Madame Justice Will Save Our Democracy: Gender Bias and Perceptions of the High Court in Transitional Regimes

Introduction

The public perception of women in positions of authority plays a critical role in how and when women are able to occupy those positions. This has prompted a burgeoning literature on the impression created by female legislators and executives, demonstrating a connection between the presence of women in these roles and improved popular evaluations of the overall political system. Much less attention has been paid to this dynamic in the context of the judiciary. Initial evidence from stable democracies shows a similar relationship between women’s presence and positive evaluations of the regime but does not address how this dynamic may change in non-democracies or those countries in a period of regime transition. In this article, we address this gap and argue that the context of regime transition cultivates a unique dynamic in which gender bias—and specifically, hostile bias—triggers unexpected support for women in the highest court. That is, we argue that the context of losing one’s democracy causes certain citizens to react to women justices in surprising ways due to their desire to maintain or return to more democratic traits. While hostile bias is negatively correlated with perceptions of legitimacy of women-dominated courts in stable democracies, we argue that the unique political context of a transitional regime causes those citizens with hostile bias to see women as potentially valuable. We test our theories using an original data set of over 600 survey respondents in Poland and Thailand and find evidence that women’s judicial presence in the context of a transitioning regime sometimes triggers citizens with hostile bias to see women as potential saviors of the democracy, though not always. These findings indicate that the presence of women on courts is sufficient to generate improved evaluations of the democratic health of the regime for some, offering additional support to the literature on the power of gender stereotypes and suggesting a potentially useful tool for transitional regimes seeking legitimacy. ¹

Perceptions of Women in Government

It is well-established that the presence of women in the executive and/or legislative branches improves trust in government and fairness across a variety of indicators, including satisfaction with democracy (Clayton, O’Brien, and Piscopo 2019; Schwindt-Bayer 2010; Schwindt-Bayer and Alles 2018) as well as expectations regarding corruption (Barnes and Beaulieu 2014; Barnes, Beaulieu, and Saxton 2018; Esarey and Schwindt-Bayer 2018; Shair-Rosenfield and Wood 2017; Valdini 2019). This work theorizes that stereotypical “communal” traits associated with women play an important causal role in this dynamic. Communal traits are characteristics associated with traditional notions of femininity, including compassion, gentleness, cooperation, honesty, and loyalty (Alexander and Andersen 1993; Eagly and Karau 2002; Huddy and Terkildsen 1993; Sanbonmatsu 2002; Spence and Buckner 2000). Because women are stereotyped with these traits, voters assume that that their presence increases the communal behavior of the government, thereby boosting the democratic appearance even of regimes that are not (Baldez 2004; Bush 2011; Valdini 2019). This association of women with democracy is powerful enough to influence the guidance of international organizations for how to increase the level of democracy in a state. The National Democratic Institute, for example, states on their website that it is essential to include women in developing democracies because “women are particularly effective in promoting honest government” (quote from Valdini 2019, 45). Existing research, in short, offers substantial evidence that women’s presence in the executive and legislative branches causes citizens to assume that their communal behavior will increase the communal—that is, democratic—nature of the entire state.

While the impact of women’s presence in the specific institution of the high court has been under-studied compared to the executive and legislative branches, there is scholarly work in this area on the effect of women’s presence on the legitimacy of the institution and the state. Kenney (2013) argues that gender plays a critical role in the courts for both legitimacy and decision-making reasons, particularly because of the benefits of diverse perspectives on the courts. Kirkpatrick, Kittilson, and Hoekstra (2020) offer a symposium of articles on diversity in the judiciary, including arguments that descriptive representation influences multiple conceptions of justice (Dovi and Luna 2020) and that intersectionality can offer a richer understanding of diversity on the bench (Kang et al. 2020). And finally, recent work from Shortell and Valdini (2022) explicitly tests the impact of women on the bench on perceptions of legitimacy in democracies, and finds that the presence of women on the high court has a strong positive impact on citizen perceptions of court legitimacy, though not among those with hostile gender bias.

