Content uploaded by Emil O. W. Kirkegaard
Author content
All content in this area was uploaded by Emil O. W. Kirkegaard on May 24, 2022
Content may be subject to copyright.
Submitted: 12th of July 2021 DOI: 10.26775/OP.2022.05.19
Published: 19th of May 2022 ISSN: 2597-324X
No Fair Sex in Academia: Evidence of Discrimination in
Hiring to Editorial Boards
George Francis∗Emil O. W. Kirkegaard †
OpenPsych
Abstract
The editorial boards of academic journals overrepresent men, even above their proportion in university faculties. We
test whether this sex disparity is caused by anti-female bias, supposing that anti-female discrimination means women
must have a higher research output than men to overcome bias against them. We collect a dataset of the research output
and sex of 4,319 academics on the editorials boards of 120 journals within four social science disciplines: Anthropology,
Psychology, Political Science and Economics. Using a transformation of the h-index as our indicator of research output, we
find male research output to be 0.35 standard deviations (p< 0.001) above female research output. However, the gap falls
to 0.13 standard deviations (p< 0.001) when years publishing is controlled for. Our results are replicated with alternative
dependent variables and using robust regression. We followed up our research with a survey of 231 academics, asking for
their attitudes towards discrimination in hiring to editorial boards. Although two-thirds of academics supported no bias,
for every 1 academic who supported discrimination in favour of men, 11 supported discrimination in favour of women. Our
results were consistent with the hypothesis that academics and journal editors are biased in favour of women, rather than
against women.
Keywords: gender, sex, discrimination, academia
1 Introduction
Academics have documented many sex disparities in their occupation that could be suggestive of pervasive
anti-female bias. Despite women being more than 50 % of undergraduates in many disciplines, they are less
likely to go into a career in academia (Ceci et al.,2014), they achieve lower pay and lower rank within academia
(Aiston,2014;Dunkin,1991;Ginther & Hayes,1999,2003;Ginther & Kahn,2004;Santos & Dang Van Phu,
2019), their papers are less likely to be cited (Abramo et al.,2009;D’Amico et al.,2011;Dion et al.,2018;Huang
et al.,2020;Maliniak et al.,2013;Bird,2011;Strumia,2021) and they are less likely to win academic awards
(Chan & Torgler,2020;Lincoln et al.,2012). Only 2 % of the individuals considered to be ‘eminent’ in science,
before 1950, are women (Murray,2003).
These disparities pose a key question: to what extent do sex biases or sex differences explain different outcomes?
Feminist scholars have argued that anecdotal reports of sexism in the lived experience of female academics
(Meyers,2013) and the fact of sex disparities themselves, suggests that academia is systemically sexist. On the
other hand, some academics have suggested psychological differences could explain sex disparities.
For example, female graduate students report being less interested in their careers than male students (Ferriman
et al.,2009), a sex difference that also increases with age. Part of this difference in careerism may be because
women have a greater interest in family and family commitments, being more likely to take time offfor parental
leave (Boston College Center For Work and Family,2019) and sick leave (Herr et al.,2020), which may have
adverse effects on 48 academic career outcomes (Ahmad,2017) and publications (Fox,2005).
∗Independent Researcher, Email: George.t.francis@protonmail.com
†Ulster Institute for Social Research, United Kingdom, Email: emil@emilkirkegaard.dk
1
Published: 19th of May 2022 OpenPsych
With regards to personality differences more generally, the only research we are aware of that attempts to
explain sex differences in academic outcomes with personality difference is (Helmreich et al.,1980). They
argued that in a sample of 196 academics, differences in motivation and masculinity/femininity could not
account for the differences in citations and publications, because there were no significant sex differences in
the personality traits. Big Five personality traits might explain sex differences in academic success. Women
score higher on extraversion, agreeableness and neuroticism (Weisberg et al.,2011). It has been speculated that
geniuses tend to be low in extraversion and agreeableness, but high in neuroticism (Dutton & Charlton,2016).
There is also the potential for intelligence differences to be driving outcomes. For example, Darwin (1871)
thought that the great success of men to achieve eminence in academic research (Murray,2003) could be
reflective of differences in intelligence. In meta-analyses (Lynn,1994,2017;Lynn & Irwing,2004), women tend
to have lower IQs than men. Furthermore, men also outperform women in general knowledge tests (Tran et al.,
2014) which may be particularly useful for academics who have to memorise and synthesis knowledge from
prior academic literature. However, the sex differences in intelligence are not clear cut; in children, boys do not
have an advantage in intelligence (Lynn,2017) and in some cognitive abilities, such as reading ability (Lynn &
Mikk,2009), women outperform men. Nonetheless, men have a higher variance in their intelligence (Baye &
Monseur,2016) which may cause more men to outperform women in intellectually elite occupations such as
academia (Nyborg,2005;O’Dea et al.,2018). For example, Baye & Monseur (2016) find the male variance in the
Programme for International Student Assessment tests is 1.17 times the female variance. If we assume aptitude
to be normally distributed, this implies that for the 98th percentile score in women, there are around 3 men for
every 1 women at or above this level of aptitude.
This paper seeks to examine whether hiring to editorial boards in academic journals is sex-biased. Many
previous studies on editorial boards show that they overrepresent male academics relative to their proportion
in university faculties (eg. Amrein et al. 2011;Cho et al. 2014;Mauleón et al. 2013;Metz & Harzing 2009,
2012;Morton & Sonnad 2007;Ioannidou & Rosiana 2015;Mazov & Gureev 2016;Morton & Sonnad 2007),
indicating hiring to editorial boards could be sex-biased. We contribute to this question by comparing the
academic output of men and women who are hired to editorial boards and through a survey of academics on
their attitudes towards women in academia.
The editors of journals hire academic experts, usually without pay, to sit on the editorial boards. Academics
sitting on editorial boards can perform three main tasks - advising on strategy for the journal, helping in
decisions on what to publish and improving the journal’s reputation through association (Wiley,2021). Some
longitudinal studies of editorial board membership show that whilst the proportion of women on editorial
boards is increasing, this is in parallel if not below the growth in the number of women in academia (Addis
& Villa,2003;Huang et al.,2020;Mauleón et al.,2013;Metz & Harzing,2012). These studies are focused
on certain niches such as journals from Spain or management journals. Nonetheless, if these studies are
generalisable, sex representation in editorial boards are not changing over time.
A sex bias in hiring to editorial boards, or anywhere else in academia, may be detrimental to the careers of those
being discriminated against and for the quality of scientific research as a whole. The Impact factor of journals
has been found to correlate with the research productivity of the members of its editorial board, although not
with its sex proportion (Hafeez et al.,2019). This means sex bias could undermine the quality of academic
journals. Not being allowed on an editorial board prevents discriminated individuals from this experience as
an academic, but it also might have knock-on effects on the careers of these discriminated individuals. Sitting
on an editorial board places an academic within a network of high-quality researchers whom you can exchange
ideas with or who can help each other in other ways.
A potential consequence of sex bias could be that it distorts scientific output. Addis & Villa (2003) suggest
that because the sexes differ in their academic interests, the proportion of women on an editorial board could
affect which articles are published. An example of sex differences in academic interest includes men preferring
’thing-oriented topics’ over ’people-oriented topics’ (Luoto,2020).
Due to concerns that women are being discriminated against, multiple publishers have asked their journal
editors to increase the proportion of women on their editorial boards. For example, Nature has been reviewing
the sex balance in its journals and asking that editors improve this balance since 2012 (Nature,2017). More
recently both the Lancet and Elsevier have been urging their editors to increase the representation of women on
their boards (Lundine et al.,2018;Bayazit,2020;Elsevier,2021a). To improve transparency, Elsevier publishes
the sex ratio for each of its journals, which may act as an incentive for editors to increase female representation
to be seen as more progressive or avoid reputation-damaging accusations of sexism (Elsevier,2021b).
2
Published: 19th of May 2022 OpenPsych
Attempts to employ affirmative for women on journal boards may be meritocratic if there is sex discrimination.
However, if there is no discrimination, affirmative action policies may counterproductive. Moreover, if affir-
mative action and sex bias support the same sex, then affirmative action may aggravate inequities. As such,
stronger evidence on whether sex bias exists is essential for judging whether affirmative action will improve
meritocracy.
Our first method for investigating the possibility of whether there is bias in hiring to editorial boards is to
compare the academic output of men and women who have been hired. A critical trait for being admitted
to an editorial board is academic expertise (Lindsey,1976) which may be measured as research output. All
other things being equal, if women are being discriminated against they would have to be more impressive
academically to compete with men.
It must be noted that a sex difference in the academic output of editorial board members can only be an
indicator, not proof of sex bias. As mentioned, the variance in intelligence is higher amongst males, and their
average also seems to be somewhat higher. This would cause men, on editorial boards, to have a higher academic
output even if there was no bias. Thus if women have a higher academic output, despite their lower variance
in IQ, we can be confident that there is anti-female bias. We can also say that the larger the sex difference in
favour of men, the lower the likelihood of anti-female bias and the higher the likelihood of anti-male bias. So if
men have a higher academic output than women we can be confident that there is no extreme anti-female bias.
Our reasoning comes from Gary Becker’s taste discrimination model of the labor market (Becker,1971). If an
employer has a distaste for one group of employees, but cannot provide them with a different wage rate, he will
only hire members of this group that are sufficiently extra productive to outweigh the cost of going against the
employer’s discriminatory tastes.
This same reasoning has been applied at least once before to editorial boards. Hafeez et al. (2019) found that
for Psychiatry journals, despite women publishing half as many papers as men, they served on journals with
the same mean impact factor. This result suggests women are not being discriminated against, when Psychiatry
journal boards hire, and may even imply that women are being favoured. The authors also found that when
women were in leadership positions the journal was less likely to include women on its editorial and advisory
boards. This should not be the case if male academics are more likely to discriminate against women. Hafeez et
al. (2019) also found that, despite women being underrepresented on journal boards relative to the proportion
of women in Psychiatry, they were represented in equal proportion to their level of seniority in academia. We
go beyond this prior paper by testing for sex differences in output, in editorial boards, in a wider range of
disciplines.
A similar test for sex bias in hiring was used by Madison & Fahlman (2021). The authors found that women had
fewer publications and citations upon becoming professors in Sweden. Likewise, Strumia (2021) found that
male physicists have a greater research output than women before being hired by a university. These results
suggest that women are unlikely to be discriminated against in hiring by universities or even a bias against
women, despite there being more male than female academics. Our paper thus applies the same logic to test
whether there might be sex bias in hiring to editorial boards.
However, other research of sex bias and hiring in academia have typically run experiments by asking faculty
members to judge the resumes are hypothetical hires. These studies have reported mixed results. Williams
& Ceci (2015) asked academics to evaluate hypothetical hires, who were identical except for sex. They found
on average university faculty preferred women to men at a 2:1 ratio. Carlsson et al. (2020), using similar
methods also found a preference for women. A follow-up study (Williams & Ceci,2015) found no preference
for women compared to better-qualified men. Quadlin (2018) also asked faculty to evaluate hypothetical hires
and found that amongst highly competent candidates with high GAs, men were preferred to women at a 2:1
ratio. Suggesting high academic achievement may be more valued in men than in women. Older studies (Foschi
et al.,1994;Steinpreis et al.,1999) focused on hiring to non-faculty positions, such as laboratory manager
(Moss-Racusin et al.,2012), and found results consistently in favour of male applicants. A caveat to these
resume studies is that sex differences in hiring may not be caused by prejudice, but by statistical discrimination.
In our test of whether editorial boards are sex-biased, we decide to use journals from the social science and
humanities. Firstly, women likely make up a higher proportion of academics in humanities than in STEM
disciplines, so getting a large sample with enough women may be easier when avoiding STEM disciplines.
