Twenty Years of Research on Women Voters

To gain a sense of how the journal moved forward the field’s understanding of gender and voting, we reviewed all research articles and critical perspectives published between 2005 (Volume 1, Issue 1) and 2024 (Volume 20, Issue 2) for insights into this topic. We excluded editors’ introductions and introductions to critical perspectives, the latter of which varied significantly in length and format, but generally did not advance an original line of inquiry. As we reviewed every article in the journal over this period, we used conservative inclusion criteria for our article selection, focusing on articles that analyzed electoral behavior or public opinion in the electorate as a whole, with voters as the primary unit of analysis. We focused on mass political behavior specifically aimed at electoral institutions and the attitudes that underpin that behavior. We opted to exclude articles on protest and social movements without explicit mentions of voting or elections. These forms of activism certainly have electoral consequences, but we only included articles in our review if they included an explicit electoral component.

In total, we considered 755 articles and perspectives pieces and identified 139 that offered insights into gender and voting (about 18%). This topic was much more likely to show up in research articles (90%) compared to critical perspectives (10%). Published work in this area disproportionately featured the American political context, with about 63% of articles and critical perspectives on this topic focused exclusively on the United States. To provide a broader sense of geographic representation, we classified locations of study based on the eight regions defined by the United Nations Sustainable Development Goals project. We included a category for “multiple regions” when an article utilized data from two or more of these regions. About 84% of the selected articles examine voters from Europe and North America. Only 20 articles focus on other regions (five span multiple regions, three analyze women voters in Eastern and Southeastern Asia, and the remainder were spread across Northern Africa and West Africa, Latin American and the Caribbean, Central and Southern Asia, Oceana, and sub-Saharan Africa).

This work is largely empirical and relies heavily on survey data. There is some methodological pluralism evident in this set of articles — which include observational, experimental work, and focus groups, along with analytic essays. Some of this work has grappled with methodological challenges associated with the study of gender, such as how to reconcile discrepant findings from experimental work on elections involving hypothetical candidates with observational work involving real candidates (Dolan and Lynch Reference Dolan and Lynch2016; Lefkofridi, Giger, and Holli Reference Lefkofridi, Giger and Holli2019), how to best measure gender bias or sexism (Setzler Reference Setzler2019), and how to integrate individual and aggregate-level analyses of gendered political action (Burns Reference Burns2007; Junn Reference Junn2009). The reliance on survey data for much of this work highlights the important role government-subsidized, publicly available datasets play in supporting timely scholarship on gender and elections (e.g., the Cooperative Election Study, World Values Survey, General Social Survey, and especially the American National Election Studies).

The citation counts for scholarship in this area are particularly telling in terms of impact on the field. Collectively, this work has been cited 6,692 times as of September 30, 2024,Footnote ¹ with an average citation count of about 48 for each article or critical perspective. Though citation counts are an imperfect metric of impact for many reasons, including the significant lag time for newly published articles to be read and cited, these figures suggest this work represents an important core of scholarship on gender and political behavior, and one that has contributed to the rising impact factor and prominence of Politics & Gender.

Major Themes in Research on Women Voters

Major themes from 20 years of research on women voters are evident in the keyword cloud depicted in Figure 1. Please note that articles published in Politics & Gender were not keyword indexed prior to 2019, and keywords for these earlier articles were derived from their titles. Keywords are weighted by frequency. Unsurprisingly, candidates and vote choice are a recurrent theme, with an emphasis on the gender gap. Issues of representation and women voters were a common focus, with 23 mentions in the set of articles, and public opinion and the attitudinal underpinnings of electoral behavior had about 34 mentions, spanning terms like “opinion,” “policy,” “beliefs,” and “preferences.”

Figure 1.

Keywords from articles on women voters.

Another cluster of articles centered around gender biases, denoted by terms like sexism (13 mentions), stereotypes (12 mentions), and bias (8 mentions). There was also a concentration of articles aimed at understanding the COVID-19 pandemic, consisting mostly of short papers solicited by the editorial team of Christina Wolbrecht and Susan Franceschet.Footnote ² Below, we outline major contributions in four areas: (1) descriptive representation and elections, (2) sources of heterogeneity among women voters, (3) gender differences in political information processing, and (4) gender bias in elections. Though these themes reflect the broad strokes of research contributions on women voters, there are standalone articles that have made an important impact on the field in other areas. Regrettably, we cannot detail every article here.

