Another Study That Does Not Show Legalizing Pot Boosts Traffic Fatalities
According to a University of Colorado at Denver press release issued last week, a study published last month by the journal Drug and Alcohol Dependence found that "the proportion of marijuana-positive drivers involved in fatal motor vehicle crashes in Colorado has increased dramatically since the commercialization of medical marijuana in the middle of 2009." Or as the headline over the press release puts it, "Marijuana use [has been] involved in more fatal accidents since commercialization of medical marijuana." The implication is that easier availability of marijuana in Colorado has led to an increase in traffic fatalities. But as with a similar analysis of data from six states that was published by the American Journal of Epidemiology in January, that is not what the study shows.
Using data from the federal government's Fatality Analysis Reporting System, pharmacologist Stacy Salomonsen-Sautel and her co-authors found that the proportion of fatal crashes involving "marijuana-positive drivers" was 4.5 percent in the first six months of 1994, 5.9 percent in the first six months of 2009, and 10 percent at the end of 2011. The upward trend accelerated after Colorado regulators loosened restrictions on medical marijuana in July 2009, and there was no similar increase in the 34 states that at the time did not have medical marijuana laws. Meanwhile, the proportion of fatal accidents in which drivers tested positive for alcohol remained about the same.
Do these data mean that legalizing marijuana for medical or recreational use results in more blood on the highways? No. What Salomonsen-Sautel et al. call "marijuana-positive drivers" actually tested positive for metabolites that linger in blood and urine long after the drug's effects wear off. "THC metabolites are detectable in an individual's blood or urine for several days and sometimes weeks for heavy marijuana users," the authors note toward the end of the article. Hence a "marijuana-positive" result does not indicate the driver was under the influence of marijuana at the time of the accident, let alone that marijuana was a factor in the crash. "This study cannot determine cause and effect relationships, such as whether marijuana-positive drivers contributed to or caused the fatal motor vehicle crashes," Salomonsen-Sautel et al. concede. "Colorado may have an increased number of drivers, in general, who were using marijuana, not just an increase in the proportion who were involved in fatal motor vehicle crashes….The primary result of this study may simply reflect a general increase in marijuana use during this same time period in Colorado."
Another reason to doubt that greater tolerance of marijuana boosts traffic deaths: "There was a decreasing trend in fatal motor vehicle crashes in Colorado since 2004." There was a similar decline in the 34 comparison states, so it does not look like readier access to marijuana has interfered with this welcome trend. In fact, there is some evidence that it has on balance reduced traffic fatalities by encouraging the substitution of marijuana for alcohol, which has a more dramatic effect on driving ability.
A recent working paper from the National Bureau of Economic Research casts some doubt on that hypothesis, finding that medical marijuana laws are associated with a 6-to-9-percent increase in the frequency of binge drinking among residents 21 or older. It is too early to say whether legalizing marijuana for recreational use will have a measurable impact on accident trends in Colorado or Washington, both of which eliminated penalties for possession at the end of 2012. Colorado also legalized home cultivation, and state-licensed pot shops began opening there in January. Washington does not allow homegrown marijuana for recreational use, and it does not have state-licensed stores yet. For what it's worth, fatal crashes in Colorado, after rising from 2011 to 2012, fell slightly (from 434 to 428) between 2012 and 2013. In Washington fatal crashes rose slightly (from 403 to 405) between 2012 and 2013.
[Thanks to Felix for the tip.]
