DOUBLE DEFERENCE IN ADMINISTRATIVE LAW

DOUBLE DEFERENCE IN ADMINISTRATIVE LAW

Administrative law presumes a neat system of agency rulemaking and adjudication followed by judicial review. But the reality of the administrative state departs starkly from this model. One such departure is the use of audited self-regulatory organizations (SROs)—private organizations comprised of specific industries that formulate binding law to regulate themselves. Although SROs operate subject to the oversight of federal agencies, their power is vast, reaching significant swaths of the national and international economies. There is little to constitutionally constrain such arrangements, and whereas the administrative law model values the norms of participation, deliberation, and transparency, the procedures that SROs use depart from these norms in important ways. Moreover, oversight agencies are deferential to SROs, and courts in turn are deferential to the oversight agencies. This doubling of deference both undermines accountability and fails to adequately guard against arbitrariness. This Article brings a much-needed administrative law lens to SROs, providing both a positive and theoretical account of SROs and exposing flaws in the model. To better ameliorate these concerns, this Article illustrates how existing administrative law can more comprehensively account for SROs and offers a series of institutional-design considerations for furthering administrative law norms in the future.

INTRODUCTION

  1. AN OVERVIEW OF THE SRO MODEL
    1. SRO Structure and Power
    2. The Theory of SROs
    3. SROs in the Constitutional and Common Law Schemes
      1. The Private Nondelegation Doctrine
      2. The State Action Doctrine
  2. SRO PROCEDURES AND OVERSIGHT-AGENCY DEFERENCE
    1. FINRA and the SEC
    2. The NFA and the CFTC
    3. NERC and FERC
  3. DOUBLE DEFERENCE: JUDICIAL REVIEW OF SRO-INITIATED ACTIONS
    1. Procedures and APA Analogies
    2. Substantive Review, Equating with Agencies, and Deference to Expertise
    3. Reviewability
    4. A Critique of Double Deference
  4. TOWARD A MORE ROBUST MODEL
    1. Situating SROs in Traditional Administrative Law
      1. The Chenery Cases
      2. State Farm and Hard-Look Review
      3. Chevron and Mead
    2. Building Better SROs
      1. Structure of the Oversight Agency
      2. The Nuclear Option
      3. Procedures Governing the SRO–Agency Relationship
      4. Facilitating External Oversight
      5. Best Practices

CONCLUSION

Introduction

Administrative law’s familiar narrative contemplates that agencies, empowered by their statutory mandates, carry out their delegated duties subject to the check of judicial review and political oversight. Each component of this narrative has a rich pedigree and doctrinal depth, reflecting normative concerns for both administrative law and agencies’ place in the constitutional structure. Innovations in regulatory inst­itutional design, however, fit only awkwardly within the traditional story, leaving practical, theoretical, and doctrinal deficiencies in administrative law. This is especially true for the regulatory state’s use of audited self-regulatory organizations (SROs), which challenge the traditional under­standing of administrative law from nearly every angle.

SROs occupy a singular place in administrative law. Composed of industry members, they are subject to agency oversight yet wield significant power: SROs are responsible for regulating the securities ex­changes, 1 The Financial Industry Regulatory Authority (FINRA) is an SRO subject to oversight by the Securities and Exchange Commission (SEC) that writes and enforces rules governing nearly 4,000 securities firms with 639,680 brokers. About FINRA, Fin. Indus. Regulatory Auth., http://www.finra.org/about [http://perma.cc/R9BF-HREY] [hereinafter FINRA, About FINRA] (last visited Aug. 2, 2016). In 2015, FINRA levied $95.1 million in fines and ordered over $96 million in restitution. See Jonathan Macey & Caroline Novogrod, Enforcing Self-Regulatory Organization’s Penalties and the Nature of Self-Regulation, 40 Hofstra L. Rev. 963, 969 (2012) (“The extent of FINRA’s authority is enormous.”); id. at 970–71 (noting FINRA has nearly the same number of employees and amount of revenue as the SEC). the $40 trillion futures market, 2 The National Futures Association (NFA) is an SRO subject to the oversight of the Commodity Futures Trading Commission (CFTC) that regulates the U.S. derivatives industry, including on-exchange futures and over-the-counter derivatives. Who We Are, Nat’l Futures Ass’n, http://www.nfa.futures.org/NFA-about-nfa/index.html [http://perma.cc/BS4A-QHDP] (last visited Aug. 2, 2016); see Oversight of Dodd-Frank Implementation: A Progress Report by the Regulators at the Half-Year Mark: Hearing Before the S. Comm. on Banking, Hous. & Urban Affairs, 112th Cong. 63 (2011) (statement of Gary Gensler, Chairman, Commodity Futures Trading Commission) (describing the $40 trillion futures market). and the reliability of the entire electric grid. 3 The North American Electric Reliability Corporation (NERC) is an SRO subject to the oversight of the Federal Energy Regulatory Commission (FERC) that regulates the reliability of the electric grid. About NERC, N. Am. Elec. Reliability Corp., http://www.nerc.com/AboutNERC/Pages/default.aspx [http://perma.cc/AS29-GU6X] (last visited Aug. 2, 2016). The grid connects 5,800 major power plants and includes over 450,000 miles of high-voltage transmission lines; weather-related outages alone cost, on average, $18–38 billion annually. Exec. Office of the President, Economic Benefits to Increasing Electric Grid Resilience to Weather Outages 3, 5 (2013), http://energy.gov/sites/prod/files/2013/08/f2/Grid%20Resiliency%20Report_FINAL.pdf [http://perma.cc/6Z6H-CGXX]. Professor Douglas Michael collects numerous other SROs and SRO-related schemes, including the fields of agricultural marketing agreements, banking, peer review of Medicare services, clinical laboratories, and higher-education financing. Douglas C. Michael, Federal Agency Use of Audited Self-Regulation as a Regulatory Technique, 47 Admin. L. Rev. 171, 203–40 (1995). Whereas administrative law has long theorized about the ill effects of agency capture, 4 See, e.g., Richard B. Stewart, The Reformation of American Administrative Law, 88 Harv. L. Rev. 1669, 1684–85 (1975) (describing the capture critique); Wendy Wagner et al., Rulemaking in the Shade: An Empirical Study of EPA’s Air Toxic Emission Standards, 63 Admin. L. Rev. 99, 123 (2011) (documenting the systemic imbalance in rulemaking participation, with regulated industry engaging in rulemaking far more frequently than public interest groups). in regard to SROs, the industry’s control is not theoretical; it is an intentional design choice. 5 For further discussion of the SRO design, see infra section I.A. The model is both counterintuitive to the traditional account of administrative law and almost completely overlooked as a component of the regulatory state.

The aim of this Article is to introduce a positive account and normative theory of SROs’ place in the administrative state. To say that administrative law has neglected SROs is not to say that they have gone entirely unnoticed. For example, SROs attract attention in the scholarly literature as part of the broader movement toward privatization of government functions. 6 E.g., Jody Freeman, Extending Public Law Norms Through Privatization, 116 Harv. L. Rev. 1285, 1286 (2003) [hereinafter Freeman, Extending Public Law Norms] (arguing privatization through contracting can serve as a means of infusing private behavior with public norms); Gillian E. Metzger, Privatization as Delegation, 103 Colum. L. Rev. 1367, 1369–76 (2003) (developing a new approach to the state action doctrine). There is also a significant body of literature concerning the role of SROs in financial regulation. 7 E.g., William A. Birdthistle & M. Todd Henderson, Becoming a Fifth Branch, 99 Cornell L. Rev. 1, 24–25 (2013) (describing the “governmentalization” of financial SROs over time); Kristin N. Johnson, Governing Financial Markets: Regulating Conflicts, 88 Wash. L. Rev. 185, 200–01 (2013) (evaluating the effectiveness of financial SROs); Saule T. Omarova, Rethinking the Future of Self-Regulation in the Financial Industry, 35 Brook. J. Int’l L. 665, 670 (2010) [hereinafter Omarova, Rethinking] (arguing reforms are needed to capture the benefits of financial SROs); Saule T. Omarova, Wall Street as Community of Fate: Toward Financial Industry Self-Regulation, 159 U. Pa. L. Rev. 411, 417–21 (2011) [here­inafter Omarova, Wall Street] (advocating for “embedded self-regulation” to improve the existing SRO paradigm in the financial sector by better incorporating public interest); Derek Fischer, Note, Dodd-Frank’s Failure to Address CFTC Oversight of Self-Regulatory Organization Rulemaking, 115 Colum. L. Rev. 69, 71 (2015). There are fewer, but nonetheless notable, sources in electricity-reliability regulation. E.g., John S. Moot, When Should the FERC Defer to the NERC?, 31 Energy L.J. 317, 317–19 [hereinafter Moot, Defer] (considering the puzzling features of FERC’s standard of review for NERC’s proposed reliability standards). Congress is active in its attention to SROs, legislating frequently in this area in recent years. 8 The most prominent examples are Dodd-Frank and the Energy Policy Act of 2005. See infra Part II (providing detailed evaluations). And scholars have argued that SRO models are appropriate for industries as diverse as food labeling, 9 See Ellen A. Black, Keep Out FDA: Food Manufacturers’ Ability to Effectively Self-Regulate Front-of-Package Food Labeling, 17 DePaul J. Health Care L. 1, 18–24 (2015). the Internet, 10 See Henry H. Perritt, Jr., Towards a Hybrid Regulatory Scheme for the Internet, 2001 U. Chi. Legal F. 215, 215–21 (discussing the advantages of private regulation of the Internet). the sharing economy, 11 See Bryant Cannon & Hanna Chung, A Framework for Designing Co-Regulation Models Well-Adapted to Technology-Facilitated Sharing Economies, 31 Santa Clara High Tech. L.J. 23, 32 (2015). and crowdfunding. 12 Karina Sigar, Comment, Fret No More: The Inapplicability of Crowdfunding Concerns in the Internet Age and the JOBS Act’s Safeguards, 64 Admin. L. Rev. 473, 500–01 (2012) (arguing the requirement that crowdfunding intermediaries register with SEC-supervised SROs provides important protections against abuse). A number of other industries are natural candidates for the SRO model because, among other things, they have already devel­oped voluntary standards of conduct and traditional regulation is not politically viable. 13 For further discussion of prerequisites to SRO formation, see Michael, supra note 3, at 192–95. The natural gas industry, for example, has already developed voluntary standards for the controversial 14 See, e.g., David B. Spence, Responsible Shale Gas Production: Moral Outrage vs. Cool Analysis, 25 Fordham Envtl. L. Rev. 141, 156–67 (2013) (describing the divergent stances among antihydraulic-fracturing groups, regulators, and industry regarding the uncertain risks associated with hydraulic fracturing). technologies involved in hydraulic fracturing, making the SRO model a logical next step. 15 See About the Center for Sustainable Shale Development (CSSD), Ctr. for Sustainable Shale Dev., http://www.sustainableshale.org/about [http://perma.cc/CN33-XUM3] (last visited Aug. 2, 2016); see also infra notes 78–79 and accompanying text (describing how self-regulation may serve norms of professionalism). In other words, the likelihood of more SROs in the future is real.

What is missing is the administrative law perspective. Specifically, this Article is concerned with how the combination of statutes, SRO and agency actions, and judicial doctrine work together to further (or hinder) the norms of participation, deliberation, and transparency. 16 See Emily Hammond & David L. Markell, Administrative Proxies for Judicial Review: Building Legitimacy from the Inside-Out, 37 Harv. Envtl. L. Rev. 313, 322–23 (2013) (describing these norms). Administrative law places great emphasis on these norms for their role in furthering accountability 17 Professor Jerry Mashaw has distinguished between public-governance accountability (which encompasses administrative law, public administration, and political accountability), marketplace accountability, and social accountability. Jerry L. Mashaw, Accountability and Institutional Design: Some Thoughts on the Grammar of Governance, in Public Accountability: Designs, Dilemmas and Experiences 115, 119–21 (Michael W. Dowdle ed., 2006). Professor Mashaw’s typology provides a helpful lens through which SRO schemes can be evaluated. Building off Professor Mashaw’s typology, this Article is largely concerned that SROs imperfectly substitute marketplace or social accountability for public-governance accountability. and guarding against arbitrariness, which in turn reinforce both democratic and constitutional legitimacy. 18 See id. See generally Lisa Schultz Bressman, Beyond Accountability: Arbitrariness and Legitimacy in the Administrative State, 78 N.Y.U. L. Rev. 461 (2003) [hereinafter Bressman, Beyond Accountability] (arguing that administrative law serves constitutional legitimacy). But as developed here, these norms are vulnerable in the world of SROs. First, as private entities that create public law, SROs raise significant con­stitutional concerns grounded in arbitrariness and lack of account­ability—but the private nondelegation doctrine offers no real check on SROs’ power. 19 See infra section I.C.1 (discussing the private nondelegation doctrine). And while SROs are not usually considered government actors for substantive constitutional purposes, 20 See infra section I.C.2 (describing the state action doctrine). they are frequently considered government actors for purposes of common law immunity. 21 See infra section I.C.3 (examining common law immunity).

