MergedFile A Dissertation Submitted to the Temple University Graduate Board In Partial Fulfillment of the Requirements for the Degree by Examining Committee Members: ii ABSTRACT Sexism, sexual harassment, and the objectification of women are issues that have gained a new level of salience in our political culture. The phrase “Me Too” has captured the pervasiveness of these experiences. Feminist theorists in particular, have long recognized the political significance of marginalization and discrimination on the basis of gender, and how even events that occur in the private sphere can have political implications. However, positivist scholars of political science have paid less attention to these seemingly non-political factors as potential predictors of political engagement. This dissertation is an effort to shed light on how gender-based discrimination affects women in the electorate and how they engage in the political sphere. Through a combination of observational research, survey experiments, and lab experiments, I demonstrate that under certain circumstances, gender-based discrimination can depress women’s political engagement and under other circumstances, gender-based discrimination can actually act as an impetus to political engagement and activism. The goal of this dissertation is two-fold. First, I argue and empirically demonstrate that sexism, sexual harassment, and the objectification of women have explicit political consequences. Second, I illuminate the moderating factors in this relationship between gender-based marginalization and political engagement. I explore how group consciousness, ideology, and emotions affect the connection between marginalizing experiences and political engagement and behavior. My findings uncover a complicated relationship between marginalizing experiences and political engagement. These experiences can depress engagement, but can also become events that galvanize political activity. The most important contribution of this dissertation is underscoring the need for iii scholars to consider how the lived experiences of marginalized groups shape the way they approach politics. iv To my mother, for everything. v ACKNOWLEDGEMENTS I first want to thank Vin Arceneaux. Without his support, advice, and belief in my work, I would not have been able to complete this dissertation. I am forever grateful for the opportunity to work in the Behavioral Foundations Lab and to use the lab’s resources to conduct my own research. His thoughtful feedback made this a better project. Having Vin as an advisor made graduate school and the process of writing a dissertation seem so much less daunting. I feel fortunate to have had an advisor who elevates the voices of junior scholars. Thank you so much, Vin. I would also like to thank Nyron Crawford and Heath Fogg Davis for serving on my committee and helping me turn my scattered ideas into a dissertation. Having them as both teachers and committee members has been a huge privilege and I am a better political scientist because of them. Thank you to Johanna Dunaway for not only serving as my outside reader, but for her unique insights, suggestions, and mentorship. I must also express my gratitude to Ryan Vander Wielen, Bert Bakker, Hillel Soifer, Robin Kolodny, Amanda Friesen, and David Nickerson for reading my work, attending my talks, and proving me with constructive feedback that helped me move forward with this project when I felt stuck. I would be remiss to not thank my fellow graduate students at Temple. I am especially grateful to my fellow cohort members Lauren Rowlands, Ethan Fried, and Joel Blaxland for their moral support and friendship. I want to thank Amanda Milena Alvarez, my “academic life partner,” for inspiring me every day to do the work that I believe in and for always reminding me that I am not in this alone. vi I also want to acknowledge my undergraduate advisor from Wilkes University, Thomas Baldino, for encouraging me to pursue my Ph.D. I would not have made it to Temple without his advice and his insistence that I was indeed capable of writing a dissertation. I want to recognize my family for their unconditional support throughout this process. My aunts, Ellen Malone and Mary Sparks, and my uncle, Tom Malone, have supported my academic pursuits since I was child, and have always made me feel like I was capable of anything. My little brother, Harry Gothreau, has rooted me on for the past five years and never once asked me when I would be done with school. I want to acknowledge my best friend, Courtney Conway, who was always there to remind me that my worth is determined by more than just my ability to succeed in academia. Thank you to Arnie, my canine best friend, who sat by my side while I finished this dissertation and Steffi Kasparek for bringing him to me. Finally, I want to thank my parents, Joan Malone and Stephen Gothreau. None of my accomplishments would have come to fruition without them. Their love and support has been limitless. My dad is the reason I became interested in politics. He sparked an intellectual curiosity in me that helped me push forward when my belief in my own work waned. My mom gave me the courage to write a dissertation that is meaningful to me, and borne of my own experiences. I love you both. vii TABLE OF CONTENTS Page ABSTRACT………………………………………………………………………..……..ii DEDICATION…….…………………………………………………………….…….….iv ACKNOWLEDGEMENTS……..……………………………………………….………..v LIST OF TABLES….………………………………………………………………….…ix LIST OF FIGURES……….……………………………………………………………....x CHAPTER 1. INTRODUCTION……………………………………………………………………...1 Sexism and Politics………………………………………………………………..2 Discrimination and Political Engagement…………………...……………………4 Plan of the Dissertation………………………………………………...………….7 2. THE POLITICAL CONSEQUENCES OF SELF-OBJECTIFICATION…….………10 Introduction. ………………………………..……………………………………11 What is Self-Objectification?.……………………………………………………12 The Consequences of Self-Objectification………………………………………17 Self-Objectification and Political Engagement……………………………….….20 Research Design………………………………..……………..………………….24 Results………………………………..……………..……………………...…….29 Implications and Conclusion………………………………..……………………34 3. ME TOO: THE IMPACT OF SEXISM ON POLITICAL ENGAGEMENT AND THE MODERATING IMPACT OF GENDER CONSCIOUSNESS…………………...…….37 viii Gender-based Discrimination and its Consequences…………………………….40 Sexist Events, Political Engagement, and Group Consciousness………..………44 Data and Measures……………………………..…………………………..…….48 The Relationship Between Gender Discrimination and Feminist Identity Development……………………………..………………………………………50 Results…………………………………………………………………………....54 Discussion…………………………..……………………………………………60 4. THE RELATIONSHIP BETWEEN GENDER DISCRIMINATION AND SOCIOPOLITICAL BEHAVIOR………………………………………………...……..61 Sexism, Discrimination, and Political Engagement……………………...………66 Research Design…………………………..………………………………...……69 Results…………………………..………………………………………………..72 Concluding Remarks…………………………..……………………………..…..76 5. CONCLUSIONS…………………………………………………………………...…78 Summary of Findings…………………………………………………………….79 Future Work………………………………………………………………...……81 BIBLIOGRAPHY. ………………………………………………………………………83 APPENDIX….....……………………………………………………………………...…93 ix LIST OF TABLES TABLE 2.1 Descriptive Statistics for Body Shame and Body Surveillance Scales…………....27 A.1 Descriptive Statistics MTurk Sample…………………………………………..…93 A.2 OLS Regression Results: Internal Political Efficacy………………………...……94 A.3 OLS Regression Results: Political Interest…………………….……….…………95 A.4 OLS Regression Results: Political Information-Seeking………………….………96 A.5 Ordered Logit Model: Robustness Check………………….……………...………98 A.6 OLS Model: Gender Egalitarianism……………….……………………………...99 A.7 Descriptive Statistics for MTurk Sample……………………………………...…103 A.8 OLS Regression: Internal Political Efficacy…………………………………..…104 A.9 OLS Regression: Political Interest…………………………………………….…105 . A.10 OLS Regression: Political Participation…………………………………………106 A.11 Ordered Logistic Regression Results………………………………………….…110 A.12 Descriptive Statistics for Laboratory Sample……………………………...…….112 A.13 OLS Regression (Model 1) ……………………………………….……………..112 A.14 OLS Regression (Model 2) ……………………………………….…………..…113 A.15 OLS Regression (Model 3) ……………………………………….…………..…114 A.16 OLS Regression (Model 4) ……………………………………….…………..…114 A.17 OLS Regression (Model 5) ……………………………………………...………115 A.18 OLS Regression (Model 6) ……………………………………….…………..…116 A.19 OLS Regression (Model 7) ……………………………………….…………..…117 A.20 OLS Regression (Model 8) ……………………………………….…………..…118 x LIST OF FIGURES FIGURE 2.1 Example of Objectifying Advertisements…………………………………………14 2.2 The Relationship Between Self-Objectification and Political Engagement...…….23 2.3 Differential Effect of Body Shame on Efficacy by Gender………...……………..30 2.4 Differential Effect of Body Shame on Political Interest by Gender…………...….31 2.5 Effect of Body Shame on Information-Seeking for Men and Women……………32 2.6 Effect of Body Surveillance on Information-Seeking for Men and Women…...…33 2.7 Differential Effect of Body Shame on Gender Egalitarianism by Gender………..34 A.1 Body Surveillance Posterior Distribution………………………………………..101 A.2 Body Shame Posterior Distribution…………………………………………..….101 3.1 The Relationship Between Sexism and Political Engagement……………………46 3.2 Reported Sexist and Objectifying Events…………………………………………54 3.3 Marginal Effect of Sexist Events and Interpersonal Objectification on Efficacy over the Range of Gender Consciousness…………………………………………………..…56 3.4 Marginal Effect of Sexist Events and Interpersonal Objectification on Political Interest over the Range of Gender Consciousness…………………………………….…57 3.5 Marginal Effect of Sexist Events and Interpersonal Objectification on Political Participation over the Range of Gender Consciousness…………………………………58 3.6 Predicted Political Participation at Varying Levels of Gender Consciousness…...59 A.3 Marginal Effect of Sexist Events on Efficacy over the Range of FIDS (Model 1)………………………………………………………………………………………..108 xi A.4 Marginal Effect of Sexist Events on Efficacy over the Range of FIDS (Model 2)……………………………………………………………………………………..…109 A.5 Marginal Effect of Sexist Events on Efficacy over the Range of FIDS (Model 3)………………………………………………………………………………………109 A.6 Marginal Effect of Sexist Events on Efficacy over the Range of FIDS (Model 4)………………………………………………………………………………………110 4.1 Audio Transcript…………………………………..……………….……………...71 4.2 OLS Regression Results…………………………………..………………………74 4.3 Estimated Coefficient of the Treatment by Gender Consciousness….……….......74 4.4 OLS Regression Results…………………………………..………………………75 4.5 OLS Regression Results…………………………………..…………………..…..76 1 CHAPTER 1 INTRODUCTION Sexism goes so deep that at first it’s hard to see: you think it’s just reality. -Alix Kates Shulman For many women, sexism, objectification, and even sexual harassment are daily experiences (Fitzgerald & Shullman 1993; Swim, Hyers, Cohen, & Ferguson 2001). The #MeToo Movement has shone a spotlight on the pervasiveness of sexual assault and harassment. In just three months, 6.5 million tweets with the hash tag were shared on Twitter, and countless women shared their stories (Chou 2018). As the 2016 presidential election underscored, the political sphere is far from being devoid of sexism and rhetoric that objectifies and dehumanizes women. One candidate was accused of sexual misconduct by several women, while the other candidate was subject to gender-based attacks and sexist rhetoric. Now more than ever, we are experiencing a cultural moment that is highlighting how gender fundamentally shapes our life experiences. The field of political science has long acknowledged the impact that sexism and sexist stereotypes can have on female candidates and their electoral success (Huddy & Terkildsen 1993; Sanbonmatsu & Dolan 2009). There is a wide breadth of literature on sexist media coverage, gender differences in the evaluation of candidates, and the impact of these factors on women’s representation in political institutions (Kahn & Goldenberg 1991; Kahn 1994). Political scientists have not shied away from addressing sexism, but there are still significant gaps. The literature is focused on the impact of sexism on political elites and neglects an important discussion about the impact on everyday 2 women. How do sexism, sexual harassment, and other marginalizing experiences, something so omnipresent in women’s lives, affect American women and their sense of political engagement and political efficacy? This dissertation is an effort to shift the focus from sexism’s effect on political elites to shed light on how sexism, as well as other forms of gender-based marginalization and discrimination, affect American women in the electorate. I conceive of marginalization as the everyday prejudiced experiences that traditionally oppressed and stigmatized groups have. This will aid us in going beyond standard predictors of political behavior such as education, income, and access to resources. This chapter will serve as a brief review of the most relevant literature. I will outline the literature on sexism and politics, discrimination and political engagement, and the prevalence of gender-based discrimination to set up the motivating question of this dissertation. I will then provide an overview of the three papers that follow. Sexism and Politics There is a wealth of literature on why women do not have parity in elected offices in the United States. Much of this research focuses on the way that gender stereotypes effect how voters perceive candidates. For example, evidence shows that voters expect female candidates and politicians to be compassionate given that this is a stereotype that applies to women broadly (Huddy & Terkildsen 1993). Women are also seen as more liberal and less capable of handling national security and foreign policy issues (Lawless 2004; Sanbonmatsu & Dolan 2009; Hayes 2011). Schneider and Bos (2014) challenge this literature by suggesting that female politicians make up a subtype within the broader 3 group of women with their own set of stereotypes. Ultimately, this can be even more damaging for female politicians as they are seen as coming up short on “masculine qualities,” as well as the positive qualities attributed to women generally.1 Preference for male candidates is evidenced in the recruitment process of political parties and politicians as well (Sanbonmatsu 2002). There is also a considerable literature on sexist media coverage of political candidates. Not surprisingly, scholars have found that women do not receive the same press treatment as male candidates. Women receive less news coverage overall, even when the competitiveness of the race is controlled for, and the coverage they did receive was often about their viability as candidates and not about issue positions (Kahn & Goldenberg 1991; Kahn 1994). Women also receive more personal coverage than their male counterparts, and often that coverage focuses on physical appearance and dress (Devitt 2002; Heldman, Carroll, & Olson 2005). The media also tends to focus on the “first woman” frame instead of the candidate’s policy issues (Kahn 1996; Heldman et al. 2005). There is also an extensive literature worth noting that explores the mechanisms in place that keep women from running for office in the first place. Before women can actually run for office, they need to be nominated by a political party. Parties can act as gatekeepers to candidacy for office, which often leads to the exclusion of potential women candidates (Norris & Lovenduski 1995; Krook 2010). This is one way in which institutional structures can lead to sexist outcomes. 1 More specifically, women are seen as coming up short on “masculine” qualities such as “assertiveness” and “self-sufficient,” but also positive “feminine” qualities like “honesty” and “warmth.” Essentially, female politicians are punished for not possessing the masculine qualities that are seen as necessary for politics, but they are also punished for not possessing positive stereotypically feminine qualities. 4 The literature briefly outlined above is valuable in elucidating the way in which women experience the political world differently from men. It’s clear that women who run for public office or who are already in public office are often subject to sexism and discrimination, from their peer, the electorate, and the media. This sexism is manifested in the application of stereotypes, media coverage, and the institutional structures that make it difficult for women to run for office and win. Discrimination and Political Engagement The studies that have linked discrimination and political participation suggest that social organizing and political activism can be collective responses to everyday prejudice (Simpson & Yinger 1985). The established consensus in political behavior research is that discrimination by political institutions motivates those who are marginalized to engage in politics in an effort to protest their status (Miller, Gurin, Gurin, & Malanchuk 1981; Pantoja, Ramirez, & Segura 2001; Barreto & Woods 2005; Cho, Gimpel, & Wu 2006; Page 2018). However, more recent work finds that interpersonal discrimination, opposed to structural or institutional discrimination, can actually be demobilizing and cause individuals who experience this type of discrimination to withdraw from political life (Oskooii 2016). With the exception of this most recent work, the extant research does not focus on interpersonal discrimination, but rather, structural or explicitly political discrimination against marginalized groups. The literature on discrimination and political engagement focuses chiefly on those with marginalized racial and ethnic identities. For example, Cho et al. (2006) find that events following 9/11 mobilized Arab-Americans in the United States and led to greater 5 political activity. The authors purport that this increased mobilization was in response to threatening and discriminatory policies against Arab-Americans. Pantoja et al. (2001), compare the voter turnout among cohorts of naturalized and native-born Latino citizens. They find that Latinos who were naturalized in California during a time of anti- immigrant rhetoric and legislation were more likely to vote in the 1996 election than Latinos who were naturalized in Texas and Florida where anti-immigrant rhetoric was not as salient. Some more recent literature has focused on groups other than racial and ethnic minorities. For example, Matilla and Papageorgeiou (2017) look at how perceptions of discrimination affect the political participation of those with disabilities. Conventional wisdom suggests that in general, disabilities depress political participation. However, the authors find that although disability decreases voting when associated with perceptions of discrimination, it increased participation in demonstrations and contact with politicians. Taking a more nuanced approach to conceptualizing discrimination and using American Muslims as the population for his analysis, Oskooii (2016) distinguishes between systemic or political discrimination and interpersonal discrimination. Oksooii (2016) notes systemic or political discrimination, “typically refers to discriminatory laws, campaign messages, policies, or practices carried out by state or private institutions and/or their affiliated actors” (p. 616). Using survey data from the 2007-2008 Muslim American Public Opinion survey and the 2007 Pew Research Center Study of American Muslims, he finds that systemic discrimination is associated with increased political mobilization, whereas interpersonal discrimination has the opposite effect and causes many to retreat from the political sphere. 6 The mechanism that mobilizes racial and ethnic groups subject to discrimination to engage in politics, in part, is a sense of group consciousness and linked fate (Jamal 2005; Ysseldyk et al. 2014; Lin 2018). Group consciousness is considered to be a political resource that explains the high levels of political participation among some marginalized groups (Sanchez 2006; Sanchez and Vargas 2016). The existing research suggests that as marginalized group members experience discrimination, their awareness of their status of being part of a group that is receiving unfair treatment is increased. This leads to a sense of linked fate or group consciousness (Pantoja et al. 2001; Barreto and Woods 2005; Cho et al. 2006; Sanchez 2006).2 On the other side of the coin, system- justifying beliefs tend to dampen reactions to inequities. Endorsement of system- justifying beliefs coupled with unequal outcomes actually undermines political mobilization (Osborne and Sibley 2013). This modest literature on the connection between discrimination and political engagement suggests that social organizing and political activism are collective responses to discrimination and marginalization. This work also delves into the mechanisms that drive the relationship between perceptions of discrimination and political mobilization. However, there is also some evidence that the type of discrimination is consequential in its effect on political engagement (Oskooii 2016). Furthermore, this work focuses on ethnic and racial discrimination, rather than gender-based discrimination. This is a useful starting point to theorize about the potential effect of gender discrimination against women on political engagement, but we also know that gender and race are not equivalent marginalized identities. 2 For more discussion on the difference between linked fate and group consciousness, see Chapter 3. 