Are Women Philosophers Underrepresented in Top Ethics Journals? (guest post)
The following is a guest post* by Maggie Dalecki (Manitoba), Meena Krishnamurthy (Michigan), Shen-yi Liao (Puget Sound), and Monique Deveaux (Guelph), based on research presented in “The Underrepresentation of Women in Prestigious Ethics Journals,” forthcoming in Hypatia.
Are Women Philosophers Underrepresented in Top Ethics Journals?
by Maggie Dalecki, Meena Krishnamurthy, Shen-yi Liao, Monique Deveaux
That philosophy has a gender problem is unlikely to be news to anyone who studies or works in the discipline. The gender imbalance is obvious in many (if not most) classrooms and in the make-up of most philosophy department faculties. The question of why philosophy has the gender imbalance that it does is complicated, and while it will probably never be possible to locate a single cause for it, there has been a lot of interest and promising research that is starting to investigate this multifaceted problem as of late.
In ‘The Underrepresentation of Women in Prestigious Ethics Journals’ (forthcoming, Hypatia), our goal was to investigate whether women are underrepresented in prestigious ethics journals relative to the number of women working in philosophy and specializing in ethics. While we recognize that the gender problem in philosophy is multifaceted, we focus on the issue of journal publishing because we think that if women are underrepresented in top philosophy journals, then this, in and of itself, is a significant aspect of the gender underrepresentation problem.
To preview, we found that women are underrepresented in ethics publishing. While the number of women who work in academic philosophy is already small, the number of female-authored articles that are ultimately published in prestigious ethics journals is even smaller. That women are more likely to specialize in ethics than in other subfields of philosophy does not suffice to mitigate this effect. Our hope is that our study opens the door to many other important questions, particularly pertaining to what sorts of things are causing this underrepresentation. In turn, answering this causal question might lead to a deeper understanding of the more general gender problem that is seen across philosophy as a whole.
What Does ‘Underrepresentation’ Mean?
We knew going into this study that there are more male than female-authored articles in prestigious ethics journals, but given that there are more men than women working in philosophy, this is to be expected. If, however, the proportion of woman- authored articles in prestigious ethics journals is lower than the proportion of woman philosophers specializing in ethics, this comparative information can help us to establish more clearly whether there is a gender problem specific to journal publishing (or whether the imbalance seen in journal publishing is just a symptom of the broader problem seen all over philosophy).
Publishing in the subfield of ethics is particularly relevant to the issue of journal publishing given that conventional wisdom has it that women are disproportionally likely to specialize in ethics (Haslanger, 2009, 3; Schwitzgebel and Jennings, Forthcoming). If the number of woman-authored articles is disproportionately small, even in a field where women are likely to specialize, the problem in philosophy publishing may be even worse than thought. The goal of our study was to determine whether women are underrepresented in prestigious ethics journals relative to their representation in the field.
We first estimated the percentage of women in philosophy who specialize in ethics. To do this, we used the faculty lists that accompanied the annual Philosophical Gourmet Report’s (PGR) ranking of the 50 departments between 2004 and 2014. We counted total faculty per department and determined the gender of each faculty member by appealing to the first names and pronouns used on departmental webpages and individual CVs. Next, we estimated the number of women specializing in ethics by looking at the stated AOS or research interest of each faculty member, again appealing to department or personal webpages and CVs. In 2004-2005 for example, women made up 19.1% of continuing faculty in the departments we looked at. 24.3% of those women had an AOS in ethics.
We then estimated the number of woman-authored articles in prestigious ethics journals. Here, we examined the table of contents of four ethics journals – Ethics, Philosophy and Public Affairs (PPA), Journal of Political Philosophy (JPP), and Journal of Moral Philosophy (JMP) – between 2004-2014, counting the total number of publications as well as the gender breakdown of the authors of those publications. We included articles, literature review essays, discussion, debates, survey articles, and introductions in our counting of publications.
| Ethics | PPA | JPP | JMP | |
| Total | (49/263) 18.6% | (31/47) 21.1% | (74/260) 28.5% | (60/273) 22.0% |
Are Women Ethicists Underrepresented in Ethics Journals?
