In good company : neighborhood quality and female employment

Using a uniquely assembled panel dataset, we estimate the impact of neighborhood and peer effects on female labor supply. Nonrandom sorting and unobserved heterogeneity at the individual and neighborhood levels make recovering these impact parameters more complicated in the absence of (quasi-)experimental variation in neighborhood attributes. Our estimation strategy rests on using a hedonic pricing model to control for neighborhood-level unobserved heterogeneity and using a fixed-effects approach to account for the correlation induced by individual time-invariant unobservables. The results suggest that women’s participation behavior is significantly associated with peer and neighborhood attributes. The extensive margin is driven by the average female employment rate; the intensive margin is driven by the average share of fulltime employed females in the neighborhood. These relationships are stronger in the subsample of mothers. However, these statistically significant associations do not survive when we control for individual time-invariant unobservable heterogeneity.