Edited by:
Reviewed by:
*Correspondence:
This article was submitted to Developmental Psychology, a section of the journal Frontiers in Psychology
This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) or licensor are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.
That philosophy is an outlier in the humanities when it comes to the underrepresentation of women has been the occasion for much discussion about possible effects of subtle forms of prejudice, including implicit bias and stereotype threat. While these ideas have become familiar to the philosophical community, there has only recently been a surge of interest in acquiring field-specific data. This paper adds to quantitative findings bearing on hypotheses about the effects of unconscious prejudice on two important stages along career pathways: tenure-track hiring and early career publishing.
That philosophy is an outlier in the humanities when it comes to the underrepresentation of women has been the occasion for a lot of discussion about possible effects of subtle forms of prejudice, including implicit bias and stereotype threat. Though real-world effects are not strongly evidenced (
These ideas have become familiar to the philosophical community, which continues to debate policy initiatives and other measures for improving diversity, such as making syllabi and conference line-ups more inclusive, adjusting the management of professional organizations, and reforming journal and hiring practices. These ongoing discussions need to be informed by the best possible evidence, and there is a growing interest in acquiring field-specific data. The investigatory model informing this study is inspired by the hiring audits used in STEM disciplines. This paper contributes data pertinent to hypotheses about the effects of prejudice on two important stages of career pathways: tenure-track hiring and early career publishing.
If women are evaluated more harshly because of unconscious bias on the part of letter writers and hiring committees, or have weaker files and perform less well in interviews because of stereotype threat, or even face conscious and explicit discrimination, then they might be expected to be less successful at finding tenure-track employment. Indeed, biases are often conjectured to be a major cause of the underrepresentation of women in philosophy.
Analysis of Jennings’ original data suggests women and men are hired at a rate roughly proportionate to their numbers for entry-level tenure-track jobs in philosophy.
Another noteworthy finding obtained from Jennings’ earlier results is that female candidates had about half as many publications as their male counterparts. The average publication counts for candidates (having no prior academic appointment) were 1.37 for men and.77 for women (medians 1 and 0;
Gender and publishing.
It is natural to wonder how important publications are when it comes to assessing job candidates, and certainly we can agree publications are not the only relevant factor. Even for a research-oriented position the quality of writing samples, the reputation of doctoral institutions, and the weight assigned to letters of recommendation will also be taken into account. As a rough proxy for reputational factors rankings of degree-granting programs obtained from the 2006–2008 edition of the Gourmet Report were utilized.
Some have claimed that prestige interacts with gender in that women from highly regarded programs tend to publish the least, whereas men from less fancy programs publish the most.
First, mean Gourmet Report scores were incorporated within Jennings’ spreadsheet revealing a disparity in average home department rankings of 3.31 for men and 2.93 for women (medians were 3.6 and 3.2). High prestige “top-20” departments have a score of at least 3.4, and so next men and women were divided into elite (“top-20”) and non-elite (“non-top-20”) subgroups to see if there would be any interesting effects.
Gourmet-ranked TT positions.
Next, turn to consider how prestige might interact with gender when it comes to publishing. As mentioned earlier, differences depended on whether candidates had a prior appointment. In considering those with no prior appointment, it was found that top-20 men stand out: they publish more, and in “better” places than the others. Meanwhile, top-20 women publish less often than non-top-20 men and women, nevertheless they tend to do better when it comes to quality (
Publishing (no prior position).
Prestige and publishing.
Now we can consider the candidates who did have a prior position. Here, men had significantly higher averages for quantity and “quality” (
Publishing (with prior position).
Returning to an earlier suggestion, might it be the case that publications are not that important in hiring? This is hard to accept given that productivity is so often tied to securing research-intensive positions in the competitive academic environment and critical to determinations about prospects for earning tenure. This seems clear when we consider lateral hires, which constitute much of the data, and as just mentioned indicate upward trends in output. That candidates from less fancy programs publish more regardless of gender also suggests a widespread presumption that publishing compensates for other deficiencies. We can also note that previous research indicates that publication records are a critical indicator of candidate strength (
In a blog comment, Jennings
Nevertheless, all HP hires were examined in order to see what proportion of their work might be attributable to enhanced opportunities.
