UC Berkeley

People of color in the United States are disproportionately and unfairly exposed to air pollution. Equity-oriented scientific evaluations quantifying these disparities often use population-average exposure metrics to capture the overall inequality within a system. Utilizing these metrics involves choices about the exposure input for assessing disparity, the study geography, and the reference population, which are critical to understanding disparities and effectively designing interventions. Here, we use a case study of exposure to fine particulate matter (PM_2.5) from California's agricultural sector to dissect the implications of these decisions. Using a reduced-complexity model and emissions of PM_2.5 and precursors, we compare estimates of racial and ethnic disparities in exposure resulting from different combinations of these methodological choices. The full population distributions highlight differences between disparities at the extremes (e.g., 90th percentile) and at the mean. Additionally, the selection of study geography and reference population can influence the magnitude and relative ordering of exposure disparities. Thus, methodological choices can lead to different conclusions for the same concentration and population surfaces; this can impact not only the findings of an individual study but also have implications for mitigation strategies. We conclude with recommendations for best practices for making, justifying, and communicating these methodological decisions.