Because of the Dutch tracking system, primary school teachers in the Netherlands can have a vocational or a higher secondary background. Policymakers and school principles worry that teachers with vocational backgrounds are less capable to teach math and reading. This study therefore examines the eects of ability tracking of future primary school teachers on the students' math and reading performance. We exploit data of 91 schools for all primary school children in grades 3, 4 and 5 and identify the tracking eect by exploiting unique information on how teachers are assigned to classes based on their teaching abilities. The estimation results for math (reading) indicate that test scores are .2 (.12) of a standard deviation lower if their teacher had a vocational background. The results
for reading are, however, not signicant at the 10% condence level and the tracking estimates appear to be less stable and precise.
JEL-codes: I20 I21 I29
Keywords: Teacher Student achievement Ability tracking Test Scores