![Sample our Law journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5942/TOC-Subject-Sample-2021-Law.jpg"})
Abstract
Backlogs and long delays in the confirmation of federal judicial nominees have become a common occurrence in Washington, leaving many federal benches understaffed and overworked. While this phenomenon has been well studied at the federal level examining political and institutional factors at play beyond senatorial courtesy, little is known if delay occurs in court systems of the individual American states, or how extensive the obstruction and the delay should it occur. We examine judicial appointments in states having the selection system of gubernatorial nomination and legislative confirmation. We focus on the duration of the judicial vacancies on state intermediate appellate courts and courts of last resort from 2000 to 2016 and explain the variation over time in the length of this process. We show that partisan division and increases in the polarization of state political parties work together to influence timely staffing of these critical institutions in the American states. Ultimately, these findings offer insight into the issues caused by political gridlock over judicial selection and offer practical institutional remedies to this problem.
Acknowledgements
The authors would like to thank Greg Goelzhause and Laine Shay for their generosity in sharing their data. The authors would also like to thank the anonymous reviewers for their valuable comments and suggestions.
Notes
1 While some attribute this phenomenon beginning with Bork, there is evidence that trends in delay and confirmation failure began before Bork’s nomination (Binder and Maltzman Citation2002).
2 State courts vary in their level of professionalism which can influence their capacity to generate and evaluate information. Further, they vary in their overall importance as some courts comparatively have greater responsibilities and may or may not have the capability such as docket control to fulfill them. As such, certain courts might experience more or less delay given their decision-making responsibilities, as a lengthy vacancy could be more or less costly for both court operations and the political actors involved in this process who could be held accountable for any delay. We include Squire and Butcher’s (2021) updated measure of court professionalism as a rough measure for the variation in these courts. The results from this supplemental analysis are found in Table SA1 of the Online Supplemental Appendix, yielding no substantive impact on the main results discussed below, nor attaining a level of statistical significance.
3 For purposes of the analysis below, “appointments” are gubernatorial appointments and does not consider legislative selection (e.g. South Carolina and Virginia which “elects” its judges via a vote within the legislature).
4 Pennsylvania, Texas, and Montana allow their governors to appoint replacements for vacancies which occur with the approval of the state senate. In a great many cases, judicial seats that are up for election are vacated by sitting judges shortly prior to the end of their terms of office and filled by judges that are appointed by the sitting governor (Glick and Vines Citation1973) who then run for reelection during the next election. The majority of the judges serving in partisan election states gained their position through an interim appointment, and often these interim appointments become permanent due to the extremely high retention rate judicial incumbents enjoy (Rottman and Strickland Citation2006). Additional robust checks controlling for these interim appointments revealed no substantive impact on the main results presented, merely attenuating the levels of significance of the estimates, and are found in Table SA2 in the Online Supplemental Appendix.
5 Vacancies which occur unexpectedly (due to a death/illness, appointment to another position, etc.) may result in a more difficult nomination and confirmation process as the involved actors are forced to begin these procedures often to comply with statutory constraints, maintain the operations of the court in the midst of important judicial proceedings, or to contend with political conditions (changing legislative majorities/minorities, upcoming elections, etc.). Additional robust checks controlling for these unexpected vacancies revealed no substantive impact on the main results presented and are found in Table SA3 in the Online Supplemental Appendix.
6 Since 2000, concerns over political credentials have surpassed professional credentials in consideration of judicial candidates (Kritzer Citation2020).
7 This relationship between the partisan makeup of the involved actors would suggest that the major parties should be concerned with variability in partisan control as potential changes in their status might increase the importance of a nomination for a party which could soon be in the political minority. Judges themselves have been observed to be strategic in their retirements, coinciding leaving the bench during favorable partisan control to maximize the chances that an ideologically compatible successor will be appointed (e.g., Hall Citation2001). Thus, nominations in states which are highly competitive and more prone to shifts in partisan control could be more contentious. As such, we conduct supplemental analysis to consider the level of partisan competition within a state by including a measure of the folded Ranney index, generously provided by Laine Shay (Citation2020), in which higher scores indicate increasing party competition within a state. The results from this analysis are found in Table SA4 in the Online Supplemental Appendix and reveal that while increased competition does lengthen this process, the findings are tenuous either failing to reach statistical significance in the unconditional models or being only marginally significant (1-tailed) in the conditional models discussed below.
