Unemployment and Labour Market Institutions: A Progress Report Olivier Blanchard (MIT), Simon Commander (EBRD & LBS) and Axel Heitmueller (LBS) 1 Introduction A common assumption in the early literature on transition was that unemployment would rise at the start of transition but would follow an inverted U shape with the speed of increase in unemployment being driven primarily by the speed of restructuring and firm closures. Indeed, data on job destruction and creation show that in Central Europe and the Baltic states there were bursts of job destructions early in transition followed - with a lag - by increases in job creation. By contrast, in much of the CIS job destruction was very restrained early on, not least because of the continuation of soft budget constraints. As such, unemployment has risen more gradually over time. In most countries whether of the CIS or Central Europe unemployment has contrary to earlier predictions stayed high and persistent. This paper is an attempt to understand why unemployment has remained high. Our plan is to explore two hypotheses. The first is that changes in labour market institutions account for high unemployment. In a manner similar to Western Europe, this line of argument would require that benefits have become too generous and/or labour taxation too high with the outcome being a high natural rate of unemployment. The second is that high unemployment is due to changes in the composition of the unemployment pool, with the share of workers least able to be matched to jobs rising over time. At this stage, we have explored only the first hypothesis.. As such, the paper is more in the nature of a progress report. 1 The authors thank the Japan-Europe Cooperation Fund and the EBRD for support
The paper is organised as follows. Section 2 provides information on the paths of unemployment, focusing on six transition countries. Section 3 starts with a brief overview of how labour market institutions could be expected to influence unemployment, and then moves on to describe the institutions present in these transition countries. Appendix 1 also provides more detailed information as to how various measures used in the analysis have been constructed. Section 4 presents the results of our initial estimates while Section 5 concludes. 2. Paths of unemployment Figure 1 plots the path of aggregate unemployment in six transition countries between 1991-2005. There is significant variation in both levels and changes in unemployment across countries. What is common, however, is the absence of an inverted U shape for unemployment. Only in Hungary, did unemployment rise sharply at the start of transition before falling steadily. In Poland the early rise was followed by an equally sharp fall in the mid-1990s before a subsequent and sustained increase. In both the Czech and Slovak Republics unemployment has tended to increase throughout the period, although the level in the former has been far lower. In Romania after a sharp increase at the start, unemployment has remained roughly constant since the mid-1990s at around 8%. Finally, in Russia unemployment has increased more gradually before stabilising at around 9% after 2000. Further, there have been very substantial declines in the employment rate. In Central Europe by 2004 the employment rate was below 60%. There has clearly been a large shift into non-participation As unemployment has increased, there have been changes in the composition of unemployment. In particular, there has been a sharp rise in share of long-term unemployment (LTU). In Central Europe and Romania by 2004 LTU rates were around 50% or higher while in Russia the share was under 40%. In terms of educational attainment, unemployment is characterised by relatively high shares of the unskilled. Figure 2 provides unemployment rates for those with primary and tertiary education respectively. With the exception of the Slovak Republic where unemployment is particularly high among people with secondary education, unemployment rates for those with primary education lie significantly above the mean rate and the reverse holds for those with tertiary education. For example, in the Czech Republic the unemployment rate for the primary educated is nearly ten times higher than for those with tertiary education.
