The Supreme Court’s much-awaited decision in Kisor v. Wilkie will have significant ramifications for the Environmental Protection Agency (“EPA”) and environmental law.  While the decision upheld the concept of Auer deference, which instructs courts to defer to agencies’ interpretations of their own regulations, it also imposed a number of limitations and restrictions on when Auer deference applies.  The decision leaves open many questions about what EPA guidance will qualify for Auer deference, and whether any statements that do qualify for deference are subject to immediate challenge as final agency action.  The decision thus presents opportunities for regulated parties to challenge EPA interpretations, but also challenges in that regulated parties may not necessarily rely on EPA’s interpretations as controlling.

EPA’s Historic Reliance on Auer

EPA, like many other agencies, implements most of its programs by regulation.  In addition, EPA issues many guidance documents, policy statements, letters to regulated parties, and other materials that explain the agency’s interpretation and application of those regulations. Historically, EPA has regularly received Auer deference from the courts.  The high-water mark was likely Decker v. Northwestern Environmental Defense Center, 568 U.S. 597 (2013), where the Supreme Court deferred to an EPA interpretation of a regulation implementing the Clean Water Act set forth in an amicus brief.

Implications of Kisor for EPA & Regulated Parties

Kisor instructs courts to defer to an agency’s interpretations of its own regulations only when (1) the regulation is genuinely ambiguous, (2) the agency’s interpretation is reasonable, and (3) the court determines deference is appropriate after making an independent, multi-part finding that the “character and context of the agency interpretation entitles it to controlling weight.”  While the first two tests draw heavily from Chevron and so are reasonably well defined, the third test is far more nebulous and the Court’s opinion raises as many questions as answers.

  • Genuine Ambiguity

Kisor permits deference only when a regulation is genuinely ambiguous, meaning there is more than one reasonable construction of a regulation even after the court carefully considers “the text, structure, history, and purpose of a regulation” de novo.  Slip. Op. at 14.  Kisor notes that courts will utilize the “same approach” they use for assessing ambiguous statutes under Chevron deference when analyzing Auer.  Id.

This test does raise the question of what role an agency’s preamble accompanying the regulations should play.  Presumably, the preamble is a relevant part of the “history” of the regulation (or another “interpretive tool” to consider).  However, the opinion does not address what weight it should be given.  Is it equivalent to legislative history or, because it presumably reflects the agency’s authoritative view at the time of promulgation, should it be given more weight?

  • Reasonable Interpretation

 Kisor states that an interpretation is reasonable when it “come[s] within the zone of ambiguity the court has identified after employing all its interpretive tools.”  Id.  This step again mirrors the inquiry courts make under Chevron.  Indeed, the Supreme Court expressly rebuked the suggestion that agency construc­tions of rules receive greater deference than agency con­structions of statutes.

  • Appropriateness of Deference

This third and final step is where parties regulated by agencies like EPA face both major uncertainty in whether EPA interpretations warrant deference, as well as opportunities to challenge EPA interpretations with which they disagree.  Stating that “[t]he inquiry on this dimension does not reduce to any exhaustive test,” the Court observed that deference is appropriate when Congress “would have wanted” courts to defer to agencies based on their “comparative attributes” and competencies.  Slip Op. at 15.  This is similar to the Chevron Step Zero analysis, where agencies are entitled to Chevron deference only if they invoke applicable authority “to make rules with force of law.”  United States v. Mead Corp., 533 U.S. 218, 237 (2001).

However, Auer Step Zero is much more ill-defined.  The Kisor Court lists several prerequisites for deference, none of which is precisely defined, and which even when taken together are not necessarily dispositive:

  1. The Interpretation must be one actually made by the agency, reflecting its official position.

Kisor requires the interpretation to be the “the agency’s authoritative or official position,” as opposed to an “ad hoc statement not reflecting the agency’s views.”  Slip Op. at 15.  This prerequisite does not require the EPA administrator or chief advisers to generate the interpretation, but it does require the interpretation to “emanate from those actors, using those vehicles, understood to make authoritative policy in the relevant context.”  Id. at 16.

The Court does not clearly locate where this line is drawn, but it does provide a few examples: official staff memoranda published in the Federal Register are authoritative, but speeches from mid-level staff, informal memoranda summarizing staff phone calls, and documents disclaiming their authoritative force are not official agency positions.  While these examples shed some light, the distance between them is quite large and major questions remain.

