Last night I posted this graphic on Facebook. (A lot of other people did, too, I think.) It purports to tell us something about the ratio of rape to conviction for rape to false accusations. It’s a snapshot, and its primary, overwhelming purpose is to give a sense of the proportions between those three numbers.
Eric needled me something fierce about posting this graphic, because:
(a) it is really, really vague about definitions, and
(b) I’ve been pretty hard on the media about sloppy definitions when it comes to guns, so why give them a pass about rape?
I think the first point is totally valid, and you can read Eric’s objections in detail here and the objections of Amanda Marcotte at Slate here.
I also think the second point is valid as a criticism of my method. What’s sauce for the gun is sauce for the gander, and all that. On the other hand, I also think that sometimes a strong, simple graphic that gives a gut sense of things about an important topic, even if it conveys little else, can be useful. I wanted to say something about how much real rape there is, how infrequently it is punished, and how rarely men are falsely accused. In part this is because two populations I’m emotionally and intellectually invested in — Native Americans and military servicemembers — are at substantially increased risk of being raped, and those rapes are particularly likely to go uninvestigated, unprosecuted, and unpunished. So I wanted to say something simple and bold and attention-getting, and I wasn’t particularly interested in definitional precision. I wanted to confront people with general scale, not specific statistics.
That said, the graphic is sloppy. Where, for instance, does the total number of rapes come from, here? The Enliven Project, which created the graphic, explains here that estimates of the ratio of reported rapes to total rapes ranges from 5% to 50%, depending on the source. “We assumed 10%, which is dramatic, but possible.”
“Dramatic, but possible” is, of course, the kind of phrase that could be equally applied to statistics such as “One in every hundred Americans has had contact with aliens!” and “Sixty-eight percent of women reported finding Axe Body Spray irresistible!” It tells us frustratingly little about why 10% was chosen. Is it the most reasonable estimate? Or just the most “dramatic”? (As a commenter on Enliven’s own website points out, the Bureau of Justice Statistics, cited by Enliven as a primary source, estimates a 35% report rate — much higher than Enliven’s estimate.)
One can also ask whether we are all working with the same definition of rape, here. My personal feeling is that we should define the crime quite broadly. Forcible rape by a stranger, rape of an adult by an adult sibling, statutory rape, prison rape, and non-violent acquaintance rape are not all equally horrific, and that’s an important point when it comes to, e.g., sentence mitigation and enhancement. But they all stem from the same moral violation — namely, the exploitation of an imbalance of power to rob someone of autonomy over their own body, for one’s personal gratification. I also don’t have a lot of anxiety about the supposed gradations of consent — as a guy in my Crim Law class said, “If you don’t know — just ask!” Words to live by. People who don’t ask, and who proceed over resistance or objections (or even, really, lukewarm reception) are, I think, not that interested in the answer. So I’m pretty comfortable defining rape broadly and inclusively for the purpose of getting a sense of the scale of our sexual assault problem. I hope that Enliven also defines rape quite broadly. Unfortunately, I don’t know, because they don’t give us a precise definition of rape either with the graphic or on the above-linked explanatory page.
Finally, the categories given in the graphic — “rapists,” “reported,” “faced trial,” “jailed,” and “falsely accused” are also not clearly defined, and in some cases the definition can’t even be guessed at. What on earth does “faced trial” mean? Indicted? Pled down? Went to trial and later pled down? Went all the way through a trial? We don’t know. And “Jailed” is not necessarily the only satisfactory outcome to a rape trial or investigation. Some percentage of rapists may not go to prison but may be put into mental institutions, deported, or otherwise dispensed with. Those numbers aren’t reflected here. And as Marcotte points out, “rapist” is not necessarily a useful category either, since many rapists are repeat offenders. Does each “rapist” in the graphic represent a rape, or an individual rapist? If the latter, the graphic like underrepresents the total number of rapes, which makes the problem look less severe than it is.
All that being said, I still don’t really object to the graphic. For one thing, I hope it can be a starting point in an information campaign about rape. I hope it will inspire better, more precise use of data. Perhaps in the next few days somebody will knock up a new infographic that remedies these problems, and then we’ll all have something satisfying to chew on, talk about, and pass along on Facebook.
But more broadly, sometimes I don’t care if a source is technically accurate or precise on every point, if it conveys something which is generally true. Marx was wrong about history and, to some extent, economics. But he was right that there’s something inhumane and fucked-up about unbridled capitalism. Malcolm X was sometimes wrong about the law and history and, most notably, he mythologized his own life. But he was right about American racism. There’s always a danger, of course, that such sources open themselves, and thus their important ideas, to ridicule and dismissal. I think that is a serious problem. And this little infographic is in no way Das Kapital or The Autobiography, so maybe the virtue of its message doesn’t outweigh its flaws. Still. I think there’s something to this as a raw sensory impression about a serious problem that we often deny or minimize. Sometimes, to slightly paraphrase Eric, I do find value in things that are shocking solely for the sake of being shocking.
