Empirical Insight into the Use of Seminole Rock Doctrine, by William Yeatman

by Guest Blogger — Wednesday, Sept. 26, 2018

Under the Supreme Court’s Seminole Rock (or Auer) doctrine, Article III courts give binding deference to an agency’s regulatory interpretation “unless it is plainly erroneous or inconsistent with the regulation.” In an effort to better understand Seminole Rock deference as judicial methodology, I recently took a deeper dive into a dataset I had created for a note published earlier this year. I started with 379 regulatory interpretations subject to judicial review under the Seminole Rock deference doctrine by U.S. Courts of Appeals from 1998 to 2013 (see the Methodology section of the note to understand how I gathered these decisions). I then reread these opinions with an eye to better understanding the mechanics of the doctrine as practiced by federal courts. What follows is a by-the-numbers summation of what I found and how I found it.

29 Non-Regulatory Interpretations Accorded Seminole Rock Deference

In 36 instances—almost 10% of the total—the government sought Auer deference for agency interpretations of non-regulatory texts, including non-legislative or informal rules, precedential administrative adjudicative decisions, and administrative orders; in 29 instances, the courts accorded binding deference. These cases are highlighted in red in the dataset

225 Seminole Rock “Step Ones”

Chevron deference famously proceeds in two “steps.” At step one, courts determine whether the statutory text is ambiguous; if so, courts continue to Chevron step two, where the court decides if the agency’s interpretation is reasonable. In a perceptive 2015 article, Prof. Michael Healy observed that the Supreme Court never has set forth a two-step analytical framework for Seminole Rock deference. Rather, under the doctrine, courts may collapse review of agency regulatory interpretations into a single query regarding reasonableness, without passing on whether the underlying regulatory text is ambiguous

For this post, I investigated whether federal circuit courts employed a distinct ‘step one’ when applying the Seminole Rock doctrine. I did so by looking for objective criteria that the court undertook an independent investigation into whether the regulatory text is ambiguous. For example, courts sometimes avowed that Seminole Rock takes place in two-steps and expressly used the Chevron framework as a referent. In other instances, discerning ambiguity may be gleaned from a court’s express recognition of either multiple possible interpretations or regulatory silence. (See columns M and AD in the dataset). In still others, the court’s textual analysis is peppered with language such as “plain meaning” or “unambiguous,” and therefore constitutes an identifiable step one. (See column N).

All told, I identified 225 “step ones,” or 65% of the total (when non-regulatory interpretations are excluded). This data is in column L of the dataset.

At Least 6 Different Standards of Review

As a standard of review, Seminole Rock is most readily associated with the understanding that an agency’s interpretation is of “controlling weight unless it is plainly erroneous or inconsistent with the regulation.” Nevertheless, lower courts limited themselves to this canonical formulation only about half the time in setting out a standard for review of agency regulatory interpretations. For the other half of interpretations subject to judicial review, courts applied a variety of review standards that emphasized different interpretive factors, including:

  • About 20% of the time, courts introduced a Seminole Rock review standard by referencing the Supreme Court’s language in Christensen v. Harris County stating that “deference is warranted only when the language of the regulation is ambiguous.” Of course, such a standard is more conducive to a two-step framework, of the sort described above.
  • About 5% of the time, the circuit courts introduced a Seminole Rock review by stressing the importance of authorial intent.
  • In about 2% of the cases, the lower courts emphasized interpretive consistency in the review standards for Seminole Rock

There were other idiosyncratic, and sometimes strange, standards sprinkled throughout the circuits, including

  • In the U.S. Court of Appeals for the Ninth Circuit, for example, multiple panels seemingly favored the government by employing a standard of review whereby an agency’s interpretation is binding unless an alternative reading is “compelled” by the rule, which is arguably a heightened standard to overturn government action.
  • Both the U.S. Courts of Appeals for the Fifth and Sixth Circuits have cases on the books featuring a standard of review that precludes binding deference to texts that do not carry the force and effect of law, in apparent contravention of the Supreme Court precedent in Auer v. Robbins.

This data is coded and recorded in column K of the dataset.

5 Most Used Interpretative Aids at Seminole Rock Step Two

Whenever the court determined that the underlying regulation was ambiguous, I recorded which interpretative aids the court employed at step two (see row 384 of the dataset). The top five most used interpretative aids at Seminole Rock step two were:

  1. Consistency: 43
  2. Sister Court Reasoning: 38
  3. Lack of Notice/Post Hoc Rationalization: 36
  4. Regulatory Purpose: 34
  5. Regulatory Preamble: 30

For More

In a follow-up post, I’ll discuss the implications of the data. Otherwise, anyone is free to use the dataset. Nevertheless, there is quantitative data I didn’t publish, and I recorded a great deal of qualitative results on Seminole Rock, so if you’re working in this area and have an empirical question, I might know the answer. I welcome any such inquiries at wmy5@georgetown.edu.

William Yeatman is a public policy professional in Washington, D.C. and a J.D. candidate at the Georgetown University Law Center.

Cite As: Author Name, Title, 36 Yale J. on Reg.: Notice & Comment (date), URL.

This entry was tagged .

Leave a Reply

Your email address will not be published. Required fields are marked *