Attorney General Jeff Sessions claimed that more car accidents were “caused” by drugs than alcohol for the first time in 2016. But the report his office cited as support didn’t show that.
The report shows alcohol — not drugs — was found in the system of more drivers killed in car crashes. It also doesn’t say the fatal accidents were “caused” by drugs. It’s more difficult to prove a person is under the influence of drugs than alcohol while driving. Unlike alcohol, testing positive for a drug — marijuana in particular — doesn’t prove intoxication. Marijuana can be detected days or even weeks after consumption.
Sessions made his claim on June 22 during a private Q&A with interns at the Department of Justice. ABC News made a video of the event public on Dec. 7, after obtaining it through a Freedom of Information Act request.
In response to a question about marijuana legalization, Sessions said there’s “this view that marijuana is harmless,” adding that car accidents were “caused more by drugs than by alcohol” in 2016.
Questioner, June 22: Since guns kill more people than marijuana, why lax laws on one and harsh laws on the other?
Sessions: Well that’s apples and oranges question I would say first and foremost. But the Second Amendment, you’re aware of that, guarantees the right of the American people to keep and bear arms and I intend to defend that Second Amendment. It’s as valid as the First Amendment. So that’s my basic philosophical view about it. Look, there’s this view that marijuana is harmless and it does no damage. I believe last year was the first year that automobile accidents that occurred were found to have been caused more by drugs than by alcohol. Marijuana is not a healthy substance, in my opinion. The American Medical Association is crystal clear on that. Do you believe that?
Questioner: I, I don’t.
The American Medical Association has publicly stated that “cannabis is a dangerous drug and as such is a public health concern.” So there is support for Sessions’ claim that the AMA says marijuana is “not a healthy substance.”
When we contacted the Department of Justice for support for Sessions’ claim about car accidents, spokeswoman Lauren Ehrsam sent us a link to a Washington Post article from April 26 that outlines the findings of a report by two nonprofits — the Governors Highway Safety Association and the Foundation for Advancing Alcohol Responsibility.
The association represents state and territorial highway safety offices that implement federal grants to improve traffic safety. The foundation, funded by eight major alcohol companies including Bacardi U.S.A. Inc. and Beam Suntory, says its mission is to “fight against drunk driving and underage drinking.”
The report actually looks at data from 2015, not 2016, as Sessions said. The data were compiled from the National Highway Traffic Safety Administration’s Fatality Analysis Reporting System. The NHTSA is part of the U.S. Department of Transportation.
The report does say that NHTSA data from 2015 show “drugs were present in 43% of the fatally-injured drivers with a known test result, more frequently than alcohol was present.” For alcohol, the report cites the figure 37.3 percent. The report doesn’t provide the actual quantities of people who tested positive for drugs or alcohol, however.
And the report doesn’t quite support the idea that drugs led to more fatal car crashes than alcohol.
The report says 70.9 percent of fatally-injured drivers were tested for alcohol. Of that 70.9 percent, 37.3 percent tested positive for some level of alcohol. That means 26.4 percent of all fatally-injured drivers tested positive for alcohol.
Only 57 percent of fatally-injured drivers were tested for drugs. Of that 57 percent, 41.7 percent tested positive for some drug. That would mean 23.8 percent of all fatally-injured drivers tested positive for a drug, which is less than those who tested positive for alcohol.
(The figure of 43 percent noted a few paragraphs up comes from dividing 41.7 percent by 97.1 percent, the proportion of fatally-injured drivers with a “known” test result, Governors Highway Safety Association spokeswoman Madison Forker told us by email.)
But Sessions was answering a question about marijuana, not drugs in general. The report’s figure for marijuana is even lower.
Of the 57 percent of fatally-injured drivers who were tested for drugs, 34.3 percent had a positive test for a drug specifically coded for in the NHTSA’s system. An additional 7.4 percent tested positive for a drug in the system, which comes to 41.7 percent who tested positive for some drug. Of that 34.3 percent, 36.5 percent tested positive for marijuana.
That would mean 7.1 percent of all fatally-injured drivers tested positive for marijuana, compared with 26.4 percent who tested positive for alcohol.