7 Plan of the Dissertation The literature outlined above may at first seem unrelated, but actually reveals a significant gap in the literature on political participation and gender and politics. The scholarly research on sexism and politics focuses primarily on those running for public office or those already in public office. The research on discrimination and political participation mainly lies in the realm of race, ethnicity, and politics. To my knowledge, there has been no systematic study of the relationship between gender discrimination and political participation. This dissertation investigates the connection between gender discrimination, broadly conceived, and political engagement, as well as the factors that moderate this relationship. Using a combination of surveys, survey experiments, and laboratory experiments, I highlight the continued relevance of personally-experienced discrimination and marginalization to scholars of political behavior. This dissertation takes a three-article format. The project’s first article uses a survey experiment to explore the effect of self-objectification on political engagement. Results from the survey experiment are presented that demonstrate how self- objectification undermines women’s political engagement, and more specifically, feelings of political efficacy. The chapter begins with a discussion of the psychology literature on self-objectification and its relationship to cognitive functioning, behavioral outcomes, and self-efficacy. Results from the survey experiment suggest that the negative consequences of self-objectification extend into the political realm as well. This article problematizes the objectification of women’s body in our culture and reveals political consequences. 8 In the second article, I go beyond just objectification to explore how other forms of gender-based discrimination and marginalization correlate with political engagement and behavior. I hypothesize that there are circumstances in which personally-experienced sexism and discrimination can increase political engagement and activism. I argue that women who are high in gender consciousness will become galvanized by sexist and discriminatory events and engage in political participation as a response. Women who are low on gender consciousness will not become mobilized by the same accumulation of discriminatory experiences. Gender consciousness, which consists of factors such as feelings of linked fate with other group members and an acknowledgement of one’s membership in a marginalized group, moderates the relationship between discrimination and political engagement. Experiences of discrimination, coupled with high levels of gender consciousness, can act as an impetus to political mobilization. In the final empirical chapter of my dissertation, I build upon the findings from my first two studies and seek to establish a causal relationship between gender-based discrimination and political engagement. In an ongoing lab experiment, I use bogus feedback on a cognitive task to prime personally-experienced sexism. I consider the role of gender consciousness and linked fate, emotional appraisal, and ideology as moderators in the relationship between exposure to sexism and political engagement. I also measure physiological response in the form of skin conductance (SCL), a common measure of emotional arousal or attentiveness generated by the sympathetic nervous system, to examine subjects’ response to exposure to sexism. The goal of this dissertation is to illuminate the ways in which seemingly non- political factors end up having explicit political consequences. Feminist political theory 9 has long recognized how gender-based marginalization is inherently political (de Beauvoir 1949; Young 1980; Bartky 1990; hooks 1991). However, discussions of everyday prejudice, sexism, and sexual harassment are absent from empirical work in political science. Marginalizing experiences, particularly ones that occur routinely, are formative in people’s lives. Political science has addressed the effect that sexism has on political elites, but has paid scant attention to how marginalization on the basis of gender affects American women. This is particularly important as we enter a cultural moment in which no topic is more salient than the treatment of women at work, in public spaces, and in politics. 10 CHAPTER 2 THE POLITICAL CONSEQUENCES OF SELF-OBJECTIFICATION Abstract Research has shown that women are significantly less politically engaged than men at both the mass and elite levels (Bennett & Bennett 1989; Verba, Burns, & Schlozman 1997; Burns, Schlozman, & Verba 2001; Lawless & Fox 2010; Pruysers & Blais 2014). More recent scholarship has found that structural factors and standard predictors of political behavior no longer sufficiently explain this persistent gap in engagement (Atkeson & Rapoport 2003; Wen 2013; Preece 2016). In the present study, I take a novel approach to exploring the discrepancy in men and women’s political engagement. I posit that self-objectification, a psychological mechanism more commonly found in women, undermines engagement. When women self-objectify, their cognitive resources are significantly diminished and they fail to see themselves as agents of change. I conduct a survey on a sample of both men and women. I find that trait self- objectification is negatively related to internal political efficacy, political interest, and propensity to seek political information. Theses results demonstrate that the negative consequences of the sexual objectification of women extend into the political sphere and in part, drive the gender gap in engagement. 11 “The sexual objectification of women produces a duality in feminine consciousness. The gaze of the Other is internalized so that I myself become at once seer and seen, appraiser and the thing appraised.” -Sandra Lee Bartky Introduction Women are objectified in popular culture, the media, and in the political sphere (Lanis & Covell 1995; Heflick & Goldenberg 2009; Galdi, Maass, & Cadinu 2014; Schooler 2015). The objectification of the female body has become normalized in American society. This persistent objectification leads to a phenomenon known as self- objectification, which occurs when individuals internalize observer’s perspectives of their physical bodies (Fredrickson & Roberts 1997). Psychologists have documented the adverse effects that self-objectification can have on a woman’s mental health, cognitive functioning, and self-efficacy (Fredrickson, Roberts, Noll, Quinn, & Twenge 1998; Gapinski, Brownell, & LaFrance 2003; Noll & Fredrickson 1998; Roberts & Gettman 2004). Building on the extensive literature on self-objectification, I explore whether or not the negative consequences of self-objectification extend to the political sphere. In particular, I look at the well-established gender gap in political engagement. Extant research has found that women are less engaged in politics at both the mass and elite levels (Burns, Schlozman, & Verba 2001; Lawless & Fox 2010; Pruysers & Blais 2014; Verba, Burns, & Schlozman 1997). Despite the fact that women vote at the same rates as men, they consistently report being less interested in politics, are less knowledgeable, express their political attitudes less than men, and are less politically efficacious (Beckwith 1986; Bennett & Bennett 1989; Delli Carpini & Keeter 2005; Huckfeldt & 12 Sprague 1995; Preece 2016; Verba et al. 1997). Men and women diverge sharply in their confidence to comprehend politics, and women are more likely to believe that politics is too complicated for them to understand (Gidengil, Giles, & Thomas 2008; Pruysers & Blais 2014). This paper presents the results of a survey study designed to test the hypothesis that self-objectification has a negative effect on political engagement in women. I go beyond measuring standard predictors of political behavior to posit this psychological mechanism, found in higher levels in women, as a predictor of political engagement. My results support the notion that state self-objectification negatively correlates with various measures of engagement. These findings underscore the relevance of the sexual objectification of the female body in the media, popular culture, and in the political sphere, to political outcomes. They also highlight the importance of psychological explanations for the gender gap in political engagement. What is Self-Objectification? The notion of objectification has been discussed by many feminist theorists writing from a social constructivist perspective. Simone de Beauvoir, Sandra Bartky, and Iris Marion Young all wrote about the objectification of the female body (Bartky 1990; de Beauvoir 1949; Young 1980). Bartky (1990) states, “Sexual objectification occurs whenever people’s bodies, body parts, or sexual functions are separated out from their identity, reduced to the status of mere instruments, or regarded as if they were capable of representing them.” Nussbaum (1995) builds on Bartky’s definition to assert that objectification involves the denial of autonomy, treating the object as lacking agency, and 13 denying the object subjectivity. She also asserts that the objectifier often treats the object as interchangeable with other objects of the same type. Objectification is to treat a human being in one or more of these ways. Ultimately, a theory of objectification was developed to understand the consequences of living in a culture that sexually objectifies women. Objectification theory posits that women are often treated as simply bodies or body parts. More specifically, they are treated as bodies that exist for the consumption and pleasure of other (Fredrickson & Roberts 1997). Fredrickson and Roberts (1997) explain, “The common thread running through all forms of sexual objectification is the experience of being treated as a body (or collection of body parts) valued predominantly for its use to (or consumption by) others” (p. 174). The crux of objectification is the degradation or even elimination of another’s agency and personhood. Building on Nussbaum’s work, Caroline Heldman developed the Sex Object Test (SOT) to determine whether images are objectifying or not. She asserts that sexual objectification is present if the image only shows parts of a sexualized person’s body, if the sexualized person is a stand-in for an object or is interchangeable, or if the image affirms the notion of violating the bodily integrity of a sexualized person that is unable to consent. The SOT test also classifies images that portray the sexualized person as a commodity as objectifying. The advertisements in Figure 2.1 are images that would qualify as sexually objectifying based on the SOT. The first image clearly shows a woman being portrayed as an object, while the second image suggests a violation of the bodily integrity of the sexualized person. The third image meets the first criteria of the SOT in which only part of a sexualized person’s body is shown. 14 Figure 2.1: Example of Objectifying Advertisements 15 Objectification can occur in interpersonal encounters, but even more often, in mainstream media. The proliferation of marketing and entertainment media has heightened the amount of sexually objectifying images and words we are exposed to (Heldman & Wade 2011). The average American is exposed to over 5,000 ads each day (Story 2007). Content analyses have concluded that the media focuses on the body and appearance as the main components of sexual desirability (Aubrey 2006). The increased access to smartphones amplifies exposure to objectifying media. Interpersonal sexual objectification refers to objectification that women directly experience. This can occur in relationships with family and friends, colleagues, or in public spaces. Kozee, Tylka, Augustus-Horvath, and Denchik (2007) developed a scale to measure these instances of interpersonal objectification that includes two factors: body evaluation and unwanted explicit sexual advances. This type of objectification captures the notion of the “objectifying gaze,” or the sexual evaluation of women’s bodies. 16 Objectification can have damaging effects on the hearts and minds of women. One of these effects is self-objectification. When women self-objectify, they internalize observers’ perspectives of their physical bodies. As Quinn et al. (2006) state, “Being in a state of self-objectification signifies that a person has moved from a subjective sense of self as agent to a sense of self as object” (p. 59). The insidiousness of self-objectification is that it can pervade every aspect of a woman’s life, including her mental health, cognitive functioning, and self-efficacy. Scholars have found that self-objectification can contribute to a host of mental health issues, including anxiety and depression (Huebner & Fredrickson 1999). It can cause feelings of shame, increased body surveillance, and intentions to have cosmetic surgery (Calogero, Pina, & Sutton 2014). Some studies have found that inducing women to self-objectify negatively impacts cognitive performance on challenging tasks (Fredrickson et al. 1998). Self-objectification is both a stable trait and a context-dependent state (Fredrickson et al. 1998). “Trait” self-objectification is a person’s overall propensity to see themselves as objects. This type of self-objectification is cultivated over a lifetime and remains relatively stable. Context-dependent or “state” self-objectification is not a stable trait. This type of self-objectification can occur in any situation where physical appearance is made salient. This could occur when an objectifying ad is viewed, when a person is subject to street harassment, or even upon receiving an appearance-based compliment. 17 The Consequences of Self-Objectification Psychologists have found that self-objectification contribute to anxiety and depression (Huebner & Fredrickson 1999). Self-objectification can also lead to sexual dysfunction and disordered eating (Calogero, Davis, & Thompson 2005). Furthermore, it has been linked to behavioral health outcomes such as self-injury and substance abuse (Calogero et al. 2014). Some studies have found that inducing women to self-objectify negatively impacts cognitive performance on challenging tasks (Fredrickson et al. 1998). Women who self-objectify have lower self-esteem and self-efficacy (Gapinski et al. 2003). Self-efficacy is the belief in one’s own capability or effectiveness in any particular situation. Research in psychology indicates that those who are more self- efficacious are healthier and generally more successful (Bandura 1997). Gapinski, et al. (2003) found that self-objectification, as both a stable and enduring trait and as a situationally-induced state, was correlated with increased negative feelings, decreased intrinsic motivation, and diminished cognitive functioning. Women are more vulnerable to self-objectification than men. This is relatively unsurprising considering our culture tends to objectify the female body more often than the male body (Fredrickson & Roberts 1997). This notion is supported by the scholarly research on the antecedents of self-objectification. Calogero et al. (2005) find that internalized media ideals for appearance contribute to self-objectification. The authors state, “This suggests that the viewing of sexually objectifying images of women in visual media (e.g., magazines, music videos, television shows) may be a contributing factor to the chronic viewing of oneself as a sexual object if those images become integrated into one’s self-perception” (p. 44). Stranger harassment, the sexual harassment of women in 18 public places by men who are strangers, is also positively related to self-objectification (Fairchild & Rudman 2008).3 Many different individual-level differences in personality contribute to whether or not someone is likely to self-objectify, but culture and socialization play a significant role as well. Women tend to be higher trait self-objectifiers than men, and show more variability across individuals than men do (Fredrickson et al. 1998). Roberts and Gettman (2004) found that when exposed to an objectifying prime, women’s ratings of negative emotions were higher and their ratings of the appeal of physical sex significantly lower than women not exposed to the prime. In contrast, men were unaffected by the objectifying prime. Huebner and Fredrickson (1999) explore the role of autobiographical memories in self-objectification from two perspectives: the original first person perspective (field images) and that of an outside observer (observer images). The authors assert that observer imagery could potentially result from the emotions and self-awareness that are associated in objectifying situations. They found support for the hypothesis that female participants report more observer imagery than male participants, both for memories in general and for memories of situations wherein women are likely to be objectified. This is consistent with the idea that self-objectification involves viewing yourself from the perspective of others. There is some evidence that self-objectification is negatively related to both internal and external political efficacy. Heldman and Cahill (2007), through survey data, 3 Men and gender non-conforming people are also the victims of stranger harassment, but the overwhelming majority of victims identify as women. Interestingly, Fairchild and Rudman (2008) found that this was only true for women who used common strategies for coping (e.g. passive, self-blame, or benign), but not women who used more active coping strategies (e.g. confronting the harasser). 19 found that respondents who were high self-objectifiers exhibited lower rates of internal political efficacy than other respondents. Specifically, they found that among the high self-objectifiers, only 10% had a high level of internal efficacy whereas among the low and medium self-objectifiers, 18.5% had a high level of internal efficacy. The survey data collected by Heldman and Cahill (2007) is an important starting point, but are always internal validity concerns with the collection of survey data. The relationship between internal efficacy and self-objectification has not been sufficiently investigated. It is only recently that scholars have begun to look at psychological mechanisms as powerful predictors of political behavior. Calogero (2013) found support for the notion that self-objectification renders women less engaged in gender-based social activism. Greater self-objectification was related to more gender-specific system justification which mediated the link between self-objectification and social activism. This meditational model was true for both trait and state self-objectification. For the purposes of this study, this research is particularly relevant as it provides some evidence that self-objectification can thwart political activism among women. Scholars have found that context-dependent self-objectification can be primed in experimental settings. Fredrickson et al. (1998) manipulated self-objectification by having participants try on either a swimsuit or a sweater while completing a difficult cognitive task. They found that women in the swimsuit condition that was intended to prime self-objectification, performed worse on the task. Tiggeman and Boundy (2008) used two subtle environmental cues to prime the emotion. One of the manipulations was an environmental manipulation in which the room where the participants had to complete 20 the study had two full-length mirrors, a set of bathroom scales, and a small display of fashion magazine covers on the room divider. The other experimental manipulation was that half of the participants received an appearance compliment, and the other half received no comment. This research shows that self-objectification can be primed in an experimental setting. Whether we are cognizant of it or not, we all participate in the act of self- objectification. It is natural for someone to self-objectify when their physical appearance is made salient. However, the psychology literature has demonstrated that chronic self- objectification can have negative consequences on emotion well-being and cognitive functioning, particularly for women. Even when men display similar levels of self- objectification as women, they tend not to experience the same psychological consequences (Roberts & Gettman 2004). The aim of this study is to explore whether or not these consequences extend to the political realm. Self-Objectification and Political Engagement There is a scant amount of literature linking self-objectification to political outcomes. Heflick and Goldenberg (2011) examined the 2008 presidential election and posited that the objectification of vice-presidential candidate Sarah Palin was detrimental to the Republican ticket. The authors suggest that the heavy national focus on Palin’s appearance led to the perception that she was less competent, warm, and moral. They also suggest, although they do not test this assertion empirically, that this intense objectification may have even undermined the competency of her own performance. In an experimental study, Heflick and Goldenberg (2009) found that asking subjects to 21 focus on and write about Palin’s appearance reduced perceptions of her humanity and competence. These findings are particularly troubling considering a candidate’s perceived level of competence plays a significant role in vote choice (Todorov, Mandisodza, Goren, & Hall 2005). These studies that explore how sexual objectification can negatively impact female political candidates is an important step in acknowledging how our culture of objectifying female bodies can pervade the political sphere. However, we have yet to fully consider how objectification, as well as self-objectification, could impact the way the average woman approaches politics. The evidence we do have suggests that self- objectification could undermine women’s political participation. Because of the pervasive forces of society and its objectification of women, in many cases, women turn inward to objectify themselves. I advance the theory that women who are high self-objectifiers fail to see themselves as agents of change. When women self-objectify their cognitive resources are significantly diminished (Fredrickson et al. 1998; Gapinski et al. 2003; Gay & Castano 2010). Self-objectification puts one in a state of vigilant body monitoring, increased anxiety, and lowered feelings of overall competence. This state is inconsistent with feelings of efficacy, interest, and overall political engagement. Internal efficacy is a feeling of competence and capability, and it is anchored by feelings of self-esteem (Clarke & Acock 1989). Possessing efficacy undoubtedly requires not only self-esteem and competence, but likely requires cognitive resources as well. Women who are high self-objectifiers self-monitor diligently. As Fredrickson et al. (1998) state, “This self-conscious appearance monitoring can disrupt an individual’s stream of consciousness, and thereby limit the mental resources available for 22 other activities” (p. 270). In this case, I am discussing self-objectification as a stable and enduring trait. If so much of one’s cognitive resources and self-efficacy are diminished, it is likely that this translates to efficacy in the political sense.4 Therefore, I hypothesize that: 𝐻!: Women who are high trait self-objectifiers will have lower internal political efficacy than those who are low trait self-objectifiers. The negative political consequences of self-objectification likely extend beyond just efficacy. If those who are high self-objectifiers have decreased self-esteem and cognitive resources, it is likely that they would be less politically engaged in other ways as well. This leads to my second and third hypotheses: 𝐻!: Women who are high trait self-objectifiers will have less political interest than those who are low trait self-objectifiers. 