3. Institutions & unemployment The persistence of unemployment begs explanation. An obvious place to start is with labour market institutions. In Western Europe and the OECD the available evidence points to a major role being played by institutions, whether directly or in combination with other factors or shocks 2. For example, Nickell, Nunziata and Ochel (2005) using data for the OECD argue that over half of the upward shift in equilibrium unemployment over the period 1960-1995 can be explained by changes in institutions. In particular, they find that (in order of importance) the benefits system; labour taxes, unions and changes in laws for employment protection have materially affected unemployment. They find contrary to Blanchard and Wolfers (2000) that the interaction of institutions and shocks adds no significant explanatory power. Replicating this type of analysis is obviously problematic for the transition countries. For a start the time series dimension of the data is small. Further, the concept of a relatively stable equilibrium unemployment rate has less relevance given the initial conditions and the need for large scale restructuring and reallocation. Nevertheless, the apparent persistence of unemployment suggests that there have been factors at work that have stopped an efficient reallocation of workers and jobs occurring. Given that the transition countries started with relatively generous systems of social benefits at least compared with countries at comparable income levels it is reasonable to ask whether it is institutions that have, as in Western Europe, contributed to maintaining high unemployment. In what follows, we pay particular attention to the level of unemployment benefits, the duration of benefits, the coverage of the benefits system defined as the ratio of beneficiaries to total unemployed as well as the strictness with which the benefits system has been operated. In addition, we consider the weight and strength of trade unions, the extent of employment protection operating in these countries and the level of labour taxation. Wherever feasible, these measures have been calculated in a way consistent with those used by the OECD. Appendix 1 gives a more detailed description of how these variables have been constructed. 2 See, Blanchard and Wolfers (2000)
3.1 Unemployment benefits: net replacement rates The level of benefits is measured as net replacement rates. 3 Figure 3 & Figure 4 give net replacement rates for both long-term and short-term unemployed in the six countries. They show the average rate for ten different family types as well as the highest and lowest net replacement rate for each. 4 It is clear that there are very substantial differences in levels across countries and unemployment types. For long-term rates, for example, Russia has had low and declining replacement rates between 2-5 percent and these have been particularly low for individuals without a labour market history. By contrast, the long-term unemployed in the Czech and Slovak Republics and Poland have received mean net replacement rates of between 40-60 percent. For the short run unemployed, the picture is somewhat different. Figure 4 shows relatively little variation in generosity across countries by 2005 mean replacement rates ranged between 40-60% - and with a common declining trend (except for Romania). The minimum values typically correspond to benefits for individuals without a labour market history. In most countries, their treatment in the long term is not different from the short term, although there are clear differences in the levels across countries. The maximum values mainly comprise lone parents and couples with children. Again there is variation across countries in the level of both short and long run replacement rates. However, over time there is a clear tendency for short and long run rates within countries to converge at lower levels of generosity than in the 1990s. Figures 5-8 do the same exercise but now distinguish between workers at 100% of the average production wage (APW) and those at 66%. There are some interesting differences across and within countries. For the Czech Republic there is little difference in the mean NRR for either type across long and short terms. There is however a clear difference in the level between 100% and 66% of APW. In the other countries, there tends to be an obvious difference between short and long term NRRs for both 100% and 66% categories. For example, in Poland short term NRRs were initially significantly higher than for long term. Over time this difference has largely disappeared for the 100% group although not for the 66% group. In Hungary, short term NRRs while similar for 100% and 66% also initially differed substantially from long term NRRs. While this difference has been maintained, the gap has narrowed, particularly for the 66% group. In 3 Net replacement rate=[benefit income of the family when unemployed-tax on benefit income]/[earned income + Benefit income of the family when employed Tax on earnings and benefits]. 4 The ten family types are: single, couple, couple with two children, lone parent with two children, for people with and without employment history; two earner couple with and without children.
