I don't understand why Susanna Schrobsdorff felt the need to lie.
An old piece I wrote about a column that appeared in the Washington Post
Last April, I saw a column in the Washington Post that I found quite offensive, so I wrote this in response and submitted it to the paper. That submission was rejected with a suggestion that I trim it down to a letter to the editor. I did so, and was rejected again. The Post did publish this more mild rebuke, which doesn’t address what was offensive about the original column or do a detailed dissection of Schrobsdorff’s non-argument, but it does at least allow someone with relevant expertise to point out that her thesis is backed up by less than zero evidence.
In any case, here is my response column, before edits for wordcount, with updates in [brackets] and some contextual footnotes added in. I also looked into the one scientific study Schrobsdorff cited; it looked like garbage and I sent an analysis to the publishing journal saying as much. This is likewise reproduced below. The letter substitute for the detailed response has a paragraph criticizing the Post for rejecting the breakdown, but isn’t really worth reproducing.
CN: Implicit devictimization1; some discussion of intimate partner abuse and sexual assault.
(Response to Schrobsdorff)
To accuse a person of lying in a national newspaper is a heavy thing - regardless of which party is publishing in the paper. Therefore, I feel I must justify my claims. On [April 21, 2022]2, a column by Susanna Schrobsdorff was published in the Washington Post on the subject of an oral contraceptive for men that [was then] soon to move to human testing. In it, Schrobsdorff does quite a bit of lying. She lies about the history of oral contraceptives for men and about the potential market for them. To be clear, none of the sentences she writes are strictly untrue. Still, she relies on a series of unconnected statements to create a fairly typical false impression: that men are, as a rule, lazy, stupid, impressionable, and impulsive, and thus have no interest in any sort of contraception. In the course of this, she cites evidence unrelated to or actively contradictory to her claims about male oral contraceptives.
First, after some vague discussion of appeals to masculinity in advertising and a brief cameo from a never-to-be-seen-again anonymous male friend, Schrobsdorff quotes an unspecified medical researcher saying that men have a "lower threshold" for side effects, and follows this up with a vague reference to earlier, hormonal "versions of the male pill", which had side effects that she compares to those of traditional female oral contraceptives. The side effects she mentions are indeed side effects of female oral contraceptives, but otherwise this claim about earlier research is false. It is not strictly Schrobsdorff's falsehood, but it is still worth correcting. The earlier study she references was of an injected drug, not a pill, for one, and the study was halted by an independent safety board3 due to side effects similar in kind to those experienced by female hormonal contraceptive users occurring at several times the rates they do in those other drugs - and it was stopped despite 75% of the participants wishing to continue.
The researcher quote is more telling. It cites a Gizmodo piece on the same research that prompted Schrobsdorff's column - and Gizmodo credits the "lower threshold" quote to none other than the lead researcher developing this new treatment[, who doesn’t attribute the low threshold to his hypothetical patients per se but rather to their circumstances, ie the fact that the absence of contraception does not pose a direct physical risk to them]. And that article also points back to an older Gizmodo piece discussing the contraceptive gap, which consists of commentary from four experts discussing the topic. Only one of these experts mentions side effects, and even he is quite explicit that "the bigger bar ... is being able to bring the sperm count low enough". And all four, in various ways, discuss the presumed lack of interest in a male contraceptive and how false that presumption actually is. I am not so uncharitable as to believe that Schrobsdorff misread or misunderstood the Gizmodo piece(s). Sadly, that leaves me with deliberate intent as the only remaining reason why the story changes so much from Gizmodo's coverage to her column.
Schrobsdorff follows up this discussion with the statement that "a 2016 study of 51 males found that men were 'less likely' to wear a condom when they judged a potential partner to be attractive or in good health." This is then used to justify the claim that "until they can suffer the consequences of an unintended pregnancy, men aren't likely to reliably take responsibility for birth control." The study she cites is itself of dubious quality - in addition to the comically small sample size, the research design and analysis have a certain Cornell-Food-and-Brand-Lab flavor that followers of Andrew Gelman, Uri Simonsohn, and their colleagues are likely to recognize. Beyond that, though, while Schrobsdorff's characterization of the conclusion is not strictly wrong, the study itself is not at all concerned with contraception. That "in good health" bit tacked on to the end of her sentence is a concession to the fact that the study is actually interested in the ways in which attractiveness interacts with male STI-risk mitigation in hypothetical one night stands. Assuming the study's conclusion is valid at all, it obviously does suggest a potential inference relating to contraception, but Schrobsdorff's claim that this study demonstrates that men can't be trusted to use an oral contraceptive is, at best, a reach. The followup sentence is particularly egregious, given the habit female rapists and groomers have of using pregnancy as a tool to regain access to and control over their victims, and the fact that reproductive coercion is the one type of intimate partner abuse that is more frequently committed by women against men than the reverse.45
Schrobsdorff concludes with a reference to a remarkably impractical contraceptive device for men that won a Dyson Award recently, alongside another vague reference to advertising and, for some reason, the aerospace industry. This is chosen in place of any reference to the other potential male contraceptive mentioned in that Gizmodo article that seems to have prompted the column. She concludes with a further dig at the fact that men are at present almost entirely reliant on their female partners for reliable contraception.
