%20-%20https://sketch.com%20--%3e%3ctitle%3e3ECBBF50-FD08-412A-943D-2CFC5E025A16%3c/title%3e%3cdesc%3eCreated%20with%20sketchtool.%3c/desc%3e%3cg%20id='Stylesheet'%20stroke='none'%20stroke-width='1'%20fill='none'%20fill-rule='evenodd'%3e%3cg%20transform='translate(-926.000000,%20-668.000000)'%20id='BUTTONS'%3e%3cg%20transform='translate(100.000000,%20597.000000)'%3e%3cg%20id='BTN-/-RIGHTS'%20transform='translate(816.000000,%2065.000000)'%3e%3cg%20id='ICON-/-Rights'%20transform='translate(10.000000,%206.000000)'%3e%3cpath%20d='M13.5752841,0%20L1.34943182,0%20C0.59375,0%200,0.586666667%200,1.33333333%20L0,22.6666667%20C0,23.4133333%200.620738636,24%201.34943182,24%20L17.6505682,24%20C18.40625,24%2019,23.3866667%2019,22.6666667%20L19,5.36%20L13.5752841,0%20Z%20M13.5752841,1.89333333%20L17.0838068,5.36%20L13.5752841,5.36%20L13.5752841,1.89333333%20Z%20M1.34943182,22.6666667%20L1.34943182,1.33333333%20L12.1988636,1.33333333%20L12.1988636,5.36%20C12.1988636,6.10666667%2012.7926136,6.69333333%2013.5482955,6.69333333%20L17.6235795,6.69333333%20L17.6505682,22.6666667%20L1.34943182,22.6666667%20Z'%20id='Shape'%20fill='%23595959'%20fill-rule='nonzero'%3e%3c/path%3e%3ccircle%20id='Oval'%20stroke='%23595959'%20stroke-width='1.5'%20fill='%23595959'%20cx='9.5'%20cy='13.5355339'%20r='5.5'%3e%3c/circle%3e%3cpolyline%20id='Path-2'%20stroke='%23FFFFFF'%20stroke-width='1.5'%20transform='translate(9.500000,%2013.035534)%20rotate(45.000000)%20translate(-9.500000,%20-13.035534)%20'%20points='8%2016.0355339%2011%2016.0355339%2011%2010.0355339'%3e%3c/polyline%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/g%3e%3c/svg%3e)
INTRODUCTION
Within the subfields of political psychology and the study of gender, the introduction of new data collection efforts, methodologies, and theoretical approaches are transforming our understandings of these two fields and the places at which they intersect. In this article we present an overview of the research that was presented at a National Science Foundation (NSF) (#SES-1014854) funded conference “New Research on Gender in Political Psychology” at Rutgers University in March 2011. This scholarship represents the expanding questions and approaches that enhance our understanding of gender within political psychology. As a result, we suggest that further innovation is needed with regard to theory and methods to understand better how gender shapes the political attitudes and actions of individuals. Our discussion here covers the use of data, interdisciplinary methods, and intersectionality to study gender. We conclude with thoughts about the theoretical implications of this recent scholarship and the future of political science research on gender.
APPROACHES TO RESEARCH, DATA, AND METHODS
Scholars who have studied the impact of women in politics (with a focus on women as voters, activists, candidates, and leaders), the gendered dynamics in political behaviors of interest (with a focus on variation in behavioral repertoires), or the difference between men and women in political attitudes and behaviors (with a focus on mean differences) have long been frustrated by the scarcity of valid data related to gender and politics. The American National Election Studies (ANES) have been the cornerstone of empirical political behavior and public opinion research in political science. The riches of these data are many: The ANES is a nationally representative sample that gives researchers a large N to work with; there is an opportunity to study trends over time; and by using the ANES data one can make a claim to external validity—a valued attribute for political scientists who study attitudes and behavior. Despite the richness of these data, the ANES has not consistently asked the kinds of questions gender scholars seek to analyze. Thus, some individuals have worked to include measures of interest to others doing work on gender in the ANES (e.g., Conover and Sapiro Reference Conover and Sapiro1992) and still other scholars, particularly those who draw from psychological theory, have been compelled to explore data options elsewhere. Here, we delineate a variety of potential approaches for nonexperimental, experimental, and quasi-experimental research and describe work that takes advantage of these approaches.
