforts by the labor offices. Eighty percent of these latter placements concern unskilled jobs, as enterprises are apparently not keen to notify skilled vacancies to the labor offices. The numerous imbalances persisting in the labor market, however, suggest that the relatively high turnover has not contributed to a better matching of supply and demand and that a more developed system of labor market information, job search assistance and placement is required.

The transition to a market economy will lead to a large number of workers searching for new jobs and of enterprises looking for personnel. This will further increase the need for the collection, analysis and publication of information on vacancies and job seekers. The availability of such information could help avoid high search costs, labor shortages, unsatisfactory working conditions or lengthy periods of unemployment. The advertising of job vacancies in the mass media is one approach to rapid dissemination of information; it is widely used in all Western countries and is considered to be an efficient method of recruitment. But for many job seekers, more developed services such as vocational counselling, retraining and the provision of work experience are needed; many enterprises will also require assistance in personnel planning and advice on recruitment and training.31 Such services have not hitherto generally been provided in the USSR.

The existing network of some 2,000 labor offices concentrates on locating personnel for state enterprises. Vocational counselling takes place in the education system. According to Goskomtrud and to officials interviewed in employment offices, students hardly ever meet with their vocational counsellor. There is on average only one counsellor available for 50,000 persons. Since 1988, some 500 employment centers have been established which are to coordinate the collection of employment information by local offices in their area and to support enterprises in their training activities. It is generally accepted, however, that the existing labor offices are understaffed and that their resources are insufficient to provide a satisfactory flow of labor market information. Goskomtrud plans a large increase in the number of placement officers and vocational counsellors. It is important to recognize, however, that the labor offices in a reform setting must not only provide a more complete range of labor market information and placement services but must also cope with the new phenomenon of unemployment. It will be difficult to concentrate on both functions at the same time and priority will have to be given to assisting the unemployed. An expanded and reformed employment service should aim, however, at collecting and disseminating information on as many vacancies as possible as this is a precondition for effective help to the unemployed. This is presumably the reason why compulsory notification of vacancies is contained in the Employment Law approved on first reading by the Supreme Soviet in October 1990. The experience of industrialized countries suggests, however, that enacting legislation requiring notification has not been particularly helpful in encouraging enterprises to cooperate with the employment offices.

(2) Effective assistance to the unemployed

Assistance to the unemployed has two elements: (1) income support and (2) assistance to get back to gainful work. Eligibility for income support raises the question of the definition of unemployment. The new law defines a person as “unemployed” if he is “capable of working but who for reasons beyond his or her control has no earnings or other legal source of income, who is registered with the state employment service as being available for work, who is actually seeking work, who is prepared to undergo training and for whom the service has not yet made a recruitment offer”. The law adds that details of this definition should be regulated at the republican level according to particular circumstances. While this definition is fairly restrictive, it is most likely that the very acceptance of the notion of unemployment will by itself increase the number of registered unemployed. Enterprises will be less reluctant to lay off workers if they have access to some income support and/or policies to assist them to find a new job. Second, individuals who are only marginally attached to the labor force may register with the employment service in order to become eligible for income support.

Under this legislation, laid-off workers are to receive up to three months wages from their previous employers. In this period they are expected to search for new employment and they have to register with the employment service if they want to safeguard eligibility for future benefits. New entrants and re-entrants have a right to benefits as well. There is a waiting period of 10 days before benefits are paid. Benefits are paid for a maximum of six months within any 12-month period. During the following six months income support is only possible through participation in public works or training programs. Thereafter, eligibility for unemployment benefits is re-established for an additional period of six months. Benefits correspond to 50 percent of the tariff wage earned in the previous job but no less than the minimum wage. The use of the tariff wage as the base implies that bonuses and other cash or in-kind benefits are not taken into consideration although they can represent a significant portion of actual earnings. The minimum benefit, which applies to first-time job seekers and others with no recent employment history, is fixed at 75 percent of the minimum wage. Benefits for the unemployed with family responsibilities are increased by 10 percent. Replacement ratios will be normally less than 50 percent depending inter alia on the difference between actual earnings and the tariff wage (on average 20 percent) and on family responsibilities.

During the first six months of receipt of unemployment benefits, beneficiaries are obliged to accept jobs considered to be “suitable” for them, with suitability judged not only in terms of age or health record but also in accordance with professional competence and skills. The latter criteria are not taken into consideration for those entering or re-entering the labor market or for those receiving unemployment benefits during a second period. Training assistance for the unemployed with family responsibilities amounts to 75 percent of previous tariff wages (50 percent if without family). For new entrants and re-entrants the support is based on 100 percent of the minimum wage. Participation in public works is paid at the going wage for the activity involved but at not less than the minimum wage.

In principle, this benefit system appears to be carefully formulated. It provides incentives for the unemployed to take on a new job or training without depriving them of benefits. It also puts pressure on public authorities to focus on the reintegration of the unemployed and to avoid prolonged dependency on income support. However, the system will require close interaction between the administration of benefits and the placement service as well as the institutions responsible for designing and providing training and public works programs. Western experience suggests that this coordination is often difficult to establish. Such a policy would also require an elaborate infrastructure, the establishment of which may be constrained for budgetary reasons.

Some unemployment scenarios suggest that inflows of job seekers may become so large that they could easily overwhelm the administrative capacity of the employment service. If the providers of training and public works or the employment services are no longer able to cope, it may become necessary to set priorities in favor of certain groups in need of intensive help. Even then, long-term unemployment on a considerable scale could emerge, in which case the limitation of the payment of unemployment benefit to six months within any 12-month period could become a serious problem.

The Employment Law does not relieve enterprises completely from their responsibilities towards their workers. It requires them to give early warning to the workers’ representatives and the employment service in the case of major redundancies. The main obligation, however, concerns the three months severance payment. The severance payment is not only a source of income for the workers laid off but may act as a disincentive for enterprises to shift short-term adjustment costs outside the firm.

The Employment Law suggests that employment offices should have the possibility to delay lay-offs for up to six months by subsidizing the enterprises from an employment fund. Experience with similar schemes in Western countries does not suggest that this would be a worthwhile investment of public funds. As soon as such a provision is established, it may become impossible not to use it in response to political pressures at the local level. If financial support for firms in temporary difficulties or in a period of restructuring has to be provided, it should be done within the framework of industrial or regional policies. Resources for averting redundancy could be used in a more forward-looking way. One option might be to establish regionally-based institutions which would provide employment measures for the workers concerned—retraining, self-employment assistance or subsidizing re-employment in expanding enterprises.

(3) Financing, resources and implementation

The key provision of the authorities’ plan for reorganizing the employment service is the decentralization of management and policy planning to the local or regional level. It is up to the republics to decide on the number and location of new offices and regional centers, as well as their links to other public authorities at the local and regional level. At the union level, general guidelines and minimum requirements are fixed. Given the size of the USSR and the economic and social differences between the regions and republics, such an approach seems appropriate. It leaves the necessary flexibility for developing regional and republican programs and policies, while making sure that policies in the whole country are based on the same principles and that certain minimum standards are provided.

There is another reason for ensuring that the central government has some say in the operation of labor market policies at the lower levels. The resources to be made available for employment policies will be raised by a one percent levy on payroll. However, the number of persons in need of help (essentially the unemployed) and the tax-base are distributed unequally over the country (Table IV.6.22). If some minimum standards are to be provided throughout the union, regional and inter-republican redistribution is needed. To finance a minimum provision of unemployment assistance, including income support, would require at officially projected levels of unemployment—according to calculations of Goskomtrud—a contribution rate on the wage fund in Uzbekistan of about 4 percent compared with less than 0.5 percent for the Western republics.

Table IV.6.22.USSR: Estimates of Wage Funds, Employment and Persons in Need of Help by Republic, 1991(In percent of USSR total)
RepublicWage FundEmploymentPersons in

Need of Help

Source: Goskomtrud.
Source: Goskomtrud.

The mechanism suggested by the union government would have all enterprises pay a contribution of 1 percent of their wage fund into the republican employment funds. Twenty five percent of this revenue would be transferred to the union fund which would be essentially used for redistribution. The envisaged transfer of surpluses to the union fund, however, would be insufficient for the required redistribution in 1991 (Table IV.6.23). Furthermore, the assumptions underlying this mechanism probably reflect existing patterns of need and unemployment more than likely projections of structural adjustment. It is likely that the adjustment would be more severe in the heavily industrialized areas of the country, and that this might generate a different regional pattern of expenditures than these official estimates suggest.

Table IV.6.23.USSR: Official Projections for the Employment Fund, 1991(In millions of rubles)

Fund 1991

in 1991
Outturn by


(- = shortfall)

to the





USSR4,734.63,839.9990.01,197 1
Source: Goskomtrud.

Shortfall made up from union budget.

Source: Goskomtrud.

Shortfall made up from union budget.

Payroll tax revenues would also be used to finance the expansion of the employment service. There are currently about 11,000 staff members working in the labor offices. If on average there were five million unemployed in 1991-92, one employee in the employment service would have to take care of 500 unemployed persons, which compares with about 100 in Portugal and Greece and about 50 in most other OECD countries. The authorities are well aware of the need to expand the employment service and plan to reach a figure of about 40,000 officers in 1992-93. The plans for expansion include an increase in salaries of placement officers to the average levels of state employees. Staff training and the provision of equipment for the employment service would be a promising area for technical assistance from the West.

In most Western countries about 10 percent or less of total expenditure related to unemployment is needed for the operation of the employment service. The driving factor behind costs is unemployment. According to projections of Goskomtrud, a one percent contribution rate would be sufficient to finance programs for about 3 million unemployed (i.e., just over 2 percent of the labor force), assuming full redistribution of revenues between the republics. If additional expenditure for training (apart from allowances) is taken into account, a 1 percent contribution rate would allow full measures for about 2¼ million unemployed. However, the limitation of full redistribution of resources projected by Goskomtrud (above) would reduce the potential coverage by about 20 percent. Taking all these factors into account, it would appear that an increase in the unemployment rate by 2 percentage points would require an increase in the contribution rate by about another one percentage point.32

This possible shortfall in financing for employment programs has already been underlined by many observers, including the official trade unions, and some republican governments were apparently planning to use their budgets to finance additional employment-related activities. To rely so heavily on the general budget does not necessarily provide a solid base either for raising the required resources for programs to cope with large-scale unemployment, or for a long-term strategy to build up the employment service. The policy objectives regarding the budget are clearly more restrictive and there will be stiff competition among government departments for scarce resources. Likewise, the increases in payroll taxes represent increases in wage costs and could be expected to reduce the demand for labor and reinforce labor shedding. An alternative to raising the tax on the wage fund may be to introduce separate contributions for employees and employers. This would be a sensible step for several reasons. First, it would introduce an element of solidarity between employed and unemployed workers, which would certainly be desirable to ameliorate the social tensions associated with rising unemployment. Second, although it may be correct that in the long-run the effects of taxes on employment are independent of whether they are paid by the employer or by the employee, it would likely make a difference in the short term.33 Third, there is a case for sharing the costs of displacement and unemployment among all employers.

(4) Employment programs

Exceptional programs might be required during the transition process. Shortages of goods, particularly food and other consumer goods as well as services, are endemic throughout the USSR. If programs could be designed to focus on these needs by providing temporary jobs, large numbers of the unemployed could find productive employment and a source of income. Two approaches are worth considering: (1) public works programs organized by local authorities; and (2) encouragement of self-employment on a large scale. Local and regional councils should be able to set up programs of this nature. The republican employment funds could offer subsidies taking into account expenditures for benefits which would otherwise have to be paid in the form of unemployment compensation.

