The Role of Judge Gender and Ideology in Hiring Female Law Clerks

Abstract

Federal law clerks play a vital role in the development and implementation of the law. Yet, women remain underrepresented in these positions. We suggest that one reason for this underrepresentation may be differences in hiring practices among judges in the federal judiciary. Specifically, we hypothesize that male judges and conservative judges may be less likely to hire female law clerks than female judges and liberal judges for two reasons. First, gendered attitudes held by judges may make some judges prone to hire women and/or others more resistant to these hires. Second, due to ideological asymmetries between the law clerk pool and judges in the federal judiciary, conservative judges and male judges may be less likely to hire women law clerks. Using data on clerks hired in the federal judiciary between 1995 and 2005, we find support for both mechanisms.¹

When looking at the gender composition of Supreme Court clerks, it has long been recognized that the institution has a gender problem, ¹ with women consistently being underrepresented in these positions. Despite women’s increasing presence in law school cohorts, ² they continue to regularly be underrepresented among law clerks. Indeed, in the most recent term, just 39 percent of the justices’ clerks were women, despite being the majority of law school graduates. ³ Notably, though the Supreme Court has taken the brunt of the criticism regarding women’s exclusion from clerkships (relative to their numbers in the legal profession and broader population), the gender gap in clerkships is present in the lower levels of the federal judiciary as well. While women fare better in the district and circuit courts, they are still underrepresented among clerks.

The underrepresentation of women in clerkships has potentially important implications for judicial outcomes, as more representative courts can be expected to better grapple with the ways in which gender matters in our society. Scholars have found, for instance, that when it comes to cases involving highly gendered issues (such as sexual harassment), women judges are far more likely to rule in favor of plaintiffs (Moyer and Tankersley 2012; Haire and Moyer 2015; Boyd 2016) and to shift the decision-making calculus of their male colleagues (i.e., Boyd, Epstein and Martin 2010). Further, the substantive importance of women’s inclusion extends down the hierarchy to clerks. Though clerks attempt to “channel” their judges/justices while doing their job (Kenney 2000), women clerks likewise appear to shift substantive outcomes in cases where gender is highly salient. Kromphardt (2017) finds that among conservative justices, the inclusion of women among their clerks is associated with more liberal voting on sex discrimination and abortion cases, suggesting that women clerks are shaping substantive outcomes in meaningful ways. Thus, women’s representation in the courts has very real ramifications for the amount of substantive representation that women in the public can expect to receive from judicial institutions.

Yet the benefits of women’s representation in the courts extend beyond judicial output. The inclusion of women in the judiciary broadly – and among clerks specifically – also has indirect downstream consequences for substantive representation in the legal profession writ large, as clerkships often serve as the entry point to positions in major law firms (Zaretsky 2018), the legal academy (Redding 2003), or judgeships (Badas 2020). Thus, when women are underrepresented as clerks, it can (potentially) indirectly influence substantive representation because it means women will have less access to elite positions where they have the ability to shape outcomes later on in their careers.

Given the substantive consequences of women’s inclusion (or exclusion) in the judiciary, understanding the factors that underpin the current gendered patterns we see among federal law clerks is vitally important. While some scholars have pointed to “supply-side” factors such as ambition (i.e., Badas and Stauffer 2023), a more thorough examination of “demand-side” factors – meaning the hiring decisions made by the justices and judges – is needed. Although several studies have attempted to shed light on gendered hiring trends in the judiciary (Brown 1996; Kaye and Gastwirth 2008; Szmer, Kaheny and Christensen 2014), these studies have focused their attention on the Supreme Court. But the small sample size of Supreme Court justices – and their clerks – has limited our ability to make clear and generalizable inferences. Moreover, in some ways, focusing on gender and hiring at the Supreme Court is examining the end of the process. In the modern era, Supreme Court clerks almost uniformly have experience clerking for judges in the District Courts or Court of Appeals. In this respect, judges in the District and Circuit courts serve as key gatekeepers in shaping the applicant pool of Supreme Court clerks. In order to more fully understand women’s access to the most elite clerkship positions, then, we need to move further down the chain and understand the hiring decisions of judges at all levels of the federal judiciary.

In this study, we advance our understanding of gender and clerkships by analyzing judicial hiring practices among federal judges from 1995 to 2005. Beyond the theoretical necessity of understanding hiring decisions among these judges, a shift to include District and Circuit Courts allows us to gain more analytic leverage by drastically increasing the number of judges and clerks in our analysis. In examining hiring practices among federal judges, we argue that more liberal judges and women judges (across the ideological spectrum) will be more likely to hire women clerks compared with more conservative and male judges. We propose two potential mechanisms for these expectations. The first has to do with the attitudes of judges related to women’s inclusion in the judiciary. Women and liberal justices may have a greater intrinsic motivation to improve descriptive representation and thus be more likely to prioritize hiring women clerks. Conversely, biases among men and conservative judges may make it more difficult for women to gain access to clerkship positions with these individuals.

