HOW DO URBAN SEGREGATION AND HIGH SCHOOL CHOICE CONTRIBUTE TO SCHOOL SEGREGATION?

This paper provides a new method for decomposing segregation indices depending on two distinct sets of unities. This method is applied to measure the contributions of residential segregation and parental school choice to social segregation amongst middle schools, using a geolocalized dataset and boundaries of the public school catchment areas for pupils living in three French urban areas. The choice of some parents to opt out of the local public school, choosing mostly private schools instead, contributes to school social segregation by as much as 37% to 49% depending on the urban area. This is mainly because the social mix in "left-behind" public schools is reduced, as parents opting out are often the most socially advantaged ones locally.