Union Density and Varieties of Coverage: The Anatomy of Union Wage Effects in Germany

Collective bargaining in Germany takes place either at the sectoral level or at the firm level; collective bargaining coverage is much higher than union density; and not all employees in a covered firm are necessarily covered. This institutional setup suggests to distinguish explicitly between the effect of union power as measured by net union density in a labor market segment and the effects of coverage at the firm level as well as coverage at the individual level. Based on standard theoretical results and based on empirical evidence in the literature, one expects that coverage is associated with higher average wages and lower wage dispersion. However, there may also exist spillover effects from the covered sector to the uncovered sector. On the one hand, uncovered firms may pay higher wages to prevent union efforts for coverage in a firm (union threat effect). On the other hand, the reduction of employment caused by higher wages in the covered sector increases labor supply in the uncovered sector, thus resulting in downward pressure on wages there. Covered and uncovered employees may coexist at the same time in one firm. This may be due to the firm paying particularly high wages to good workers under an individual contract or due to the firm only having to pay part of its workforce according to the wage bargaining contract. The latter may for example occur when the firm leaves the employer association and only has to apply the contract to its incumbent workforce but not to the newly hired or when the firm is a merger of a formerly covered and a formerly uncovered firm. Because of legal constraints, wage bargaining can not discriminate between union members and nonmembers. Thus, there is no individual membership effect on wages. Nevertheless, the effect of coverage should depend upon union density in a labor market segment, taking union density as an indicator of union bargaining power. It is to be expected that higher union density increases the effects of coverage on wages because a higher union density should move the bargaining outcome closer to union preferences. Furthermore, union density can also affect wages in the uncovered sector because of the aforementioned spillover effects. This is the first empirical paper which distinguishes the three dimensions of union influence on wages, namely (1) coverage by firm-level or industry-wide agreements, (2) coverage at the firm level or the individual level, and (3) union density. Using a unique linked employer-employee data set for Germany, the German Structure of Earnings Survey 2001, and applying both OLS and quantile regressions for wages, we simultaneously analyze the respective effects on the structure of wages. Ceteris paribus, a higher share of employees in a firm covered by industry-wide or firm-level contracts is associated with higher wages. Yet, individual bargaining coverage in a covered firm ceteris paribus shows a negative impact on the wage level. The negative impact of individual coverage is stronger at higher quantiles of the conditional wage distribution. Collective bargaining coverage thus reduces wage inequality. The findings are in line with the hypothesis that firms apply collective agreements in order to follow a transparent wage policy. However, a risk premium is paid to workers in a covered firm who are not paid according to a collective contract, i. e., who are not covered individually. These workers tend to be the more successful workers in the firm and tend not to be unionized. Their wages are determined by negotiations of the individual worker with the firm. The wages of these workers are particularly high in successful firms for which the coverage effect is also high since collective agreements can extract higher rents from the firms. Having controlled for the varieties of coverage considered, we also find significant effects of net union density on the wage level and on wage dispersion. A higher share of union members in a relevant labor market segment is ceteris paribus associated with lower wages, and the effect is strongest among uncovered individuals and at the upper end of the wage distribution. At the same time, a higher union density reinforces the positive wage effects of coverage at the firm level. A higher union density also reduces wage dispersion. This is in line with an insurance motive of union representation. However, the uniform decline of wages in the uncovered sector across the entire wage distribution in response to an increase in union density requires a further explanation. This finding may be due to a decline of firms’ capital investment when union density increases, or to downward pressure on wages because of the increase in labor supply in the uncovered sector broughtabout by higher wages—and thus lower employment—in the covered sector. The analysis does not allow us to distinguish between these two hypotheses. Our results highlight the importance of using linked employer-employee data in order to control for worker as well as firm characteristics when evaluating the varieties of union effects on wages. Unfortunately, our estimations can not take account of the apparent endogeneity of union density and collective coverage, and so the results should not be interpreted as causal effects. The cross-sectional data do not provide adequate instruments for exclusion restrictions. However, the endogeneity problem is reduced by controlling for both individual and firm characteristics.

Fitzenberger, Bernd, Karsten Kohn and Alexander C. Lembcke (2008), Union Density and Varieties of Coverage: The Anatomy of Union Wage Effects in Germany, ZEW Discussion Paper No. 08-012, Mannheim.