If an environmental justice policy cleans up a local environment and leads to re-sorting, will that widen or narrow inequality? We combine an equilibrium sorting model characterizing renter location choices with a new approach to causally estimate the impact of a cleaner environment on expected welfare to answer this question. We estimate the model leveraging a plausibly exogenous change in local air quality due to shale gas propensity, together with spatially-granular bilateral migration, air quality, and emissions data. We find that while improving local air quality can reduce race-based pollution disparities, it can also widen welfare disparities. We demonstrate that a partial equilibrium approach to answering this question substantially over estimates the expected welfare benefits compared to a general equilibrium approach.
