We explore the importance of modeling strategies when estimating the emissions-income relationship. Using U.S. state-level panel data on nitrogen oxide and sulfur dioxide emissions, we estimate several environmental Kuznets curves using the standard parametric framework as well as a more flexible semiparametric alternative. Formal statistical comparisons of the results overwhelmingly reject the parametric approach. Moreover, the differences, particularly for sulfur dioxide, are economically significant.

This study examines the effects of air quality regulation on economic activity. Anecdotal evidence and some recent empirical studies suggest that an inverse relationship exists between the stringency of environmental regulations and new plant formations. Using a unique county-level data set for New York State from 1980 to 1990, we revisit this conjecture using a seminonparametric method based on propensity score matching. Our empirical estimates suggest that pollution-intensive plants are responding to environmental regulations; more importantly, we find that traditional parametric methods used in previous studies may dramatically understate the impact of more stringent regulations.