Secondly, it has been argued that women prefer these less quantitative disciplines (Kahn & Ginther,2017), and
may have less aptitude for STEM disciplines (Reilly & Neumann,2013;Lord,1987). If this were true, the effect
of higher male performance would be more likely to obscure the effect of anti-female discrimination, making
3
Published: 19th of May 2022 OpenPsych
non-STEM disciplines more appropriate for our test. Whether or not women have less interest or aptitude for
STEM disciplines, we chose to study social sciences just in case this would bias our results. Thus although we
are concerned with sex bias in academia as a whole, our method only focuses on testing this hypothesis within
social science disciplines.
We thought it was also important to choose disciplines within a wide range of political persuasions in case
politics influences bias in hiring to editorial boards. Some research has suggested that right-wing individuals
exhibit an anti-female bias (Austin & Jackson,2019;Christopher & Mull,2006;Hodson et al.,2017). Other
research finds that left-wing individuals may be prone to bias towards groups with low status, including women
(Winegard et al.,2018). Overall this body of research indicates that as one moves politically right one becomes
less pro-female and more pro-male. Whilst a wide range of disciplines with a very large sample size would be
necessary to test whether politics did create biased hiring, having a range of disciplines allows us to be sure
that our results are not due to the political confounds of any particular discipline.
We chose four social science disciplines to study: Anthropology, Psychology, Political Science and Economics.
These disciplines vary widely in their political persuasions, with economics being the least left-wing and
Anthropology being the most left-wing (Langbert,2018). The ratio of Democrat to Republican faculty members
in each discipline is presented in Table 1below.
Table 1: Political Affiliation of University Faculty
Discipline Democrat - Republican Ratio in Faculty
Economics 5.5:1
Political Science 8.2:1
Psychology 16.8:1
Anthropology 133:1
Source: Langbert (2018)
There have been many studies on sex representation on editorial boards including in Anthropology (Bruna et
al.,2017), Psychology (Evans & Robinson,2005;Hafeez et al.,2019;Over,1981;Palser et al.,2021;Robinson
et al.,1998), Political Science (Fraga et al.,2011;Palmer et al.,2020) and Economics (Addis & Villa,2003;
Gibbons & Fish,1991;Mumford,2016). Anthropology, Psychology and Economic editorial boards tend to
slightly underrepresent women relative to the number of academic staffin these fields. This could suggest there
is anti-female bias in these journals’ boards.
However in Political Science (Fraga et al.,2011;Palmer et al.,2020), Economics (Mumford,2016) and Psychiatry
(Hafeez et al.,2019) editorial board sex proportions have been compared to the sex proportion amongst senior
academics, not just the totality of junior and senior staff. When this is done editorial boards have a similar sex
proportion to that of senior academics.
2 Data
To choose which journal’s editorial boards to study, we employed the website Scimagojr (SCImago Journal &
Country Rank,
https://www.scimagojr.com/
) which contains a dataset of 34,346 journals on their website
based on Scopus, Elsevier’s abstract and citations dataset. We ranked journals in each of the disciplines we
studied according to the number of citations per document they had in the previous two years. From this
ranking, we then took the top 30 journals from each discipline.
We disagreed with the discipline label of some of the journals on Scimagojr. For example, some of the ’Economics’
journals such as the ’Journal of management’ were deemed closer to Business Studies than Economics. Likewise,
’Politics’ journals such as the Journal of Political Economy’ typically only had economists as authors. Nonetheless,
the Journal of Political Economy was also classified as an Economics Journal by Scimagojr, a classification we
agreed with. Journals not obviously in the correct disciplines were ignored. In Table 9of appendix A, we
present a list of all 120 journals used in this study and their respective disciplines.
From the websites of the journals, we recorded members of their editorial boards. The term "editorial board’
had slightly different meanings for different journals. Some used the term to describe everyone working for the
journal. Most however used it to label a subsection of the editorial team involved in more advisory work. When
4
Published: 19th of May 2022 OpenPsych
there was no subsection of a journal’s stafflabelled the ’editorial board’ we took the relevant subsection that
seemed closest in meaning such as ’advisory board’. As such our methodology did not include journal chief
editors as part of the editorial board.
In line with the practice of previous research on sex representation on editorial boards, we coded the sex of
academics according to whether their names were clearly male or female (e.g. Ioannidou & Rosiana 2015).
When this was not obvious we used Google Search to find their sex from pictures or left the sex variable missing
when this was insufficient. Of the 5,625 editorial board members in our dataset, we were unable to determine
the sex of 7 individuals. To measure the productivity of academics on editorial boards we obtained relevant
statistics from their Google Scholar page when it was available. These statistics included the publication count,
h-Index, i10 Index, citation count, h-Index since 2016 and the citation count since 2016. Furthermore, to control
for years publishing in academia we also recorded the year of the researcher’s first publication. When the
researcher did not have a page on Google Scholar we left these statistics missing.
For ease of interpretation, our measures of academic output were first log
10
transformed and then then Z-
transformed into standard deviation units within each academic discipline. This allows us to compare the
relative performance of researchers across disciplines. For example, a transformed h-index of 1 means the
researcher’s h-index is one standard deviation above the mean of the editorial board’s members in the respective
discipline. Nonetheless, we also used raw data in the appendix.
All our data was collected between March and June 2021
1
. Although 5,625 editorial board members were
recorded, 7 individuals couldn’t be identified by sex and a further 1,098 individuals did not have Google Scholar
pages. Of the board members recorded 40 % were women, but 42 % of researchers without Google Scholar
pages were women, meaning women were slightly less likely to have a Google Scholar page.
Sometimes Google Scholar pages for individual academics contained errors in them. Some papers had the
wrong date on them and others were attributed to the wrong author. When a Google Scholar Page included five
or more articles with citations that the author had not written, we noted the page as overattributing research to
the author. We excluded these ’over-attributed individuals’. When the earliest paper on a Google Scholar page
appeared to be of the wrong date or by a different author we made use of the next earliest paper that appeared
to be correct.
Despite our attempt to remove academics with exaggerated publication metrics, some unusual results remained.
Some academics had higher hand i10 indexes for the period after 2016 compared to their all-time hand i10
Indexes. We removed 21 academics because they had higher indexes of academic output for the period since
2016 than they had over all-time Furthermore, some academics had very large academic outputs. For example,
one academic had 2,876 publications, possibly suggesting either errors with Google Scholar, plagiarism or that
they mostly relied on co-authors to write the papers. To deal with these extreme values we applied Tukey’s
Fences2to identify and remove positive outliers and removed 44 observations from the dataset.
In deleting observations our data cleaning approach loses information and degrees of freedom in our results
and thus may be critiqued. As such, we re-ran our main results, in Table 12 of Appendix B, without omitting
any observations for over-attribution, being outliers, or having inconsistent metrics post-2016 and for all-time.
After excluding observations we went from having 4,520 complete cases to 4,319 complete cases. This moved
the sample from being 39.4 % female to 40.2 % female. As such, in removing the academics with the greatest
publication metrics we were more likely to exclude men making our results slightly biased in finding a female
advantage in academic output. The descriptive statistics for this complete dataset are in Table 2.
In Table 3we present a correlation matrix of our recorded variables, with the dependent variables in their
raw and transformed versions. Notably, our measures of research output strongly correlate with each other.
This suggests that any of the dependent variables will work similarly well as a measure of research output.
For simplicity, we thus focus on the popularly used h-index. The h-index is the largest value of ’h’ for which
an author has published ’h’ articles with ’h’ citations each. The h-index has the advantage of being easy to
understand (Rørstad & Aksnes,2015) and having high external validity (Ruscio et al.,2012) in its association
with academic rank e.g. professor versus assistant professor. However, the differences between the indexes for a
researcher’s entire career versus just what they have done since 2016 may be related to sex, especially since
women have been increasingly joining academia.
1
In this time period, journal rankings by citations changed from the default year of 2019 to 2020. This can be verified with the
Internet Archive (Internet Archive,
https://web.archive.org/web/*/https://www.scimagojr.com/journalrank.php
). During data
gathering, this change was not noticed meaning journals were ranked by citations in different years depending upon when the data was
gathered.
2
Tukey’s Fences identifies positive outlier hindex values as equal or greater than the following Q
3
+ 1.5
×
(Q
3
-Q
1
), where Q
3
and Q
1
represent the third and first outlier respectively
5
Published: 19th of May 2022 OpenPsych
Table 2: Descriptive Statistics.
Statistic Mean
Standard
Deviation
Minimum
25th
Percentile
75th
Percentile
Max Skew Kurtosis
Years Pub-
lishing
24.2 11.1 2.0 16.0 31.0 70.0 0.6 2.8
h-Index 30.5 21.4 1.0 15.0 40.0 136.0 1.8 7.8
Transformed
h-Index
0.0 1.0 -4.1 -0.6 0.7 2.7 -0.2 3.1
h-Index
since 2016
23.4 14.5 0.0 13.0 30.0 96.0 1.8 8.5
h-Index
Since 2016
0.0 1.0 -5.6 -0.6 0.7 2.7 -0.3 3.5
i10 Index 56.6 59.9 0.0 18.0 71.0 504.0 3.8 26.7
Transformed
i 10 Index
0.0 1.0 -4.1 -0.7 0.7 2.8 -0.2 3.4
Publication
Count
128.8 132.4 1.0 45.0 163.0 1,151.0 6.0 57.1
Transformed
Publication
Count
0.0 1.0 -4,.2 -0.7 0.7 2.9 0.0 3.4
Citation
Count
8,406.1 13,415.8 1.0 1,382.0 9,356.0 159,016.0 4.7 35.3
Transformed
Citation
Count
0.0 1.0 -5.0 -0.6 0.7 2.6 -0.4 3.7
Citation
Count since
2016
3,895.1 5,427.3 0.0 861.0 4,626.5 58,699.0 5.9 64.0
Transformed
Citation
Count since
2016
0.0 1.0 -6.6 -0.6 0.7 2.7 -0.5 4.3
3 Results
To begin with we follow previous literature in simply comparing the sex proportions on editorial boards to
comparison samples. In Table 4we show the sex proportion in journal boards in each discipline. To see whether
these proportions are representative of the field they should be compared with the population of academic
researchers, be it for example faculty members or published researchers. We use the terms overrepresent and
underrepresent to denote whether the fraction of women on editorial boards in a discipline is greater or less
than female representation in the relevant population of academics who could be placed on editorial boards (ie.
active authors and university faculty members).
For comparison, we found a range of datasets representing the sex proportion amongst academics in the
disciplines studied here. Our first source of comparison is the sex proportion of active authors with at least two
publications during the years 2014-2018. The figures are provided for the USA and the EU28 (The European
Union plus the United Kingdom). These figures are reported by Elsevier (Kleijn et al.,2020) in their 2020
Gender Report and are derived from the Scopus dataset. Unfortunately this data does not have sex proportions
specifically for Anthropology or Political Science, so we use the proportions for the closest related discipline
groups ’Arts and Humanities’ and ’Social Sciences’. From the UK we have the sex proportions amongst academic
stafffrom the Higher Education Statistics Agency (2021). We use the proportions from 2016 because newer
staffmight be too early in their research career to get on a journal board. For economics, we also record the
proportion of published economists registered with the Research Papers in Economics Author Service as of
2021 (Research Papers in Economics Author Service,2021).
