Whereas research on gender and politics has been historically marginalized within political science, the subfield has grown over time, and explicit overtures have been made to remedy the underrepresentation of this work in the discipline’s top journals. For instance, the new all-women editorial team at the American Political Science Review (APSR) stated their commitment “to responding to the concerns of many colleagues—including women, people of color, scholars of race, gender, and sexuality, and scholars who employ qualitative methods—who feel that the APSR has been unreceptive to them and to their work” (APSR 2020).
Recent work explores the descriptive representation of women in political science by examining authorship and citation patterns in academic journal articles. We introduce a new dataset of gender and politics research articles (Barnett et al. Reference Barnett, FitzGerald, Krumbholz and Lamba2022) that makes possible analysis of the substantive and symbolic representation of gender in the discipline through examination of the topics typically covered in such work and in which outlets it appears. The dataset includes research published in 37 political science journals through 2019. Our data facilitate research on the relationships between these different forms of gender and political science research representation over time and across journals. Because gender-related research is conducted disproportionately by women, patterns of publication of gender-related research matter for the gendered composition of the discipline.
Following a brief history of gender and politics research within the discipline, we describe the data-collection, cleaning, and coding procedures that we used to construct a comprehensive dataset of related journal articles from 37 political science journals. Additional details about the data-collection and coding process are in the online appendix that accompanies this article. We discuss theoretical and methodological challenges raised during this process, how we addressed them, and what opportunities they create for further research. Finally, we present findings from our dataset on the number and proportion of gender-related articles appearing in each journal and how they have increased over time. We conclude by previewing additional research that we plan to conduct using this dataset, which will be a public resource for the benefit of scholars working on substantive gender and politics research questions and those interested in examining trends in this subfield.
GENDER RESEARCH IN POLITICAL SCIENCE
Women’s representation within the discipline of political science was sparse until the 1960s. Between the 1960s and the 1980s, feminist women political scientists pursued both institutional change in the American Political Science Association (APSA) and research agendas that focused on women, gender, and politics (Tolleson-Rinehart and Carroll Reference Tolleson-Rinehart and Carroll2006). In 1980, the journal Women & Politics—later renamed Journal of Women, Politics, and Policy—began publishing some of this research and, in 1986, APSA’s Organized Section on Women and Politics Research was formed. Three additional gender-focused journals have emerged since then: International Journal of Feminist Politics in 1999, Politics & Gender in 2005, and European Journal of Politics and Gender in 2018.Footnote 1 Furthermore, women and politics became a formal field of doctoral study in Rutgers University’s political science program in 1986 (Tolleson-Rinehart and Carroll Reference Tolleson-Rinehart and Carroll2006).
Early research on gender within political science began with critiques of political science scholarship for excluding women as political actors and research that merely incorporated women into existing frameworks within the discipline—dubbed the “add-women-and-stir” approach (Carroll and Zerilli Reference Carroll, Linda and Finifter1993). Subsequent research argued that frameworks must be reconceptualized to accommodate the inclusion of women. Gender and politics scholars also increasingly shifted from using “sex” as a demographic variable to using “gender” as an analytic category (Carroll and Zerilli Reference Carroll, Linda and Finifter1993; Cassese, Bos, and Duncan Reference Cassese, Bos and Duncan2012; Hawkesworth Reference Hawkesworth2005; Tolleson-Rinehart and Carroll Reference Tolleson-Rinehart and Carroll2006). Such an approach “illuminates gender power and gendered institutions” rather than simply examining differences between women and men (Hawkesworth Reference Hawkesworth2005, 147). Without diminishing the importance of these distinctions, our data collection and analysis conducted for this article takes as broad a view as possible of what “counts” as “gender and politics research.” Parsing research that uses “gender” more analytically versus additively is an important question for future research that we introduce and describe the contours of through methodological choices involved in dataset construction.
