Diversity, consensus, and decision making: evidence from the U.S. Courts of Appeals

ABSTRACT

In this article, we identify key theoretical perspectives from the literature in social and organizational psychology on diversity and workgroups and apply these concepts to an analysis of decision making in the U.S. Courts of Appeals. Using data from twelve circuits over two decades, we leverage the diversification of the federal appellate bench to investigate the nature of the relationship between changes to a court’s gender and racial composition and levels of consensus. The results suggest that the impact of a diversified court depends on the size of the court.

KEYWORDS:

• Judicial behavior
• judicial decision making
• gender
• race
• diversity
• consensus
• dissent

Disclosure statement

No potential conflict of interest was reported by the author(s).

Notes

1 Another perspective (Dovi 2020) argues that “[j]ustice demands that one group not possess exclusive control over the judiciary.” Kirkpatrick (2020) argues that the study of descriptive representation has often overlooked the ways in which all judges (including white men) engage in descriptive representation, not simply judges from underrepresented groups. With practical application to specific policy areas, Boyd and Rutkowski (2020) examine whether women and racial minority judges are more likely to exercise discretion in favor of disability claimants’ positions, relative to white and male judges.

2 This is not to say that dissents cannot serve positive purposes. Indeed, Justice William Brennan (1985) argued that dissents hold accountable the majority in the instant case, while serving as a clarion call to future courts that may revisit and overrule the initial decision.

3 At the case level, dissent may also serve a signaling function, depending on the composition of the panel and the dynamics of the judicial hierarchy (Beim and Kastellec 2014). The presence of a dissenting opinion may be best thought of as the tip of the iceberg, disagreement that rises to the level where incurring additional work is worth the cost, though other disagreements may lie beneath the surface (Bowie, Songer, and Szmer 2014). Indeed, reduced efficiency in disposing of cases is also an indicator of group conflict (Christensen and Szmer 2012).

4 This conceptualization is often referred to as “social category diversity.” As Kang et al. (2020) note, social categories also operate in an intersectional way, because of the multiple and overlapping ways that privilege and disadvantage are produced. Similarly, indicators of task diversity like differences in professional backgrounds and legal training may affect the ability of panels to reach consensus. Previous work has noted that the career trajectories of white male judges have often been different from women and minorities, who tend to have more public sector and judicial experience (Haire and Moyer 2015; Slotnick 1983–1984). Thus, task diversity and social category diversity may sometimes work in tandem.

5 Information on judges comes from the Federal Judicial Center. The calculations are restricted to judges on active service in the U.S. Courts of Appeals, excluding the Federal Circuit.

6 Westlaw searches for dissents only include sitting judges from the circuit. While visiting judges certainly do file dissents, this occurs at levels less than the already-low dissent rates of sitting judges (Epstein, Landes, and Posner 2013).

7 Lindquist (2007) uses a slightly different measure of workload but finds that it is not significantly related to circuit dissent rates. We also test whether docket composition affects consensus, following other work (Posner 1996; Lindquist 2007). In supplemental models (not shown), the inclusion of a variety of different measures of the percentage of criminal, administrative, and bankruptcy cases on the docket did not affect the statistical or substantive results for the main independent variables.

8 We interpret our substantive findings consistent with Cameron and Trivedi’s (2010, 88) guidelines for log-linear models. We also estimated the untransformed dissent rates using beta regression. The results were consistent with those presented here, except that the coefficient for the number of judgeships was only significant at the 0.07 level.

9 With the exception of the First Circuit, the Fifth Circuit, the Sixth Circuit, and the Ninth Circuit, the remainder of the circuits had between 11 and 15 seats during the 1997–2016 period.

10 The responsibilities and workload of senior judges differ substantially from active judges, as dictated by federal law and court policy. By taking senior status, judges gain more control over the quality and quantity of their workload. Unlike active judges, they may decide which months of the year they will decide cases (Yoon 2005) and, according to one senior judge, “can decide that they no longer wish to preside over certain types of cases” (Block 2006, 540). Senior judges also have different obligations about participation in court governance activities (Yoon 2005; Hooper, Miletich, and Levy 2011). Second, senior judges as a group are not uniform in terms of their judicial participation. There is wide variation in merits terminations by senior judges both within circuits and across circuits (Benesh 2006; Hooper, Miletich, and Levy 2011). Whether and how senior judges participate in the en banc process also can vary quite a bit by circuit, as well as over-time within circuits (Hooper, Miletich, and Levy 2011).

Additional information

Funding

Data for this project were collected with support from the National Science Foundation (NSF-SES #1655159, 1654614, 1654559, 1654697). Any opinions, findings, and conclusions or recommendations expressed are those of the authors and do not necessarily reflect the views of the National Science Foundation.