Why this view that sexual harassment allegations should be treated as their own roped-off category? I think the mistake can be attributed, in part, to our quick association between rape and criminal law—the incorrect assumption that sexual harms are inherently criminal and therefore that every adjudication must look like a criminal trial. An article in the Atlantic, for instance, complained that a certain university’s sexual harassment policies were “inconsistent with due process for accused sex criminals,” as though that were the relevant standard. If a criminal trial is the expectation, every institutional process will seem inadequate by comparison, no matter if it is fair and consistent with the institution’s policies for handling other misconduct.
The civil libertarian critic Judith Levine made a similar error when discussing how schools have responded to pressure from Know Your IX and related groups. “Colleges and universities,” she lamented, “have instituted codes of sexual conduct and quasi-judicial tribunals that ignore the constitutional protections afforded defendants in the criminal justice system.” But of course university disciplinary systems will not mirror criminal prosecution. That is not what they are called to do—legally, practically, or ethically. And the same goes for workplaces. When Minnesota Public Radio fired Garrison Keillor, the longtime host of A Prairie Home Companion, because of his alleged repeated sexual harassment of a former employee, a fan wrote in to protest that “Keillor was convicted without a trial.” Another MPR listener ended her membership because, in her view, the radio network was “turning [Keillor] into a criminal.” I have no idea if MPR’s process for investigating the harassment complaints was sufficient. But a trial, certainly, was not required. In firing Keillor, MPR did not “convict” him of a crime.
A similar set of assumptions came to the fore in the uproar over the Obama administration’s Dear Colleague Letter and its guidance on evidentiary standards. Even before the 2011 letter, most schools were already using “preponderance of the evidence”—the lower of the two common standards. The Dear Colleague Letter made that standard an official requirement for sexual harassment investigations into students. Personally, I think the preponderance is the right choice. It best balances the interests of both people whose educations or careers will be deeply affected by the outcome, promotes accuracy, and is used in comparable proceedings in and out of schools. That said, I do understand the objection that discipline can come with serious stakes, which perhaps suggests a higher standard. But regardless of what you might think about it as a matter of policy, there’s no question that long before the Dear Colleague Letter, the preponderance standard was used every single day in schools and civil trials for a wide range of harms, including sexual assault. Courts have confirmed that its use in student discipline is constitutional. Even the Trump administration required schools to use the preponderance standard when investigating harassment based on race or disability.
Yet from the response to the Dear Colleague Letter, you might have thought that preponderance of the evidence was a special kind of torture device. In an article titled “When Does a Watershed Become a Sex Panic?” Masha Gessen presented the preponderance standard as an example on the wrong side of that divide and decried that it would contribute to a reduction in consensual sex. (Never mind that someone could be sued for the same conduct under the same standard in a civil proceeding.) A professional organization representing professors wrote that the preponderance “conflicts with due-process protections.” In the Wall Street Journal, conservative political scientist Peter Berkowitz decried the preponderance as an “egregiously” low standard and bemoaned that schools now imposed “a presumption of male guilt.” (Regardless of the standard of evidence, a complainant always has the burden of proving her case.) Perhaps Berkowitz had forgotten that, when he had unsuccessfully sued Harvard for denying him tenure, his claim was subject to that same standard.
In a public statement, the libertarian Foundation for Individual Rights in Education, known as FIRE, wrote: “The preponderance of the evidence standard does not sufficiently protect an accused person’s right to due process. While this standard is acceptable for lawsuits regarding money, allegations of sexual violence or sexual harassment are far more serious than disputes that can be resolved by transferring money from one individual to another.” Accordingly, it urged, schools should use a higher standard of evidence, closer to that used in criminal trials. Note this: FIRE’s argument was not that the preponderance was the wrong standard for student discipline in general. It said that the preponderance was inappropriate for sexual assault.
The basis for FIRE’s argument is patently incorrect. The preponderance standard is not only used for suits for money damages. But even if that were our metric, lots of sexual harassment disputes are resolved through the transfer of money! That might come in the form of a jury verdict at the end of a civil lawsuit, or as a settlement to avoid litigation. Indeed, financial damages are often the only remedy available when a survivor brings a civil case against a perpetrator or institution. (And that payment might be more useful to her than her assailant’s incarceration.) Clearly, the seriousness of a sexual assault allegation does not foreclose a monetary remedy, any more than the seriousness of murder foreclosed the civil suit seeking money from O. J. Simpson. Here is the clincher: a civil suit for sexual assault will go through the same procedure as any other civil suit. This means that, like nearly all civil cases, it will be subject to the preponderance of the evidence standard. The gravity of the underlying harm does not dictate a higher standard of proof.
