Burt Hubbard is a wide-eyed, gray-bearded data lord.
He was a top-10 finalist for the Pulitzer Prize at the Rocky Mountain News and is enshrined in the Scripps Howard Journalism Hall of Fame. His numerous awards include Reporter of the Year in Colorado.
For the past dozen years, he has taught computer-assisted reporting and internet research at the University of Colorado School of Journalism and Mass Communication. He has led data journalism symposiums throughout Colorado, and he is The Gazette’s resident Jedi in analyzing vast data sets and spreadsheets, ferreting out their inner secrets.
To Hubbard, reporting a story without good data is like driving with your eyes closed.
Yet, according to Hubbard, that is exactly what Colorado is doing when it comes to its experiment with legalized marijuana: driving with its eyes closed. Now that 29 other states have followed Colorado’s lead, that means there’s a lot of blind leading the blind when it comes to understanding the legalization of marijuana and its ultimate impact.
The Gazette has been trying all year to report on the impact of marijuana on Colorado five years after it was legalized. Time and again, our reporters have been frustrated by an infuriating lack of reliable data measuring that impact, despite state laws that require it.
The recent report released by the Colorado Division of Criminal Justice titled “Driving Under the Influence of Drugs and Alcohol” is an aggravating case in point.
The data lord took a really close look at the report.
“The report by the state was done as a result of legislation so lawmakers could get a sense of how much marijuana use was associated with DUIs and dangerous driving. But it sounds like the data is not there to really document the presence of cannabis,” he concluded.
There is accurate and plentiful data when it comes to alcohol’s effect on driving, but hardly any data on marijuana’s, and what is there is inconsistent and unreliable, Hubbard reports.
The first line of the report’s executive summary painfully illustrates this point: “Very little is known about drug-involved driving when alcohol is not involved.”
(A reminder — that is the first line of a report mandated by law to tell us exactly what the impact of “drug-involved driving is.” And we’ve had five years to study it, 18 if you count medical marijuana.)
So why don’t we have good information on driving while under the influence of marijuana yet? Authorities aren’t collecting it.
“I looked at the report and my initial reaction is that the numbers are not apples to apples and the criminal justice system has included a lot of caveats,” Hubbard said. “First, for all DUI cases, it says that police generally don’t even order a test for cannabis if the driver already has high alcohol levels. The cannabis test is much more expensive. It also says there is no guarantee that if a test showed no cannabis that that wasn’t recorded by authorities (but the division does not know for sure.)”
“I am not sure what to make of the numbers.”
The report puts it this way: “Alcohol is faster, easier and cheaper to screen for compared to other drugs thanks to preliminary roadside breath screenings. Once alcohol is detected, law enforcement generally has enough evidence to reliably achieve a conviction. Therefore, agencies do not consistently spend the additional money and time requesting toxicology blood testing for substances beyond alcohol.”
So one reason we don’t have good data is because it’s just too expensive. I’d hate to explain that to someone who was killed by a drugged driver.
Recent stories about the DUI report concluded contradictory things because the data in the report did, too.
The Denver Post reported that about 73 percent of some 4,000 drivers charged with driving under the influence in 2016 tested positive for marijuana, according to the new report.
“However,” notes Hubbard, “the report found that only about 4,000 DUI defendants were tested for marijuana out of 27,000 DUI cases in 2016. One reason cited was the higher cost of tests for marijuana vs. alcohol, especially if the driver had already violated the .08 standard. The report also found that far more officers have been trained to look for signs of drinking and driving than for signs of marijuana use and driving.”
So we have no idea if the other 23,244 people stopped for DUI used marijuana or other drugs. That makes that 73 percent number worthless.
Other media, including The Gazette, led with this factoid that came straight from the report’s summary: More than 91 percent of suspects who had a toxicology screen done had alcohol in their systems, while only 6.2 percent had marijuana.
But of course, many of those toxicology tests didn’t check for marijuana or drugs, and there’s no way to find marijuana if you don’t check for it.
