Overview
Chevron — the stalwart doctrine of administrative deference — is under political, judicial, and academic attack. Our newest Justice, Neil Gorsuch, has denounced Chevron as “a judge-made doctrine for the abdication of the judicial duty.” Chief among his complaints is that Chevron deference enables an agency to “reverse its current view 180 degrees anytime based merely on the shift of political winds and still prevail [in court].” Justice Gorsuch minces no words: “We managed to live with the administrative state before Chevron. We could do it again.”
When agencies act with the force of law in interpreting statutes, the canonical Chevron Two-Step framework for judicial review asks: first, whether Congress has answered the precise issue at hand, and, only if not, second, in the face of Congressional silence or ambiguity, directs courts to defer to “permissible” or “reasonable” agency interpretations.
In a forthcoming article in Fordham Law Review, “Cutting in on the Chevron Two-Step,” I advocate for the incorporation of State Farm “hard look” review into the Chevron framework. It is too soon to stop the Chevron dance. But it is high time for State Farm to cut in. Justice Gorsuch did not mention State Farm, and this omission is telling; State Farm’s demand for “reasoned decisionmaking” from agencies mitigates the judicial rubberstamping that Chevron’s critics decry.
My proposed Chevron–State Farm conceptual framework acknowledges the inextricable link between questions of statutory interpretation (typically the domain of Chevron) and questions of discretionary policy-making authority by agencies (typically the domain of State Farm). Incorporation of State Farm into Chevron’s second step bolsters judicial oversight by requiring agencies to provide fact-based policy reasoning when deciding between viable interpretations permitted by a statute.
Such a framework impacts a wide domain of agency rules, requiring agencies to provide factual support and reasoned explanation to support policy-based choices and determinations before a court will defer to their statutory interpretations. It has significance for questions pertaining to the role of cost-benefit analysis in agency regulation and many other fact-based interpretive questions as well, with courts applying arbitrary and capricious review to agencies’ policy-based preferences for one statutory reading as opposed to another. The model is premised on the idea that effective judicial oversight leading to optimal agency regulation depends upon forcing agencies to justify their policy-based choices when issuing rules that interpret statutes. This will prevent agencies from using legal statutory interpretation arguments — which are relevant at Step One — to evade their obligation to justify their policy-based interpretation arguments at Step Two.
In Part I of this blog series, I describe how Catskill Mountains Chapter of Trout Unlimited, Inc. v. EPA — a recent Second Circuit case, now awaiting the Government’s response in a pending petition for writ of certiorari before the U.S. Supreme Court — shines a light on the issue of judicial acquiescence to agency statutory interpretation. This high-profile case involves a decades-long legal challenge to the Environmental Protection Agency’s (EPA) promulgation of the Water Transfers Rule, which interprets language from the Clean Water Act. As is typical when a regulatory agency exercises its delegated authority to interpret ambiguous statutory language, the EPA made a difficult policy choice in choosing one plausible statutory interpretation over another. But in Catskill Mountains, the agency defended its interpretation on solely legal grounds. The case thus provides a vivid illustration of the core problem — namely, an agency effectively uses Chevron Step One legal statutory interpretation arguments to justify its implicitly policy-based interpretation at Step Two. This example is part of a wider trend of the weakening, if not effective nullification of, judicial oversight at Chevron Step Two. This perfunctory Chevron Step Two allows agencies to shroud policy choices in legalistic reasoning, thereby avoiding a need to demonstrate factual support for their policy decisions. And the Catskill Mountains case offers a compelling example of the need to resuscitate Chevron’s anemic Step Two — a need that could be addressed by the State Farm approach.
In Part II of this blog series, I will evaluate the normative significance of the Chevron–State Farm conceptual framework at the dawn of what may be a new age for hard look review at Step Two.
Part I: Catskill Mountains: A Spotlight On The Problem of Judicial Acquiescence To Agency Statutory Interpretations, and the Chevron-State Farm Solution
The legal dispute in Catskill Mountains arises from the EPA’s interpretation of ambiguous statutory language based on its choice of one permissible interpretation of the Clean Water Act (CWA) over another. More specifically, the case concerns an environmental dispute regarding whether the EPA must issue water pollution permits for systems that transport water from one waterbody to another. Since the 1972 enactment of the CWA, which prohibits “the discharge of any pollutant by any person,” the EPA has taken the position that water transfers were exempt from permitting requirements. The EPA, however, only formalized its position in 2008 by promulgating the Water Transfers Rule in response to numerous legal setbacks, in which courts refused to defer to the agency’s informal, albeit longstanding, position regarding water transfers.
To justify its rule, the EPA provided explicitly “legal” statutory interpretation reasoning — an analysis focused exclusively on inferring Congressional intent from the language and structure of the statute — in the preamble and response to comments. In the EPA’s own words, it “reasonably based the Water Transfers Rule on . . . Congress’s intent . . . to avoid unduly burdening state authority over water allocations and water resources, including for critical drinking water, agriculture, and flood control needs.”