Nonexperimental Research
Virginia Sapiro, Pamela Conover, Kira Sanbonmatsu, Kathleen Dolan, Nancy Burns, and others have worked to add valuable questions on gender to the ANES (e.g., Conover and Sapiro Reference Conover and Sapiro1992; Sanbonmatsu and Dolan Reference Sanbonmatsu and Dolan2007), and indeed, there are still many fruitful applications of ANES data. For instance, Susan Hansen, Kris Kanthak, and Jennifer Victor (Reference Hansen, Kanthak and Victor2011) used ANES data to show that in 2008, perceptions of the candidates were significantly affected by respondents' views of the role of women in society. The authors argued that there were significant gender differences in the ways in which voters evaluated presidential candidates Barack Obama and John McCain's individual morality. Voters' evaluation of the candidates' morality was found to be strongly affected by voters' views of gender roles. To be more specific, men who strongly believed in the equality of the sexes and who viewed McCain as more moral than Obama, voted for McCain; but the same men who saw Obama as more moral voted for Obama. The authors used these findings to show that contrary to much of the existing literature, there were significant gender effects in the 2008 election when accounting for voters' perceptions of candidates' morality and voters' personal views of gender roles. Through creative thinking, as exemplified by Hansen, Kanthak, and Victor (Reference Hansen, Kanthak and Victor2011), the ANES data can answer important questions about gender and politics.
Other researchers have turned to datasets outside mainstream political science research. For example, Erin Cassese and her co-authors (Weber, Johnson, and Cassese Reference Weber, Johnson and Cassese2011) are engaged in analyses of twins using the Adolescent Heath (AdHealth) dataset, and this interdisciplinary approach to studying gender is discussed in the next section. J. Celeste Lay, whose work on differences in political knowledge among boys and girls in rural and nonrural communities, has made use of the National Household Education Survey data (Lay Reference Lay2011). These efforts to use the questions in the ANES pertaining to gender and to use surveys other than the ANES illustrate how existing data can be leveraged to answer innovative questions at the nexus of political psychology and gender studies.
Despite many options for the exploration of research questions available from survey data, many questions that interest gender scholars require new data. Original data collection is challenging and expensive. In the current economic climate many colleges and universities do not provide sufficient financial support for data collection. Researchers often must seek external funding or collaborate with colleagues within and outside of political science who have access to data or funds. Still, the collection of large-scale data for nonexperimental work is possible. For example, at the Rutgers conference it was noted that Melissa Marschall and Paru Shah recently obtained NSF funding to engage in a large data collection effort on local elections in the United States (the Local Elections in America Project, or LEAP). Because many women are elected to office on a local level, understanding local elections could provide insight into the election of women at various levels of office.
In another effort to collect original survey data, Sarah Fulton (Reference Fulton2011) conducted a mail survey of state assembly-members to examine how sex-based differences in perceived discrimination deters women from running until they have reached a threshold level of quality. Her research suggests that perceptions of bias shape the quality of female candidates who emerge. Indeed, original data collection that focuses on state and local levels complements studies that use the ANES and other existing datasets.
Zoe Oxley (Reference Oxley2011) gathered new data on media coverage of candidates to assess whether the newspaper reporting patterns seen when women are running for high prestige offices (presidency, US Senate, governorship) extend to lower prestige state-level offices (such as state treasurer and attorney general). By coding newspaper articles on all nongubernatorial state executive office elections between 2006 and 2009 when there were both female and male candidates, Oxley finds that gender stereotypes do shape coverage: women running for offices that were connected to stereotypical female issues received higher quality coverage, while men running for offices that were connected to stereotypical male issues also received better coverage. Thus, data collection efforts such as this, as well as the innovative use of datasets not traditionally used by political scientists, create new opportunities for scholars to see the influence of gender on political behavior at both the individual and the institutional levels.
Finally, partnering with political consultants, perhaps in the context of a field experiment, also may be a way to access data concerning women as candidates and the “difference” factor in campaigning. During the conference at Rutgers, a panel of political consultants—Kelly Anne Conway, Manolo Gonzalez-Estay, Celinda Lake, and Candy Straight—shared their expertise with the conferees, as well as suggested ways to access the individuals and processes necessary to answer research questions of interest. Kelly Dittmar's (Reference Dittmar2010) recent work involves surveying active campaign consultants regarding their experiences dealing with gender stereotypes and identities within the prevailing electoral landscape. She also focuses on the campaign strategies aimed at negotiating gender in the context of the institutional processes of campaigns and elections and concludes that campaign institutions are gendered.
