Minnesota sheriff asserts “approximately 54 percent of males arrested for violent crime test positive for marijuana in Hennepin County.” Really?

In this post last week, I wondered aloud whether “anyone has yet tried (or if it really would even be feasible) to develop effective and sophisticated empirical studies to explore if there have been any statistically significant changes in violent crime rates or unemployement rates in states that have legalized medical marijuana.”  With that query still in mind, I was especially struck by one statistical claim made by Hennepin County sheriff Rich Stanek in this recent op-ed headlined “Lax marijuana enforcement is bad.”   Here are excerpts from the commentary with the data claim of special interest in bold:

The U.S. Department of Justice (DOJ) recently announced that it does not intend to challenge policies in Colorado and Washington state that legalized the sale and recreational use of marijuana to adults — despite the fact that these state laws are in opposition to federal law.

As president of the Major County Sheriffs’ Association, I have joined a broad coalition of law enforcement officers from across the country to express our extreme disappointment with this unprecedented decision.

As law enforcement officials with decades of experience, we know that keeping neighborhoods safe will become more difficult for our men and women on the front lines because of the DOJ’s decision…..

Marijuana is an addictive gateway drug that harms Minnesota’s children and public safety in every community in our state. As sheriff of Hennepin County, I am concerned that legalization of marijuana in other states and reduced federal prosecution will increase the availability of marijuana in Minnesota.

I have seen firsthand in Hennepin County that there is a direct connection between marijuana and violent crime. Drug task forces here have linked marijuana to assaults and homicides. In the Hennepin County Adult Detention Center, marijuana is the most commonly detected drug among the 36,000 inmates who are booked into the facility each year.  According to our most recent data, approximately 54 percent of males arrested for violent crime test positive for marijuana in Hennepin County.

The student who sent me the link to this claim by Hennepin County sheriff Rich Stanek remarked that this number seems very high.   I share this reaction, in part because I think advocates against modern marijuana reforms would be frequently stressing a link between marijuana and violent crime if it was common to find in more than a few justidictions that over half of all males arrested for violent crimes tested positive for marijuana. (Of course, we all know that correlation does not prove causation and that prohibition rather than legalization might be the reason marijuana users turn to crime, but such a statistic still struck me as potentially quite valuable for the anti-reform forces in the broader debate.)

Intrigued by the data claim made here by Sheriff Stanek, I have now written via e-mail directly to the Hennepin County Sheriff’s Office seeking more information about the basis for these claims linking marijuana use and violent crime.  (I quickly got this automated response to the e-inquiry: “Emails are answered when staff is available to view them and respond. It may take several days before your email is viewed. Thank you for your patience.”)

While patiently waiting for more information concerning these violent crime data claims from the Hennepin County Sheriff’s Office, I did a little digging about drug testing of arrestees and found that the Office of National Drug Control Policy released this 2012 Annual Report on arrestee drug abuse.   With a focus on five major US cities (none in Minnesota), this report found that in 2012, “the proportion of [big city] arrestees testing positive for marijuana ranged from 37 percent in Atlanta to 58 percent in Chicago,” but it also reported that only “17 percent (Atlanta) to 27 percent (Chicago) of [these] arrestees in 2012 had a violent crime as one of the charges recorded for the current arrest.”

In other words, though there appears to be extant (and seemingly rigorous) data from the ONDCP report to support a claim that a majority of total arrestees in some urban areas may test positive for marijuana, there is still reason to suspect that nonviolent drug arrestee (particularly those arrested for marijuana offenses) will be disproportionately among those testing positive for marijuana, as opposed to those who are arrested for violent crimes.   (And there are other interaction and intersectionality concerns with the data here too, as a significant percentage of arrestees in the ONDCP groups tested positive for multiple drugs in their system and reported alcohol and prescription drug abuse, too.)

Interestingly, two of the five cities that are the focus of the ONDCP report are in medicial marijuana states (Denver and Sacremento), while the three others are in prohibition states (Atlanta, Chicago and New York). It might be interesting (though surely challenging) for a sophisticated empiricist to use the ONDCP data to explore whether big-city arrestee drug use data is potentially impacted by marijuana laws in a particular state.