Finally, we compared the percentage of woman philosophers specializing in ethics to the percentage of articles authored by women and found that women are underrepresented in prestigious ethics journals relative to their representations in the field of ethics. Overall, we found that the mean proportion of women specializing in ethics from 2004-2014 was 27.1% (SD = 2.0%) while the mean proportion of women-authored articles published in prestigious ethics journals was 22.6% (SD = 2.3%). This difference was found to be statistically significant.
To look a bit deeper into the data, we compared the proportion of women in ethics to the proportion of woman-authored articles published by each of the four journals individually. Notably, the only journal with a statistically significant discrepancy in proportions was Ethics. Philosophy and Public Affairs, Journal of Political Philosophy, and Journal of Moral Philosophy did not show statistically significant differences between the proportion of woman-authored articles and woman ethicists. The numbers for Ethics (which is probably the most prestigious of the journals examined), when examined in isolation, are rather striking: in 2004 for example, we estimate that the percentage of women working in philosophy who specialized in ethics was 24.3%. The number of woman-authored articles published in Ethics in 2004 was just 8.0%. In 2014, the percentage of women specializing in ethics was 29.2% while women authored only 20.8% of the articles published in Ethics.
Our central finding was that yes, overall, women are underrepresented in prestigious ethics journals. In stating our central finding, it seems worth emphasizing that our study was carried out in a rather conservative way, meaning that the underrepresentation of women in prestigious ethics journals might actually be worse than we found. First of all, we counted any article with at least one woman author as ‘woman-authored’; many of the articles that we ended up counting also had male authors. This way of counting articles likely caused an overestimation of the proportion of women authored publications in the journals we considered. There is also evidence that our estimation of the number of women working in ethics is on the low side: Schwitzgebel and Jennings estimated the percentage of women specializing in value theory to be 34% (Schwitzgebel and Jennings, Forthcoming). Lastly, it seems safe to assume that, given the effects of prestige bias, women in the top 50 philosophy departments are more likely than those from non-top 50 departments to publish in prestigious ethics journals.
But Why?
Our main goal in this study was to determine whether or not there is a discrepancy between the proportion of woman ethicists and the number of woman-authored articles being published in prestigious ethics journals. We found that there is. While our study doesn’t have anything to say about the cause of the discrepancy, we include in our paper a few suggestions for further research that might help to eventually answer the causal questions:
- Journals’ Definition of Ethics. While we used a very broad definition of ethics in estimating the number of women specializing in the field, it might be the case that prestigious journals use a more narrow definition of ‘ethics’. This is particularly concerning if the way that journals are defining ‘ethics’ discounts the subspecialties in which women are more likely to work. For example, do prestigious ethics journals view feminist social criticism and feminist political philosophy as falling outside of the scope of ethics? Such an effect has been noted in others fields of philosophy (Rooney, 2011) . If something like this is in fact contributing to the gender problem in ethics publishing, it would be important to look at this more closely.
- Professional Status. Are women in continuing positions more likely to publish in prestigious ethics journals? If so, women not holding such positions might face structural barriers to publishing in prestigious journals. Also, while we did not differentiate between assistant, associate, and full professors in determining the number of women who publish in ethics journals, there is evidence that woman philosophers are clustered in the assistant and associate professor ranks, and are underrepresented among full professors (Schwitzgebel and Jennings, Forthcoming). The four journals we studied do have anonymous peer review processes and so, in theory at least, professional status shouldn’t have any effect on the likelihood that someone gets their work published. However, as both editors and reviewers of journals articles are well aware, there are numerous ways in which the identity of authors might be revealed in the review process. This should also be looked at more closely.
