Inconsistencies in Risk Analyses for Ambient Air Pollutant Regulations

28 September 2016
By Dr. Anne E. Smith

In an article in a special issue of Risk Analysis on air pollution health risks (the September 2016 issue), NERA Managing Director Dr. Anne E. Smith identifies inconsistencies between the assumptions that the US Environmental Protection Agency (EPA) uses to estimate an air pollutant’s health risks when the agency decides how stringent that pollutant’s National Ambient Air Quality Standard (NAAQS) should be and the assumptions it uses to estimate that same regulation’s benefits in EPA’s accompanying Regulatory Impact Analysis (RIA). As a result, as Dr. Smith’s analysis demonstrates, very large fractions of the regulatory benefits that EPA attributes to a tightening of a NAAQS are projected to be occurring in areas of the country with air quality that the same rulemaking is simultaneously deeming clean enough to be protective of the public health with a margin of safety. This raises a question of the degree of confidence that should be assigned to EPA’s total benefits estimates. The article provides a quantitative example based on the 2012 revision of the fine particulate matter (PM2.5) primary NAAQS. It also explains how the degree of overconfidence is amplified for estimates of co-benefits from PM2.5 in RIAs for other rulemakings. Quantitative examples are provided for co-benefits from the RIAs for the Mercury and Air Toxics Standards and the Clean Power Plan.

In the same issue, Dr. Smith also contributes an “Author Synthesis and Response” to a set of invited commentaries by other air pollution risk assessment professionals. She notes their general consensus on the need to improve quantitative characterization of risk estimate uncertainty beyond statistical error bounds reported in epidemiological studies, and their recognition that the most important non-statistical uncertainties to address are differential potencies of the many ambient PM2.5 constituents and whether a causal relationship continues to exist at ever-lower ambient concentrations. Dr. Smith suggests that such advancement might be facilitated by more explicitly distinguishing between the needs for hazard identification and for dose–response assessment in future discussions of methods for incorporating epidemiological evidence into the risk analysis process.

Smith, Anne E. (2016, September). Inconsistencies in Risk Analyses for Ambient Air Pollutant Regulations. Risk Analysis 36/9, ©2016 Wiley Periodicals, Inc., a Wiley company.