Unemployment insurance and cigarette smoking.

We provide evidence for the causal relationship between unemployment insurance (UI) and individuals' smoking behavior using the 1995-2011 Current Population Survey-Tobacco Use Supplement data. Our identification relies on the exploitation of the exogenous variations of the maximum UI weekly benefits across states and over years. Instead of focusing on all unemployed people, we concentrate on those who are eligible for UI benefits. We find that when the maximum UI weekly benefit level increases by $100, smoking cessation increases by approximately 2.9 percentage points among the UI-eligible unemployed. The results are robust to various model specifications. We also explore the mechanism and find suggestive evidence that the increase in the maximum UI weekly benefit leads to a decrease in the probability of over-work of the respective spouse. Moreover, the unemployed who are less educated are more responsive to the increasing UI benefits.