Goskomtrud has made proposals to the Council of Ministers for a public works program. The main areas of economic activity would be agricultural production, food processing, infrastructure renewal, social, cultural and health services, and child-care services. It is envisaged to leave the organization to local authorities who would also have to contribute to the financing. Not only full-time jobs but also part-time or seasonal jobs are envisaged. The Employment Law provides for two-month contracts with the possibility of a further two months for participation in public works. Short periods of employment of this nature have the advantage that temporary economic activity and seasonal production can be supported effectively.

Such special measures for the unemployed, if needed, should be concentrated on economically meaningful activities and pursued with the objectives of reintegrating displaced workers as quickly as possible into gainful work and preventing long-term unemployment. Goskomtrud intends to minimize the substitution of public works for regular employment by requiring that funded projects be separated from regular activities of local authorities.

More recently, the authorities have considered one- to two-year contracts for public works programs in order to provide longer-term professional orientation. The provision of work experience is an appropriate way to achieve professional re-orientation for older workers or those who have less capacity for retraining. Longer participation may also be necessary in order to achieve reintegration of those who have been unemployed for long periods or who belong to particularly disadvantaged groups. The direct costs of such programs would be considerably higher, although the net results would depend on the savings of unemployment benefits and on induced labor supply effects in the longer-term. It might be more difficult to avoid substitution effects in the case of long-term programs. However, these programs would have a strong social and redistributional objective and aim at improving the employability of severely disadvantaged groups such as older unemployed workers with obsolete skills.

Active encouragement of self-employment initiatives is a second promising route to pursue. Presently, self-employment is still inhibited by legal restrictions and by a generally unfavorable climate for entrepreneurial activities. Many people are engaged in such activities which, however, are not generally legalized and are often used as a second job. The first step to encourage self-employment would be to legalize all types of entrepreneurial activities, notably in the areas in which shortages exist. The second step would be to give some support to people who are interested in developing these activities further: access to premises, supplies of raw materials and other inputs, as well as some financial help and social security.

As a result of active encouragement of small-scale enterprise and self-employment activities, black market activities would tend to be moved into the regular economy. Moreover, many individuals have a legal job but also an undeclared business at the same time. If their second activity were to be legalized and given public support, they might decide to drop the first jobs and concentrate on expanding their businesses, which could lead to the creation of additional jobs. At the same time, this would contribute to relieving the overmanning problem in the state sector. The latter consideration also suggests that it might be in the interests of state enterprises to assist their workers who want to set up their own businesses by providing them with financial support or supplies of necessary inputs.

It is difficult to estimate the additional opportunities in the area of self-employment. As noted in section 2.a.(2), self employment in the USSR accounts for only 0.2 percent of the labor force as compared with about 10 percent on average in OECD countries. This suggests that considerable job opportunities may arise in this area once the transition process has gained momentum.

b. Education and training

Rapid industrialization in the USSR required a massive expansion in the supply of skills and technical knowledge, which was provided by an extensive network of educational and training institutions. The number of students involved in education, training, and retraining has multiplied several times since 1940 (Chart 4). The result has been a steady rise in the average level of education of the labor force, with no significant gender differences in this respect. About 12 percent of the population aged 20 and over is estimated to have completed higher education. International comparisons of educational attainment are difficult to make because of immense differences in the structure of educational systems in general, and especially in the role of non-university post-secondary training; nonetheless, 12 percent is at or above the level of several countries of Western Europe, and much higher than the countries of Southern Europe. In 1984, the legislated entrance age into Soviet primary schools was lowered from seven to six years. As a result, most children should eventually receive eleven years of schooling. The last two years are not compulsory, and 16 percent of the labor force is described as having only “incomplete secondary school” (Table III.6.24). Nevertheless, very few join the labor force with less than ten years of schooling. Some 60 percent of students complete secondary school on an academic track.


(in millions)

Source: Education and Culture in the USSR: Statistical Compendium (1989).

Table IV.6.24.USSR: Persons with Secondary and Higher Education, 1970-89(Share of employed labor force)












Source: Goskomstat, based on census data.
Source: Goskomstat, based on census data.

(1) Vocational, technical secondary and higher education

Two types of vocational and technical schools provide alternatives to the last two years of regular secondary school: “professional-technical schools” (PTUs) and “specialized secondary schools” (SSUZs). There are now about 8,000 PTUs, 90 percent of which provide regular day-time classes, to which about 20 percent of those completing eight (possibly nine) years of general education attend (Table IV.6.25). The PTUs have traditionally trained skilled manual workers and were particularly important in rural areas. Some years ago, to improve their standing, the PTUs began to include academic training and increased their course length to three years so their graduates are now also recorded as completing secondary education. In addition, students who have completed general secondary school may take one-year courses in PTUs—in 1989, these amounted to 18 percent of such graduates (Table IV.6.26). At any one time there are about 4 million students in these schools (including some 10 percent studying at night) and about 2.4 million graduate annually. Each PTU is associated with at least one base enterprise, and all enterprises with over 2,000 employees are supposed to be associated with a PTU.

Table IV.6.25.USSR: Transition from Incomplete Secondary Education,” 1970-89(In millions of persons)
“Incomplete secondary” output4.34.14.2
Proceeded to:
General secondary2.62.62.5
Daytime PTUs0.10.80.8
Daytime SSUZs0.50.40.5
Labor force1.10.30.5
Source: Goskomstat.
Source: Goskomstat.
Table IV.6.26.USSR: Transition from “Complete Secondary Education,” 1970-89(In millions of persons)
“Complete secondary” output2.02.72.2
Proceeded to:
Higher education0.30.40.4
Daytime PTUs0.20.60.4
Daytime SSUZs0.30.40.4
Labor force1.31.31.1
Source: Goskomstat.
Source: Goskomstat.

All training of the PTUs is carried out within the context of contracts concluded with a particular enterprise, and although individual students are not directly contracting parties, they are expected to take a job in the designated enterprise on graduation. Contracting enterprises provide some practical training on their premises (though schools also have their own workshops), and in cases in which this produces saleable output, the enterprises will pay an appropriate allowance usually similar to the wage of an entry level worker. Other educational costs, including meals, clothing, and boarding where necessary, are paid by the state. This close association with local enterprises reduces the danger that training will be dissociated from enterprise needs; indeed, the practice of contracting with enterprises arose because many of the graduates of vocational schools were having difficulty in getting appropriate employment. This has not, however, prevented criticism that training is poorly organized and fails to respond to new needs.34 There are also complaints that many of the PTUs are old, their own equipment is obsolete, they are short of qualified teachers, and they are not competitive in the market for scarce skills.

A further 10 percent of students with eight or nine years of regular school attend 4,500 specialized secondary schools (the SSUZ—also known as tekhnikomy.) These provide four-year courses to train for careers in elementary education, or for such occupations as paramedics or laboratory technicians. Increasingly, the SSUZs have also taken in complete secondary graduates. In 1988, some 63 percent of the 1.5 million admissions to the SSUZs had completed general secondary school. A high proportion are nighttime and correspondence students.

Graduates of both SSUZs and PTUs are eligible to compete for places at higher educational institutions (VUZs), but in 1986, only 4 percent from SSUZs and 2 percent from PTUs, in contrast to 15 percent from general secondary schools, successfully applied. The VUZs include both universities and specialized institutes and typically offer five-year courses. Those who take at least half of such a course but do not complete it are classified for statistical purposes as having “incomplete higher education.” In the past, graduates of both VUZs and SSUZs were subject to administrative assignment for their first three years, though this has not always been followed and now has been dropped, at least as a formal requirement.35 About five million students are enrolled in some 900 institutions of higher education, of which 69, with nearly 600,000 students, are classified as universities. The majority are directed towards technical careers—over 40 percent of graduates are classified as engineers. Many institutions teach at night or by correspondence; these account for nearly 40 percent of students.

(2) Relationships between training and occupation

A survey conducted in January 1988 suggests that only half of all industrial specialists had jobs requiring their particular qualifications. Four million out of 37 million (11 percent) “specialists with higher and secondary education” in industry were in jobs which required less education than they had received. While complaints of mismatches are so widespread that there is no reason to doubt their validity, it might be noted that the same problem could probably be found in most countries. Moreover, these numbers are very difficult to interpret. Soviet training systems prepare students in a very large number of separate specializations, which presumably implies a very large extent of overlapping content among “adjacent” specializations. PTUs currently train individuals for more than 1,500 occupations, with about 80 percent of the training for 300 occupations. In contrast, the vocational training system in the Federal Republic of Germany trained for 440 occupational categories in 1985, with 60 percent of boys being trained in the 20 most popular categories; a comparable figure for girls was 80 percent.36 The number of occupations for which vocational training takes place in Western Europe continues to fall.

The fact that many individuals are in positions that normally require much less education or training than they have received is a potentially more serious problem. To some degree it is likely to reflect supply factors. First, there is the common tendency for the output of the upper levels of the educational system to grow faster than the demand for its specialized skills; the proportion of those completing secondary education rose from less than one-half in the early 1960s to around 80 percent by the late 1970s.37 Yet, in the 1970s, 40-50 percent of industrial jobs were still manual rather than mechanized. Second, there are the inevitable imperfections of a rigid and excessively ambitious system of central planning. In particular, the need for workers in service industries is neglected compared with training for the so-called productive sectors. The problem of over-qualification reflects excessively egalitarian wage policies in the past which have resulted in pay structures which tend to provide higher incomes for many skilled manual workers than for those with engineering or technical training.

Regarding overqualification as a supply-side phenomenon, some analysts have seen it as a cause of poor worker motivation; successive cohorts are disappointed to find their qualifications mean somewhat less than those of their predecessors and so are discouraged from trying to work hard.38 Far larger proportions of younger cohorts have technical qualifications than do older cohorts which, combined with the rigidity of the labor market, means that there may be many underqualified people in senior technical positions, blocking the promotion of overqualified people at lower levels. This distortion has potentially adverse effects on efficiency and worker motivation as well as being a waste of educational investment.

Given the egalitarian wage structure, the economic return on the expansion of upper levels of secondary education has probably been very low. However, as noted earlier, the link between wages and salaries and productivity is very unclear, and conclusions in this regard must be very tentative. While perverse pay differentials between technical and some manual workers suggest that the private financial rate of return from education will be low, the generous system of free education and student grants, which has lowered the opportunity cost of education, offset this effect to some extent.

The 1986 wage reform had some success in improving the relative pay of salaried workers. This suggests that at least part of the compression of differentials that had taken place over the previous thirty years was induced by institutional factors rather than by the market. On the other hand, in many branches of industry, wage earners continue to earn more than professional personnel, suggesting that to some degree such perverse differentials also represent excessive training. Even if part of the problem may have been poor information, and may suggest to some degree waste of training, it has not resulted in emerging general shortages of technically qualified workers. Moreover, even if there has been a drop in the relative social status of certain occupations requiring higher level training, there is no evidence that this has yet led to reluctance of qualified students to undertake such training.

Although of considerable concern to the authorities, the stock of trained manpower appears large enough to prevent any large-scale emigration—following removal of legal constraints in this respect—from seriously weakening the economy as a whole. Given linguistic and other obstacles abroad to the acceptance of such emigrants, future emigration of highly skilled manpower may not be significant; this does not necessarily apply to individual republics, however, if local conflicts were to lead to substantial internal migration.

