Saturday, January 26, 2013

Don't underestimate import of false arson convictions

Maurice Chammah at the Texas Tribune has coverage of the arson review at the Texas Forensic Science Commission titled "Arson reviewers expect small number of cases," describing a report from my IPOT colleague Nick Vilbas who told commissioners he expected perhaps 6-8 cases to be sent through the process that had a serious chance of the convictions being overturned. Commissioner Sarah Kerrigan suggested that, "When you have six to eight cases out of a thousand, you're really restoring public confidence in the fire investigation system.”

Your correspondent interjected, though, to point out that one shouldn't draw too many hard conclusions from those numbers. Yes, IPOT sent 1,085 questionnaires to TDCJ prisoners convicted of arson, but only 173 returned them. Anyone who's ever done direct mail knows you'd need to do more than one pass to get a higher response rate. It's possible there are other innocents in prison who for whatever reason simply didn't return the form. Out of the 173 who did, said Vilbas, 33 warranted further review, and it's out of that subset that he estimated 6-8 of the cases might turn out to be false convictions based on junk science. In addition, as Chammah's story pointed out:
At the meeting, Scott Henson, a policy consultant for the Innocence Project and author of the criminal justice blog Grits for Breakfast, cautioned the commission that the arson review only addressed convictions in which someone is currently serving prison time. They did not look at cases in which the convicted arsonist is on parole or probation, or has finished their sentence. In addition, Henson said, many false arson accusations did not enter the criminal justice system, but instead were handled by the Texas Department of Insurance as a part of questions over fire damage claims.
So while I agree with Commissioner Kerrigan that the arson review should go a long way toward "restoring public confidence," IMO it would be incorrect to imply that those 6-8 cases are the only ones out there. If there turn out to be six innocence cases out of the 173 prisoners who responded, that'd be a rate of 3.5%. IMO it's probably not correct to apply a larger denominator that includes cases never vetted. That said, IPOT and the state fire marshal have set up a process whereby if more cases are identified in the future, there's now a method for having experts vet them for potential innocence claims, as fire marshal Chris Connealy reminded me outside the meeting room..

Despite Grits' caution about minimizing preliminary numbers, Professor Kerrigan is surely correct that the arson review process will contribute to rebuilding public confidence that was justifiably shaken by the Todd Willingham fiasco. But there's a long way to go before the public can be sure these cases will be justly resolved. Indeed, it remains unclear whether the Court of Criminal Appeals will provide habeas relief based on new scientific evidence; in past opinions, several members of the court have said they shouldn't.

Texas is the first state to undertake this sort of comprehensive review of old arson cases and no doubt the participants will learn a great deal as they go through the process, which is only just beginning. In fact, Connealy said other states have expressed interest in copying Texas' process and may launch similar efforts. Wouldn't that be something?

2 comments:

Barry Scheck said...

Excellent analysis of the error rate. A further distinction should be made between those arson cases that were based on either chemical detection of an accelerant or direct evidence of criminal activity (witnesses who saw defendant light the fire, etc) and "expert" evaluation of the crime scene from which conclusions were drawn an accelerant was used when none was found.

Gritsforbreakfast said...

Thanks Barry, that means a lot coming from you. Good point on the distinction between cases involving direct evidence and those based solely on conclusions of an arson investigator. The latter are definitely the category where the possible false convictions are coming from.