Rebecca Allen, Simon Burgess and Tomas Key
This paper uses the pupil census in England to explore how family house moves contribute to school and residential segregation. We track the moves of a single cohort as it approaches the secondary school admission age. We also combine a number of cohorts and estimate a dynamic nonlinear model for house moving with unobserved effects. These approaches yield the same result: moving is significantly negatively correlated with school quality, and segregation does increase as a cohort reaches age 11. However, this relationship is weak: the increase in segregation is slight and quantitative significance of the estimated relationship is low.
