Kent Barnett & Christopher J. Walker, Chevron in the Circuit Courts
, 115 Mich. L. Rev. (forthcoming 2017), available at SSRN
Kent Barnett and Chris Walker begin this fascinating article by describing the Chevron doctrine and its history. In its landmark 1984 opinion in Chevron v. NRDC, the Supreme Court announced a new, seemingly more deferential doctrine that it instructed lower courts to apply when they review agency interpretations of the statutes they administer. The Chevron opinion is one of the most cited opinions in history. It has been cited in “nearly 15,000 judicial decisions and in over 17,000 law review articles and other secondary sources.” (P. 2.)
Barnett and Walker agree with most scholars that the Supreme Court’s “choice to apply Chevron deference, as opposed to a less-deferential doctrine or no deference at all, does not seem to affect the outcome of the case.” (P. 4.) They note that the Supreme Court did not even mention Chevron in three-quarters of the cases in which it reviewed agency statutory interpretations during the twenty-two-year period immediately after it issued its opinion in Chevron. They then report the findings of their study—the largest empirical study of circuit court applications of Chevron ever undertaken. As they characterize the results of their study, what they call Chevron Regular seems quite different from Chevron Supreme.
Barnett and Walker read, analyzed, and coded 1330 opinions issued by circuit courts between 2003 and 2013. Their dozens of findings are surprising in many ways. I will discuss just the five that I found most surprising. First, “agency statutory interpretations were significantly more likely to prevail under Chevron deference (77.3%) than Skidmore deference (56.0%) or, especially, de novo review (38.5%).” (P. 5) (footnote omitted). Second, circuit courts upheld agency interpretations more frequently when they applied Chevron to interpretations adopted through informal means (78.4%) than to interpretations adopted in notice and comment rulemakings (74.2%). Third, when circuit courts applied Chevron, they upheld longstanding interpretations far more often (87.6%) than recent interpretations (74.5%) or changed interpretations (65.6%). Fourth, circuit courts varied greatly with respect to the proportion of cases in which they applied Chevron to agency statutory interpretations—from a high of 88.9% for the D.C. Circuit to a low of 60.7% for the Sixth Circuit. Fifth, circuit courts also varied greatly with respect to the proportion of cases in which they upheld agency statutory interpretations, albeit not with a high correlation between their rates of outcomes and their rates of invocation of Chevron. The First Circuit upheld interpretations most frequently (83.1%); the Ninth Circuit upheld interpretations least frequently (65.5%), while the D.C. Circuit was around the middle (72.6%).
Barnett and Walker are appropriately cautious in drawing inferences from their findings. Their findings raise far more questions than they answer. Here are just a few.
First, the findings are a major disappointment to those of us who initially saw in Chevron the potential for greater consistency and predictability in the process of judicial review of agency statutory interpretations. We have long been disappointed with the massive inconsistencies in the Supreme Court’s approach to Chevron, but many of us believed (or at least hoped) that circuit courts were applying Chevron in a relatively consistent and predictable way. We were wrong. Circuit court applications of Chevron are at least as inconsistent, unpredictable, and incoherent as Supreme Court applications of Chevron. Those findings raise the question of whether Chevron can, or should, continue to exist as a review doctrine.
Second, whatever Chevron means in circuit courts, the circuit court version differs from the Supreme Court version in many ways. The glaring inconsistencies between the Supreme Court’s approach to Chevron and the approach (more accurately the approaches) of the circuit courts raise the question whether a doctrine can, or should, survive in circuit courts when it bears no relation to the version of the doctrine that exists in the Supreme Court.
Third, it is impossible to determine whether, or to what extent, circuit courts use Chevron as a review doctrine rather than as a tool in writing opinions. The findings of the Barnett and Walker study are consistent with a legal regime in which courts first decide a case based on some undisclosed reasoning process and then use Chevron as a means of rationalizing that decision. In fact, some of the findings are more consistent with the use of Chevron as an after-the-fact justification for decisions than as a framework for making decisions. Thus, for instance, the finding that circuit courts uphold interpretations adopted through informal means more frequently than interpretations adopted in notice and comment rulemakings if, but only if, the court cites Chevron, suggests to me that circuit courts are first using some other method to decide to uphold an interpretation adopted through an informal means and then citing Chevron to justify that decision.
Fourth, the findings tell us nothing coherent about the views of the circuit courts with respect to the important debate the Supreme Court began when it issued its opinions in Mead, Christensen, and Barnhart—when, if ever, a court should apply the Chevron doctrine and when, if ever, a court should apply the putatively less deferential Skidmore doctrine. The finding that circuit courts uphold agency actions much more frequently when they apply Chevron than when they apply Skidmore suggests that circuit courts agree with the Supreme Court’s characterization of Chevron as more deferential than Skidmore. However, the finding that courts uphold longstanding interpretations far more frequently than recent or changed interpretations when they apply Chevron suggests strongly that courts actually use a decision-making framework like Skidmore and then cite Chevron when that method of decision-making yields a decision to uphold an agency interpretation of a statute. The Skidmore test identifies duration and consistency as important criteria in deciding whether to uphold an agency interpretation, while the Supreme Court has often said that neither duration nor consistency are important factors when a court applies Chevron.
Barnett and Walker have provided an extremely valuable database and set of findings, but they have created a situation in which they, and the rest of us, must do a tremendous amount of work to draw useful inferences from their pathbreaking study.
Editor’s Note: Reviewers choose what to review without input from Section Editors. Jotwell Administrative Law Section Editor Christopher Walker had no role in the editing of this article.
The passing of Justice Antonin Scalia removes from the Supreme Court its most strident modern advocate of the “unitary executive” idea—specifically, the view that Article II’s vesting of law execution power in the President forbids Congress to extend any such authority to individuals or entities not subject to “meaningful presidential control.” Printz v. United States, 521 U.S. 898, 922 (1997). I have long argued that this interpretation cannot be reconciled with our constitutional history. But an insightful, tightly argued new article by Leah Litman, a Harvard Law School Climenko Fellow and Lecturer in Law, demonstrates that this view of the separation of powers can also not be reconciled with the Court’s contemporaneous preemption jurisprudence. Put simply, despite the Court’s occasional pronouncements in separation of powers cases that “Article II requires the President alone to execute federal law,” the “preemption cases suggest that nonexecutive actors may likewise vindicate the public interest in seeing federal law enforced.” (P. 1293-94.)
Professor Litman’s thesis rests on an astute recognition of the relationship in separation of powers jurisprudence between two core ideas. One is the familiar truth that federal law execution is policy-laden at every stage. Implementing federal law entails the exercise of significant discretion, both in legal interpretation, Chevron v. Natural Resources Defense Council, 467 U.S. 837 (1984), and in deciding whether to move forward in individual cases, Heckler v. Chaney, 470 U.S. 821 (1985). Indeed, but for the ubiquitous presence of discretion in federal law execution, the unitary executive ideal would presumably carry very little real-world punch.
The second is that the President’s control of law enforcement discretion is implicated not only in disputes about agency design, e.g., Free Enterprise Fund v. Public Company Accounting Oversight Board, 561 U.S. 477 (2010), but also in cases involving Congress’s decisions to enlist the services of persons outside government in fulfilling government’s law execution function, i.e., cases on citizen standing, such as Lujan v. Defenders of Wildlife, 504 U.S. 555 (1992). The latter cases are important because, barring some contractual relationship, persons outside the federal government are also outside presidential control, thus raising the same problem for unitary executive theory that federal independent administrative agencies pose.
Lujan is the Court’s clearest attempt to root the restrictiveness of Article III standing doctrine in Article II’s vesting of executive power in the President. Finding the environmentalist plaintiffs’ allegations of injury too general and speculative to support standing, the Lujan majority denied them a right to sue even in the face of an explicit citizen suit provision of the Endangered Species Act. The ESA authorizes “any person [to] commence a civil suit on his own behalf . . . to enjoin any person, including the United States and any other governmental instrumentality or agency . . . who is alleged to be in violation of any provision of this chapter.” 16 U.S.C. § 1540(g). Speaking through Justice Scalia, the Court insisted, however, that the category of persons authorized to sue could not constitutionally include plaintiffs asserting no more than the general “citizen’s interest in [the] proper application of the Constitution and laws.” To hold otherwise would ignore that “[v]indicating the public interest (including the public interest in Government observance of the Constitution and laws) is the function of Congress and the Chief Executive.” 504 U.S. at 576. By necessary implication, if Congress wants help via the judicial process in upholding the general public interest in observance of the law—that is, in the policy-laden activity of law execution by government—it can turn only to the executive branch of government, not to private citizens.
