Politics
The truth about the lie detector
Published 16 October 2006
Critics claim that polygraph testing is as credible as the tooth fairy or witchcraft. Yet the US government still relies on it to identify terrorists and vet FBI agents. Andrew Stephen on America's alarming love affair with junk science
Did ex-Representative Mark Foley have sex with teenage male congressional pages? Was Wen Ho Lee, an American nuclear scientist, guilty of espionage by passing nuclear secrets to the Chinese? Did John Mark Karr kill six-year-old JonBenet Ramsey? Was the British nanny Louise Woodward guilty of the involuntary manslaughter of the baby in her care? Was Aldrich Ames, a senior CIA official in charge of analysing Soviet intelligence, actually a Soviet double agent? Was Leandro Aragoncillo, an FBI analyst with top-secret clearance who was based in the White House under Vice-Presidents Gore and Cheney, also a spy? What about Dr Ignatz Theodor Griebl, a Nazi ringleader who fled New York on the SS Bremen in 1938?
I do not care about Foley, or Karr - who was innocent of JonBenet Ramsey's murder, as it turned out - but all the other cases have a thread in common. They illustrate a century-old American fallacy which, at long last, is beginning to crumble: that polygraph (aka lie-detector) tests actually work. Evidence is mounting that, far from being the infallible tools of world-beating American investigative procedures that Hollywood would have us believe, they have actually been responsible for countless miscarriages of justice and have ruined lives.
Ames, for example, sailed through three polygraphs before the CIA discovered that he was actually one of the worst US traitors in history. Woodward "passed" one but was then convicted on other evidence. Lee both "failed" and "passed" polygraphs, resulting in him being imprisoned and then released before being awarded $1.65m in damages by the federal government. Aragoncillo "passed" a pre-employment FBI polygraph but pleaded guilty to espionage in May. Griebl "passed" an FBI polygraph test and promptly returned to Hitler's side.
And Foley? Given what I now know about polygraphs - not least from reading a 151-page report, issued by the US government last month and entitled Use of Polygraph Examinations in the Department of Justice - I suspect a polygraph test would have a 50:50 chance of digging the truth out of him. As a thick-skinned and smooth-talking politician who lived decades pretending to be somebody he isn't, he would probably have cruised through like Ames. In a letter from his prison cell in Pennsylvania, no less an expert than Ames himself described polygraphy as "junk science" comparable with astrology; Ted Kennedy likens it to "20th-century witchcraft".
Yet nothing illustrates better the dysfunctional operation and inbuilt contradictions of US government in the 21st century than its attitude towards polygraphing. The justice department report solemnly outlines how, between March 2001 and February 2005, the FBI expanded the number of its employees liable to be polygraphed from 550 to 18,384; that, between 2002 and 2005, it conducted 1,994 polygraph exam inations specifically regarding terrorism and counter-terrorism; and that, in the same period, the FBI and two related government departments conducted 28,000 pre-employment polygraphs for job applicants. Yet I spotted a six-line item in the Washington Post a few days ago that began: "The energy department is ending required polygraph tests for thousands of its workers at its nuclear weapons facilities . . ."
In short, the left hand of government does not know what the right hand is doing, despite the mounting evidence. Eighteen years ago, Congress passed the Employee Polygraph Protection Act, which made it illegal for private sector employers to force employees to take lie-detector tests or to sack them for refusing to do so (except in certain fields, such as where private security firms were involved); Congress was thus in effect saying that procedures which do not work satisfactorily in the private sector can none the less be used by the US federal government itself.
"It's totally insane"
Last month, Professor David Lykken - behavioural geneticist, emeritus professor at the University of Minnesota and the world's leading expert on polygraphs - died at the age of 78, leaving a lifetime's work of studying polygraphy (which meant, mostly but not entirely, debunking it) behind him. His definitive verdict? "There's something about us Americans that makes us believe in the myth of the lie detector," he said. "It's as much of a myth as the tooth fairy."
The ancient rituals of reading entrails by Roman priests, the dunking of witches in medieval England, the use of rice by the Chinese to see whether a suspect's mouth goes dry during questioning - all, he suggested, were as efficacious as polygraphy. The irreproachably respec table National Academy of Sciences reported in 2002 that accuracy in polygraphy was "insufficient to justify reliance on its use in employee security screening in federal agencies".
