How often do women falsely cry rape? Because of the 18-year-old Hofstra student who recanted after telling police that five men had tricked her into a bathroom and then gang raped her two weeks ago, that question has been flying around the Internet. As Cathy Young notes in Newsday, the answers often fall into one of two camps. "Many feminists argue that the problem of false accusations is so minuscule that to discuss it extensively is a harmful distraction from the far more serious problem of rape. On the other side are men's-rights activists, claiming that false accusations are as much of a scourge as rape itself."

But isn't the rate of false rape charges an empirical question, with a specific answer that isn't vulnerable to ideological twisting? Yes and no. There has been a burst of research on this subject. Some of it is careful, but much of it is questionable. While most of the good studies converge at a rate of about 8 percent to 10 percent for false rape charges, the literature isn't quite definitive enough to stamp out the far higher estimates. And even if we go by the lower numbers, there's the question of interpretation. If one in 10 charges of rape is made up, is that a dangerously high rate or an acceptably low one? To put this in perspective, if we use the Bureau of Justice Statistics that show about 200,000 rapes in 2008, we could be looking at as many as 20,000 false accusations.

Legal scholars used to be routinely suspicious of rape victims. "Surely the simplest, and perhaps the most important, reason not to permit conviction for rape on the uncorroborated word of the prosecutrix is that the word is very often false," a Yale Law Journal article opined in 1952, echoing a view voiced since at least the 17^th century. These views remained mainstream into the 1970s, if not later. As Marcia Clark said yesterday  recalling the 1977 rape charges against Roman Polanski, "Those were the days when folks still believed rape was 'easy to charge and hard to disprove.' " And that old adage couldn't have been further from the truth. Prosecutors well knew that unless the victim was Snow White, the case was toast."

You can see what Susan Brownmiller was up against when she wrote her path-breaking feminist tract, Against Our Will: Men, Women and Rape, in 1975.

In her book, Brownmiller said that only 2 percent of rape allegations are false, citing findings by the female police in a New York City rape squad. The problem is that while this statistic has been widely repeated, with dutiful mentions of New York-based "research," no one has ever tracked down its source. This we learned from a comprehensive review of the literature on false rape charges published in the Cambridge Law Journal in 2006. The author, Philip Rumney, finds a couple of small studies that back up the 2 percent claim but isn't confident of their methodology.

Rumney's survey of the terrain is the best we found. He also takes aim at the findings on the other end of the spectrum—the research that purports to show that the rate of false allegations of rape is in the range of 40 percent, as well as the flawed (but often cited) work that makes a crazy high jump to as high as 90 percent. The 40 percent figure is usually attributed to a 1994 article by E.J. Kanin in the Archives of Sexual Behavior. Kanin looked at 109 reports of rape to police in one small Midwestern metropolitan area over nine years. His pool was small. The police he studied always offered the victim a polygraph—perhaps signaling they doubted her veracity. And Kanin himself "warns against generalising from his findings" and points to reasons for questioning them, as Rumney explains.

The hugely high 90 percent false rate is several degrees more suspect. The citation for it is usually a study in Scotland by police surgeon N.M. MacLean of only 34 rape complaints made from 1969-74. Complaints were labeled false if they were made after a delay. Or if the victim didn't look "disheveled" or upset or seriously injured. But those factors don't necessarily indicate that a rape charge is trumped up. When police use stereotypes about rape to sort real allegations from false ones, they can do victims a real disservice, as this model paper from the Oregon Attorney General's Sexual Assault Task Force explains. In a 1981 study of 16 reports that claimed the victim admitted to making it up in 14 of them, one case was disproved because the police decided the woman was too large for the alleged rapist to have taken off her "extremely tight undergarments" against her will. Need we say that this not the critical eye we want from the cops?

Rumney's smart debunkings leave us with a group of American, British, Canadian, and New Zealand studies that converge around a rate of 8 percent to 10 percent for false reports of rape. Not all of these studies are flawless, but together they're better than the rest of the lot. They include a massive 1997 report on sexual assault by the U.S. Department of Justice, which includes data from 16,000 local, county, and state law enforcement agencies. The DoJ found that "in 1995, 87% of recorded forcible rapes were completed crimes and the remainder were classified as attempts. Law enforcement agencies indicated that about 8% of forcible rapes reported to them were determined to be unfounded and were excluded from the count of crimes."

If 8 percent to 10 percent is about right for false reporting of rape, based on what we know so far, how should we think about that number? Rumney says he's not sure whether crying wolf is more or less likely over rape than over other crimes, because the comparative research is even less conclusive. So that's a question that appears to have no answer at the moment. (A 2001 Department of Justice report says that the rate of false reports is similar for other crimes, but it also gives the 2 percent figure without a source, so we're skeptical.)

What is clear, however, are two problems that are the flip side of the same coin. False charges of rape are an absolute nightmare for the men caught in their net. And the specter of made-up allegations is a real problem for law enforcement—which means they are also a problem for women who are telling the truth. Let's take the men first. We've heard from many of them in e-mails and comments since the Hofstra incident. Here is one story, equal parts heartbreaking and thoughtful:

My girlfriend was raped several years ago.  I was falsely accused of rape less than a year ago.  I contacted her (I had known her before her incident) because I was desperate for someone to talk to who would understand what I was going through.  To my great relief, it turned out that we understood each other very well.  From the initial stages of suicidal thoughts and not being able to function to the long-term fear, mistrust, and guilt that are facts of our lives, it turns out that her experience of being raped and mine of being falsely accused of rape were very similar. … One important difference, though, is that when she was violated, she received a great deal of help (medical, legal, psychological).  Apart from family and friends, I was on my own.  My legal and psychological problems had to be dealt with by me at a time when I couldn't eat, sleep, or think (except, of course, about killing myself).

On the law enforcement end, we heard from Steve Cullen, an Army attorney who's worked extensively as a prosecutor. He offered this cogent—and dire—explanation of the reverberations when women cry wolf about rape:

False reports have an incredibly corrosive impact on how sexual assault accusations are policed.  Police treat sexual assault accusers badly—much worse than the lawyers do—much worse than the courtroom does.  Forget what you see on "Law and Order SVU," the police end absolutely discourages victims from reporting.  Why is this so? Because cops suspect just about every victim is another false accuser, because either he/she has personally dealt with such a problem, or has heard stories from his or her cop buddies to this effect (and yes, in my experience female cops can be even worse offenders).  This police behavior is bad, and counterproductive—but it's real.  Putting a real stigma on false reports might combat this a bit—and make it a little easier for actual victims at the police station.

False reports also have a disproportionate impact on juries. How I'd hate to be prosecuting a sexual assault right now. Often in sexual assault prosecutions there's no debate as to the sex, but everything falls on proving lack of consent—and can only be proven through a convincing and persuasive victim's testimony.  Often, that victim's testimony has to overcome some less than ideal circumstances—she was drinking, people observed her flirting with the perpetrator etc. That's something she can own up to, and overcome on her own.  What she can't do on her own is extinguish jury members' memory of reading of some spectacular false accusation case in the newspaper last month. Every false accusation that makes it into the news makes it that much harder for the real victims to receive justice.

If police and juries are influenced by false reports, especially high-profile instances of false charges, like the Duke lacrosse case or the Hofstra case, why wouldn't those reports influence victims, too? Up to 60 percent of rapes go unreported. The Hofstra story will only make more women wonder if the police will believe them.

This is sobering. As, of course, is the whole topic. We're left to draw the following conclusion: False allegations of rape aren't rampant. But they don't have to be to cause terrible trouble. This is a problem that a men's rights movement shouldn't trump up. And also one that feminists can't dismiss.