Second, the mere fact that SROs operate subject to the oversight of a federal agency is what bears the entire weight of SROs’ uncomfortable position in the constitutional scheme. With so much riding on the relationship between SROs and their oversight agencies, a closer look is merited. There has been one study of SROs, conducted by Professor Douglas Michael for the Administrative Conference of the United States (ACUS) in the early 1990s. 22 Michael, supra note 3. This important effort collected all of the SRO schemes existing at the time and suggested various prerequisites for the successful formation of such schemes. 23 Id. at 191–203. But it did not test the schemes through a normative administrative law lens, nor did it examine the many assumptions inherent in the doctrinal account of SROs’ place in the constitutional structure. As it turns out, many traditional assump­tions about how SROs operate and where they find their legitimacy are simply wrong.

If one delves more deeply—taking a careful look at how SROs are constituted, how they develop rules or bring enforcement actions, and especially how they are reviewed by their oversight agencies—what emerges is a deference regime. Specifically, oversight agencies are defer­ential either in practice or as a matter of statutory design to both rules and orders originating from SROs. 24 See infra Part II (providing an overview of agency deference to SRO procedures). This is cause for concern: A regulated industry that regulates itself has numerous incentives to dampen participation of statutory beneficiaries, engage in anticompe­titive conduct toward weaker members, and ratchet regulatory law toward the lowest common denominator. 25 See Sidney A. Shapiro, Outsourcing Government Regulation, 53 Duke L.J. 389, 404–05 (2003) (describing concerns when private interests are involved in agency standard setting). These concerns are at the forefront of the regulatory-capture literature. See, e.g., Roger Noll, Reforming Regulation 40–43 (1971) (presenting the classic view of regulatory capture); Sidney A. Shapiro & Rena A. Steinzor, Capture, Accountability, and Regulatory Metrics, 86 Tex. L. Rev. 1741, 1753–55 (2008) (collecting sources showing disproportionate participation of regulated entities in rulemakings, one of which sug­gested agencies alter final rules favorably to regulated entities in response to such participation); Wagner et al., supra note 4, at 114–15 (describing opportunities for industry groups to ratchet down regulations following rulemaking). Further, many aspects of SRO activity are opaque, decreasing opportunities for oversight and thereby under­mining accountability.

If deference by the oversight agency in the SRO’s favor is of concern, consider now the impact of judicial deference. Because SRO actions have their oversight agencies’ imprimatur, when such actions are challenged in court, the ordinary rules of administrative law apply. That is, courts are simply reviewing agency actions. 26 See Chevron U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 843 (1984) (holding “if the statute is silent or ambiguous[,] . . . the question for the court is whether the agency’s answer is based on a permissible construction of the statute”). And a hallmark of administrative law is judicial deference to the agency—whether on grounds of comparative political accountability, 27 See id. at 865 (“While agencies are not directly accountable to the people, the Chief Executive is . . . .”). See generally Emily Hammond Meazell, Presidential Control, Expertise, and the Deference Dilemma, 61 Duke L.J. 1763 (2013) [hereinafter Hammond, Deference Dilemma] (arguing expertise and presidential-control justifications for deference fit awkwardly into statutory schemes involving overlapping or competing jurisdictions). superior expertise, 28 See, e.g., Fed. Power Comm’n v. Hope Nat. Gas Co., 320 U.S. 591, 602 (1944) (“[T]he product of expert judgment . . . carries a presumption of validity.”). This author has criticized the judicial practice of affording “super deference” to agencies. Emily Hammond Meazell, Super Deference, the Science Obsession, and Judicial Review as Translation of Agency Science, 109 Mich. L. Rev. 733, 749–50 (2011) [hereinafter Hammond, Super Deference]. or implicit legislative intent. 29 See United States v. Mead Corp., 533 U.S. 218, 219 (2001) (stating that legislative intent to delegate interpretive authority can “be apparent from the agency’s generally conferred authority”). Too much deference, however, undermines administrative law norms because it fails to adequately incentivize agencies both ex ante and ex post. It does not incentivize the use of legitimizing procedures in the first place, nor does it provide a sufficient check after the action has taken place. 30 Hammond, Super Deference, supra note 28, at 737–38 (listing these and other critiques of super deferential judicial review of agency actions); Hammond & Markell, supra note 16, at 321–27 (describing the theory of how judicial review enhances agency legitimacy).

In the SRO context, deference is doubled. Put plainly, administrative law ignores the SROs’ role in the administrative state, blindly equates SROs with their oversight agencies, and countenances even greater deference than usual owing to the fact of the SROs’ presence. Combin­ing this judicial deference with the oversight agency’s deference obscures the many participatory, deliberative, and transparency-related short­comings of the overall scheme. Thus, the final aims of this Article are twofold. First, this Article suggests ways to reorient administrative law to the realities of SROs, illustrating how common doctrines can be used to accommodate the SRO model while addressing the concerns that double deference raises. Second, it sets forth a number of institutional-design considerations for either amending existing SROs or creating new ones. Before turning to these matters, however, a note about the scope of this project may be helpful.

Specifically, this Article focuses on audited self-regulation, which means that the self-regulated industry has power to issue binding law but operates subject to the oversight of a federal agency. 31 This definitional approach is consistent with that already used in the literature. See Michael, supra note 3, at 175–76 (employing a substantively identical definition); Omarova, Rethinking, supra note 7, at 698 (same). The agency’s oversight role distinguishes SROs from voluntary or purely private self-regulatory efforts. 32 This criterion excludes the “sanctioned” self-regulation approach such as the television-program rating system, which, although approved by the Federal Communications Commission pursuant to the “V-Chip Law,” is not enforced by that agency. See Joel Timmer, Television Violence and Industry Self-Regulation: The V-Chip, Television Program Ratings, and the TV Parental Guidelines Oversight Monitoring Board, 18 Comm. L. & Pol’y 265, 269–70 (2013) (describing the history and current state of self-regulation within the television industry). Moreover, this Article distinguishes SROs here from standards development organizations (SDOs), which often formulate standards that become federal law in a rulemaking process but lack the enforcement powers of SROs. 33 When standards become federal law, the enforcement power resides with the promulgating agency. Michael, supra note 3, at 177–78. An important body of literature identifies the transparency issues related to SDOs in administrative law; when standards are incorporated by reference into administrative rulemakings, they can be difficult to ascertain. See, e.g., Emily S. Bremer, On the Cost of Private Standards in Public Law, 63 U. Kan. L. Rev. 279, 279–80 (2015); Nina A. Mendelson, Private Control over Access to the Law: The Perplexing Federal Regulatory Use of Private Standards, 112 Mich. L. Rev. 737, 739 (2014) [hereinafter Mendelson, Private Control]; Peter L. Strauss, Incorporating by Reference: Knowing the Law in the Electronic Age, Admin. & Reg. L. News, Winter 2014, at 36, 36 (2014). Copyright issues may complicate the availability of standards. See Emily S. Bremer, Incorporation by Reference in an Open-Government Age, 36 Harv. J.L. & Pub. Pol’y 131, 159–76 (2013). For an insightful look at the SDO process involved in interoperability standards for the Smart Grid, see Joel B. Eisen, Smart Regulation and Federalism for the Smart Grid, 37 Harv. Envtl. L. Rev. 1, 30–31 (2013) (describing traditional standards-development process for organizations accredited by the American National Standards Institute); see also Michael P. Vandenbergh, Private Environmental Governance, 99 Cornell L. Rev. 129, 133 (2013) (exploring broadly defined private-governance aspects of environmental law). Also distinguishable are licensing boards that regulate the entry into, and practice of, professions; these entities are typically a feature of state and local law and lack both the self-regulatory and audited features of SROs. 34 Michael, supra note 3, at 178. Nor are SROs within the many collaborative public–private partnerships governed by contract. 35 See, e.g., A. Michael Froomkin, Wrong Turn in Cyberspace: Using ICANN to Route Around the APA and the Constitution, 50 Duke L.J. 17, 142–43 (2000) (criticizing the Department of Commerce’s contractual delegation of power over the Internet root to ICANN); see also Jody Freeman, The Contracting State, 28 Fla. St. U. L. Rev. 155, 156 (2000) (cautioning against “dire consequences” of “[w]idespread contracting out of services or arguably ‘public’ functions”); Mark Seidenfeld, An Apology for Administrative Law in The Contracting State, 28 Fla. St. U. L. Rev. 215, 224–38 (2000) (arguing administra­tive law has accommodated bilateral contracting arrangements to promote accountability); Paul R. Verkuil, Public Law Limitations on Privatization of Government Functions, 84 N.C. L. Rev. 397, 424 (2006) [hereinafter Verkuil, Public Law Limitations] (arguing policy­making functions should be held back from the scope of privatization); cf. Mendelson, Private Control, supra note 33, at 739 (distinguishing the SDO process from private contractual arrangements). Al­though SROs can be understood as a form of privatization, they are uniquely distinguishable from these contract-based forms. 36 Cf. Metzger, supra note 6, at 1370 (defining privatization broadly as “government use of private entities to implement government programs or to provide services to others on the government’s behalf”). The procedures that SROs use are also distinguishable from negotiated rulemaking. See Freeman, Extending Public Law Norms, supra note 6, at 1328–29 (describing negotiated rulemaking).

The reasons for carving out SROs from these other forms of private involvement in the government relate to their functions. Of all the models, SROs most closely mimic regulatory agencies in their broad powers of rulemaking and adjudication. They enjoy deference from their oversight agencies but are nearly invisible to administrative law. This invisibility has led to a hidden layer of government that raises serious accountability and legitimacy concerns. It is deserving of independent treatment, but it also holds insights for how one understands the admin­istrative state more generally. 37 This Article is therefore of a piece with other efforts at painting a more accurate picture of the administrative state. See, e.g., Daniel A. Farber & Anne Joseph O’Connell, The Lost World of Administrative Law, 92 Tex. L. Rev. 1137, 1142 (2014) (showing how ultimate decisionmaking authority often rests outside agencies); Miriam Seifter, Second-Order Participation in Administrative Law, 63 UCLA L. Rev. 1300, 1304 (2016) (revealing how interest groups diverge from participatory assumptions).

Part I of the Article first provides a more detailed overview of SROs, which further defines the scope of the project and engages the primary benefits and drawbacks of the model. Next, this Part considers the relative lack of constraints on SROs, focusing on the private nondele­gation doctrine, the state action doctrine, common law immunities, and preemption. As this discussion shows, these sources of law protect administrative law values only tangentially and in a piecemeal fashion, if at all. This conclusion leads to the task of Part II, which is to dig deeply into the relationship between SROs and their oversight agencies. Using three concrete examples, this Part exposes the many design flaws in existing SRO models when viewed from a process-based perspective. Part III shows why judicial review fails to ameliorate these flaws. Indeed, double deference is supremely unsuited to the SRO model and fails to serve the signaling function that judicial review offers in administrative law more generally. With this analytical work in place, Part IV considers doctrinal puzzles that SROs present and suggests concrete ways that administrative law could account for the realities of SROs while remaining consistent with governing doctrine. Next, this Part suggests institutional design features that would better align the SRO model with the norms of administrative law. The Article concludes with observations about what this study of SROs offers for understanding the modern regulatory state more generally.

