Persistence of the school entry age effect in a system of flexible tracking

In Germany, the streaming of students into an academic or nonacademic track at age 10 can be revised at later stages of secondary education. To investigate the importance of such revisions, we use administrative data on the student population in the German state of Hessen to measure the persistence of school entry age's impact on choice of secondary school track. Based on exogenous variation in the school entry age by birth month, we obtain regression discontinuity estimates for different cohorts and grades up to the end of secondary education. We show that the effect of original school entry age on a student's later attending grammar school disappears exactly at the grade level in which educational institutions facilitate track modification.