Republic of Poland: Selected Issues

This Selected Issues paper examines the driving factors for the broad movements of investment over the past decade in Poland. After booming for half a decade, Poland’s investment plummeted during 2001–03 and has recovered only marginally since then. The paper provides a historical perspective on the evolution of economywide and sectoral investment in Poland, and briefly summarizes possible determinants of investment. It analyzes the determinants of investment more systematically using panel regressions based on sectoral data, and reports the results of in-sample and out-of-sample simulations.

Abstract

This Selected Issues paper examines the driving factors for the broad movements of investment over the past decade in Poland. After booming for half a decade, Poland’s investment plummeted during 2001–03 and has recovered only marginally since then. The paper provides a historical perspective on the evolution of economywide and sectoral investment in Poland, and briefly summarizes possible determinants of investment. It analyzes the determinants of investment more systematically using panel regressions based on sectoral data, and reports the results of in-sample and out-of-sample simulations.

II. The Labor Market in Poland8

A. Introduction

19. High unemployment is one of the dominant problems of the Polish economy. Having declined from 14 percent in 1993, after the early stages of transition, to almost 10 percent in late 1997-early 1998, the unemployment rate has since soared. It peaked at 20½ percent in 2002, the highest level in the region (Figure 1 and Table 1). Restructuring in the wake of the Russia crisis, together with privatization and the EU slowdown, contributed to this rise. However, the strong growth rebound in Poland during 2003–04 was associated with only a modest fall in the unemployment rate.

Figure 1.
Figure 1.

Labor Market Trends in Selected Countries, 1993-2004

Citation: IMF Staff Country Reports 2005, 264; 10.5089/9781451831986.002.A002

Sources: Eurostat data; OECD data; GUS, Labor Force Survey data; and IMF staff calculations.
Table 1.

Principal Labor Market Developments by Age and Sex Groups, 1993–2004

(In percent, unless otherwise stated) 1/

article image
Sources: GUS, Labor Force Survey bulletins; and IMF staff calculations.

For males and females, the data are for persons aged 15 and above.

Notes: * Unlike for the remaining age and sex categories, this peaked in 1993Q1.** Unlike for the remaining age and sex categories, this peaked in 2003Q4.

20. Related adverse developments in employment and labor force participation are also a serious concern. Despite continuous growth in the working-age population since the early 1990s, employment has dramatically fallen and, at 52 percent of the working-age population, remains the lowest in the region. At the same time, labor force participation has steadily declined (Figure 1). These adverse trends stand out in comparison with other countries in the region. Employment and labor force participation rates did decline in the Czech Republic but by smaller magnitudes, and both remain at significantly higher levels than in Poland. Although unemployment remains almost as high in the Slovak Republic as it is in Poland, labor force participation is stronger and gradually improving, and the employment rate has slightly risen in recent years.

21. Are these trends in Poland entirely due to the persistence of transition and restructuring shocks? Or are there ingrained structural problems and market distortions that will prevent labor market clearance at a satisfactory employment level even after the effects of these shocks have waned? This paper explores these hypotheses by reviewing the main features of labor supply in Poland. Section B outlines the main problems of Poland’s labor market, largely drawing on existing studies and on the 2005 Article IV consultation discussions. Section C uses labor force survey data to examine in greater depth supply-side factors that hinder labor market improvements. Section D presents forward-looking scenarios that show the difficulty of achieving a major recovery of employment, given current structural problems. Section E concludes.

B. A Synopsis of Poland’s Labor Market Problems

22. Poland’s high unemployment rate is at least in part a legacy of economic transition and restructuring. Initially, restructuring and the closing down of old industries led to massive job destruction during 1990–93 and resulted in double-digit unemployment rates. While there was also job creation, improvements in efficiency and continued labor shedding in the mid 1990s, help explain the lack of net employment growth (World Bank, 2004). Large scale job losses recurred in the second wave of restructuring after the 1998 crisis in Russia, which seems to have mostly affected exporters and heavy industries, such as mining and steel mills. Privatization contributed to these job losses, as did the slowdown that gripped the country in 2001–02. Therefore, by 2004, nearly half of the unemployed who had previously held jobs lost them because of the closure of their workplace or redundancy.

23. The uneven impact of these factors across the country has contributed to regional labor market disparities. The process of industrial destruction at the onset of the transition has been uneven across regions, setting the stage for unbalanced regional economic growth (Daseking, Jiang and Summers, 2000). But limited labor and capital mobility and regional differences in skills and education have contributed to the persistence of the regional disparities in labor market performance (Boni, 2002; UNDP, 2004; and World Bank, 2004). Also, uniform benefits and minimum wages across regions appear to have been binding in some regions, discouraging labor supply (see below). Lack of regional labor cost differentiation to reflect differences in average labor productivity—with the lack of such differentiation itself also partly resulting from uniform minimum wages across regions—has impeded growth of labor demand.