Experimental Research
In her influential study on candidate gender, Virginia Sapiro (Reference Sapiro1981) used an experimental design on a student sample to determine if the sex of the candidate influences voter perceptions (see also Huddy and Terkildsen Reference Huddy and Terkildsen1993a, Reference Huddy and Terkildsen1993b). Experimental studies have oft been criticized among political scientists for using undergraduate student samples that are convenient but not necessarily ideal for understanding political psychological phenomenon (Sears Reference Sears1986, but see also Druckman and Kam Reference Druckman, Kam, Druckman, Green, Kuklinski and Lupia2011). To overcome this bias, gender and politics research may consider sources other than undergraduate samples that are nearly as convenient and perhaps as cost effective for conducting experimental research.
Deborah Jordan Brooks (Reference Brooks2011a) collected original data through an Internet-based survey experiment to understand how candidate behavior affects the public's perceptions of female and male politicians. She used Polimetrix, a firm that keeps a panel for conducting research and uses geographically and demographically representative samples. In a series of experimental studies (e.g., Brooks Reference Brooks2011b), Brooks found no evidence for the “double bind,” the idea that women are negatively evaluated more than men for “unladylike” behaviors such as toughness, anger, and crying (Brooks forthcoming).
The Cooperative Congressional Election Study (CCES) is a national stratified sample survey that can be used to address both experimental and nonexperimental questions. In particular, it is an excellent vehicle for studies that examine the gender of candidates because the sample is stratified by state and type of congressional district. Although not free, the CCES does enable scholars to pool their resources to obtain high-quality survey research data. The CCES is a promising new resource for gender scholars, as illustrated by work on cue taking and gendered racial subgroups presented at the Rutgers conference by Corrine M. McConnaughy and Ismail K. White (Reference McConnaughy and White2011).
Also, at the conference Leonie Huddy recommended that scholars use Amazon's Mechanical Turk (MTurk), which is an online tool for the collection of experimental data for a nominal fee. Scholars examining the quality of samples generated from MTurk have concluded that study participants recruited through MTurk are more representative than participants recruited through more traditional convenience sample methods (Berinsky, Huber, and Lenz Reference Berinsky, Huber and Lenz2010). This tool opens up many possibilities for researchers investigating questions regarding gender and political psychology who have limited research budgets. Indeed, Schneider and Bos are piloting an experimental study in 2012 with MTurk to examine how either gender or party stereotypes can be primed during a campaign. Another option, Time-sharing Experiments for the Social Sciences (TESS), which is free to scholars whose experiment-based proposals are accepted, contracts with Knowledge Networks to produce a nationally representative, probability-based Internet panel.
Other creative approaches to data collection can be used. For instance, Schneider and Bos (Reference Schneider and Bos2011) received permission from the Hamilton County Jury Commissioner to collect data from potential jurors seated at the courthouse awaiting jury duty. Although this is not a nationally representative sample, potential jurors are randomly chosen from lists of registered voters. Hamilton County, Ohio includes the diverse city of Cincinnati, so the likely pool is as diverse as sampling from citizens living in Cincinnati or many other major US cities. Schneider and Bos used these participants to help them understand stereotypes about political candidates based on both gender and party. A host of gender and politics questions might be addressed using a similar type of sample in the future.
Finally, field experimental techniques have been used successfully in understanding Get Out the Vote (GOTV) campaigns (e.g., Gerber and Green Reference Gerber and Green2000; Green, Gerber, and Nickerson Reference Green, Gerber and Nickerson2003; and see Gerber, Green, and Larimer Reference Gerber, Green and Larimer2008) and could be applied to understand questions pertaining to gender and voter turnout: for example, Do social-emotional appeals increase turnout more for women or men? Does “competitive” framing promote voter turnout more strongly among men or women? In the pursuit of such questions, we see field experiments as a fruitful tool for gender scholars.
Quasi-Experimental Research
Similar to field experiments, quasi-experimental methods offer promising opportunities to observe carefully the impact of gender in organic research settings. For example, Rebecca J. Hannagan and Christopher W. Larimer (Reference Hannagan and Larimer2011) are involved in a large-scale data collection effort funded by the NSF (#SES-1015406 and #SES-1015391) to study gendered group dynamics in local political decision-making bodies. Their study began when the State of Iowa passed a law requiring all local boards and commissions to be “gender balanced” by January 2012. Hannagan and Larimer saw this requirement as an opportunity for a natural experiment and to test their laboratory findings in the real world. Their findings from a laboratory experiment illustrated that the gender composition of a group influences what individuals prefer to do in a strategic decision-making scenario. Men, in particular, were influenced by the experimental manipulation, which was random assignment to all male, all female, or mixed groups (Hannagan and Larimer Reference Hannagan and Larimer2010).
