Use of Questionable “Lie Detectors” by Law Enforcement Expands Nationwide
by Matt Clarke
In the aftermath of the Sept. 11, 2001 terrorist attacks, law enforcement and other government agencies implemented new practices to obtain information from suspects during investigations and interrogations. The use of torture and torture-like techniques to extract information from suspected terrorists and “enemy combatants” has been widely publicized in the mainstream media. Less well known is the government’s use of so-called lie detectors as routine interrogation tools in our nation’s criminal justice system.
Generally, two types of devices are used as lie detectors – polygraphs and voice analyzer systems. Polygraphs are more familiar; they use sensors on the fingers, hands, arms, chest and abdomen to measure as many as 21 stress indicators such as sweating, breath rate, blood pressure and heart rate. These indicators are graphed and interpreted by a trained examiner who then purports to determine whether the person being questioned is telling the truth or being deceptive. Voice analyzers are usually computer-based programs that purport to measure stress levels in a person’s voice.
The big lie about lie detectors is that they can detect lies. At most they measure stress. Studies have generally shown that polygraphs can successfully measure stress levels; however, they do not differentiate stress due to the testing itself from deceptive or untruthful responses. Dr. Ken Alder, author of The Lie Detectors: The History of an American Obsession (Free Press 2007), calls the practice of measuring physiological responses to detect lies “a farce.”
In 1988, Congress passed the Employee Polygraph Protection Act (29 U.S.C. § 22), which prohibits private employers from using polygraphs for pre-employment screening and during the course of employment. Employers are still allowed to use polygraphs to investigate workers suspected of wrongdoing. The law does not apply to state, local or federal agencies – including the military – or to certain security-sensitive industries.
Polygraphs are not admissible as evidence in criminal prosecutions in most states or in federal court, though test results can be introduced by stipulation or at the discretion of the court in some cases. The use of polygraph results in court is governed by the Daubert standard, which requires that scientific evidence must be both relevant and reliable. See: Daubert v. Merrell Dow Pharmaceuticals, 509 U.S. 579 (1993). It is the “reliable” part that has proven problematic for polygraph tests.
In 1998 the Supreme Court held, “To this day, the scientific community remains extremely polarized about the reliability of polygraph techniques. Some studies have concluded that polygraph tests overall are accurate and reliable ... others have found that polygraph tests assess truthfulness significantly less accurately.” The Supreme Court upheld the Air Force Code of Criminal Procedure’s blanket exclusion of polygraph evidence against a defendant who wanted to use the government’s test results to bolster his claims of innocence regarding drug use. See: United States v. Scheffer, 523 U.S. 303, 118 S.Ct. 1261, 140 L.Ed.2d 413 (1998).
The National Academy of Sciences reviewed polygraph research studies in 2003 and determined that most of the research was unreliable and unscientific. The reviewers concluded there was “no direct scientific evidence assessing the value of the polygraph as a deterrent, as a way to elicit admissions and confessions, or as a means of supporting public confidence.”
A study conducted by the British Psychological Society in 2005 found that polygraph examiners classified up to 17 percent of guilty subjects as innocent and up to 47 percent of innocent subjects as guilty. Noting that subjects who tell the truth may exhibit the same stress indicators as deceptive subjects, the Society warned against using polygraphs as a means to detect guilt. The reasons given for stress in innocent subjects include a fear of not being believed, being upset about the phrasing of test questions, and concerns about the underlying crime for which they are being questioned (and implicitly accused of committing).
The Society also warned that guilty subjects can suppress their physiological reactions to questions using a variety of techniques such as meditating, drug use, or causing certain physiological responses during the control questions used to establish a subject’s baseline, which can distort the polygraph results. In fact, professional spies, guerrillas and others who may be subjected to polygraph testing in the course of their activities regularly receive training in doing just this: fooling the polygraph.
Some law enforcement officers are specifically trained to beat lie detector tests. According to an October 30, 2008 Associated Press article, undercover ATF agents were able to infiltrate the Mongols motorcycle gang despite being subjected to polygraphs conducted by a private investigator hired by cautious gang leaders. The resulting federal investigation led to 64 arrests in multiple states. “Our guys are highly trained, and they were pretty much hand-selected to do this mission and for their ability to think fast under pressure and beat the box [polygraph],” said ATF agent John Torres.
Regardless, polygraph proponents claim that the tests accurately detect lies about 90 percent of the time if a qualified examiner is used and the polygraph is competently performed. They argue that problems result from poor examiners, not a fundamental flaw in the polygraph process itself.
