The Messy Truths About False Rape Allegations
Currents
"Almost no one is falsely accused of rape.” So declared The Cut in 2018. And they’re not alone. In recent years, it has become common to hear that only a tiny fraction of rape accusations are false, typically in the low-to-mid single digits. Particularly since the 2017 #MeToo movement, repetitions of such claims are often made with a kind of moral certainty — the implication being that we essentially need not worry ourselves overmuch about due process. If an accusation is made, this attitude suggests, the accused is almost certainly guilty and should be treated as such. But where do these numbers come from? What do we even mean by “false report?” When evidence is murky, or when there isn’t an official investigation, how does anyone know whether a report is false or not?
As the nonprofit End Violence Against Women International dutifully put it, “In reality, no one knows — and in fact no one can possibly know — exactly how many sexual assault reports are false.” Nonetheless, they provide several figures for journalists, academics, and activists to cite authoritatively, including their estimated false report range of 2–8%. But before we get into why they are exactly right about our collective uncertainty and inability to know precisely how many reports are true or false, first the inevitable throat clearing. Without question, the vast majority of women (or men) do not falsely accuse others of rape, and the majority of rape reports are unlikely to be malicious false reports. It should go without saying that rape and sexual assault are very serious issues and that all women deserve to live free of sexual violence. Nothing that follows here is intended to argue otherwise. Nonetheless, the truth is that the proportion of rape accusations that are mistaken is probably higher than advocates would like us to believe.
When single-digit prevalence figures for false reports are cited, they often appear to be used to suspend any disbelief about future rape allegations or a specific allegation in question, even when the circumstances differ from those in which the statistics were gathered. Part of the confusion seems to stem from failing to recognize that the guidelines for what constitutes a false report are fairly narrow. For example, a 2014 National Institute of Justice study of rape reports investigated by the Los Angeles Police Department considered a report to be false if “deliberately fabricated by the victim” and “only if a thorough investigation led the police to conclude that the allegation was false and that no crime occurred.”
This definition is consistent with other studies producing similar single-digit prevalence figures, in which false reports involved recantations by the alleged victim or evidence that the alleged perpetrator could not possibly have been with the accuser at the time of the crime (i.e., an alibi). The infamous Duke Lacrosse Case, in which a black stripper suffering from mental health issues falsely accused three white college students of rape, is a well-known example (and one whose lessons the mainstream media seem stalwartly reluctant to learn). That particular case hit on the dynamics not only of sex but also race and class which may have accelerated the rush to judgment.
It’s important to note that the definition of “false report” used in both the research literature and in popular cases are, to be exact, malicious false reports. Further, the false report figures only apply to careful investigations — not to accusations made on social media. When it comes to sexual assault allegations that undergo proper legal investigations, the single-digit malicious false report figures are probably fairly accurate. And again, it’s also safe to say most rape reports are not maliciously false. But does this necessarily mean the other 90–95% of accusations are accurate?
Much of the debate over rape allegations tends to present the issue as an overly simplistic binary. Either the accuser is a malicious liar (which is rare, given the significant legal and social penalties) or the accused is guilty. To be reluctant to immediately endorse the latter invites accusations of misogyny and rape apologism. It is this latter fear, particularly since the #MeToo movement, that has likely led so many workplaces and organizations to throw individuals to the baying wolves and angry mobs the moment an accusation is made, without regard to due process.
Yet we know that no trivial number of good faith allegations prove to be wrong. The most dramatic examples involve men who have been convicted of rape based on victim testimony, only to be later proven innocent with DNA analysis. One recent study by the Urban Institute and funded by the US Department of Justice found that this figure may be as high as 11.6%. In most of these cases, the victim was not malicious, she simply misidentified her attacker. Intuitively, this may seem astounding, but it turns out that human memory is highly fallible, alterable, and unreliable. Police investigations themselves may create false memories in victims.
Just last September, Leonard Mack was exonerated by DNA evidence for a rape he was convicted of in 1975 and for which he served 47 years in prison. As with many such false convictions, good-faith but erroneous eyewitness testimony was key in the wrongful conviction. The Innocence Project, which helped to exonerate Mack, states that at least 375 wrongful convictions have been overturned in the United States using DNA evidence. Rape and sexual assault, along with murder, account for 97% of those wrongful convictions. When we take a more complete view of false sexual assault and rape reports that account for both malicious false reports as well as mistaken good-faith reports, the error rate may be closer to 20%.
Source: The Innocence Project.
Apart from long lapses of time, misidentification is more likely to happen in the context of rape committed by a stranger (the proverbial man leaping out of the bushes or breaking in with a mask and a knife) rather than rape between acquaintances or dating partners, particularly when an accusation is made fairly quickly. In cases where memory is not an issue, often murky circumstances are.
Consider a situation in which two individuals meet at a bar. Both consume alcohol beyond the legal threshold of intoxication. That night they have sex, both verbally consenting at the time. In the sober light of the morning after, however, one of the individuals regrets the encounter and believes they would not have consented had they not been drunk. Did one person rape the other, despite them both being equally intoxicated at the time? Does post-hoc regret make one person a victim and the other a predator? Does it matter if the regretful person is the man or the woman?