6
Published: 19th of May 2022 OpenPsych
Table 3: Correlation Matrix
Variable
Years Pub-
lishing
h-Index
Transformed
h-Index
h-Index
since 2016
Transformed
h-Index
Since
2016
i10 Index
Transformed
i10 Index
Publication
Count
Transformed
Publica-
tion
Count
Citation
Count
Transformed
Citation
Count
Citation
Count
since 2016
Transformed
Citation
Count
since 2016
Years
Publish-
ing
1
h-Index 0.62 1
Transformed
h-Index
0.65 0.88 1
h-Index
since 2016
0.58 0.96 0.86 1
Transformed
h-Index
Since 2016
0.65 0.85 0.97 0.89 1
i10 Index 0.6 0.94 0.79 0.87 0.74 1
Transformed
i10 Index
0.68 0.86 0.98 0.84 0.94 0.82 1
Publication
Count
0.5 0.81 0.71 0.73 0.66 0.89 0.76 1
Transformed
Publication
Count
0.63 0.78 0.86 0.74 0.81 0.76 0.89 0.84 1
Citation
Count
0.5 0.83 0.66 0.81 0.64 0.77 0.62 0.66 0.56 1
Transformed
Citation
Count
0.63 0.82 0.93 0.81 0.92 0.71 0.9 0.63 0.77 0.69 1
Citation
Count since
2016
0.41 0.82 0.66 0.85 0.68 0.75 0.62 0.63 0.55 0.95 0.7 1
Transformed
Citation
Count since
2016
0.51 0.78 0.9 0.82 0.93 0.68 0.87 0.59 0.72 0.67 0.97 0.72 1
7
Published: 19th of May 2022 OpenPsych
Table 4: Proportion female of editorial board members, active authors and university faculty.
Discipline
Sampled Edito-
rial Boards
Active Authors
(USA)
Active Authors
(EU28)
Academics in
UK Universi-
ties as of 2016
Registered au-
thors with the
Research Papers
in Economics
Author Service
Anthropology
49 %
43 % (Arts and
Humanities)
43 % (Arts and
Humanities)
51% N/A
Psychology 41 % 56 % 58 % 61 % N/A
Political
Science
39 %
47 % (Social
Science)
44 % (Social
Science)
37% N/A
Economics 28 % 24 % 34 % 30 % 26 %
Editorial boards in Anthropology, Political Science and Economics seem to be broadly representative of their
fields. Anthropology editorial boards are 49 % female which is close to the proportion of UK Anthropologists
who are female - 51 %. Although Anthropology has a greater percentage of women than active authors in the
Arts and Humanities these may not be an accurate match for the disciplines. Political Science overrepresented
women relative to their role in UK Universities but not compared to active authors in social science. Whether
this is because other Social Sciences have more women, or because the UK has an unusual lack of women in
their Political Science departments is unclear because the data reported by Elsevier (Kleijn et al.,2020) does not
give a sex breakdown for individual disciplines within the Social Sciences. Compared to every comparison, our
sample of Psychology editorial boards underrepresents women.
In previous research Anthropology underrepresented women (Bruna et al.,2017) but we find women propor-
tionally represented in editorial boards. Political Science (Fraga et al.,2011;Palmer et al.,2020) and Economics
(Mumford,2016) were only representative of senior academics, however in our sample here they appear broadly
representative of all academic staff. Only our results from Psychology (Evans & Robinson,2005;Hafeez et al.,
2019;Over,1981;Palser et al.,2021;Robinson et al.,1998) were in line with prior research suggesting women
are under-represented.
One possibility could be that publishers, at least in Anthropology, Politics and Economics, have been successful
in encouraging their journals to increase female representation in recent years. Nonetheless, whether these
proportions are meritocratic will depend on the research output of women. Assuming no underlying differences
in ability, if the sex disparities found here represent anti-female bias, women would need to substantially out-
perform men to get on Psychology editorial boards. Moreover, female research output should be approximately
equal to men’s in Anthropology, Political Science and Economics.
Our first method for testing whether women need a higher level of research productivity than men to get on
editorial boards is to simply compare research productivity between men and women on editorial boards. As
stated in the data section, our measures of research productivity are standardised relative to the mean research
productivity of academics in editorial boards of journals residing in the same discipline. This ensures that there
is no bias from differential sex interest in disciplines that may be associated with higher or lower absolute levels
of research productivity.
Before using regression to compare sex differences whilst using controls, we present the sex distributions of
research productivity by discipline in Figure 1. This is to create a clear visualisation of the results of our
study. Test results for Welch’s t-tests and their pvalues for the difference between male and female research
productivity are reported in Table 5.
In each discipline, men have a higher level of research productivity in terms of our transformed h-index. The
female disadvantage in research output is between 0.28 standard deviations below men in economics to 0.44
standard deviations below men in political science. Moreover, this difference is statistically significant in each
discipline (p< 0.001). It should be noted that despite including just as many journal boards in Economics as
we have included in Anthropology and Psychology, it has substantially fewer degrees of freedom because the
economics journals had fewer editorial board members.
Women are under-represented in psychology editorial boards, and yet the women who do manage to get on the
editorial boards dramatically underperform relative to the men that are on the board by 0.44 standard deviations.
8
Published: 19th of May 2022 OpenPsych
Figure 1: Distributions of Log10 then Z-Transformed h-Index of female and male editorial board members
Table 5: Sex Differences in log10 transformed h-Indexes of editorial board members.
Statistics
Discipline Mean Difference t value P value Degrees of Freedom
Anthropology 0.34 5.23 p <0.001 928.17
Psychology 0.31 6.12 p <0.001 1439.83
Political Science 0.44 6.48 p <0.001 757.80
Economics 0.28 4.10 p <0.001 535.46
9
Published: 19th of May 2022 OpenPsych
In other words, women are underrepresented on Psychology editorial boards relative to their proportion on
faculty but are still overrepresented relative to their merit. Likewise, women may be overrepresented relative to
their merit in Economics, Political Science and Anthropology. Despite women being proportionally represented
in these disciplines, male research output is still higher.
Also of note is that there is no monotonic relationship between sex differences in research output and how
right-wing a discipline’s faculty is (disciplines are ordered in Table 5from the most left-wing to least left-wing).
To properly test for any sex bias arising from political opinion between disciplines we would need to include
more disciplines.
We again analyse the differences between male and female research productivity now using ordinary least
squares regression. This has multiple advantages. Firstly, we can combine our samples from different disciplines,
using dummies to control for any discipline effect, giving us a larger sample size. Nonetheless, we also run
regressions for each discipline separately. Secondly, we can control for the number of years a researcher has
been publishing. More years in publishing allows an academic to increase their publication count and receive
additional citations for old articles, boosting metrics of research output. This means a brilliant academic
may have a lower h-index than a mediocre academic who has been publishing for longer. Thus a meritocratic
editorial board should take into account the length of an academic’s career when judging their research output.
Since men tend to have had longer careers in academia (Huang et al.,2020;Martinez et al.,2007) whilst women
are joining academia at greater rates we should control for the length of academics’ publishing years to see
whether women are held to a higher standard. On the other hand, time in academia is itself an indicator of
knowledge and experience which could help as a member of an editorial board. Time in academia is correlated
at 0.62 with the h-index in our sample. Thus controlling for years publishing could be partially controlling for
the variable we are trying to model - merit to be on a journal board. This possibility becomes more severe if
younger and less experienced scholars are less intelligent. Akcigit et al. (2020, p. w27862) have shown that
there are more academics today than before. The authors show that reduced selectiveness for joining academia
has reduced the IQ of the average PhD student. This is corroborated by the fact that scientists are becoming
less productive (Huang et al.,2020). Given arguments for and against this control variable, we decide to run
regressions with and without it.
Our regression models of the transformed h-index are presented in Table 6. Models using only sex as an
independent variable find women perform worse in terms of research output in each discipline (p< 0.001).
When we control for the years publishing we find it has a consistently positive effect (p< 0.001) on research
output regardless of what disciplines are studied. Every 10 years of experience in academic publishing is
associated with a research output increase of between 0.6-0.7 standard deviations. This is in accordance with
our expectation that academics with less experience tend to have a lower research output. Years publishing
moderates the effect size of sex in every discipline, more than halving sex’s effect size in every regression.
Without the years publishing control, men perform better than women between 0.28 and 0.51 standard
deviations, but with the control men only perform better by 0.1-0.21 standard deviations.
The moderating effect of years publishing is to be expected given sex and years in academia are confounded;
female academics tend to have less experience because they are becoming more represented in academia over
time (Miller & Wai,2015) and they are more likely to quit their academic career (Huang et al.,2020) . Thus
a partial cause of low female representation in editorial boards may be their lower levels of experience, as
evidenced by the fact that years publishing correlates with the h-index and it moderates the sex difference
in academic output. This result is consistent with the findings that female scholars, and particularly the
worst-performing female scholars (Rørstad & Aksnes,2015), are more likely to drop out of academia and thus,
presumably, editorial boards.
When we combine all the disciplines together in regression models 9-12 we find sex still has a statistically
significant effect on research output. In regressions 11 and 12 we use the interaction terms between discipline
and sex, indicating whether some disciplines significantly differ in their respective sex effects. In these
regressions, we find no statistically significant interaction terms. Log-likelihood ratio tests were used to judge
whether models 11 and 12 are superior to models 9 and 10. The chi-square values were insignificant so the
discipline sex interaction terms do not improve the models. Thus we cannot reject the null hypothesis of sex’s
effect being homogenous across disciplines.
To test whether our results are robust we ran the same set of regressions for alternative dependent variables
representing academic output. These variables were the non-transformed raw h-index, the h-index score since
2016, the publication count and citation count. We also reran our regressions without cleaning our data, to see
10
Published: 19th of May 2022 OpenPsych
Table 6: Regression model of Log10 Transformed h-Index, Standardised as Z scores.
Disciplines Used in Models Anthropology Psychology Political Science Economics All disciplines
Model Number (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
Sex
Female = 1
Male = 0
-0.34***
(0.06)
-0.10*
(0.05)
-0.30***
(0.05)
-0.14***
(0.04)
-0.51***
(0.07)
-0.21***
(0.06)
-0.28***
(0.07)
-0.12*
(0.05)
-0.35***
(0.03)
-0.14***
(0.02)
-0.30***
(0.05)
-0.14***
(0.04)
Years Publishing 0.06***
(0.002)
0.06***
(0.002)
0.06***
(0.003)
0.07***
(0.002)
0.06***
(0.001)
0.06***
(0.001)
Anthropology 0.03
(0.04)
-0.10**
(0.03)
0.03
(0.06)
-0.13**
(0.04)
Economics -0.04
(0.04)
0.15***
(0.03)
-0.04
(0.05)
0.15***
(0.04)
Political Science -0.00
(0.04)
-0.16***
(0.03)
0.08
(0.05)
-0.14**
(0.04)
Sex X
Anthropology
-0.03
(0.08)
-0.06
(0.06)
Sex X
Economics
0.02
(0.09)
-0.01
(0.07)
Sex X
Political Science
-0.20*
(0.08)
-0.05
(0.06)
Constant 0.17***
(0.05)
-1.41***
(0.07)
0.12***
(0.03)
-1.36***
(0.05)
0.21***
(0.04)
-1.39***
(0.08)
0.08*
(0.04)
-1.47***
(0.06)
0.14***
(0.03)
-1.38***
(0.03)
0.12***
(0.03)
-1.38***
(0.04)
Observations 935 935 1,612 1,612 836 836 936 936 4,319 4,319 4,319 4,319
R2 0.03 0.46 0.02 0.47 0.06 0.38 0.02 0.48 0.03 0.44 0.03 0.45
F Statistic 28*** 400*** 37*** 672*** 56*** 257*** 16*** 439*** 32*** 692*** 19*** 432***
*p<0.05; **p<0.01; ***p<0.001
11
Published: 19th of May 2022 OpenPsych
whether our results were the artefact of our cleaning method. We also employed robust regression, using Huber
weights, to test whether our results were robust to outliers. To test for whether a possible confound, between-sex
differences in subdiscipline and subdiscipline citations, drives our results, we also tried dummy variables
for each academic journal. The results of all these robustness checks were extremely similar to the results in
Table 6. As such, we present these results in appendix B. For the regressions in Table 6, we also tried robust and
clustered standard errors. The pvalues for all regression coefficients remained within the same thresholds for
statistical significance. These results are not reported but are in the code within the supplementary files.
