Separation of Powers and the Middle Way

Nov 11, 2011 Jack Beermann

Jack Beermann

Composer Arnold Schoenberg famously once quipped that “the middle way is the one that surely does not lead to Rome.”  The idea behind this thought, I gather, is that intellectual compromise does not lead to the truth.  John Manning’s recently published article, Separation of Powers as Ordinary Interpretation, 124 Harv. L. Rev. 1940 (2011), proves Schoenberg’s principle wrong, at least with regard to separation of powers.  In this article, Manning, the Bruce Bromley Professor of Law at Harvard Law School, persuasively demonstrates that neither extreme in current debates about separation of powers is correct, and that a true understanding of separation of powers in the United States requires a more nuanced view of the subject than either extreme is willing to undertake.  In my view, Manning’s article is the best published American law review article about separation of powers.  It states a coherent theory of separation of powers clearly and elegantly, and it explains, just as clearly and elegantly, exactly why separation of powers extremists on both sides are wrong.  The only problem I have with the article is that at the time I first read it in draft, I was working on my own separation of powers article, and Manning stole, improved and expanded upon much of my thunder.

There is too much of value in this article to capture in a brief review, so I focus on two aspects, namely Manning’s disagreement with the extremes in separation of powers and the middle way that Manning charts, based on his view that separation of powers in the United States embodies a constitutional compromise.  Along the way, I explain what Manning means by “ordinary interpretation” and how that differs from the interpretive methods employed by separation of powers extremists on both sides.

Manning’s article confronts and refutes the two extremes we find today in separation of powers analysis.  On one side is the formalist insistence on strict separation of powers doctrine that reads each of the Constitutions three Vesting Clauses as assigning a defined sphere of governmental power to a particular branch and applies a broad “strict separation” principle when a particular procedural or structural provision does not apply.  This methodology is usually associated with expansive views of the President’s power to execute the law free from congressionally imposed restrictions on appointment, removal and direction of federal officials and is exemplified by Justice Scalia’s dissent in Morrison v. Olson and my colleague Gary Lawson’s view, published in an important article written together with Patricia Granger, that the word “proper” in the Necessary and Proper Clause imports strict separation norms into the Constitution.

On the other side lies the functionalist approach that eschews strict application of structural constitutional limits in favor of a doctrine designed to “ensure that Congress has respected a broad background purpose to establish and maintain a rough balance or creative tension among the branches.”  Ironically, the functionalists also place primary reliance on the Necessary and Proper Clause, which they read to grant Congress wide ranging authority to rearrange government institutions and reallocate power in pursuit of legitimate federal goals.

The heart of Manning’s critique of separation of powers extremism is his demonstration that both formalism and functionalism depend on overarching theories of separation of powers that are not actually embodied in the Constitution.  This is where the concept of “ordinary interpretation” enters the picture.  As Manning explains, in recent decades, the Supreme Court has shied away from broad purpose-based statutory interpretation methodologies.  This is based on the recognition that statutes are the product of compromises that make it unrealistic to assume that the legislature as a whole would favor non-textual applications that advance a court’s view of the statute’s underlying purposes.  Manning imports this understanding into the realm of constitutional separation of powers principles.  As Manning explains:

the intricate detail within the constitutional structure in fact reflects the fruits of quite particular compromises over how to divide and structure the various powers.  No overarching theory of separation of powers can explain the document’s many elaborately specified procedures. (P. 1973)

Manning supports this view by pointing out that, contrary to the practice in most state constitutions, the Framers chose not to include a general separation of powers clause in the Constitution.  In Manning’s view, this is consistent with the way the Framers thought about it—separation of powers under the United States Constitution is the result of faithful application of the document’s numerous procedural and structural provisions and not based on an overarching theory.  The Framers did not conceive of separation of powers in a more absolutist way because, as Manning explains, their design involved a careful compromise between separation and checks and balances.

What about the extremists’ reliance on the Necessary and Proper Clause?  Manning answers both arguments the same way—reading the Necessary and Proper Clause to either require or rule out strict separation assumes that the Framers hid an “elephant in a mouse hole,” concealing a fundamental constitutional principle in obscure language.

Manning’s critique of the extremists foreshadows the middle way that he charts for separation of powers analysis under the U.S. Constitution.  The first principle of separation of powers, with which I wholly agree, is that the Constitution’s highly specific structural and procedural provisions ought to be honored and enforced.  Those are the easy cases.  The hard cases are those that arise outside the purview of any particular clause where resolution has significant ramifications for the distribution of power within the government.  The best example of this is removal of executive officials.  The strict separationists derive, from the Necessary and Proper Clause and the Vesting Clause of Article II, an overarching separation of powers principle that the President should have unlimited power to remove all officials involved in the execution of the law.  Manning doesn’t see it that way but he doesn’t rule out some limits on Congress’s power to restrict removal even though there is no specific constitutional clause that addresses the matter.  The solution, says Manning, should be arrived at through a process of ordinary interpretation, deriving principles from the particular constitutional provisions that grant and shape the relative powers of Congress and the President, rather than by applying an overarching theory of separation of powers.

One great virtue of Manning’s portrayal of separation of powers is that it is much closer to the Supreme Court’s jurisprudence than either of the extremes.  Rarely, if ever, does the Supreme Court act in accordance with an overarching theory of separation of powers that requires either strict separation or pure functionalism.  Rather, separation of powers law under the United States Constitution involves relatively strict application of the Constitution’s procedural and structural provisions (not including the Vesting Clauses), and when no such clause applies, as Manning explains, the preferred decision methodology is to draw structural inferences via a careful consideration of the relationship “between the Vesting Clauses and the more precise clauses that, with them, create the constitutional structure.”  This is “ordinary interpretation.”

I heartily recommend this article to anyone interested in separation of powers or administrative law more generally.  While the middle way may not lead to Rome, it does lead to a better understanding of the structure of government in Washington, D.C., as demonstrated in this excellent article by John Manning.