𝐻!: Women who are high trait self-objectifiers will display less political information- seeking behavior than those who are low trait self-objectifiers. Taken together, efficacy, interest, and information-seeking behavior, are a solid operationalization of political engagement. Decades of research in political science emphasize the importance of internal efficacy in political agency and democratic participation (Almond and Verba 1963; Pinkleton and Austin 2001). To be engaged in politics, it stands to reason that one must be efficacious, interested, and willing to seek out politically relevant information. Figure 2.2. summarizes the proposed relationship between self-objectification and political engagement. 4 The notion that increased self-objectification increases cognitive has been established by the psychology literature. I am not claiming that increased cognitive load is the causal mechanism that decreases efficacy. However, it is likely a contributing factor. 23 Figure 2.2: The Relationship Between Self-Objectification and Political Engagement The extant literature suggests that self-objectification is more common in women than it is in men (Fredrickson et al. 1998). Furthermore, there is some evidence that even when men display similar levels of self-objectification as women, they do not suffer the same negative psychological consequences (Roberts & Gettman 2004). I hypothesize that the differential impact of self-objectification on men and women likely carries over into the political realm as well. As self-objectification increases, women will become less engaged, whereas for men, self-objectification will have no effect on their level of engagement. Self-Objectification − political efficacy, interest, and information-seeking − cognitive resources, self-efficacy, feelings of competence + body monitoring and anxiety 24 Recent research on self-objectification and women’s political consciousness lends supports for my theory. Calogero (2013) finds that greater trait self-objectification was related to more gender-specific system justification and less engagement in gender-based social activism. In other words, women who are high self-objectifiers were more likely to be content with the status quo in terms of gender relations and less likely to engage in efforts to improve the status of women. Similarly, Calogero, Tylka, Donnelly, McGetrick, and Leger (2017) find that the belief that beauty is a type of currency, self- objectification, and support for the gender status quo were negatively related to gender- based activism. This research indicates that there is a relationship between self- objectification and political consciousness and activism as it relates to gender-based attitudes. The psychology literature has thoroughly demonstrated the negative impacts that self-objectification can have on a woman’s mental health and cognitive functioning. Women who are high self-objectifiers are more likely to feel ashamed of their bodies, pursue cosmetic surgery, develop eating disorders, experience sexual dysfunction, perform poorly on cognitive tasks, and feel negative emotions (Calogero et al. 2005; Calogero et al. 2014; Tiggeman & Boundy 2008). However, we have yet to explore the political consequences of something so omnipresent in many women’s lives. Research Design Participants To test the proposed hypotheses, in June of 2017 I recruited 244 male and female respondents to participate in an online survey through Amazon’s Mechanical Turk (See 25 Appendix A.2 for more detailed demographic information). Six respondents were dropped for completing less than 50% of the survey instrument. Participants ranged from age 20 to age 79 (M=35.9, SD=11.3), and 60.0% of the sample had a bachelor’s degree or higher. Out of the 244 participants, 55.9% identified as white, 10.2% as black, 5.7% as Hispanic, 22.4% as Asian, and 5.7% as more than one race. The median reported yearly income was $35,000 to $64,999. Participants were compensated $1.00 for completing the study. While the sample was quite diverse, it is younger, more educated, and more liberal than the United States population. While this limits external validity somewhat, MTurk samples tend to be more representative than in-person convenience samples (Berinsky et al. 2012). Procedures and Measures After consenting to participate in the study, all respondents were asked to take a pre-test which included McKinley and Hyde’s Objectified Body Consciousness Scale (OBCS). This questionnaire is a measure of trait self-objectification. Two of three sub- scales were included; 1) body surveillance, and 2) internalization of cultural body standards (body shame). This scale was derived from feminist theory and is designed to measure the behaviors and attitudes that contribute to women’s negative body experience (McKinley and Hyde 1996).5 The body surveillance sub-scale includes statements like, “I rarely compare how I look with how other people look,” and “I am more concerned with what my body can do than how it looks.” The body shame sub-scale includes statements 5 Following Moradi and Varnes’s (2017) research on the factor structure of the OBCS, I only utilized the Body Surveillance and Body Shame sub-scales. In their analysis they found that the factor loadings for the Control Beliefs items were low, indicating that these items were poor measures of the Control Beliefs factor. The Control Beliefs Scale also yielded a theory- inconsistent negative correlation with the Body Surveillance factor and the Body Shame factor, as well as a low Cronbach’s alpha. 26 such as “I feel like I must be a bad person when I don't look as good as I could,” and “I would be ashamed for people to know what I really weigh.” Response categories were on a scale from 1 (strongly disagree) to 7 (strongly agree), with a “not applicable” option. Not applicable responses were treated as missing values, and appropriate items were reverse coded such that higher scores indicated higher levels of body surveillance and body shame. Using these items on the sub-scale, the goal was to generate scores for each latent variable. Scholars have typically used confirmatory factor analyses to uncover the factor structure of the OBCS. Using factor analysis allows items on the scale to have different weights. However, standard factor analyses treat each factor score as deterministic, and does not take into consideration that these estimates are uncertain. To better account for potential measurement error, I used Bayesian ordinal factor analysis to estimate scores for each respondent on separate body shame and body surveillance scales (Treier & Jackman 2008). This is a more rigorous statistical technique that allows for the incorporation of measurement error into the regression models. Specifically, I used a Monte Carlo Markov Chain to estimate the model 10,000 times, each time randomly sampling a factor score from each respondent’s posterior distribution.6 Distributions of the posterior means can be found in Appendix A.2. 6 Various diagnostic tests, including the Geweke diagnostic (Geweke 1992) were run to confirm that two parts of the chain (first 10% and the last 50%) were asymptotically independent of each other. Raftery and Lewis’s diagnostic, as well as traceplots to ensure that the chain did not get stuck in certain areas of the parameter space. 27 Table 2.1: Descriptive Statistics for Body Shame and Body Surveillance Scales Average Posterior Distribution Statistics: Body Shame Body Surveillance Range 5.49 5.64 Mean .0017 -.0007 Standard Deviation .914 .927 Cronbach’s Alpha on Scale Items .81 .87 Correlation Between Factor Scores .983 .978 & Posterior Means The survey also included a variety of demographic questions. The primary dependent variables are questions tapping political engagement. The four internal efficacy items were the standard items used on the National Election Study developed by Niemi, Craig, and Mattei (1991). These items include statements like “I consider myself to be well-qualified to participate in politics,” and “I feel that I could do as good a job in public office as most other people.” Response categories were on a scale from 1 (strongly disagree) to 7 (strongly agree). Items were coded such that higher values indicated a more efficacious answer. Scholars have found evidence that these four items display high internal consistency, and both confirmatory and exploratory factor analyses demonstrated that the items measure a single concept that is robust across major subgroups (Niemi et al. 1991; Morrell 2003). In this sample, these four items displayed a high level of internal consistency (𝛼=.854). Confirmatory factor analysis using promax rotation was used to render factor scores for the efficacy items with higher scores indicating more efficacious answers. 28 Political interest was measured with two items tapping interest in both political campaigns and current events.7 Items were coded such that higher values indicated more interest. Confirmatory factor analysis using promax rotation was also used to render factor scores for the political interest items with higher scores indicating more political interest.8 The two items displayed high internal consistency (𝛼=.763). The post-test also measured information seeking behavior.9 Participants were required to read a brief news article.10 They were then asked how likely they would be to look for more information related to the story, or to read another news story about related issues if they encountered one. This measure gets at one’s propensity to not only be interested in politics, but to actively seek out more political information. Finally, respondents completed a battery of questions from the World Values Survey on gender egalitarianism (McDaniel 2008). This included six questions about women’s equality in jobs, education, and politics, as well as questions about women’s role in the workplace and as mothers. The six items displayed an acceptable level of internal consistency (𝛼=.724). Items were coded such that higher 7 Respondents were asked, “Some people don't pay much attention to political campaigns. How about you? Would you say that you have been very much interested, somewhat interested, or not much interested in the political campaigns this year?” and “Some people seem to follow what's going on in government and public affairs most of the time, whether there's an election going on or not. Others aren't that interested. Would you say you follow what's going on in government and public affairs most of the time, some of the time, only now and then, or hardly at all?” 8 Summated rating scales were also created with the interest and efficacy items. Substantive results were unchanged when the models were run using the summated rating scales instead of the factor scores. 9 The information seeking question was measured on a 7-point scale from extremely likely to extremely unlikely. I treated this as a continuous variable, although as a robustness check, I also ran this model as an ordered logit model, which can be found in the appendix. Substantive results were unchanged. 10 The article involved a discussion of a Texas legislative session on sanctuary cities, and a bill that would ban them and punish local governments that do not comply with the law. 29 responses indicated more gender egalitarianism, and confirmatory factor analysis was conducted on all six items. Results To test the hypothesis that women who are high trait self-objectifiers will have lower internal political efficacy than those who are low trait self-objectifiers (𝐻!), I ran an OLS model regressing the internal political efficacy scale on education, ideology, income, gender, the body shame scale, and the body surveillance scale, as well as an interaction between gender and both self-objectification scales.11 I chose this set of controls as they are all potential predictors of political engagement. Regression results can be found in Appendix A.2, Table A.2. Contrary to theoretical expectations, body shame has a significant (p<.05) positive impact on efficacy, however, the significant interaction (p<.1) between body shame and gender shows that for women, the effect of body shame is negative. Figure 2.3 illustrates the differential effect of body shame on internal political efficacy on men and women. We see that for men, body shame actually has a positive impact on efficacy, whereas for women, there is a negative impact. Body surveillance has no discernable impact on efficacy levels. These results display partial support for 𝐻!. 11 Education was coded on a scale of 1 through 5, 1 being “less than a high school diploma,” and 5 being “post-college degree.” Ideology was coded on a scale of 1 through 6, 1 being “very liberal” and 6 being “very conservative. Party was dummy coded 0 for Republicans and 1 for Democrats. Gender was dummy coded 0 for those identifying as men and 1 for those identifying as women. 30 Figure 2.3: Differential Effect of Body Shame on Efficacy by Gender To test the hypothesis that women who are high self-objectifiers will also be less politically interested, I regressed the political interest factor scores on the same set of controls, as well as the surveillance and shame scales interacted with gender. Support was also found for 𝐻!. Although there was no main effect of either body shame or body surveillance, the interaction between body shame and gender was significant (p<.1), showing that body shame has a negative effect on interest, but only for those who identify as women. We see in Figure 2.4 that for women, increased body shame decreases political interest, whereas for men, there is no significant impact. Full OLS regression results can be found in Appendix A.2, Table A.3. 31 Figure 2.4 Differential Effect of Body Shame on Political Interest by Gender To test the final hypothesis that women who are high self-objectifiers will show a decreased propensity to seek out political information than those who are low self- objectifiers, I regressed the 7-category information-seeking variable on the same set of covariates as in the models for 𝐻! and 𝐻!. Consistent with my hypothesis, the interaction between the body shame scale and gender was significant (p<.1), as well as the interaction between body surveillance and gender (p<.1), indicating that shame and surveillance have a negative impact on information-seeking, but only for women. It’s worth noting that these interactions were the only significant predictors in the model. Figures 2.5 displays the predicted levels of information-seeking over the range of body shame and body surveillance scores for men and women. In Figure 2.5, we see that if anything, there is a positive relationship between body shame and information-seeking for men. Figure 2.6 also shows the clear negative impact of shame on information- 32 seeking for women. Figure 4 illustrates a similar pattern with the body surveillance scores. Figure 2.5: Effect of Body Shame on Information-Seeking for Men (Top) and Women (Bottom) 33 Figure 2.6: Effect of Body Surveillance on Information-Seeking for Men (Top) and Women (Bottom) Finally, with no a priori hypothesis, I wanted to explore the impact of self- objectification on gender egalitarianism, which was measured with the gender egalitarianism items from the World Values Survey. I used OLS to regress the gender 34 egalitarianism factor scores on the interaction between gender and the self-objectification scales, as well as the controls. Interestingly, there was a significant negative main effect of body shame on gender egalitarianism (p<.01), but a significant positive main effect of body surveillance on gender egalitarianism (p<.01). However, the significant interaction term between body shame and gender (p<.01) indicated that this negative relationship only holds for men in the sample. In other words, as body shame increases for men, they display less egalitarian attitudes. For women, there is actually a slight positive relationship. Figure 2.7 displays these differential effects. Figure 2.7: Differential Effect of Body Shame on Gender Egalitarianism by Gender Implications and Conclusions The literature on the gender gap in political engagement is expansive. However, much of this research centers around structural factors such as access to education, income, access to other resources, and other macro-level forces as drivers of this gap. 35 These explanations become problematic as women are now increasingly integrated into the workforce, have surpassed men in terms of educational attainment, and have access to more resources than ever. Furthermore, recent scholarship shows that even when controlling for existing structural disparities, the gender gap in engagement persists (Atkeson & Rapoport 2003; Preece 2016; Wen 2013). More recently, scholars have put forth psychological explanations for the gap in engagement. For example, one reason that women may be less engaged is that they lack the confidence to fully participate in politics (Preece 2016). Building off of this notion, I posit that self-objectification is one such psychological mechanism found in higher levels in women, that undermines their engagement in politics. The extensive literature on objectification theory teaches us that self-objectification decreases feelings of overall efficacy and competence, likely exacerbating the confidence gap between men and women in the political realm. Self-objectification is also a reminder to women that their value and self-worth is derived from their physical appearance, underscoring their sense that they do not belong in politics. Essentially, women with high trait self-objectification do not think of themselves as agents of change. There are of course limitations to this study. Due to the observational nature of this work, we cannot be completely confident that the relationship between self- objectification and political engagement is causal. Additionally, the sample is not totally representative of the broader population of Americans, which limits this study’s generalizability. In the future, it would be fruitful to replicate these findings using a larger and more representative sample in the U.S., as well as outside of the U.S. in countries with various of female objectification. Despite these limitations, this study provides 36 compelling evidence of a robust relationship between self-objectification and political engagement that holds up to a variety of controls. Body shame was negatively related to internal political efficacy, interest, and information-seeking. Interestingly, body surveillance had no significant impact on any of the dependent variables except information-seeking. This suggests that shame, the more affective component of self- objectification, plays a larger role in impacting political engagement. Body surveillance, which is theoretically linked more to appearance monitoring and the disruption of an individual’s attentional resources, seems to have a less significant effect on engagement. The most important contribution of this study is shedding light on how seemingly non-political factors like objectification, end up shaping political engagement. The sexual objectification of women has been a salient topic for many years with both academics and the public speculating about the ramifications of a culture that emphasizes female beauty standards. There is a wealth of evidence that shows that objectification, and consequently, self-objectification, have dire consequences for women’s mental health and cognitive functioning. The results in the present study suggests that the negative consequences of self-objectification for women extend into the political realm as well. 37 CHAPTER 3 ME TOO: THE IMPACT OF SEXISM ON POLITICAL ENGAGEMENT AND THE MODERATING EFFECT OF GROUP CONSCIOUSNESS Abstract Sexism, sexual harassment, and the objectification of women are issues that have gained a new level of salience in our political culture. The phrase “Me Too” has captured the pervasiveness of these experiences. The present study is an effort to demonstrate how the marginalizing experiences that women have in their everyday lives affect their engagement in the political sphere. I test the primary hypothesis that gender-based discrimination has an effect on both political engagement and behavior utilizing data from a sample of 311 women living in the United States. Findings indicate that here is a robust connection between gender-based marginalization and political engagement. Women who experience more discrimination and marginalization become more engaged in politics. However, this relationship is conditional on levels of group consciousness. These results underscore the relevance of discrimination to the study of political engagement, as well as the importance of group consciousness in understanding how women perceive and cope with discrimination. 