(3) Further training and retraining

As in any other industrial economy, there has been a steady process of retraining and upgrading of the labor force, not only at the workplace and in enterprise training centers but in evening and correspondence courses and studyleave programs as well.39 In 1988, some 6 million workers received enough further training, or retraining, to have their occupations reclassified. A further 37.7 million people (nearly 30 percent of whom were managers and specialists) had further training to improve their skills. PTUs retrain about half a million workers a year. There is also an extensive network of state institutes devoted to the upgrading and advanced training of government officials, enterprise managers and other specialists such as medical personnel and teachers. There are more than 600 departments in higher educational institutions devoted to this, some focusing on technical courses, others on managerial training. At their apex is the Academy of the National Economy, which runs courses for very senior officials from across the ministerial and enterprise spectrum, as well as providing training for the staff of the other upgrading institutions.

(4) Effects of recent reforms

The principle of self-financing has spread to educational institutions, and this, together with the increased profit-orientation of enterprises, should encourage even greater closeness between enterprises and vocational schools. There is the risk, however, that the schools may attempt to carry out too much practical training on their own premises in order to enhance the employability of their students. This could raise costs substantially, since equipment in vocational schools is typically used much less intensively than that in factories, and it could also isolate the schools from the needs of the enterprises.

Concern has been expressed that the harder budget constraints and greater profit orientation associated with economic reform might make enterprises reluctant to contribute as much to the cost of training workers, especially if the mobility of workers increases (although as noted earlier, turnover is already fairly high). The 1987 Law on State Enterprises appears to have reduced the demand for newly-trained workers—as evidenced by the difficulty that some training institutions have had in placing their graduates—but not yet to an alarming degree. This situation does, however, require monitoring.

Five hundred of the larger job placement centers have been designated as Centers for Professional Orientation, Employment and Retraining. These have some responsibility for obtaining retraining for the unemployed, if work is not available in the previous occupation, using existing centers such as the PTUs. The process is only just beginning, and the small amount of involuntary unemployment until now has meant that the requirements for retraining have not been very large. In contrast to normal vocational training, state funds have not been available to support workers undergoing retraining, and this has therefore taken place only when enterprises have been prepared to provide the necessary finance. The proposed Employment Fund should take care of this particular problem.

(5) Prospective needs for training and retraining

The institutional basis for future technical and vocational training is well established. A balance has to be struck between an emphasis on general education, which provides a base for the future acquisition of specific skills, and direct vocational and technical training. It is clear that there is no unique solution. There are great differences among successful industrial countries; for example, the role of formal pre-employment vocational training, in contrast to higher levels of general education, is much greater in Germany than Japan. It is important that the USSR preserve at least the existing emphasis on general education, which means a continuation of state financing for part of the program of training institutions. It is also important that enterprises continue, and indeed increase, their involvement in the training process.

The content of training programs and courses will need to change to reflect the needs of a market-oriented economy. This is true both for pre-employment training and for the retraining that will be needed as part of the transition. The amount of such retraining is expected to be large, since a significant proportion of existing activities are expected to prove internationally uncompetitive, and because the structure of the economy will change, in particular with a shift out of heavy industry towards consumer goods and services. The very slow growth of the labor force, especially in the European part of the RSFSR, means that this change in structure cannot be met simply by shifting the training of new entrants to the labor market—it will be necessary to retrain an increased number of adult workers. Estimates run as high as 30-35 million workers changing their jobs in the years to come.

At first glance, the numbers of those expected to be displaced as part of the transition to a market economy seem very large compared to formal training capacity. It is estimated that the PTUs could be used to train some 1.5-2 million additional workers each year, if they were used more intensively in the afternoons. Other facilities could also be used, and also specially created, but they could serve only a fraction of the workers who would be displaced under most projections. However, all the redundancies will not occur at once. It may take time to restructure the stock of physical capital, and it is not essential to restructure the entire stock of human capital in advance. Most worker retraining can be accomplished relatively quickly as new investment is taking place.40 What is desirable is for the two processes to take place at the same time. Special efforts may be needed in particular fields in which Soviet technology has lagged well behind the state of the art elsewhere. A very considerable effort, for instance, will be needed to broaden and deepen all levels of knowledge involving computers—whether programming, electronics maintenance, or computer-aided design. But individuals who have the necessary background for this work may not be those who have difficulty in finding other employment—in these fields, the availability of facilities for part-time upgrading may be more important than retraining the unemployed.

Training in the areas of finance and management is perhaps the most important of all. Only about 5-7 percent of state enterprise directors have any training in economics. Moreover, such training would in any case have little relevance to managerial decision-making in a market economy. Such areas as modern methods of accounting, as an aid to managerial decisions rather than simple bookkeeping; banking and finance; investment choice; the assessment of risks; marketing decisions; and the interpretations of price movements and trends in the world economy are completely unknown questions for most Soviet enterprise directors. A number of courses have been started, both within established institutions and in new cooperative schools, some in conjunction with foreign institutions. But the needs are massive; to provide a very limited training to a handful of people in each of the 46,000 state enterprises could easily involve more than a million man-months of training.

There has been substantial experience in industrial countries with the use of distance learning techniques, predominantly using video cassettes, in the training of corporate managers and of the staff of financial institutions. This approach has a great deal more flexibility than the use of television. While home-ownership of VCRs is not yet wide enough to focus these on home use, they could be used at special local training centers, which may often be within individual corporations themselves. The material should be based on Soviet examples. An early priority must be the training of the teachers themselves.

c. Industrial relations

(1) The role of the official trade unions

Throughout its existence, the USSR has retained institutions for regulating labor relations which, at least in name, resemble those found in industrialized market economies. In particular, trade unions always have had the nominal role of protecting workers from managerial excesses. In fact, however, they were fulfilling quite different functions and this has so compromised them in the eyes of many workers that it is an open question what role the official union movement will play in a truly competitive market economy.

The origins of this ambiguous role lie in Lenin’s dictum that under Communist Party rule the trade unions were to be the “transmission belt” between the “vanguard” party and the mass of the people. This meant that while the unions did have the nominal function of “transmitting” grievances from below, they also had the role of “transmitting” decisions from above. Trade unions exercised quite conventional functions during the New Economic Policy (NEP) period of the 1920s: they organized workers in the still extensive private sector, and strikes also occurred even in the state sector, which was subject to market competition. From 1930 to 1953, the industrialization drive eliminated unemployment and a general scarcity of labor became the norm. During this period, the union movement lost most of its role of protecting workers’ rights, and came to be almost solely an instrument of the state assisting in the drive for greater production, and in particular in enforcing labor discipline.

This extreme distortion of the unions’ role was eliminated shortly after Stalin’s death, and since the mid-1950s the unions have been allowed a distinct role within the Soviet system, though until 1989 it was exercised under the explicit control of the Communist Party. This role essentially has been to administer to the welfare needs of workers. It has included taking responsibility for the administration of both state and enterprise funded welfare benefits and acting as the formal mechanism for employee representation within the enterprise.41

As a result, the funds controlled by the All Union Central Council of Trade Unions (AUCCTU) fell into two major categories. One category comprised the union dues transmitted to the center. This totaled rub 14 billion for the three years 1987-1989. The other was the welfare fund collected from a payroll levy and supplemented by a budget subsidy—this totaled rub 175 billion over the same period.42 This fund, controlled by the trade unions, represented a significant part of the total social consumption fund (SCF) discussed in section 3.

(2) Reform of trade union government

The AUCCTU is seeking to escape from the legacy of poor representation of workers’ interests by changing the constitution and the name of the organization. At a Congress held in October 1990, it was decided that the organization would change from the present highly centralized structure to a General Confederation of Trade Unions (GCTU), in which the branch and/or republican level organizations would be autonomous.

The ownership of the assets accumulated over the years by the AUCCTU remains a controversial issue. Since many entitlements (such as full sick pay) were in the past only available to union members, membership and payment of 1 percent of wages as union dues was de facto compulsory. The assets accumulated from these dues (8 percent of which were transmitted to the center) mean that even though entitlements are not now conditional on membership, loss of access to union owned facilities is so costly that workers do not effectively have “free choice of representatives.”43 The official unions remain responsible for the distribution of discretionary funds, and in particular for deciding who will have access to the sanitaria and rest homes.

There are several possible ways to deal with this issue. One would be to take into account that the AUCCTU was essentially a part of the state since the dues were de facto compulsory and its leadership appointed by the state. On this argument, union assets are state property and should be disposed of as part of the general privatization process. In the interim, such property would presumably be leased by the state to potential users including, possibly, trade union bodies.

An alternative process, for possible consideration at a republican level, might be the one pursued in the Czech and Slovak Federal Republic. In that country, in preparation for its movement to individual unions, the property of the former Revolutionary Trade Union Movement was vested in a “property union”. If there is opposition to the dismantlement of the property,44 another option would be to vest it in a “foundation of labor” to which all workers would belong, no matter to which (if any) union they belonged.45

The case for establishing a “foundation for labor” becomes much stronger if it is decided to continue the operation of a separate welfare fund, paid for out of payroll taxes. If such a separately identified fund is to continue, it will be absolutely necessary to insulate the operations of the fund and its management from the representation of workers. Here, the European examples of authorities operating under tripartite management and using the state’s taxation authority, such as the French Caisse Nationale d’Assurance Maladie, could offer some models. However, for some time it will not be possible to establish true tripartite management, as no clearly identifiable employer groupings will be available. Hence a foundation could well be a more appropriate framework.

The GCTU, as successor to the AUCCTU, is pressing for the welfare fund to continue to be under its administration. However, if the successor bodies to the AUCCTU unions continue to receive and to control such funds, and to maintain discretion over use of existing union facilities, this would compromise the free choice of workers in choosing their representatives.

(3) Collective bargaining structures

The move to a market economy involves the dismantling of the old procedures for determining wages and conditions of employment. In fact some key decisions have already been taken. According to the new Law on Enterprises, enacted in June 1990 and effective January 1, 1991, enterprises other than those totally dependent on the state budget are to be free to determine their wages, constrained only by an “incomes policy” which taxes excessive wage increases and by a floor which will be provided by the existing tariff wage structure.

As of December 1990, there were two alternative legislative frameworks under consideration for collective bargaining on wages and conditions. One was a draft decree of the Council of Ministers, which had been proposed by Goskomtrud. The other was a draft law proposed by the AUCCTU to the Supreme Soviet. The Goskomtrud draft differed from the AUCCTU draft in two crucial respects. First, it extended the right to choose the body to represent workers in collective bargaining which was conferred on the work collective by the 1987 Law on State Enterprises. The draft envisaged that a general assembly (conference) of the work force shall decide: whether it is necessary to conclude a collective agreement; and whether the trade union committee or any other body so empowered shall: (1) engage in collective bargaining on behalf of the work force; (2) be a party to the collective agreement; and (3) ratify an agreement on behalf of the work force.

Both drafts specified that any agreement shall apply to all wage earners, specialists and salary earners, irrespective of whether they are members of the trade union. The draft decree also modified the law passed in 1989 legalizing strikes by specifying that “the parties may also decide that workers shall refrain from striking on matters that fall within the obligations set out in the collective agreement as long as the agreement remains in effect”.

The provision in the 1987 Law on State Enterprises which empowers the work collective to choose collective bargaining representatives, and which has occasionally resulted in the Workers’ Council (STK) or rival “strike committees” being chosen to represent workers, is considered as discriminatory by the AUCCTU, which views this provision as contradicting its reason for existence. It maintains that workers always have the right to elect new union leaders if they are not satisfied with their present ones. The draft legislation presented to the Supreme Soviet by the AUCCTU reflected this view. It specified the trade union committee as the party to collective bargains at the enterprise, specified that the trade unions, through their corresponding elected bodies, “have the right” to conduct collective bargaining on behalf of workers, and conceded only that trade unions “organize the collection of workers and work collectives’ proposals for inclusion in collective agreements”. However, the draft did specify that the draft agreement is to be approved by the general assembly (or conference) of the work collective before being finalized.