Yet as Professor Litman goes on to explain, Congress commonly and with the Supreme Court’s apparent approval does invite voluntary help from outside the federal executive branch in enforcing federal law, namely, from the states. Among the prominent examples she cites is the federal Immigration Reform and Control Act, which authorizes states to enforce IRCA’s prohibitions through “licensing and similar laws.” 8 U.S.C. § 1324a(h)(2). In Chamber of Commerce of the United States v. Whiting, 563 U.S. 582 (2011), the Supreme Court upheld an Arizona statute that permitted state officials to revoke state-issued business licenses where an entity had violated federal law. As summarized by Professor Litman:
The Court reasoned that although IRCA expressly preempted “any State or local law imposing civil or criminal sanctions . . . upon those who employ . . . unauthorized aliens,” Congress specifically authorized [state laws like Arizona’s]. Congress therefore intended to “preserve the ability of the States to impose their own sanctions through licensing” for violations of federal law. Whiting noted approvingly that Arizona had adopted the federal definition for unauthorized persons and relied on the federal government’s determination of who qualifies as an unauthorized person, thereby eliminating the possibility of “conflict . . . either at the investigatory or adjudicatory stage.” The Court explained, “Congress . . . . in IRCA . . . ban[ned] [the] hiring [of] unauthorized aliens, and the state law here simply seeks to enforce that ban.” (P. 1309-10.)
She goes on cite other significant statutes, including the Dodd-Frank Wall Street Reform and Consumer Protection Act, which specifically authorize states and state attorneys general to bring civil actions to enforce federal law against private individuals, even without a showing of particularized harm to the state itself. Why Congress may constitutionally invite voluntary state efforts to help enforce federal law, but not the voluntary efforts of citizen litigants is a puzzle the Court never addresses.
Professor Litman anticipates and convincingly refutes a number of possible doctrinal ripostes to her thesis, but one is especially important. The preemption cases do not impose any constitutional requirement that the states respect presidential policy preferences in enforcing the laws that Congress has invited states to help execute. In other words, the fact that IRCA works to eliminate “conflict [between state and federal authority] either at the investigatory or adjudicatory stage” does not insure the absence of conflict at the enforcement stage. Nothing in IRCA or the Arizona law in question precludes Arizona’s enforcement of federal law against a business that the federal executive branch has independently decided not to pursue.
Justice Scalia, without elaboration, joined Chief Justice Roberts’s opinion upholding the business licensing law at issue in Whiting. His apparent indifference in Whiting to the potential conflict between state and presidential prosecution policies seems not to have been inadvertent. A year after Whiting, Justice Scalia dissented from those parts of a different ruling which invalidated portions of another Arizona immigration law, S.B. 1070, Arizona v. United States, 132 S. Ct. 2492 (2012). Among the provisions that the Court invalidated was the creation of a state misdemeanor for noncitizens who failed to comply with federal alien registration law. The Court deemed the law preempted because Congress had entirely occupied the field of noncitizen registration regulation. It further observed with disapproval that, were the statute upheld, Arizona “would have the power to bring criminal charges against individuals for violating a federal law even in circumstances where federal officials in charge of the comprehensive scheme determine that prosecution would frustrate federal policies.” Id. at 2503. Justice Scalia was scornful of the argument: “[T]o say, as the Court does, that Arizona contradicts federal law by enforcing applications of the Immigration Act that the President declines to enforce boggles the mind,” Id. at 2521 (Scalia, J., dissenting). So much for any unitary policy control over federal law enforcement that Justice Scalia might otherwise have attributed to Article II.
Professor Litman’s thorough analysis goes on to catalogue benefits that may flow from state involvement in the execution of federal law, notwithstanding the absence of “meaningful presidential control.” The bottom line seems to be that the Supreme Court regards it as within the discretion of Congress to weigh the advantages and disadvantages of letting states in on federal law execution. If Congress is constitutionally entitled to conclude that state involvement in vindicating the general public interest in law enforcement is a good thing, it is not obvious why reaching the same judgment as to citizen suits is constitutionally problematic. The basis for restrictive standing does not lie in Article II, which does not demand a unitary executive when it comes to taking care of federal law.
Cass R. Sunstein and Adrian Vermeule, The Unbearable Rightness of
Auer, U. Chi. L. Rev.
(forthcoming 2016), available at SSRN
In 1945 the Supreme Court decided the case of Bowles v. Seminole Rock & Sand Co., in which it stated without citation to precedent or other explanation that, when the meaning of the words in an agency’s regulation are in doubt, “the administrative interpretation . . . becomes of controlling weight unless it is plainly erroneous or inconsistent with the regulation.” Over the years, this language has been often quoted by the Supreme Court, including in 1997 by Justice Antonin Scalia in Auer v. Robbins. Subsequently, courts and commentators have usually referred to this doctrine as Auer deference, and until recently the doctrine generally occasioned little discussion in the courts except in some cases where there was a suggestion of a possible exception from the doctrine when the regulation in question was itself hopelessly vague. But recently, there has been a frontal attack on the Auer doctrine led by the late Justice Scalia and Justice Thomas and apparently viewed sympathetically by Justice Alito and the Chief Justice. Moreover, leaders in the House and Senate have introduced a bill essentially to overrule Auer.
Now come Professors Sunstein and Vermeule in The Unbearable Rightness of Auer to take up the cudgel in defense of Auer. Their article is the starting point for any further discussion of the Auer doctrine.
The first intimation of a problem with Auer occurred in 2011, when Justice Scalia wrote a concurring opinion in Talk America, Inc. v. Michigan Bell Telephone Co., in which he admitted to having uncritically accepted the doctrine in the past, but he indicated that, while in the case before him no one had asked for reconsideration of the doctrine, in the future he would “be receptive to doing so.” He summarized the structural constitutional arguments against the doctrine contained in Professor John F. Manning’s 1996 law review article, Constitutional Structure and Judicial Deference to Agency Interpretations of Agency Rules. Talk America was quickly followed by Christopher v. SmithKline Beecham Corp., Decker v. Northwest Environmental Defense Center, and Perez v. Mortgage Bankers Ass’n. In SmithKline, the Court refused to accord Auer deference to an agency’s interpretation of its regulation when first applied in an enforcement action, citing both to the Manning article and a more recent article by Professor Matthew Stephenson and Miri Pogoriler, Seminole Rock‘s Domain, suggesting limits to when Auer deference should apply.
In Decker, Justice Scalia dissented from the Court’s deference to the EPA’s interpretation of its own regulation, taking up the invitation by the respondent to reconsider the Auer doctrine and concluding that it should be overruled. The Chief Justice and Justice Alito noted that, while the respondents had in one sentence in a footnote invited the Court to reconsider Auer, the issue had not been briefed or argued. Consequently, although they believed Justice Scalia had raised “serious questions about the [Auer doctrine],” they said they would await a case when the issue was fairly raised.
Finally, in Mortgage Bankers, Justice Scalia, concurring, again indicated his desire to overrule the Auer doctrine, but now his position was joined by Justice Thomas, who wrote separately to explain his constitutional bases for objecting to the doctrine. Justice Alito also wrote separately to indicate his concern with the doctrine, but again he said that he would “await a case in which the validity of Seminole Rock may be explored through full briefing and argument.” And last month, leaders in the House and Senate introduced the Separation of Powers Restoration Act, which would require courts in reviewing agency action to interpret laws and regulations de novo, without showing any deference at all. In short, Auer has been under relentless attack recently.
Against this attack, Professors Sunstein and Vermeule provide a response. They begin by positing that there are three possible responses to an ambiguous regulation: first, the agency’s interpretation prevails, with certain exceptions; second, judges resolve the ambiguity without any deference to the agency’s view; and third, in the face of ambiguity the private sector is allowed to do what it wants. The first response is the current Auer doctrine. The second is what Justices Thomas and Scalia have suggested and what the Separation of Powers Restoration Act would require. The third, the authors point out, no one has suggested. It would be sort of a rule of lenity applicable to agency regulations. The third option is something of a strawman, which the authors put away rather quickly, noting that ambiguity does not mean authorization. There might be another option the authors don’t mention, which I will discuss later.
The authors take us over familiar ground with respect to Chevron v. NRDC, citing and quoting from a law review article by former Justice Scalia. First, Justice Scalia saw no conflict between affording Chevron deference and the text of the Administrative Procedure Act, which states that a “reviewing court … shall interpret constitutional and statutory provisions, and determine the meaning or applicability of the terms of agency action.” He found it self-evident that Congress could explicitly delegate to an agency the determination of the meaning of a statutory provision. Then, it would be for the courts to determine whether the agency acted reasonably within that delegation, not to determine for themselves the meaning of the provision. When a law is ambiguous, however, it is either because Congress had a particular intent, but failed to express it, or Congress had no particular intent, but meant to leave its resolution to the agency. While the former situation would call for judicial resolution of the ambiguity, the latter would be an implicit delegation to the agency.