Coordinating Agencies

Oct 7, 2011 Kathryn Watts1 comment

Kathryn Watts

Areas of fragmented and overlapping delegations of power to administrative agencies are common today. For example, fifteen federal agencies play roles in the American food safety arena. Similarly, twelve different agencies deal with exports, and numerous agencies regulate the financial sector, including the SEC, CFTC, OCC, FHA, FDIC, OTS and the Federal Reserve. In addition, as President Obama recently quipped during a State of the Union Address, we have one agency (the Department of the Interior) that is in charge of salmon while they are in fresh water, but a different one (the Department of Commerce) that handles them when they are in saltwater.

Despite the prevalence of these sorts of overlapping delegations in the regulatory arena, legal scholars generally have approached administrative law through a single-agency lens. In a forthcoming Harvard Law Review article titled Agency Coordination in Shared Regulatory Space, Professors Jody Freeman and Jim Rossi seek to change this picture. Specifically, Professors Freeman and Rossi depart from what they call the “single-agency focus that is so foundational to administrative law” by offering the “first comprehensive discussion in the legal literature of the problem of fragmented and overlapping delegations of power by Congress to administrative agencies.”

Freeman and Rossi’s aim is largely descriptive. They begin by describing various theories that scholars have articulated to explain why the legislative process creates agency coordination problems in the first place, including game-theoretic models that explore potential strategic benefits to Congress of creating overlapping or redundant agency jurisdiction. Then they move on to describe what they see as the primary challenge raised by overlapping jurisdiction:  the problem of “coordination” in a “shared regulatory space.” Freeman and Rossi explain that conceptualizing the challenge as one of “coordination” is preferable to conceptualizing it as a problem of “redundancy.” This is so, they explain, because collapsing or eliminating agencies to reduce duplicative functions would likely produce the same coordination problems in “ever-larger bureaucracies.” In addition, given the scope and complexity of congressional delegations to agencies, Freeman and Rossi suggest that there is an “irreducible minimum” of overlap and fragmentation that calls out for coordination tools and strategies.

The remainder of Freeman and Rossi’s article focuses on how we might mitigate the “stubborn and serious crisis of coordination.” Specifically, they describe how both Congress and the President already have numerous tools they can use to promote coordination. For example, Congress can resort to structural integration of agencies, inter-agency consultation provisions, and congressionally required joint rulemaking to facilitate agency coordination, and the Executive branch can utilize memoranda of understanding, White House oversight, and coordinated rulemaking. The hard question, according to Freeman and Rossi, is determining which of these tools have the most promise in a given setting. In other words, which tools will achieve coordination benefits that will justify the investment of time and resources and the increase in agency decision costs?

In thinking about which coordination tools might offer the most promise, Freeman and Rossi conclude that the more binding, transparent and substantive the tool, the more likely it will be to “control bureaucratic drift and the easier for principals to monitor,” but with the concomitant effect of raising agency decision costs. As a result, different tools might be more or less appropriate in different circumstances. Take joint rulemaking, for example. It will drive up agency costs and will take away agency flexibility by resulting in a durable agency policy choice that can be changed only via an act of Congress or another notice-and-comment rulemaking. Hence, Freeman and Rossi argue that joint rulemaking might be “especially beneficial in instances where agencies are establishing substantive regulatory standards for financial markets or industry, or where certainty and uniformity are more important to a regulatory program than flexibility.” In contrast, inter-agency consultation provisions might be most useful in areas where expertise, data or additional perspectives are needed. 

In the end, in addition to its significant descriptive contributions, Freeman and Rossi’s article stakes out a strong normative commitment in favor of more coordination among agencies. In Freeman and Rossi’s view, more effective coordination among agencies has the potential to “generate valuable expertise and information, improve the quality of agency decision making, harmonize potentially inconsistent approaches, and reduce both public and private transaction costs.” Members of the legislative and executive branches will be well-served by carefully reading Freeman and Rossi’s article and giving serious thought to which tools each branch might use to most effectively achieve the coordination benefits that Freeman and Rossi describe. Scholars too will benefit from following Freeman and Rossi’s lead and giving more sustained attention to the trade-offs of various coordination tools rather than continuing to think of administrative law largely through a single-agency lens.

What Actually Happens in a Rulemaking?

Aug 22, 2011 Richard Pierce

Richard Pierce

Wendy Wagner, Katherine Barnes, and Lisa Peters provide a wealth of data and insights with respect to the rulemaking process in this well-researched and well-written article. They engaged in intensive empirical analysis of the ninety rulemakings in which EPA issued air toxic emission standards. For each of the rulemakings they studied three stages of the decision making process—pre-NPRM, notice and comment period, and period after issuance of a final rule.

I cannot do justice to the authors’ excellent discussion of their methodology and the implications of their findings. I will instead simply recite some of their most important findings in the hope that my summary of some of their findings will induce everyone to read this important article with care.

The authors found that the pre-NPRM period averaged a bit less than four years. During the pre-NPRM period, EPA engaged in an average of 178 contacts per rulemaking—more than twice the average number of comments EPA received after it issued the NPRM. Contacts with industry were 170 times greater than contacts with public interest groups and 10 times greater than contacts with state regulators during the pre-NPRM period.

The period between issuance of the NPRM and issuance of a final rule averaged 18 months. Industry accounted for 81% of the comments filed, while public interest groups accounted for 4% and governments accounted for 7%. The authors found that comments generally led to changes. 83% of the changes made between the proposed rule and the final rule had the effect of weakening the rule. Based on the changes made during the comment process, the authors drew the inference that EPA is influenced more by industry comments than by public interest group comments.

About 70% of the rules were changed after EPA issued a final rule—typically in response to a petition for reconsideration or a petition for review. Public interest groups participated almost as much as industry during this stage of the decision making process.

The authors also found that, while public interest groups did not participate actively or effectively with regard to the substance of the rules during the pre-NPRM period or during the comment period, they played “a forceful role … with regard to the timeline” of many of the rulemakings. Thus, 73% of the rules were issued after a court issued an order with an action deadline at the behest of a public interest group.

