How often have you heard a mention of a machine called a “lie detector?” According to pop culture, members of law enforcement organizations, and even the U.S. government, a polygraph or lie detector is a machine that records certain vital functions of a person suspected of lying. The person is asked a series of questions while the polygraph reads their physiological responses, and the resulting readings are supposedly highly accurate evidence as to whether that person was lying or telling the truth.
Over the decades, many people have protested the use of the polygraph in the justice system, as a tool for detecting disloyalty in intelligence agencies, or as a condition of employment. Some arguments have been based mainly on philosophical objections to the concept of a lie detector: Preemptive polygraph tests contradict the principle that people are innocent until proven guilty in a fair justice system, or they violate a person’s constitutional right to be protected from self-incrimination, or they prevent them from “remaining silent” regarding criminal charges. These arguments have merit, and they are undoubtedly made out of genuine concern for human rights. However, there is a much more practical, relevant, and immediate reason to protest the use of polygraphs as lie detectors: They don’t work.
The basic premise that is used to promote the accuracy of polygraphs is that all humans – regardless of age, sex, ethnicity, health, personality, and all other factors – display some uniform, measurable physiological reaction when saying something they know to be false which is clearly different from their response to stating something that is true. This reaction is usually said to be increases in pulse, blood pressure, and breathing rate and sometimes changes in skin conductance and/or resistance. The only problem? There is no evidence to support the existence of such a universal reaction. No study has ever shown that a person’s reaction to lying is measurable, and people don’t always react to lying, stress, or the feeling of being nervous in the same way.
The accuracy of polygraphs has been disputed and many studies on the topic are heavily flawed. Researchers who already believe in the polygraph’s accuracy unsurprisingly tend to find the highest estimates, claiming guilty accuracy rates above 80% or 90%, and an innocent accuracy rate of a perfect 100%. That is, they claim that people who are truthful will always be found innocent, and liars will be caught often enough that if the polygraph was a student, it would average an A or a B. However, skeptical research teams have found more modest guilty accuracy rates of around 74-76% (Tavel 2016, Synnott 2015). Even worse, the innocent accuracy rate plummets when questioned; only about 52% of innocent people will be judged truthful (Tavel 2016).
False Negatives
The first type of invalid result from a polygraph involves people who lied during the test but are judged truthful, a result called a false negative. This may stem from a number of factors, including a number of “countermeasures” that can influence a person’s physiological state, but the basic idea behind the polygraph may create errors as well. If a guilty person doesn’t feel nervous, the test may not indicate abnormal vital signs.
There are a number of well-known cases of liars passing polygraph tests, especially in the area of national defense and intelligence. Despite the extensive use of polygraphs by the CIA during the Cold War to screen its agents for dishonesty and disloyalty, there are six known people who have been judged innocent while actively delivering U.S. secrets to foreign governments. One of the most high-profile cases was that of Aldrich Ames. Ames had a long career in the CIA, starting with summer jobs as a high school student. He applied to his first full-time position in 1962, and was also given his first polygraph at this time. During the test, he admitted to the petty crime of stealing a bicycle while drunk, and was deemed honest.
In 1985, Ames began passing information the Soviet Union regarding the activities of CIA agents and assets. Initially, he planned it as a one-time payoff to satisfy his debts, including the money he owed his first wife per their divorce settlement, but the arrangement quickly became extensive and permanent. The information he disclosed included the identities of CIA assets, endangering their safety and freedom, and in some cases costing them their lives. During this period, he was given three more polygraph tests, one in 1986 and two in 1991. During the initial 1991 polygraph, the results showed a seemingly deceptive response from Ames in reaction to a question of whether he was “concealing contacts with foreign nationals,” but not when he was asked if he was “working for a foreign intelligence service” (Senate Select Committee on Intelligence 1994). The response to the “foreign nationals” question was concerning enough to the polygraph examiner that Ames was asked to come in for a follow-up polygraph a few days later. That test showed no indication of deception at all. Ames was declared honest.
Ames was arrested in 1994 after his connections to the leaks were noticed, and an investigation was immediately begun into how one man had been able to do so much damage to American intelligence interests and avoid suspicion for so long. An assessment was commissioned for the Senate Select Committee on Intelligence, and it briefly discussed the possibility that Ames had been specially “trained” to beat the polygraph by his Soviet handlers. In truth, Ames received no training; his contacts merely told him to get a good night’s sleep and avoid stress on the day of the test. One might think that the US government would be quite concerned about this fact, but the report waves it off and instead faults other factors for the CIA’s failure to prevent or immediately stop Ames’ actions. No recommendations were made regarding the use of the polygraph in internal investigations, even though the report contains ample evidence of the fallibility of polygraph testing and nonchalantly mentions a number of incidents that imply CIA overconfidence in the inability of employees to conceal information. One quote says that Ames became so rich from selling state secrets that he was able to “purchase a new Jaguar automobile and a $540,000 home, with cash, in Arlington. Apparently, these seemingly large expenditures by an employee making less than $70,000 a year had not raised questions at the CIA” (Senate Select Committee on Intelligence 1994).
False Positives
The other type of inaccurate result happens when polygraphs “prove” that truthful people are lying. This scenario is everyone’s worst nightmare. Picture it: you are accused of a terrible crime and agree to take a lie detector test to clear your name. Despite being truthful and telling the police everything you know, you are told that the polygraph readings show that you were lying. Instead of being released, you are taken to an interrogation room, where officers press you for more information. You must have been lying, they tell you, because you were tested by a scientific instrument that measures specific human reactions to withholding the truth. What do you do then?
