Lie Detection, Non-Transparency, and Power: Judge Kavanaugh Defending the Emperor with No Clothes

As Mark Harris recently observed in his Wired article featuring some of my research — “The Lie Generator: Inside The Black Mirror World of Polygraph Job Screenings” — federal agencies have generally attempted to keep their polygraph program data from researchers, especially vis-a-vis bias and efficacy. Federal courts have fairly consistently backed them up, citing the vital role polygraph programs play in law enforcement. However, at the same time, it is longstanding government and scientific consensus that there is no such thing as a lie detector, and polygraph programs actually jeopardize the very national security that they are intended to promote. Researchers can thus neither purport to comprehend nor effectively contest this supposedly vital nature of polygraph programs to law enforcement, without greater transparency of the federal program polygraph data that apparently disproves all of the canonical governmental reports on polygraphs—from the 1965 U.S. Committee on Government Operations’ to the 2003 National Academy of Sciences’—as well as most of the scientific literature on psychophysiological deception detection for the last fifty years.

This unified front of non-transparency, and the Catch-22 of secret justifications for keeping this secrecy intact, is exemplified by Sack v. United States Department of Defense. Attorney Kel McClanahan argued this case on my behalf, Circuit Judges Tatel, Griffith, and Brett Kavanaugh of the U.S. Court of Appeals, D.C. Circuit Court decided it, and Judge Kavanaugh authored the ruling on May 20, 2016. Judge Kavanaugh’s defense of non-transparency here is important in the context of his current status as a prospective U.S. Supreme Court Justice. Much has been made recently, and incorrectly, of his regard for polygraphs (e.g., “Brett Kavanaugh Once Said Polygraphs Are A Good Tool. Now He Says They’re Unreliable“; “A False Charge on Polygraphs“). Mr. McClanahan has already corrected the record on some of these confusions: This case was primarily about researchers’ access to federal polygraph program data that was denied under Exemption 7(E), as well as students’ access to educational requestor status when they file Freedom of Information Act (FOIA) requests.

The latter is not directly an issue of transparency, but rather of access to public records requesting processes—a means of promoting transparency. Thanks to Mr. McClanahan’s persistence, Judge Kavanaugh kept the Office of Management and Budget (OMB), the Department of Defense (DOD), and the District Court before him from unfairly stratifying student access to FOIA without statutory basis in a way that would have hurt students’ access to public records. The DOD had refused to classify me, when I was a graduate student requesting records, as an educational requestor, citing an OMB Guideline on student requests. Kavanaugh ruled out that this guideline lacked statutory basis, and that it even seemed OMB was trying to manufacture a reason to charge students more than teachers. Kavanaugh called out corruption where he saw it, telling them that he knew FOIA was “grossly overburdened,” but they couldn’t “provide relief for the FOIA bureaucracy on the backs of students.” That was logical, fair, and egalitarian. This set an important precedent for future students to have more effective access to public records requests.

However, without transparency in the form of responsive records disclosures, FOIA access itself is ineffectual or worse. It could even be conceived of as a resource-capturing diversion, if responsive records are not disclosed when they should be through request processes that take time, money, and intellectual resources. On this score, Kavanaugh wrongly enshrined non-transparency in the law, undermining the principle of accountability of government to which public servants in democracies owe their highest commitment.

Exemption 7 lets the Government withhold records compiled for law enforcement purposes, if they can show that producing those records would cause a specified harm. Under 7(E), that harm is disclosing “techniques and procedures for law enforcement investigations.”

In section III paragraph 2 of his opinion ruling that withholding the records fits under 7(E), Kavanaugh says the Government made that case. But there is no evidence that they did. Disclosing federal polygraph program data would not harm law enforcement. Rather, polygraph programs themselves, according to the National Academy of Sciences’ 2003 polygraph report, harm law enforcement by generating too many inaccurate results to be good for security in contexts such as screening National Lab employees for spies—both missing too many real spies and implicating too many innocent employees.

Bias and efficacy data on federal polygraph programs that was not made available to the National Academy scientists or other researchers might show problems with bias and efficacy. But this would hurt polygraph programs, not law enforcement in the sense the Court means. Perhaps because most polygraphers are current or former law enforcement, law enforcement seem to have trouble telling the difference. Yet it is bias and backfiring programs such as polygraph programs, not disclosing information about these problems, that undermine rule of law and national security.

In section III paragraph 4, Kavanaugh says “The reports at issue in this case assist law enforcement agencies in taking ‘proactive steps’ to deter illegal activity and ensure national security. As the Government notes, law enforcement agencies use polygraphs to test the credibility of witnesses and criminal defendants.” But rather than deterring illegal activity, polygraphs make a mockery of law enforcement who use them because it is widely known that they do not work. This makes law enforcement look unprofessional and ineffective, encouraging illegal activity.

Moreover, to the extent that law enforcement agencies really use polygraphs to test witness and defendant credibility, releasing data on polygraphs’ bias and efficacy should either help bolster law enforcement’s credibility—or cause law enforcement to question their use of polygraphs. If the data are so good that they prove polygraphs are much better than scientists have long thought they are, then law enforcement should be excited to release that data in order to deter illegal activity and promote national security, according to Kavanaugh’s own logic.

In section III paragraph 5, Kavanaugh says “The Government has satisfactorily explained how polygraph examinations serve law enforcement purposes.” But explaining that would require showing data on polygraph programs that contradicts everything scientists know about how they undermine security. That is exactly what Kavanaugh permits the Government to refrain from doing.

In the same paragraph, he goes on to claim “It has also explained how the reports assessing the efficacy of those examinations and identifying needed fixes likewise serve law enforcement purposes” by helping “ensure that law enforcement officers optimally use an important law enforcement tool.” But again, unless the government’s data on polygraph efficacy contradicts everything scientists have known about polygraphs for fifty years, this non-transparency in efficacy data merely covers up inefficacy. Kavanaugh is effectively helping law enforcement cover up their incompetence, protecting “the man behind the curtain” with judicial non-transparency.

In section III paragraph 9, Kavanaugh repeats his faith in non-transparency (and the District Court’s before him). He says that releasing even “reasonably segregable” parts of reports on polygraph efficacy “whether pertaining to the strengths of polygraphs, their weaknesses, or anything else—would create ‘at least a risk that subversive individuals will be armed with advanced knowledge of the procedures used by the United States to screen applicants for sensitive employment positions and security clearances’ ” (Sack v. Dept. of Defense, 6 F. Supp. 3d 78, 91 (D.D.C. 2013)). But this is utterly illogical. Letting researchers evaluate programs consistently charged with bias and inefficacy does not help subversives. Rather, promoting freedom of information, scholarship, and the advancement of science in a free society improves the functioning and accountability of government including law enforcement in that society. Kavanaugh does not seem to understand how democracy works.

Government secrecy around federal polygraph programs guards weakness, not strength. That weakness is already publicly known. Defending it with non-transparency undermines rather than strengthening national security.

It is the role of federal judges to check and balance governmental power in other branches when citizens ask them to. In this case, Judge Kavanaugh was asked to check several executive branch abuses of power. In one area—OMB’s imaginative FOIA fee structure guideline limiting student access to public records requests—he forced the executive to back down and stop denying students their rightful educational requestor status. But in another—the application of Exemption 7(E) to permit DOD to withhold polygraph program data—he upheld non-transparency in a way that hurt national security… But bowed down to its chant in the name of the status quo.

The form of national security that Kavanaugh defended in this decision was an emperor with no clothes.