Judicial influence and the importance of intersecting identities

Abstract

Judges, like all of us, possess multiple intersecting identities. Drawing on social identity theory, we examine how gender, race, and partisanship jointly influence judicial citation practices. We hypothesize that in-group favoritism motivates judges to preferentially cite peers with whom they share salient identities, and that this effect intensifies as the number of shared identities increases. Using an extensive dataset of discretionary citations from search and seizure cases between 2000 and 2010, we find that sharing a single identity yields no significant effect. However, as the number of shared identities increases, so does the probability of citation. Judges who share race, gender, and partisanship exhibit a 25% higher probability of citation relative to those with no shared identity. These findings highlight the cumulative impact of intersecting identities on judicial influence and behavior, emphasizing that diversity is critical for the development of legal precedent and the law.

Identity plays a pivotal role in shaping judicial behavior, as scholarship has demonstrated that race and gender significantly influence judicial decision making (see, e.g., Szmer et al. 2023; Boyd et al. 2010). Recently, scholars are taking an increasingly intersectional approach to examine how multiple identities—including race, gender, and other individual characteristics—“intersect” or overlap to affect judicial behavior (see, e.g. Collins and Moyer 2008; Haire and Moyer 2015; Hinkle and Nelson 2018). Despite these advances, our understanding of how intersecting identities interact remains incomplete.

We build on this research by examining intersectional identities through the lens of citation behavior. Judicial citations not only enhance a precedent author’s reputation, but also reflect its influence on the development of law (Klein and Morrisroe 1999). We investigate how federal appellate judges, who enjoy broad discretion in their citation practices, may preferentially cite peers with whom they share one or more salient identities. Drawing on social identity theory, we hypothesize that in-group favoritism will lead to higher likelihood of citation as the number of shared identity traits increases.

To test our hypothesis, we analyze an original dataset of discretionary out-of-circuit citations among all published search and seizure cases from 2000 to 2010. Beause these citations are purely discretionary and unconstrained by legal doctrine (Szmer et al. 2023), they provide a window into the influence of intersecting identities on judicial behavior. Our results show that as the number of shared identities increases, so does the probability of citation. Judges simultaneously sharing race, gender, and partisanship are 25% more likely to cite one another than those with no shared identities. This finding suggests the importance of diversifying both liberal and conservative judges rather than simply examining the overall composition of the federal judiciary. Exploring judges’ intersecting identities provides significant insight into why some judges’ work languishes in obscurity while others’ repeatedly impacts the development of law. More broadly, our findings show how identity influences how we treat information from other actors, which could potentially inform scholars in other subfields, such as legislatures, or even other disciplines. In short, the nature of intersecting identities has important implications for the courts and beyond.

Judicial identities

Social psychologists argue that a person’s sense of self is significantly shaped by shared social identity, which can lead to preferential treatment towards those who share these identities (Tajfel 1970). Specifically, social identity theory holds that in-group favoritism is an unconscious psychological process rooted in common identity traits. Recent work in the judicial context applies social identity theory to examine how identity-based relationships shape judges’ decisionmaking and behavior (e.g. Haire and Moyer 2015; Hinkle and Nelson 2018; Tillman and Hinkle 2018).

We argue there are two primary reasons judges may favor those with whom they share an identity over their out-group colleagues. First, individuals gain a sense of pride and esteem from group membership and favoring one’s own group may boost that group’s profile, reputation, or standing (Haire and Moyer 2015). For example, citing a judge’s opinion increases that judge’s influence and also benefits the group(s) with which they both identify. In this way, judges may perceive a direct advantage in citing colleagues with shared identities (Hinkle and Nelson 2018).

Second, well-known cognitive biases reinforce the tendency to view in-group members as more intelligent, competent, and persuasive. ¹ This phenomenon aligns with broader psychological findings that people tend to seek and trust information that favors their own group (e.g., Hinkle and Nelson 2018). Consequently, judges, like others, may be particularly receptive to the arguments of colleagues whose identities overlap with their own.

