We exploit an unusual policy reform to estimate the causal effects of tuition fees in higher education at the intensive and extensive margin. In 2007, fees for university studies were introduced at a large number of German universities and, critically, also applied to incumbent students who were already enrolled at the time. This feature allows us to estimate causal effects of fees at the intensive margin, net of selection. Using census data on the universe of students enrolled in higher education, we find that modest fees of about 1,000 EUR/year have large and significant impacts on study duration and completion for incumbent students. Combining these findings with causal estimates on enrollment effects at the extensive margin, we discuss new implications for policy.

Joint work with Jan Marcus and Felix Weinhardt