“We’d like the accuracy rate to be better,” acknowledged Donald Krapohl, special assistant to the director of the Defense Academy for Credibility Assessment (DACA, formerly the Dept. of Defense Polygraph Institute), the government agency that oversees federal polygraph examiners. “But it is simply the best we have.” Supporters and critics both agree that the quality of the examiner is critical to the trustworthiness of the test results.
“There are just some examiners you need to beware of,” said Mike Chimarys, Chairman of the Texas Association of Polygraph Examiners (TAPE). “There are some I would not feel comfortable with – but those are few.”
“We’re still trying to recover from the polyester polygraphist back in the ‘80s,” stated Donald Ramsey, president of TAPE, referring to problems with self-taught polygraph examiners who were used to screen potential employees, which led to federal standardization of polygraph training and procedures.
Even more controversial than physiological lie detectors are voice analyzers, which purport to detect microtremors associated with the muscles used when speaking. Critics argue there is no scientific basis for voice analyzer manufacturers’ claims that they can identify lies.
“In our evaluation, voice-stress analysis detected some instances of deception, but its ability to do so was consistently less than chance – you could have gotten better results by flipping a coin,” said Mitchell S. Sommers, a Washington University associate professor of psychology. “We tested one of the more popular voice-stress lie detection technologies and got dismal results, both in the system’s ability to detect people actually engaged in deception and in its ability to exclude those not attempting to be deceptive,” she stated.
Sommers noted that high stress levels do “not necessarily correlate with deception.” Her research was conducted in cooperation with DACA, presented at the World Congress of International Conference of Psychophysiology in July 2002, and later published in a peer-reviewed academic journal.
A March 2006 study by Harry Hollien and J.D. Harnsberger, researchers at the University of Florida, evaluated voice stress analyzers using a double-blind experiment to detect truth and deception, and found an overall false positive rate of 41-49% and false negative rate of 62-81% under field-based conditions. The study, which was funded by the U.S. Dept. of Defense, determined that voice stress analysis “performed at chance-level for deception, truth and stress.” That is, about the same as a 50-50 coin toss.
According to the website for the Dept. of Defense’s Counterintelligence Field Activity, which recently merged with the Defense Intelligence Agency, the department does not sanction the use of voice stress technology because it “has yet to be validated by research and replicable scientific studies.” Instead, 26 federal agencies, including the ATF, FBI, CIA, BOP, NSA and DEA, use standard polygraphs.
Both polygraphs and voice analyzers are used by law enforcement, prison and parole officials to intimidate suspects during questioning. Although an interrogator screaming at a subject that he failed a lie detector test and needs to “come clean” may result in some legitimate confessions from the guilty, there is also the danger of coercing innocent suspects who are vulnerable, stressed and confused into confessing their “guilt” for crimes they did not commit.
Psychology and criminology professor Richard Leo at the University of California at Irvine calls computer voice stress analyzers “complete nonsense.”
“It’s junk science with a capital J,” said Leo. “I think these CVSA [Computer Voice Stress Analyzer] machines are dangerous and they are contributing to the process that elicits false confessions.”
Innocence Project co-director Keith Findlay notes there is a correlation between failed lie detector tests, false confessions and wrongful convictions. Steven Drizen, legal director of the Northwestern Law School’s Center on Wrongful Convictions, and an Innocence Project consultant, agrees.
“The problem is that an innocent suspect volunteers to take the test and, when told that he failed, reaches the point of hopelessness where he can be easily persuaded to confess,” said Drizen.
Indeed, one common police interrogation technique is to tell a suspect that he failed a polygraph when in fact he passed or the results were inconclusive. Ironically, such lies told by law enforcement officials regarding the results of lie detector tests are completely legal. The Supreme Court has condoned coercive police interrogation techniques when questioning suspects. See, e.g., Chavez v. Martinez, 538 U.S. 760 (2003).
In June 2003, a female jail prisoner told sheriff’s detectives in Maricopa County, Arizona that Robert Louis Armstrong, a 51-year-old machinist, was the shooter in an Easter 1998 triple homicide. The woman passed a voice analyzer test, and Armstrong was arrested. He claimed to have been with his mother in Oregon when the murder occurred. After hours of interrogation, the detectives gave Armstrong a lie detector test, told him he failed, and insisted he had the dates of the visit to his mother wrong.
Armstrong then allegedly “confessed” by answering questions posed by the detectives. He spent a year in jail awaiting a capital murder trial before his defense team uncovered bus company records proving he was in Oregon when the murders took place. Faced with this evidence his accuser recanted, and Armstrong was released. He then sued. The case settled under undisclosed terms in August 2006. See: Armstrong v. Maricopa County, U.S.D.C. (D. Ariz.), Case No. 2:05-cv-01563.
“It’s so easy to abuse the voice-stress analyzer,” said Buddy Rake, Armstrong’s attorney. “We know of other confessions from people who were not guilty.”