Clearly, FIRE was only thinking about sexual assault as a crime, ignoring the various other contexts in which those same claims may be adjudicated. It should have known better. But FIRE’s flawed assumptions are widely shared. That’s the rub: FIRE’s exceptionalist argument persuades many not in spite of its mistaken focus on rape as a crime and crime alone, but because of it.
SCRAPE THE SURFACE of exceptionalism and you’ll usually find misogyny beneath. Our collective fear of scorned, lying, irrational women who falsely “cry rape” has been ingrained in our culture and law for centuries, no matter the baselessness of the archetype. It serves as the unified, animating force behind the casual rape apologias, explicit doubts, and flawed policies that stand in the way of justice for survivors and an end to the violence. If we look at how rape victims have been treated historically, especially in criminal law, we can recognize contemporary exceptionalism as just the latest iteration of a long, terrible tradition. We believe we need special protections from sexual harassment allegations because we think that these accusations are particularly likely to be untrue.
The seventeenth-century English jurist Sir Matthew Hale did not invent this exceptionalism. But through his influential writings, he gave it form and force that shaped centuries of Anglo-American law for the worse. Rape, he said, “is an accusation easily to be made and hard to be proved, and harder to be defended by the party accused, tho never so innocent.” For this reason, Hale insisted, “we may be the more cautious upon trials of offenses of this nature, wherein the court and the jury may with so much ease be imposed upon without great care and vigilance.” As anthropologist Peggy Reeves Sanday has written, Hale’s views were animated by the idea that many women accused men of rape in order to extort blackmail. That fear of women as manipulative liars was a common one during Hale’s time and shaped rape law from the ground up. Women victims, for example, were expected to defer to the judgment of their husband or father as to whether to levy charges. A woman who instead came forward on her own behalf “conjured the image of the aggressive Eve, whose complicity with the serpent made all mortal women susceptible to the wiles of the devil: they were prone to be sexually voracious and enticing, spiritually weaker, and less rational,” writes Sanday of seventeenth-century British attitudes. “More likely than not, a woman bringing rape charges on her own was either a scorned woman or one out to blackmail the man who seduced her.”
Hale’s biases were not limited to theories about lying accusers. He did not believe a woman had the right to deny her husband sex, and thought men should lock up wives who spent too much money. Presiding over a 1662 witchcraft trial, Hale dismissed evidence that proved the innocence of the two accused women, and then sentenced them to death. “Contrary to Hale’s well-known concern to give rape defendants the benefit of the doubt,” Sanday notes, he showed “no evidence of giving defendants in witchcraft trials the same benefit.”
Hale’s warning about rape accusations was also empirically incorrect. “Rape is not, and never has been, an easy charge to make and a difficult one to rebut,” writes criminologist Gilbert Geis. “It certainly was not so before or in Hale’s time.” Although British law prescribed harsh penalties for rape, including capital punishment, convictions were so unheard of that one judge “entertained friends with jokes about it.” In Hale’s seventeenth century, far more women were executed for witchcraft than men were executed for rape. In eighteenth-century London, the 17 percent conviction rate for rape was the lowest of any capital crime, less than a third of the 56 percent conviction rate for burglary.
Nonetheless, Hale’s warning was adopted into British common law, which deeply shaped the law of its American colonies. Many U.S. courts instructed jurors in rape cases accordingly, telling them to be unusually suspicious of the charges before them, and continued to do so well into the second half of the twentieth century. To this day, the Model Penal Code—an authoritative source, originally published in 1962, to which states look in drafting their own laws—encourages the use of a Lord Hale instruction. It recommends that “in any prosecution before a jury [for sexual offenses], the jury shall be instructed to evaluate the testimony of a victim or complaining witness with special care in view of the emotional involvement of the witness and the difficulty of determining the truth with respect to alleged sexual activities carried out in private.” Extensive amendments to the Model Penal Code’s treatment of sexual assault were officially proposed by a formal committee of experts in 2019, but at the time of my writing, the American Law Institute has not adopted them.
