https://orcid.org/0000-0002-3875-5519
![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
This article analyzes public attitudes toward replacing lifetime tenure with term limits for federal judges, including U.S. Supreme Court justices. We employ novel data that we collected from a nationwide survey experiment. We find that although partisans are less supportive of proposals from their opponents, the magnitude of this effect is much smaller than one might expect in today’s polarized environment. We also find that a respondent’s support for term limits is a function of his or her subjective ideological agreement with the Supreme Court. Finally, we demonstrate that although support for term limits is generally high, only a modest subset of reform supporters believe that term limits should be a top political priority. These supporters also tend to exhibit weaker levels of support for the rule of law more generally. Taken together, the results contribute to our understanding of an issue of significant importance.
Acknowledgments
An earlier iteration of this project was presented at the 2022 Annual Meeting of the American Political Science Association. We thank Logan Strother for the comments provided there. Miles Armaly also provided helpful feedback on an earlier draft. We thank Matt Cota, Rusty Johnson, Jonathan King, Elizabeth Lane, Amanda Savage, and Jessica Schoenherr for being guinea pigs and helping us stress test the survey experiment itself. Last but not least, at New Political Science, we thank Courtenay Daum and the anonymous reviewers for their thoughtful feedback and suggestions.
Disclosure statement
No potential conflict of interest was reported by the authors.
Notes
1 William Blackstone, The Oxford Edition of Blackstone’s Commentaries on the Laws of England, 1st ed., ed. Ruth Paley (Oxford University Press, 2016), 1: 259–60.
2 Bruce Peabody, ed. “The Politics of Judicial Independence,” In Chap. Introduction (The Johns Hopkins University Press, 2011), 1.
3 Alexander Hamilton, “Federalist Number 78,” 1788.
4 Brutus, “Anti-Federalist Papers: Brutus 15,” 1788, http://www.constitution.org/afp/brutus15.htm.
5 Jeff Shesol, Supreme Court: Franklin Roosevelt v. the Supreme Court (W. W. Norton & Company, 2010); Robert Scigliano, The Supreme Court and the Presidency (Free Press, 1971).
7 Stephen G. Calabresi and James Lindgren, “Term Limits for the Supreme Court: Life Tenure Reconsidered,” Harvard Journal of Law and Public Policy 29 (2006): 770–877.
12 Roger C. Cramton and Paul D. Carrington, Reforming the Court Reforming the Court: Term Limits for Supreme Court Justices (Carolina Academic Press, 2006).
13 Kevin T. McGuire, “Are the Justices Serving Too Long? An Assessment of Tenure on the U.S. Supreme Court,” Judicature 89, no. 1 (July 2005): 8–15.
14 Calabresi and Lindgren, “Term Limits for the Supreme Court”; Justin Crowe and Christopher F. Karpowitz, “Where Have You Gone, Sherman Minton? The Decline of the Short-Term Supreme Court Justice,” Perspectives on Politics 5, no. 3 (September 2007): 425–45.
15 Calabresi and Lindgren, “Term Limits for the Supreme Court.”
16 Hemant Sharma and Colin Glennon, “A Case for Supreme Court Term Limits? The Changing Ideological Relationship between Appointing Presidents and Supreme Court Justices,” Politics and Policy 41, no. 2 (2013): 267–97.
17 Joshua C. Teitelbaum, “Age and Tenure of the Justices and Productivity of the U.S. Supreme Court: Are Term Limits Necessary?” Florida State University Law Review 34 (2006): 161–82.
18 Brandon L. Bartels and Christopher D. Johnston, Curbing the Court: Why the Public Constrains Judicial Independence (Cambridge University Press, 2020).
19 Joseph Daniel Ura and Patrick C. Wohlfarth, “Greater Public Confidence in the U.S. Supreme Court Predicts More Jurisdiction Stripping,” Political Science Research and Methods 10, no. 4 (October 2022): 831–9.