- Types of articles. While our study did not differentiate between article types, it will be important to establish whether women are more likely than men to publish author-invited articles.
- Submission rate. Are women less likely than men to submit articles for peer review? Establishing the existence of this sort of phenomenon would be very difficult, especially given that the journals we surveyed do not collect submission data, but it is nevertheless an important question to ask.
References:
Haslanger, Sally. 2009. Preliminary report of the survey on publishing in philosophy. Presented at the American Philosophical Association Committee on the Status of Women in the Profession Session, Eastern APA, December 2009. http://www.mit.edu/%7Eshaslang/papers/HaslangerPRSPP.pdf
Rooney, Phyllis. 2011. The marginalization of feminist epistemology and what that reveals about epistemology “proper.” In Feminist epistemology and philosophy of science: Power in knowledge, ed. Heidi Grasswick. New York: Springer.
Schwitzgebel, Eric, and Carolyn Dicey Jennings. 2017. Women in philosophy: Quantitative analyses of specialization, prevalence, visibility, and generational change. Public Affairs Quarterly 31: 83-105.
Art: Laurie Frick, “7 Days of a Man Age 25” (2015)
I fear I must be missing something. Isn’t the significant number to compare to the proportion of papers by women in top ethics journals the percentage of ethicists who are women rather than “the percentage of women who specialize in ethics”? If there were only one women in philosophy and she was an ethicist, the percentage of women who specialize in ethics would be 100% and she could publish half of the papers in ethics and still be underrepresented by these criteria because they don’t publish 100% of the papers in ethics. That is, unless your grammar is confusing me and that’s what was meant.
“On a weighted average over the period surveyed, in a given year there are roughly 100 women and 266 men who specialize in ethics.”
That’s about 27%, the same figure used for comparison with the “the mean proportion of women-authored articles in prestigious ethics journals” (22.6%).
Thank you for the grammar correction, William Bell. We do mean percentage of ethicists who are women. And those are the numbers reported in Table 1 and Table 3. (I also just noticed that Table 3 did not print correctly; the scope of ‘% of women in philosophy’ should go over ‘All’ and ‘In Ethics’.) Hope that clarifies what we mean.
The figure that Jonathan quoted is correct. (The small discrepancy is in the rounding of ‘roughly’.) But, as Table 3 shows, what we did in this study was compare the percentages on an yearly basis. Just want to emphasize that.
While I agree that there is an underrepresentation problem in philosophy, and no doubt this spills over to every AOS, this approach makes things a bit messy. An easy indicator is that there are just a lot fewer women faculty, a lot fewer women graduate students, and a lot fewer women philosophy majors. This is something that we as a discipline should work to improve.
When we look at something like differential publication in high-end journals, there end up being a bunch of potential confounds. The most obvious is that this study didn’t look at the rate of submitted papers versus for either gender versus the rate of accepted papers, but the discussion pre-supposes that the rate of submitted papers is equivalent. This may not be true, due to different publishing strategies (that may or may not be affected by biases related to gender). The other big thing (that I take to be a general problem in our discipline) is that the acceptance rates, especially for PPA and Ethics, are miniscule, and so almost assuredly Type II errors are going to be huge relative to the acceptance rate. While no doubt the editors and referees do their very best to avoid this, it’s not inconceivable that these errors have a directionality to them. It wouldn’t be shocking if it was some sort of prestige bias (whether by person, by institution, or by topic). So, it may well be the case that those men that do disproportionally publish in these top places aren’t a random sample of men in the profession, but instead those that are already highly placed in the discipline. Again – this shouldn’t be that much of a surprise – one would expect highly regarded people to publish in highly regarded places. So, one would get the gender bias in publication as a side effect of a prestige bias in publication, combined with the disproportionate number of men in prestigious places (and with prestigious cv’s).
