Using administrative data on more than six million German high school graduates from 1995 to 2018, this paper studies how labor market conditions around high school graduation affect postsecondary skill investments. Exploiting cyclical deviations from secular state-specific trends, I find procyclical overall university enrollment, especially at academic universities. Long-run attainment effects are procyclical for academic universities and countercyclical for applied universities. Using large-scale survey data, I present changes in expected returns to different degrees as a potential mechanism. During downturns, graduates expect lower returns to university degrees, while expected returns to the outside option, apprenticeships, are stable.

This paper studies matching efficiency in higher education systems with substantial stratification within universities. Exploiting unique quasi-experimental variation in the staggered introduction of field-specific university rankings and administrative data on the universe of German students from 1995-2018, we show that top ratings based on student satisfaction attract more mobile students. This finding is consistent with a choice model where mobility is costly, and information frictions create inefficient matching outcomes. Using survey data on university graduates, we find that top ratings based on research reputation reduce achievement mismatch with classmates, but that neither peer nor research ratings have a positive effect on labor market outcomes.

Using survey data on seven cohorts of German university graduates from 1989 to 2013 and administrative data on first-year students, we study whether a state label for elite universities changes what employers pay and where students enroll. Germany’s Exzellenzinitiative designated nine universities in 2006/07, after the 2009 graduates had chosen their university. We find no effect on the monthly entry wages of the 2009 and 2013 graduates and rule out premiums above 8 percent. First-year enrollment at the designated universities did not change either.