Just Another Day at the Office

Fingerprints, shoe prints, bite mark comparisons, hair and fiber comparisons, ballistics. They're all bullshit.  Oh, the claimed matches might be right.  Then again, maybe not.

The folks who do them claim perfection.

When I've concluded that the fingerprint couldn't belong to anybody else in the world, then it's a match. . . . There is no error rate because once I'm certain it can't be anyone else in the world, then it's done.

That's from an actual trial transcript.

As I say, it's bullshit.

It's not that the folks who do the comparisons are dishonest (though too many are) or incompetent (though too many are).  It's that there are no standards.  All those forensic examinations? They're just the work of paid lookers. 

You know, they get a fancy microscope and look at two things.  If they look enough alike, they're a match.  You may not see the similarity, but you aren't a trained looker.  And you don't understand that there can't be errors because each fingerprint, each fiber and hair and striation and mark is unique and unchanging.

Except, well, we don't actually know that.  And even if it's true, we don't know how to tell when two of these things are so similar that they match.  I mean, how close is close enough?  And, of course, we're not talking about matching two pristine exemplars.  At least one was taken in the rough, after all.  That's the one they want to match, to be able to say

This guy left this smudged partial finger/palm/boot/tire/bitemark/whatever at the scene of the crime and we know it was him because the smudge has similarities to this carefully made version with vastly more detail made in the lab.  You can't tell that just by looking, but I can.  Because I'm a trained looker.

Or something.

Which, as I said, is bullshit.

Just ask Brandon Mayfield.  He's the lawyer in Oregon who didn't leave a fingerprint on the bag holding the detonating devices cops in Spain found after the Madrid train bombing.  You know, the guy that three of the FBI's top fingerprint analysts identified because his print matched the smudge.

It was, they said, 100%.

Which is, as it happens, exactly the percentage by which those crackerjack experts were wrong.

Because it's bullshit. (Have I mentioned that?)

The Philadelphia police department has 12 trace evidence analysts in their crime lab.  Six of them were recently given a routine competency test.  Three flunked.  The agency that gives the test hasn't told the Philly PD which three.

So we know that at least a quarter of the time the analyst's reports are altogether unreliable. We just don't know which ones those are. No doubt that will make the folks who are doing time as a result of the quackerie feel better.

Michael Short worked as a criminalist at the Stark County Crime Lab in Canton, Ohio for 21 years.  He was fired last week for, according to Ed Balint in the Canton Repository, 

falsified reports . . .  improper job performance and insubordination.

The thing is, this stuff isn't really news.  It happens regularly, and it's a real problem.  Hell, I've written about it repeatedly.

And of course, it isn't just forensics.

Consider confessions.  

The night of April 19, 1989, Patricia Meili went for an evening run in Central Park. Along the way, she was attacked, raped, tied up, viciously beaten, and left for dead, which she nearly was when her moaning body was found several hours later lying in a pool of mud and blood. Antron McCray, Kevin Richardson, Korey Wise, Yusef Salaam, and Raymond Santana, Jr. all confessed to participating in the rape and beating.  

Sarah Burns, in The Central Park Five: A Chronicle of a City Wilding (different subtitle for the paperback) tells how the cops gulled them into the confessions.  Their stories were inconsistent, didn't match the facts, made almost no sense.  Oh, and the DNA wasn't from any of them. But they confessed. The cops believed the confessions. The prosecutors believed the confessions. The jurors believed the confessions. The media believed the confessions.

As it happens, the confessions were so much bullshit.

The cops didn't beat them (though there may have been a slap).  The cops didn't torture them.  The cops just preyed on their gullibility and their youth.  (Korey was 16, the others 14 and 15.) 

Tell us you're guilty and you can go home.

So they told.

And they went to prison.

For years.  And years.

For a crime committed by Matias Reyes, which the cops would have figured out had they been paying attention at the time.  And if they could have imagined they could have been mistaken. Burns writes, tellingly.

Despite a prosecutor's obligation to seek justice, it seems that at that moment [when the DNA results came back], winning the case trumped investigating the evidence.

After all, they had these five kids.

Who confessed.

A true story

The state's whore expert had just testified that he'd done a chemical analysis of the powder and it was, in fact, cocaine. He was, he explained, experienced.  He did about 10,000 tests a year.  During cross-examination, here's what happened.

Defense Attorney:        Ten thousand tests a year?  Is that what you said?

Expert:                         Yeah, about that.

Defense Attorney:        Did you do any of them right?

YOU HAVE THE RIGHT TO BE QUESTIONED

Once you have counsel, police can't ask you questions unless counsel is present. That's what SCOTUS said in 1986 in Michigan v. Jackson. Seems simple enough. It's clear, easy to follow, let's everyone know the rules of the game.

OK, maybe the rule was a bit less clear in the 5th Circuit where the rule was understood to apply only after you ask for counsel. Montejo never had a chance to ask; the court just appointed counsel for him, so he didn't have that protection. But that's essentially just the quibble that got the issue back to the Supreme Court after 23 years.

Jackson was a 6-3 decision, Justice Stevens writing the majority opinion. He's the only one of the Justices still on the Court and it seems clear to him that it was right then and is right now. But the composition of the Court has changed. And so did the vote. Today, in Montejo v. Louisiana, Jackson was overruled by a 5-4 vote with Justice Scalia providing the majority opinion.

Jackson should be overruled, Scalia said, because it isn't that well-established, nobody much relies on it, it's hard to apply, and it was a bad idea to begin with since it makes it harder for police to get confessions. None of that is true, of course. But when you want to reverse, you say what you must.

Stevens read his dissent from the bench, asserting bitterly (albeit fatuously), that the "decision can only diminish the public’s confidence in the reliability and fairness of our system of justice."

*********

We interrupt this blog entry to bring you a word from the, uh, real world of actual cops and lawyers and judges, the real world of trial courts, the real world of criminal defendants who confess to stuff all the time - even when it isn't true (no need for waterboarding; just ask my client who confessed to the rape of a 2 month old child who everyone including the police who secured the confession eventually agreed was not, definitively not, raped). In the real world, confessions are almost never suppressed. Certainly, Jesse Montejo's wasn't.

*********

But wait, there's more to the opinion. The Court recognized that it was changing the landscapte, that by relying on Jackson, Montego was acting perfectly reasonably, but he was giving up on arguments that seemed silly to make when he had a sure winner. So they remanded the case for Montejo to try proving that there's some other reason his confession should be suppressed.

Don't hold your breath.