Abstract

Many communities and several states prohibit smoking in bars or restaurants. Using county-level data on employment from across the US, we find that communities where smoking is banned experience reductions in bar employment compared with counties that allow smoking. Smoking bans have a larger detrimental impact on bars in geographic areas with a high prevalence of smokers. The relative effect on restaurant employment is neutral or mildly positive. The positive effects are concentrated in areas with fewer smokers. We also find that bans have a positive effect on restaurant employment in warmer regions of the country, especially during the cooler winter months, and in the summer in colder regions. This suggests the prevalence of outdoor seating might influence the policy's effect.