The source is NOT “generally true”. You miss the point entirely, and then go on to justify the shoddy statistics because the cause is good; or as you put it, “I don’t care if a source is technically accurate or precise on every point,if it conveys something which is generally true”. It is NOT generally true, and borders on outright fabrication.
Here is what I attempted to post on Enliven’s site, and which, as of yet, has not passed moderation:
Preface: Rape is a heinous crime that, in my opinion, should merit the strictest penalties that the law allows.
That said, this graphic epitomizes Mark Twain’s quote of “Lies, damned lies, and statistics”. If you use distortions and misrepresentations to support your cause, no matter how noble it may be, then you poison your own cause.
It is quite apparent that *no one*, to date, has actually looked at the “source” for this graphic. The ONLY listed source is the “National Crime Victimization Survey: 2006-2010″. The other listing of”FBI reports” is meaningless as there is no way to verify the data used, without a specific citation.
So, lets examine the only listed source. For brevity’s sake, lets only look at the “not reported to police” portion of this graphic, since that portion is the most dramatic, and compare this to what is shown in this graphic:
First, let’s just take the source of the data at face value. There is a lot to discuss about this source, but let’s just assume that the source used means “all rapes”. The graphic states that 90% of rapes are not reported. Yet, their *only* cited source says that the number of unreported “rapes” is 65%.
That means that at the very basest of levels, this graph is inaccurate to the only cited source to the tune of 350%. Yes, this graph is inaccurate by 350%. But it is much worse than that.
However, let’s first look at how they came up with a 90% figure. Well, they just assumed it. Yep, they just made it up. They say so themselves. I quote: “We assumed 10%” .
So, a factor of 3.5 times wrong is not so bad, is it? Well, looking closer at the data’s source, “rape” includes actual rape, sexual assault, attempted rape, and the verbal threat of rape. So, this graphic is saying that the mere verbal threat of rape is as good as rape itself.
If we look at actual physical rape, then the data source states that only 35% of those cases are actually unreported. That means that 65% of actual, real, physical rapes (not verbal threats) are actually reported, meaning this graphic has now distorted the data by 650%.
There is much more to be said about how bad this graphic is, but this will do for now. The comparison of “false accusation” (which is in their words also “assumed”) is an apples-to-oranges comparison, meant only to minimize the real pain felt by those falsely accused of rape. This graphic is a disgrace to those falsely accused, and to real rape victims. The use of lies to achieve your ends will only result in negative consequences for the real victims.
So I reread your post, and thought more about the consequences of it. You state: “All that being said, I still don’t really object to the graphic. For one thing, I hope it can be a starting point in an information campaign about rape. I hope it will inspire better, more precise use of data”.
Have you really thought about the consequences of your non-objections? If you are willing to give a pass to a graphic which, at best, distorts the truth from somewhere between 350% o 650%, then why should you be believed wen you cry “rape”?
First, I already made the point about the 10 vs. 35%. Look at the post again. On the other hand, it’s not clear that the NCVS number is the correct one, so we should be careful, perhaps, about saying that the graphic is “inaccurate by 350%.” But I agree — and said in the post — that Enliven’s estimate is arbitrary.
Second, you’re right that BJS, which compiles the statistics cited, defines “sexual assault” to include threats:
I think this is probably because of the wide variation of state definitions. (Sexual assault is generally defined by state law.) Unless you have a reliable source saying otherwise, I find it extremely unlikely that such threats constitute a significant percentage of sexual assault reports, and therefore would not constitute a meaningful percentage of the projected unreported assaults, either.
Third, and forgive me if I’m wrong, but I’m getting a whiff of “men’s rights” silliness from you. (I’m looking specifically at references to “the real pain felt by those falsely accused of rape” — a “problem” which, as Marcotte points out, is actually less prevalent than the Enliven graphic purports, because the Enliven graphic has conflated false reports of a rape happening at all with false accusation against a specific man.) Maybe that’s not right, but if that’s your game, really, no need to reply.
Finally, I’ll rise to your ridiculous baiting and answer: I suppose if I’m ever raped I’ll just have to hope the police investigating the rape haven’t been spending their downtime at the precinct house reading my blog, which obviously paints me as a hand-over-fist liar and someone who would totally make up a rape charge for fun and giggles.
Also, I’m not sure where you’re getting the 65% report figure for rape alone. Table 2, page 5 of the report linked above shows that “rape without other injury” goes unreported at a rate of 59%, but that’s the only place I see rape alone in the NCVS source.
Thank you for your reply.
You may be glib in your response regarding believability of rape based on your comments here, but if you think a good defense attorney would not be able to dig up your support for poor data in the glaring light of facts to the contrary, then you are mistaken. My point was aimed more in the general sense of how society perceives rape and rape victims, but in the event you were ever a victim (which I truly hope and pray would never be the case!), then your comments on a public blog can be used against you. Ridiculous? At you personally, probably. As a public perception – not-at-all.
