Liar, Liar: Polygraph Evidence, The Jury As Lie Detector, Cross-Racial IDs & Quantifying Probable Cause & Reasonable Doubt

In United States v. Scheffer, 523 U.S. 303 (1998), 

Air Force airman Edward Scheffer volunteered in March 1992 to work as an informant on drug investigations, which rendered him subject to drug testing and polygraph examinations. Soon after taking a drug test that April, but before the results were known, Scheffer agreed to take a polygraph test, which “indicated no deception” when he denied using drugs since enlisting. If Scheffer was lying, the drug test was not, as it revealed the presence of methamphetamine.

During his ensuing court-martial on charges of methamphetamine use, Scheffer raised an “innocent ingestion” theory, but the military judge precluded him from introducing the polygraph evidence in support of this claim pursuant to Military Rule of Evidence 707, which stated in relevant part: “(a) Notwithstanding any other provision of law, the results of a polygraph examination, the opinion of a polygraph examiner, or any reference to an offer to take, failure to take, or taking of a polygraph examination shall not be admitted into evidence.”

In finding that the application of Rule 707 to the polygraph results did not violate Scheffer's right to present a defense, Justice Thomas, writing for the Court, found that "there is simply no way to know in a particular case whether a polygraph examiner's conclusion is accurate, because certain doubts and uncertainties plague even the best polygraph exams." Accordingly, the Court held that Rule 707 rationally and proportionally advanced the government's "legitimate interest in ensuring that reliable evidence is presented to the trier of fact in a criminal trial."

It was another portion of Justice Thomas' opinion that got the most ink even though this portion of the opinion was not joined by a majority of the Court. In this portion of the opinion, Justice Thomas averred that the problem with polygraph evidence is that "[a] fundamental premise of our criminal trial system is that 'the jury is the lie detector.'" And, according to Justice Thomas,

Jurisdictions, in promulgating rules of evidence, may legitimately be concerned about the risk that juries will give excessive weight to the opinions of a polygrapher, clothed as they are in scientific expertise and at times offering, as in respondent's case, a conclusion about the ultimate issue in the trial. Such jurisdictions may legitimately determine that the aura of infallibility attending polygraph evidence can lead jurors to abandon their duty to assess credibility and guilt. Those jurisdictions may also take into account the fact that a judge cannot determine, when ruling on a motion to admit polygraph evidence, whether a particular polygraph expert is likely to influence the jury unduly.

Before my recent blog exchange with Erica Goldberg (see here, here, and here), I never understood why a majority of the Court did not join in this portion of Justice Thomas' opinion. Now, I think I do, and this understanding further informs the issue of whether courts should quantify probable cause and reasonable doubt.

The answer for why the majority of Justices did not join the "jury is lie detector" portion of Justice Thomas' opinion seems to be the collateral effect that this portion of the opinion could have on polygraph evidence in bench trials. If the problem is that jurors could be taken, hook, line, and sinker by polygraph evidence, doesn't that mean that defendants should be able to present polygraph evidence in bench trials because judges aren't subject to the same infirmities as jurors? After all, that is the reasoning that courts use in finding that the Bruton doctrine doesn't apply in bench trials.

And, if we strip away the jury argument, all we are left with in Scheffer is the "polygraphs aren't perfect" argument. But how imperfect? In Bennett v. City of Grand Prarie, Tex., 883 F.2d 400, 405 (5th Cir. 1989), the Fifth Circuit found that "Polygraph exams, by most accounts, correctly detect truth or deception 80 to 90 percent of the time." Bennett, however, was not a case in which the court was deciding whether to admit polygraph evidence at trial; it was a case in which the court was trying to determine whether a magistrate properly considered polygraph evidence in making a probable cause determination.

And, like many other courts, the Fifth Circuit concluded that the magistrate's reliance of the polygraph evidence was proper, concluding that

The fear that a jury may overestimate the probative value of such evidence when considering an individual's guilt or innocence-the factor that led some courts to preclude and other courts to limit the use of polygraph exams as evidence at trial-is absent when a magistrate relies on such an exam to determine whether there is probable cause to issue an arrest warrant. Unlike a lay jury, a magistrate possesses legal expertise; when determining probable cause, he is unlikely to be intimidated by claims of scientific authority into assigning an inappropriate evidentiary value to a polygraph report or to rely excessively on it.

Okay, but then why not allow defendants to present polygraph evidence in bench trials? This takes me back to Professor Goldberg's most recent post, and, specifically, one of the comments she made in response to a question by Orin Kerr: 

At the highest level of generality, I believe that quantifying probable cause will add integrity and accountability to the jurisprudence. It’s hard for us citizens to determine if judges are making fair decisions about probable cause if we don’t even know how much suspicion is required. Proof beyond a reasonable doubt is less in need of quantification because it’s so far out there on the spectrum of proof. I think people intuitively understand what a reasonable doubt is because it is such an extremely high standard, but the probable cause decisions are all over the place. This makes them elusive to students, scholars, and citizens, who then lose faith in the decisions.

Below that, quantifying probable cause will give courts some understanding of how to determine if probable cause exists. Sure, there is precedent and rules for judges to apply. For example, running from the police is not enough on its own to constitute probable cause, but running plus some other suspicious behavior is sufficient. However, these rules are almost beside the point because they say nothing about what the actual standard is.

Below that, quantifying probable cause will give police something more standardized to shoot for, besides untethered rules, when making probable cause determinations.

Finally, because so much evidence has a quantitative component (DNA evidence, drug dog alerts, fingerprint matching), we can use numbers to assess the probative value of this evidence. Until now, the presumption has been that a positive alert automatically gives rise to probable cause, but we cannot know that unless we grapple with the numbers. This, in turn, will make quantifying probable cause inevitable, in situations where all of the suspicious evidence has numerical error rates (false DNA matches, for example).

