From Urban Form to Friending Bias: Testing Jane Jacobs’ Hypotheses

Abstract

Recent work has shown that social integration can influence a variety of outcomes ranging from educational attainment to overall health to income inequality. This paper examines the extent to which urban design can foster cross-class interactions and promote social integration in American cities. Measuring cross-class interactions using individual-level GPS cellphone mobility data and cross-class social ties using large-scale Facebook friendship data, I show that three features of the urban built environment — small city blocks, entertainment place density, and mixed primary-use buildings — promote interactions between individuals of different socioeconomic groups, leading to greater cross-class social ties. The findings remain significant with instrumental variable estimates. Consistent with the ideas of Jane Jacobs and Gordon Allport, these results suggest that the urban built environment can reduce cross-class social segregation by promoting greater cross-class interactions.