Aside from the 1997 capital murder conviction of Joe Louis Dansby, 61, of Redland in Nevada County, hundreds of other convictions nationwide have been called into question since the science behind what is known as “compositional bullet lead analysis,” (CBLA) was discredited by a 2004 National Academy of Sciences report commissioned by the Federal Bureau of Investigation.

Third in a series. Aside from the 1997 capital murder conviction of Joe Louis Dansby, 61, of Redland in Nevada County, hundreds of other convictions nationwide have been called into question since the science behind what is known as “compositional bullet lead analysis,” (CBLA) was discredited by a 2004 National Academy of Sciences report commissioned by the Federal Bureau of Investigation. The FBI's use of the analytical technique reflected in testimony in the 1998 trial of a Florida man accused of killing his wife was first successfully challenged in 2009 in the case of Jimmy Ates, according to the Innocence Project of Florida. “At Jimmy's trial, FBI analyst Kathleen Lundy testified that the chemical composition of the bullets found in Norma Jean's body matched the chemical composition of bullets found in Ates' utility room,” the IPF notes in a statement of facts. “Lundy therefore concluded that the bullets must come from the same unique batch.” Ates successfully challenged his conviction on appeal upon the discovery of new evidence, and the conviction was vacated based upon an agreement with the special prosecutor who tried the case, according to the IPF. “Bill Cervone deserves praise for righting this wrong, and we hope and expect that other Florida prosecutors will follow his example in other CBLA cases,” Seth Miller, IPF executive director, said in a web posting of the story. Local reliance Evidence suggests that an understanding of the underlying flaws of CBLA was relatively widespread in the legal and scientific communities after the original challenges by former FBI chief forensic metallurgist William Tobin beginning in 1999. But, there were legal challenges as early as 1974. In State v. Krummacher, 523 P. 2d 1009, 1017-18 (Or. 1974), an Oregon state trial judge ruled that CBLA evidence was admissible in court based on the fact that its probative value outweighed its prejudicial impact, irrespective of whether it met any other standards of evidence. And, it has been that kind of acceptance in the state court venues that has fostered other legal challenges to the concept of CBLA, not merely as evidence, but also as science. In 2009, Oregon state inmate Philip Scott Cannon successfully challenged the prejudicial acceptance of CBLA by prosecutors and judges in Oregon. Cannon was tried in state court in January, 2000, in connection with a 1998 triple homicide at a mobile home in West Salem, Ore., where he had been contracted to repair the plumbing. His convictions were set aside in a stipulated judgement in September, 2009 and all charges dismissed in December, 2009. According to pleadings in a 2010 civil lawsuit brought by Cannon and his family, in 2000, the prosecutor and an attorney for the state sought to have the Oregon State Police laboratory perform a CBLA. “Nyhus and Fisher attempted to have OSP crime lab perform the CBLA analysis and were advised by Kenn Meneely at OSP that bullet lead analysis was of questionable forensic value and OSP would not perform this test,” the filing states. At that time, the only facilities in the nation equipped to perform CBLA were the FBI Laboratory in Quantico, Va., and the Oregon State University, the filing notes. An analyst for the OSU lab, Michael Conrady, testified as an expert witness for the state at trial, according to the filing. “Conrady inaccurately determined that Plaintiff possessed ammunition in his residence that was 'analytically indistinguishable' from the lead found in the bodies of the victims,” the filing states. And, in closing arguments at trial, the prosecutor asserted that the findings were indisputable. “Polk DA/DOJ used this testimony that the bullets found in the victims and the ammunition from Plaintiffs homes must have come from the same 'batches' and that the chances of that occurring by coincidence were 1 in 64,000,000.,” the filing states. “This statistic was presented to the jury by Nyhus during closing rebuttal arguments, and no evidence to support this statistic had ever been generated during the trial, nor has any scientific study ever been conducted that supports this conclusion.” Despite the clear rejection of CBLA as a scientifically accepted forensic tool by the 2004 NAS report, the filing states that OSU and Conrady continued to use the technique. In a further effort to buttress the CBLA evidence, a retired engineer for an ammunition manufacturer testified for the state. “Birch testified about bullet distribution and the smelting of lead used in the manufacture of .22 caliber bullets,” the filing states. “Birch's testimony about the bullet lead found in Plaintiff's garage being unique was inaccurate, because he did not realize until trial that the bullets in Plaintiff's garage were mispackaged by Remington.” Federal standards Such reliance by state prosecutors on CBLA has been possible over the years, as Tobin noted in his research, that it has become almost legendary as a prosecutorial tool. But, that has not been the case in every instance, as some state courts have adopted aspects of federal evidence standards. In Clemons v. State of Maryland, the Maryland Court of Appeals held that the conclusions drawn from the CBLA technique did not fit the standards of admissible evidence under the federal precedent of Frye v. United States, 293 F. 1013 (D.C. Cir. 1923) as adopted in Reed v. State of Maryland, 283 Md. 374, 391 A. 2d 364 (1978). That standard held that a forensic science had to be generally accepted within the relevant scientific community to be admitted as evidence. Gemar Clemons was arrested in 2002 and tried in connection with the robbery and shooting death of a man, and CBLA testimony was offered by FBI analyst Charles Peters in connection with the case. According to the opinion of the Maryland high court, Peters testified in detail as to the bullet manufacturing process as well as the CBLA process and its alleged accuracy, and to his experience with both: “Q: And how many bullet lead analyses have you performed over your years with the FBI?” “A: I have done tens of thousands of analyses.” ...”Q: And have you ever testified as an expert in the field of comparative bullet lead analysis?” “A: Yes.” “Q: Approximately how many times?” “A: Over 80.” At this juncture, the defense attorney objected to Peters' qualification as an expert witness, to which the trial court judge replied, in part: “...So I'm going to admit him as an expert in this field because he's been in this field for a long time. He's done tens of thousands of these analyses. He's been around since the '70s... And, the fact that there are some challenges, it doesn't mean that it's such a novel and scientific kind of test that the court finds as a matter of law that it shouldn't be submitted to the jury.” The Maryland high court pointed out several exceptions, including rebuttal testimony by Tobin, a 1991 study presented at the International Symposium on the Forensic Aspects of Trace Evidence, a 2002 study by Tobin and three other scientists, and a current article in the Oklahoma City Law Review. “Thus, as these studies indicate, the assumption that an ingot or vat of lead is homogenous as required for CBLA to be valid is not generally accepted by the scientific community,” the court wrote. The high court went on to point out that the FBI had discovered flaws in the underlying assumptions on bullet distributions. “Furthermore, at least one study conducted by Dr. Robert D. Koons, a research chemist with the FBI Laboratory in Quantico, Virginia, and Dr. Diana M. Grant, a forensic examiner with the FBI Laboratory in Washington, D.C., observed an error rate, which includes false positives and negatives, of twenty-five to thirty-three percent,” the court wrote. “Although scientific unanimity is not required to satisfy the Frye-Reed test's requirement of general acceptance... it is clear that a genuine controversy exists within the relevant scientific community about the reliability and validity of CBLA. “Based on the criticism of the processes and assumptions underlying CBLA, we determine that the trial court erred in admitting expert testimony based on CBLA because of the lack of general acceptance of the process in the scientific community.”