Now, this isn't the type of quantification that Goldberg is talking about, but let's say that we were really trying to numerically quantify probable cause and reasonable doubt. If I'm just spitballing, I might put probable cause at about 33%. If there's about a 1/3 chance that, for example, the fruits of a bank robbery are at a suspect's house, a judge should issue a search warrant for the search of that house. And, again, if I'm spitballing, I might put reasonable doubt at 10%. In other words, if a jury or a judge in a bench trial thinks that there is a 10% chance that the defendant did not commit the crime(s) charge, the verdict should be "not guilty."

Now, you might quibble with the percentages that I have attributed to probable cause and reasonable doubt, but I doubt that many people would think that the percentage of reasonable doubt required for an acquittal should be higher than the percentage of probable cause required to authorize a search. In fact, if you think about it, reasonable doubt is more analogous to "reasonable suspicion" than it is to "probable cause." For instance, in the Terry context, an officer can conduct a stop if he has reasonable suspicion, supported by specific and articulable facts, that criminal activity my be afoot. I think that most would agree that if a defendant could present enough evidence at trial pointing to an alternate suspect such that a Terry stop of that alternate suspect would be warranted, the defendant would have created reasonable doubt. And, of course, establishing reasonable suspicion requires a lesser showing than establishing reasonable doubt.

So, back to Bennett and quantifying probable cause and reasonable doubt. Here's why I think it might be valuable to quantify both in addition to quantifying different types of evidence admitted and excluded at trial. Without quantifying either standard, we can look at Bennett and Scheffer and think that both make intutive sense. Bennett makes sense because if polygraphs are 80-90% reliable, of course they are reliable enough for magistrates to consider them in deciding whether to issue warrants. Of course, we also need to consider the fact that 80 or 90% reliability doesn't mean that it is 80% or 90% likely that police will find, say, evidence of the bank robbery at the suspect's home. Even if the suspect fails a polygraph question asking whether he robbed a bank, that doesn't mean that evidence of that bank robbery was ever at his house or that it is still there. Meanwhile, Scheffer makes sense because criminal trials require guilt beyond a reasonable doubt, and the government has a legitimate interest in only allowing the factfinder to consider reliable evidence in making the decision to convict or exonerate

But when we dig into the numbers, these intuitions collapse. What is the magistrate in Bennett actually doing? He's trying to decide whether it is 33% (or whatever percentage) likely that evidence of the bank robbery is at the suspect's house. And what is the court-martial in Scheffer actually doing? It is trying to decide whether it is 10% (or whatever percentage) likely that the defendant did not commit the crime charged. So, how is it that we allow the magistrate to consider polygraph evidence in deciding whether to issue a warrant while we don't allow the court-martial to consider the polygraph evidence in deciding whether there is reasonable doubt given these percentages? And keep in mind that, unlike in the probable cause context, there is likely to be a perfect correlation between polygraph accuracy and actual innocence. In other words, if the polygraph is 85% accurate in determining that Scheffer honestly didn't know that he ingested drugs, then there is an 85% chance that he did not know that he ingested drugs.

Of course, Justice Thomas' answer is that the court-martial in Scheffer and a jury in criminal court could be unduly swayed by polygraph evidence. But what about in a bench trial? "The federal government, especially its defense and intelligence arms, uses polygraph testing tens of thousands of times a year in screening prospective and current employees for espionage." And yet, you're telling me that a judge in a criminal trial shouldn't be allowed to use polygraph evidence to determine whether there is reasonable doubt as to whether the defendant committed the crime charged?

So, what evidence can jurors consider? Well, for one, they can consider eyewitness testimony. And how accurate is eyewitness testimony? Well, studies vary, but, according to an expert in Commonwealth v. Alston, 1999 WL 1204002 (Mass.Super. 1999), it is 75%. And what about cross-racial identifications? According to the same expert, if the eyewitness has a history of interacting with people of other races, at best, the accuracy rate is 55%. If not, the number drops. It is thus not surprising that "[a]mong the first 200 DNA-based exonerations, almost all involved 'stranger' misidentifications, and nearly 48% of the misidentifications were made by a witness of a different race than the suspect." Cynthia E. Jones, The Right Remedy for the Wrongly Convicted: Judicial Sanctions for Destruction of DNA Evidence, 77 Fordham L. Rev. 2893, 2930 (2009).

And yet, despite these percentages, we allow eyewitness identifications and cross-racial eyewitness identifications as evidence to prove guilt beyond a reasonable doubt while we preclude polygraph evidence to prove reasonable doubt. Indeed, some courts even preclude defendants from presenting expert testimony on the inaccuracy of cross-racial identifications because, you guessed it, it would usurp the role of the jury as lie detector. Well, guess what,

Although the psychological research has not focused on cross-racial lie detection or remorse detection per se, there is some evidence that cross-cultural lie detection is less accurate, and there are reasons to think that the same phenomenon would be observed in cross-racial settings. Michael M. O'Hear, Appellate Review of Sentences: Reconsidering Deference, 51 Wm. & Mary L. Rev. 2123, 2147 (2010).

The point of all of this is that there is a good argument for quanification. Quantifying probable cause. Quantifying reasonable doubt. Quantifying the accuracy of polygraphs. Quantifying the accuracy of cross-racial identification. Quantifying the accuracy of excited utterances. Without quantification, a judge relying on a polygraph to find probable cause while another judge cannot rely on a polygraph to find reasonable doubt might make sense. But when we dig into the numbers, it does not.

-CM