A second potential mechanism has less to do with attitudes about descriptive representation and more to do with ideological preferences. On average, the women sitting on the federal bench tend to be more liberal compared to men (Sen 2017). Likewise, women in the public – particularly younger women – tend to be more liberal (Norrander and Wilcox 2008). ⁴ Perhaps more importantly, when compared to men, women are often perceived as more liberal (Koch 2000, 2002a). As past research has made clear, judges often attempt to hire clerks who are ideological “matches” (Bonica et al. 2017; Bonica and Sen 2020). Thus, the gender of potential clerks may serve as a heuristic for women and liberal judges who may perceive women clerks as more ideologically proximate to their own preferences. In contrast, conservative and male judges may be less likely to hire female clerks because they experience – or perceive – a greater ideological mismatch.

We test both of these mechanisms and find evidence of both. Women judges – regardless of ideology – are more likely to hire women clerks than men, and liberal judges are more likely to hire women clerks than conservative judges. These findings have important implications for understanding how to improve women’s representation in the ranks of federal law clerks. First, these results shed more light on why a gender gap exists among federal law clerks. While prior research by Badas and Stauffer (2023) identified a supply side component to women’s under representation, our findings show that there is also a demand side component. This means that any intervention focused on increasing gender diversity among law clerks cannot focus on only one side of the equation – it must address both supply and demand side factors. Second, our results suggest that by nominating more women judges, regardless of ideology, presidents can secure a second dividend in terms of greater women’s representation among law clerks. This greater representation, in turn, will have a feeder effect into the legal profession more broadly. Third, aspiring women law clerks who want to work for conservative judges may need to strongly signal their conservative ideology, particularly when applying to male conservative judges.

The judicial hiring process

Every year, federal judges in the United States seek to fill somewhere between one and five term law clerk positions, or judicial clerkships. ⁵ Judicial clerkships are highly prestigious and sought-after. Within the legal profession, a judicial clerkship is widely seen as a significant career stepping stone. Law firms heavily recruit former clerks, often with significant hiring bonuses, and clerks are more likely to go on to careers as law professors, ⁶ appellate litigators, ⁷ or even judges. ⁸ Not surprisingly, then, the application process for judicial clerkships is exceedingly competitive.

Law clerk hiring is a decentralized process (Avery et al. 2001). Applicants apply directly to individual judges and must tailor their application materials to indicate both why they want to clerk for each judge and why they have the skills to be a successful clerk. Because the application process is so competitive, and because judges differ when it comes to what they think constitutes the “ideal” applicant, most applicants cast a wide net and apply to multiple judges. This means that judges are receiving a large number of applications for each open position. Existing data suggests that on average there are roughly 110 applicants per open position, but this number is likely higher for the most sought-after clerkships, particularly those that are viewed as more likely to lead to a subsequent clerkship at the Supreme Court (Baum and Ditslear 2010). In 2017, for instance, there were at least 2,433 applicants for just 13 spots in the Eastern District of Pennsylvania, while the District of Columbia, perhaps the most prestigious district when it comes to clerkships, saw at least 4,731 applications for only 32 positions. ⁹ Judges therefore face considerable choice when it comes to staffing their chambers each year.

Gendering the judicial hiring process

Because the judicial hiring process offers considerable autonomy to judges, the question naturally becomes whether some judges are more or less inclined to hire female clerks. Of course, it is important to acknowledge that the ability of judges to hire women is conditional on the pool of applicants. Work by Badas and Stauffer (2023) suggests that the supply of female applicants for clerkships will be shallower than the pool of male applicants due to a gender gap in ambition to hold these posts. Moreover, they find that women hold themselves to a higher standard of qualification than men, believing that they must be especially qualified before they report considering an application. The implication then is that the women who do emerge as applicants will be of higher quality than men. While Badas and Stauffer (2023) are unable to test this directly, such a conclusion would be consistent with the research on legislative candidates, which finds that while women are less likely to express ambition than men, the ones who do emerge are higher quality (Anzia and Berry 2011; Fulton 2012; Lazarus and Steigerwalt 2018).

Research on supply-side factors of women clerks indicates that the pool of female applicants will be smaller, yet likely higher quality, than the pool of male applicants. This may imply an advantage for the women that do decide to apply. However, such a conclusion cannot speak to the gendered patterns that might be present on the demand-side of the equation (i.e., the factors that underpin the hiring decisions of judges). If some judges are more likely or less likely to recognize women’s qualifications, this creates a gendered terrain after the initial emergence process. In other words, even after clearing the hurdle of deciding to apply for a position, women applicants may still be subjected to gendered hiring practices based on the preferences of the judges to whom they apply.

We propose two characteristics that may influence a judge’s predisposition to hire female clerks: gender and ideology. Specifically, we argue that women and liberal judges will be more likely to hire female clerks than male and conservative judges, respectively. There are a few potential reasons for this expectation. First, women and liberal judges may have stronger preferences when it comes to advancing women’s descriptive representation within the federal judiciary and/or be more likely to recognize women applicants as competent and qualified. On the other side of the coin, men and conservative judges may be more actively resistant to women’s advancement in the judiciary and more prone to down-weight or question their qualifications. Another explanation is that any gendered patterns we observe in hiring practices may actually be more about ideological congruence rather than gender per se. If judges seek to match with law clerks of a similar ideological bent, the ideological distribution within the judiciary – where women are generally more liberal (Bonica and Sen 2020)– and among prospective clerks may lead women and liberal judges to hire more women and male and conservative judges to hire less. We consider each of these characteristics in turn.

Attitudes about women’s inclusion

One reason to expect gendered differences in clerk hiring patterns may lie in individual judges’ attitudes about the value of women’s inclusion in the judiciary. In particular, women and liberal judges may have a greater commitment to incorporating more women into the judiciary, while male and conservative may be more resistant or antagonistic toward women’s inclusion.

Turning to gender first, there are two theoretical reasons to suspect judge gender may influence judges’ propensity to hire women clerks. First, women – across the ideological spectrum – may have a greater intrinsic motivation to advance the representation of their “gender group.” Though some scholars have questioned the strength – or even existence – of a gender consciousness that motivates women’s political behavior (Conover 1984, 1988), other research suggests that women elites often feel pressure to act as representatives of their gender (Reingold 2003). In a legislative context, examples abound of women “acting for” women through bill (co-)sponsorship (Osborn 2012; Swers 2016, 2019), policy support (Sanders 2018), speaking patterns (Pearson and Dancey 2011; Dietrich, Hayes and O’brien 2019), and policy outputs (Reingold, Haynie and Widner 2020). While judges serve a different function than legislators, research still finds that women judges substantively represent women in some instances. In cases related to the highly gendered area of sexual harassment and discrimination, women judges are more likely to rule in favor of plaintiffs (Boyd, Epstein and Martin 2010; Moyer and Tankersley 2012; Haire and Moyer 2015; Boyd 2016), and women’s presence on judicial panels induces more pro-women behavior from male judges as well (Boyd, Epstein and Martin 2010). Though behavioral differences between men and women judges are not widespread, to the extent they do exist, the gender of potential clerks may serve as a heuristic that the applicant has positions and priorities in line with the judge’s preferences on these issues. ¹⁰

Another way that women may “act for” women is through working to incorporate more women into political institutions. The unique experiences that women in positions of power face may lead them to have a distinct understanding of the importance of women’s inclusion in positions of power as a policy issue in its own right. Women only gained admittance to law schools in sizeable numbers beginning in the 1970s, and as such, many of the more senior women judges are likely to have built their careers during a time in which women were very much a minority in the legal profession. ¹¹ Moreover, although increasing numbers of women have entered the legal profession over the past several decades, gender parity in the profession remains a distant prospect. Entering associate classes have been comprised of approximately 45% women for several decades now, ¹² but women continue to be underrepresented as they move up the law firm ladder. ¹³ We see similar statistics in the judiciary: women still account for only slightly more than a quarter of federal judges. ¹⁴ Thus, a leaky pipeline problem clearly remains within the legal profession.

Given their success and possible obstacles faced along the way, women judges might be particularly attuned to the challenges women face when aspiring to professional careers, and this awareness may filter into their hiring decisions. ¹⁵ Moreover, at the mass level, scholars have identified a “base-line gender preference” such that voters prefer to be represented by candidates who share their gender, and this preference is stronger among women (Sanbonmatsu (2003); see also Badas and Stauffer (2018), Badas and Stauffer (2019), Stauffer and Fisk (2022)). While there are of course differences between voters selecting candidates and judges hiring clerks, if women have an underlying preference for women’s descriptive representation, all else equal, this may manifest in the hiring process.