In short, public interest groups influence the rulemaking process almost exclusively through their activities in court before and after the agency decision making process. I agree with the authors that this limited indirect role in the agency decision making process reflects the resource constraints of the public interest groups. They lack the resources required to participate effectively in the actual agency decision making process. This insight underlines the significance of the longstanding debate with respect to the standing of public interest groups to enforce public laws in courts. If they are denied access to the courts, public interest groups may have no means through which they can influence the agency decision making process.

Anyone who is interested in the rulemaking process must read this article. It is the most valuable source of data I have found on the subject.

The Judicial Playing Field: Courts as Lawmakers

Jul 11, 2011 Lisa Bressman

Lisa Bressman

To what extent is a court just another lawmaker in our governmental structure? Professor Amnon Lehavi argues that the U.S Supreme Court has given a surprising answer in a surprising place: Stop the Beach Renourishment Inc. v. Florida Dept. of Envtl. Protection. The question in Stop the Beach was whether a decision of the Florida Supreme Court altering state property law deprived property owners of their rights in violation of the Fifth and Fourteenth Amendments. The U.S. Supreme Court held that it did not, but four justices expressly recognized that a judicial decision could constitute an unconstitutional taking.¹ Lehavi contends that the case is significant beyond its ramifications for constitutional property law: four justices would treat state courts as lawmakers, indistinguishable from legislatures in this context. As Justice Scalia wrote: “It would be absurd to allow a State to do by judicial decree what the Takings Clause forbids it to do by legislative fiat . . . the particular state actor is irrelevent.”²

Lehavi considers the implications of this view. For example, Lehavi asks, if state courts are lawmakers, are they entitled to the same sort of judicial deference as legislators and agencies? Specifically, should the U.S. Supreme Court engage in deferential review of their determinations rather than more aggressive de no review? If so, should the same doctrines apply or apply in the same manner? Lehavi observes many of those doctrines are adapted for legislatures or agencies, such as the regulatory taking doctrine. Consider Penn-Central’s three-prong test, which asks a reviewing court to consider: (1) the economic impact of the regulation on the claimant;” (2) “the extent to which the regulation has interfered with distinct investment-backed expectations;” and (3) the character of the governmental action.” Lehavi notes that the third prong is awkward.  In some cases, the Court has examined whether the invasion is part of a broader governmental program adjusting benefits and burdens among citizens. But courts do not maintain such programs. Perhaps, then, substantive due process supplies the proper analysis, asking whether the invasion is effective at achieving some legislative public purpose. Courts have been understandably deferential to legislatures and judges on the means-ends connection. But, Lehavi inquires, should they maintain the same posture for state courts?

In the end, Lehavi argues for a model that sees state courts as recipients of delegated authority. When legislatures and agencies enact broad legal standards or leave open certain terrain within a common law field, courts (through common law adjudication) specify the rules in this area. Returning to Stop the Beach and the question of when a judicial decision altering the common law constitutes an unconstitutional taking, Lehavi argues that the Court should view a suspect judicial decision as flowing from the underlying delegation of authority from the political branches and thus ascribe the motives and policies of the judicial decision to the delegation. As such, the judicial decision is not merely a “judicial wrong” but a governmental wrong, and the Court should assess that wrong in reference to the motives and policies of the original legislative and executive acts. The idea, Lehavi says, is not one of vicarious liability but of situating the judicial decision in its broader, realistic lawmaking context. In Lehavi’s words, “for purposes of identifying the underlying motives, goals, and background, socioeconomic facts, a federal reviewing court could rely on the way in which the institutional collaborative scheme has unfolded over time, while independently reviewing the federal law questions implicated by the development of the common law doctrine.” (P. 40).

Lehavi’s article is part of a broader discussion about the nature of the judicial power in the modern regulatory state. We have long held onto the fiction that courts, particularly Article III courts, do not make law. Chief Justice Roberts endorsed the fiction in his confirmation hearings: “Judges are like umpires. Umpires don’t make the rules; they apply them.” Justice Scalia has been more self-conscious about the fiction, but he has nonetheless adhered to it: Judges “make” the law “as judges make it, which is to say as though they were ‘finding’ it–discerning what the law is, rather than decreeing what it is today changed to, or what it will tomorrow be.” The fiction has been widely criticized by legal realists who acknowledge that judges do make law as well as by political scientists who have provided empirical support for the view of judges as lawmakers, and politically predictable ones at that.

Yet the old umpire uniform still hangs around.  Judges may be political, but we are reluctant to treat courts as lawmakers, subject to the same legal doctrines that govern legislators or agencies. For example, courts implement broadly-worded statutes, such as the Sherman Act or Title VII, but we regard those statutes differently from the ones that agencies implement. As Professor Maggie Lemos has argued, those statutes have not traditionally been subject to the same sort of nondelegation limits as statutes that agencies implement. Nor do we generally view courts as recipients of delegated authority, asking why Congress might design statutes with a preference for courts over agencies as implementing institutions. Moreover, we do not evaluate different judicial practices with a realistic understanding of the judicial power. For example, should stare decisis apply to questions of statutory interpretation in judicially-administered statutes? Should the absurdity doctrine apply? We either take the answers for a given (stare decisis: of course), or we finesse the answers to avoid harsh results (absurdity doctrine: judges should not remake the law except when the law needs to be remade). That is not to deny differences between courts, on the one hand, and legislatures and agencies, on the other. It is to recognize the similarities.

Lehavi shows us that the Court itself may be beginning to acknowledge the similarities, albeit still maintaining a safe distance from the plate. State courts are not federal courts, and the Court has always been more willing to acknowledge that state courts possess the power to make the law by altering the common law. (P. 7). Still, Lehavi shows us, the Court has gone further than ever before. Judicial changes to the common law can have the same legal character as legislative or administrative changes to statutory or regulatory law. The implications are far reaching.

1. Four other justices, in two different concurring opinions (Kennedy, J. and Breyer, J.), left open that determination.
2. Id. at 2602.

The Dilemma of Nonlegislative Rules

Jun 3, 2011 William Funk

William Funk

So-called nonlegislative rules, rules adopted as interpretative rules or statements of policy without notice and comment, have posed problems for courts and scholars for a number of years. In addition to myself,³ in recent years professors Robert Anthony,⁴ Peter Strauss,⁵ Elizabeth Magill,⁶ Nina Mendelson,⁷ Donald Elliott,⁸ Jacob Gersen,⁹ Ronald Levin,¹⁰ and John Manning¹¹ have all attempted to bring coherence to the questions raised by nonlegislative rules.

