Succession, Opportunism, and Rebellion on State Supreme Courts: Decisions to Run for Chief Justice

Abstract

We examine decisions to seek promotion on state courts of last resort, focusing on the conditions when an associate justice will run for the position of chief justice. We analyze data including all chief justice elections from 1990 to 2014 in the states that elect this position. We construct a taxonomy of associate justices who seek the chief justice position, then use regression analysis and postestimation techniques to better understand these choices. Our findings indicate that judicial actors who seek electoral promotion are strategic and motivated by ideological preferences rather than institutional features or raw ambition.

Keywords:

• State supreme courts
• chief justice
• judicial elections

Notes

1 See Clark (2005) for more information about Roy Moore’s removal from the Alabama Supreme Court and the legal battles about the Ten Commandments monument he installed in the state judicial building.

2 Nabers defeated Parker by a 61.4–38.6 percent margin in the Republican primary. Parker remained on the Supreme Court of Alabama as an associate justice.

3 The other methods of chief justice selection include chosen from the sitting members on the court by the court (18 states); gubernatorial or legislative appointment (15 states); independent judicial commission (one state); and random or rotation (nine states). See Langer et al. (2003) for further discussion of chief justice selection systems.

4 The number of seats on state courts of last resort ranges from 5 to 9.

5 Scholars have also examined discrete ambition at state high courts, primarily with the goals of explaining judicial turnover and accountability. Hall (2001) argued that judges’ exits are strategic, and they step down when they believe they are vulnerable to electoral defeat. Curry and Hurwitz (2016) further analyzed the relationship between judicial selection systems and the risk of state high court departures. They found that justices in partisan systems were most likely to quit relative to judges chosen by other selection methods. They explain this dynamic is due to partisan selection systems’ tendencies to result in more contentious campaigns and draw more qualified challengers. More recently, Hughes (2019) studied the role of economic incentives in state high court judges’ decisions to step down. He found that their exits are driven primarily by their pension vesting status, work-related factors (including salary), and age and benefits rather than electoral vulnerability.

6 See Nixon (2003) for an example of this behavior by the Chief Justice of the United States in his capacity as Chairman of the Judicial Conference.

7 Similar to research on the emergence of legislative and executive candidates (Abramson, Aldrich, and Rohde 1987; Rohde 1979), we restrict our analysis to states’ associate justices. Obviously, the pool of potential candidates for an office does not come solely from one type of lower office. While restricting analysis to associate justices limits our understanding of who runs for the office of chief justice, it allows examination of how likely candidates decide whether to emerge as candidates.

8 Notably, no state in our data applies a “resign-to-run law” to high court judges. Texas does apply this requirement to many local judges. The other four states with these laws are Arizona, Florida, Georgia, and Hawaii (see http://ballotpedia.org/Resign-to-run_law).

9 Other institutional characteristics that we considered (but found no significant support for) were the size and workload of the state supreme court. The size of the state’s high court determines the number of associate justices who occupy the pool of highly qualified candidates for the chief justice. We also considered the level of electoral competition in state elections overall but found no support that it affects the likelihood of emergence.

10 We also considered that an associate justice who will reach pension eligibility during the next term might consider the salary increase of the chief justice position and decide to run. The higher salary will ultimately lead to a higher pension, as they are tied to the salary earned during the final year(s) of service. Importantly, justices at any stage of their career may consider or be motivated by the salary differential between an associate and chief, which can range from $1,000–$15,000 in the states we examine. We included a measure of the chief–associate salary differential to control for this possibility, as well as a control for pension eligibility. Neither variable achieved statistical significance.

11 We also investigated alternative modeling techniques to estimate our results. Penalized maximum likelihood estimation did not permit us to use clustered standard errors but reproduced the statistically significant effects of open seats and ideological distance identified by our logit model (but not their interaction). A multilevel logistic regression model with states nested within election years produced similar coefficients to the logit model while increasing our standard errors. Open seats and ideological distance remain statistically significant. Note, however, that the effects of interaction terms in a logistical regression model must be interpreted relative to the other term in the interaction (Norton, Wang, and Ai 2004). A likelihood-ratio test comparing the output of the multilevel model to our logit model confirms that the former does not improve the model fit or substantively change the results with chi² (0) = 0.00.