24. Recent analyses identify a rather large number of institutional and policy factors that have magnified these labor market problems:

  • Overly easy access to early retirement and social benefits has inhibited labor supply (UNDP, 2004; and World Bank, 2001 and 2004). At the onset of the transition, generous provision of preretirement benefits was geared to contain the unemployment surge (Keane and Prasad, 2000). Generous disability pensions also helped to discourage labor supply. By allowing 30 percent of Polish households to subsist today without having any economically active member, these benefits, together with social assistance benefits, have contributed to a structural increase in unemployment. Indeed, Estevao (2003) provides evidence for such an increase by showing that the wage curve has shifted over time, likely associated to some extent with the benefits system in place.

  • The relatively high minimum wage is believed to have constrained job creation for youths, for the relatively unskilled, and in high-unemployment regions (Selassie, 2001; Estevao, 2003; and World Bank, 2004). Although the ratio of the minimum to average wage has declined, from 42 percent in 1997 to 36 percent in 2004, minimum wages remain uniform across regions and across trades; hence, the constraint seems to remain binding for the groups just mentioned.

  • Some studies have noted that lack of flexibility in sectoral wages may have contributed to large job losses in sectors that were hardest hit by the restructuring, as employees did not accept wage cuts commensurate with the drop in prices in these sectors (Pujol, 1996). While estimates of the wage-unemployment elasticity in Poland are sensitive to the methodology used, they generally suggest that this elasticity is not different from those reported for industrial countries (Estevao, 2003). However, there is also evidence that wages may not be entirely flexible in Poland despite the limited influence of unions. Yamaguchi (2005) finds that the wage curve is flat (and, hence, wages are inelastic) when unemployment is high (exceeding 14 percent), although at low levels of unemployment the wage curve becomes steeper. This finding suggests that wages in Poland have been inflexible over the past few years, hampering job creation.

  • Several studies argue that relatively high payroll taxes (about 45 percent of gross income in 2004) have discouraged employment (Boni, 2002; UNDP, 2004; World Bank, 2001 and 2004; and Estevao, 2003). Indeed, the tax wedge in Poland seems among the highest in OECD countries (Table 2), although this comparison may be biased: countries may appear to have a lower tax wedge simply because of generous cash benefits and/or tax credits. For example, changes to such benefits and credits in Hungary partly explain the improvement in the tax wedge in 2001–04 for a couple with two children and one earner. Nevertheless, payroll taxes remain relatively high in Poland, reducing both labor supply and labor demand to the extent that the incidence is on employers as well as employees.

  • Restrictive elements in the employment protection legislation are also believed to be a problem (Boni, 2002; and World Bank, 2001). Such legislation, although more flexible than that of other countries in the region, such as the Slovak Republic or the Czech Republic (Tables 34), and despite recent reforms (see the Appendix for details), appears to remain an impediment. In particular, although there seem to be no difficulties in hiring, legislation governing collective dismissals and regular employment remains very strict, in contrast to the increasing flexibility of legislation governing temporary employment. Consequently, job creation is driven by an increase in temporary jobs, the share of which in total paid employment had risen above 30 percent by end-2004, from close to 10 percent four years earlier.

  • Some studies also note that administrative barriers to entrepreneurship in Poland have hampered job growth (Boni, 2002; UNDP, 2004; and World Bank, 2001 and 2004). Indeed, it is often argued that both financial and legal barriers to the development of small and medium-sized enterprises (SMEs) in Poland are binding9, whereas Schiff and others (2005) note that a vibrant SME sector is key for job creation.

  • Gaps in the education system are also believed to impede employment growth (Boni 2002; Daseking, Jiang and Summers, 2000; and UNDP, 2004). The Polish schooling model is often referred to as not flexible enough, as it has not responded to changing market needs. Vocational education, in particular, is believed not to have adjusted to market requirements. Anecdotal evidence suggests that, in some regions, vocational schools that were set up about 15 years ago continue to provide education following the same curricula used at the time of their inception. Also, the institutional setup for retraining the unemployed, especially those above 50 years of age, is weak, and there is a lack of own-career management skills.

  • The low efficiency of labor offices is sometimes argued to be a hurdle for employment growth. Policies implemented by these offices do not seem to be well-targeted, and no adequate incentives are provided for their effective intermediation of employment information (Boni, 2002).

Table 2.

OECD Countries: Total Tax Wedge, 1993–2004 1/

(In percent of Average Production Worker’s Earnings (APW))

article image
Sources: OECD (1999), “Taxing Wages”; OECD (2001), “Taxing Wages”; and OECD (2004), “Taxing Wages”.

Total tax wedge is personal income taxes plus social security payments by employers and employees, less cash benefits, as a percentage of labor costs.

For 1993, data corresponds to a single worker without children earning 100 percent of APW.

