The 2009 Report by the National Academy of Science on Forensic Science may finally be having an impact. While popular TV shows give the impression that forensic crime lab analysts can divine the perpetrator based on almost anything, the reality is much much different. Setting aside the problem of crime lab dysfunction, during the past year I have mounted two full scale assaults on what can charitably be called junk science. The first assault was against the fire investigation technique known as negative corpus. The second assault was fought in the fantasy land known as forensic document examination, in particular, handwriting/printing comparisons. Both assaults were successful and the more I looked at these two issues the more offended I became as to their supposed legitimacy.
Now a state largely known for wrongful convictions, Texas, had passed a law which allows defendants to challenge their convictions that were secured by now questioned forensic identification methodologies. While hair comparisons and bite mark comparisons are front and center, the law apparently applies to any forensic method that is now being questioned due to scientific advances. The Wall Street Journal recently reported on the new Texas law and its impact. See http://on.wsj.com/1adHWs3
Tuesday, November 5, 2013
I am ashamed to say that it had taken almost 18 years to mount a Daubert/Kumho challenge in Federal Court for the Western District of Wisconsin. On the other hand, it was the first time that a defendant had done so in this District. Gerald Johnsted was indicted for sending a threatening communication by use of the U.S. Mail. The main evidence for the government was expert testimony by a "forensic handwriting expert" for the US Postal Service opining that the printed threats in two letters were written/printed by the defendant. A full day evidentiary hearing was held in July and the parties subsequently briefed the issue. The district court found that under the Daubert and Kumho standards "that the science or art underlying handwriting analysis falls well short of a reliability threshold when applied to hand printing analysis." The case is on hold while the government decides whether or not to appeal.