Free Minds & Free Markets

Stossel: Junk Science Locks Up Innocent People

DNA testing reveals that long-used forensic methods are error-riddled.

TV shows like CSI, Law & Order, and NCIS depict incredible technology for identifying criminals.

In one NCIS scene, a 3D hologram identifies a person's teeth, precisely matching the killer's bite to a bite mark on a victim. "A little 3D magic for clarity and I give you—the killer's incisors!" NCIS character "Abby" announces proudly.

John Stossel loves crazy Abby, but notes that in the real world, court-approved experts reach similar conclusions—without good science to back them up.

Alfred Swinton was convicted of murder after "bite mark expert" Dr. Gus Karazulas said Swinton's teeth matched a bite on a victim.

Karazulas told the TV show Cold Case Files: "We look at the evidence and we make sure that if we are going to make a decision it's gonna be a truthful decision."

But a decade later, DNA testing showed it was not Swinton's DNA on the bite mark. He was released after 18 years in prison.

To his credit, Karazulas now admits he was wrong. "Bite mark evidence is junk science," he tells us by email. He testified to that in court and resigned as a member of the American Academy of Forensic Sciences.

The verdict was wrong because bite mark analysis has never been scientifically proven, says Chris Fabricant of the Innocence Project, which represented Swinton.

Fabricant tells Stossel that bite mark analysis "is similar to you and I looking at a cloud, and then I say, 'John, doesn't that cloud look like a rabbit to you?' And you look at it and say, 'yeah, Chris, I think that does look like a rabbit.'"

Many experts agree. "Bite mark analysis is a subjective method," reads a 2016 report by the President's Council of Advisors on Science and Technology, which notes that even the most positive studies on it find an error rate of 1-in-6.

NYPD detective Harry Houck tells Stossel that he agrees bite mark analysis has problems, but that it can be used as one piece in a puzzle to convict people.

Other methods have flaws, too. Even fingerprint analysis sometimes goes wrong, as do carpet fiber evidence, gun tracing, and hair matching.

Stossel asks Fabricant: "Why do judges admit this stuff? Why don't defense attorneys get it thrown out?" Fabricant answers: "We all went to law school because we don't know science, we don't know math. And if somebody comes in in a white lab coat...that's good enough for government work."

That shouldn't be the standard, says Stossel. Jurors and judges should be much more skeptical of "scientific" evidence.

Subscribe to our YouTube channel.

Like us on Facebook.

Follow us on Twitter.

Subscribe to our podcast at iTunes.

The views expressed in this video are solely those of John Stossel, his independent production company, Stossel Productions, and the people he interviews. The claims and opinions set forth in the video and accompanying text are not necessarily those of Reason.

Editor's Note: We invite comments and request that they be civil and on-topic. We do not moderate or assume any responsibility for comments, which are owned by the readers who post them. Comments do not represent the views of or Reason Foundation. We reserve the right to delete any comment for any reason at any time. Report abuses.

  • Don't look at me.||

    The science is settled.

  • R. K. Phillips||

    Yeah, ask the FBI about bullet analysis.

  • TrickyVic (old school)||

    ""Alfred Swinton was convicted of murder after "bite mark expert" Dr. Gus Karazulas said Swinton's teeth matched a bite on a victim."'

    Where did Karazulas get his training? Haynes?

  • gaoxiaen||

    We do't need no stinking training/badges.

  • John||

    Shame on judges for this. This science only got before juries because judges were too stupid and gullible to stop it. It is incumbent upon judges to understand the evidence and claims being presented to them. Sadly, judges were too often too lazy or too stupid to see through this nonsense and just let the state introduce anything they wanted, provided a guy in a lab coat was saying it.

  • R. K. Phillips||

    Did you say that it's incumbent upon a judge to understand the evidence? In my neck of the woods, it appears that it's incumbent upon a judge to remain the incumbent, so the will make whatever decisions work to that end.

  • TSTB||

    I am an expert on experts and you can trust me when I say; Never trust experts.

  • $park¥ leftist poser||

    Chris Fabricant

    That name looks made up.

  • Juice||

    All names are made up, but it's totally a name from Blade Runner.

  • No Longer Amused||

    A perfect opportunity to discuss the junk science of "global warming"...

  • Jerryskids||

    "Why do judges admit this stuff? Why don't defense attorneys get it thrown out?"