However, existing work on women’s presence on the high court has overlooked two important factors that may change how people respond to women on the bench. First, to the best of our knowledge, there is no existing research that considers the impact of women on the high court on the perception of the likelihood that the court will preserve or increase the level of democracy of the state. That is, research in the general area of the impact of women on the bench has made valuable contributions in understanding the effect of women justices’ presence on the broader concepts of legitimacy, justice, and diversity, but scholars have yet to consider the unique issue of the perception of protecting the democracy. This is an oversight because courts can play a central role in supporting the establishment of democracy (Issacharoff 2015) or in providing cover to regimes becoming more authoritarian (Landau and Dixon 2020; Scheppele 2018; Urribarri 2011). Even in environments where citizens express low levels of confidence in courts, they still turn to the judiciary as the only viable alternative to consider their rights (Taylor 2018). Thus, this institution plays a unique role in preserving/dismantling the democracy, and therefore the stereotype that women are inherently democratic may be particularly resonant with citizens who are considering the potential of the court to protect the democratic elements of the state.

The second factor that has been overlooked by existing research on the impact of women justices on perceptions of the state is the condition of regime change or fluidity. That is, while existing research has engaged the unique contexts of courts in democracies or in authoritarian regimes, to the best of our knowledge, there is no scholarship on the impact of living in a regime that is in the midst of change. The context of change/uncertainty is particularly relevant for an analysis on the perception of women due to the well-documented phenomena of “think crisis-think female.” The context of threat or crisis increases peoples’ preference for women leaders due to their stereotypical association with change (Brown, Diekman, and Schneider 2011) or their stereotypical association with traits that are useful for managing people in a crisis situation (Ryan, Haslam, and Kulich 2010). Further, Valdini (2019) demonstrates that under conditions of rapid change or chaos, the stereotyped traits associated with women—such as honesty and stability—increase in value, and thus the likelihood that party elites select women as candidates increases because of the change in value of these presumed traits. Thus, it is reasonable that conditions of change or instability in the regime may alter both how people perceive women justices as well as the importance or usefulness of their presence.

Impact of Sexism in a Transitional Regime

Research into the impact of women’s presence in positions of authority has long noted the role that bias can play in evaluations of women and the institutions in which they serve. For example, recent work on the impact of bias on voter behavior offers substantial evidence that sexism has a direct effect on voters’ support for women candidates (Barnes, Beaulieu, and Saxton 2020; Cassese and Barnes 2019; Cassese and Holman 2019; Ditonto 2019; Fulton and Dhima 2021).

However, while we know that general bias against women can have a powerful effect on how women in power are perceived, there is also evidence that each type of bias triggers a unique response to women. At this point, two types of bias against women are widely recognized in the literature and, while both forms serve to justify women’s subordinate status, they do so in different ways (Glick and Fiske 1996). The first type, hostile bias, consists of blatantly negative feelings toward women, commonly resting in the perception that women are trying to control or overpower men. The second type, benevolent bias, often presents as concern or care, which masks the ways it frames women as childlike, both in need of protection and less capable than their male counterparts (Glick et al. 2000; Glick and Fiske 1996). Further, existing literature in political psychology demonstrates that the type of bias affects the reaction to women in power. For example, only benevolent bias—not hostile—is correlated with support for upholding traditional gender roles in government positions that are gendered masculine, such as the presidency (Cassidy and Krendl 2019; Glick and Fiske 2001; Jost and Kay 2005; King et al. 2012). In short, while both types of bias serve to maintain women’s oppression, they do so in dramatically different ways and thus suggest that the behavior of bias-holders may change depending on the type of bias. ²

Building from the foundation that not all bias has the same effect, as well as that communal traits increase in value in times of instability, we assert that people with hostile bias behave in unexpected ways in transitional regimes. Specifically, we argue that people with hostile bias—and only those with hostile bias—see women on the high court as potential saviors of the democracy, while people with benevolent bias or no bias will not. It is straightforward to explain why we believe those without bias against women will not see women as saviors of democracy: those citizens who do not demonstrate any gender bias are less likely to activate the traditional stereotypes associated with women (Lepore and Brown 1997; Wittenbrink, Gist, and Hilton 1997), and thus are unlikely to assume that women on the bench have an innate desire to increase democratic rule. Less straightforward is our expectation that only those with hostile bias, not benevolent bias, will see women as saviors of democracy.

