Try having your students spot the problems in this article by Christina Hoff-Summers. As Igor, who sent it to us notes:
Maybe the best point is when the author criticizes not taking variables like education into account when computing the gender gap, and then proceeds to cite a result that urban young women make slightly more dough than their counterpart men due to higher average education level [as evidence that women now have it better than men]
Feminist Philosophers 
Um, CHS is odious and I disagree with her politics, but I am just missing something here. Why is it a fallacy to criticize those that say there is a general pay gap for not taking education into account, and then give an example where if you do take it into account the gap goes away? Her argument seems empirically weak and light on context, but the move described in the post doesn’t seem like a fallacy to me.
I had to get up at 5:00 am to catch a flight and I’m in an airport – perhaps I am tired and distracted enough that I am missing something?
I thought: the stat only makes it look like women are doing better if you fail to take education into account. Once you do that, it turns out women aren’t doing better.
Does that make sense? I’m getting worried now.
I’m thinking there may be a problem with the quote from Igor. If she does not mention education level when she says women now earn more, then she is guilty of the fault she finds in others.
Of course, I could read her stuff and see just what’s going on, but then it is so bad for one’s blood pressure.
She mentions it, but fails to see that it undermines her point about the pay gap being reversed.
I took it her argument was:
If you ignore education altogether, it looks like women do worse. But in the demographic corners of the population where women are more educated than men, they actually do better (generalizing from her one example). So if we were to control for education in general, the gap would not show up.
Empirically shaky indeed, but not exactly a fallacy I don’t think.
I agree that it’s not a fallacy. Not sure why it would undermine her point, either. And is she really guilty of generalizing from the one Reach Advisors study? Her point – that “When these factors are taken into account the gap narrows considerably” also seems to find support in the 2009 DoL/CONSAD metastudy she links.
Yeah I don’t see why it undermines her point either. Can you explain, Jender?
OTOH, as we all know from the Chronicle stats, in academia women make reliably and significantly less than men, controlled for rank, discipline, and institution, even though we generally all have the ‘same education’ in this line of work. So the point can’t be generalized indefinitely.
Jender- I think the stat makes it look better when you do take education into account. I *think* though, that particular number may be coming from a very recent study that looked at wages for men and women in certain metro areas. If it is from that study- it may not be an indicator of women’s success so much as in indicator of the economic situation. In a number of lay-off situations, a disproportionate number of men were laid off- because they were paid more- leaving behind a higher number of women.
More men have been laid off, not just because they were paid more, but because traditionally male jobs such as construction worker, home renovator, and factory worker have been much harder hit by the recession than traditional female jobs such as those in the service and care-taking industries.
So I agree with the spirit of Kathryn’s point – but even here, it’s not all about women being paid less for the same work.
Rebecca- I think the study I had in mind only compared wages of employed men and women so I wasn’t thinking about the overall situation- but rather the differences within the workforce after layoffs.
Rebecca, I see how you’re reading her argument. Yes. it’s better that way.
Quoting CHS:
When these factors [differences in education, experience, and job tenure] are taken into account the gap narrows considerably — in some studies, to the point of vanishing. A recent survey found that young, childless, single urban women earn 8 percent more than their male counterparts, mostly because more of them earn college degrees.
That’s an instance of the metonymic fallacy: not all women are young, childless, single, and live in urban areas.
Her overall argument appeals to either a false dichotomy or something that I think is called the simple explanation fallacy. The previous paragraph:
But that wage gap isn’t necessarily the result of discrimination. On the contrary, there are lots of other reasons men might earn more than women, including differences in education, experience and job tenure.
First, she’s leaving out number of children, which makes a huge statistical difference: if I remember right, whatever else you control for, women with children make less money than men with children. Anyways, the options she presents are not mutually exclusive, as `[o]n the contrary’ seems to imply; and interactions between several of these factors (along with race, class, &c.) could very well be the best explanation: that discrimination leads to women being passed over for promotions or other opportunities to gain experience doing higher-paying kinds of work, for example. Along with that, she’s assuming that the pay gap has one explanation across the board — in all industries, within all races and classes, &c. I think it’s quite likely that there’s a complex web of factors, some of which are more active than others in different situations. That’s a more general version of the simple explanation fallacy.
I don’t know why I didn’t realize this until Dan quoted it again, but her stat is not really taking education into account. She says women are earning more than their male counterparts- because more of them are getting degrees. Which means it’s not comparing men and women with equal educational background.
Kathryn: You’re correct that the 8% figure does not take education into account, but as Hoff-Summers alludes in the same sentence, the authors of the study said that the 8% difference was accounted for almost entirely by educational differences. I saw in another source where the primary author did assert this; I’m assuming this means that when when they then controlled for educational attainment, the men and women studied earned about the same – hard to say for sure since I don’t think the full study has been published. So it seems to me that it does actually support what she said: in some studies there is virtually no gap if you control for education (and the certain other factors she highlights). I think that’s what Hoff-Summers had in mind by introducing the Reach Advisors study.
Dan Hicks: I don’t think she’s actually committing a metonymic (synechdochic?) fallacy there. She’s citing an example in support of the statement “When these factors are taken into account the gap narrows considerably — in some studies, to the point of vanishing” – but does not appear to be reasoning from the one study to her more general assertion about the gap narrowing. The metastudy she cites just afterwards also provides support for the statement.