Numerous theoretical and historical analyses of the “political construction” of judicial independence in the United States and elsewhere have proposed a connection between public support for courts and both legislative decisions to enhance or undermine judicial power and courts’ propensity to decide cases contrary to legislators’ preferences (e.g., Graber, Reference Graber2005; Friedman, Reference Friedman2009). In contrast, studies of jurisdiction stripping have found statutory limitations on American federal courts’ ability to decide cases are principally related to judicial workloads rather than political factors, such as ideological differences between courts and Congress (e.g., Chutkow, Reference Chutkow2008). Here, we bridge these studies and test the hypothesis of a negative relationship between public support for the US Supreme Court and jurisdiction-stripping legislation. Contrary to prior studies, though, we find a robust, positive relationship between Americans’ confidence in the Supreme Court and jurisdiction stripping.
In this research note, we briefly introduce prior research on public opinion and judicial independence and on jurisdiction stripping, explain the data and methods we use to evaluate the association between public opinion and jurisdiction stripping, and report and discuss these unexpected results. We conclude by speculating about the origins of the political processes that generate more jurisdiction-stripping laws as positive public evaluations of the Supreme Court increase. To be clear, these results were contrary to the expectations we derived from prior theoretical and empirical studies of public opinion and judicial independence, and we do not test a new positive theory here. Our work here is descriptive, presenting a robust and substantively important empirical anomaly that may spark further study.
1. Public opinion and judicial independence
Although legislatures often have the authority to limit courts’ independence, public opinion can constrain election-minded legislators to accept or expand judicial power. When support for courts is sufficiently high, legislators—who would otherwise prefer to make policy without the constraint of judicial review—are induced “to respect judicial decisions as well as the institutional integrity of a court…[by] fear of…a public backlash” against court curbing (Vanberg, Reference Vanberg2001, 347). The theory of judicial independence rooted in public opinion—what (Ura and Wohlfarth, Reference Ura and Wohlfarth2010, 942) call the “public support hypothesis”—is substantiated by a growing body of empirical research showing a robust, positive association between public support for the US Supreme Court and judicial independence.
Americans’ confidence in the Supreme Court predicts Congress's propensity to acquiesce to judicial decisions striking down federal laws (Nelson and Uribe, Reference Nelson and Uribe2017) as well as Congress's allocation of discretion and resources to the Supreme Court (Ura and Wohlfarth, Reference Ura and Wohlfarth2010). Positive views of the Court also predict Supreme Court decisions invalidating federal laws (Clark, Reference Clark2009, Reference Clark2011; Merrill et al., Reference Merrill, Conway and Ura2017). Similar connections between public support for courts and elements of judicial independence are evident in other developed and developing democracies (Gibson et al., Reference Gibson, Caldeira and Baird1998; Vanberg, Reference Vanberg2005; Helmke, Reference Helmke2010).
2. Jurisdiction stripping
At the same time, another stream of research investigates Congress's decisions to limit federal courts’ jurisdiction. Congressional control of federal courts’ jurisdiction is an important element of the American checks and balances system. Article III, Section 2 of the US Constitution grants Congress authority to create “regulations” and “exceptions” to the Supreme Court's appellate jurisdiction. Chutkow (Reference Chutkow2008) shows that Congress makes frequent use of this power. Between 1943 and 2004, “Congress passed 248 public laws containing 378 provisions expressly stripping jurisdiction from the federal courts” (Chutkow, Reference Chutkow2008). Yet, analyses of the frequency of jurisdiction-stripping legislation show that statutory limits on federal courts’ ability to decide cases are associated with administrative concerns (i.e., judicial workloads) rather than ideological disagreement between Congress and the Supreme Court (Chutkow, Reference Chutkow2008; Heise, Reference Heise2011; Greenfest, Reference Greenfest2013).
