The Supreme Court's Deference Continuum, An Empirical Analysis (Chevron to Hamdan)
William N. Eskridge Jr.
Yale University - Law School
affiliation not provided to SSRN
Georgetown Law Journal, April 2008
Eskridge and Baer conducted an empirical study of all 1014 Supreme Court cases between 1984 and 2006 (inclusive) in which an agency interpretation of a statute was at issue. Each case was coded for 156 variables. This article, presented in preliminary form as the 2007 Ryan Lecture at the Georgetown University Law Center, will be the first comprehensive empirical analysis of the Supreme Court's actual practice in agency interpretation cases. The study was motivated by academic debates about how extensive the Supreme Court's Chevron Revolution has been and ought to be, debates typically conducted without any systematic grounding in the Court's actual practice.
There are dozens of descriptive contributions the article tries to make. Consider three that illuminate larger doctrinal debates within the Court and among commentators.
First, the data demonstrate that the academic obsession with Chevron is doctrinally misguided, at least as regards Supreme Court practice. Indeed, from the time it was handed down until the end of the 2005 term, Chevron was applied in only 8.3 percent of Supreme Court cases evaluating agency statutory interpretations. During this time frame, the Court employed a continuum of deference regimes, to wit:
Curtiss-Wright Super-Deference, the super-deference afforded to executive interpretations involving foreign affairs and national security;
Seminole Rock Strong Deference, a strong form of deference afforded to agency interpretations of their own regulations;
Chevron Deference, the famous two-step approach that permits reasonable agency interpretations so long as the statute has not clearly spoken to the issue;
Beth Israel Deference, pre-Chevron tests, associated with particular subject areas, permitting reasonable agency interpretations that are consistent with the statute;
Skidmore Deference, giving agency interpretations respect proportional to their power to persuade;
Consultative (Skidmore-Lite) Deference, where the Court, without invoking a named deference regime, relies on some input from the agency (e.g. amicus briefs, interpretive rules or guidance, or manuals) and uses that input to guide its reasoning and decision-making process; and
Anti-Deference, which invokes a presumption against the agency interpretation in criminal cases and in some cases in which the agency interpretation raises serious constitutional concerns.
This continuum is more complicated than the literature or the Court's own opinions suggest, and it is a continuum in which Chevron plays a modest, perhaps even minor, role. Skidmore, the New Deal decision many thought had been rendered obsolete by Chevron, not only survives but flourishes; indeed, Skidmore is the unexpected star of the Chevron Revolution. However, Eskridge and Baer also caution that in the majority of cases - 53.6 percent of them - the Court does not apply any deference regime at all. Instead it relies on ad hoc judicial reasoning of the sort that typifies the Court's methodology in regular statutory interpretation cases.
Second, the study demonstrates that the Court applies none of these deference regimes in a doctrinally consistent or predictable manner. According to the data, the Court usually does not apply Chevron to cases that are, according to the Court's own criteria, Chevron-eligible. Analyzing how Chevron is applied in the cases where it is invoked by the Court, Eskridge and Baer find little doctrinal consistency, except for the proposition that legislative history is relevant in the Court's determinations whether Congress has delegated lawmaking authority to an agency or has directly addressed the issue in the statute. According the data, the Court is more likely to cite and follow Chevron in Chevron-eligible cases when it agrees with the agency interpretation, a finding we also made for the Curtiss-Wright, Seminole Rock, Beth Israel, and Skidmore deference categories. Apart from that regularity, the Court's explicit invocation of a particular deference regime is hard to predict.
Third, although there is little doctrinal consistency, the Court's practice does reveal regularities in cases where there is an agency interpretation. Although noting that the Solicitor General wins most of his cases before the Court whatever the criterion, Eskridge and Baer report that the likelihood of the Justices' agreeing (for whatever reason) with an agency interpretation are particularly high when:
(1) the statutory subject matter involves the instrumental solution to technical economic or scientific problems by an expert technocratic agency (e.g., bankruptcy, intellectual property, pensions, telecommunications), as opposed to social or economic legislation involving issues the Justices believe to be more normative or political (e.g., civil rights, entitlements, labor relations);
(2) pursuant to congressional delegation, the agency has adopted a legislative rule through a notice-and-comment process, and especially if the agency is interpreting its own established rule;
(3) the agency has consistently adhered to its interpretation over time.
