The authors considered unemployment, an often overlooked covariance of highway fatalities, hypothesizing that (1) as unemployment rises, aggregate driving decreases, especially among the unemployed, and as driving decreases, fatalities should decrease; (2) unemployment may influence drinking - some among the unemployed may drink less due to lower incomes, while others may drink more due to stress so the net effect would be ambiguous; and (3) unemployment may increase aggregate levels of stress and unhappiness, which can result in poor concentration on driving and thus, in turn, should result in more accidents and deaths. The authors used data from fifty states and the District of Columbia from 1976-1980, representing 255 observations. Using econometric models of the data, the authors present evidence for two of the three hypothesized effects of unemployment.
Abstract