Death and taxes. An opportunity for substitution.

The risks of fatal vehicle crash for young drivers rise precipitously with more drinking, reaching a 100-fold increase in risk for youths with 6+ drinks. Considerable external damage arises from drunk driving, providing firm logic for government intervention. If the 'cost' of an alcohol tax is the foregone consumer surplus of drinkers, then the cost per life saved using higher alcohol taxes is less than +10,000 per life at low taxes, even ignoring the lives of the drinking drivers. Thus, alcohol taxes are one of the cheapest possible life saving techniques available. Preliminary evidence suggests that youthful drinking drivers badly under-perceive the risks of their own behavior. This increases the measure of 'external damage' and hence the optimal tax for alcoholic beverages. It may be impossible to administer an 'optimal tax' on alcohol, but 'second-best' taxes appear to lie in the range of 25 to 40 percent or more.