Abstract
Introduction
States and municipalities are increasingly restricting tobacco sales to those under age 21, in an effort to reduce youth and young adult smoking. However, the effectiveness of such policies remains unclear, particularly when implemented locally.
Methods
Analyses use 2011–2016 data from the Behavioral Risk Factor Surveillance System’s Selected Metropolitan/Micropolitan Area Risk Trends dataset. Difference-in-differences and triple-difference regressions estimate the relationship between local tobacco-21 policies and smoking among 18- to 20-year-olds living in MMSAs (metropolitan/micropolitan statistical areas and metropolitan divisions).
Results
Current smoking rates fell from 16.5% in 2011 to 8.9% in 2016 among 18- 20-year-olds in these data. Regressions indicate that a tobacco-21 policy covering one’s entire MMSA yields an approximately 3.1 percentage point reduction in 18- to 20-year-olds’ likelihoods of smoking (confidence interval [CI] = –0.0548 to –0.0063). Accounting for partial policy exposure—tobacco-21 laws implemented in some but not all jurisdictions within an MMSA—this estimate implies that the average exposed 18- to 20-year-old experienced a 1.2 percentage point drop in their likelihood of being a smoker at interview relative to unexposed respondents of the same age, all else equal.
Conclusions
Local tobacco-21 policies yield a substantive reduction in smoking among 18- to 20-year-olds living in MMSAs. This finding provides empirical support for efforts to raise the tobacco purchasing age to 21 as a means to reduce young adult smoking. Moreover, it suggests that state laws preempting local tobacco-21 policies may impede public health.
Implications
Although states and municipalities are increasingly restricting tobacco sales to under 21-year-olds, such policies’ effectiveness remains unclear, particularly when implemented locally. Using quasi-experimental methods, this article provides what may be the first evidence that sub-state tobacco-21 laws reduce smoking among 18- to 20-year-olds. Specifically, considering metropolitan and micropolitan areas from 2011 to 2016, the average 18- to 20-year-old who was exposed to these policies exhibited a 1.2 percentage point drop in their likelihood of being a current established smoker, relative to those who were unexposed. These findings validate local tobacco-21 laws as a means to reduce young adult smoking.
The Triple Difference Estimator