The other issue on which there is disagreement is that the official trade unions want legislative recognition of their right to conclude “conventions” with union, republican and regional government bodies. These conventions would cover at the all-union level both basic guidelines for wage formation and social security protection including, for example, a specification of the minimum wage, indexation procedures for other wages, and tax concessions for enterprises which engage in socially constructive activities. The breadth of these proposals seems to have been inspired by the phenomenon of “strike committees” making demands on regional governments as well as on enterprises; it appears that the union confederation is seeking to have itself recognized as the spokesman for workers in such negotiations.

The proposals of Goskomtrud, for these reasons, seem to provide a much better basis for the establishment of a collective bargaining system than do those of the trade union federation. They provide an institutional context for bargaining on the terms and conditions of employment in enterprises that will be changing their orientation from fulfilling central plan directives and filling state orders, to serving the uncoordinated needs of consumer, producer goods and export markets. This bargaining process will require the abandonment of the doctrine that managers and workers have the same interests as joint beneficiaries of the development of state property, and the recognition of the creative tension between the role of the enterprise in serving the markets to which it sells and the desire of workers for improved living standards and job security. This tension is the reason that collective bargaining in a market economy involves a clear divide between the management side and that representing the workers, although the way such tensions are addressed varies widely between, and within, Western market economies.

Nonetheless, such bargaining cannot take place in each enterprise in isolation. Workers’ demands will inevitably be affected by general price rises and shortages and by changes in collective provisions of basic facilities such as housing. There will thus be a commonality in issues arising across enterprises and unless some common approach is reached a “wage-wage” spiral between enterprises could easily be generated. Such commonalities of approach could be discussed through regional or branch negotiations between organizations of employers and federations of workers’ representatives, which would result in guidelines for workplace negotiators.46 However, formally binding centralized negotiation, common to some Western European countries, should be approached with caution. For the time being, only the existing state-sponsored unions would be in a position to participate in such negotiations. Moreover, the size and diversity of a country like the USSR would make centralized bargaining impractical.

The new Enterprise Law clearly distinguishes between “owners” and “workers”, and provides for the establishment of an enterprise board with equal representation from each side. However, these boards could easily be dominated by the management side if the employees are disunited and disorganized. The reason the latter is likely is that workers’ experience with official unionism over the last half century may make them skeptical and apathetic about any attempt to “organize” the workers’ side. Furthermore, the stringency of the economic climate and the threat of unemployment will make them very reluctant to defy managerial authority, particularly if the enterprise has been privatized and that authority is therefore not under challenge as being a continuation of traditional state control.

In summary, the structure of collective bargaining must be adapted to a wide range of conditions. Decentralized collective bargaining at the individual enterprise level, although a flexible instrument, can lead to leap-frogging wage settlements in situations in which enterprises compete among themselves in accommodating labor demands. Centralization of bargaining may therefore be beneficial in limiting wage price spirals but may not be practical in a country as large and diverse as the USSR. To provide market signals and incentives, wages must reflect productivity performance and regional differentials in cost of living. More appropriate, therefore, would be regional or economic branch negotiations between employers’ organizations and federations of workers’ representatives to establish guidelines for plant-based negotiations.

(4) Wage bargaining

Some features of the current Soviet wage system have clearly influenced the development of enterprise labor relations, and will now need to change. In particular, the failure to properly account for capital usage, and also for environmental and personal health degradation, resulted in a tendency to emphasize material output increases alone—both in the setting of production targets, and in the construction of bonuses within the enterprise to encourage the attainment and surpassing of such targets. In this context, local union officials came to identify their joint functions of defending workers and maximizing production through cooperation in the design and implementation of bonus systems that were intended to maximize output and earnings.

The local union officials were not particularly successful in this respect. First, failure to account for other input costs meant that capital goods and other inputs were used inefficiently. Second, any success in raising output above the norm carried the risk that the norm would be revised upwards through the process of “planning to the achieved level”. Third, the provision of many goods and services (in particular housing) outside the price system meant that cash wages were often not particularly important for achieving a better life.47 Finally, recent reforms have resulted in a growth of aggregate wages at a time when output has stagnated, leading to a general shortage in goods on the market which has further discredited cash wages as a means of motivating workers.

In a market economy, bargaining between workers’ representatives and management occurs in a context in which enterprises are required to pay the full cost of capital, and to implement adequate environmental and health and safety standards. Western experience suggests that efficiency and productivity will be enhanced if there is a move away from payment systems based on individual and even group bonuses (particularly if they are based on material production targets), towards a system of time-rate payments which are adjustable on the basis of the overall factor productivity of the enterprise.48 However, in the transitional period, such a bargaining system will have to evolve within the context of overall wage controls, i.e., the pursuit of an incomes policy (see section d).

(5) Occupational health and safety

The consequences of placing a priority on material production over both the health and safety of workers and the environment in general is now evident throughout the USSR. Western experience has shown that health and safety regulation at the workplace generally requires worker involvement if it is to be effective. It also means giving local authorities the responsibility required to enforce national standards, and relying on workplace health and safety committees as a “front line” inspectorate. In the light of Soviet history on this issue, there must now be a new, relatively uncompromising outlook. It needs to be accepted that work sites which are unsafe must be brought up to standard, and that if this cannot be done at a cost that is consistent with the financial viability of the workplace, it will have to be closed down.

In order to bring about even this very basic representational function, a fundamental change in existing attitudes to worker compensation will be needed. The current paradigm for representing workers’ interests is to associate them with the production process by basing a large part of earnings on production levels. Given the low level of real wages, and their possible further reduction as a result of structural dislocation in the move to a market-oriented economy, such a wage system encourages and almost requires workers to ignore health, safety and environmental standards in order to exceed production targets and thereby earn large bonuses. It will be necessary to re-orient production processes away from this system to one in which managers will attempt to maximize total factor productivity. If defined with proper breadth, this will also involve taking account of the true cost of imperiling workers’ health and safety, and of the cost of environmental degradation. This probably requires that enterprises be required themselves to meet the full cost of compensation for industrial safety and accidents, and to pay stiff levies for breach of environmental standards (see Chapter V.1). A workers’ compensation fund may be necessary, but it should only be drawn on to compensate workers who come from enterprises that have gone out of business. Western experience is that workers’ compensation insurance which is not performance rated encourages employers and workers to ignore long-term health and safety risks, relying on the general insurance pool for compensation.49

(6) Industrial democracy

The confusing nature of social changes in the USSR, and the absence of a clearly defined and legitimate focus of authority, has led in many instances to workers’ movements taking on both a political and an industrial nature. One by-product of this has been a tendency to include the demand for workers’ control of their enterprises in more general calls for democratization. In this, the Soviet workers’ movement has followed in the steps of some workers’ movements in other East European countries, most notably Solidarity in Poland.

In principle, the 1987 Law on State Enterprises did in fact confer on workers, through their enterprise council (STK), considerable authority over their enterprises, including the right to vet the appointment of the Managing Director. In practice, these rights have been very largely formal, with the existing apparatus controlling the elections in most instances. However, this has not been the case everywhere, and there is already a perception—within the official union movement at least—that some of the decisions of the workers’ councils were not in the interests of the enterprises. These reservations have now been accepted in the new Enterprise Law, which will cover both state and private sector enterprises. It provides for the establishment of enterprise boards, whose composition will be 50 percent from the owners and 50 percent from the work collective (including managers). These enterprise boards will have some functions similar to those of the supervisory boards in the German system.

The crucial issue is who will constitute the “owners” side, both on the enterprise boards and in any collective bargaining which takes place. This goes to the heart of the privatization debate. Some proposals for privatization envisage part or all of the ownership rights to enterprises being given or sold (presumably under credit terms) to the workers in those enterprises. Such proposals are particularly seductive when the scale of the privatization task is considered. They raise serious equity issues. Under the existing system, workers in successful enterprises have already been relatively privileged compared with those in the service sector or in enterprises with less favored positions. Moreover, the distinction between the workers’ side and management disappears—both on the enterprise board and in a collective bargaining context—if the owners who appoint the bargaining agents are the workers themselves. This is particularly the case if ownership rights are not transferable, and are vested purely in workers as long as they are employed by the enterprise. A system of labor management is prone to overconsumption and underinvestment at the enterprise level. For this reason, it is very important that any attempt to involve workers in the privatization process should clearly distinguish their role as owners from their role as employees. This would require that any share allocation to workers be made on an individual rather than on a collective basis.

d. Wage determination and incomes policy

(1) Incentive structure and wage reforms

In the past, the centrally determined tariff wage structure had little or no relation to market forces, productivity differences or availability of skills. Instead, it appears to have been based on an egalitarian notion of fairness, together with a rather ideological hierarchy of sectors and skills, but with little weight attached to the incentive function of wages. As noted in section 3, past trends in money wage differentials may be broadly characterized by a relative stability of sectoral differentials—in particular, within a given skill category—and by a gradual narrowing of skill differentials until 1986.

The lack of a proper incentive structure has long been a concern of the central authorities and has inspired many reform attempts since the mid-1960s. In particular, labor requirements in certain sectors, occupations or regions have been accorded a corresponding raise in pay, i.e. so-called “compensating differentials” to offset the disutility of working and living conditions in these sectors or regions. In addition, poor productivity performance and the need to enhance workers’ motivation have led to a proliferation of so-called material incentives—bonuses, piece rates, benefits—that alter the structure of wages from that given by the tariffs. Material incentives have been tied to quantitative indicators of productivity and to the fulfillment of plan targets in physical terms. The focus on volumes has frequently given rise to perverse incentives, notably in terms of product quality. It has also influenced the evolution of the industrial relations system, with local union officials viewing their function as one of defending workers through cooperation in the design and implementation of systems that maximize earnings. But the criteria for receiving bonus awards were often so easy to attain that material incentive payments became a fixed addition to basic pay, without requiring much effort on the part of the work force.

The lack of real incentives in the traditional system for workers to enhance efficiency and for managers to economize on labor has prompted several attempts to reform the wage remuneration system. The Shechekino experiment, which began in the late 1960s and had been applied to half the industrial labor force by the mid-1980s, assigned the enterprise a fixed wage fund, instead of having the wage fund depend on the number of employees. This meant that the enterprise was entitled to reallocate part of its saving on wage costs to the bonus fund, making salary increases dependent on productivity increases. The brigade system, which was encouraged in the first half of the 1980s, was an attempt to apply the Shechekino method to small groups within enterprises. While the incentives embedded in these remuneration systems could in principle stimulate workers to increase productivity, in practice they had limited effects. Efficiency bonuses were subject to official regulation, which restricted supplementary payments to 30 percent of the base wage, thus blurring the incentives for marginal productivity increases. A more fundamental reason for the ineffectiveness of these schemes was the tendency of the central planners to raise the output norm in line with the efficiency increases actually obtained, thereby all but eliminating the incentives for workers to be more productive.

