“Deaths of despair” over the business cycle: New estimates from a shift-share instrumental variables approach

Abstract

This study presents new evidence of the effects of short-term economic fluctuations on suicide, fatal drug overdose, and alcohol-related mortality among working-age adults in the United States from 2003–2017. Using a shift-share instrumental variables approach, I find that a one percentage point increase in the aggregate employment rate decreases current-year non-drug suicides by 1.7 percent. These protective effects are concentrated among working-age men and likely reflect a combination of individual labor market experiences as well as the indirect effects of local economic growth. I find no consistent evidence that short-term business cycle changes affect drug or alcohol-related mortality. While the estimated protective effects are small relative to secular increases in suicide in recent decades, these findings are suggestive of important, short-term economic factors affecting specific causes of death and should be considered alongside the longer-term and multifaceted social, economic, and cultural determinants of America’s “despair” epidemic.