A court finally took notice of my reporting—but for all the wrong reasons.
A little while ago, a reader sent me a Nevada Supreme Court ruling. In the unanimous opinion, the justices cited my work from the mid-2000s criticizing the use of bite-mark identification in criminal trials. For a journalist who writes and reports on the criminal-justice system, getting cited in a court opinion can be gratifying. You want to feel like your work matters; you aren’t shouting into a void.
Even so, a citation isn’t always a win. Of the dozen-odd times I’ve seen my work cited by a court, all but a few were either in a dissenting opinion or in the majority but followed by a nevertheless—a minor concession en route to an opinion that comes down the other way. That’s frustrating, but not surprising. My beat is reporting on flaws in the criminal-justice system, and it’s an institution that can be reluctant to admit its mistakes.
But the Nevada decision the reader flagged for me was something different—and “frustrating” doesn’t quite capture what’s going on. Maddening is more like it, because in this case my work was used to justify an ongoing injustice.
Over the past 15 years, I’ve written at length about how, despite near-universal agreement in the scientific community that matching marks on human skin to a suspect’s teeth is forensic quackery, the courts have continued to allow this sort of analysis to be used in criminal trials. Dozens of people convicted or arrested because of this “science” have been exonerated.
The maddening part of the Nevada opinion, which came down in December 2022, is that the justices appeared to agree with what I’ve reported, and went on to argue that the defendant should have noticed my reporting years ago. Because he failed to do so then, he is prohibited from using it in his appeal now.
That’s all bad enough. Here’s the punch line: While the Nevada Supreme Court says that criminal defendants should have been aware of the reporting I and others were doing on bite-mark analysis back in the mid-2000s, some defendants did notice it then, and begged the courts to do something. The courts ignored them—including the Nevada Supreme Court.
The defendant in the Nevada case was David Middleton, a former police officer who in 1997 was convicted and sentenced to death for the murders of two women in Reno. At Middleton’s trial, prosecutors called the forensic dentist Ray Rawson to testify. Rawson told the jury he had matched a bite mark on the breast of one of the victims to Middleton’s teeth. He also said he could tell that the bite had been inflicted while the woman was still alive, and that it had been painful.
This was significant: Prosecutors argued that biting the victim while she was still alive was a form of torture—a factor in establishing one of the aggravating circumstances they needed to seek the death penalty. Middleton is still on Nevada’s death row.
There were two major problems with relying on Rawson’s testimony. The first was that bite-mark analysis is pure charlatanism. The field relies on three core premises: that every human being has unique dentition that leaves unique bite marks, that human skin is capable of recording bite marks in a way that preserves their uniqueness, and, finally, that analysts like Rawson can use the first two premises to match a bite mark to one person, to the exclusion of all other people.
I’ve never come across any scientific research that supports these premises. And the research I have seen strongly suggests that they simply aren’t true.
The second major problem was Rawson’s track record. The dentist—a prolific expert witness and a former state senator—was an early practitioner of bite-mark identification, and helped write the guidelines that many bite-mark analysts used. You might call him a pioneer, if a person can pioneer a field that has no business existing. (Rawson could not be reached for comment.)
Rawson is among a handful of expert witnesses whose testimony has helped persuade more than one jury to send an innocent person to prison. In the 1985 Wisconsin murder trial of Robert Lee Stinson, Rawson testified in support of the conclusions of another dentist who had said that tooth marks left on the victim were a match to Stinson’s teeth. In a 1986 opinion upholding Stinson’s conviction, a unanimous Wisconsin Supreme Court concluded that “the credibility of the witnesses” and “the weight of the evidence” were persuasive enough “to exclude to a moral certainty every reasonable hypothesis of innocence.” Stinson was exonerated by DNA testing more than 23 years later.
Less than a decade after Stinson’s trial, Rawson testified in the Arizona trial of Ray Krone. According to a local news article published in 2005, Rawson told the jury that because of Krone’s unusual teeth (reporters dubbed him “the Snaggletooth Killer”), he had no problem matching them to the bite mark on the breast of a murdered waitress. Krone, too, was later exonerated by DNA testing.
You might think that with this history, any court that had allowed Rawson to testify would be anxious to fix its mistakes. You’d think there might be some embarrassment, even a rush to review old cases to ensure that Rawson’s testimony hadn’t been used to convict other innocent people. But that isn’t what the Nevada Supreme Court did.
David Middleton’s case is now at a stage known as post-conviction, which comes after a prisoner has exhausted his normal appeals. From what I’ve seen, this is the phase in which prisoners are most likely to find exculpatory evidence, but it’s also the phase in which getting back into court is most difficult.
In most states, for a court to even consider new evidence, a prisoner must show that the evidence could not have been discovered at the time of trial and that it would likely have led to an acquittal. They also face strict deadlines, based on when a court thinks the new evidence should have been found.
