Mead and the Prospective Exercise of Discretion
Posted: 17 Jan 2002
This article offers a skeptical assessment of United States v. Mead Corp., 121 S.Ct. 2164 (2001). In Mead, the Court undertook to clarify the manner in which courts should review administrative agency pronouncements that lack the force of law, such as interpretive rules, general statements of policy, and advice letters. Mead exempts those pronouncements from the familiar test of Chevron U.S.A., Inc. v. NRDC, 467 U.S. 837 (1984), under which an agency's statutory interpretation should be upheld if the statute is ambiguous and the agency's interpretation is reasonable. Instead, according to Mead, courts should review interpretations rendered in these nonbinding "formats" using the so-called Skidmore test, under which the court may construe the statute on its own authority but should give "weight" to the agency's views (Skidmore v. Swift & Co., 323 U.S. 134 (1944)).
The article maintains that this rule fits together poorly with the Chevron formula as it has come to be understood. Chevron has evolved into an amalgam of ordinary statutory construction (used to determine whether the statute is "ambiguous") and oversight of the agency's exercise of discretion (used to determine whether the interpretation is "reasonable"). In practice, the latter step overlaps the traditional inquiry into whether the agency's position was arbitrary or capricious.
At the first stage of Chevron analysis -- which deals with whether the statute is ambiguous -- the shift to Skidmore probably will not matter very much. Under either test, the court is expected to apply "traditional tools of statutory construction," giving a small edge to the agency. Minor variations in emphasis as to how deferential the court should be while performing this task are unlikely to affect outcomes very much.
Instead, the distinction between Chevron and Skidmore deference is likely to matter more at the second stage -- where the statute has been deemed ambiguous. In that situation, someone will have to use creativity to ascribe a meaning to the statute. At least in theory, it should make a difference whether the court allows the agency to be the primary decisionmaker (per Chevron) or retains that responsibility for itself (per Skidmore).
However, the article goes on to claim that, in most of these situations, the answer given by Mead is wrong. The agency's resolution of what is acknowledged to be a statutory ambiguity is closely intertwined with its exercise of statutory discretion. The reviewing court should not use Skidmore review to displace that discretion. Regardless of whether the agency has already exercised its delegated authority or has merely announced the manner in which it intends to exercise it in the future, Chevron should apply.
Mead mistakenly treats nonbinding actions as an inferior form of administrative behavior -- to which, therefore, the court owes less respect. Those actions should rather be seen as fundamentally similar to binding actions, even if reviewed at an earlier stage of implementation. It is true that many nonbinding pronouncements contain only interim interpretations, as to which the agency has not fully thought out its position. However, finality and ripeness defenses should keep those pronouncements from being reviewed at all. Where, on the other hand, an interpretation is settled enough within the agency that the challenger can overcome finality and ripeness problems, the argument that the interpretation is only a work in progress, and thus deserves little deference, loses much of its force.
The foregoing analysis applies primarily to regulatory schemes in which the statute is implemented exclusively by the agency in the first instance, and the court's only function is to review the agency's decisions. The article concludes by examining administrative schemes in which the agency shares implementation authority with courts or other entities. In these latter situations (one of which underlay the Skidmore case itself), the Skidmore test can appropriately play a substantial role.
Suggested Citation: Suggested Citation