By MICHAEL J. SNIFFEN
The Associated Press
Thursday,
September 1, 2005; 9:37 PM
WASHINGTON -- The FBI decided Thursday to stop tests that match bullets by
lead content, a practice criticized as producing a high rate of false matches
between crime scene bullets and bullets taken from suspects. The bureau said it was informing 300 state, local and foreign law enforcement
agencies that had received positive match reports from the FBI Laboratory since
1966. The FBI said it had not determined those results were wrong but informed
them so they could take whatever action they deem appropriate. Criminal defense attorneys have contended that re-evaluation of these tests
could affect some convictions on appeal. One New Jersey defendant has been
granted a new trial because of questions about the bullet test analysis, and a
convicted double murderer in New Zealand has requested a pardon based on
questions about the test. Jack King, National Association of Criminal Defense Lawyers spokesman, said
defense attorneys will track down most defendants in those 300 cases. "Now, we
need the FBI to provide a live witness, a scientist from the FBI Lab, to testify
at post-conviction hearings on these old cases," King said. "Some of these guys
have sat in jail for decades, and it's about time they got a fair hearing." The bureau spent more than a year reviewing recommendations of a scientific
panel that last year criticized the reliability of conclusions based on the
tests. The panel was assembled by the National Research Council, principal
operating arm of the National Academy of Sciences, a private institution
chartered by Congress to advise the government on scientific matters. The FBI said it dropped the tests primarily because "neither scientists nor
bullet manufacturers are able to definitively attest to the significance of an
association made between bullets in the course of a bullet lead
examination." Examinations of the lead content of bullets have been performed in limited
circumstances, the FBI said. Typically they are done when a firearm has not been
recovered or when a fired bullet from a crime scene is too small or damaged to
compare with the marks left on a slug by the barrel of a firearm. These bullet lead exams determine the amounts of trace elements, like copper,
arsenic, antimony, or tin, in the bullets. That analysis allows crime-scene
bullets to be compared to bullets found in a suspect's possession or weapon, the
FBI said. Since the early 1980s, the bureau has performed bullet lead examinations in
2,500 cases submitted by federal, state, local and foreign law enforcement
agencies. But the results were introduced at trials in fewer than 20 percent of
those cases, the FBI said. The NRC study, requested by the FBI and released in February 2004, found the
FBI's chemical analyses and equipment were sound. However, the study questioned
a statistical analysis method known as chaining that compares trace elements in
a series of bullets. It noted that bullets sold in one package are not
necessarily all from the same batch of lead. The study was first reported by The Associated Press in November, 2003. Kenneth O. MacFadden, a Chestertown, Md., research and management consultant
who chaired the NRC study, said its most urgent recommendation was to have FBI
witnesses in criminal cases more clearly explain the limits of the test. MacFadden said chaining is like saying bullet A is like bullet B and B is
like C and C is like D and so on, and then concluding that means A is the same
as E because they are part of the same chain. The committee said this can lead to an artificially large group of bullets
being considered identical, "when this would not be true if other statistical
methods were used." Last year, King, of the defense lawyers group, had predicted the NRC study
would affect cases on appeal. "They cannot match lead samples like fingerprints"
but an FBI scientist's testimony gets more credibility from the jury than it
deserves, King said. Last March, a New Jersey appeals court ruled the FBI lab technique is based
on "erroneous scientific foundations" and overturned a 1997 murder conviction.
The court ordered a new trial for defendant Michael S. Behn because the FBI
analysis that used chaining to link bullets found at Behn's residence with those
used in the killing was the only expert testimony not countered by Behn's
lawyers. "The integrity of the criminal justice system is ill-served by allowing a
conviction based on evidence of this quality, whether described as false,
unproven or unreliable, to stand," the judges said. That is believed to be the only U.S. conviction overturned on this basis
since the NRC report, King said.
