The argument for privatizing crime labs, could prevent errors and analyst bias.

Scandals have plagued state crime labs in North Carolina, California, Virginia, Illinois, Maryland, West Virginia and Mississippi; the city crime labs in Houston, Cleveland, Chicago, Omaha, Oklahoma City, Washington and San Francisco; the county lab in Nassau County, New York; and even at the FBI and Army crime labs.

What's going on? Most of these scandals were exposed after DNA testing cleared someone who was convicted based on testimony from crime lab analysts. DNA testing, which is actually grounded in solid science, is showing that forensic analysis isn't as certain or as scientific as it is often claimed to be. It's also showing us that forensics is plagued by bias, both intentional and unintentional, and that bias is caused by poorly structured incentives that often reward crime lab workers for helping win convictions, not for sound analysis. The Innocence Project estimates that bad forensic science contributed to about half of wrongful convictions that were later exposed by DNA testing.

Consider the scandal in North Carolina, uncovered last year after a state investigation and follow-up series by the Raleigh News & Observer. The initial investigation found at least 230 cases in which crime lab workers failed to turn over potentially exculpatory evidence, including three cases that resulted in the defendant's execution.

The News & Observer follow-up found even more, often stunning, bias at the lab, including training manuals that taught workers to consider defendants and their attorneys as the enemy. Many lab workers' performance reviews were actually written by prosecutors. In one case, two blood-spatter specialists were caught on video high-fiving one another after running through multiple experiments until they found one that supported the prosecution's theory of a case.

It isn't difficult to see how having experts who present themselves in court as objective, unbiased analysts report directly to prosecutors or police agencies could present some problems. But that's exactly what's happening. According to a 2009 report on forensic science by the National Academy of Sciences, more than half the crime labs in the U.S. report directly to a law enforcement organization. In some cases, this can lead to overt pressure from police officers and prosecutors to produce desirable results. But most of the time the bias is more subtle, and unintentional.

Link:
http://www.huffingtonpost.com/2011/06/14/the-case-for-private-crime-labs_n_876963.html