I. An Overview of the SRO Model

Modern SROs in the United States are traceable to the New Deal. 38 Although the New Deal formalized the SRO model, self-organization and self-regulation in commercial and financial groups already had a strong historical tradition dating back hundreds if not thousands of years. Johnson, supra note 7, at 199–200 & n.60 (describing hundreds of years of financial-market participants’ organized efforts and collecting sources regarding ancient community-based regulations governing commercial trading); see also Omarova, Rethinking, supra note 7, at 671 (noting self-regulation in Medieval European merchant guilds); Mark D. West, Private Ordering at the World’s First Futures Exchange, 98 Mich. L. Rev. 2574, 2580 (2000) (describing seventeenth-century Japanese self-regulatory rice exchanges). During this period, national policymakers were concerned with both remedying widespread economic and social failures—not the least of which was the stock market crash of 1929 39 See Birdthistle & Henderson, supra note 7, at 16 (describing conventional wisdom at the time as attributing the crash and Great Depression to unregulated securities speculation). —and ensuring adequate checks on new agencies with broad statutory mandates. 40 The history of the Administrative Procedure Act (APA) reflects these concerns. See generally Walter Gellhorn, The Administrative Procedure Act: The Beginnings, 72 Va. L. Rev. 219 (1986) (describing the American Bar Association’s strong opposition to the New Deal’s broad statutory mandates). Even as the Lochner era was closing, 41 Named for Lochner v. New York, which implied liberty of contract in the Due Process Clause of the Fourteenth Amendment and struck down a labor law, 198 U.S. 45 (1905), the Lochner era is often described as a period of judicial activism favoring industry. See generally Barry Friedman, The History of the Countermajoritarian Difficulty, Part Three: The Lesson of Lochner, 76 N.Y.U. L. Rev. 1383, 1389–428 (2001) (critiquing both conventional and revisionist accounts of the Lochner era). The era closed with West Coast Hotel Co. v. Parrish, 300 U.S. 379 (1937), which upheld the constitutionality of a minimum-wage law. industry groups wielded significant power, and their involvement in self-regulation offered a politically acceptable means of introducing federal regulation to major economic sectors. Below, section I.A describes SROs’ typical characteristics related to structure and power, section I.B turns to the theory of SROs, and section I.C considers SROs’ place in the constitutional and common law structures.

A. SRO Structure and Power

The most enduring SRO models to emerge from this time period were those under the oversight of the Securities and Exchange Commission (SEC) and Commodity Futures Trading Commission (CFTC). 42 See Johnson, supra note 7, at 201–04 (providing background of the self-regulatory model for financial markets). Other variants persist, such as market orders under the oversight of the Secretary of Agriculture. See Michael, supra note 3, at 234–39 (describing the Secretary of Agriculture’s authority under the Agricultural Marketing Agreement Act of 1937 “to facilitate, or in some cases impose” agreements between handlers and pro­ducers to regulate “the quality and quantity” of market products). These schemes provided the original blueprint for, and continue to shape, the defining features of SROs. First, Congress devises nearly all SRO schemes and mandates that SROs implement authority delegated to an administrative agency. 43 Michael, supra note 3, at 175–76. In other words, SRO schemes are creatures of the legislature, not discretionary choices of agencies. For concrete examples, see infra Part II. Although the courts have not directly tested the necessity of such legislative enabling acts, 44 Cf. Pub. Citizen v. Nuclear Regulatory Comm’n, 901 F.2d 147, 158 (D.C. Cir. 1990) (interpreting § 306 of the Nuclear Waste Policy Act of 1982 to reject the Nuclear Regulatory Commission’s (NRC’s) reliance on the policy statement endorsing industry-developed standards for training nuclear-power-plant operators). Strictly speaking, the NRC in this case did not attempt to create an audited self-regulated entity; it encouraged but did not compel licensees to comply with industry standards developed by the Institute of Nuclear Power Operations, which was the industry’s self-regulatory body. Id. at 149. that they are functionally required is both descriptively and normatively accurate: Explicit legislative authority stands behind nearly every SRO, 45 Agencies rarely create SROs, but in those unusual circumstances, the agencies’ organic statutes have authorized such action. See Michael, supra note 3, at 244–45 (describing these unusual schemes). and the legislative process both lends legitimacy to the model’s use and offers opportunities for oversight once the model is implemented. 46 See id. at 245 (“[I]t is likely that explicit congressional authority is necessary in any event, and is certainly a practical requirement.”). The involvement of the oversight agency is also a critical component of the model. As described in more detail below, the private nondelegation doctrine prevents direct delegations of authority to entities not within the federal government. 47 Infra section I.C.1 (discussing the private nondelegation doctrine). As such, an SRO’s authority is properly viewed as derivative of the oversight agency’s authority. 48 See Free Enter. Fund v. Pub. Co. Accounting Oversight Bd., 561 U.S. 477, 527 (2010) (Breyer, J., dissenting) (describing SRO schemes generally).

Second, the SROs are comprised of individuals or entities from within the regulated industry and are frequently funded by membership dues and enforcement fees. 49 E.g., Who We Are, Nat’l Futures Ass’n, supra note 2. For an illustration of a parti­cularly problematic conflict of interest in a non-SRO banking context, see Kent Barnett, Codifying Chevmore, 90 N.Y.U. L. Rev. 1, 22–26 (2015) [hereinafter Barnett, Codifying Chevmore] (describing the Office of the Comptroller of the Currency’s funding from chartered members and its resulting misuse of the preemption doctrine). Because SROs regulate their own indus­tries, they act as gatekeepers to activities within those industries. 50 In re Series 7 Broker Qualification Exam Scoring Litig., 510 F. Supp. 2d 35, 38 (D.D.C. 2007) (identifying the National Association of Securities Dealers (NASD) as “a gatekeeper to participation in the securities industry”). And statutory schemes require that persons or entities undertaking a regul­ated activity be members of the relevant SRO, which has the powers to approve and expel members. 51 See id. at 38–39 (describing the relevant statutory provisions for NASD); see also Macey & Novogrod, supra note 1, at 963–66 (arguing FINRA’s ability to expel members is its most important disciplinary tool and noting its ability to do so is dependent on the market power of the SRO).

Notwithstanding these common membership attributes, SROs differ significantly in organizational form. Many securities exchanges and clearinghouses, for example, are private businesses owned by share­holders. 52 Johnson, supra note 7, at 204. Other SROs are not-for-profit organizations run by executive officers and governed by multimember boards. 53 Notable examples are FINRA, NERC, and the NFA. See FINRA, About FINRA, supra note 1; Who We Are, Nat’l Futures Ass’n, supra note 2; Governance, N. Am. Elec. Reliability Corp., http://www.nerc.com/gov/Pages/default.aspx [http://perma.cc/4JP8-98UN] [hereinafter NERC, Governance] (last visited Aug. 2, 2016). SROs generally have significant authority to determine their own governance structures, 54 Justice Douglas, a former SEC chair, has referenced the autonomy of financial SROs, with the SEC playing “an essentially passive role, allowing the securities industry to govern itself in its own wisdom.” See William I. Friedman, The Fourteenth Amendment’s Public/Private Distinction Among Securities Regulators in the U.S. Marketplace—Revisited, 23 Ann. Rev. Banking & Fin. L. 727, 740 (2004) [hereinafter Friedman, Revisited] (describing this view and collecting sources). though some are subject to statutory or regulatory requirements aimed at ensuring fair stakeholder representation, such as board-composition criteria. 55 E.g., 16 U.S.C. § 824o(c)(2)(A) (2012) (requiring that FERC’s reliability SRO establish rules “assur[ing] its independence of the users and owners and operators of the bulk-power system, while assuring fair stakeholder representation in the selection of its directors and balanced decisionmaking in any ERO [electricity reliability organization] committee or subordinate organizational structure”). But see Johnson, supra note 7, at 201 (noting exchanges and clearinghouses “retained significant autonomy to determine the fundamental elements of their operating policies and governance structure”). In the parlance of Professor Mashaw, such schemes impose accountability regimes borrowed from public governance on imperfect market- or social-accountability regimes. See Mashaw, supra note 17, at 136 (“Public law accountability might, of course, be imposed by statute, regulation, or contract.”).

Third, for carrying out their rulemaking and enforcement duties, SROs have governance layers that in some respects mimic those present in most federal agencies. 56 For a fine-grained structural description of the EPA’s rulemaking process illustrating the many layers of an agency’s rule development, see generally Thomas O. McGarity, The Internal Structure of EPA Rulemaking, Law & Contemp. Probs., Autumn 1991, at 57. For a description of governance layers attendant to formal agency adjudication, see William F. Funk, Sidney A. Shapiro & Russell L. Weaver, Administrative Procedure and Practice 207–08 (4th ed. 2010). Below the highest level of governance are compliance and rules-development bodies (among many others), which are further broken into subcommittees and working groups. 57 E.g., NERC, Governance, supra note 53 (listing NERC’s committees, subcom­mittees, working groups, and task forces); see also Committees, Fin. Indus. Regulatory Auth., http://www.finra.org/about/committees [http://perma.cc/GMX3-9NHP] (last visited Aug. 2, 2016) (describing various committees); Committees, Nat’l Futures Ass’n, http://www.
nfa.futures.org/NFA-about-nfa/committees [http://perma.cc/GKH7-7MR8] (last visited Aug. 2, 2016) (providing links to each committee).
As described in more detail below, for specific SROs, these lower governance levels are responsible for early rule development, including drafting and vetting, as well as early investigatory and initial enforcement activities. 58 For details, see infra Part II. These activities filter up through the SRO’s ranks until one or more boards vote (in the case of rulemaking) or the internal adjud­icatory process is complete (in the case of enforcement). 59 This means a full appeals process can take years before a court ultimately resolves the matter. See, e.g., Paz Sec., Inc. v. SEC, 494 F.3d 1059, 1062, 1066 (D.C. Cir. 2007) (showing a nearly four-year period from initial action to appellate ruling remanding the matter for further proceedings before the SEC).

Finally, the SROs’ actions are subject to the oversight of a federal agency. For example, SRO rules must be filed with the oversight agency, which retains at least some authority to approve or disapprove the rules—though, as discussed below, there is significant and meaningful variation in the applicable presumptions and standards. 60 See infra Part II (providing case studies). Further, whether the oversight agency has plenary authority to issue or modify rules itself varies by statutory scheme—a nuance that courts and scholars alike often overlook. 61 The apparently inaccurate presumption seems to be that agencies always retain their plenary authority. See, e.g., Michael, supra note 3, at 176 (stating the delegating agency retains “the powers of review and independent action”). Similarly, aggrieved persons may appeal from an SRO adjudication to the oversight agency, which can reject or approve the disposition. 62 See Macey & Novogrod, supra note 1, at 970 n.42 (giving an illustrative example of the appeals process for FINRA disciplinary actions). Again, however, there is significant variation in this regard, particularly relating to the standard of review and the scope of the over­sight agency’s powers to modify the SRO’s disposition. 63 See infra Part II (giving a variety of examples). Only aggrieved persons—which do not include the SROs themselves—may seek review of an adverse agency decision before a federal court. 64 See, e.g., Nat’l Ass’n of Sec. Dealers, Inc. v. SEC, 431 F.3d 803, 810 (D.C. Cir. 2005) (finding NASD has no right under the Securities Exchange Act to appeal SEC reversal of its disci­plinary actions). This is a general rule in administrative law; Congress must expressly say so if it intends for an agency acting in its governmental capacity to have standing to challenge a reviewing body’s decision. See Dir., Office of Workers’ Comp. Programs, Dep’t of Labor v. Newport News Shipbuilding & Dry Dock Co., 514 U.S. 122, 127–30 (1995) (“[T]he phrase ‘person adversely affected or aggrieved’ does not refer to an agency acting in its governmental capacity.” (quoting 33 U.S.C. § 921(c) (1994))); cf. United States v. Interstate Commerce Comm’n, 337 U.S. 426, 430 (1949) (holding the federal government as a market participant may be “aggrieved” because of its nongovern­mental capacity).

B. The Theory of SROs

With this understanding of SRO structure in place, this Article now turns to the rationales for the SRO model and explores the normative concerns expressed in the literature. As is evident from the history of New Deal compromises described above, 65 See supra text accompanying notes 38–40 (describing history). SROs are favored for their political expediency. 66 See Michael, supra note 3, at 181 (explaining regulated entities are more likely to accept self-regulation); see also Friedman, Revisited, supra note 54, at 738 (“Through [the Securities Exchange Act of 1934], Congress achieved a historic compromise between the public and private powers of Washington and Wall Street over the future of regulation in the securities markets.”). Not only was this true during that time of immense changes for the regulatory state, but modernly it offers an alternative to regulatory heavy handedness by providing the compromise of regulation carried out by the regulated industry itself. 67 See Omarova, Wall Street, supra note 7, at 427 (describing SROs as consistent with the challenge “New Governance” poses to “the old dogma that the administrative state is . . . the sole locus of power to regulate and that the private sector is a passive recipient” of regulation); Richard B. Stewart, Madison’s Nightmare, 57 U. Chi. L. Rev. 335, 338 (1990) (exploring how relaxation of separation-of-powers limitations frees the govern­ment to meet demands for “economic stabilization, growth, and economic justice”). In fact, SRO legislating tends to happen in response to crises, when a quick and politically expedient solution is especially appealing. 68 See, e.g., John C. Coffee, Jr., The Political Economy of Dodd-Frank: Why Financial Reform Tends to Be Frustrated and Systemic Risk Perpetuated, 97 Cornell L. Rev. 1019, 1020 (2012) (“A good crisis should never go to waste.”); Joel Seligman, Cautious Evolution or Perennial Irresolution: Stock Market Self-Regulation During the First Seventy Years of the Securities and Exchange Commission, 59 Bus. Law. 1347, 1348 (2004) (noting changes occur “when stock market discipline has broken down”); see also Omarova, Rethinking, supra note 7, at 695 (describing securities SROs as “a historical product of political compromise and economic expediency”).

