Skidmore deference—the standard of judicial review of agency legal interpretations articulated in Skidmore v. Swift & Co.1—is an old doctrine. The Supreme Court’s decision in Chevron U.S.A. Inc. v. Natural Resources Defense Counsel, Inc.2 arguably rendered Skidmore irrelevant; but more recent Court pronouncements in Christensen v. Harris County3 and United States v. Mead Corp.4 resurrected Skidmore deference as an alternative to Chevron. Since Christensen and Mead, Skidmore has reemerged as a unique and frequently used standard of review. Yet while reams of scholarly articles evaluate Chevron and Mead, contemporary scholarship offers comparatively little analysis of Skidmore as a doctrine.
Where it applies, the Skidmore standard contemplates that courts will give weight to agencies’ interpretations of statutes that they administer according to their persuasiveness, as determined by reference to “the thoroughness evident in its consideration, the validity of its reasoning, its consistency with earlier and later pronouncements, and all those factors which give it power to persuade, if lacking power to control.”5 Courts and scholars seem to agree that Skidmore is less deferential than Chevron, but how much less and in what way is unclear. The scope of Skidmore’s applicability is equally unsettled. The Supreme Court has applied Skidmore in several cases, but with tremendous rhetorical inconsistency that further muddles rather than clarifies.
In our Article, we analyze Skidmore from several angles. Principally, we evaluate empirically how the federal courts of appeals have conducted Skidmore analysis post-Mead. Our aim is not to judge when these courts do or should apply Skidmore as opposed to Chevron or no deference doctrine at all, although we do end the Article by summarizing some emerging issues of Skidmore’s scope that we identified in the course of our study. Rather, our primary goal is to consider what the courts of appeals are doing when it is clear that they are, in fact, attempting to apply Skidmore as a doctrine.
The Nature of Skidmore—Sliding Scale vs. Independent Judgment
We start from the premise that, at least in theory, Chevron and Skidmore are different deference doctrines that reflect distinguishable, if related, concerns. Neither deference doctrine is relevant if a statute’s meaning is clear because both courts and agencies must give effect to the unambiguous intent of Congress.6 Where a statute is ambiguous, however, we accept Chevron deference as a creature of congressional intent that agencies rather than the courts should be the primary interpreters of certain designated statutes. By contrast, we see Skidmore deference as a doctrine of judicial prudence which appreciates that, sometimes, agencies bring to bear superior resources and expertise in interpreting statutes.
Starting with this basic understanding, our review of scholarly analysis and Supreme Court opinions identifies two distinct visions of what Skidmore represents: the independent judgment model and the sliding-scale model. The independent judgment model of Skidmore, most closely associated with Colin Diver, anticipates that a court will assign an administrative interpretation “weight” depending upon its “persuasiveness,” meaning the court’s evaluation of the interpretation’s merits. Hence, while the independent judgment model of Skidmore falls under the heading of judicial deference, such an approach in effect allows a reviewing court to treat an agency’s interpretation much like any other litigant’s view, which is hardly deferential at all.
The sliding-scale model of Skidmore is conceptually more deferential. Most closely associated with Kenneth Culp Davis and Thomas Merrill, the sliding-scale approach expects that courts will evaluate several factors in deciding how much weight to give the agency’s view rather than allowing a reviewing court to reject a disfavored agency interpretation as merely “unpersuasive.” Because agencies typically have special expertise related to their respective regulatory schemes, applying the Skidmore factors requires courts to assess the extent to which an interpretation reflects an exercise of the agency’s potentially superior interpretive competency. As we see it, a reviewing court that feels compelled to frame its conclusions in such terms perceives some real limitation on the scope of its own discretion.
Both the independent judgment model and the sliding-scale model of Skidmore find support in post-Mead Supreme Court opinions.7 Because most cases do not result in Supreme Court review, we were curious to determine what the federal appellate courts have done with the Supreme Court’s muddled guidance. From a starting sample of the 450 federal appellate court cases decided in the five years after Mead that cited any of Skidmore, Christensen, or Mead, we identified and evaluated 106 “Skidmore applications” in which the reviewing court applied Skidmore to evaluate an agency’s interpretation after finding the statute ambiguous.8
Quantitatively, we asked whether the case law demonstrated a judicial preference for the independent judgment model or the sliding-scale model.9 We found that the courts applied the sliding-scale approach in 79 of 106, or 74.5%, of Skidmore applications. Related to that inquiry, we observed also that the agency’s interpretation won in 64 of 106, or 60.4%, of the Skidmore applications. Consistent with our perception of the independent judgment model as less deferential than the sliding-scale model, the agency’s interpretation prevailed less often when courts applied the former approach (50% of independent judgment model applications versus 59.5% of sliding-scale applications).10 A comparison of our study with others suggests that, while less deferential than Chevron, Skidmore nevertheless represents a clear thumb on the scale in favor of administrative agencies.
