The statistics are surprising. Scary, even.

Fatal car crashes involving Washington drivers who recently used marijuana doubled in just over a year, according to a new AAA study. Shortly after the state legalized the drug in 2012, 8 percent of these motorists had THC, the “active ingredient” in cannabis, in their systems. In 2014, 17 percent had it in their systems.

Utah is dealing with a similar situation. There, the Department of Public Safety recently reported the amount of fatal crashes involving drivers who tested positive for pot jumped from 6 percent to 15 percent between 2012 and 2015.

‘Recently’ Doesn’t Mean Within the Last Hour

One factor that softens the blow of these figures a bit: the “recently” used above doesn’t mean a driver used marijuana a few hours before the crash. He or she could have used it within the last week or even month.

Also, the current practice of looking at THC levels in the blood to determine if someone drove while impaired is far from perfect. In fact, AAA says “legal limits for marijuana and driving are arbitrary and unsupported by science.”

Regarding that second point, Jake Nelson, AAA’s director of traffic safety advocacy and research, recently told The New York Times “there is no concentration of the drug that allows us to reliably predict that someone is impaired behind the wheel in the way that we can with alcohol.”

AAA added in a press release that “depending on the individual, drivers with relatively high levels of marijuana in their system might not be impaired, while others with low levels may be unsafe behind the wheel.” That’s very different from alcohol, “where it is clear that crash risk increases significantly at higher BAC levels.”

Another problem is “frequent users of marijuana can exhibit persistent levels of the drug long after use.” When it comes to occasional users, on the other hand, “drug levels can decline more rapidly.”

More on Why Current Legal Limits Don’t Mean Much

Those and other issues make it likely officials are wrongfully convicting some motorists for impaired driving. They also make it likely some unsafe drivers are going free. And they make it tough for states to develop guidelines related to all of this too. 

There is “a strong desire by both lawmakers and the public to create legal limits for marijuana impairment, in the same manner as we do with alcohol,” Marshall Doney, AAA’s president and CEO, said in a release.  Unfortunately, “it’s simply not possible today to determine whether a driver is impaired based solely on the amount of the drug in their body.”

"Alcohol and cannabis are very different drugs. They behave in the body in very different ways,” Doney’s colleague, Nelson, recently told livescience.com. “Trying to use the system from alcohol is not the way to go."

As a result, AAA is pushing states to “use more comprehensive enforcement measures to improve road safety.” Rather than rely on legal limits, it suggests they use a system that requires:

  • A positive test for recent cannabis use
  • Behavioral and physiological evidence of driver impairment

High Drivers Less Likely to Crash Than Drunk Ones

Should some states follow that advice, it’ll be interesting to see how many people are pulled over and convicted of a DUI a year or two down the road.

After all, a 2015 National Highway Traffic Safety Administration (NHTSA) study found that “drivers who use marijuana are at a significantly lower risk for a crash than drivers who use alcohol.” Even more surprising: after adjusting for age, gender, race, and alcohol use, those same people “were no more likely to crash than those who had not used any drugs or alcohol prior to driving.”

Don’t take those findings to mean it’s a good idea to ingest marijuana in some form or fashion and then get behind the wheel. The drug still impairs "one measure of driving performance"—by giving motorists tunnel vision--the NHTSA says.