38 Sexism goes so deep that at first it’s hard to see: you think it’s just reality. -Alix Kates Shulman Sexism, sexual harassment, and objectification are all ways in which women are marginalized in which men are generally not. Sometimes these events and experiences are relatively minor, and other times they are significant. Gender-based discrimination is commonplace and makes up the “lived experience” of women (Feagin & Sikes 1994). These experiences range from seeing a sexually objectifying advertisement, to hearing a sexist comment, to being the victim of sexual violence. Although these events are different from each other, they are all rooted in a patriarchal system that fundamentally denies women of the same freedoms that men enjoy. The few studies that have linked discrimination and political participation suggest that social organizing and political activism can be collective responses to the type of everyday prejudice described above (Simpson & Yinger 1985). However, recent work finds that interpersonal discrimination can actually be demobilizing and cause individuals who experience this discrimination to withdraw from political life (Oskooii 2016). With the exception of this study, the extant research does not focus on interpersonal discrimination, but rather, structural or explicitly political discrimination against marginalized groups. Moreover, there are virtually no studies exploring personally- experienced gender discrimination and its effect on political behavior. How do the everyday marginalizing experiences that women have affect the way in which they approach politics? 39 Sexism toward political candidates, rather than citizens, has largely occupied the focus of the political science research. This is likely due to the fact that it is easier to see the logical connection between sexism and electoral success. However, we have yet to consider how sexism impacts women in the electorate. The limited research on discrimination and political behavior would suggest that discrimination and unfair treatment can spur individuals to become engaged and active in politics (Barreto & Woods 2005; Cho, Gimpel, & Wu 2006). This work focuses on ethnic and racial discrimination, rather than gender-based discrimination. This is a useful starting point to theorize about the potential effect of gender discrimination against women on political engagement, but we also know that gender and race are not equivalent marginalized identities. Furthermore, the extensive psychology literature has documented the negative effects of sexual harassment and sexism on women. Discriminatory events take a toll on women’s mental health and cognitive functioning (Klonoff & Landrine 1995). Social psychologists have found that experiences with interpersonal discrimination can lead to depression, as well as feelings of inferiority and lowered self-esteem (Branscombe, Schmitt, & Harvey 1999; Verkuyten 1998). These are all consequences that are inconsistent with high levels of political engagement. I argue that gender-based discrimination can both act as an impetus and an impediment to political engagement. Sexist experiences are impactful events that women internalize (Klonoff, Landrine, & Campbell 2000; Miles-McLean et. al. 2015; Swim, Hyers, Cohen, & Ferguson 2001). They are constant reminders of the power differentials between men and women. I posit that group consciousness moderates the relationship between sexist events and engagement. For women who are high in gender 40 consciousness, gender-based discrimination leads to higher levels of engagement. For women who are low in gender consciousness, gender-based discrimination will have no effect on engagement, and could potentially even lead to decreased engagement. I explore this theory empirically using data collected from an online sample of 311 women living in the United States. I also consider how these findings relate to the #MeToo movement and the political mobilization around combatting sexual harassment and sexual violence. I find support for the notion that sexist events spur political engagement when paired with higher levels of gender consciousness. Taken together, these results suggest that scholars of gender and politics, as well as political behavior more generally, should consider the role of discrimination and everyday prejudice in the study of political engagement. Gender-based Discrimination and its Consequences Before explicating the connection between gender-based discrimination against women and political engagement, it is first essential to understand the nature and prevalence of discrimination. The crux of discrimination is that it reinforces existing systems of dominance and subordination. Krieger (1999) states, “Discrimination is a socially structured and sanctioned phenomenon, justified by ideology and expressed in interactions, among and between individuals and institutions, intended to maintain privileges for members of dominant groups at the cost of deprivation for others” (p. 201). Discrimination can broadly be classified into two categories. As Oksooii (2016) notes systemic or political discrimination, “typically refers to discriminatory laws, campaign messages, policies, or practices carried out by state or private institutions and/or their affiliated actors” (p. 616). In the context of discrimination against women, examples of 41 systemic discrimination include laws that limit women’s reproductive rights and institutional features in politics that keep women from running for public office. Interpersonal discrimination occurs between individuals. It can come from family, friends, or strangers in private or public settings. The focus of this work is on women’s experiences with interpersonal discrimination as I am interested in the effect of seemingly non-political factors on political outcomes. Furthermore, interpersonal discriminatory events such as street harassment, sexist jokes, and workplace harassment are near daily experiences for many women (Fitzgerald 1993; Klonoff & Landrine 1995; Pew 2017). Importantly for this study, they are also experienced directly in a way that systemic discrimination is often not. Interpersonal gender-based discrimination can take a multitude of forms. Scholars have documented the ubiquity of sexual harassment and discrimination in the workplace (Fitzgerald 1993), sexist comments (Swim et al. 2001), and both interpersonal and media sexual objectification (Fredrickson & Roberts 1997; Kozee, Tylka, Augustus-Horvath, & Denchik 2007). Everyday sexist events include sexist degradation such as being called a sexist name, hearing sexist jokes, or being disrespected because of one’s identity as a woman. A sexist event can also include discrimination in both distant and close relationships, as well as in the workplace (Klonoff & Landrine 1995). Utilizing the Schedule of Sexist Events that measures these experiences, Klonoff and Landrine found that 99% of the 631 women surveyed reported experiencing sexist events at least once in their lifetime, and 97% reported experiencing sexist events at least once within the past year. The most commonly reported events were being forced to listen to sexist or sexually degrading jokes (94%), being sexually harassed (82%), being called sexist 42 names (82%), and being treated with a lack of respect (83%). More recent data shows similar patterns. According to a 2015 survey of 2,235 full-time and part-time women employees, 1 in 3 women had experienced sexual harassment (Huffington Post). A 2018 survey conducted by the non-profit, Stop Street Harassment, found that 81% of women experienced some form of sexual harassment (NPR 2018)12 and a 2017 survey conducted by Pew found that 42% of women in the United States say that they have faced discrimination on the job because of their gender. These estimates are likely rather conservative given that women are often reluctant to label an event as sexist or as constituting sexual harassment (Magley, Hulin, Fitzgerald, & Denardo 1999). Perceived sexist events are considered to be stressors and are linked to psychological issues for those who experience these events (Klonoff & Landrine 1995; Swim et al. 2001). Counseling psychologists have long theorized about the potential for psychological distress due to experiences of discrimination and oppression (Clark et al. 1999; Landrine & Klonoff 1996). Klonoff & Landrine (1995) purport that sexist events over a lifetime are distal causes of psychological distress, and more recent sexist experiences are proximal predictors of distress.13 The authors find that both recent and lifetime perceived sexist events were related to obsessive-compulsivity and anxiety. Lifetime sexist experiences were also related to somatic symptoms, and recent events were related to depressive symptoms. Klonoff and Landrine (1995) find that sexist events actually have a greater negative impact on women’s physical and mental health than 12 The Stop Street Harassment Survey broadly defined sexual harassment to include verbal sexual harassment, unwelcome sexual touching, cyber-sexual harassment, being physically followed, unwanted genital flashing, and sexual assault. 13 The authors also note that more significant events, such as physical sexist discrimination like rape and sexual assault, regardless of the time of occurrence, also act as proximal predictors of psychological distress. 43 generic stressors, even when controlling for appraisal, social support and coping style. The authors (1995) state, “This is because sexist events are inherently demeaning, degrading, and high personal; they are attacks upon and negative responses to something essential about the self that cannot be changed: being a woman. Sexist discrimination thereby has a higher potential to erode women’s physical and mental health” (p. 442). Victims of workplace sexual harassment experience negative consequences as well, such as job dissatisfaction and absenteeism (Fitzgerald et al. 1997; Schneider, Swam, Fitzgerald, & Murphy 1997). They also experience nervousness, anger, low self- esteem, and elevated stress (Crocker & Kalemba 1999). Everyday sexism has ramifications as well. Swim et al. (2001) found that women experience approximately one to two impactful sexist incidents per week.14 These incidents affect women’s psychological well-being by decreasing their comfort, increasing feelings of anger and depression, and decreasing self-esteem. Simply put, more experiences with sexist events leads to more psychological distress for women (Fischer & Holz 2007; Klonoff, Landrine, & Campbell 2000; Szymanski, Gupta, Carr, & Stewart 2009). The psychology literature shows that accumulated marginalizing experiences have consequences on women’s mental health, cognitive functioning, and their behavior. Everyday incidents make up the fabric of people’s lives. For members of traditionally oppressed groups, everyday prejudiced events are a significant part of these experiences. These incidents take place at home, with family or friends, at work, or out in public. Like another other daily hassle or stressful life event, they have a non-trivial impact on psychological well-being (Swim et al. 2001). 14 The authors conceptualize sexist incidents as gender role stereotypes and prejudice, degrading comments and behaviors, and sexual objectification. 44 Sexist Events, Political Engagement, and Group Consciousness The scant research on discrimination and political engagement would lead us to think that perhaps gender discrimination mobilizes women to become more involved in politics. We know from the literature on race and ethnicity that perceptions of prejudiced treatment can spur political action (Simpson & Yinger 1985; Barreto & Woods 2005, Cho et al. 2006). On the other hand, it’s possible that gender discrimination dampens political engagement among women. As previously discussed, these events have negative psychological consequences. They are often demoralizing experiences that lower women’s self-esteem, cause anxiety, and remind women of their subjugated status in society. These are all things that seem inconsistent with a state of high political engagement. This conclusion is supported by recent work on Muslim-Americans’ experiences with interpersonal discrimination and sociopolitical behavior (Oskooii 2016). This study finds that interpersonal discrimination actually causes some to retreat from the political sphere, whereas political or structural discrimination is associated with mobilization. I advance the theory that marginalizing events can either stifle political engagement or be a motivating force that increases engagement. Women are not a monolith and the effect of discrimination on political engagement may not be constant across all women. I argue that the key moderator in the relationship between discrimination and engagement is gender consciousness. Essentially, to become mobilized by discrimination, one has to acknowledge that they are indeed subject to 45 discriminatory events that are based in sexism and patriarchal societal norms. Appraisal of discriminatory events and how women perceive their status as part of a marginalized group is key. The mechanism that mobilizes racial and ethnic groups subject to discrimination to engage in politics is a sense of group consciousness and linked fate (Jamal 2005; Lin 2018; Ysseldyk et al. 2014). Group consciousness is considered to be a political resource that explains the high levels of political participation among some marginalized groups (Sanchez & Vargas 2016). I argue that this is the case for women as well, although there is more heterogeneity in levels of group consciousness.15 Group consciousness involves three major factors (Miller, Gurin, Gurin, & Malanchuk 1981). First, it involves identification with a particular group or social stratum. Secondly, group consciousness requires a level of political awareness about the group’s position in society relative to others. Finally, it involves a commitment to collective action with the goal of realizing the group’s interest. Gender consciousness is a sense of belonging to women as a social group or social identity, as well as having a psychological connection to one’s gender identity (Conover 1988). Cassese and Holman (2016) explain, “Gender consciousness, and social identity more generally, is a multifaceted construct consisting of factors like perceptions of group-based discrimination and feelings of linked fate with other group members” (p. 516). For group consciousness to be present, one must have a sense that their group is subject to unfair discrimination and that their own well-being is inextricably tied to the well-being of the entire group. This feel of linked fate, which is conceptually very similar to group 15 Past findings indicate that the varied level of group consciousness among women is in part due to the high levels of contact and close relationships that most women have with men (Henderson- King & Stewart 1994; Gurin, Miller, & Gurin 1980). Henderson-King and Stewart (1994) also note that gender functions as both a personal and social identity and because gender is central to one’s personal identity, women are less likely to develop high levels of group consciousness. 46 consciousness, has been shown to motivate political engagement. Members of marginalized groups are more aware of group membership than are dominant group members. This often leads to the development of group consciousness (Duncan 1999; Gurin 1985). The relationship I propose is illustrated in Figure 1. Figure 3.1: The Relationship Between Sexism and Political Engagement I argue that women who are high in gender consciousness will become galvanized by sexist and discriminatory events and engage in political participation as a response. If one acknowledges their own status as a marginalized person, they are more likely to attribute sexist events to a society that systematically discriminates and disadvantages women. They are also more likely to see political action as a way to cope with and Experiences of Sexism and Gender-based Discrimination Political Engagement/Efficacy/Behavior Moderator: Gender Consciousness 47 confront marginalization.16 Additionally, there is evidence that recent sexist experiences combined with a commitment to social change is related to lower levels of self-silencing (Watson & Grotewiel 2016). Marginalizing events combined with high levels of gender consciousness is likely to be positively associated with political engagement. Women who are low on gender consciousness will not become mobilized by the same accumulation of discriminatory experiences. They likely do not attribute these events to a larger system that disadvantages and marginalizes women. It is also possible that they do not see these events as constituting sexism or discrimination. For these women, sexist events could potentially depress political engagement. These experiences can lower self-esteem, cause feelings of inferiority and powerlessness, and depression. My argument emphasizes the protective role of feelings of linked fate, an acknowledgement of one’s membership in a marginalized group, and the belief that women have less access to power and resources than men. Scholars have found that even the perception of gender inequality can promote political participation and engagement among (Bernstein 2005). It is likely that actually experiencing discrimination and sexism would have the same effect. However, I hypothesize that this effect is conditional on gender consciousness. 16 Of course, political activism is not the only way in which women respond to or cope with gender-based discrimination. There is a sizeable literature on how women cope with sexist experiences (Fitzgerald 1990; Fitzgerald et al. 1993). These coping mechanisms can include more active mechanisms like confronting the harasser, discussing the event with friends, or filing a formal complaint. Passive coping mechanisms include ignoring the behavior, engaging in self- blame, and deeming the event as benign. However, literature suggests that group consciousness spurs political activism (Duncan 1999). Women who are high in group consciousness are more likely to see engaging in politics as a way to cope with or address discrimination. 48 Data and Measures Participants To test the proposed hypotheses, in March of 2018, I recruited 311 participants through Amazon Mechanical Turk (See Appendix A.1 and A.2 for details regarding the recruitment process and more detailed demographic information). MTurk provides a more representative sample than in-person convenience samples (Berinsky, Huber, & Lenz 2012). All participants were pre-screened such that only those who identified as women were able to complete the survey. Participants ranged from age 22 to 75 (M=39.67, SD=11.31), and 57.5% of the sample had a bachelor’s degree or higher. Out of the 311 participants, 224 (72.03%) identified as White, 33(10.61%) as Black, 33 (10.61%) as Asian, 4 as Latino/a or Hispanic (1.29%), and 21 (6.75%) as multi-racial. According to United States Census data (2017), this sample is relatively representative of the population, although significantly more educated. The median reported yearly income was “$35,000 to $64,999.” Participants were compensated $1.20 for completing the survey. Gender Discrimination After consenting to participate in the study, participants first completed the Interpersonal Sexual Objectification Questionnaire (Kozee et al. 2007). The ISOS is a measure of objectification that occurs through personal interactions and relationships. The first 11-items in the scale are a measure of body evaluation (i.e. “How often have you heard a rude, sexual remark made about your body?”, “How often have you overheard inappropriate sexual comments made about your body?”) and the last 4-items are a measure of unwanted explicit sexual advances (i.e. “How often have you been 49 touched or fondled against your will?”, “How often has someone made a degrading sexual gesture towards you?”). Response categories were “never,” “rarely”, “occasionally,” “frequently,” and “almost always.” Maximum Likelihood Factor Analysis was conducted on the items from the ISOS, and latent factors were allowed to correlate. I opted to conduct factor analysis to allow items on the scale to have different weights. The sample size of 311 exceeds recommendations that sample sizes of at least 200 are sufficient for CFA (Kline 2005). Maximum likelihood estimation is also robust to multivariate non-normality (Weston & Gore 2006). Scores on the scale ranged from -1.49 to 3.32 with a median score of -.075 and a standard deviation of .92. Participants then completed 14-items from the Schedule of Sexist Events (Klonoff & Landrine 1995).17 The SSE was designed to measure daily sexist events across a wide range of different domains (i.e. “How often have you been treated unfairly by your employer, boss, or supervisors because you are a woman?”, “How often have you heard people making sexist jokes, or degrading sexual jokes?”). The scale was developed to be analogous to other stressful and negative life events.18 Maximum Likelihood Factor Analysis was also conducted on the items from the SSE. Latent factors were allowed to correlate. Scores on the sale ranged from -1.20 to 3.46 with a median score of -1.39 and a standard deviation of .92. These two scales are two slightly different operationalizations of gender discrimination. The ISOS scale taps objectifying events, body evaluation, and 17 Factor analysis has revealed that the SSE yields three reliable factors, “Sexist Degradation and It’s Consequences,” “Unfair and Sexist Events at Work/School,” and “Unfair Treatment in Distant and Close Relationships” (Matteson & Moradi 2005). I also used only the SSE-recent as to not fatigue participants. 