School density and inequality in student achievement

In the us, test score gaps by socioeconomic status and race increase with city size. This paper examines to what extent residential sorting on school quality can explain this fact. We combine 15 years of data on public elementary school students in North Carolina with geocoded school locations and proxy city size with a measure of school density in a local labor market. Assortative matching between student advantage and school quality markedly increases with city size, accounting for 10% of the city-size gradient in test score inequality. Assortativeness is strongest in the high-income neighborhoods of large cities.