Relationships between US state alcohol policies and alcohol outcomes: differences by gender and race/ethnicity.

Abstract

Alcohol taxation and availability restrictions are among the most effective methods for reducing alcohol use and problems, yet may affect demographic subgroups differently. Understanding who responds to specific policies can inform approaches for reducing disparities. We examined how state-level beverage-specific taxes and availability restrictions in the United States are associated with consumption and alcohol-related problems across subgroups defined by gender and race/ethnicity. Data came from the 2000-15 National Alcohol Surveys (n=28 251), computer-assisted telephone cross-sectional surveys of United States residents aged 18+. African Americans and Hispanics were oversampled. Primary outcomes were beverage-specific (beer, wine, spirits and total) volume, DSM-IV alcohol dependence and alcohol-related consequences. Analyses entailed survey-weighted log-log and logistic regressions adjusting for state-level beer tax, spirits tax, government-controlled spirits sales and sales tax; respondent ZIP-code-level density of off-premise beer outlets, off-premise spirits outlets and on-premise bars; respondent individual-level age, marital status, education, employment and income; and fixed effects for wet/moderate/dry US region and year. Higher beer tax was significantly (P<0.05) associated with lower odds of any drinking among white women [odds ratio (OR)=0.98] and lower beer volume (price-elasticity=-0.40), total volume (price-elasticity=-0.50) and odds of alcohol-related consequences (OR=0.84) among African American women. Higher spirits tax was significantly (P<0.05) associated with both lower beer and total volume among Hispanic women (price-elasticities=-0.73 and-1.04, respectively) and men (price-elasticities=-1.19 and-0.92, respectively) and decreased wine volume among Hispanic women (price-elasticity=-0.62). Apparent protective effects of living in a state with government-controlled spirits sales or a neighborhood with lower bar density was greater among white men than other groups. The effects of beverage-specific taxes and alcohol availability policies may vary across subgroups, highlighting the importance of considering differential policy impacts in future research and intervention.