A rich recent literature on patterns in publication and citation in political science by author gender finds that women tend to receive fewer citations than men (Atchison Reference Atchison2017; Maliniak, Powers, and Walter Reference Maliniak, Powers and Walter2013; Masuoka, Grofman, and Feld Reference Masuoka, Grofman and Feld2007; Mitchell, Lange, and Brus Reference Mitchell, Lange and Brus2013). Explanations suggested for lower citation counts among women scholars include low descriptive representation (Chibnik Reference Chibnik2014; Mitchell, Lange, and Brus Reference Mitchell, Lange and Brus2013); gendered topic or methodological preferences (Djupe, Smith, and Sokhey Reference Djupe, Smith and Sokhey2019; Ferber and Brün Reference Ferber and Brün2011; Maliniak, Powers, and Walter Reference Maliniak, Powers and Walter2013; Teele and Thelen Reference Teele and Thelen2017); gendered tendencies of self-promotion, including self-citation (Atchison Reference Atchison2017; Maliniak, Powers, and Walter Reference Maliniak, Powers and Walter2013); and gendered citation patterns via homophilous networks (Atchison Reference Atchison2018). The dataset we introduce facilitates analysis of publication and citation patterns within gender and politics scholarship rather than differences between men and women scholars of political science. However, such gender differences are relevant because the majority of gender and politics research is conducted by women. Although gender and politics research has worked its way into all of political science’s major subfields (Fox Reference Fox2011; Krause Reference Krause2011; Krook Reference Krook2011; Prügl Reference Prügl2011)—largely as a result of the influx of female scholars into all subfields, bringing gender questions with them—there are disparities in the amount and types of gender scholarship that have been produced across subfields. Ritter and Mellow (Reference Ritter and Mellow2000) argued that these disparities are the result of stark divisions and insularity among the subfields as well as between quantitative and qualitative research. More recently, this can be seen in the dramatic differences in the gender composition of the discipline’s subfields (APSA 2019). Our dataset facilitates systematic inquiry into the relationship between gendered citation practices and topic emergence and diffusion that underpins the descriptive and substantive representation of gender and politics in political science.
The dataset we introduce facilitates analysis of publication and citation patterns within gender and politics scholarship rather than differences between men and women scholars of political science.
DATA COLLECTION AND CODING
We collected original research articles from 37Footnote 2 peer-reviewed journals using the Web of Science (2021) and Scopus (2021) databases in July 2020. We restricted our sample to journals analyzed in recent similar investigations of substantive representation in political science journals (Cammett and Kendall Reference Cammett and Kendall2021; Wilson and Knutsen Reference Wilson and Knutsen2020), in journals that have an affiliation with a professional association, and in well-established journals focused on gender and politics. We gained a substantial benefit to the completeness of our dataset by drawing from both Scopus and Web of Science, given that more than 1,200 of the articles that we identified were indexed in only one of the two databases.Footnote 3 The final dataset contains metadata and abstracts for 5,487 articles published between 1913 and 2019. The online appendix elaborates on how the dataset was compiled and prepared for coding, and it provides an extended discussion of the operationalization of the different coding categories. The following discussion summarizes the inclusion and exclusion criteria used to hand-code each article by the contents of its abstract.
We included as unambiguous 3,083 original research articles that expressed a central focus on gender theoretically and/or empirically. These articles were motivated explicitly by some aspect of gender and their research and argument were organized around this concern. Unambiguous research included articles about women’s rights, gendered patterns of political participation and representation, gender identity, sexuality and reproductive health, LGBTQ+ rights, masculinities, and intersectionality, among many other topics of gender and politics.
We included as ambiguous 485 articles that mentioned something relevant to the study of gender and sexuality in the abstract but for which it was unclear whether the full research article maintained a theoretical focus on this dimension. Coding articles as ambiguous due to the author’s framing raised a methodological challenge related to authorial intent that this study cannot resolve but which future research should investigate: we questioned in such cases whether the authors’ primary interest was, in fact, the gendered dimensions of their topic but the articles nonetheless were framed around less explicitly gendered theories and research questions to facilitate publication. Particularly in instances in which the empirical case was clearly about a women’s organization or group but the framing of the article was entirely divorced from gender considerations, it seems plausible that this may have occurred. Whereas research on gender increasingly is mainstreamed within the discipline of political science, this was not always the case. Authors may have strategically chosen to downplay the gender-related theoretical import of their work to ensure that it was more easily accepted in disciplinary conferences and publications.
A subset of the ambiguous category was composed of articles that mention sex or gender control variables in the abstract. In some articles, it appeared that “gender differences” was an alternative explanation for an outcome that the authors sought to contrast with their favored explanation. In other cases, gender was mentioned among a “laundry list” of control variables, with little indication that it was of any theoretical interest. Given our decision to read only article titles and abstracts, our ability to distinguish between these two case types was limited. As a result, we adopted a deliberately inclusive approach that extended the ambiguous category to research that might have few or no implications for gender research.
Finally, given our own substantive interests, the research team was highly cognizant throughout the process of the temptation we faced to “read” gender into various topics. For some topics, such as care work—even in cases in which gender was not explicitly mentioned—we believed that viewing it as a gendered topic was justifiable given the typically strong association between care work and gendered social roles. For other topics—such as research into personality traits associated with individuals in elected office or positions of power—we refrained from imposing a gendered lens on the topic when the authors did not clearly have one themselves. In general, we attempted to adhere as closely as possible to a “literal” reading of article abstracts, setting aside our own proclivities to see gender implications across a wide range of topics in favor of coding articles from the information explicitly presented by authors. In summary, the articles included via the ambiguous category spanned a wide range of possible connections to gender and politics. This invites further research into the relationships among authorial intent; journal prestige, aims, and scope; and research methods.
We excluded by content 1,796 articles that had no explicit relevance to or focus on gender-related issues. These articles often were captured in our database search due to keywords with overlapping meanings. For example, the word “engender” appeared in many abstracts that were excluded due to lack of gender-related content. Some items that we excluded might be considered ambiguous cases by other scholars. When a gendered dimension of analysis was not apparent, we excluded articles focused solely on children, spouses or couples, and families. We also excluded records in which the only gendered component was that the population under study was exclusively male or female for reasons incidental to the research focus—for example, studies that investigated outcomes among men subject to the draft during the Vietnam War, a population of interest largely because the draft was random.
We excluded by type 126 book reviews, literature reviews, bibliographies, critical responses to or forums on existing published work, introductions to special issues of journals, and personal reflections. Although we recognize that these forms of scholarship contribute to the body of political scientific knowledge, we limited our inquiry to original research articles because publication of these types is the primary focus of graduate training, hiring and promotion committee evaluations, and the core product of journals that are cited and inform future research. Articles excluded by content or type are included in the dataset made available with this publication but are excluded from the following analysis.
TRENDS IN GENDER RESEARCH PUBLICATION
This section analyzes trends in the publication of gender research between 1980 and 2019Footnote 4 in political science journals not explicitly oriented toward gender research.Footnote 5 Table 1 displays the total number of articles from non-gender–dedicated journals coded as either unambiguous or ambiguous gender research, as well as the proportion of the total number of articles that gender research represents.Footnote 6 The table also lists full journal titles and the corresponding abbreviations that are used throughout this section.
Notes: The combined count is the sum of articles coded as unambiguous or ambiguous. The proportion is the quotient of this combined count and the estimated total number of articles published by the journal (see the online appendix for details).
Aggregating over time by journal, PRQ appears to be the most prolific publisher of gender and politics research among non-gender–dedicated journals in terms of total count and proportion, with a total of 234 articles published and an overall proportion of 10.8%. Indeed, PRQ is the only journal in our sample to rank in the top five in both the raw number of gender-related research articles published and as a proportion of total articles. In general, however, the distribution is heavily skewed. Eleven of the 33 journals publish gender research less than 3% of the time and only three journals more than 10%. Notably, the volume of gender research published does not necessarily correlate with its proportion. Gender and politics research published by JOP and AJPS accounted for about 4% of all articles from each journal despite being in the top five of total gender research articles. However, JREP—as a relatively new journal—published a total of only six gender research articles to date, yet this accounted for 21% of its research articles published through 2019.
On a yearly basis, the estimated publication rate of gender research articles displays substantial variation within a limited range. On average, non-gender–dedicated journals published 1.71 gender research articles per year, with a standard deviation of 1.14. All but one journal (i.e., PRQ) averaged less than four such articles published per year. Ten journals published at an estimated rate of less than one gender research article per year and the next 14 journals at less than two per year.Footnote 7 Excluding PRQ, non-gender–dedicated journals published, on average, 1.58 gender research articles per year, with a standard deviation of 0.876.