8 Theories of Conditional Party Government suggest that when the majority party is ideologically homogeneous and ideologically distinct from the minority party, more power tends to be delegated to the leadership (Aldrich and Rohde 2001). As such, we investigate the degree to which variation in party leadership might influence the length of judicial vacancies, through the inclusion of Mooney’s (Citation2013) additive index of Speaker power, which was updated and generously provided by Laine Shay (Citation2020). Results for this supplemental analysis are found in Table SA5 in the Online Supplemental Appendix and revealed no significant impact nor any substantive influence on the main results.
9 In Pennsylvania, the governor selects a nominee from list compiled by the General Counsel, who is a member of the governor’s cabinet.
10 The year 2016 serves as the end date for data inclusion given the availability of the Birkhead scores, discussed below, derived from the Shor-McCarty polarization data.
11 In 2013, Kansas changed its method of judicial selection when the legislature voted to replace merit selection for the Court of Appeals with a system of gubernatorial appointment and senate confirmation.
12 The Rhode Island Superior Court hears appeals from trials held in District Courts, though these result in a trial de novo. Yet, the Superior Court has appellate jurisdiction of decisions of local municipal and probate courts. Further, appeals from various boards and commissions, such as a zoning board and the state Ethics Commission, as well as petitions to review disciplinary action imposed upon a state or local police officer by the chief of that department, are heard in Superior Court. Given this appellate jurisdiction, the Superior Court is included to avoid omitting important data.
13 Rhode Island utilizes a slightly different confirmation procedure for members of its Supreme Court whereby both houses of the General Assembly confirm a judicial nominee (N = 6). However, nominees for the Rhode Island Superior Court are confirmed by the traditional procedure addressed above (N = 16). Further, in 2014 Tennessee amended its Constitution empowering the governor to appoint judges to its Supreme and Appellate Courts with confirmation by the joint houses of the General Assembly (N = 2). Given that analysis omitting Rhode Island Supreme Court appointments and Tennessee appointments does not substantively alter the main findings, these cases are thus included to avoid omitting potentially important data.
14 Vacancies to replace New Jersey Supreme Court Justices John Wallace (2167 days) and Virginia Long (840 days) present potentially significant outliers given their length. Supplemental analysis (Table SA6 in the Online Supplemental Appendix) omits these cases to determine if their vacancy length would bias the analysis. The results from this analysis reveal that their inclusion imparts little substantive impact to those results presented, mostly yielding some attenuation of conventional levels of statistical significance for ancillary variables. Thus, to avoid omitting useful data, the analysis below includes these cases. We further consider the impact of pooling within states where prior norms concerning filling vacancies might impact the duration of future vacancies by estimating a shared frailty model (Table SA7 in the Online Supplemental Appendix). This analysis reveals only minor substantive impact. Given that the group size however varies, with some states only having a few observations present in the dataset (Kansas N = 1, Montana N = 2, and Tennessee N = 2), some caution should be taken in interpreting differences between these results and those from the main models presented below.
15 Additional robustness checks controlling for appointments informally made before a vacancy occurs revealed no substantive impact on the main results presented.
16 The legislatures in both Montana and Texas operate under biennial sessions. Additional robust checks controlling for biennial sessions revealed no substantive impact on the main results presented.
17 To consider the importance of considering partisanship within measures of polarization, supplemental analysis substitutes another measure of polarization (the distance between the Republican and Democratic party medians, averaged across the two chambers), the results of which are presented in Table SA8 in the Online Supplemental Appendix. Although similar results are found, these are estimated with much less precision. Clearly, consideration of the various other characteristics of partisanship is necessary to understand the impact of polarization. If the political parties are overall are far apart but the individual legislatures within the party have considerable spread among them, there can be an incentive for these individual members to go it alone rather than defer to party leadership, impacting judicial delay.