The 1986 wage reform attempted to strengthen the link between the performance of an enterprise and the basic wages of its work force. The reform called for a general increase in tariff wages of 20-25 percent, with qualified workers and engineers receiving higher pay raises. The changing structure of tariff wages was intended to reverse the previously mentioned compression of the skilled-unskilled wage differential, which the authorities had judged to have had adverse effects on labor allocation. In contrast to previous tariff wage revisions, enterprises were only allowed to raise tariff wages if they could pay the increasing costs out of their own funds. This meant that the tariff wage increase had to be financed by reducing the bonus fund. But if this did not suffice, the tariff wage could only be raised by increasing labor productivity or by saving on other inputs. The wage reform had been tried out on an experimental basis on the Belorussian railway system during 1985-1986, with large gains in labor productivity, partly due to labor displacement, and steep gains in average wages. As discussed earlier, the 1986 wage reform is credited as one of the factors behind the reduction of employment in state industrial enterprises since the beginning of 1987. However, the impact of the reform in stimulating labor productivity was weakened by the fact that it was implemented in an environment set by the 1987 Law on State Enterprises.

The 1987 law established profit-related pay with the aim of forging closer ties between salaries and economic results. The crucial element was that all state enterprises should be self-financed, i.e., cover their costs (including the user cost of capital and cost of product development) from their revenues. With respect to wage determination, the law offered enterprises a choice between two alternatives, depending on the share of salaries which was related to profits. One possibility was to relate total salaries to profits, and the other was to continue with tariff wages but to supplement them with a profit-related bonus component. The overwhelming majority of enterprises opted for the second alternative. With profit-related pay, increased labor productivity showed up in higher salaries for the work force.

Other provisions in the 1987 law, as well as its administration, nevertheless weakened the incentives for enterprises and the work force to economize on labor use. The law gave enterprises some freedom to set their own prices (see Chapters II.2 and IV.1). With lack of competition in the product markets, this opened the way for enterprises to finance wage increases by increasing their prices rather than using labor more efficiently. The pressure on management to raise wages without shedding labor was also increased by the law giving the work collective a greater say in management decisions. As noted earlier, the work collective was given the right to vet the appointment of the manager of the enterprise, and the manager was made answerable to the general meeting of the work collective for his activities. A serious weakness in the implementation of the law was that branch ministries did not allow the share of the profit-related pay component to become a significant element in workers’ salaries. As a result, salaries responded only modestly to increases in profits, weakening incentives for the work force to become more efficient. To some extent, the law also met with political and social resistance due to a low tolerance for large income differentials. More fundamentally, however, it failed to produce real improvements in incentives because in a shortage economy pay raises do not necessarily entail greater command over consumer goods and services.

(2) The current situation and policy tasks

The twin objectives of the 1986 wage reform were to restore skill differentials by granting higher pay raises to skilled workers—especially managers and engineers—and to favor some top-priority sectors. As noted earlier, this policy has resulted in a partial reversal of past trends. But the gap between average wages in the “productive” sectors and some of the more labor intensive “non-productive” ones—education, health, culture—is still very large. More generally, in the process of adjusting the relative wage structure and trying to improve incentives, the authorities have lost control of the wage determination process. Recent wage inflation may be partly attributed to continuing labor shortages in some branches of the economy, as well as to social unrest. In addition, competition for the best workers stemming from newly created private activities (cooperatives, self-employment and black market activities) has increased the pressure on wages in the state sector. More fundamentally, however, increases in money wages that have outpaced both price inflation and real productivity growth have reflected growing enterprise liquidity and weakened central control over enterprise wage payments (see Chapter II.2).

From a policy point of view the current situation, therefore, is particularly complex and difficult. The transition to a market economy would require, in principle, the withdrawal of the government from the wage determination process, thus giving room to decentralized wage setting, a widening of wage differentials and the inclusion of non-tariff elements in the actual market wage. However, the urgency of the macroeconomic stabilization goal and the need to cushion social hardship during the transition period require the government to regain some control over wage determination. At least in the short term, there is a need to set a ceiling for wage increases in order to prevent a wage-price spiral and to maintain minimum wage floors in order to protect the economically weak against inflation.

Is there nevertheless some room for decentralization and wage flexibility? The Employment Law and the presidential guidelines foresee the maintenance of the current tariff wages as minima which cannot be undercut. Otherwise enterprises would be free to set wages. This amounts to turning the current basic wage structure into a complex structure of minimum wages. Actual earnings, however, might decrease if bonuses and/or benefits are cut. Since this provision does not protect real wages, some form of partial indexation has been envisaged by the authorities as well. This approach would permit a degree of wage flexibility, however, only within the lower bound given by the indexed minima and the upper bound imposed by an incomes policy involving a cap on maximum wage increases (see below). Under the circumstances, this policy could be viewed as a reasonable solution. Retaining the current tariff wage structure as a floor for future wage settlements would also have the advantage of enforcing some stability in the transition process by providing basic guidelines for wage setting—pending the development of a proper collective bargaining structure as discussed in section c.

(3) Incomes policies

The need to halt inflationary pressures and prevent a wage-price spiral in a situation of pent-up demand and price liberalization lies at the heart of the stabilization problem. This raises the question of the role, if any, an incomes policy might play as part of the macroeconomic stabilization discussed elsewhere in this study. The experience from some other countries is reviewed in Appendix IV.6-1.

In the USSR, a tax-based incomes policy has been pursued since 1989. The initial formula of the so-called Abalkin tax was that enterprises would become subject to a tax penalty if their total wage bill—money and in-kind wages—increased in excess of 3 percent. The tax failed to limit the steep rise of wages for two reasons. First, it was not effectively enforced. Since it was an across-the-board formula, which would have entailed a considerable degree of rigidity, several exemptions had to be granted. Once this process had started, it was impossible to refuse a proliferation of exemptions. Second, the arrangement did not stop enterprises from granting taxable pay increases since they did not operate in competitive markets or under hard budget constraints. They could in some cases avoid losses by increasing prices; in other cases, losses were covered by state subsidies or credits from the banking system.

The “consumption fund tax”—adopted under the Act of June 14, 1990 to become effective January 1, 1991—is also intended to curb excessive increases in the total wage bill of enterprises. Under this scheme, an increase in the wage bill will be liable to tax unless it is accompanied by a reduction in the share of labor remuneration in value added (excluding profit taxes) by a certain margin. The authorities can vary this margin across sectors, but the legislation contains central guidelines on determining the wedge. The tax rate schedule has 4 brackets, with very steep increases in marginal rates for wage increases which exceed norms by up to 3 percent, the top rate being 200 percent for increases in excess of 3 percent (for details see Appendix III. 1-1).

At first sight, this formula looks like an improvement compared to the previous attempt in that it does not freeze inter-firm differentials. But while this may be an advantage in the very short term, it could make the policy more vulnerable over time. The higher remuneration of workers in well-performing enterprises is likely to induce workers in other enterprises with the same or similar skills to seek wage increases in order to catch-up with their fellow workers. These forces will be stronger, the more developed the sense of equity and solidarity in the work force.

Other shortcomings may arise as well. First, the policy may not be easily enforceable, and collection lags are likely to dampen its efficacy. Second, even if the policy is enforced toughly and widely, and supposing that subsidies and noncommercial bank credits will no longer be available, firms may seek to pass on wage increases into higher prices for their products, and may be successful within limits at doing so, especially given the degree of monopolization in the economy. In other words, the current formula provides a mechanism for almost fully indexing the enterprise wage bill on own-product prices. Finally, a potential problem with an incomes policy based on the total wage bill is that it may introduce an incentive for “insiders” (workers with a strong position within the enterprise) to increase their own wages by reducing the number of workers with whom they have to share, or to retard the hiring of new workers in enterprises that should be expanding. In other words, this formula may encourage labor shedding and discourage hiring which is unrelated to productivity developments. The alternative would be to base the wage norm on average wages. This would be more advantageous in terms of protecting employment, but would create an incentive for enterprises to hold on to lower skilled workers and thus retard the upgrading of the human capital stock.

Whatever the precise formula chosen, it would be desirable to detach the wage norm from the enterprise’s own results. Another essential element is that the policy must be enforced rigorously and consistently. Since it will come at the expense of flexibility, it should be regarded as a purely transitional device. Finally, the experience of Eastern European countries suggests that even in the short term it will only work if it is accompanied by macroeconomic stabilization, the imposition of hard budget constraints and a decisive move towards competition in product and factor markets.

e. Social protection

Relative price adjustments affect the real living standards of all, including those not in the labor market. This results from administered wages and transfers being determined in the context of a fixed set of prices. As discussed in section 3, there is a fairly equal distribution of incomes at a relatively low mean in the USSR, given the extensive free, subsidized and in-kind provisions. While this has not been adequate in preventing poverty, it is clear that major price adjustments without regard to wages would lead to further dramatic increases in poverty. Issues of social protection also arise in respect of existing programs, particularly for pensions and allowances to cover various contingencies. These programs had deteriorated in terms of adequacy of cover well before the formulation of the present reform proposals, and the transition to a market economy will put them under additional pressure. Moreover, provisions will be needed for new contingencies, particularly the unemployed.

(1) Income protection during the period of price liberalization

It is expected that a market oriented price system in the USSR would lead to an increase in the prices of most agricultural goods and raw materials, but also a decrease in the prices of some manufactures. This report concentrates on food prices, which are crucial in establishing standards of living, although the discussion could be extended to other items of essential consumption such as housing.

(a) Reform proposals

The Abalkin reform program had suggested an administered adjustment in producer and consumer prices, with subsidies on meat and milk products being phased out. However, a full “lump-sum” compensation was proposed for consumers. The Shatalin plan was more cognizant of the market mechanism, arguing for the free interplay of demand and supply, although during the initial transition period prices of 20-30 percent of commodities were to be frozen to limit the possibility of a potentially inflationary wage compensation-price spiral. While the presidential guidelines echo many of the Shatalin proposals on price reform, they recognize that freezing producer prices for essential goods could have adverse supply responses and argue for fundamental changes in the “method of paying out subsidies,” with considerable leeway for adjustments for producers and rationing at the local and republican levels.

The presidential guidelines advocate the full indexation of pensions and transfer payments. There would be partial indexation for those on low and fixed incomes. While the precise formulae to be used for the latter were not disclosed, the guidelines suggest that the range of indexation, for example for those in the so-called nonproductive public sector such as teachers, would be between 50-70 percent with the degree of indexation declining as income levels rise. The extent of indexation would depend on the choice of commodities to be included. The basket of goods to index would include a heavy weighting of essential items provided at administered prices. This would limit the budgetary outlays on account of minimum wages and transfers such as pensions, allowances and unemployment compensation. It had been determined by the Government, in agreement with trade unions, that if the prices in the chosen basket increased by more than 2 percent, there would be automatic compensation, but not otherwise. The different proposals concerning the initial level of minimum wage to which indexation would apply range between rub 70-100. If no adjustment is made to the minimum wage, then indexation would apply to the current figure of rub 70 per capita per month.

(b) Measures for the transition

There are number of different methods of ensuring social protection in the transition period. They are not mutually exclusive, nor is any given solution likely to be suitable for all regions of the USSR, as consumption patterns and administrative capabilities vary. For a given adjustment in prices, attempts to compensate households fully would likely exacerbate inflationary trends; moreover, less than full compensation could be attractive on distributional grounds. The options include: (1) undifferentiated benefits which (e.g., through a system of coupons) could serve to ensure the provision of a limited quantity of goods at predetermined prices, or the cash equivalent, to all in a particular region; (2) food stamps, which could be either means-tested, or targeted by individual characteristics, such as age or disability; and (3) means-tested cash compensation. Amongst these alternatives, there are two basic questions: whether the benefits should be universal or restricted in some fashion; and the method of administering the chosen option.

(c) Universal provision

An important method of protecting consumers is to continue to provide certain quantities at subsidized prices. Market clearing prices would obtain for producers and for quantities demanded in excess of the ration.

