https://orcid.org/0000-0002-3663-6356
![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
ABSTRACT
In the wake of George Floyd’s murder in May 2020, some state supreme courts decided to issue public comments to address disparities in racial justice, while others did not. We explain the decision to issue these voluntary statements as a strategic choice by state supreme court justices motivated by their retention audience and state demography. We demonstrate that state supreme courts appear responsive to their state audience when deciding to issue statements, specifically based on the number of racial justice protests by state residents and the liberalism of their state. Furthermore, state judicial retention systems and race identity in the public and on the state supreme court appear to influence word choice and sentiment in the racial justice statements issued. This extends understanding of motivations behind judges’ decisions and behavior in extra-judicial settings beyond the courtroom.
Acknowledgements
To be provided upon acceptance in order to maintain anonymity.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
3 Numerous states in the past had canons of judicial conduct that would have prevented courts from making these types of statements on political issues. With the decision in Republican Party of Minnesota v. White (2002), these judicial canons were determined to run afoul of the First Amendment. However, the tradition of following these ethical cannons continues, and even in 2020, some judges sitting in state courts were sanctioned for their participation in peaceful protests. Judicial ethics committees in many states determined that statements of the sort we analyze in this manuscript were acceptable but warned that they could be perceived as reducing the unbiased perception of judges (see: https://ncscjudicialethicsblog.org/2020/07/28/in-response-to-recent-events-judicial-participation-in-demonstrations-protests-marches-and-rallies/).
4 For a more complete account of the nature of judicial representation see Curry and Romano (Citation2022), Romano and Curry (Citation2020).
5 Not only is the origin of the BLM movement stem from the shooting of Travon Martin in 2012, but recent research also finds that BLM protests specifically reduced police performance evaluations among black and white respondents (Wright et al. Citation2023).
6 This timeframe is inclusive of all racial justice statements made by state supreme courts and chief justices in 2020, which ranged from May to July.
7 We focus on judicial retention mechanisms because those issuing racial justice statements have passed their initial selection and only face judicial retention methods for their future on the bench. We use life tenure as our reference category since judges in life tenure systems do not harbor substantial concern over backlash or reprisal as they do not possess a defined retention audience.
8 Illinois, New Mexico, and Pennsylvania select judges through competitive elections and retain through retention elections.
10 South Carolina and Virginia use legislative reappointment retention systems. For all other states, the partisanship of the governor was as a proxy for government partisanship. Fortunately, in 2020, South Carolina and Virginia enjoyed unified control of government, making the partisan determination consistent across states regardless of retention audience.
11 Oklahoma and Texas each have two courts of last resort, one of which is the state’s supreme court and the other is the court of criminal appeals. All other states have one court of last resort.
12 The National Center for State Courts hosts this clearinghouse of racial justice statements on their website here: https://www.ncsc.org/newsroom/state-court-statements-on-racial-justice. Two other websites were used to cross-verify the universe of statements, including the Self-Represented Litigation Network (https://www.srln.org/node/1442/race-justice-statements-courts-2020) and Rethink.org (https://rethinkmedia.org/resource/tracking-judicial-statements-racial-injustice?authkey=bc6fe68b2230debb2eee700e074f8fbebcb89a92f6472edfe9547c5ac8cfce9c).
13 Protests concerning policy misconduct, police brutality, and discrimination against people of color in the criminal justice system were ongoing throughout the summer 2020. We focus solely on the month of May due to the timing of judicial statements. The vast majority of statements made by the judiciary, however, occurred in early June starting with Chief Justice Beasley’s June 2 statement. We therefore presume that the initial wave of protests that occurred throughout the month of May had more of an impact on the decision to make a statement by judges than protests in the month of June. As a robustness check, however, we included number of protests in both May and June separately, and received similar results for May, with June being insignificant to the models.
14 Crowd Counting Consortium, https://sites.google.com/view/crowdcountingconsortium/home?authuser=0
15 It is more appropriate to utilize nonwhite population than the African American population because a number of statements specifically reference other nonwhite populations of prominence within the state. For example, the Alaskan Supreme Court specifically calls out the abuse of Alaskan Natives by the criminal justice system in their statement.
16 We define the South as states in the former Confederacy, along with Kentucky, Missouri, Oklahoma, and West Virginia. Robustness checks using alternative measures that define Southern states as 1) solely states in the former Confederacy or 2) states as defined by the Census did not have a substantive impact on the results.
17 The Brennan Center’s State Supreme Court Diversity — February 2020 Update can be found here: https://www.brennancenter.org/our-work/research-reports/state-supreme-court-diversity-february-2020-update
18 Given the partisan candidate selection processes, in spite of the nonpartisan general election, we run a check for robustness for Ohio and Michigan (Nelson, Caufield, and Martin Citation2013). Estimating our auxiliary analysis with both states as partisan election retention types yields similar results as our main model.
19 As a robustness check, we ran the following models also including interaction terms for public ideology and government partisanship, finding insignificant results for the interactions. Public ideology consistently remained significant in our models regardless of the inclusion of the interaction, and an ANOVA analysis of the interaction model and the models presented here were insignificant, indicating that the models presented here best represent the data. We are happy to provide the results of this robustness check upon request.
20 We tested several alternative models to check the robustness of these findings. Particularly, we tested an alternative hypothesis that southern states with greater nonwhite populations might drive such results. We find no statistical evidence that this is the case, and the dummy variable for southern states is insignificant in both the model presented and the alternative robustness models. We choose to report the simpler model without this interaction for ease and clarity of the findings.
21 These two steps – tokenizing and removal of stopwords – are common in automated textual analysis as it allows us to better focus on terms that have higher meaning and value to a document’s main themes or discussions. Stopwords include commonly used terms like “a,” “the,” “is,” which occur commonly across all speech.
22 While the bulk of judicial statements focus predominantly on the events surrounding George Floyd’s murder, some judges also chose to highlight other recent tragedies involving the death of Black civilians in the United States and chose to say their names as well.
23 Determining the optimal number of topics for judicial statements concerning the death of George Floyd and Black Lives Matter was done by analyzing the residual output from multiple iterations of the topic model and “semantic coherence,” (Mimno et al. Citation2011). The held-out likelihood predicting how well a model does at predicting words within a document (Roberts, Stewart, and Tingley Citation2019) optimizes by the fourth topic before degrading.
24 Topic 1 includes references to the COVID-19 pandemic. As this historical artifact was occurring at the same time as the George Floyd protests, administrative concerns were split between protecting employees from exposure while also recognizing how judicial staff are also important to promoting equality and racial justice.
25 The Illinois Supreme Court later hired its first CDIO, Deanie Brown, in 2020.
26 Hawaii HB1567 was introduced in the Hawaii State House of Representatives in 2022 (see bill summary: https://www.capitol.hawaii.gov/measure_indiv.aspx?billtype=HB&billnumber=1567&year=2022). HB1567 has passed both chambers of the Hawaii state legislature but in different versions. As of April 15, 2022, the differences in the bill have not been resolved. Hawaii HB SB1260 was introduced in the Hawaii State Senate in 2021 (see bill summary: http://www.capitol.hawaii.gov/Archives/measure_indiv_Archives.aspx?billtype=SB&billnumber=1260&year=2021). SB1260 did pass both chambers of the Hawaii state legislature but in different versions. There has been no activity on the bill since April 20, 2021 (as of April 15, 2022).