Trade unions bargain for higher wages, equal pay, fair working conditions, or employment protection (Freeman and Medoff 1984). Classical models such as monopoly unions, right-to-manage models, or efficient bargaining predict a monotonic positive relationship between union power and the level of bargained wages; see the surveys of Farber (1986), Oswald (1985), and Naylor (2003). Some more recent studies emphasize effects on higher moments of the wage distribution.
In line with an insurance motive for union representation of risk-averse workers (Agell and Lommerud 1992, Burda 1995), union impact compresses the wage distribution relative to the distribution of productivities. By enforcing “equal pay for equal work” unions further seek to limit favoritism and discrimination by superiors and colleagues, and to encourage solidarity among the work force (Freeman 1982). Union-bargained wages may serve as wage floors, thereby narrowing the distribution of wages from below.
Collective agreements reflecting unions’ bargaining objectives then have two effects on the structure of wages. First, differences between covered and non-covered segments would increase as the result of the unions’ strive for higher wages. Second, wage compression induced through a collective contract would reduce within-segment inequality. The question which effect would prevail has been discussed for some time in the Anglo-Saxon context; see the survey of Card, Lemieux, and Riddell (2003).
However, the Anglo-Saxon concept of union gaps or membership premia is inappropriate for Germany because collective agreements constituting discriminatory wage policies with disadvantages for non-members are forbidden by constitutional law (negative freedom of association, negative Koalitionsfreiheit, Grundgesetz Art. 9). The scope of collective agreements goes beyond the organized parties. Wages set at the firm level as well as individually bargained wages are adapted towards collective bargaining agreements, be it in order to reduce transaction costs or not to create incentives for employees to join a union. Collective bargaining coverage thus is considerably higher than union density. The decision whether to apply a collective contract or not is basically left to the firms. In the interpretation of Dustmann and Schönberg (2004), firms use collective agreements as a commitment device.
Employees are paid either according to individual contracts between the employee and the firm or according to a collective agreement. The collective agreement can be negotiated between a union and an employers’ association, a union and a firm, or a works council and a firm. Arrangements between firm and works council are only allowed to govern wages or salaries if the firm is not subject to a collective contract or if the collective contract explicitly allows for this type of arrangement. Firm-level agreements involving a union are allowed to set wages even if a collective agreement exists, as long as the firm-level agreement is more specific than the collective agreement. No more than one collective wage agreement must apply at the same time, but not all employees working in a firm applying a collective agreement are automatically covered. Collective contracts may also contain an opening clause explicitly allowing deviations from the terms of the contract under particular circumstances (Heinbach 2006).
Collective bargaining coverage, as measured by the share of employment contracts following collective agreements, was relatively stable in West Germany until the mid 1990’s but has been declining since. By the year 2003, 70% (45%) of West German employees (firms) were covered by a collective agreement (Schnabel 2005). With respective shares of 47% and 26%, coverage in East Germany was markedly lower. The “erosion” towards more decentralized wage setting is examined by a group of studies using firm-level data, and is reconfirmed by survey evidence from works councils.