While there is theoretical reason to believe that women judges may be predisposed to hiring women clerks, it is also possible that men on the bench are more antagonistic toward increasing women’s inclusion and that these attitudes make them less likely to hire women. In their work on confirmation hearings, Boyd, Collins and Ringhand (2023) use social identity theory (i.e., Tajfel and Turner 1978) to argue that male senators view women as an “out-group” and apply negative stereotypes related to competency and experience to female nominees seeking to gain access to the bench. Indeed, they find male senators – particularly those who do not share a nominee’s party – are more likely to question female nominee’s qualifications, more likely to interrupt female nominees, and use different types of language when interacting with female nominees. Similar in-group vs. out-group thinking may occur among hiring judges leading them to be more skeptical of female applicants and their qualifications. Indeed, evidence indicates that judges are more likely to interrupt female attorneys and permit them less speaking time relative to men (Patton and Smith 2017; Boyd, Collins and Ringhand 2023). Patton and Smith (2017) attribute these patterns to gendered schemas which hold up the court as a masculine domain, in which women are seen as less of a “fit.” Patton and Smith (2020) argue that while the view of the court as a masculine space likely impacts all judges, women judges hold more egalitarian attitudes that serve as a counterweight, though they do not find strong evidence of this among Supreme Court justices. A similar argument may apply to the hiring decisions made by judges, with gendered views of the court being more likely to bias men’s decision-making against women.

Ideology may also shape the attitudes and behaviors of judges relating to the importance of women’s inclusion in the judiciary. Just as women may view increased representation as important, so too may liberal judges (of all genders). Grossmann and Hopkins (2016) argue that the Democratic party is properly understood as a coalition of social movements wherein most Democrats are motivated by specific policies designed to benefit particular social groups, while the Republican party is best viewed as the agent of an ideological movement whose members are united by a common commitment to limited government. Moreover, Stauffer (2023) finds that women’s inclusion legitimizes political institutions for Democratic men and women, while Republican men are relatively ambivalent to women’s inclusion. ¹⁶ These party differences may lead more liberal judges to view the advancement of “group interests” as more important than conservative judges.

While liberal judges may have an underlying preference for hiring women, bias may come into play when conservative judges make their hiring decisions. Just as Patton and Smith (2020) argue that women hold more egalitarian views that can disrupt gendered schemas upholding the court up as a masculine institution, so too do they argue that liberal judges hold these attitudes. These attitudes serve as a counterweight to views of the court as a masculine domain, and as a result, liberal judges are less likely to view women as ill-fitting for the institution. Indeed, Patton and Smith find evidence that the higher interruption rates faced by female attorneys are almost exclusively driven by conservative judges. Other research similarly suggests that conservative judges may be uniquely antagonistic toward women in the judiciary with conservative judges exerting more “verbal control” over female attorneys (Phillips and Carter 2009) and being less likely to support litigants represented by women (Szmer et al. 2013). These patterns suggest a bias against women who attempt to break into the space of the judiciary, and may likewise manifest in hiring decisions, where conservative judges may likewise view women applicants for clerkships as less of a “fit” for the role.

Beyond their attitudes about the value of women’s inclusion in the judiciary broadly, liberal and conservative judges may view the legality of taking gender into account during the hiring process quite differently. While Democratic-appointed and liberal-leaning judges have held that affirmative action programs are constitutionally sound efforts to redress systematic inequality, Republican-appointed and conservative-leaning judges have been much more likely to view such programs as unconstitutional discrimination. Indeed, it was no surprise when the Supreme Court’s recent decision striking down affirmative action programs ¹⁷ was split along ideological lines. Although much of the debate has centered on the use of race and ethnicity in admissions and hiring decisions, there is reason to believe these partisan divisions may also extend to attitudes about the use of gender. While the tests used to establish state-based racial and gender discrimination differ (state action based on race receives a higher level of scrutiny), the same arguments made in opposition to affirmative action based on race can be made (and have been made) with respect to affirmative action based on gender. ¹⁸

Desire for ideological proximity

It is also possible that judges’ hiring decisions have less to do with preferences for descriptive representation and more to do with the ideological proximity between a judge and a potential clerk. First, existing research – and comments from judges themselves – suggests that many judges take ideology into account when hiring law clerks (Bonica and Sen 2017, 2020; Ditslear and Baum 2001). This is true across the different levels of the judiciary and across the ideological spectrum (Bonica et al. 2017). Liberal Supreme Court justices, for instance, are much more likely to hire clerks who previously clerked for judges appointed by Democratic presidents, while conservative justices are more likely to hire clerks who clerked with a Republican-appointed judge. Justice Thomas, in fact, once famously said that he “won’t hire clerks who have profound [ideological] disagreements with” him because it’s “like trying to train a pig. It wastes your time, and it aggravates the pig.” ¹⁹

While we expect all judges to consider ideology when selecting clerks, there is also reason to suspect that ideological bona fides play a greater role in hiring decisions among conservative judges. First, as noted earlier, Grossmann and Hopkins (2016) have argued that Republicans and Democrats think about partisan politics quite differently, such that Republican party members care much more about ideological purity than Democrats. And while, in theory, partisanship should play no role in judicial appointments or judicial behavior, reality has never been so simple (Segal and Spaeth 2002; Epstein and Segal 2005). Judges are citizens just like anyone else, and they possess political opinions. Those who align more strongly with the Republican party, the party driven most by ideology, may be more likely to seek ideological proximity in their law clerks than judges who align more strongly with the Democratic party and its focus on promoting particular social groups. Second, the rise of the Federalist Society, and its increasing influence in conservative legal circles, has effectively created a gatekeeper for aspiring clerks on the political right, with only the most ideologically pure being given the group’s endorsement (Hollis-Brusky 2015; Bird and McGee 2023).

