Introduction
Historically marginalized within the discipline of political science, research on gender and politics has become increasingly mainstream — both numerically and in terms of the placement of research articles in top field journals — over time (Barnett et al. Reference Barnett, FitzGerald, Krumbholz and Lamba2022; Barnett et al. Reference Barnett, FitzGerald, Krumbholz and Lamba2025). While attention to gender has grown in general-interest journals, journals specifically committed to advancing knowledge about how gender operates in and through political phenomena, including Politics & Gender, have created spaces where such research does not have to compete with all other topics examined by political scientists, or justify the relevance of gendered or feminist theoretical approaches or empirical phenomena related to gender. To date, however, no one has quantified the relative contribution of Politics & Gender and similar journals to the body of scholarship on gender and politics, or evaluated how different journals have contributed to types of gender research published in political science. How has the volume and content of research articles published in Politics & Gender evolved over time, and how does it compare to trends in gender and politics research published in other leading journals? To address these questions, in this article, we analyze a comprehensive dataset of metadata from original research articles substantively related to gender and politics published in 37 political science journals, including Politics & Gender, through the end of 2023.
The data come from 33 general-interest or subfield-specific political science journals (henceforth “non-gender-dedicated” journals) and four journals focused specifically on gender and politics (“gender-dedicated” journals). We constructed the dataset using a combination of theoretically informed hand-coding of systematically gathered abstracts available through 2019 (Barnett et al. Reference Barnett, FitzGerald, Krumbholz and Lamba2022, Reference Barnett, FitzGerald, Krumbholz and Lamba2025), machine learning models trained on the human-coded data to add information about articles’ methodologies and incorporate work published through the end of 2023, and computational topic modeling techniques to identify the substantive topics most prevalent in the published research. Our analysis focuses on original research articles and excludes other types of publications in Politics & Gender, such as Critical Perspectives pieces and “Notes from the Field.”
In this article, we first briefly describe the methodologies used to construct the original dataset covering research published through 2019, update it with articles published from 2020–23, and code the primary methodological approach of each article. We then analyze basic patterns in the relative number of articles published in Politics & Gender and other journals over time, comparing trends in the publication of articles using primarily quantitative or primarily qualitative methodologies. Finally, we present the results of a computational topic modeling process, which inductively categorizes the titles and abstracts of gender-related research articles in our dataset. While the topic modeling algorithm struggled to classify a large percentage of the gender research articles, the common topics that it did identify enable us to draw comparisons across journals.
We find that the volume of research published by Politics & Gender has increased over time, as has the volume of gender-related research published in numerous general-interest political science journals. In fact, in the dataset overall, the annual number of published gender and politics research articles tripled from the mid-2000s, when Politics & Gender was founded, to 2023. While Politics & Gender has always published a mix of qualitative and quantitative research articles, qualitative work was more prevalent than quantitative work for much of the journal’s history. That has changed since 2019, after which the journal has published more quantitative work than qualitative work. This reflects broad patterns in political science publishing of work about gender and politics, which has tended to be more quantitative than qualitative, although the volume of both types of work published has grown in recent years. Patterns in Politics & Gender also resemble those in another gender-dedicated journal, The Journal of Women, Politics and Policy (JWPP). In contrast, the other gender-dedicated journals in our dataset (Social Politics [SP] and the International Feminist Journal of Politics [IFJP]) have tended to publish more qualitative research.
The most common topics addressed in gender research articles varied across journals. In Politics & Gender, JWPP, and non-gender-dedicated journals (pooled together), the most common research subjects related to women running for office (including work on ambition, stereotypes, quotas, and other aspects of women’s candidacies) and the nature of women’s political representation in legislative and executive office. Other common topics included feminist theory and concepts, women’s movements, and various dimensions of the political gender gap. In contrast, SP and the IFJP featured more research on care work, the diffusion of equality norms, and conflict, among other topics.
