Intimate Partner Violence in the US

The Center for Disease Control issued a report on Dec. 14th. The figures are astonishingly high. I haven’t been able to access their site for some time, but here are some significant points that have shown up on Medscape:

The survey found that nearly 1 in 5 women (18.3%) and 1 in 71 men (1.4%) have been raped at some time in their lives. …

One in 6 women (16.2%) and 1 in 19 men (5.2%) experienced stalking at some point during their lifetime. Most female victims (66.2%) and about 40% of male victims were stalked by a current or former intimate partner.

Among victims of intimate partner violence, more than 1 in 3 women experienced multiple forms of rape, stalking, or physical violence; 91% of male victims experienced physical violence alone.

Severe physical violence at the hands of an intimate partner was experienced by roughly 1 in 4 women (24.3%) and 1 in 7 men (13.8%).

Given these figures, it can seem odd to see the report as generally saying that roughly 1 in 3 women, and 1 in 4 men, have experienced rape, stalking and/or physical violence by an significant other. However, there is a difference between the unqualified physical violence of the summary and severe physical violence, with the latter occurring for 1 in 7 men. Even that figure is horrible.


18 thoughts on “Intimate Partner Violence in the US

  1. Logoskaieros, thanks so much! I think that’s the link I’d tried, but Two pc’s failed. Now it’s on my ipad.

    On seeing it, I’m reminded of some questions being asked. One is whether the sample included cell phones. It looks at a quick glance as though it did. Another question raised is about their definition of rape and it’s relation to penetration.

    I’m rushing here, and haven’t looked yet at all carefully, but now we can, since we’ve got the report

  2. Does it seem as though some of the questions were phrased, perhaps unwittingly, in such a way as to increase greatly the risk of “false positives”? For example, one of the questions at the beginning of Appendix C appears to me to be likely to sweep up all manner of possibly non-assaultive sexual contact while the respondent was intoxicated.

    The CDC’s newly-published data are so far out of line from those generated by the much larger annual National Crime Victimization Survey (NCVS) that one has to wonder if the discrepancy is at least in part the result of methodological flaws on CDC’s part (which, if true, is of course not mutually exclusive of possible flaws in NCVS’ methodology).

    For example, the latest published NCVS results (see suggest that solely during the course of 2010, roughly 1 in 1,400 (0.7:1000) Americans age 12 and older (both female and male) were the victims of rape or sexual assault – sexual assault being defined to include
    attacks or attempted attacks generally involving unwanted sexual contact between the victim and offender that may or may not involve force. Now, according to NCVS historical data, this rate has been dropping pretty consistently for the last couple of decades. But even if one factors in higher rates in past years in order to come to a picture of “lifetime likelihood”, how does one begin to reconcile the NCVS’ findings with the CDC’s conclusion that nearly 1 in 5 women and 1 in 71 men in the United States have been raped (not just sexually assaulted, mind you, in the NCVS’ broad sense, but raped) at some point in their lifetime? Am I correct in deducing that *at least one* of these studies must be seriously defective?

  3. Nemo, I’m not sure what you are concerned about. Is your problem with the idea that if someone was so drunk that they were unable to assent and the other person still managed to have vaginal sex with them, then that could be non-assualtive sexual contact? (Just to be clear for people who haven’t looked at the report, vaginal sex involves a vagina, but it is allowed that the women may be the rapist.)

    It seems pretty clear that for the report, intercourse without consent is rape. Do you have a problem with that? Or is it something else?

  4. BTW, I think there are all sorts of ways in which physical agression needn’t involve what I/we think normally of as agressive contact(i.e., use of force and/or involving pain or injury).

    You probably know that many people have complained about the past statistics being way too low, and perhaps this is an important difference.

  5. Anne, let me be more specific as to the first point raised in my comment. According to Appendix C of the CDC report, respondents were asked:

    “When you were drunk, high, drugged, or passed out and unable to consent, how many people ever [for women, put his penis in your vagina/anus/mouth, or put their mouth on your vagina or anus]?”

    I would submit that this question, as worded, is likely to induce people to give a nonzero (i.e. positive) answer even if the incident in question relates to times when the sex did not constitute rape.