Table 3.

Selected Countries: Indicators of Strictness of Employment Protection Legislation, 1998 and 2003 1/

(Scale: 0-6)

article image
Source: OECD, Employment Outlook, 2004.

Italics indicate countries with a ranking better than Poland’s. Higher values indicate stricter regulations.

EPL version 1 is an unweighted average of summary measures for regular and temporary contracts only. Version 2 is a summary indicator based on three subcomponents: strictness of regulation for regular contracts, temporary contracts, and collective dismissals, with the latter receiving 40 percent of the weight assigned to the former two.

Table 4.

Indices Measuring the Strictness of Employment Regulation, 2004 1/

(Scale: 0–100, 100 being the most restrictive)

article image
Source: http://rru.worldbank.org/DoingBusiness/ExploreTopics/HiringFiringWorkers

Conditions covered by the indices include availability of part-time and fixed-term contracts, working time requirements, minimum wage laws, and minimum conditions of employment.

25. The current state of the Polish labor market seems, therefore, to be the product of the interaction among shocks, policies, and institutions. The transition and two-stage restructuring shocks have created a large pool of unemployed, rendering certain skills obsolete and affecting regions to different extents. However, institutions and policies have combined with these shocks to delay labor market adjustment. High reservation wages—partly associated with the easy access to benefits—have reduced labor force participation, while a relatively wide tax wedge and strict employment protection laws have dampened job creation. With time, and in the absence of corrective measures, structural problems, such as obsolete skills and regional mismatches, have been exacerbated. Therefore, although persistently low employment growth partly reflects overpriced labor, structural factors have also delayed labor market adjustment.

C. Structural Impediments to Labor Market Adjustment

26. Structural labor market problems in Poland exist both on the demand and supply side. This section complements the studies described in the previous section by exploring labor force survey data to shed some light on hypotheses about impediments to job creation.10 Three specific hypotheses are explored: (i) benefit systems and changing demographics have reduced labor supply; (ii) skills and education mismatches have reduced “effective” labor supply; and (iii) barriers to the mobility of labor and capital have fostered the persistence of regional employment and unemployment disparities.

The Decline in Labor Supply

27. The trend decline in labor force participation over the past 15 years is an important manifestation of structural labor supply problems. The labor force participation (LFP) rate declined from 69 percent in 1992 to 64 percent in 2004, but the decline was uneven across age groups (Figure 2). It was sharply higher for youths (ages 15–24), whose labor force participation dropped from 45 percent to under 34 percent over that period, than for persons in the prime working age (ages 25–54) whose participation remained robust, easing only slightly from 84 percent to 82 percent. The LFP for the older population (ages 55–64) also declined from an already-low 38 percent in 1992 to 32 percent in 2004. Falling LFP therefore reflects a shift in demographics—with working-age population growth of between ½ and 1 percent a year resulting in an increasingly young population over the sample period—combined with a delay in the participation of youth in the labor force at least in part because of growing involvement in education. More worrisome, the trend decline in LFP is also due to the significantly lower participation of the older population, which reflects easy access to disability and preretirement benefits.

Figure 2.
Figure 2.

Labor Force Participation Rate by Age Groups, 1992:Q2-2004:Q3

(In percent)

Citation: IMF Staff Country Reports 2005, 264; 10.5089/9781451831986.002.A002

Sources: GUS: Labor Force Survey bulletins; and IMF staff calculations.

28. These observations are echoed in data that identify education, disability, and retirement as main reasons for inactivity(Figure 3).11 Education has been the almost only cause of inactivity among younger persons and, with a growing working-age population, explains about 26 percent of inactivity today compared to about 23 percent in 1992. This may have positive implications inasmuch as the youths involved are expected to eventually add to labor supply and become relatively more employable as they obtain higher education. Disability, a second cause for inactivity, explained on average during 2000–04 about 23 percent of inactivity among persons aged 15 years and over. This is equivalent to over 3 million persons, or 10 percent of the population in that age group, suggesting generous access to disability benefits in the past, the effect of which has not yet been reversed. Finally, retirement explains almost 34 percent of inactivity among persons aged 15 years or above. About 70 percent of inactive persons on retirement had exceeded the working–age limit of 65 by the third quarter of 2004. Still, almost half of the inactive persons with ages between 55 and 64 attribute their inactivity to retirement, which suggests that part of the low activity rate among older groups is attributable to early retirement. About another one-third of this age group attribute their inactivity to disability, lending support to the hypothesis that the latter has facilitated the low activity rate among older persons.

Figure 3.
Figure 3.

Inactive Persons by Reason, Average 2000-04

(In percent of persons aged 15 or more)

Citation: IMF Staff Country Reports 2005, 264; 10.5089/9781451831986.002.A002

Sources: GUS: Labor Force Survey bulletins; and IMF staff calculations.