4 Survey
To see if the sex disparity in research output reflects anti-male bias we decided to run a survey of academics. If
academics said they supported discrimination in favour of women that would support the theory that hiring to
editorial boards is biased in favour of women. If this was not the case, the survey results would indicate that
sex disparities on editorial boards are best explained by sex differences alone.
We designed our survey using Alchemer (
https://www.alchemer.com/
). We created four questions on attitudes
towards gender bias
3
in hiring to journals and four questions on attitudes towards age bias in hiring to journals.
We asked questions on age bias for two reasons. The first reason was to test if years publishing’s effect on
research output was partly due to age bias. The second reason was that given the younger age of female
academics, an age bias may inadvertently cause a gender bias. We asked a further two questions on general
attitudes to meritocracy in hiring. All questions were on a 0-10 scale. When questions offered a choice between
two extremes (eg. pro-male bias to pro-female bias), the question stated that option 5 was a neutral answer. For
questions on gender bias and age bias, higher numbers indicated a pro-female bias or a pro-young bias
4
. We
achieved this by creating labels for each side of our 0-10 scale. Pictures of the questions asked can be found in
the supplementary materials.
We gathered a sample of survey respondents using Prolific (
https://www.prolific.co/
) Individuals are paid
to answer surveys through this website. Our inclusion criteria were for all individuals to have a PhD giving us
425 respondents. We employed a question asking respondents whether or not they worked in academia or were
publishing academic papers. After excluding individuals not in academic publishing we had a sample size of
231. All respondents were from Western countries such as The United States, The United Kingdom and Israel.
Summary statistics from our survey are shown in Table 7and density plots of question responses are presented
in Figure 2. The red dashed lines in Figure 2 indicate the 95 % confidence intervals for the mean response. We
used a t-test on the mean response to each question to see whether it differed significantly from 5. On question
4, academics were asked, "Should journal editors have a sex preference in hiring to editorial boards?". To ensure
all respondents correctly interpreted the question as implying that a sex preference would be discriminatory
and anti-meritocratic, we labelled the right end of responses "They should favor females above their academic
accomplishments" and the left the same but for males.
The mean response to this question was 5.6 which is significantly different from 5 (p< 0.001). Moreover,
one-third of academics said journals should have a pro-female bias and nearly two thirds (64 %) said journals
should have no age preference. This meant for every academic preferring men, there were eleven who preferred
women. Although most academics were against a sex bias, they were overwhelmingly more likely to support
journals preferring women than the reverse. This suggests there is a large minority of academics that would act
to discriminate against men in hiring to editorial boards.
Only 3 % of respondents believed that journal editors should be biased in favour of men. Such a low response for
this option could indicate academics only chose this option by mistake in answering the question or were lying
for the sake of humour. For comparison, an opinion poll found 4 % of Americans indicated that they believed
reptilians ran the world (Public Policy Polling,2013). This 4 % figure has been dubbed by blogger Alexander
(2013) as the ’Lizardman’s Constant’ to be used as a rule of thumb indicating the maximum survey response
3
In our survey of academics we used the term ‘gender’ rather than ‘sex’, although the rest of the paper is focused on sex. These two
concepts may have different interpretations and connotations, with sex implying biology and gender implying a ‘social construct’.
Transgender people constitute 0.6 % of all US adults (Jones,2021), so we suppose that in practice the concepts gender and sex mostly
overlap. As such we do not think changing terminology should change the interpretation of our results.
4
For questions 5 and 7, our survey responders were told higher numbers indicate a pro-old preference instead of a pro-young preference.
For ease of interpretation across different questions, answers for questions 5 and 7 were mirrored around point 5. Thus a raw answer of
3 became an answer of 7 and vice versa.
12
Published: 19th of May 2022 OpenPsych
Table 7: Survey Results.
Question Mean Response
t value (A mean
response of 5 is
the null hypoth-
esis)
Percent of re-
sponses below
5
Percent of re-
sponses at 5
Percent of re-
sponses above
5
number of re-
sponses
Q1. Is age diversity in
editorial boards impor-
tant?
6.8*** 11.9 13% 8% 79% 231
Q2. Is sex diversity in
editorial boards impor-
tant?
7.5*** 15.3 13% 5% 82% 231
Q3. Should journal edi-
tors have an age prefer-
ence in hiring to edito-
rial boards? (Pick 5 for
no age preference)
5.3*** 3.8 8% 71% 21% 231
Q4. Should journal ed-
itors have a sex prefer-
ence in hiring to edito-
rial boards? (Pick 5 for
no sex preference)
5.6*** 6.6 3% 64% 33% 231
Q5. Do older academics
have a greater aptitude
for academic research
than younger academics
(Pick 5 for no age differ-
ence)
5.1 1.1 21% 55% 24% 231
Q6. Do female aca-
demics have a greater ap-
titude for academic re-
search than men? (Pick
5 for no sex difference)
5.1 1.7 4% 87% 9% 231
Q7. Do you think jour-
nal editors have an age
preference in hiring to
editorial boards? (Pick
5 for no age preference)
3.8*** -9.9 62% 24% 13% 231
Q8. Do you think jour-
nal editors have a sex
preference in hiring to
editorial boards? (Pick
5 for no sex preference)
3.9*** -10.0 55% 35% 10% 231
Q9. How important
do you think academic
merit *should be* for hir-
ing to editorial boards?
8.1*** 26.2 3% 4% 93% 231
Q. 10 How important
do you think academic
merit currently is for hir-
ing to editorial boards?
6.8*** 14.2 13% 10% 77% 231
Notes: Critical values p<0.05, |t| > 1.96; p < 0.01, |t| > 2.60 ; p < 0.001, |t| > 3.3
13
Published: 19th of May 2022 OpenPsych
Figure 2: Density plots of survey responses. Note: red dashed lines indicated 95 % confidence intervals for the mean
response
that may be explained by mistakes or malice on the respondents’ behalves. Since support for anti-female
discrimination is lower than the Lizardman’s Constant we should be sceptical whether any respondents actually
support bias against women at all.
The results suggest that there is a large minority of academics that want to act to discriminate against men to
serve on editorial boards. The reverse case of academics willing to discriminate against women seems extremely
rare.
In our model of research output on editorial boards, we found scholars with more years of publishing performed
better. This might not just be due to older scholars having more experience but a result of biased lower
requirements for younger scholars. In question 3 academics were asked, "Should journal editors have an age
preference in hiring to editorial boards?". The mean answer was 5.3 indicating an average pro-young bias. It
was significantly different from the no bias response of 5 (p< 0.001). 21 % supported a pro-young bias, 71 %
supported no bias and 8 % supported a pro-old bias. These results, whilst not as extreme as the sex responses,
indicate a moderate pro-young bias in academia; nearly three academics preferred young academics for every
one that supported older academics.
These results indicate that academics are far more likely to be biased in favour of women and younger scholars.
As such, female academics are likely advantaged over men not only because of their sex but also because of
their relative youth.
In addition to asking academics whether they had an age or sex preference, we asked them whether they thought
journal editors were biased. For the sex question, the mean answer was 3.9 and for age 3.8. These differed
significantly from 5 (p< 0.001), suggesting that academics thought journals were biased in favour of men and
older scholars. So whilst academics are biased in favour of women and young people, they simultaneously
believe other academics have the opposite bias. This result seems somewhat paradoxical. We speculate in the
discussion that academics have such a strong anti-male bias that it deludes them into thinking that academia
has the opposite bias.
What motivates the academics to prefer young and female academics? In Question 2 we asked, "Is gender
diversity in editorial boards important". Question 1 asked the same but age diversity. A response of O meant
diversity was "not important", whilst a response of 10 indicated that diversity was "very important". Mean
responses were 7.5 for sex diversity and 6.8 for age diversity. 82 % and 79 % gave responses above 5 for sex and
age diversity respectively. With responses overwhelmingly closer to 10 than 0, it seems academics place much
value on diversity.
We also asked academics whether they believed men and older scholars have greater aptitude for academic
research than female and young scholars. The mean response to both questions was 5.1 which was not
significantly different from 5. This indicates academics believed neither sex had a greater aptitude for research,
14
Published: 19th of May 2022 OpenPsych
despite the fact men tend to receive more citations (Abramo et al.,2009;D’Amico et al.,2011;Dion et al.,2018;
Huang et al.,2020;Maliniak et al.,2013;Bird,2011), academic awards (Chan & Torgler,2020;Lincoln et al.,
2012) and are more likely to be considered eminent in their field (Murray,2003). It also suggests academics
believe that young scholars are just as good as older scholars
In Table 8we present a correlation matrix of all our survey questions to better examine what makes scholars
prefer women. Concern for sex diversity (Question 2) correlates at 0.34 (p< 0.001) with belief that journal
editors should prefer women (Question 4). Curiously, concern for age diversity (Question 1) does not appear to
correlate with belief that journal editors should prefer younger scholars (Question 3). This could suggest that
whilst academics prefer women for the sake of diversity, preference for younger scholars is not to do with a
general concern for age diversity. This could be because some scholars that believe in age diversity think this
requires more older scholars to be represented on journal boards.
In our survey, we found no statistically significant belief that younger or female scholars had a greater aptitude
than older or male scholars. This could indicate that bias against men is so strong amongst academics that
they refuse to believe in greater male academic ability. We find belief in higher female aptitude (Question 6)
correlates at 0.22 (p< 0.001) with a preference for hiring women (Question 4). This would support the idea that
bias in favour of women is motivating bias regarding their ability and also discrimination in favour of women.
The belief that journals are biased against women (Question 8) had a small negative correlation (-0.12) with a
preference to discriminate in women (Question 4). This could suggest that discrimination in favour of women
is motivated by countering perceived injustices against women. However, this correlation was not statistically
significant.
5 Discussion and Limitations
Our results have shown that men substantially outperform women on editorial boards in Political Science,
Psychology and Anthropology between 0.10-0.45 standard deviations in research output depending on model
specification. These results are robust, remaining stable when different measures of research output are used,
when journal effects are controlled for, when robust regression was used in addition to OLS and whether or not
we cleaned our data to discard outliers (including clearly erroneous data). In regression results, we found that
controlling for years publishing reduces the male advantage in research output, implying men in our sample
have been publishing for longer. We are uncertain about the reasons for this, but suggest that:
(1) older scholars have had more time to publish papers;
(2) younger cohorts of scholars are worse than older ones and;
(3) journals could have a pro-old age bias.
Overall we can be confident that male research output is higher than women’s on editorial boards. This is
unlikely under the hypothesis of anti-female bias which predicts that women have a higher research output. The
regression results update our prior beliefs away from anti-female discrimination and towards the possibilities of
anti-male discrimination and higher performance amongst male academics. To further explore the hypothesis
of anti-male bias, we surveyed academics on their attitudes to gender bias. We found that whilst most academics
were opposed to discrimination, they were eleven times more likely to support discrimination in favour of
women than against with regards to hiring to editorial boards Moreover, support for anti-male discrimination
represented only a trivial 3 % of our sample. This further supports the idea that there is anti-male bias in hiring
to editorial boards Academics also supported discrimination in favour of younger scholars. This means the
moderating effect of years publishing on the sex disparity in research output may be because age bias indirectly
creates a sex bias.