Returning to market outcomes, the previous results were augmented by placement data obtained from two additional sources: the American Philosophical Association’s Guide to Graduate Programs
Data was transcribed about gender and hiring found in the APA’s 2013 and 2014 Guides to Graduate Programs for two 5-year periods: 2008–2013 and 2009–2014. Only programs that allowed for a comparison between hiring outcomes and how many men and women went to market were included in the calculations. The Guides provided data for 64 schools in the 2013 edition and 65 for 2014 (37 schools provided data twice, so there is placement information available for 92 distinct programs for these mostly overlapping timespans). For 2008–2013 it was found that 40% of men who went on the market eventually landed a tenure-track job compared to 50.6% of the women, meaning a woman’s probability of obtaining tenure-track employment was about 25% better (
One might wonder if schools with good placement records, and, especially, good records for placing women might be overrepresented in the APA data. However, this concern is not realistic. Consider what it would take for the schools where we didn’t have data to close the gender gap. According to the APA, the 2008–2013 period comprised 530 junior tenure-track placements, and yet we would have to suppose an additional 200 competitions went unreported in which a man won every single time—in that case the chances equalize to 50%. There would have to be more than 500 unreported tenure-earning jobs going solely to men for the disproportion to be reversed (i.e., for men to have a 25% greater chance). In such a small profession, there are probably not enough unreported jobs for this to be the case: while Philjobs reported 816 junior tenure-track placements for the same period, many of these are lateral moves that placement officers would not normally pass on to the APA.
Adding to uncertainty about possible unreported hires, one might also wonder if these results would hold up for periods other than 2009–2014, and what the year-to-year results look like. With these concerns in mind we can turn to data provided by Philjobs. This process began by examining the 2014 hiring season, which was arbitrarily defined as spanning July 1, 2014 to June 30, 2015. Over the course of the year information was gathered about individual tenure-track hires, including those who had a previous academic appointment as well as those going to market straight from graduate school. For 2014 it was found that 56 out of 148 hires (37.8%) went to women. While the number of doctorates awarded to women as a percentage of the total doctorates in philosophy fluctuates somewhat from 1 year to the next, it was assumed that the year immediately prior would give a reasonable approximation of how gender is distributed on the job market; in 2013, for example, 27% were awarded to women.
To add more depth to the investigation hiring outcomes for nine further years (2005–2013) were also examined using data provided by the Philjobs website. In order to make this information useable certain corrections and additions to their spreadsheet were necessary, including the elimination of duplicated entries, filtering out senior appointments and non-tenure-track hires, spot-checking for accuracy, and using Google searches to ascertain gender where it was missing or in doubt. Next, year-by-year comparisons were made between placement and the distribution of philosophy PhDs using the NSF’s Survey of Earned Doctorates. The relationship between awarded PhDs and junior hiring from 2005 to 2014 is depicted in
Earned PhDs and TT hiring.
Earned Doctorates by gender.
Gender and tenure-track hiring.
Finally, these results were compared to updates found in the 2016 APDA report. First, my list of successful job candidates for the 2012 season was merged with the APDA’s. Although these mostly overlapped, there were some differences. In order to seek greater accuracy every candidate was re-checked, one-by-one, in attempts to verify gender and success in a tenure-track competition in 2012 (e.g., by consulting cvs, locating welcome messages at hiring Departments, etc.). Both data sets contained errors resulting in 56 changes to my list (37 additions and 19 deletions) and 36 changes to the APDA’s (29 additions and 7 deletions), thus bringing the two into harmony.
