On December 7, the Department of Justice held an unusual “summit” on reforming the Administrative Procedure Act in its Great Hall. The keynote speaker was Deputy Attorney General Jeffrey A. Rosen. The major theme of its three panels was that it was past time for the APA to be “modernized.” Various suggestions were heard ranging from some “consensus” fixes supported by the ABA Section on Administrative Law and Regulatory Practice, to support for the much more drastic rewrites of the APA’s rulemaking provisions that would be wrought by the Regulatory Accountably Act and REINS Act, each of which was passed by the GOP-controlled House several times but stalled in the Senate. The first of those Acts would amend the APA Section 553’s notice-and-comment provisions to require many additional procedural steps and additional analytical requirements for informal rulemaking. The second would require both houses of Congress to affirmatively vote to approve any major rule issued by an agency—a recipe for stasis whenever we have divided government.
Most panelists, who were largely drawn from the current administration and the Republican legislative staff, continued to express support for such bills, notwithstanding the concerns that have been expressed by the ABA and many scholars that these bills would essentially make rulemaking almost impossible. Indeed I and many other commentators have expressed the view that the many procedural and analytical requirements loaded onto the APA procedures by Congress in other statutes, the courts in their judicial review of rules, and by presidential executive orders have made rulemaking too arduous already. Justice Kavanaugh brought up this concern in oral argument [pp 31-32] in the Kisor v. Willkie case with this question:
Can I get your reaction to the thought that the lower courts have made notice-and-comment rulemaking too difficult through various requirements, requiring detailed explanations, making it hard to change regulations that have gone through notice and comment? Do you have a reaction to that? Because that may be one of the reasons that has pushed [agencies] into the more guidance rather than notice and comment in the first place.
Given these debates over major rewrites of the APA, it may seem rather picayune to focus on one other current that ran through the DOJ meeting—the view that it is time to return to a greater use of formal rulemaking, i.e., use of the adjudicative procedures in Sections 556-57 of the APA to issue rules. To be sure, the APA does allow for this in section 553(c), which provides that “[w]hen rules are required by statute to be made on the record after opportunity for an agency hearing, sections 556 and 557 of this title apply instead of this subsection.”
When the APA was enacted in 1946, there were a fair number of statutes that required formal rulemaking, but after a 1972 study of the FDA’s protracted formal rulemakings on food standards by Professor William Hamilton, the Administrative Conference (ACUS) stated that it “emphatically believes that trial-type procedures should never be required for rulemaking except to resolve issues of specific fact.”
That next year, in, United States v. Florida East Coast Ry. Co, 410 U.S. 224 (1973), the Supreme Court made clear that a rulemaking statute calling for a “hearing” did not trigger sections 556 and 557 unless it also prescribed that the hearing be “on the record.” In 1978 the Court struck another blow against creeping formality in rulemaking in Vermont Yankee Nuclear Power v. NRDC, 435 U.S. 519 (1978), by instructing the lower courts not to require additional procedural requirements beyond those in the APA when they reviewed rules.
However, although Congress eventually began to scrub the U.S. Code of formal rulemaking requirements, later in the 1970s, it did enact a series of major regulatory statutes that required additional procedures in rulemaking. These so-called hybrid rulemaking statutes included those administered by the Federal Trade Commission, Consumer Product Safety Commission, and Occupational Safety and Health Administration. Not surprisingly, these agencies’ began to be criticized for their inability to issue regulations in a timely fashion. One of these statutes, section 202(d) of the Magnuson-Moss Warranty—Federal Trade Commission Improvement Act, also mandated an ACUS study of the procedure. An ACUS recommendation growing out of the resulting study took a very dim view of the how the process had worked out, saying it had “not proved to be effective in controlling the agency’s discretion,” and reiterating that “Congress should not ordinarily require, for agency rulemaking, procedures in addition to those specified by §553 of the [APA], although the agencies should have the discretion to utilize them.” Few hybrid rulemaking statutes have been enacted since then.