18 An important feature of the SSE and ISOS is that the majority of the items do not explicitly ask about “sexism.” This alleviates concerns that perhaps women who are more conservative and do not interpret these experiences as sexist would then underreport. Indeed, research shows that women are often reluctant to label an event as sexist or as constituting sexual harassment (Magley et al. 1999). 50 unwanted sexual advances whereas the SSE taps unfair treatment in the workplace, gender stereotypes, and workplace harassment. In the subsequent analyses, I look at sexist and objectifying events separately to see if there are differential effects based on the type of gender discrimination. After completion of the SSE, participants answered questions from the Coping with Sexual Harassment Scale (Fitzgerald 1990; Fitzgerald et al. 1993) and a Big Five Questionnaire. Measuring Gender Consciousness: Feminist Identity Development Participants completed a modified version of the Feminist Identity Development Scale (Bargad and Hyde 1991).19 The full FIDS scale is 48-items, so in the interest of not fatiguing participants, 21-items that best tapped gender consciousness, linked fate, and perceived discrimination were included (i.e. “I’ve never really worried or thought about what it means to be a woman in this society.”, “I used to think there wasn’t a lot of sex discrimination, but now I know how much there really is.”, “I feel angry about the way that women have been left out of the history text books.”). Responses were coded such that a higher value indicated a “more gender conscious” response. The items displayed a high degree of internal validity (𝛼=.921). Maximum likelihood factor analysis was also conducted that revealed that most of the variance was explained by one factor. Scores on the scale ranged from -1.98 to 1.80 with a median score of -.019 and a standard deviation of .95. Importantly, all but one of the items on the scale did not explicitly ask about “feminism.” This was done for a few reasons. First, the goal was to tap gender 19 There is a relatively extensive literature on the convergent validity, temporal stability, and internal consistency and reliability of the FIDS (Bargad and Hyde 1991; Gerstmann and Karmer 1997; Moradi and Subich 2002) 51 consciousness and a sense of linked fate more so than identification with the feminist label. Secondly, because the term “feminist” is politicized, there was concern that including it in some of the question items would simply be capturing the fact that highly identified feminists are simply more likely to be politically engaged in the first place. Finally, this takes care of some concern that this operationalization is just a correlate of ideology or partisanship.20 This sub-set of questions from the FIDS captures a variety of individuals who psychologically identify with women as a group while avoiding any obvious ideological connotations. Political Engagement The primary dependent variables are questions tapping political engagement. I utilized commonly used measures of internal political efficacy (Niemi et al. 1991) and political interest (Shani 2012). Responses were coded such that higher scores corresponded to more efficacious and interested responses. Maximum likelihood factor analysis was conducted on the internal political efficacy items. Scores on the scale ranged from -2.26 to 1.68 with a median of .18 and a standard deviation of .94. A summated rating scale was created with the two items tapping political interest. Scores ranged from 2 to 7 with a median score of 3.13 and a standard deviation of .83. A measure of political participation was created using four dichotomous questions about engaging in various acts of political participation within the past year (signing an online petition, attending a public rally or demonstration, donating money to a political campaign or cause, and posting on social media about an issue that matters to you). Maximum likelihood factor 20 Factor scores from the FIDS scale still correlated modestly with ideology and party identification. Ideology and party identification are included as control variables in all regression models. 52 analysis was conducted on the four political participation measures. Scores ranged from - 1.1 to 1.31 with a median of .18 and a standard deviation of .84. As a robustness check, I created an additive scale with the four dichotomous political participation measures and estimated the model as an ordered logit (Appendix A.5). Substantive results were unchanged. Control Variables Education was measured on a four point scale (“high school grad,” “some college,” “college grad,” and “post-college grad”). Income was measured on a five point scale (“under $15,000,” “$15,000 to $34,999,” “$35,000 to $64,999,” “$65,000 to $149,999,” and “over $150,000”). Ideology was measured on a six point scale coded such that higher scores indicate more conservative political ideology (“very liberal,” “liberal,” “somewhat liberal,” “somewhat conservative,” “conservative,” “very conservative”). Race was dummy coded such that 1 indicated those who identified as white. Party identification was dummy coded such that 1 indicated those who identified as Democrats. Age was measured in years. The Relationship Between Gender Discrimination and Feminist Identity Development The sections above introduced the argument that gender discrimination leads to increased political engagement for women high in gender consciousness. There is plausible concern that perhaps women who are high in gender consciousness are already more politically engaged and simply report experiencing more gender discrimination because they are more attuned to and sensitive to sexism. To address this potential 53 confound, it is useful to look at the reported levels of gender discrimination across the feminist identity development scale. Figure 2 depicts the reported levels of both sexist and objectifying events over the range of the feminist identity development scale. We see that for those higher on the feminist identity scale seem to report slightly more sexist events. The reported amount of objectifying events is relatively similar across varying levels of feminist identity development. This provides relatively solid evidence that an over-reporting of gender discrimination among those high in gender consciousness is not driving the results of the analyses. It is also worth noting that there are only weak positive correlations between sexist events and feminist identity development (r=.330) and objectifying events and feminist identity development (r=.202). 54 Figure 3.2: Reported Sexist and Objectifying Events Over the Range of Gender Consciousness Results To test the hypothesis that there is a relationship between sexist events and political engagement and the moderating effect of feminist identity development, I regressed internal political efficacy on the Schedule of Sexist Events (SSE) factor scores, the gender consciousness factor scores, and the interaction between the two. I controlled for education, income, ideology, race, party identification and age. The full models can be found in Appendix A.3, Table1. Strong support was found for Hypothesis 1. Although the sexist events term was not significant, the interaction between sexist events and gender consciousness was positive and significant (p<.05). 55 Due to the difficult of interpreting interaction terms, this finding is better illustrated graphically in Figure 3. Essentially, Model 1 shows that as gender consciousness increases, so does the estimated coefficient for sexist experiences. In other words, for women who are relatively high in gender consciousness, sexist experiences actually become galvanizing events that increase their levels of internal political efficacy. For women who are low in gender consciousness, sexist experiences have no impact on efficacy. To check the robustness of this finding, I ran the same model, but with the Interpersonal Sexual Objectification Scale (ISOS) factor scores.21 This model also provides support for my hypothesis. Gender consciousness moderates the relationship between sexist/objectifying events and internal political efficacy. Tests of linearity assumptions can be found in Appendix A.4 using diagnostics suggested by Hainmueller, Mummolo, and Xu (2016). 21 The ISOS taps the same underlying concept as the SSH, but gets at more explicitly sexually objectifying experiences like unwanted sexual advances and body evaluation. 56 Figure 3.3: Marginal Effect of Sexist Events (left) and Interpersonal Objectification (right) on Efficacy over the Range of Gender Consciousness22 To further evaluate the relationship between discrimination and political engagement, I regressed political interest on sexist events (SSE) and objectifying events (ISOS) factor scores, the gender consciousness factor scores, the interaction between the two, and the same set of covariates specified in the first two models. A similar pattern emerges in which women who are high in gender consciousness experience more political interest as sexist events increase. Women who are low in gender consciousness seemed to display lowered interest as sexist events increase, although this relationship is not statistically significant. This relationship is less pronounced when we look at 22 These figures are showing us that as gender consciousness increase (x-axis), the coefficient for sexist/objectifying increases (y-axis). In other words, gender consciousness bolsters the positive effect that sexist/objectifying experiences have on internal political efficacy. I also looked at the marginal effect of sexist/objectifying experiences on the FIDS and there is a symmetrical effect. 57 objectifying events, but the general pattern holds. The full regression models can be found in Appendix A.3, Table 2. Figure 3.4: Marginal Effect of Sexist Events (left) and Interpersonal Objectification (right) on Political Interest over the Range of Gender Consciousness Internal political efficacy and political interest are measures of political engagement, but can experiences with sexist events actually affect political behavior? To assess this question, I created a variable using four different dichotomous measures of political activity.23 The political activity factor scores were regressed on the same set of covariates specified in the previous models. Full results can be found in Appendix A.3, Table 3. Once again, there is a strong relationship between sexist experiences and 23 The question wording was, “People engage in social, civic, and political activity in different ways...... Over the past 12 months, have you done any of the following, or not?” The four activities were “signed an online petition,” “attended a public rally or demonstration,” donated money to a campaign or cause,” and “posted on social media about an issue that matters to you.” 58 political activity (p<.01). Women who experience more sexist and objectifying events engage in more political activity. Again, this relationship appears to be moderated by gender consciousness. Women who are high on the gender consciousness scale are the most mobilized by sexist experiences. This pattern does not hold for objectifying events, although there is a positive and significant main effect of objectifying events on political participation (p<.01). Figure 5 shows that as gender consciousness increases, so does the estimated coefficient for sexist experiences. This relationship is much less pronounced when using the objectifying events scale. Figure 3.5: Marginal Effect of Sexist Events (left) and Interpersonal Objectification (right) on Political Participation over the Range of Gender Consciousness To results presented so far have demonstrated that there is a strong relationship between sexist and objectifying events and political participation and engagement that is 59 moderated by feminist identity development. To further illustrate this moderating effect, Figure 6 depicts the level of participation across sexist and objectifying events for women at varying levels of gender consciousness. We see a particularly sharp difference between women who are one standard deviation about the mean gender consciousness score and those who are one standard deviation below the mean gender consciousness score. Figure 3.6: Predicted Political Participation at Varying Levels of Gender Consciousness 60 Discussion The goal of this study was to investigate the connection between gender discrimination and political engagement, and the moderating impact of group consciousness. Political scientists have focused almost exclusively on the impact that sexism has on the electoral success of elite female candidates, while paying scant attention to how marginalization on the basis of gender affects women generally. Results indicate that personally-experience gender discrimination can mobilize women to become politically engaged. However, this relationship is conditional on gender consciousness. Women who were low on gender consciousness were not mobilized by sexist events. These findings are in line with literature on black Americans experiences with everyday prejudice. Prejudice and marginalization spur social organizing and political activism (Simpson & Yinger 1985; Swim, Cohen, & Hyers 1998). These results are also consistent with the limited data on personally-experienced sexism and political engagement (Bankert & Williamson 2017).24 There are limitations to this study. Due to the observational nature of this work, we cannot be completely confident that the relationship between sexist and objectifying events and the dependent variables is causal. In the future, studies could employ experiments to further investigate the causal link between marginalizing events and political engagement and participation. Additionally, the sample is not totally representative of the broader population of American women, which limits this study’s 24 These authors find that women who have experienced gender discrimination report higher levels of political interest and participation, however, this finding only holds for liberal women. 61 generalizability. It is significantly more liberal and educated, both factors that could be associated with a higher likelihood of reporting marginalizing events. Despite these limitations, the connection between marginalizing events and political engagement and participation is robust. The correlation holds up to a variety of controls and across several different measures of political engagement with two different operationalizations of gender discrimination. This research highlights the continued relevance of personally-experienced discrimination and marginalization to scholars are political behavior. The everyday lived experience of marginalized groups impacts the way in which they approach the political sphere. The evidence presented here comports with the literature on the racial and ethnic discrimination that shows how everyday prejudice and marginalization can increase political engagement and social activism. However, these findings also underscore the way in which women are not a monolith that behave politically in a cohesive way. For women, gender consciousness is an important variable that makes sexist and discriminating events mobilizing. Women who have a high degree of gender consciousness recognize that discriminatory events relate to a larger system that disadvantages and marginalizes women. They are more likely to see politics as an avenue for redress. Women who do not possess a high degree of gender consciousness do not see their own fate as linked to the fate of women generally. Consistent with system justification theory, they may not even see these events as problematic or discriminatory (Jost & Kay 2005). These results suggest that scholars of political behavior and gender and politics should consider personally-experienced discrimination as a predictor of political engagement. 62 CHAPTER 4 THE RELATIONSHIP BETWEEN GENDER DISCRIMINATION AND SOCIOPOLITICAL BEHAVIOR Abstract The extant literature on discrimination and political behavior suggests that perceptions of marginalization and prejudice increase political engagement and activism. The vast majority of this literature focuses on marginalized racial, ethnic, or religious identities. There is virtually no research on the impact of gender-based discrimination on sociopolitical behavior. Furthermore, much of this work is observational in nature, limiting the causal claims we can make regarding personally-experienced discrimination and political behavior. This paper explores the relationship between personally- experienced gender discrimination and political engagement and activism. Using a laboratory experiment, participants received bogus feedback after a cognitive task that led them to believe they performed poorly. Women in the treatment group were told that their gender accounted for their poor performance, priming personally experienced sexism. I consider the role of gender consciousness and linked fate, emotional response, and ideology as moderators in the relationship between exposure to sexism and political engagement. I also measure physiological response in the form of skin conductance (SCL), a common measure of emotional arousal or attentiveness generated by the sympathetic nervous system, to examine subjects’ response to exposure to sexism. 63 The role of sexism and sexual harassment in shaping women’s everyday lives has never been a more salient topic in the media and in our political culture. We are only beginning to understand the prevalence of gender discrimination, sexism, and sexual violence. Despite the acknowledgement that discrimination and gender-based marginalization is pervasive, there is very little work done that seeks to understand how this discrimination impacts political behavior. How does gender discrimination, something that constantly punctuates the lives of so many women, affect how they approach the political sphere? It is well-established in the race and politics literature that experiences of discrimination can motivate individuals to take part in politics (Simpson & Yinger 1985; Barreto & Woods 2005; Cho, Gimpel, & Wu 2006; Oksooii 2016). However, as Oskooii (2016) states, Yet, despite the historical present prevalence of prejudice in both social and political domains, the direct relationship between discrimination and sociopolitical behavior is still relatively understudied. Discrimination is rarely the focal point of the most comprehensive studies related to the civic and political engagement of minorities. Consequently, our understanding of how discrimination affects the political behavior of marginalized individuals is limited (p. 613). This dearth of literature on the relationship between discrimination and political engagement is particularly true in the case of gender-based discrimination. This is 64 surprising given the fact that women are a historically marginalized group in the United States that have been discriminated against and oppressed in both the political sphere and the private sphere. The study of the “gender gap” in political engagement has received a significant amount of attention from political science scholars (Verba, Burns, & Schlozman 1997; Burns, Schlozman, & Verba 2001; Lawless & Fox 2010; Pruysers & Blais 2014) and some scholars have made vague reference to the notion that sexism and the way in which women are socialized undoubtedly affects their levels of political engagement (Bennet & Bennet 1989; Burns et al. 2001; Atkeson 2003; Gidengil, Giles, & Thomas 2008; Preece 2016). However, there has been no systematic study of the way in which gender-based discrimination, an omnipresent force in the lives of many women, affects political engagement. Rather, political scientists have focused on the way in which structural disadvantages have driven the gender gap in engagement, and how sexism affects political elites. In one of the only studies linking gender-based discrimination to political action, Duncan (1999) finds that experiences with sexual harassment, which she includes as a measure of “life experiences with oppression,” were associated with higher levels of feminist consciousness, and in turn, increased propensity to engage in women’s rights activism. There are significant limitations to this study. The measurement for sexual harassment was a yes or no question asking if the respondent had experienced sexual harassment at work or school. The dichotomous coding does not allow for variation in how often the respondent had experienced harassment, and the inclusion of only one question does not cover the myriad of ways in which women are discriminated against and oppressed. Furthermore, this work explored the development of feminist 65 consciousness and women’s rights activism. The present study is focused on how group consciousness, opposed to a politicized gender identification like feminist consciousness affects political engagement and participation generally. This article investigates whether gender-based discrimination affects women’s engagement in politics. The race and politics literature would lead us to believe that gender-based discrimination would increase political engagement. However, there are important and distinct differences between race and gender as identities, and the way in which women are marginalized verses the way people of color of marginalized. The majority of the race and politics literature has found that discrimination is politically mobilizing (Barreto & Woods 2005; Cho et al. 2006; Simpson & Yinger 1985). However, most of this literature does not differentiate between systemic discrimination and personally-experienced discrimination. More recent literature has found that personally-experienced discrimination in particular can be politically demobilizing (Oskooii 2016). To test whether personally-experienced discrimination impacts political engagement, I employ a randomized laboratory experiment in which personally- experienced discrimination is simulated. I consider the role of emotional response, physiological response, gender consciousness, and ideology in the relationship between discrimination and political engagement. In line with the race and politics literature, I find that discrimination can be politically mobilizing. However, this relationship is conditional on gender consciousness.25 These findings have broad implications for the 25 For a detailed discussion of gender consciousness and how it related to political engagement and participation, see Chapter 3. 