The literature evaluating the public support hypothesis suggests the model of jurisdiction stripping emerging from these studies is incomplete, though. In particular, the model takes no account of public support for the Supreme Court as a rival explanation for variance in the frequency of Congress's decisions to limit federal courts’ jurisdiction. The public support hypothesis anticipates a negative relationship between Americans’ confidence in the Supreme Court and laws limiting federal courts’ jurisdiction; greater public support should be associated with fewer limits on courts’ deciding cases. In contrast, studies of jurisdiction stripping would anticipate no significant relationship between public opinion and courts’ jurisdiction; Congress limits courts’ jurisdiction mostly for administrative reasons.
3. Does public opinion influence jurisdiction stripping?
To examine the linkages between jurisdiction stripping and public confidence in the Supreme Court and Congress, we analyze the frequency with which Congress limited the jurisdiction of a federal court from 1973–2014.Footnote 1 The unit of analysis is the calendar year, and the dependent variable is the annual number of public laws passed by Congress that contain a jurisdiction-stripping provision. Chutkow (Reference Chutkow2008) identifies every public law with explicit language that stripped jurisdiction from the federal courts from 1943 through 2004. We extend the data through 2014, applying Chutkow's (Reference Chutkow2008) search criteria in Westlaw. That is, we identified every public law (during each session of Congress) that contained at least one of the terms “court,” “judicial,” “review,” “jurisdiction,” or “conclusive,” and then reviewed each of those laws to identify provisions that stripped a court's jurisdiction.
The authorization of the Medicare Advisory Board in the Patient Protection and Affordable Care Act is an example of legislation with a jurisdiction-stripping provision. The enacting Congress designed the Board to monitor Medicare growth and propose changes to Medicare spending. While Congress specified the manner in which it may review a “proposal” from the Board, it empowered the Secretary of Health and Human Services to implement a proposal should it fail to act. Yet, Congress also exempted proposals’ implementation from judicial review. Specifically, §3403(e)(5) of the Act states:
(5) LIMITATION ON REVIEW. There shall be no administrative or judicial review under section 1869, section 1878, or otherwise of the implementation by the Secretary under this subsection of the recommendations contained in a proposal.
Jurisdiction-stripping laws reveal a different aspect of congressional interactions with the federal judiciary than court curbing bill introductions, another measure of Congress's inclination to exert political control of courts (e.g., Clark, Reference Clark2009).Footnote 2 Figure 1 displays the frequency of jurisdiction stripping over time. Panel (a) shows the annual percentage of public laws with jurisdiction-stripping provisions. The presence of jurisdiction stripping has varied significantly from 1973 to 2014, ranging from 0.41 to 6.62 percent of public laws.Footnote 3 Panel (b) of Figure 1 compares the annual number of jurisdiction-stripping laws from 1973 to 2014 to the number of court curbing bills proposed by members of Congress (see e.g., Clark, Reference Clark2009; Mark and Zilis, Reference Mark and Zilis2018, Reference Mark and Zilis2019). Temporal variation in the annual frequency of jurisdiction-stripping laws is negatively related to the introduction of court curbing bills (r = −0.53).
Congressional jurisdiction stripping laws and court curbing bills, 1973–2014.
Note: Panel (a) displays the annual percentage of public laws with a jurisdiction stripping provision; Panel (b) compares the annual number of jurisdiction stripping laws against the number of court curbing bills proposed in Congress.