Russia, not only are there particularly large changes over time but the difference between long and short term is huge for both 100% and 66% categories. In short, these figures show that over time mean NRRs have tended to decline for both long and short-term cases (with the exception of Romania). For the period 2000-2005 most countries report NRRs in the range of 0.4-0.6. 3.2 Strictness of benefits The impact of the generosity of the unemployment benefit system on labour market outcomes depends very much on the enforcement of the benefit rules. For example, a generous system does not necessarily lower employment incentives if the access rules are strict and targeted at the most deserving. Our strictness indicator has been collected over time and is based on eight questions capturing aspects such as job search requirements, requirements on occupational and geographical mobility, acceptability of jobs, and sanctions attached to job refusals. The index ranges from 1 (low) to 5 (high) and is unweighted. An important limitation of the indicator is that it relates to the strictness of the rules rather than the strictness of actual enforcement. For example, local employment offices may have the ability to enforce rules differently, depending on local circumstances 5. If so, the variation in the strictness indicator may underestimate or overestimate the actual changes in enforcement over time. Figure 9 gives reports the index for the six transition countries between 1992 and 2005. There are significant differences in levels of enforcement. For example, Hungary has relatively strict rules throughout while the index is lowest for Russia. The other countries exhibit quite similar levels of strictness, except for the Slovak Republic which has seen a sharp increase over the period. However, there are clear differences in the underlying factors driving the differences in levels. For example, in Hungary the strictness measure is mainly driven by sanctions for rule violations rather then requirements on occupational or geographical mobility or proof of job search, while in Russia there are higher requirements for job search but lower sanctions and requirements for mobility. With the exception of the Slovak Republic, there has been a high degree of persistence in strictness over time. Where changes have occurred they have generally been through more stringent requirements to provide evidence of job search activity 5 There is some evidence for Hungary that the frequency of visits to local employment offices varies substantially by region and is often left to the discretion of the employment office (Micklewright and Nagy, 2005)
(Czech and Slovak Republics and Romania) or the requirements for occupational mobility (Poland and Russia). 3.3 Unions & bargaining It is well established that the way in which wages are set in the economy can have a bearing on unemployment rates. Trade unions role in wage setting can have a significant impact on both wages and employment 6. Further, it has been argued that the locus of bargaining can be important - a high degree of centralisation or coordination in wage setting on the one hand or plant level bargaining on the other being viewed as having superior employment implications than intermediate loci of bargaining (such as industry level bargaining) 7. Given that wages in Eastern Europe are generally set by collective wage bargaining for the majority of employees, we construct three different measures; a co-ordination index 8, a centralisation index 9, and an index of the number of central union confederations 10. We have also collected information on union density and union coverage 11. Figure 10 plots the three indices over time. Several things stand out. First, there is significant variation across countries in both centralisation and coordination scores. In the majority of countries the level of co-ordination and centralisation has been extremely low. Yet, in Hungary, co-ordination has substantially increased over time and is currently high despite its low degree of centralisation for collective wage bargaining. Second, the number of trade union confederations between 2-5 union confederations and/or private sector organisations - is similar in most countries, with the exception of the Slovak Republic where wage bargaining is dominated by one union confederation. Third, while in Poland, Slovak Republic and Russia there have been limited changes over time, in both Hungary and the Czech Republic changes have been quite frequent. With 6 See Layard, Nickell and Jackman (1995) 7 There is a large literature on this topic; see, inter alia, Calmfors and Driffill (1988) 8 Co-ordination index (characterising the degree of consensus between the actors in the collective bargaining system. 1:low, 2:medium, 3:high) 9 Centralisation index (characterising the degree of centralisation of the collective bargaining system according to the privileged level of bargaining within national union organisation and national employer organisation: 1:firm level, 2:industry level, 3:national level) 10 Number of central union confederations (3:one dominating union confederation and one dominating private sector employer organisation, 2:the existence of 2-5 union confederations and/or 2-5 central employer organisations, 1:the absence of a central organisation on one or both sides of the labour market) 11 This information is not available on a consistent basis for most countries from official sources. Survey data is often unreliable as many employees are not aware of whether they are covered by a union wage agreement.
