Abstract
In this paper we approach the regional unemployment dynamics in Poland. Using policy relevant NUTS4 level data from 1999 to 2006, we employ tools typically applied to income convergence analyses to inquire the patterns of unemployment distribution. We apply diverse analytical techniques to seek traces of convergence, including β and σ convergence as well as pass-through analysis.
We demonstrate that the unemployment rate distribution is highly stable over time, while only weak “convergence of clubs” is supported by the data and only for the high unemployment regions. Results suggest no support in favour of β-type convergence, i.e. convergence of levels. Even controlling for nation-wide labour market outlooks (conditional convergence) does not provide any support to this hypothesis. Further, regions with both very high and very low unemployment show signs of high persistence and low mobility in the national distribution, while those in the middle tend to demonstrate higher mobility and essentially no persistence of regional unemployment differentials. This diagnosis is confirmed by σ-convergence analysis which indicates no general divergence or convergence patterns. Transitions seem to be slightly more frequent, but at the same time less sustainable for middle range districts, while movements up and down the ladder occur predominantly for the same districts.
This methodology allows to define the patterns of local labour market dynamics, pointing to differentiated divergence paths. Importantly, these tendencies prevail despite cohesion financing schemes, which allocate relatively more resources to deprived regions.
This is a preview of subscription content, log in via an institution.
Buying options
Tax calculation will be finalised at checkout
Purchases are for personal use only
Learn about institutional subscriptionsNotes
- 1.
Kernel density estimates (KDE) were employed, among others, by Bianchi and Zoega (1999), Lopez-Bazo et al. (2002, 2005). In fact, our approach differs significantly in that we have KDE conditional on the distribution (again, KDE) in the previous period, which makes this technique so suitable for analysing σ-convergence.
- 2.
In Poland the entitlement to unemployment benefit is temporary and lasts only 12 months after the registration (18 months in regions with more labour market hardships). After this period, unemployment benefits may be replaced with social assistance benefit (which is lower and based on family income rather than labour market status). The entitlement to unemployment benefit is re-established for an unemployed who obtains legal employment for a period above 6 months. No publicly available statistics report the effective share of long-term unemployed or unemployment duration. For example, if one does not confirm `willingness to undertake employment’, one is de-listed from the unemployment registries. However, 3 months after this occurrence, one may register again (the basic incentive is free access to public health care system for the unemployed and his/her family) and then, the unemployment tenure is calculated from the scratch. However, benefit entitlements are not. Consequently, the share of unemployed still retaining the right to the benefit is a reliable measure of actual rate of long-term unemployment.
- 3.
Testable hypothesis of local and national unemployment rates cointegration can be formulated as \( \forall t:{\lim_{s \to \infty }}E({U_{i,t + s}} - {U_{j,t + s}}|{I_t}) = constant \), where U i,t denotes respective unemployment rates and I t is the conditioning information set. To be precise, this is a conditional stochastic convergence formula. Unconditional versions would require the limit to approach 0. However, such a condition would discriminate between dispersion convergence scenarios to differentiated levels (so called convergence of clubs) classifying it as non-convergence. Allowing a non-zero constant, permits to account for regional differentiation. This is empirically approached by testing for a unit root in \( {u_{i,t}} = ln{U_{i,t}} - ln{\bar U_t} \), where \( {\bar U_t} \) is the corresponding national average.
- 4.
With the large number of observations (over 400 units for Poland) we uniformly used the Gaussian kernel function, thus implicitly assuming normal distribution. However, Gaussian assumption is by far the most frequently used one, while it only concerns the properties of the nearest surrounding of each point (within the bandwidth windows) and not the distribution as a whole.
- 5.
Fixed window kernel estimate.
- 6.
An approach similar to ours was taken by Overman and Puga (2002) with the main difference that they consider two distinct points in time – namely 1986 and 1996 – for NUTS2 level EU regions.
- 7.
Due to the administrative changes in Poland in 1999 no prior data are available at NUTS4 level.
- 8.
An administrative reform of 1999 has introduced the current structure of NUTS4 levels with the exemption of large cities, whose administrative units were separated from the non-agglomerations only as of January 2001 onwards. Consequently, prior to 2001 for some districts data cover both municipal and rural areas, while after 2001 in each of these cases two districts were formed instead of one, with two separate unemployment rates reported. Since units comprising cities and rural areas were divided to two separate poviats, each with a different labour market structure and potential. Therefore, subsequent to the change, both these units are treated as new in our sample.