Descriptive Representation and Elections

An important area of inquiry regarding women voters in this journal pertains to descriptive representation. Women have historically lagged behind men in elective office, and research on women voters has sought to investigate demand for gender-based descriptive representation. Some of this work has searched for evidence of a gender affinity effect — a tendency for women to support women candidates at higher rates than men by virtue of their shared gender identity, and the expectations about the quality of representation they will receive stemming from this shared identification. Research suggests that demand for descriptive representation is not universal, but instead conditioned on a number of different factors.

In the United States, partisanship constrains gender affinity, such that women voters will rarely cross party lines to vote for a woman (Dolan Reference Dolan1997; Reference Dolan2004). In other contexts, institutional factors moderate gender affinity (Golder et al. Reference Golder, Stephenson, Van der Straeten, Blais, Bol, Harfst and Laslier2017; Marien, Schouteden, and Wauters Reference Marien, Schouteden and Wauters2017). For instance, elements of Canadian electoral systems discourage candidate-based voting, which dampens gender affinity in vote choice (Goodyear-Grant and Croskill Reference Goodyear-Grant and Croskill2011). In their study of women’s representation in Japan, Kage, Rosenbluth, and Tanaka (Reference Kage, Rosenbluth and Tanaka2019) argue that the dearth of women leaders is not a function of gender bias in the electorate but attributed instead to an electoral system that demands intense, round-the-clock legislative work and constituent service, which suppresses the supply of women available for this work. The salience of gender issues in a given election can play a role as well (Campbell and Heath Reference Campbell and Heath2017).

In addition, scholars have considered the impact of being descriptively represented on women voters. Working from the point of view that representation does not merely influence policy and other government outputs, scholars demonstrate this representation shapes women’s attitudes toward politics and the political system in several ways, including: their level of trust (Kreutzer Reference Kreutzer2023; McDermott and Jones Reference McDermott and Jones2022), political efficacy (Atkeson and Carrillo Reference Atkeson and Carrillo2007), increased confidence in women’s leadership abilities (Alexander Reference Alexander2012), and reduced perceptions of corruption in the political system (Barnes and Beaulieu Reference Barnes and Beaulieu2014). Some of this research suggests that descriptive representation bolsters women’s attentiveness to politics, and women voters show greater awareness of the legislative activity of their descriptive representatives compared to those who do not share a gender identity (Jones Reference Jones2014). Beyond this, descriptive representation is associated with a subjective sense of having one’s preferences represented (Montoya et al. Reference Montoya, Bejarano, Brown and Gershon2022), and evidence from 27 European countries suggests women legislators share policy preferences with women citizens, demonstrating substantive representation indeed follows descriptive (McEvoy Reference McEvoy2016). Collectively, this work has offered important insights into factors influencing the pursuit of descriptive representation among women voters, as well as the broad consequences of women’s political incorporation on women in the electorate.

Sources of Heterogeneity among Women Voters

Two decades of gender and politics research have established that the political implications of gender are often dependent on and even secondary to other social identities. This work largely takes an intersectional perspective, maintaining that gender consciousness and feelings of linked fate between women do not operate in isolation but instead function as part of a constellation of other politicized identities, including, but not limited to, race and ethnicity, religion, sexual orientation, citizenship, disability, and parenthood. For instance, in the early years of the journal, Wendy Smooth (Reference Smooth2006) demonstrated the importance of an intersectional look at voting behavior, particularly for African American women in the wake of the Voting Rights Act. The through line of this research is evident in a recent article about Black women’s critical role as the keepers of American democracy (Slaughter, Crowder, and Greer Reference Slaughter, Crowder and Greer2024).

Work in this vein has also offered critical insights into the intersection of gender, race, and class in subnational American elections (Bedolla and Scolla Reference Bedolla and Scola2006), explored how national origins and subnational cultures may also influence gendered political behavior (Bejarano, Manzano, and Montoya Reference Bejarano, Manzano and Montoya2011), and how beliefs about race and gender intersect to shape policy preferences (Cassese, Barnes, and Branton Reference Cassese, Barnes and Branton.2015). Religious identity is often tied to race and ethnicity, and this intersection with gender can have important implications for vote choice and policy preferences. For example, Muslim identity in Indonesia is related to opposition to gender quotas and women candidates (White et al. Reference White, Warburton, Hendrawan and Aspinall2024).

Though Politics & Gender has published a large number of studies of women voters in countries around the world, the majority of the published identity articles are centered in the United States, with particular emphasis on African American and Latina women. There are important studies on the voting behavior of other ethnic groups in North America, including Native Americans (Herrick Reference Herrick2018) and Indigenous peoples in Canada (Harell and Panagos Reference Harell and Panagos2013). But gaps in the literature remain, both in terms of other key racial and ethnic groups in the United States (e.g., Asian women voters), and in terms of racial and ethnic heterogeneity among women voters in comparative political contexts, particularly outside of North America.