A tendency on judges’ part, particularly conservative judges, to hire clerks who are ideological matches has significant implications for women’s ability to secure federal clerkships for several reasons. First, public opinion polling has long shown that women tend to be more liberal than men (Chaney, Alvarez and Nagler 1998; Wolbrecht 2010; Kaufmann and Petrocik 2020). Moreover, whether women are more liberal than men or not, they are frequently perceived as such due to belief stereotypes about women as more left-leaning compared to men (Huddy and Terkildsen 1993; Koch 2002b; King and Matland 2003). These belief stereotypes may mean that conservative judges are less likely to see women as ideological “matches.” Second, and in contrast, recent scholarship by Bonica and Sen shows that federal judges trend rightward in terms of ideology, such that “the average American judge is slightly right of center (Bonica and Sen 2020). And the average judge becomes even more conservative as one moves up the judicial hierarchy, with circuit judges more likely to be conservative than district judges. Third, the majority of women federal judges were appointed by Democratic presidents, meaning that there are far more liberal women judges than conservative women judges. ²⁰ Thus, as Figure 1 demonstrates, the ideological distributions of clerks and judges vary considerably, particularly once one accounts for gender. ²¹ Male judges and male clerks have average scores of .14 and -.42, respectively, while female judges and female clerks have average scores of -.36 and -.85, respectively.

Figure 1. Ideological Distributions by Position and Gender, 1995 - 2004. Data from Bonica et al. (2017). Figure generated by the authors.

Taken together, these demographic trends suggest that, if judges are truly focused on hiring ideological matches, it will be much harder for conservative and male judges to find qualified female clerks who are ideological “matches” than it will be for liberal judges, a group that will include the majority of female judges. This hypothesis is supported by the recent findings from Fogel and coauthors’ study of circuit judges, which found that Republican appointees had a more difficult time recruiting female clerks than Democratic appointees. One Republican appointee stated that conservative students tended to apply to conservative judges and that, because “far fewer female law students are conservative,” her applicant pool was largely male (Fogel, Hoopes and Liu 2022, 29).

On the flip side, because there are more conservative-leaning judges on the federal courts than liberal-leaning judges, this would also suggest that there are fewer clerkship positions that are realistically attainable for aspiring female clerks. In other words, liberal judges may face a surplus of quality female applicants, while conservative judges face a drought. Notably, even conservative women clerks may face a harder time being hired by conservative judges due to the perception that they are more liberal (i.e., Koch 2002; King and Matland 2003) Because judges often must sort through hundreds of applications to identify potential candidates worth interviewing, the process may be particularly susceptible to the influence of various mental shortcuts (i.e., biases). This may lead judges to view female candidates as more likely to be liberal, on average, than male candidates. It may therefore be particularly important for conservative female applicants to vigorously signal their ideological bona fides to make it on (some) conservative judges’ short list.

Data and analysis: Selecting female law clerks

To assess our expectations of gendered patterns in law clerk hiring, we rely on data generated by Bonica et al. (2017) which provides information on all law clerks and the judges who hired them between 1995 and 2005. ²² The data include law clerk hiring data for the District Courts, the Circuit Courts of Appeals, and the Supreme Court. In total, the data include information on 14,429 hiring decisions. Included in the law clerk data is the clerk’s gender, the law school the clerk attended, and the clerk’s CF score. ²³ CF scores are measures of ideology. CF scores are estimated using the Bonica (2014) Database on Ideology, Money, and Elections (DIME). DIME leverages campaign contributions to estimate ideology for individuals who have donated in local, state, or federal elections (Bonica 2014). The Bonica et al. (2017) law clerk data include the same information on the hiring judge. For judges who have made no campaign contributions, ²⁴ their CF score is imputed based on their judicial common space score (Epstein et al. 2007) and demographic traits of the judge. ²⁵ The imputed scores are highly correlated with other measures of judicial ideology (Bonica and Sen 2017).

We first consider whether women and liberal judges are more likely to hire women law clerks relative to men and conservative judges. Our dependent variable is whether the hired clerk was a woman or not. Our primary independent variables are the hiring judge’s gender and the hiring judge’s ideology. If our expectations are met, we anticipate women judges will be more likely to hire women clerks relative to male judges, and that liberal judges are more likely to hire women clerks relative to conservative judges.