The government’s economic expediency is another pragmatic benefit of SROs. Put frankly, today’s major oversight agencies could not themselves assume the responsibilities of their SROs without extra­ordinary increases in their staffing and budgets. 69 See Fischer, supra note 7, at 80 (“From a rational perspective, it would be practically impossible for government regulators alone to adequately oversee the industry considering the scope of their mandate.”). Even during the New Deal, this benefit of the SEC–SRO relationship provided an important incentive to lawmakers in choosing the SRO form. 70 See Onnig H. Dombalagian, Self and Self-Regulation: Resolving the SRO Identity Crisis, 1 Brook. J. Corp. Fin. & Com. L. 317, 323 (2007) (“[T]he [SRO] concept has endured because lawmakers have generally regarded self-regulation to be a practical and efficient way to outsource the burdens of regulation to the private sector.”); cf. Hammond & Markell, supra note 16, at 332 (explaining the EPA’s inability to take over state imple­mentation of federal environmental programs should the agency withdraw the states’ authority to implement those programs). Other benefits, however, have a more theoretical basis. Because SROs are industry composed, their expertise is one of their important virtues. 71 See Johnson, supra note 7, at 189 (“SROs have unique expertise and sophistication.”). After all, in the parlance of the regulatory state, the discourse typically concerns comparative expertise. The term “expertise” is worth further exploration here. 72 Cf. Hammond, Super Deference, supra note 28, at 739–56 (describing the comparative institutional expertise of courts and federal agencies); Sidney A. Shapiro, The Failure to Understand Expertise in Administrative Law: The Problem and the Consequences, 50 Wake Forest L. Rev. 1097, 1105–17 (2015) (providing a meticulous account of meaning of agency expertise). Importantly, the unstated assumption underlying the expertise rationale is that the SRO has greater expertise than the oversight agency. This view is evident, for example, in the observation that SROs have better and timelier access to information, as well as better capabilities for understanding and analyzing information. 73 Omarova, Wall Street, supra note 7, at 433 (asserting “private industry actors have an important potential advantage over government regulators” to “identify, analyze, and assess systemic implications of underlying trends in the financial markets”). Given this closeness to the data, SROs may be more adept than their agencies are at developing carefully tailored rules, achieving compliance, and exercising enforcement discretion. 74 Michael, supra note 3, at 181 (listing several distinct advantages that SROs have in self-regulation over agency regulation). Overall, these benefits translate to efficiency and nimbleness.

SROs also promote industry self-preservation. In addition to developing rules of behavior, when SROs undertake enforcement actions, they protect their industry’s credibility. 75 One commentator has noted the SRO model may be especially effective for emerging industries and new technologies—such as crowdfunding—because SROs are highly incentivized to protect their industry’s credibility and are likely more zealous in their regulatory roles. Sigar, supra note 12, at 501. Indeed, enforcement is particularly noteworthy because it holds the potential to tell an anticap­ture story. If an SRO has monopoly power over an industry and is a true gatekeeper to activity within that industry, it holds the ultimate stick: expulsion from membership. 76 Indeed, Professor Jonathan Macey and Caroline Novogrod have argued that the power of expulsion is the most important disciplinary tool available to FINRA. Macey & Novogrod, supra note 1, at 965–66. However, the SRO must have market power over its industry by virtue of a robust membership; otherwise, expelled members can shift to another market. Id. at 966–67. Professors William Birdthistle and Todd Henderson, for example, emphasize that the self-regulatory model is well suited to circumstances in which an entire industry bears the costs of bad actors but the benefits of bad behavior go only to a few. 77 Birdthistle & Henderson, supra note 7, at 8.

There are softer norms at work in the self-preservation rationale. SROs can foster community-minded ethics, particular behavioral cultures, or special attributes of professionalism. 78 See id. at 62–64 (describing FINRA Rule 2010 requiring “high standards of commercial honor”); see also FINRA Rules Section 2010: Standards of Commercial Honor and Principles of Trade, Fin. Indus. Regulatory Auth., http://finra.complinet.com/en/
display/display_main.html?rbid=2403&element_id=5504 [http://perma.cc/NK3W-58ZA] (last visited Aug. 2, 2016). In this way, FINRA trades some public-governance accountability for a social-accountability regime. See Mashaw, supra note 17, at 134–35 (describing contracted-out governance as sometimes imposing a social-accountability regime).
These norms are very difficult for agencies to achieve through traditional regulatory means because they are somewhat inchoate, but they can be part of an industry’s culture and perhaps are a prerequisite to the development of voluntary standards. Professor Saule Omarova has demonstrated, for example, how self-regulation can develop to promote an industry’s need to further “morality,” particularly for high-risk areas involving health and safety in which the industry has a common fate. 79 See Omarova, Wall Street, supra note 7, at 447–50 (describing the Institute of Nuclear Power Operations (INPO), an industry effort to promote nuclear safety following Three Mile Island, and Responsible Care, the chemical manufacturing industry’s global response to the Bhopal accident). Professor Omarova’s examples are not purely the audited self-regulatory regimes treated in this Article, but they are helpful both for elaborating the potential origins of such models and for suggesting how they can be retained in a transition to a full SRO approach.

SRO schemes may also be helpful in alleviating international regulatory issues, though this benefit is highly context dependent. The North American electric grid, for example, is interconnected from Canada to the United States and Mexico; 80 Key Players, N. Am. Elec. Reliability Corp., http://www.nerc.com/AboutNERC/
keyplayers/Pages/default.aspx [http://perma.cc/5W3E-6AY3] (last visited Sep. 19, 2016).
the legislative history of the Energy Policy Act of 2005 suggests that the drafters intended partly that the SRO model would assuage concerns that the Federal Energy Regulatory Commission (FERC) could not dictate reliability standards to other sovereign nations. 81 148 Cong. Rec. 3217–18 (2002) (statement of Sen. Thomas). Similarly, Professor Omarova has argued that the thorny problems of global governance in the financial sector suggest an advantage for SROs, which can better monitor and mitigate financial risk across borders. 82 Omarova, Wall Street, supra note 7, at 436–37.

SROs’ strengths are also their weaknesses. The same industry involvement promoting expertise can also lead to inadequate enforce­ment and anticompetitive conduct, potentially leading to suboptimal standards. 83 See Michael, supra note 3, at 189; see also Concept Release Concerning Self-Regulation, 69 Fed. Reg. 71,256, 71,256 (Dec. 8, 2004) (codified at 17 C.F.R. pt. 240) (“Inherent in self-regulation is the conflict of interest that exists when an organization both serves the commercial interests of and regulates its members or users.”); Birdthistle & Henderson, supra note 7, at 10 (describing the possibility that larger firms will exert disproportionate influence within SROs and squeeze out smaller firms); Johnson, supra note 7, at 189–90 (“Members’ incentives frequently diverge from SROs’ regulatory objectives.”). Scholarly literature on financial regulation describes another important concern: cartelization—the transfer of wealth from the investors, who are the beneficiaries of regulation, to the brokers, who are the SRO members. 84 See Birdthistle & Henderson, supra note 7, at 12. Further, shifting some of the work of government to the private sector can create special accountability problems, 85 See Freeman, Extending Public Law Norms, supra note 6, at 1304 (“Public law scholars worry that privatization may enable government to avoid its traditional legal obligations, leading to an erosion of public norms and a systematic failure of public accountability.”). including procedures that are more difficult to access, lack of transparency, and reduced opportunities for public (as opposed to regulated parties’) engagement. 86 See infra Part II (detailing membership composition and voting rights); see also Angela J. Campbell, Self-Regulation and the Media, 51 Fed. Comm. L.J. 711, 715–19 (1999) (providing a general summary of arguments in favor of and in opposition to self-regulation).

Given these concerns, there is broad consensus in the literature that SROs need strong agency oversight. 87 See, e.g., Michael, supra note 3, at 243–44 (“The agency also should have independent enforcement authority over all regulated entities and independent rulemaking authority for the self-regulatory organization.”); Omarova, Wall Street, supra note 7, at 483 (“[T]here must be a strong regulatory and supervisory framework in whose shadow such self-regulation operates.”). And indeed, agency oversight forms the entire basis of SROs’ constitutional legitimacy, as described next.

C. SROs in the Constitutional and Common Law Schemes

SROs fit awkwardly within the constitutional structure because they spring from congressional delegations of power to both agencies and private entities. At first glance, the private component of such arrangements seems to directly run afoul of the prohibition on private delegation. 88 See Dep’t of Transp. v. Ass’n of Am. R.Rs., 135 S. Ct. 1225, 1237 (2015) (Alito, J., concurring) (“Even the United States accepts that Congress cannot delegate regulatory authority to a private entity.” (internal quotation marks omitted) (quoting Ass’n of Am. R.Rs. v. U.S. Dep’t of Transp., 721 F.3d 666, 670 (D.C. Cir. 2013))). The private nondelega­tion doctrine is distinguishable from the nondelegation doctrine concerning whether Congress has set forth intelligible principles in an agency’s statutory mandate. The latter relates only to the relationship between courts and agencies, while the former adds a private party to the mix. Cf. Whitman v. Am. Trucking Ass’ns, 531 U.S. 457, 474 (2001) (describing the nondele­gation doctrine’s intelligible-principle standard and collecting examples); J.W. Hampton, Jr., & Co. v. United States, 276 U.S. 394, 409 (1928) (establish­ing the intelligible-principle standard). As shown below, however, this doctrine provides no mean­ingful constraint on modern SROs. Because SROs are private entities, moreover, they are not typically considered government actors and are therefore not answerable for constitutional violations. 89 See infra section I.C.2 (describing the state action doctrine). But counter­intuitively, they are not sufficiently private to be answerable for common law claims. 90 See infra section I.C.3 (discussing common law parameters). In other words, none of these sources of law offers a meaningful constraint on SROs. 91 Antitrust law offers a possible, but in reality unlikely, constraint. See Credit Suisse Sec. (USA) LLC v. Billing, 551 U.S. 264, 279–85 (2007) (giving the SEC broad authority to displace antitrust law); see also Michael, supra note 3, at 198–201 (writing prior to Credit Suisse and describing the limited relevance of antitrust law as a restraint).

1. The Private Nondelegation Doctrine. — The private nondelegation doctrine provides that Congress may not delegate to private entities the authority to create binding law. 92 See Carter v. Carter Coal Co., 298 U.S. 238, 311 (1936) (opining that legislative delegation to a private party is the “most obnoxious form” of delegation). Given the nature of SROs, the doctrine’s prohibition seems directly applicable. But courts and scholars have recently questioned both the history of the doctrine and its continuing existence. Section I.C.1.a below describes the historical roots of the doctrine and largely attributes the current debate to an unfortunate lack of linguistic precision. The best view is that the private nondelegation doctrine serves important purposes distinct from other related theories, procedural due process in particular. Section I.C.1.b then turns to the modern contours of the doctrine, concluding that notwithstanding its important animating principles, it serves as only a very limited check on SROs.

a. Historical Roots and Modern Relevance. — The private nondelega­tion doctrine is typically attributed to the 1936 decision Carter v. Carter Coal Co., in which the Supreme Court struck down a statute permitting certain members of the coal industry to create binding law regarding, among other things, coal prices. 93 Id. Some readers may wonder how the opinion in A.L.A. Schechter Poultry Corp. v. United States, 295 U.S. 495 (1935), bears on private nondelegation. In Schechter Poultry, the Court struck down provisions of the National Industrial Recovery Act providing for the President to either approve industry-developed codes regulating competition or develop such codes himself. Id. at 521–23. The Court held that these provisions violated the nondelegation principle and did not consider either private nondelegation or due process arguments. Id. at 537–42. Nevertheless, in dictum the Court foreshadowed its decision in Carter Coal. Id. at 537 (“Such a delegation of legislative power [to industry or trade groups] is unknown to our law and is utterly inconsistent with the constitutional prerogatives and duties of Congress.”).
This Article aims at federal-level governance, but there is an analogous prohibition at the state-government level grounded in the Fourteenth Amendment’s Due Process Clause. U.S. Const. amend. XIV, § 1; Metzger, supra note 6, at 1437. Private-citizen-enforcement suits have also been criticized on private nondelegation grounds. See Matthew C. Stephenson, Public Regulation of Private Enforcement: The Case for Expanding the Role of Administrative Agencies, 91 Va. L. Rev. 93, 95 (2005) (arguing the executive branch should exercise more oversight of private-citizen suits); cf. Friends of the Earth, Inc. v. Laidlaw Envtl. Servs. (TOC), Inc., 528 U.S. 167, 197 (2000) (Kennedy, J., concurring) (observing citizen-suit provisions that exact fines and enforce federal law raise “[d]ifficult and fundamental” questions about delegating executive power).
The Court described this scheme of legislative delegations to private entities to be the “most obnoxious form” because it permitted a private majority to impose its will on an “unwilling minority”—who were also business competitors. 94 Carter Coal, 298 U.S. at 311.

