Air pollution has been considered in the literature an important driving force underlying the long-lasting puzzle of the negative relationship between unemployment and mortality rates in the US. I use a newly developed satellite-based dataset on local pollutant concentrations and demographic characteristics at a county level in the US to explore the relationship between the two factors in different contexts and to identify the conditions under which this effect is present. I find evidence that the mechanism “Unemployment increases -> local pollution decreases -> mortality decreases” may partially explain the identified puzzle. Yet, as shown in this paper, it is only present when pollution levels and/or the decrease in the pollutants’ concentration exposure are close to or above the average levels. These results suggest that to better estimate the impact of economic recessions on health outcomes and to implement the appropriate environmental policies to reduce local populations’ air pollution exposure, a clear understanding of the underlying local conditions, spatially and over time, that may drive those interactions is required. (work in progress)
Supplementary notes can be added here, including code and math.