As a matter of statutory interpretation, the EPA’s construction of the Clean Water Act (CWA) is plausible. What is missing, however, is any fact-finding, critical analysis, or even explicit acknowledgement of, the underlying policy considerations embedded in its interpretive decision. By relying solely on its view that it was Congress’ intent to avoid unduly burdening the states, the EPA did not engage in any kind of scientific analysis of water transfers in order to consider the competing policy considerations, namely the potential environmental, health, and economic harms caused by some inter-basin transfers. In fact, the EPA explicitly disclaimed reliance on any type of cost-benefit analysis or assessment of these competing considerations.
At Chevron Step One, the district court agreed with the EPA that the CWA was ambiguous as to whether water transfers require permits. But then, applying State Farm at the second step of Chevron, the court concluded that the agency had acted arbitrarily and capriciously by failing to undertake some sort of analysis of its policy choice, namely, to consider alternative policies and to explain why it chose one over others.
The agency’s rare Step Two loss was short lived. On appeal, the Second Circuit reversed, holding that the district court erred by incorporating the stricter State Farm standard into its Chevron Step Two analysis. Having decided that State Farm should fall out of the equation in the case of its review of EPA’s interpretation, the court readily found that the EPA’s rationale “was sufficiently reasoned to clear Chevron’s rather minimal requirement that the agency give a reasoned explanation for its interpretation.”
Catskill Mountains illustrates the all-too familiar scenario, whereby the agency’s articulation of a permissible legal interpretation of a statute triggers deference that, in effect, amounts to a wholesale endorsement of the agency’s policy-based choice between two or more potential readings of the statute without sufficient judicial oversight. The EPA’s starting point was that the Water Transfers Rule was within the agency’s discretion and thus must be afforded deference under Chevron, for the agency had reasonably explained why it chose the permissible interpretation that it did. The EPA was quite forthright that it had authority to choose any possible “legal” interpretation of the Act without engaging in fact-finding. As the district court noted (and the EPA did not dispute), the EPA “based its interpretation on an analysis of many of the same provisions the Court analyzed at step one. . . [namely] parts of the statute and its legislative history that the Court discussed in its step-one analysis . . . .” Thus, the court continued, the “EPA effectively attempted to use Chevron-step-one arguments to justify its interpretation at step two.”
Under the Chevron–State Farm model (which the district court implicitly applied), the EPA would have had to justify its interpretive choice with support in the administrative record. In the existing administrative record, the EPA, in its response to public comments, merely asserted various “beliefs” regarding the potential harms from inter-basin transfers and the potential burdens upon the states from permitting. Specifically, the EPA asserted that exempting inter-basin transfers from permitting would avoid “unnecessarily” or “unduly” burdening States’ water allocations.
But determining whether the burdens of permitting justify the costs requires factual findings that the EPA never made. Indeed, the “EPA has conceded that the Rule is not based on any ‘scientific analysis of’ inter-basin transfers or any study of the effects of such unregulated transfers ‘on the costs of drinking water treatment, recreation, or commercial fishing.’ Nor did EPA ever evaluate or find any facts regarding the burdens of compliance with . . . permitting.” In fact, the EPA took the position that it need not put forth any such policy-based evidence; instead, it took cover behind its legal analysis, relying wholly on the very same materials (statutory text, purposes, and legislative history) that led the court to conclude that there was ambiguity at Chevron Step One. Again, the EPA was honest about this, complaining that the district court “erroneously undertook a searching and skeptical examination of EPA’s reasoning” and “improperly required EPA to explain its choice in far more detail than the law requires.” Under the Chevron–State Farm model, such explanations would be mandatory.
The Chevron–State Farm model is premised on the principle that courts, via judicial review, “retain a role, and an important one, in ensuring that agencies have engaged in reasoned decisionmaking.” With respect to ambiguous statutory language, agencies are often called upon to make their interpretive choices based on policy-driven considerations. With State Farm incorporated into Chevron Step Two, the agency is forced to identify those considerations and provide factual support to show how its legal interpretation effectuates them. Contrary to the EPA’s prevailing argument on appeal in Catskill Mountains, the district court did not believe that the mere “exercise [of the EPA’s] Chevron-triggering power to make rules through notice and comment” should suffice. In its view, it would not be “per se” reasonable when an agency chooses — based on unarticulated and thus unvetted policy variables — between two permissible statutory interpretations.
The implications of adopting the Chevron–State Farm model go beyond their impact on the outcomes of particular disputes. The model actualizes the normative justification for Chevron deference at Step Two, namely that the agency has made a policy choice to which the courts should defer. It highlights the inextricable link between the implied delegation rationale and expertise rationale. It has always been the case that among “[t]he principles underlying” Chevron deference is the need for an agency to apply “more than ordinary knowledge” — i.e., “agency expertise” — when “fill[ing] . . . gap[s] left, implicitly or explicitly, by Congress.” The Chevron–State Farm model casts courts in an independent role to analyze the means by which agencies choose statutory interpretations, and the revised framework will affect agency decisionmaking, I submit, for the better. By requiring that the “reasoned decisionmaking” by agencies includes fact-finding or cost-benefit analysis of underlying policy choices vetted in the administrative record, the model will ensure that agency expertise is at the center of the discussion and will produce more effective regulatory decisions. It would also put the brakes on what Justice Gorsuch decried as agency actions driven solely by “shift[ing] political winds.”
This post is the first in a two-part series by Professor Sharkey. The second post can be found here.