To evaluate our expectations about public opinion's effects for jurisdiction stripping, we estimate a model of the number of jurisdiction-stripping laws passed each year controlling for the total number of public laws.Footnote 4 The first two substantive predictors of jurisdiction stripping reflect the degree of public support for the Supreme Court and Congress. We measure these concepts using the GSS data on respondents’ confidence in each institution. The GSS asks: “I am going to name some institutions in this country. As far as the people running these institutions are concerned, would you say you have a great deal of confidence, only some confidence, or hardly any confidence at all in them? (Congress/US Supreme Court).” This item is useful because it has been administered since 1973.Footnote 5 Also, studies of public perceptions of Congress and the Supreme Court have often relied on the GSS data (e.g., Clark, Reference Clark2009; Ura and Wohlfarth, Reference Ura and Wohlfarth2010). We measure Court Confidence and Congress Confidence with the proportion of respondents each year saying they have a “great deal of confidence” in each institution.Footnote 6, Footnote 7
Federal courts’ workload is another predictor. Chutkow (Reference Chutkow2008) and others show that greater case volume is associated with more jurisdiction stripping (Heise, Reference Heise2011; Greenfest, Reference Greenfest2013). We take two approaches to measuring judicial workload. First, we follow Chutkow (Reference Chutkow2008) and measure federal courts’ workload—Judicial Workload—as the number (in thousands) of civil cases filed the previous year in federal district courts by private parties seeking action against the United States. We retrieve these counts from the Federal Judicial Center (FJC)'s annual Federal Judicial Caseload Statistics (Table C-2).Footnote 8 Given the upward trend in the judiciary's workload, suggesting a unit root (confirmed by several diagnostics), we model the effect of workload's first difference (i.e., the change in workload from t − 1 to t).Footnote 9 We count civil cases where the United States is the defendant since the decision to initiate litigation is not under the US government's direct control (Chutkow, Reference Chutkow2008).Footnote 10
We next modify this workload variable to account for the increase in Article III federal judges over time. Adjusted Judicial Workload is the number of civil district court cases filed by private parties seeking action against the US government divided by the number of Article III district court judges serving each year.Footnote 11 This alternative workload predictor captures the judiciary's workload relative to the number of judges available to resolve cases.
The logic of separation of powers models suggests Congress strips jurisdiction as courts become increasingly ideologically distant (e.g., Segal et al., Reference Segal, Westerland and Lindquist2011). To account for this, we measure Ideological Distance as the absolute difference between the median Supreme Court justice's ideal point and the midpoint between the two congressional chamber medians’ ideal points using Bailey's (Reference Bailey2007) ideology measures.Footnote 12 Although jurisdiction-stripping laws may limit the authority of many federal courts, we use the spatial location of the Supreme Court because it sits atop the judicial hierarchy and is the final arbiter of disputes appealed from the lower courts (Chutkow, Reference Chutkow2008).Footnote 13
A final control variable is the annual number (in hundreds) of public laws passed by Congress (Chutkow, Reference Chutkow2008). This variable accounts for changes in Congress’ productivity that may affect the frequency of jurisdiction-stripping laws.
4. Methods and results
We use negative binomial regression to predict the annual number of jurisdiction-stripping laws.Footnote 14 Using this approach, we specify: (1) a baseline model including only the two public confidence measures; (2) a second specification including the unadjusted workload variable and all controls; and (3) a third model that substitutes the adjusted workload predictor for the unadjusted one.Footnote 15 Table 1 reports the negative binomial regression model estimates (Models 1–3).
Public Opinion and Jurisdiction-Stripping Laws (Count), 1973–2014
Table entries are negative binomial regression estimates with standard errors in parentheses; *p <.05 (two-tailed tests).
First, the impact of Court Confidence on passage of jurisdiction-stripping laws is consistent across all three models. In contrast to prior research, our data show a positive relationship between confidence in the Supreme Court and laws stripping federal courts’ jurisdiction. Panel (a) in Figure 2 illustrates these results across the observed range of public Court Confidence. The count model estimates (Model 3) predict an increase of one standard deviation (0.034) away from the mean (0.326) confidence in the Supreme Court would yield roughly 40 percent more jurisdiction-stripping laws (i.e., from approximately five to eight laws). Likewise, a shift from the minimum to maximum observed level of public confidence in the Court predicts an increase from roughly two to ten jurisdiction-stripping laws. This expected increase in jurisdiction stripping exceeds the magnitude of the effect of a min-to-max increase in Adjusted Judicial Workload, which predicts an increase of about six jurisdiction-stripping laws.Footnote 16 Thus, Court Confidence is both a statistically and substantively significant predictor of jurisdiction-stripping legislation, but in a surprising direction.
The impact of public opinion on jurisdiction stripping.