Our reasoning to support the argument that those with hostile—but not benevolent bias—are more likely to see women as potential saviors of democracy rests on the crucial differences in these types of bias. Benevolent sexism, at its core, is about a lack of competence; those with these views believe that women are less competent than men, and thus that they need extra care and protection. As Cassidy and Krendl deftly explain, “despite seemingly more positive characterizations, benevolent sexism assumes that women comprise the weaker, and thereby less competent, sex” (2019, 505). Hostile sexism, on the other hand, has a core of fear, and specifically a fear of women’s power and control. Existing literature demonstrates that men who endorse hostile sexism tend to be afraid of being dependent on women because they fear manipulation, control, and exploitation (Cross, Overall, and Fletcher 2017, Cross et al. 2019). In this type of bias, therefore, there is no question of competence; women are viewed as so competent, in fact, that they are inherently threatening to the entire power structure. This is echoed in the work of Fiske (2012) in which she notes that:

“[h]ostile sexism targets nontraditional women who threaten men’s dominance in various ways: female professionals, intellectuals, and trades-women compete for men’s traditional roles; lesbians and vamps reject heterosexual intimacy; and feminists challenge male power. These women are stereotyped as threateningly capable but not nice” (37, emphasis added).

A hostile sexist, therefore, views women in authority as potentially powerful, while a benevolent sexist sees women as weak and incapable. These are very different views of women and their competence, and thus it is reasonable that a person with hostile bias would respond differently to women on the bench than a person with benevolent bias.

In addition to the unique effect of different types of bias, our theory also engages the unique effect of the context of living in a regime that is actively increasing/decreasing its democratic elements. In this unique context of a transitional regime, citizens are aware of the fluid nature and instability of their regime, and thus are looking for political figures who can stabilize the country and potentially increase the level of democracy of the regime. ³ This is very different than a stable democratic or authoritarian regime; it is the instability that triggers the value of these communal traits, as well as the perception that the regime can change. The association between women’s perceived communal traits and specifically, the assumption that women are inherently more democratic because of those traits, should increase in value when citizens are concerned about the survival of democratic ideals in their transitional regime. However, this change in the value of women’s stereotyped traits will not occur for all citizens. Those who are not sexist—and thus, we assume, are least likely to rely on traditional gender stereotypes—will not view women as saviors of democracy in this (or any) context. Further, we argue that because benevolent sexists do not view women as competent, they are unlikely to see women as capable of defending the democracy and thus will also not view women justices as potential democracy saviors. It is, we argue, the hostile sexists who, while they may despise women in power, simultaneously view them as powerful enough to increase the level of democracy of the regime, and thus they see women on the court as the potential saviors of the democracy. We lay out these theoretical expectations in Figure 1.OPEN IN VIEWER