Ironically, these factors - comparative expertise, open process, and consistency - are those associated with Skidmore, but they show up in the Chevron cases as well. Eskridge and Baer also report a significant correlation, for all the current Justices, between voting patterns and ideology, with conservative agency interpretations generating more support from Justices appointed by conservative Republican Presidents and liberal agency interpretations generating more support from Justices appointed by moderate Democratic and Republican Presidents.
In a lengthy normative part, Eskridge and Baer weigh in on the deference debates within and outside the Court, especially as they relate to the contours of the (perhaps misnamed) Chevron Revolution. Again, three conclusions deserve to be highlighted.
First, Eskridge and Baer join those scholars and judges who believe that Chevron is a proper approach, so long as the agency is acting pursuant to a congressional delegation of lawmaking authority. They support this conclusion by reference to constitutional and statutory sources, as well as functional considerations that have dominated judicial and academic arguments seeking to expand Chevron's domain. Eskridge and Baer, however, maintain that the Court's chaotic application of Chevron needs to be addressed, if for not other reason than to provide better guidance for lower courts. Specifically, they urge the Court to clarify which statutes the Justices believe constitute congressional delegation of lawmaking authority, to jettison most of the pre-Chevron deference regimes that have survived Chevron. Like barnacles on a rusting ship, most of these regimes (specifically, Curtiss-Wright, Seminole Rock, and Beth Israel) should be scraped away.
Second, Eskridge and Baer support the notion that Skidmore survives Chevron and, indeed, would carry that thesis further. Skidmore is, and should be, the universal default approach: any statutory issue as to which the agency is not acting pursuant to delegated lawmaking authority is one where the Supreme Court (and of course lower courts) ought to be receptive to agency inputs not just previous interpretations, but also valuable agency-generated information about the history of the statute and of administrative activity, factual materials about the social and enforcement context, and analysis of the policy consequences of different interpretations. This suggestion is by no means revolutionary; it reflects the deep practice of the Court, which for institutional reasons relies and perhaps has little choice but to rely on agency inputs. The primary concern, according to Eskridge and Baer, is that the Solicitor General is, for these institutional reasons, playing too much a role in bankruptcy etc. cases; to prevent policy from becoming unbalanced, the Court should consider a greater diversity of policy views.
Third, and most important, Eskridge and Baer's study concludes that both the practice and the theory of statutory interpretation must view the enterprise through the lens of the modern administrative state. Such a lens requires rethinking such judicial doctrines as the plain meaning rule, purposive interpretation, legislative history, stare decisis, and the substantive canons of statutory construction. For one example, traditional theories justify consultation of legislative history as evidence of Congress's original intent, including compromises; critics claim that such history is unreliable at best, corrupting at worst. Eskridge and Baer's examination of 1014 Supreme Court cases, and the government's briefs in most of those cases, revealed to them that, from the agency's point of view, legislative history is more valuable as evidence of the problem Congress was trying to solve, including values and trade-offs; the range of discretion Congress expected the implementing agency to have; and political feedback to the agency as it adopted various policies and strategies. This last value, dismissed by doctrinalists as bootless subsequent legislative history, is revealed dramatically in many of the Chevron or Skidmore cases handed down by a doctrinally conservative Court. Eskridge and Baer suggest that the Justices' practice is sounder than the abstract theorizing of both judges and law professors in this and other areas.
At some point during the Winter of 2007-2008, Eskridge and Baer will make their database available to other researchers. (This summer, a leading legislation scholar is doing an independent reality check of the authors' coding of a random sample of the 1014 cases.)
Date posted: June 20, 2007