18 For most states the chamber responsible for confirming the nominee is the state’s upper chamber, thus Divided Government is a simple indicator if the party of the Governor and the party in control of the state’s upper chamber are different, whereas % Legislature Shared Gov Party is the percentage of the seats in the state’s upper chamber held by the members of the Governor’s party. However, for Rhode Island Supreme Court nominees and nominees to Tennessee’s intermediate appellate and court of last resort after 2014, both legislative chambers are responsible for confirmation a nominee, thus for these cases the aforementioned variables are evaluated across both legislative chambers.
19 These measures are negatively correlated with each other at 0.83, and each correlates with the measure of polarization at -0.31 (Divided Government) and 0.39 (% Legislature Shared Gov Party).
20 Given the importance of nominations to these positions, it could be expected that politicians would be concerned with controlling outcome of this process (increasing the duration of a vacancy), ensuring that a vacancy does not stretch on leaving either the high court shorthanded or the court without its leadership, or some combination of the aforementioned. As such, no hypothesized relationships are posited for these variables. Additional robust checks controlling for state high courts compared to lower appellate courts revealed no substantive impact on the main results presented. Results are found in Tables SA9 & SA10 in the Online Supplemental Appendix.
21 Substitution of Legislative Election years yields very little substantive change in the main results. See Table SA11 in the Online Supplemental Appendix.
22 Within the sample Female and Person of Color nominees were present in almost every state, though several states had larger numbers of nominees: New York (12.2% Female/13.3% Person of Color), Maryland (18.4% Female/26.7% Person of Color) and Texas (20.4% Female/23.3% Person of Color). To examine if pooling effects were present, supplemental analysis omits these states and is presented in Table SA12 in the Online Supplemental Appendix.
23 Using Cox-Snell Residuals (Binder and Maltzman Citation2004; Box-Steffensmeier and Jones Citation2004), we assessed the adequacy of each model. Further, we tested the proportional hazards assumption implicit in a Cox model (Box-Steffensmeier and Jones Citation2004, Binder and Maltzman Citation2004).
24 For a comparison of parametric and proportional hazard rate models, see Box-Steffensmeier and Jones (Citation1997).
25 We utilize the Cox ED method developed by Jonathan Kropko and Jeffrey Harden to calculate duration-based quantities from Cox model results, such as the expected duration (or survival time) given covariate values and marginal changes in duration for a specified change in a covariate. See Kropko and Harden (Citation2020) for more information detailing this procedure.
26 While inconsistent with theoretical expectations, these results are supported by the literature regarding increased institutional capacity which allows political actors to more effectively perform their respective functions (Mooney Citation1995; Huber and Shipan Citation2002), whereby the legislature can more efficiently vet and confirm nominees to the state courts. Given these finding, we further examine the role of legislative professionalism as a proxy for the level of patience a legislature exhibits in their ability to contend with a governor. Consistent with Kousser and Phillips (Citation2012), patience is operationalized as the number of days a legislature is in session and is taken from Bowen and Greene (Citation2014). The results from this analysis are found in Table SA13 in the Online Supplemental Appendix. The inclusion of this measure, which reduces the number of observations due to missing data (N = 165 to N = 135), reveals results for professionalism which though appropriately signed are estimated much less precisely.
27 As noted by Shor and McCarty (Citation2017), polarization at the state level is highly varied, with some states since 1996 polarizing rapidly, some remaining relatively stable, and one (Kansas) even depolarizing.
28 As discussed above, there remains a long-running academic debate over the merits of selecting judges via appointment or election (e.g. Ducat and Flango Citation1979; Bonneau and Hall Citation2009). Given wider fears of the politicization of the courts and the potential consequences for lengthy vacancies, the findings here would suggest support for the arguments advanced by proponents of judicial elections. As Olszewski (Citation2004) argues, under both appointment and elections politics inherently plays a role, but at least under elections the political influence originates with the people. Non-partisan and partisan elections could inherently be more efficient selection and replacement mechanisms because of their inherent pro-democracy foundations.