School choice, segregation, and forced school closure

We exploit the forced closure of three segregated primary schools in Amsterdam to establish the determinants of school choice of ethnic minority pupils. The schools were closed due to mismanagement and poor assessment from the Education Inspectorate. Most of the affected students were of socially disadvantaged and non-western migrant background. Our analysis contrasts the respective school choice decisions of the early movers who had voluntarily changed schools within two years before the forced closure and the forced movers who had to move to other schools after the closure. Using a conditional logit model and a nested logit framework, we find that i students of segregated schools tend to re-concentrate into the same schools rather than disperse into different schools; ii primary school choice is nested upon school type; and iii the forced movers prefer schools with more peers of own non-western and low socioeconomic background, less peer truancy, and shorter residence-to-school distance.