Last week, the Office for Civil Rights in the Department of Education (DOE) withdrew the Obama administration's Dear Colleague Letter of 2011. The letter had outlined a set of procedures that all educational institutions receiving federal funds-under Title IX of the 1972 Civil Rights Act-must use to "investigate, adjudicate, and resolve allegations of student-on-student sexual misconduct." The Obama rules had required that sexual misconduct cases be tried under a "preponderance of the evidence" standard instead of the higher standard of "clear and convincing evidence," which are traditionally used in such cases.
The first standard allows for a guilty finding where the evidence is more certain than not; with the second standard, the accuser must show, roughly speaking, a 75 percent certainty. The letter permitted the accuser who lost at the university or college tribunal to bring appeals to overturn the initial determination. It strongly discouraged the use of cross-examination by parties or their attorneys, and refused to allow covered institutions to involve law enforcement officials in these investigations of criminal conduct.
The DOE's recent reversal of policy under Betsy DeVos rests on the conviction that the regulations set forth in the letter violate the procedural due process of the accused students, who face heavy sanctions after a final judgment of guilt, including suspension or expulsion from their universities, which, for foreign students, could lead to deportation. Defenders of the 2011 Obama Guidance sharply denounced the new policy, which DeVos outlined in a speech at George Mason University several weeks ago. Thus the group End Rape on Campus wrote: "Today, Betsy DeVos and the Trump Administration chose to tip the scales in favor of rapists and perpetrators. Rolling back this guidance is an affront to the students, survivors, and allies who have fought to bring the sexual assault epidemic out of the shadows."
There are a number of responses to this statement. On the theoretical side, the great challenge of any set of rules is how to best minimize the two errors that arise when the facts are uncertain: making wrong convictions of innocent people and allowing those who have committed crimes to go free of the charges.
Most disputes between private parties involve the payment of money or the transfer of property, so the lower preponderance of evidence standard makes sense. But at the same time, in order to minimize the overall level of error, the parties are given extensive rights to present their own evidence through their attorneys, to examine documents, to cross examine witnesses offered by the other side, to exclude many forms of hearsay and prejudicial evidence, and to have judges give juries clear instructions on the applicable legal rules. In contrast, in criminal cases, where the consequences of incarceration are always grave, it is today a constitutional requirement in both state and federal cases that the government prove each element of the case beyond a reasonable doubt.
Academic hearings on sexual harassments are not criminal cases, but the weight of an erroneous judgment falls more heavily on the accused than the accuser. Within this system, the accuser whose claim is rejected suffers no penalty for wrongful prosecution. The accused, however, is routinely subjected to severe sanctions even before the final judgment is issued. And once suspension and expulsion are part of the record, it is difficult if not impossible for the convicted student to enroll in another institution.
The severity of sexual misconduct cases falls midway between civil and criminal cases, which is why most universities had prior to 2011 settled on the higher clear and convincing standard of proof. This more demanding standard blocks a finding of guilt when someone charges rape after a consensual hookup, sometimes (a fact that can be disputed) involving the use of alcohol by either or both parties. Under the bare preponderance standard, either verdict is defensible in such an ambiguous case. Under the standard of clear and convincing evidence, the accuser would need to present more independent evidence. Instead of disputes going to litigation, they can be resolved by more informal mechanisms, like counseling and mediation.
The discussion of these issues often proceeds at a highly abstract level. But the miscarriages in justice come into clearer focus by looking at actual litigation stemming from suspensions put into place under the 2011 Obama rules. There are many books and articles that document the risks in these cases. But on this occasion, I shall recount the 2016 decision in John Doe v. Alger. The decision was entered by Judge Elizabeth K. Dillon, an Obama appointee to the Western District of Virginia in 2014. Her detailed opinion found James Madison University in violation of the Due Process Clause of the Fourteenth Amendment: "nor shall any state deprive any person of life, liberty or property without due process of law."
The case turned on the evidence and the procedures used to evaluate it. The first warning sign in this case was that Doe received notice on November 6, 2014 of the charges filed by Roe for events that had occurred either late August 22 or early August 23 of that same year. Roe, like Doe, was a first-year student in the same dormitory but on a different floor from Doe. Doe did not dispute that sexual intercourse had taken place, but strongly insisted that Roe had given her full consent. Roe claimed that Doe had taken advantage of her when she was under the influence of alcohol. Doe claimed, and Roe disputed, that the couple had sexual intercourse several days later when both were sober.
One risk in these cases is that a party who regrets the encounter afterwards might deny consent at some later late. To avoid these problems, serious rape charges are best made right after the incident so that physical and testimonial evidence can be preserved. Indeed, police departments are generally loath to investigate such tardy charges, which is one reason why the Obama guidelines treated law enforcement as irrelevant to university proceedings. Nonetheless, Roe's charges were sufficient to trigger a JMU order to Doe to change dormitories before any determination of guilt had been made, even though there had been no untoward interactions between the two in the ten weeks after the August incident.
On December 5, 2014, both sides repeated their rival accounts before a three-member JMU panel. Perhaps the decisive piece of evidence leading to a verdict of "no responsibility" for Doe was testimony from Roe's roommate that "she did not recall seeing Roe drink that evening and that Roe was "completely fine" when they separated earlier that evening." She further noted that although Roe was "droopyish" and "tired" when she escorted Roe back to her room, she did not seem drunk.
But two weeks later, Roe appealed the adverse decision and presented in writing new evidence, as she was allowed to do under the Obama rules. By this point, however, the students had gone home for break, so Doe did not learn of Roe's new submissions in time to respond. That procedural glitch, however, didn't stop the university's three-person appellate disciplinary panel from summarily upending the previous decision without officially hearing either party or any witness.
It issued a "sanction increased" order during the break that barred Doe from returning to campus, even to take classes. The two pieces of evidence on appeal were Roe's claim that her roommate had lied and a voice mail Roe left with her friend noting that she was drunk. But that voice mail was dated August 21, about 24 hours before the alleged assault had taken place. The point was ignored. Eventually, the case made it to the district court, where Judge Dillon held that Doe had presented a valid claim that his constitutional rights had been abridged by the faulty procedures of JMU.
The appellate proceedings were a farce that went forward only because of the defective procedures introduced by the 2011 Obama directive. No new evidence should ever be allowed in on an appeal because it gives the accuser two bites of the apple. All evidence that comes in on any rehearing should be subject to cross-examination. Proceedings should not take place during academic breaks. Here, the evidence suggests that the initial panel got it right even under the preponderance of the evidence standard.
But the impact of these frequent procedural errors is softened if the clear and convincing evidence standard is used. So when critics of the DeVos policy worry about "survivors" of the "epidemic" of rape, they should soberly look at the evidence in these cases one at a time. Many times, they will be forced to come to a different conclusion about what happened.