I thought before I posted my thoughts on Emily Glassberg Sands's study on sexism in the theatre, I would first post the e-mail I sent to her on Sunday (to which she has not yet responded):
Hi Emily -
Thanks for talking with me yesterday, however "contentious" the conversation might have been!
To follow up on our discussion, I wanted to offer you (as promised) the chance to respond to a few questions I have about your study via e-mail.
Let me say upfront, however, if I never got a chance to during our conversation, that I greatly admire the amount of work you put into the study, and I think the "audit survey" portion of the paper is brilliant.
But as I mentioned in our talk, I'm having some trouble with your final arguments regarding Broadway productions.
My questions about that portion of the study are the following:
First, I wanted to return to my concerns about your "female writer" sample, because I admit I got a little confused during our discussion and didn't quite ascertain the following: did you code musicals with a female book writer as "entirely" written by a woman? That's my interpretation of the text of your study, but I just wanted to check to make sure, because I think most theatre people would consider the composer the salient "writer" of most high-profile musicals. And if you are considering a musical with a female book writer as a "female-written show," could that possibly be a driver of the "higher weekly revenue" numbers you discovered?
A slide from Emily Glassberg Sands's presentation in New York.
Second, I remain concerned about the fact that your study relies on a proxy (show type) to approximate profitability of Broadway shows, rather than actual hard data on that profitability. This bothers me particularly because in your NY presentation, despite the use of the show-type proxy in your tables, you nevertheless titled a key slide, "Female-written shows on Broadway are 18% more profitable than male-written shows," [slide reproduced above] while another slide title referenced "higher weekly profits for female-written shows," statements which could reasonably lead the casual reader to assume that you actually have hard data on those profits (when you don't). Will you be changing those titles in any future presentations?
A further question about those proxies. I can see them as indicators in Eq. 6.3 on page 97 (PDF) of your study, and you note below the equation that these should "control partially for production costs" (a key component of profitability). My question is, can you more accurately define "partially"? Did you have average production costs figures for your show type categories, and if so, were these applied to the analysis?
Finally, you mentioned repeatedly during our talk that you weren't going to discuss the "implications" of the study, but would limit your comments to the study itself. My question is, then, whether it concerns you that many descriptions of the study seem to be omitting the problems with the "proxy" you used for profitability in your final chapter? The NY Times "Economix" blog, for instance, said that you "found that shows written by women earned more money than shows written by men," without mentioning that those shows still might not have been more profitable. The New York magazine article about you didn't make this distinction either. I understand your reluctance to comment on the "implications" of your study, but do you worry that the discussion may be being driven by an inaccurate comprehension of your conclusions? And do you think there is a way in which you could address that seeming recurrent mistake?
Again, thanks for your thoughts, and I look forward to hearing from you -