Protecting teens from the dangers of alcohol use and abuse:
wishful thinking versus science
Convened by Mothers Against Drunk Driving, American Medical Association, National Transportation Safety Board, and Insurance Institute for Highway Safety to support 21 minimum drinking age
Presentation by Adrian Lund, IIHS president
October 9, 2007 • Washington, DC
In 1972, at a conference on road safety in Canberra, Australia, William Haddon Jr., M.D., the first head of what is now the National Highway Traffic Safety Administration and President of IIHS from 1969-1985, talked about the beginning of a transition
"away from a pre-scientific period. That is, from a period in which folk culture has dominated — in which virtually everyone, both in and out of public life, has been a self-certified expert with his own pet, dogmatically advanced panacea — in which the notion has been virtually absent that public and private conclusions, pronouncements and measures to reduce the losses should be based on well-done, carefully scientific determinations of relevance and efficacy rather than on the unsubstantiated assertions of some individual or group."
We are here today because this transition to science-based approaches to reducing the deaths and injuries from motor vehicle crashes is not yet complete. Thirty-five years later, John M. McCardell, Jr. has mounted a campaign to reduce the drinking age from 21 to 18 in the United States. His justification — a desire to reduce the clandestine and sometimes biologically dangerous levels of alcohol consumption among 18-20 year-olds — is laudable. However, his reasoning about what works is quintessentially pre-scientific. Highway safety policies need to be grounded in solid research, not wishful thinking. His arguments are demonstrably wrong. My comments today are limited to his two central theses:
- that the benefits of the 21-year-old drinking age are unproven; and
- that alcohol education for teens promises to be more effective in dealing with the problem of teen alcohol use and abuse.
Both theses are contradicted by fact.
Teen crashes vary with drinking age laws
On his website, Mr. McCardell says, "Advocates of the 21-year-old drinking age have long argued that the decrease in fatalities was a result of the lowered drinking age but cannot offer a cause and effect relationship."
The truth is, the cause and effect are clear. lf we lower the drinking age, we will be killing more teens on the highway. Actions among the states in lowering, raising, lowering, and raising again the age at which it is legal to purchase alcohol have provided a series of opportunities to evaluate the effects of these changes on motor vehicle crashes.
IIHS's first study in 1974 looked at two states and one Canadian province that lowered the drinking age, carefully comparing their experience to that of adjacent states that did not change. That study showed that the number of 15-20 year-olds involved in fatal crashes increased in the jurisdictions that lowered the drinking age.
Subsequently, in the late 1970s, states began to increase drinking ages again. Again, it was possible to compare states that made this change to states that didn't. Again, we saw a change related to the drinking age — this time, fatal crash rates declined as teen drinking and teen drinking and driving declined.
IIHS has been a leader in studying the effect of drinking age, but it hasn't been alone.
Although there's variation the effects are consistent: deaths go up when the drinking age is lowered and they go down when it is raised. The 21-year-old drinking age is saving lives.
Drinking education won't counteract easier availability of alcohol
While ignoring the vast literature confirming the public health benefit of the 21-year-old minimum drinking age, Mr. McCardell asserts that drinking education could effectively supplant and improve upon 21-year-old drinking laws in combating the problem of alcohol among 18-20 year-olds. What's the evidence?
Mr. McCardell offers none. Nor is there much in the way of evidence about what effect drinking education might have on 18-20 year-olds. However, there is evidence about the effects of driver education, which offers some insights about how drinking education and a drinking license, as recommended by McCardell, might affect teens.
Again, IIHS has done much of the research on driver education, a fair task since, in years past, much of the funding for driver education in high schools came from insurers. However, when IIHS studied the effects of driver education carefully in the 1970s, a main finding was that teen crashes tended to be higher when high school driver education was available. In the late 1970s, this correlation was confirmed when Connecticut stopped state funding for high school driver education, and many schools in the state dropped the course. The result was fewer teen crashes, based on our study that compared those schools with schools that continued to fund driver education locally.
In response to criticism that these driver education courses were too simplistic, the US DOT spent millions to develop a model course. It was called the Safe Performance Curriculum and was submitted to a proper study in DeKalb County, Georgia. When compared to other high school students who received either no driver education or a more basic information course, the result was the same: SPC increased the number of teens getting licensed and the number involved in crashes. This is an unintended consequence of driver education — it can encourage earlier licensure that is not offset by any improvement in knowledge or skill.
Driver education can help drivers learn to operate vehicles and to understand why traffic laws are what they are. Driver education, however, is not itself an effective public health strategy. Drinking education will teach teens about alcohol, but it may only produce better educated drinking and driving teenagers while at the same time making our highways more dangerous. McCardell offers no scientific evidence to the contrary.
The scientific evidence is clear.
Others too would die in crashes involving drinking teenagers.
Experience with driver education suggests that drinking education wouldn't counteract this effect. In fact, one implication of driver education experience is that exposing students to drinking education could increase the number drinking. Receiving a license to drink could cause teens and some parents to conclude that the school thinks their teens will drink safely.
This is not the path to reducing the problem of teenage drinking — it is a proven formula for increasing the number of dead teens. Clandestine underage drinking is a problem, but lowering the drinking age is not a solution.