We conclude by discussing the implications of the topic modeling algorithm’s failure to classify many of the research articles into the topics that the algorithm identified. Examining the content of some of these unclassified articles that were published in Politics & Gender, we argue that the high volume of “unclassifiable” research both reflects the inclusivity of our data-gathering procedure and is a testimony to the diversity and nuance of topics studied by scholars of gender and politics. This article thus demonstrates the contribution that Politics & Gender has made to advancing research on gender within political science not only through the volume of scholarship it has published, but through the diversity of methods, theoretical approaches, and topics addressed in that scholarship. Ultimately, Politics & Gender’s commitment to methodological pluralism and diverse conceptions of “gender and politics” research has facilitated the growth of a subfield and its integration into the discipline while also compensating for the comparatively narrow range of gender and politics research articles published in general-interest journals.
Data & Methods
Data Preparation
We draw on an original dataset compiled in 2020 and updated in 2024 that includes metadata on 4,504 research articles published in 37 political science journals that can substantively be considered “gender and politics” research. Of these articles, 3,273 were published during the period when Politics & Gender has existed (2005–23). We explained the procedure for compiling this dataset in detail in a 2022 article published in PS: Political Science and Politics (Barnett et al. Reference Barnett, FitzGerald, Krumbholz and Lamba2022, Reference Barnett, FitzGerald, Krumbholz and Lamba2025) and its supplementary material. In brief, we compiled the original dataset, covering work published through 2019, by first identifying relevant journals and then conducting wide-ranging keyword searches in the Web of Science and SCOPUS databases for items published in those journals classified as “research articles” (versus, for example, book reviews, symposia introductions, etc.). We exported metadata for all of these items into a dataset, for which we then undertook the following cleaning and categorizing steps:
-
1. We identified and removed duplicate items.
-
2. We identified items missing abstracts and checked to see whether those abstracts existed on the journals’ own websites.
-
3. Where abstracts were missing, we used a text-summarizing function in Python to scan the article text and generate an “auto-abstract,” when feasible.
-
4. As a research team, we read every title and abstract and hand-coded all items in the dataset, assigning them to one of the following categories: unambiguous gender research; ambiguous gender research; excluded by content, meaning we judged the article was not actually gender research; and excluded by type, meaning the item was not an original research article. Items in the former two categories were ultimately included in our dataset as “gender research,” which we define in expansive and inclusive terms.
Each item was assigned to two coders, and discrepancies were adjudicated by assigning a third coder or eventually reviewing the item as a whole team. Our 2022 article discusses the “ambiguous” and “unambiguous” categorizations at length; in essence, these categories distinguish between articles that have something to do with gender at least tangentially and those we theoretically believe qualify as “gender and politics” research. At the end of this process, we had identified 3,568 gender and politics research articles published between 1913 and 2019. Our 2022 article presented the dataset and analyzed some basic trends in this data across journals.
For this article, we updated the dataset, gathering metadata on articles published in the same set of journals from 2020 through 2023. After completing the deduplication process and eliminating from this set items that were either not actually research articlesFootnote 1 or were missing abstracts (11 articles), we were left with 1,479 new research articles.
Supervised Machine Learning
After our initial data collection, we decided to additionally code whether articles’ primary method of analysis was quantitative or qualitative.Footnote 2 Two coders hand-coded a sample of 400 abstracts from 1978–2019, with disagreements resolved through additional coding rounds or team discussions. We then used a machine learning model, with this set of abstracts as a training dataset, to identify whether the remaining articles (in both the old and new data) should be considered primarily quantitative or qualitative. The Appendix contains details about the machine learning model and its output. The results of this classification are presented in the next section.
In addition, after gathering the new metadata from 2020–23, to determine which among these articles should be considered “gender research” under the definition we applied during our hand-coding of the original dataset, we applied supervised machine learning again. We consolidated our previous coding categories, combining both “unambiguous” and “ambiguous” gender research into a new unified “gender research” category; all other items were marked as “not gender research.” This data was used to train the machine learning classifier. We split the data 80-20, where the training dataset consisted of 3,944 articles and the test data consisted of 987 articles. We tested five classifiers on the articles’ abstracts for the classification task and found that the AdaBoost classifier performed the best for our data in comparison to others (see Appendix Figure A.1 and Table A.1). We applied our machine learning model to the abstracts from the new articles gathered (published from 2020–23) to predict if they should be considered gender research or not gender research. Of articles in the new dataset, 1,131 were classified as gender research, whereas 348 were classified as non-gender research.