In fact there are numerous cases of detectives and examiners using polygraphs and voice analyzers as tools to coerce innocent suspects into confessing. Lie detectors are likewise used to extract confessions from guilty suspects, which is one of the reasons law enforcement personnel like them so much. But is the power to intimidate a guilty suspect into confessing worth the price of coercing confessions from the innocent? Many law enforcement agencies apparently think so. A false confession and a wrongful conviction still amounts to a “solved case.”
In 1998, Michael Crowe, Joshua Treadway and Aaron Houser, all juveniles, were accused of murdering Crowe’s 12-year-old sister in Escondido, California, after police used a Computer Voice Stress Analyzer to bully Crowe into falsely confessing. The victim’s blood was later discovered on a homeless man’s shirt as the boys were preparing for trial, and all charges against them were dropped. The boys’ families then sued the police, prosecutors, other government officials and the West Palm Beach, Florida-based National Institute for Truth Verification (NITV), which manufactured the $9,995 voice analyzer. NITV settled in June 2005 after acknowledging that their machine was “not capable of lie detection.” Claims against the other defendants are still pending. See: Crowe v. City of Escondido, U.S.D.C. (S.D. Cal.), Case No. 3:99-cv-00241-JM-RBB.
The inadmissibility of polygraphs in court can cut both ways; some defendants, as in Scheffer, want to use the tests to show they are being truthful and are innocent. Due to the risk of false positive polygraph results, however, that may not be a wise course of action.
In October 2008, Trent Lee Banks, who had been convicted of attempted murder in Baltimore, Maryland the previous year, asked to take a polygraph in an attempt to reduce his prison sentence for probation violations on other charges. The court agreed, but the results weren’t what Banks had expected. “The examinee did not appear to be truthful,” the examiner found. The court therefore denied Banks’ motion to modify his 45-year sentence.
A recent increase in the use of polygraphs and voice analyzers is due to the supervision of a growing number of probationers, parolees and registered sex offenders. Lie detectors are also being used to investigate prison disciplinary cases.
One state with a high number of polygraph examiners is Texas. A majority of the 235 state-licensed examiners in Texas are former law enforcement officers or current officers who have an off-duty polygraph business. Most of their business is sex offender monitoring, which became mandatory in the 1990s. The approximately $200 cost of the test is paid by the person taking the test. A parolee or sex offender who fails or refuses to take a polygraph can be returned to prison. Some examiners claim that the system is insular, with sex offender treatment providers choosing which polygraph testers to use and giving preference to members of the law enforcement community.
A May 2008 Texas Sunset Commission report sharply criticized the state’s Polygraph Examiners Board, and recommended that it be dissolved and reconstituted under the Texas Department of Licensing and Regulation. The report found that the Board’s enforcement efforts did not adequately protect the public. Further, the Board had made several decisions based upon Board members’ self-interest rather than protection of the public, adopted rules that created an apparent conflict of interest, and used overly-subjective licensing examinations.
The Sunset Commission report also noted that the Board almost never took disciplinary action against polygraph examiners despite numerous complaints, was the approving authority for polygraph schools owned by Board members, and was the certifying authority for students from those schools. Furthermore, Board members were allowed to grade the examinations of their own interns. The Commission staff reported interviews with numerous examiners who told them that “one of the chief purposes of a polygraph exam is to pressure an allegedly guilty party into making a confession.” So much for innocent until proven guilty.
A recent Second Circuit decision allowed the use of polygraphs to verify that sex offenders are complying with the terms of their probation. While the appellate court acknowledged that studies had questioned the reliability of polygraph tests, since “the bottom range [of test accuracy] is still more-likely-than-not,” polygraph testing was permissible because it “produces an incentive to tell the truth, and thereby advances the sentencing goals.” Not discussed was the high percentage of false positives and false negatives in polygraph tests. See: United States v. Johnson, 446 F.3d 272 (2nd Cir. 2006), cert. denied.
Alaska began using polygraphs to test sex offenders on parole and probation starting in 2006. Forty other states require sex offenders to submit to polygraphs upon leaving prison. The New Jersey parole board began polygraph testing of registered sex offenders in 2007; periodic polygraphs are required as part of their supervision.
In 2006, England completed a pilot program that used polygraphs to monitor sex offenders on probation. The program, which involved hundreds of subjects, reportedly resulted in “admissions” from 80% of the offenders when they were confronted with the test results. It was announced in September 2008 that under an expanded three-year pilot program, sex offenders in the U.K. will be required to take lie detector tests to determine if they are complying with the conditions of their release. This is despite the earlier study by the British Psychological Society which found significant problems with polygraph testing.