The problem was how to tell what type of ambiguity was involved in a particular case. Justice Scalia believed this case-by-case resolution was “a font of uncertainty and litigation.” He believed the better approach was Chevron, which “represents merely a fictional, presumed intent [to confer discretion on the agency], and operates principally as a background rule of law against which Congress can legislate.” Even if this fiction was not 100% accurate, which the prior case-by-case approach certainly had not been, at least it provided a relatively clear rule, relatively easily applied. Antonin Scalia, Judicial Deference to Administrative Interpretations of Law, 1989 Duke L.J. 511, 516-17. Sunstein and Vermeule agree with Justice Scalia on all these points, “so long as it is understood that the choice of fiction depends on the consequences of adopting one or another.” In other words, if one is to adopt a legal fiction, it should be done in a matter that maximizes benefits (or minimizes costs) compared to the alternative. Justice Scalia, at least in 1989 and for years afterward, believed Chevron was indeed that best alternative.
In addition, Justice Scalia saw no conflict between Chevron and the separation of powers generally, because Congress was always in control. It could eliminate the doctrine by statute if it wished, and it was always able to make clear its intent if an agency, upheld by a court applying Chevron, adopted an interpretation not to its liking.
In light of this general agreement with Chevron, Professors Sunstein and Vermeule ask: Why isn’t Auer deference equally justified? First, it seems self-evident that Congress could explicitly authorize an agency to issue interpretations of its regulations with the force of law. Indeed, it does so occasionally without noticeable consternation. See, e.g., 12 U.S.C. § 2617(b). As with Chevron, however, usually Congress does not explicitly delegate that authority, but then, also like Chevron, it has not explicitly withheld that authority. So, why can’t Auer deference represent just another fictional, presumed intent? Isn’t Auer also just the better alternative to independent judicial interpretation?
And the advantages of Auer deference are clear and no doubt explain why the doctrine has for so long been unquestioned. The first is the “agency’s comparative epistemic advantage as interpreter.” That is, the agency is in the best position to know what the agency’s intent was underlying the legislative rule. The second is the agency’s comparative advantage as policymaker. If the ambiguity is such that it requires an exercise of policymaking to some degree, then just as with Chevron, the agency with its at least tangential political responsiveness is better placed to make policy judgments than the courts.
The authors then turn to the bases for the recent critiques. The first is that Auer creates an incentive for agencies to adopt legislative rules that are vague and broad so that the agencies may retain the flexibility to interpret them in the future, even with retrospective effect. The authors acknowledge this concern as theoretically valid, but it strikes them “as a phantasmal terror.” They say that no one has provided an example in the history of American regulation where an agency has designed a vague and broad regulation because of the advantages provided by Auer deference. Professor Sunstein notes that in his four years as director of the Office of Information and Regulatory Affairs, in which he dealt with well over two thousand rules, “he never heard even a single person suggest, or come close to suggesting, that a regulation should be written ambiguously in light of Auer.” I might add that in my three-plus years as the lawyer responsible for an agency’s regulations, albeit more than 30 years ago, I can say the same. This concern, the authors say, “is a reflection of a pervasive error within the economic analysis of law, which is to identify the likely sign of an effect and then to declare victory, without examining its magnitude – without asking whether it is realistic to think that the effect will be significant.” The authors dub this error “the sign fallacy.”
The other basis for the recent critique of Auer involves a question of the separation of powers. Basically, Auer “produces a constitutionally suspect combination of power to make law with the power to interpret law.” In Justice Scalia’s words, concurring in part and dissenting in part in Decker, “He who writes a law must not adjudge its violation.” In this way he can distinguish Chevron, where Congress enacts and agencies adjudge the violation, compared to Auer, where the agency adopts a regulation and then adjudges its violation. The authors spend some time on this critique, but their argument boils down to the fact that such a critique of Auer is a critique of almost all administrative law, which epitomizes the combination of functions the separation of powers is supposed to keep apart. Large numbers of agencies write binding rules, bring enforcement actions, and adjudicate violations. It is to combat that combination of functions that administrative law erects procedural requirements to mitigate the combination and provides judicial review of the administrative results. The authors recognize that Justice Clarence Thomas’s recent adoption of Professor Philip Hamburger’s thesis would indeed sweep away most of the modern administrative state, but aside from Justice Thomas they see little inclination or reason to overthrow a century of administrative law. In other words, Auer is a very little tail on the dog of a combination of functions and should not overcome its agreed-upon and identified real benefits.
Professors Sunstein and Vermeule end up stressing that Auer deference, no more than Chevron deference, is not an abandonment of judicial review. There is still a real check on agency interpretations of its own regulations. First, as Justice Scalia wrote with respect to Chevron, where the lawmaker has made a clear line, the agency cannot go beyond it in interpreting regulations, and where the lawmaker has left an ambiguous line, the agency cannot go further than the ambiguity will allow. See City of Arlington v. FCC, 133 S. Ct. 1863, 1874 (2013). The same applies to Auer deference. Moreover, as the authors point out, “the Court emphasized in Perez v. Mortgage Bankers, the ‘most notable’ constraint on agency decisionmaking is ‘the arbitrary and capricious standard,’ which serves to promote ‘procedural fairness’ by requiring agencies to give good reasons for their procedural choices – and, of course, for their interpretations.” And it is well to reiterate that Auer may not even be applicable when an agency’s regulation merely parrots the statute. See, e.g., Gonzales v. Oregon, 546 U.S. 243, 257 (2006).
All well and good. A frontal attack on Auer in the Supreme Court would not be likely to prevail, even if Justice Scalia were still on the Court. But wait. There still may be some smoldering embers below the smoke. What is missing from the Sunstein and Vermeule article is a different critique of Auer, one that has some empirical basis. An agency faced with the recognition that one of its regulations is ambiguous has options. It may issue an amendment to the regulation, clarifying it. Or it may issue an interpretive rule clarifying the regulation. If the agency is challenged in court, it does not matter which option the agency takes. In either case it will receive strong deference: Chevron deference under the former option, Auer deference under the second option. But it will be infinitely faster and easier for the agency to use the second option. It will avoid the requirements for notice-and-comment rulemaking under the APA and any requirements of the Regulatory Flexibility Act. It will as a practical matter avoid the requirements for regulatory review under E.O. 12866 and the requirements of the Congressional Review Act. Unlike the incentive to write ambiguous regulations in order to retain flexibility for later interpretation, for which there is no empirical support for agencies acting on that basis, the incentive to avoid notice-and-comment rulemaking altogether is strong and there is a wealth of empirical support for the fact that agencies indeed try to cut corners, especially given the number of cases challenging agency interpretive rules as improperly adopted legislative rules. Does Auer deference play a role? Yes, because there is no cost associated with adopting the clarification as an interpretive rule, rather than as a legislative rule.
What this may suggest is that Auer deference should not be at one with Chevron deference but instead should be equated with Skidmore deference – a lighter deference, one with the power to persuade if not to bind. Agencies then could make a choice between taking the longer route with resulting greater deference, or the shortcut and less deference. The more obviously correct (or inconsequential) the interpretation is, the stronger the case for the shortcut. The more the interpretation raises important policy issues that may be highly contested, the stronger the case might be for notice-and-comment rulemaking. And shouldn’t that be the correct direction for the incentives to work?
Professors Sunstein and Vermeule have provided a valuable service in starting the defense of Auer deference. They have provided convincing rebuttals of the critiques of Auer that have appeared in recent Supreme Court opinions and pointed out the obvious benefits of deference to an agency’s interpretation of its own regulations. Perhaps the only piece missing in the puzzle now is the effect of strong Auer deference on an agency’s choices of whether to clarify regulations by notice-and-comment rulemaking or by interpretive rule.
Cite as: William Funk, Saving Auer
(June 23, 2016) (reviewing Cass R. Sunstein and Adrian Vermeule, The Unbearable Rightness of
Auer, U. Chi. L. Rev.
(forthcoming 2016), available at SSRN), https://adlaw.jotwell.com/saving-auer/
Margaret B. Kwoka, FOIA, Inc., Duke L.J.
(forthcoming 2016), available on SSRN
Congress may be gridlocked on many issues, but both parties are working hard to strengthen the Freedom of Information Act. Motivations differ, of course. According to the New York Times, Republicans are displeased with the State Department’s response to requests for then-Secretary of State Hillary Clinton’s emails while Democrats favor a stronger transparency statute.
