The Truth About Polygraph Tests

 They're junk science, inadmissible in court, and about as reliable as a pack of Tarot cards [or Scientology's E-meter]

By Claire Berlinski

City Journal, September 19, 2018

News organizations would render a valuable service if, whenever they report that someone has taken or proposes to take a polygraph, they reminded readers (or explained to them) that polygraphs are voodoo. Junk science. They are no more reliable than a pack of Tarot cards. Polygraph evidence is inadmissible in court. There is a good reason for that. To check Brett Kavanaugh's qualifications for the Supreme Court, Congress would do well to ask him whether he believes Frye v. United States and United States v. Scheffer were correctly decided. This would be far more illuminating (and meaningful) to anyone trying to discern his qualifications for the Court than asking him whether he assaulted Christine Blasey Ford.

Journalists who report that Mike Pence has offered to take a polygraph (to prove that he was not the author of the anonymous New York Times op-ed), or that Ford has taken one, without explaining that polygraphs cannot discern truth from falsehood are wasting an opportunity to educate their readers.  If you promulgate the idea that there's a machine that can tell when someone is lying, you shouldn't be surprised to find yourself living in a culture so hostile to science that kids go unvaccinated and measles break out in the First World.

A polygraph measures your heart rate, breathing, and galvanic skin response. There is no evidence that any pattern of physiological responses is unique to deception. Polygraphs are useful to investigators trying to elicit a confession, however: if you convince suggestible people that these measurements are associated with lying, they are more likely spontaneously to confess when you tell them, "The machine says you're lying."

It works as follows:

I say, "We've got your endotrygliceride levels from the doorknob you're touching. We'll match those up against the steering wheel and that'll tell us the whole story right there. Son, why on earth wouldn't you want us to match up those endotrygliceride levels if you're not involved in this? If you're afraid of what those endotrygliceride levels will tell me, you should sit right back down. If the truth comes out later and you've been wasting my time, I won't be able to help you."

I was right. It was the moment. The endotrygliceride levels never lie. Then I tell the suspect he is guilty. Full stop. He is guilty and I know he is guilty. I tell him all the evidence we have against him, piling it up later after layer until he feels entombed by his misdeeds, until the suspect is well-nigh positive he cannot escape.

That's a fictional scene, from my brother's novel Peacekeeping. But it is exactly how the polygraph works. You can't discern truth or falsehood from "endotrygliceide levels." There is no such thing as an "endotrygliceride level." Suspects who don't know better can, however, be coerced into confessions by a procedure that purports to be scientific.

The only value of a lie-detector test is as an interrogation tool—if you're dealing with someone who believes in it. But it's insanity to rely on it. People can be trained not to crack upon interrogation. If the polygraph is administered by someone friendly, there will be no interrogation. Anyone whom Mike Pence or Christine Ford hires to administer a lie-detector test will "discover" that he or she is telling the truth.

So why, then, does our intelligence community use them? The short answer: they shouldn't. Reposing our country's security in a tool of no more proven utility than a dowser has is an exceptionally bad idea. No doubt it prompts a few true confessions among the guilty, but the value is outweighed by the harm: that believing in polygraphs, or pretending to believe in them—or in astrology, or homeopathy, for that matter—makes us collectively more stupid.

Appeal to the polygraph may help interrogators catch criminals, but it does so in a way that encourages magical thinking in the public. Using polygraph results to make hiring decisions, or pretending that the polygraph is anything but an interrogation tool, is a terrible idea in government. By necessity, it sorts government employees into two categories: those credulous enough to believe in polygraphs—that is, people who are too credulous and scientifically illiterate to be qualified for a significant job—and those who know better, who are forced to pretend that they believe in this nonsense to keep their jobs. It thus makes employees who are smart enough to know it is nonsense into liars from their first day at work.

Polygraphs also return false results that prevent qualified people from taking important jobs. They give employers confidence in employees who don't warrant it. It's true that if educated people pretend to believe in the polygraph, then its value is enhanced as an interrogation tool. But the consequences of living in a culture that is profoundly untethered from empiricism and reality are obvious enough.

That Mike Pence has offered to take a polygraph tells us exactly what we'd know from studying his horoscope—nothing. That he would appeal to the polygraph tells us that he is credulous or cynical. That Christine Ford has "passed" one tells us what we'd know from learning that she is a Virgo whose ascendant is in Libra—nothing, again. That she is appealing to it tells us that she is suggestible, manipulative, or manipulated. Reporters who discuss her polygraph results uncritically tell us exactly the same thing about themselves. None of this means that they're lying—it just means that they're dumb.