Beyond our primary independent variables, we also need to be aware of potential confounders. Judges often hire clerks from the law school they attended (Peppers 2006). Thus, we include a variable that captures whether the judge and the hired clerk attended the same law school. Gendered hiring patterns may also vary between levels of the judicial hierarchy. For example, Badas and Stauffer (2023) find that women are less likely to articulate ambition for Supreme Court and Circuit Court clerkships. So it may be the case that these judges are less likely to hire women due to gendered differences in the applicant pool. For this reason, we include fixed-effects for each level of the judicial hierarchy. Finally, we include fixed-effects for hiring year. Women’s inclusion in law school cohorts has increased steadily over time (Moyer and Haire 2015). The year fixed-effects will allow us to account for the fact that there are less women in the potential pool of law clerks early in the time period analyzed – and therefore less opportunity to hire women clerks. By including fixed-effects for both level of the judiciary and year, we account for potential gender differences in the clerk application pool to the best of our abilities.

Since our dependent variable is a binary outcome, we estimate a logistic regression model, which is presented in Table 1. ²⁶ Overall, the results support both of our expectations. Women judges are more likely to hire women law clerks compared to male judges. The substantive effect is presented in Figure 2. Women judges are estimated to hire women 50.7% of the time, while male judges are expected to hire women 47.6% of the time. The difference of 3% is statistically distinguishable from zero ( $ p. $ =.003). Based on these probabilities and the gender makeup of the judiciary, ²⁷ the results imply that there are roughly 34 fewer female clerks per term. Over the ten-year period studied here, that would mean a total of roughly 340 fewer female clerks were hired.

Table 1. Logit Regression Model: Selection of Female Clerk

Standard errors in parentheses.

* p<0.05,

^** p<0.01,

^*** p<0.001.

Figure 2. Probability of Hiring Female Clerk by Judge Gender.

We also observe that liberal judges are more likely to hire women law clerks than conservative judges. The substantive effect across the range of judge ideology is presented in Figure 3. The histogram embedded in the figure displays the overall distribution of judge ideology. The average probability of a judge with a CF score less than zero hiring a female clerk is .503, while the average probability of a judge with a CF score greater than zero hiring a female clerk is .465. Based on these probabilities and the ideological makeup of the judiciary, the results imply that roughly 33 fewer female clerks are hired per term due to ideological preferences. ²⁸ For the entire ten-year period studied here, this would imply that 330 fewer women clerks are hired.

Figure 3. Probability of Hiring Female Clerk by Judge Ideology.

Histogram on x-axis displays distribution of judge ideology.

Data and analysis: Testing potential mechanisms

We have observed that women judges are more likely than male judges to hire women as law clerks, and we have observed that conservative judges are less likely than liberal judges to hire women as law clerks. Previously, we outlined two potential mechanisms that might explain our results. First, the gender of a judge may lead to gendered preferences in hiring patterns. Likewise, ideology may lead to liberal judges preferring to hire women and/or conservative judges being more resistant to increased diversity. Second, considering the ideological-gender asymmetries in the pool of federal judge and the pool of law clerks, it may be the case that conservative judges and male judges are less likely to hire women as law clerks because they have a more difficult time finding an ideological match between themselves and a female clerk than they do with a male clerk.

Gender and hiring decisions

To determine whether the gender attitudes mechanism is valid, we reestimate the model presented in Table 1 but include an interaction between judge gender and judge ideology. Our reasoning is that if the gender attitudes mechanism is valid, we should observe that women across the ideological spectrum are more likely than men to hire female law clerks and such an effect should not be limited to liberal women. The results of the model are presented in Table 2.

Table 2. Logit Model: Gender Representation Mechanism

Standard errors in parentheses.

* p<0.05,

^** p<0.01,

^*** p<0.001.

The substantive effect of the interaction is presented in Figure 4. The left panel displays the predicted probability of hiring a female clerk across the range of ideology for female and male judges, and the right panel displays the marginal effect of a female judge across the range of ideology. The results support the gender attitudes mechanism, as we observe a consistent gender effect across the ideological range. Women judges have a .031 higher probability of hiring a women law clerk than male judges, and the effect is statistically distinguishable from zero between the fourth percentile and sixty-eighth percentile of the CF ideology score. Thus, the majority of women judges have a preference for hiring women law clerks. Only extremely liberal women and very conservative women do not display a preference for hiring women law clerks. We cannot say whether this is due to a difference in preference among these two groups or because the limited sample of women judges at these threshold limits our ability to draw conclusive inferences. Our findings confirm, however, that women across the ideological spectrum are more likely to hire women as law clerks. Thus, it does appear that judges of different genders have different orientations to hiring women clerks, though we are unable to tell whether it is a preference among women, a bias among men, or some combination of the two that is driving this result.

