Determinants of gender-specific wages in Germany
new evidence from linked employer-employee data
This thesis consists of empirical and methodological contributions to the literature on the German gender wage gap. Due to a lack of appropriate data, previous studies have been unable to document the potential impact of employers and establishments on the gender wage gap. We investigate these issues using a newly available Linked Employer-Employee Dataset. The first study pays attention to the gender wage gap within establishments. We find that the gender wage gap varies tremendously across establishments. Establishments with work councils and those covered by collective wage agreements are found to have a smaller wage gap. Furthermore, establishments operating under strong product market competition behave in a more egalitarian way. In the second study, the key issue is to disentangle gender differences in human capital endowment and the segregation of men and women in different types of establishments as sources of wage inequality. Using an approach of Machado and Mata (2005), we apply a decomposition method extended to four terms across the entire wage distribution. The analysis shows that the gap is highest in the lower part of the wage distribution. This is largely explained by segregation of women into less successful and productive firms. Gender differences in the human capital endowment have a lesser impact upon the gap. The third study comprises an investigation of the relationship between the share of women in establishments and the wages of both sexes. For this correlation, hypotheses are formulated as to what a high proportion of women in an establishment can indicate: attractive working conditions for women, lower qualification requirements or less discrimination against women. The results show, that even when including worker and establishment covariates, a higher share of female employees reduces wages for both males and females. In particular, attractive working conditions lead to lower wages in female dominated establishments.
School:Humboldt-Universität zu Berlin
Source Type:Master's Thesis
Keywords:Geschlechtersegregation Linked Employer-Employee Datensatz gender-specific wage differentials gender segregation data
Date of Publication:06/04/2009