Skills and Education Mismatches

29. Skills and education mismatches represent another important structural problem affecting labor supply. About one-fifth of the unemployed have at most a primary education, whereas demand for employees with such a background is relatively small, at 13.4 percent of the total (Table 5). Besides, this category of unemployment is characterized by the longest average duration by far (18½ months on average in 2004), an indicator that the concerned persons are likely to lose their skills to a greater extent than other unemployed persons. Table 5 shows that the majority of the unemployed have a vocational education, a qualification that ought to make them “employable” insofar as labor demand is also strongest for persons possessing this type of education. However, these persons are obviously finding it hard to get a job, given that the average duration of their unemployment status is more than 16 months. While part of the problem may be the price of labor, these data also point to significant weaknesses in vocational education, which seems not to have kept up with changing market needs. This phenomenon may be an important contributory factor to youth unemployment, as about two-thirds of the unemployed youth have a vocational education.

Table 5.

Employment and Unemployment by Education Level, 2000–04

(In percent)

article image
Sources: GUS, Labor Force Survey bulletins; and IMF staff calculations.

30. The professional distribution of labor demand and supply also suggests that skills mismatches may tend to reduce job creation (Table 6). Nearly 18 percent of the unemployed had last held occupations requiring only an elementary education, but demand for such skills captures a mere 7½ percent of total labor demand. A similar situation holds for the unemployed whose last jobs were in crafts or related trades. The average duration of unemployment for persons in these occupations at their last jobs is relatively high (above 15 months), attesting to their difficulties in finding jobs relative to other unemployed with different work experiences. However, demand for professionals and technicians is relatively strong, at 25 percent of total labor demand, but only one-tenth of the unemployed have held their last jobs in that domain.

Table 6.

Professional Distribution of the Employed and Unemployed, 2004:Q2–2004:Q3

(In percent)

article image
Sources: GUS, Labor Force Survey bulletins; and IMF staff calculations.

Although seasonal factors may explain part of the large share of employed in the agricultural sector, the sizeable discrepancy between this share and the share of unemployed whose last employment was in that sector also supports the widely held view that there is significant disguised unemployment in agriculture. Indeed, although the share of agriculture in total employment has declined from 30 percent in 1994 to 17 percent in 2004, it remains very high compared with the OECD average (about 7 percent).

The Persistence of Regional Disparities

31. Labor market patterns differ widely and persistently across regions. Regional unemployment (UMP) rates ranged from 7.4 to 15.6 percent in 1998 (Table 7); this range widened in 2004 to 14.7–26.7 percent. The employment (EMP) rate varied within narrower ranges—48.2–54.9 percent in 1998, widening and declining to 39.1–48 percent in 2004. The same holds for the LFP rate, which ranged between 53.2 and 60.3 percent in 1998, and 51.2 and 58.3 percent in 2004. Differences in employment and unemployment rates across regions have been particularly persistent across time, as evidenced by the high rank correlations (although the lower, albeit positive, rank correlation for the LFP rate indicates a minor degree of persistence over time). Indeed, there is no increased migration into regions where the employment rate is higher, and there seems to be relatively less migration out of regions where unemployment is higher (Figure 4). These data suggest the presence of barriers to mobility of supply and demand of labor across regions.

Table 7.

Labor Market Statistics by Voivodships, 1998 and 2004

(In percent)

article image
Sources: GUS, Labor Force Survey bulletins; GUS, Voivodships from 1995 till 2003 (available online at http://www.stat.gov. pl/english/opracowania_zbiorcze/wojewodztwa/index.htm); and IMF staff calculations.
Figure 4.
Figure 4.

Employment, Unemployment, and Migration Across Regions, 1995 and 2003 1/

Citation: IMF Staff Country Reports 2005, 264; 10.5089/9781451831986.002.A002

Sources: GUS; and IMF staff calculations.1/ The R -squareds of the fitted linear trends are shown in the upper-right corner of the graphs.

32. Insufficient regional wage differentiation appears to contribute to the persistence in regional employment and unemployment disparities. Such differentiation would help close the labor market gap between regions. If wages were low enough in slack labor markets relative to tighter markets (and abstracting from other barriers to labor mobility), labor supply would move to tighter markets and labor demand would move to markets where the price is cheaper, thereby equalizing employment and unemployment rates across regions. The lack of labor mobility therefore suggests that regions with higher employment (unemployment) rates do not seem to have sufficiently higher (lower) wages to induce such mobility. Indeed, there seems to be little, if any, positive (negative) correlation between the regional employment (unemployment) rates and average regional wages, and this correlation weakened in 2004 relative to 1999 (Figure 5). These observations echo the points made in Section B that wages are insufficiently flexible and that minimum wages are relatively too high to prevent a deep enough fall in average wages to generate the needed labor mobility across regions (see also Selassie, 2001).

Figure 5.
Figure 5.

