Association for Public Policy Analysis & Management

Use of electronic cigarettes (e-cigarettes), particularly among youth, has been rising exponentially. Within a four-year period (2011-15), e-cigarette use increased from 1.5 to 16 percent among high school students and from 0.6 to 5.3 percent among middle school students. This has raised concerns regarding possible negative health effects over the longer term from using e-cigarettes, as well as concerns that e-cigarettes may be a gateway for youth use of other addictive substances, particularly cigarettes. In response, state governments have launched a wave of regulations limiting youth access to e-cigarettes, the most common being the Minimum Legal Sale Age (MLSA) law for e-cigarettes. New Jersey became the first state to ban e-cigarettes sales to minors in 2010 and all but two states had a comprehensive ban in place at the time of a federal MLSA mandate in 2016. Very little is known about the workings of these regulations, and whether, by restricting e-cigarette purchases among minors, they have any spillover effects on smoking and the use of the other substances. The present study provides some of the first evidence on these issues.

We use individual-level data from the national and state Youth Risk Behavior Surveillance System (YRBSS) to examine the effects of e-cigarette MLSA laws on youth smoking, drinking, and marijuana use. Standard economic theory suggests that enhanced non-pecuniary costs associated with the restrictions could induce substitution away from e-cigarettes and toward other substances, making them economic substitutes. However, complementarity is also possible, and there may be heterogeneity in this relationship depending on the substance (for instance, between alternate products that both deliver nicotine, or between nicotine products and alcohol or marijuana). Furthermore, the addictive nature of the substances may also lead to important inter-temporal effects that may differ from the contemporaneous effects.

We exploit variation in the timing of state adoption of the laws for affected age groups, under the difference-in-differences and event study framework, controlling for relevant policy and macroeconomic factors as well as individual characteristics, state and year fixed effects and proxies for unobserved state trends. Preliminary results suggest that the passage of the law increases youth smoking participation by 1 to 1.4 percentage points, half of which is attributable to smoking initiation. An relevant question is whether this increase in smoking persists beyond when youth have aged out of the purchase restriction. We therefore assess the inter-temporal link between exposure to the law when under-age and smoking when aged out of the law, and find little evidence of persistence and such inter-temporal substitution. Our initial results also show little effect of the law on youth drinking or binge drinking but do provide suggestive evidence that e-cigarettes and marijuana are economic complements. Taken together, our findings suggest a possible unintended effect of e-cigarette MLSA laws — raising cigarette use in the short term while youth are unable to purchase e-cigarettes. We discuss some welfare implications of these effects and how they may inform efforts of other regulations aimed at restricting e-cigarette uses.