Everyone agrees that agencies must be able to issue certain interpretations and policy statements, generically guidances, without having to follow the notice-and-comment process applicable to legislative rules. On the other hand, everyone also agrees that agencies can abuse the ability to avoid notice and comment rulemaking through invocation of the exceptions for “interpretative rules” and “general statements of policy.”  How to police the line between those rules requiring notice and comment and those that do not is what has stymied courts and commentators. Now there are two more attempts in this regard, and while both are worthwhile additions to the field, Professor Seidenfeld seems to this author to come closest to hitting the mark.

Nonlegislative rules raise three distinct but interrelated problems.  The first question, and the one that has probably received the greatest amount of attention, is how to determine whether a rule adopted without notice and comment is indeed an interpretative rule or statement of policy, or whether it is instead a legislative rule invalidly promulgated because it was not adopted following notice and comment. The second question is when should a nonlegislative rule be subject to preenforcement judicial review or, stated otherwise, whether a nonlegislative rule can be final and ripe for review. The third question is what deference should a policy or an interpretation of law in a nonlegislative rule receive from a reviewing court.

The questions are related, because how they are answered may greatly affect the incentives for an agency to utilize nonlegislative rules in place of a legislative rule.  An agency almost always has the authority to issue a nonlegislative rule as a legislative rule, but given the existing statutory and non-statutory hurdles to adopting legislative rules and the relative ease with which regulated entities and regulatory beneficiaries can obtain preenforcement judicial review of legislative rules, there is already a bias in favor of using nonlegislative rules rather than legislative rules. If, however, an agency can both receive the same substantial deference to its interpretations and policies contained in nonlegislative rules as it can in legislative rules and avoid preenforcement judicial review of those nonlegislative rules, the agency will have an even greater incentive to substitute nonlegislative rules for legislative rules.

Both Franklin and Seidenfeld relate the history of the controversy and understand the interrelationship of the three questions. Both recognize the problem with the judicial doctrine that if a rule is legally binding it cannot be a nonlegislative rule: “difficult to apply consistently” (Franklin, P. 288) or “so confused that courts and commentators alike describe the doctrine as engulfed in smog.” (Seidenfeld, P. 14.) Franklin, however, sees his task to be the rebuttal of what he terms the “short cut” approach to dealing with this problem. As Franklin describes the short cut:

[R]ather than asking whether a challenged rule was designed to be legally binding in order to determine whether it must undergo notice and comment, courts should simply turn the question inside-out and ask whether the rule has undergone notice and comment in order to determine whether it can be legally binding….  No longer would a rule’s substantive nature dictate its procedural governance; instead, its procedural provenance would determine its substantive effect. (Franklin, P. 279.)

A number of commentators have argued for this approach, suggesting that substantive judicial review of nonlegislative rules, combined with the lesser deference generally accorded nonlegislative rules, provide a sufficient safeguard against agency abuse.

Franklin, however, questions this conclusion. For one, he doubts that agencies merit any difference between the deference received under Chevron or Skidmore as worth avoiding nonlegislative rulemaking. A recent article by Richard Pierce might support that conclusion, in which he surveys a number of empirical studies of judicial review of agency action, finding that “the choice of which doctrine to apply… is not an important determinant of outcomes in the Supreme Court or circuit courts.”¹² Nevertheless, agencies still spend a lot of energy litigating in favor of Chevron versus Skidmore deference when the issue comes before a court. Perhaps more important, Franklin doubts the efficacy of substantive review of nonlegislative rules because often, perhaps overwhelmingly, the nonlegislative rules will escape any substantive review. Because of the hurdles imposed by courts to preenforcement review of nonlegislative rules, they are likely to be reviewed only in enforcement actions, but that assumes that a regulated entity will violate a nonlegislative rule. If the regulated community is sufficiently coerced into following the rule, it will never be tested. Equally important, if the nonlegislative rule favors the regulated community and it is the regulatory beneficiaries who wish to challenge the rule, the absence of preenforcement review will mean no review can ever take place. These gaping holes in the availability of substantive review of nonlegislative rules, Franklin believes, renders the check of substantive review an illusion.

Finally, as Franklin says, the “most fundamental” objection to the short cut is that it substitutes judicial review for public scrutiny and participation in making policy as reflected in the requirement for notice and comment. Franklin places a high value on public participation, but the Administrative Procedure Act only requires notice and comment in certain situations. Franklin seems to want to reach the same end the D.C. Circuit wanted to reach in its pre-Vermont Yankee line of cases requiring notice and comment for nonlegislative rules that had a substantial impact on persons. That is, if persons are to be adversely affected by a nonlegislative rule, they should be able to participate in the formulation of that rule. Of course, the APA’s response to persons adversely affected by agency action is to provide the ability to challenge the lawfulness of that action in court, not to comment on its formulation.

Franklin concludes by approving of the very inconsistency of judicial decisions that he acknowledged in the beginning. This inconsistency means that agencies are never sure how far they can go in making nonlegislative rules; to avoid judicial invalidation they will err on the side of restricting their output of nonlegislative rules and will have a greater incentive to adopt legislative rules that will be insulated from procedural invalidation. This assures greater public participation.

Seidenfeld’s article, although apparently written mostly before Franklin’s article was published, is a direct response to Franklin’s proposed solution – the status quo – as well to proposals made by Elizabeth Magill and Nina Mendelson. Seidenfeld clearly does not value public participation as highly as Franklin. Perhaps as a former regulator Seidenfeld appreciates what the literature overwhelmingly supports: that the formal public participation involved in the notice-and-comment process usually does not in fact have much influence on agency decision-making. Also, perhaps as a former regulator, Seidenfeld values more highly the importance of agencies being able to issue guidance documents without either the procedural hurdles of notice-and-comment rulemaking or the uncertainty of outcome resulting from ad hoc judicial review as to whether the guidance is an invalidly adopted legislative rule.