None of this is to deny that there are very real underrepresentation problems for women (and minorities) in the discipline. Nor is this to suggest that there aren’t issues with publication. But this does suggest that it may be that there’s just a general problem that we’re asking way too much out of journals – they can’t both be conferers of prestige and effective disseminators of useful ideas, especially in a discipline that has many different approaches and areas, even within sub-topics. If journals like Ethics and PPA published more papers in a given year, and acceptance rates ticked up to, say, 10%, Type II errors might not be as large of a factor, and we might see a broader array of publications (by field and author). Right now, the filtering is so high, it’s going to have a conservative bias on a bunch of dimensions.
Thanks Ryan. Speaking just for myself and not my co-authors, I don’t think there’s one unique way to think about underrepresentation. So we try to be super explicit that we’re thinking about underrepresentation with respect to faculty ratio, rather than underrepresentation with respect to submission rate.
Indeed, we posit that differential submission rate is one explanation of the phenomenon we’re interested in. If, for example, there’s differential submission rate due to self-selection, it might still be worth asking why that is. (The article makes some conjectures about this under ‘fourth’ potential explanation, p. 10-11.)
FYI people interested in submission rate by gender can see http://jonathanweisberg.org/post/A%20Look%20at%20the%20APA-BPA%20Data/ by Jonathan Weisberg, from which no easy conclusion can be reached.
” Lastly, it seems safe to assume that, given the effects of prestige bias, women in the top 50 philosophy departments are more likely than those from non-top 50 departments to publish in prestigious ethics journals.”
Maybe I’m misreading, but shouldn’t blind review defeat any effects from prestige bias? (I’m assuming you mean the prestige bias that attends membership in a high PGR department, rather than the prestige that attends publication in a venue like Ethics.)
Shen-yi,
You seem to be responding to queries, so I will direct this question to you.
Would you mind explaining why you think that it is still appropriate to conduct these sorts of studies of underrepresentation in philosophy that invoke a group “women” as if all the philosophers who identify in that way are similarly situated, that is, as if the group is internally homogeneous?
Disabled women philosophers make up an estimated 1% of the profession. So who do these figures that you have presented really refer to?
Thank you Shelley. Again just speaking for myself, I don’t think we assume anywhere that all women are similarly situated. Insofar as we’re doing something like sociology of academia, we are conducting analyses at the group level, but that does not presuppose the group is homogeneous. I don’t, by the same token, take that your expression ‘disabled women philosophers’ implies a homogeneity in that group, despite other situated differences such as race, class, employment status, etc.
I’d be happy to agree that the conclusions we draw, given that it only examines one dimension of oppression, should be read as preliminary and open to further moderation via examining other oppressive dimensions.
I would have liked to know about impact of the work produced by women philosophers. Some years ago, an analysis similar to this was done for women psychologists, and the results indicated that although women published fewer papers, the papers they published had greater impact. One implication was that women are more likely to take their time in developing their work so that it yields meaningful contributions to their fields whereas men are more likely to focus on amassing quantity, perhaps at the cost of quality. Unfortunately, the latter strategy seems to impress tenure and promotion committees (as well as conference organizers).
Does anything change if you just choose the last five years of data, 2010-2014? It seems to me that things have gotten better even in the short time I’ve been in the profession but that’s not backed up by any actual knowledge, so it’d be interesting to see whether the second half of data is any different than the first.
