In this excellent post, Rosalynde talks about the gender differences in subject material among Deseret Book writers. This renews the discussion brought up by Taryn Nelson-Seawright on the same difference existing in other Mormon outlets. Explanations abound for this phenomena, ranging from differing preferences to piggy discrimination, but most of them are sort of boring. Here’s one that is at least slightly more interesting:
Suppose there are two kinds of writing, which we’ll call A and B. When men and women talk about A in an informed way, what they say is pretty much the same. An example would be a proof of the intermediate value theorem. The proof runs pretty much the same no matter the gender of the speaker. Type B things on the other hand, are much more subjective, and so tend to vary some by gender. Poetry or fiction tends to vary depending on the author’s personal characteristics, of which gender is a big one. What this means is that type A is unified, but type B can be divided into Bm and Bf, to be very simplistic. Thus there are now three kinds of writing, A, Bf, and Bm.
Given this world, let’s suppose that there are more men publishing than women, owing to the fact that more women are caring for children and more men are supporting families. Where should woman writers concentrate her efforts? Well, since there are few Bf books around, but plenty of men in Bm and A, this scarcity drives up the return to Bf, making it easier for her to get published, to earn more when she does, and to feel like she has made a unique contribution.
As a variation on the model. suppose that men don’t actually write very much poetry, so that there really is not all that much demand or supply of Bm books. Well the man are all going to end up in A books. This will tend to push the women out, because they have a high opportunity cost of writing an A book– they could be writing a Bf book.
This idea (which I got from somebody but can’t remember who) surely does not explain the entire gender gap. But then, I doubt any one thing does.
Off the topic (or is it?), but I would question the existence of absolute, non-gender subjective, type A writing, at least not beyond some fairly narrow parameters.
Frank,
Correct me if I\’m wrong, but this rational choice model seems to depict women\’s publishing choices in terms of perceived rates of success in the market. This is seen as the only reason that women actively choose Bf. That is, women\’s writing interests are determined only by the market value that they see in these interests. What room is there, if any, in your model for gender stereotypes in publishers choices or the social construction of \”women\’s interests\”?
The problem that I have with this model is that it attempts to account for the gap by noting that there is a gap. It really doesn\’t tell me anything.
Russell F.,
I believe D&C 93:24 provides adequate basis for the existence of type A writing.
RAF,
To the extent that genderishness varies by subject, the basic issue in the post re-asserts itself. But I am curious how you think a prrof of the intermediate value theorem would look by gender :). I have read a fair bit of writing in economics by men and women, and what is written does not (as best I can tell) typically pivot on the gender of the author.
TT,
We could easily, in a larger model, throw in all sorts of other things. But that isn’t really needed in order to make the point that what I am noting is one influence going on.
In other words, I hope it is clear that I am not claiming that because I wrote down this particular simple model, I don’t think any other issue matters. Rather I hope to add to the list of plausible and interesting influences on gender segregation in publishing, but without drowning the audience in a set of first order conditions and optimization constraints that would be needed if I wrote down a model that included all the possibilities at once.
As for explaining the gap by noting that there is one, I don’t think that is what I am doing, in that I posit a set of fundamentals, which may or may not lead to a gap, and then show that a gap can emerge based on the goals of the actors.
Frank,
But this model prioritizes one explanation over others by highlighting it. If you really are saying only that it is only one possible factor of many then I am not sure what its value is. It wouldn’t tell me that this is more or less influential than the color of a persons hair as a factor for choosing to publish on a certain topic. I suspect that you think that this is not just one factor among many, but maybe even one of the most important factors.
I am not denying that this sort of cold, rational, market analysis may be behind some women’s choices to publish or research on certain subject matters. I am suggesting that this is probably not as influential of a cause as you may think.
As for you assuming a gap in order to demonstrate that there is one, maybe I still don’t understand your point. As I understand it you say that the reason that men publish on certain topics and women publish on other topics is because when they look at the market, they see that men publish on certain topics and women publish on other topics. To me this doesn’t suggest a functioning rational market as a causal explanation, but the influence of cultural values which produce these realities to which one conforms.
Very interesting, Frank. The opportunity cost model might work well to describe male dominance of the historical fiction niche: historical fiction is probably of type A, not highly responsive to the gender of the author, and since a few men already dominate the niche, a woman genre fiction writer would turn to romance or mystery (and we see from the data that some have done so).
On the other hand, it probably wouldn’t work very well to explain what’s happening in Mormon studies, where many of the publishers would LIKE to publish more women in order to boost their credibility. The lack of women in the pages of the journals must derive from something else (I have my hunches).
“But this model prioritizes one explanation over others by highlighting it.”
Perhaps you misunderstand the purpose. I am writing about an interesting idea, not a grand unified theory. In economics, we systematically try to write down models to highlight one point at a time. Then we right down another model to illustrate another point. If the two interact in interesting ways, we’ll right down models with both. But the goal of a model is to strip away other issues to see one more clearly. Not to ignore the others, but rather to be systematic about it. Models of preferences or discrimination are trivially easy to write down (and have been extensively discussed). So there is no gain in cluttering up this model with those other issues. It just makes the math harder. “What math? you ask.” Exactly, in this model I can hide all the math because everything is so simple and one dimensional. If I add in all the other explanations I can’t do that anymore.