Seidenfeld’s solution is to adjust judicial doctrine to respond to the practical problems Franklin identifies with using the short cut.  First, he endorses the short cut, that is, the use of substantive judicial review of nonlegislative rules in place of courts attempting to discern the procedural validity of such rules. He recognizes the limitations existing in current doctrine on preenforcement review of nonlegislative rules and would make appropriate changes to that doctrine to enable the substantive review to be a real “check” on potential agency abuse of the use of nonlegislative rules. Perhaps it would be more accurate to say that he suggests courts have been applying current doctrine incorrectly. For example, he points out that courts finding nonlegislative rules non-reviewable because they are not “final agency action” are really just wrong in their application of the finality doctrine, albeit they have some excuse because the Supreme Court has not been clear as to whether the second prong of finality doctrine actually requires the agency action to have binding legal consequences or whether it merely changes the legal landscape. He further argues that ripeness as appropriately applied should not bar review of most guidance documents that adversely affect either regulated entities or regulatory beneficiaries.

Finally, Seidenfeld addresses at some length an issue no one has previously dealt with precisely – how should substantive judicial review of an agency policy statement proceed in the absence of an administrative record to which outside parties had no opportunity to contribute. His answer is simple; courts should simply apply traditional “arbitrary and capricious review,” or stated another way, use a requirement of reasoned decision-making. That is, in essence his proposal would have courts review that agency’s nonlegislative rule on the basis of the information the agency had available to it when it adopted it, supplemented by a requirement that the agency provide a contemporaneous explanation for its action – both of which are standard, current practice for judicial review of agency action. And, Seidenfeld adds, drawing from the Court’s decision in Massachusetts v. EPA,¹³ where the Court was reviewing an agency action without a traditional administrative record, courts could, in assessing whether the agency had “considered the relevant factors,” also take into account what the agency should have known and considered. This might in some degree make up for the lack of third party input to the record.

With these modifications or clarifications to existing doctrine, Seidenfeld suggests the dilemma of nonlegislative rules can be solved.  Whether he is right I expect will lead to further articles; whether the courts will adopt his suggestions I fear is as unlikely as the legislative solutions I proposed years ago.

1. William Funk, Legislating for Nonlegislative Rules, 56 Admin. L. Rev. 1023 (2004), available at SSRN; When is a “Rule” a “Regulation”? Marking a Clear Line Between Nonlegislative Rules and Legislative Rules, 54 Admin. L. Rev. 659 (2002), available at SSRN; A Primer on Nonlegislative Rules, 53 Admin. L. Rev. 1321 (2001), available at SSRN.
2. Robert Anthony, Interpretive Rules, Policy Statements, Guidances, Manuals and the Like –Should Federal Agencies Use Them to Bind the Public?, 41 Duke L.J. 1311, 1372 (1992).
3. Peter L. Strauss, Publication Rules in the Rulemaking Spectrum: Assuring Proper Respect for an Essential Element, 53 Admin L. Rev. 803, 807 (2001), available at SSRN.
4. Elizabeth Magill, Agency Choice of Policy Making Form, 71 U. Chi. L. Rev. 1383 (2004), available at SSRN.
5. Nina Mendelson, Regulatory Beneficiaries and Informal Agency Policymaking, 92 Cornell L. Rev. 398 (2007), available at SSRN.
6. E. Donald Elliott, Re-Inventing Rulemaking, 41 Duke L.J. 1490 (1992).
7. Jacob Gersen, Legislative Rules Revisited, 74 U. Chi. L. Rev. 1705, 1709 (2007).
8. Ronald M. Levin, Nonlegislative Rules and the Administrative Open Mind, 41 Duke L.J. 1497 (1992).
9. John F. Manning, Nonlegislative Rules, 72 Geo. Wash. L. Rev. 893, 926-27 (2004).
10. Richard J. Pierce, Jr.,What Do the Studies of Judicial Review of Agency Action Mean?, 63 Admin L. Rev. 77, 85 (2011), available at SSRN (showing an affirmation rate of 60%-81.3% for Chevron review and of 55.1%-73.5% for Skidmore review).
11. Massachusetts v. EPA, 549 U.S. 497 (2007).

Designing Agency Independence

May 2, 2011 Gillian Metzger

Gillian Metzger

How do we structure an agency to be independent?  Not surprisingly, the answer to that question depends on what we want the agency to be independent from.  The traditional legal view, exemplified most recently by the Supreme Court’s decision in Free Enterprise Fund v. PCAOB, is that Congress intends independent agencies to be independent of the President and it achieves this goal primarily by imposing limiting the President’s power of removal.  Not so fast, say Rachel Barkow, Lisa Bressman, and Robert Thompson.  In two separate recent articles—one written before the Court handed down its decision and one after—these scholars argue that agency independence means both more and less than independence from the President.

Barkow begins her article by arguing that what often has motivated creation of independent agencies is not presidential insulation but fear of agency capture, which she defines as the desire to protect an agency from one-sided political pressure from the well-financed industry interests that the agency regulates.  Barkow then assesses how well traditional indicia of independence—such as removal, multimember heads, bipartisan requirements, and exemption from OIRA regulatory review—help to limit capture.  She concludes that these features provide important insulation but are often not sufficient to create an adequate buffer against one-sided interest group pressure.  Instead, Barkow emphasizes the value of other structural mechanisms that have received less attention in discussions of agency independence:  guaranteed agency funding, substantive expertise requirements and revolving door limits, relationships with other agencies and the states, and an agency’s ability to independently gather and disseminate information, provide congressional testimony, and represent itself in court.  According to Barkow, these insulating features may be particularly helpful in equalizing the pressure that interest groups can otherwise bring to bear.

Bressman and Thompson also critique the traditional emphasis on removal and presidential insulation, but with an alternative aim:  to challenge the binary division between independent and executive agencies.  Focusing on financial regulation, an area where independent agencies are rife, they identify a number of mechanisms by which the President can have direct and formal involvement in policymaking, in particular agency consultation and collaboration requirements of the kind embodied in Dodd-Frank’s structural reforms.  They argue not only that these mechanisms can translate into substantial presidential input into independent agency decisonmaking, but also that such input mechanisms are enough to satisfy political and constitutional demands for presidential control.  Like Barkow, they emphasize a broader view of agency independence, one that focuses on developing expertise and the ability to resist short-term political pressures, and to rebut claims that independence undermines agency accountability.

