Saturday, July 15, 2006

Get the Kids Drunk

So, returning to my one man campaign to roll back Liddy Dole's assault on freedom I decided to take a look at teen drunk driving fatalities, as that's the primary reason used to justify the 21-year-old drinking age.

It is true that teen drunk driving fatalties have fallen significantly since the law was passed in 1984, but it's also the case that all drunk driving fatalities have fallen then due a combination of various factors including cultural change, education, enforcement, etc.

Since the law was passed (I'm not sure how quickly every state changed their laws), teen drunk driving fatalties fell from about 3600 in 1984 to 1536 in 2004 (.pdf) , about a 57% decline.

Over that same period all drunk driving fatalties fell from about 82,000 to 44,000, a 46% drop. (.pdf link)

So, yes, over that time period teen fatalties have decreased by a bit more than for that of the general population, but not by all that much (if I were an economist instead of a lazy blogger I'd normalize these things for population cohort size, and separate one set of numbers from the other, but such is the joy of being a lazy blogger). In other words, at a glance the law doesn't seemed to have improved things all that much over and above over what has been achieved through a general increase in societal hostility to drunk driving.

As a sort of compromise I'd propose the option for the under-21 crowd to choose between a drinking license and a driving license. You couldn't have both until you become 21. We could figure out if this was a one time choice irreversible choice or if switching were possible, and there are some other logistical issues, but it would seem to make sense.

Also it would help to achieve a better policy goal which is reducing the amount of teen driving. I spent way too much time looking for decent historical data and couldn't find good clean numbers, but from what I can tell total deaths involving young drivers declined during the 80s but have been pretty flat since.