Ever heard the old saying, ‘You can make statistics say anything you want?’ A new federal drug study is vainly hoping to do just that. In interviewing and urine testing ‘random’ arrestees from 10 major cities, half of the results came up positive for at least one type of drug.
The study [Warning: .pdf], conducted by the Arresting Drug Use Monitoring Program (ADAM), maintained that marijuana was the most common drug found in respondents (between a third and a half of the men tested positive) in all the cities except for Atlanta. It also found high rates of cocaine, crack use, and methamphetamines, among the arestees.
Newly minted drug czar Gil Kerlikowski noted that the study,
...shows a clear link between drugs and crime
Really, Gil? Does it? The word ‘link’ here is incredibly misleading. As any statistician knows, there is a universe of difference between correlation and causality. In other words, just because there is a ‘link’ between drugs and crime, doesn’t mean that using drugs causes crime, or that criminals are more likely to use drugs than non-criminals. For example, one could correctly say that 85% of US criminals have brown eyes. There is definitely a link between brown eyes and crime. If this were a causal link, we might have to start putting Lojack bracelets on all brown-eyed people.
Even outside the misleaing semantics, the study fails on several other levels. First of all, any arrestee who tested positive for pot smoked at least once in the last six weeks. There is absolutely no way to know whether the person was high at the time of the crime, arrest, or booking. Given that 42% of Americans have tried pot at some point or other, there’s a very good chance you’ll be able to get a positive test from a significant number of them.
The kicker though, can be found in the study itself. Some poor ADAM intern forgot to expunge the following passage from the study:
In all sites, a minimum of one in five arrestees had a drug crime amongst the charges on the current arrest, with larger percentages in Sacramento and Washington... and Chicago having the highest percentage (60%).
In other words, these federal researchers went out of their way to disproportionately interview people arrested on DRUG CHARGES. Hmmm, do you think it’s possible that people arrested for possessing drugs might be pretty likely to test positively for them?
Poor Chicagoans. National newspapers this month are rife with stories of the throngs of addict criminals in their prison system, and no one thought it fitting to mention that of the few hundred Chicago men surveyed, 60% of them had been arrested for possessing or selling drugs.
Apparently the original executive summary: "People Arrested for Ingesting Drugs Tend to Test Positive for Them," didn’t fly with the higher ups. Congratulations to the ONDCP on another mountain of taxpayer dollars well spent.