National Crime Victimization Survey: A new report finds that the Justice Department has been undercounting instances of rape and sexual assault.

We’ve Been Measuring Rape All Wrong

We’ve Been Measuring Rape All Wrong

What women really think.
Nov. 19 2013 11:19 AM

We’ve Been Measuring Rape All Wrong

The Justice Department survey that tracks instances of sexual assault has been vastly undercounting.

Anonymous Woman
The most often cited U.S. data about rape doesn’t ask a single question about whether a victim was unable to consent because of drugs or alcohol.

Photo by Shutterstock

How do you measure rape and sexual assault? It’s a tricky and loaded question, and the answer impacts a couple of highly charged debates. As in: If you believe the measurements that say sexual violence against women is significantly on the wane—as one prominent national survey shows—then you might argue against spending a lot of money fighting it. Or you might argue, as Slate’s Amanda Hess does, that binge drinking among women doesn’t really explain the problem of sexual assault, since the drinking has increased even as the rape numbers have fallen.

Emily Bazelon Emily Bazelon

Emily Bazelon is a staff writer at the New York Times Magazine and the author of Sticks and Stones

On the other hand, if you’re worried that the same measurement tool—the Justice Department’s National Crime Victimization Survey—is vastly undercounting sexual violence against women, especially when it comes at the hands of men they know and in the company of drinking or drugs, then you might agree with Emily Yoffe that it’s time to stop letting “a misplaced fear of blaming the victim” prevent college educators—and the rest of us—from warning “inexperienced young women that when they get wasted, they are putting themselves in potential peril.”

How helpful, then, that the Justice Department asked the National Research Council (part of the National Academies, which also includes the National Academy of Sciences and the Institute of Medicine) to study how successfully the federal government measures rape. The answer has just arrived, in a report out Tuesday with the headline from the press release: “The National Crime Victimization Survey Is Likely Undercounting Rape and Sexual Assault.” We’re not talking about small fractions—we’re talking about the kind of potentially massive underestimate that the military and the Justice Department have warned about for years—and that could be throwing a wrench into the effort to do the most effective type of rape prevention.


The NCVS statistics show the rate of completed and attempted rape in the United States declining from a high of 5 percent of girls and women victimized annually in 1995 to a low of about 2 percent from 2005 to the present. Sounds good, right—men behaving better, women protecting themselves more. But here are the flaws that call the nice-sounding stats into doubt: The NCVS is designed to measure all kinds of crime victimization. The questions it poses about sexual violence are embedded among questions that ask about lots of other types of crime. For example:

(Other than any incidents already mentioned,) has anyone attacked or threatened you in any of these ways: a) with any weapon, for instance, a gun or knife, b) with anything like a baseball bat, frying pan, scissors, or stick, c) by something thrown, such a rock or bottle, d) include any grabbing, punching, or choking, e) any rape, attempted rape or other type of sexual attack, f) any face to face threats, OR g) any attack or threat or use of force by anyone at all?

That’s not a good way to prompt women (or men) to report nonconsensual sex, broadly speaking, especially if they haven’t previously gone to the police—as most rape victims don’t. As the new report puts it: “This context may inhibit reporting of incidents that the respondent does not think of as criminal, did not report to the police, or does not want to report to police.”

The NCVS also doesn’t include scenarios in which a victim is unable to consent to sex because she or he is “drunk, high, drugged, or passed out.” And the NCVS doesn’t do enough to provide survey-takers with privacy. They can’t quietly check off a box on a self-administered questionnaire—they have to answer questions out loud over the phone. These features of the survey have also been shown to inhibit victims from responding.


Here’s how to fix this, the National Research Council panel says: Conduct a survey of rape and sexual assault separately from other kinds of crime. The best way to get an accurate count is to frame the questions in a “neutral context, such as a health survey.” Instead of asking, Have you been raped? the survey tool should ask questions about specific behavior, for example: When this incident happened, were you passed out from drinking or taking drugs? This gives room for survey-takers who might not call what happened to them “rape” to provide a more accurate measure of how many people are actually victims of nonconsensual sex. And they should be able to enter their answers on their own, on a computer, rather than over the phone.

There is, in fact, an existing survey that has many of the attributes the NCVS currently lacks. It’s administered by the Centers for Disease Control and Prevention, and it’s called the National Intimate Partner and Sexual Violence Survey. (NISVS is the acronym. Apologies for the alphabet soup.) NISVS “represents the public health perspective,” as Tuesday’s report puts it, and it asks questions about specific behavior, including whether the survey-taker was unable to consent to sex because he or she had been drinking or taking drugs. NISVS was first conducted in 2010, so it doesn’t go back in time the way the NCVS numbers do. But here’s the startling direct comparison between the two measures: NISVS counted 1.27 million total sexual acts of forced penetration for women over the past year (including completed, attempted, and alcohol or drug facilitated). NCVS counted only 188,380 for rape and sexual assault. And the FBI, which collects its data from local law enforcement, and so only counts rapes and attempted rapes that have been reported as crimes, totaled only 85,593 for 2010.

It’s a real cause for alarm that there is such a huge discrepancy between the national survey that most closely follows the approach recommended by the experts and the ones that don’t, yet are more often cited. The bottom line is that women are still experiencing date rape or acquaintance rape or gray rape—whatever you want to call it—in dismayingly large numbers. As Christopher Krebs, a sexual violence researcher at RTI International, puts it, “We all know that rape and sexual assault are the most underreported crimes in the world, and it’s very hard to say that the problem is declining. The NCVS data could be missing a lot.” And especially critical: The NCVS doesn’t directly capture the instances in which drugs or alcohol leave women less able to defend themselves. Let me say that again: The national data about rape that gets cited over and over again doesn’t ask a single question about whether a victim was unable to consent because of drugs or alcohol, even though that is a major risk factor. The NCVS fails to see the full range of nonconsensual sex that should concern us. It also doesn’t accurately reflect the circumstances in which this kind of rape occurs—another important function that a tool like this should serve.

In the vast majority of sexual assaults, Krebs says, the victim knows the offender. And sometimes she may not remember exactly what happened—because her memory is blurred by intoxication. That’s why Emily Yoffe called for rape prevention education that reaches women as well as men. Not instead of men—of course we need to make clear that men who force sex are fully responsible for their violence, no matter what the circumstances. But we also should treat women as fully capable of agency by giving them the information they need to understand that binge drinking is a risk factor for sexual assault. I’m the mother of sons, not daughters. It is absolutely my responsibility to teach my boys that there is no excuse—none—for having sex unless they can be absolutely sure the other person wants to. But if I had girls, I would want to open their eyes to the reality that drinking to the point of passing out will make them more vulnerable. That doesn’t mean blaming them. It means arming them. The number in this piece that’s probably the most accurate count—1.27 million women sexually assaulted per year—underscores the urgency.