There are some limitations to our research methods. There may be potential errors in our data gathering
because of human error or Google Scholar making errors. Nonetheless, we do not believe any such data errors
could substantially alter our results. This is because our results were extremely similar when using different
dependent variables, both when we included and excluded outliers and when we used robust regression.
Furthermore, when citations on Google Scholar have been compared with citations on the Web of Science
database no sex bias was found (Andersen & Nielsen,2018). This suggests any errors from Google Scholar are
unlikely to cause bias in our results.
15
Published: 19th of May 2022 OpenPsych
Table 8: Survey Correlation Matrix.
Question Q1. Q2. Q3. Q4. Q5. Q6. Q7. Q8. Q9 Q10.
Q1. Is age diversity
in editorial boards
important?
1
Q2. Is gender di-
versity in editorial
boards important?
0.54*** 1
Q3. Should journal
editors have an age
preference in hiring
to editorial boards?
(Pick 5 for no age
preference)
0.05 0.005 1
Q4. Should jour-
nal editors have a
gender preference in
hiring to editorial
boards? (Pick 5
for no gender prefer-
ence)
0.14* 0.23*** 0.34*** 1
Q5. Do older
academics have a
greater aptitude for
academic research
than younger aca-
demics (Pick 5 for
no age difference)
0.02 0.07 0.04 0.03 1
Q6. Do female
academics have a
greater aptitude for
academic research
than men? (Pick 5
for no gender differ-
ence)
0.14* 0.17* 0.06 0.22*** -0.004 1
Q7. Do you think
journal editors have
an age preference in
hiring to editorial
boards? (Pick 5 for
no age preference)
-0.04 -0.03 -0.06 -0.11 0.03 -0.20** 1
Q8. Do you think
journal editors have
a gender preference
in hiring to edito-
rial boards? (Pick 5
for no gender prefer-
ence)
-0.11 -0.18** 0.04 -0.12 -0.15* 0.004 0.18** 1
Q9. How impor-
tant do you think
academic merit
*should be* for
hiring to editorial
boards?
-0.04 -0.05 -0.10 0.02 0.03 0.06 -0.13 0.07 1
Q10. How impor-
tant do you think
academic merit cur-
rently is for hiring to
editorial boards?
-0.15* 0.01 0.07 -0.04 -0.17** -0.07 -0.11 0.17* 0.16* 1
*p< 0.05; **p< 0.01; ***p< 0.001
16
Published: 19th of May 2022 OpenPsych
A limitation of our survey work of academics is that the respondents may not be a representative sample.
Respondents were people who supplemented their income by answering online surveys, suggesting our
respondents were disproportionately poor and possibly poorly performing academics. It could be that academics
near the bottom of the career ladder have different attitudes to discrimination than those higher up, such as
journal editors. We sampled ’elite’ journals, with the greatest citations per paper, creating further differences to
the academics in our survey sample. It is possible that whilst our respondents wanted to discriminate against
men, journal editors may discriminate against women. Nonetheless, this hypothesis seems very unlikely. The
fact that top publishers and journals are supporting affirmative action in favour of women (Bayazit,2020;
Elsevier,2021a,b;Lundine et al.,2018;Nature,2017) would suggest that high performing academics share the
same attitudes to sex bias as our surveyed academics who are likely poor performing Moreover, academics
at elite institutions are overwhelmingly left-wing, which is associated with having pro-female preferences
(Winegard et al.,2018), suggesting editors of top journals are likely to share the same preferences. For example,
39 % of elite American liberal arts colleges have no registered Republican professors (Langbert,2018).
Another limitation, pointed out by a reviewer, of our survey is the possible ambiguities of our questions. In our
questions we gave a 0-10 scale, with 0 and 10 labelled as extreme responses and 5 as intermediate. For example,
in question 4 on whether editors should have a preference for women, 10 was labelled. "They should favor
females above their academic accomplishments", 0 was given the same label but for men and 5 was labelled as
no preference. As such, the difference between 1-4 and 6-9 was not defined although we meant higher numbers
to represent more pro-female preferences. Some respondents may not have realised that these intermediate
values represented different points on the dimension of pro- male to pro-female preferences. Nonetheless, we
do not think any ambiguity in our questions have distorted our results. Respondents were given the opportunity
to gave feedback, but did not make comments about the scale of our questions being confusing. Furthermore, a
visual inspection of the results in Figure 2show smooth distributions, with modal answers not always being 0,
5 or 10, suggesting respondents correctly interpreted the other values on our 0-10 scale.
The fact that many academics and publishers are concerned that academia has an anti-female bias would seem
to make the theory of anti-male bias unlikely if these academics were rational in their claims. However, this also
poses a paradox, if so many academics are publicly against anti-female discrimination how can academia still be
so biased against women? For example, in our survey results, whilst academics on net supported discrimination
in favour of women and younger scholars they believed other academics who ran journals had the opposite
biases.
Clark & Winegard (2020) explain this paradox by arguing that the pervasive narrative of misogyny could itself
be caused by academia and society at large having an anti-male bias This could be an example of preference
falsification (Kuran,1997), whereby individuals lie about their true preferences, or self-deception (Trivers,
2011) whereby individuals lie to themselves about what is true or desirable to avoid the reputational costs of
breaking social taboos. After all, there are large incentives to believe in the value of diversity and affirmative
action. Academics who do not support affirmative action for women or diversity might be shunned or even
’cancelled’ by their overwhelmingly left-wing colleagues, if they are hired at all. For example, Cern physicist
Alessandro Strumia lost his job for publicly arguing that higher male performance in academia was not a result
of discrimination. This theory would also explain why, in our survey results, academics do not believe in sex
differences in academic aptitude despite greater male average intelligence (Lynn,1994;?;?;Lynn & Irwing,
2004;Nyborg,2005), greater variance in male intelligence (Baye & Monseur,2016) and the overwhelming
representation of men as eminent figures in science (Darwin,1871;Murray,2003).
We find some evidence supportive of the Clark & Winegard (2020) view, those who were more strongly biased
against men, more strongly believed academia was biased against women. Although this could be a rational
desire to balance the scale, it could also illustrate anti-male bias making scholars biased in their evaluation
of academia. We also found that academics’ female sex preference was associated with a belief in greater
female aptitude, despite lower female publication metrics. Indicating that sex biases can distort academic’s
non-normative beliefs about sex in academia.
We cannot determine whether editorial boards have previously exhibited a bias against women because our
data are not longitudinal, but we can be reasonably confident that they do not now. As such, affirmative action
policies for editorial boards may be undermining meritocracy. In Gary Becker’s taste discrimination model of
the labour market (1971), profit-seeking firms should employ discriminated groups because they are accepting
of lower wages Likewise, journals looking for top talent could do well in recruiting men other editorial boards
have ignored.
17
Published: 19th of May 2022 OpenPsych
References
Abramo, G., D’Angelo, C., & Caprasecca, A. (2009). Gender differences in research productivity: A bibliometric
analysis of the italian academic system. Scientometrics,79(3), 517–539. doi: 10.1007/s11192-007-2046-8
Addis, E., & Villa, P. (2003). The editorial boards of italian economics journals: Women, gender, and social
networking. Feminist Economics,9(1), 75-91.
Ahmad, S. (2017). Family or future in the academy? Review of Educational Research,87(1), 204–239. doi:
10.3102/0034654316631626
Aiston, S. (2014). Leading the academy or being led? hong kong women academics. Higher Education Research
& Development,33(1), 59–72. doi: 10.1080/07294360.2013.864618
Akcigit, U., Pearce, J., & Prato, M. (2020). Tapping into talent: Coupling education and innovation policies for
economic growth. Cambridge, MA: National Bureau of Economic Research. doi: 10.3386/w27862
Alexander, S. (2013). Lizardman’s constant is 4%. Retrieved from
https://slatestarcodex.com/2013/04/12/
noisy-poll-results-and-reptilian-muslim-climatologists-from-mars/
Amrein, K., Langmann, A., Fahrleitner-Pammer, A., Pieber, T. R., & Zollner-Schwetz, I. (2011). Women
underrepresented on editorial boards of 60 major medical journals. Gender Medicine,8(6), 378-387. doi:
https://doi.org/10.1016/j.genm.2011.10.007
Andersen, J. P., & Nielsen, M. W. (2018). Google scholar and web of science: Examining gender differences
in citation coverage across five scientific disciplines. Journal of Informetrics,12(3), 950-959. doi:
https://
doi.org/10.1016/j.joi.2018.07.010
Austin, D. E., & Jackson, M. (2019). Benevolent and hostile sexism differentially predicted by facets of right-
wing authoritarianism and social dominance orientation. Personality and Individual Differences,139, 34-38.
doi: https://doi.org/10.1016/j.paid.2018.11.002
Bayazit, K. (2020). Kumsal bayazit at eu gender in science symposium: “we must make progress across all
dimensions of diversity.”. Elsevier. Retrieved from
https://www.elsevier.com/connect/kumsal-bayazit
-at-eu-gender-in-science-symposium
Baye, A., & Monseur, C. (2016). Gender differences in variability and extreme scores in an international context.
Large-scale Assessments in Education,4(1), 1. doi: 10.1186/s40536-015-0015-x
Becker, G. (1971). The economics of discrimination (2nd ed.). Chicago: University of Chicago Press (Economics
research studies of the Economics Research Center of the University of Chicago).
Bird, K. S. (2011). Do women publish fewer journal articles than men? sex differences in publication
productivity in the social sciences. British Journal of Sociology of Education,32(6), 921-937. doi:
10.1080/
01425692.2011.596387
Boston College Center For Work and Family. (2019). Expanded paid parental leave measuring the impact of leave
on work & family. Retrieved from
https://www.prweb.com/releases/boston_college_center_for_work
_family_examines_expanded_paid_parental_leave_in_the_workplace/prweb16700674.htm
Bruna, S., Fullerton, H., Király1, G., & Bruna, E. (2017). The gatekeeper project: Crowdsourced examination of
the gender composition of anthropology journals. Retrieved from
https://www.seanbruna.com/wp-content/
uploads/2017/12/AAA-Poster-2017-Final_WATERMK.pptx.pdf
Carlsson, M., Finseraas, H., Midtbøen, A. H., & Rafnsdóttir, G. L. (2020). Gender Bias in Academic Recruitment?
Evidence from a Survey Experiment in the Nordic Region. European Sociological Review,37(3), 399–410. doi:
10.1093/esr/jcaa050
Ceci, S. J., Ginther, D. K., Kahn, S., & Williams, W. M. (2014). Women in academic science: A changing
landscape. Psychological Science in the Public Interest,15(3), 75-141. doi: 10.1177/1529100614541236
Chan, H., & Torgler, B. (2020). Gender differences in performance of top cited scientists by field and country.
Scientometrics,125(3), 2421–2447. doi: 10.1007/s11192-020-03733-w
18
Published: 19th of May 2022 OpenPsych
Cho, A. H., Johnson, S. A., Schuman, C. E., Adler, J. M., Gonzalez, O., Graves, S. J., . . . Bruna, E. M. (2014).
Women are underrepresented on the editorial boards of journals in environmental biology and natural
resource management. PeerJ,2, e542. doi: 10.7717/peerj.542
Christopher, A. N., & Mull, M. S. (2006). Conservative ideology and ambivalent sexism. Psychology of Women
Quarterly,30(2), 223-230. doi: 10.1111/j.1471-6402.2006.00284.x
Clark, C., & Winegard, B. (2020). The myth of pervasive misogny. Retrieved from
https://quillette.com/
2020/07/27/the-myth-of-pervasive-misogyny/
D’Amico, R., Vermigli, P., & Canetto, S. (2011). Publication productivity and career advancement by female
and male psychology faculty: The case of italy. Journal of Diversity in Higher Education,4(3), 175–184. doi:
10.1037/a0022570
Darwin, C. (1871). The descent of man, and selection in sex.
Dion, M. L., Sumner, J. L., & Mitchell, S. M. (2018). Gendered citation patterns across political science and
social science methodology fields. Political Analysis,26(3), 312–327. doi: 10.1017/pan.2018.12
Dunkin, M. (1991). Determinants of academic career advancement at an australian university. Higher Education
Research & Development,10(2), 115-131. doi: 10.1080/0729436910100201
Dutton, E., & Charlton, B. (2016). The genius famine: Why we need geniuses, why they’re dying out, why we must
rescue them. University of Buckingham Press.
Elsevier. (2021a). Elsevier’s journals’ now displaying editors’ gender in support of diversity. Elsevier. Re-
trieved from
https://www.elsevier.com/about/press-releases/corporate/elseviers-journals-now
-displaying-editors-gender-in-support-of-diversity
Elsevier. (2021b). Inclusion & diversity advisory board. Elsevier. Retrieved from
https://www.elsevier.com/
about/inclusion-diversity-board
Evans, H. P.-H., J., & Robinson, D. (2005). Women’s involvement in educational psychology journals from 1976
to 2004. Educational Psychology Review,17(3), 263–271. doi: 10.1007/s10648-005-5619-1
Ferriman, K., Lubinski, D., & Benbow, C. (2009). Work preferences, life values, and personal views of top
math/science graduate students and the profoundly gifted: Developmental changes and gender differences
during emerging adulthood and parenthood. Journal of Personality and Social Psychology,97(3), 517–532. doi:
10.1037/a0016030
Foschi, M., Lai, L., & Sigerson, K. (1994). Gender and double standards in the assessment of job applicants.
Social Psychology Quarterly,57(4), 326. doi: 10.2307/2787159
Fox, M. F. (2005). Gender, family characteristics, and publication productivity among scientists. Social Studies
of Science,35(1), 131-150. doi: 10.1177/0306312705046630
Fraga, L. R., Givens, T. E., Pinderhughes, D. M., Avalos, M., Covin, W. D., Hagopian, S. F., . .. Wallace, S. L.
(2011). Political science in the 21st century: Report of the task force on political science in the 21st century.
Gibbons, J. D., & Fish, M. (1991). Rankings of economics faculties and representation on editorial boards
of top journals. The Journal of Economic Education,22(4), 361–372. Retrieved 2022-05-01, from
http://
www.jstor.org/stable/1183356
Ginther, D. K., & Hayes, K. J. (1999, May). Gender differences in salary and promotion in the humanities.
American Economic Review,89(2), 397-402. doi: 10.1257/aer.89.2.397
Ginther, D. K., & Hayes, K. J. (2003). Gender differences in salary and promotion for faculty in the humanities
1977-95. The Journal of Human Resources,38(1), 34–73. doi: 10.2307/1558755
Ginther, D. K., & Kahn, S. (2004). Women in economics: Moving up or falling offthe academic career ladder?
Journal of Economic Perspectives,18(3), 193–214. doi: 10.1257/0895330042162386
Hafeez, D. M., Waqas, A., Majeed, S., Naveed, S., Afzal, K. I., Aftab, Z., ... Khosa, F. (2019). Gender
distribution in psychiatry journals’ editorial boards worldwide. Comprehensive Psychiatry,94, 152119. doi:
https://doi.org/10.1016/j.comppsych.2019.152119
19
Published: 19th of May 2022 OpenPsych
Helmreich, R., Spence, J., Beane, W., Lucker, G., & Matthews, K. (1980, 11). Making it in academic psychology:
Demographic and personality correlates of attainment. Journal of Personality and Social Psychology,39,
896-908. doi: 10.1037/0022-3514.39.5.896
Herr, J., Roy, R., & Klerman, J. A. (2020, November). Gender differenes in need-
ing and taking leave. ABT Associates. Retrieved from
https://www.dol.gov/
sites/dolgov/files/OASP/evaluation/pdf/WHD_FMLAGenderShortPaper_January2021.pdf?
_hsenc=p2ANqtz-8_pduc8RO3z2ye9aSgmKVzgKw5iEj7eoo9wvEz1KHmjEuL86968lsbCPpQ_NWZeW1
-zlzsZm8t0YeUJ8gjqXX26qJRiAR7tBz8ldt4hvagpsU4eGM
Higher Education Statistics Agency. (2021, 30 April). Academic staffby academic cost centre, sex and academic
year. Retrieved from https://www.hesa.ac.uk/data-and-analysis/staff/chart-6
Hodson, G., MacInnis, C. C., & Busseri, M. A. (2017). Bowing and kicking: Rediscovering the fundamental link
between generalized authoritarianism and generalized prejudice. Personality and Individual Differences,104,
243-251. Retrieved from
https://www.sciencedirect.com/science/article/pii/S0191886916309199
doi: https://doi.org/10.1016/j.paid.2016.08.018
Huang, J., Gates, A. J., Sinatra, R., & Barabási, A.-L. (2020). Historical comparison of gender inequality in
scientific careers across countries and disciplines. Proceedings of the National Academy of Sciences,117(9),
4609-4616. doi: 10.1073/pnas.1914221117
Ioannidou, E., & Rosiana, A. (2015). Under-representation of women on dental journal editorial boards. PLOS
ONE,10(1). doi: 10.1371/journal.pone.0116630
Jones, J. M. (2021). Lgbt identification rises to 5.6% in latest u.s. estimate. Retrieved from
https://news.gallup
.com/poll/329708/lgbt-identification-rises-latest-estimate.aspx
Kahn, S., & Ginther, D. (2017). Women and stem. Cambridge, MA: National Bureau of Economic Research, w23525.
Retrieved from http://www.nber.org/papers/w23525 doi: 10.3386/w23525
Kleijn, M. D., Jayabalasingham, B., Falk-Krzesinski, H. J., Collins, T., Kuiper-Hoyng, L., Cingolani, I., . . .
et al. (2020). The researcher journey through a gender lens: An examination of research participation,
career progression and perceptions across the globe. Elsevier. Retrieved from
https://www.elsevier.com/
connect/gender-report
Kuran, T. (1997). Private truths, public lies: the social consequences of preference falsification (1st ed.). Harvard
University Press.
Langbert, M. (2018, 04). Homogenous: The political affiliations of elite liberal arts college faculty. Academic
Questions,31, 1-12. doi: 10.1007/s12129-018-9700-x
Lincoln, A. E., Pincus, S., Koster, J. B., & Leboy, P. S. (2012). The matilda effect in science: Awards and prizes in
the us, 1990s and 2000s. Social Studies of Science,42(2), 307-320. doi: 10.1177/0306312711435830
Lindsey, D. (1976). Distinction, achievement, and editorial board membership. American Psychologist,31(11),
799–804. doi: 10.1037/0003-066X.31.11.799
Lord, T. (1987). A look at spatial abilities in undergraduate women science majors. Journal of Research in Science
Teaching,24(8), 757–767. doi: 10.1002/tea.3660240808
Lundine, J., Bourgeault, I. L., Clark, J., Heidari, S., & Balabanova, D. (2018). The gendered system of academic
publishing. The Lancet,391(10132), 1754–1756. doi: 10.1016/S0140-6736(18)30950-4
Luoto, S. (2020). Sex differences in people and things orientation are reflected in sex differences in academic
publishing. Journal of Informetrics,14(2), 101021. doi: https://doi.org/10.1016/j.joi.2020.101021
Lynn, R. (1994). Sex differences in intelligence and brain size: A paradox resolved. Personality and Individual
Differences,17(2), 257-271. doi: https://doi.org/10.1016/0191-8869(94)90030-2
Lynn, R. (2017). Sex differences in intelligence: The developmental theory. Mankind Quarterly,58(1), 9-42. doi:
10.46469/mq.2017.58.1.2
Lynn, R., & Irwing, P. (2004). Sex differences on the progressive matrices: A meta-analysis. Intelligence,32(5),
481-498. doi: https://doi.org/10.1016/j.intell.2004.06.008
20
Published: 19th of May 2022 OpenPsych
Lynn, R., & Mikk, J. (2009). National iqs predict educational attainment in math, reading and science across 56
nations. Intelligence,37(3), 305-310. doi: https://doi.org/10.1016/j.intell.2009.01.002
Madison, G., & Fahlman, P. (2021). Sex differences in the number of scientific publications and citations
when attaining the rank of professor in sweden. Studies in Higher Education,46(12), 2506-2527. doi:
10.1080/03075079.2020.1723533
Maliniak, D., Powers, R., & Walter, B. F. (2013). The gender citation gap in international relations. International
Organization,67(4), 889–922. doi: 10.1017/S0020818313000209
Martinez, E. D., Botos, J., Dohoney, K. M., Geiman, T. M., Kolla, S. S., Olivera, A., . . . Cohen-Fix, O. (2007).
Falling offthe academic bandwagon: Women are more likely to quit at the postdoc to principal investigator
transition’. EMBO reports,8(11), 977–981. doi: 10.1038/sj.embor.7401110
Mauleón, E., Hillán, L., Moreno, L., Gómez, I., & Bordons, M. (2013). Assessing gender balance among journal
authors and editorial board members. Scientometrics,95(1), 87–114. doi: 10.1007/s11192-012-0824-4
Mazov, N., & Gureev, V. (2016). The editorial boards of scientific journals as a subject of scientometric
research: A literature review. Scientific and Technical Information Processing,43(3), 144–153. doi:
10.3103/
S0147688216030035
Metz, I., & Harzing, A. (2012). An update of gender diversity in editorial boards: a longitudinal study of
management journals. Personnel Review,41(3), 283–300. doi: 10.1108/00483481211212940
Metz, I., & Harzing, A.-W. (2009). Gender diversity in editorial boards of management journals. Academy of
Management Learning & Education,8(4), 540–557. doi: 10.5465/amle.8.4.zqr540
Meyers, M. (2013). The war on academic women: Reflections on postfeminism in the neoliberal academy.
Journal of Communication Inquiry,37(4), 274-283. doi: 10.1177/0196859913505619
Miller, D. I., & Wai, J. (2015). The bachelor’s to ph.d. stem pipeline no longer leaks more women than men: a
30-year analysis. Frontiers in Psychology,6. doi: 10.3389/fpsyg.2015.00037
Morton, M. J., & Sonnad, S. S. (2007). Women on professional society and journal editorial boards. Journal of
the National Medical Association,99(7), 764–771.
Moss-Racusin, C. A., Dovidio, J. F., Brescoll, V. L., Graham, M. J., & Handelsman, J. (2012). Science faculty’s
subtle gender biases favor male students. Proceedings of the National Academy of Sciences,109(41), 16474-16479.
doi: 10.1073/pnas.1211286109
Mumford, K. (2016, 04). On gender, research discipline and being an economics journal editor in the uk. , 15–
19. Retrieved from
https://eprints.whiterose.ac.uk/97744/1/Gender_Composition_of_Editorial
_Boards_in_Economics_March_17_2016.pdf
Murray, C. (2003). Human accomplishment: the pursuit of excellence in the arts and sciences, 800 bc to 1950 (1st
ed.). New York, NY: HarperCollins.