Wages and Labor Market Characteristics by Region, 1999 and 2004

Employment Rate (Percent) and Average Monthly Wages (Zloty)

Citation: IMF Staff Country Reports 2005, 264; 10.5089/9781451831986.002.A002

Sources: GUS; and IMF staff calculations.

33. Infrastructure gaps are another likely barrier to labor mobility. The lack of adequate roads and railways to facilitate access to high-unemployment areas discourages entrepreneurs from establishing their operations in these areas, thus thwarting a shift in demand to regions where labor supply seems relatively abundant. This problem seems prevalent, notwithstanding improvements since the mid-1990s (Figure 6). A number of high-unemployment regions remain inaccessible by motorways and still have less developed networks of other roads and railways.

Figure 6.
Figure 6.

Unemployment and Means of Transportation Serving the Regions, 1995, 1999, and 2003

Citation: IMF Staff Country Reports 2005, 264; 10.5089/9781451831986.002.A002

Sources: GUS; and IMF staff calculations.Note: Railways data were unavailable for 1995.

34. Housing market problems also seem to hamper labor mobility. Restrictive regulations, such as rent control and overly protective tenant rights, have made it extremely difficult for landlords to withdraw from rental contracts and have discouraged the construction of rental units, as reported to IMF staff during the 2005 Article IV consultation mission. This should tend to lessen the availability of rental spaces and deter the mobility of labor across regions, particularly for job seekers who consider moving to big cities, where these problems seem most acute.

D. Forward-Looking Labor Market Scenarios

35. The economic rebound in 2003–04 helped improve labor market performance. With real GDP growth averaging 4.6 percent in 2003–04, employment rose by an annual average of 0.9 percent (mostly driven by the creation of temporary jobs), and unemployment declined. In an indication of improved labor market perceptions, the number of employed looking for another, better-paying job is rising again (although it remains below its 2002 level), while the number of employed looking for another job because of a fear of losing their current job has declined significantly from 2002 (Table 8).

Table 8.

Employment Indicators, 2002 and 2004

article image
Sources: GUS, Labor Force Survey bulletins; and IMF staff calculations.

Average of the first three quarters of 2004. Growth rates are relative to the average of the first three quarters of 2003.

36. But the amelioration in labor market indicators remains modest. The decline in unemployment and increase in employment over the past two years do not measure up to the improvements in the early-to-mid-1990s (Table 1). Employment rates for all age groups remain below their levels in the early 1990s, let alone the peak level in 1998.12 Unemployment rates exceed by far their levels in the early 1990s. Economic growth is strongly driven by productivity gains, and, with employment elasticity of output growth averaging 0.2 in the past two years, the employment rate had edged up only slightly to 52 percent by end-2004. This employment performance falls behind those of EU-15 and OECD countries, for which the employment elasticity of output growth averaged 0.27 and 0.35, respectively, in the past two years. Employment gains are even more wanting, given the large gap that Poland needs to bridge to reach the OECD and EU-15 average employment levels (both at 65 percent). Improving the employment-elasticity of growth is therefore essential.

37. The importance of raising the employment elasticity of growth is illustrated in two forward-looking scenarios. Given population projections and changing demographics, the scenarios show the overall employment growth needed to reach an assumed employment rate ten years from now; they also show, for an assumed labor force participation rate at that same point in the future, the unemployment rate associated with this employment level. The purpose of the scenarios is not to project what levels employment and labor force participation rates are expected to reach ten years from now, which is beyond the scope of this paper. The aim is rather to illustrate the imperative of undertaking structural reforms needed to raise the employment elasticity of growth in Poland.

38. The scenarios are based on actual 2004 data and assumptions on labor market indicators in 2015. The upper panel of Table 9 shows the actual 2004 data, while the next two panels show the scenarios. In both scenarios, the 2015 shares of disabled persons in the prime and old-age groups are calculated from the actual 2004 distribution of disabled persons by age, for which detailed information by deciles for ages 15 and above is available from GUS.13 In both scenarios, the possibility of labor migration is ignored.

Table 9:

Labor Market Scenarios, 2004 and 2015 1/

article image
Sources: GUS; and IMF staff calculations.

Data for 2004 are actual labor force survey averages for the first 3 quarters of the year. Population projections for 2015 are taken from GUS. In both scenarios, the share of youths disabled is assumed to be 1 percent, an innocuous assumption for completeness, while disabled shares for other age groups are based on 2004 data. The LFP and EMP rates (4th and 5thcolumns of the table) are assumed to return to their 1992 levels. These rates then allow to calculate the UMP rate (last column, bolded for each scenario) and the EMP and LFP rates for the working-age population (last row of each scenario).