On the surface, this explanation appears to be directed at anticompetitive behavior, but it is better understood as a constitutionally rooted concern about fundamental fairness. This conclusion is logical: If a private group imposes its will on others using the force of law—esp­ecially in a self-interested way—the action lacks fundamental fairness. 95 Indeed, the D.C. Circuit espoused this reading in Association of American Railroads v. U.S. Department of Transportation (Amtrak II), 821 F.3d 19, 28 (D.C. Cir. 2016) (“[W]hat primarily drives the Court to strike down this provision is the self-interested character of the delegatees’ [sic].”). But the Court’s focus on fairness blurs the doctrinal basis for its decision. Although the case is still cited as the source of the private nondelegation doctrine, 96 Dep’t of Transp. v. Ass’n of Am. R.Rs., 135 S. Ct. 1225, 1252 (2015). its wording reflects a Lochner-era fixation on due process. The relevant passage adds the following:

The difference between producing coal and regulating its production is, of course, fundamental. The former is a private activity; the latter is necessarily a governmental function, since, in the very nature of things, one person may not be entrusted with the power to regulate the business of another, and especially of a competitor. And a statute which attempts to confer such power undertakes an intolerable and unconstitutional interference with personal liberty and private property. The delegation is so clearly arbitrary, and so clearly a denial of rights safeguarded by the due process clause of the Fifth Amendment, that it is unnecessary to do more than refer to decisions of this court which foreclose the question. 97 Carter Coal, 298 U.S. at 311.

As is evident, the Court directly referred to “due process,” but it did so in connection with “personal liberty and private property.” 98 Id. Given its Lochner heritage, this passage has led at least one court to describe the doctrinal basis as fitting most comfortably within substantive due process. 99 See Synar v. United States, 626 F. Supp. 1374, 1383 n.8 (D.D.C. 1986) (per curiam, joined by Scalia, J.) (noting “the [Carter Coal] holding appears to rest primarily upon denial of substantive due process rights”). Under a modern understanding of due process, however, this passage is better viewed as relating to procedural due process because it references protected interests 100 See U.S. Const. amend. V (protecting “life, liberty, or property”); Bd. of Regents of State Colls. v. Roth, 408 U.S. 564, 569 (1972) (“The requirements of procedural due process apply only to the deprivation of interests encompassed by the Fourteenth Amendment’s protection of liberty and property.”). and the need for neutral decision­making. 101 See Withrow v. Larkin, 421 U.S. 35, 47, 51–52 (1976) (noting a presumption of honesty in favor of administrative adjudicators as well as the variety of means employed by Congress and agencies to avoid separation of functions issues).

Unfortunately, courts and scholars do not necessarily distinguish between substantive and procedural due process when discussing this understanding. 102 See, e.g. Alexander “Sasha” Volokh, The Shadow Debate over Private Nondelegation in DOT v. Association of American Railroads, 2015 Cato Sup. Ct. Rev. 359, 372 & n.74 (describing dictum referencing substantive due process); id. at 374–75 (arguing Carter Coal is best understood as a due process case but not specifying which kind); id. at 376 (“[T]he presence of procedures can satisfy due process.”); see also Amtrak II, 821 F.3d 19, 27–31 (D.C. Cir. 2016) (striking down the statute on due process grounds without specifying which kind); Froomkin, supra note 35, at 153 (describing Carter Coal as “rooted in a prohibition against self-interested regulation that sounds more in the Due Process Clause than in the separation of powers”); Michael, supra note 3, at 196 (“The fundamental issue is . . . due process.”). This lack of specificity only further obscures the mod­ern relevance of Carter Coal. The bottom line is that, notwithstanding the procedural due process rationale, the private nondelegation doctrine should not be considered superfluous. 103 Several scholars have suggested otherwise. See Harold I. Abramson, A Fifth Branch of Government: The Private Regulators and Their Constitutionality, 16 Hastings Const. L.Q. 165, 209 & n.260 (1989) (noting scholars have suggested “courts should replace the nondelegation doctrine with the Due Process Clause”); Volokh, supra note 102, at 370 (referring to “the “[]nonexistence of the[] private nondelegation doctrine”). For one, classical procedural due process is limited in its scope: It restricts individualized decision­making (that is, adjudication in modern administrative law parlance 104 Although it is theoretically possible that procedural due process would constrain some agency rulemaking, adjudication more typically fits the requirement of individualized decisionmaking. See Bi-Metallic Inv. Co. v. State Bd. of Equalization, 239 U.S. 441, 445–46 (1915) (distinguishing legislative actions, not subject to procedural due process strictures, from actions that may trigger due process because of characteristics including, among other things, individualized decisionmaking); cf. Vt. Yankee Nuclear Power Corp. v. Nat. Res. Def. Council, Inc., 435 U.S. 519, 542 (1978) (reversing the court of appeals’s imposition of non-APA procedures on agency rulemaking and noting the possibility that there could be due process constraints). ), and violations are typically remedied with better procedures. 105 See Mathews v. Eldridge, 424 U.S. 319, 334 (1976) (“[W]hether the administrative procedures provided here are constitutionally sufficient requires analysis of the governmental and private interests that are affected.”); see also Gary Lawson et al., “Oh Lord, Please Don’t Let Me Be Misunderstood!”: Rediscovering the Mathews v. Eldridge and Penn Central Frameworks, 81 Notre Dame L. Rev. 1, 5 (2005) (referring to due process law’s “ultimate touchstone” as “fairness”). On the other hand, there is some support for the view that a pecuniary interest in a proceeding has a much higher burden to overcome. Withrow, 421 U.S. at 47 (listing, in dictum, pecuniary interest as a situation “in which experience teaches that the probability of actual bias on the part of the judge or decisionmaker is too high to be constitutionally tolerable”). Moreover, Carter Coal’s emphasis on fairness fits well with administrative law’s concern for checking arbitrariness because a process that is arbitrary is at its core unfair. 106 See Bressman, Beyond Accountability, supra note 18, at 496 (“At a basic level, arbitrary administrative decisionmaking is not rational, predictable, or fair.”).

Arbitrariness is a helpful lens through which to view Justice Alito’s much more recent comments on private nondelegation in Department of Transportation v. Ass’n of American Railroads. 107 135 S. Ct. 1225, 1234 (2015) (Alito, J., concurring). There, a freight railroad industry group challenged a statute giving Amtrak and the Federal Railroad Administration (FRA) joint authority to issue metrics and stan­dards governing the passenger railroad’s performance and schedules. 108 Id. at 1235. The D.C. Circuit held that Amtrak was a private entity and invalidated the scheme on private nondelegation grounds. 109 Ass’n of Am. R.Rs. v. U.S. Dep’t of Transp., 721 F.3d 666, 677 (2013), rev’d and remanded, 135 S. Ct. 1225 (2015). Acknowledging that it may be permissible for private entities to help agencies make a decision—a point to which this Article returns in a moment—the court considered Amtrak’s authority much too drastic. 110 Id. at 671 (“[T]he Supreme Court has never approved a regulatory scheme that so drastically empowers a private entity in the way § 207 empowers Amtrak.”). Here, Amtrak crafted its own regulations and held authority equal to the Administration, leaving the agency “impotent” to act without Amtrak’s permission. 111 Id. In short, the court called the statute “as close to the blatantly unconstitutional scheme in Carter Coal as we have seen.” 112 Id. at 673. The court also suggested that the lack of a historical antecedent for such a scheme gave reasons to suspect its constitutionality. Id.

The Supreme Court reversed the circuit’s determination that Amtrak was a private entity, but it did not completely obviate the lower court’s analysis. 113 Ass’n of Am. R.Rs., 135 S. Ct. at 1231 (majority opinions). Although leaving the lower court to parse the issues on remand, it noted that there perhaps remained due process or nondel­egation concerns related to Amtrak’s authority over its own industry. 114 Id. at 1234 (describing the parties’ arguments). Justice Alito’s concurrence, moreover, expressed concern with a portion of the statute permitting an arbitrator to resolve disputes between the agency and Amtrak, emphasizing his view that if the arbitrator could be a private person, such delegation would also be unlawful. 115 Id. at 1237–38 (Alito, J., concurring). His analysis primarily invoked the vesting clauses of the Constitution, but he advanced a process-oriented rationale in noting that the Framers purposefully vested the deliberative process of the legislature within that body. 116 Id. at 1237. Indeed, he argued that private entities lack “even a fig leaf” of constitutional justification because they are not vested with any of the legislative, executive, or judicial powers. 117 Id. at 1237–38.

In addition to the process-oriented concerns, one might understand Justice Alito’s reasoning as rooted in an accountability rationale. If a private actor can make law but is not subject to the structural protections of the Constitution—because the actor is not part of the constitutional scheme at all—the constitutional accountability of the actor is simply nonexistent. 118 See Froomkin, supra note 35, at 146 (noting the private nondelegation doctrine’s “concern for proper sources and exercise of public authority promotes both the rule of law and accountability”). Taking this understanding together with that of Carter Coal, then, the private nondelegation doctrine is concerned with both arbitrariness and accountability.

These constitutional principles are evidenced somewhat curiously in the D.C. Circuit’s opinion on remand in Amtrak II. 119 821 F.3d 19 (D.C. Cir. 2016). The court struck down the relevant statute, holding that giving self-interested entities regulatory authority over their competitors violates due process. 120 Id. at 23. The court also held that the arbitrator provision violated the Appointments Clause. Id. at 37–38. Rely­ing once again on Carter Coal, the court explained that although the case could be read as either a private nondelegation or a due process decision (or perhaps both), the most important basis for the holding was the latter. 121 Id. at 28–29.

In applying this fairness rationale, the court emphasized Amtrak’s profit motives and role in developing standards to directly regulate its competitors. 122 Id. at 32–34. In other words, the court seemed to suggest that the statutory scheme lacked a neutral decisionmaker. 123 See id. at 29–31 (adopting Carter Coal’s position that “delegating legislative authority to official bodies is inoffensive because we presume those bodies are disinterested,” whereas private actors “may be and often are adverse to the interests of others” (emphasis omitted) (internal quotation marks omitted) (quoting Carter v. Carter Coal, 298 U.S. 238, 311 (1936))). But if the decision was truly grounded in procedural due process, the court might have grappled with that doctrine’s applicability as well as the actual proce­dures at issue. 124 Carter Coal did not consider whether there was individualized decisionmaking, but recall that the Court was far more concerned with substantive due process at the time. See supra text accompanying notes 98–99. The court rejected the government’s argument that the federal government retained significant authority over Amtrak, providing accountability. 125 Amtrack II, 821 F.3d at 30–31. As the court reasoned, the FRA was not required to safeguard Amtrak’s competitors’ interests, nor did it have the power to overrule Amtrak. 126 Id. at 32–34.

What is interesting about the Amtrak II opinion is that it reads as a private nondelegation case dressed up as procedural due process. The principles of fairness and accountability—central to the private nondelegation doctrine—took center stage in the court’s analysis. The court’s formalistic approach to fairness looked far more like a private nondelegation analysis (elaborated in more detail momentarily) than the typical due process analysis, which tends to be more functional. 127 The court did not engage in any real discussion of whether the relevant procedures and structural crafting of the institutional arrangement might actually meet the requirements of due process. Cf. Londoner v. City & County of Denver, 210 U.S. 373, 385–86 (1908) (holding due process applied and explaining “something more” was required for procedural due process, without specifying precise rectifying procedures).