Figure 4. Predicted Probability of Hiring a Female Law Clerk by Judge Ideology and Gender.

Histogram on x-axis displays distribution of judge ideology.

Importantly, these results also suggest that liberal male judges may be motivated by a preference for descriptive representation. The left panel of Figure 4 indicates that liberal male judges are more likely to hire a female law clerk than a conservative female judge. That is, while female judges are more likely to hire a female law clerk compared to their male ideological counterparts, both female and male liberal judges are more likely to hire a female law clerk compared to their conservative colleagues.

Ideological Compatibility

For the ideological compatibility mechanism, we predict the ideological distance between the judge and their law clerk. If the ideological compatibility mechanism is valid, we anticipate observing greater ideological distance between judge and clerk when a conservative judge hires a female law clerk and when a male judge hires a female law clerk. The greater ideological distance would imply that a conservative judge would have to incur greater ideological costs when hiring female clerks, which should generally dissuade conservative judges from hiring women. We again rely on the data provided by Bonica et al. (2017). Both the judge’s ideology score and the clerk’s ideology score is measured using their CF score. We take the absolute difference between the two scores to produce a measure of ideological distance. For these analyses, we have a total of 5,880 observations. The reason for the more limited set of observations is because many of the clerks in the sample do not have CF scores. For this reason, we cannot measure the ideological distance between these clerks and the judge who hired them. ²⁹ The distribution of ideological distance between a judge and their hired clerk is summarized in Figure 5.

Figure 5. Distribution of Ideological Distance between Judge and Clerk. One value of 5.675 truncated from figure.

We estimate a linear regression model (OLS) predicting the ideological distance between a judge and their clerk. Our main independent variables of interests are the interactions between the judge’s ideology and the gender of the clerk and the interaction between the judge’s gender and the clerk’s gender. We also control for whether the judge and clerk went to the same law school, as it may be the case that judges are willing to sacrifice some ideological compatibility to hire clerks from their alma mater. We control for whether the clerk had previously held a lower court clerkship. These clerks come with recommendations from sitting judges and therefore may lead to better ideological congruence with the current judge. While most clerks serve for one year (73% of clerks in the data set clerked for a single year), some clerks are rehired by their judge. For example, one clerk clerked for Judge David Russell each year included in the data. Judges likely decide to retain clerks because they have a good working relationship with them. This likely includes an ideological match. So we control for the cumulative number of years a clerked as served with a judge. We also control for the clerk’s CF score. Purely extreme clerks likely lead to greater distance between themselves and their judges. The models also include fixed effects for level of the judicial hierarchy and year. The results to our model are presented in Table 3. Overall, the results demonstrate that the ideological compatibility mechanism is valid.

Table 3. Linear Regression: Ideological Distance between Judge and Clerk

Standard errors in parentheses.

* $ p<0.05 $ ,

^** $ p<0.01 $ ,

^*** $ p<0.001 $ .

The results of the interaction between judge gender and clerk gender are presented in Figure 6. For women judges, we observe no differences in expected ideological distance between themselves and their clerks based on gender. However, the same is not true for male judges. When a male judge hires a male clerk, the expected ideological distance is .957 compared to 1.021 when a male judge hires a female clerk. The difference of .064 is statistically distinguishable from zero and represents 9% of a standard deviation in the ideological distance measure. This implies that when male judges hire female law clerks, they are required to incur greater ideological costs than they would if they hired a male clerk. ³⁰ This additional ideological cost may lead to male judges hiring fewer female law clerks.

Figure 6. Predicted Ideological Distance between Judge and Clerk by Judge and Clerk Gender.

The results of the interaction between judge ideology and clerk gender are presented in Figure 7. For liberal judges, we observe that there is no ideological penalty for hiring a female clerk. In fact, it seems that liberal judges are more likely to experience ideological compatability when they hire female clerks compared with when they hire a male clerk. For example, a liberal judge (20% of CF score) is expected to have an ideological distance between themselves and a male clerk of .627 while only observing an ideological distance of .477 when hiring a female clerk. The difference of .15 is statistically distinguishable from zero and represents roughly a 21% standard deviation in the ideological distance measure. For conservative judges, we find the opposite effect. A conservative judge (80% of CF score) who hires a male clerk is expected to have an ideological distance of 1.335 between themselves and their clerk, while that difference is expected to be 1.517 when the conservative judge hires a female law clerk. The difference of .182 is statistically distinguishable from zero and represents a 25% of a standard deviation.