    I remember years ago reading (might have been in OMNI magazine if you remember that one) about the gatekeeper function of judges in admitting expert witnesses to testify becoming more difficult as science and technology advanced and the average judge had no qualifications to judge what was and was not accepted science. I don't remember if they offered a policy prescription, or if they did what it was, but what struck me in the article was how casually they mentioned the junk science of so much of forensics, as if everybody knows that stuff's a load of crap. A big part of the problem is that the FBI is at the forefront of this stuff and they self-certify that their bullshit isn't bullshit and the system trusts the FBI to know bullshit when they see it.

  • John||

    The FBI is a big part of the problem. How many state district judges in Peoria are going to tell the big bad FBI to get stuffed? And how many will be re-elected after they do?

    Prosecutors stopped caring if they got the right result if they ever did care, which is very debatable. In an ideal world, prosecutors would not have introduced junk science in the first place. But we don't live in such a world. And prosecutors at every level didn't care if this stuff was valid and just believed any "expert" who told them what they wanted to hear.

    It is a total mess. I don't know how you fix it to be honest.

  • Unicorn Abattoir||

    Prosecutors stopped caring if they got the right result if they ever did care, which is very debatable.

    Exhibit A - Janet Reno as AG of Florida.

  • D-Pizzle||

    Exhibit B - Martha Coakley as AG of Mass.

  • sarcasmic||

    It is a total mess. I don't know how you fix it to be honest.

    End immunity for starters.

  • NotAnotherSkippy||

    And make all awards come out of pensions instead of whacking taxpayers.

  • Sometimes a Great Notion||

    Don't know if you've seen Evil Genius on Netflix about the collar bombing case, but one FBI agent has one of the suspects do a polygraph and when he doesn't get the results he wants; he says they can be beaten. At that point I yelled at the TV ,"why then waste your fucking time if you knew how fucking shitty those things are!" Good watch by the way.

  • Illocust||

    It seems one of the best thing justice advocates could do, is provide a database of properly vetted forensic techniques and their limitations to judges. 60 year old judge's don't havethe expertise to know when a prosecutor is forensically bullshitting them.

    Having a resource of good evidence acquiring methods could give them the yes/no answer they need to decide.

  • DaveSs||

    Expert: The justice advocates are wrong, I am an expert in this. Trust me
    Judge: Sure thing

  • Citizen X - #6||

    I think you found the flaw in Illocust's plan.

  • Rossami||

    The problem is not the existence or lack of "properly vetted forensic techniques". The problem is that perfectly valid techniques are being applied to situations for which they were never designed. The problem in most situations is that these techniques are being used for positive identification when they are only valid for negative identification.

    Consider bite-mark analysis. There is no possible way a bite-mark can say person A did it and no other. A bite-mark could, however, say that person B did not do it if, for example, the bite marks clearly shows two incisors but person B lost his left incisor in some prior accident.

    Likewise with blood-type analysis. If you've got a critical blood sample from the scene that tests as O-positive, you can't say which of the roughly half the population with that blood type did it but you can definitely clear the suspect with B-negative blood.

  • Paloma||

    Or "expert" psychologist testimony, e.g. recovered memory.

  • R. K. Phillips||

    My S/O is a shrink, and laughs at some of the stuff that's tried in court.

  • Forrest Grump||

    You generalize: SOME 60-year-old judges might not have the expertise. SOME 40-year-old judges also might fall into the same basket. But judges in both groups also might have science degrees that would make them more than qualified to be as astute as anyone else.

  • ||

    It looks like science.

    I've seen photos of UFOs, ghosts, the Loch Ness monster, and Bigfoot. All of them were more credible than the photos I've seen of science.

  • Citizen X - #6||

    I once saw a picture of science coming out of a bar at 11 AM, drunk as hell.

  • sarcasmic||

    That wasn't science. That was me.

  • Johninsd||

    Seems that most of what the legal system does is bullshit. For example, drug dogs - the Supreme Court has said that the drug dogs can do a judge's job and issue a search warrant - maybe we should replace all the judges with dogs - they work cheap and are mostly honest - their handlers are liars but we can't have everything -

  • DaveSs||

  • Johninsd||

    Sums it up perfectly -

  • D-Pizzle||

    I'm convinced that most "drug" dogs are not trained to sniff out drugs. They are trained to "signal" on a subtle cue by their handler(s). That the police are often right in these cases does not prove the efficacy of the dogs.

  • Johninsd||

    That's exactly what's happening, though often the handlers cue the dogs subconsciously. And I don't know that the police are "often" right - they may keep statistics that show them to be mostly right but they rarely even file a report when a dog falsely alerts - I learned this from personal experience following a pretext stop and dog justified illegal search ( during which nothing illegal was found). When I wrote to the PD and asked for a copy of the incident report the chief hisself called me and said that in Texas, no report is filed unless an arrest or citation happens. Awfully convenient - don't count the false alerts but brag about the true ones - can claim 100% accuracy when in truth the accuracy is unknown. The few controlled studies that have been done have shown the dogs' accuracy to be dismal - look up the study done at UC Davis. The dog industry and the police and their enablers have pretty much halted any real testing or training requirements as they are fully aware that it's all bullshit but it lets them intimidate and humiliate people at random and that seems to be what they live for.

  • Enjoy Every Sandwich||

    John Stossel loves crazy Abby

    Not on account of the science, I suspect.

  • sarcasmic||

    The point of a trial is not to get to the truth, but to get a conviction. If the truth mattered then cops would be held accountable when they testilie. I imagine it's the same thing with junk science. Any amount of bullshit is acceptable if it gets the desired result.

  • loveconstitution1789||

    If one scientist says one thing about bite marks and another scientist says another thing about bite marks, you literally have reasonable doubt right there. If the case was hinged on that scientific fact, NOT GUILTY should be the verdict.

    The problem is that people on juries dont understand what "presumption of innocence" means and "guilt beyond a reasonable doubt"

  • Paloma||

    If the glove don't fit, you must acquit.

  • Mickey Rat||

    Here I thought being skeptical as to the claims of scientists was hostility to science itself.

  • mtrueman||

    "Stossel: Junk Science Locks Up Innocent People"

    Maybe because these innocents are too cheap to hire their own scientific experts. Claus Von Bulow was not afraid to splash his money around and was freed on the strength of their undermining the case against him.

  • RevengencerAlf||

    If only every defendant had thousands of dollars of their own money free and available to spend on legal experts.

    Do you seriously think people don't put up a robust defense to felony charges because they're "too cheap" to and not because they CAN'T AFFORD IT? Jesus Tapdancing Christ that's insane.

  • PaulTheBeav||

    Based on that mustache I've conclusively determined that Stossel is either a 70s porn star or an 80s private detective. Possibly both.

  • Hank Phillips||

    Pseudoscience made LSD illegal and has by now damaged the lives of thousands... and robbed millions of potential benefits. It made marijuana illegal and got Thalidomide on the market as a replacement. Even today purveyors of pseudoscience struggle to preserve the perpetration. If Republicans they seek to ban harmless drugs, if Dems they seek to ban electrical energy. Both want men with guns to make their bans stick.

  • ||

    "We all went to law school because we don't know science, we don't know math. And if somebody comes in in a white lab coat...that's good enough for government work."

    There's so much wrong with this.

    So much.

  • Alan@.4||

    I find myself curious as to the following. When the "experts" screw up, and the innocent who are later absolved suffer, do the experts or might EGGSPURTS be a more accurate description ever suffer real consequences? They do not? How does that come to pass?

  • Galane||

    Bite marks can only have any relevance when the biter has teeth that are significantly out of place from perfect. Missing, twisted, badly chipped etc.

    Telling if a person has an over or under bite from bite marks isn't possible. The Human jaw can move back and forth quite a bit, and soft squishy surfaces (like the human body) will move around when bitten, even if the bite-ee is dead.

    With teeth that are naturally perfect or have been jacked around with braces and other devices to be perfect, forget being able to have any definitive identification from bite marks.

  • RevengencerAlf||

    Even then the relevance is purely exclusionary. It's just too imprecise for anything else. You can look at a bite mark with a crooked tooth and probably eliminate everyone with near-perfect teeth, but there are hundreds if not thousands of people within an hour's driving time of the suspect who probably have a similarly crooked tooth.

    The big problem here is that a lot of evidence that should only have ever been used to eliminate suspects is now being purported to confirm them.

  • Amir Najam Sethit||

    Thank you sharing the article

  • Mark22||

    Looking at that picture, what comes to mind is...

    "Sorry, John, I'm not into that kinky stuff."

  • Vindex||

    Junk science also awards billions in damages by gullible juries in civil litigation. Just look at these recent awards claiming talcum powder contains asbestos and causes cancer, for example. Simply preposterous!


Get Reason's print or digital edition before it’s posted online