There are many reasons why we have such poor data, according to the report:
• First, there is no criminal charge for driving impaired by marijuana. The current law applies to driving under the influence of alcohol, drugs, or a combination of the two. Everything is smooshed together in one vague category, which makes precise data, not to mention enforcement, that much harder.
• Second, there is no central repository of toxicology test results that would allow for an analysis of trends, according to the report.
• The technology for detecting marijuana at the roadside does exist but has not yet been accepted in the courts as evidence of impairment.
• Law enforcement officers may choose not to pursue additional toxicology testing if the driver’s blood alcohol content is at or above the legal limit. Once the burden of proof is met, there’s no motivation to do further toxicology testing.
• Testing for marijuana takes longer, because the officer has to transport the suspect to a location where blood can be drawn, usually a jail or a police station.
• Toxicology results are difficult to interpret due to the variation in procedures followed at multiple labs.
Beyond the challenges of testing, there are huge research challenges in trying to assess the impact of pot on driving, in part because it is still an illegal drug according to federal law and access to the drug is restricted and the requirements for researchers are strict and time-consuming.
Since 1968, the National Institute on Drug Abuse has contracted with the University of Mississippi as the single grower for all U.S. marijuana research. Generally, this Mississippi marijuana is of a much lower potency than that in the recreational market, especially in Colorado, which has the most potent marijuana in the country. That discrepancy makes it difficult to generalize the study findings.
There are also multiple methods of consumption that make research difficult. Cannabis can be smoked, vaped, dabbed, inhaled, ingested and even puffed through an inhaler. The onset of a high from edible cannabis can take from 45 minutes to two hours, while the onset of smoked or vaped cannabis happens within the first 10 minutes.
“There is a wealth of information available on alcohol-impaired driving while there is a dearth of research on the problem of drug-impaired driving,” the report concludes. “As the national landscape of marijuana legalization continues to expand, it is critical to gain a better understanding of driving impairment associated with drugs, especially cannabis.”
Other black holes of data
The state’s failure to collect data on this grand social experiment isn’t limited to DUIs and traffic.
Reporting on the sharp rise of arrests for selling black market pot in Colorado, we discovered that no one is keeping statistics on the size of the black market statewide. The government of Colorado doesn’t accurately know if legal marijuana has reduced or increased the illegal trade in pot here.
What about crime and pot? Since 2013, Denver has seen its crime rise as the national average has fallen. In 2016, the crime rate was up 4 percent, with violent crime rising 9 percent, according to the Colorado Bureau of Investigation. We know that half of marijuana cultivation arrests are in the Denver area and that drug investigators consider the metro area a hub in investigating thousands of home grows. The former Denver district attorney, Mitch Morrisey, at one point said, “The Denver Police Department is busier enforcing marijuana laws and investigating crimes directly related to marijuana, including murders, robberies, and home invasions, than at any other time in the history of the city.”
But Denver police say the data is “inconclusive” linking pot to that crime spike and have refused to release data to The Gazette about pot-related homicides or police time devoted to marijuana-related crimes. Denver police have even refused to share such information with other metro-area drug investigators, who are so frustrated by their lack of cooperation they have turned to the Colorado Open Records Act to help them get data.
Why is it so hard to get good data on marijuana-related crimes? Because no legal statute identifies them.
What about homelessness and pot? The state’s rate of homelessness rose 5.3 percent from 2013 to 2017, while the national rate dropped 8.6 percent in the same period, according to data from the Department of Housing and Urban Development. Is the Colorado increase linked to marijuana?
“We’re trying to get data on it. That’s a difficult one to measure,” Gov. John Hickenlooper told CNN. “More data is the only way we’re going to figure this out,” he said.
And the state does not even have a database accessible to the public that documents who owns the state’s various pot businesses and licenses. They have a database of pot companies, and a database of people who have licenses, but not one that tells the public which companies are owned by which people.
The problem with all this, the problem with driving with your eyes closed, concludes Hubbard: We, as a state, really don’t know where the hell we’re going.