66 literature on political behavior and the predictors we consider when studying political engagement and mobilization. Sexism, Discrimination, and Political Engagement Women are discriminated against in a multitude of ways. Gender discrimination encompasses things like sexism and harassment, but can also include discrimination on the basis of gender identity or sexual orientation. For the purposes of this work, I am focused on the ways in which those who identify as women experience discrimination. Even more specifically, I am interested in the everyday discriminatory events and micro- aggressions that punctuate women’s lives. Interpersonal discrimination is particularly insidious as it cannot always be dealt with in a legal fashion in the way that overt forms of discrimination can. Research shows that for many women, events such as street harassment, sexist comments and jokes, and workplace harassment are comment experiences (Fitzgerald 1993; Klonoff & Landrine 1995; Pew 2017). At first the relationship between personally-experienced discrimination and political participation may seem like an unlikely one, particularly because marginalized groups tend to have fewer resources associated with political engagement (Brady et al.1995). However, the literature on race, ethnicity, and politics has found a robust correlation between discrimination and political mobilization. Scholars find that a sense of group consciousness and linked fate, in part, explains the mobilization of racial and ethnic groups subject to discrimination (Jamal 2005; Lin 2018; Ysseldyk et al. 2014). Group consciousness is considered to be a political resource that explains the high levels of political participation among some marginalized groups that we would otherwise expect not to participate (Sanchez & Vargas 2016). Given these findings, I hypothesize 67 that group consciousness moderates the relationship between sexism and political engagement.26 𝐻!: The relationship between exposure to sexism and political engagement is moderated by gender consciousness. 𝐻!!: For women who are high in gender consciousness, exposure to sexism will increase political engagement and political activity. 𝐻!!: For women who are low in gender consciousness, exposure to sexism will have no effect on political engagement and political activity. I also propose that ideology moderates the relationship between sexist events and political engagement. On the 2016 ANES, liberal women were more likely to report that they experienced gender discrimination in comparison to conservative women (Bankert & Williamson 2017). This is not likely to be due to actual differences in experiences, but rather, perceptions of those experiences as sexist. Research indicates that some women are indeed reluctant to label an event as sexist or as constituting sexual harassment (Magley, Hulin, Fitzgerald, & Denardo 1999). Liberal women might be more likely to perceive the feedback as sexist and therefore also be more likely to be mobilized by it. This leads to my second set of hypotheses: 𝐻!: The relationship between exposure to sexism and political engagement is moderated by ideology. 26 I use the terms “discrimination” and “sexism” as interchangeable as I consider discrimination to reinforce existing systems of dominance and subordination. Krieger (1999) states, “Discrimination is a socially structured and sanctioned phenomenon, justified by ideology and expressed in interactions, among and between individuals and institutions, intended to maintain privileges for members of dominant groups at the cost of deprivation for others” (p. 201). Sexism is one manifestation of discrimination. 68 𝐻!!: For liberal women, exposure to sexism will increase political engagement and political activity. 𝐻!!: For conservative women, exposure to sexism will have no effect on political engagement and activity. It is likely that women who perceive the feedback on the cognitive task as sexist will be more emotionally aroused by it. Perceived sexist events are stressors and are linked to psychological issues and negative emotions like anger and obsessive- compulsivity (Klonoff & Landrine 1995; Swim, Cohen, & Hyers 2001). We know from the political science literature that some emotions are associated with increased political mobilization. Specifically, we know that anger is a mobilizing emotion (Valentino et al. 2011). Anger tends to boost political efficacy (Valentino, Gregorowicz, and Groenendyk 2009; Weber 2008). This leads to my third hypothesis: 𝐻!: Women who report increased feelings of anger after exposure to sexism will be more politically engaged than women who do not feel increased anger. Although self-reported emotional responses are useful, people are often motivated to misrepresent how they really feel, or they have unconscious emotional responses to stimuli that they may not even be cognizant of. For this reason, I also collect a physiological measure in the form of skin conductance levels (SCL), an unobtrusive measure of emotional arousal (Dawson, Schell, & Filion 2007). Although our physiological measure will not give us the valence of the emotion, both strong positive, 69 like enthusiasm, and strong negative emotions, like anger, can motivate political participation (Valentino et al. 2011). Therefore, my final set of hypotheses are:27 𝐻!!: Women who are more physiologically aroused by exposure to sexism will be more politically engaged. 𝐻!!: Women who are less physiologically aroused by exposure to sexism will be less politically engaged. Research Design To explore the relationship between personally-experienced gender discrimination and political engagement we could ask a representative sample to report how often they experience discriminatory events, as well as questions about political engagement and participation. However, we would not be sure that this relationship is causal. It is possible that women who are more politically engaged tend to systematically report more sexist and discriminatory experiences. Therefore, to test my hypotheses I employ an experiment in which personally-experienced sexism is simulated in a laboratory setting. The study was conducted in an experimental lab at a large urban university in the Northeast in the Winter of 2018-2019. A total of 71 participants who identified as women were recruited for the study.28 Participants were compensated with a $5 Amazon gift card for the 20-minute study. Upon signing the informed consent, participants completed a pre-test survey with demographic items, items from the Schedule of Sexist Events (Klonoff & Landrine 1995) and the Interpersonal Sexual Objectification Scale (Kozee, 27 This set of hypotheses will not be presented in the results section of this dissertation as the physiological data has not been analyzed yet. 28 Four participants were excluded from the analyses for not completing the post-test. 70 Tylka, Augustus-Horvath, & Denchik 2007). They also completed items from the Coping with Sexual Harassment Questionnaire (Fitzgerald 1990; Fitzgerald et al. 1993), as well as the Feminist Identity Development Scale (Bargad & Hyde 1991), a Big Five Personality measure, and an Issue Attitudes questionnaire. This inclusion of the latter two batteries were added to create some space in between the ISOS and the SSE and the treatment to reduce any possible demand effects.29 Upon completion of the pre-test survey, participants were connected to the Biopac system by trained research personnel. Two sensors were placed on the middle and pointer fingers to measure individual difference in electrodermal response to the sexist feedback using skin conductance levels (SCL). This is a standard approach to gauging physiological sensitivity (Dawson, Schell, & Filion, 2007). SCL is a measure of electrodermal activity or the degree to which people sweat, which is an index of people’s emotional arousal generated by the sympathetic nervous system. Because the sympathetic nervous system is difficult to control through conscious effort, SCL provides an unobtrusive measure of emotional arousal caused by the feedback. After being connected to the Biopac, participants completed a brief cognitive task called the Nonsense Syllogisms Test, which is used as a measure of logical reasoning from the Kit of Factor-Referenced Cognitive Tests (Ekstrom et al. 1976).30 After completion of the cognitive task, participants were randomly assigned to either the treatment or control group and were told that they would receive audio feedback on their 29 I used items from the SSE, ISOS, and FIDS scales with the highest factor loadings. To decrease participant fatigue and potential demand effects, I reduced the number of items from each scale. 30 I choose the Nonsense Syllogisms Test because it makes no reference to politics. I am interested in whether or not sexism in women’s daily lives affects their levels of political engagement. This closer approximates these experiences. I would also argue that this is a more difficult test of the hypotheses. 71 performance on the task.31 Participants in the treatment group were given bogus feedback that indicated they did poorly on the task and their poor performance was likely due to their gender.32 Participants in the control group were given bogus feedback that indicated they did poorly on the task, but there was no mention of gender. The feedback statements were pre-tested on Amazon’s MTurk to ensure that the statements were perceived as both sexism and realistic. The full audio feedback transcripts are displayed in Figure 4.1. Figure 4.1: Audio Transcript Control Condition You received a score of 77% on the cognitive task. This is a below average score for this Syllogisms Test. Psychology research shows that this Nonsense Syllogisms Task is a reliable and valid measure of logical reasoning skills. Treatment Condition You received a score of 77% on the cognitive task. This is a below average score for this Syllogisms Test. Psychology research shows that women perform worse on logical reasoning tasks than men. Men tend to have superior logical reasoning skills. After receiving the feedback on the cognitive task, participants completed the brief Discrete Emotions Questionnaire (Harmon-Jones, Bastian, & Harmon-Jones 2016) to measure self-reported emotions and questions measuring political engagement. I measured political interest (Shani 2012) and internal efficacy, (Niemi, Craig, & Mattei 1991). As a measure of political participation, respondents were asked how likely they would be to sign a petition, donate to a political campaign, attend a rally, and post about a 31 Audio feedback was used opposed to written feedback to ensure that we would know exactly when the treatment started and ended for the purposes of analyzing the physiological data. This also ensures that the participants were actually exposed to the treatment. 32 This was done to ensure that the two groups were parallel and the only difference is the mention of gender in the treatment group. 72 political issue on social media. I also included a measure of propensity to engage in a future study that involved discussing politics with a focus group.33 Finally, participants were debriefed regarding the purpose of the survey study, and provided with resources to refer to if they are experiencing emotional distress. Results Measures To measure gender consciousness, I created a summated rating scale with the thirteen gender consciousness items from the Feminist Identity Development Scale (𝛼=.77). The scale ranged from 0 to 1. Higher scores indicated higher levels of gender consciousness. Ideology was measured on a six-point scale with higher numbers indicating a more conservative ideology. Race was dummy coded (1=white, 0=non- white), and education was measured on a five-point scale from “no high school diploma” to “post-graduate degree.” Emotions, including anger, were coded on a seven-point scale with higher scores indicating higher levels of emotion. Finally, summated rating scales were created with the four internal efficacy items (𝛼=.73) and the four items tapping political participation (𝛼=.73) to operationalize political engagement and participation, as well as the two items measuring political interest (𝛼=.83). Therefore, there are three outcome variables of interest: internal political efficacy, political participation, and political interest. 33 The question read, “The Behavioral Foundations Lab is conducting another upcoming study. We are interested in the political attitudes of college students. You will share our political attitudes and opinions have a discussion about these attitudes among your focus group. How interested would you be in participating in this study?” 73 To test the first hypothesis that gender consciousness moderates the relationship between exposure to sexism and political engagement, I regressed my measures for political engagement on gender consciousness, as well as the interaction between gender consciousness and the treatment. I controlled for ideology, race, and education. Figure 4.2 depicts the regression results for all three models with the full set of controls. Full regression results can be found in Appendix A.4. Strong support was found for the first set of hypotheses. There was no main effect of the treatment on internal political efficacy, the interaction between the treatment indicator and gender consciousness is positive and significant (p<.1). However, the model with political participation as the outcome variable displayed very interesting results. The main effect of the treatment was negative and significant (p<.05). However, the interaction between the treatment and gender consciousness was positive and significant (p<.05). Figure 4.3 displays the differential effect that the treatment had on those who were high in gender consciousness verses those who were low in gender consciousness. Again, those who were high in gender consciousness seemed to be mobilized by the sexist feedback. For those who were on the very low end of gender consciousness, the treatment had no effect and is actually trending in a negative direction. With more statistical power, it is likely that the treatment would have had a significantly negative impact. Finally, the variables of interest had no significant effect on political interest. 74 Figure 4.2: OLS Regression Results Figure 4.3: Estimated Coefficient of the Treatment by Gender Consciousness 75 To test my second set of hypotheses, I regressed my measures for political engagement on my treatment variable, ideology, the interaction between the treatment and ideology, as well as my controls. Figure 4.4 shows the results of the regression models with the full set of controls. Regression results can be found in Appendix A.4. Very little support was found for 𝐻!. Ideology did not moderate the relationship between the treatment and political engagement. Although, the interaction between the two variables was negative and significant (p<.1), suggesting that more conservative women in the treatment group displayed less propoensity to engage in politics, the correlation was not robust with the addition of controls. Figure 4.4: OLS Regression Results 76 Lastly, to test the hypothesis that anger moderates the relationship between sexism and political engagement, I ran a pair OLS regressions to determine if anger moderates the relationship between sexism and political engagement. No support was found for 𝐻!. Those who self-reported higher levels of anger after the treatment did not display significantly higher levels of political engagement. Figure 4.5 displays the results of the regression models. Figure 4.5: OLS Regression Results Concluding Remarks Gender-based discrimination and sexism are salient topics in our current political discourse. The #MeToo Movement, the election of Donald Trump, and the nomination of Brett Kavanaugh to the Supreme Court have made this topic more relevant to politics than ever. However, discrimination, harassment, and sexual assault have always been part 77 of women’s lives. Although work in political science has addressed the way sexism affects women political candidates and politicians, there has been no systematic study of how gender-based discrimination and marginalization affects women in the electorate and how they engage in politics. The study presented here is one of the first steps towards establishing a research agenda that explores the role of discrimination in shaping women’s political lives. 78 CHAPTER 5 CONCLUSION “The personal is political for so many.” -Tarana Burke An unprecedented number of women are running for president in the 2020 race. Despite the sexist attacks against Hillary Clinton in the 2016 presidential race, a historical number of women won in the 2018 mid-term elections. Women, many angered by the sexism present in the 2016 election, are running for political office and they are winning. The presence of women in politics at the elite level is important and worthy of study. However, sexism and gender discrimination impacts all women. This dissertation has highlighted the ways in which gender-based discrimination can affect how women engage in politics. The world’s first mass movement against sexual abuse and violence, #MeToo, has shifted the way in which we think about harassment, gender, and privilege. It has reminded us that the personal is inherently political. Sexual harassment and sexual violence have been dealt with in the legal system for many years. As Catherine MacKinnon (2017) notes, the legal breakthroughs that defined sexual harassment and violence as sex discrimination were in many ways a necessary precondition for the #MeToo Movement. However, #MeToo has not been based in legal battles. The battleground of this social movement has been based in the mainstream and social media, as well as on the political world stage. Feminist political theorists have always recognized that the personal is political, and that the multitude of ways in which women are oppressed in the private spheres 79 affects how they engage with the public sphere. In fact, the division of the public and private sphere has often led to issues of violence against women to be “privatized” (Pateman 1985). Under the guise of the protection of liberty, the private sphere became a place of oppression for women. In private, women can still be marginalized, powerless, and be subject to violence (Young 1990). Through a combination of survey and experimental research, I show that the consequences of gender-based marginalization extend to the political sphere in an empirically demonstrable way. Summary of Findings In the first paper, I explored how self-objectification impacts political engagement. Drawing from an extensive literature in psychology on the mental, physical, cognitive, and behavioral impacts of self-objectification, I argue that self-objectification undermines women’s political agency, efficacy, and propensity to engage in political activity. I find support for the hypothesis that state self-objectification is related to lower levels of internal political efficacy, interest, information-seeking, and gender egalitarianism. Consistent with the psychology literature, the negative consequences of self-objectification only held for women in the sample. The strongest correlation was found between body shame and the outcome variables. Body surveillance had no significant impact on any of the dependent variables with the exclusion of information- seeking behavior. This suggests that shame, the more affective component of self- objectification, plays a larger role in impacting political engagement. Body surveillance, which is theoretically linked more to appearance monitoring and the disruption of an individual’s attentional resources, seems to have a less significant effect on engagement. 