The Other CSI Effect

There has been a lot of moaning and groaning recently from prosecutors' corners about the so-called "CSI effect" where juries, presumably having watched to many episodes of CSI, are demanding high-tech forensic science in cases where they shouldn't. Most of the stories I have read about this were purely anecdotal, dealing with a prosecutor here or there citing a case here or there where the jury didn't convict because they hadn't analyzed the dirt and plant life found on the body or something ridiculous. In reality, I have found that if there is any real CSI effect, it tends to favor the prosecution. If jurors hear scientific testimony from a crime lab, they believe it, even when it shouldn't be believed. In fact, a lot of forensic "science" is not really science at all. It has never been subjected to rigorous scientific review standards such as a verifiable rate of error. DNA evidence is an example of forensic science that is real science, but even that is no good when it is performed by incompetent lab technicians. Things like ballistics, fingerprint analysis, hair analysis, and others would never pass muster in the non-forensic scientific world. Fingerprint analysis does not even have a uniform standard for determining when two prints are a match. Some experts require six points of comparison, others 8, still others 12. Many experts don't have any set number. They just use their "professional judgment" to pronounce something a match. This is a real problem, and it has been a major factor in wrongful convictions. The Houston Chronicle had an excellent column this weekend on the perils of forensic science, pointing out that false or flawed forensic science was a factor in 63% of a sample of wrongful convictions.

So, what is the solution? If the criminal courts in Texas ever seriously applied the Daubert standard like it is in civil cases, a lot of this mess could be avoided. All sorts of expert testimony is admitted in criminal cases here that would never get close to a jury in civil cases. Another idea is to separate crime labs from law enforcement. When forensic scientists actually work for the police or the state troopers, it's a lot harder to be certain about the work being that of independent scientists, and not agents of the police.