39. The first scenario shows the implications of raising employment and labor force participation rates to their early 1992 levels. The fourth and fifth columns show these levels, which imply for the overall working-age population an employment rate of 59.2 percent and a labor force participation rate of 71.4 percent. Assuming these rates are reached in 2015, the overall unemployment rate would fall only to 17.1 percent. Demographic factors play a major role: as the population ages, the share of the old-age group in the total population in 2015 would be significantly larger than it is today, producing a much larger labor force at the assumed LFP rate, and, hence, high unemployment in this age category at the assumed employment rate. Despite the still high unemployment rate, the scenario implies an employment growth exceeding 1.1 percent per year over the next ten years, which, at the 2004 employment elasticity of output (0.25), would require 4¾ percent annual real GDP growth over the same period.

40. The second scenario modifies the first by assuming a higher employment rate among the older population. The purpose of this exercise is to derive the growth implications of a greater improvement than depicted in the first scenario, an improvement that would be expected if all the workers currently in the 45–54 age group remain employed ten years from now. In this case, the employment rate for the 55–64 age group would be about 45 percent in 2015, which would raise overall employment rates and lower overall unemployment rates to 61.3 percent and 14.1 percent, respectively. This scenario implies an annual employment growth of nearly 1½ percent over the next decade, which, at the 2004 employment elasticity of output, would require 6 percent annual real GDP growth. Even more dramatic, improving the employment rate to the current OECD/EU-15 level of 65 percent over the next ten years would imply annual employment growth of about 2 percent per year, requiring, with an unchanged elasticity, annual GDP growth rates of greater than 8 percent.

41. The scenarios highlight the need to raise the employment elasticity of output. This elasticity could increase with higher output growth rates, as suggested by actual data for a number of countries, including Poland itself for which a scatter plot of the employment elasticity of growth against growth rates over the past decade points to a positively sloped fitted trend (Figure 7). For Poland in particular, as the young increasingly acquire higher education, some of the existing skills mismatches should be reduced over time thereby improving the employment elasticity of growth. Nonetheless, the substantial fall in the unemployment rate and the strong increase in employment and labor force participation that Poland needs to generate still require policy action directly aimed at raising this elasticity for current growth levels. This purpose would probably best be served by developing a concerted approach to resolving the problems, discussed in Section B, that are hampering labor market adjustment.

Figure 7.
Figure 7.

Employment Elasticity and Growth in Poland, 1993-2004

(In percent)

Citation: IMF Staff Country Reports 2005, 264; 10.5089/9781451831986.002.A002

Source: GUS; OECD; and IMF staff calculations.

E. Conclusions

42. This paper explores evidence suggesting that poor labor market performance in Poland is the result of shocks whose impact has persisted because of ingrained policy and institutional problems and market distortions. Transition, privatization, and economic restructuring, in the midst of demographic pressures, generated a large pool of unemployment, as several factors hindered job creation and fostered the persistence of high unemployment and low labor force participation. These factors include but are not restricted to regional and skills mismatches resulting from shortcomings in the education and training systems, in infrastructure, and in incentives for labor mobility; easy access to benefits; high labor costs; and an insufficient degree of flexibility in regular employment contracts.

43. The recently improved labor market performance falls short of what is needed to achieve strong and durable progress. Continued economic growth alone would generate employment, and sustaining the momentum of economic reform and restructuring (including by easing the process of doing business in Poland) would go a long way toward supporting that growth. However, the current low elasticity of employment growth with respect to output growth would not be enough to generate sufficiently quickly the level of job creation required in Poland. With the continued growth of the working-age population through the end of this decade, and given this low elasticity level, raising employment and labor force participation rates at an adequate pace to reach EU-15 levels over the next 15–20 years, say, would require annual growth rates of 4½ to 6 percent; however, such growth rates seem difficult to sustain over time.

44. Policies are therefore needed to raise the employment elasticity of growth. Recent reforms (described in the Appendix) should continue raising employment and its elasticity vis-à-vis output growth. But more efforts are needed to further improve the functioning of labor markets in Poland by reducing remaining labor market rigidities.

APPENDIX I: Recent Labor Market Reforms

This appendix summarizes recent policy actions taken by the Polish authorities to improve labor market performance.

Labor code revisions14

The 1974 labor code was first amended in 1996; while this usefully clarified many employment rules, the amendments did not seem to remove rigidities from the system overall. The reform was rather favorable to employees in several respects—such as the prohibition of firing employees in case of mergers; the mandatory transformation of a third fixed-term contract into an open-ended contract when it is signed within a month of the date of expiration of the second fixed-term contract; the shortening of the weekly working time (from 42 to 40 hours) and increase in the length of the annual leave entitlement for employees who have worked for at least one year but less than six years in the same company. Some changes were particularly unfavorable to employers, such as the obligation to write a code for work in case the enterprise has a minimum of 5 employees—this used to be 50. Others were more favorable, such as lengthening the maximum probation period from two weeks to three months.