This played out in a real-life scenario at the University of Cincinnati in 2018. Two ROTC students returned from hard drinking, both intoxicated, and engaged in some sexual fumbling. Several days later a complaint was filed… but with a twist. In this case, it was the man who filed the complaint against the woman, who was subsequently suspended “indefinitely.” She sued the university for lack of due process but her case was eventually dismissed. This raises uncomfortable questions about how we decide who has victimized whom. Is it simply the case that, as Reason editor Robby Soave once put it, whoever wins the race to accuse the other becomes the designated victim?
These types of scenarios are often at the center of Title IX investigations — non-criminal investigations of sexual misconduct on college campuses — which present controversial civil liberties concerns. Some commentators such as the journalist Cathy Young argue that the biased and sloppy ways in which universities often carry out Title IX investigations risk trampling due process rights for both accuser and accused. These investigations involve university tribunals whose rigor and standards of evidence fall far short of a court of law, sometimes resulting in rushes to judgment, including expulsions, without a fair chance for the accused to defend themselves. These cases have led to an influx of largely successful lawsuits in which defendants sued their universities to clear their names and secured undisclosed settlements.
Opaque circumstances may also show up in sexual assault accusations made through legal channels or simply on social media. The point, of course, is not to treat such cases flippantly. Rather, the solution might be something more nuanced than deciding if one party is a predator and the other a victim. Exactly what proportion of assault accusations fall into this category is unclear, and probably varies by context (i.e., likely fewer in police reports, more in Title IX investigations, and with social media accusations).
One main argument is that false rape accusations are rare because there is considerable deterrence against them — indeed even good-faith accusations may be deterred. Victims of rape are often subjected to public scrutiny, skepticism, and moral judgment; may have to devote time to lengthy investigations and anxiety-provoking trials; and have very personal details revealed to the public. If making a legal allegation is so emotionally draining for good-faith accusations, why would anyone be tempted to make a bad-faith malicious report for which there are non-trivial legal consequences?
These are fair observations, and finding ways to make good faith allegations less taxing on victims is a worthwhile endeavor. However, it is critical to realize that any estimates regarding rates of false allegations are based specifically on a legal context where there is a significant deterrence. If we as a society were to eliminate that deterrence by taking #BelieveVictims to its logical endpoint, the prevalence figures for malicious false reports would change. More troubling is that pundits, commentators, and activists routinely apply the data derived from legal investigations to other contexts such as social media, where the landscape of incentives and deterrence are wildly different from legal contexts.
Unlike the process of going through official channels and courts of justice, the Internet is a free-for-all governed mostly by outrage algorithms and the court of public opinion. In the online culture war arena, fact-checking, evidence-gathering, and the ethos of “innocent until proven guilty” — the very mechanisms that help to keep the proportion of false reporting low in “real life” — have been essentially rendered taboo. More than that, alleged victims are treated as heroes and showered with sympathy, which may invert the reward/deterrence structure, incentivizing rather than deterring false claims.
The example of the retracted Rolling Stone article about an alleged (but false) rape at the University of Virginia highlights how mentally ill individuals, overzealous reporters, incompetent school administrators, faculty, and illiberal online social norms can combine to unwittingly incentivize false allegations. Again, this is unlikely in the legal context, where the penalties for false reporting, perjury, and obstruction of justice remain powerful deterrents. But online or on campus, where the only real deterrence — counter-claims of defamation — are often extraordinarily difficult to prove, and where garlands of acclaim await any accuser deemed to be on the right side of an intersectional “power imbalance”, the dynamics can be profoundly different. This is why using the false reporting prevalence data gathered from legal cases in a social media (or work/campus) context is not an apples-to-apples comparison.
Given everything discussed, how then are we to navigate the thorny issues surrounding sexual assault and rape allegations? Rather than the reductive slogan #BelieveWomen, perhaps we might adopt #TakeAccusersSeriously as our mantra and subject every rape and sexual assault allegation to an independent due process investigation. Due process is, after all, a pillar of democracy and is invoked twice in the US Constitution. Instead of immediately rushing to (and broadcasting) conclusions as news of allegations hit our timelines, as spectators we can reserve judgment until all the facts come in and the investigation is concluded.
In addition, we can become more informed about the differences between legal cases, social media accusations, Title IX investigations, and workplace harassment allegations. We can take particular circumstances into consideration instead of taking shortcuts via a one-size-fits-all heuristic. For example, the proportion of false abuse allegations that come out of divorce and child custody battles may be higher for obvious reasons of human psychology. At the same time, that doesn’t mean we should assume such accusations are necessarily false. Above all, general prevalence figures on false rape allegations tell us little about whether a specific rape allegation is true. The best approach is case-by-case.
Those are some things we can all do. Employers have a larger role to play and may need to recalibrate their response to sexual assault allegations, adopting policies that neither ignore alleged victims nor cave to uninformed public pressure to terminate accused employees without proper evidence. Universities, too, have an onus to ensure that Title XI investigations treat every accused as innocent until proven otherwise and that men accused of sexual assault don’t have campus housing, scholarships, or their ability to complete their degrees summarily revoked without due process.
Ultimately, dutifully attending to the possibility of mistaken reports and the due process rights of the accused is a critical safeguard to women’s rights. Maintaining the highest standards in the maintenance of civil liberties in no way leads us back to an era in which women’s claims are brushed aside. Instead, broad public confidence in an investigative system that is fair and rigorous only helps bring justice for victims of rape. We can “believe all women” and thereby erode the norms and institutions that have ushered in the safest period in history, or we can build on past progress by ensuring the system upholds the rights of everyone, including the accused.
Published June 28, 2024
