Highways and segregation

This paper examines the impact of the Interstate Highway System, constructed between 1950 and 1990 in the United States (US), on racial segregation. To provide causal estimates, I use the 1947 plan of the Interstate Highway System, a variant of the 1947 plan that connects city center pairs in this plan through shortest-distance and exploration routes in the 16th-19th century, as the instruments for actual highways built. Empirical results from census tracts in the US show that the construction of highways led to sorting along racial lines. I find strong evidence of heterogeneous effects based on the initial black population. I do not find any impact of highways on neighborhoods which have a lower share of initial black population. However, there is an increase in the share of the black population for neighborhoods located in close proximity to highways, and which have a higher share of the initial black population. This increase is driven by the white population moving out and black population moving into these neighborhoods. I demonstrate that whites that moved out of neighborhoods, now commute to work, made possible due to access provided by highways. These residents were on average better educated and had a higher income than the residents moving into the neighborhoods. The reasons for this movement are disamenities emanating from highways, and racial preferences for social interactions. Finally, I show that this relationship between highways and segregation is also observed at the aggregate level. Empirical estimates indicate that one new highway passing through the central city leads to 0.02 units increase in the dissimilarity index for the metropolitan area.