This paper makes the following point: “detracking” schools, that is preventing them from allocating students to classes according to their ability, may lead to an increase in income residential segregation. It does so in a simple model where households care about the school peer group of their children. If ability and income are positively correlated, tracking implies that some high income households face the choice of either living in the areas where most of the other high income households live and having their child assigned to the low track, or instead living in lower income neighbourhoods where their child would be in the high track. Under mild conditions, tracking leads to an equilibrium with partial income desegregation where perfect income segregation would be the only stable outcome without tracking.
Fraja, G. D., & Martínez-Mora, F. (2014). The desegregating effect of school tracking. Journal of Urban Economics, 80, 164-177. https://doi.org/10.1016/j.jue.2014.01.001