- 9.
Yearly rolling change means a change in a given month with respect to the same month in previous year for all months over subsequent years.
- 10.
Over the analysed time horizon Polish unemployment rate moved between 10 and 20% thresholds.
- 11.
Please note that after the initial period the boundaries for decimal groups may change together with the distribution.
- 12.
Graphs for NUTS2 regions were located in order which resembles their boundaries in Poland (a four by four quadratic shape) with the intention facilitating the reader locating NUTS2 aggregates to the maps of NUTS4 presented earlier in the paper.
- 13.
We also observe, that graphs are definitely thicker than in the case of nation-wide analyses, which suggests there is more mobility in the “intra-regional” rankings than in the national ones.
- 14.
This is due to the fact that NUTS4 labour markets averaged for NUTS2 aggregation produce statistically less diversified outcomes.
- 15.
Transition matrices available upon request.
- 16.
Detailed results of Monte Carlo experiment available upon request.
- 17.
Monthly data (relatively high frequency) may exhibit seasonality and autocorrelation. In addition, since units of analysis differ substantially in unemployment levels and changes observed over time, one risks heterogeneity as well. Therefore, our preferred econometric specification is feasible generalised least squares (FGLS) with heteroscedasticity and autocorrelation consistent standard errors and panel-specific autocorrelation structure. More explicitly, we calculate panel-corrected standard error (PCSE) estimates for linear cross-sectional time-series models where the parameters are estimated by Prais-Winsten regression. When computing the standard errors and the variance-covariance estimates, this method assumes that the disturbances are heteroscedastic and contemporaneously correlated across panels.
- 18.
To avoid problems with statistical quality of the estimates, rolling window analysis was performed on data post December 2003 – this significant “shock” to unemployment levels was disastrous to the quality of estimates.
References
Armstrong, H., & Taylor, J. (2000). Regional Economics and Policy, 3rd edn., Blackwell.
Bayer, C., & Juessen, F. (n.d.).
Bianchi, M., & Zoega, G. (1999). A nonparametric analysis of regional unemployment dynamics in Britain. Journal of Business and Economic Statistics, 17(2), 205–216.
Blanchard, O., & Katz, L. (1992). Regional evolutions. Brookings Papers on Economic Activity, 1(1), 1–75.
Boeri, T., & Terrell, K. (2002). Institutional determinants of labor reallocation in transition. Journal of Economic Perspectives, 16(1), 51–76.
Boldrin, M., & Canova, F. (2001). Inequality and convergence in Europe’s regions: Reconsidering European regional policies. Economic Policy, 16(32), 205–253.
Buettner, T. (2007). Unemployment disparities and regional wage flexibility: Comparing EU members and EU-accession countries, Empirica, 34(2), 287–297.
Camarero, M., Carrion-i Silvestre, J. L., & Tamarit, C. (2006). Testing for hysteresis in unemployment in OECD countries: New evidence using stationarity panel tests with breaks. Oxford Bulletin of Economics and Statistics, 68(2), 167–182.
Carlino, G. A., & Mills, L. O. (1993). Are U.S. regional incomes converging? A time series analysis. Journal of Monetary Economics, 32(2), 335–346.
Decressin, J., & Fatas, A. (1995). Regional labor market dynamics in Europe. European Economic Review, 39(9), 1627–1655.
Egger, P., & Pfaffermayr, M. (2005). Spatial β and σ convergence: Theoretical foundations, econometric estimation and application to the growth of European regions. Kiel, Germany: Spatial Econometrics Workshop.
Ferragina, A. M., & Pastore, F. (2008). Mind the gap: Unemployment in the new EU regions. Journal of Economic Surveys, 22(1), 73–113.
Fihel, A. (2004). Polscy pracownicy na rynku Unii Europejskiej (Polish Workers in EU Labour Markets), Warsaw: Wydawnictwo Naukowe Scholar, chapter Aktywnosc ekonomiczna migrantów na polskim rynku pracy (Economic Activity of Migrants on Polish Labour Market), pp. 115–128.
Gomes, F. A. R., & da Silva, C. G. (2006). Hysteresis vs. nairu and convergence vs. divergence: The behavior of regional unemployment rates in Brazil, Technical report.
Gorzelak, G. (1996). The regional dimension of transformation in central Europe. London: Jessica Kingsley Publishers.
