May 2015 Update | Christina Hoff Sommers has posted a rebuttal to this post on her Facebook page, and I’ve written a reply to that rebuttal. In the course of composing that reply I discovered that one claim I made in the post below was in error — I apologized for that error in the reply linked above, and I have corrected it in the blogpost here.

If you were around for the so-called Culture Wars of the mid-1990s, you probably remember Christina Hoff Sommers — her 1994 book Who Stole Feminism? was a centerpiece of right-wing attacks on mainstream feminist theory and organizing at the time. Recently Sommers has re-emerged as the “mom” — that’s literally what they call her — of #GamerGate, that weird movement of video game fans obsessed with “ethics in gaming journalism” and what they see as feminist attacks on their hobby.

I haven’t paid more than desultory attention to Sommers since the nineties, so when I somehow wound up at her Twitter feed on Saturday I was surprised to see her supportively retweeting this:

The assertions in this tweet — that Rolling Stone “invented” its recent story on an alleged gang rape and that this supposed invention single-handedly discredits broader “rape culture” arguments — struck me as even more ridiculous than I’d remembered Sommers being back in the day, so I fired off a quick tweet expressing my surprise.

Too quick, as it turned out, because when I went back to Sommers’ timeline, I saw that it was stuffed with even weirder stuff, much of it in Sommers’ own words. Mildly embarrassed by my ignorance of her current mindset, I deleted the tweet, but as I did so I noticed that several people had already responded to it, so I figured I should explain:

This second tweet wasn’t directed to her, as you can see — I didn’t include her screen name in it, didn’t @ her on it. It was a heads-up to my own Twitter folks about why the previous tweet had disappeared. But she found it anyway, and RTed it, along with a followup declaring that “Much of the data on sexual violence is flawed. Victims need good research & smart policies—not hype.”

Sommers only has about 32,000 followers, but those two tweets unleashed a flood of responses — all, sadly, while I was on my way to my kid’s birthday party. A few of the tweets were over-the-top repulsive. Most, though, just took issue — often abusively — with my charge that Sommers is a “rape denialist.” It’s those that I wished I’d had time and space to reply to as they came in, and those that I’d like to respond to today. Because I do consider Sommers a rape denialist, and I think it’s important to say exactly what I mean.

So. Why do I say it, and what do I mean?

I mean this: Christina Hoff Sommers, in her many recent public statements about rape and sexual assault in America, understates the prevalence of rape in this country in ways that are unsupported by the evidence. She analogizes America’s rape crisis to entirely invented “crises” of the past while wildly overstating the evidence for the existence of an epidemic of false rape claims. To read her writing, watch her videos, and follow her on Twitter is to be given a wholly unrealistic impression of the scale and seriousness of rape in America. And that’s the case — and this is crucial — even if you agree with her contention that rape reporting data is seriously flawed.

Let me say that again, because this was a core claim that her supporters made on Saturday afternoon: I am not calling Christina Hoff Sommers a rape denialist because we rely on different statistical estimates of the prevalence of rape. I am calling her a rape denialist because the way she deploys even her own preferred statistics is fundamentally bogus.

Enough introduction. Let’s get down to cases.

On Saturday Sommers tweeted me a link to a video in which she critiqued a 2011 CDC study that concluded that about 1.3 million women were raped in 2010, saying that “the agency’s figures are wildly at odds with the official crime statistics, the Justice Department’s annual crime survey.” She’s referring to the National Crime Victimization Survey (NCVS), which, she says, “reports that there were about 188,000 rapes and sexual assaults in 2010.”

What’s behind this discrepancy, to Sommers’ mind? Beyond unspecific methodological concerns, Sommers offers a number of concrete criticisms of the study’s questioning and statistical analysis.

None of these criticisms hold up.

To begin with, Sommers claims that “no-one interviewed” by the CDC “was asked if they had been raped or sexually assaulted,” but this claim is ridiculously misleading. The study’s core rape estimate was generated by respondents’ answers when asked whether anyone had “used physical force or threats of physical harm to make you have” anal, oral, or vaginal sex — a clear, straightforward, unambiguous description of rape. Some 620,000 of the women that the CDC reported as having been raped in the previous year — almost exactly half of the total — were the result of affirmative answers to that question.

And what about the rest? They answered in the affirmative to questions about either attempted rape or rape facilitated by drugs or alcohol. There too, Sommers dramatically misstates the statistical evidence.

