Aeon, of course I can speak only for myself and not for Roderick, but if he’s citing the statistic that appears in our essay (which seems likely), the figure comes from the CDC / NIJ National Violence Against Women Survey, whose results are reported on at length by Patricia Tjaden and Nancy Thoennes (2000) at http://www.ncjrs.org/pdffiles1/nij/183781.pdf. The full Methodology Report is apparently available only in print, but there is a discussion of the methodology on pp. 13-15 of the research brief at http://www.ncjrs.org/pdffiles/172837.pdf.
The report is horrible stuff; what’s more horrible is that the experiences tha I have learned about from my friends don’t even leave me with cognitive dissonance about its truth; the survey methodology is solid but I can also believe the numbers because they have been proven true in my life. Which is a horrible thing to realize.
That said, here’s a dry and abstract discussion of the figures that we know.
Section 5 of the report discusses “Intimate Partner” Violence (including a survey of the past literature), and reports that NVAWS found (p. 38, exhibit 9):
24.8% of women surveyed had suffered, at the hands of an intimate partner, rape, attempted rape, or physical assault, in their lifetime.
25.5% of women surveyed had suffered, at the hands of an intimate partner, rape, attempted rape, physical assault, or stalking, in their lifetime.
“Intimate partners” are current and former spouses, male or female cohabiting partners, boyfriends or girlfriends, and dates. (Figures elsewhere demonstrate that the overwhelming majority of the intimate partner violence against women reported in the survey is committed by male partners. Although findings of widespread violence against women in lesbian relationships would hardly be reassuring, either, from a feminist standpoint.)
“Rape” and “attempted rape” are defined quite conservatively—that is, the categories include significantly fewer activities than would be prosecutable under existing sexual assault laws. The screening questions can be found on p. 13 of the research brief. Part of the reasons for the narrowness of the screening questions is, I think, to avoid the controversy around Mary Koss’s findings (controversy which I happen to think is unwarranted and which I know in some cases — Katie Roiphe’s and Christina Hoff Sommers’ attacks, for example — to be based on mischaracterizations, but which is beside the point here).
“Physical assault” was screened using the Conflict Tactics Scale, with assaults ranging from slapping or hitting to using a knife or gun. (Abuse was usually repeated: women who reported being physically assaulted by a partner in the past 12 months had been assaulted, on average, about 3-4 times that year.)
As we mention in the footnotes to the essay, a full discussion of the validity of common statistics on violence against women was beyond the scope of our essay; our effort was to make some progress on the philosophical issue of how radical feminist class analysis and libertarianism can be reconciled, and to mostly bracket the empirical question of whether the evidence and arguments commonly given to support feminist class analysis are cogent. Of course, that’s an awfully important issue, but we ran long on talk as it was, and for myself I think it’s an issue that is already thoroughly discussed in the existing feminist literature.
That said, it is an important issue for the claims made in our paper, even if it’s one we bracketed at the time, and I’d be glad to discuss it further here as you like.