Some of the scholarship in this area dovetails with the focus on descriptive representation outlined in the previous section. This work highlights the need to consider descriptive representation from an intersectional vantage point, in that voters often seek representation on multiple axes of representation simultaneously (Montoya et al. Reference Montoya, Bejarano, Brown and Gershon2022). The impact of such representation is not always straightforward, however. For instance, Albaugh and Baisley’s (Reference Albaugh and Baisley2023) investigation of the impact of a lesbian premier (a provincial-level executive) in Ontario, Canada, found her leadership had mixed effects on vote choice depending on the gender and LGBTQ+ identity of voters.

Voters also see factors like parenthood as important representational criteria, as childless women candidates are often penalized, and mothers of young children are evaluated differently than male candidates with young children (Stalsburg Reference Stalsburg2010). Motherhood itself can shape citizen preferences, where in the case of the US, mothers may take up more conservative viewpoints (Greenlee Reference Greenlee2010; see also Carroll Reference Carroll2006 and Elder and Greene Reference Elder and Greene2006), and lower rates of participation as well. Using data across 25 European countries, Shore (Reference Shore2020) demonstrates the importance of parental leave, cash assistance, and childcare access policies for boosting participation for single mothers. This work highlights the need to take an expansive perspective on the social factors shaping women’s lives in order to gain a granular understanding of women as political actors.

Gender Differences in Political Information Processing

An additional area of research emphasis worth mentioning involves evidence of gender differences in the processing and retention of political information. Though this topic is more difficult to detect from the word cloud, it addresses a common underlying question about the electoral implications of knowledge acquisition, which provides the basis for issue attitudes, candidate evaluations, and ultimately, vote choice. Early work on the gender gap in political knowledge came at the issue from a deficit perspective and focused on the implications of women’s lower average political knowledge for their civic competence.

Yet, research published in this journal came to qualify this perspective in important ways. Gidengil, Giles, and Thomas (Reference Gidengil, Giles and Thomas2008) uncovered a confidence gap at the root of the knowledge gap. This confidence gap stemmed from gender differences in concrete socioeconomic resources rather than a lack of interest or engagement in politics. Similarly, Lizotte and Sidman (Reference Lizotte and Sidman2009) identified risk aversion, uncertainty, and women’s tendency to select “don’t know” on survey questions as factors significantly inflating the gender gap in political knowledge. Evidence accumulated to suggest measurement choices related to political knowledge are not gender neutral (Hannagan, Littvay, and Popa Reference Hannagan, Littvay and Popa2014), and contributed to normatively problematic conclusions about gender and civic capacity in previously unforeseen ways (for a comprehensive review, see Ferrin, Fraile and García-Alabcete Reference Ferrin, Fraile and García-Albacete2018).

Research in this area also points to a gender difference in reactions to negativity in political communication and campaigns generally, suggesting the tone of campaigns may exacerbate these differences in how men and women encounter and absorb political information. Women are typically less receptive to campaign negativity compared to men (Philips Reference Phillips2021), and there is some evidence that men are both more receptive to and more mobilized by negativity compared to women (Brooks Reference Brooks2010). Women voters are not completely turned off by negativity. Under some circumstances, women actually attend more to negative news coverage than men (Soroka et al. Reference Soroka, Gidengil, Fournier and Nir2016), and there is some evidence that women are more receptive, compared to men, to fact checking targeted at correcting inaccuracies presented in negative ads (Fridkin et al. Reference Fridkin, Courey, Hernandez and Spears2016). Given trends toward heightened polarization and campaign negativity in the United States and globally, this is likely a fruitful area for continued research.

Gender Bias in Candidate Evaluations

Research in Politics & Gender has systematically explored the ways that gender bias might manifest in elections, with an emphasis on expressly prejudicial beliefs about gender as well as gender stereotypes. The primary goal of much of this scholarship has been to understand how these factors impact the electoral fortunes of men and women candidates, and whether they operate in ways that undermine women’s political incorporation. Beyond this, researchers seek to advise candidates on their strategic self-presentation, exploring questions about how to balance societal expectations about gender and power. Notably, much of this research shows that gender stereotypes and sexist beliefs operate in a similar fashion for men and women voters. To the extent that women voters endorse sexist attitudes, they think and act in ways similar to men who share these same beliefs. Thus, voters’ beliefs about gender, rather than their own gender identification, prove to be electorally consequential.

