Subscribe for 17¢ / day

Montana Attorney General Mike McGrath said Thursday he does not believe other criminal defendants were wrongly convicted based on the hair analysis work done by former state Crime Lab director Arnold Melnikoff.

"I'm comfortable that we've done more than anyone wanted us to do," McGrath said at a Missoula meeting of the Crime Lab advisory board. "I've reviewed all of these cases now and I don't believe we've got any more people who are wrongly convicted based on Melnikoff's work."

The work of Melnikoff, who worked at the lab from 1975 to 1989, came under heightened scrutiny in October 2002, when Jimmy Ray Bromgard was freed after serving 15 years for a rape that DNA evidence proved he did not commit.

Melnikoff testified about hair evidence in the case, using a self-generated theory to conclude that there was a 1-in-10,000 chance that the hair belonged to someone other than Bromgard. Similar testimony in a Richland County case led to another rape charge being dropped against Paul Kordonowy in May 2003.

The state Department of Justice and the Crime Lab commenced an investigation into Melnikoff's work, a probe that eventually reviewed every case in which Melnikoff generated a report on hair analysis. All together, the cases of about 30 current prison inmates and 244 older cases were reviewed.

"I'm confident based on this that we don't have any other individuals wrongly convicted," McGrath said.

Although McGrath's report to the advisory board was well-received, it didn't sit well with University of Montana law professor Jeff Renz, who argued that the lab should have done a scientific audit of Melnikoff's actual lab work.

"There's only one way to know that, and that's to do an audit," Renz said.

Renz's comments clearly angered McGrath, who claimed that the review was more thorough than a scientific sampling because it included every hair case Melnikoff ever dealt with.

The Innocence Project, a New York nonprofit organization that played a role in Bromgard's release, also criticized the report.

"The review itself violated the most fundamental principles of science and justice," attorney Peter Neufeld told the Associated Press on Thursday.

However, McGrath and others on the advisory board felt the review was thorough. And, as board members noted, the problem associated with Melnikoff's work has not generated a host of appeals or complaints.

Only two current prison inmates have filed for any sort of review, and one's petition for relief has already been dismissed.

Missoula County Attorney Fred Van Valkenburg, a member of the advisory board, said the state had done a thorough job of reviewing Melnikoff's work, and that's all it needed to do. If former defendants or their lawyers want to raise questions, raise them, he said. The state is better served by focusing its resources on cases currently in the judicial system, Van Valkenburg said, rather than reviewing "what ifs" from cases 20 years old.

"If you want to be a do-gooder group and get out there and work for these people, go for it," Van Valkenburg said in comments clearly targeted toward Renz.

When problems first surfaced in Melnikoff's work, the first cases reviewed were those of people currently in prison. What investigators were primarily looking for was cases where Melnikoff testified at trial and used his probability theory to state the likelihood that a hair belonged to someone involved in a case. Melnikoff's theory often involved a multiplier effect, which he based on his experience. Occasionally that led Melnikoff to set out lofty odds like those found in the Bromgard and Kordonowy cases.

After the prison cases were reviewed, McGrath hired a former law enforcement officer, Sam McCormack, to research all of Melnikoff's cases involving hair analysis. McCormack found and reviewed 244 cases, looking at records from the Crime Lab, prosecutors and courthouses around the state. Of that total, McCormack found 118 cases where Melnikoff made what's called a "hair association." That means he found some hair-related connection between a defendant and a victim. Melnikoff testified at trial in 19 of those cases, and 12 of those 19 defendants were found guilty.

Nine of those 12 defendants had felony convictions, and eight of nine appealed their cases to the Montana Supreme Court. None raised issues regarding Melnikoff's testimony.

"In the same eight cases, five were cases in which the defendant either claimed to have had consensual sex with the victim or admitted the victim had been in his bed," McGrath wrote in his report to the advisory board. "Since the purpose of hair association evidence is to place the defendant and the victim together, in these five cases Melnikoff's testimony is not pertinent to the ultimate issue of fact."

In the other cases, McGrath said, the defendant was either known to the victim or the Montana Supreme Court found "substantial evidence" to corroborate the victim's eyewitness identification.

"We were trying to find people who were wrongly convicted," McGrath said. "I am convinced, based on this analysis Š that nobody was."

Even so, as the board reviewed the 19 cases where Melnikoff testified, board members found three where they wanted to see a little more done. Randi Hood, a defense attorney who serves on the board, actually found one of her old clients on the list and agreed to re-review the case.

If past defendants want to review their cases, McGrath said the justice department would accommodate them, but he believes the state has done all it needs to do to make sure people weren't wrongly convicted based on Melnikoff's work.

Melnikoff's troubles continue, however. He has worked for the Washington State Patrol crime lab, but his casework there has been under investigation. A review of 99 cases Melnikoff worked on from 1999 to 2002 found some sort of problem in 29 cases.

"The review of the cases did not necessarily reflect any mistakes that would have changed the basic conclusions drawn from the analysis," a report written by Washington State Patrol Capt. Brian Jones stated. "It is just that often the work product was weak or unsupported by insufficient data to reach clear conclusions Š"

The report concluded that Melnikoff, who is on paid leave, should be fired.

Reporter Michael Moore can be reached at 523-5252 or 370-3330, or at

You must be logged in to react.
Click any reaction to login.