Limiting subsidized quantities consumed to the average consumption level of, say, the poorest 10-15 percent of the population, would restrict the budgetary outlays. In providing fixed supplies at given prices, undifferentiated benefits would also ensure that the prices of essentials are not bid up beyond the reach of the poor. A similar system adopted during World War II in the United Kingdom ensured that nutritional standards actually improved with the assurance of an adequate provision of a balanced diet of proteins and carbohydrates, despite the disruptions of supply. High market prices of milk and meat would otherwise have forced a substitution of consumption by the poor towards less nutritious starches and potatoes. Rationed benefits in kind would lead to considerable reductions in budgetary outlays, compared to a general subsidy, as only limited quantities of a few goods would be involved. Targeting would reduce costs further. However, in the Soviet case, universal rations would minimize additional costs that arise through the indexation procedure. This is because the prices of rationed commodities would be included in the weighting of a basket of goods used to determine changes in the index.

While cash transfers are recommended on grounds of simplicity and administrative ease, resalable in-kind benefits are equivalent in principle. If there are sharp price movements and the eventual price level is uncertain, then in-kind provision may have the advantage of shielding consumers from the uncertainty involved. Cash transfers may be inadequate in ensuring guaranteed consumption minima, but have the advantage that budgetary outlays are likely to be known in advance. However, if there are additional cash payments to compensate for unforeseen (or unplanned) price rises, then this advantage would be lost. Moreover, outlays with respect to in-kind transfers would be limited by the international price of importables.

(d) Targeted options

If it were relatively costless to identify the poor and vulnerable, then the preferred approach would be targeted transfers to persons below or around the poverty line. Universal provision could actually become more targeted in practice through the public supply of relatively low quality types of meat or grains as part of an in-kind transfer. The method of delivery would, however, vary according to administrative capabilities at the local or republican levels. Rising prices for agricultural products, to the extent that these were not offset by input price changes, should lead to rising incomes for farmers; in this case it would not be necessary to compensate such groups for the price changes. Moreover, provision of rations in rural areas may pose administrative difficulties.

Given the regional differences in consumption across the union, it makes little sense to determine all-union norms for in-kind transfers. To limit the annual public provision of meat or dairy products to the levels consumed by the poor (defined as those with less than rub 75 per capita) in Uzbekistan, or 18 kilograms per capita of meat and 145 kilograms of dairy products, would lead to considerable hardship in the RSFSR, since the poor in the latter consume more of these items than the average for all households in Uzbekistan (Table IV.6.15). Conversely, a provision determined by the minimum quantities of such items consumed by the poor in the RSFSR (38 kilograms of meat and 305 kilograms of dairy products) would be overly generous in Uzbekistan, and would be tantamount to a general subsidy in some regions.

One approach to targeting would be to issue food stamps to the needy. If, say, 30 percent of the population in the USSR is below the current poverty line, limiting the transfers to these households could, in principle, reduce the expenditures on universal provision correspondingly. However, as discussed in section 3, there is a concentration of individuals at or just above the current poverty line. Therefore, the necessary adjustments of the poverty line in consonance with price increases would spread the net much wider, reducing the budgetary savings relative to uniform provisions. Administrative costs and feasibility are also important, as is the need to generate wider, political support. Another difficulty with means-testing, as noted earlier, is the high marginal tax rate that is involved at the point of withdrawal of the benefit, which could result in a “poverty-trap”.

Food stamps or targeted benefits provided on the basis of household characteristics, such as to the elderly, families with a large number of children (defined in the Soviet context as exceeding two), or the disabled, might be easier to administer than the means test, and would tend to avoid the poverty-trap disincentives. Such individuals could be easily identified on the basis of current benefit programs. Another category of persons in need is likely to be the unemployed and those in retraining, or those employed under the safety net of last resort—public works. Such persons also could be provided with the food stamps along with other benefits.

The lower direct cost of the targeted options could be offset (as the example below suggests), however, by additional expenditures on account of the full indexation of pensions and benefits, as well as the adjustment of wages of those in the nonproductive sector such as teachers, the military, doctors and nurses and so on. Compensation cannot be made only to those on the poverty threshold, as this would lead to a further compression of relativities (see section d). Thus, adjustments might entail fairly major additional expenditures on (non-enterprise) public sector wages.

Where the majority of the population in a region is at a relatively low income level, and is in danger of poverty after the price changes, or if incomes cannot be easily verified, then universal provision (perhaps targeted by quality differentials) may be the most cost-effective option.

(e) Budgetary implications

Despite difficulties in determining regional variations in consumption patterns, and the local solutions that are likely to emerge, it is important to establish overall budgetary implications for the social programs. Under present rules, social protection expenditures on pensions and allowances would be limited to 26 percent of the wage fund of enterprises (plus one percent for unemployment insurance) and the one percent on individual wage receipts (see below). Increased expenditures arising from indexation would have to be accommodated through a reduction in contingencies covered or benefit levels, or if this were not possible, would emerge as a charge on the budget. In the following, the magnitude of the likely direct subsidies are estimated.

Producer price increases from January 1991, in the absence of adjustments in consumer prices, were estimated to increase food subsidies from approximately rub 100 billion, or around 10 percent of GDP, to rub 160-180 billion (with some estimates as high as rub 200 billion). Methods of reducing this subsidy, without excessive burdens on the poor, are discussed below.

On average, the consumption of key food items in quantity terms by those under the official poverty line is about 43 percent of average per capita consumption (Table IV.6.15). Guaranteeing the maintenance of the existing consumer basket for the poorest in the USSR at existing prices, and no more than this to richer households, would entail a fairly limited subsidy. A system of universal benefits in kind could reduce the direct subsidy to around 43 percent of projected outlays, or to around rub 70 billion.

A system of food stamps or targeted cash transfers, based on a poverty line with around 30 percent of the population classified as poor, would reduce direct budgetary outlays to about one third of the universal benefit, or to around rub 24 billion. However, with the full price rise, it is likely that the number of persons classified as poor would increase, reducing the savings through the targeting.

The full price changes may also trigger adjustments in (non-enterprise) public sector wages that would have to be met through the budget, as discussed above. Even if concentrated initially on low income wage earners, the indirect adjustment of wages could be expensive if relativities were maintained. With a wage bill in the non-productive sector of around rub 180 billion, a 60 percent increase in the price level would involve an additional expenditure of rub 108 billion. Adjustments on account of pensions and transfers would also be more than in the rations case, with lower aggregate price effects.

The choice of an appropriate mechanism to protect the vulnerable is more likely to turn on administrative feasibility rather than overall cost, and for this reason should be decided at a republican or local level. Food stamps may be easier to administer in Estonia, for example, than in Uzbekistan. It is apparent, however, that there is significant scope for reductions in the subsidy bill, regardless of the option adopted, without adversely affecting the living standards of the poor in the USSR.

(2) Reform of pensions and other transfers

There are a number of policy issues relating to the system of social protection for existing contingencies, including the elderly and households in need of assistance. Some of those relating to living standards were discussed in section 3. In this section the provisions for established coverage of life-cycle contingencies are examined: old age, disability, maternity and child birth and the special needs of children and particular types of households.50

(a) Pensions

New pension provisions designed to protect old-age pensioners from declining standards of living were to come into effect on January 1, 1991. Minimum pensions, aligned to minimum wages, are to be determined on an all-union basis, while the pension law gives individual republics a free hand in increasing old-age pensions beyond the centrally guaranteed level. Changes in minimum pensions would be reflected at all pension levels, thus maintaining relativities. Minimum pensions are to be indexed to prices as well as to minimum wages, so that price rises would be reflected in the minimum pension even if there were no change in minimum wages. Finally, the new law provides coverage to all individuals irrespective of source of incomes, provided contributions will have been made to the Pension Fund. A guiding rationale for the reform of the Soviet social security system is the need to separate pensions, which should be organized on social insurance principles, from other allowances for households. To this end, two separate funds are to be established, the Pension Fund and the Social Security Fund. The intention is to make these funds viable without calls on general revenues, taking social expenditures off-budget.

Given the demographic profiles of various republics, a Pension Fund would redistribute resources from poor (but relatively “young”) Central Asian republics to richer (and demographically more “mature”) regions such as the Baltic republics. The initial proposal was, therefore, for the Pension Fund to include not only items that are normally covered under social insurance, such as old age pensions, disability and loss of breadwinner pensions, but also two groups of child allowances, one universal and the other apparently means-tested. The inclusion of these allowances would change the direction of resource flows towards poorer regions. Thus, the Ukraine, with potential collections of rub 20 billion, for example, and expenditures including allowances of rub 15 billion, would contribute rub 5 billion in principle to the all-union fund. The inclusion of allowances in Uzbekistan, however, would more than offset the surplus on pension account, resulting in a deficit of around rub 2 billion to be met out of the central fund.

In principle, this appears to be a sensible and well thought out scheme. However there are difficulties which may preclude the effective establishment of an all-union fund and vitiate the objective of moving these expenditures off-budget. The major difficulty is partly administrative and partly political. As noted earlier, the USSR does not have a system of individual social security identification numbers, and Goskomtrud thus far has no administrative powers in this area. Collections, verification of work records, and disbursement of benefits are in the republican domain. Even the proposed Pension Fund would in fact be composed of 15 republican funds, with surpluses transferred to and deficits met from the all-union fund. However, it is not clear whether republics with surpluses would be willing to part with them. Elimination of aggregate surpluses with a reduction of the contribution rate, so that the overall system remains essentially pay-as-you-go, does not remove the problem as interregional transfers would still be involved. The difficulty is that different republics may have varying objectives, with the Ukraine, for example, probably keen to build up reserves to cater for increasing numbers of the aged, but with Uzbekistan likely to be more anxious to use surpluses on pension account to provide benefits to poor children.

The current legislation setting up an all-union fund would be the preferred option on economic grounds, during the transition and subsequently. However, if this were politically infeasible, then the second-best option would be only to legislate minimum benefit levels on an all-union basis. Republics would then be free to add to these (as at present), but would determine funding options and contribution rates. Effectively there would be 15 republican funds for pensions, but without the possibility of inter-fund transfers. Allowances could then be separated from the pension account and be treated analogously, although recourse to republican or union revenues might be needed in particular cases.

To the current payroll tax for pensions and social security, it would be necessary to add the taxes for employment measures and unemployment compensation (see section a), and eventually health care.51 As noted earlier, there are doubts about whether the level of payroll tax initially proposed was in line with the realities of a changing labor market situation, with considerable pressure on enterprises and the likelihood of substantial short-run austerity and unemployment. Given hardening budget constraints, the tax would eventually have to be borne by adjustment within enterprises, possibly accelerating the labor shedding process. High contribution rates may also be difficult to implement, particularly with an expanding private sector, and may have the effect of reducing the contribution base as has happened in several Latin American countries. There are also some doubts as to the ability of the currently underdeveloped capital market to maintain the value of surpluses. Thus expenditures for future retirees may still require additional increases in contribution rates or subsidies from the budget. In October 1990, the Government announced that the proposed enterprise contribution rate had been reduced from 37 percent to 26 percent of the wage fund. 52 This drastically reduces planned reserves, and returns the overall system towards pay-as-you-go funding, but without Government contributions. The ageing of the population will increase outlays fairly rapidly, however, necessitating rising contribution rates. An alternative, akin to the French repartition system, would be to fix the contribution rates for a period of time and to adjust benefit levels, eligibility and scope of allowances.