respect to the locus of bargaining, until recently collective bargaining has tended to occur at firm level. In the Czech Republic and Hungary, which started out with collective bargaining at the national level, they have gradually moved towards less centralised systems in the mid and late 1990s. Other available evidence also suggests that in general, there has been an increase in wage flexibility over time 12. 3.4 Employment protection We construct an index of employment protection based on hiring and firing practices. Figure 11 indicates some differences in the extent of flexibility across countries: Russia is, for example, more flexible in terms of hiring and firing rules than most of the other countries. There are also some clear changes over time with the Slovak Republic becoming more flexible as its neighbour the Czech republic - has tended to become less flexible. To put this in context, Riboud et al (2000) and Micevska (2003) have estimated that in Central Europe there was generally less or comparable employment protection to that existing in the pre-accession EU countries. Hiring temporary workers has generally been easier in Central Europe (but harder in Romania) than in the EU while collective dismissal (>5 workers) has everywhere been more difficult. Suranyi and Korosi (2003) have used firm level data for Hungary to estimate adjustment costs. They found that such costs amounted to around 3.5 months wages: a level considerably lower than in most of the Western European economies. 3. 5 Payroll taxes Obviously the main taxes of interest are those that determine the wedge between product and consumption wages. In principle, these include income and consumption taxes as well as payroll taxes. Although empirical evidence of the impact of labour taxes on employment has been ambiguous, Nickell et al (2005) find for the OECD not only have they increased sharply since the 1960s but they are also a significant explanatory factor in explaining the rise in unemployment. Figure 12 plots the payroll tax burden and the employer contribution. What is clear is that payroll tax rates have generally been high in the transition countries; they have with the notable exception of Romania been mostly stable. In Romania we observe a very strong increase between the mid 1990s and now. Figure 13 distinguishes 12 For example, responses from employers the Executive Opinion Survey reported in the World
further by type of worker 100% or 66% of APW using OECD data (2005). Figure 13 shows that the tax burden is generally lower for the low wage category but the margin of difference is also small. There seems to be no difference by type in either Romania or Russia. To put this in context, the average tax wedge for single people in France would be 38.5%, in Germany 45.5%. The wedge in these six transition countries is mostly roughly comparable to France and below Germany. 3.6 Real Interest Rates A regression of unemployment on institutions should also include the other relevant determinants of unemployment, initial conditions and shocks. For the time being, we have introduced only one shock variable, the real interest rate. A number of f empirical papers have found interest rates to be relevant in explaining unemployment in Western Europe. The channel is typically argued to be through capital accumulation and hence to labour demand 13. Figure 10 plots the evolution of real interest rates in the six countries. There is large variation across countries Russia, for example, has only infrequently run a positive real interest rate with in addition significant variation over time within countries. Polish real interest rates, for example, jumped from around 2.5 percent in 1995/96 to average over 8 percent between 1997-2001. 4. Estimation some initial results We now explore the relationship between unemployment rates, the labour market institution variables and the real interest rate variable. In order to distinguish types of unemployment we disaggregate unemployment into two categories - namely unemployment for those with low and high educational attainment respectively. Table 1 brings the results together. In the first instance, country and time dummies are excluded. The idea is to see whether institutions can explain time and cross country evolutions. Country dummies are then included to see whether institutions can explain time evolutions. Finally, both country and time dummies are included to see whether institutions can explain specific country evolutions beyond common time evolutions and differences in levels. In the base runs (1-3), the unemployment measure is related to the short and long term net replacement rates alone with and without country and time dummies. Competitiveness Report
Without either country or time dummies, the coefficient on the short term NRR is positive and significant at the 10% level while the long term NRR coefficient is negative and significant at the 5% level. The explanatory power of the estimate is very low. In (2) we add country dummies. The sign, size and significance of the coefficient on the short term NRR remains very similar, but that for the long term NRR switches sign and is significant.(3) repeats the exercise but this time including both time and country dummies. Estimates (4-6) then include the other institutional variables, namely the strictness index, employment protection and tax wedge. None enter significantly with the exception of the tax wedge measure that is actually negatively signed. With country and/or time dummies, the coefficient on short term benefits increases in size and is highly significant. In Table 2 the same exercise is performed but this time the real interest rate variable specified in logs - has been added. In all instances it enters positively and highly significantly. Finally, there is a large degree of inter-country heterogeneity the country dummies are mostly significant. By contrast, there are no significant time effects. 