Fortunately, on this point the Model Penal Code’s impact on law has diminished with time. Over the last decades of the twentieth century, anti-rape advocates and scholars successfully pushed states to stop requiring judges to echo Lord Hale’s words. In 1975, a defendant convicted of rape appealed to the California Supreme Court, arguing that the trial judge had improperly refused to read a Lord Hale instruction. In response, California’s highest court not only overturned its previous rule that the warning was required, but decided it was, in all circumstances, inappropriate. “Criminal charges involving sexual conduct are no more easily made or harder to defend against than many other classes of charges, and those who make such accusations should be deemed no more suspect in credibility than any other class of complainants,” the court wrote. “When such prosecutions present close evidentiary questions, they do so not because a victim—generally a woman—claims to have been sexually assaulted or abused, but because the alleged crime took place in evanescent circumstances difficult to reconstruct in court, a happenstance … which is indeed a typical occurrence in such non-sexual crimes as fraud and narcotics transactions.” In other words, the challenges that sometimes arise in sex crimes prosecutions arise in plenty of other prosecutions, too.
ADJUDICATING SEXUAL OFFENSES may indeed be harder than parsing some other kinds of wrongdoing. If a defendant goes on a videotaped shooting spree, or a student prints out a Wikipedia article and presents it as his own essay, those offenses will be easy to prove. The evidence there is clearer than, say, in a sexual assault investigation where the accused says that the alleged victim consented. In such a case, the distinction between sex and rape may turn on what was said and done behind closed doors without a witness. But the common instinct that sexual accusations are somehow uniquely unknowable overlooks two vital points.
First, there’s often much more evidence in a sexual harassment case than one might assume. Sometimes a coworker hears the “jokes” about an underling’s skirt that a boss makes at a meeting. Sometimes a college roommate hears the assault. And often the two parties create their own evidentiary trail. In some cases, the assailant directly admits to the misconduct, sometimes in the form of an apology and sometimes a boast. (Soon after my client Darbi’s rape, her assailant told classmates, “That’s just what I do”—“that” being sexual assault.) In other instances, a victim might seek medical care and a “rape kit,” or might tell his friends. He might suddenly stop going to work, class, or church—signals that something happened.
Second, even where there isn’t much evidence beyond the two parties’ stories, many adjudications that have nothing to do with sex or violence are effectively “he said, she said” situations, where the decision turns entirely on the credibility of the people involved. Take a fender bender with no video footage or witnesses, for example. I say I stopped at the stop sign before you barreled into me; you say I didn’t, so you had no chance to brake as I zoomed past. Either way, there are dents in both our cars. A judge, or an insurance agent, will have to decide which one of us seems more likely to be telling the truth. To do so, they will consider factors like the consistency of our stories and our apparent trustworthiness.
The same dynamic plays out, with much higher stakes, in self-defense cases. You say I came at you outside the bar in a drunken rage and shot you in the foot; you barely had time to draw the knife from your pocket. I say you charged at me with your knife drawn, threatening to kill me; I shot your foot to save my own life. If no one witnessed the altercation, the jury will be left to consider our testimony and weigh how credible we each seem.
And it’s not only courts that have to deal with such “evanescent circumstances,” as the California Supreme Court put it. Institutions handle them all the time. What if a school catches us with identical answers on a test? We sat next to each other during the exam. Did I peek at your papers or did you cheat off me? If neither of us was the clear academic superstar beforehand, the discipline board will have to make a decision based on, well, which one of us seems more likely to be the liar.
It is telling, then, that “he said, she said” has become shorthand for an unsolvable case. When it’s “he said, he said,” we are perfectly comfortable adjudicating. Add in a woman, and add in sex, and suddenly the waters are impossibly muddy.
What’s more, in a hypothetical case where there is truly no evidence beyond the parties’ testimony, and no basis at all for finding either person more credible than the other, the decision-makers have a clear mandate to follow: they have to find in the accused’s favor. In any conflict and any context, it is the complainant (or the prosecuting state or institution) who always carries the burden of proof, by whatever standard of evidence governs the proceeding. If there is no way to determine whether the harassment probably occurred, it is the alleged victim who will lose.
So why, then, the insistence on special warnings about “easily made” rape allegations? Hale and his legacy make clear that the logic of exceptionalism is that sexual assault allegations must be so specially guarded against because they are uniquely likely to be false. But we have no actual evidence that sexual violence is misreported at especially high rates. As I mentioned earlier, it’s impossible to know exactly what percentage of reports are false, because there is no failproof method of distinguishing the real from the fake. Rarely do fraudsters fess up, and any outside observers’ guesses at whether a report is true or not will be based on incomplete information and filtered through their own biases. To reach an informative estimate, researchers have to scrutinize police classifications of false reports according to consistent criteria. At the least, they must apply a clear definition of a false report that doesn’t lump untrue complaints together with credible complaints that were never prosecuted. Studies that follow these guidelines converge on rates between 2 and 8 percent, consistent with rates of false reporting for other crimes. To put this in perspective, by any of these counts a man is significantly more likely to be sexually assaulted than to be falsely accused of sexual assault.