Methodology

Due to our interest in the impact of women on courts outside the stable democratic context, we focused our testing on two transitional regimes: Poland and Thailand. We selected these two countries for three reasons: first, at the time of our survey (February 2020), they were both undergoing elements of regime transition to/from democracy and thus both are easily categorized as having an unstable or fluid regime type. Second, we chose these particular states because of the significant role the constitutional high courts have played in supporting authoritarian rule in recent years. The specifics of this for each of our cases is discussed in greater depth in the analysis section for each country. Though our survey specifically indicates that we are asking about a court in a different, unspecified country, we expect that the responses to the questions will draw on existing attitudes shaped by respondents’ domestic political situation. Thus, if countries with courts that are prominently associated with rulings favorable to authoritarian policies can still prompt positive reactions to the presence of women for democracy, then we can have a high degree of confidence about this relationship in a wide variety of transitional regimes. These are, in short, both hard cases in which to test our theories. Finally, we chose Poland and Thailand due to the fact that they are very different countries. For example, while the regimes in both Poland and Thailand are changing, their 2020 Freedom House scores demonstrate that they had very different levels of authoritarianism at the time of our survey. Poland’s democracy has been slowly but steadily declining in recent years but still maintains several democratic attributes (reflected by their score of 84), while Thailand has been immersed in military coups, increasing control by the monarchy, and very few protections for civil rights (demonstrated by their Freedom House score of 32). These cases also provide clear cultural and regional variation to help with the generalizability of our findings. There is also variation regarding attitudes towards women in these two countries. In the most recent World Values Survey (2017–2020), the public in Thailand had significantly higher scores for traditional views about women’s roles in society than Poland, even though Poland is a traditionally Catholic country. ⁴

To assess how people in transitional regimes evaluate the democratic health of a country when there are women justices on the high court, we utilized Qualtrics to conduct surveys in Poland and Thailand in February, 2020. ⁵ All questions were in Polish or Thai, respectively, and were translated by native speakers in close consultation with the authors. Qualtrics facilitated an equal distribution of respondent age, education, and gender. A total of 327 surveys were completed in Thailand and 321 were completed in Poland. After accepting participation in the survey, our respondents were each randomly assigned to one of three groups. Each of these groups was then given a one paragraph description of a current high court in an anonymous country in the process of transitioning away from democracy. These descriptions were exactly the same with the exception of one key difference: we varied the gender balance of judges on the anonymous high court. We offered three different justice gender scenarios: in the first, there were 6 men and 1 woman on the court. In the second, there were 4 men and 3 women on the court. And in the third, there were 6 women and 1 man on the court. Thus, our independent variable is categorical and has three possible values. In the description of the court, we did not strongly emphasize the gender of the judges, but rather mixed that material in with supplementary information about the judicial selection mechanism and term length. ⁶ This variable is labeled “treatment” in our model, and the majority male court is our baseline comparison category.

After reading the assigned paragraph, the respondent was then given a question designed to elicit their opinion on the level of democracy in the regime, as well as questions to track their opinion about the general level of legitimacy of the court and the likelihood of justice being served by the high court. These were, in short, designed to pick up different facets of the idealized high court; while not the focus of our analysis, we included opportunities to state their opinion on general legitimacy and justice in order to see if their responses to the democracy question mirrored those on other potential ideals. Due to the nature of our analysis, our primary focus was on responses to a single question: “Based on this description, how likely do you think it is that the court will contribute to the overall health and well-being of democracy in this country?” Respondents were given a six point Likert scale to express their opinion, and we condensed their responses into three categories for our dependent variable. ⁷ In addition, we collected control variables including age, gender, ideology, and education, as well as two sets of questions designed to capture the respondent’s other personal beliefs about high courts. ⁸

In addition, the final section of the survey utilized 4 questions from the benevolent/hostile sexism battery created by Glick and Fiske (1996). In our analysis, we offer four models—two with measures of benevolent bias and two measures of hostile bias—thus each model tests both an effect of bias as well as the effect of unique measurement of that bias. For example, in order to test the effect of hostile bias, one model utilizes data from a question that asked how much the respondent agreed or disagreed with the statement that “women tend to exaggerate problems at work.” And in another model, designed to test benevolent sexism, we utilize data from a question that asked how much the respondent agreed or disagreed with the statement that “women should be cherished and protected by men.” We chose to design this study using different measurements of bias (and specifically chose to not combine them) because a foundational premise of our work is that the effect of bias is nuanced, specific, and more difficult to track than much of the existing literature acknowledges.