Margaret B. Kwoka’s forthcoming article, FOIA, Inc., in the Duke Law Journal already has a place in the policy discussions (and in the NY Times). It should also have a place in research and teaching in Administrative Law. I am a strong proponent of teaching something about FOIA in the core Administrative Law class, focusing on its potential use as an oversight mechanism and as an information tool in the many cases that are excluded by the Federal Rules of Civil Procedure and the presumption of regularity from discovery. I warn students, however, that they should not be swayed by tales of disinfecting sunlight, mentioning briefly old studies about the use of FOIA by private parties to get information about other private parties.
My warning has not been strong enough. From considerable original empirical work, Kwoka supports three important points. First, commercial requesters dominate in FOIA practice. Second, some of these requesters are taking information obtained from FOIA and making a profit selling the information. Third, the fees agencies take in pay only a small fraction of the costs of processing FOIA requests. (Kwoka also argues that these realities have disconcertingly crowded out media requesters—that’s a more contested and less interesting (to me) point, so I do not spend time on it here.)
Empirical work takes time. It takes even more time if you collect your own data. Kwoka requested FOIA logs from nearly two dozen agencies, all of whom had reported more than 1,000 requests in FY 2013. Only six of those agencies provided “complete data in a usable form,” so Kwoka turned to an in-depth examination of the Defense Logistics Agency (4,420), the Environmental Protection Agency (9,737), the Federal Trade Commission (1,538), the Food and Drug Administration (10,167), the National Institutes of Health (1,198), and the Securities and Exchange Commission (12,091). The number of requests to each agency is in parentheses.
From her careful, detailed look at these logs, Kwoka unearths some fascinating insights. To name some important ones:
- Although a smaller FOIA operation, all but four percent of DLA’s requests were categorized as commercial. Day & Day, one information reseller, charges $1800 for an annual subscription to an online database of FOIA documents from DLA (specifically, procurements and contracts).
- Nearly eighty percent of EPA’s requests were submitted by commercial requesters. Compared to other agencies, frequent requesters were, well, less frequent. Only six sources made more than 100 requests and none put in more than 180.
- By contrast to four of the other agencies, only about one-third of FTC requests fell into the commercial category. Over half of those commercial requests (and twenty percent of all requests) came from law firms. Nearly half of the total requests were made by individuals, almost all of whom wanted information about their own consumer complaints to the agency.
- Three-quarters of FDA’s requests came from commercial sources. Contrary to what I had been teaching, the “most frequent requesters are not . . . pharmaceutical companies, but information resellers.” These resellers make good money. FDA News charges $997 for a one-year subscription for FDA Form 483s and $117 for a particular form. The only high-volume pharmaceutical requester, Merck & Co., overwhelmingly asked (more than 80 percent of its 373 requests) about others’ FOIA requests.
- Like the FTC, a little more than one-third of NIH’s requests were labeled commercial. Fifteen percent came from educational institutions.
- SECProbes accounted for 12 percent of all SEC requests. These 2498 requests from SECProbes were labeled as coming from news media but should have been placed in the commercial category (from impressive online sleuthing by Kwoka). If they are removed from the news media pile, only 309 requests remain in that category.
Agencies are not getting reimbursed for this work. For example, according to Kwoka, the cost of FDA’s FOIA operations came to $33.57 million but the agency collected only $327,075 from its commercial requesters, which make up three-quarters of the agency’s workload. It adds up, with the government paying “nearly half a billion dollars on FOIA.”
Kwoka suggests some possible reforms, mainly placing pressure on agencies to affirmatively disclose more information. The DLA, for instance, could run a fuller database of contracts, including bids and bid abstracts. When the NY Times asked Day & Day about this proposal, Vice President John Day was refreshingly honest: “If they did that, a good part of our business would go away. So I think it’s a bad idea.” In addition, the FDA could put Form 483s on-line.
Disclosure about disclosure has critical implications for our teaching and research as well as for public policy. Kwoka’s important study deservedly has already generated attention in the public sphere. It also deserves a close look by scholars. Along with a new study by David Lewis and (one of my former Ph.D. students) Abby Wood, we are benefitting from some very interesting empirical work on a powerful statute.
Dan R. Meagher, The Principle of Legality and a Common Law Bill of Rights—Clear Statement Rules Head Down Under
(2015), available on SSRN
I decided to think outside the box this year with my recommendation, or more accurately, outside of our Country’s academy. About a year ago, an Australian Law Professor Dan Meagher contacted me about presenting his paper to our faculty at Mercer University School of Law. I’m very grateful that he did. Professor Meagher ended up visiting with us for a week this past fall as a visiting scholar. During that time, he provided one of the best development presentations that I have seen. His topic was interesting yet completely outside of most of our expertise. His presentation style was relaxed and fostered the interaction of the entire faculty. Perhaps the relaxing part should not be surprising: Australians are not necessarily known for being uptight. I chose to recommend his article to Jotwell readers because I found the topic interesting, the paper well-written, and the application of the legal doctrine a bit contradictory to the way we do things here in the U.S.
The title of his paper is The Principle of Legality and a Common Law Bill of Rights—Clear Statement Rules Head Down Under. In his article, Professor Meagher traces the evolution of the Australian Courts’ approach to protecting fundamental rights. This evolution is fascinating, controversial, and directly connected to both our Constitution and statutory interpretation principles. This history lesson begins with a simple point: “the Australian Constitution is a redraft of the American Constitution of 1787 with modifications found suitable for the more characteristic British institutions and for Australian conditions.” Our system of a government with separated powers was adopted. Importantly, however, the Australian framers consciously rejected, even deleted from a draft version, the American Bill of Rights. The framers rejected the American approach, believing that common law and a parliamentary form of government offered a superior and more democratic way to protect these rights. Professor Meagher describes the Australian Constitution’s development and the strong role that our Constitution played in the drafting process. That part of the paper should be interesting enough to Administrative Law Scholars who teach this aspect of the Constitution. But the story is much more interesting.
Despite this deliberate rejection of a bill of rights, Australia’s High Court (the equivalent of our Supreme Court) has morphed an old friend, the clear statement rule, to temper and invalidate legislation openly hostile to fundamental rights. This judicial response has been both remarkable and controversial.
Let me provide just one example: immigration. Beginning in the early 2000s, the Australian Government sought to limit and even prevent immigrants arriving in Australia by boat from accessing the courts to seek asylum. The government intercepted the boats at sea, transported those on board to processing centers in small pacific nations, refused to resettle them in Australia, and enacted legislation specifically prohibited legal challenges by these individuals.
Despite the clear legislation, the asylum seekers flooded the Australian courts. The Australian Constitution contains a mandamus/original jurisdiction provision (that arose in response to the U.S. in Marbury v. Madison, 5 U.S. 137 (1803)). That provision provides: “[i]n all matters…in which a writ of mandamus or prohibition or an injunction is sought against an officer of the Commonwealth . . . the High Court shall have original jurisdiction.” The High Court concluded that Parliament could not restrict the Court’s jurisdiction in this area absent unmistakably clear statutory language (a super strong clear statement rule, if you will). Even though Parliament was relatively clear in the legislation that it intended to abrogate the Court’s jurisdiction, the Court required Parliament to be “crystal clear;” a standard that despite its best efforts, Parliament seems unable to reach.
Professor Meagher concludes that as a result of this morphed clear statement rule, the High Court has turned “an historically loose collection of rebuttable presumptions [regarding fundamental rights] . . . into a common law bill of rights that is strongly resistant to legislative encroachment, maybe defiantly so.” The fight between Parliament and the High Court reminds me of a wonderful statutory interpretation piece Professor Hillel Y. Levin wrote. The fictional piece begins with the Supreme Lawmaker, MOTHER, proclaiming that “I am tired of finding popcorn kernels, pretzel crumbs, and pieces of cereal all over the family room. From now on, no food may be eaten outside the kitchen.” Litigation then arose, and the “courts” issue a series of cases culminating in a number of exceptions; to which, MOTHER once again decries:
Over the past few months, I have found empty cups, orange juice stains, milkshake spills, slimy spots of unknown origin, all manner of crumbs, melted chocolate, and icing from cake in the family room. I thought I was clear the first time! And you’ve all had a chance to show me that you could use your common sense and clean up after yourselves. So now let me be clearer: No food, gum, or drink of any kind, on any occasion or in any form, is permitted in the family room. Ever. Seriously. I mean it!
Hillel Y. Levin, Everything I Needed to Know About Statutory Interpretation I Learned by the Time I was Nine, 12 The Green Bag 357 (2009). In the case, Parliament has tried to be clear; the High Court ignores the clarity.