Interdisciplinary Research
As empirical political scientists, the questions we ask might be better answered with theories and methods “borrowed” from other disciplines. Political scientists have increasingly looked to other disciplines and their methodologies to apply to political science questions. Twin studies and physiological analyses coupled with more traditional survey data analyses are two such examples.
Cassese, as cited earlier, is analyzing twin data using a classic twin design using structural equation modeling (SEM) to decompose aspects of the trait in question (i.e., political trust) into additive genetic, common environmental, and unique environmental influences. The authors then use SEM to identify a gene's x environment interaction. The findings illustrate that although there is a modest genetic effect on political trust, the unique environmental influence make the variance in political trust different for women and men (Weber, Johnson, and Cassese Reference Weber, Johnson and Cassese2011). This type of analysis is no longer exotic in political science but used by many scholars (e.g., Alford, Funk, and Hibbing Reference Alford, Funk and Hibbing2005; Fowler, Baker, and Dawes Reference Fowler, Baker and Dawes2008; Hatemi et al. Reference Hatemi, McDermott, Bailey and Martin2011). Techniques borrowed from behavior genetics are a fairly recent import to studies of political behavior, as are studies using methods from psychology as it pertains to the study of emotions. Hannagan, for example, explores the differences in emotion-motivation systems between men and women that underpin ideology (Hannagan and Oxley Reference Hannagan and Oxley2010a), and in a laboratory experiment also explores the relationship between empathy and perceptions of fairness in policy preferences (Hannagan and Oxley Reference Hannagan and Oxley2010b). Further, Hannagan and Larimer employ the Facial Action Coding System (FACS) to measure the implicit emotions of men and women in decision-making groups in their field study (Hannagan and Larimer Reference Hannagan and Larimer2011).
RECOGNIZING THE IMPORTANCE OF INTERSECTIONALITY
One theme that emerged from several presentations at the Rutgers conference was the consideration of how gender intersects with other identities. Conference participant Corrine McConnaughy (Reference McConnaughy2007) wrote eloquently in an earlier piece about the issue of measuring identification with gender as a group to capture more accurately how women and men relate to their groups. Not only might gender be a “metacategory” but, McConnaughy argues, we must consider the “subcategories” within the broader constructs of men and women. This applies to women as voters and citizens, as well as to women as candidates. Thus, we review several pieces that consider the “subcategories” of women; the findings support a notion that political behavior may be fundamentally gendered—inasmuch as gender is a social role that differs not only between men and women but varies within gender categories—and has implications for future research.
With work by theorists such as Crenshaw (Reference Crenshaw1991) and Hancock (Reference Hancock2004), empirical researchers have become increasingly aware of the importance and complexities of studying the intersection of race and gender. Much of the early research in this area focused on the voting behavior or legislative decisions of individuals whose identity embodies this intersection (Barrett Reference Barrett1995, Reference Barrett1997; Gay and Tate Reference Gay and Tate1998). However, the work of McConnaughy and White (Reference McConnaughy and White2011) take this work in a new direction. They address intersectionality in a different context by examining the ways in which race-gender subgroup specific categories structure racial stereotypes and affect political decision making. Using data from a survey experiment in the 2008 CCES and data gathered from college students using a modified version of the Affect Misattribution Procedure (AMP), McConnaughy and White show that the use of gendered racial discourse, which is sometimes used to normalize racial discourse, merely taps into or reinforces subgroup specific stereotypes when making political evaluations.
In another exploration of intersectionality and stereotypes, Monica Schneider and Angela Bos (Reference Schneider and Bos2011), building on Huddy and Capelos' work (Reference Huddy, Capelos, Ottati, Posavac, Edwards, Bryant, Tindale, O'Connell, Heath and Suarez-Balcazar2002), examine the psychological mechanisms and process involved in the use of combining party and gender stereotypes to evaluate political candidates. They find that the dominance of one identity over another depends on the issue or trait in question. In particular, the authors show that female Republican politicians are in a unique position whereby they embody cross-cutting gender and party stereotypes. This suggests that the political context may be particularly consequential for Republican female candidates, as the greater salience of gender or party may influence the public's evaluations of them.
