Nice feature in the Post about the continuing problems with forensic science. This is in the news this week because the FBI, prodded by an earlier Post investigation, has re-opened 27,000 cases in which data from its crime lab helped to secure a conviction. In particular, they discovered that their lab technicians had been testifying in court that hair from crime scenes was "certainly" from the suspect, when there is no certainty at all about this method. This was proved conclusively in the 1970s and the FBI modified its own guidelines to say that hair "matches" should not be used except as supporting evidence. But their own "experts" never got the memo and never stopped testifying about supposed hair matches, which have helped convict hundreds of people. The Post's summary of this method says, "Hair and fiber cannot be identified to one source." The same summary also notes that the marks guns leave on bullets are not unique and matching them is difficult, elemental analysis of bullet lead is "unreliable and potentially misleading," bite mark analysis lacks any supporting evidence whatsoever, and DNA works if you do it right but police and crime labs often mess up the samples, nobody knows how often.
By way of relating this to cases in the news, let me say I am personally convinced that George Zimmerman brought on the confrontation that led to Trayvon Martin's death, seriously mishandled the situation, and then lied about being attacked out of the blue. The verdict was nonetheless correct. We have no way of knowing beyond reasonable doubt what really happened. The scandal is that we send tens of thousands of poor black men to prison without giving any consideration to the real weight of evidence in their cases.
Given what sociologists have shown about the unreliability of eyewitness testimony, the ongoing scandal of what we call forensic science, and the steady stream of revelations of police and prosecutor misconduct, I doubt the standard of beyond reasonable doubt could be met to my satisfaction in half of criminal cases. Maybe not in 10 percent. It may be that our society could not survive if we really took our own laws and standards seriously, but if so we should acknowledge that and think hard about what it means.