Professor Meagher argues that the Australian courts have applied the clear statement canon not to discern congressional intent, as that canon is arguably used in the U.S., but rather to thwart legislative intent. Reminiscent of Justice Scalia, the High Court has concluded that “legislative intention . . . is a fiction which serves no useful purpose.” Lacey, 242 CLR 573, 592 (2011). But the Court then does something that would surprise even Justice Scalia. The Court suggests that legislative intent is not something that exists independently of judicial interpretation, but rather is the product of the court’s process of construction. The High Court reconceptualized the interpretive duty of judges as one of determining legislative intent as the product of rather than the goal of statutory interpretation.
Professor Meagher concludes his paper by noting that the High Court has transformed the clear statement canon into a principle of legality that acts as a protector of fundamental rights and grounded its new principle in that Country’s constitution. In so doing, the Court has constructed (and then robustly protected from legislative encroachment) a quasi-constitutional common law bill of rights. While he may support the idea that fundamental rights are important, the High Court’s approach “has shaken the very foundations of—and the principles that attend to—the proper judicial role in the construction and application of statutes in a constitutional system of separated powers.”
Ever since courts have recognized the legitimacy of political influence on agency policymaking, scholars have struggled to formulate a model of Administrative Law that describes an appropriate balance between such influence and agency expertise. The current reigning consensus – the Presidential Control Model – fails to satisfy many critics, especially in light of recent Presidential assertions of greater and greater power over the apparatus of administrative government. More recently, the heightened partisanship of federal government has added to concerns that presidential control does not assure that the administrative state is sufficiently responsive to the general polity and the public interest. Thus, it is surprising that up until now few scholars have explicitly analyzed the role of political parties in the operation of the federal administrative state, and none have tried to use the workings of contemporary parties to formulate a normative account of how politics should inform agency policymaking. Political Parties and Presidential Oversight by Michael Livermore takes a large and impressive first step to fill that analytic vacuum.
Livermore begins by reviewing the replacement of the local, patronage-driven party system that existed prior to the Kennedy Administration, with the modern national, professional and programmatically driven party system. He then summarizes arguments that the modern party system, along with candidate-centered politics, will drive Presidential elections towards candidates that implement the policy preferences of the majority or, more precisely, the median voter. Livermore rejects the candidate-centered model because Presidents do not seem to implement unifying policy agendas that reflect the position of the median voter. He therefore reinvigorates a theory of “responsible party government.”
This theory posits that parties provide programmatic platforms that distinguish between their respective presidential candidates. Federal elections become a competition for each program to attract voters. The Party that wins the White House gets four years to implement its programs. The party out of the White House uses its position in Congress and more generally its party platform to criticize current administration programs, with the ultimate goal of convincing the polity to reject the incumbent party’s presidential candidate in the next election. In essence, parties provide competing programs, and help those elected to implement the programs on which they run.
Applying responsible party government to the administrative state, Livermore notes that the President can personally participate in a very small percentage of decisions made by the executive branch, and hence the President depends on loyal and capable appointees. Livermore notes that choosing officials from those active in programmatically driven parties helps assure that the appointee will share the party’s and hence the President’s policy vision. He also notes that party connection with issue networks can help the President choose effective officials. More importantly, the theory of responsible party government provides a sounder justification for the presidential control model than does the median voter hypothesis. Even though the President’s party’s programs usually will not represent the preferences of the median voter, the party whose candidate wins the White House must be given the opportunity to implement its programs, and this is best accomplished by close White House supervision of the apparatus of the administrative state.
Livermore, however, recognizes that the theory of responsible party government can also lead to problems with administrative governance depending on the role parties play in Congress, and the power of the out-of-White-House party in the legislature. One role of parties in Congress is to solve collective action problems created by intraparty disputes among members of Congress. For the out-of-White-House party, leadership limits the party’s legislatively proposed programs to those consistent with the party’s programmatic vision that competes with that of the President. But Livermore notes that such mechanisms work less effectively for administrative oversight, which is more decentralized than the process of enacting legislation. For oversight activities, party dominance will obtain only when there is “sufficiently homogenous party membership together with institutional context that places power in the hands of a leader who is responsive to the median-party member.” When such conditions are not present, the nonincumbent party cannot propose a coherent programmatic alternative to that of the President, making the theory of responsible party governance a less persuasive justification for presidential control.
Livermore also addresses a problem that can arise even when the nonincumbent party does adopt a coherent alternative to the President’s programs. Essentially, partisanship in Congress can lead the out-of-White-House party to do everything in its power to sabotage the President’s programs, even if sabotage is against the best interests of the nation or even of the constituents of the opposition party members of Congress. The nonincumbent party’s proposed programs will look better if the President’s programs are made to fail. Thus, without an alternative positive vision of party influence in Congress, Livermore concludes that “the best that can be done is to limit congressional influence [over agencies] to the absolute minimum.”
Finally, Livermore proposes an alternative normative project, which he calls “responsible party administration,” that “seeks to integrate party government with traditional administrative values [of coherence and vigor, representativeness, and legality].” Livermore’s project is pragmatic: it does not provide an overarching set of principles to implement responsible party administration. Rather, his project seeks to highlight and clarify tradeoffs that must be balanced between competing priorities for the administrative state. He proposes doing so both by improving party responsibility and explicitly vindicating administrative values in the context of particular questions of administrative law, such as structuring judicial review and executive review.
What is most impressive about “Political Parties and Presidential Oversight” is its endeavor to incorporate the structure and role of current political parties into the debate on fundamental questions of administrative law. Livermore does not have all the answers, and the answers he provides are often contingent on current party practices. But, this Article provides an excellent start to this endeavor first by cogently reviewing and analyzing various theories of political parties in government, and then carefully identifying ways in which the theory of responsible party government can illuminate tradeoffs inherent in the administrative state. It is a must read for those who think about how to incorporate politics into the current administrative
Cynthia Barmore, Auer in Action: Deference After Talk America
, 76 Ohio St. L.J.
813 (2015), available at SSRN
Administrative law geeks know that Auer deference has been in trouble. This doctrine, which used to go by the much better name of Seminole Rock deference, instructs courts to defer to an agency’s interpretation of its own rule so long as the interpretation is not plainly erroneous. Its primary supporting intuition is that an agency should be better than anyone else at interpreting a rule that it drafted and implements. During the last five years of his life, Justice Scalia mounted a strong campaign to eliminate this doctrine, which he had come to regard as a terrible affront to separation of powers. Although Justice Scalia is now gone, his critique of Auer retains substantial support on the Court. Justice Thomas agrees with it; Justice Alito has expressed strong sympathy; and the Chief Justice might be on board, too.
But, before rushing off to dump Auer in the ashbin of administrative law history, those who prefer to take their separation of powers with a dash of functionalism might like to know: Just how are courts applying this deference doctrine these days, anyway? Fortunately, Cynthia Barmore has shed considerable light on this question in her article, Auer in Action: Deference after Talk America, which was just published in the Ohio State Law Journal. Her hard work reveals that affirmance rates under Auer have declined in recent years and are in line with the rates for other so-called “deference” doctrines. Courts do not, in short, seem to treat Auer as granting agencies free rein to abuse regulated parties with aggressive (mis)interpretations of their regulations.
Justice Scalia first expressed doubts concerning the validity of Auer deference in his brief concurrence in Talk America, Inc. v. Michigan Bell Telephone Co., 131 S. Ct. 2254, 2266 (2011). The core problem is that it allows an agency both to promulgate and interpret law, and “[w]hen the legislative and executive powers are united in the same person, or in the same body of magistrates, there can be no liberty; because apprehensions may arise, lest the same monarch or senate should enact tyrannical laws, to execute them in a tyrannical manner.” Id. (quoting Montesquieu, Spirit of the Laws bk. XI, ch. 6, pp. 151–152 (O. Piest ed., T. Nugent transl.1949)).
A couple of years later, in Decker v. Northwest Environmental Ctr., Justice Scalia thundered, “[e]nough is enough. For decades, and for no good reason, we have been giving agencies the authority to say what their rules mean, under the harmless-sounding banner of deferring to an agency’s interpretation of its own regulations.” 133 S. Ct. 1326, 1339 (2013) (quotation marks omitted) (concurring in part and dissenting in part). The Chief Justice and Justice Alito conceded that Justice Scalia “raise[d] serious questions” about the validity of Auer deference and that “[i]t may be appropriate to reconsider that principle in an appropriate case.” Id. at 1338 (Roberts, C.J., concurring).
