Sarah Kliff reported yesterday on research that shows, Drunk Eyewitnesses Are Just as Good as Sober Ones. The research was performed by researchers at the University of Gothenberg. They specifically looked at whether people who watch a video of a crime could identify the perpetrator in a line-up a week later. Half the subjects were drunk when they watched the video. And what a surprise! The drunk viewers were as good as the sober ones.

The reason this is the case is that people are really bad eyewitnesses—whether sober or drunk. But strangely, this does not seem to be the take away that the researchers get from the results. They note that many crimes happen where people are drunk. So now, in court we can use drunk witnesses without the fear of (mostly) defense attorneys arguing that the eyewitness testimony ought to be discarded.

I take the opposite from this research: sober people are no better eyewitnesses than drunks! Juries find eyewitness testimony extremely powerful. The last thing our justice system needs is more unreliable, often racist, eyewitness testimony.

It boggles my mind that criminal justice researchers would pitch their results in such an evil direction. Look, I’m all for convicting the guilty—at least when it comes to real crime, as opposed to moralistic “crime.” But our criminal justice system is too often more like a random conviction generator than anything else. But not completely random. If you’re white in the suburbs, you’re probably safe. If you’re black in a city: good luck! Because if you have bad luck, you’ll likely be picked out of a lineup by some drunk the police grabbed off the street.