This paper uses cross-state and cross-regional time series to estimate the effect of the variation in shares of alcohol beverage types on traffic fatalities in young people, with particular emphasis on consumption of beer versus spirits, or “hard liquor”. Depending on the specification, consumption shares matter for traffic fatalities in the 15-19 year age group, but results are not conclusive. Initial state median income is however a strong predictor of lower traffic mortality rates in ensuing years.
