In addition to state and year fixed effects, we controlled for selected time-varying state economic and demographic variables that could influence both tobacco control policy and alcohol consumption rates. We considered state beer excise tax per barrel, wine and spirits excise taxes per gallon (obtained from W. Ponicki, Statewide Availability Data System II: 1933–2003, Pacifica Institute for Research and Evaluation), state per capita income (from the Bureau of Economic Analysis), annual unemployment rate (from the Bureau of Labor Statistics), proportion of population in rural areas (from the U.S. Census Bureau), percent of the population that was African American, Hispanic, or Other (not Caucasian) race (calculated from U.S. Census population estimates), age distribution (14–17 years, 30–49 years, 50–64 years, and 65 years or older, calculated from U.S. Census population estimates), and percent of population affiliated with primarily Judeo-Christian denominations, including Mainline Protestant, Evangelical Protestant, Catholic, Orthodox, and Other denominations (church membership data from the Association of Religion Data Archives). Beer excise tax and per capita income were adjusted for inflation to reflect 2012 dollars. For rural population and religious adherents, data were not available for all years; data for missing years were imputed via linear interpolation.