Examining the time trends in publication of gender research across journals revealed a general trend toward an increase in published gender and politics research in recent years. Figure 1a displays the counts of gender research articles published each year when including gender-dedicated journals and when restricting the data to non-gender–dedicated journals. Since the mid-1990s, there has been a notable increase in gender and politics research published in political science journals and, during the past decade, this increase has been especially dramatic. However, figure 1a also demonstrates that much of the growth over time has been driven by the increased output of the four gender-dedicated journals included in our sample. Figure 1b highlights that the growing number of published gender-related articles in absolute terms appears to be driven by an overall expansion in publication volumes, given that the proportion of articles published in the journals examined that we code as gender research remained relatively flat over time.
Time trends for individual journals identified which outlets are driving these overall trends. Historically, journal prestige has been negatively correlated with the publication of gender and politics research. The correlation coefficient for the journal H-index (i.e., a measure of productivity and citation count) and the overall proportion of articles related to gender is -0.365 (p<0.05). The upward trend of greater representation for gender research in top subfield and generalist political science journals is an important departure from this negative relationship. As shown in figure 2a, some of the journals with the clearest “upticks” in recent years—AJPS, BJPS, CPS, JOP, NPS, POP, PB, and PRQ Footnote 8—are among the most prestigious general-interest or subfield journals in the discipline, which suggests that gender research has become increasingly “mainstream” in terms of the outlets where it is published. However, as shown in figure 2b, the increasing volume of gender research published in these journals generally is not indicative that gender research comprises an increasing proportion of published research articles. As noted previously, the average counts and proportions confirm the historical difficulty of publishing gender research in political science, and they highlight the critical role that gender-dedicated journals have in facilitating gender research output.
Historically, journal prestige has been negatively correlated with the publication of gender and politics research.
Research related to gender was almost nonexistent in political science journals as recently as 1980. Since then, and particularly in the past decade, the volume of research related to gender has grown substantially across a wide range of journals, including several of the discipline’s top general-interest journals. The dataset we constructed of all unambiguously and ambiguously gender-related articles across 37 political science journals demonstrates this growth; the substantial variation across subfields (as represented by their flagship journals) in the quantity of gender-related research that is published; and the relatively consistent proportion of overall research articles that gender-related research comprises.
…particularly in the past decade, the volume of research related to gender has grown substantially across a wide range of journals, including several of the discipline’s top general-interest journals.
Future research using this dataset will (1) investigate the evolution of the topics addressed in gender-related articles over time and across journals using computational topic modeling methods; (2) map the structure of scholarly networks within the gender and politics subfield, using methods of network and citation analysis;Footnote 9 (3) investigate how social media is shaping scholarly conversation in the gender and politics subfield, by combining traditional citation analysis with analysis of engagement with and circulation of gender and politics research on Twitter, which has become an important venue for the promotion of academic research (Klar et al. Reference Klar, Krupnikov, Ryan, Searles and Shmargad2020); and (4) conduct a systematic review of research in the subfield. We also plan to build a machine-learning model trained on our hand-coded dataset that automatically will tag future articles in the field of gender and politics research to add them to the database as they are published.
The full dataset described herein is available with the publication of this article, providing a resource for further investigating publication patterns and trends in gender and politics as well as for undertaking literature reviews on subjects within this field. This will expand and diversify the citation of existing research.
The authors are listed in alphabetical order and contributed equally. The authors acknowledge the hosts of and other participants in the Summer Institute in Computational Social Science at Rutgers University (2020), during which this project was conceived. We also thank Kathleen Rogers for her contribution to the project in its early stages.
DATA AVAILABILITY STATEMENT
Research documentation and data that support the findings of this study are openly available at the Harvard Dataverse at https://doi.org/10.7910/DVN/UVWRTV.
To view supplementary material for this article, please visit http://doi.org/10.1017/S1049096522000385.
CONFLICTS OF INTEREST
The authors declare that there are no ethical issues or conflicts of interest in this research.