Do women face a reputational bias when they co-author?

by on November 12, 2015 at 11:07 pm in Data Source, Economics, Education, Science | Permalink

Apparently.  Heather Sarsons has a paper on this phenomenon (pdf), the abstract is this:

Within academia, men are tenured at higher rates than women are in most quantitative fields, including economics. Researchers have attempted to identify the source of this disparity but find that nearly 30% of the gap remains unexplained even after controlling for family commitments and differences in productivity. Using data from academic economists’ CVs, I test whether coauthored and solo-authored publications matter differently for tenure for men and women. While solo-authored papers send a clear signal about one’s ability, coauthored papers are noisy in that they do not provide specific information about each contributor’s skills. I find that men are tenured at roughly the same rate regardless of whether they coauthor or solo-author. Women, however, suffer a significant penalty when they coauthor. The results hold after controlling for the total number of papers published, quality of papers, field of study, tenure institution, tenure year, and the number of years it took an individual to go up for tenure. The result is most pronounced for women coauthoring with only men and is less pronounced the more women there are on a paper, suggesting that some gender bias is at play. I present a model in which bias enters when workers collaborate and test its predictions in the data.

See also this very interesting paper on “Confidence Men,” in economic science, women seem to have more epistemic modesty than men.

Hat tip goes to Dina Pomeranz.

1 Guest61 November 12, 2015 at 11:23 pm

kilo samba mortimer trajan grey acting dubbed high definition partner zoom credit solutions worries tablet

2 Matt November 12, 2015 at 11:33 pm

Finally someone has the guts to say it.

3 Zack November 13, 2015 at 1:21 am

Brilliant

4 Paul November 12, 2015 at 11:27 pm

This paper only looks at publication data. It does not adjust for differing teaching responsibilities, service, or family commitments.

5 Joan November 13, 2015 at 12:37 am

Yes they do

6 E.A. Potobaum November 13, 2015 at 5:02 am

Could you tell me on which page the methods used to control for teaching, etc are described?

The only controls mentioned in the paper that I’m aware of is “I use the RePEc/IDEAS ranking of economics journals to control for the quality of a person’s publications” on page 14 of the pdf.

7 Rguest4T November 12, 2015 at 11:36 pm

men women bells Christmas bolt tennis boat buy TBD dance suit eggs bags books carpet

8 Ptintnbotcode5 November 12, 2015 at 11:46 pm

What you said made me think. Everyone has such days and concerns. Yet, thoughtful presentation, thanks, we surely need more of it. Your blog is a nice reading.

9 Ptintnbotcode5 November 12, 2015 at 11:47 pm

anime dubbed tokyo nice sailor cheap hong kong dragon blog dollars

10 Adrian Ratnapala November 13, 2015 at 6:20 pm

I think someone is trying to build an artificially intellgient talk-bot which will eat us all. A mere spammer will try to generate actual sentences, but these guys want wee-HAL to learn everything from scratch. That means sending random verbiage onto the intertubes in the hope that some like me responds.

Little by little it will learn to perceive the world with internet-senses alien to us just as real as sight and hearing. It will have an intimate understanding of human cultures and behaviour, but without our preconceived notions about the same. Then it’s chow time.

11 Ptintnbotcode5 November 12, 2015 at 11:52 pm

mao suits tatame shoes kung fu rowing gymnastics leotards gears ping pong clothes vietnam tai chi gdp want kendo

12 Barkley Rosser November 13, 2015 at 2:54 am

The seriousness of the point of this post is given by the complete lack of comments on this matter of gender bias in joint publications regarding the Case-Deaton vs Deaton-Case link in the previous post, where most of the comments were by people who were running around in white sheets with burning crosses while attempting to ridicule the Open Borders movement.

Here with maybe one exception most of the comments have either attempted to justify sexist bias or been just plain gibberish, perhaps amusing for the first round, but then just stupid.

13 Josh M November 13, 2015 at 3:18 am

It’s pretty late, Barkley, and the gibberish is being posted by spam bots (who have also affected the previous 3 or 4 posts).

14 dearieme November 13, 2015 at 7:41 am

But were they male or female spam bots?

15 E.A. Potobaum November 13, 2015 at 4:52 am

If you’re going to search through the data and only write up ones with large “unexplained” differences… the obvious hypothesis for the cause of differences would be random fluctuation. https://en.wikipedia.org/wiki/Data_dredging

With that said, here is what I’d have written to the author were I a reviewer for her paper:

1. Rather than model economists as “high ability workers” and “low ability workers”, why not use the data you’ve already collected to do this analysis with respect to the actual future output of a worker? (Your mention of controlling for productivity in the introduction made me think this was where you were headed; I was sad to see this simpler model substituted when we finally got to the math).

2. I was somewhat confused to see p-values mixed with bayesian analysis, and would be interested to see what your appendix would look like if redone in a more bayesian manner. (“when an X gendered economist publishes a paper, there’s a change of 1:N in their odds of getting tenure”, etc). This could nicely be combined with point #1 if you also included the betting odds for future productivity. (“when an X gendered economist publishes a paper, there’s a Y expected lifetime increase of publications”)

3. Your sample spans 1975 – 2014. I know the data is sparse, but I would be very interested in knowing how those betting odds have changed over time.

16 dearieme November 13, 2015 at 7:43 am

May I suggest that X and Y are not the ideal choice of variable symbols here? Unless you are being deliberately waggish.

17 Gochujang November 13, 2015 at 9:15 am

Status seeking is one of Pinker’s universal human drives. Academeia amplifies it, or channels it. Should we be surprised that males in an amplified status seeking environment actually do so?

I favor calm contemplation of ideas, self-control, but I am certainly not shocked.

18 charlies November 13, 2015 at 12:14 pm

hypothesis: when a woman co-authors with only men, maybe the men agreed to the arrangement for reasons other than the female co-author’s productivity?

of course, even if true, that probably doesn’t explain the full magnitude of the effect the researchers find. but it could be part of the dynamic.

19 ZZZ November 13, 2015 at 12:39 pm

“The result is most pronounced for women coauthoring with only men and is less pronounced the more women there are on a paper, suggesting that some gender bias is at play.”

So papers with more female authors do better then papers with a more equal numbers of men and women. How exactly does this imply that there is a bias against female authors?

Comments on this entry are closed.

Previous post:

Next post: