Women are underrepresented in the tables of contents of the political science profession’s most prestigious journals (Breuning and Sanders Reference Breuning and Sanders2007; Evans and Moulder Reference Evans and Moulder2011; Teele and Thelen Reference Teele and Thelen2017). Because men are more likely to publish in these journals, they gain greater visibility, which leads to gender gaps in citation patterns (Dion, Sumner, and Mitchell Reference Dion, Sumner and Mitchell2018; Maliniak, Powers, and Walter Reference Maliniak, Powers and Walter2013) and underrepresentation of women on syllabi (Colgan Reference Colgan2017). As long as scholars are evaluated by the quantity of publications in prestigious journals, the gender publication gap puts women at a disadvantage for tenure and promotion (Box-Steffensmeier et al. Reference Box-Steffensmeier, Cunha, Varbanov, Hoh, Knisley and Holmes2015), which may contribute to the discipline’s leaky pipeline.
Several recent studies published in PS examine potential sources of this gender publication gap but find no evidence of gender bias in the review process (Breuning et al. Reference Breuning, Gross, Feinberg, Martinez, Sharma and Ishiyama2018; König and Ropers Reference König and Ropers2018; Nedal and Nexon Reference Nedal and Nexon2018; Peterson Reference Peterson2018; Samuels Reference Samuels2018; Tudor and Yashar Reference Tudor and Yashar2018).Footnote 2 Instead, they find that submission pools are distorted by gender. Women, compared to men, comprise a far smaller share of authors submitting to most of the journals evaluated in their studies, and this submission gap starts among graduate students (Tudor and Yashar Reference Tudor and Yashar2018).Footnote 3 A large survey of political scientists’ professional activities (Djupe, Smith, and Sokhey Reference Djupe, Smith and Sokhey2019) similarly affirms that men report submitting more papers than women, especially among assistant professors and scholars with quantitative approaches.
To contribute to this important debate, we fielded an original survey to a sample of American Political Science Association (APSA) members. Specifically, we asked participants whether they are likely to submit to and get published in each of 13 political science journals, including American Journal of Political Science (AJPS); American Political Science Review (APSR); Comparative Political Studies (CPS); International Organization (IO); Journal of Political Science Education (JPSE); Journal of Politics (JOP); Perspectives on Politics (PoP); Political Analysis (PA); Political Behavior (PB); Political Research Quarterly (PRQ); Political Theory (PT); Politics, Groups, and Identities (PGI); and Polity.
Results reveal that the gender submission gap is accompanied by a gender perception gap at some but not all leading journals. Among journals with a gap, they run in both directions. Men state that they are more likely to submit work to and get published in some journals, such as APSR, AJPS, and PA. Women state that they are more likely to submit work to and get published in other journals, such as PGI. As this suggests, perception gaps do not always align with submission gaps. Existing studies show that submission gaps are consistent across journals, whereas the perception gaps that we find vary substantially.
After describing the survey design and our results, we speculate about the reasons for the gender perception gaps we find. Our data permit us to rule out a common claim: a gender gap in methods. Our results reveal perception gaps even among scholars with the same methodological (i.e., quantitative or qualitative) approach. In the conclusion, we argue that the alignment—or lack thereof—of perception gaps and submission gaps should inform journal editors’ outreach strategies.
SURVEY DESIGN
Are women more inclined than men to submit their manuscripts to some journals and less inclined to submit to others? Do women perceive the chances of publication differently than men? If so, to what extent can these differences be attributed to different methodological approaches?Footnote 4 To answer these questions, we invited a sample of 2,440 APSA members to participate in an online survey from October 10 to November 6, 2017.Footnote 5 We received 810 completed surveys, for a response rate of 33.2%. The journals included in our study are the two APSA flagship journals, APSR and PoP; each of the main journals of the major US regional political science associations, AJPS, JOP, Polity, and PRQ; and one important journal or two from different subfields of research and practice, CPS, IO, JPSE, PA, PB, PGI, and PT.
For each journal, we asked (displayed on the same page): “Thinking about [the name of a journal], what is...[t]he likelihood you will submit your work to this journal...[and t]he likelihood your work will be published in this journal?” Participants responded to these questions on five-point ordinal scales with the following response options: “very likely,” “somewhat likely,” “neither likely nor unlikely,” “somewhat unlikely,” and “very unlikely.” The journal order was randomized for each participant. At the end of the survey, we asked participants about their gender identity and primary research method, as well as several participant-specific questions and an open-ended question about the survey.
Are women more inclined than men to submit their manuscripts to some journals and less inclined to submit to others? Do women perceive the chances of publication differently than men? If so, to what extent can these differences be attributed to different methodological approaches?
For each likely–unlikely question for each journal, we measure the difference in average scores in three ways: (1) the difference between women and men; (2) the difference between quantitative and qualitative scholars; and (3) the difference between women and men, within each methodological orientation.Footnote 6
RESULTS
Figure 1 shows these results for APSR; detailed results for all other journals are included in figures C.1 through C.8 in the supplementary materials. Two bars in each panel represent average scores for men versus women or for quantitative versus qualitative scholars. The 95% confidence intervals are shown in error bars. We also test differences statistically and report the results above the bars. The numbers in parentheses are standard errors. The differences that are statistically significant at the 0.05 level are highlighted in black.
Figure 1 American Political Science Review
Note: The differences that are statistically significant at the 0.05 level are highlighted in black.
Regarding APSR, the top row of panels reveals that women state that they are less likely to submit their work and be optimistic about publication than men. The middle set of panels shows that the same is true for qualitative scholars. Most importantly, the bottom set of panels demonstrates that the gender perception gap persists even when we control for participants’ methodological approach. The only exception is in the case of women quantitative scholars, whose intention to submit is lower than quantitative scholars who are men; however, the difference is only barely insignificant at the 0.05 level (i.e., p<0.06).
Figure 2 summarizes the results for all journals. We focus on examining gender gaps within the same methodological orientation for each journal. Dots that are to the left (right) of the vertical dotted line indicate that women state that they are less (more) likely than men to submit a manuscript or believe that it will be published in the journal. The top two panels are for quantitative scholars; the bottom two are for qualitative scholars.
Figure 2 Results of Difference-of-Means Tests
Note: The effects that are statistically significant at the 0.05 level are highlighted in black.
Journals with a “Masculinized” Perception Gap
A few journals exhibit a “masculinized” perception gap, in that men are significantly more likely than women to state that they would submit a manuscript and/or be optimistic about their chances of publication, even within the same broad methodological approach. In addition to APSR, these journals include AJPS and PA. Regarding AJPS and PA, quantitative scholars are much more likely than qualitative scholars to submit their manuscripts and believe that they will be published (see figures C.1 and C.2 in the supplementary materials). Similar to APSR, however, we see gender perception gaps among quantitative scholars. Specifically, gender differences in the likelihood of submission—but not the likelihood of publication—are significant for AJPS. Both differences are significant and slightly larger for PA.
Journals with a “Feminized” Perception Gap
Views on two journals in our survey, PGI and JPSE, are characterized by gender perception gaps that run in the opposite direction. After controlling for methods, women scholars are more likely than men to submit manuscripts and to be optimistic about their chances of publication. These are journals with a “feminized” perception gap. The journal showing the clearest pattern is PGI (figure 3). Regardless of their methodological orientation, women are more likely to state that they will submit a manuscript and get it published at PGI. These effects are large and highly significant. Another journal with a similar pattern is JPSE (see figure C.3 in the supplementary materials), in which qualitative women scholars are more likely to state that they will submit their work and get it published compared to qualitative men scholars. There is no gender gap among quantitative scholars.
[T]he gender perception gap persists even when we control for participants’ methodological approach.
Figure 3 Politics, Groups, and Identities
Note: The differences that are statistically significant at the 0.05 level are highlighted in black.
Journals with No Gender Perception Gap
At most other journals in our survey, no gender perception gap exists among quantitative or qualitative scholars. At CPS, JOP, PB, and PRQ, however, there is a methods gap: quantitative scholars are significantly more likely to state that they will submit and get published (see figures C.4 through C.7 in the supplementary materials). At IO, quantitative scholars are significantly more likely than qualitative scholars to state that they will submit their manuscripts (see figure C.8 in the supplementary materials). However, there is no gender perception gap within the methodological approach at any of these journals. By contrast, non-quantitative scholars state that they are more likely to submit and get published in PoP, PT, and Polity (see figures C.9 through C.11 in the supplementary materials). However, for the most part, there is no gender gap when we hold methods constant.Footnote 7
Robustness Tests
Figure 2 shows the association between participants’ gender identity and the outcome variables among quantitative and qualitative scholars. Readers may wonder whether some of the gender perception gaps could be attributable to other factors. As a robustness test, we run our analysis using additional control variables, including participants’ subfield, position, and institution type (see supplementary materials A). We add each categorical variable to regression analysis as a set of dichotomous variables with one category as the baseline. We also include our main variables—each participant’s gender identity (i.e., man or woman), methodological approach (i.e., quantitative or not), and their interactions—in the estimation. Results reported in figure C.12 in the supplementary materials demonstrate that our findings remain robust.Footnote 8
DISCUSSION
These results suggest the existence of gender perception gaps for some but not all journals but do not directly answer two important questions. We discuss them in turn.
What Explains Perception Gaps?
What underlies the perception gaps of these journals? Several studies argue that gender publication and submission gaps are related to gender gaps in methods. Women are underrepresented at some journals, the argument claims, because these journals emphasize quantitative articles. That is, because women’s work tends to be relatively more qualitatively oriented than men’s, women prefer to avoid quantitatively-oriented journals and submit their work elsewhere (Breuning and Sanders Reference Breuning and Sanders2007; Evans and Moulder Reference Evans and Moulder2011; Teele and Thelen Reference Teele and Thelen2017).
Do men and women differ in their choice of methods? Our survey asked participants about their primary methodological approach. We find that 59% of women and 57% of men select some type of quantitative method as their main approach. Figure 4 shows the distribution of methodological approaches by gender among survey participants using the more-refined categories set by the APSA publications committee.Footnote 9 The high share of women who engage in various types of quantitative research makes it unlikely that women’s underrepresentation in some quantitatively oriented journals is due exclusively to their choice of method. As demonstrated in our analysis, some of the gender perception gaps persist among scholars within the same methodological orientation. In other words, gender matters even after controlling for methods.
Figure 4 Distribution of Methodological Approaches by Gender
What could be other reasons for the perception gap? Because women tend to be underrepresented among journal editors and editorial boards, women lead editors at journals could serve as role models and send welcoming signals to potential women authors (Topaz and Sen Reference Topaz and Sen2016). In our sample, however, the gender identity of the lead editor is not clearly associated with differences between journals with “masculinized” perception gaps (e.g., APSR and AJPS) and those with no gender perception gap (e.g., JOP and CPS). All of these journals had men lead editors at the time our survey was in the field, and three had women lead editors in the not-too-distant past.
Another possible explanation centers on journal prestige and perceived risk associated with submission. Djupe, Smith, and Sokhey (Reference Djupe, Smith and Sokhey2019) find that women are more likely than men to state that they submit their work to the journal most likely to accept it, whereas men are more likely than women to state that they first submit their work to a top-tier journal. Gender differences in risk aversion thus imply that women initially may be less willing than men to submit their work to journals with high rejection rates.Footnote 10
In our view, gender differences in risk aversion are the likely product of socially conditioned publication strategies. Individual decisions about where to submit, how often—and whether and where to resubmit a paper that has already been rejected—tend to be shaped by interpersonal contacts and information communicated through social networks, as well as habits acquired through collaborative relationships, which often differ by gender.
For example, Djupe, Smith, and Sokhey (Reference Djupe, Smith and Sokhey2019) show that men and women are equally likely to collaborate but that men gain many more submissions and publications as a function of collaboration. This pattern suggests that something happens to men in their collaborative networks that bolsters their likelihood to submit papers and to resubmit previously rejected papers. Younger men scholars in highly productive collaborative groups may be more likely to receive encouragement for and to develop early habits of frequent article submission. They also may feel more inoculated against the morale-damaging effects of rejection from top-tier journals.
Do Perception Gaps Explain Submission Gaps?
Gender publication and submission gaps exist at most journals (Djupe, Smith, and Sokhey Reference Djupe, Smith and Sokhey2019; Teele and Thelen Reference Teele and Thelen2017), but not all journals have gender perception gaps. At some journals, the gaps align: APSR, for example, has both a publication and a submission gap—women authors comprised an average 23% to 24% of authors of published and submitted articles in the 2000s (König and Ropers Reference König and Ropers2018; Teele and Thelen Reference Teele and Thelen2017)—as well as a perception gap that we reveal in this article. The same is likely true at AJPS and PA, although we lack data on submissions by gender. By contrast, although the editors of both CPS and PB acknowledge gender submission gaps (Peterson Reference Peterson2018; Samuels Reference Samuels2018), our study does not find any difference, by gender, in the perceived likelihood of submitting to either journal.
As this suggests, differences in attitudes toward journals do not always align with differences in behavior. For some journals, women and men may state that they are equally likely to submit their manuscripts, even though men ultimately submit more papers. In these cases, the gender submission gap may stem from habit rather than gender differences in preferences or aversions to a journal. However, in other cases, men are more likely than women to state that they will submit a manuscript to a journal and actually do so. Whether or not attitudes toward journals align with actual behavior may have implications for strategies to close the gender publication gap, as discussed in the conclusion.
CONCLUSION
Our analysis generates several important findings about gender perception gaps. First, gender gaps characterize perceptions of some but not all journals. Second, gender gaps run in both directions, with men more favorable to some journals (e.g., APSR) and women to others (e.g., PGI). Third, perception gaps cannot be explained primarily by gender differences in methodological orientations. Men and women within quantitative and non-quantitative research communities tend to have different perceptions of some journals in our study—and likely more throughout the discipline. Fourth, although submission gaps are relatively constant across journals, perception gaps are not. In other words, behavior is not always aligned with attitudes.
To promote diversity and inclusion in the profession, journal editors should try to alter the gender composition of the submission pools, which will improve gender balance among published authors as long as review processes remain unbiased. The alignment—or lack thereof—between submission and perception gaps may be consequential for journals’ outreach strategies. When gender gaps in submissions are not matched by gender gaps in attitudes toward a journal, editors simply may need to change habits and encourage men or women (depending on the direction of the perception gap) to submit more. When gender gaps in submissions align with gender gaps in perceptions, outreach also may need to focus on changing the substance of journals’ reputations. Expectations about potential biases in the types of articles that a journal publishes could affect scholars’ decisions to submit (Breuning et al. Reference Breuning, Gross, Feinberg, Martinez, Sharma and Ishiyama2018; Nedal and Nexon Reference Nedal and Nexon2018).
[A]lthough submission gaps are relatively constant across journals, perception gaps are not. In other words, behavior is not always aligned with attitudes.
As Breuning et al. (Reference Breuning, Gross, Feinberg, Martinez, Sharma and Ishiyama2018) point out, however, editors can do only so much to influence impressions. Scholars in multiple positions throughout the discipline—advisers, colleagues, and peers—also play an important role in shaping views about what reviewers and journal editors want to publish as well as what type (and whose) work is worth submitting where. Promoting greater gender balance in submission pools requires a collective effort in which we all encourage one another to take risks and to remain resilient when we are rejected. At the same time, political science should join other scholarly disciplines in the movement to evaluate research primarily by its substance, novelty, and rigor rather than narrower metrics such as journal impact factors and numbers of citations (Heckman and Moktan Reference Heckman and Moktan2018; Jensenius et al. Reference Jensenius, Htun, Samuels, Singer, Lawrence and Chwe2018).Footnote 11
SUPPLEMENTARY MATERIAL
To view supplementary material for this article, please visit https://doi.org/10.1017/S1049096519001227