But not just any shared group membership matters; the social identity must be sufficiently salient to elicit recognition. We focus on three of the most studied salient identities: race, gender, and partisanship. Prior research indicates that judges favor colleagues who share their racial/ethnic or gender-based identities. For example, circuit judges belonging to minority racial/ethnic groups positively cite in-group colleagues more frequently than others (Hinkle and Nelson 2018). Similarly, judges are more likely to assign their peers of the same race to write more published decisions (as opposed to unpublished rulings), whereas the pattern is reversed for racial out-group members (Tillman and Hinkle 2018). Gender can likewise trigger in-group favoritism, particularly in cases that explicitly activate shared gender identity, such as gender-based discrimination cases (Haire and Moyer 2015).

Political identity similarly shapes individuals’ beliefs, attitudes, and behaviors. Partisanship has become a social identity in which individuals view copartisans favorably and discriminate against members of the opposing party (Mason 2018). Federal judges, nominated in a highly partisan process, operate in an environment where their partisanship identity remains salient throughout their careers, and these identities continue into other facets of their work. Both U.S. Supreme Court justices and federal appellate judges are more likely to cite opinions written by their ideological allies, and partisanship influences which precedent they follow (Benjamin et al. 2024). Judges at both levels are more inclined to dissent when the author of the opinion is ideologically distant (e.g., Hazelton et al. 2023).

Thus, social identity theory predicts in-group favoritism to shape how judges vote, assign opinions, cite, and dissent. Yet, the intersection of multiple identities remains relatively unexplored. There is good reason to simultaneously examine gender, race, and partisanship as salient identities that shape judges’ behavior. Scholars have long embraced the importance of judges’ politics, ideology, gender, and race in shaping judicial behavior. A growing body of literature demonstrates that shared individual identities significantly influence judicial decisionmaking. However, focusing on a single facet of identity is inherently incomplete, as no individual possesses only one identity.

Although a judge’s race, gender, or partisanship individually may not inherently outweigh their other identities, we anticipate that as judges share more traits, they will be more inclined to cite one another than they will cite colleagues who do not share those traits. We argue that partisanship—given its well-documented significance in political science and judicial decisionmaking—merits particular attention, which may explain why previous literature suggests that identities' impact may be contextual.

In this study, we extend the analysis of in-group citations advanced by Hinkle and Nelson (2018) and other scholars in two critical ways. First, we introduce partisanship as an essential salient identity—a factor absent from their original focus—which may account for their finding of only partial evidence of identity-based favoritism among women and minority judges, with no comparable effect among men or white judges. Second, while previous work considered at most two shared identities, we examine the joint effect of all three identities—gender, race, and partisanship—in combination to determine whether having more shared identities produces even greater levels of in-group(s) favoritism, as their interplay profoundly shapes individuals’ perceptions and experiences (Collins and Moyer 2008; Hancock 2007). Yet, little attention has been paid to how such intersectional nuances emerge in the realm of relational identities.

We hypothesize that shared identities are most likely to affect judges in the more nuanced aspects of judicial behavior, particularly in the selection of persuasive precedents to cite. Unlike the binary decision to vote for or against a party, the act of citation involves hundreds of discretionary choices, with opinion authors selecting among numerous relevant precedents (Niblett and Yoon 2015). This selection process affords judges considerable latitude, especially when it comes to non-binding precedents, as binding precedents must be discussed as a matter of legal doctrine, while other sources fall entirely within a court’s discretion. Although substantive revisions may be subject to collegial negotiation, the authoring judge maintains substantial control over the content of the majority opinion (Bowie et al. 2014), and citation decisions typically escape such scrutiny. As we discuss next, these citation choices are crucial indicators of a judge’s influence over the development of the law.

Judicial influence

Judicial citation practices have long served as a critical lens for understanding patterns of policy adoption among elites. Citation decisions also provide an opportunity to examine individual level policy influence (Szmer et al. 2023). Citations—whether of judicial decisions or academic articles—signal that the referenced work has shaped the development of new legal arguments or frameworks. In this sense, when a judge acknowledges and builds upon another’s legal reasoning, it reflects that judge’s influence on the evolution of the law. Although judges work closely with law clerks, this collaborative process does not diminish the validity of citation-based influence measures, as judges retain ultimate responsibility for the opinions issued under their names and, thus, exercise close supervision (Bowie et al. 2014). Despite their inherent limitations (e.g. Klein and Morrisroe 1999), citations remain a valuable metric for assessing the relative influence of one judge over another (Landes et al., 1998).