We validated these results two different ways. First, we checked the classification of articles published between 2020–23 in the gender-dedicated journals, which we assume are gender-related. Of the 181 research articles published by Politics & Gender during 2020–23 in our dataset, all were correctly classified as gender research. The model also correctly classified more than 97% of the research articles in the other gender-dedicated journals as gender research.
Second, we reviewed the abstracts and classification of a randomly selected 10% (N = 89) of articles from non-gender-dedicated journals in the new data. Of these 89 randomly selected articles, 59 were classified by the model as gender research, and 30 as non-gender research. Overall, we found that our hand-coding procedure would have produced a different coding for six of these articles (approximately 7% of the validation sample): two articles classified as gender research should not have been, according to our criteria, while four articles classified as non-gender research should have been classified as gender research. Notably, all four such articles would have fallen into the category of “ambiguous” gender research that we included in our dataset only to be as inclusive as possible. Specifically, they are all cases of articles whose focus theoretically and empirically is not gender, but which highlight findings related to sex or gender within the abstract. We thus conclude that while further manual review of the dataset will be required before we can claim to have correctly classified every individual article, we have confidence that the aggregate data presented below describing trends in the publication of gender research are generally accurate. If anything, the results presented here may modestly undercount articles that we would classify as gender research.
Topic Modeling
Once we classified all the articles as gender or non-gender research, we used topic modeling to identify the latent semantic concepts among the set of articles identified as gender research. We used the BERTopic algorithm (Grootendorst Reference Grootendorst2022) to perform topic modeling. The BERTopic algorithm identifies coherent topics in the corpus of study through a class-based Term Frequency-Inverse Document Frequency (TF-IDF) procedure. Figure 1 shows the three main steps to create the topic representation using the BERTopic algorithm. Initially, each document is transformed into an embedding using a pretrained language model. For this study, we used the paraphrase-MiniLM-L12-v2 sentence transformer. Next, the dimensionality of these embeddings is reduced to improve the efficiency of the clustering process using UMAP and HDBSCAN. Finally, topic representations are derived from the document clusters using a class-based version of TF-IDF. We present the results of this analysis and discuss its limitations later in the article.
Results
Publication Trends: Quantity and Methodologies
We first analyze trends in the quantity of articles published over time and their primary methodologies.
Figure 2 shows the number of research articles in our dataset by year (2005–23) that appeared in Politics & Gender. The solid line indicates articles coded as using primarily qualitative methodologies, while the dotted line indicates articles coded as using primarily quantitative methodologies. The total quantity of research articles published annually by Politics & Gender has increased over time. From a low of 14 articles in our dataset in 2013, the journal increased output to a high of 61 articles in 2020 (driven, in part, by publications specifically about the gendered impacts of COVID-19; see below). There are 52 articles in our dataset published by Politics & Gender in 2023. For much (but not all) of the journal’s history, qualitative work has slightly outpaced quantitative work. Since 2019, however, that pattern has reversed. In 2020, 2022, and 2023, more quantitative than qualitative articles were published, and an almost equal number of each type were published in 2021.
Publication trends in Politics & Gender.
Note: The figure shows the number of original research articles published each year by Politics & Gender, separately plotting articles using primarily quantitative methodologies (dotted line) and those using primarily qualitative methodologies (solid line).
