HOW can government best discourage drunk driving?
Raise beer taxes. And suspend or revoke a driver's license for a year by mandatory administrative action when caught driving with a specific blood alcohol content.
These are the two most effective means of reducing drunk driving in the United States, according to a study by three economists, Frank Chaloupka of the University of Illinois at Chicago, Michael Grossman of the City University of New York, and Henry Saffer of King College in Union, N.J.
For more than a decade, federal, state, and local governments have campaigned to reduce drunk-driving accidents. The campaign is working. The National Highway Traffic Safety Administration estimates that 21,000 people died from alcohol-related traffic accidents last year. But the NHTSA also figures some 7,500 lives were saved through alcohol enforcement efforts.
Such a reduction is of considerable economic importance. Traffic fatalities and injuries of all kinds cost the nation about $100 billion a year in medical costs, insurance, vehicle damage, etc. This sum is equal to one-third of the defense budget.
Alcohol is involved in almost half of fatal traffic accidents, which killed 42,350 people in 1991, the NHTSA calculates. That 1991 projection was down 7 percent from the actual number in 1990. It was the lowest in total in almost three decades and the lowest fatality rate per 100 million vehicle miles for all time.
The anti-drunk-driving campaign "has been successful by any measure," says Michael Brownlee, who is in charge of NHTSA's traffic safety programs.
One part of the campaign was a 1983 law providing financial incentives for states to enact and enforce new, more stringent drunk-driving laws. Then a 1984 law withheld federal highway funds from states that did not raise the minimum legal drinking age to 21.
As a result of these bills, over 500 new drunk-driving laws have been passed in the country. The study by the three professors is the first comprehensive examination of the effectiveness of these laws. It was sponsored by the National Institute on Alcohol Abuse and Alcoholism (NIAAA). An outline has just been published by the National Bureau of Economic Research (NBER).
At the federal level, Congress doubled the beer tax from 16 cents to 32 cents a six-pack as of the start of 1991. As part of the budget reduction package passed in the fall of 1990, the beer tax change was more of a revenue-raiser than an anti-alcohol measure. Nonetheless, combined with recession and the Gulf war, it did cut beer consumption last year by 2.2 to 2.4 percent, according to industry publications.
If the tax had been quadrupled instead and thus raised to its inflation-adjusted value in 1951, it would have cut the number of drunk-driving fatalities by 11.5 percent, the NBER study says.
Statistics on total alcohol consumption for 1991 are not yet available. However, the NIAAA has just released a report indicating that the annual per capita consumption of alcohol contained in beer, wine, and spirits by those 14 and older declined in 1989 to 2.43 gallons from 2.49 gallons the year before. Alcohol consumption peaked in 1980 and 1981 at 2.76 gallons.
If a state suspends or revokes drunk drivers' licenses for a year, independent of later court action, it shrinks drunk-driving auto deaths by 9 percent a year, the NBER study finds. At present, 29 states have laws imposing such administrative action. Mr. Brownlee says such laws "should be across the board."
The next most effective policies (5 or 6 percent each) are a 21-year-old legal drinking age, laws permitting police to administer a preliminary breath test for alcohol without the assistance of medical personnel, dram shop laws that allow a person injured by a drunk to bring suit against the person or establishment serving the alcoholic beverages, and relatively large mandatory fines on drivers. Laws that do not permit those charged with driving under the influence to plea bargain and mandatory license s anctions upon conviction have some deterrent effect.
However, mandatory jail sentences and community service options and "illegal per se" laws requiring conviction of drivers with specific alcohol blood content at a trial have no deterrent effect. Apparently such penalties are not quick or certain enough.
Brownlee hopes the NBER study will help cut fatalities further.