No matter how many studies confirm low rates of false rape reporting, men’s rights activists—anti-feminists who believe that men, not women, are the disadvantaged sex—insist the number is far higher. These critics often point to a 1994 study by Eugene Kanin, which reported that 41 percent of the 109 sexual assault reports made to a police agency were false. But the determination of whether reports were true or not was made by the detectives, and not reevaluated by Kanin—or anyone else. Given what we know about law enforcement biases, there’s good reason to double-check their classifications. Sexual assault researcher David Lisak explains that Kanin’s approach “violates a cardinal rule of science, a rule designed to ensure that observations are not simply the reflection of the bias of the observer.” As Lisak notes, “the greater the scrutiny applied to police classifications, the lower the rate of false reporting detected.”
Another common error in statistical analyses is eliding the difference between people who lie and victims who lack sufficient evidence to confirm their account. During Brett Kavanaugh’s confirmation process, for example, some of his defenders latched on to a then recent Pentagon study showing that the military dismissed nearly a quarter of the rape complaints it received for lack of proof. But, of course, a lack of proof does not mean an assault did not occur—it just means a decision-maker concluded that the complainant did not have the evidence necessary to pursue the case.
HALE WAS NOT the only English jurist to shape U.S. rape law for the worse. For centuries, English common law required victims of violent crimes to “hue and cry”—basically, to immediately call out to their neighbors to get the assailant. Over time, courts largely abandoned that requirement, noticing that it was not, in fact, a particularly helpful way to catch criminals. They long retained it, though, for sexual violence alone. In the nineteenth century, U.S. courts adopted a modified approach: a rape case was allowed to proceed if the victim had not “hued and cried,” but her failure to come forward quickly could be used against her by the defense.
The idea behind what is known as the “prompt complaint requirement” comes from what judges thought rape victims would always do, not from how actual victims often act. In 1900, for example, the all-male Utah Supreme Court postulated that “the natural instinct of a female thus outraged and injured prompts her to disclose the occurrence, at the earliest opportunity, to the relative or friend who naturally has the deepest interest in her welfare.” A woman who delayed her report, then, was unnatural—a liar. Ninety-one years later, the Supreme Court of Illinois reinforced the rule, writing that “it is entirely natural that the victim of forcible rape would have spoken out regarding it, and the fact that she did not do so would in effect be evidence of the fact that nothing violent had occurred.”
We now know the many reasons a victim might not report to the police at all, let alone promptly. But the Model Penal Code still contains a uniquely short statute of limitations of only three months for sexual offenses. Its statute of limitations for other first-degree felonies, meanwhile, is six years—except for murder, which has none. The code’s justification for this choice, as formally memorialized in 1980, explicitly invokes the myth of widespread false rape accusations by scorned women: “The requirement of prompt complaint springs in part from a fear that unwanted pregnancy or bitterness at a relationship gone sour might convert a willing participant in sexual relations into a vindictive complainant.” In a critique, early rape law scholar Susan Estrich noted that the Model Penal Code commentary “is startlingly attentive to the problem of the vindictive, spurned woman, but silent about the woman who legitimately worries about the receptiveness of police, prosecutors, juries and even friends or employers to a report that she was raped.” Luckily, the states have outpaced the Model Penal Code in abandoning the prompt complaint requirement.
We cannot blame the Brits for everything. One of the exceptional features of American rape law is very much of our own making: the requirement of corroborating evidence—that is, evidence other than a victim’s testimony—to support a conviction. The first law to require corroborating evidence for sexual abuse convictions was passed by the New York legislature in 1886. The result, yet again, was a unique burden placed on allegations of rape. In one bizarre 1960s case, a defendant was convicted of attempting to rape a woman and stealing her coin purse. He appealed the conviction, pointing out that the victim had no corroborating evidence for either of the claims. The rape verdict was overturned, but the state’s highest court held that the theft verdict could stand; the opinion explained that corroboration was required only for the rape itself, not for other, nonsexual crimes committed at the same time. An article published the next year in the prestigious University of Pennsylvania Law Review explained that the corroboration requirement was necessary because “women often falsely accuse men of sexual attacks to extort money, to force marriage, to satisfy a childish desire for notoriety, or to attain personal revenge. Their motives include hatred, a sense of shame after consenting to illicit intercourse, especially when pregnancy results, and delusion.” The author cited one psychiatrist’s assertion that “most women” have rape fantasies, and that “neurotic individuals” are unable to distinguish those fantasies from reality.