5. Preliminary conclusions What can we conclude from these results? First, we need to be cautious - there are still gaps in the dataset and this has limited the number of observations. Second, we do not adequately cover a number of institutional features, such as the presence or absence of extension agreements. Third, we offer no robustness checks at this stage. Fourth, we have experimented with a number of other variables such as the political orientation of the ruling party, centralisation of wage bargaining, a wage flexibility index but they were all insignificant. The main storyline that emerges at this point is that unemployment in this group of transition countries cannot be explained by the evolution of labour market institutions alone. If institutions matter, it must be in combination with other factors in explaining unemployment. An open issue is whether we would have more success explaining nonemployment rather than unemployment. For example, disability and early retirement pensions may well have provided incentives to leave employment and ultimately the labour force. For example, in Hungary disability benefits have been in place throughout the 1990s and there has been little change in entitlement rules over time 13 Blanchard and Wolfers (2000)
Appendix 1 Data appendix Unemployment rate: Aggregate unemployment rate from the EBRD database as well as disaggregated unemployment rates by education and where possible by previous earnings. Unemployment rates by educational attainment come from Labour Force Surveys and Statistical Offices. Annual net replacement rates: NRR=[Benefit income of the family when unemployed-tax on benefit income]/[earned income + Benefit income of the family when employed Tax on earnings and benefits]. Gross earnings in work for an Average Production Worker (APW) defined as an adult full-time production worker in the manufacturing sector whose wage earnings are equal to the average wage earnings of such workers. It is assumed that the worker is fully employed during the (no sick leave or unemployment). White-collar workers are excluded. Manufacturing is defined as in Division D of the International Standard Industrial Classification of All Economic Activities (ISIC Revision 3, UN). Income includes average amounts of overtime and regular cash supplements (such as Christmas bonuses, thirteenth month payments, vacation month payments). Regular annual bonuses are to be included unless they are dividend payments. Fringe benefits are to be excluded. Where possible annual earnings have been calculated by referring to the average of hourly earnings in the manufacturing sector in each quarter or month, weighted by the hours worked during each period, multiplied by the average number of hours worked during the, assuming that the worker is neither unemployed nor sick (see OECD Taxing Wages, Table V3, p 438 for various examples for different countries). NRR are available for six different family types (single, couple, lone parent, two-earner couples with and without 2 children). They are also available at 100% and 66.7% of the APW. We also have information on the initial (unemployment insurance or unemployment assistance benefits) and long-term (social assistance) rates. NRRs are based on the OECD definitions where possible. The NRR is calculated for an overall category with a full employment record assumed to qualify for the full amount of benefits and a youth category without any employment record. Strictness indicator: Information on the strictness with which unemployment benefit rules are enforced (on paper) based on a questionnaire by the Danish Ministry of Finance, used by Nickell et al. (2005). However, there is a potential problem as the enforced strictness is likely to differ from the legislative strictness in some countries (e.g. Hungary). The index ranges from 1 to 5. Union density: Union members as percentage of employees
Collective bargaining coverage: Percentage of employed labour force whose pay is determined by collective bargaining Co-ordination index: Characterising the degree of consensus between the actors in the collective bargaining system. 1: low, 2: medium, 3: high, intermediate scores are possible Centralisation index: The degree of centralisation of the collective bargaining system according to the level of bargaining within national union organisation and national employer organisation: 1: firm level, 2: industry level, 3: national level, intermediate scores are possible Number of central union confederations: 3: one dominating union confederation and one dominating private sector employer organisation, 2: the existence of 2-5 union confederations and/or 2-5 central employer organisations, 1: the absence of a central organisation on one or both sides of the labour market, intermediate scores are possible Tax Burden: Taken from the OECD publication Taxing Wages 2004 and other local sources. Consist of average tax rates including income tax, employer and employee contribution less cash benefits. This differs from other definitions that have used either marginal tax rates or/and included consumption taxes (see e.g. Nickell, 2005). Currently we have this information from 1995-2004 for Hungary, Czech, Poland and Slovak (2000-2004). Furthermore, the information is available at 100% and 66.7% of APW. Employer Contribution: Taken from the OECD publication Taxing Wages 2004 and other local sources. For the Slovak Republic 1996-1999 Czech Republic values have been used. Hiring and Firing index: Constructed from responses by executives to questionnaires sent out by the World Economic Forum and reported in various s in the World Competitiveness Report Log real interest rate: Taken from the EBRD database mostly using the money market rate and consumer prices. Largest Government Party Policy Direction: Taken from the Database of Political Institutions (DPI) and refers to the political direction of the largest party in government (right, central, left).
