In only a few years, unemployment has become one of the most pressing issues facing the transition economies in Eastern Europe and increasingly in the states of the former Soviet Union as well. As of early 1994, 10-15 percent of the labor force in Eastern Europe lacked work, and over 40 percent of the unemployed had been jobless for more than a year (Table 1). The significance of this development cannot be overstated. Under central planning, workers were virtually guaranteed full employment. This was achieved by placing workers on the payrolls of state enterprises, which were, as a result, chronically overstaffed and unproductive. With the collapse of central planning, however, the objective of maintaining full employment was abandoned, and, for the first time in decades, large numbers of people found themselves out of work.
While unemployment is a necessary result of the resource reallocation that must occur as state-owned enterprises are privatized and restructured, the high levels of unemployment that are now observed have increasingly emerged as an obstacle to restructuring. Indeed, the economic and political costs of high unemployment have affected the overall reform process, slowing the restructuring of the state sector and discrediting reform-minded governments. Thus, the rate of growth of unemployment and, in particular, ensuring that the long-term unemployed are brought back into the active labor force has become a major concern for economies in transition.
Olivier Blanchardfrom France, is Professor of Economics at the Massachusetts Institute of Technology. He holds degrees from the University of Paris and MIT.
Simon Commanderfrom the United Kingdom, is Principal Economist in the Bank’s Economic Development Institute. He holds degrees from Oxford and Cambridge Universities.
Fabrizio Coricellifrom Italy, is Associate Professor in the Department of Political Economy at the University of Siena, Italy. He holds degrees from the Universities of Modena and Pennsylvania.
Patterns of joblessness
Unemployment has primarily been driven by reductions in employment in the state sector. Unemployment would have been even higher if it were not for large flows out of the labor force, which partly reflects the reversal of the artificially high labor force participation rate—particularly for women—that was a characteristic of the planned economy. One way of thinking about this evolution is that, in theory, as people lost their jobs in the state sector and unemployment rose, wages would fall, and with this decline, the incentives for the private sector to create jobs would grow. This has indeed happened, but with several important qualifications.
The flow into unemployment has been low relative to OECD comparators, however, the flow out of unemployment has been lower still. Once workers become unemployed, they have great difficulty finding a job. Thus, over time, unemployment has become like a stagnant pool with new entrants trickling in but with few exiting. In Poland or Hungary today, less than 3 percent of the unemployed find jobs each month (the monthly exit rate); by comparison, the rate in the United States in 1992 was nearly 25 percent.
Unemployment: the long wait
Why has unemployment remained high? A number of explanations have been put forward.
• Workers commonly search for and find other jobs while still employed;
• Employers may prefer “job-to-job” transitions believing that they can select better quality workers among those already employed rather than among the unemployed, about whom they have imprecise information;
• Workers who remain unemployed for long periods tend to become discouraged and give up looking for work;
• Duration and levels of benefits are too high, deterring the unemployed from looking for work. This explanation may not be valid, however, because eligibility has been universally tightened over time—for example, the Czech and Slovak Republics now provide benefits for only six months—while the ratio of benefits to the worker’s last wage (the replacement rate) has also been reduced. The fact that benefits are not generous suggests that, on average, they are unlikely to have been a powerful factor in restraining job search by the unemployed; and
• There appears to be a major imbalance in the distribution of the unemployed and vacancies across regions. This geographic mismatch—exacerbated by constraints on mobility, including a poorly functioning housing market—has led to severe inefficiencies in matching workers with jobs.
This last point is especially significant. It partly reflects the effect of earlier decisions on the structure and location of industry. Regional concentration in output and employment and the associated phenomenon of the one-company or industry town—particularly important in Russia—has been a significant factor in these regional imbalances. Faced with new relative prices and competition, these industries and towns have declined, and the shocks to local employment have been profound. For example, this has been the case for military producers in the Slovak Republic, and heavy industry in northeast Hungary. Similarly, rural areas throughout Eastern Europe have continued to experience high unemployment. The growth of jobs in the private sector has not compensated for these shocks to any great extent so far.
