A new report from the Washington State Traffic Safety Commission, entitle “Driver Toxicology Testing and the Involvement of Marijuana in Fatal Crashes, 2010-2014”, has generated some frightening headlines in the mainstream media. [I’ll reference tables from this report in brackets like these, so you can look it up yourself.]
More WA drivers test positive for THC in deadly crashes since 2010, warns KING5 News. Washington Marijuana-Related Traffic Fatalities Double After Pot Legalized, intones Breitbart News. Fatal crashes involving marijuana doubled in Washington after legalization, scolds The Oregonian.
Oh no, legalization has led to stoned mayhem on the freeways! At least, if you trust the news media’s reporting, which ignores one very salient point.
There was not much difference in the total number of traffic fatalities from 2010 to 2014.
In 2010, there were 460 drivers who died on Washington roads. In 2014, there were 462 fatalities. [Table 35: Total Fatalities Involving Any Alcohol or Any THC].The average of the three years before legalization (2010-2012) comes up to a shade over 450. The average of the two years since (2013-2014) comes up to 449.
So if no more people are dying on the freeways than before legalization, what is the problem?
Well, according to the media, it’s that more stoned drivers are dying and fewer drunk drivers are dying. While traffic fatalities over the alcohol limit of 0.08 BAC have dropped almost a quarter since 2010, fatalities involving THC alone or with alcohol and/or drugs have doubled or tripled! [Table 15: Driver Comparison Groups by Year]
Digging deeper into those numbers reveals the statistical sleight-of-hand. Indeed, alcohol-only (>0.08 BAC) fatalities dropped from 67 in 2010 to 51 in 2014. But why should fatality deaths involving THC+alcohol or THC+alcohol+drugs get blamed on the THC? Studies show that alcohol is a far more impairing substance than marijuana or drugs and that marijuana and drugs found in one’s system aren’t as indicative of impairment like a 0.08 BAC for alcohol is.
So when THC+alcohol (>0.08 BAC) jumps from 16 to 23 from 2010 to 2014, that just shows us an increase in drunk driving deaths among people who also happen to smoke pot. The THC+drugs+alcohol (>0.08 BAC) increase from 2 to 6 shows the same thing. We can’t know whether the THC or drugs were affecting the driver, but 0.08 BAC does tell us alcohol was affecting the driver. Add all the alcohol-related deaths at above 0.08 BAC and it went from 85 in 2010 to 80 in 2014.
Of these THC-related fatalities, understand we are not talking about the inactive carboxy-THC metabolites found in urine screens. All the carboxy-THC fatality statistics declined, including carboxy-THC only (down 45.5 percent), carboxy-THC+alcohol (down 75 percent), carboxy-THC+drugs (down 50 percent), and carboxy-THC+drugs+alcohol (down 100 percent). [Table 14: Categorization of Drivers Testing Positive for Alcohol or Drugs by Year.]
Where the scare data lie are among the fatalities testing positive for active THC. In 2010, there were 81 fatalities that had tested positive for cannabinoids, but only 36 of those for active THC, and only one-third of those were over the 5 nanogram per se DUID limit in Washington. In 2014, there were 89 fatalities positive for cannabinoids, but a whopping 75 of them were for active THC, and almost half were over the 5 nanogram limit.
In other words, active THC-involved fatalities more than doubled since 2010, and those over 5 nanograms more than tripled! Should we be terrified now?
Not really. First, even though active THC is a better measure than inactive metabolites, it is still not indicative of impairment. A 2015 study in the Journal of Analytical Toxicology tested 11 frequent cannabis consumers and found that most of them had detectable THC above five nanograms even 8-to-14 hours after last smoking.
So, when we test a driver following a fatality crash, we’re more likely to find that they are a frequent marijuana consumer, but that doesn’t tell us that frequent marijuana consumers are more likely to be in fatality crashes. We’re probably more likely to find married gay drivers in fatality crashes these days, too, now that there is access to marriage for gay people, just as we’re more likely to find marijuana in the system of a fatality-involved driver, now that there is better access to marijuana for tokers.
But that’s not the only confounding part about this report. The other problem is that when you come back from the river with twice as many fish, it doesn’t necessarily mean there are twice as many fish in the river. It could mean you switched from using a fishing pole to a fishing net. That’s what Washington State has done with respect to detecting THC in its fatality crashes. The report warns:
On January 1, 2013, the WSP Toxicology Lab reset the THC reporting threshold to one ng/mL and began conducting full panel (alcohol and drug) tests on all traffic crash blood sample submissions. Prior to this date, the Lab tested blood for the presence of alcohol first. Only if blood alcohol concentrations were under .10, the Lab then conducted drug testing. In addition, full panel alcohol and drug testing was only performed when a driver was involved in vehicular homicide/assault and/or underwent a Drug Recognition Expert examination.
In other words, prior to 2013, when legalization took effect, a drunk driver over 0.10 BAC wouldn’t be drug tested at all, versus now when any driver can be drug tested. Prior to legalization, only drivers charged with vehicular homicide/assault were tested, versus now when any driver can be drug tested. Prior to legalization, only drivers suspected of impaired driving by a drug recognition expert were drug tested, versus now when Washington State has more trained drug recognition experts. And prior to legalization, anybody under two nanograms of THC wasn’t counted, versus now when one nanogram is the threshold.
What a surprise, then, that before legalization when the state was using a fishing pole for drug testing, there were lower numbers of marijuana-related fatalities than after legalization when the state was using a fishing net.
The most important data point is this: since marijuana has been legally for sale in Washington State, traffic fatality rates have not budged. Even the state itself tells us “the information in this report cannot be used to determine if marijuana is or is not causing fatal crashes.”