(b) Allowances

Under the reform provisions, most social allowances are to be paid from the Social Security Fund, although two major allowances would continue to be paid out of the Pension Fund. The first is a universal child allowance for all children from birth to 18 months, with benefits at 100 percent of the minimum wage paid to working mothers, or 50 percent for non-working women. The second is a means-tested child benefit which would be paid from 18 months to age 6, at 50 percent of the minimum wage, irrespective of whether the mother works outside the household or not. The means test is to be based on family per capita income less than twice the minimum wage (or rub 140, currently), to be determined and administered at the local level.

Other means-tested benefits would be contracted with republics and local authorities with support from the central fund. However benefits and minimum criteria would be allowed to vary, given the differences in demographic characteristics. There are major incentive problems in administering a means test at the local level with union or republican funding, as there is no sanction against a proliferation of expenditures (see section 3).

The other allowances proposed to be paid out of the Social Security Fund include (percentage of minimum wages in brackets): (1) maternity leave, 18 weeks before birth (100 percent); (2) lump-sum benefits after birth, equal to 3 months minimum wages; (3) unmarried mothers’ allowance (50 percent) for a child until 16, or 18 if the child pursues studies at a higher education establishment; (but until the child is 6 years old the mother receives full allowance (100 percent)); (4) wives of military personnel (100 percent); (5) widows with children and persons taking care of orphans (100 percent) to age 16, or to 18 if studying; (6) guardians of children (100 percent) to age 16 or 18; and (7) an allowance for looking after a sick child under the age of 14, for not more than 14 consecutive days, depending on the salaries of the person taking leave.

These allowances consolidate the myriad of ad hoc provisions which currently prevail, and offer cover for most life-cycle contingencies. However, these are seen as enabling provisions to prevent poverty. A solution to the basic problem of existing poverty depends primarily on growth prospects to generate the resources which would be required for the social protection provisions to operate effectively. The tradeoffs may be seen in the reduction in the proposed payroll tax to 26 percent. This would reduce the resources for the allowances from around rub 25 billion to rub 16 billion in 1991. The response has been a proposal to move the grant for the birth of a child (allowance 2) to the Pension Fund, and the allowances for unmarried mothers, military wives and guardians (allowances 3, 4, and 6) to republican budgets. The guardians’ allowance would be reallocated to the educational budget of the republics. Clearly, birth grants are viewed as essential, and the other benefits to be reallocated to republican budgets are perceived as “dispensable”, since they cannot be forced on the republics.

The effects of minimum guarantees on the expenditures to be incurred on pensions and allowances are not easy to quantify, not least because of the uncertainty concerning the establishment of the minima. The actuarial balances between potential costs and likely resources are hazardous to determine in a situation in which the indexation procedures are still to be finalized, and possible changes in the base could occur as a result of the overall reform program. The danger to be avoided is to construct an elaborate and expensive social security edifice that provides benefits to a privileged subset of the target population, in the presence of poverty and need. If the state can actually ensure minimum standards of living for the entire population, it will have achieved a major objective of any social security system. To this purpose, it will be necessary to evaluate the needs of various groups of the population, including those such as the unemployed that become vulnerable as the system changes.

Appendix IV.6-1 Experience With Incomes Policies In Other Selected Countries


Incomes policies were very prominent in many OECD countries in the late 1960s and early 1970s. They were viewed as a means to contain wage pressures in an environment of creeping inflation, stop-go cycles in demand management and a rising tide of worker militancy. It was also a period of full employment and a rapid spread of generous welfare provisions. With the onset of the world recession following the first oil price shock, the concern of governments shifted from inflation to unemployment. Incomes policies were now predominantly viewed as a possible means to reduce high wage costs, which were regarded as discouraging job creation. During the 1980s, the policy emphasis shifted to structural reforms of labor markets, their deregulation and improved flexibility. These micro-economic reforms came to be viewed as a more effective approach to ensuring that labor markets, including the wage formation process, contributed to non-inflationary and job-creating economic growth.

The most extreme and interventionist form of an incomes policy is a statutory one. In this case the government or the legislator fixes a certain wage target, such as a total wage freeze, over a specified period of time. Wage settlements above this target are unlawful. It is generally felt that statutory incomes policies are not compatible with industrialized market economies because they run counter to the principle of free collective bargaining between organized labor and management. Only under exceptional circumstances, such as war conditions or abrupt changes in the economic environment, can statutory incomes policies on a temporary basis be justified in free societies. They were nevertheless tried occasionally, with the result that typically stabilization gains achieved during the freeze were reversed by efforts to catch up once the period of restraint had ended—unless there had been a marked weakening of labor market pressures during the freeze.

“Voluntary” incomes policy has been tested more widely as governments have sought ways to influence collective bargaining outcomes indirectly and with the consent of the parties involved. One such option was simply to announce “wage guidelines” and to use “moral suasion” to convince negotiators to stick to them. However, workers and their representatives who accepted the guidelines became quickly frustrated when others behaved less responsibly or when they discovered that non-wage incomes, such as profits or transfer incomes, had increased more rapidly than wages.

One important lesson learned from early experiences with wage guidelines was that it is essential to ensure that the burden of a stabilization program is shared in a just and equitable way between the various groups of income earners. However, “equality of sacrifice” means different things to different people. Should low and high wage earners forego the same amount of income? If so, should it be measured in absolute terms or in percentage terms? The general view—strongly pushed by the trade union movement—was that low-wage earners should be less severely affected by wage restraint. Most stabilization policies therefore contained a significant redistributional element.

Another significant development was the attempt to link wage outcomes explicitly with fiscal measures. One variant was a wage/tax bargain where a decrease in tax rates was partially substituted for an increase in wages. Another similar version was a wage/expenditure bargain which substituted an increase in “social income” arising from government expenditure for wage income. Also widely discussed, but never implemented in a single OECD country, were so-called tax-based incomes policies (TIP). The general idea behind TIP was to impose tax penalties on employers who exceeded wage guidelines and, in some versions, to grant tax rewards or rebates for those who undercut them. The proponents of TIP suggested that if the policy were designed to relate not to individual but to total wage bills, employers could still have scope to vary differentials within a prescribed increase in overall wages. This would permit some wage flexibility which is usually not possible under other forms of incomes policy.

A crucial question became how to implement a consensus-based incomes policy. Since the idea was to preserve free collective bargaining, a subtle policy machinery was needed. Since industrial relations systems vary among OECD countries, it is not possible to point to a standard approach. The common feature, however, was the need to base the policy, not on enforcement, but on consensus between governments and the social partners. Typical approaches attempted in various OECD countries consisted of social contracts, concerted action, tripartite bargaining, and similar devices. In this context it is often difficult to draw a borderline between “corporatist” arrangements and institutional links between governments and the social partners, on the one hand, and the pursuit of an incomes policy on the other. A typical example of a “corporatist” approach is Austria, which will be discussed below.

These consensus-based attempts worked somewhat better than statutory incomes policies and voluntary wage guidelines. But in general, and judging retrospectively, it can be said that they mostly worked in countries which had a good industrial relations climate, centralized bargaining structures and strong union and business leadership. A general problem which all forms of incomes policies seem to encounter is that, with the passage of time, powerful forces start to work against them. These include conflicts and competition between trade unions, and within trade unions between leadership and membership; efforts of some groups to re-establish wage differentials; and decentralized bargaining structures resulting in leads and lags of wage settlements and subsequent attempts to “catch up” on previous settlements. Most importantly, it is clear that any effective incomes policy cannot impose outcomes unduly at variance with market conditions and therefore has to be pursued in conjunction with appropriate demand management policies.


A tax-based incomes policy has been pursued in Poland since the early 1980s, and it formed one of the cornerstones of the stabilization program put in place at the beginning of 1990. Attempts to limit “excessive” wage increases in the 1980s, through making enterprises liable to a progressive tax on wage growth in excess of government established norms, were not successful, mainly because of widespread exemptions from the tax. Indeed, this experience clearly shows the difficulties in implementing a tax-based wage policy in an environment in which enterprises are lacking a hard budget constraint. The incomes policy pursued after January 1, 1990 generally avoided exemption for specific sectors or enterprises. Increases in average wages or the wage bill beyond a certain norm were subject to progressive taxes, the top marginal tax being 500 percent on wage increases in excess of 3 percent over the norm. The norm was determined every month as a percentage of the increase in consumer prices, and unused permissible wage increases could be carried forward. The norms applied in 1990 implied a sharp fall in real wages, as well as a marked slowdown of the rate of increase in nominal wages. Developments in the first half of 1990 were consistent with these objectives, but this does not necessarily reflect the effectiveness of the incomes policy measures: an abrupt easing of the labor market with the unemployment rate rising from virtually zero to over 8 percent should have generated disinflation on its own. Indeed, in the first half of 1990 wage increases were below the norms established by the Government, suggesting that demand deflation had rendered the incomes policy measures superfluous. However, in spite of the unused scope for wage increases gained in the first half-year, wage increases exceeded the norm in the second half-year.

During the early 1980s, the growth of enterprise wage funds in Hungary was linked to indicators of economic performance. Apart from encouraging productivity gains, this policy was aimed at containing inflationary wage pressures. The economic performance of firms was assessed relative to output targets, that of workers in terms of time standards. Experience with these wage regulations was rather disappointing, partly because of flaws in the indicators and partly due to firms using bonus funds to evade restraints on the growth of their wage bills. In conjunction with the abolition of direct controls on wage funds in the mid-1980s, the authorities introduced a tax-based incomes policy to contain the rise in nominal wages. Taxes were levied on profits if the growth of the wage bill exceeded a productivity related ceiling. In 1989, wages were further liberalized, but taxes on “excessive” wage increases were strengthened. Under the new regime, the difference between the rate of growth of wages and value-added became subject to a 50 percent profits tax (adjusted in 1990 to 35-40 percent, depending on the size of enterprises). So far, the policy has not been very successful as enterprises have resorted increasingly to non-cash forms of remuneration of employees.

Government intervention in wage determination in Yugoslavia was frequent in the 1980s, but it failed to moderate wage trends or inflation. Successive stabilization packages over the decade generally included a mixture of price and wage controls, with wage controls assuming a greater role over time. Common to all the wage regulation measures was that an abolition of controls was followed by a step increase in wages to restore real incomes, and such increases were accommodated by the banking system. The 1990 stabilization program differed from its predecessors in that most prices were exempt from controls, while no nominal wage increases were allowed for the first six months of the year unless the Government departed from its commitment to keep the dinar/deutsche mark exchange rate fixed. Real wages fell sharply in the first half of the year, but the abolition of wage controls was accompanied by a step increase in wages which, however, was less than on similar occasions in the past. This can be attributed mainly to policy-induced demand compression in the economy which worsened the labor market situation and moderated price increases, as well as a continued commitment of the authorities to maintain a fixed exchange rate between the dinar and the deutsche mark. The relative failure of the incomes policy was due to the fact that a new framework to set wage increases had not been established. A key condition for the success of incomes policy would seem to be the reform of the self-management system.

Incomes policy has traditionally played an important role in economic management in Norway. The authorities have sought to facilitate discussion among the social partners by creating permanent institutions, where the social partners could come to an agreed view on the state of the economy and its likely evolution. This institutional setting would seem to have contributed both to keeping the equilibrium rate of unemployment low as compared with other countries, and to enhancing the responsiveness of wages to changes in unemployment. In the 1970s, the authorities repeatedly offered a broad range of measures, such as increased public spending and tax benefits, in exchange for moderate wage settlements. This form of tripartite bargaining was abandoned as it yielded disappointing results, generally adding fiscal stimulus to an already overheated economy. Faced with acute imbalances in the economy and the need to prevent a wage-wage spiral, the authorities put statutory limitations on wage increases in 1988 and 1989, effectively imposing on all labor market participants the moderate wage agreements between the main social partners. The period of statutory wage limitations was accompanied by a sharp easing of wage increases. However, as unemployment was growing very fast at the same time, it is uncertain what independent role incomes policy played in the disinflation. It is likely that the policy brought about the disinflation more rapidly than otherwise.