80 This first empirical article problematizes the objectification of women’s bodies in popular culture and in the media. The evidence presented suggests that women who internalize cultural beauty ideals and turn inward to see themselves as objects, are less likely to become engaged in politics. This work builds on other research that finds that self-objectification disrupts involvement in gender-based activism by increasing support for the status quo (Calogero 2013). Evidence presented here shows that self- objectification does not undermine only gender-based political activism, but political engagement generally. Future research should further uncover the mechanism that links self-objectification and political engagement. The second empirical article focuses on gender discrimination broadly, and its impact on political engagement. Utilizing commonly used psychological measures, I collected large sample survey data on women’s experiences with various forms of gender-based marginalization and discrimination, as well as political engagement measures. Building on the race, ethnicity, and politics literature, I posit that gender discrimination can galvanize political engagement and activity, although the relationship is conditional on gender consciousness. I find support for the notion that sexist events spur political engagement when paired with higher levels of gender consciousness. Personally experienced discrimination and sexism were associated with higher levels of political efficacy, interest, and political activity. This work underscores the continued relevance of personally-experienced discrimination and marginalization to scholars of political behavior. Evidence from the second empirical article comports with the literature on racial and ethnic discrimination and political engagement and social activism. Taken together, 81 these results suggest that scholars of gender and politics, as well as political behavior more generally, should consider the role of discrimination and everyday prejudice in the study of political engagement. These results may also shed light on the increased political activism spurred by the #MeToo Movement. Future research could explore the role of ambient sexism on political engagement, as well as how multiple intersecting identities interact to shape the way in which women respond to discrimination. It would also be fruitful to parse out exactly what type of gender-based marginalizing events are the most mobilizing. The final dissertation article builds upon the findings from the second empirical study. Using a novel experimental design to “prime” personally experienced sexism in a laboratory setting, I find that gender discrimination spurs political engagement. As in the second study, this relationship is conditional on gender consciousness. There was some evidence that ideology also moderated this relationship, with liberal women being more mobilized by the treatment than conservative women, although this finding was not robust to the additional of control variables. These experimental findings make us particularly confident that the relationship between discrimination and political engagement is indeed, causal. Future Work The first contribution of this dissertation is shedding light on the political nature of sexism, sexual harassment, and all forms of gender-based marginalization. These events punctuate women’s lives and have an impact on how women see their role in politics. I present some evidence that when women internalize sexism and turn inwards to 82 see themselves as objects instead of full people with agency and power, political consciousness and engagement can be undermined. Other evidence I offer suggests that marginalizing events can also be mobilizing. Survey and experimental research demonstrates that gender discrimination increase political engagement, although this relationship is moderated by gender consciousness. Women with a stronger sense of linked fate are more mobilized by discrimination. The second contribution of this dissertation is expanding the variables we consider to be predictors of political engagement. The study of political behavior occupies an important place in the field of political science, and many theories have been developed to understand why some people engage in politics while others do not. We know that factors like education, time, and resources are important for political participation. However, these variables do not explain all of the variation in political activity across groups. The exploration of how marginalization shapes political behavior is understudied. As Oskooii (2016) notes, societal discrimination has been the purview of social psychologists. The work presented in this dissertation underscores the need for scholars of political science to consider the impact of discrimination on sociopolitical behavior. 83 BIBLIOGRAPHY Almond, G. A. & Verba, S. (1963). The Civic Culture: Political Attitudes and Democracy in Five Nations. Princeton: Princeton University Press. “American Trends Panel.” Pew Research Center, Washington, D.C. (2017). Atkeson, L., & Rapoport, R. The more things change the more they stay the Same: Examining Gender Differences in Political Attitude Expression, 1952-2000. The Public Opinion Quarterly 67.4 (2003): 495-521. Aubrey, J. S. (2006). Exposure to sexually objectifying media and body self-perceptions among college women: An examination of the selective exposure hypothesis and the role of moderating variables. Sex Roles, 55 (14), 159. Bandura, A. (1997). Personal efficacy in psychobiologic functioning. In G. V. Caprara (Ed.), Bandura: A leader in psychology (pp. 43-66). Milan, Italy: Franco Angeli. Bankert, A., & Williamson, S. (2017). The Differential Effects of Gender Discrimination on Liberal and Conservative Women’s Political Engagement. Presented at the Annual American Political Science Meeting, San Francisco, CA. Bargad, A., & Hyde, J. (1991). Women's Studies: A Study of Feminist Identity Development in Women. Psychology of Women Quarterly, 15(2), 181-201. Barreto, M., & Woods, N. (2005). The Anti-Latino political context and its impact on GOP detachment and increasing Latino voter turnout in Los Angeles county. In G. Segura & S. Bowler (Eds.), Diversity in democracy: Minority representation in the United States (pp. 148–169). Charlottesville: University of Virginia Press. Bartky, S. (1990). Femininity and Domination. United Kingdom: Routledge. Beckwith, Karen. (1986) American Women and Political Participation: The Impacts of Work, Generation, and Feminism. New York: Greenwood Press. Bennett, L., and Bennett, S. (1989). Enduring gender differences in political interest: the impact of socialization and political dispositions. American Politics Research, 17 (1), 105–22. Berinsky, A. J., Huber, G. A., & Lenz, G. S. (2012). Evaluating online labor markets for experimental research: Amazon. com's Mechanical Turk. Political Analysis, 20(3), 351- 368. http://dx.doi.org.libproxy.temple.edu/10.1177/1532673X8901700106 84 Bernstein, A. (2005). Gendered characteristics of political engagement in college students. Sex Roles, 52 (5-6), 299-310. Branscombe, N. R., Schmitt, M. T., & Harvey, R. D. (1999). Perceiving pervasive discrimination among African Americans: Implications for group identification and well- being. Journal of Personality & Social Psychology, 77 (1), 135–149. Burns, N., Schlozman, K., & Verba, S. (2001). The Private Roots of Public Action: Gender, Equality, and Political Participation. Cambridge, MA: Harvard University Press. Calogero, R. M., Davis, W. N., & Thompson, J. K. (2005). The role of self- objectification in the experience of women with eating disorders. Sex roles, 52(1-2), 43- 50. Calogero, R.M. (2013). Objects don’t object: Evidence that self-objectification disrupts women’s social activism. Psychological Science, 24, 312-318. Calogero, R. M., Pina, A., & Sutton, R. M. (2014). Cutting Words Priming Self- Objectification Increases Women’s Intention to Pursue Cosmetic Surgery. Psychology of Women Quarterly, 38(2), 197-207. Calogero, R.M., Tylka, T.L., Donnelly, L.C., McGetrick, A., & Medrano Leger, A. (2017). Trappings of femininity: A test of the “beauty as currency” hypothesis in shaping college women’s gender activism. Body Image, 21, 66-70. Calogero, R.M. (2017). Political consciousness and gender collective action: A case and place for self-objectification. In A.L. Bos & M.C. Schneider (Eds.), The Political Psychology of Women in U.S. Politics (pp. 93-110). New York: Routledge. Carpini, M. X. D., & Keeter, S. (1996). What Americans Know About Politics and Why it Matters. Yale University Press. Cassese, E.C. & Holman, M. R. (2016). Religious Beliefs, Gender Consciousness, and Women’s Political Participation. Sex Roles, (75), 514-527. Chatterjee, R. (2018, February 21). A new survey finds 81 percent of women have experienced sexual harassment. NPR, Retrieved from www.npr.org. Cho, W., Gimpel, J., & Wu, T. (2006). Clarifying the role of SES in political participation: Policy threat and Arab American mobilization. Journal of Politics, 68(4), 977–991. Chou, S. (2018, January). Millions say #MeToo. But not everyone is heard equally. PRI, Retrieved from https://www.pri.org https://www.pri.org 85 Clarke, H.D. & Acock, A.C. (1989). National elections and political attitudes: The case of political efficacy. British Journal of Political Science 19, 551–62. Clark, R., Anderson, N. B., Clark, V. R., Williams, D. R., & Fowler, R. D. (1999). Racism as a Stressor for African Americans. American Psychologist, 54 (10), 805-816. Conover, P. J. (1988). The role of social groups in political thinking. British Journal of Political Science, 18(1), 51–76. Crocker, D., & Kalemba, V. (1999). The Incidence and Impact of Women's Experiences of Sexual Harassment in Canadian Workplaces. The Canadian Review of Sociology and Anthropology, 36 (4), 541-558. Dawson, M. E., Schell, A. M., & Filion, D. L. (2007). The electrodermal system. Handbook of psychophysiology , 2 , 200–223. de Beauvoir, S. (1949). The Second Sex. London: Vintage Classics. Devitt, J. (2002). Framing gender on the campaign trail: Female gubernatorial candidates and the press. Journalism and Mass Communication Quarterly, 79(2), 445-463. Duncan, L. (1999). Motivation for Collective Action: Group Consciousness as Mediator of Personality, life Experiences, and Women’s Rights Activism. Political Psychology, 20(3), 611-635. Ekstrom, R. B., French, J. W., Harman, H. H., & Dermen, D. (1976). Manual for kit of factor-referenced cognitive tests. Princeton, NJ: Educational Testing Service. Fairchild, K., & Rudman, L. A. (2008). Everyday stranger harassment and women’s objectification. Social Justice Research, 21(3), 338-357. Feagin, J., & Sikes, M. (1994). Living with Racism : The Black Middle-Class Experience. Boston: Beacon Press. Fischer, A., Holz, K., & Mallinckrodt, Brent. (2007). Perceived discrimination and women's psychological distress: The roles of collective and personal self-esteem. Journal of Counseling Psychology, 54 (2), 154-164. Fitzgerald, L. F. (1993). Sexual harassment: Violence against woman at work. American Psychologist, 48, 1070-1076. Fitzgerald, L. F. & Shullman, S.L. (1993). Sexual harassment: a research analysis and agenda for the 1990s. Journal of Vocational Behavior, 42, 5-27. 86 Fitzgerald, L.F., Drasgown, F., Hulin, C., Gelfand, M.J., Magley, V.J. (1997). Antecedents and consequences of sexual harassment in organizations: A test of an integrated model. Journal of Applied Psychology, 82 (4), 578-589. Fredrickson, B. L., & Roberts, T.A. (1997). Objectification theory: Toward understanding women’s lived experiences and mental health risks. Psychology of Women Quarterly, 21 (2), 173–206. Fredrickson, B. L., Roberts, T.A., Noll, S.M., Quinn, D. M., & Twenge, J. M. (1998). The swimsuit becomes you: Sex differences in self-objectification, restrained eating, and math performance. Journal of Personality and Social Psychology, 75 (1), 269-284. Galdi, S. Maass, A. & Cadinu, M. (2014). Objectifying media: Their effect on gender role norms and sexual harassment of women. Psychology of Women Quarterly, 38(3), 398-413. Gapinski, K. D., Brownell, K. D., & LaFrance, M. (2003). Body objectification and fat talk: Effects on emotion, motivation, and cognitive performance. Sex Roles, 48 (9-10), 377–388. Gay, R. K., & Castano, E. (2010). My body or my mind: The impact of state and trait objectification on women's cognitive resources. European Journal of Social Psychology, 40(5), 695-703. Gerstmann, E.A., & Kramer, D.A. (1997). Feminist Identity Development: Psychometric Analyses of Two Feminist Identity Scales. Sex Roles, 36, 327-348. Geweke, J. (1992) “Evaluating the Accuracy of Sampling-Based Approaches to the Calculation of Posterior Moments,” in J.O. Berger, J.M. Bernardo, A.P. Dawid, and A.F.M. Smith (eds.), Bayesian Statistics 4, 169-194. Oxford: Oxford University Press. Gidengil, E., Giles, J., & Thomas, M. (2008). The gender gap in self-perceived understanding of politics in Canada and the United States. Politics & Gender, 4(04), 535- 561. Gurin, P., Miller, A., & Gurin, G. Stratum identification and consciousness. Social Psychology Quarterly, 43(1), 30-47. Gurin, P. (1985). Women's gender consciousness. Public Opinion Quarterly, 49(2), 143- 163. Hainmueller, J., Mummolo, J. & Xu, Y. How much should we trust estimates from multiplicative interaction models? Simple tools to improve empirical practice. Political Analysis, 27 (2), 163-192. 87 Harmon-Jones C., Bastian, B., & Harmon-Jones, E. The discrete emotions questionnaire: A new tool for measuring state self-reported emotions. PLoS ONE, 11 (8). Hayes, D. (2011). When gender and party collide: stereotyping in candidate trait attribution. Politics & Gender, 7(2), 133-165. Heflick, N. A., & Goldenberg, J. L. (2009). Objectifying Sarah Palin: Evidence that objectification causes women to be perceived as less competent and less fully human. Journal of Experimental Social Psychology, 45, 598–601. Heflick, N.A. & Goldenberg, J.L. (2011). Sarah Palin, a nation object(ifie)s: The role of appearance focus in the 2008 U.S. presidential election. Sex Roles, 65 (3) 149-155. Heldman, C., Caroll, S.J., & Olson, S. (2005). “She brought only a skirt”: print media coverage of Elizabeth Dole’s bid for the republican presidential nomination. Political Communication, 22(3), 315-335. Heldman, C., & Cahill, M. (2007, March). The Beast of Beauty Culture: An Analysis of the Political Effects of Self-Objectification. In Western Political Science Association conference, Las Vegas. Heldman, C. & Wade, L. (2011). Sexualizing Sarah Palin. Sex Roles, 65 (3), 156-164. Henderson-King, D. & Stewart, A. (1994). Women or feminists? Assessing women’s group consciousness. Sex Roles, 31(9), 505-516. hooks, bell. (1991). Ain’t I a Woman? Black women and feminism. Boston: South End Press. Huckfeldt, R. & Sprague, J. (1995). Citizens, politics, and social communication. Cambridge: Cambridge University. Huddy, L. & Terkildsen, N. (1993). The consequences of gender stereotypes for women candidates at different levels and types of office. Political Research Quarterly, 46(3), 503-525. Huebner, D. M., & Fredrickson, B. L. (1999). Gender differences in memory perspectives: Evidence for self-objectification in women. Sex Roles, 41(5-6), 459-467. Jamal, A. (2005). The political participation and engagement of Muslim Americans: mosque involvement and group consciousness. American Politics Research, 33(4), 521- 544. Jost, J.T., & Kay, A.C. (2005). Exposure to benevolent sexism and complementary gender stereotypes: Consequences for specific and diffuse forms of system justification. Journal of Personality and Social Psychology, 88, 498-509. 88 Kahn, K. & Goldenberg, E. (1991). An examination of gender differences in U.S. senate campaigns. Public Opinion Quarterly, 55(2), 180-899. Kahn, K. (1994). The distorted mirror: press coverage of women candidates for statewide office. Journal of Politics, 56(1), 154-173. Kline, T. J. (2005). Psychological Testing: A Practical Approach to Design and Evaluation. Thousand Oaks: Sage Publications. Klonoff, E., & Landrine, H. (1995). The schedule of sexist events: A measure of lifetime and recent sexist discrimination in women's lives. Psychology of Women Quarterly, 19(4), 439-472. Klonoff, E., Landrine, H., & Campbell, R. (2000). Sexist Discrimination May Account for Well-Known Gender Differences in Psychiatric Symptoms. Psychology of Women Quarterly, 24(1), 93-99. Kozee, H., Tylka, T., Augustus-Horvath, C., & Denchik, A. (2007). Development and psychometric evaluation of the interpersonal sexual objectification scale. Psychology of Women Quarterly, 31(2), 176-189. Krook, M. L. (2010). Beyond supply and demand: a feminist-institutionalist theory of candidate selection. Political Research Quarterly, 63(4), 707-720. Lanis, K. & Covell, K. (1995). Images of women in advertisements: Effects on attitudes related to sexual aggression. Sex Roles, 32(9-10), 639-649. Lawless, J. (2004). Women, war, and winning elections: gender stereotyping in the post September 11th era. Political Research Quarterly, 53(3), 479-90. Lawless, J. L., & Fox, R. L. (2010). It Still Takes A Candidate: Why Women Don't Run for Office. Cambridge: Cambridge University Press. Lin, M. (2018). From Alienated to Activists: Expressions and Formation of Group Consciousness Among Asian American Young Adults. Journal of Ethnic and Migration Studies, 1-20. MacKinnon, C. A. (2018, February). #MeToo Has Done What the Law Could Not. The New York Times. Retrieved from www.nytimes.com. Magley, V.J., Hulin, C.L., Fitzgerald, L.F., & Denardo, M. (1999). Outcomes of Self- Labeling Sexual Harassment. Journal of Applied Psychology, 84(3), 390-402. Magley, V., & Diener, Ed. (2002). Coping With Sexual Harassment: Reconceptualizing Women's Resistance. Journal of Personality and Social Psychology, 83(4), 930-946. http://www.nytimes.com 89 Matteson, A., & Moradi, B. (2005). Examining the Structure of the Schedule of Sexist Events: Replication and Extension. Psychology of Women Quarterly, 29(1), 47-57. Mattila, M. & Papageorgiou, A. Disability, perceived discrimination, and political participation. International Political Science Review, 38(5), 505-519. McDaniel, A. E. (2008). Measuring gender egalitarianism: The attitudinal difference between men and women. International Journal of Sociology, 38 (1), 58-80. McKinley, N. M., & Hyde, J. S. (1996). The objectified body consciousness scale development and validation. Psychology of women quarterly, 20(2), 181-215. Miles-McLean, H., Liss, M., Erchull, M., Robertson, C., Hagerman, C., Gnoleba, M., & Papp, L.J. “Stop Looking at Me!”: Interpersonal Sexual Objectification as a Source of Insidious Trauma. Psychology of Women Quarterly, 39(3), 363-374. Miller, A., Gurin, P., Gurin, G., & Malanchuk, O. (1981). Group consciousness and political participation. American Journal of Political Science, 25(3). Moradi, B. & Subich, L. M. (2002). Feminist identity development: comparing the psychometrics of three instruments. The Counseling Psychologist, 30 (1), 66-86. Moradi, B., & Varnes, J. R. (2017). Structure of the Objectified Body Consciousness Scale: Reevaluated 20 Years later. Sex Roles, 1-13. Morrell, M. (2003). Survey and experimental evidence for a reliable and valid measure of internal political efficacy.” Public Opinion Quarterly 67 (4): 589-602. Niemi, R.G., Craig, S.C., & Mattei, F. (1991). Measuring internal political efficacy in the 1988 National Election Study.” American Political Science Review 85 (4), 1407-1413. Noll, S. M., & Fredrickson, B. L. (1998). A mediational model linking self- objectification, body shame, and disordered eating. Psychology of Women Quarterly, 22 (4), 623–636. Norris, P. & Lovenduski, J. (1995). Political recruitment: gender, race, and class in the British parliament. Cambridge: Cambridge University Press. Nussbaum, Martha (1995). Objectification. Philosophy & Public Affairs 24 (4), 249–291. Osborne, D. & Sibley, C.G. (2013). Through rose-colored glasses: system-justifying beliefs dampen the effects of relative deprivation on well-being and political mobilization. Personality and Social Psychology Bulletin, 39(8), 991-1004. 90 Oskooii, K. (2016). How discrimination impacts sociopolitical behavior: a multidimensional perspective. Political Psychology, 37(5), 613-640. Page, D. (2018). When does sexuality-based discrimination motivate political participation? Political Psychology 39(5), 1013-1030. Pantoja, A. D., Ramirez, R., & Segura, G. (2001). Citizens by choice, voters by necessity: patterns in political mobilization by naturalized Latinos. Political Research Quarterly 54(4), 729-750. Pateman, C. (1985). The problem of political obligation: a critique of liberal theory. Cambridge: Polity. Pinkleton, E. W. & Austin, B. (2001). Individual motivations, perceived media importance, and political disaffection. Political Communication, 18(3), 321-334. Preece, J. R. (2016). Mind the gender gap: an experiment on the influence of self- efficacy on political interest. Politics & Gender 12(1). Pruysers, S. & Blais, J. Anything women can do men can do better: an experiment examining the effects of stereotype threat on political knowledge and efficacy. The Social Science Journal 51, 341-349. Quinn, D. M., Kallen, R. W., Twenge, J. M., & Fredrickson, B. L. (2006). The disruptive effect of self-objectification on performance. Psychology of Women Quarterly, 30 (1), 59-64. Roberts, T. A., & Gettman, J. Y. (2004). Mere exposure: Gender differences in the negative effects of priming a state of self-objectification. Sex Roles, 51(1-2), 17-27. Sanbonmatsu, K. (2002). Political parties and the recruitment of women to state legislatures. Journal of Politics 64, 791 – 809. Sanbonmatsu, K. & Dolan, K. (2009). Do gender stereotypes transcend party? Political Research Quarterly, 62, 485-494. Sanchez, G. (2006). The role of group consciousness in political participation among Latinos in the United States. American Politics Research, 34(4), 427–450. Sanchez, G. & Vargas, E. (2016). Taking a closer look at group identity: the link between theory and measurement of group consciousness and linked fate. Political Research Quarterly, 69(1), 160-174. Schneider, K., Swan, S., Fitzgerald, L., & Murphy, Kevin R. (1997). Job-Related and Psychological Effects of Sexual Harassment in the Workplace: Empirical Evidence From Two Organizations. Journal of Applied Psychology, 82 (3), 401-415. 