Pot luck: marijuana and driver fatalities.

OBSERVING THE CARNAGE CAUSED by drunk drivers, opponents of drug legalization worry that making marijuana easier to obtain would multiply the number of traffic fatalities. A study released this year by the National Highway Traffic Safety Administration suggests that such fears are overblown.

The study examined 1,882 accidents in seven states, all of which resulted in the driver's death. In each case, blood drawn from the driver was tested for traces of 24 drugs. Based on official records, a researcher who did not know the results of the drug test independently assessed the driver's responsibility for the accident.

Drivers under the influence of alcohol alone (about 40 percent of the sample) were significantly more likely than drug-free drivers to be deemed responsible for their accidents. But drivers under the influence of marijuana alone (about 1 percent of the sample) were no more likely than drug-free drivers to be deemed responsible. "There was no indication that cannabis by itself was a cause of fatal crashes," the report says.

This result does not prove that driving under the influence of marijuana is safe. But it does indicate that the drug's current role in fatal accidents is minor, and it jibes with laboratory research finding that marijuana impairs motor coordination less than alcohol does.

Marijuana legalization probably would mean more stoned drivers. The number of users would increase, and current users would be more likely to smoke pot away from home. But given the evidence that people use alcohol less when they use pot more (see Trends, July 1993), marijuana legalization could also mean fewer drunk drivers. If drunk drivers are a lot more likely to cause accidents--a conclusion supported by the NHTSA study--the number of traffic fatalities might stay about the same, or even decline.