Market outcomes starting in 2014 and going back 10 years offer no evidence women are at a disadvantage in tenure-track competitions. The same can be said for the other objective measures that were examined including publishing and the reputations of home and hiring departments. No statistically significant evidence that pervasive dysfunction in departmental cultures is harming early career market outcomes of budding women philosophers could be found. Meanwhile, the biggest drop in women’s participation appears to occur almost immediately, right after first exposure to philosophy’s themes, methods, and traditions (
All the same, we can be somewhat reticent to draw strong conclusions about the extent of philosophy’s climate problems, and it might be premature to say that there is no systemic anti-female prejudice. Bias that was present but somehow neutralized by measures departments have taken or coping strategies adopted by women might have been overlooked. Then again it seems doubtful that explicit policy changes and coping strategies were adopted more than 10 years ago, long before there was wider awareness of the issue of unconscious bias. It is also conceivable that bias shows up elsewhere, affecting outcomes for tenure and promotion, though keep in mind this conjecture is not supported even by mainstays of the implicit bias literature, such as
While counter thoughts are not to be dismissed lightly, the hypothesis that unconscious bias works against women in hiring and early career publishing is not well supported. Although it is conceivable implicit bias initially reduces perception of a woman’s cv. and then “affirmative actors” reverse its impact, this proposal strikes one as overly complicated: why not just assume people are not downgrading the accomplishments of talented women?
The suggestion that there is a shyness effect making bias hard to detect is also hard to square with the evidence about pre-market publishing opportunities. Why doesn’t bias reveal itself in disparities for special invitations to publish where there are no equity policies or structures, little to no collegial oversight, and it is hard to conceive of coping strategies? We should also worry that efforts to improve the representation of women could even backfire, e.g., if committees adopted blind review of candidates under the dubious assumption that more accomplished women are systematically undervalued.
The author confirms being the sole contributor of this work and approved it for publication.
The author declares that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
The author is indebted to the referees, Calvin K. Lai, Shen-yi Liao, Paul Draper, Neb Kujundzic, Matthew Mosdell, Monika Piotrowska, and Louis-Philippe Hodgson, as well as audiences at the Atlantic Region Philosophers Association, the University of Louisville, and the University of Miami for their insightful feedback and suggestions.
Raw data is available at
“Implicit bias and stereotype threat…will make it harder for women to do well…to be recognized in graduate school, less likely to get strong letters of recommendation, and less likely to be hired. The women who, despite this, get hired at strong research departments are likely to be especially exceptional philosophers” (
The Academic Placement and Data Analysis project’s results can be found here:
Here and there minor corrections were made for errors, such as duplicate entries.
All of our raw data can be accessed at
See p. 11 of the 2015 report:
Meaning a 99.9192% confidence that the result is not due to chance.
It was reasoned that the reputational ranking should reflect the fact that candidates take around 5 or 6 years to obtain their degrees. But the results do not differ much if a slightly more recent or less recent edition is used instead.
See
Note that candidates from top-20 schools appear to fill half of new tenure-line positions.
Women seem to be slightly more likely to obtain degrees from unranked programs, which were scored as a “0.” Hence, if we ignore unranked programs these small differences in the overall rankings disappear.
Against this the reader is asked to compare the data in
Though we note that enthusiasm for stereotype threat theory is in rapid decline thanks to concerns about ecological validity, experimental design and interpretation, replicability, and evidence of publication bias, see
This finding of a gender gap in publishing is in step with
HP candidates were very likely to have had at least one prior position (median for both genders = 1; means were 0.88 for men and 0.86 for women).
In this case, we also included 10 highly productive individuals with post-doctorates in order to increase the population.
We didn’t count book reviews as peer-reviewed publications. If we do, the proportion of work attributable to enhanced opportunities falls to 31%.
There are about a third as many as one would expect given that women account for 32.2% of the hires.
Falling to 31% with book reviews.
Falling to 32.8% with book reviews.
Using Jennings’ data, we found that the highly productive men (with at least five publications) had an average of 1.98 publications in top journals. Women had an average of 1.14. The median was 1 in both cases.
Philjobs
Calculated using a Fisher Test.
Fisher Test.
See the 2013 edition of the NSF’s Survey of Earned Doctorates.
The final tally for 2012 was 140 men and 61 women. Note that there was only one instance where a gender was assigned incorrectly due to a limitation of the software used by the APDA project.