By 1994, when ACUS issued a comprehensive recommendation on rulemaking procedures, it cautioned Congress:
Specifically, we recommend that it refrain, as it generally has done since the 1970s, from imposing program-specific rulemaking requirements that go beyond the APA’s basic notice-and-comment procedures. Statutory “on-the-record” and “hybrid” rulemaking provisions that require adjudicative fact-finding techniques such as cross-examination, or more stringent provisions for judicial review (in particular, use of the “substantial evidence” test instead of the normal “arbitrary and capricious” test), can be unnecessarily burdensome or confusing and should be repealed.
So it was rather surprising when two attempts to rehabilitate formal rulemaking occurred place out of the blue. The first was buried in President Bush II’s amendments to President Clinton’s Executive Order 12,866 on Regulatory Planning and Review. In his Executive Order 13,422, President Bush made several amendments to the Clinton order, including that “In consultation with OIRA, each agency may also consider whether to utilize formal rule-making procedures under 5 U.S.C. 556 and 557 for the resolution of complex determinations.” At the time of this Order, Jeffrey Rosen was general counsel and senior policy advisor at OMB, and surely had an influential role in this Executive Order. But because President Obama quickly reverted back to the original Clinton order, this change had a short shelf life.
The next revival was in the form of a law review article, In Defense of Formal Rulemaking, 75 Ohio St. L.J. 237 (2014) by Professor Aaron Nielson, a frequent perceptive blogger on this site and a self-described contrarian. Professor Nielson, who was a prominent panelist at the DOJ event, confided that he had noticed the paucity of articles in support of formal rulemaking and had decided to fill that void. At the DOJ, he commented that because of the lack of “on-the-record” rulemaking statutes, we don’t have much recent empirical data about its alleged complexity and delays, and, therefore, some pilot projects or experimentation in formal rulemaking might be salutary.
The idea of resurrecting formal rulemaking has been criticized by the ABA Administrative Law Section and some leading scholars. See e.g., William Funk, Requiring Formal Rulemaking is a Thinly Veiled Attempt to Halt Regulation, Reg. Rev. (May 18, 2017) and Richard J. Pierce, Jr., A Good Effort, With One Glaring Flaw, Reg. Rev. (May 8, 2017). At the DOJ meeting, Professor Ronald Levin commented that if the summit’s goal was to pursue modernization of the APA, it hardly made sense to revive a procedure that has essentially been dead for almost 50 years.
As for Professor Nielson’s comment that we lack relevant empirical data, I would point to the experience of the Federal Trade Commission under its Magnuson-Moss Act rulemaking statute—which is perhaps the closest existing analogue to APA formal rulemaking. As I attempted to show in my own article, It’s Time to Remove the “Mossified” Procedures for FTC Rulemaking, 83 Geo. Wash. L. Rev. 1979 (2015), it is instructive to compare the FTC’s experience in making rules under Magnuson-Moss with its experience under special statutes allowing it to use the APA’s informal procedures in particular rulemakings.
As detailed in my article, the Magnuson-Moss procedures include about 20 procedures and analysis requirements not found in §553 of the APA, (but which are similar to many of the provisions in the proposed Regulatory Accountability Act). According to my research, before the Magnuson-Moss procedures had been enacted, the FTC was able to issue trade regulation rules in 2.94 years, on average. After Magnuson-Moss, it took the agency 5.57 years, on average, to issue the seven rules it managed to issue using these formalized procedures (and the agency gave up on other rulemakings subject to those procedures after even longer periods).
Tellingly it completely abandoned the idea of issuing additional new rules under these procedures only a few years after it began trying to implement the Act. It still occasionally had to amend some existing rules using the Magnuson-Moss Procedures, however, and the nine amendments also took, on average, about the same time (5.26 years) per rule.
But when given the chance by Congress to use regular APA procedures to issue a dozen rules between 1993 and 2009, the FTC was able to do so in an average of 287.25 days.
To me, this was clearly a failed experiment (though it is still on the books). We don’t need to revive it under another name.
Jeffrey S. Lubbers is Professor of Practice in Administrative Law at Washington College of Law, American University.