Reducing unemployment requires creating jobs in the private sector, particularly as one can expect further employment losses from privatized firms as they restructure, as well as from the remaining state-owned firms. Understanding the behavior of enterprises, especially with regard to wage and employment decisions, is thus essential for determining the likely future rise of employment in these economies.
The way unemployment has evolved is primarily the result of the asymmetric behavior of state and private firms in their employment decisions. So far, job losses in the state sector have been much greater than the number of jobs created by the private sector. Moreover, the level of unemployment itself has had an important impact on the behavior of firms.
State firms. When the transition to a market-based economic system began, state firms moved slowly to reduce their work forces, relying primarily on attrition, early retirement, and other voluntary measures. Only where the shocks to the economy were sufficiently large, as in Bulgaria, did involuntary separations figure significantly early in the transition process. A second phase, in which involuntary separations increased, was largely attributable to reductions in credit to state firms, announcements of changes in ownership, and the growing autonomy of managers. Surprisingly, the evidence from Eastern Europe suggests that firms in general have operated ex ante as if facing a hard budget constraint. Even when subsidies and other forms of support have been maintained, as in Romania and Russia, there have still been clear declines in employment.
Nevertheless, avoiding or limiting layoffs has remained an important objective for both managers and workers in state firms. For instance, evidence from Romania shows that those industries receiving subsidies and other assistance adjusted employment the least. These protected industries also experienced relative wage increases. Further, even when unions or other labor organizations have exercised little direct say in firms’ decisions, their implicit voice may not have been small.
Evidence from a number of countries—Bulgaria, Poland, Romania, and the Slovak Republic—suggests that the process of employment reduction is by no means linear. Phases of increased layoffs have been followed by episodes of “labor hoarding.” This appears to be explained primarily by political factors, as governments—sensitive to high national or regional unemployment rates—have selectively sanctioned labor hoarding by relaxing financial constraints. This has often been done not by explicitly increasing subsidies to firms—these have declined very significantly almost everywhere—but by facilitating credits from the banking system to heavily indebted firms. In Hungary, for example, this practice has subsequently required large, repeated government bail-outs of the financial system. In addition, tax relief, toleration of nonpayment of social security contributions, and other mechanisms have been used to support state firms.
Why unemployment rates differ
Aggregate unemployment rates mask significant changes over time and across regions. Indeed, unemployment rates vary widely among economies in transition and are directly related to how much employment in the state sector has fallen and to job creation in the private sector. Constraints on labor mobility accentuate the difference among countries.
Table 2 puts employment and unemployment changes together, and shows the ratio of the change in unemployment to the absolute value of the change in employment. Under the benchmark case of no change in working age population, and all workers going from employment to unemployment, the ratio is equal to 1. Values much below 1 indicate that much of the employment decline has translated into nonparticipation rather than unemployment.
Consider first the three major Central European countries, Poland, Hungary, and the Czech Republic. The table suggests two proximate reasons for the lower unemployment rate in the Czech Republic. The first is a larger offset of the decline in state employment by private sector growth. The second is a much lower ratio of the change in unemployment to the change in employment. Tough unemployment eligibility rules clearly have led many more workers in the Czech Republic to become nonparticipants rather than unemployed. And the labor force participation rate at the start of transition was the highest in Eastern Europe. In Poland by contrast, the ratio is greater than 1, indicating that, despite the importance of early retirement soon after the transition, much of the decrease in employment since then has translated into unemployment rather than non-participation. Were the Czech Republic to have the same ratio as Poland, the unemployment rate would be close to 8 percent. Were the Czech Republic to have the same ratio as the Slovak Republic, the rate would be close to 7 percent, compared to 11.6 percent for the Slovak Republic. Thus, unemployment is lower in the Czech Republic because of both stronger private employment growth and a larger increase in nonparticipation. In Bulgaria, the primary cause of high unemployment is the combination of large losses in state employment and a struggling private sector. A large part of the employment decline has translated into nonparticipation. Were Bulgaria to have a ratio of 1 rather than 0.5, its unemployment rate would be close to 30 percent.
