I've been turning this question of gender and racial diversity over for a while, in my own head, in conversations with oldcharliebrown and the_flea_king, and lately, watching it erupt in our little corner of the blogosphere.
For example, there's lately been a kerfuffle over whether F&SF under editor Gordon van Gelder has developed or expanded a bias against women writers. (GVG has been handling this question with admirable grace and professionalism, in my opinion.) Numbers provided by oldcharliebrown indicate a declining trend since he took over the magazine.
Source: oldcharliebrown, compiling from historical tables of contents.
It's very easy to look at that chart and conclude that editorial bias — conscious or unconscious — is in play. I don't believe this. More on that below.
From the same source, Asimov's under Sheila Williams has bumped to a multi-year average of 27% women authors from about 24% women authors under Gardner Dozois. Oddly, this compares well to about 25% women authors for F&SF under GVG. By comparison, Realms of Fantasy runs about 44% women authors over a multiyear average. This averaging ignores trending, obviously, but it's still a useful discussion point. Over the past three years, all three markets under discussion have trended roughly flat or slightly up. Taken over larger time spans, the swings are more dramatic, per the chart above.
Other markets have provided data. Polyphony has always had a goal of striving for gender balance in our tables of contents. At the same time, our submissions always ran about 35/65 female-to-male ratio. (Casual observation and anecdotal evidence suggests this ratio holds true in a lot of markets, though Jed Hartman in particular and Broad Universe in general have done a lot of work to run these numbers out in detail.)
Shimmer data provided by maryrobinette shows that over the history of the market, the magazine has published 21 stories by women and 34 stories by men. Due to their blind submission process, she was not able to tell me what the submission ratio was, but this 21/34 roughly maps to Polyphony's 35/65.
I don't think the numbers themselves provide a conclusive evidence of editorial bias against women. If nothing else, the statistical universe is too small for general conclusions, and each market must be taken on its own. Anecdotal evidence would seem admissible in this discussion, and anyone who's ever met GVG, for example, knows him to be a committed, engaged and honorable man who's passionate about the field and passionate about his magazine. Hardly the transparent case of the male oppressor by any stretch of the imagination.
So where do the numbers come from? How do we get from, say, 35% women in the slush pile to 25% in the pages of F&SF and Asimov's, yet 44% in the pages Realms of Fantasy?
To me a strong candidate answer is both obvious and somewhat facile — market bias. Each of those markets has a distinct demographic that it addresses, and that demographic influences editorial selection. The starkest example of this was when Ellen Datlow edited SCI FICTION — though she is a lover of horror fiction, with a very broad and deep understanding of horror and dark fantasy, that wasn't the editorial remit of her market, and so the horror that was very much to her taste was dramatically de-emphasized in her selections for SCI FICTION, given the demographic she was serving.
ktempest asked about "vagina stories". Certainly there are, just as there are "parent stories", "dog stories" and so forth. Yet if the readership is largely male (slightly less than 30% of subscribers to F&SF are female, I believe), if there is a gender bias in the stories selected for publication, it's serving the interests of the paying customers. Is this a problem? It depends on your view of how the world should work. Reading isn't prophylactic, and people who buy magazines for entertainment aren't seeking involuntary enlightenment.
And go back to the numbers for a second. Recapping a few assumptive leaps about F&SF from above, we've tossed out 30% female subscriber ship, 35% female submission ratio and 25% female publication ratio.
Those numbers actually don't seem that whacked out to me. Should that 25% be 30% or 35%? I don't know. Maybe so. But as someone who's edited a competitive market, and as someone who knows Gordon Van Gelder, I can tell you there's no way in hell he's making his selections based on the apparent gender of the author. And his numbers, from readers to submitters to publication, are clustered. Which doesn't scream 'bias' to me.
The racial bias discussion as raised by ktempest, Tobias Buckell and others, is more difficult and more touchy. I don't have numbers to tear into, as race is almost impossible to discern from a submission, except by using the Barnes Principle — "Science fiction is about white people and their imaginary friends." Most of the submitters are white, most of the published authors are white. But due the nature of the process, the bias can't be on the editorial or market side. It has to be on the developmental side, which is a topic for another time.
'Nuff said for now. Fire up the comments and tell me where I got it wrong.