Note: Panels (a) and (b) display the predicted count of public laws with a jurisdiction stripping provision across the range of GSS confidence in the Supreme Court and Congress using the results from Model 3 in Table 1.
Additionally, and also in contrast to prior research, Americans’ confidence in Congress is negatively associated with jurisdiction stripping. In other words, the higher public confidence in Congress, the fewer jurisdiction-stripping laws Congress produces. Panel (b) in Figure 2 shows predicted counts of jurisdiction-stripping laws across the observed range of GSS confidence in Congress using results from Models 3. A decrease from the maximum to minimum level of public confidence in Congress predicts an increase from two to twelve jurisdiction-stripping laws. Public confidence in Congress is also a statistically significant and substantively important predictor of congressional action to limit federal courts’ jurisdiction.Footnote 17
We find mixed support for the influence of judicial workloads on jurisdiction stripping. Contrary to prior research, we find no significant relationship between the unadjusted number of federal cases filed against the United States and the number of laws passed limiting courts’ jurisdiction. However, we do find a significant relationship between adjusted workload and jurisdiction stripping. Increases in workload relative to the number of Article III judges predicts more jurisdiction-stripping laws.
The models also show that ideological distance between the Supreme Court and Congress has no significant association with jurisdiction-stripping laws. Finally, the number of jurisdiction-stripping laws is significantly, positively related to the total number of laws passed by Congress in the count models.
5. Discussion and conclusions
Several studies find that public support for the Supreme Court predicts both expressions of judicial power and congressional support for (or acquiescence to) courts’ authority and decisions. However, the association between public support for courts and judicial independence is inverted for jurisdiction stripping. Greater confidence in courts predicts more jurisdiction stripping, and greater confidence in Congress predicts less jurisdiction stripping.
Vanberg's (Reference Vanberg2001) theory of transparency in legislative-judicial interactions suggests an explanation for this anomaly. Perhaps legislation limiting courts’ jurisdiction is so far removed from public view that an unpopular Congress turns to jurisdiction stripping as a way to control a popular judiciary when a more visible confrontation between the branches would be costly. So, when Congress has less leverage to obtain favorable outcomes from courts, it stealthily retreats to limiting their ability to decide cases at all.
The literature on “internal” legislative support for judicial review suggests a second possibility (e.g., Rogers, Reference Rogers2001). Judicial review by allied courts can advance policy goals preferred by legislators (Whittington, Reference Whittington2005). So, as long as a legislature and a court are sufficiently politically aligned, an unpopular legislature might have incentives to empower a relatively popular court to pursue policy goals (at least temporarily) outside its reach.
Another possibility is that some limits on deciding cases may enhance judicial independence. For example, the Supreme Court Case Selections Act of 1988 removed litigants’ right to appeal certain state court decisions to the Supreme Court (although these cases could still reach the Court via writs of certiorari). The law is often interpreted as an expansion of judicial independence since it gave the Court greater freedom to select more legally and politically salient cases without the burden of deciding as many mandatory appeals. It is possible that some jurisdiction-stripping laws likewise clear out classes of unimportant cases, freeing judges to act in other areas.
Finally, it is also possible that the relationship between public opinion and jurisdiction stripping observed here is spurious and some additional factor influences public perceptions of the Court and Congress as well as Congress's production of jurisdiction-stripping legislation, although few plausible mechanisms are yet evident to us. Whatever the case, the positive association between confidence in the Court and jurisdiction-stripping legislation confounds the straightforward prediction of the public support hypothesis. These results suggest a need for continued attention to the role of public opinion in shaping interactions between federal courts and Congress (Rogers and Ura, Reference Rogers and Ura2020). Addressing this anomaly is an important challenge for scholars of judicial independence.
Acknowledgments
The authors are grateful to Mike Bailey, Dawn Chutkow, Alyx Mark, and Mike Zilis for providing data used in this paper and to Mike Bailey, Brandon Bartels, and John McIver for offering valuable comments on earlier versions of this project.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/psrm.2021.14.
