Association for Public Policy Analysis & Management

Recent research examining whether there are systematic race- and/or class-based disparities in the enforcement of environmental laws has come to mixed conclusions. Some research shows disparities in regulatory actions such as compliance inspections and administrative sanctions for violations (Konisky and Reenock 2013; Gray and Shadbegian 2004), while other studies find few such disparities (Gray and Shadbegian 2012).  One possible explanation for these mixed findings is that researchers typically do not make distinctions among minority and lower-income communities. Specifically, some communities are more over-burdened from pollution risks than others, either due to hosting a disproportionate number of pollution facilities or because the facilities in their area are especially large sources of pollution.
Given limited resources, one might expect governments to direct their enforcement activities in these areas. The U.S. Environmental Protection Agency, for example, has long had policies in place that direct the agency to focus their enforcement efforts in these communities (Konisky and Reenock 2014). Many states have similar policies in place (Ringquist and Clark 2002). Alternatively, these communities may face additional hurdles in overcoming collective action problems, and as a consequence be less effective in pressuring government officials to pursue strong enforcement.

We test these competing hypotheses using a novel empirical approach that combines fine-grained geographic data on environmental risk with facility-level regulatory enforcement data. Specifically, we use data from the EPA’s Risk-Screening Environmental Indicators (RSEI) model to develop a geographic “riskscape” (Abel and White 2011; Morello-Frosch, Pastor, and Sadd 2001) for the entire United States. The RSEI is a screening tool that combines information on the amount of toxic chemical releases, the degree of their toxicity, and the size of the exposed population to determine a numerical risk score of environmental risk for a given facility or geography. We use data from the RSEI at a geographic resolution of one kilometer-by-one kilometer, which provides a high-level resolution measure of environmental risk. We then use GIS software to create two types of neighborhood-level (i.e., one-mile radius around a facility) measures for each of approximately 15,000 major air polluters regulated under the U.S. Clean Air Act: 1) a risk score; and 2) indicators of the percentage of the population that is African-American, Hispanic, and below the poverty line.

We will then estimate regression models to test if government enforcement effort is directed at polluting facilities in minority and low-income areas that experience high levels of risk. This will be done by creating and interpreting multiplicative interaction terms between the neighborhood-level measures of risk and demographic indicators. Controlling for other economic and political contextual factors, the coefficients on these interaction terms provide tests of whether enforcement disparities are exacerbated or mitigated by underlying levels of community risk. The findings from this analysis will advance our understanding of the nature of the race and income-based disparities in regulatory enforcement, and in so doing contribute to the environmental justice literature.