Gender stereotypes can potentially disadvantage women candidates, to the extent that political leadership is culturally tied to masculine rather than feminine stereotypes. This expectation has led to debate over whether women candidates are better off “running as a man” and stressing their stereotypically masculine bona fides. Yet, this strategy runs the risk of violating cultural expectations about femininity, so that a woman candidate is viewed as competent but not warm. Some gender scholars have referred to this struggle to find a precise balance of both masculine and feminine traits as “the double bind” (see, for example, Wang, Merolla, and Manganiello Reference Wang, Merolla and Manganiello2023). Yet, gender stereotypes are not always disadvantageous. Work published in this journal reports that positive evaluations of women senators often stem from their gender stereotypic strengths (Fridkin and Kenney Reference Fridkin and Kenney2009; Lucciola Reference Lucciola2023). Beyond this, Bauer’s (Reference Bauer2020) research argues that understanding gender stereotypes requires distinguishing between trait and issue-based stereotypes, as it is the former and not the latter that can sometimes disadvantage women candidates.

Research in this journal highlights that the electoral implications of stereotyping are hard to pin down — not stable over time and highly dependent on candidate characteristics (e.g., Doan and Haider-Markel Reference Doan and Haider-Markel2010) as well as contest-level factors (Glaurdić and Lesschaeve Reference Glaurdić and Lesschaeve2023; Ondercin and Bernstein Reference Ondercin and Bernstein2007). Partisanship is a particularly important confound in that gender stereotypes are conditioned and even constrained by party stereotypes in significant ways (Bernhard Reference Bernhard2022; Cormack and Karl Reference Cormack and Karl2022). The Republican Party is associated with stereotypically masculine traits and issues, and the Democratic Party with stereotypically feminine traits and issues. Candidates then face pressure to conform to these aspects of party culture, regardless of their gender identification (Winter Reference Nicholas2010). And as polarization increases, so does this pressure, such that gender stereotypes are increasingly secondary to party stereotypes (Hayes Reference Hayes2011).

Much of the early work in this journal on electoral gender bias was motivated by Hillary Clinton’s first presidential bid in 2008. This research debated whether Clinton’s electoral fortunes were constrained by competing expectations regarding communal and agentic traits (Carroll Reference Carroll2009) and how gender consciousness conditioned vote choice in the primary (Lawless Reference Lawless2009), along with the relative impact of racism and sexism among voters choosing between Hillary Clinton and Barack Obama (Hancock Reference Hancock2009; Huddy and Carey Reference Huddy and Carey2009; Junn Reference Junn2009; McThomas and Tessler Reference McThomas and Tesler2016). Later work illustrated that sexism operates in elections even when a woman is not on the ticket. For instance, Simas and Bumgardner (Reference Simas and Bumgardner2017) demonstrated that the GOP “War on Women” narrative in the 2012 presidential campaign heightened the salience of gender-relevant political issues and sharpened the effect of modern sexism on vote choice. Further scholarship has clarified this point, uncovering evidence that the hostile variant of sexism depresses support for women candidates, but the benevolent variant is tied not to a candidate’s gender identification, but rather to their gendered leadership style, reflecting a preference for candidates with a more masculine self-presentation (Winter Reference Winter2023).

Of course, gender bias does not just impact women candidates, but also the experiences of women officeholders. Hostile attitudes toward women explain why women leaders are singled out for threats of violence and harassment more so than similarly-situated men (Håkansson Reference Håkansson2024). Sexism also shapes perceptions of leaders’ responsiveness to their constituents (Costa Reference Costa2021), with broader implications influencing factors ranging from attitudes toward key political issues (Cizmar and Kalkan Reference Cizmar and Kalkan2023; Smilan-Goldstein Reference Smilan-Goldstein2024), to compliance with government-issued public health mandates (Chen and Farhart Reference Chen and Farhart2020; Kalaf-Hughes and Leiter Reference Kalaf-Hughes and Leiter2020), to support for right-wing parties (Christley Reference Christley2022). Taken together, this research suggests that sexism exerts an ongoing influence on electoral politics and the political culture more broadly.