It’s this last requirement that trips up many, and it’s what doomed Middleton in his recent litigation. In most states (and the federal courts), any claim based on new evidence must be brought within a year of when the evidence could have reasonably been discovered. Nevada’s rule is more ambiguous; prisoners must file within a “reasonable time.”
These deadlines are difficult to navigate even for skilled lawyers, and even when dealing with clearly new evidence that can be tied to a specific date, such as a confession. But exculpatory evidence rarely comes out in big, revelatory chunks. More often, defense teams find evidence in pieces, then put the pieces together. Another problem they face is that most states offer only one shot at a post-conviction request to reopen a case.
So these prisoners and their lawyers face a balancing act: If they file a claim before they’ve accumulated enough new evidence to decisively prove their client’s innocence, the court will reject it, and their client will be barred from using any of that evidence again—even if they later find other evidence to corroborate it. If they file too late, they’ll be rejected for missing the deadline.
These dilemmas get even more perplexing when the issue involves “science” that has since been discredited. For about a century now, criminal courts have been wildly permissive, allowing quackish forensic disciplines, dubious experts, and theories untested by scientific method into criminal trials. In the past couple of decades, the scientific community has begun to expose the failures of these fields, but the process of reaching a scientific consensus is an incremental one. A study casts some doubt. More studies follow. A scientific body may review the literature and issue an opinion. A series of exonerations or proficiency tests—in which forensic examiners’ analytical skills are tested on cases in which the “ground truth” is known—might further chip away at the field’s credibility.
How can someone convicted with bogus expert testimony know when his one-year deadline to file a claim based on “new evidence” begins? Is it after the first skeptical study? The fifth? The tenth? These prisoners and their attorneys have no choice but to guess when judges might finally be convinced.
The consequences for guessing wrong are dire, as Middleton learned.
The Nevada Supreme Court now seems to agree with scientific bodies such as the National Academy of Sciences that bite-mark analysis is not a credible discipline. The court cites several cases in which people convicted with such evidence were later exonerated, and favorably quotes papers concluding that the practice is “highly questionable,” and that the science behind it is “murky at best.” The court also doesn’t raise any defenses of the discipline. But when did the court reach this view?
The justices don’t say. They say only that Middleton was too late. “Although Middleton represents that [the] changed landscape with respect to bite-mark-identification testimony came to a head around 2013 or 2014,” the court wrote, “we are convinced he could have raised this claim earlier.”
It’s here that the court cites my work. To support their conclusion that Middleton waited too long, the justices refer to numerous media investigations, academic papers, and scientific bodies that have raised concerns about bite-mark analysis. One of the oldest of these is a 2009 article I wrote for Reasonmagazine.
Reason is a fine magazine, and it publishes terrific reporting on the criminal-justice system. But for a state supreme court to punish a prisoner because, 15 years ago, his attorneys failed to notice an article in a monthly libertarian magazine with a circulation of about 50,000 seems a bit much.
Anyway, if the justices had a better grasp on the history of this issue, they could have gone back a lot further than 2009. Critics have been noting the inherent subjectivity of bite-mark analysis since it first caught on, in the 1980s, but the courts have mostly ignored them.
Despite unanimous agreement by scientific bodies such as the National Academy of Science, the Texas Forensic Science Commission, and the President’s Council of Advisors on Science and Technology that bite-mark identification is hokum, every single defendant I’m aware of thus far who has asked a court to exclude bite-mark evidence because of the lack of scientific validity has lost, including rulings in Indiana (2014 and 2015), Pennsylvania (2017 and 2018), and Ohio (2018), well after such organizations started criticizing the field.
Incredibly, in at least three states (Wisconsin, Mississippi, and Connecticut) the controlling case establishing the admissibility of bite-mark identification—the case that prosecutors cite when they want to use this sort of testimony—is a case in which the defendant has been exonerated.
The Nevada Supreme Court’s ruling in Middleton’s case reasoned that defendants convicted with bite-mark testimony should have challenged the validity of such evidence years and years ago. At least three Nevada defendants did exactly that—and the Nevada Supreme Court rejected them.
The first two cases were both in 1982. In those cases, the justices evaluated other evidence to convict the defendants. But the court also refused to consider the validity of bite-mark identification, which allowed prosecutors to continue presenting it to juries.
The court didn’t revisit the issue until 2016, when a defendant, Thomas Collman, challenged the credibility of bite-mark evidence, noting in particular that other dentists believed Rawson’s findings in his case to be “bogus.” As in the Middleton case, Rawson not only claimed to match the bite mark to Collman, but claimed that the bite had been inflicted prior to death and was painful, which helped prosecutors seek the death penalty. (This seems to be typical of how prosecutors use Rawson’s work—he gave similar testimony in another death-penalty case in Arizona.)