We estimate the models using ordered probits due to our categorical dependent variable. In addition, based on the work of Williams (2021), we used LR chi-square tests in order to determine whether to code our ordinal independent variables as categorical or continuous, and discovered that the more parsimonious continuous coding was justified for our independent variables that track hostile and benevolent bias. Finally, we include an interaction term between bias and the presence of women justices because we expect the effect of these variables to be conditional, not constant; that is, we expect that respondents will be more likely to positively assess the likelihood that the court will contribute to democratic health when there are women justices on the court, but only for those respondents that express hostile bias against women. While the coefficients cannot be directly interpreted because of the presence of interaction terms (Kam and Franzese, Jr. 2007), the significance and direction of the coefficients offer insight into the relationships among these variables. Following a brief discussion of these baseline results, we will offer an analysis of marginal effects that allows us to provide a more specific discussion of predicted probabilities. ⁹

Analysis: Thailand

Thailand has seen a series of military coups, an increasingly active monarchy, and widespread street protests in the recent past (Kanchoochat and Hewison 2016). Five new constitutions, three permanent and two temporary, have been implemented since 2006 and two recent elections were annulled by courts (McCargo 2019, 120). Thailand is not purely authoritarian, though. The so-called “People’s Constitution” of 1997 introduced an array of democratic reforms. There have been elections interspersed between periods of military rule, most recently in 2019 which at least nominally transferred power from the military junta to civilian control. Much of the civil unrest from 2006 to present has been driven by conflicts between supporters of the monarchy, designated as yellow shirts, and supporters of deposed prime minister Thaksin Shinawatra, designated as red shirts, although this division splintered to some degree in the most recent election (McCargo 2019, 124–28). Throughout these conflicts, courts have played a visible role. The Constitutional Court is regarded as a critical tool of the royalist yellow shirts and was instrumental in actions such as dissolving certain political parties, evaluating the role of the monarchy in politics, and directly overturning elections (Dressel 2010; Dressel and Tonsakulrungruang 2019; Mérieau 2016). Thus, while Thailand has many authoritarian characteristics, they have also experienced an ongoing push for greater democratization that is sometimes successful.

Figure 2 offers the results of four models, each examining the impact of a different type and measurement of bias. In most of these models, bias and the gender balance of the court has no effect on our respondents’ views on whether or not the court will contribute to the health and well-being of the democracy; none of the interactions of benevolent bias and court gender balance are significant, and neither is one of the measures of hostile bias. However, Model 2 is the exception: in this model, the results suggest that when a respondent is faced with a majority-women high court and they have high levels of hostile bias (measured using the “women exaggerate” question), then they believe that the court is more likely to contribute to the well-being of the democracy than a majority-male or gender-balanced court. In an effort to more easily and precisely interpret these results (because we cannot directly interpret interaction variables), we generated predictive margins that allow us to see the predicted probabilities based on differing levels of bias and court composition.OPEN IN VIEWER

Figure 3 presents the results from marginal effects that we calculated based on Model 2; in other words, this figure demonstrates the effect of a change in the level of hostile bias on the probability that a respondent agrees or strongly agrees that the high court will contribute to the health and well-being of the democracy. As Model 2 demonstrated, the impact of hostile bias is conditional on the gender-balance of the court, but seeing the predicted probabilities depicted visually helps dramatically in understanding what these results mean. The y-axis represents the probability that a respondent will give a certain answer to the following question: “Based on this description, how likely do you think it is that the court will contribute to the overall health and well-being of democracy in this country?” The x-axis represents the level of hostile bias expressed by the respondent, where the low end of the scale indicates low bias and the high end indicates high bias against women. These results suggest support for our main hypothesis; for those respondents who received the treatment of a majority women high court, the probability that they would believe that the high court will contribute to the overall health and well-being of the democracy increases with the respondents’ hostile bias. In other words, this suggests support for our hypothesis that those with hostile bias—but not benevolent—are more likely to see women as potential saviors of democracy.OPEN IN VIEWER