The campaign continued in Perez v. Mortgage Bankers Ass’n, in which Justice Scalia argued that Auer deference violates the APA by giving agencies the power to imbue their interpretations with the force of law. 135 S. Ct. 1199, 1212 (2015) (concurring). Justice Thomas concurred separately and at some length, praising John Locke and Baron de Montesquieu while condemning Auer deference for eroding judicial power and transferring it to the executive. Id. at 1217 (Thomas, J. concurring).
Notably absent from these condemnations of Auer is any empirical discussion of its practical effects. This gap is especially striking given that various empirical studies suggest that deference doctrines in administrative law may not have much impact on the rates at which courts affirm agencies. The affirmance rates for arbitrariness review, substantial evidence review, Chevron deference, and Skidmore deference all hover around 60-70%. David Zaring, Reasonable Agencies, 96 Va. L. Rev. 135, 169 (2009). Admittedly, one study did find that agencies win at a 91% rate under the Auer doctrine at the Supreme Court. William N. Eskridge, Jr. & Lauren E. Baer, The Continuum of Deference: Supreme Court Treatment of Agency Statutory Interpretations from Chevron to Hamdan, 96 Geo. L.J. 1083, 1142 & tbl.15 (2008). This result has lent Auer something of a reputation as a super-potent form of deference. A 2011 study of Auer in the lower courts, however, found affirmance rates of about 76% during 1999-2001 and 2005-2007. Richard J. Pierce, Jr. & Joshua Weiss, An Empirical Study of Judicial Review of Agency Interpretations of Agency Rules, 63 Admin. L. Rev. 515, 519 (2011).
Another notable gap in the conservative justices’ condemnations is any recognition that the Court in recent years has tightened Auer deference, hemming it in with new limits. For instance, in Gonzales v. Oregon the Court declared an “anti-parroting” canon that blocks application of Auer to rules “that do little more than restate the terms of the statute itself.” 546 U.S. 243, 257 (2006). In Christopher v. SmithKline Beecham Corp. the Court held that Auer deference should not apply in cases where the new interpretation would sandbag regulated parties by “unfair surprise.” 132 S. Ct. 2156, 2167 (2012).
The imposition of these new limits made it all the more important to look again at how lower courts are applying Auer. To this end, Cynthia Barmore collected and reviewed 190 opinions in which circuit courts applied Auer deference since issuance of Talk America in 2011. She found, consistent with the Pierce & Weiss study, an overall affirmance rate of about 75%. More intriguingly, she also found that this affirmance rate declined over time as the Court turned the screws on Auer. Between Talk America (2011) and SmithKline (2012), the affirmance rate was 82.3%; between SmithKline (2012) and Decker (2013), the affirmance rate was 74.4%; and between Decker (2013) and the end of 2014, the affirmance rate was 70.6%.
Shucks, that last number looks an awful lot like the affirmance rate that other deference doctrines typically yield, doesn’t it?
Of course, all this data does not interpret itself. On its face, it does, however, suggest that agencies are not, armed with Auer deference, roaming the land abusing regulated parties with unreasonable interpretations of regulations. It also suggests that the lower courts are getting the Supreme Court’s message to be a bit more careful when applying Auer. One should think that data like this should inform the Supreme Court’s decision in whatever test case it takes to determine whether to get rid of Auer based on centuries-old abstractions concerning separation of powers. Cynthia Barmore is to be applauded for gathering and analyzing this data.
Ms. Barmore has lots more to say both descriptively and prescriptively based on her study. To pique your interest, here are a few of her observations: The court that knows agencies and administrative law the best, the D.C. Circuit, has one of the lowest affirmance rates among the circuits when applying Auer—65%. Practice may not make perfect—the Department of Labor and the Bureau of Immigration Affairs were among the agencies that invoked Auer most often, but their affirmance rates were among the lowest at 62% and 61%. And, although the good Baron de Montesquieu, were he still with us, might not care for Auer, lower courts do not seem to mind applying it. In only one case out of the 190 did a court suggest that Auer deference compelled it to accept an interpretation that the court would otherwise reject.
To find out more, by all means check out Cynthia Barmore’s article at the link provided above.
Some readers value an article for logical rigor, some for sound judgment, some for immediate utility, some for originality, and so on into N dimensions. (We may value more than one dimension, of course, but not “all of the above,” because the desirable traits may trade off against one another, at a frontier; no one piece can display all of them simultaneously and to a maximum degree). The peculiar excellence of richness is on display in Administrative War by Tino Cuellar, formerly of Stanford, now molted into a higher form of life as Justice Cuellar of the California Supreme Court. Cuellar recounts the history of the administrative state during the Second World War, and connects it to the surrounding political conflicts and developments in legal theory. There is no single thesis, no one-sentence nugget. Rather we are treated to a kind of legal-historical cornucopia. Cuellar’s story undermines conventional wisdom on a number of critical issues in administrative law. Let me attempt to lay out some of the wealth of interesting points that emerge.
1. The New Deal and the War. Cuellar’s basic narrative recounts the arc of the administrative state just before and during the Second World War. Administrative lawyers, particularly critics of the administrative state, still talk about “the New Deal” as though it were the moment when the Rule of Law gave way to the administrative state (and as though “the New Deal” were all one thing or era, as opposed to a pastiche of movements and developments). Distilling, synthesizing and translating-for-lawyers a library of background literature, Cuellar explains that the war, rather than the New Deal, represented the key “inflection point” in the growth of the administrative state. Furthermore, unlike World War I, which gave rise to a number of more or less temporary bureaucracies, the burgeoning administrative state was cemented into place during and by World War II, and by the odd political consensus that created the Administrative Procedure Act of 1946—a key legitimating mechanism for Leviathan. The pedagogical implication of all this is that the constitutional conflicts of the 1930s, which occupy so much space in public law courses, should at a minimum be supplemented and probably partly displaced by a study of the bureaucratic developments of the war years. Less time on the National Industrial Recovery Act (NIRA), which did not provide an enduring model for the American administrative state; more time on (entities like) the War Production Board (WPB) and the Office of Price Administration (OPA), which did.
2. Ratchets are Contingent. The contrast with World War I finally lays to a peaceful rest that exhausted old meme, the “ratchet effect,” according to which agencies, once created, are immortal. Cuellar persuasively explains that the inflection point around World War II stuck, whereas the earlier one didn’t (at least to the same degree), because of the differing contingencies of the political situation at the two moments, including the imperatives of veteran’s benefits and continued geopolitical competition in the late 1940s. This is not to say that the administrative state is only as sticky as the current political consensus at any given moment; institutions are more resistant than that. But it is also true that institutions created on the spur of the moment need not attain stickiness, just in virtue of their existence. Whether they will do so depends on whether political contingencies set into motion a further process that cements them into place.
3. Flexible Bureaucracy, Sclerotic Markets. There is another trope, according to which bureaucracies are sclerotic and markets adjust rapidly, thanks to the decentralized nature of the price system. In fact private industry was incapable of organizing sufficient war production with sufficient speed, incapable of generating a rational allocation of resources in rapidly changing circumstances. Hence the WPB, OPA and their kin. The American state created and deployed complex new bureaucracies with astonishing rapidity in the war years. Those bureaucracies organized an unprecedented surge in industrial and military production and in logistical capacity, and generated an unprecedented ability to deliver a given quantity of force to a given area at a given time—unprecedented not only for the United States, but in all of human history. (This is also a standing problem for glib claims about the relative efficiency of European bureaucracies). The British economic historian Alan Steele Milward observed that in World War II, political regimes of different types—democratic, fascist, communist—reached consensus on a fundamental proposition of political economy: the ordinary operation of market capitalism was too sclerotic in wartime conditions. Wrote Milward, “[e]verywhere the price mechanism came to be regarded as a method of allocating resources which was too slow and too risky.” (War, Economy and Society 1939-1945 (1977), at 99). Leviathan had such a good war in part because markets couldn’t keep up.
4. Self-imposed Administrative Procedure. Cuellar shows that, well before the Administrative Procedure Act went into force, agencies were subject to a range of legal and procedural constraints. Interestingly, these were self-imposed in many cases. Agencies themselves wrote procedural constraints, and were subject to judicial review on that ground. The threat of review, although constraining ex post in particular cases, only came into being because agencies had bound themselves to a procedural system ex ante. Agencies were willing to impose constraints on themselves because “for senior agency administrators, there was always more at stake than the outcome of any individual case” (1429). Agencies used procedures to obtain information from participating actors, very much including regulated entities and interested groups as well as the public at large. Agencies would self-impose procedures in order to more effectively carry out their missions. Bureaucracy was not only flexible, but adept at generating or acquiring information.
