The NRC's Verdict on Polygraph: A Plain-Language Summary of the 2003 Report

In 2003, the National Research Council — the operating arm of the National Academy of Sciences, one of the most authoritative scientific bodies in the United States — published a comprehensive review of the polygraph commissioned by the Department of Energy. The conclusion was unambiguous: the scientific basis for polygraph testing is weak, and its use for security screening creates an unacceptable trade-off between false positives and false negatives.

That report, titled The Polygraph and Lie Detection, is freely available from the National Academies Press. What follows is a plain-language summary of its core findings.

The Central Finding

The NRC's overarching conclusion, stated in Chapter 8, is this: "Almost a century of research in scientific psychology and physiology provides little basis for the expectation that a polygraph test could have extremely high accuracy."

The committee found that the physiological responses measured by the polygraph — blood pressure, respiration rate, galvanic skin response — are not uniquely associated with deception. The same physiological changes occur in response to anxiety, nervousness, indignation, medical conditions, and deliberate countermeasures. There is no single physiological signature of lying.

The Accuracy Problem

The NRC reviewed the available scientific literature on polygraph accuracy and found it "scanty and scientifically weak." Most studies were methodologically flawed: conducted in artificial laboratory conditions, with subjects who had no real consequences attached to detection, or with samples too small to draw reliable conclusions.

Based on the available evidence, the committee estimated polygraph accuracy at an area under the ROC curve of approximately 0.89 — which sounds impressive until examined in context. Applied to a realistic security screening scenario, the report modelled the following outcome:

"If the test were set sensitively enough to detect about 80 percent or more of deceivers [in a population of 10,000 employees containing 10 spies], about 1,606 employees or more would be expected to 'fail' the test; further investigation would be needed to separate the 8 spies from the 1,598 innocent employees who failed."

In other words: for every genuine security threat detected, approximately 200 loyal employees would be subjected to investigation and stigma. The committee described this as "an unacceptable choice."

Why the Theoretical Basis Is Weak

The polygraph rests on the theory that deception produces distinctive physiological arousal — specifically, that lying triggers a fear response measurable through autonomic signals. The NRC identified three fundamental problems with this theory:

• Multiple causes, single measurement. The same physiological responses triggered by lying are also triggered by anxiety about being falsely accused, medical conditions, anger, and deliberate countermeasures. The instrument cannot distinguish between these causes.
• Countermeasures work. The NRC explicitly noted that "some simple countermeasures can apparently be learned in a few minutes and may be effective in causing innocent-seeming responses." Research subsequently confirmed that physical countermeasures — tensing muscles, performing mental arithmetic, biting the tongue — can suppress or amplify polygraph responses strategically.
• Examiners introduce uncontrolled variation. The report found that polygraph test administration involves "uncontrolled variation" in examiner behaviour, question framing, and emotional climate — all of which affect results independently of the subject's truthfulness.

The 2018 Update

A 2018 paper in Law and Human Behavior — the official journal of the American Psychology-Law Society — revisited the NRC findings 15 years on. The authors, Iacono and Ben-Shakhar, concluded that the scientific landscape had not materially improved. They found that polygraph proponents' claims of 90%+ accuracy remained unsupported by rigorous methodology, and that the NRC's conclusion about weak scientific foundations remained accurate.

Legal Status

The Employee Polygraph Protection Act of 1988 prohibits most private employers in the United States from requiring or requesting polygraph tests from employees or applicants. Federal agencies and certain security-cleared positions are exempted. The results of polygraph tests are inadmissible as evidence in federal courts and most state courts. Canada prohibits polygraph results from being used in criminal proceedings.

The gap between the legal and employment use of polygraphs — where they retain significant institutional power — and their scientific status — where they are widely recognised as unreliable — is the central subject of the Polygraph Pass guide.