20 Bartels and Johnston, Curbing the Court.
21 Ibid.
22 We pre-registered the underlying survey experiment from which these data are drawn with the “AsPredicted” platform at the University of Pennsylvania’s Wharton Credibility Lab. See Mandatory Retirement Age and Term Limits, registration #105578 (https://aspredicted.org/ck5si.pdf). Results from that planned analysis are reported in Ryan C. Black, Ryan J. Owens, and Patrick C. Wohlfarth, “The Public’s Support for Judicial Reforms: An Experiment” (paper presented at the 2022 annual meeting of the American Political Science Association in Montreal, Quebec, Canada. September 2022). What follows below constitutes a secondary exploratory analysis of these data. That is, these analyses were not pre-registered.
23 Note that this experimental manipulation took place after we asked a number of other questions that provide the basis for control variables we include in our statistical model described below. We do this so as to avoid contaminating those answers with potential post-treatment bias. See, Jacob M. Montgomery, Brendan Nyhan, and Michelle Torres, “How Conditioning on Posttreatment Variables Can Ruin Your Experiment and What to Do about It,” American Journal of Political Science 62, no. 3 (July 2018): 760–75. There was an additional level of random assignment that took place prior to being assigned to a partisan frame. In particular, respondents were randomly assigned to either the term limits proposal, which we discuss in this paper, or another proposal to establish a mandatory retirement age for all federal judges. Because whether a respondent was assigned to term limits or retirement age was random, we are able to treat the two samples as independent. We focus here on the term limit proposal as it is the only reform to be formally proposed, as of this writing, in Congress (multiple times, in fact).
24 Dennis Chong and James N. Druckman, “Framing Theory,” Annual Review of Political Science 10 (2007): 103–26.
25 We excluded those who expressed opposition to term limits because it did not make substantive sense to seek their response to this particular question. Future studies could consider crafting a question that is adaptable to both supporters and opponents but that would likely introduce other dimensions that would need to be manipulated experimentally or otherwise held constant.
26 See https://lucidtheorem.com/.
27 See https://lucidtheorem.com/faq. We dropped respondents who did not correctly answer two attention check questions included in our survey.
28 Alexander Coppock and Oliver A. McClellan, “Validating the Demographic, Political, Psychological, and Experimental Results Obtained from a New Source of Online Survey Respondents,” Research and Politics 6, no. 1 (2019): 1–14.
29 Albert H. Fang and Gregory A. Huber, “Perceptions of Deservingness and the Politicization of Social Insurance: Evidence from Disability Insurance in the United States,” American Politics Research 48 (2019): 543–59; Dieter Stiers et al., “Candidate Authenticity: ‘To Thine Own Self Be True’,” Political Behavior (2019); Lilla V. Orr and Gregory A. Huber, “The Policy Basis of Measured Partisan Animosity in the United States,” American Journal of Political Science 64 (2019): 569–86.
30 We also include an overall estimate that comes from FiveThirtyEight, which does not field its own sample but rather produces an estimate based on a proprietary aggregation and weighting procedure it uses on the results of existing surveys.
31 AP-NORC Center for Public Affairs Research, “Americans Have Lost Confidence in the Supreme Court: Most Americans Favor Constitutional Reforms to the Supreme Court to Impose Term and Age Limits for Justices,” July 2022, https://apnorc.org/ (accessed November 13, 2022).
32 Our measure of partisanship comes from demographic data collected by Lucid Theorem, who allowed respondents to identify as one of ten values. Eight of those values include a party identity: Strong Democrat, Not Very Strong Democrat, Independent Democrat, and Other–Leaning Democrat (and the same for Republicans). In what follows, we code all of four of these values as being Democrat (and their Republican analogs as Republican). Independents are individuals who answered one of the two other values: Independent–Neither or Other–Neither.
33 This, too, is consistent with the aforementioned AP/NORC survey, which suggested favorability levels of 82%, 51%, and 57% for Democrats, Independents, and Republicans, respectively.
34 Future research might examine whether asking respondents about term limits for Supreme Court justices as opposed to asking them, as we did here, about their support for term limits on federal judges. It could be that when people hear “federal judges” they do not necessarily think of Supreme Court justices.
