Stuck with a Bad Choice? on the Relation Between Field of Study and the Dynamics of the Gender Pay Gap

This paper deals with gender-specific career dynamics of German university graduates. We focus on the relation between field of study and the dynamics of the gender pay gap over the first 16 years after graduation. Our analysis is based on a new administrative panel dataset containing individual-level data on recent German university graduates and their job market history. The data combine information from a mid-size German university with exceptionally rich German labor market data, namely the Integrated Employment Biographies (IEB) of the German Institute of Employment Research (IAB).

Using fixed effects regression and standard decomposition methods, we show that the field of study explains a significant part of the gender pay gap among graduates in different stages of their careers. As commonly observed in the literature, field of study is a deciding factor when explaining gender differences in salaries at the outset of individual careers. At one year after graduation, we find that female graduates earn 12% less than male graduates due to their field of study. However, our results suggest that the absolute wage penalty due to the chosen field of study is constant over time. In fact, more than a decade after graduation, female graduates' choice of field of study still explains about 10% higher wages of male over female graduates. This holds true even after controlling for individual work experience.

In addition, we provide three extensions to our analysis of the relationship between field of study and the dynamics of the gender pay gap. First, we show that our results do not change qualitatively due to the inclusion or exclusion of occupation and industry affiliations of graduates’ jobs in our model. This is especially important, since it is commonly agreed upon the fact that such variables may incorporate potential discrimination against women. Second, we provide evidence for the necessity to control for disparities in employment biographies such as differences in the timing of both work experience and unemployment. Third, we shed light on the role of current and past labor market conditions in our gender-specific context. Most importantly, we raise the question whether wages of female graduates respond differently than those of males to changes in human capital-related match qualities. Preliminary results from all three extensions strengthen the hypothesis that the wage penalty induced by the distribution of females across fields of study is indeed irreversible.