References Addison, J. T. and J.-L. (1996). Job security provision and unemployment: revised estimates. Industrial Relations 35: 585-603. Aghion, Philippe and Olivier Blanchard (1994), On the Speed of Transition in Central Europe, NBER Macroeconomics Annual Belot, M. and J.C. van Ours (2000). Does the recent success of some OECD countries in lowering their unemployment rates lie in the clear design of their labour market reforms? IZA Discussion Paper 147 Bentolila, S. and G. Bertola (1990). Firing costs and labour demand: how bad is euro sclerosis? Review of Economic Studies 57: 381-402. Blanchard, O. and J. Wolfers (2000). The role of shocks and institutions in the rise of European unemployment: the aggregate evidence. Economic Journal (conference papers). 110:C1-33 Boeri, Tito and Kathy Terrell, (2002), Institutional determinants of labour reallocation in transition, Journal of Economic Perspectives, 16 Boeri, Tito (2000), Structural Change, Welfare Systems and Labour Allocation, Oxford University Press Bornhorst, Fabian and Simon Commander (2004), Region unemployment and its persistence in transition countries, IZA Working Paper Calmfors, L. and J. Driffill (1988). Bargaining structure, corporatism and macroeconomic performance. Economic Policy: 14-61. Cseres-Gergely (2006). Inactivity in Hungary the effect of the pension system. Institute of Economics, Budapest, mimeo Danish Ministry of Finance (1999). The Danish Economy: Medium Term Economic Survey. Copenhagen Disney. R. (2000). Fiscal policy and employment 1: a survey of macroeconomic models, methods and findings. Washington D.C. IMF Feldmann, Horst, (2004), Labour Market Institutions and Labour market Performance in Transition Countries, University of Tubingen, mimeo Lazear, E.P. (1990). Job security provisions and employment. Quarterly Journal of Economics 105: 699-726.
Micklewright, J. and G. Nagy (2005). Job Search Monitoring and Unemployment Duration in Hungary: Evidence from a Randomised Control Trial. Corvinus University of Budapest, mimeo Nickell, S. and R. Layard (1999). Labour market institutions and economic performance. In (O. Achenfelter and D. Card, eds) Handbook of Labor Economics 3, Amsterdam, North-Holland. Nickell, S., L. Nunziata, and W. Ochel (2005). Unemployment in the OECD since the 1960s. What Do We Know? Economic Journal, 115: 1-27. OECD (2005). Taxing Wages. Paris OECD (1994). The OECD Job Study, Evidence and Explanations, Vol. II. Paris Pissarides, C. (2000). Equilibrium Unemployment Theory, 2 nd edition, Cambridge, MA: MIT Press Riboud, Michelle, C. Sanchez-Paramo and C. Silva-Jauregui, (2002), Does Eurosclerosis Matter? Institutional Reform and Labour Market Performance in Central and Eastern European Countries in the 1990s, World Bank, July Svejnar, Jan (2002), Labour Market Flexibility in Central and East Europe, University of Michigan World Economic Forum (2005). The Global Competitiveness Report (various issues). Geneva, Switzerland
Figure 1 : Aggregate unemployment rates aggregate unemployment rate 0.1.2 0.1.2 1990 1995 2000 20051990 1995 2000 20051990 1995 2000 2005 Figure 2: Unemployment rates by education level and overall 0.1.2.05.1.15.2 0.1.2 0.1.2.3 0.05.1.15 0.1.2.3 primary tertiary aggregate unemployment rate
Figure 3: Net replacement rates long term (averages, max and min).2.4.6.8 0.05.1.15.2 0.5 1 1.5.2.4.6.8 0.2.4.6.2.4.6 Mean long-term Max long-term Min long-term Figure 4: Net replacement rates initial phase (averages, max and min).2.4.6.8 1 0.5 1.4.6.8 1 1.2.2.4.6.8 1 0.5 1.2.4.6.8 Mean short-term Max short-term Min short-term
