Welcome to our discussion of “The Past 110 Years: Historical Data on the Underrepresentation of Women in Philosophy Journals” by Nicole Hassoun, Sherri Conklin, Michael Nekrasov, and Jevin West, published in the most recent issue of Ethics. You can find the paper here. Eric Schwitzgebel’s critical précis is immediately below. Please join the discussion!!
Nicole Hassoun, Sherri Conklin, Michael Nekrasov, and Jevin West present data that should shock no one who has been paying attention to gender issues in mainstream Anglophone philosophy. Women constitute only a small percentage of authors of philosophy journal articles – about 19% of authorships in the most recently available decade. Although not surprising, Hassoun and collaborators’ work is novel for the depth with which it explores this issue.
Hassoun and collaborators examined authorship records for 56 philosophy journals in the JSTOR database from 1900 through 2009. They classified eighteen journals as “top” philosophy journals, based on online prestige rankings; twenty-two as “non-top” philosophy journals; and sixteen as “interdisciplinary”. It is plausible that this is a representative sample of journals. In particular, the “top” group includes most of the journals widely perceived as elite by the mainstream Anglophone philosophy community, such as Philosophical Review, Noûs, Mind, and Ethics.
Hassoun and collaborators classified authors into the binary categories of “woman” or “man” based on data from the top 1000 gendered baby names in the U.S. Social Security system from 1880 to 2010, excluding names commonly used for both boys and girls. Unfortunately, this procedure disregards nonbinary people, presumably misgenders some others, excludes authors publishing under first initials (who might disproportionately be women), and disproportionately excludes people from non-Anglophone backgrounds whose names are likely to be uncommon in the United States. Nonetheless, when dealing with very large datasets, it is impossible to wholly avoid such shortcomings.
In this manner, then, the gender distribution of 52,865 authorships was estimated for a large number of philosophy journals over the course of 110 years – an impressively large and valuable dataset.
Hassoun and collaborators report the following:
(1.) The proportion of women authors was only about 5% through the 1960s. From the 1970s to the 1990s, the proportion rose to about 19%, not materially increasing between the 1990s and the early 2000s. (See Appendix C, Figure C1.)
(2.) Through most of the period, the “top” philosophy journals had a lower proportion of women authors than the “non-top” and “interdisciplinary” journals.
(3.) In the 1990s and 2000s, the top journals had about 10% women authors, while the non-top and interdisciplinary journals had about 20-25% women authors. (See Figure 6.) By the early 2000s, women constituted about 23%-25% of faculty in the U.S. and other major Anglophone countries. Thus, their underrepresentation in top philosophy journals is striking even compared to their underrepresentation among faculty as a whole.
(4.) Based on estimated proportions of women faculty in various subareas in the U.S., women are especially disproportionately underrepresented in top general philosophy journals and top journals in value theory. (See Figure 9.) (Hassoun and collaborators acknowledge, however, that faculty subarea proportions might differ between the U.S. and the global population of authors publishing in English-language philosophy journals.)
(5.) Journals with anonymous review practices do not in general publish higher proportions of women. (See Figure 11.)
Why are women so underrepresented in elite philosophy journals? Hassoun and collaborators suggest that women might be more hesitant than men to submit to these journals. Hassoun and collaborators don’t speculate in detail on the possible sources of such hesitancy if it exists, but sources might include expectations of being held to higher standards, anticipation that their work will not be valued, a preference for different or more specialized journals, or less encouragement from colleagues or journal editors. Alternatively but not incompatibly, Hassoun and collaborators suggest that women might disproportionately work on topics (such as value theory or specific subareas in value theory) that are not seen as valuable by the editors and referees of elite journals.
I will raise two questions, then discuss implications.
One question concerns the extent to which we can determine whether these are field- or subfield-level effects versus journal-level effects. For example, the category of “top” value theory journals contains only two journals: Ethics and Philosophy and Public Affairs. Thus, the underrepresentation of women authors among top value theory journals might be due either to factors specific to one or both of those journals (such policies of one or a few particular editors), or to field-wide or subfield-wide factors, such as a tendency for mainstream Anglophone ethicists in general to privilege certain approaches to or subtopics within value theory. To the extent journal-level effects dominate, change might be effectively achieved through specific changes at a few individual journals. To the extent field- or subfield-effects dominate, change might require more general policy or attitude shifts.
Given the statistical approaches that Hassoun and collaborators use, it is plausible that some of the results are field-level effects not dependent on factors specific to individual journals (see for example Figure 1, which shows top journals disproportionately among those with lower proportions of women authors). For other analyses, such those of top value theory journals, there’s plainly no statistical means to separate subfield-level effects from journal-level effects. For still other analyses, it’s unclear how much the results are driven by one or a few individual journals. Multi-level statistical modeling or journal-level analyses might help clarify the issues.