Analysis: Poland

Poland has long served as the exemplar of democratic transition among post-communist states (Taras 1995; see Rose-Ackerman 2005; Schwartz 2000 for caution about this view), but since 2015 the actions of the Law and Justice (PiS) ruling party have carved away at that reputation. This was precipitated by a constitutional crisis in December 2015 when the newly elected parliamentary majority replaced five members of the constitutional court (Sadurski 2019, 61–70). In addition to capturing the Constitutional Tribunal through procedural irregularities, the National Council of the Judiciary was reformed so that it was selected by the legislature rather than judges, the retirement age for Supreme Court judges was lowered by 5 years while the number of seats was increased from 82 to 120, and the Minister of Justice gained significantly increased control of the common courts (Sadurski 2019, 96–126). While reforms to the courts have been the most extensive and visible, prompting a score of 1 out of 4 on possessing an independent judiciary from Freedom House, Sadurski (2019) provides an extensive review of all the changes, both in the legal realm and in terms of policy. The media, NGOs, the right to free assembly, the electoral commission, and the European Union have all been the targets of reforms that track those seen in Hungary under Orbán (Puddington and Roylance 2017). These changes have prompted widespread academic reconsideration of Poland’s democracy and suggest that it is less stable than previously understood (Bunikowski 2018; Folvarčný and Kopeček 2020; Nalepa 2019; Sata and Karolewski 2020; Vachudova 2019). While it remains technically characterized as a democracy, it is more accurately understood to be a transitional regime.

Figure 4 offers the results of four models using Polish respondents, each examining the impact of a different type and measurement of bias. Similar to the results from Thailand, most of these models show that bias and the gender balance of the court has no effect on our respondents’ views on whether or not the court will contribute to the health and well-being of the democracy; again, none of the interactions of benevolent bias and court gender balance are significant, and neither is one of the measures of hostile bias. However, just as was the case for Thailand, one of the measures of hostile bias does have a significant effect when coupled with a majority-women high court. In Model 1, hostile bias is measured with a question that asked how much the respondent agreed or disagreed with the statement that “women are too easily offended,” and this model demonstrates a significant effect of this variable on assessments of the court’s contribution to democratic health. Again, in an effort to more easily interpret these results, we generated predictive margins that allow us to see the predicted probabilities based on differing levels of bias and court composition.OPEN IN VIEWER

Figure 5 offers a visual representation of predicted probabilities generated from Model 1, and presents a surprising twist. Our Polish respondents behaved differently than those in Thailand; when faced with a majority women high court, the probability that they believe that the high court contributes to democratic health did not increase with hostile bias. Instead, they appear to have the more traditional and expected response: those with highest hostile bias against women were most likely to believe that the majority men high court will contribute to the health and well-being of the democracy. It is important to note that there is only a significant difference at the highest level of bias; for all other levels of hostile bias, the gender composition of the court has no effect. While these results do not support our hypothesis, we caution against using this case as definitive evidence that our theory is wrong; rather, we assert that the Poland results illustrate a further complication when investigating the effect of bias that scholars should consider in future analyses.OPEN IN VIEWER