Second, you are referring to a BJS survey. It incorporates state crime statistics, but the threat of rape is part of the survey function. At least where I live, the threat of rape is not a crime – and thus would not be recorded in reported crime data. However, that is immaterial as the graphics data is based on surveys, not crime data itself. And based on survey questions, which is where the “unreported rape” statistics come anyhow, and that my comment was primarily about, they formulate the threat of rape. So, it is not as insignificant as you make out. In fact, it is quite significant, as the source, which I will cite below, shows that “attempted rape” and “sexual assault” including “verbal threats” is 67% of the reported “rapes”. Actual physical rape accounts for 33% of the the data the survey.
You ask for the source: It uses the most recent year of the BJS data, which is 2008, which is part of the source cited by the original graphic. I used: Table 91. Personal and property crimes, 2008:
Percent distribution of victimizations, by type of crime
and whether or not reported to the police You may find the data here: http://bjs.ojp.usdoj.gov/content/pub/pdf/cvus08.pdf
Oh, I forgot to mention about “men’s rights”. Am I a proponent of “men’s rights”? Absolutely.Unabashedly and wholly yes.
I am equally (or at least I hope) a proponent of women’s rights.
Yes, very egalitarian of me. Yep, I get it wrong quite frequently. But I see injustices to both genders – merely because of their genders, and I spout off because of them.
Again, Richard, there is no number in that report showing that a significant percentage of the sexual assault figure includes threats. The NCVS survey form, given here, does not define “sexual assault,” and given that the definition including threats is far from universal, it’s not at all clear to me that anyone reporting on this survey would include threats. There are separate questions about general threats of violence, so it’s possible that BJS is aggregating any threats with a sexual component into the “sexual assault” column. But we have no information on that (I find nothing in their methodology to suggest it), and we certainly have no information showing that threats constitute a significant percentage of either the reported or the unreported sexual assaults.
Also, your figure on rape alone from 2008 appears to be anomalous. The figure I cited (59% unreported for rape without other injury) is from 2006-2010 — a broader range of years.
Moreover, as I’ve already said, we have no way of knowing whether the BJS estimates (based, as you point out, on surveys) are accurate. They’re one source.
But all of that is beside the point. It’s clear, and we agree, that for the number of rapes/rapists Enliven didn’t rely on the BJS numbers anyway. They looked at several sources (some given in the links above, others, as you point out, not named directly), and came up with a guess. A guess which I explicitly said was given without any hint of their methodology or reasoning. WE AGREE ON THIS POINT.
For some reason you seem really stuck on this idea that I’m a champion of bad data, when in fact the vast majority of my original post was devoted to pointing out flaws in the graphic. At the end, I did note that the graphic points to a more general truth, despite its methodological flaws. First, we don’t know that Enliven is wrong about its 10% guess. The simple truth is that nobody, not even the BJS, knows how what percentage of rapes go unreported. BUT even by the numbers I’ve given above from the BJS report (59% unreported for rape alone), the Enliven numbers are only off by about a factor of six. (I.e., if Enliven was counting 90 unreported rapes for every ten reported, those same ten would now be accompanied by about 15 unreported rapes, or about 1/6th of Enliven’s estimate.) A big mistake, and one that I’VE BEEN AT PAINS TO POINT OUT in the original post and here below. But it’s not off by even an order of magnitude. That’s what I mean by “scale” — it captures a sense, an intuitive sense, of the real problem.
When I say I don’t object to the graphic, that doesn’t mean I sign off on its validity, either. It means that it’s useful for starting a discussion — including the one I had on Facebook with Eric and the one I’m having with you now. In fact, its incorrectness is valuable to me. I posted it, people criticized it, I had an opportunity to examine it critically. Having done so, I find it certainly flawed in its presentation and perhaps (we don’t know) flawed in its data. But it is not wrong about the general truths that: (a) some very large number of rapes go unreported or unprosecuted, and (b) false accusations of a specific person are very, very rare.
That was the point I have been making from the beginning, which you seem hell-bent on shaping into some narrative in which I’m (a) promoting falsehoods to stoke an agenda, and (b) apparently going to “cry rape” at some point in the future. “Cry rape,” of course, is a phrase itself implying that I would falsely report or accuse someone of rape. That’s a direct personal attack. I responded with sarcasm the first time in the hope that you would take the hint and drop that. But no — now you’ve decided to imagine a scenario in which I’m under the hot lights of a defense attorney’s interrogation, and all of a sudden I’m confronted with my blog post about how I’m a lying liar who lies! Or if not me, some other poor victim, who will bear the consequences of my lying lies and casual way with the truth! Even though this blog post is devoted almost entirely to criticizing the graphic and pointing out its flaws, and merely suggests that the general thrust of the graphic remains correct, somehow it is translated (in this fantasy future courtroom) into a sweeping indictment of the trustworthiness and honesty of all those who “cry rape”! Ah, me. I didn’t realize what I was doing.
For heaven’s sake. I’m going to blacklist you now, on the general principle that this is my house and I don’t have to argue with people who make sly accusations about my character in my own house, and I don’t want to read more of your comments in the morning. Good night.
Pingback: toward a better rape graphic, pt. 2 | The Handsome Camel
Pingback: Steubenville, some add’l notes | The Handsome Camel