(The report actually includes the figures 36.5 percent and 35.6 percent for marijuana, but Forker confirmed that the latter figure was a typo.)
But here’s the thing about drug tests for marijuana — unlike alcohol breathalyzers, they don’t prove a driver was intoxicated at the time of the accident.
As we wrote back in August 2016, marijuana can remain in the body for days, even weeks, after a person consumes the drug. The same goes for some other drugs, such as prescription pain killers. In fact, the report referenced by Sessions and provided to us by the DOJ cites this fact, stating that “marijuana metabolites can be detected in the body for weeks after use.”
“Alcohol is far simpler because it is quickly absorbed into the body and impairment is directly related to BAC [blood alcohol concentration],” the report by the Governors Highway Safety Association and the Foundation for Advancing Alcohol Responsibility says.
This is why experts don’t say marijuana, and some other drugs, caused traffic fatalities. Instead, they say marijuana use was associated with traffic fatalities. Sessions used the term “caused,” and that’s misleading.
The report cited by the DOJ adds that the data it used have other “critical limitations.” For example, the data come from individual states, and, “States vary considerably in how many and which drivers are tested, what tests are used, and how test results are reported.”
For all of these reasons, Sessions’ claim that drugs “caused” more car accidents than alcohol was inaccurate. The report DOJ cited doesn’t support that.
What about other research on the topic?
The National Academies of Sciences, Engineering and Medicine concluded in a January 2017 report: “There is substantial evidence of a statistical association between cannabis use and increased risk of motor vehicle crashes” (our italics).
The report emphasized that this is not necessarily marijuana use at the time of the accident, but use generally. The increased risk also isn’t necessarily large. The report didn’t consider whether marijuana was associated with more crashes than alcohol.
One meta-analysis published in the journal Addiction in April 2016, which weighed heavily on the report, found that marijuana use increased crash risk by 1.04 to 1.18 times, after controlling for alcohol use. This analysis took into account studies that had been conducted both within and outside of the United States.
Conversely, a February 2015 study by the National Highway Traffic Safety Administration found that a statistically significant crash risk for drivers who tested positive for THC disappeared after the study authors controlled for a number of variables, including alcohol concentration level. The academies’ 2017 report didn’t consider this study.
For comparison, the NHTSA says a blood alcohol level of 0.07, which is below the legal limit of 0.08, increases crash risk by 3.22 times. That would be the equivalent of two drinks for a 140-pound woman or three drinks for a 160-pound man.
There’s also some evidence that suggests medical marijuana laws may be associated with decreasing traffic fatalities.
One study published in The Journal of Law & Economics in 2013 reviewed traffic fatalities in the 19 states that had passed medical marijuana laws by 2010 and found that “legalization is associated with an 8–11 percent decrease in traffic fatalities” for the year after the laws took effect.
Another study, published in the American Journal of Public Health in July found: “Three years after recreational marijuana legalization, changes in motor vehicle crash fatality rates for Washington and Colorado were not statistically different from those in similar states without recreational marijuana legalization.”
Still, there is evidence that “marijuana significantly impairs judgment, motor coordination, and reaction time,” according to the National Institute on Drug Abuse.
Overall, experts disagree about the relationship between marijuana and driving. But there’s little to no disagreement over whether drinking alcohol before driving increases crash risk. And at this point, there isn’t sufficient evidence to conclude more drivers are dying in car crashes because of drugs than alcohol.
Clarification, Dec. 26: We changed one word in this sentence: “The report shows alcohol — not drugs — was present in the system of more drivers killed in car crashes.” It now reads, “The report shows alcohol — not drugs — was found in the system of more drivers killed in car crashes.”
Update, Jan. 18: The National Academies of Sciences, Engineering and Medicine released a report on Jan. 17 that says, “Despite widespread public attention to distracted and illegal drug-impaired driving in recent years, alcohol-impair driving remains the deadliest and costliest danger on U.S. roads today.” However, the report does acknowledge that “available data on driving fatalities due to impairment from drug use and impairment from co-use of drugs and alcohol is limited.”
Editor’s Note: SciCheck is made possible by a grant from the Stanton Foundation.