As an example, Sommers makes the following claim: “Sixty-one point five percent of the women the CDC projected as rape victims in 2010 experienced what the CDC called ‘alcohol and drug facilitated penetration.'” Here she leaves the clear impression that more than three-fifths of the incidents of rape reported fell into this category, but that’s not the case. Although 61% of the women the CDC says were raped did report alcohol and drug facilitated penetration, 49% reported forced penetration, and another 41% reported attempted forced penetration. Many, in other words, reported multiple types of assault, and alcohol and drug facilitated penetration accounts for 41%, not 61%, of the reports.

But that statistical sleight-of-hand is only a small part of Sommers’ misrepresentation in this area. She suggests that the CDC counts consensual “sex while inebriated” as rape — indefensible, if true — but she does so by selectively and tendentiously quoting from the questionnaire. In fact, that section of the questionnaire — read to all respondents, but never mentioned by Sommers — states specifically that the questions within it concern sexual contact that “happens when a person is unable to consent to it or stop it from happening because they were drunk, high, drugged, or passed out from alcohol, drugs, or medications.”

Sommers knows this, but she deliberately excludes it from her writing and speaking on the topic in order to facilitate her misrepresentation of the CDC report.

Want more? Here’s more. Sommers claims that the NCVS found that there were “about 188,000 rapes and sexual assaults in 2010.” But the CDC figure of 1.3 million, as we have seen, includes both completed and attempted rape, which Sommers’ 188,000 does not. Adding together the NCVS stats for completed and attempted rape, to make it an apples-to-apples comparison with the CDC, gives us a total of a quarter million victims.

May 2015 Note | I have discovered that the above claim was erroneous. Details here.

So yes, the CDC found more sexual assaults than the NCVS, and yes, they found it by asking broader questions. (The NCVS asks about “rape, attempted rape, or other type of sexual attack.”) But the CDC’s questions are far more robust than she claims and the one category of sexual assault that she singles out for mockery is both more reasonable and a smaller portion of the whole than she would have us believe.

Sommers urges us to reject the CDC data as preposterous, in other words, but the arguments she puts forward against it are factually weak and intellectually dishonest.

And if her data-based arguments are spurious, what she does with them is even worse. In the video linked above she describes the “women’s crisis” portrayed by research such as the CDC’s as “manufactured,” and as “madness.” On Saturday morning she tweeted that the United States is currently in the midst of a “rape panic” analogous to the wholly invented Satanic ritual abuse scare of the 1980s, driven by a “false accusation culture on campus.” And this wasn’t just a poorly phrased tweet — in it, she linked to a Time magazine column from earlier this year in which she went on at length about the Satanic abuse panic and its “striking similarities” to the rape “panic” of today. And after she tweeted that link she went even further, approvingly retweeting two readers who likened our current dialogue around rape to the Salem witch trials.

There were no witches in Salem. There were no Satanic ritual abusers running preschools in the 1980s. But even the NCVS, which Sommers cites as the “gold standard” for such statistics, concludes that nearly a quarter of a million women experience rape or attempted rape each year. Our country’s rape crisis is real, not imaginary, and it is the millions of American women who are raped, not the comparatively tiny number of men who are falsely accused, who bear the overwhelming majority of its burden.

To claim otherwise can only be described as denial.

Note | As an anti-spam measure, the first comment on this blog from any commenter is automatically sent to moderation. (Subsequent comments are typically approved automatically.) If you comment on this post, and you haven’t commented before, be patient — your comment will go up as soon as I see it.

Update | A little more about my use of the word “denialism.” Some have claimed that since Sommers doesn’t deny that rape exists at all, she can’t fairly be called a rape denialist. I address this in the post itself, but to underscore and clarify:

If one misrepresents the true scope or scale of a real problem one can be fairly described as a “denialist” in that arena. If you admit that climate change is real, but offer cooked, false data to misrepresent its extent, you’re a climate change denialist. If you insist in the face of the evidence — as some widely-quoted pseudo-scholars do — that the Nazis “only” killed a few hundred thousand Jews, you’re a holocaust denialist. And if you continually, willfully misrepresent the statistics and scholarship surrounding rape and sexual assault in order to foster a false narrative that minimizes the extent of that crisis, then yes, you’re a rape denialist.

May 13, 2015 Update | Sommers posted a reply to this five-month-old blogpost on Facebook this morning. My response can be found here.