Góra, M., & Lehman, H. (1995), How divergent is regional labour market adjustment in Poland? In S. Scarpetta & A. Wörgötter (Eds.), The Regional Dimension of Unemployment in Transition Countries. A Challenge for Labour Market and Social Policies, OECD.
Grotkowska, G. (2006). The case of Poland – Recent changes of non-standard employment and labour market flexibility. In C. Kohler, K. Junge, T. Schroder, & O. Struck (Eds.), Trends in Employment Stability and Labour Market Segmentation, Jena: SFB 580 Mitteilungen, Universitaet Jena.
Huber, P. (2007). Regional labour market developments in transition: A survey of the empirical literature. European Journal of Comparative Economics, 4(2), 263–298.
Kaczmarczyk, P., & Tyrowicz, J. (2008). Wspólczesne procesy migracyjne w Polsce (Contemporaneous Migration Processes from Poland). Warszawa: Fundacja Inicjatyw Spoleczno-Ekonomicznych.
Lehmann, H., & Walsh, P. (1998). Gradual restructuring and structural unemployment in Poland: A legacy of central planning. mimeo, LICOS, Centre for Transition Economies, Katholieke Universiteit Leuven, Leuven.
Lopez-Bazo, E., del Barrio, T., & Artis, M. (2002) The regional distribution of Spanish unemployment: A spatial analysis. Papers in Regional Science, 81(3), 365–389.
Lopez-Bazo, E., del Barrio, T., & Artis, M. (2005). Geographical distribution of unemployment in Spain. Regional Studies, 39(3), 305–318.
Marelli, E. (2004). Evolution of employment structures and regional specialisation in the EU. Economic Systems, 28(1), 35–59.
Newell, A., & Pastore, F. (1999). Structural unemployment and structural change in Poland. Studi Economici, 69(3), 81–100.
Obstfeld, M., & Peri, G. (1998). Regional nonadjustment and fiscal policy: lessons from EMU, NBER Working Paper 6431, NBER.
Overman, H., & Puga, D. (2002). Regional unemployment clusters. Economic Policy, 17(34), 115–148.
Paas, T., & Schlitte, F. (2007). Regional income inequality and convergence processes in the EU-25. Tartu, Estonia: Nordic Econometric Meeting.
Perugini, C., Polinori, P., & Signorelli, M. (2005). GDP growth and employment in EU’s member regions: Convergence dynamics and differences for Poland and Italy, EACES Discussion Paper 2, EACES.
Scarpetta, S., & Huber, P. (1995). Regional economic structures and unemployment in Central and Eastern Europe. An attempt to identify common patterns. The Regional Dimension of Unemployment in Transition Countries. A Challenge for Labour Market and Social Policies, OECD.
Silverman, B. (1986). Density Estimation for Statistics and Data Analysis, Chapman & Hall.
Svejnar, J. (2002a). Labour market flexibility in Central and East Europe, William Davidson Institute Working Paper 496, William Davidson Institute.
Svejnar, J. (2002b). Transition economies: Performance and challenges. Journal of Economic Perspectives, 16(1), 3–28.
Acknowledgement
The authors are grateful for the very valuable comments of Badi Baltagi, Roger Bivand, Ryszard Kokoszczynski, Francesco Pastore, the participants of NEM 2007 in Tallinn, CAPE 2007 in Nuremberg and XXII National Conference of Labour Economics in Naples as well as two anonymous referees. The remaining errors are, of course, ours. Part of the work has been performed while Joanna Tyrowicz was a Visiting Researcher at IZA in Bonn, whose support is gratefully acknowledged.
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2010 Springer-Verlag Berlin Heidelberg
About this chapter
Cite this chapter
Tyrowicz, J., Wójcik, P. (2010). Regional Dynamics of Unemployment in Poland A Convergence Approach. In: Caroleo, F., Pastore, F. (eds) The Labour Market Impact of the EU Enlargement. AIEL Series in Labour Economics. Physica-Verlag HD. https://doi.org/10.1007/978-3-7908-2164-2_6
Download citation
DOI: https://doi.org/10.1007/978-3-7908-2164-2_6
Published:
Publisher Name: Physica-Verlag HD
Print ISBN: 978-3-7908-2163-5
Online ISBN: 978-3-7908-2164-2
eBook Packages: Business and EconomicsEconomics and Finance (R0)