Tempest in a bottle

"That is a really important measure of success," says Simon. "Back in the 1970s when states started lowering the drinking age to 18, that's when this experiment began. There were increases in traffic fatalities, and people said let's go back to the way it was. We forget there is so much science and historical context here. We have been down this road before."

* Among those studies comparing the years before 1984 with the current era was a 2001 report from the National Institute on Drug Abuse, which found that college students who reported drinking in the last month fell from 82% in 1980 to 67% in 2000.

* In 2007, the University of Michigan's annual Monitoring the Future survey found that annual alcohol use by high school seniors has dropped from 77% in 1991 to 66% last year.

Perhaps the strongest evidence for the harmful health effects of drinking at a young age come from studies on biology and addiction, Foster says.

* A 2002 report from the American Medical Assn., citing numerous studies, concluded that alcohol use during adolescence and young adulthood causes damage to memory and learning capabilities.

* A study in the 2006 Archives of Pediatrics & Adolescent Medicine found that teens who began drinking before age 14 had a lifetime risk of alcohol dependence of 47% compared with 9% for those who began drinking at 21. For each additional year under age 21 of drinking, the greater the odds he or she would develop alcohol dependence. Though the cause of this correlation is unknown, some experts believe pure biology -- priming the young brain to need alcohol -- is involved.

"This is a public health problem and a medical problem," Foster says. "It's about the national failure to recognize addiction as a disease. If we think of it as kids behaving badly or breaking the rules, that gets in the way."

* And in a 2002 analysis of 33 high-quality studies on the age-21 drinking law's effects, University of Minnesota researcher Traci L. Toomey found that all but one study showed the higher age resulted in lower rates of alcohol consumption and traffic crashes.

"It is the most well-studied alcohol control policy we have in this country," says Toomey, an associate professor in the school of public health. "Usually we find no effect when we do policy studies. Here we have this policy effect that is very consistent -- a big chunk of the studies showing this inverse relationship."

Evidence for change