Keywords: PM2.5, coal combustion, propensity score matching, source-oriented, IHD
There is increasing evidence that particulate air pollution originating from different sources impacts health differently. Existing methods have typically focused on identifying health impacts of broad categories of sources (coal combustion, traffic, etc). While valuable, focus on broad source categories limits the utility of these methods to address the relationship between emissions from specific point sources (a single power plant, for example) and health. Assessing the health effects of specific point sources of emissions is essential to evaluating public policy interventions. In this paper, we deploy a novel statistical analysis method to more directly isolate the association between fine particulate matter originating from coal-fired power plants and health. We use a reduced-form chemical transport model to assess the exposure of 21,351 U.S. ZIP codes to emissions from each of 783 coal-fired power generating units operating in 2005. Then, we use propensity score matching to match highly exposed locations to suitable controls and then estimate the impact of coal emissions on ischemic heart disease hospitalizations in the Medicare population.