The next figure compares the trends in Politics & Gender to other journals. In Figure 3, the dotted lines show the same trendlines for Politics & Gender that we displayed above in Figure 2, while the solid lines show the total number of gender and politics articles published in all other journals in our dataset, pooled. For both sets of articles, the triangular points indicate quantitative work, while the circular points indicate qualitative work. Complementing the increased volume of articles published in Politics & Gender is a dramatic increase in gender-related work published in political science overall over the last 10 years. In our prior work (Barnett et al. Reference Barnett, FitzGerald, Krumbholz and Lamba2022, Reference Barnett, FitzGerald, Krumbholz and Lamba2025), drawing on data covering 1980–2019, we documented this emerging trend and argued that gender research was becoming increasingly mainstreamed in political science. That trend has only accelerated since 2019. In fact, in the dataset overall, the annual number of gender and politics research articles published tripled between the mid-2000s and 2023.
Publication trends in Politics & Gender vs. other journals.
Note: The figure plots, separately for Politics & Gender (dotted lines) and all other journals (solid lines) in the dataset combined, the number of original gender research articles published each year, separately plotting articles using primarily quantitative methodologies (triangles) and those using primarily qualitative methodologies (circles).
Figure 3 also shows that over this period, the volume of both quantitative and qualitative work published has increased, although quantitative research has accounted for more articles published each year from 2018 on. Figure 4 shows that the relatively equal attention given to qualitative research, and a large part of the total increase in published gender research, is driven by trends in the gender-dedicated journals. These journals — Politics & Gender, along with the JWPP, SP, and the IFJPFootnote 3 — have increased their collective output of research over time, from an annual low of 42 articles in 2007 to 169 in 2023 (more than 300% growth). Among these journals, qualitative research has tended to be published more than quantitative research, but publication of both types of research has increased. Figure 4 also shows a relatively steady gap in the volume of qualitative research published compared to quantitative research in the gender-dedicated journals. As we noted above, however, for Politics & Gender specifically, quantitative work has recently outpaced qualitative work. The non-gender-dedicated journals have consistently published more quantitative research than qualitative research related to gender and politics.
Publication trends in gender-dedicated versus other journals.
Note: The figure plots, separately for gender-dedicated journals (dotted lines) and non-gender-dedicated journals (solid lines) in the dataset combined, the number of original gender research articles published each year, separately plotting articles using primarily quantitative methodologies (triangles) and those using primarily qualitative methodologies (circles).
Finally, we offer a snapshot of how the trends in Politics & Gender compare to individual non-gender-dedicated journals. Figure 5 shows trends in the eight non-gender-dedicated political science journals that we identified as having the biggest “jumps” in publishing gender research, which we define as the largest increase from the average per-year count of gender articles prior to 2020 to their average per-year count of gender articles from 2020 on. The increased output of Politics & Gender and the other gender-dedicated journals has been matched by increases — in some cases large increases — in gender research published by journals including the American Political Science Review (APSR), the British Journal of Political Science (BJPS), Comparative Political Studies (CPS), and the Journal of Politics (JOP). Notably, the APSR was led by an all-female editorial team from 2020 to 2024. Whereas Politics & Gender has increased its output over time of both qualitative and quantitative research, Figure 5 shows that other political science journals have expanded their publication of gender-related research almost exclusively by publishing quantitative scholarship.
Trends in specific non-gender-dedicated journals.
Note: The figure shows, for each journal, the number of original gender research articles published each year, separately plotting articles using primarily quantitative methodologies (dotted lines with triangles) and those using primarily qualitative methodologies (solid lines with circles).
Gender Topics in the Journals
We next turn to the results of the topic modeling process. Because the topic modeling algorithm requires many observations to function well, and to enable comparability across journals, we applied the algorithm to the entire dataset of research articles identified as gender research in all years from 1913–2023 (N = 4,504). The algorithm output identified 52 distinct conceptual clusters to which it could assign articles. We combined these into 21 consolidated clustersFootnote 4 based on our substantive knowledge of the discipline and the relationships between different topics. Around half of the gender research articles in the dataset were not automatically assigned to any of these clusters — a point to which we return below. As a result, the specific percentages discussed below are useful mainly for the purposes of comparing patterns across journals in the publication of research clearly related to these topics, rather than drawing any conclusions about the absolute number of articles published by each journal on each topic. The Appendix contains the full list of topics identified by the BERTopic algorithm and indicates how we grouped them.
