If you see a rape allegation in the news, those words aren't far behind. They are talismans, touchstones for the idea that we must never, ever forget that women lie about rape. These women lied; therefore, women lie.
The truth is, of course, that some women do lie about having been raped. That shouldn't surprise us. People make false accusations about every type of crime, even murder, where it is excruciatingly difficult to do. If no woman ever lied about being raped, the gender might have some collective claim to sainthood.
The difference with rape is the reminder. Name someone who gave an acquaintance a gift then accused them of robbery. Find me a blog post about a robbery where one of these people is mentioned. Name someone who is used to demonstrate that insurance fraud occurs--every time a large insurance payout for theft makes the papers. Name one of those audacious people who tried to frame someone for a murder that never happened, even in fiction, then show me how their name comes up every time a body isn't found.
It doesn't happen. We're not told that people lie about these things. We're told that women lie about rape.
The implication in the "women lie" narrative is that we must be particularly on our guard against false accusations of rape, that any particular accusation is unlikely to be true. But is it?
The Rate of False Report
The standard figure passed around by victim advocates suggests a rate of false reports of 8% based on FBI crime statistics from 1997. This is comparable to rates for other crimes. However, citations can be found for rates as low as 1.5% and as high as 90%. In other words, huh? How do we deal with a range that big?
Luckily for those who want to sort out the truth of the matter, two papers came out in 2010 that shed considerable light by examining how false rape report rates are generated. David Lisak, Lori Gardinier, Sarah C. Nicksa, and Ashley M. Cote collected those prior studies that had the best (and most transparent) processes for sorting between false and merely unproven allegations. They also used a similar process for determining the rate of false reports of rape at a U.S. college.
Their results were interesting in two respects. The first is that all the credible studies produced rates close to the standard figure. Rates ranged from 2.1% to 10.9%, with the college study falling in the middle at 5.9%. The numbers on rape just don't support the idea that extra vigilance is required for this crime over others.
The second finding of the study is even more striking. In the authors' own words, "It is notable that in general the greater the scrutiny applied to police classifications, the lower the rate of false reporting detected." Those studies that relied on sorting done by the police produced the highest rates of rape. Those that examined the details of the cases labeled as false and required evidence of lies, rather than merely suspicion, produced the lowest rates. The 2.1% represents accusers who were charged with making false reports, the strictest criteria. (See this post for some thoughts on applying the presumption of innocence equally to accusers and accused in cases of rape.)
What Is A False Report?
The paper from Lisak and his coauthors discusses the criteria that must be met in order for a police report to be classified as false, noting that official statistics frequently include cases not meeting the criteria. Liz Kelly, in a separate paper released in 2010, examines two "attrition" studies, studies that track the ways in which rape cases fall out of the criminal justice system. Aside from convictions, how can rape cases end up classified?
- Declined to prosecute: Not enough evidence has been turned up to comfortably persuade a jury. Although the legal standard here is "reasonable doubt," prosecutors also face human prejudices when making these decisions, their own and those of potential jurors.
- Uncooperative victims: The victim has stopped cooperating with investigators or refuses to testify. This will include cases where the victim doesn't want to be responsible for the rapist going to prison, common in all types of domestic assault. It will also include cases in which dealing with the criminal justice system has become too traumatic. More discussion on that later.
- Victims deemed not credible: The police or prosecutor have decided that the victim is not to be believed. The victim may have personal characteristics that are considered untrustworthy (such as mental illness), or may have memory impairments (such as intoxication) making a clear picture of the rape difficult. This may also cover victims who withhold details of the rape or events leading to the rape out of embarrassment or fear of incrimination. I'll also discuss this in more detail in the next section.
- No crime occurred: This is different than false reports. This category includes incidents that may have occurred but did not rise to the level of a crime in the jurisdiction involved (for example, failure to stop sex when consent is withdrawn is not codified as a crime in legal statutes in the vast majority of the U.S.). It also involves complaints by third parties that did not hold up when the "victim" was interviewed and cases in which someone went to the police for help determining whether they had been sexually assaulted, usually after a period of unconsciousness, but no evidence of sex was found.
- False reports: The reporter has plausibly recanted, or substantial evidence exists to show that the accusation is unfounded.
Only the last of these is actually a false report. The rest of them either don't involve an accusation, or they exist in that murky land where we don't know what happened. So how do so many of them end up being included in the false report statistics?
Making the Numbers
It is notable that in general the greater the scrutiny applied to police classifications, the lower the rate of false reporting detected.