Even in the milestone United States v Scheffer case in 1998, none other than the mighty US Supreme Court ruled that "there is much inconsistency between the government's extensive use of polygraphs to make vital security determinations and the argument it makes here, stressing the inaccuracy of these tests". Justice Clarence Thomas added that "there is simply no consensus that polygraph evidence is reliable". R James Woolsey, director of central intelligence from 1993-95, says that polygraphing should be "radically curtailed"; his immediate successor, John Deutch, dismisses it as "totally insane".
Yet old habits, however ludicrous, die hard. A primitive machine to measure pulse rates during questioning was developed in 1895, but credit for modern polygraphy - consisting, these days, not of jiggering needles drawing lines on revolving rolls of paper, but data that goes straight to laptops - is usually given to an oddball peripatetic academic called William Moulton Marston (1893-1947). He became better known in 1941 as the creator of the comic-book character Wonder Woman, possessor of the Lasso of Truth - which forced anybody under her command not just to tell the truth, but to obey her.
Thanks to Marston and his fantasies, the FBI started using polygraphs in 1935; according to last month's report, today everybody applying for a job with the FBI must undergo a lie-detector test. Yet the FBI itself previously told the Senate judiciary committee that a staggering 20 per cent of job applicants were "determined to be withholding pertinent information", clearly a preposterous finding. The Journal of Applied Psychology, contradicting the 95-98 per cent success rate claimed by the polygraphists, reported in 1997 that polygraphy worked just 61 per cent of the time.
This means, perversely, that it is innocent people who are overwhelmingly the victims of polygraphy. The National Research Council, a wing of the National Academies, reported last year that if a group of 10,000 people including ten spies was polygraphed, 1,600 innocent people would "fail" and two of the spies "pass". A faithful spouse who agrees to take a test because his partner wrongly accuses him of cheating - this happens frequently because of a sordid commercial polygraphy business - will see his marriage collapse if he "fails" the test. And so on.
Like all dubious science, polygraphy abounds with gobbledegook to give it spurious validity: the polygrapher, we are told sombrely, will measure pneumographic, electrodermal and cardiovascular responses by asking CQTs, DLTs and GKTs. Briefly, this means that a polygraph will measure the subject's heartbeat, breathing rate, blood pressure and sweating while he or she is being asked a series of questions that fall into specific categories. The first is a string of irrelevant questions ("Is today Tuesday?"), designed to baffle the subject, but the responses to which will not even be measured.
Then come the "control questions", which may comprise simple things such as: "Have you ever told a lie to get you out of trouble?" The assumption is that any truthful person will answer "yes", but that the very exchange will make him or her uncomfortable and elicit physical reactions against which responses to the serious "guilty knowledge" questions ("Have you ever committed espionage?" or, in the commercial sector, "Have you ever cheated on your wife?") can then be measured. Bingo! There it is: you've unmasked your thieves, adulterers and spies.
I would fail any such test
The trouble, as Lykken so eloquently spelled out in A Tremor in the Blood, is that all polygraphs do is to measure physiological reactions to verbal stimuli: fear or anger or embarrassment might just as easily produce these reactions as lying. I am convinced I would fail any polygraph test, for example; some of us are just made that way. In my case, words as harmless as "In our Washington studio to discuss this now is Andrew Stephen . . ." are guaranteed to make my heart palpitate, my breathing quicken, my blood pressure soar and my clothes soak with sweat; I would thus immediately fall into the "DI" (deception indicated) classification, yet no lie would have passed my lips.
If you are Aldrich Ames (or Mark Foley?) or the like, however, you are made differently and "NDI" - no deception indicated - would be the verdict. It helps not to have a conscience and, I gather, even to drill yourself into believing that you tell the truth. Furiously doing mental arithmetic during questioning can throw the measurements; biting your tongue, clenching the sphincter, thinking exciting or frightening thoughts, and consciously controlling breathing are all techniques that apparently can work, too.
Fortunately for the rest of the world, polygraphy has been largely confined to the US. Britain, I'm told, was put under pressure by Washington to start using polygraphy to root out its Blunts and Philbys in the 1980s, but the request was turned down after the British Psychological Society (among others) gave it a decided thumbs-down. Notwithstanding the current enthusiasms of the FBI, I suspect that Lykken's passing and the unheralded decision by the department of energy will come to be seen as watersheds in the history of a doomed American pseudo-science. Just the cue, I presume, for the Blair government to announce that it is introducing American-style polygraphing in Britain.
Post this article to
We want to encourage people to comment on our content and to exchange views with other readers and hope this will be done on a courteous basis. However, if you encounter posts which are offensive please let us know by emailing comments@newstatesman.co.uk and we will take swift action where necessary.
