drinking age: 18 vs. 21

On ethical grounds, the choice of drinking age is largely arbitrary. The matter, then, should rest on practical grounds and the data...

Again, a hat tip to Jacob Sullum in this month's Reason...

The primary argument for 18 (vs. 21) is saving lives. A recent NBER's study by Harvard's Jeffrey Miron and Yale's Elina Tetelbaum find little improvement in the statistics, resulting from the changes in the laws. (This is much easier to study since states passed laws about this over a period of years.)

M&T find a 8% drop in fatalities for 18-20 year olds. And interestingly, most of the drop was in states that voluntarily reduced their drinking ages-- before the federal government started to twist the arms of the remaining states, by threatening to withhold federal funding for roads. (As an aside, this is further evidence of federalism's effectiveness. States that "needed" to do this the most, had already done so.)

Of course, auto fatalities have decreased quite a bit since the late 1960s-- despite higher speed limits re-introduced in the 1990s. Almost all of these gains can be credited to technological advance in auto safety and medical care.

One other interesting finding: a lower drinking age has led to a modest increase (3-4%) in drinking and "heavy drinking".


And then, on a related matter, from one of the lead articles in this month's Reason, David Harsanyi has an article on the same topic which is based on his recent book, Nanny State: How Food Fascists, Teetotaling Do-Gooders, Priggish Moralists, and other Boneheaded Bureaucrats are Turning America into a Nation of Children. (What a long but entertaining title!)

Harsanyi cites a study by Donald Freeman in Contemporary Economic Policy on the negligible effects of dropping Blood/Alcohol Content (BAC) standards from .10% to .08%. Whether the states adopted the new standard on their own or were coerced into it by the Federal government, "there's no evidence that lowering the BAC limits...reduced fatality ates, either in total or in crashes likely to be alcohol-related."

According to Harsanyi, in 1982, 60% of traffic-related fatalities were related to alcohol and this number fell steadily until 1997 when it reached about 40% (and has stayed steady since). He credits the work of MADD and others efforts to educate the public and stigmatize drunk driving. But the new, tighter restrictions were of no evident, practical use.