Wage Rigidities in Western Germany? Microeconometric Evidence from the 1990s
Social Science Research Network
The fact that in the 1980s and in the early 1990s wage inequality has increased significantly in the United States, but hardly, if at all, in Germany, is well-established empirically. Furthermore, there is convincing empirical evidence that the different experiences of the United States on the one hand and Germany on the other can be traced back to differences in supply and institutional factors. As to supply, the rise in wage inequality in the United States has substantially been driven by
... been driven by increasing returns to skills. Yet higher increases in the relative supply of skilled vs. low-skilled labour in Germany than in the U.S. have contributed to the ability of Germany to prevent wage inequality to rise to U.S. levels. As to institutional differences, there is empirical support for the hypothesis that the system of wage bargaining in Germany has prevented wages of low-skilled workers to fall to levels that would be justified by supply and demand developments. The existence of institutional barriers to greater wage dispersion at the lower end of the wage distribution raises the question of whether these barriers cause unemployment of less skilled workers in Germany. In order to test the hypothesis that rising unemployment rates of the low-skilled in Europe are related to the development of wages of this group, I propose an empirical methodology which makes less restrictive identifying assumptions than some previous related studies. In this paper, a 'strong wage rigidity dynamic' is defined as a rising wage rate and a rising unemployment rate in a labour market (one may also call this case 'wage push', as the word 'rigidity' often implies that the wage does not change). A 'weak wage rigidity dynamic in a decreasing market' is defined as a constant wage rate and a rising unemployment rate. The analysis further distinguishes between rigidities related to average wage developments and those related to the wage structure (i.e. relative wage developments). The proposed methodology could serve collective bargaining partners as an information tool in the form of a 'wage monitor'. With a large and detailed data set, one can give advice to collective bargaining partners as to whether (the lack of) recent developments in the wage structure have been warranted by market forces. Collective bargaining partners may pay special attention to sections of the labour market where wage behaviour has been identified as having been 'rigid' in the very recent past. In an application of my methodology to west German individual data from the German Socio-Economic Panel (GSOEP), I find that the west German wage structure exhibited no 'strong rigidities' between 1992 and 1998. However, 'weak wage rigidities in a decreasing market' are found for the characteristic young and for the characteristic white-collar worker: the ceteris paribus analysis demonstrates that these groups experienced an increase in the unemployment likelihood but no change in the wage relative to the 'average person'. What do my results suggest for the validity of Krugman's hypothesis that 'the European unemployment problem and the U.S. inequality problem are two sides of the same coin' (Krugman, 1994, p. 37), namely a fall in the relative demand for low-skilled workers? Krugman's hypothesis implies that -in the face of falling demand -inflexible wages for the low-skilled in Europe have been responsible for rising unemployment rates for this group. The evidence for western Germany in the 1990s demonstrates that this is only partly true: considering different dimensions of skill (e.g. age (experience), education, occupation, and others), I find that there has been no ceteris paribus fall in the relative demand for low-educated and blue-collar workers in western Germany in the 1990s. Hence, the comparatively stable wage structure with respect to low educational achievement and blue-collar occupation seems to have been justified by market forces. Support for Krugman's hypothesis, on the other hand, is provided by the finding that the west German wage structure did not respond to a negative relative demand shock for young workers (who have little labour market experience). This has led to a ceteris paribus increase in relative unemployment for this group. Moreover, at a 'macroeconomic' level, one observes that average real wages as well as the average unemployment rate have risen in the 1990s. Thus, average wages have been 'strongly rigid' (as defined in this paper) during the observation period. Abstract: This paper investigates whether and in what sense the west German wage structure has been 'rigid' in the 1990s. To test the hypothesis that a rigid wage structure has been responsible for rising low-skilled unemployment, I propose a methodology which makes less restrictive identifying assumptions than some previous related work. I find that the relative stability of educational wage premia was justified by market forces. However, relative wages did not respond to negative net demand shocks for young workers, as well as white-collar workers.