Just as with false negatives, there are known examples of polygraphs giving false positive results, and they have had terrible consequences on some lives. Here are a few examples:
- A man who was a suspect in the Green River Killer cases was deemed a liar after a polygraph test. Later, DNA evidence proved that the culprit was actually Gary Ridgway, who had escaped suspicion after passing a polygraph.
- For almost two decades, the main suspect in the death of Vicki Wegerle was her husband Bill, who failed not one but two polygraphs during the course of the investigation. The real murderer was Dennis Rader, another serial killer known as BTK, who was also caught with DNA evidence.
- Numerous people have applied for jobs at the FBI, CIA, and other government agencies that are allowed to require polygraph testing as a condition of employment and have been rejected despite telling the truth. The website AntiPolygraph.org maintains a list of statements made by these people, who describe abusive treatment, “guilty until proven innocent” attitudes, extreme suspicion, leading questions, and (ironically) lying on the part of their examiners. This process has also been used as a pretext for firing employees, with the failed polygraph results, or even just the examiner’s intuition, as the only evidence whatsoever.
Going beyond individual cases, there is evidence that polygraphs and other attempts at detecting a person’s level of honesty are actually biased against truth tellers. Questions may lead to an increased response in an examinee because they are unusually guilt-prone and feel nervous and guilty in response to hearing details of a crime even if they did not commit it. Such people are sometimes called “guilt grabbers” (Lilienfeld 1993).
In addition, test questions may be phrased in a way that makes them likely to provoke a response in anyone, guilty or not. The behavior of the examiner likely has an influence on the results as well. One of the statements from Antipolygraph.org, made by an applicant to the Arizona Department of Public Safety, is a great example of these effects. In his statement, “Derek” notes that he was pressured by the examiner to answer in certain ways and was repeatedly asked about his sex life, including the question, “Have you ever fondled a girl while she was sleeping?” (Antipolygraph.org, 2022) Since any person with a conscience would react to this statement with feelings of shock and disgust, a polygraph would no doubt almost always record a “lie” response to this question, picking up the heightened blood pressure and heartbeat caused by such feelings. This treatment seems egregious, especially since (if the polygraph truly detected lies) a simple “Have you ever committed a sex crime?” should be sufficient to catch sexual predators and prevent them from working in public service. As it was asked, though, the only type of person who would consistently be judged truthful would be someone who felt calm and unconcerned about child molestation!
Conclusions
The myth of the polygraph as a lie detector has persisted for decades. Why? Aldrich Ames can offer an educated guess. In a letter written in 2000 to the Federation of American Scientists, he notes a number of factors that may be at play – polygraph examiners profit from their work, the polygraph is easily abused to force confessions, etc. (Ames 2000) The ideas that he focuses on, however, are the systematic suspicion of US intelligence services and the tremendous risk of their putting trust in the wrong person. With the polygraph, politicians and bureaucrats can claim that they are doing their due diligence for national security by using a “scientific” technique to root out liars, and if someone slips through the cracks, they can avoid blame by faulting a “rare” machine-made mistake rather than human error. This practice has enabled what Ames calls “bureaucratically-driven flights from accountability,” and it will likely continue into the foreseeable future unless our government decides to follow the science and face the polygraph’s long history of false positives and false negatives (Ames 2000).
In sum, the idea that the polygraph can accurately detect lies is a complete myth. It has no established scientific basis, and has failed repeatedly when applied in the real world. Thanks to its continuing use in our society, innocent people have been accused of serious crimes, and spies have been declared honest while actively working against US government and intelligence interests. It’s time for the polygraph to be seen for what it truly is: wishful thinking and willful ignorance masquerading as science.
References:
Ames, Aldrich. “A Letter from Aldrich Ames on Polygraph Testing.” Sgp.fas.org, 28 Nov. 2000, sgp.fas.org/othergov/polygraph/ames.html.
Lilienfeld, Scott. “Do “Honesty” Tests Really Measure Honesty?” Skeptical Inquirer, 1993, cdn.centerforinquiry.org/wp-content/uploads/sites/29/1993/10/22165136/p34.pdf.
“Personal Statements of Polygraph Victims | AntiPolygraph.org.” AntiPolygraph.org, 2025, antipolygraph.org/statements.shtml. Accessed 19 Apr. 2025.
“Polygraph Statement of Derek (Arizona Department of Public Safety).” AntiPolygraph.org, 2022, antipolygraph.org/statements/statement-046.shtml. Accessed 19 Apr. 2025.
Senate Select Committee on Intelligence. “An Assessment of the Aldrich H. Ames Espionage Case and Its Implications for U.S. Intelligence – Senate Select Committee on Intelligence – 01 November 1994 – Part One.” Irp.fas.org, 1 Nov. 1994, irp.fas.org/congress/1994_rpt/ssci_ames.htm.
Synnott, John, et al. “A Review of the Polygraph: History, Methodology and Current Status.” Crime Psychology Review, vol. 1, no. 1, Jan. 2015, pp. 59–83, www.tandfonline.com/doi/full/10.1080/23744006.2015.1060080, https://doi.org/10.1080/23744006.2015.1060080.
Tavel, Morton. The “Lie Detector” Test Revisited: A Great Example of Junk Science | Skeptical Inquirer. 30 Mar. 2016, skepticalinquirer.org/2016/03/the-lie-detector- test-revisited-a-great-example-of-junk-science/.
Wikipedia Contributors. “Polygraph.” Wikipedia, Wikimedia Foundation, 21 May 2019, en.wikipedia.org/wiki/Polygraph.