Figure 5: Net replacement rates (100%) long term Czech Hungary Poland.2.4.6.8 0.05.1.15.2 0.5 1 1.5.2.4.6.8 0.2.4.6.2.4.6 Romania Russia Slovak Mean NRR long-term Min NRR long-term Max NRR long-term Figure 6 Net replacement rates (66%) long-term.2.4.6.8 1 0.1.2.3 0.5 1 1.5.4.6.8 1 0.2.4.6.2.4.6.8 Mean NRR long-term Max NRR long-term Min NRR long-term
Figure 7 Net replacement rates (100%) short-term.2.4.6.8 1 0.5 1.4.6.8 1 1.2.2.4.6.8 1 0.5 1.2.4.6.8 Mean NRR short-term Max NRR short-term Min NRR short-term Figure 8 Net replacement rates (66%) short-term.2.4.6.8 1 0.5 1.4.6.8 1 1.2.4.6.8 1 0.5 1.4.6.8 1 1.2 Mean NRR short-term Max NRR short-term Min NRR short-term
Figure 9: Strictness indicator of the enforcement of unemployment benefit rules strictness index 2.5 3 3.5 4 2.5 3 3.5 4 1990 1995 2000 20051990 1995 2000 20051990 1995 2000 2005 Figure 10: Measures of wage determination Czech Hungary 1 1.5 2 2.5 3 1 1.5 2 2.5 3 Poland Slovak 1990 1995 2000 20051990 1995 2000 2005 Year Co-ordination index Centralisation index Number of central union confederations Graphs by Country
Figure 11: Hiring and firing indices Hiring and Firing 3 4 5 6 3 4 5 6 1990 1995 2000 20051990 1995 2000 20051990 1995 2000 2005 Figure 12 Tax burden (Overall and employer) 30 40 50 60 25 30 35 40 45 20 25 30 35 40 20 30 40 50 20 30 40 50 10 20 30 40 Overall tax burden Employer tax burden Note: Values 1996-1999 for Slovak are from Czech
Figure 13 Tax burden by APW type 30 40 50 60 25 30 35 40 45 39 40 41 42 43 41 42 43 44 40 45 50 55 41 42 43 44 45 Tax burden Tax burden 66% Note: Values 1996-1999 for Slovak are from Czech
Figure 14 Real interest rates Real interest rate -20-10 0 10 0 5 10 15-5 0 5 10-200 -100 0 100-600 -400-200 0 200-10 0 10 20
Table 1 Mean NRR shortterm Mean NRR longterm (1) (2) (3) (4) (5) (6) 0.0580 0.0726 0.1881 0.0035 0.2606 0.2882 (1.67)* (2.46)** (4.00)*** (0.05) (2.89)*** (3.13)*** -0.0467 0.1163 0.0346 0.0112 0.0186 0.0457 (2.04)** (2.15)** (0.60) (0.21) (0.17) (0.40) Strictness index -0.0273 0.1242 0.0356 (0.99) (1.00) (0.25) Hiring and Firing -0.0114 0.0090 0.0075 (1.21) (1.28) (0.94) Tax burden -0.0019-0.0044-0.0004 (1.82)* (2.93)*** (0.18) Constant 0.0712-0.0467-0.1092 0.3092-0.3627-0.2795 (4.14)*** (1.51) (2.78)*** (2.92)*** (0.84) (0.60) Country dummies No Yes Yes No Yes Yes Time dummies No No Yes No No Yes Observations 130 130 130 81 81 81 R-squared 0.02 0.45 0.54 0.09 0.60 0.64 Robust t statistics in parentheses * significant at 10%; ** significant at 5%; *** significant at 1% Note: Dependent variable is aggregate unemployment rate for average over primary and vocational educated or secondary and tertiary educated individuals. See data appendix for variable definitions and explanations 22
Table 2 (1) (2) (3) Mean NRR short-term 0.1383 0.2156 0.2271 (1.91)* (2.41)** (2.44)** Mean NRR long-term -0.0716 0.1008 0.1019 (1.22) (0.83) (0.84) Strictness index -0.0000 0.2506 0.2724 (0.00) (1.63) (1.48) Hiring and Firing 0.0062 0.0111 0.0127 (0.66) (1.56) (1.51) Tax burden -0.0025-0.0041-0.0037 (2.36)** (3.01)*** (1.46) Log real interest rate 0.0465 0.0266 0.0334 (3.34)*** (2.30)** (2.53)** Constant 0.1046-0.8281-0.9490 (0.97) (1.54) (1.61) Country dummies No Yes Yes Time dummies No No Yes Observations 77 77 77 R-squared 0.28 0.63 0.66 Robust t statistics in parentheses * significant at 10%; ** significant at 5%; *** significant at 1% Note: Dependent variable is aggregate unemployment rate for average over primary and vocational educated or secondary and tertiary educated individuals. See data appendix for variable definitions and explanations 23