Today Jason Kottke linked to an article in Reason Magazine advocating a reduction or elimination of the federal minimum drinking age. He pulled this quote from it:

The age at highest risk for an alcohol-related auto fatality is 21, followed by 22 and 23, an indication that delaying first exposure to alcohol until young adults are away from home may not be the best way to introduce them to drink.

Is it just me, or is this an extremely odd argument? I read it over, trying to figure out how even the author intended it to be interpreted. Finally I guessed that they meant this statistic indicates that after being prohibited from it for too long, people turning 21 binge irresponsibly for three years. But much of the rest of the essay complains that people under 21 drink more irresponsibly because it's illegal, drinking "furtively and dangerously" rather than in social settings. In any case, to me the far more obvious reading of the statistic is that people of these ages cannot handle drinking and driving, that is, without doing them at the same time.

Now I am hardly advocating that the drinking age be increased to 24. But imagine if it were reduced to 18, and then you heard this statistic with the most deadly ages being 18, 19, and 20. Would you take it as a sign that the age should be further reduced to 15?