Finally, in a project that examines the intersections of ethnicity and gender, Christina Bejarano uses the World Values Survey and the 2006 Latino National Survey to compare the size and direction of the US Latino ideological gender gap (Bejarano Reference Bejarano2011). She finds evidence of a modern gender gap, shaped by structural and cultural factors, with Latinas demonstrating a realignment in their political attitudes.
Another subcategory of interest represents the intersection of gender and social roles and identities. The question of how motherhood shapes women's political attitudes and behaviors has garnered increasing interest from scholars who seek to understand how this role produces a new dimension of difference among women (Howell and Day Reference Howell and Day2000; Eagly et al. Reference Eagly, Diekman, Johannesen-Schmidt and Koenig2004; Elder and Greene Reference Elder and Greene2007, Reference Elder, Greene and Whitaker2008). Jill Greenlee (Reference Greenlee2010) shows that not only are there cross-sectional attitudinal differences between women who have children and women who do not, but that motherhood produces modest attitudinal changes within the individual that likely contribute to the gap between mothers and nonmothers. By pairing longitudinal data from the Political Socialization Panel Study with cross-sectional data from the ANES, Greenlee shows that it is not simply self-selection into motherhood that produces these attitudinal distinctions, but that the role itself alters women's political views. This work reveals yet another politically consequential intersection of gender, social roles, and identity. Literature from both political science (e.g., Conover Reference Conover1988; Conover and Hicks Reference Conover, Hicks and Taras1998) and social psychology (Duncan and Stewart Reference Duncan and Stewart2007; Duncan Reference Duncan, Deaux and Snyder2011) illustrate the importance of the various social roles that underpin identity and, in turn, predict certain political attitudes and behaviors.
THEORETICAL IMPLICATIONS AND NEW DIRECTIONS
A theoretical framework either guides or emerges from data collection and rigorous testing. Although the deductive approach to empirical political science is the dominant approach, a more inductive approach should be considered. Many of the discussions throughout the Rutgers conference resulted in some agreement that men and women are different enough such that, as a discipline and not only as a subfield, we ought to more carefully consider the ways in which we measure political attitudes and behavior. Because the ANES lacks questions targeted at studying the impact of gender on the American political landscape some changes about the study gendered nature of politics are necessary. To improve gaps in our knowledge we need to recraft and add new questions to existing datasets that are widely used. Sapiro, Dolan, Sanbonmatsu, and others have done excellent work in furthering gender research via the ANES. Looking at the theories and methods from fields other than political science, we may need to measure men and women's attitudes and behaviors separately if not differently (see McDermott and Hatemi Reference McDermott and Hatemi2011; Hannagan Reference Hannagan2008). For example, if men and women perceive their role as legislator differently, asking the same questions of men and women candidates and politicians may not adequately capture gender differences. A one-size-fits-all approach to investigating a gendered subject matter may result in empirical inconsistencies that cannot build useful theory.
In addition to potential for an inductive approach to studying men and women's political attitudes and behaviors in the future, much of the discussion throughout the conference hinged on the issue of the study of gender as fundamental to an understanding of political psychology. At the conference the scholars brainstormed approaches to making gender a unifying framework for understanding human political behavior. The group concluded that more effort must be made to “mainstream” the study of gender in political psychology. Approaching panel chairs for major conferences to ensure that work on gender appears on a wide range of panels and is not limited to only “gender” and “women and politics” panels is key. Measuring both men and women and undertaking the difficult task of theory-building based on attitudinal and behavioral similarities and differences is also critical.
What can political scientists learn from this new direction in work on gender in political psychology in American politics? The promise of understanding and predicting political behavior is great if the scholars who attended the Rutgers conference collaboratively build databases from their innovative data collection efforts, continue to use and develop new methodological techniques, work beyond the traditional sources for data, employ interdisciplinary frameworks, and more carefully consider intersectionality in their work. We also hope our conference will encourage and inspire others to do the same. In addition to producing more interesting and increasingly rigorous research, political scientists may gain a better theoretical understanding of what political behavior entails and how it may be fundamentally gendered. This “new direction” poses nothing less than a paradigm shift in how we as a discipline view and study public opinion and political behavior.
ACKNOWLEDGMENTS
We thank all the participants who presented at the 2011 “New Research on Gender in Political Psychology” conference and especially Leonie Huddy and Kira Sanbonmatsu for their support and feedback. We also thank the National Science Foundation for funding the conference (#SES-1014854).