The 2002 reform of the labor code, however, did much to increase labor market flexibility, in ways particularly favorable for small businesses. For example, the third fixed-term contract rule mentioned above was suspended pending EU accession; a new type of contract was introduced to ease the hiring of temporary workers to replace employees on extended leave (maternity leave, for example); sickness pay coverage was removed for the first day of a sick-leave period shorter than six days; mass layoff rules were softened; and the minimum number of employees for which a code for work had to be written was raised to 20 employees. The amendments also introduced flexibility in working arrangements, such as working from home, and cut by half the pay for overtime on nonholiday workdays. Most employee entitlements became a function of the duration of employment with the current employer, rather than total lifetime employment. The resulting increase in labor market flexibility is reflected in the rise in the share of temporary contracts in total contracts from 12 percent in 2001:Q1 to about 30 percent in 2004:Q3.

Additional amendments to the labor code in 2003, largely needed for compliance with EU regulations, represented some backtracking from the 2002 reforms. These amendments, effective from 2004, introduced daily and weekly statutory rests and limits on overtime work, limiting the flexibility of work arrangements. The annual leave allowance was raised for all employees with serving time under ten years, and the right for it would accrue after the first month of work (previously, after the first six months). Leave conditions were also improved for temporary workers by granting them the right to two days of paid vacation for each month worked. The 2002 amendment of sick leave payments was reversed (but this was somewhat justified by abuse that had resulted when the average duration of sick leave had risen above six days). Also, the third fixed-term contract rule was reinstated, effective from the date of EU accession (except for seasonal jobs and temporary job replacement contracts). Mass layoff procedures were lengthened, and entitlements for severance payments were extended in cases of employment restructuring.

Other reforms

The 2002 labor code reform was followed by efforts to increase youth employment. A law passed in October 2002 set the minimum wage for new entrants at 80 percent of the nationally set minimum wage in the first year of work and at 90 percent in the second year. This law, also called the “first job” program, aimed at curbing high youth unemployment. The authorities have reported that, since its inception, 522,600 youth received activation benefits, mostly starting their careers as interns. About 700,000 youth visited counselors in labor offices for assistance in finding a job—the authorities’ estimates indicate that the number of such visits would have been lower by 200,000 without the first job program.

A number of benefit reforms were prepared or introduced in the late 1990s that were expected to raise labor force participation:

  • A reform to the disability system effective 1998 tightened eligibility requirements and introduced frequent reviews of the disability certificate and measures to prevent active accidents. While the reform was effective in reducing claims and limiting abuses of the system, and the number of disabled declined from nearly 4.7 million at end-1997 to 4.4 million at end 2000, this number remained high—at over 4 million in 2004.

  • Measures to tighten the access to early retirement were introduced in the context of the 1999 pension reform and applied only to persons covered by the new system established with the reform. The retirement age became compulsory but remained at 65 for men and 60 for women. A proposal to make retirement ages more flexible for certain groups is under consideration.

  • Some forms of preretirement allowances 15 were eliminated effective January 2002 (albeit with grandfathering of persons receiving these allowances before this date). Further reforms to preretirement benefits were undertaken as part of the Hausner Plan.

  • The 50+ program, in effect from the second half of 2004 onwards, aimed at raising the labor force participation of the population over age 50 by financing the promotion of specific projects by labor offices and NGOs to encourage employment of that age group. However, the results of that program have been disappointing.

  • The law on “Employment Promotion and Labor Market Institutions”, passed by parliament in April 2004, tightened eligibility for registering as unemployed and receiving social benefits. It also redefined responsibilities of labor offices with the aim of improving their efficiency and, hence, the impact of labor market measures. Labor offices were discharged from the role of paying family allowances and preretirement benefits, and were obliged to confirm every month the claimants’ readiness to work—this was previously done every three months. The duration of unemployment benefits in regions where the unemployment rate falls between 100 and 125 percent of the national average was cut by half to six months. The unemployed were obliged to take offered work and to take part in labor offices programs. This likely contributed to the decline in registered unemployment, as 406,000 persons were removed from registers in 2004:Q4 but did not take up work—these persons represented more than 58 percent of total persons removed from unemployment registers in 2004:Q4. This improvement notwithstanding, more time is needed to assess the overall impact of this law on the labor market.

  • In line with the above law, the government adopted a National Plan for Employment in 2005. This plan should provide training for the unemployed or workers threatened by unemployment. It also entails analyzing long-term unemployment, including by examining the role of social assistance in creating unemployment traps, and by assessing future labor demand. This would be a welcome diagnosis of labor market problems and should form the basis of a detailed action plan to promote employment. The Plan also provides for continuing ongoing labor market programs, such as the first job and 50+ programs, as well as the use of EU structural funds to support rural employment.

  • The government also uses active labor market programs (ALMPs) to promote employment. Spending on these programs rose to about Zl 1.3 billion annually during 2003–04, about two-and-a-half times its level in the previous two years; the activation of new graduates captured more than one-third of that amount (Table 1). The effectiveness of most ALMPs has often been questioned outside the government, however.