Future Directions for Research on Women Voters in Politics & Gender

Past research in this journal situates us to understand the outcomes of future elections and provides a strong foundation for additional research on women voters. Recently, Vice President Kamala Harris, only the second woman in the history of the United States to be a major party nominee in the general election, lost her presidential bid to Donald Trump. Naturally, comparisons can be made with Hillary Clinton’s previous campaigns, but prior work on gender and elections in the journal also helps us explore questions related to campaign finance (Tolley, Besco, and Sevi Reference Tolley, Besco and Sevi2022), intersectional representation (Montoya et al. Reference Montoya, Bejarano, Brown and Gershon2022), and even the potential for a “gender backlash” (Sanbonmatsu Reference Sanbonmatsu2008). It also orients us toward understanding gender gaps on key issues voters are citing as critical to their vote choice, such as immigration (Corral Reference Corral2024), abortion (Cizmar and Kalkan Reference Cizmar and Kalkan2023), and to a lesser degree, violence associated with terrorism (Haider-Markel and Vieux Reference Haider-Markel and Vieux2008). And because the election is closely tied up in bigger questions about democratic backsliding and public support for democratic norms in the United States, it presents new opportunities for exploring gender differences in perceptions of democratic governance (for related work in Europe, see Hansen and Goenaga Reference Hansen and Goenaga2021).

Given the dearth of research on women voters outside of Europe and North America in past issues of Politics & Gender, the editors may wish to encourage submissions of critical perspectives on women voters in other regions of the world. Another option might be to invite submissions organized around critical elections globally to complement past perspectives focused on voter behavior in American elections (e.g., the recent presidential election in Mexico). Other gaps we have identified include research into the intersection between gender, religious identity, and religious beliefs, with a focus on downstream electoral consequences (but see Friesen and Djupe Reference Friesen and Djupe2017 for a look at the development of civic skills in Christian congregations). Masculinity is often overlooked as well. Past research suggests that men’s ambivalence toward advances in gender equity impacts their political behavior (Jennings Reference Jennings2006), and this finding seems important for understanding, among other things, conservative trends in ideological identification among young men and the impact it has on their votes (Deckman Reference Deckman2024).

Our review also highlighted that although sexual orientation is a politically-relevant identity, there is less research on the role of LGBTQ+ politics in elections (but see Bell and Borelli Reference Bell and Borelli2024). This is no doubt due to limited data availability and lack of sub-sample sizes sufficient to draw meaningful inferences. With the increased frequency of LGBTQ+ identification in younger generations and the destigmatization of these identities for older generations, barriers to LGBTQ+ voting behavior research are lowering. For instance, a recent survey in Canada found that the voting behavior and party affiliation of nonbinary individuals differs compared to men and women, even when analyzing a subset of only LGBTQ+ identifying respondents (Albaugh et al. Reference Albaugh, Harell, Loewen, Rubenson and Stephenson2024). This work highlights the importance of scholarship encompassing a broad understanding of gender inclusive of queer, transgender, nonbinary, feminine, masculine, and other gendered conceptions of self (Murib Reference Murib2024). Such work will encourage more nuanced development of the social and psychological underpinnings of gender difference.

The presence of gender and politics research has dramatically risen in the top journals in our field and the general social sciences, but for much of political science, “gender” is simply a control variable. The richness of the scholarship reviewed here highlights the utility of a dedicated subfield journal. As this brief review demonstrates, the study of women voters within political science crosses many major subfields, with scholarship on women voters primarily situated in country specific subfields (e.g., American, Canadian, etc.), but also explored comparatively. Theoretical and methodological approaches vary widely and often incorporate interdisciplinary perspectives. We expect this trend to only increase with time as political scientists continue to draw upon work from sociology, psychology, economics, and communication science. The majority of the articles in our corpus for this review leverage secondary data from sources such as the American National Election Studies, the Canadian Election Study, European Social Survey, and other established national datasets collected both online and over the phone. But increasingly authors are collecting novel data, which incorporate experimental designs and new measures to better understand the relationship between gender and electoral behavior. Still others rely upon different methodological perspectives offering contributions grounded in critical theory, discourse analysis, in-depth interviews, focus groups, and case studies.

Advances in access to behavioral data, such as online voting records and campaign donations, will sharpen our explanations and predictions if we no longer have to rely only upon self-report. Technological advances in big data processing and artificial intelligence models also may change the scope of our questions, such as content analysis of large bodies of text from legislative speeches (Clayton, Josefsson, and Wang Reference Clayton, Josefsson and Wang2017). Access to under-studied populations continues to increase as scholars from North America and Europe collaborate with counterparts in other regions of the world (Dieng, Haastrup and Kang Reference Dieng, Haastrup and Kang2024; Medie and Kang Reference Medie and Kang2018). Applied research and partnerships with political parties, NGOs, government agencies, and others have opened the door to raising the impact of our scholarship while simultaneously providing field work opportunities. In sum, we enter the next decade of Politics & Gender scholarship well-grounded in theoretical and empirical research from the past 20 years, with promising new directions to advance the field’s understanding of the political implications of gender in elections.