“I am not denying that this sort of cold, rational, market analysis may be behind some women’s choices to publish or research on certain subject matters. I am suggesting that this is probably not as influential of a cause as you may think.”
As for cold and rational, I would guess people who actually get published (as opposed to those who dream of it) are those who think rather carefully and coldly about what exactly will get published. Perhaps you think only one’s personal preferences matter, and not market returns. But note that the model may also work through publishers, who should, seeing the scarcity, prefer publishing Bf books to yet _another_ A book. Now, in that world, your authors can be as sentimental and flighty as you wish, but the outcome will still be coldly rational :).
“the reason that men publish on certain topics and women publish on other topics is because when they look at the market, they see that men publish on certain topics and women publish on other topics.”
This is not, actually, the point. Read through it again and you see that the women can choose between two markets: A and Bf. The man chooses between two markets, A and Bm. Many people have noted that A is saturated with men, disproportionate to the number of male over female authors. They question is, why do women spend their scarce writing time on Bf, rather than A? And one answer would be that, owing to diminishing returns, the low number of Bf books (compared to A books) gives them a higher return. Thus, it is exactly the opposite of “women do this because other women do this”. Rather, women do this because not enough other women do this, so the returns are still high. If there were more women authors, or a higher demand for Bm (so men spent more time there), then women would publish more in A.
RW,
They may “like to” but perhaps at the end of the day one more boring exegetical work just doesn’t sell as well as poetry unique to a female perspective. Thus the poetry keeps getting accepted because it has a higher marginal return, even factoring in the political preferences of the journal editor.
That said, this is not the grand unified theory of gender selection in publishing. As I said in my post, there are lots of things going on and no one explanation explains every nook and cranny.
“If you really are saying only that it is only one possible factor of many then I am not sure what its value is. It wouldn’t tell me that this is more or less influential than the color of a persons hair as a factor for choosing to publish on a certain topic.”
I forgot to respond to this. I am not sure why my proffered idea must meet a standard that none of the other theories has– namely being able to rank it compared to others. Ranking theories’ importance is the realm of statistical inquiry and is step 2. First we line all the theories up against the wall…
“But I am curious how you think a proof of the intermediate value theorem would look by gender.”
I can’t even imagine how it would look. Quite possibly this is one of those topics that truely are an absolutely non-subjective gender neutral topic. However, I doubt a history of the economists who worked on and developed the intermediate value theorem, or a sociology of the research-sharing patterns and modelling experiements which led to that development, or an anthropology of the social dynamics and pressures which made coming up with an intermediate value theorem seem worth pursuing in the first place, would all fit in the same type-A, gender-analysis has nothing to tell us here category.
But as you say, your post doesn’t turn how just how much gender variation is out there, but rather merely that it exists. So mine is a nitpicky point.
FYI, Russell, the Intermediate Value Theorem comes from mathematics, not economics. It basically says that if you have a continuous function that goes from negative to postive, it must cross zero somewhere along the way. Sounds completely obvious, but much progress in mathematics has come from carefully proving seemingly obvious things.
And I agree with Frank…math is pretty much just about the math.
“And I agree with Frank…math is pretty much just about the math.”
Ah, you’ve never read Robert Heinlein’s “And He Built a Crooked House,” where an earthquake collapses a three-dimensional representation of a tesseract into a fourth-dimension genuine tessaract. Or somebody else’s story about a subway system so complex that a newly opened stretch of track turns the system into a Mobius strip. Or the one about the guy who goes crazy because his 100 monkeys on 100 typewriters really are producing 100 classic novels letter perfect. Or the one where …
/she wanders off, lost in a pleasant contemplations of beauty bare/
Frank,
In comment # 7 you write: “Models of preferences or discrimination are trivially easy to write down (and have been extensively discussed). So there is no gain in cluttering up this model with those other issues.”
Isn’t your variation (4th paragraph in the post) essentially a model of preference? It seems to me that the model without the variation makes your basic point. Plus, I think the basic model would have men gravitating away from Bm towards A anyway, relative to a situation where men weren’t the disproportionate suppliers of writing.
More generally, the model might be suited to derive other interesting implications, as well. For instance, suppose you allowed for a joint distribution of type A and type B writing abilities across potential authors. I suspect you’d find that decreasing the pool of available female authors relative to male authors (because of child care, etc.) would, in most cases, imply that the writing quality of females doing type A writing is–on average and all else equal–superior to that of males doing type A writing.
Dan,
Yes, I am guilty, the variation I added is unneeded, but it was so easy I wrote it down anyway. I hoped it would help people see what was going on. As for the argument at the end, that’s a roy model sort of argument and I think you are right. Only the ones really good at it would write in A. You get the same result, I think, from a model of discrimination, where only the truly exceptional pass the raised bar.
We see some of this in the econ major, there are comparatively few females, but the ones we get tend to be on average better than the men.
My DW opines that the basic ideas of A/Bm/Bf are probably a decent explanation of why so many women leave careers that could be done by just about anyone else and choose to stay home with their children…
Great post, I see racial self-segregation all the time, and I want to investigate the issue more thoroughly.
I always find something new and interesting every time I come around here – thanks.