    The questioners probably intended for the phrase “and unable to consent” to modify everything preceding it and not just “passed out”, such that not every incident of vaginal/anal/oral sex while drunk, but only those while intoxicated to the point of being unable to consent to sex, were intended to be covered. However, I think this is unlikely to have been discerned by all respondents, as the question is not unambiguously clear. If a statistically significant portion of the respondents included *any* sex while they were to some degree drunk – even if not deprived of the judgment or ability to consent – then it is reasonable to expect that the resulting figure for rape incidence could skyrocket. As we know, the relevant consideration is not simply whether the victim was drunk, but whether the level of the victim’s intoxication was such as to deprive the victim of legal capacity to give consent (and there also sometimes other highly fact-dependent considerations that determine whether a crime has occurred). Do you agree?

  6. Nemo, so the idea is that a gddly-tiddly romp in the hay might lead someone to an affirmative answer on this question. (I apologize for the Masterpiece Theater construel of fun sex, but it is Sunday night and Downton Abbey is on.)

    That seems to me to be an empirical problem. In my experience, researchers with any knowledge of the problems survey questions can bring would have been well aware of the question you are asking. So I think your question is really about their basic competence. And I can’t vouch for that.

    At the same time, I do want to note that we can always interject such doubts. I’ve seen too many times when philosophers have basically been asking researchers in others fields, ‘but are you competent’.

  7. So, I take it that the [possible] problem is that respondents might not know whether or not they were ‘capable of consent’ according to either (a) the survey designers’ conception and/or (b) some legal definition?

    I think this is an important question in many ways. However, I suspect that – even if individual notions of ‘capcity for consent’ vary – most respondents are able to report whether they, as individuals in real situations, believe they were able to consent or to refuse consent.

    It is necessary that we have an agreed-upon legal conception of capacity and incapacity (if only within a single jurisdiction), but such ‘objective’ standards should not obscure individualized facts and conditions.

  8. Sissystars, I think Nemo’s concern was a bit different. His problem may have been that the wording of the question, with consent relegated to the end of a list of conditions, could prompt a positive response in those who were drunk but not VERY drunk. I.e., in those drunk but still were able to consent and in fact did consent.

    Still, I think we’re all interested in a similar issue, and it is important to think about what consent means. In addition to Nemo’s concern, we could add that in many jurisdictions it is possible to fail the sobriety test for driving and still be sober enough to be counted as consenting to drive. So the relation between the ability to consent and ‘being under the influence’ is murky.

  9. Anne, the “romp in the hay” makes it sound too benign, but I believe that there is a nontrivial likelihood that affirmative answers to that question would have picked up some number of noncriminal (e.g. drunken but legally consensual) sexual encounters – “false positives”, as I referred to them initially. Just how many (ranging from zero to some statistically significant number) is, I agree, an empirical question. But because the question appears on its face to create a plausible risk of this, it seems to me that in the absence of an empirical answer to that question, the reliabillity of the finding based on the responses to the question cannot be established.

    I’m not sure I would agree that my question can be reduced to one about the researchers’ basic competence, which I’m happy to grant. The list of competent experts who have at some point carried out research that at the time of its first publication turned out to contain an unaddressed vulnerability (methodological, inductive or otherwise) is very long and very distinguished. I confess to not being entirely sure, either, what you mean to say by noting that experienced researchers knowledgeable about the problems survey questions can cause would have been aware of the point I’ve raised. Do you mean to suggest that the problem is merely an apparent one, and that the surveyors were probably able to ascertain that the question, when posed to a broad cross-section of laypersons, would not trigger a statistically significant number of false positives? Or that some other part of the methodology (not obvious from the report) probably controlled for this possibility? Given that we know what the question was as a starting point (and that the finding seems to be a direct reflection of answers to that question), could even all the competence in the world make a really reliable dataset out of the responses to that question?

    What’s your take on the question – elevated risk of false positives, or no real risk?

  10. Nemo, I really disagree and do so in a way that seems to me to make your point much more empirically questionable. Here’s why: in the prior sections of questions, they twice say “Remember, we are only asking about things that you didn’t want to happen.” So those questioned will have been primed to notice the “without consent,” and even to assume it was there if it wasn’t explicitly put. That makes a huge difference, and I think invalidates your claim that the wording of that question in that context would prompt false positives. It might be that it would in another context, but that’s a different issue. And these were phone interviews, I think, so the order was set.