Table 1.

Expenditure on Active Labor Market Programs and Participants, 2000–04

article image
Source: Ministry of Economy and Labor.

References

  • Boni, Michal, 2002, “The Labor Market in Poland: Current Trends and Challenges,” in Labor, Employment, and Social Policies in the EU Enlargement Process: Changing Perspectives and Policy Options, ed. by Bernard Funck and Lodovico Pizzati, (Washington: World Bank), pp. 5777.

    • Search Google Scholar
    • Export Citation
  • Daseking, Christina, Guorong Jiang, and Victoria Summers, 2000, Republic of Poland: Selected Issues, Country Report No. 00/60, (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Estevao, Marcello, 2003, “Structural and Cyclical Labor Market Changes in Poland,” in Republic of Poland: Selected Issues, Country Report No. 03/188, (Washington: International Monetary Fund), pp. 1443.

    • Search Google Scholar
    • Export Citation
  • GUS, Labor Force Survey in Poland, various issues (Warsaw: Główny Urząd Statystyczny).

  • Keane, Michael, and Eswar Prasad, 2000, “Inequality, Transfers and Growth: New Evidence from the Economic Transition in Poland,” Working Paper No. 00/117 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • OECD, 2001, Benefits and Wages: OECD Indicators (Paris: OECD).

  • OECD, 2004, Benefits and Wages: OECD Indicators (Paris: OECD).

  • Pujol, Thierry,The Role of Labor Market Rigidities During the Transition: Lessons From Poland,” Working Paper No. 96/77 (Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Schiff, Jerald, and others, 2005, Labor Market Performance in Transition—The Experience of Central and Eastern European Countries, IMF Occasional Paper (forthcoming).

    • Search Google Scholar
    • Export Citation
  • Selassie, Abebe, 2001, “Unemployment, Wage Flexibility and the Minimum Wage in Poland: Some Evidence from Micro Data,” (Unpublished; Washington: International Monetary Fund).

    • Search Google Scholar
    • Export Citation
  • Surdej, Aleksander, 2004, “Managing Labor Market Reforms: Case Study of Poland,” (Washington, DC: World Bank), available via internet: http://siteresources.worldbank.org/INTWDR2005/Resources/bp_poland_labor_market_re form.pdf

    • Search Google Scholar
    • Export Citation
  • UNDP, 2004, “Working Out Employment: Human Development Report for Poland” (Warsaw: UNDP).

  • World Bank, 2001, “Poland’s Labor Market: The Challenge of Job Creation,” Country Study No. 23033 (Washington: World Bank).

  • World Bank, 2004, “Poland: Growth, Employment and Living Standards in Pre-Accession Poland,” Report No. 28233-POL (Washington: World Bank).

    • Search Google Scholar
    • Export Citation
  • Yamaguchi, Shintaro, 2005, “Measures of Wage Flexibility and Their Application to Central and Eastern European CountriesIMF Working Paper (forthcoming).

    • Search Google Scholar
    • Export Citation
8

Prepared by Nada Choueiri.

10

The data on which Section C draws are taken from the quarterly Labor Force Survey in Poland bulletins published by the Polish statistical office, GUS, and from information on voivodships available on GUS’s website, http://www.stat.gov.pl/english/index.htm.

11

The category “Other” in Figure 3 includes discouragement with job search (only about 3 percent of the inactive population), and family and household responsibilities (explaining between 8 and 10 percent of inactivity during 2000–04).

12

This is also true for the 25-54 age category but rounding hides this fact in Table 1.

13

It is assumed that only 1 percent of the youth population in 2015 will be disabled. This assumption is made only for completeness and is not crucial for the results.

14

This subsection partly draws on Surdej (2004).

15

Preretirement allowances were granted to older workers—women with 30 years of work experience or men with 35 years of work experience—upon becoming unemployed to bridge the gap to retirement.

Republic of Poland: Selected Issues
Author: International Monetary Fund
  • View in gallery

    Labor Market Trends in Selected Countries, 1993-2004

  • View in gallery

    Labor Force Participation Rate by Age Groups, 1992:Q2-2004:Q3

    (In percent)

  • View in gallery

    Inactive Persons by Reason, Average 2000-04

    (In percent of persons aged 15 or more)

  • View in gallery

    Employment, Unemployment, and Migration Across Regions, 1995 and 2003 1/

  • View in gallery

    Wages and Labor Market Characteristics by Region, 1999 and 2004

    Employment Rate (Percent) and Average Monthly Wages (Zloty)

  • View in gallery

    Unemployment and Means of Transportation Serving the Regions, 1995, 1999, and 2003

  • View in gallery

    Employment Elasticity and Growth in Poland, 1993-2004

    (In percent)