Moreover, scholars have shown that judges’ citations of each other are quite varied (Hinkle and Nelson 2018). One of the many dynamics at play when a judge cites a decision is the personal relationship between the judges (Hazelton et al. 2023). Judges often cite their friends (Niblett and Yoon 2015) and their former colleagues (Szmer et al. 2020), especially when they served together for a long time (Pryor 2017). Moreover, circuit court judges from prestigious law schools tend to be cited more frequently (Szmer et al. 2020), and judges are more likely to cite one another when they share professional backgrounds (Hinkle and Nelson 2018). As discussed earlier, shared characteristics such as race, gender, and partisanship further contribute to citation behaviors. Together, these dynamics underscore the complexity of personal and professional relationships that shape the evolution of legal doctrine. They also provide evidence that judges are aware of many characteristics of their peers.

A judge’s capacity to shape the law is intrinsically linked to the willingness of other judges to cite their work. By leveraging social identity theory, we examine how intersecting judicial identities combine to influence the development of law. In-group citations may serve both as a means to reinforce the esteem of a particular group and as an outcome of cognitive biases favoring in-group members. These mechanisms are likely to operate simultaneously across multiple identity dimensions, producing a cumulative effect. Therefore, we hypothesize that judges who share a greater number of salient traits will cite each other more frequently than they cite judges with whom they share fewer traits.

Data & research design

We examine the effect of shared group membership on citation patterns in the U.S. Courts of Appeals. Analyzing when judges cite one another requires not only a set of opinions and their citations, but also a well-defined set of potentially applicable precedents that might have been cited, but were not. To delineate this choice set, we focus on Fourth Amendment search and seizure law—a discrete legal issue that routinely arises in both civil and criminal contexts and that may activate multiple identities. Although Fourth Amendment doctrine is not explicitly race-based, its application often intersects with racial considerations, potentially heightening race-based identities for some judges. While examining cases of explicit race- or gender-based discrimination might offer a more direct test of our hypothesis, our selection of search and seizure cases present a more challenging test while remaining broadly relevant, as many legal issues are experienced differently based on one’s race and gender.

Prior to 2000, there was unfortunately insufficient variation across all three dimensions of gender, race, and partisanship to test our hypothesis. Accordingly, we analyze citation patterns in search and seizure panel opinions from 2000 to 2010, while incorporating cases going back to 1995 as possible precedents. ² The unit of analysis is the opinion-precedent dyad, representing each instance in which an opinion author could have chosen to cite a precedent. Our outcome variable is a measure of whether they did so. It equals one if there is a non-negative citation and zero otherwise. We estimate a probit model and use robust standard errors clustered on the opinion.

Our primary explanatory variable captures the extent to which the identities of the opinion’s author and the precedent’s author overlap. ³ Using data provided by the Federal Judicial Center, we code each judge’s gender, race, and party. In the absence of publicly available data on federal judges’ political party, we assume that a judge’s partisanship aligns with that of the appointing president. At one extreme, two judges may share neither gender, nor race, nor political party. At the other extreme, the judges may share all three. We operationalize identity in the most parsimonious way—by simply counting the number of identities (ranging from zero to three)—which directly tests our hypothesis that a greater number of shared identities leads to more frequent citation. ⁴

We further control for a variety of factors that capture both the relational dynamics between the opinion and the precedent, as well as intrinsic characteristics of each. For instance, judges who belong to the same generational cohort may share formative historical experiences that shape their perspective in similar ways. Additionally, the substantive similarity of a precedent to the case at hand is a critical determinant of citation, as is the ideological alignment between opinion author and the cited decision. We account for ideology by controlling for whether an opinion author opposes the ideological direction of the cited decision, anticipating that Republican (Democratic) authors are less likely to cite precedents that reach liberal (conservative) outcomes. Including this control means that any observed effect of two judges sharing partisanship is present even after accounting for how much an opinion author is likely to agree with the precedent outcome. Finally, we include controls for various other aspects of the opinion and precedent that are known to affect citation practices. Further details on control variables and data collection are available in the supplemental material, along with summary statistics and full regression results.