It is important to remember that Amtrak II is not a private nondel­egation case, nor did the Supreme Court’s decision in Amtrak I topple the doctrine. A modern procedural due process reading of Carter Coal, however, confuses two distinct doctrines and obscures the primary points: One ought to be concerned about delegations to private entities both because they risk unfair (that is, arbitrary) decisionmaking in rulemaking and adjudication and because they risk diminishing government accountability. In short, the private nondelegation doctrine ought to have something to say about privatization in the form of SROs.

b. Modern Contours of the Private Nondelegation Doctrine. — Having established the theoretical underpinnings and continuing relevance of the private nondelegation doctrine, this section turns to the doctrine’s limitations as a meaningful constraint on SROs. The Court has constructed a formalistic approach to applying the doctrine that obscures the actual function of SROs. 128 In this sense, the private nondelegation doctrine is like some of its cousins at the boundary between constitutional and administrative law. See, e.g., Mistretta v. United States, 488 U.S. 361, 381–84 (1989) (applying a formalistic review of a legislative attempt to aggrandize power); Myers v. United States, 272 U.S. 52, 127–28 (1926) (holding Congress may not limit removal of purely executive officers on formal separation-of-powers grounds). More recently, the Court has deviated from formalism in various contexts. See, e.g., Free Enter. Fund v. Pub. Co. Accounting Oversight Bd., 561 U.S. 477, 484 (2010) (applying a functional approach to removals analysis); Morrison v. Olson, 487 U.S. 654, 670–72 (1988) (using the same analysis for both appointments and removals). However, it has not done so with private nondelegation. The Court’s approach, moreover, is only superficially correlated with the accountability and arbitrariness ration­ales of the doctrine. As a result, the private nondelegation is not a particularly meaningful source of restraint on SROs.

Consider two examples. In Sunshine Anthracite Coal Co. v. Adkins, the Court upheld Congress’s answer to Carter Coal. 129 310 U.S. 381, 396–98 (1940). The statute directed industry groups to propose minimum prices for coal and related stan­dards to the National Bituminous Coal Commission (a federal agency). The Commission would then approve, disapprove, or modify the proposals to ensure standards consistent with the statutory criteria. 130 Id. at 397–98. In upholding the scheme, the Court emphasized that the industry’s role was subordinate to the Commission; the Commission did not have final rulemaking authority, and the industry operated under the Commission’s surveillance. 131 Id. at 399.

Similarly, the Court in Currin v. Wallace upheld a statute conditioning a tobacco regulation’s operation on a two-thirds vote of tobacco growers. 132 306 U.S. 1, 6 (1939). Because the Secretary of Agriculture had already chosen the content of the regulation, the vote was not a private impos­ition of the majority over the minority but rather a procedural hurdle to Congress’s chosen policy as expressed in the statute. 133 Id. at 15.

These cases reflect an on–off approach that asks only whether the privatization at issue is subject to formal oversight by a federal agency. 134 See also Verkuil, Public Law Limitations, supra note 35, at 451 n.302 (noting “delegations to private hands seem to require only a formal set of oversight mechanisms”). Arguably, we might add a second criterion that the delegation not amount to private imposition of the majority over the minority, borrowing from Currin and Carter Coal. But this seems simply to reaffirm the arbitrariness and accountability rationales of the doctrine itself, and the criterion has never been a meaningful part of the analysis. Essentially, private involvement is permissible provided that involvement is subject to agency oversight and approval. The brief overview of SROs set forth above demonstrates that such schemes easily meet this test. 135 See supra Part I (providing overview). By definition, the statutory schemes within the scope of this Article provide that the oversight agencies retain at least some powers of review. And of course, the oversight inquiry does not probe what kind of oversight or whether the oversight is sufficient to guard against arbitrariness or promote accountability. 136 This point offers a way of reconciling the Amtrak II analysis with other analyses typical of private nondelegation doctrine challenges. In Amtrak II, the court looked more specifically at the form of the FRA’s oversight of Amtrak in a way that is distinguishable from the approach taken in other private nondelegation examples. Amtrak II, 821 F.3d 19, 30–32, 35 (D.C. Cir. 2016) (evaluating the degree of oversight to which Amtrak is subject and whether it is a self-interested entity given its regulatory authority over competitors). In short, the perfunctory “oversight” analysis glosses over the vast authority such entities actually exercise. 137 See e.g., Metzger, supra note 6, at 1440–41. For SROs, this means that positive constitutional law does not vindicate the const­itutional accountability and arbitrariness concerns animating the private nondelegation doctrine.

2. The State Action Doctrine. — Because SROs carry out government functions, it seems possible that their activities (as opposed to their structure) may be constrained by the rights-granting provisions of the Constitution. Ordinarily, of course, private actors are not so con­strained. 138 See, e.g., Shelley v. Kraemer, 334 U.S. 1, 13 (1948). But if their actions are “fairly attributable” to the govern­ment, they may be treated as state actors for institutional purposes. 139 Lugar v. Edmondson Oil Co., 457 U.S. 922, 937 (1982). The term “state actor” here is synonymous with “government actor.” In another case involving Amtrak, for example, the Supreme Court held that the railroad was part of the government for First Amendment purposes. 140 Lebron v. Nat’l R.R. Passenger Corp., 513 U.S. 374, 394–95, 399 (1995) (considering features of the nature of government-created and -controlled corporations). Therefore it seems logical that SROs—which have authority to make and enforce federal law—ought to be considered government actors for constitutional purposes. Indeed, the few scholars to have con­sidered the matter have simply presumed that this is the case. 141 See, e.g., Michael, supra note 3, at 197 (“[S]elf-regulatory organizations clearly are [state] acting.”). Professor Michael cites Abramson, supra note 103, at 213, which in turn cites Gibson v. Berryhill, 411 U.S. 564 (1973), which did not consider the issue.

Notwithstanding the intuitive appeal of this presumption, in practice the private status of SROs often shields them from constitutional claims, even if their alleged violations take place within the scope of their regulatory duties. 142 E.g., Desiderio v. Nat’l Ass’n of Sec. Dealers, Inc., 191 F.3d 198, 206–07 (2d Cir. 1999) (holding the NASD is not a government actor for constitutional purposes). A prominent example is Desiderio v. National Ass’n of Securities Dealers, Inc. 143 Id. at 198. In that case, a securities broker challenged an arbitration clause located in a form that she was required to sign to register with the National Association of Securities Dealers (NASD), an SRO. 144 Id. at 200. She argued that the clause violated her due process rights, among other things, and sought a declaratory judgment. 145 Id. The Second Circuit held that the SRO was not a government entity for constitutional purposes because (a) it is an entirely private entity 146 Id. at 206. and (b) its actions were not “fairly attributable” to the government. 147 Id. at 206–07. Regarding the latter point, the court determined that even though the SEC had approved the clause, that approval was insufficient to create the necessary nexus between the NASD and the SEC because the SEC had not exercised its coercive power or provided significant encouragement. 148 Id. at 207; see also Perpetual Sec., Inc. v. Tang, 290 F.3d 132, 138 (2d Cir. 2002) (reaching the same result on a different NASD arbitration clause and distinguishing Amtrak in Lebron because the government created Amtrak); Santos-Buch v. Fin. Indus. Regulatory Auth., 32 F. Supp. 3d 475, 484–85 (S.D.N.Y. 2014) (reaching a similar result with respect to FINRA’s SEC-approved amendment to a disclosure rule).

This example illustrates how difficult it is to predict outcomes around the state action doctrine, particularly in the SRO context. What if, for example, the SEC had authority to modify the clause but had not done so? What if it had? What if SEC regulations had required dispute-resolution clauses? Courts have not examined these questions, but they suggest underappreciated implications associated with SRO design. It is true that some courts have indeed suggested that SROs might at least be constrained by the procedural due process requirements of the Constitution. 149 Cf. Cody v. SEC, 693 F.3d 251, 257 (1st Cir. 2012) (declining to reach the issue because the statute required that FINRA provide “the substance of procedural due process”); Rooms v. SEC, 444 F.3d 1208, 1214 (10th Cir. 2006) (stating, without analysis, that due process requires the NASD to give fair warning prior to disciplining and holding that due process was met). But the point for our purposes is that the law is far from clear, requires a case-by-case analysis, and seems to shield SROs from con­stitutional liability. 150 Professor Gillian Metzger has offered a thoughtful rethinking of the private nondelegation doctrine as a way of remedying the inchoate and problematic aspects of the state action doctrine. See Metzger, supra note 6, at 1374.

Indeed, in Professor Gillian Metzger’s important study of privatization and constitutional accountability, she emphasizes that not only are the jurisprudential approaches to determining state action inchoate, but they also do not adequately account for the risk that privatization amounts to an illegitimate government effort to bypass the usual constitutional constraints. 151 Id. at 1412, 1420. As she explains, when governments remain closely involved in private entities’ implementation of programs, both the governments and the private entities risk being considered state actors with respect to how programs are implemented. 152 Id. at 1432. But when governments cede more authority to the private entities, both govern­ments and the private entities avoid the state actor label. 153 Id. This, of course, raises significant constitutional-accountability concerns. 154 Id. Professor Metzger also argues that the state action doctrine has been courts’ primary tool for monitoring nondelegation issues. Id. at 1443–44.

Professor Metzger’s prescription involves rethinking state actor jurisprudence in private delegation terms, such that the role of government oversight is emphasized. 155 Id. at 1470–71. Her thoughtful analysis has much to recommend it, but it does not really remedy the concerns identified in this Article with respect to SROs. First, Professor Metzger directs her strongest criticism at privatized government benefit programs, through which independent contractors provide important government bene­fits. 156 Id. at 1462 (stressing that a private entity’s role as part of a government program may cause significant harm, “particularly when privatization occurs in contexts where program participants or applicants have a great need for the government benefits and services at issue”). To capture these types of programs, she argues, agency law should inform the oversight analysis. 157 Id. at 1464–65. For SROs, identifying hypothetical power is not the problem; what is more concerning is how the power is used.

She also pays attention to how private delegations are structured, emphasizing that there must be some mechanism by which individuals can hold the government accountable for constitutional violations, even if they cannot directly assert such violations against a private entity exercising government power. 158 Id. at 1470. This includes statutory surrogates for constitutional claims and encompasses the adequacy of administrative-review mechanisms. 159 Id. at 1486, 1490. She emphasizes that government oversight is a critical part of ensuring accountability and suggests, among other things, that in the absence of other such constraints, courts might directly apply constitutional constraints to private delegates. 160 Id. at 1473, 1480–81. She acknowledges the many drawbacks of this approach. Id. at 1481.

With respect to SROs, government oversight is retained. Yet the reality of SROs is that they are constitutionally constrained only when an impacted individual uses the Constitution as a shield and only after the oversight agency has made the action its own. There is no opportunity to use the Constitution as a sword to affirmatively protect constitutional rights. For example, if a business contends that an SRO’s enforcement action violates the Constitution, it can raise that argument before the oversight agency on appeal and later through judicial review of the oversight agency’s action. 161 See supra section I.A (describing the usual appeals process in the SRO adjudica­tory setting). But if a business argues that an SRO-required form—issued pursuant to a rule approved by an oversight agency—violates the Constitution, the business, as described above, has no constitutional recourse against the SRO. It must wait for an enforce­ment action, assuming pre-enforcement review against the agency itself is not available.

And in any event, a constitutional-rights-based method of legit­imizing SROs is only partially satisfactory because (a) it takes place on a case-by-case basis and (b) it does little for the broad accountability and arbitrariness concerns identified by private nondelegation. Further, it does not at all attack the administrative law concerns that SROs raise about participation, deliberation, and accountability. 162 Professor Metzger notes that regulatory reforms may better address these concerns than constitutional reforms. Metzger, supra note 6, at 1452.