91 Schneider, M. & Bos, A. (2014). Measuring stereotypes of female politicians. Political Psychology, 35(2), 245-65. Schooler, D. (2015). The woman next to me: Pairing powerful and objectifying representations of women. Analyses of Social Issues and Public Policy, 15(1), 198-212. Shani, Danielle. "Measuring Political Interest." Improving Public Opinion Surveys. Princeton: Princeton University Press, 2012. 137-57. Simpson, G. E., & Yinger, J. M. (1985). Racial and cultural minorities: An analysis of prejudice and discrimination (5th ed.). New York: Plenum. Swim, J.K., Cohen, L.L, & Hyers, L.L. (1998). Experiencing Everyday Prejudice and Discrimination. In J.K. Swim, & C. Stangor, C. (Eds). Prejudice: The target’s perspective (pp. 37-60). New York, NY: Academic Press. Swim, J., Hyers, L., Cohen, L., & Ferguson, M. (2001). Everyday sexism: evidence for its incidence, nature, and psychological impact from three daily diary studies. Journal of Social Issues, 57(1), 31-53. Szymanski, D., Gupta, M., Carr, A., & Stewart, E. (2009). Internalized Misogyny as a Moderator of the Link between Sexist Events and Women’s Psychological Distress. Sex Roles, 61(1), 101-109. Tiggemann, M., & Boundy, M. (2008). Effect of environment and appearance compliment on college women's self-objectification, mood, body shame, and cognitive performance. Psychology of Women Quarterly, 32(4), 399-405. Todorov, A., Mandisodza, A., Goren, A., & Hall, C. C. (2005). Inferences of competence from faces predict election outcomes. Science, 308 (5728), 1623-1626. Treier, S. & Jackman, S. (2008). Democracy as a latent variable. American Journal of Political Science, 52 (1), 201-217. Vagianos, A. (2017, December 6). 1 In 3 Women Has Been Sexually Harassed At Work, According to Survey. Huffington Post, Retrieved from www.huffingtonpost.com Valentino, N.A., Gregorowicz, K. & Groenendyk, E. (2009). Efficacy, emotions, and the habit of participation. Political Behavior, 31(3): 307-330. Valentino, N.A., Brader, T., Groenendyk, E., Gregorowicz, K., & Hutchings, V.L. (2011). Election night’s alright for fighting: The role of emotions in political participation. The Journal of Politics, 73(1): 156-170. http://www.huffingtonpost.com 92 Verba, S., Burns, N., & Schlozman, K. L. (1997). Knowing and caring about politics: Gender and political engagement. The Journal of Politics, 59(04), 1051-1072. Verkuyten, M. (1998). Perceived discrimination and self-esteem among ethnic minority adolescents. Journal of Social Psychology, 138(4), 479–493. Watson, L., & Grotewiel, B. (2016). The protective role of commitment to social change in the relationship between women’s sexist experiences and self-silencing. Sex Roles, 75(3), 139-150. Wen, N., Hao, X., and Cherian, G. (2013). Gender and political participation: news consumption, political efficacy and interpersonal communication. Asian Journal of Women's Studies, 19(4). Weston, R., & Gore, P. A. (2006). A Brief Guide to Structural Equation Modeling. The Counseling Psychologist, 34, 719–751. Young, I. M. (1980). Throwing like a girl: a phenomenology of feminine body comportment motility and spatiality. Human Studies, 3(2), 137-156. Young, I. M. (1990). Justice and the politics of difference. Princeton, New Jersey: Princeton University Press. Ysseldyk, R., Talebi, M., Matheson, K., Bloemraad, I., & Anisman, H. (2014). Religious and ethnic discrimination: Differential implications for social support engagement, civic involvement, and political consciousness. Journal of Social and Political Psychology, 2(1), 347-376. 93 APPENDIX Demographic Information Table A.1: Descriptive Statistics MTurk Sample Overall N= 244 Age (mean (sd)) 35.94 (11.28) Education (%) Less than HS 2 ( 0.8) HS grad 17 ( 6.9) Some college 65 (26.5) Currently in college 14 ( 5.7) College grad 107 (43.7) Post-college grad 40 (16.3) Democrat (%) 111 (65.3) Income (%) Under $15,000 27 (11.0) $15,000-$34,999 74 (30.2) $35,999-$64,999 80 (32.7) $65,000-$149,999 50 (20.4) Over $150,000 14 ( 5.7) Female (%) 110 (44.9) Ideology (mean (sd)) 3.04 (1.33) White 137 (59.3) Study Recruitment Between June 21st and June 28th of 2017, 250 male and female respondents living in the United States were recruited via the survey recruitment platform, Amazon Mechanical Turk. The Internet panel was recruited to participate in a study about “body image, demographic information, and gender attitudes.” Participants were compensated $1.00 for completing the 20-minute survey. 94 Table A.2: OLS Regression Results: Internal Political Efficacy Dependent variable: Internal Political Efficacy Education (Ref=less than high school grad) High school grad -0.497 (0.890) Some college -0.072 (0.862) College grad 0.083 (0.861) Post-college grad 0.412 (0.868) Ideology -0.030 (0.042) Income (Ref=less than $15,000 a year) $15,000 to $34,999 -0.288 (0.200) $35,000 to $64,999 0.020 (0.196) $65,000 to $149,999 -0.273 (0.212) Over $150,000 -0.481 (0.299) Gender (1=Women) -0.355*** -0.354*** (0.121) (0.122) Shame Scale 0.241*** 0.182** (0.091) (0.090) Surveillance Scale -0.030 -0.016 95 (0.090) (0.088) GenderXShame -0.294** -0.234* (0.143) (0.142) GenderXSurveillance -0.183 -0.165 (0.147) (0.145) Constant 0.213*** 0.407 (0.081) (0.896) Observations 232 232 R2 0.097 0.190 Adjusted R2 0.077 0.138 Residual Std. Error 0.876 (df = 226) 0.846 (df = 217) F Statistic 4.843*** (df = 5; 226) 3.637*** (df = 14; 217) Note: *p<0.1; **p<0.05; ***p<0.01 Table A.3: OLS Regression Results: Political Interest Dependent variable: Political Interest Education (Ref=high school grad Some college 0.287 (0.269) College grad 0.277 (0.269) Post-college grad 0.585 ** (0.294) Party ID (1=Democrat) -0.157 (0.197) Ideology -0.082 (0.064) Income (Ref=less than $15,000 a year) $15,000 to $34,999 0.212 (0.221) 96 $35,000 to $64,999 0.227 (0.219) $65,000 to $149,999 0.210 (0.238) Over $150,000 -0.514 (0.360) Gender (1=Women) -0.057 -0.266** (0.111) (0.134) Shame Scale 0.013 0.037 (0.081) (0.106) Surveillance Scale 0.061 0.075 (0.081) (0.097) GenderXShame -0.173 -0.304* (0.130) (0.159) GenderXSurveillance -0.081 0.00004 (0.134) (0.156) Constant 0.028 0.025 (0.074) (0.456) Observations 233 165 R2 0.020 0.120 Adjusted R2 -0.001 0.038 Residual Std. Error 0.801 (df = 227) 0.785 (df = 150) F Statistic 0.934 (df = 5; 227) 1.465 (df = 14; 150) Note: *p<0.1; **p<0.05; ***p<0.01 Table A.4: OLS Regression Results: Political Information-Seeking Dependent variable: Political Information-Seeking Education (Ref=less than high school grad) High school grad -0.772 (1.797) Some college -0.626 (1.745) 97 College grad -0.049 (1.741) Post-college grad -0.298 (1.756) Ideology -0.016 (0.085) Income (Ref=less than $15,000 a year) $15,000 to $34,999 -0.176 (0.405) $35,000 to $64,999 -0.320 (0.398) $65,000 to $149,999 -0.026 (0.426) Over $150,000 -0.425 (0.605) Gender (1=Women) 0.146 0.229 (0.236) (0.247) Shame Scale 0.257 0.182 (0.173) (0.177) Surveillance Scale -0.049 -0.067 (0.174) (0.177) GenderXShame -0.563** -0.492* (0.277) (0.284) GenderXSurveillance -0.538* -0.486* (0.286) (0.293) Constant 4.860*** 5.366*** (0.157) (1.813) Observations 231 231 R2 0.073 0.101 Adjusted R2 0.052 0.042 Residual Std. Error 1.703 (df = 225) 1.712 (df = 216) F Statistic 3.521*** (df = 5; 225) 1.725* (df = 14; 216) Note: *p<0.1; **p<0.05; ***p<0.01 98 Table A.5: Ordered Logit- Robustness Check Dependent variable: Political Information-Seeking Education (Ref=less than high school grad) High school grad 0.321 (1.586) Some college -0.232 (1.511) College grad 0.261 (1.507) Post-college grad 0.163 (1.527) Ideology -0.038 (0.090) Income (Ref=less than $15,000 a year) $15,000 to $34,999 -0.123 (0.420) $35,000 to $64,999 -0.408 (0.403) $65,000 to $149,999 0.250 (0.450) Over $150,000 -0.364 (0.611) Gender (1=Women) 0.272 0.339 (0.251) (0.262) Shame Scale 0.190 0.123 (0.187) (0.187) Surveillance Scale -0.087 -0.143 (0.193) (0.194) GenderXShame -0.576** -0.523* (0.292) (0.302) GenderXSurveillance -0.592* -0.486 99 (0.309) (0.314) Observations 231 231 Note: *p<0.1; **p<0.05; ***p<0.01 Table A.6: OLS Model: Gender Egalitarianism Dependent variable: Gender Egalitarianism Education (Ref=high school grad) Some college 0.120 (0.250) College grad -0.039 (0.249) Post-college grad 0.167 (0.271) Ideology -0.154 *** (0.041) Income (Ref=less than $15,000 a year) $15,000 to $34,999 0.367** (0.185) $35,000 to $64,999 0.640 *** (0.186) $65,000 to $149,999 0.578 *** (0.197) Over $150,000 0.486 * (0.266) Gender (1=Women) 0.509*** 0.387*** (0.119) (0.120) Shame Scale -0.600*** -0.562*** (0.086) (0.085) Surveillance Scale 0.344*** 0.295*** (0.087) (0.084) 100 GenderXShame 0.538*** 0.543*** (0.137) (0.134) GenderXSurveillance -0.108 -0.085 (0.141) (0.137) Constant -0.311*** -0.294 (0.080) (0.310) Observations 189 189 R2 0.307 0.400 Adjusted R2 0.288 0.356 Residual Std. Error 0.782 (df = 183) 0.744 (df = 175) F Statistic 16.177*** (df = 5; 183) 8.984*** (df = 13; 175) Note: *p<0.1; **p<0.05; ***p<0.01 101 Figure A.1 Body Surveillance Posterior Distribution Figure A.2 Body Shame Posterior Distribution 0.0 0.1 0.2 0.3 0.4 -3 -2 -1 0 1 2 3 values de ns ity ind postmeans 0.0 0.1 0.2 0.3 0.4 -2 -1 0 1 2 3 values de ns ity ind postmeans 102 Study Recruitment Between March 8th and March 15th of 2018, 311 female adults living in the United States were recruited via the survey recruitment platform, Amazon Mechanical Turk. The Internet panel was recruited to participate in a “Gender and Political Attitudes Study.” Participants were compensated $1.20 in Amazon.com credit for completing the 20- minute survey. MTurk provides a more representative sample than in-person convenience samples (Berinsky, Huber, and Lenz 2012). HIT Advertisement: We invite you to take part in a research study about gender identity and political engagement. This survey will take approximately 30 minutes to complete. You will be compensated $1.20 for participation in the study. 103 Table A.7: Descriptive Statistics for MTurk Sample Overall N= 311 age (mean (sd)) 39.67 (11.31) education (%) Less than high school grad 1 ( 0.3) High school grad 35 (11.3) Some college 96 (30.9) College grad 146 (46.9) Post-college grad 33 (10.6) Party identification Democrat 143 (46.0) Independent 82 (26.4) Republican 80 (25.7) Other 6 ( 1.9) Income Under $15,000 22 ( 7.1) $15,000 to $34,999 79 (25.4) $35,000 to $64,999 108 (34.7) $65,000 to $149,999 92 (29.6) Over $150,000 10 ( 3.2) White(%) 224 (74.4) Ideology (%) Very Liberal 60 (19.3) Liberal 65 (20.9) Somewhat Liberal 60 (19.3) Somewhat Conservative 74 (23.8) Conservative 37 (11.9) Very Conservative 15 ( 4.8) 104 Table A.8: OLS Regression: Internal Political Efficacy Dependent variable: Internal Political Efficacy OLS (Model 1) (Model 2) Education (Ref=high school Grad) Some college 0.553*** 0.559*** (0.179) (0.178) College graduate 0.926*** 0.931*** (0.178) (0.175) Post-college graduate 0.899*** 0.967*** (0.232) (0.231) Ideology 0.016 0.008 (0.053) (0.053) Age 0.007 0.008 (0.005) (0.005) Party (1=Democrat) -0.132 -0.127 (0.137) (0.137) Race (1=White) 0.167 0.196 (0.125) (0.126) Income(Ref=under $15,000) $15,000 to $34,999 0.402* 0.357 (0.229) (0.229) $35,000 to $64,999 0.358 0.312 (0.224) (0.223) $65,000 to $149,999 0.194 0.134 (0.225) (0.225) Over $150,000 0.094 -0.005 (0.345) (0.345) SSEscale 0.072 (0.060) ISOSscale 0.099 * (0.058) FIDS 0.229*** 0.231*** (0.072) (0.071) 105 SSExFIDS 0.117** (0.056) ISOSxFIDS 0.100 * (0.058) Constant -1.415*** -1.369*** (0.344) (0.341) Observations 294 295 R2 0.202 0.208 Adjusted R2 0.162 0.168 Residual Std. Error 0.859 (df = 279) 0.859 (df = 280) F Statistic 5.037*** (df = 14; 279) 5.240*** (df = 14; 280) Note: *p<0.1; **p<0.05; ***p<0.01 Table A.9: OLS Regression: Political Interest Dependent variable: Political Interest OLS (Model 2) (Model 3) Education (Ref=high school Grad) Some college 0.387 0.352 (0.268) (0.266) College graduate 0.618** 0.563** (0.267) (0.263) Post-college graduate 0.747** 0.768** (0.347) (0.346) Ideology 0.003 -0.0004 (0.079) (0.079) Age 0.028*** 0.028*** (0.007) (0.007) Party (1=Democrat) -0.067 -0.072 (0.204) (0.205) Race (1=White) 0.293 0.304 (0.187) (0.189) Income(Ref=under $15,000) 106 $15,000 to $34,999 0.128 0.074 (0.342) (0.344) $35,000 to $64,999 0.152 0.102 (0.334) (0.334) $65,000 to $149,999 -0.001 -0.051 (0.337) (0.337) Over $150,000 0.039 -0.018 (0.516) (0.517) SSEscale 0.014 (0.089) ISOSscale 0.110 (0.086) FIDS 0.454*** 0.438*** (0.108) (0.107) SSExFIDS 0.168** (0.084) ISOSxFIDS 0.029 (0.087) Constant 3.428*** 3.561*** (0.514) (0.512) Observations 294 295 R2 0.189 0.186 Adjusted R2 0.148 0.145 Residual Std. Error 1.284 (df = 279) 1.287 (df = 280) F Statistic 4.644*** (df = 14; 279) 4.563*** (df = 14; 280) Note: *p<0.1; **p<0.05; ***p<0.01 Table A.10: OLS Regression: Political Participation Dependent variable: Political Participation OLS (Model 4) (Model 5) Education (Ref=high school Grad) Some college 0.044 0.053 107 (0.160) (0.156) College graduate -0.013 0.003 (0.159) (0.154) Post-college graduate 0.174 0.296 (0.207) (0.203) Ideology -0.036 -0.041 (0.047) (0.046) Age 0.002 0.001 (0.004) (0.004) Party (1=Democrat) -0.086 -0.071 (0.122) (0.120) Race (1=White) -0.137 -0.091 (0.112) (0.111) Income(Ref=under $15,000) $15,000 to $34,999 0.363* 0.254 (0.204) (0.202) $35,000 to $64,999 0.233 0.134 (0.200) (0.196) $65,000 to $149,999 0.191 0.083 (0.201) (0.198) Over $150,000 0.474 0.371 (0.308) (0.304) SSEscale 0.163*** (0.053) ISOSscale 0.242 *** (0.051) FIDS 0.287*** 0.282*** (0.064) (0.063) SSExFIDS 0.084* (0.051) ISOSxFIDS 0.013 (0.051) Constant -0.138 -0.036 (0.307) (0.301) Observations 294 295 108 R2 0.226 0.253 Adjusted R2 0.187 0.216 Residual Std. Error 0.768 (df = 279) 0.756 (df = 280) F Statistic 5.804*** (df = 14; 279) 6.771*** (df = 14; 280) Note: *p<0.1; **p<0.05; ***p<0.01 Tests for linearity Hainmueller, Mummolo, and Xu (2016) Figure A.3: Marginal Effect of Sexist Events on Efficacy over the Range of FIDS (Model 1) 109 Figure A.4: Marginal Effect of Objectifying Events on Efficacy over the Range of FIDS (Model 2) Figure A.5: Marginal Effect of Sexist Events on Interest over the Range of FIDS (Model 3) 110 Figure A.6: Marginal Effect of Sexist Events on Interest over the Range of FIDS (Model 4) Table A.11: Ordered Logit Results (Models 5 and 6) Dependent variable: Political Participation Education (Ref=high school grad) Some college -0.360 -0.729 (0.485) (0.487) College graduate -0.249 -0.509 (0.378) (0.378) Post-college graduate -0.338 -0.613* (0.359) (0.360) Ideology -0.042 -0.081 (0.095) (0.097) Age 0.005 0.004 (0.010) (0.010) Race(1=white) -0.391 -0.261 (0.263) (0.265) Income (Ref=under $15,000) 111 $15,000 to $34,999 0.754 0.477 (0.515) (0.517) $35,000 to $64,999 0.436 0.173 (0.496) (0.495) $65,000 to $149,999 0.334 0.058 (0.501) (0.499) Over $150,000 1.046 0.616 (0.828) (0.820) SSEscale 0.499*** (0.134) ISOSscale 0.685 *** (0.128) FIDS 0.710*** 0.734*** (0.158) (0.157) SSExFIDS 0.326** (0.137) ISOSxFIDS 0.196 (0.135) Observations 299 300 Note: *p<0.1; **p<0.05; ***p<0.01 112 Table A.12: Descriptive Statistics for Laboratory Sample Overall N= 71 Education (%) High school diploma 9 (12.9) Currently in college 56 (80.0) College graduate 1 ( 1.4) Post-college graduate 4 ( 5.7) Ideology (%) Very Liberal 16 (22.9) Liberal 28 (40.0) Somewhat liberal 21 (30.0) Somewhat Conservative 5 ( 7.1) Income (%) Under $15,000 5 ( 7.2) $15,000 to $34,999 7 (10.1) $35,000 to $64,999 19 (27.5) $65,000 to $149,000 30 (43.5) Over $150,000 8 (11.6) White (%) 31 (44.3) Republican (%) 2 ( 2.9) Table A.13: OLS Regression (Model 1) Dependent variable: Internal Political Efficacy Gender Consciousness 1.321 -0.681 (3.215) (3.472) Treatment -4.096 -3.130 (2.994) (3.189) Education 0.149 (0.801) Race (1=white) 1.033 (0.988) Ideology -1.105 (0.695) GenderConsciousnessXTreatment 8.139* 6.537 113 (4.439) (4.762) Constant 11.071*** 14.131*** (2.116) (3.651) Observations 66 66 R2 0.161 0.200 Adjusted R2 0.120 0.119 Residual Std. Error 3.738 (df = 62) 3.740 (df = 59) F Statistic 3.953** (df = 3; 62) 2.465 ** (df = 6; 59) Note: *p<0.1; **p<0.05; ***p<0.01 Table A.14: OLS Regression (Model 2) Dependent variable: Political Participation Gender Consciousness 1.397 0.614 (3.595) (3.919) Treatment -7.772** -7.152** (3.254) (3.471) Education -0.139 (0.886) Race (1=white) -0.914 (1.083) Ideology -0.577 (0.756) GenderConsciousnessXTreatment 12.494** 11.776** (4.876) (5.208) Constant 14.498*** 16.912*** (2.321) (3.971) Observations 65 65 R2 0.231 0.255 Adjusted R2 0.193 0.178 Residual Std. Error 4.023 (df = 61) 4.060 (df = 58) F Statistic 6.094*** (df = 3; 61) 3.311*** (df = 6; 58) Note: *p<0.1; **p<0.05; ***p<0.01 114 Table A.15: OLS Regression (Model 3) Dependent variable: Political Interest Gender Consciousness 3.240** 2.654 (1.498) (1.637) Treatment -2.008 -1.886 (1.395) (1.504) Education -0.171 (0.377) Race (1=White) -0.144 (0.466) Ideology -0.321 (0.328) GenderConsciousnessXTreatment 2.905 2.793 (2.068) (2.245) Constant 4.031*** 5.508*** (0.986) (1.721) Observations 66 66 R2 0.273 0.291 Adjusted R2 0.238 0.219 Residual Std. Error 1.742 (df = 62) 1.763 (df = 59) F Statistic 7.758*** (df = 3; 62) 4.034*** (df = 6; 59) Note: *p<0.1; **p<0.05; ***p<0.01 Table A.16: OLS Regression (Model 4) Dependent variable: Internal Political Efficacy Ideology -2.053** -2.180** (0.821) (0.828) Treatment -0.365 -0.755 (2.487) (2.582) Education 0.641 115 (0.764) White 0.952 (0.920) IdeologyXTreatment 0.692 0.821 (1.032) (1.077) Constant 16.751*** 15.396*** (2.012) (2.446) Observations 70 70 R2 0.169 0.193 Adjusted R2 0.132 0.130 Residual Std. Error 3.639 (df = 66) 43 (df = 64) F Statistic 4.485*** (df = 3; 66) 3.057** (df = 5; 64) Note: *p<0.1; **p<0.05; ***p<0.01 Table A.17: OLS Regression (Model 5) Dependent variable: Political Participation Ideology -0.559 -0.558 (0.942) (0.961) Treatment 4.820* 4.561 (2.862) (2.999) Education 0.282 (0.904) White -0.182 (1.071) IdeologyXTreatment -2.034* -1.914 (1.183) (1.249) Constant 16.906*** 16.417*** (2.322) (2.913) Observations 69 69 R2 0.174 0.176 Adjusted R2 0.136 0.111 Residual Std. Error 4.163 (df = 65) 4.225 (df = 63) 116 F Statistic 4.575*** (df = 3; 65) 2.690** (df = 5; 63) Note: *p<0.1; **p<0.05; ***p<0.01 Table A.18: OLS Regression (Model 6) Dependent variable: Political Interest Ideology -0.812* -0.829* (0.414) (0.423) Treatment 0.586 0.607 (1.251) (1.314) Education 0.005 (0.391) Race (1=white) 0.177 (0.469) IdeologyXTreatment -0.231 -0.248 (0.521) (0.550) Constant 7.855*** 7.814*** (1.009) (1.246) Observations 71 71 R2 0.182 0.184 Adjusted R2 0.145 0.121 Residual Std. Error 1.840 (df = 67) 1.866 (df = 65) F Statistic 4.963*** (df = 3; 67) 2.924** (df = 5; 65) Note: *p<0.1; **p<0.05; ***p<0.01 Table A.19: OLS Regression (Model 7) Dependent variable: Internal Political Efficacy Anger 0.493 0.661 (0.419) (0.398) Treatment 2.594 1.108 (2.022) (1.956) Education 0.695 117 (0.732) White 1.081 (0.910) Ideology -1.896 *** (0.541) AngerXTreatment -0.486 -0.238 (0.558) (0.530) Constant 10.820*** 12.872*** (1.289) (2.389) Observations 71 70 R2 0.050 0.235 Adjusted R2 0.007 0.162 Residual Std. Error 3.863 (df = 67) 3.575 (df = 63) F Statistic 1.174 (df = 3; 67) 3.221*** (df = 6; 63) Note: *p<0.1; **p<0.05; ***p<0.01 Table A.20: OLS Regression (Model 8) Dependent variable: Political Participation Anger 0.395 0.648 (0.491) (0.476) Treatment 2.225 1.318 (2.377) (2.344) Education 0.793 (0.895) White -0.390 (1.091) Ideology -1.781 *** (0.646) AngerXTreatment -0.601 -0.490 (0.654) (0.633) Constant 14.694*** 16.611*** (1.527) (2.951) Observations 70 69 118 R2 0.015 0.171 Adjusted R2 -0.030 0.091 Residual Std. Error 4.520 (df = 66) 4.271 (df = 62) F Statistic 0.337 (df = 3; 66) 2.137* (df = 6; 62) Note: *p<0.1; **p<0.05; ***p<0.01 Table A.21: Regression (Model 9) Dependent variable: Political Interest Anger 0.149 0.234 (0.215) (0.204) Treatment 0.743 0.063 (1.039) (1.005) Education 0.124 (0.380) Race (1=white) 0.166 (0.471) Ideology -1.044 *** (0.281) AngerXTreatment -0.197 -0.088 (0.288) (0.273) Constant 5.630*** 7.484*** (0.651) (1.227) Observations 72 71 R2 0.010 0.205 Adjusted R2 -0.034 0.130 Residual Std. Error 2.009 (df = 68) 1.856 (df = 64) F Statistic 0.230 (df = 3; 68) 2.743** (df = 6; 64) Note: *p<0.1; **p<0.05; ***p<0.01 TITLE LINE ONE: THE POLITICAL CONSEQUENCES OF GENDER-BASED TITLE LINE TWO: MARGINALIZATION TITLE LINE THREE: TITLE LINE FOUR: DOCTOR OF XXX: DOCTOR OF PHILOSOPHY Author's Name: Claire Malone Gothreau Diploma Date: May 2019 Mentor's Name and Dept: Kevin Arceneaux, Advisory Chair, Political Science Committee Member's Name and Dept 1: Nyron Crawford, Political Science Committee Member's Name and Dept 2: Heath Fogg Davis, Political Science Committee Member's Name and Dept 3: Johanna Dunaway, External Member, Texas A&M University External Committee Member's Name and Affiliation: