This paper estimates linear probability models for drinking prevalence and binge drinking by youth, young adults, and adults by using state-level estimates for 1999–2003 from the National Survey on Drug Use and Health. The regression models contain explanatory variables for economic influences (alcohol taxes, outlet density, income), demographics, and regulatory variables. The main results are, first, a positive relationship exists among youth and adult alcohol behaviors. Second, state-to-state variation in real beer taxes does not negatively affect youth behaviors. Third, higher outlet densities positively affect behaviors by young adults and adults, but do not affect youth. Fourth, several regulatory variables have a negative effect on drinking prevalence and bingeing by youth and young adults, including state liquor monopolies, Sunday closing laws, and 0.08 BAC laws for drunk driving. Fifth, attendance at sports events does not increase drinking prevalence or bingeing.
Read full abstract