In challenging traditional views of agency independence, both deepen legal academic views on the impact of institutional design.  They also advance interdisciplinary exchange between law and political science about the forces behind agency structure, acknowledging the central role of politics at the same time as resisting reductionist accounts that view agency independence solely in terms of presidential-congressional struggles for influence.  As interesting is the perspectives these articles bring to bear on recent congressional efforts to balance independence with more direct political accountability in the Dodd-Frank Wall Street Reform and Consumer Protection Act.   Indeed, studying the operation of many of the measures enacted by Dodd-Frank—the new Financial Stability Oversight Council and Consumers’ Bureau, as well as the protections granted state regulation and enforcement—offers an unusual opportunity to test how well theoretical intuitions about agency structure play out in practice.

Strengthening Intelligence Through Administrative Law

Apr 4, 2011 Peter Shane

Peter Shane

Having taught some version of “separation of powers law” since 1982, I think I can say with some certainty that few problems of democratic accountability are more vexing than the general subject of “intelligence oversight.”  For half a century, scandal after scandal has exposed an intelligence apparatus that is too often unreliable and susceptible to gross abuse.

Against this background, one might be forgiven a certain amount of pessimism for the future of reform.  But it is not as if we are lacking for ideas.  Samuel Rascoff’s article, Domesticating Intelligence, 83 S. Cal. L. Rev. 575 (2010), takes an especially thoughtful and creative approach with regard to domestic intelligence gathering, basically urging the application of familiar administrative law principles to achieve both “full compliance with the law, but also intelligence that is accurate, efficient, and useful to policymakers.”  Professor Rascoff’s core argument is that “an expansive approach to cost-benefit analysis that [he refers] to as rationality review, judicial review, and public participation made possible by increased transparency ought to play significant roles in reconfiguring the governance of domestic intelligence.”  Taking administrative law into this unaccustomed domain is an important scholarly contribution.

Professor Rascoff’s article begins by showing how the structures currently in place for preventing intelligence abuse are poorly positioned to do so, and how intelligence gathering operates with little prospect for either judicial review or meaningful day-to-day operational oversight.  But, in a move that ought to be taken as a significant conceptual breakthrough, Professor Rascoff argues that the business of intelligence oversight has also been afflicted by a critical conceptual misdirection.  That is, intelligence gathering has been thought of as something akin to criminal law enforcement, and reform efforts have focused on “on the prevention of illegality and the politicization of” the process.  What Professor Rascoff urges is that domestic intelligence gathering be viewed as a form of quintessential administrative activity – namely, risk assessment – and that a governance regime ought to be developed for this administrative activity that aims not just to prevent abuse, but actually to produce good administration, i.e., good intelligence.

Insofar as Professor Rascoff’s governance regime relies on processes internal to the intelligence community, there is a real kinship between his approach and the “bureaucratic justice” framework within which Professor Mashaw urged us in the 1980s to understand and reform the activity of social security adjudication.  Having myself deployed the Mashaw framework in recommending reforms for one familiar aspect of domestic intelligence – namely, the use of antiterrorist watch lists – I can only applaud Professor Rascoff’s essential insight (beautifully captured in his title).   That is, we may not need to create exotic approaches to intelligence reform if we understand intelligence as part and parcel of the administrative state and apply, to its governance, carefully tailored tools of the sort we use to maintain the legality and effectiveness of other administrative activity.

Although it is impossible in a few words to do justice to Professor Rascoff’s multifaceted argument, I can highlight the core of his approach to each of the three key tools he recommends.  First, he would have the Office of the Director of National Intelligence subject programs of domestic surveillance to a kind of rigorous cost-benefit analysis he calls “rationality review.”  According to Professor Rascoff, “[R]ationality review would help promote more accurate and cost-effective intelligence. Second, and somewhat more controversially, … rationality review may actually prove to be a more effective tool for the protection of basic rights than the current governance regime.   Third, … rationality review will help supply the methodological foundations of a centralized regulatory review process in the intelligence sphere akin to the role that OIRA has come to play in the regulatory state.”

Second, he would have internal rationality review policed by the courts: “At some regular interval after an agency has implemented a particular intelligence program (following successful rationality review), a court should review the agency’s program for fidelity to the agency’s own stated (and previously approved) objectives. In focusing on how the agency has implemented the intelligence program in practice, a court could determine whether, in view of empirical evidence, the actual costs and benefits of the program are roughly in line with those that were anticipated prior to the program’s implementation. Even more basically, the court could determine whether the agency was remaining true to the stated goals and limitations of the program’s mandate.”

Third, he would subject the regulation of intelligence gathering to some form of increased public input.  Professor Rascoff mentions both the interest group vetting of the 2008 Attorney General’s Guidelines for Domestic FBI Operations and the prospect of external expert peer review of intelligence as suggesting possible directions.  The core idea would be to subject the regulation of intelligence to something like the kind of pluralistic evaluation that is made possible in principle by the requirement of a public comment period for most agency regulations.

Serious questions obviously loom about all of this.  I have previously been critical of what I think may already be our overreliance on cost-benefit analysis in more conventional contexts.  Although Professor Rascoff offers many useful insights into how it might be designed in the intelligence context, his framework does not purport to offer a completely clear blueprint. Likewise, with regard to public input, the endorsement in principle falls short of a precise operational prescription.  His discussion of judicial review does not address questions of standing and whether review of the kind he recommends is consistent with the role of Article III courts. And then, looming over all these recommendations is the obvious question: would they actually make a difference in improving agency performance and reducing actual abuse?

Nonetheless, Professor Rascoff’s article frames beautifully what ought to be a productive new discussion about the future of intelligence governance.  Agency efficacy, accountability and fidelity to law are foundational themes in administrative law.  Deploying administrative law insights to help strengthen intelligence gathering and keep it within legal confines is a compelling project that Professor Rascoff has launched with real – I have to say it – intelligence.