5. The Great Charter of the Administrative State. Cuellar lays out with clarity and subtlety the political dynamics that led to the enactment of the Administrative Procedure Act in 1946. Like other framework statutes, the Act was in effect a plausible second choice for each side, fearing that the alternative would be worse—the first choice of the other side. In Justice Jackson’s formulation, it was in effect a treaty of peace, a compromise “upon which opposing social and political forces have come to rest.” (Wong Yang Sung v. McGrath, 339 U.S. 33 (1950)). That point is broadly familiar, but Cuellar excels at sketching the fine grain of the political calculations and bargaining dynamics. The relevant actors had to consider not only the immediate proposal before them, but the risks of all relevant alternatives as well, and the dynamic risks that success in one period might produce a backlash in a future period that would result in a worse outcome overall. For business for and defenders of free markets, the war reframed the space of alternatives, suggesting that if regulation were blocked, the alternative might not be noninterference, but instead outright government ownership of industry. While “supporters of a robust federal administrative state could see in the APA relatively limited constraints on agencies, as well as a chance to head off more draconian procedural limits on agency power … Republicans and Southern Democrats skeptical of the administrative state could see the bill as a step in the right direction, even if they might have preferred a more drastic change” (1432).
6. An Unstable Equilibrium. Cuellar, whose story more or less ends in the later 1940s, is hardly to be faulted for not continuing the story farther still. If he had, he might perhaps have qualified his account of the constraints placed on agencies by the Administrative Procedure Act. As detailed by an insightful student of the APA’s shifting hydraulics, Antonin Scalia (see especially his “Vermont Yankee: The APA, The D.C. Circuit, and the Supreme Court,” 1978 Supreme Court Review 345), many of the assumptions underpinning the APA, and many of the constraints it assumed would govern agencies, have given way over time. Perhaps the most fundamental constraint—stemming from Crowell v. Benson (1932)—was that courts would declare what the law meant. So when the Court later began to defer to reasonable agency decisions even on questions of law, law’s self-abnegation gathered headway. Vermont Yankee itself squelched judicial development of administrative procedures as a matter of common law; a recent decision, Perez v. Mortgage Bankers (2015), emphatically reaffirms that position. As for the putative constraint of “hard look” review, developed in the 1960s in part to shape and constrain agency rulemaking, there is much less there than meets the eye, as Jacob Gersen and I have detailed.
Thus Cuellar’s conclusion that the procedural constraints of the APA legitimated the newly cemented administrative state, like Dan Ernst’s similar conclusion, is more timebound and placebound than the reader might gather from the four corners of his narrative. But this is an ungrateful and intemperate demand for even more richness than Cuellar already gives us.
Jacob E. Gersen & Matthew C. Stephenson, Over-Accountability
, 6 Journal of Legal Analysis
185 (2014), available at SSRN
Many an administrative law article ends with a simple and appealing recommendation: “just add accountability!” Accountability, along with institutional expertise and democratic legitimacy, is one of the key yardsticks that frames evaluations of the legal rules and institutions of the regulatory state. Why might judicial deference to agency interpretations of statutes be desirable? Because agencies are more politically accountable than courts. Why might privatization be worrisome? Because corporations are less accountable than agencies. Accountability, like motherhood and apple pie, is something we can all safely get behind.
Or is it? In Over-Accountability, Jacob Gersen and Matthew Stephenson look at the downsides of augmenting the accountability of political institutions. Lots of ways exist to add accountability to governmental decision-making: one could have more elections, or concentrate power in a “unitary” executive, or reduce the power of politically unaccountable Article III courts. As the authors point out, these and other such accountability-enhancing moves might actually have a surprising and perverse consequence: they might exacerbate bad behavior by the government.
The basic problem, as Professors Gersen and Stephenson explain it, is a consequence of a simple principal-agent dynamic. Agents (say, the Mayor of New York or the head of the FDA) want to look like good agents to their principals (say, the citizens of New York or the President), because the agents want the principals to retain them as their agents in the future. Under certain conditions, the agent’s desire to appear good may cause the agent to adopt a course of action that will convince the principal that it is a good agent, even if the agent knows that that course of action is not actually in the principal’s best interest. Put another way, an agent will sometimes have the perverse incentive to make decisions that it knows to be contrary to the principal’s best interests in order to avoid resembling a bad agent; because the principal is relatively less informed than the agent, the principal can’t tell the difference.
The authors explain that this basic overaccountability dynamic can manifest in a variety of ways. To signal their competence, agents can pander (take action that “cater[s] excessively” to the public); they can posture (take bold action to look good to the public); or they can be persistent (act stubbornly, so that the public doesn’t perceive them to have made a mistake). To signal their lack of capture, agents can act in way that is populist (in a way that unduly burdens a minority or interest group); conversely, to signal their lack of bias, agents can act in way that is politically correct (in way that markedly illustrates an absence of animus towards some given group). In a lively and clear fashion, the article marries its analysis of formal models of such behavior with concrete examinations of how these varieties of overaccountability might emerge in different legal contexts, from debates over affirmative action to the sentencing behavior of elected state court judges.
The article’s second part, on mitigating overaccountability, is equally interesting. In it, the authors point out that there are three strategies for coping with the phenomenon of overaccountability: (1) do nothing; (2) reduce or eliminate accountability; and (3) add oversight from other agents who might be able to monitor or constrain the primary agent for overaccountability problems. Some of the potential responses they analyze—e.g., relying on judicial review by politically insulated courts—will be familiar moves. But other potential responses—randomly timing elections, for example, or reducing the principal’s ability to know which agent was responsible for a particular choice, or “altering the laws and institutional structures that govern the media and other watchdog organizations” so as to weaken their ability to report on government action—are so facially dystopian (at least to me) that it was delightful to see them drily catalogued as among the potential avenues for getting to a system that has the right quantity of accountability “on net.”
This is one of the great rewards of this paper—it forces one to see what was once familiar in an entirely new way. If you are someone who has made use of the “workhorse concept” (p.186) of accountability (as I have done), Professors Gersen and Stephenson will make you want to look that “workhorse” in the mouth.
Cite as: Mila Sohoni, Too Much of a Good Thing
(February 1, 2016) (reviewing Jacob E. Gersen & Matthew C. Stephenson, Over-Accountability
, 6 Journal of Legal Analysis
185 (2014), available at SSRN), https://adlaw.jotwell.com/too-much-of-a-good-thing/
Connor Raso, Agency Avoidance of Rulemaking Procedures
, 67 Admin. L. Rev.
1 (2015), available at SSRN
It is puzzling. Administrative agencies continue to produce thousands of rules each year in the face of an accumulation of procedural requirements that administrative law scholars say have ossified rulemaking and even led some agencies to retreat from rulemaking altogether.
How can this be? How can federal regulatory output be “rising steadily for decades” notwithstanding procedures that have created a supposedly “confusing labyrinth through which agencies seeking to adopt rules must grope”? As someone who has long been puzzled by the seeming contradiction between expectations and reality, I liked reading Connor Raso’s recent article, Agency Avoidance of Rulemaking Procedures, because it offers a persuasive, even if partial, answer to a core conundrum about rulemaking, along with thoughtfully-analyzed, supportive empirical evidence.
To see how much there is to like about Raso’s study, the reader will need to be familiar with the concept known as rulemaking ossification. For decades, administrative law scholars have lamented ossification and the corresponding loss of an earlier era of simple rulemaking. They have believed that agencies, besieged by procedural steps and intrusive, uncertain judicial review, have come to be nearly as burdened by so-called informal rulemaking as they had previously been with its more formal adjudicatory counterparts.
Rulemaking has purportedly grown slow and cumbersome. And yet, despite the nearly universal acceptance of the view of an ossified regulatory process, empirically-oriented scholars such as Anne Joseph O’Connell, Susan Yackee and Jason Yackee, and me have shown quite clearly that agencies continue to produce a large number of rules even in a supposed era of rulemaking retreat.
Again, how can this be? If agencies have retreated from rulemaking, or even if rulemaking has just slowed down substantially, then this should mean fewer rules, not a continued outpouring. Rulemaking’s puzzle persists. How do we solve it? This is where Raso’s interesting new study comes in. He shows that rulemaking procedures in practice are not as burdensome as they seem because agencies are able to avoid certain procedural steps. In other words, what looks to be layer upon layer of procedures on the books does not equate to heavy procedural burdens in action.
Consider that Raso finds that:
- In more than 92% of federal rulemakings from 1996-2012, agencies did not produce the analyses that the Regulatory Flexibility Act (RFA) requires for rules that will have a “significant economic impact on a substantial number of small entities.” Sixty-two percent of “major rules”—those that by definition should have the most substantial impacts—also lacked these analyses.