Table 1 provides an overview of the topics most frequently appearing in the gender-dedicated journals and the non-gender-dedicated journals in the dataset, pooled together. Among gender-dedicated journals (including Politics & Gender), welfare and care work, women’s representation and women in office, and women’s movements are the most common topics. Among the non-gender-dedicated journals, women running for office, women’s representation and women in office, and feminist theory and concepts were the most common topics.
Most common topics in gender-dedicated vs. other journals
Note: The table shows the topics to which articles were most frequently assigned using the BERTopic algorithm, separately for the gender-dedicated journals and non-gender-dedicated journals.
Table 2 shows the topics and the percentage of gender research articles identified as belonging to each one from work published between 2005 and 2023 in Politics & Gender, each of the other gender-dedicated journals individually, and all non-gender-dedicated journals, pooled. In Politics & Gender, the two most common topics are work on women running for office — which encompasses research related to political ambition, female candidates, and gender quotas (9.2%) — and women’s representation and women in office (7.5%). Work on women’s movements, women in specific regional or national contexts, feminist theories and concepts, gender gaps (including work on political knowledge gaps), and COVID-19Footnote 5 also each account for at least 3% of the published articles. Topics that are prominent in other journals but which have received comparatively little attention in Politics & Gender include sexual harassment and abuse, religion, welfare and care work, and migration.
Prevalence of topics in Politics & Gender vs. other journals (2005–23)
Notes: Table 2 shows the percentage of gender research articles assigned to each topic within each journal in the period 2005–23: Politics & Gender (P&G); the Journal of Women, Politics and Policy (JWPP); Social Politics (SP); the International Feminist Journal of Politics (IFJP); and all non-gender-dedicated journals, pooled. The top row shows the percentage of articles not assigned to any topic. The bottom row indicates the total number of gender research articles in each journal or group during this period. Four topics are omitted from this table for brevity, but are included in the Appendix: Socialization, Social Policies, The Media, and Miscellaneous.
The overall distribution of topics in the pooled group of non-gender-dedicated journals is similar. In these political science journals, work on women running for office is the single most-covered topic (6.5%), with women’s representation and women in office in second place (4.8%). Other relatively frequent topics include LGBT issues and queer theory (4.2%) and feminist theories and concepts (3.8%). Abortion and reproductive health make up 2.2%; however, abortion is frequently a topic in political science journals when scholars use related laws and court cases as examples or case studies in the study of courts, the judiciary, and state politics, as we discuss in the supplementary materials to our previously published article (Barnett et al. Reference Barnett, FitzGerald, Krumbholz and Lamba2022, Reference Barnett, FitzGerald, Krumbholz and Lamba2025). Feminist theories and concepts, the judiciary, war and violence, religion, gender gaps (including work on political knowledge gaps), and welfare and care work also each account for around 2% or more of the published gender research articles in the non-gender-dedicated journals.
The distribution of topics in Politics & Gender is also similar to the distribution for JWPP. Research articles in JWPP are dominated by work on women’s representation (8.7%) and women running for office (7.9%). Other frequent topics in JWPP include welfare and care work (6.2%), sexual harassment and abuse (3.9%), Black women (3.7%), and the judiciary (3.7%).
SP and the IFJP tend to publish different kinds of research. SP is dominated by research related to welfare and care work (25.8%) and also publishes often on the diffusion of equality norms and policies (11.5%), while publishing very little research on women running for office or women’s representation. IFJP publishes frequently on topics related to war and violence (15.6%), women’s movements (5.7%), women in specific regional or national contexts (4.8%), and the diffusion of equality norms and policies (4.6%).
What Makes Gender Research Unclassifiable?