Many EPA regulatory programs rely heavily on various forms of guidance, and it is far from clear whether all of them are entitled to Auer deference under this revised approach.  For example, are all of the EPA letters to regulated parties setting forth interpretations of the Resource Conservation and Recovery Act (“RCRA”), collected in EPA’s “RCRA Online” Database, entitled to deference?  Or EPA’s Hazardous Waste Frequent Questions, which appear to be a less formal repository of information?   Do only interpretations from authorized EPA Headquarters officials receive deference, or can interpretations provided by EPA regional officials also warrant deference?

Moreover, how are regulated parties and courts to know if a document meets this test?  Many EPA websites provide helpful guidance, but their provenance and authoritativeness is unclear.  For example, EPA has a lengthy and detailed webpage on pesticide questions, but the authority or identity of the persons providing answers is unexplained, and the website itself notes that “[t]hese answers are not intended to create significant new guidance.”  Is there a presumption that material on an agency’s website satisfies this test, in the absence of contrary information?

Many guidance documents also disclaim any binding nature.  How will Kisor apply to such materials?  For example, EPA has issued guidance under its pesticide program providing that “[e]very [Office of Pesticide Programs, (“OPP”)] policy guidance document should clearly state that it is not binding on the public or OPP, and that OPP decision-makers and staff should not treat any policy guidance document as if it were a legally binding rule.”  Do all such documents not qualify for Auer deference, given the Court’s apparent approval of an opinion “declining deference when the agency had itself ‘disclaimed the use of regulatory guides as authoritative’”?  Slip Op. at 16.

  1. The Interpretation must implicate the agency’s substantive expertise.

Kisor requires the agency issuing the interpretation to have “comparative expertise,” as compared to an Article III court, “in resolv­ing a regulatory ambiguity.”  Slip Op. at 17.  The Court again clarifies this inquiry by way of examples: technical determinations and policy decisions like weighing costs and benefits fall within an agency’s substantive expertise, but courts are better suited to interpret legal concepts like the meaning of a common-law property term.

Questions remain about how broadly this requirement will be interpreted.  For example, will interpretations of procedural requirements be granted deference?

  1. The Interpretation must reflect fair and considered judgment.

The interpretation must also reflect the agency’s “fair and considered judgment.”  Kisor does not define that term.  Slip Op. at 17.  Instead, the Court lists examples of interpretations that fail to meet this standard: those that are merely convenient litigating positions, post hoc rationalizations advanced to defend past agency action against attack, unfair surprises that fail to account for reliance on existing interpretations, or attempts to correct an agency’s failure to anticipate an issue when crafting a rule.  Notably, the Court observes that the need to avoid unfair surprises means Auer deference is rarely appropriate when an agency has changed its regulatory interpretation.  Id. at 18.

The lack of clarity regarding what constitutes fair and considered judgment creates significant uncertainty.  For example, must an agency provide an articulation of the reasoning underlying its interpretation?  If so, this might call into question the adequacy of many guidance documents providing only cursory analysis.  For example, many of the RCRA online decisions contain arguably conclusory and equivocal language.  Conversely, does merely stating that, after consideration, an agency reached a certain conclusion suffice?  Or will courts apply a “show don’t tell” approach instead?

Finally, the decision leaves open certain questions regarding how Auer deference applies to changed agency interpretations.  For example, if a new interpretation conflicts with a prior one, is it never entitled to Auer deference?  Or would deference attach after the passage of an adequate amount of time?  Presumably, the Court did not declare that an agency’s changed interpretation can never obtain Auer deference: at some unspecified point in time, a changed interpretation would presumably become the norm, providing “fair warning” to regulated parties.  But the Court leaves unanswered how courts are to approach this issue.

  • Auer Deference & Final Agency Action

Kisor also has potential implications for the relationship between agency guidance, final agency action, and whether the action qualifies for deference.  The Court holds that Auer deference is only appropriate if the agency has made “authoritative policy” after exercising “fair and considered judgment.”  The Court also notes that agency documents that themselves “disclaim[] the use of regulatory guides as authoritative” will not receive deference.

These statements raise a significant question: will agency guidance that qualifies for Auer deference now by definition qualify as “final agency action” subject to judicial review?  The Supreme Court has held that agency action is final if it (1) marks the consummation of the agency decisionmaking process, and (2) is one from which legal consequences will flow.  Bennett v. Spear, 520 U.S. 154, 177-78 (1997).  One could argue that agency interpretations entitled to Auer deference under Kisor necessarily meet the first criterion, as they will constitute considered, official agency positions.  These interpretations may also automatically meet the second criterion if their ability to generate Auer deference creates or changes legal obligations for regulated parties.  Likewise, policies that do not amount to final agency action arguably should not be granted Auer deference, for example, in a subsequent enforcement proceeding where the agency attempts to rely on that document as an authoritative basis for its action.