3. Common Law Liability. — If it is surprising to see that SROs are not government actors for constitutional purposes, it is baffling to discover that they are treated as government actors when they are sued for money damages arising out of common law claims. Courts regularly hold that SROs enjoy absolute or regulatory immunity from such suits for harms arising out of their government functions. 163 See, e.g., Standard Inv. Chartered, Inc. v. Nat’l Ass’n of Sec. Dealers, Inc., 637 F.3d 112, 116 (2d Cir. 2011) (holding the NASD was immune from liability in a lawsuit directed at misstatements in a proxy statement amending NASD bylaws to effectuate transferal of regulatory powers to FINRA because the consolidation was an exercise of the SRO’s delegated functions); D’Alessio v. NYSE, Inc., 258 F.3d 93, 106 (2d Cir. 2001) (“[T]he NYSE, when acting in its capacity as a[n] SRO, is entitled to immunity from suit when it engages in conduct consistent with the quasi-governmental powers delegated to it pursuant to the Exchange Act and the regulations and rules promulgated thereunder.”); Sparta Surgical Corp. v. Nat’l Ass’n of Sec. Dealers, Inc., 159 F.3d 1209, 1215 (9th Cir. 1998) (stating SROs “enjoy freedom from civil liability when they act[] in their regulatory capacity”), abrogated on other grounds by Merrill Lynch, Pierce, Fenner & Smith Inc. v. Manning, 136 S. Ct. 1562 (2016). See generally In re Series 7 Broker Qualification Exam Scoring Litig., 510 F. Supp. 2d 35 (D.D.C. 2007) (providing a detailed and thoughtful immunity analyses of NASD’s gatekeeping function). Commentators also regularly speak in immunity terms. See e.g., Jennifer M. Pacella, If the Shoe of the SEC Doesn’t Fit: Self-Regulatory Organizations and Absolute Immunity, 58 Wayne L. Rev. 201 passim (2012) (criticizing immunity); Jaclyn Freeman, Note, Limiting SRO Immunity to Mitigate Risky Behavior, 12 J. on Telecomm. & High Tech. L. 193, 212 (2014) (arguing for sharply circumscribed immunity). These principles are under considerable tension: How can an SRO be both a private actor and a government actor for the same action?

Consider Scher v. National Ass’n of Securities Dealers, Inc., in which the plaintiff sought money damages from the NASD for allegedly violating her Fifth Amendment privilege against self-incrimination and due process rights when it interviewed her under oath in connection with an investigation. 164 386 F. Supp. 2d 402, 404–06 (S.D.N.Y. 2005). The court rejected her claim on two competing grounds. First, the court held that the plaintiff had failed to show that the NASD—a private actor, whose creation was not mandated by statute and for which the government does not appoint members or serve on any of its boards—was a government actor constrained by the Constitution. 165 Id. at 407. Second, the court held that the NASD was shielded from liability for money damages for precisely the opposite reason. 166 Id. That is, the SRO engages in quasi-governmental conduct under the supervision of the SEC pursuant to the powers granted to the SEC under the Securities Exchange Act. 167 Id. at 406. When it interviewed the plaintiff, the NASD was investigating alleged violations of an SEC order—conduct going to the core of the government function. 168 Id. at 407.

The seeming inconsistency of these two grounds has not been lost on litigants, but the courts frequently reject invitations to reconsider either principle on that basis. The Scher court simply stated that it was “by no means ‘inconsistent’” because the labels “private” and “governmental” had different meanings in the different contexts. 169 Id. at 408. As another court has recognized, the “dual public/private status requires caution in reading precedents because a ruling that an exchange is pri­vate for a particular purpose does not necessarily mean that it is private for all purposes.” 170 Martens v. Smith Barney, Inc., 190 F.R.D. 134, 138 (S.D.N.Y. 1999) (citing United States v. Solomon, 509 F.2d 863 (2d Cir. 1975)).

The better view is that such suits are preempted—not that SROs enjoy governmental immunity. 171 See In re Series 7 Broker Qualification Exam Scoring Litig., 510 F. Supp. 2d 35, 46–47 (D.D.C. 2007). In other words, when SROs act pursuant to a carefully crafted federal statutory scheme, common law claims that would conflict with such schemes are preempted. 172 Id. at 47. Therefore, the organic statute setting forth the SRO scheme will provide the exclusive means of challenging the SRO’s actions. 173 Id. at 47–48. This view suggests that savings clauses in SRO statutes would be the most efficient way to preserve some common law liability when creating SRO schemes.

Regardless of the theoretical underpinnings, the practical result of preemption from common law claims and the state action doctrine is that SROs enjoy more freedom from suit—whether common law or constitutional—than their oversight agencies do. 174 Id. at 44 (“[T]he result of a sovereign-immunity based rationale may be that an SRO enjoys greater protection from suit than would the government.”). The result is that only statutory law and implementing regulations cabin SROs’ discretion. Stating generalized preferences for effective agency oversight and fair SRO procedures is easy; operationalizing these preferences is much more difficult. Moreover, this high level of generality obscures the many design details that both reinforce the problematic aspects of SRO governance and reveal new concerns. Identifying these issues is difficult without a careful look at the SRO–agency relationship across statutory schemes. It is to that task that this Article now turns.

II. SRO Procedures and Oversight-Agency Deference

To truly assess SROs from a normative administrative law perspective, it is necessary to roll up one’s sleeves and look closely at SRO procedures and SROs’ relationships with their oversight agencies. This section uses three case studies to illustrate a spectrum of arrangements: the SEC and FINRA, the CFTC and the NFA, and FERC and the North American Electric Reliability Corporation (NERC). This Article high­lights these examples for several reasons. First, they provide good tempo­ral coverage. As described above, financial regulation’s SRO heritage is long and deep, so the SEC and CFTC models provide historical coverage, while the FERC example is a relative newcomer. Notwithstanding their older pedigrees, the SEC and CFTC models diverge in important respects, and the FERC example is yet again different from the others. The unique constructions of these SRO schemes illustrate a full spectrum of design choices currently in operation. Finally, these examples have major implications for extraordinary swaths of the economy. 175 See supra notes 1–3 (describing the scope of these SRO schemes). The scope of the SROs’ regulatory authority is breathtaking on this metric alone.

As noted in section I.C.1 above, oversight of SROs by government agencies is of considerable importance to the constitutional validity of such privatization. But the superficial analysis of agency oversight obscures the many forms it takes. In reality, agency oversight is often limited in important ways, leaving more discretion—and perhaps, less accountability—to the SROs than the nondelegation cases might suggest. In describing the three examples here, this Article pays particular attention to the composition of the SROs’ members and governing bodies, the SROs’ rulemaking and adjudicatory procedures, and the role of the oversight agency with respect to either type of action. While a full assessment of the effectiveness of each scheme must be left to the subject-matter-specific literature, this Article also notes a few of the perennial issues arising within each scheme to give a sense of how the SROs work.

A. FINRA and the SEC

The SEC’s oversight of financial SROs dates to the Securities Exchange Act (the “Exchange Act”) and the New Deal 176 Securities Exchange Act of 1934, 15 U.S.C. § 78 (2012); see Birdthistle & Henderson, supra note 7, at 12–17 (describing major periods of financial self-regulation); Roberta S. Karmel, Should Securities Industry Self-Regulatory Organizations Be Considered Government Agencies?, 14 Stan. J.L. Bus. & Fin. 151, 159–70 (2008) [hereinafter Karmel, SROs as Agencies?] (providing a detailed overview of changes in financial regulation since New Deal). and continues today primarily in the relationship between the SEC and FINRA. 177 Previously, the NYSE Group, Inc. (NYSE) regulated exchanges, while the NASD regulated broker-dealers; both were SROs under SEC oversight. Congress consolidated both of these functions into FINRA in 2007. Order Approving Proposed Rule Change to Accommodate the Consolidation of the Member Firm Regulatory Functions of NASD and NYSE Regulation, Inc., Exchange Act Release No. 34–56145, 72 Fed. Reg. 42,169, 42,169–70 (Aug. 1, 2007). FINRA is an independent, not-for-profit organization tasked with investor protection and ensuring securities industry integrity. 178 See FINRA, About FINRA, supra note 1; Fin. Indus. Regulatory Auth., Get to Know Us 3 (2012) [hereinafter FINRA, Get to Know Us], http://www.finra.org/web/sitess/default/
Corporate/p118667.pdf [http://perma.cc/SQ63-SWMU] (last visited Aug. 2, 2016).
In addition to writing and enforcing rules for securities firms and brokers, it conducts compliance examinations and oversees trading activity for stocks, bonds, and options on the New York Stock Exchange and NASDAQ Stock Market, among others. 179 FINRA, About FINRA, supra note 1; FINRA, Get to Know Us, supra note 178, at 3. FINRA’s governing board must fairly represent members and include outsiders. 180 Here, “outsiders” means persons “representative of issuers and investors and not . . . associated with a member of the exchange, broker, or dealer.” 15 U.S.C. § 78f(b)(3); see also Karmel, SROs as Agencies?, supra note 176, at 160 (noting the SEC has recently required greater numbers of independent directors).

Proposals for new rules may come from any number of sources, including FINRA members and staff, as well as the SEC. 181 FINRA Rulemaking Process, Fin. Indus. Regulatory Auth., http://www.finra.org/industry/finra-rulemaking-process [http://perma.cc/Z2QU-6DQD] [hereinafter FINRA, Rulemaking Process] (last visited Aug. 2, 2016). It was only in 1975 that Congress granted the SEC the power to initiate rulemaking; before then, the SEC’s authority had been limited to approving or disapproving rules. See Karmel, SROs as Agencies?, supra note 176, at 159–60 (describing amendments to the Exchange Act). The proposals undergo rounds of drafting and revisions among FINRA’s departments and committees before they are submitted to the FINRA Board, which may then authorize a notice-and-comment period. 182 FINRA, Rulemaking Process, supra note 181. As already noted, the Board consists of members of the public as well as industry representatives; some of the public members, however, appear to have very close industry ties. See FINRA Board of Governors, Fin. Indus. Regulatory Auth., http://www.finra.org/about/finra-board-governors [http://
perma.cc/H3GA-KLZA] (last visited Aug. 2, 2016) (listing public members of the FINRA Board of Governors, including, among others, a former SEC commissioner and several “retired” individuals).
Following the comment period, FINRA staff will either revise the rule for the Board’s further action or file it with the SEC. 183 Note that FINRA also issues guidance documents. See, e.g., Fin. Indus. Regulatory Auth., Regulatory Notice 16-08, Contingency Offerings (Feb. 8, 2016), http://www.finra.org/sites/default/files/Regulatory-Notice-16-08.pdf [http://perma.cc/
F9ZG-GTB9].

When FINRA files a rule with the SEC, the statute contemplates that the SEC will review it, conduct notice and comment, and determine whether it conforms with the Exchange Act. 184 15 U.S.C. § 78s(b)(1) (articulating the notice-and-comment procedure); id. § 78s(b)(2)(C) (describing the standards for approval or disapproval). The SEC’s review also includes an assessment of anticompetitive impacts. See Lanny A. Schwartz, Suggestions for Procedural Reform in Securities Market Regulation, 1 Brook. J. Corp. Fin. & Com. L. 409, 419 (2006). Within forty-five days of publication for notice and comment, the SEC is directed to either approve or disapprove the rule or “institute proceedings to determine whether the rule should be disapproved.” 185 15 U.S.C. § 78s(b)(2)(A). The agency may not approve a proposed rule earlier than thirty days after such publication without good cause. 186 Id. § 78s(b)(2)(C)(iii). If the SEC institutes proceedings to consider the disapproval of a rule, it must provide notice and an opportunity for a hearing to the SRO within 180 days. 187 Id. § 78s(b)(2)(B). The timeframes are important; the statute provides that a proposed rule shall be deemed approved if the SEC fails to meet its deadlines—though the agency can extend the deadlines upon a finding that doing so is appropriate, accompanied by a statement of reasons, or with consent from the SRO. 188 Id. § 78s(b)(2)(A)(ii).