- In more than 99% of rulemakings during the same period, agencies did not produce written statements that the Unfunded Mandates Reform Act (UMRA) requires for rules that would impose annual private sector costs greater than $100 million. Even about 80% of all major rules lacked these statements. (“Major rules” are defined as those that would have $100 million in annual economic impacts, which encompass both benefits as well as costs.)
- Agencies even avoided the defining feature of informal rulemaking—its notice and comment requirements under the Administrative Procedure Act (APA)—in nearly 52% of the proceedings completed during this same period.
Such seemingly widespread avoidance behavior arises, according to Raso, because of exceptions and ambiguities in the procedural “requirements” themselves. It is not necessarily the case that agencies are violating the law. On the contrary, by their very terms administrative law statutes often contain exceptions; they are not intended to apply to each and every rule. As already noted, RFA and UMRA only apply to rules with certain kinds of expected impacts (although Raso suggests that agency avoidance occurs even with high-impact rules). The APA also contains express exceptions to its notice-and-comment requirements, such as a catch-all exception for “good cause.” Especially with vague terms like “good cause,” procedural standards and exceptions sometimes admit room for interpretation. Raso suggests that agencies at times exploit ambiguities to claim that they have no obligation to follow a procedure they would rather not follow.
Raso is certainly not the first to call attention to agency avoidance of rulemaking requirements. But with this latest study of his, and with another study of his, he has clearly done more than anyone else to document systematically the extent to which procedural requirements are not applying to—or are at least not being applied to—a substantial number of agency rulemakings.
In this respect, Raso’s contribution may well be viewed as compatible with a recent defense Richard Pierce has made of the ossification thesis. In contrast with earlier scholarship, Pierce argues there is no general ossification of rulemaking; rather, he argues, there exists only an ossification of a subset rulemakings that “raise controversial issues where the stakes are high.”
Pierce’s qualification marks an important shift in nearly twenty years of legal scholarship about rulemaking’s ossification, which from the earliest days appeared to have been about rulemaking generally and not just a subsample of the most difficult rulemakings. For example, Jerry Mashaw and David Harfst, in their widely admired research on rulemaking at the National Highway Traffic Safety Administration (NHTSA), compared NHTSA’s “total rulemaking output” before and after the agency’s loss in the courts.
But we need not dwell on how much of a retreat or concession Pierce appears to have made because, even on its face, it is not clear what a narrowing of ossification’s domain accomplishes in terms of solving the basic puzzle. Agencies, after all, continue not just to produce a large number of rules, but also to produce many rules with great economic impact. The rules adopted in the Obama Administration, for example, have been estimated to generate notably greater costs and benefits than those adopted in the preceding Bush Administration. In the absence of evidence that the economic impacts of regulation have systematically declined, skepticism about ossification, even in its qualified sense, would seem to be justified.
Moreover, if procedurally-induced ossification applies only to rules that, in Pierce’s words, “raise controversial issues where the stakes are high,” then we have to struggle mightily to accept that the “long time” and “extensive commitment of agency resources” that these rules appear to demand—in other words, their ossification—stems from procedural hurdles and judicial review rather than their controversial, high-stakes nature. Heightened controversy and high stakes, after all, matter greatly in slowing down or impeding decisions in any institutional setting, even those lacking in rulemaking’s procedural hurdles and judicial oversight.
Of course, Raso does not sort out how much rulemaking may be delayed by controversy, by procedures, or by an interaction of both—nor does he pretend to. Rather he argues, appropriately, for “a more nuanced empirical analysis of how rulemaking procedures may plausibly contribute to ossification.” This means scholars should not be too quick to overstate the extent to which adding new procedures will seriously impede agencies from regulating.
A central part of Raso’s nuanced account centers on litigation risk, so in this respect his work may seem to resonate well with proponents of the ossification hypothesis. Raso argues that agency avoidance decreases as the threat increases of successful litigation against agencies on procedural grounds. For each of the statutes under examination—the APA, RFA, and UMRA—Raso admirably seeks to go beyond anecdotes and identify the number of reported cases raising issues of procedural compliance as well as tally up the percentage of agency losses in these cases. He concludes that the litigation risk associated with procedural compliance with the APA is “moderate,” while the litigation risks under RFA and UMRA are “little.” These characterizations seem to correspond in roughly inverse fashion to the levels of avoidance Raso reports: 52% of rules do not follow notice and comment under the APA, compared with 92% and 99% of rules lacking analyses called for under the RFA and UMRA, respectively.
Raso’s conclusions are intuitively appealing and his evidence suggestive. Of course, the precise relationship between litigation and avoidance is not necessarily so easy to nail down. Despite data on tens of thousands of rules and hundreds of lawsuits, Raso’s litigation risk theory is essentially tested with only a “sample” size of three, the cases of the APA, RFA, UMRA. Moreover, his characterization of litigation risk as “little” or “moderate” is highly qualitative. Raso does report that courts entertained claims of procedural noncompliance with RFA in 72 cases, which “constituted less than one-third of 1% of the 24,787 finalized rules listed in the Unified Agenda during this period” (.002). But then, the 156 APA procedural avoidance cases in this same period seem extremely “little” too, amounting to less than two-thirds of one percent of all the rules (.006). Raso also tells us that, among this tiny fraction of APA cases, the agency prevailed 67% of the time. Even though qualitatively the courts’ reasoning in the APA cases may have seemed somewhat less predictable, the line separating “little” litigation risks and “moderate” ones appears to be a very fine one indeed.
Virtually any thoughtful and innovative empirical study, like Raso’s is, will raise questions and invite further analysis. Some more puzzles await further exploration. For example, even though Raso refers to a study showing that the EPA and FCC face the highest litigation risk, he reports that these agencies have surprisingly different APA avoidance rates: 15% for the FCC and 45% for EPA. (His treatment of litigation risk also never controls for the underlying potential number of lawsuits, a limiting factor in understanding actual risk, and it relies on those cases that resulted in reported decisions, not also those resulting in settlements.) Moreover, despite the adoption of amendments to the RFA in 1996 that authorized judicial review of agency avoidance of the statute’s procedural requirements, the rate at which most agencies conducted “reg-flex” analyses basically remained constant over time.
Still, what is refreshing about Raso’s approach is the careful, dispassionate way he marshals data to tackle an important question. We could benefit from more research like this throughout the field of administrative law. Raso also offers a valuable lesson about the power of procedures: as with a metal chain, they will only be as strong as their weakest link—or their most flexible exception. In addition to his empirical findings, Raso provides his readers with an especially thoughtful account of the jurisprudential and political challenges in achieving legal precision in procedural exemptions, and he further explains how the ambiguities in many procedural statutes contribute to inconsistent judicial outcomes.
If Raso is right that litigation risk is key to explaining compliance with rulemaking procedures, this should matter greatly in contemporary debates over regulatory reform. Most critics of the House-passed Regulatory Accountability Act (RAA) have fixated on its benefit-cost requirements and the ways it would impose still more formal procedures on an already procedurally-burdened regulatory state. Yet following Raso’s focus on litigation risk, critics of the RAA might instead wish to call more attention to the provision it contains which would impose both the opportunity for immediate judicial review as well as additional procedural obligations whenever an agency invokes an exception to notice-and-comment requirements to issue an interim rule. That provision in the RAA, perhaps more than any other one, could well be what most determines whether the bill, if ever signed into law, would deliver the benefits its advocates promise—or would result in the costs its opponents fear.
Then again, the relationship between procedural change and administrative behavior remains far from well understood. After all, as I already noted, the addition of a statutory provision authorizing judicial review of exceptions to RFA requirements apparently made very little difference in decreasing agency avoidance under that statute. Raso suggests that part of the explanation may stem from case law under the RFA that remains, after nearly two decades, still “lenient, undeveloped, and ambiguous.” We cannot be sure that the RAA, if it ever became law, would not succumb to similar slippage. Political scientist Stuart Shapiro’s research even suggests that procedural regulatory reforms never work as intended.
Overall, Raso’s recent article serves as another reminder that administrative law operates in a highly complex, dynamic organizational and political environment. It really shouldn’t be surprising that things are not always as they should seem, at least not if expectations are based on simplistic conceptions of the behavioral effects of administrative procedures. Perhaps in the end, the real puzzle to be solved is why so many lawmakers, advocates, and scholars continue to place so much confidence in their beliefs about how rulemaking procedures will affect regulatory outcomes, whether for good or for ill.
Cite as: Cary Coglianese, Rulemaking’s Puzzles
(December 16, 2015) (reviewing Connor Raso, Agency Avoidance of Rulemaking Procedures
, 67 Admin. L. Rev.
1 (2015), available at SSRN), https://adlaw.jotwell.com/rulemakings-puzzles/