35 The chi-square statistic for the test of statistical independence among proposer partisanship and a three-level variable of support for term limits among Independents is 3.06 with 4 degrees of freedom, which has a corresponding p-value of just 0.55. This means we cannot reject the null hypothesis that the two variables are statistically independent.
36 Brandon L. Bartels and Christopher D. Johnston, “On the Ideological Foundations of Supreme Court Legitimacy in the American Public,” American Journal of Political Science 57, no. 1 (January 2013): 222.
37 Dino P. Christenson and David M. Glick, “Chief Justice Roberts’s Health Care Decision Disrobed: The Microfoundations of the Supreme Court’s Legitimacy,” American Journal of Political Science 59, no. 2 (April 2015): 403–18.
38 Ibid., 410.
39 Alex Badas, “Policy Disagreement and Judicial Legitimacy: Evidence from the 1937 Court-Packing Plan,” Journal of Legal Studies 48, no. 2 (2019): 377–408.
40 James L. Gibson and Michael J. Nelson, “Is the U.S. Supreme Court’s Legitimacy Grounded in Performance Satisfaction and Ideology?” American Journal of Political Science 59, no. 1 (2015): 162–74.
41 James L. Gibson, “The Legitimacy of the U.S. Supreme Court in a Polarized Polity,” Journal of Empirical Legal Studies 4, no. 3 (November 2007): 529.
42 The graded response model is the equivalent to a two-parameter item response model that has been extended for the purpose of working with ordinal variables like ours. It is preferable to the more traditional additive index because it allows for variation among items and responses within an item. We estimated these values using the irt grm command in Stata 16.1. The resulting scores range from −2.56 to 1.25 with a mean and standard deviation of 0.01 and 0.85, respectively.
43 Alex Badas, “The Applied Legitimacy Index: A New Approach to Measuring Judicial Legitimacy,” Social Science Quarterly 100, no. 5 (2019): 1848–61.
44 The Knowledge Score variable ranges between −1.13 and 0.68 with a mean and standard deviation of 0.00 and 0.67, respectively. The average difference in Knowledge Score between respondents who answered an item incorrectly versus correctly is −1.01 (tenure length), −1.26 (selection process), and −0.16 (average age). All of these differences are statistically significant (p < 0.05, two-tailed test), which suggests that even if they are providing different levels of information about a respondent’s knowledge, they are still tapping into the same latent quantity of interest. Admittedly, the question about judge age is more tenuous than the others in terms of substantive knowledge of courts. The item response model we employ, however, alleviates this concern to the extent that it simply devotes (much) less weight to this answer than to the other two answers.
45 Lucid Theorem separately asks respondents whether they identify as being Hispanic. Such individuals were not automatically coded as non-white, but could have been coded as such if they so identified to the race question. As an empirical matter, about 64% of self-identified Hispanics did so (i.e., 37 out of 58).
46 The results do not change if we instead treat them as categorical predictors with a dichotomous indicator for each category (minus a baseline).
47 Gibson and Nelson, “Is the U.S. Supreme Court’s Legitimacy Grounded.”
48 Ibid., 163.
49 Ibid.
50 Individuals with the median score for rule of law attitudes (−0.1) strongly support term limits with a probability of 0.24.
51 Gibson and Nelson, “Is the U.S. Supreme Court’s Legitimacy Grounded,” 167.
52 Our specific parameterization treats the concept of support for term limits as an interval-level variable and also includes its squared value. This approach can be seen as something of a “statistical compromise” between treating the concept as being purely interval and estimating separate coefficients for each level of the variable. We use our approach as measures of model fit like the Bayesian Information Criterion (BIC) suggest that it best fits the data versus both the simpler and more complicated alternatives.
53 Individuals with the median score for rule of law attitudes (0.00) assigned low importance with a probability of 0.42 and high importance is 0.31. Also, the data uncover little variation in the probability of assigning medium importance as a function of support for the rule of law.