The hardest part of her column to understand, for me, is the why of it. Why write this? What purpose could such a column serve? I had not heard of this new contraceptive option under research, which is understandable. People tend not to ask whether I want a child, now or in the future. I get the sense it doesn't occur to them I might have thought about the question. This column, though, is not a report on the state of the problem or even a call to action for other men who may not have realized that this choice could be on the horizon. It seems to be nothing more than a preemptive declaration of failure for any and all attempts to change the status quo, and I just don't see the point of that.
Below I present the critique of the condom use study that I sent to the BMJ. This was published in relatively short order and can still be found attached to the original study. I have not received any other response on this subject from BMJ or the authors, as far as I am aware (and to be clear I don’t necessarily expect one).
(Breakdown of issues in condom use study)
Dubious Noise Mining Used to Denigrate Genuine Medical Research
I write because this study was cited by a rather offensive opinion piece published on the morning of April 21, 2022 in the Washington Post. While that opinion made no particular argument beyond a rather vague claim that research into hormonal contraceptives for men is without merit on the grounds that men have no interest in contraception, it did cite this study in support of the claim that "men are 'less likely' to wear a condom when they judge a partner to be attractive or in good health." While the relationship between this claim and the rest of the column is tenuous, it seems to me like a fair characterization of the study authors' conclusion.
However, on reading the study, I have difficulty trusting even that claim. My concern is that this study looks like little more than noise-mining: questionable data with superficial resemblance to the purported research question, hammered with a grab-bag of statistical techniques for significant-seeming results with an arguable relationship to the researchers' conclusion. I should admit here at the outset that I do not have the statistical background to give an in-depth critique of the mathematical analysis of the results. Still, there are certain numerical hallmarks that lead me to question the confidence that any conclusion about men in general can be drawn from these data.
First, the sample size in this study is absurdly small. As a rough metric, we can observe some quirks in the data on lifetime sexual partners. For one, there appears to be a single individual - the one who reported 60 lifetime sexual partners, itself a rather dubious number given that it implies a minimal average of one new sexual partner every nine months for a period of 45 years, based on the other stated demographic extremes - who accounts for a 90% increase in the average number of past sexual partners in the sample. Alternately, converting from mean and standard deviation to a 95% confidence interval, the authors appear to predict that men similar to their sample will have had between -10 and 20 lifetime sexual partners. That these nonsensical summary statistics are presented without comment leads me to doubt the validity of all of the statistical models described in the paper.
Second, the paper does not appear to make any attempt to scale its significance tests based on the number of comparisons it makes. By my count, the authors attempt to model 80 distinct relationships among thirteen different variables with no mention, at any point, of the additional degrees of freedom introduced this way. The authors claim that their samples provide "over 1000 data points for each measure", but this is
flatly incorrect[a misleading characterization of the data]. Depending on whether one analyzes the subject men or the photo being rated, there are either 20 or 51 data points across 51 or 20 measurements on five different measures, and 51 data points on each of 8 other measures. [A more reasonable statement might be that their samples provide 51 groups of 20 data points on their primary measures, with 8 other measures to characterize the groups.]
All of this, though, is not even my deepest concern with this paper. That deep concern is with the basic experimental design. This is a study that wishes to examine a possible relationship between the attractiveness of a sexual partner and the likelihood that a man will choose to use a condom in an unspecified sexual encounter with that partner. At no point in the study is any attempt made to examine or control for the relationship between the man and the hypothetical partner. Indeed, the only concession to the possible existence of other partners is the presumption that the man has none in the hypothetical! Based on the way the question was asked, there appears to be an unstated assumption that there is no connection between whether a man chooses to use a condom in any given sexual encounter and any sort of prior interactions between the man and his partner. Every sexual encounter is treated as the first sexual encounter between partners. Furthermore, the measure of attractiveness being used here is a single photo of the potential partner's face. Every sexual encounter is thus actually treated as if it is the first direct encounter of any kind between partners - with absolutely no attempt to adjust how either the historical or the experimental data is treated as a result.
Likewise there is no effort to restrict the various questions asked to any single hypothetical where the men being studied could consider each option and make a choice. Participants are asked how likely they would be to have sex with the woman "if the opportunity were to arise", how likely the participant would be to use a condom if they were to have sex with the woman, and how likely other men similar to them would be to have condomless sex with the woman. But the various possibilities are never presented as a choice about which the participant is making a decision. They are only asked to predict their own behavior, on the basis of nothing more than a photograph of a woman they have never met. It seems to me like it would be difficult to frame a series of questions about these hypotheticals in a way that better obscures the actual decision-making process of the person answering.
