Members of the Science Advisory Board (SAB), thank you for taking the time to hear to my comments this morning. Today’s topic—how to measure the impact of Environmental Protection Agency (EPA) regulations on low-income and minority citizens in the United States—is both timely and important. At the research center where I work, we have begun to explore the consequences of regulations on vulnerable populations. I appreciate the opportunity to share some of our findings and to contribute to this important discussion.
I am currently a program manager at the Mercatus Center at George Mason University. We are a university-based research center dedicated to bringing economic scholarship to policymakers. In my role, I manage our portfolio of research related to federal regulations. At Mercatus, we focus a great deal on the economic analyses that federal agencies produce, known as regulatory impact analyses or RIAs. In addition to my role at the Mercatus Center, I am also a doctoral student studying economics at George Mason University.
In this statement, I plan on addressing the following issues. First, environmental justice is not just about the state of the environment in which vulnerable populations live, but also their health, which is considered an important component of environmental justice according to the relevant presidential executive order. Unfortunately, the draft version of the EPA’s guidance for incorporating environmental justice into regulatory analysis does not adequately address the impact of regulatory costs on the health of low-income and minority populations. Regulatory costs, and the income losses associated with those costs, impact health every bit as much as regulatory benefits do.
Environmental justice is also related to the ability of citizens to mitigate risks in their own lives. This has been explicitly stated by the EPA in the new draft guidance. For this reason, it is vitally important that the EPA focus on mitigating those risks that the poor are willing to pay for given the costs that are imposed on them. There are many risks that people must address using their own income, such as their choice of neighborhoods, diets, and cars, for which additional health and safety is purchased at higher prices. This means that regulations that purchase health- and safety-risk reductions at very high costs may be crowding out private purchases that would lower risk much more efficiently.
Unfortunately, the way the EPA currently measures benefits is biased in a way that can overestimate the benefits of EPA rules to vulnerable populations. The EPA uses a mean estimate of the populations’ willingness to pay (WTP) to reduce risks when it calculates many types of regulatory benefits. While this metric is appropriate to measure the overall efficiency of a rule, it is not sufficient to measure benefits to subgroups in the population whose WTP may differ from the population mean WTP. Since environmental justice is concerned with distributional effects, then in addition to using a mean estimate of WTP, the EPA should use a WTP estimate for the individual subpopulations being impacted by a particular regulation. Otherwise, the EPA will systemically overestimate the benefits of a rule to those with modest incomes while underestimating the benefits to the rich. This results from the fact that the wealthy are generally willing to pay more for risk reduction given their higher incomes.
Lastly, the EPA should consider more closely the impacts of its regulations on employment, which also has important distributional consequences. For example, losing one’s job due to a regulation can affect lifetime earnings as well as health and may also contribute towards issues like income inequality if the compliance jobs created by rules require higher skills than production jobs that are lost.
I conclude this statement with recommendations about ways to improve upon the EPA’s draft guidance, including ways the EPA can provide more information about the benefits and costs of its rules to vulnerable populations and gather meaningful feedback from these groups in a transparent manner.