Collman argued that the dentist should never have been allowed to testify at trial. He also argued that his trial attorney was deficient for not mounting a stronger challenge to Rawson’s testimony and bite-mark analysis, and that prosecutors violated their Brady obligations—the requirement to provide any exculpatory evidence to the defense—when they failed to disclose that several forensic analysts they had consulted prior to the trial expressed doubts about the reliability of Rawson’s testimony.
The Nevada Supreme Court rejected all of those arguments, reasoning that Collman would likely have been convicted on other evidence. And they also again refused to specifically rule on the challenges to bite-mark identification, and to Rawson.
This time, there was one dissent. Justice Michael Cherry chided his fellow justices for being so dismissive of the fact that a murder conviction and possible death sentence had been won with the help of a testimony relying on discredited methods.
The chronology here is crucial. The Middleton and Collman rulings were just six years apart, and the facts were remarkably similar. They involved the same dubious forensic methods, the same dubious analysis, and the same strategy of using Rawson’s testimony to help obtain a death sentence. Yet although the court rejected challenges to Rawson’s testimony and bite-mark identification in 2016, by 2022 the court appeared to accept that both were no longer credible. But the justices did so while also ruling that it was now too late for any of this to matter.
There’s no point in time in which Middleton could have won. If he had filed in the late 2000s, not only would he have lost; he would’ve been barred from raising these issues later. The court is essentially saying that both Rawson’s expert testimony and bite-mark evidence should never have been allowed into criminal trials. But it’s also saying that nothing can be done about it.
If they had wanted to, the Nevada justices could have seized on several developments in the bite-mark debate over the past decade to put prisoners convicted with bite-mark evidence on notice—to say, “We’ve changed our minds on this, and you’re now on deadline to file.” In 2015, for example, the American Board of Forensic Odontology—the professional organization for bite-mark identification—gave a proficiency test to its members. The results were damning: In the majority of cases, the certified members couldn’t even agree on which of the test photos depicted human—rather than animal—bites. The ABFO subsequently changed its guidelines in 2016, recommending that bite-mark analysis be used only to “exclude” suspects, not to “match” them, as was done to Middleton and dozens of others.
That same year, the Texas Forensic Science Commission—formed after revelations that the state had likely convicted innocent people, including a man executed in 2004, on the basis of junk arson science—recommended that bite-mark analysis be barred from the courtroom. Also in 2016, the President’s Council of Advisors on Science and Technology made a similar recommendation. Most recently, in October 2022, the National Institute of Standards and Technology concluded that “forensic bitemark analysis lacks a sufficient scientific foundation because the three key premises of the field are not supported by the data.”
Instead of giving prisoners an explicit out, however, the Nevada Supreme Court shifted its stance so quietly that you’d have to be, say, a journalist who is slightly obsessed with this stuff to even notice it.
It’s possible—probable even—that the justices weren’t fully aware of what they were doing. Still, there’s no mistaking the ruling’s impact: After decades of failing to keep bite-mark identification out of criminal trials, the court finally seems to have conceded that it isn’t a credible field—only to slam the courthouse door onthe people such evidence has sent to prison.
Bite-mark cases are relatively rare. The total pool of convictions is probably in the hundreds. But other scientifically dubious areas of forensics, such as firearms identification and blood-spatter analysis, are quite common. Getting the courts to acknowledge the inadequacies of such fields—and of experts who have testified hundreds of times—is going to be even more difficult than it has been with bite-mark analysis, especially because doing so would mean revisiting thousands of old convictions.
Many of the laws that make post-conviction claims so difficult to win today were passed in the 1980s and ’90s, an era in which politicians and law-and-order pundits claimed that too many violent offenders were filing frivolous appeals to delay their execution or had duped liberal courts into releasing them “on a technicality.” Although there were certainly some examples of this, neither practice was widespread. More commonly, the courts had stepped in because in that era’s frenzied rush to convict and carry out executions, states had taken unconstitutional shortcuts, provided inadequate indigent defense, and tolerated police and prosecutor misconduct.
Although the onset of DNA testing in the ’90s showed that many of these prisoners were, in fact, innocent, the laws restricting their access to the courts remained. The result is that many prisoners convicted at the conclusion of trials tainted by forensic evidence now have no way back into court, not because the evidence of their innocence or the unjustness of their trials isn’t persuasive, but because artificial deadlines and procedural roadblocks prevent the courts from ever considering the merits of that evidence.
It’s hard to not be overcome with cynicism when, after you’ve spent years writing about these issues, the courts have finally started to notice your work—only to weaponize it in the service of one of those technicalities.
Radley Balko is an investigative journalist who writes about civil liberties and the criminal-justice system on his Substack newsletter, The Watch.
https://www.theatlantic.com/ideas/archive/2024/03/courts-junk-science-forensics-evidence/677739/