The Wisconsin Department of Corrections purchased seven voice analyzers in 2004 at a cost of $14,000. Training cost an additional $2,100 per person. Wisconsin state law prohibits the use of voice stress analyzers by employers, with limited exceptions. Not to worry, announced the DOC – the analyzers were to be used on prisoners in disciplinary investigations and no disciplinary conviction would be upheld without “corroborating information.” Furthermore, there would be no retribution against prisoners who refused to submit to voice stress analysis. It was not revealed whether prison officials shared that information with prisoners before requesting (or ordering) that they submit to voice analyzer tests. Court decisions have held that polygraph results constitute “some evidence” which will result in courts upholding prison disciplinary convictions. However, unlike parolees for whom polygraphs may be a parole condition, prisoners can not be compelled to take polygraph tests against their will. Given the inherent inaccuracy of polygraphs, there is no reason prisoners should consent to polygraphs, whether innocent or guilty of the accused offense.
Over 20 law enforcement agencies in Arizona use voice stress analyzers. NITV claims that over 1,800 local, state and federal law enforcement agencies utilize their devices. Many of these agencies use lie detector tests for pre-employment screening and internal investigations.
Sometimes the mere threat of a polygraph is sufficient to indicate wrongdoing. On October 24, 2008, an unnamed Vigo County, Indiana jail guard was accused of inappropriate conduct. He was scheduled to take a polygraph test but was a no-show; he was then fired for disobeying a direct order.
Not all law enforcement agencies have embraced lie detectors – at least not for investigating their own. It was reported in November 2008 that the El Paso, Texas police department was discontinuing the use of polygraphs for internal investigations. Police chief Greg Allen stated that polygraph testing was a “piece of junk,” while Bobby Holguin, president of the El Paso Municipal Police Officers Association, said the tests were “garbage.”
“I don’t like intimidating people. I don’t like the threatening attitude of the machine,” said Allen. “It’s a good tool if you want to scare someone to tell the truth, but often the truth comes to the surface anyway.”
“If it was an exact art, I’d be all for it, but it’s not even close to an exact art,” Holguin remarked.
Notably, however, criminal suspects in El Paso are still subject to polygraph testing, which indicates a certain hypocrisy: Lie detectors are “junk” and “garbage” when used on police officers suspected of wrongdoing, but are acceptable for use on members of the public accused of crimes.
Many other law enforcement agencies have embraced the use of lie detectors – unless they indicate a suspect is telling the truth. In a 2007 criminal prosecution in Ohio, the trial court overruled objections from the district attorney’s office and allowed a polygraph examiner to testify regarding three lie detector tests that the suspect had passed. The court noted that prosecutors routinely used polygraphs against defendants, and thus had apparently accepted “the validity and reliability of the polygraph test process.”
However, in this case the prosecutors objected to the introduction of polygraph results – including a test conducted by a polygraph examiner regularly used by the state – that indicated the defendant was being truthful about his innocence. The defendant was subsequently acquitted at trial. See: State of Ohio v. Sharma, 143 Ohio Misc.2d 27, 875 N.E.2d 1002 (Ohio Com.Pl., 2007).
It appears that the use of lie detectors will continue to increase in our nation’s criminal justice system, with support from vested interest groups such as the American Polygraph Association and National Polygraph Association, a growing industry of polygraph examiners, companies like NITV that have a financial stake in the profitable lie detector market, and accommodating law enforcement agencies. This holds true at both ends of the system. For police interested in getting people into prison and at the back end where polygraphs are increasingly a condition of parole and probation, especially for sex offenders.
After all, it’s easier to browbeat and intimidate a polygraph suspect, and rely on dubious lie detection technology, than it is to conduct a real investigation. The prevailing attitude seems to be, “who cares if a few innocent people falsely confess along the way, or if polygraph tests produce highly questionable results? The main thing is we have a nifty machine to play with.”
And that’s the plain and simple truth.
Sources: Dallas Morning News, Sunset Advisory Commission Staff Report - Texas Polygraph Examiners Board (May 2008), Wisconsin State Journal, San Diego Union-Tribune, Arizona Republic, Associated Press, sjnewsco.com, Times Online (U.K.), bps.org.uk, wthitv.com, El Paso Times, Maryland Daily Record, www.cifa.mil
As a digital subscriber to Prison Legal News, you can access full text and downloads for this and other premium content.
Already a subscriber? Login
Related legal cases
Crowe v. City of Escondido, U.S.D.C. (S.D. Cal.)
|Cite||Case No. 3:99-cv-00241-JM-RBB|
State of Ohio v. Sharma
|Cite||143 Ohio Misc.2d 27, 875 NE2d 1002 (OH ComPl 2007)|
|Level||State Trial Court|