Nature. (2017). Gender imbalance in science journals is still pervasive. Nature,541(7638), 435–436. doi:
10.1038/541435b
Nyborg, H. (2005). Sex-related differences in general intelligence g, brain size, and social status. Personality and
Individual Differences,39(3), 497-509. doi: https://doi.org/10.1016/j.paid.2004.12.011
Over, R. (1981). Representation of women on the editorial boards of psychology journals. American Psychologist,
36(8), 885–891. doi: 10.1037/0003-066X.36.8.885
O’Dea, R. E., Lagisz, M., Jennions, M. D., & Nakagawa, S. (2018). Gender differences in individual variation
in academic grades fail to fit expected patterns for stem. Nature Communications,9(1), 3777. doi:
10.1038/
s41467-018-06292-0
Palmer, B., van Assendelft, L., & Stegmaier, M. (2020). Revisiting the presence of women in political science
journal editorial positions. PS: Political Science & Politics,53(3), 499–504. doi:
10.1017/S1049096520000190
Palser, E. R., Lazerwitz, M., & Fotopoulou, A. (2021). Gender and geographical disparity in editorial boards of
journals in psychology and neuroscience. bioRxiv. doi: 10.1101/2021.02.15.431321
21
Published: 19th of May 2022 OpenPsych
Public Policy Polling. (2013). Democrats and republicans differ on conspiracy theory beliefs. Public
Policy Polling. Retrieved from
https://www.publicpolicypolling.com/wp-content/uploads/2017/09/
PPP_Release_National_ConspiracyTheories_040213.pdf
Quadlin, N. (2018). The mark of a woman’s record: Gender and academic performance in hiring. American
Sociological Review,83(2), 331-360. doi: 10.1177/0003122418762291
Reilly, D., & Neumann, D. (2013). Gender-role differences in spatial ability: A meta-analytic review. Sex Roles,
68(9–10), 521–535. doi: 10.1007/s11199-013-0269-0
Research Papers in Economics Author Service. (2021). Female representation in economics, as of march 2021.
Retrieved from https://ideas.repec.org/top/female.html
Robinson, D. H., McKay, D. W., Katayama, A. D., & Fan, A.-C. (1998). Are women underrepresented as authors
and editors of educational psychology journals? Contemporary Educational Psychology,23(3), 331-343. doi:
https://doi.org/10.1006/ceps.1997.0967
Ruscio, J., Seaman, F., D’Oriano, C., Stremlo, E., & Mahalchik, K. (2012). Measuring scholarly impact using
modern citation-based indices. Measurement: Interdisciplinary Research and Perspectives,10(3), 123-146. doi:
10.1080/15366367.2012.711147
Rørstad, K., & Aksnes, D. W. (2015). Publication rate expressed by age, gender and academic position –
a large-scale analysis of norwegian academic staff.Journal of Informetrics,9(2), 317-333. doi:
https://
doi.org/10.1016/j.joi.2015.02.003
Santos, G., & Dang Van Phu, S. (2019). Gender and academic rank in the uk. Sustainability,11(11), 3171. doi:
10.3390/su11113171
Steinpreis, R., Anders, K., & Ritzke, D. (1999). The impact of gender on the review of the curricula vitae of
job applicants and tenure candidates: A national empirical study. Sex Roles,41, 509–528. doi:
10.1023/A:
1018839203698
Strumia, A. (2021, 04). Gender issues in fundamental physics: A bibliometric analysis. Quantitative Science
Studies,2(1), 225-253. Retrieved from
https://doi.org/10.1162/qss_a_00114
doi:
10.1162/qss_a
_00114
Tran, U., Hofer, A., & Voracek, M. (2014). Sex differences in general knowledge: Meta-analysis and new data on
the contribution of school-related moderators among high-school students. PLoS ONE,9(10), e110391. doi:
10.1371/journal.pone.0110391
Trivers, R. (2011). Deceit and self-deception: fooling yourself the better to fool others. London: Allen Lane.
Venables, W., & Ripley, B. (2010). Modern applied statistics with s (4th ed.). New York: Springer (Statistics and
computing).
Weisberg, Y., DeYoung, C., & Hirsh, J. (2011). Gender differences in personality across the ten aspects of the big
five. Frontiers in Psychology,2. doi: 10.3389/fpsyg.2011.00178
Wiley. (2021). Editorial office guidelines. Retrieved from
https://authorservices.wiley.com/editors/
editorial-office-guidelines/editorial-board.html
Williams, W. M., & Ceci, S. J. (2015). National hiring experiments reveal 2:1 faculty preference for women
on stem tenure track. Proceedings of the National Academy of Sciences,112(17), 5360-5365. doi:
10.1073/
pnas.1418878112
Winegard, B., Clark, C., & Hasty, C. (2018). Equalitarianism: A source of liberal bias. SSRN Electronic Journal
[Preprint]. doi: 10.2139/ssrn.3175680
22
Published: 19th of May 2022 OpenPsych
Appendix A
Table 9: List of Journal Editorial Boards.
Anthropology Journals Economics Journals
Political Science and Interna-
tional Relations Journals
Psychology Journals
Journal of Consumer Research
Quarterly Journal of Eco-
nomics
American Journal of Political
Science
The Annual Review of Psychol-
ogy
Journal of Peasant Studies
Journal of Economic Perspec-
tives
American Political Science Re-
view
Psychological Bulletin
American Ethnologist
Brookings Papers on Economic
Activity
International Organization
Psychological Science in the
Public Interest
Journal of Human Evolution Journal of Political Economy
British Journal of Political Sci-
ence
International Review of Sport
and Exercise Psychology
Annual Review of Anthropol-
ogy
Journal of Economic Literature
Political Analysis
Annual Review of Clinical Psy-
chology
Science, Technology & Human
Values
Journal of Financial Economics
International Security
Annual Review of Organiza-
tional Psychology and Organi-
zational Behavior
Journal of Marriage and Family
Review of Environmental Eco-
nomics and Policy
International Affairs
Personality and Social Psychol-
ogy Review
American Journal of Physical
Anthropology
Energy Economics
Review of International Orga-
nizations
Social Issues and Policy Review
Journal of Cross-Cultural Psy-
chology
American Economic Review
Geopolitics, History, and Inter-
national Relations
Journal of Personality and So-
cial Psychology
Evolutionary Anthropology Economic Policy Critical Social Policy
Journal of Occupational Health
Psychology
Games and Culture Journal of Finance
European Journal of Interna-
tional Relations
Clinical Psychology Review
Evolutionary Human Sciences
Cambridge Journal of Regions,
Economy and Society
Journal of Peace Research
Educational Psychology Re-
view
Archaeological and Anthropo-
logical Sciences
American Economic Journal:
Applied Economics
Policy and Society Educational Psychologist
Journal of Racial and Ethnic
Health Disparities
Econometrica Global Environmental Politics
Current Directions in Psycho-
logical Science
Race and Social Problems Economic Geography
Chinese Journal of Interna-
tional Politics
Trends in Cognitive Sciences
Anthropological Theory
Review of Economics and
Statistics
East European Politics Developmental Review
Cross-Cultural Research Small Business Economics Research and Politics Behavior Research Methods
Sexualities Review of Economics Studies Journal of Conflict Resolution
Behaviour Research and Ther-
apy
Journal of Anthropological Sci-
ences
The Review of Financial Stud-
ies
Security Dialogue Neuropsychology Review
Human Ecology
Journal of Business & Eco-
nomic Statistics
Cooperation and Conflict Psychological Methods
Culture, Medicine, and Psychi-
atry
Annual Review of Economics World Politics
Perspectives on Psychological
Science
Medical Anthropology: Cross
Cultural Studies in Health and
Illness
Finance Research Letters European Union Politics
European Journal of Psychol-
ogy Applied to Legal Context
Discourse Studies World Development
Political Science Research and
Methods
Computers in Human Behavior
Chinese Sociological Review
Journal of Accounting and Eco-
nomics
Perspectives on Politics Psychological review
Anthrozoas
American Economic Journal:
Economic Policy
Democratization
Journal of the Learning Science
Journal of Contemporary
Ethnography
Ecological Economics Political Studies Review
European Review of Social Psy-
chology
American Journal of Human Bi-
ology
Annual Review of Resource
Economics
Journal of Contemporary
China
Trauma, Violence & Abuse
Journal of Eastern African
Studies
Journal of Asian Finance, Eco-
nomics and Business
Politics
Journal of Business and Psy-
chology
Journal of Human Trafficking
American Economic Journal:
Macroeconomics
International Studies Quar-
terly
Journal of Applied Psychology
Culture and Psychology Oeconomia Copernicana Geopolitics
Journal of Behavioral Addic-
tions
23
Published: 19th of May 2022 OpenPsych
Appendix B
In Table 10 we re-run the results of Table 6with dummy variables for journals. This is to check whether women
have a lower academic output because they prefer subdisciplines that receive fewer citations. Some of the
sex coefficients are lower and some higher after controlling for journal effects. In model 2, controlling for
journal effects make the sex coefficient lower from -0.10 to -0.09. This makes the coefficient lose its statistical
significance at the 5 % level. Given the close consistency of the Table 10 results and the low p values for
coefficients in the other 11 models, it is very likely that model 2 is a false negative.
As a robustness test, we use the robust regression with Huber weights. This approach puts lower weights on
observations with a high residual. This is useful for seeing whether lessening the effect of outlier values changes
our results. For example, this helps us to be confident that human errors in data gathering or random errors by
Google Scholar have not distorted the results. Our robust regressions are created using the rlm() function in the
R package MASS. For details on the robust regression see Venables & Ripley (2010). The Robust regression
results are shown in Table 11.
The use of robust regression does not seem to change our results substantially. The predicted sex disparity
appears approximately the same and is still statistically significant in every model. Likewise, the coefficients
for years publishing are the same, rounded to two decimal places.
There are still no significant sex discipline interaction terms. Overall this suggests that outlier observations are
not distorting our regression results.
In table 12 we rerun our regression analyses but with the inclusion of individuals that Google Scholar has
misattributed 5 or more papers to and without removing outlier observations. We do this to see whether our
exclusion of these individuals may have biased our results. The results are almost indistinguishable from the
regression results in table 6. Some of the coefficients on sex are slightly different - within 0.03 of the coefficients
in table 6. This means our exclusion of ‘overattributed individuals’ has only changed our estimates of the sex
gap in research productivity by a maximum of 0.03 standard deviations. This suggests that our results are not
an artefact of our data cleaning process.
In Tables 13,14 and 15 we use alternative dependent variables for research output instead of our transformed
h-index. The variables employed are the raw h-index and transformed citation and publication counts. There
are no notable differences between these regressions and our main results in Table 6. This suggests the sex
difference in academic output is measurement invariant and not a coincidence or p-hacked result of relying on
our transformed h-index.
24
Published: 19th of May 2022 OpenPsych
Table 10: Regression models of Log10 Transformed h-Index, Standardised as Z scores.
Anthropology Psychology Political Science Economics All disciplines
Model
Number
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
Sex
Female = 1
Male = 0
-0.23***
(0.06)
-0.09 (0.05)
-0.33***
(0.05)
-0.19***
(0.04)
-0.44***
(0.07)
-0.24***
(0.06)
-0.23***
(0.07)
-0.13*
(0.05)
-0.31***
(0.03)
-0.17***
(0.02)
-0.33***
(0.05)
-0.19***
(0.04)
Years Pub-
lishing
0.06***
(0.002)
0.05***
(0.002)
0.05***
(0.003)
0.06***
(0.003)
0.06***
(0.001)
0.06***
(0.001)
Anthropology
-1.24***
(0.27)
-0.97**
(0.21)
-1.29***
(0.27)
-1.02**
(0.20)
Economics
-1.46***
(0.27)
-0.60***
(0.21)
-1.49***
(0.25)
-0.62***
(0.20)
Political Sci-
ence
-1.25***
(0.28)
-0.83***
(0.22)
-1.21***
(0.26)
-0.82***
(0.21)
Sex
×
An-
thropology
0.10 (0.08) 0.10 (0.06)
Sex
×
Eco-
nomics
0.10 (0.08) 0.05 (0.06)
Sex
×
Politi-
cal Science
-0.11 (0.08) -0.05 (0.06)
Journal
Dummy
Variables
✓✓✓✓✓✓✓✓✓✓✓✓
Constant
-0.42**
(0.04)
-1.92***
(0.12)
1.30* (0.50)
-1.07***
(0.05)
0.04 (0.04)
-1.39***
(0.07)
-0.14***
(0.04)
-1.63***
(0.06)
0.43***
(0.03)
-1.06***
(0.03)
0.43***
(0.03)
-1.05***
(0.03)
Observations
935 935 1,643 1,643 843 843 941 941 4,362 4,362 4,362 4,362
R2 0.19 0.53 0.24 0.55 0.24 0.48 0.29 0.53 0.24 0.53 0.24 0.53
F Statistic 7*** 33*** 17*** 63*** 8*** 24*** 13*** 33*** 11*** 39*** 11*** 38***
*p<0.05; **p<0.01; ***p<0.001
25
Published: 19th of May 2022 OpenPsych
Table 11: Robust Regression models of Log10 Transformed h-Index, Standardised as Z scores.