Qualitatively, we evaluated the 79 Skidmore applications applying the sliding-scale model to determine how that approach actually functions in practice. We found much evidence of courts attempting to apply several factors that the Court’s Skidmore jurisprudence identifies as relevant: thoroughness, formality, validity, consistency, expertise, and duration/contemporaneity. We observed, however, that the appellate courts do not exhibit consistent understandings of what the factors signify individually or how they relate to one another.
For example, some courts invoke the thoroughness factor to assess an agency’s explanation of its interpretation, while others invoke thoroughness to assess the procedures an agency followed to produce the interpretation. Agency procedures are also the focus of the formality factor, and we therefore suggest that courts would do well to separate their assessments of an agency’s explanations and procedures. Further, in assessing an agency’s procedures, some courts calibrated the formality factor, improperly in our view, by counting against an agency the fact that its interpretation had not issued from notice-and-comment rulemaking, a fact true of almost any Skidmore-eligible interpretation.
We also noted confusion among the courts regarding the relevance of an agency’s consistency in holding an interpretation. Especially since Chevron’s ascendance, courts have been more attuned to agencies’ need for flexibility in administering statutes, which may require changing positions on a given policy. Accordingly, some courts question whether such a change should preclude deference, as Skidmore seems to suggest. We suggest that the consistency factor assesses whether an agency’s changed position upsets regulated parties’ reliance interests, undermines rule-of-law values by treating similarly situated parties differently, or reflects arbitrary or ill-intentioned action. In theory, those harms are less likely to be present, and deference therefore not inappropriate, when the agency undertakes relatively involved interpretive procedures to produce a thorough explanation for a valid, albeit new, interpretation. We found some evidence to support that theory, as an explicitly negative finding on the consistency factor was less often correlated with a rejection of the agency’s interpretation than were negative findings on the other factors.
Finally, applications of the expertise factor were often underwhelming. This factor, more than any other, lacked teeth as courts rarely found that it weighed against deferring to the agency. Indeed, courts often seemed to invoke agency expertise as a throwaway line, tacking it onto an independent decision to defer to the agency. A few courts, however, understood this factor, appropriately in our view, to require a careful evaluation of the agency’s interpretive competence vis-à-vis the court, and whether the agency applied that expertise in forming the given interpretation.
Reconceptualizing the Modern Skidmore Standard
In an effort both to clarify and simplify Skidmore analysis, we offer the following conceptualization. As we see it, in applying Skidmore’s factors, the courts often miss the forest for the trees. In other words, courts tread their way through the individual factors without asking what those factors are meant to reveal. Stepping back, we suggest the heart of Skidmore analysis lies in two competing concepts: comparative agency expertise and the potential for arbitrariness in the exercise of agency power.
Generalist judges are often prudent and wise in deferring to a specialized agency that has focused its resources and superior expertise carefully and deliberately upon a particular interpretive conundrum. Yet not all agency interpretations are the product of such care and deliberation. The informal interpretive formats for which Skidmore provides the appropriate standard often lack the procedural safeguards that the Administrative Procedure Act imposes on notice-and-comment rulemaking and formal adjudication. The Administrative Procedure Act specifically calls upon the courts to serve as a bulwark against arbitrary and capricious agency action.
Skidmore’s factors should thus be understood as ferreting out two things: first, the extent to which agencies have deliberately employed their expertise and resources in evaluating the statutory ambiguity at hand; and second, the potential for arbitrariness in the agency action in question. Considering the factors as assisting in this task rather than as isolated individual inquiries should help clarify their meaning and relative weight.
In addition, we suggest that courts view Skidmore’s sliding scale as a choice among three identifiable moods or attitudes that they may adopt toward agencies’ legal interpretations. Conceptualizing Skidmore’s sliding scale as consisting of three moods rather than infinite degrees along a spectrum should ease the burden of its application while also fulfilling its promise of tailoring deference to the unique interpretation at hand. Each of the three Skidmore moods turns on an evaluation of the contextual factors, guided by the overarching principles of expertise and arbitrariness.
The first mood is highly deferential, even Chevron-like, and is appropriate when the court finds that the contextual factors more or less uniformly fall in the agency’s favor, suggesting applied expertise rather than arbitrariness as the basis for the agency’s action. In such a case, a court should defer to a reasonable agency interpretation even if the court might prefer an alternative view. The second mood lies at the opposite end of the sliding scale and consists of little or no deference toward the agency’s interpretation. This mood is appropriate when the contextual factors fall more or less uniformly against the agency, suggesting that an interpretation is an ad hoc, arbitrary conclusion rather than the result of the agency’s applied expertise. True intermediate Skidmore deference occupies the middle ground, where a court may legitimately assert its own preferences but should be wary of doing so for fear of intruding too deeply upon agency policy prerogatives or entrenched reliance interests. Intermediate deference is appropriate when assessment of the factors leaves the court with some question of whether the interpretation resulted from the agency’s applied expertise or from arbitrary action.
