The effect of employment protection and product market regulation on labour market performance: substitution or complementarity?

Abstract

This article provides new evidence on the linkages between various forms of market regulation and joblessness and its components, unemployment and inactivity. One crucial contribution of this article is the analysis of the interdependence across Product Market Regulation (PMR) and labour markets regulation (Employment Protection Legislation (EPL)). With the help of a dynamic fixed effect model with an interaction term, we estimate the marginal impact of EPL and PMR at different levels of the other interacted variable. To cope with problems related to the inclusion of time-invariant institutional variables in fixed effect models, we present results of regressions based on a new procedure, specifically designed to treat slowly changing variables. We build time-series data to account for the annual evolution of EPL, and use new data for unemployment insurance net replacement rates. Among other results, we find evidence of a positive (negative) effect of EPL (PMR) on employment performance as well as of substitution, rather than complementarity, between the two forms of regulation.

Acknowledgements

This research was supported by funds from the ESEMK European project (FP6 STREP1). The authors wish to thank Thomas Plümper and Vera Tröger for making their FEVD code available, as well as A. Clark, L. Faggio, Y. L’Horty for their helpful remarks. We gratefully acknowledge the access to the FRDB Social Reforms Database. We are also grateful for the comments provided by participants in research seminars at CEPN and PSE as well as by participants in the IZA-FRDB Workshop ‘Measurement of labour Market Institutions’, EEA, AFSE, EALE and SASE annual conferences.

Notes

¹ See, for instance, Nickell (1997), Siebert (1997), Blau and Kahn (1999) and Bertola et al. (2002).

² See Bertola et al. (2007), for a recent article focusing on employment.

³ See Bertola et al. (2007), for the effects of institutions on employment for various subgroups of the population.

⁴ The FRDB Social Reforms Database has been developed by the FRDB (http://www.frdb.org).

⁵ Only few empirical contributions address the issue of the employment consequences of financial constraints. This is partially due to lack of satisfactory empirical measures of financial market imperfections. Nickell and Nicolitsas (1999) address this issue by studying the effects of interest rate increase on a panel of UK firms: the authors show that credit constraints yield negative effects on employment.

⁶ Soskice and Iversen (2000) show that large unions have an incentive to moderate sectoral money wages, and thereby expected real wages, if the central bank is nonaccommodating. The result is an increase in the real money supply, and hence higher employment. However, Lippi (2003) shows that if wage setters are nonatomistic, more conservatism may either increase or decrease equilibrium unemployment, depending on the the number and size of trade unions.

⁷ Note that the real interest rate is not included because it turned out to be nonsignificant in all estimations.

⁸ Monte Carlo experiments suggest that the Fixed Effect Vector Decomposition (FEVD) estimator is the least biased estimator when time-variant and time-invariant variables are correlated with the unit effects.

⁹ Results can be obtained from the authors.

¹⁰ Estimations with GLS and PCSE estimators are available upon request. Regressions’ results are consistent over the different estimation techniques.

¹¹ However, one should interpret this result with caution due to the lack of a control variable controlling for structural features of financial markets.

¹² This result is not specific to the FEVD estimation.