More information about text formats
¶ I enjoyed reading this paper. I appreciate the author's use of difference-in-difference (DD) methodology. There were some things I found unclear that I would like to ask the authors to comment on.
¶ First, could the authors provide greater clarity on the model for column 1 of Table 1? Is the dependent variable here a yes/no for current cigarette use? The authors write, "Adolescents reported lifetime and prior month use of cigarettes, which we combined into a count variable of days smoked in the past month (0–30)." How does lifetime cigarette use help the authors to code the current number of cigarette days? The authors later state that they show that "increasing implementation of flavoured tobacco product restrictions was associated not with a reduction in the likelihood of cigarette use, but with a decrease in the level of cigarette use among users." Do the authors mean lifetime cigarette use here, or current cigarette use? The authors estimate this equation with an "inflation model," which I am not aware of. Could the authors provide more information on this modelling technique? This is not discussed in the "Analysis" section.
¶ Second, I felt like this statement is too strong. "Our findings suggest that[...] municipalities should enact stricter tobacco-control policies when not pre-empted by state law." Municipalities need to weigh many factors in making these decisions, including the effects of popu...
¶ Second, I felt like this statement is too strong. "Our findings suggest that[...] municipalities should enact stricter tobacco-control policies when not pre-empted by state law." Municipalities need to weigh many factors in making these decisions, including the effects of population health (not just to youth tobacco use). This study provides evidence from a single state that may not be generalizable to other states without preemption policies. Other studies have found unintended negative effects of flavor policies, and these studies should be referenced to balance the discussion section.
¶ I applaud the authors for providing an early data point on the effect of these policies, but certainly more work in this space is needed before policy recommendations can be made. The authors may also wish to consider for future difference-in-difference papers whether there is evidence in support of the parallel trends assumption , which is a crucial assumption underpinning the reliability of the model.
¶ Friedman, Abigail S. "A difference-in-differences analysis of youth smoking and a ban on sales of flavored tobacco products in San Francisco, California." JAMA pediatrics 175, no. 8 (2021): 863-865.
¶ Xu, Yingying, Lanxin Jiang, Shivaani Prakash, and Tengjiao Chen. "The Impact of Banning Electronic Nicotine Delivery Systems on Combustible Cigarette Sales: Evidence From US State-Level Policies." Value in Health (2022).