Should Review Doctrine Be Simplified or Restated?

Feb 3, 2011 Richard Pierce

Richard Pierce

David Zaring first makes two contributions to the growing empirical literature on judicial review of agency actions and then suggests a dramatic change in doctrine in light of his findings. Based on a study of 226 cases, Zaring found that courts uphold about 70% of agency actions when they apply either the substantial evidence test or the arbitrary and capricious test to agency findings of fact. He then combined his study with over a dozen other empirical studies of judicial review of agency actions to create a meta study of 5081 cases.

In his meta study, Zaring found that courts at all levels uphold about 70% of agency actions no matter what doctrine a court applies. Since choice of review doctrine has no apparent effect on the outcome of a case in which a court reviews an agency action, Zaring argued that courts should simplify review doctrine by replacing the six tests courts now apply with a single simple test—a court should uphold any reasonable agency action.

When I first read Zaring’s article, I found his proposal compelling. It doesn’t make a lot of sense for courts to expend a lot of energy deciding which of six review doctrines to apply in a particular case if a court’s choice of review doctrine has no effect on the outcome of a case. On further reflection, however, I realized that Zaring’s proposed test is incomplete. It is devoid of content unless it is coupled with some set of criteria for determining what makes an agency action reasonable or unreasonable.

If we step back and look at the six review doctrines courts now use in the aggregate, it is apparent that a reviewing court considers three factors in deciding whether to uphold an agency action—(1) the relationship between the action and the statute it purports to implement, (2) the relationship between the action and the available evidence that is relevant to the action; and, (3) the quality of the reasoning the agency uses to link its action to the relevant statute and the available evidence. Each of the six discrete doctrines courts say they apply is just a reference to one or more of those criteria.

I continue to find Zaring’s proposed change in doctrine compelling, but it would have to be restated in a longer form to be susceptible to judicial application. Thus, courts might replace the six doctrines they now use with the following single doctrine—a court will uphold an agency action if the agency explains adequately why the action is consistent with the statute it is implementing and why it is supported by the available evidence relevant to the action.

Of course, courts would then need to explain the manner in which they evaluate the adequacy of the agency reasoning process, the manner in which they evaluate the adequacy of the relationship between the action and the statute it implements, and the manner in which they evaluate the adequacy of the relationship between the available evidence and the action. The resulting doctrinal environment might be an improvement over the existing doctrinal environment in the sense that it is a more accurate and a more complete description of the role of a reviewing court. I doubt, however, that it would represent a simplification of existing doctrine. It might more accurately be characterized as a restatement of existing doctrine.

Whether you call Zaring’s proposal a simplification or a restatement, it is appealing. The article is well worth reading independent of his proposed change in doctrine. It is packed with data, analysis, and useful insights.                

The Truth Might Set Your Statutory Interpretation Free

Dec 17, 2010 Richard Murphy1 comment

Richard Murphy

There is something silly about Supreme Court decisions in which five justices explain that the conventional tools of statutory interpretation—e.g., legislative intent, objective textual meaning, and judicial rules for discerning and applying them—plainly indicate that a statute means A but the other four justices deploy the same tools to explain that the statute plainly means B.  After all, if the relevant meaning were all that clear, wouldn’t all nine of the extraordinarily capable legal minds on the Court come to quick agreement?  And isn’t their disagreement strong evidence that the statutory question has no pre-existing, determinate legal answer?   One might expect that under such circumstances, the rules of reasoned legal discourse would require justices to make remarks like, “Wow.  My dissenting colleagues’ arguments are really very good—they almost persuaded me—and I’m no pushover. But, on balance, I still think it is a better idea to choose interpretation A instead of B.”  Instead, the more usual practice is for both sides to insist that the other is just plain wrong.

In his elegant essay, Professor Foy suggests courts dispense with such nonsense and instead tell the truth.  Suppose, for instance, a judge determines that conventional tools of statutory interpretation do not compel a choice between readings A and B.  The judge happens to think that the world would be a better place were she to choose B.  Under the current rules of the game, the judge should write an opinion that tries to justify choosing B based solely on conventionally acceptable tools—which might include, say, old dictionaries.  This sort of exercise can generate judicial explanations that are strained, arbitrary, or untruthful.  According to Professor Foy, the judge should: (a) instead concede that interpretations A and B both seem pretty darn reasonable as a matter of conventional legal analysis, and then (b) truthfully explain whatever reasons of policy, equity, or justice moved the judge to choose one interpretation over the other.  In short, judges should admit that they must exercise discretion when choosing among reasonable interpretations of ambiguous statutes and then exercise that discretion as prudently and transparently as they can.

I enjoyed Professor Foy’s essay for several intertwined reasons.  First, it is a fun read.  If statutes were written as clearly and lucidly as Professor Foy’s essay, then he would not have had to write it.  The quality of the writing lends it persuasive force in a way that calls to mind that lawyers—and by extension, law professors—are supposed to be good at rhetoric.  My favorite turn of phrase from the essay comes in Professor Foy’s dissection of Chapman v. United States (1991)—the case in which a majority of the Supreme Court relied in part on dictionary definitions to determine that LSD and blotter paper form a “mixture” for sentencing purposes.  As Professor Foy remarked in a fairly devastating riposte, “[a] reasonably intelligent English speaker would not ordinarily use the word ‘mixture’ to describe a necktie stained with soup or a napkin soaked in cod liver oil.” Professor Foy’s essay is wise and wry.

Second, Professor Foy’s essay provides a novel angle for thinking about a very old problem.  It is not, of course, news that statutory interpretation, in the hard cases, includes a discretionary element.  Indeed, recognition of this point is a foundational element of the Chevron doctrine, which in essence instructs courts to affirm an agency’s discretionary choice among reasonable statutory constructions provided the agency gives a reasonable explanation for the choice.  One can understand Professor Foy’s essay as insisting that courts apply the same standards of reasoned decision-making to themselves as they apply to agencies—what is good for the discretionary administrative goose is also good for the discretionary judicial gander.