While these findings enable some comparisons across journals, they should be treated with caution. In addition to the substantive topics, Table 2 shows the percentage of articles that were not allocated to any topic (“Unclassified”). Many articles were unclassified, ranging from around 40% (in SP) to more than half of the articles in the other individual journals and in the pooled set of non-gender-dedicated journals. Across all the journals in the dataset, the lowest “unclassified” rates were found in two journals clearly linked to defined topics: Politics & Religion, and the Journal of Law & Courts, each of which had around 15% of their gender research articles unclassified. At the other end of the spectrum, journals with high rates of unclassified gender research articles included New Political Science (76%), Political Analysis (79%), and Presidential Studies Quarterly (100%).
One drawback of BERTopic, a generative probabilistic algorithm, is that any individual item is probabilistically assigned to only one topic, and the algorithm may struggle to find a “home” for items that do not clearly belong to one of the latent clusters identified. The coherence of the topics that did emerge, combined with the high volume of unclassified articles, demonstrates both the advantages and drawbacks of this method. Upon running the algorithm, we were immediately struck by how the categories reflected easily identifiable, popular subfields of gender and politics scholarship. At the same time, the computational techniques need to be supplemented by human expertise to classify all the gender research articles comprehensively and coherently. Hand-coding the remaining articles to identify their topics is beyond the scope of this article, but planned for future work.Footnote 6 In the remainder of this article, we explore what some of these unclassified articles look like, and why the algorithm may have failed to link them to defined topics.
As the above statistics about “unclassified” rates across journals suggests, the inclusivity of our data collection may have made some articles that are only peripherally gender-related difficult to classify in relation to the corpus of more clearly gender-related research. For example, from Political Analysis, there are 14 articles (total, across all years) that we consider gender research by our broad criteria, and the majority of these were unclassified by the topic model. Given the journal’s focus, all of these are articles with a primary focus on exploring questions of quantitative research methodology that happen to apply their techniques to a gender-related issue. The abstracts, accordingly, include little information that would help link the articles to one of the defined topics.
This does not explain, however, why so many articles in the gender-dedicated journals, including Politics & Gender, were unclassified. We argue that the nuance, diversity, and cross-cutting nature of much published gender and politics scholarship helps account for why it is “unclassifiable” by the algorithm. To illustrate this, we selected 10 articles published in Politics & Gender that were unclassified by the topic modeling algorithm. Table 3 lists the titles, publication years, and authors of these articles. The example articles sometimes span multiple topics identified in Table 2, but also move beyond them, demonstrating the diversity of issues that gender and politics scholars investigate.
Examples of “unclassified” articles in Politics & Gender
Notes: Table 3 lists the titles, year of publication, and authors of 10 articles published in Politics & Gender that were unclassified in the topic modeling procedure that we discuss in the text. Full citations for these articles are listed in the references.
Several of the articles relate to political leadership, which is related to the topics of women running for office, women’s representation, feminist theories and concepts, and socialization, without falling neatly into any of them. Lay et al. (Reference Lay, Holman, Bos, Greenlee, Oxley and Buffett2021), for example, study the content of educational materials and how they shape children’s gender stereotypes about political leaders. Erikson and Josefsson (Reference Erikson and Josefsson2023, p. 1061) explore the “leadership ideals, evaluations, and treatment of men and women leaders in the numerically gender-equal Swedish parliament”; this article could potentially be grouped with work about women’s representation, but the contribution of the article is really to work on the nature of leadership and gendered experiences of it. Snipes and Mudde’s (Reference Snipes and Mudde2020) study of Marine Le Pen’s leadership in France spans the topics of intersectionality, leadership, and radical right politics, with none of these topics emerging from the topic modeling algorithm.
Other articles address the position and status of women around the world along various dimensions. Cassola et al. (Reference Cassola, Raub, Foley and Heymann2014) take stock of the extent of gender equality as enshrined in constitutions around the world. Cabeza Pérez, Alonso Sáenz de Oger, and Gómez Fortes (Reference Pérez, Laura and Fortes2023) examine the manifestos of regional political parties in Spain to identify the preferences of political parties on salient gender issues. Doğangün (Reference Doğangün2020, p. 258) catalogues the “revival of traditional gender norms” and resurgent narratives of conservatism in Turkey and Russia and their detrimental effects on gender equality. All, perhaps, could be considered part of a related topic — the status of gender equality — yet their diverse approaches would make this difficult to discern via mechanical parsing of the abstracts.