Results

Modeling potential citation dyads necessarily involves a tremendous amount of data. Analyzing such a large number of observations creates the possibility that very small effects can be estimated with statistical significance. Therefore, it is especially important to focus on substantive significance as well. As hypothesized, judges are increasingly likely to cite a precedent the more identities they share with its author and that effect is statistically significant. But is it large enough to be substantively notable? The overall probability of citation in the dataset is only 0.0007. The effect of shared identities must be evaluated in light of the reality that less than one-tenth of a percent of precedents are cited. Figure 1 illustrates the impact of shared identities by showing the predicted probability of citation for each of the possible values of Number of Shared Identities. When the opinion and precedent authors share neither the same race nor gender nor political party, the precedent is cited an estimated 0.062% of the time. That increases to 0.078% when all three of those identities are shared. This represents a relative increase of 25%.

Figure 1.

Predicted probability of citation: this graph provides the predicted probability of citation for each value of Number of Shared Identities. Other variables are held at their mean. The whiskers depict the 95% confidence interval around the predicted probabilities.

In circumstances like these where the underlying probability of citation is very low, but such decisions are being made by the dozens nearly daily, it is useful to consider relative changes as informative with respect to the substantive size of the effect. An estimate of the aggregated absolute impact is also possible in broad strokes, although necessarily imperfect. For example, the average precedent in this dataset is available for citation 1028 times. Using the estimates from our model, that means a single precedent would be cited 0.82 times if always considered by a judge of all three shared identities and only 0.62 times if none were shared. If a judge writes a thousand opinions over the course of their career, the maximum discrepancy in citations would accumulate to a difference of 205 citations. This is a notable number of out-of-circuit citations that reflects considerable influence on one’s peers around the country.

Discussion & conclusion

Our analysis demonstrates that diversity in race, gender, and partisanship plays a critical role in the evolution of legal doctrine. We contribute to the rich literature on citations and the burgeoning body of work on intersectionality by revealing that a judge’s influence is closely linked to shared identity traits. In fact, our evidence indicates that as the number of shared identities increases, so does the likelihood of citation, with judges sharing all three salient traits experiencing a 25% increase in citation probability. In an environment replete with available precedent, a complex web of personal and professional relationships—shaped by identity-based affiliations—directly influences citation practices.

These findings have profound implications for which judges are cited—and which are not. The underrepresentation of conservative judges of color, for example, means that their precedents are cited less frequently, allowing their conservative white male counterparts to exercise greater influence over the development of the law. Although both Republican and Democratic administrations have sought to diversify the federal judiciary, Democratic presidents have been most successful in promoting racial and gender diversity. Our findings suggest diversifying the federal judiciary should not be left to one party alone. The current imbalance implies that there are relatively few racial/ethnic conservative judges, constraining the opportunities for reciprocal in-group citation among these judges.

Finally, our work underscores the nuanced ways in which identities permeate judicial behavior. Whereas prior research has often found judicial identities only appear to have an impact in cases directly implicating those identities (e.g., Boyd et al. 2010), our study reveals that intersectional identities also affect the subtler dimensions of legal decisionmaking (Haire et al. 2013; Means 2022), including citation practices. By examining race, gender, and partisanship simultaneously, we break new ground in demonstrating the cumulative impact of shared identities on judicial influence. Although our evidence comes from a time and institution that contained a relative dearth of intersectional diversity, and should be interpreted with caution, it establishes a robust foundation for future research. Subsequent studies might explore how these dynamics differ across various types of judicial behavior or within distinct identity configurations, recognizing that the most salient identity for a white male conservative, for example, may differ from that of a liberal black woman. Shared identities apply in all facets of life, from elite actors in legislatures to citizens. While answering such empirical questions undoubtedly presents serious data challenges, our work demonstrates the importance of continued efforts to explore the impact of intersectional identities in the judiciary and beyond.

Supplemental Material

Supplemental material - Judicial influence and the importance of intersecting identities

Supplemental material for Judicial influence and the importance of intersecting identities by Abigail A. Matthews, Rachael K. Hinkle in Research & Politics