Over the course of the twentieth century, New York’s corroboration rule spread. In 1904, the Supreme Court of Georgia declared that a corroboration requirement was legally necessary. Without such a requirement, the court explained, “every man is in danger of being prosecuted and convicted on the testimony of a base woman in whose testimony there’s no truth.” Absent such a protection, the court continued, “the man is powerless.” Interestingly, the court cited Lord Hale for support—even though he had never endorsed such a rule requiring corroborating evidence. In fact, in a rare moment of insight, Hale had explained that an alleged rape victim’s own testimony does constitute evidence on its own; whether it is credible is a question for the jury. But the Georgia justices could perhaps be forgiven for forgetting the particulars: the big takeaway from Hale, after all, is that women lie and men must be protected.
In 1965, the Supreme Court of Minnesota explained in clear terms why it, too, required corroboration for sex crimes. In an opinion overturning the conviction of a man accused of sexually abusing his daughter, the court quoted at length from Dean John Henry Wigmore’s treatise on evidence, the most influential work on the subject in American jurisprudence for most of the twentieth century. Special rules were required in rape trials, the court said, because, according to Wigmore, the “psychic complexes” of “errant young girls and women … are multifarious, distorted partly by inherent defects.… One form taken by these complexes is that of contriving false charges of sexual offences by men. The unchaste (let us call it) mentality finds incidental but direct expression in the narration of imaginary sex-incidents of which the narrator is the heroine or the victim.… The real victim, however, too often in such cases is the innocent man.”
The U.S. Court of Appeals for the D.C. Circuit, the country’s most powerful tribunal after the Supreme Court, agreed in a 1973 opinion. “It is established law in this jurisdiction that a person may not be convicted of a ‘sex offense’ on the uncorroborated testimony of the alleged victim,” the court wrote. “Complainants all too frequently have an urge to fantasize or even a motive to fabricate.” As with the other exceptional rules for sexual assault, the corroboration requirement is still recommended by the Model Penal Code, though only a handful of states today include it in their statutes.
These three exceptionalist rules—the Lord Hale Instruction, the prompt complaint requirement, and a requirement of corroborating evidence—were obvious targets for feminist legal reformers in the late twentieth century. But even as states revised their rape laws, the biases and myths upon which they relied continued to circulate in our culture. As the feminist legal theorist Michelle Anderson puts it: “After scrubbing down the foul Model Penal Code and tidying up a dozen or so state codes on the bookshelf, one might toss these relics—prompt complaint, corroboration, and cautionary instructions—into the dustbin of historical misogyny, declare a victory for the second wave of feminism, and go home. But the cultural dirt from the criminal law has drifted into an adjacent room previously assumed to be uncontaminated.”
That adjacent room includes institutional responses to sexual harassment and public debates about what fair process should look like. Today, as throughout history, exceptionalist policies treat allegations of sexual harassment as uniquely suspect. We see the continuation of this long tradition in the demands that students accused of sexual harassment must receive extra protections, in the special rights for such students that DeVos guaranteed and Missouri legislators drafted, and in the vilification of a perfectly common standard of evidence simply because it was applied to accusations of sexual assault. We see it, unsubtly, in the “justifications” that some critics provide for their preferred policies, sweeping declarations that the truth of sexual assault allegations is uniquely unknowable and the risk of wrongful blame uniquely high. Invariably, these assertions are accompanied by no citation, as though they are self-evident—as though we should base national policy on these writers’ instincts, which might be another word for biases. There isn’t much distance between ancient misogynists and some of today’s critics, such as Berkowitz, for example, who insisted in his attack on the Dear Colleague Letter that rape investigations are particularly prone to erroneous outcomes because “where erotic desire is involved, intentions can be obscure, passions conflicting, the heart murky and the soul divided.” In other words, women fling around rape allegations because they don’t know what they want.
It is easy to see the ghost of Potiphar’s wife—the lying, scorned woman who cries rape—in those sweeping, inaccurate, age-old generalizations. And the impact of long-standing legal traditions on contemporary policy means that misogyny has an even broader effect than one might initially realize. In Lord Hale’s time, rape was generally defined as a crime against a woman; there’s no question that his warnings about “an accusation easily to be made and hard to be proved” had women in mind. In recent decades, rules against sexual harms have expanded to encompass victims of all genders, but the exceptionalism persists. As a result, those ancient sexist myths now haunt every allegation, not only those brought by women. Stereotypes about one sex have become stereotypes about all survivors.