Methods: Studies were identified by systematic database searches and screening references of reviews and relevant studies. Google and web-pages of tobacco control agencies were also searched. The review identified 56 studies using absolute sales, sales ratio or employment data and employing regression methods to evaluate the impact of smoking bans in the United States, Australia or in countries in South America or Europe. The meta-analysis included 39 comparable studies, with 129 cases identified based on the outcome measure, scope of the ban, type of establishment and geographical location. Methodological quality was assessed based on four pre-determined criteria. Study and case selection and data extraction were conducted independently by two researchers.

Results: Random-effects meta-analysis of all cases showed no associations between smoking bans and changes in absolute sales or employment. An increase in the share of bar and restaurant sector sales in total retail sales was associated with smoking bans [0.23 percentage-points; 95% confidence interval (CI) 0.08–0.375]. When cases were separated by business type (bars or restaurants or wider hospitality including bars and restaurants), some differential impacts emerged.

Conclusions: Meta-analysis of the economic impact of smoking bans in hospitality sector showed overall no substantial economic gains or losses. Differential impacts were observed across individual business types and outcome variable, but at aggregate level these appear to balance out.