In this paper, I focus on the impact of one of the most substantial social changes of the past 50 years – the educational expansion – on the selection of high school teachers. This is important, as it could have long-run effects on learning outcomes of students which is the primary goal of this paper.

Exploiting the German educational expansion for both, identification as well as to trace out highly policy relevant effects that are important for shaping future reforms, I find that teachers are adversely selected in times of a high demand for teachers. These effects are still detectable today, decades after the expansion was rolled out. I find small but significant negative effects in student test scores for math and reading competence. To separate initial teacher selection effects from effects due to school sorting, I make use of a triple difference design that can hold potentially confounding parental background characteristics and class mate peer effects fixed – making teacher quality the most likely channel where the effect operates through.