Many an eyebrow will be raised by our contention that, in some cases, Skidmore should prescribe strong, Chevron-like deference. That contention, however, follows directly from Skidmore’s underlying principles of expertise and arbitrariness. Simply put, where the evidence strongly shows that the agency deliberately applied its superior interpretive competence and avoided arbitrariness, the court’s deference should be also be strong. One might respond that Skidmore deference should never equal Chevron deference in strength because Skidmore applies, in theory, when Congress intended the court, rather than the agency, to hold ultimate interpretive authority. The fact that Congress vested interpretive authority in courts says nothing, however, about how courts should exercise that discretion. At bottom, Skidmore represents a prudential standard that, in some cases, should lead courts to be very reluctant to reject the agency’s view.
Skidmore’s Scope
Finally, in Part IV, we noted that questions remain regarding the range of Skidmore’s applicability. Like other scholars who have examined Mead’s test for deciding whether Skidmore or Chevron provides the appropriate review standard, we found that courts sometimes avoid the Mead inquiry and employ Skidmore even though Chevron arguably would apply. Few scholars, though, have examined the opposite boundary of Skidmore’s domain to determine when pure de novo review, rather than Skidmore, is appropriate. In truth, very few cases so far have addressed true scope-of-applicability issues between Skidmore and pure de novo review. Yet the few cases involving such issues lead us to suspect that others will follow.
For example, interpretations advanced for the first time in the course of litigation represent agency views that arguably do not deserve Skidmore deference. Another issue of Skidmore’s reach pertains to the relationship between Skidmore and stare decisis. While the Supreme Court in the Brand X case addressed the primacy of Chevron versus judicial precedent,11 the Brand X Court neither discussed nor cited Skidmore, even though Skidmore deference shares the same tension with stare decisis as Chevron previously did. Additionally, the Supreme Court has given conflicting signals regarding how to address agency interpretations that are arguably unconstitutional.
Finally, Skidmore’s domain is also being extended to agency interpretations of the agency’s own regulations—interpretations which arguably are governed by the highly deferential standard of Seminole Rock & Sand Co.12 While we offer preliminary thoughts on the extent to which Skidmore should apply in these various circumstances, we believe this is an area ripe for further study.
* Associate Professor of Law, University of Minnesota Law School.
** Bristow Fellow, Office of the Solicitor General. All of the views expressed in this Essay and the Article it discusses are our own, and none represent the positions or views of the United States or the Department of Justice.
1. 323 U.S. 134 (1944). 2. 467 U.S. 837 (1984). 3. 529 U.S. 576, 587 (2000).
4. 533 U.S. 218, 234–38 (2001).
5. 323 U.S. at 140.
6. Mead’s two-part inquiry for determining whether Chevron or Skidmore applies assumes the relevant statute’s ambiguity. We discuss in the Article competing conceptualizations of Mead as a “step zero” or “step one-and-a-half” as it relates in particular to Chevron but also to Skidmore.
7. Compare Raymond B. Yates, M.D., P.C. Profit Sharing Plan v. Hendon, 541 U.S. 1, 17–18 (2004) (applying independent judgment model of Skidmore by determining statute’s meaning without regard to agency interpretation and then briefly noting that agency concurred), with Alaska Dep’t of Environmental Conservation v. EPA, 540 U.S. 461, 487–88 (2004) (applying sliding-scale model of Skidmore by assessing agency interpretation’s reasonableness, long history, and reflection of applied expertise before deciding to afford it weight).
8. In other words, the cases we identified as Skidmore applications excluded those in which the reviewing court concluded, either explicitly or implicitly, that the statute’s meaning was clear. Consistent with our beginning understandings of the interrelationship of Mead, Chevron, and Skidmore, discussed supra note 5, we found that the courts generally apply the equivalent of Chevron step one review before proceeding to Skidmore analysis.
9. Of course, evaluating judicial opinions toward such ends is inevitably at least somewhat subjective. The Article contains an extensive discussion of the methods we employed to make our inquiry as objective as possible.
10. Seven of the 106 cases could not be categorized as either independent judgment or sliding scale. The government’s view of the statute prevailed in all of those cases, pushing the total deference rate across the sample to 60.4%.
11. Nat’l Cable & Telecomms. Servs. Ass’n v. Brand X Internet Servs., 545 U.S. 967 (2005).
12. Bowles v. Seminole Rock & Sand Co., 325 U.S. 410 (1945). Many refer to this standard as Auer deference in honor of Auer v. Robbins, 519 U.S. 452, 461 (1997).