Thanks for this. I think it’s really valuable work. At the same time, I’m wondering if you can say more about why you used only faculty from the PGR Top 50 as the guide to estimate the percentage of ethicists who are women. Off the top of my head, I can think of a few reasons to hesitate about this: (1) the gender ratios may be different at those 50 schools than more widely in the profession (or is that wrong?); (2) ethics jobs may be distributed differently (say, more frequently than LEMM jobs) outside those 50 schools; (3) at least some of these four journals seem to commonly publish authors outside of those 50 schools (maybe PPA is an exception?)–in fact, it would not be unreasonable to speculate that faculty at those 50 schools receive more invitations to publish elsewhere, and so are less likely to publish in those four venues (or indeed in any unsolicited venues, even if when they do submit unsolicited work, they are more likely to go for top journals), making them a potentially non-representative subset of the profession. So I guess (3) calls into doubt this: “Lastly, it seems safe to assume that, given the effects of prestige bias, women in the top 50 philosophy departments are more likely than those from non-top 50 departments to publish in prestigious ethics journals.” There are a lot of people publishing these days, from a very wide variety of institutions (including those outside the US, sub-top-50 PhD programs in the US, and non-PhD-granting programs). I’m not so sure that assumption is safe.
Again, I really think this is important news you all are sharing, so thanks again! I hope this is taken in the spirit of understanding the research better, rather than as an attempt to pile on.
“The four journals we studied do have anonymous peer review processes and so, in theory at least, professional status shouldn’t have any effect on the likelihood that someone gets their work published.”
This seems mistaken to me. Surely we’d expect a full Professor to publish more and be better at writing publishable work than an early career assistant professor. We’d also expect full professors to be older and have entered the workforce earlier, and so to be have higher male representation. If women make up a lower proportion of full professors then it seems like this should be controlled for before concluding women are under represented in these journals.
An obvious hypothesis seems to be missing: the reason that female ethicists are better-represented in departments than in journals is that there is more gender-based affirmative action in departmental appointments than in journal publication. Many departments have a policy that when faced with male and female candidates who are otherwise equally qualified with respect to publications and so on, they should appoint a woman. Journals don’t have a corresponding policy. That departmental policy seems the right one to me, and the assumption that departments actually follow such a policy provides a potential explanation of the data. I don’t know whether this hypothesis is the correct explanation, but it should be on the table.
Even more pertinent: it’s well known that men who are successful on the job market have more publications than women who are successful, especially in recent years. Moreover, if I remember the data correctly, most hyper-productive publishers are men. In other words, the job market seems to select for men who are productive publishers, and this selection pressure is stronger than it is for women.* Hence, it would not be surprising if there is under-representation of women in journals, nor would it be surprising if the under-representation actually increases in the years ahead.
*This doesn’t mean that the men who get jobs are “better philosophers” than the women who get jobs; however, it does mean you can expect men junior scholars to publish more.
It would seem that, by your own criteria, only 1 of the 4 journals you investigated showed any under-representation of women. So it would seem to be a mistake to offer broad-brush causal explanations that concern ethics journals in general when, if anything, what you’ve identified is a problem specifically with the journal Ethics.
Moreover, Ethics clearly had a much more even gender ratio in 2014 than it did in 2004, so averaging over that whole time period may be misleading. Is the discrepancy for Ethics still statistically significant if you use a shorter and more recent time span?
More generally, what happens if you use other time spans for all the journals? What happens if you include more or different ethics journals? What happens if you include ethics papers published in generalist journals?
These are all sensible points. These are all reasonable questions to ask. We are really intentional in shying away from broad-brush causal explanations. Rather, we want to adduce some evidence for a phenomenon, but of course this evidence might be undercut or bolstered by future investigations.
Let me highlight a few paragraphs from the discussion section of our article, in case people have a hard time accessing it.
“We also explored whether underrepresentation appeared in any particular prestigious ethics journal. We did find that underrepresentation occurred in Ethics, which is perhaps the most prestigious of the journals we surveyed. But we did not find that underrepresentation occurred in Philosophy & Public Affairs, Journal of Political Philosophy, or Journal of Moral Philosophy.
