Smoking behavior in 18-20 year-olds after tobacco 21 policy implementation in California: A difference-in-differences analysis with other states.

In June 2016, California implemented a Tobacco 21 (T21) policy that increased the minimum sale age of tobacco products from 18 to 21. This study examined the association between California's T21 policy and smoking behavior (ever, current, daily, and nondaily) in 18-20 year-olds using data from the 2012-2019 Behavioral Risk Factor Surveillance System (n = 15,863). The annual change in odds of smoking among 18-20 year-olds post-policy (July 2016-December 2019) was compared with the pre-policy period (January 2012 - June 2016) 1) within California and 2) compared with states without a T21 policy. As a sensitivity analysis, 21-23 year-olds in California were used as the referent. Difference-in-difference estimates (D-I-D) were calculated using adjusted logistic regression and compared the post to pre-policy change in trends in California to the referent groups. Before California's T21 policy, there was an 11% annual decrease in the odds of ever smoking among 18-20 year-olds in California and a 6% decrease in the referent states. After the policy, these trends did not change significantly. Results for current smoking were similar. For daily smoking, there was an 8% annual decrease before the policy and a 26% annual decrease after the policy among 18-20 year-olds in California; D-I-D estimates were 0.80 (95% CI: 0.57, 1.14) using referent states as the comparison and 0.62 (95% CI: 0.41, 0.95) using 21-23 year-olds in California as the comparison. There was an association between California's T21 policy and a decrease in daily smoking among 18-20 year-olds, compared with 21-23 year-olds, more than three years post-implementation.