Thursday, July 31, 2014

Forensic sciences in the criminal justice system: still a complete mess

Remember 5 years ago, when the National Academy of Sciences published a comprehensive report about the state of forensic crime labs and their impact on the criminal justice system?

Among the systemic failures noted by the NAS (and others, notably defense attorneys):

  • that crime labs are too inextricably linked with the prosecution and investigators, thus infusing bias into the testing process
  • that no national standards existed for many testing disciplines
  • that results are overstated in court
  • that junk science is passed off as conclusive evidence

Well, this might surprise you, but nothing much has changed. Why? Well, I'd suggest it's because the institutions that are currently in charge of the crime labs and the types of testing done don't want things to change. Because admitting things need to change is tantamount to admitting past mistakes. And we know prosecutors, the FBI, and pretty much everyone else involved in prosecuting people don't much care to admit past mistakes.

Now we learn that

Nearly every criminal case reviewed by the FBI and the Justice Department as part of a massive investigation started in 2012 of problems at the FBI lab has included flawed forensic testimony from the agency, government officials said. Washington Post
And the FBI's initial response was to stop the review. A see no evil kind of response, perhaps? If we stop digging, we'll stop seeing all the bad things we've done in the name not of science but of convicting people.

Some of the stuff that was reviewed for that 2012 investigation, and that still needs to be reviewed, includes "hair matching" which I put in quotes because it's such bunk. Tip if you're ever on a jury: if a lab analyst gets on the stand at your trial and starts talking about "hair matching," know said lab analyst is full of it. Unless a hair left at a crime scene has a follicle from which DNA can be extracted, there's no way for a person to use any scientific method to claim to match that hair to the hair of a particular person.

At yet, the FBI did it all the time. The first questions about the conduct of some FBI examiners were raised in 1997. And yet, we still don't have a thorough review of all the cases those examiners touched. People are on death row. People have been executed. But still, to this day, the FBI is dragging its feet about honestly assessing the scope of the problem. With that, I will presume, they're not being as insistent and urgent about making changes as I would like to see. (As I would hope anyone interested in justice and good science would like to see.)

No one benefits when the FBI, or its state and local counterparts, play fast and loose with science in the name of nailing the perp. That kind of mentality leads to wrong results. The evidence, when properly and neutrally examined, should lead investigators to suspects. Too often, though, crime labs that have any connection to, and thus a vested interest in, the investigation look at the evidence with the hope of using it to convict the pre-determined suspect. When crime labs and investigators work for the same team, investigators can get a little too tempted to tell the lab analysts what result they need to see. All forensic sciences are capable of having the results cajoled, massaged, overstated, or just flat manipulated. Yes, even DNA isn't exempt from this because there is always a human element of interpretation to any test result. This is a large part of what the FBI uncovered when it was conducting this review, finding that analysts in court were grossly overstating results, indicating far more confidence in matches than science called for. Heck, in many situations, declaring a match at all isn't supported by the science. But when your colleague is really sure this suspect is the guy and there are no protocols telling you you can't declare a match...

This is why national standards matter and were recommended 5 years ago. So we can't have an overzealous fingerprint analyst in Oklahoma be willing to declare a match based on 3 points of similarity instead of 8 or 9. So we can do away with scientists with all their sciencey credentials coming into court and telling jurors they did "microscopic hair analysis" and the crime scene hair matched the defendant's hair.

This is why the NAS recommended crime labs be made independent of police and prosecutors offices 5 years ago. So that crime lab analysts wouldn't be part of the prosecution team, but would be free to analyze the evidence free of that bias. Scientists left to their own science devices and purposes would never have come up with that microscopic hair analysis nonsense. Instead, it was thought up by FBI employees who were part of the team trying to catch the bad guys.

It is disheartening to see the FBI is still dragging its feet about acknowledging problems with bad evidence, overstated testimony from its analysts, and the bias that comes from working for the agency that's charged with arresting the bad guys. It is horrifying to realize how many people have been executed or incarcerated for decades based on this type of evidence.

There has been far too much delay already. We desperately need to clean up our approach to forensic sciences in the criminal justice system. One more day of delay is inexcusable.

No comments:

Blog Designed by : NW Designs