Two familiar adages about statistical inference are worth emphasizing in this con- text. First, a statistically nonsignificant result is not itself evidence for the null hypothesis (Hoenig and Heisey 2001). So, in this context, one should not interpret the nonsignificant results as indicating appropriate gender representation in Philosophy & Public Affairs, Journal of Political Philosophy, or Journal of Moral Philosophy. Second, the difference between statistically significant and nonsignificant may not be statistically significant: in comparing various results, it is a mistake to do so via their statistical significance versus nonsignificance as if there were a sharp difference between the two (Gelman and Stern 2006). So, in this context, though the underrepresentation was statistically significant for Ethics but not for Philosophy & Public Affairs and Journal of Moral Philosophy, there is no statistical difference among the three journals. However, there is a statistically significant difference between the proportions of women-authored articles in Ethics and in Journal of Political Philosophy.
We wanted to present these exploratory analyses to acknowledge the complexity of this phenomenon. Honest examinations of real-world phenomena rarely offer a cut-and-dried picture, especially given standard concerns about variation and sampling. Given how few articles each journal publishes per year, it is to be expected that there is considerable variation from year to year within any given journal. It is, for this reason, more difficult to draw any firm conclusions on the basis of the disaggregated data about particular journals. Nevertheless, we do want to emphasize the central finding: women are undoubtedly underrepresented in prestigious ethics journals as a whole.”
Further to Shen-Yi’s reply – it’s easy to come up with a hypothetical data set where *no* journal is statistically different from the profession average, but the set of journals together *is*.
Imagine a scenario where 1/3 of ethicists are women, and there are four journals that each publish 10 papers, 2 of which are by women. If I’m calculating correctly, it looks like 2 out of 10 papers is not significantly different from 1/3 (the only way to be more representative would be to have exactly 3 or 4 papers by women, so assuming equal chances, there’s a 15% chance of a journal having exactly 2, or some other more non-representative number), while 8 out of 40 papers *is* significantly different from 1/3 (anywhere between 9 and 17 papers by women would be closer to 1/3 than 8 is, and there’s only a 2% chance of getting a number so far from 1/3 by pure chance). In that case, we could say there’s probably some systematic bias, but we can’t say whether all four journals share it, because it could easily be systematic bias in some journals while the other journal just got unlucky.
It would be interesting to conduct a similar study of the representation of women in edited collections of papers on ethics from top publishers. I have been party to two conversations with journal editors who have reported that, having heard really good conference presentations, they’ve encouraged the women who presented the work to submit papers to their journal. In reply, they’ve been told that the work is already promised for an edited collection.
That’s utterly anecdotal evidence, of course. But anyone who’s attuned to Philosophy’s gender problem and who is putting together a collection of papers will seek to have a good mix of contributors. So it is certainly within the realm of possibility that under-representation of women in one publication type (the refereed journal) might be partly explained by their improved representation in another publication type (the edited collection). Perhaps this should be the follow-up study to the present one.
Even if this were so, it certainly does nothing to diminish the fact that our discipline suffers from under-representation of women (as well as being depressingly homogeneous along other dimensions as well). But it would be mildly encouraging if the results of this particular study were partly explained by efforts to increase women’s prominence in other publication types.
It seems quite plausible that the bias here is mediated by prestige, or institutional support, or other factors that might be related to the question of whether the top 50 departments is really the representative pool for authorship in these four journals. For instance, someone might think that people employed at particularly prominent universities might just publish more in the top 4 journals, and that this effect exists even *within* the 50 universities you look at (maybe people at the top 10 institutions publish in these journals ever more than people at the next 40?) Or someone might think that certain universities exert pressure on their faculty to publish in prestigious peer-reviewed journals while others instead expect more teaching and administrative service, or publication in other venues.
One test you might already have the data to perform would imagine a hypothetical version of these four journals with the same institutional distribution of authorship, but representative gender distribution. The idea would be that for every paper published in these four journals, instead of looking at the *actual* gender of the authors, attribute gender based on the independently-collected gender distribution of the faculty at that institution (or perhaps just the ethicists). If this distribution is also significantly different from the observed actual gender distribution, this would be clear evidence that there’s something more going on than just what’s above.