The impact of institutions on economic performance in general, and on wage setting in the labor market in particular, is currently under debate (OECD 2006). In times of increasingly heterogeneous economic conditions the catchword is eurosclerosis, stating that institutional rigidities restrain labor market performance and the dynamics of economic development. A major focus in this context is on the impact of trade unions; see, e.g., the handbook of Addison and Schnabel (2003).
The main channel for unions to influence the wage structure is through collective bargaining. In Germany, this influence goes beyond mere negotiation of wage premia for union members since collective agreements on individual membership premia are forbidden by constitutional law. Given the high rate of collective bargaining coverage in the German labor market, union-bargained wages apply to the better part of all employees and unions influence the wage structure of members as well as of non-members. The design of the German wage-setting system thus offers the possibility to explicitly distinguish between the effects of union density and collective bargaining coverage. We argue that net union density as a proxy for union power governs the union’s threat point in the collective bargaining process and therefore determines the bargaining outcome. Collective bargaining coverage, on the other hand, captures the actual application of bargained agreements. So density and coverage offer a pre-bargaining and a post-bargaining indicator for unions’ influence in the labor market.
The empirical literature on the impact of unions on the German wage structure has so far been confined to using either union membership (e. g., Fitzenberger and Kohn (2005)) or collective bargaining coverage (e.g., Stephan and Gerlach (2003, 2005)). Our study extends upon this literature and analyzes both effects simultaneously.
We use a newly available linked employer-employee data set, the German Structure of Earnings Survey (GSES, Gehalts- und Lohnstrukturerhebung) 2001, which provides individual and firm-level information including an indicator for collective bargaining coverage. Since there is no detailed information on net union density available for Germany, we impute propensities to be a union member for the individuals in the GSES from GSOEP-based estimations of Fitzenberger, Kohn, and Wang (2006). Taking averages of these propensities, we project net union densities for homogeneously defined labor market segments. We then employ OLS and quantile regressions in order to estimate the impact of both union density and bargaining coverage on wage levels and wage dispersion.
From a methodological point of view the analysis involves the challenge to estimate the asymptotic distribution of a weighted quantile regression estimator accounting for clustering, as the estimations contain regressors from different levels of aggregation.
While substantiating the need to employ firm-level as well as individual-level data, our results show a positive effect of firms’ decisions to apply collective or firm-level contracts on the level of wages. Given the share of covered employees in a firm, however, employees with individual contracts ceteris paribus earn higher wages. Moreover, collective bargaining coverage is found to reduce wage inequality. On average, the impacts of net union density on the wage level and on wage dispersion are also negative. While striving for equal wages, powerful unions even make concessions regarding the wage level. Our findings thus are in line with an insurance motive of union representation.
The course of the paper is organized as follows. Section 2 outlines the design of the German system of collective wage bargaining and reviews empirical literature on the links between union density, bargaining coverage, and the structure of wages. Our econometric investigation is presented in section 3. Section 4 concludes.
Contents
1 Introduction
2 The German System of Collective Bargaining
- 2.1 Union Membership and Union Power
2.2 Union Power and the Wage Structure
2.3 Collective Bargaining Coverage
2.4 Bargaining Coverage and the Wage Structure
3 Econometric Investigation
- 3.1 Data
3.2 Descriptive Evidence
3.2.1 Net Union Density and Collective Bargaining Regimes
3.2.2 Wage Levels and Wage Dispersion by Bargaining Regime
3.3 OLS and Quantile Wage Regressions
4 Conclusions
References
A German Structure of Earnings Survey 2001
B Standard Errors for Quantile Regression with Sampling Weights and Clustering
C Tables and Figures
Download
PDF Ebook Union Wage Effects in Germany: Union Density or Collective Bargaining Coverage?