Consider Hassoun and collaborators’ analyses of anonymous review policies. Only nine journals engage in triple-anonymous review and only six use non-anonymous review. The remainder, the large majority, use double-anonymous review. Without multi-level statistical analysis, it remains unclear whether the discouraging results regarding anonymity are a consequence of atypical results at one or a few journals in the triple-anonymous or non-anonymous categories, perhaps for reasons idiosyncratic to those individual journals. Because of this ambiguity in the results, I recommend caution regarding general conclusions about the value of anonymous review.
(Hassoun and collaborators acknowledge a version of this issue in footnote 70, where they discuss the fact that methodological choice to include Hypatia and Feminist Studies in the “history” subfield might be largely responsible for the relatively larger proportion of women authors in that subfield.)
My second question concerns the fact that the analysis runs only through 2009 (due to data availability). This is unfortunate, since one might naturally wonder whether the situation has changed in the past thirteen years. Hassoun and collaborators argue that it’s unlikely that the situation has changed much, since change in such matters is typically slow. It is perhaps notable that the nine authors of the other articles included in this particular issue of Ethics are all men, if I have gendered them correctly. (Please correct me and accept my apologies if I am mistaken.) Looking at the most recent issue of Philosophical Review, plausibly the most elite general philosophy journal, it appears that nine of the ten authors are men. This at least anecdotally suggests that Hassoun and collaborators are correct that the situation has not changed dramatically.
In general, the gender situation in academic philosophy has not changed much since the early 2000s. Nails and Davenport estimate that women were 26% of philosophy faculty in the U.S. in 2017. Beebee and Saul estimate that women were 30% of permanent academic philosophy staff in the U.K. in 2021. These numbers might be important increases over the approximately 23%-25% of the early 2000s (Beebee and Saul estimate a 6% increase over comparable data in 2011), but change has been slow at best. Looking at data from the National Science Foundation on philosophy PhDs earned in the United States, my collaborators and I also found remarkably little change in gender proportions since the 1990s, with women receiving on average about 27% of philosophy PhDs in the 1990s, the early 2000s, and the early 2010s, with perhaps a slight uptick to about 30% in the most recent available years.
The lack of gender diversity – but not only gender diversity – in philosophy is regrettable. Philosophy should be among the most diverse academic disciplines, not among the least diverse. Philosophy explores the most important general features of the human situation, challenging our presuppositions and critiquing culturally dominant understandings. It drives toward the roots of what we take for granted. To do this well, philosophy needs to welcome people from as wide as possible a range of backgrounds, especially people who might bring a new or different perspective. People from minority cultural backgrounds, people who have unusual life experiences at odds with the cultural mainstream, and people who have been systematically disadvantaged by mainstream society, should be disproportionately overrepresented in philosophy, not underrepresented.
What can we do to rectify the situation? Given their results, Hassoun and collaborators are not optimistic that anonymous refereeing alone will bring journal authorships closer to gender balance. More active and creative measures are required. To start, we need more knowledge: Hassoun and colleagues encourage journals to collect data on the diversity of their authors so that they can measure and track progress in improving diversity and knowledgeably evaluate the effectiveness of new policies. Appendix J contains a concrete list of specific proposals developed in consultation with a large number of editors, authors, and experts on diversity, which I hope we might discuss further.
In his helpful commentary Eric asks about the extent to which we can determine whether these are field- or subfield-level effects versus journal-level effects. I will leave it to our CS colleagues to talk about exactly what we can and cannot conclude from our data in some detail, but I think the answer is basically that we don’t always know – further analysis is necessary to tell us (though some of our results are more plausibly due to field level effects and others journal effects).
We might not need to answer all of these questions, however, to figure out how to improve diversity in the discipline. If the problem is just that a few (top-ranked ethics) journals have bad practices (e.g.) – then they need to change. The fact that a few journals (like the Proceedings of the Aristotelian Society) also seem to publish a lot of work by women – points to some obvious strategies for improvement (https://women-in-philosophy.org/data). The editors might, for instance, simply invite women to publish more in their journals.
Eric also asks if things might have changed since 2009 – he concludes that they probably have not, but this is also a question for further research. I hope experimental philosophers will pick up the torch and take the inquiry from here!
Ultimately, we agree with Eric that “people from minority cultural backgrounds, people who have unusual life experiences at odds with the cultural mainstream, and people who have been systematically disadvantaged by mainstream society, should be disproportionately overrepresented in philosophy, not underrepresented.” Wouldn’t philosophy be so much richer for their ideas? Why think that a small mostly rich white male European-descended English speaking section of humanity holds most of the world’s knowledge? Surely it cannot be so!
What can we do to rectify the situation?
One thing that we believe has helped improve the situation for women in the UK is the SWIP Good Practices schema. Having not only looked at all the data we could find but talked about best practices with editors of dozens of journals and heads of departments at the APA, we believe it is important to replicate that effort (https://women-in-philosophy.org/).