A final set of articles we examine features intersectional approaches to research, which may have obscured their affinity with some of the broader identified topics. Herrick (Reference Herrick2018) looks at the gender gap in identity and political attitudes specifically among American Indians. This work could in principle be added to the topic on the gender gap and political knowledge, our understanding of which the article extends in an important way through its examination of a specific population. Tolley (Reference Tolley2023) compares the aspirations to political office of white and racialized women in Canada and argues that parties’ efforts to diversify their candidate pools have primarily benefited white women. This could be categorized with the literature on “women running for office.” Other work defies simple categorization. Doan and Haider-Markel (Reference Doan and Haider-Markel2010) define and explore how “intersectional stereotyping” affects the prospects of gay and lesbian political candidates, finding that whether stereotypes based on candidates’ gender or perceived sexual minority status matter more varies by context. This article would fit within either the “women running for office” or LGBT issues topics — yet which topic should be its primary “home” is subjective. Finally, Sparks (Reference Sparks2016) investigates the role of “intersectional formations” in dissident practices, focusing on the welfare rights movement in the United States in the 1960s–1970s. Sparks’s use of the term “militants” may have made it difficult for the algorithm to determine whether this article should be part of the topic of war and violence, which includes research on female combatants, or something else. Sparks’s article could feasibly be grouped with the topics of welfare and care work or women’s movements, or potentially a new topic on intersectional research (as could the above articles).
In future work, we plan to manually inspect more of the unclassified articles to determine when articles such as these could easily be added to existing topics, and when their substance is more ambiguous. However, the fact that several of the works discussed here would slot into topics that are already identified as relatively prevalent in the data suggests that we can still take the existing results of the topic modeling as indicative of broad patterns in the subjects most frequently addressed in gender and politics research.
Conclusion
Twenty years after the establishment of Politics & Gender, the journal has facilitated the contribution of hundreds of original research articles to an increasingly broad, deep, and high-profile literature on gender and politics. From our original, comprehensive dataset gathering research substantively about gender and politics from 37 political science journals, we demonstrate that both Politics & Gender and other journals have increased their volume of output in recent years, while publishing work that covers a wide range of issues and themes, and generally balancing the publication of quantitative and qualitative work, although the former has outpaced the latter in recent years in Politics & Gender. Notably, however, the balance of methodologies in Politics & Gender contrasts with the fact that general political science journals have expanded their attention to gender almost exclusively through the publication of quantitative research. Thus, Politics & Gender appears to have both helped spearhead the increasing prominence of gender and politics topics within political science research (after decades of marginalization) while also compensating for general-interest journals’ relatively narrow interest in work that examines gender and politics quantitatively.
We also demonstrated that the most frequent clearly identifiable topics addressed by authors in Politics & Gender, the JWPP, and general interest political science journals are women running for office and women’s political representation. At the same time, the diversity and nuance of the field made a large proportion of the articles in our dataset difficult for an algorithm to classify. As scholars of gender and politics continue to innovate and extend their gaze beyond “typical” political science topics, the subfield’s landscape is likely to become only more complex over time. Politics & Gender’s commitment to pluralism in the topics, theories, and methods employed by its authors makes it well-positioned to continue to identify and publish a wide range of scholarship within this rich field.
Supplementary material
The supplementary material for this article can be found at http://doi.org/10.1017/S1743923X24000540.
Acknowledgments
The authors are listed in alphabetical order and contributed equally. We thank the editors of Politics & Gender and an anonymous reviewer for their feedback and for the opportunity to contribute to this special issue. We also thank Kathleen Rogers for her contribution to the project in its early stages.
Competing interest
The author(s) declare none.
Open access