Anthropology Psychology Political Science Economics All disciplines
Model
Number
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
Sex
Female = 1
Male = 0
-0.34***
(0.07)
-0.11*
(0.05)
-0.33***
(0.05)
-0.14***
(0.04)
-0.53***
(0.07)
-0.25***
(0.05)
-0.26***
(0.07)
-0.13*
(0.05)
-0.36***
(0.03)
-0.15***
(0.02)
-0.33***
(0.05)
-0.14***
(0.04)
Years Pub-
lishing
0.06***
(0.002)
0.06***
(0.002)
0.06***
(0.003)
0.07***
(0.002)
0.06***
(0.001)
0.06***
(0.001)
Anthropology
0.04 (0.04)
-0.11**
(0.03)
0.04 (0.06)
-0.13**
(0.04)
Economics
-0.06 (0.04)
0.15***
(0.03)
-0.07 (0.05)
0.15***
(0.04)
Political Sci-
ence
0.02 (0.04)
-0.14***
(0.03)
0.09 (0.05)
-0.01*
(0.04)
Sex
×
An-
thropology
-0.01 (0.08)
0.04 (0.06)
Sex
×
Eco-
nomics
0.07 (0.09)
-0.01 (0.06)
Sex
×
Politi-
cal Science
-0.19*
(0.09)
-0.09 (0.06)
Constant
0.20***
(0.05)
-1.40***
(0.06)
0.16***
(0.03)
-1.37***
(0.05)
0.25***
(0.04)
-1.34***
(0.08)
0.08* (0.04)
-1.46***
(0.06)
0.17***
(0.03)
-1.38***
(0.03)
0.16***
(0.03)
-1.38***
(0.03)
Observations
935 935 1,612 1,612 836 836 936 936 4,318 4,319 4,319 4,319
Residual
Standard
Error
1.02 0.66 1.06 0.71 0.96 0.72 0.94 0.62 1.01 0.69 1.00 0.68
*p<0.05; **p<0.01; ***p<0.001
26
Published: 19th of May 2022 OpenPsych
Table 12: Regression models of Log10 Transformed h-Index, Standardised as Z scores. Includes individuals with erroneous Google Scholar pages.
Anthropology Psychology Political Science Economics All disciplines
Model Number (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
Sex Female = 1 Male = 0 -0.36*** -0.10* -0.34*** -0.15*** -0.49*** -0.20** -0.30*** -0.11* -0.37*** -0.14*** -0.34*** -0.15***
(0.06) (0.05) (0.05) (0.04) (0.06) (0.05) (0.07) (0.07) (0.03) (0.02) (0.05) (0.04)
Years Publishing 0.06*** 0.06*** 0.05*** 0.07*** 0.06*** 0.06***
(0.002) (0.002) (0.002) (0.002) (0.001) (0.001)
Anthropology -0.42*** -0.53*** -0.41*** -0.55***
(0.04) (0.03) (0.05) (0.04)
Economics -0.55*** -0.33*** -0.55*** -0.34**
(0.04) (0.03) (0.05) (0.04)
Political Science -0.42*** -0.56*** -0.36*** -0.56***
(0.04) (0.03) (0.05) (0.04)
Sex ×Anthropology -0.03 0.06
(0.08) (0.06)
Sex ×Economics 0.04 0.06
(0.08) (0.06)
Sex ×Political Science -0.16 0.001
(0.08) (0.06)
Constant 0.01 -1.57*** 0.42*** -1.07*** 0.06 -1.39*** -0.13*** -1.64*** 0.43*** -1.07*** 0.42*** -1.07***
(0.04) (0.06) (0.03) (0.05) (0.04) (0.07) (0.04) (0.06) (0.03) (0.03) (0.03) (0.03)
Observations 961 961 1,707 1,707 884 884 970 970 4,522 4,522 4,522 4,522
R2 0.03 0.47 0.03 0.47 0.07 0.40 0.02 0.50 0.08 0.49 0.08 0.49
F Statistic 33*** 426*** 48*** 754.85*** 68.5*** 296*** 19*** 476*** 100*** 858*** 58*** 536***
*p<0.05; **p<0.01; ***p<0.001
27
Published: 19th of May 2022 OpenPsych
Table 13: Regression models of Raw h-Index.
Anthropology Psychology Political Science Economics All disciplines
Model
Number
(1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
Sex
Female = 1
Male = 0
-7.53***
(1.23)
-3.23***
(0.96)
-7.21***
(1.23)
-3.30***
(0.93)
-8.53***
(1.09)
-3.92**
(0.93)
-5.60***
(1.35)
-2.67**
(1.02)
-7.24***
(0.64)
-3.03***
(0.50)
-7.21***
(1.04)
-3.94***
(0.80)
Years Pub-
lishing
1.03***
(0.04)
1.44***
(0.04)
0.87***
(0.04)
1.30***
(0.05)
1.21***
(0.02)
1.21***
(0.02)
Anthropology
-8.69***
(0.84)
-11.33***
(0.65)
-8.54**
(1.15)
-11.7****
(0.89)
Economics
-11.75***
(0.85)
-7.79***
(0.66)
-12.21***
(1.04)
0.14**
(0.04)
Political Sci-
ence
-10.56***
(0.87)
-13.68***
(0.68)
-10.03***
(1.14)
-14.42***
(0.88)
Sex
×
An-
thropology
-0.32 (1.70) 1.45 (1.31)
Sex
×
Eco-
nomics
1.61 (1.81) 1.07 (1.40)
Sex
×
Politi-
cal Science
-1.32 (1.78) -1.81 (1.47)
Constant
31.3***
(0.86)
2.55 (1.31)
39.87***
(0.79)
3.68**
(1.19)
29.85***
(0.70)
4.85***
(1.37)
27.67***
(0.73)
-0.52 (1.18)
29.89***
(0.58)
9.11***
(0.72)
30.87***
(0.67)
9.44***
(0.76)
Observations
935 935 1,612 1,612 836 836 936 936 4,319 4,319 4,319 4,319
R2 0.04 0.43 0.02 0.45 0.07 0.37 0.02 0.45 0.08 0.45 0.08 0.45
F Statistic 38*** 359*** 34*** 647*** 61*** 245*** 17*** 379*** 95*** 714*** 55*** 446***
*p<0.05; **p<0.01; ***p<0.001
28
Published: 19th of May 2022 OpenPsych
Table 14: Regression models of Log10 Publication Count, Standardised as Z score.
Anthropology Psychology Political Science Economics All disciplines
Model Number (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
Sex
Female = 1
Male = 0
-0.36***
(0.06)
-0.12***
(0.05)
-0.29***
(0.03)
-0.13***
(0.04)
-0.53***
(0.07)
-0.20**
(0.05)
-0.23***
(0.07)
-0.06
(0.05)
-0.34***
(0.03)
-0.34***
(0.02)
-0.29***
(0.05)
-0.12***
(0.04)
Years Publish-
ing
0.06***
(0.002)
0.06***
(0.002)
0.06***
(0.002)
0.07***
(0.002)
0.06***
(0.001)
0.06***
(0.001)
Anthropology
0.03 (0.04)
-0.11***
(0.03)
0.06 (0.06)
-0.12***
(0.04)
Economics
-0.04
(0.04)
0.16***
(0.03)
-0.05
(0.05)
0.15**
(0.04)
Political Science
-0.002
(0.04)
-0.16***
(0.03)
0.10 (0.05)
-0.13**
(0.04)
Sex
×
Anthro-
pology
-0.07
(0.08)
0.02 (0.06)
Sex
×
Eco-
nomics
0.07 (0.09) 0.04 (0.06)
Sex
×
Political
Science
-0.34**
(0.09)
-0.08
(0.06)
Constant
31.3***
(0.86)
2.55 (1.31)
39.87***
(0.79)
3.68**
(1.19)
29.85***
(0.70)
4.85***
(1.37)
27.67***
(0.73)
-0.52
(1.18)
29.89***
(0.58)
9.11***
(0.72)
30.87***
(0.67)
9.44***
(0.76)
Observations 935 935 1,612 1,612 836 836 936 936 4,319 4,319 4,319 4,319
R2 0.04 0.43 0.02 0.45 0.07 0.37 0.02 0.45 0.08 0.45 0.08 0.45
F Statistic 38*** 359*** 34*** 647*** 61*** 245*** 17*** 379*** 95*** 714*** 55*** 446***
*p<0.05; **p<0.01; ***p<0.001
29
Published: 19th of May 2022 OpenPsych
Table 15: Regression models of Log10 Citation Count, Standardised as Z scores.
Anthropology Psychology Political Science Economics All disciplines
Model Number (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12)
Sex
Female = 1
Male = 0
-0.34***
(0.06)
-0.12*
(0.05)
-0.25***
(0.03)
-0.09*
(0.04)
-0.43***
(0.07)
-0.14**
(0.05)
-0.25***
(0.07)
-0.10 (0.06)
-0.31***
(0.03)
-0.11***
(0.02)
-0.25***
(0.05)
-0.10*
(0.04)
Years Publish-
ing
0.05***
(0.002)
0.06***
(0.002)
0.05***
(0.003)
0.07***
(0.003)
0.06***
(0.001)
0.06***
(0.001)
Anthropology 0.03 (0.04)
-0.10***
(0.03)
0.06 (0.06)
-0.10**
(0.04)
Economics
-0.04 (0.04)
0.15***
(0.03)
-0.03 (0.05)
0.16**
(0.04)
Political Science
-0.002
(0.04)
-0.15***
(0.03)
0.07 (0.05)
-0.14**
(0.04)
Sex
×
Anthro-
pology
-0.09 (0.08) -0.00 (0.06)
Sex
×
Eco-
nomics
0.00 (0.09)
-0.03 (0.07)
Sex
×
Political
Science
-0.17*
(0.09)
-0.02 (0.07)
Constant
0.17***
(0.05)
-1.34***
(0.07)
0.10**
(0.03)
-1.37***
(0.05)
0.17***
(0.04)
-1.39***
(0.09)
0.07 (0.04)
-1.37***
(0.06)
0.13***
(0.03)
-1.35***
(0.03)
0.10**
(0.03)
-1.35***
(0.04)
Observations 935 935 1,612 1,612 836 836 936 936 4,319 4,319 4,319 4,319
R2 0.03 0.43 0.02 0.44 0.04 0.35 0.01 0.42 0.02 0.41 0.02 0.41
F Statistic 28*** 353*** 25*** 631*** 38*** 221*** 12*** 334*** 25*** 606*** 15*** 379***
*p<0.05; **p<0.01; ***p<0.001
30
