Declines in air passenger volume lead to substantial excess deaths from auto accidents, as consumers substitute driving for less dangerous air travel. The variables are pretty hard to isolate and the linked paper examines only one unusually dramatic case, but this effect could have far-reaching implications; for example, if the paper's accurate in citing a 5-8% decline in American air travel linked to stricter and less convenient passenger security screening, then these screening procedures may be actively counterproductive.