Yesterday afternoon the Office of National Drug Control Policy released a report (which you can read below) that found a majority of arrestees in five metropolitan areas had marijuana in their system when they were booked, and that many others had recently used harder drugs. During his speech before the Urban Institute yesterday, Drug Czar Gil Kerlikowske argued that the report is evidence that we shouldn't legalize marijuana. Members of the media naturally jumped on his claims. Here's a typical headline and lede, courtesy of McClatchy:
MARIJUANA IS DRUG MOST OFTEN LINKED TO CRIME, STUDY FINDS
WASHINGTON — Marijuana is the drug most often linked to crime in the United States, the U.S. drug czar said Thursday, dismissing calls for legalization as a "bumper-sticker approach" that should be avoided.
Gil Kerlikowske, the White House director of national drug-control policy, said a study by his office showed a strong link between drug use and crime. Eighty percent of the adult males arrested for crimes in Sacramento, Calif., last year tested positive for at least one illegal drug. Marijuana was the most commonly detected drug, found in 54 percent of those arrested.
We're going to see versions of this story everywhere, and I wouldn't be surprised if we saw most of them written up the way McClatchy's was, which is to say, without any indication that reporter Rob Hotakainen actually read the 2012 Annual Report on the Arrestee Drug Abuse Monitoring Program II (or ADAM II in ONDCP shorthand), which is 122 pages long–far too long for Hotakainen to have examined it before firing off a dispatch about Kerlikowske's speech. And yet, reading the report is the only way to tell whether Kerlikowske is spinning the results. (He is.)
First, here's how the study was conducted: Over the course of 21 days in 2012, researchers in five metropolitan areas–Atlanta, Chicago, Denver, Sacramento, and NYC–asked male arrestees in large urban jails to submit to voluntary interviews and urine tests. The requests were made as close to booking as possible, "and always within 48 hours of arrest." Arrestees were asked about their education, living arrangements, employment, and health insurance. They were also asked about their drug use. It's here that we encounter the report's first problem: In addition to asking arrestees how often they used illegal drugs in the last year, interviewers also asked them how often they consumed five or more alcoholic drinks in a sitting. The drug use statistics are included in the report, but the alcohol data is not.
According to a 2006 report from the Pacific Institute for Research and Evaluation, "crimes attributable to alcohol cost 84 billion dollars, more than two times the 38 billion dollars attributable to drugs." I'm not saying that's why the alcohol data–which we know ONDCP collected–was omitted from the report; but if you're timing the release of your report to coincide with a speech about the need to force users into treatment while keeping drugs illegal, you don't want alcohol getting all the attention. (I've filed a FOIA request for this data, btw, and will post it when I get it.)
The report's problems don't end there. While ADAM II is incredibly detailed with regard to the types and frequency of drug use, it's far less clear why suspects were arrested in the first place. The most detail ADAM II provides is whether the arrest was for a violent crime, a property crime, a drug crime, or "other." If you're going to argue (as Kerlikowske has) that the link between marijuana and crime is so troubling that it precludes the possibility of legalization, it certainly matters whether an arrest is the result of a traffic stop in which officers claimed to smell weed, an unconstitutional stop-and-frisk, or an undercover officer convincing an autistic student to buy him a joint. These are not hypotheticals, by the way: In New York last year, where ADAM II says 29.8 percent of arrests were for drug crimes, the NYPD stopped, frisked, and arrested 5,000 people for marijuana, according to the New York Civil Liberties Union. Which means that If Kerlikowske were being honest, he would've said there was a strong correlation between using marijuana and getting arrested for marijuana.
That's not to say that the ADAM II research doesn't tell us anything noteworthy. Thanks to the ONDCP's report, we know that in Atlanta, 82 percent of arrestees last year were black, 39 percent of arrestees did not finish high school, 52 percent had been arrested 1-2 prior times in the past year, and that 74 percent did not have health insurance. We know that in Chicago, 76.2 percent of arrestees were black, only 3.4 percent of arrestees had a college degree or higher, 56.7 percent of arrestees did not own or rent their own home, 41.8 percent of arrestees were unemployed, and 78.7 had no health insurance. We know that in Denver, 53.8 percent of arrestees were white, 43.8 percent were over the age of 36, 18.4 percent lived in a shelter or had no fixed residence. We know that in Sacramento, 41 percent of arrestees were over the age of 36, 48 percent were white, 41 percent were unemployed and looking for work, and 19 percent were homeless.
If there's a correlation between drugs and crime, there's an even stronger one between being poor and getting caught.
Oh, and there's one other thing we know from ADAM II: "One of the questions of interest to policymakers is whether there are significant differences between arrestees who have illegal drugs in their systems at the time of arrest and those who do not. With the exception of citizenship status and employment, there were few significant differences between these two groups across all or almost all sites."
Simply put, this is a report about poverty.