The Effects of State-Level Legislative Smoking Bans in Private and Government Workplaces on Smoking Prevalence

Abstract

This paper examines the effects of state-level legislative smoking bans in private and government workplaces on smoking prevalence. Using data from the Behavioral Risk Factor Surveillance System and other sources, a difference-in-difference method is used. The main findings are that private and government workplace smoking bans are not consistently found to have an effect on whether a person is a smoker. However, when regressions are run on subsets of the population, or when the definition of smoker is changed to only include more intensive smokers, the coefficients for private bans, government bans, natural logarithm of real cigarette tax and bar bans are sometimes statistically significant. None of the effects are of large economic significance, and some have unexpected signs.