So, building off the little list we put together and more importantly all the great work on good practices done in the UK, Canada, and elsewhere, we have put together a very rough draft of promising strategies we believe departments should implement to advance all kinds of diversity in the discipline. We welcome comments both on what we should add but also on what might not really work https://docs.google.com/document/d/1tF8-bwOgJx-1voOLfm4_i1TGss6DOaGU/edit?usp=sharing&ouid=103967904913139590630&rtpof=true&sd=true. Our hope is to revise this in the coming weeks and eventually ask folks to take something along these lines to their departments, journals, research collaborations, and professional organizations to get sign on.
Those resistant to change sometimes act as if it is very difficult to improve diversity in academia, but we don’t buy it. There are many ways to increase diversity in the discipline and many promising possibilities are backed by significant experience from implementing departments and other organizations (if not the kind of evidence we present here). I have no doubt that those of us in leadership positions in our institutions can set targets for improvement, come up with plans to meet them, and implement those plans (and then revise our strategies and improve our performance if we do not succeed at first). Please join us in this effort!
The long arch of history only bends towards justice if we make it.
Fabulous work from all involved. This is really important and I’m glad to see it get the attention it deserves in PEA Soup. I found very plausible the authors’ claim that journal undersubmission from women (especially in the ‘top’ journals) could be a significant contributor, but they also add that less women co-authoring could be a factor as well (p 704).
My question is: Do we have any information about whether co-authored papers are more commonly accepted? I have no systematic knowledge, but I always felt that philosophers tacitly (or even overtly) look down on co-authorship in philosophy, so I would not have thought this would be a major concern, but I would be happy to learn if others know or feel that this looking down is correct–and if it inviting women to co-author could help address this disparity.
I’m not sure – Sherri or Jevin might be able to speak to this – I believe I have seen evidence that women get less credit for coauthoring with men. I suppose that means we should work together?
Its hard to tell whether co-authored papers are more commonly accepted, since we don’t have data on the proportion of co-authored papers submitted. Co-authored papers count for a very low proportion of publications in philosophy, which is only common in a handful of disciplines today. I suspect that disciplines with higher levels of co-authorships would also have larger proportions of women authorships because there would be more opportunities for publication. As it stands, its already very difficult to publish in our discipline, and it seems like work by men is prioritized (through some mechanism we don’t fully understand). If we had the same number of published articles but increased the total number of authors on each publication (via valuable contributions to the paper’s development), we might reasonably expect to see some of those numbers tick up. For example, Nicole and I are both collaborative authors. I would have no publications if I were not also working with Nicole. Anecdotally, collaborative co-authorships on our work alone resulted in a 9:4 ratio of women to men authors across 4 articles. This wouldn’t necessarily bypass issues relating the philosophy “boy’s club”, but it might be a start in the right direction.
Fascinating and important work!
A couple small points:
1. ‘Louis Anthony’ urgently needs to be corrected to ‘Louise Antony’ in the paper.
2. A lot of complexity is necessarily lost in a huge study like this. E.g. Analysis switched to triple anonymous review in 2008 or 2009 as I recall, so was not practicing triple anonymous for most of the years studied. It’s data then don’t actually tell us anything about the effects of triple anonymous review. Before then it was officially double anonymous but had very high rates of desk rejection and also of desk acceptance, so much of the decision making was non-anonymous. I find myself wondering how many journals’ stories that I don’t know harbour such complicating factors.
Thanks Jenny – and apologies to Louise! I’m not sure if the journal can correct errors at this point but we can reach out. As far as your second point, it is well taken. I believe we did try to track some historical data – Sherri has the details of that survey – however, tracking review status back to 1900 would be very difficult and I agree that journals often fail to do exactly what they say they do, even today. Some journals even ff the same “type” (e.g. double anonymous review) rely significantly on editorial discretion while others utilize referee reports to a much greater extent. I have seen some interesting work on “prestige effects” – where even some journals claiming strong anonymity seem to publish much more by authors with PhDs from certain institutions. And its clear, even on a cursory review, that some journals will accept papers by editors’ colleagues or students unlike almost anything else they publish (whatever they say about review processes). Consider how those at “elite” institutions have an incentive to maintain that work from those at “elite” instiuttions is of great quality in our economy of esteme – does this amount to the kind of conflict of interest against which there should be safegaurds? How can we realign these incentives to provide more opportunity for all? I suppose diversifying editorial boards and ensuring representation in editorial positions might be a good first step. Again, see the good practices guidelines for some other suggestions for improvement (and thanks for sharing your thoughts!)…
Hopefully it can at least be corrected in online versions– it’s an especially unfortunate error given the topic!
FWIW: I did actually get data from some “top” journals (big scare quotes– I hate rankings) on gender and submissions as well as acceptances. My hope was to compare the situation before and after switch to double and triple anonymous review. But the data folks I spoke to said the percentage of submissions by women was so tiny that they didn’t think we would be able to learn anything useful.
It is corrected! And yes we found the same thing – the data was too scarce to arrive at any reasonable conclusions. I hope that editors will make collecting this data into the future more of a priority.
Great that you could correct it! We used to be able to collect these data at Analysis, but the OUP started telling us we couldn’t due to privacy regulations. (This seems plainly false to me.) We are now on our way to finding a way around that, but’s taken years.