Entry and Exit Patterns of "Toxic" Firms

We pair an establishment-level dataset from Texas with public information available in the Toxics Release Inventory (TRI) to evaluate the standing of dirty industries in Texas census tracts with a focus on environmental justice concerns. The share of nonwhite residents in a tract is positively correlated with the number of TRI-reporting firms, and an inverse-U-shaped relationship characterizes the number of TRI-reporting firms and a tract's median income. Even after controlling for factor prices and other covariates that might drive firm location decisions, entrants that report to the TRI are more likely to locate in areas with a higher share of nonwhite residents. Firms that report to the TRI are also more likely to enter areas with a low share of college graduates. In contrast, the number of entrants from industries that do not have TRI reporters is negatively related to the percentage of nonwhite residents in a tract. Firms in these non-reporting industries are also more likely to enter areas with a high share of college graduates. Polluters appear to agglomerate, raising concerns about both chemical releases being concentrated in certain tracts, and also affecting nonwhite-dense areas disproportionately. The strength of these effects often depend on an urban/rural classification, with rural areas experiencing the most pronounced concerns. Moreover, TRI-reporting firms are less likely to exit the market relative to their peers that operate in the same industry but do not need to file TRI reports, suggesting that releases may affect a region in the long run.