We believe that recent events in Polish politics, and particularly the actions of one woman justice on the Polish high court—Constitutional Tribunal (CT)—explain the unexpected reaction by Polish respondents to women justices. Julia Przylebska was one of five judges elected to the CT by a PiS majority immediately upon their majority taking control of the Sejm in December 2015. Przylebska’s appointment was less controversial than those of three of her colleagues; she was appointed to a seat that was actually vacant, unlike three others who were appointed to seats that already had judges elected by the previous PO/PSL majority (Sadurski 2019, 63–64). And yet she quickly became acting president of the CT, which carried its own controversies. The position had just been created by statute and did not exist in the constitution (there was already a constitutionally recognized vice president position). In her selection by the General Assembly of the CT, eight judges refused to vote and another was unable to make it back to Warsaw in time to cast a vote. She was elected by a majority by including the three “quasi-judges” who were incorrectly appointed by the PiS majority (Sadurski 2019, 65–66). Once she was in this position, she allowed the three controversially selected judges to sit on panels and hear cases. She also established new rules for the court that limited what dissenting judges could say in their opinions and reformulated the composition of panels of judges almost a hundred times through 2019 (Sadurski 2019, 68–69). By 2018, some of the older judges on the CT wrote a joint letter that was leaked to the press challenging her composition of panels in ways that were exclusively in favor of PiS positions (Sadurski 2019, 70). Judge Przylebska, in short, actively countered the stereotype that women are “naturally” democratic and advocate for democracy; in fact, her behavior against democratic principles was so extreme and blatant that we believe she caused the people of Poland to update their stereotypes and expectations about women on the high court. Existing literature on exposure to counter-stereotypic behavior finds that the presence of women in high prestige jobs, for example, leads to people updating their stereotypes about women’s competence (Eagly et al. 2020), and thus it seems reasonable that exposure to an anti-democratic woman judge could alter how people in that country perceive the likely behavior of women on the high court. In addition, because the Constitutional Tribunal has been so instrumental in undermining the democracy, there is reason to believe that people have little faith in its ability to save the democracy, no matter the gender balance of the judges.

Conclusion

As the first study to explore the role of bias in perceptions of women judges in transitional regimes, this project offers a number of insights, but also raises significant questions for further exploration. Our evidence suggests that this relationship is neither simple nor straightforward; there does appear to be a relationship between the presence of women justices and perceptions of better democratic health, but that relationship is mediated by the amount of hostile sexism held by respondents and is only found in Thailand. In Poland, the gender-balance of the court has a very minor impact on the respondents’ beliefs that the court will save the democracy and the relationship is reversed.

While our results are not as consistent as one might hope, we see this as an important step forward in better understanding the unique impact of different types of bias on women in politics. Our argument and evidence on hostile sexism adds a dimension that is largely absent from the extant literature; by accounting for the interactions between different types of sexism and regime type, our findings suggest a need for a more nuanced understanding of how and when women’s presence shifts public perceptions. And there is no reason to believe that this dynamic is limited to the judiciary, opening the door to a broad reassessment of the relationship between the public and women officials.

The implications of this research suggest that not only do transitional regimes benefit from the increased presence of women on the court but a particularly receptive audience for their presence is, ironically, the hostile sexists in society. Since hostile sexism is associated with traditional and conservative views, this may be of particular value to right-leaning parties that target these voters as supporters. As Araya, Hughes, and Pérez-Liñán (2021) find in Latin America, it is leftist governments who are more likely to appoint women to courts, but these gains in gender diversity are difficult to sustain over time. Our findings here may offer both an explanation for why leftist governments receive less political benefit from appointing women and a reason for right-leaning governments in transitional regimes to appoint more women strategically. At the same time, the case of Poland highlights that there may be limits to the regime gains of appointing women if those women are perceived as central to the movement away from democracy.

Future literature on this topic has a wide variety of avenues to explore. We encourage scholars to dive more deeply into understanding both how sexism manifests differently in countries around the world, as well as to reconsider expectations that sexism means an automatic rejection of women’s power. Now that powerful women are relatively common, sexism may be manifesting less as “I don’t like her and want her excluded” and more as “I don’t like her but I can use her to my advantage.” That is, the sexism remains, but the rational actor has realized that exclusion is impossible and thus may be reacting to women in more nuanced and self-interested ways, as Valdini (2019) suggests. In addition, there is very little existing work on the impact of the environment of transitional regimes, and yet more and more countries are demonstrating regime instability in recent years. It is critical to understand how the unique context of regime fluidity and uncertainty may trigger people in these countries to react in unexpected ways, and particularly how this impacts the political power of traditionally marginalized groups.

Supplemental Material