John Oliver on unreliable forensic science

Forensic science is not a magic bullet, no matter what popular crime shows try to tell you. John Oliver explains why on the latest Last Week Tonight.

In the world of television, crime is big business. Some of the most successful and long-running shows currently on TV, like Law & Order: SVU (currently in its 20th season) or NCIS (now in its 15th season), use crime as their bread and butter. Or, rather, these shows focus on the forensic experts who, with their expensive-looking backlit equipment and quipping skills, catch criminals every week.

Yet, real-life forensic science is far more complex than the fancified television version. And comparisons between real life and the television version can make things even direr in a court of law. For instance, prosecutors often complain about the “CSI effect”, wherein jurors expect rock-solid forensic evidence in every case. However, a federal study quoted by Oliver found that expert witnesses “[go] far beyond what the relevant science can justify”. This mix of entertainment-fueled expectation and sometimes shaky delivery can make for a dangerous situation.

“It’s not that all forensic science is bad,” said Oliver. “But too often, its reliability is dangerously overstated”.

Here’s the full clip. (Warning: there’s some NSFW language.)

For example, take the phrase “a reasonable degree of scientific certainty”. Expert witnesses are often asked about this certainty in court, but what does it actually mean? It sounds good, but in fact there is no objective standard for “reasonable certainty”.

Even more importantly, said Oliver, “there are people behind those numbers”. For all that people get excited about forensic evidence, that false confidence can ruin lives. Take Santae Tribble. He served 23 years for a murder he didn’t commit, all because an analyst claimed that his hair matched a sample collected at the crime scene. Tribble was eventually exonerated by DNA analysis, which proved that none of the hairs were his. Also, FBI analysts in the original case had missed the fact that some of the hairs actually came from a dog.

The problem with bite marks

It’s not just hair samples that cause trouble, though. There’s uncertainty in many other avenues of inquiry, like bite mark analysis. “In the real world, bite mark analysis is highly subjective and unreliable,” said Oliver.

Dr. Michael West, a former proponent of bite mark analysis, now says that it should not be used in court. West, by the way, is directly linked to a number of wrongful convictions thanks to his phony forensics. Despite the evidence, however, courts still consistently rule bite mark evidence to be admissible. That’s thanks to already established precedent. Never mind that such a precedent shouldn’t have been allowed to take root in the first place.

Often, the people who decide whether or not evidence is scientific enough to hold up in court are not scientifically trained themselves. “Trials can often be a situation where no one really knows what they’re doing,” said Oliver.

Now, this isn’t to say that we should throw all forensic techniques out the window and rely on spectral evidence or the coke-fueled deductive reasoning of some dude in a deerstalker hat. DNA evidence has been used to exonerate people. In fact, Santae Tribble is now a free man thanks to DNA analysis.

Brandon Mayfield

Still, it’s important for jurors and anyone else in court to understand that no forensic evidence is infallible. What about fingerprint analysis? For yet another example,  it turns out that one in 306 fingerprint analyses can result in a false positive.

This, too, can have dramatic real life consequences. In one case, an American named Brandon Mayfield was accused of participating in the 2004 Madrid train bombings — despite the fact that he had never been to Spain. Experts claimed that they had matched fingerprints from a bag at the Madrid scene to Mayfield, who lives in Oregon.

Mayfield was later exonerated after experts determined that the fingerprint match was a false positive. That was only after Mayfield had been arrested by U.S. authorities. He’s since received a formal apology from the government and a $2 million settlement.

So, some fingerprints or partial fingerprints can be so alike that even seasoned experts can be fooled. Oliver says that this may even lend credence to his theory that there is only one Olsen “twin”. Said Oliver: “She’s just moving very fast, back and forth. She confuses your eye”.

Even DNA — “the gold standard in forensic science” — is not necessarily a perfect tool. DNA is actually quite fragile and can degrade under the right conditions. Lower-quality DNA samples can make the conclusions of expert witnesses and lawyers far less reliable.

It’s can also be difficult to determine which DNA samples came from a given source. For a particularly gross example, imagine you have to tease out unique DNA samples from a large pool of blood. This kind of cross-contamination makes producing a good analysis spectacularly difficult, if not practically impossible.

Subconscious bias

Meanwhile, forensics labs don’t always work independently. Many actually have connections with local law enforcement. This means that analysts may know more about a case than they strictly should. A subconscious bias to work on the side of the “good guys” can nudge analysts to indulge in bias when they make reports or testify in court.

It’s not all terrible. Some states have made some big changes to combat these issues. Texas — yes, seriously, Texas — passed “junk science” laws to address this very issue.

There’s also the National Commission on Forensic Science. The NCFS was founded to work with the Department of Justice to work against faulty forensic evidence.

Keith Harward, a man who was imprisoned but eventually exonerated, even spoke to a large group at an NCFS meeting. During that appearance, he said that “How many people have to be wrongly convicted before they realize: this stuff’s all bogus. It’s all made up.”

In April, Attorney General Jeff Sessions shut down the commission. The Department of Justice attempted to reframe this closing as part of “new initiatives to advance forensic science”, but it’s awfully hard to feel confident about the move.

Oliver argued that thanks in part to Sessions’ decision and a lack of scientific rigor in courts, we may, in fact, be going backward. And while innocent people are locked up or worse for a crime, the actual guilty party walks free.

What to do?

“We should, at the very least, do more to educate potential jurors” about these literal life-or-death problems, he claimed. To this end, Oliver and company aired a pilot for a new kind of crime drama. Except, instead of being a quip-fest or showing off inaccurate and oftentimes completely fantastical scientific equipment, it features a doofus detective.

Namely, said doofus is one who thinks that baloney sandwiches are evidence and that phrases like “reasonable degree of scientific certainty” are, well, reasonable. Thankfully, he’s surrounded by far more intelligent and empathetic colleagues (including a cast of familiar faces).

Next: John Oliver on corporate consolidation

Even a sack full of magic 8 balls and a crime-sniffing pony isn’t enough to help him. Though, apparently, at least one of them got someone convicted in a court already. Hopefully, at least in this small niche of television, cooler heads will still prevail.