Best of all, Professor Foy’s essay is thought-provoking.  After reading it, I found myself mulling over, among other things:

• Would it really be better for courts to be more truthful about the discretionary nature of the task of interpreting ambiguous statutes?
• Are courts actually lying about this process even if no one believes them?
• Would greater judicial truthfulness about statutory interpretation tend to exacerbate or mitigate the role of ideology in this process?
• What would be the effect of judicial truth-telling on stare decisis?  Where a judge admits that he or she chose an interpretation based on a personal policy preference, should that choice be entitled to weight in other cases?
• And, last but not least, what exactly do we mean by “discretion,” anyway?

So, there you have it:  Professor Foy’s essay on discretion in judicial statutory construction is wise, wry, novel, and thought-provoking.  You will enjoy reading it.

Jettisoning Chevron

Nov 19, 2010 Linda Jellum1 comment

Linda Jellum

As one academic, among many, who has made my scholarly reputation based in part on the landmark case of Chevron U.S.A., Inc. v. Nat. Resources Defense Council, 467 U.S. 837 (1984), I noted with some concern Professor Jack Beermann’s latest  work entitled:  End the Failed Chevron Experiment Now:  How Chevron Has Failed and Why it Can and Should be Overruled. Overrule Chevron?  End the experiment? Say it ain’t so! Hadn’t Chevron offered “a wonderful new world … full of promise for administrative-law professors in need of tenure articles….?”  National Cable & Telecommuns. Ass’n v. Brand X Internet Serv., 545 U.S. 967, 1019 (2005) (Scalia, J., dissenting).

I approached his article with some trepidation but also with great interest.  Why would anyone want to overrule Chevron?  Professor Beermann succinctly answers this question in his abstract:  “Chevron has complicated judicial review and, at best, it is uncertain whether it has resulted in increased deference to agency interpretation. In fact, for numerous reasons, Chevron has been a failure on any reasonable measure and should be overruled.”  Intrigued, I forged ahead.

His article begins by identifying a number of reasons why the Chevron doctrine has failed.  Specifically, (1) the Chevron doctrine violates 5 U.S.C. § 706, the statute that generally governs judicial review of administrative decisions and requires that courts review questions of law de novo, (2) the Chevron doctrine was based on faulty presumptions regarding congressional intent to delegate by ambiguity and regarding political accountability, (3) the Chevron doctrine is highly unpredictable, (4) the Chevron doctrine has not lead to increased deference to agency interpretations, (5) the Chevron doctrine is often not cited by the Supreme Court in cases in which it should apply, (6) the Chevron doctrine has increased litigation costs, and (7) the Chevron doctrine encourages agencies and judges to act irresponsibly.  Perhaps most importantly, however, Professor Beermann notes that the Chevron doctrine is simply unclear in application.  First, it is unclear whether Chevron applies:  Professor Cass Sunstein famously labeled the question of whether Chevron applies as Chevron step zero.  Second, it is unclear how Chevron applies; in other words, is step one just a textual analysis or is it a full statutory interpretation analysis and does step two apply differently depending on whether Congress was explicit or implicit when it delegated?  Third, it is unclear why Chevron applies given that the doctrine was contrary to the established distribution of interpretive power among the three branches.  Finally, it is unclear when Chevron applies; in other words, does the doctrine apply to agency policy decisions or just to agency statutory interpretations?

While Chevron initially promised simplicity, it has delivered only chaos.  Indeed, the Justices have developed four variants of Chevron:  (1) the “original directly spoken” variant, (2) the “traditional tools” variant, (3) the “plain meaning” variant, and (4) the “extraordinary cases” variant.  The number and diversity of the variants epitomizes all that is wrong with Chevron; hence, Professor Beermann concludes that the case should be overruled and the doctrine replaced with a more consistent and simple one.

After describing Chevron’s failings, Professor Beermann very briefly explains why overruling Chevron would not violate the Supreme Court’s stare decisis principles.  In short, overruling Chevron would be permissible under Pearson v. Callahan, 129 S. Ct. 808, 811 (2009), in which the Supreme Court explained when overruling a case would be consistent with stare decisis principles.  At bottom, overruling Chevron would not affect settle expectations because the doctrine’s application has been so unsettled.  Additionally, it is a judge-made doctrine that time and experience have proved unworkable.

In the final section of his article, Professor Beermann offers two alternatives: reforming or replacing Chevron.  Both alternatives would be more consistent with the APA’s assignment to the courts to review questions of law de novo and questions of policy under arbitrary review.

Alternative one:  if Chevron were not jettisoned, Professor Beermann begs that it at least be reformed.  Specifically, he argues that the narrow, step one variants be rejected.  Courts should be free to interpret statutory language using all the traditional tools of interpretation.  Such a change would return interpretive power to the judiciary.  Second, he recommends that step two be limited to occasions when Congress explicitly delegated a range of choices to the agency and be expanded to include arbitrary and capricious review.  Supporting this additional step, he says that it is possible for an interpretation to be a reasonable interpretation of the statute, but still be arbitrary and capricious choice.  Third, he suggests that Chevron step zero die a quick and painless (painful?) death.  This step simply complicates the analysis unnecessarily; rather, Chevron should apply universally, whenever Congress specifically delegated interpretive authority to the agency.

Alternative two:  Instead, if Chevron were jettisoned, Professor Beermann recommends returning to pre-Chevron practice.  Under this practice, assuming an agency had the power to make legislative rules, the reviewing court would determine whether the agency had interpreted the statute reasonably by paying close attention to the language of the statute, its purpose, and its history.  In addition, the court would consider the wisdom of the agency’s policy choice under the arbitrary, capricious standard.  Pre-Chevron, judges were the final arbitrators of statutory meaning, while agencies played an advisory role.

Regardless of whether you agree with Professor Beermann’s provocative demand to “End the Failed Chevron Experiment Now,” he has written a persuasive article that identifies and explains the problems with continuing with Chevron in its current form, whatever variant that might be.  He has identified two alternatives, either of which would vastly simplify judicial review of agency interpretations of statutes.  In short, while I expect many law professors might decry the loss of Chevron, this article suggests we would be crying alone.