Incomes policy is widely regarded as having played a crucial role in Spain in its transition to a modern market economy. Against the background of a disorderly labor market in the mid-1970s, in which real wages and unemployment rose in tandem, the authorities issued wage guidelines to stabilize the purchasing power of wages in the context of slower wage increases. The guidelines were only indicative for the private sector, but were backed-up by threats of more restrictive policies if they were not adhered to. This intervention paved the way for a national agreement between the social partners and facilitated the development of a modern industrial relations system. In conjunction with measures to deregulate the labor market, the Government continued to issue wage guidelines in the 1980s to encourage wage disinflation. It also entered into direct negotiations with the social partners, offering commitments to substantial job creation and comprehensive changes in labor and social policy in exchange for wage moderation. The incomes policy in the first half of the 1980s is judged to have assisted the deceleration of price increases during this period. However, wage guidelines have not succeeded in moderating wage increases more recently. This is partly because the Government has failed in its aim to create a permanent institutional framework for the pursuit of incomes policy, but it also reflects the difficulty of moderating wage increases in a booming economy.

The Austrian system of “social partnership” grew out of a long history of strong unionism and in response to controversial relationships between trade unions, employers and the Government during the inter-War period. Since the early 1950s, trade unions and employers’ organizations have been involved in virtually all aspects of economic policy. The Government consults regularly the leaders of these organizations, many decisions on economic policies are made in tripartite bodies, and the social partners are represented and play a leading role on the board of the central bank. There are regular gatherings on economic research and forecasting issues with the view to developing a common perspective on the state of the Austrian economy and its likely evolution. Collective bargaining is carried out by individual trade unions, not by the Confederation of Austrian Unions. However, even without issuing official wage guidelines, the Confederation has a considerable influence through discussions with negotiators which take place prior to the start of the negotiations. Formally, negotiations cannot take place without the consent of a Social Partnership Commission (“Paritätische Kommission”). This strengthens the position of the Confederation vis-à-vis the individual unions. Since the late 1970s, Austria has followed a hard currency policy by pegging the schilling to the deutsche mark. This has kept the rate of inflation down due to the imposition of a severe budget constraint on export and import competing firms. This policy is strongly supported by the union movement. To satisfy the external constraint imposed by this policy without impinging on domestic employment, wage moderation is required—an argument widely used by the union leadership and accepted by the membership. This internalization by labor of the discipline involved in the maintenance of a hard currency has been a key factor in shaping the wage setting environment in Austria during the 1980s.

The Italian Government regularly intervened in wage determination during the 1980s with a view to dampening wage increases. Indicative wage guidelines in 1981, which were accepted by the social partners, were accompanied by a deceleration in wage growth. These were followed by a tripartite agreement in 1983, with the Government promising, inter alia, changes in tax policy in exchange for wage restraints, both in the form of moderate increases in negotiated wages and changes in the wage indexation mechanism. With continuing erosion of international competitiveness, the authorities requested further limitations in the coverage of indexation and imposed this by decree when the main union federations failed to agree among themselves on the issue. The decree was later supported in a national referendum in 1985, and the automaticity of the wage indexation mechanism was further weakened subsequently. While not changing the indexation process fundamentally, these modifications are generally considered to have assisted disinflation. However, the main factors behind the disinflation in Italy have been the stance of macroeconomic policy and, in particular, the commitment to a stable exchange rate within the EMS.

Australia has a comprehensive and complex system of minimum wages differentiated by occupation and industry, in some ways similar to the Soviet system of tariff wages. Since 1983 an accord between the (sole) central union federation and the Government has drawn on the centralized arbitration procedures which govern this system to implement an anti-inflationary incomes policy. The access of individual unions to centrally determined wage increases has been made dependent on their commitment to making “no extra claims” at a sectoral level. The central agreements were initially based on indexation for price increases, but since 1985 have increasingly used reductions in direct taxation as a complementary means of compensation for increases in prices. These policies have resulted in reductions in real wage costs of the order of one percent per annum, and this has contributed to Australia having enjoyed one of the highest rates of employment growth in the OECD area. However, the contraction in real wages has taken place in an environment of high nominal wage growth, spurred, inter alia, by exchange rate movements and buoyant aggregate demand. Recently, the central decisions have sought to improve labor market functioning by requiring the amalgamation of previously segregated skill classifications, the removal of age limits on training access, and the elimination of inhibitions on labor market flexibility, such as constraints on part-time employment.


The labor force in Western statistics is conventionally defined as the sum of employment plus unemployment. However, since the concept of unemployment does not exist yet in official Soviet labor force statistics, Soviet participation rates are essentially equivalent to employment/population ratios. The international comparisons in Chart 2 are based on a “standardized” labor force participation rate which relates the labor force to the size of the age group 15-64, not to the number of adults below retirement age. This accounts for the lower standardized rate for the USSR in Chart 2 than is usually recorded, since the Soviet retirement age is relatively low.


According to a recent Goskomstat survey, the main reason women cite for working full-time is that they need the income to support their families. This suggests that continued high participation rates among women may be related to relatively low wages of the main breadwinner. Given the choice, the majority of women stated their preference for part-time work. But many respondents also added that they could not find part-time jobs because their employers were unwilling to create such positions.


It is worth noting that the current Soviet share of around 20 percent of total employment in agriculture is similar to the OECD average in 1950; the average share of agriculture in the OECD countries has since declined to around 7 percent. See Maddison (1989).


The average employment in Hungarian industrial enterprises is 2,500 (Hare (1990)).


OECD (1989), p. 190.


See Oxenstierna (1990), pp. 31-34, for a review of this issue.


Goskomstat estimated that, according to ILO definitions, there were 2 million unemployed persons in 1990. Goskomtrud had cited a figure of 2.5 million for the end of 1990. Non-official Soviet sources have quoted estimates ranging from 4 to 6 million.


See Oxenstierna (1990), pp. 222-28, for a review of this literature.


The estimates supplied by the Soviet authorities for the scale of such employment cut-backs are, however, quite small. Available data show that the central administration employs around 2 million persons, or less than 1.5 percent of total employment. Compared with Western countries this proportion is very small, suggesting that not all employment in branch ministries is allocated to the central administration in the statistics.


A recent review of cross-country econometric estimates of the unemployment elasticity of real wages concluded that it is small, of the order of -0.1. See Blanchflower and Oswald (1990).


Assuming that inflows into unemployment revert back to their late-1980s value.


See, however, the discussion in Ofer and Vinokur (1990), based on a sample of Soviet migrants to Israel in the 1970s.


If the collective farm is unable to pay the minimum wages, the state lends it money for this purpose. Such loans have rarely been repaid. See Oxenstierna (1990), p. 126.


An effective minimum wage was reintroduced only in 1957, the first since 1937. This was initially rub 27-35 a month, then rub 40. It has been held at rub 70 a month since 1977.


See Oxenstierna (1990), pp. 144-145.


For a discussion of the issues, see Chapman (1990), pp. 16-21.


For evidence in this regard, based on surveys of Soviet emigrants to Israel, see Ofer and Vinokur (1990).


See Chapter III.1 for a detailed discussion of the budgetary implications of food subsidies.


Commonly used inequality indices are sensitive to transfers in particular segments of the income distribution. The coefficient of variation, for instance, is particularly sensitive to changes amongst higher income groups, reflecting inequality due to the presence of rich individuals. The Gini coefficient is more sensitive to changes in the middle ranges of the size distribution. The Atkinson index permits explicit weighting of a social welfare function, ranging from equal treatment of transfers to all households (zero inequality aversion), cases where relatively greater weight is placed on transfers to the poor than to others (inequality aversion parameters in the range 1-2), toward Rawlsian preferences when only the welfare of the poorest is considered (inequality aversion parameters of 5 or more). Thus, the Atkinson index with an inequality aversion parameter of 2 will represent inequality due to large numbers of the poor relative to the mean, whereas that implied by an inequality aversion parameter of 5 will concentrate on the poorest households relative to the mean.


To some extent, the per capita size distributions overstate the extent of poverty in the Central Asian republics, relative to the Baltic republics for example, as children and adults are treated the same. However, this procedure is quite commonly adopted. See, for example, World Bank (1990). Equivalence scales and household observations were not available.


This probably reflects limitations of access to supplies for the poorer income groups rather than taste factors, since the upper income households tend to have more similar consumption patterns across republics. The data are based on household consumption surveys conducted by Goskomstat.


See Rumer (1989), p. 137.


Goskomstat, Social Statistics (1990), p. 40.


Carey (1989), pp. 1-38.


Old age pensions doubled on average and for the lowest paid they increased three-fold. The minimum pension was set at 50 percent above the previous maximum, and the minimum disability pension was raised by 600 percent. See Madison (1990), p. 166.


For a more detailed description of the various social security benefits and their present and prospective financing, see Appendix III. 1-3.


See, for example, Atkinson (1990).


For a discussion of the role of employment services in raising labor market efficiency, see OECD (1990b).


See also the discussion in Chapter III.1.


See OECD (1990a), pp. 153-171, for a full discussion of this issue.


See Yanowitch (1984-85).


See Yanowitch (1981) and Marnie (1986).


Kostakov (1990) estimates that a Soviet worker spends 15 percent of his time in training, although the estimate of this may include pre-employment education and training of students of working age.


For further information about the functioning of trade unions in the centrally planned Soviet economy see Lowit (1971) and Ruble (1981).


These figures were contained in the report of the Control Commission to the inaugural Congress of the successor body to the AUCCTU, the General Confederation of Trade Unions (GCTU), which was held in Moscow on October 23-25, 1990.


For example, in April 1990, the ICFTU reported that since the (then) AUCCTU was determined to keep its major assets such as recreation centers and health centers, “Many workers are therefore reluctant to resign from the established unions and at least some of the independent groups accept that their members or supporters will stay in these unions at least for the time being.”


The issue of the disposition of the property of the AUCCTU generated considerable discussion at the Congress held in October 1990. The final decision was that the assets should not be dismantled, which ensured that it would remain under the central control of the successor body to the AUCCTU, the GCTU.


The Austrian Chamber of Labor provides one example of such an arrangement.


The National Council for Coordination of Interests, established in Hungary in 1989—and in which all union groups are represented—is one example of such a nonbinding coordinating body. The Hungarian Government now plans to establish such councils at the regional level.


For a particularly stringent critique of these distortions in the wage system, see the paper delivered to the Joint ILO/USSR Conference, October 1-4, 1990, in Moscow, by the Chairman of Goskomtrud (Shcherbakov (1990)).


Such adjustments are frequently achieved through so-called thirteenth or fourteenth month bonuses which are related to overall enterprise performance.


For a discussion of these issues in health and safety regulation, see OECD (1989).


Also see Appendix III. 1-3 for a detailed discussion of social security reform and its budgetary implications.


The authorities confirmed that medical insurance and health care spending were not covered under the proposed social security funds. There were proposals to form a special health insurance fund, which would be independent of the state budget. This fund would be financed by a special insurance premium, and if it were to cover current health care spending of around rub 19 billion, the rate would have to be set around 2-3 percent of income. Within the present framework of state health guarantees to all, it might prove difficult to switch over to an insurance-based system.


See Appendix III. 1-3 for a more detailed discussion.

    Other Resources Citing This Publication