Additionally, all of the experimental responses are input via a slider from 0 to 100. Together with the total lack of any sort of contextualization of the hypothetical sexual encounter or relationship, this raises serious doubts in my mind about the test-retest validity of any of the experimental data. This, in turn, leads me to suspect that the experimental data discussed here is meaningless in any sort of practical sense. Women, after all, are not fully described by a single black and white photograph of their face, and men are as a rule aware of this fact and able to adjust both their behavior and their assessment of the women in question accordingly.
Incidentally, it appears that one of the heterosexual subjects of the study was in fact bisexual. How this interacted with the questions asked about said subject's sexual history or the analysis of the subject's opinions about the women in the pictures is never addressed.
The commentary of the peer reviewers on this paper appears to be concerned exclusively with the clarity and readability of the paper, and with the inclusion of citations to other papers on similar subject matter. The logic, or lack thereof, connecting subject matter to experiment and experiment to conclusion is never commented upon, and the validity of the statistical methods used for analysis are likewise of no apparent concern. The value of peer review over a simple editorial pass in this situation seems, to me, questionable, especially when this study is subsequently used to give a veneer of scientific backing to an effort to devalue genuinely novel medical research.
I may return to this topic in the future, but I am nervous about doing so. The attitudes exemplified here by Schrobsdorff (and others like them) are a real problem, but talking about such things has a resonance with the rhetoric of people like Andrew Tate and other toxic grifters with a masculine focus. The absence of dedicated discussion about the ways in which sexism hurts and dehumanizes men, specifically, is one of those pressing concerns for which I have no satisfactory answer. It is a difficult situation to live with.
Finally, a message to Ms Schrobsdorff, should this post find its way to her:
If you now find it more difficult to get columns published in prominent national newspapers, well, the thing is that when you say something so profoundly stupid, it makes everything else you have to say a little bit suspect, especially if you’ve never addressed it.
If, on the other hand, someone uses this post as an excuse or a method to harass you, I sincerely apologize, and state plainly that if there is something I can do that might reduce or mitigate such harassment, I would like to be told. What offended me were the words of your column, the ideas they expressed, the blithe assumptions about my thinking and my desires - not anything specific to yourself. I believe that we can do better, and I hope that you do too.
I just made up the term “devictimization”. I think it better fits the profile of negative sexist responses experienced by male victims (of many things, but especially sexual assault and intimate partner abuse), as compared to “victim blaming”. My impression is that women are more likely to hear, “the bad thing that happened to you was actually your fault,” and men are more likely to hear, “the thing that happened to you was not actually bad.” For an example separate from gender issues, Ye’s comments on slavery being a “choice” would be victim blaming, while Lost Cause assertions that slaves were happy in or benefited from their subjugation would be devictimization. Of course men, women, and every other flavor of person can and do experience either or both response, but I am by nature a Splitter and this is my understanding of the overall trends and phenomena.
The Dobbs draft was leaked on May 2, a bit more than a week after I first wrote this essay. At the time, I recall serious concerns about the Roberts court endorsing the 15-week ban at issue in that case, and about the possibility that the Texas abortion bounty law would lead, one way or another, to an end of what we might call the “Roe/Casey era”. That said, the arrogance, self-righteousness, unscrupulousness, and sheer naked prejudice of Alito’s decision came as a surprise. In retrospect it was a long time coming, but the shift from at least attempting to disguise misogyny to open contempt for women among American reactionaries was nevertheless quite fast. I freely admit I was caught off-guard when it actually happened.
No, I can’t find any first-party documentation about the shutdown of this trial. Yes, the links in that Vox piece are broken. I agree, this is deeply annoying.
This particular claim is based on the statistics reported in the 2010 NISVS report (p 48) which gives lifetime general reproductive coercion victimization rates of 10.4% for men and 8.6% for women. It also gives the lifetime rates of a partner explicitly trying to cause a pregnancy over the victim’s objections at 8.7% for men and 4.8% for women. It’s hard to say which of these comparisons is more accurate, since contraception and STI mitigation are more entangled in the sexual behaviors of men than in those of women.
More recent NISVS data seems to combine reproductive coercion with other behavior into broader coercive subcategories of “psychological aggression”, and as far as I can tell this specific form of abuse can’t be readily broken out. It is nevertheless worth noting that in the most recent report (p 6), those broader categories of coercive behavior tend to impact women somewhat more than men, by a bit less than 5% (with about 40% or 20% of men experiencing the behavior in question, depending on the scope of comparison). Finally, of course, all of this data (more or less; that most recent report is from 2022) is pre-Dobbs.
As with most forms of intimate partner abuse, reproductive coercion is discussed almost exclusively in terms of female victims, with a heavy presumption of male perpetrators. Outside of the NISVS, it can be hard to find any data on male victims at all. For a rather striking example of this phenomenon, look at statistics on rape, where the overwhelmingly most common form of rape experienced by male victims is excluded by definition from more or less every sample.