Figure 7. Predicted Ideological Distance between Judge and Clerk by Judge Ideology and Clerk Gender.

Histogram on x-axis displays distribution of judge ideology.

Conclusion

The underrepresentation of women in the most elite legal positions remains an important topic of discussion in the legal community. ³¹ Inherent in these discussions is the tacit acknowledgment that for institutions to be perceived as legitimate and just, it is important that they be open to the inclusion and participation of a wide range of groups – not just a select few. Indeed, many often evoke the argument that positions in the judiciary – such as attorneys, clerks, and judges – should “look like America” (Fogel 2022). Beyond issues related to fairness and access, the dearth of women in the most elite legal positions has substantive ramifications for legal outcomes as well. Clerks often play a vital role in shaping the opinions produced by their judges, and women clerks in particular have been shown to influence how judges approach issues related to women’s issues (Kromphardt 2017). In this sense, the lack of women in the federal judiciary is troubling not only for reasons related to descriptive representation, but also the substantive representation that women in the public can expect to receive from the courts.

Despite the implications of women’s descriptive representation (or lack thereof) in the federal courts, our understanding of why women tend to be underrepresented in these posts remains limited. To the extent that we do have systematic analyses of women’s inclusion, research tends to focus on women’s access to Supreme Court clerkships. While these positions are among the most elite and influential, in order to gain access to these posts, women (and men) applicants must have already passed through a rigorous process, as nearly all Supreme Court clerks must have first held a clerkship with a lower-level court. In this sense, in order to truly understand women’s access to clerkships at the most elite level, we must understand their access to lower-level (but still elite) positions earlier in the chain. While past research has broadened its examination to include clerkships in all federal courts (i.e., Badas and Stauffer (2023)), this work has focused on supply-side factors (i.e., women’s ambition to apply for and hold clerkships) as opposed to demand-side factors (i.e., structural barriers in the hiring process that might impede women’s inclusion).

In this article, we advance our understanding of the factors underpinning women’s inclusion in clerkships by examining the hiring decisions of judges across the federal judiciary. In this respect, we are able to understand when and how gender influences clerkship hiring decisions, giving us insights into when and how women are able to break into the pipeline. Using data on federal law clerks and judges from 1995 to 2005, we explore how two factors in particular shape the prospects of women applicants: hiring judge gender and ideology. The results of our analysis underscore that gendered terrain facing female candidates who choose to apply for these positions. Indeed, we find that would-be female applicants have a much greater chance at success when the hiring judge is a woman and is more liberal.

After uncovering these initial patterns, we dug deeper to understand the reasons underpinning them. We find evidence for two mechanisms. The first appears to be a rooted in the orientations that men and women judges have toward women’s inclusion in the judiciary. Indeed, we find that women across the ideological spectrum are more likely than comparable men to hire women clerks, suggesting the presence of a gender-affinity effect in hiring. At the same time, that does not mean that ideology plays no role in the hiring process. Indeed, we uncover evidence that male judges often incur a loss in ideological congruence when hiring female law clerks. This means that for male judges, their relative reticence to hire women clerks may have less to do with bias against women and more to do with achieving an ideological fit with their clerks.

Importantly, although our analyses relied on data from 1995 to 2005, there are several reasons why our findings are likely to be replicated using more current hiring data. Not only have Democratic presidents continued to outpace Republican presidents when it comes to appointing women judges, but women in the public continue to lean more liberal than men. Moreover, while the Federalist Society appears to play a more prominent role in the clerkship process today (for conservative judges), this would suggest that conservative judges are better equipped to identify ideological matches in the hiring process.

The patterns we uncover in this study represent structural barriers for women’s inclusion in three ways. First, the hesitance of male judges to hire women clerks (at least relative to women judges) is a hindrance to women’s inclusion because men continue to be overrepresented in federal judgeships, ³² meaning a greater number of judges are less likely to hire women clerks. Second, because conservative judges are more likely to hire male clerks, this means women’s access is asymmetric across the ideological spectrum. More generally, in the current political landscape, the majority of “feeder judges” (those that most frequently send their clerks onto to even more elite positions) tend to be more conservative. ³³ Not only does this mean women’s access to conservative clerkships is limited, but it also means that their access to the posts that are likely to beget future positions higher up in the judicial hierarchy are limited as well. In this sense, the patterns we uncover show that while in some areas of the judicial system women are likely to be able to successfully make inroads, there are other areas where women are likely to be systematically excluded.

The research presented here highlights the continued need for scholars to examine when and how women gain access to positions of power. Indeed, as our findings show, women face a gendered terrain when attempting to access some of the most elite positions in the American judiciary – a pattern that mirrors women’s experiences in other aspects of American politics. Future work should continue to interrogate when and how women face structural barriers in their attempts to access the judiciary.