Both the Lisak and Kelly papers include multiple studies that compare actual police classification procedures to international standards. To put it briefly, they don't measure up. Depending on the location, any of those other classifications, aside from reports ending in convictions, might end up being included in official figures on false reports.
Some of this may be sloppy paperwork or coding, but part of the problem is the officers themselves. Kelly reports that even among those who are supposed to be experts in rape, the following attitudes can be found:
We have a lot of allegations that are then retracted, we have a lot of allegations that it comes out in the wash one way or another that it was consensual. He says it’s consensual and she doesn’t, or they’ve been together for like hours beforehand, she’s gone back to his flat. . . . But stranger rape, you immediately start to think, “Oh God, this could be a real proper sort of drag you in the car,” absolutely nothing beforehand has happened. I think subconsciously you would consider it more serious. . . . I think I’d have more belief in the victim, that that was saying it was by a stranger, that . . . it was a proper rape, rather than perhaps someone who said “It’s my ex-boyfriend, he came round,” because then you start to think things like, “Oh, she’s just getting back at him now.” (Female detective constable 2)
I have dealt with hundreds and hundreds of rapes in the last few years, and I can honestly probably count on both hands the ones that I believe are truly genuine. (Male detective constable 2)
In addition to finding that coding procedures weren't followed, the attrition studies Kelly reviewed also uncovered investigation techniques that violated international standards. The most egregious of these was offering lie-detector tests to victims, a practice widely viewed as hostile and accusatory toward victims. Using procedures such as these is one way to inflate the number of cases in which victims stop cooperating.
The prevalence of rape myths among the police forces coding reports as false should also be cause for concern when looking at their uncorrected numbers. When the women they consider untrustworthy match the profiles for those most at risk of rape (mentally ill, developmentally disabled, intoxicated, previously victimized--although the papers don't mention it, racial and sexual minority status fall here too), or those exhibiting rape trauma (scattered, faulty memory, embarrassed, ashamed), they are making decisions that push these cases out of the system on a prejudicial basis, not a factual one.
Then there is the fact that law enforcement is under continual pressure to reduce crime rates. That can lead to situations like that uncovered in Baltimore last year by the Baltimore Sun:
The problem in Baltimore is striking, but Baltimore is hardly the only city affected. If you need to change your crime rate, deciding that more rape accusations are "unfounded" is a simple administrative solution.
These examples are why we can't trust raw law enforcement numbers, which provide the citations for "women lie" arguments. If a police force doesn't know what is and isn't rape, how can it decide which rapes are falsely reported? If a police force decides that in he-said, she-said situations, "she" is arbitrarily not to be trusted, how can we trust their decisions on whether or not she lied? If a police forces continue to endorse rape myths, why would we trust their reporting numbers uncritically?
These are decisions that have consequences. They have consequences for the recovery of victims, as being disbelieved is a risk factor for poor recovery after rape (Ullman 1996). And they have consequences for the rest of us as well. The re-offence rate for rapists isn't entirely clear, but the low estimates put it around 20%. Kelly cites three cases of serial rapists in which early victims were recorded as having filed false reports.
Given the small number of false reports found in well-built studies, and the large number of repeat rapists, it might be time to give some serious thought to how well our societal strategy of disbelief serves us.
A Note on Real False Reports
Kelly's paper provides some interesting detail on a sample of reports that were accurately coded as false. Unlike the stereotype, most of the false reports did not involve direct accusations of a particular person. They were stranger-rape scenarios.
Also, in both the stranger-rape and acquaintance-rape scenarios, the false accuser was generally a victim of some sort. Some had been otherwise abused by those people they accused, including prior sexual abuse. Some were reporting rape to avoid abuse they would have otherwise received, as is suspected to the case for Tawana Brawley.
None of that excuses the false reporting of rape. It simply provides an opportunity to think about what might be done to reduce the rate of false reports even further.
Kelly L (2010). The (in)credible words of women: false allegations in European rape research. Violence against women, 16 (12) PMID: 21164212
Lisak D, Gardinier L, Nicksa SC, & Cote AM (2010). False allegations of sexual assualt: an analysis of ten years of reported cases. Violence against women, 16 (12), 1318-34 PMID: 21164210
Ullman, S. (1996). SOCIAL REACTIONS, COPING STRATEGIES, AND SELF-BLAME ATTRIBUTIONS IN ADJUSTMENT TO SEXUAL ASSAULT Psychology of Women Quarterly, 20 (4), 505-526 DOI: 10.1111/j.1471-6402.1996.tb00319.x