This paper uses a state-level alcohol sales dataset to analyze how marijuana legalization affects alcohol consumption. I employ a difference-in-differences model to investigate the relationship between alcohol and marijuana in the short-term and long-term. In addition, marijuana legalization effects are estimated for different alcohol types – beer, wine and spirits. Overall, the results indicate a negative, yet insignificant relationship between the two narcotic drugs. There is not enough evidence to firmly conclude substitutability or complementarity of the two goods, therefore, leaving the debate unsolved.