    The reason I think you are questioning their competence is that a HUGE stress in setting the questions for surveys is on whether the wording will bias the outcome. The one time I worked with professional level people on the actual wording of a survey, it was possible to spend a very long time on a question. Perhaps one hour, with the idea that one would come back to it when one had some of the others set. Of course, one mistake wouldn’t show incompetence, but the sort of problem you see in it might well lead one not to hire those people again.

  11. Anne, while I agree with you that the presence of those notices probably reduces somewhat the likelihood of false positives and is thus relevant, I disagree that it disposes of the problem. While in one strict sense you could say that anything to which you consented you ipso facto desire to happen; yet people can effectively consent to things about which they have substantial contemporaneous misgivings and/or about which they subsequently experience substantial feelings of regret, and in either case there is a decent chance that they will at times think of or describe those things as being in a sense things they didn’t want to happen. I have known myself to do this, and I have also seen it happen many times with other people. That’s partly why I remarked that the “roll in the hay” formulation sounded too benign; I was not so very much concerned about the kind of false-positive incidents likely to be placed by their recollectors in the category of “wanted it to happen”. Focusing on consent rather than desire would have been better, but for the reasons I mentioned previously I think the question was ambiguously worded as to the relation of drunkenness to capacity to consent. (And linguistically, I think it’s hard to deny that there is at least a formal ambiguity there.)

    I do understand your point about the emphasis placed in survey development and administration on the crafting of questions. But I still maintain that I’m unconcerned by the researchers’ competence here. The competence matter makes it sound as though the finding is based on some kind of underlying argument from authority that I’m disputing, but it’s based simply on the survey responses. It would be one thing if we did not have direct information about the wording of the questions, in which case a presumption of competence might lead us to assume that there were no issues of the type I’ve raised. In this case, however, we actually have the survey question before us, and there appears to me to be a potentially significant problem posed by it. You’ve suggested that the competence of the researchers is at least incidentally implicated by my observation, but what are you suggesting follows from that? Are you suggesting that if we grant the competence of the researchers, we ought to mistrust our lyin’ eyes and regard any perceived problems with the question as illusory – that the prima facie plausibility (which, I allow, is apparently not universally granted) of that survey question triggering false positives is somehow reduced?

  12. Nemo, if we worry about false positives due to memory, then I think there’s a problem with any of these surveys.

  13. True enough, Anne, but that’s not really what I was getting at, and I may have explained myself poorly. I was referring there to the fact that a respondent might well identify a drunken sexual encounter as, in whole or in part, something that s/he hadn’t wanted to happen, even though the same respondent might not (if the question were asked clearly) identify the same incident as having occurred under circumstances where s/he was so incapacitated as to be unable to give consent. Thus, I don’t think that the presence in the survey of reminders that “we are only asking about things that you didn’t want to happen” ultimately obviates the risk posed by the fact that the question is phrased in such a way as to weaken the link between the incapacity qualifier and the “when you were drunk” concept.

    I’m still not clear on the implications of the competence thing.

  14. “Rape” is a legal term. The US has 50 state jurisdictions, plus federal, plus military jurisdictions. All of these have varying definitions and degrees of rape.

    Adding a 53rd definition by the CDC or anyone else only confuses matters more. If you use a different definition of rape than the law, then of course you will come up with “unreported” or “unprosecuted” numbers because some of the included cases legally constitute no rape.

    Also, I find it highly dubious to use the number of rape cases reported by “victims” without asking the alleged perpetrator in question if they could maybe explain or try to defend their act. It might well be that the story looks completely different once you listen to the other side. That’s why we have trials with a right to a defence and cross examination.
    A social worker or the CDC can NOT decide what a rape was and what wasn’t. Only courts can.

  15. The “1 in 71 men have been raped″ stat from the CDC survey doesn’t tell the whole story. It defines “rape” as the attacker penetrating the victim, which excludes women who use their vagina to rape a man (rape by envelopment) which is counted as “made to penetrate”. The very same survey says “1 in 21 men (4.8%) reported that they were made to penetrate someone else.” Therefore, if you properly include men who are forced to engage in PIV sex against their will, from 1 in 16 to 1 in 21 men have been raped, using the numbers in the CDC study you cited. Interestingly, the study says that 79.2% of male victims of “made to penetrate” reported only female perpetrators, meaning they were raped by a woman.

Comments are closed.