There’s a hot story today about some new research by the American Automobile Association on marijuana in fatal accidents. As usual, media coverage is misleading.
The best example of this media abuse comes from Tom Costello and The Today Show, with a headline asserting marijuana is causing a spike in fatal accidents.
The AAA study (pdf) absolutely does not say that. From page 16:
Drivers who had detectable THC in their blood at the time of the crash were not necessarily experiencing impairment in their ability to drive safely, nor were they necessarily at fault for the crash. Determination of actual impairment or fault status was beyond the scope of the study. Relatedly, many of those who were positive for THC were also positive for alcohol and/or other drugs, which in some cases likely contributed more significantly to the crash than did the THC. Research on the relationship between THC presence and risk of crash involvement has been inconclusive.
In fact, the most recent and thorough study on crash risk from drugs (pdf) found that for both marijuana and other illegal drugs: “analyses … did not show a significant increase in levels of crash risk associated with the presence of drugs.” AAA’s study acknowledged that, but Tom Costello and The Today Show didn’t mention it. Did they actually read the article? Or did they deliberately mischaracterize it?
Other media are covering marijuana DUI research more accurately, but still misunderstand it. Ashley Halsey III of the Washington Post and Joan Lowy of the Associated Press both did a somewhat more responsible job by covering a related AAA study (pdf) that shows that tests measuring the amount of THC in the blood do not reliably indicate impairment.
What both articles get wrong (thanks to misleading or misquoted statements from AAA’s Jake Nelson) is that using police officers trained as “Drug Recognition Experts” (DRE) will solve the problem. First of all, the study does not recommend that approach but rather includes it as one of several options:
In the absence of a scientifically based cannabis per se law, there are several options. One is to train officers to detect the signs and symptoms of cannabis use in drivers stopped at roadside. Initial suspicion of cannabis use would lead to a field sobriety test (SFST). This process could be coupled with rapid, on-site oral fluid screening for evidence of drug use. The technology to detect certain drugs (including cannabis) in a specimen of oral fluid quickly at roadside is improving and could be used in a manner comparable to preliminary breath testing devices currently used to test for alcohol. The suspect would then be taken for a complete drug evaluation by a DRE. This approach requires enhancing the complement of DRE officers available to conduct assessments for impairment.
The … approach, however, does have limitations ….
While the AAA study does discuss some of those limitations, it leaves out the question as to whether police officers with a few hours of training are really qualified to perform neurological and ophthalmological testing and to draw medical conclusions from such testing. They’re not.
It also ignores the evidence mentioned above that marijuana and other illegal drugs do not increase crash risk. The option not mentioned by any of these journalists is to recognize that DUI laws should focus on alcohol. We should stop wasting time and money – and stop arresting and prosecuting innocent drivers – over things that do not increase crash risk.