The mechanisms of action for PM2.5 are not well understood including its contribution to increase frailty (long term exposures) and harvesting among those already frail (short term exposures) as possibilities. While empirical evidence points to a causal relationship between long term exposure to total PM2.5, without understanding mechanisms, the result is an inefficient and misguided regulatory policy that over regulates some types of emissions and correspondingly under regulates others.
Our paper examines the case for differential toxicity of PM2.5 constituents in both the US and Europe as well as different mechanisms that can be used to combine this fragmented (including tools such as meta-analysis, integrated exposure assessment and expert elicitation). To that end, we combine syntheses of the literature such as Janssen et al (2012); Grahame et al (2014); and Wyzga and Rohr (2015) as they pertain to long term exposure.
Using this evidence, we re-evaluate nine recent air quality regulations promulgated by US EPA from 2010 through 2015 where US EPA steadfastly adheres to an equal toxicity assumption. The rules we examine exclude those where the resulting emissions profiles are left to the States or other entities such as the National Ambient Air Quality Standards (NAAQS) or the Cross State Air Rule. Our examined rules cover a variety of emissions sources including power plants, solid waste incinerators, spark ignited and compression ignited standby generators, wood stoves and other major source boilers. We evaluate the regulatory impact analyses from these rules and find that EPA tends to overstate the benefits of large rules (for example with the Mercury Air Toxics Standards (MATS) our analysis suggests benefits on average of $billion per year, and possibly as low as $-7billion instead of the $30-80 billion indicated by EPA). At the same time, the net benefits of smaller rules such as woodstoves and compression ignited internal compression (CI-RICE) have substantially more life saving potential than suggested by EPA. We conclude that as a minimum, US EPA should provide and uncertainty and or sensitivity analysis associated with black carbon.