A third pattern, seen in Romania and Russia, was continued low unemployment at the start of the transition, primarily because of the smaller decline in state employment and smaller increase in private employment. In Russia, the ratio of unemployment to employment changes is again very low; were it equal to 1, the unemployment rate would stand around 4 percent. The subsequent increase in Romania in 1992 and Russia in 1994 reflects some tightening of the budget constraint, although policy changes have been frequent and unpredictable.
|Change in total employment||Change in state employment||Change in private employment||Change in unemployment||Ratio change in unemployment/ employment|
Although the process has been uneven, state firms have shed labor, and often on a major scale. Many of those workers who lost their jobs either left the labor force entirely or found new jobs directly. The rest became unemployed. But unemployment (or its political fall-out), in turn, has affected the rate of jobless in the state sector, so that as unemployment rose, labor shedding by state-owned firms tended to decrease.
It is still too soon to assess the implications for unemployment of privatizing major parts of the state-owned enterprise sector. Even in the Czech Republic, where the first phase of voucher privatization has been completed, a very small number of bankruptcies have resulted, and restructuring has remained small. However, given the extensive restructuring that would likely be required to make many formerly state-owned firms viable, privatization, whether from above, as in the Czech case, or more commonly, from below, respecting the rights of workers, is likely to entail further employment losses. This can arise from active restructuring decisions or equally from the response of firms to a difficult financing environment and inability to attract outside investors—a particular problem when firms have been privatized by insiders. Indeed, the cost of restructuring the privatized firms remains a major obstacle in almost all countries.
Private firms. The private sector has generally grown significantly, and not simply as a result of privatization. The relative rate of expansion of the private sector has been closely related to the rate at which the state sector has declined. Where there has been policy ambiguity or where maintaining the state sector has been given priority—as in Romania—growth in private firms has clearly been held back. Private sector growth has also been held back in Bulgaria, but for rather different reasons. In Bulgaria, both output and employment in the state sector declined drastically because of the size of the initial output shock, which was magnified by a large fiscal correction and associated deflationary policies. These effects hit the whole economy, limiting the growth of private firms. In other countries, such as the Czech Republic, Hungary, and Poland, private sector growth in output and employment has been substantial. But in all countries, private sector growth thus far has largely come from filling the gaps, particularly in services, that were inherited from the previous system. This can be viewed as a once-and-for-all adjustment and, as such, it raises the difficult question of whether much more growth from this part of the private sector can be expected.
The path of wages
The employment level that these economies ultimately attain will be sensitive to the evolution of wages. At the start of the transition many feared that, freed from earlier constraints and given uncertainties over ownership, those who still had jobs would behave short-sightedly, and push for aggressive wage settlements. This has not happened to any great extent. But despite large declines following price liberalization, over time, real wages have been recovering. Unit labor costs, which measure wage costs per unit of output, show a clear increase over pre-transition levels. Part of this may reflect declining output and continued overstaffing in state firms. However, as profits in the state sector have generally evaporated, the evolution of unit labor costs may also reflect the ability of insiders—those still employed—to set wages higher than would be warranted in a competitive setting. Even so, it is hard to attribute much of the employment losses to date to selfish and expansive wage behavior by those still working. However, it is clearly important that wage restraint remain a characteristic of transition, for without it, employment would be lower.
Wage setting in state and privatized firms is also very important because of the interaction between wages in such firms and the wages offered by the private sector. For instance, if wages remain higher in the former, this differential wage will likely feed back to private sector wages and, ultimately (negatively), to employment. Until now, the available data indicate that private sector wages are generally lower than in state firms. However, the private sector appears to have several constituent parts; a larger low-skill, low-wage end (primarily in services), as well as a much smaller, high-wage component. The fact that wages in the private sector are not higher, on average, provides little incentive for workers to quit the state sector.
Quite striking is the evidence reported for almost all of the countries surveyed; workers, including those unemployed, generally express a preference for state employment. In the case of the unemployed, it may be a result of their attaching a low probability to finding a private sector job, but more generally it reflects the perception that job conditions, including wages and benefits, are superior in the state sector. And, of course, there is the legacy of values and expectations from the planned economy. These factors have generally combined to ensure that the decline of the state sector has been more gradual than initially expected and that insiders have also managed to preserve their privileges.
The recent experience of unemployment in Eastern Europe and the states of the former Soviet Union has been diverse and prospects are mixed. In some cases—Russia and Romania are the most obvious examples—the break with the soft budget constraint has yet to be achieved and macroeconomic instability has resulted. In these countries, unemployment has only begun to emerge. And in almost all cases, unemployment to date has mainly been driven by macroeconomic shocks, suggesting that the restructuring necessary to transform the economy to a market-based system will bring further job losses.
In countries where reforms are more advanced, there is still a common fear that the private sector may not be able to create employment on a scale required for robust economic growth. In addition, unemployment may stay high as regional imbalances become entrenched and many of the unemployed become too discouraged to continue seeking work. In such cases, the political viability of reform may be compromised, as well as the fiscal stance of the government because of the cost of providing additional benefits to the unemployed.
“In countries where reforms are more advanced, there is still a common fear that the private sector may not be able to create employment on a scale required for robust economic growth.”
While such a scenario seems unlikely, at least in the Czech Republic, Hungary, and Poland, much depends on whether unemployment effectively triggers lower wage claims as workers seek to preserve their jobs in the state sector. If this occurs—and here the evidence is still fragmentary—downward pressure on wages should promote job creation in the private sector. Further, with privatization and restructuring, deteriorating employment options will likely result in slower restructuring in state or privatizing firms, rather than an increase in unemployment. This can, however, cut two ways. In Bulgaria, for example, the sharp initial jump in unemployment was not accompanied by any coherent program of privatization. Restructuring of state firms likewise dwindled, but financial supports were made available to arrest further growth in open unemployment. In turn, this acted as a brake on the growth of the private sector as well as provoking a deterioration in government finances. And a more general danger is that in countries where the state sector remains large, wage and employment decisions will be taken consistent with the old “guaranteed employment” mentality that prevailed under central planning. These effects will, of course, be magnified if many of the unemployed have been out of work for a long time and may no longer be seriously searching for jobs. As in OECD countries, such so-called duration effects can weaken the impact of unemployment on wages, with possibly severe long-run consequences for the overall level of employment.
In this context, policymakers need to face some hard realities:
• First, it is important to recognize that high unemployment has occurred alongside limited restructuring and reallocation of resources. There is more unemployment to come.
• Second, unemployment varies widely among regions in all countries. This is partly because of barriers to mobility, such as housing. But it also suggests a possible role for infrastructural projects that can increase mobility. Stimulating a housing market and improving infrastructure will take time, however, and the benefits are unlikely to be seen quickly. In addition, mechanisms and financing for the transfer of social benefits provided by firms to local governments—a major feature in Russia—need to be put in place. Without these reforms, firms will hesitate to release labor, and workers will lack appropriate access to services outside of firms.
• Third, by end-1994 over half the unemployed will be long-term unemployed with few chances of recovering a job. It is important to reduce the share of the long-term unemployed. Only in the Czech Republic have significant steps been taken to adopt active labor market programs, including training and wage subsidies for hiring the unemployed. Elsewhere, fiscal and institutional constraints on implementing active labor market policies have prevailed. Shifting the balance toward a larger active component will be desirable and schemes—such as marginal wage subsidies—that encourage employers to hire from the pool of unemployed can help in breaking the cycle of long-run joblessness, even if they impose additional short-run costs.
• Finally, policies that respond to unemployment by simply postponing restructuring and going slow on ownership reform in the state sector may yield short-run political benefits. But these are also likely to have other less desirable effects that may paradoxically result in the longer term in a lower level of employment.
A longer version of the paper by the authors can be found in Unemployment, Restructuring and the Labor Market in Eastern Europe and Russia, World Bank, 1994.