In practice, following the notice-and-comment period, the SEC requests FINRA’s responses to comments received. 189 FINRA, Rulemaking Process, supra note 181. If the agency has concerns about the rule, FINRA consents to an extension of time and the two entities enter into extended negotiations. 190 Karmel, SROs as Agencies?, supra note 176, at 172. The SEC almost never disapproves a rule; the “understanding” is that SEC review is deferen­tial. 191 See Omarova, Rethinking, supra note 7, at 695 (arguing that while the SEC has an independent statutory authority to regulate activities of broker-dealers and other market intermediaries directly, in reality it fully delegates these regulatory functions and merely “function[s] as the watchful guard and supervisor”); David G. Tittsworth, H.R. 4624: The Pitfalls of a Self-Regulatory Organization for Investment Advisers and Why User Fees Would Better Accomplish the Goal of Investment Adviser Accountability, 87 St. John’s L. Rev. 477, 486 (2013) (noting that “the SEC’s oversight of SRO rulemaking may have been largely deferential,” due to the SEC’s not being required to weigh in on the merits of SRO rules). In the usual case, in which the SEC approves the rule, the Commission finalizes the rule as it would any rule issued independently by the agency, accompanied, for example, by a statement of basis and purpose. 192 See 5 U.S.C. § 553(c) (2012). FINRA rules approved by the SEC have the force of law. 193 See Desiderio v. Nat’l Ass’n of Sec. Dealers, Inc., 191 F.3d 198, 201 (2d Cir. 1999) (holding FINRA rules approved by the SEC preempted conflicting state law); Charles Schwab & Co. v. Fin. Indus. Regulatory Auth. Inc., 861 F. Supp. 2d 1063, 1065 (N.D. Cal. 2012) (noting FINRA rules have “the force and effect of a federal regulation”). Although little used, the SEC also has authority to issue a rule that would “abrogate, add to, or delete from” an SRO rule following notice-and-comment procedures and an opportunity for an oral hearing; the new rule must conform to the APA’s requirements. 194 15 U.S.C. § 78s(c) (2012); see Roberta S. Karmel, Realizing the Dream of William O. Douglas—The Securities and Exchange Commission Takes Charge of Corporate Governance, 30 Del. J. Corp. L. 79, 92–93 (2005) (describing an SEC attempt in 1990 to adopt a rule under the Exchange Act that ultimately failed judicial review). Members can also petition the SEC for changes to SRO rules. Charles Schwab, 861 F. Supp. 2d at 1065.

The Exchange Act also requires SROs to enforce their rules, again subject to SEC oversight and the SEC’s ability to indepen­dently enforce the SROs’ or its own rules. 195 15 U.S.C. § 78s. The enforcement proceed­ings at an SRO begin with a hearing before a Hearing Panel, 196 Id. § 78o-3(h). the outcome of which either side may appeal to the FINRA National Adjudicatory Council. 197 See Charles Schwab, 861 F. Supp. 2d at 1066 (describing the procedures for reviewing the Hearing Panel’s decision). The FINRA Board has discretion to review the Council’s decision, and an aggrieved person may ultimately seek review from the SEC. 198 15 U.S.C. § 78s(d). The SEC reviews FINRA orders de novo and has further authority to affirm, modify, set aside, or remand an SRO’s sanction. 199 Id. § 78s(e). Thereafter, an aggrieved person may seek review in a federal court of appeals. 200 Id. § 78y(a).

The relationship between the SEC and its SROs has endured significant criticism over the years, and Congress has increasingly tightened the constraints on the SROs by granting additional powers to the SEC. 201 Birdthistle & Henderson, supra note 7, at 42–43; see also Karmel, SROs as Agencies?, supra note 176, at 154 (noting the Exchange Act “has been repeatedly amended to grant the SEC more control of SROs”). Of interest, the increasing SEC oversight over time has been motivated by recurring concerns about market abuses and scandals. Id. at 162–65. In Dodd-Frank, moreover, Congress added a new check on the entire scheme: It provided for mandatory reviews of the SEC–FINRA relationship. First, the Comptroller General (housed within the Government Accountability Office (GAO)) must submit a detailed report to Congress every three years that evaluates the SEC’s oversight of FINRA. 202 15 U.S.C. § 78d-9. Second, the SEC itself must hire an independent consultant “of high caliber and with expertise” to examine among other things the SEC’s relationship with all its SROs. 203 Dodd-Frank Wall Street Reform and Consumer Protection Act, Pub. L. No. 111-203, § 967, 124 Stat. 1376, 1913 (2010). Following the expert’s report, the SEC must provide a report to Congress every six months for a two-year period describing how the agency is responding to the consultant’s report. 204 Id. § 967(c).

These studies continue to paint a picture of deference and reveal ongoing challenges to effective agency oversight. A Comptroller General study concluded that although the SEC regularly reviews FINRA rules, it almost never reviews SRO governance-related matters, like executive compensation and transparency of governance. 205 U.S. Gov’t Accountability Office, GAO-12-625, Securities Regulation: Opportunities Exist to Improve SEC’s Oversight of the Financial Industry Regulatory Authority 7 tbl.1 (2012). And between 2009 and 2011, for example, the report identified only one instance of the SEC dis­approving a rule. 206 Id. app. I at 30.

Moreover, neither the SEC nor FINRA has a formal process for evaluating the effectiveness of existing rules. 207 Id. at 25. And the independent consultant’s report generally emphasized that the SEC should strengthen its oversight of SROs, develop a centralized and coordinated approach to its interactions with SROs, and strengthen its processes for SRO rule proposals. 208 Bos. Consulting Grp., U.S. Securities and Exchange Commission: Organizational Study and Reform 8 (Mar. 10, 2011), http://www.sec.gov/news/studies/2011/967study.
pdf [http://perma.cc/VR5U-2ZDV].
The report noted that the SEC has a challenging task in reviewing rules: Their volume and complexity has grown, and it is diffi­cult to ensure that they are described and analyzed sufficiently to ensure meaningful notice and comment. 209 Id. at 240.

In addition, the report noted that the SEC lacks statutory authority to ensure consistency of the rules across markets, such that SRO rules may conflict but still be compliant with the Exchange Act. 210 Id. In its fourth and final report on progress responding to the consultant’s study, the SEC stated that it has improved coordination of its oversight of rulemaking, for example, by improving electronic records and creating a dashboard for the most important rules. 211 SEC, Fourth Report on the Implementation of SEC Organizational Reform Recommendations 28–29 (2013), http://www.sec.gov/news/studies/2013/sec-organizational-reform-recommendations-043013.pdf [http://perma.cc/VR5U-2ZDV].

Overall, there are a number of points to keep in mind from the SEC–FINRA scheme when comparing the next two case studies. On its face, the SEC scheme seems to have the greatest oversight role for the agency, especially considering that it reviews both FINRA rules and orders de novo. But a deeper look reveals a number of disruptions to that story. For example, although the FINRA rulemaking procedures have the appearance of being similar to those of agencies, the participants are clearly stacked in favor of industry. And when those rules go to the SEC, there is a negotiation process between the agency and FINRA. FINRA responds to comments received, and the agency almost always approves rules, generally not making use of its authority to modify rules. Furthermore, proposed rules can become effective—binding law—if the SEC fails to meet its review deadlines. 212 See supra notes 193–198 and accompanying text.

This entire deferential interaction, moreover, is obscured by the procedures officially undertaken by the SEC, which amount to a variant of the APA’s notice-and-comment rulemaking. An interested party, authorized to comment on the proposed rule during the SEC’s notice-and-comment period, faces the daunting prospect of significant regul­atory inertia favoring FINRA’s rule. To be sure, proposed rules developed within agencies are also criticized on this basis, 213 See Wagner et al., supra note 4, at 128–32 (showing empirically that comments on complex rules “come predominantly from regulated industry, and the changes made to the . . . final rule[s] will track this imbalance and generally favor industry”). but here the cards are stacked all the more strongly because the force of FINRA is behind them.

Likewise, FINRA enforcement involves layers of internal appeals, similar to those often found at agencies, but with the disincentive to act too strongly to check members because the oversight agency could view an SRO enforcement action as anticompetitive. 214 See, e.g., 5 U.S.C. § 557 (2012) (providing for review within the agency of initial formal adjudicatory proceedings). If the SRO is indeed behaving anticompetitively, this is an important check, but over time it will result in reduced penalties, thereby decreasing the compliance incentives for all members. But consider a case in which the SRO is acting in good faith to check violations of its rules but then faces an order reducing the penalty. The SRO does not count as an aggrieved person and therefore cannot appeal an adverse decision at the agency before a federal court. 215 See supra note 64 (providing several cases discussing the issue of standing). Thus, there is no opportunity to correct a shift toward diminished compliance incentives through judicial review.

B. The NFA and the CFTC

As noted previously, 216 See supra note 2 (describing the roles of the CFTC and the NFA). the CFTC oversees the NFA pursuant to the Commodity Exchange Act and amendments thereto. 217 Commodity Exchange Act, 7 U.S.C. §§ 1–27f (2012). The CFTC oversees two types of SROs—exchanges such as the New York Mercantile Exchange and associations such as the NFA. To maintain analytical consistency with the other case studies, the focus here is on the relationship between the CFTC and the NFA, with similarities involving exchanges documented in the footnotes. NFA members are all industry participants, 218 NFA Manual/Rules, Nat’l Futures Ass’n, at art. VI, § 1, http://www.nfa.futures.
org/nfamanual/NFAManual.aspx (on file with the Columbia Law Review) (last visited Aug. 5, 2016).
but the governing board is comprised of a mix of market representatives, merchants, brokers, dealers, and public representatives. 219 Id. at art. VII, § 2A. Executives must also include public representatives. Id. at art. VIII, § 3(c)(v). No less than twenty percent of the governing board must be comprised of “qualified nonmembers or persons . . . not regulated by the association.” 7 U.S.C. § 21(b)(11). Although the Commodity Exchange Act requires that the NFA’s rules “assure a fair representation of its members in the adop­tion of any rule of the association or amendment thereto,” 220 7 U.S.C. § 21(b)(5). the actual processes by which rules are developed within the NFA are far more opaque than those of the other SROs in this set of case studies. 221 There is no description of the rulemaking process on the NFA’s website, nor are such procedures set forth in the NFA manual or bylaws.

Moreover, the statute itself sets a far more deferential tone than does the Exchange Act. When the NFA submits rules to the CFTC, the NFA “may make such rules effective ten days after receipt of such submission.” 222 7 U.S.C. § 21(j). The provisions for exchanges are similar, though perhaps even more deferential to the exchanges. See id. § 7a-2(c)(1) (stating that the exchanges may implement new rules or amendments by providing the CFTC a written certification that the rule complies with the relevant statute). New rules become effective ten days after the CFTC receives the certification unless the CFTC stays the certification “because there exist novel or complex issues that require additional time to analyze, an inadequate explanation by the submitting registered entity, or a potential inconsistency” with the statute and applicable regulations. Id. § 7a-2(c)(2). Although the CFTC may decide to review the rules, triggering a notice-and-hearing period and extension of time, 223 Id. § 21(j) (providing that the CFTC has 180 days or such time as the NFA agrees to conduct proceedings in review of the rule and up to one year within which to conclude said proceedings, after which the rule becomes effective). this ten-day presumption in favor of effectiveness is stunning. It places the burden of inertia on the CFTC and essentially presumes that the NFA’s proposed rules will be consistent with the statute. In practice, it is extremely rare that the CFTC finds problems with proposed rules. In a study of NFA rules proposed over a ten-year period, for example, one commentator determined that those rules went into effect without CFTC intervention nearly 100 percent of the time. 224 Fischer, supra note 7, at 97–98. However, the CFTC can abrogate and request alterations to futures-association rules, which it may not do for rules of registered entities. 7 U.S.C. § 21(k); Fischer, supra note 7, at 90–91.

Similar to adjudications by FINRA, the NFA’s enforcement actions have an internal appeals process and thereafter must be filed with the CFTC, 225 See MBH Commodity Advisors, Inc. v. Commodity Futures Trading Comm’n, 250 F.3d 1052, 1058–59 (7th Cir. 2001) (providing an example of this appeals process). which may affirm, set aside, or remand the action. 226 7 U.S.C. §§ 21(h)–(i). The CFTC can also reduce any penalty if it finds the penalty is “excessive or oppressive.” 227 Id. § 21(i)(2). Unlike the SEC’s review of FINRA adjudications, however, the CFTC has interpreted its statute to require use of the more deferential weight of the evidence standard when reviewing NFA ad­judications. 228 MBH Commodity Advisors, 250 F.3d at 1060–63 (applying Chevron deference to the CFTC’s interpretation of statutory procedures regarding standard of review); see also 17 C.F.R. § 171.34 (2016) (announcing this interpretation). In a decision upholding that interpretation, the Seventh Circuit reasoned that because the CFTC operates in an appellate cap­acity, a de novo standard would be inappropriate because it would ignore the SRO’s expertise in evaluating industry-related facts. 229 MBH Commodity Advisors, 250 F.3d at 1063. The court acknowledged that the SEC conducts a de novo review under similar statutory language in the Exchange Act, but suggested that if anything, the SEC’s interpretation might be a better candidate for unrea­sonableness because the legislative history suggested some deference would usu­ally be appropriate. 230 Id. at 1063–64.

Overall, the picture painted with respect to the CFTC is very different from the popular account of recent financial reform, which emphasizes increased accountability and transparency in the financial sector. 231 Of course, Dodd-Frank also significantly expanded the CFTC’s authority, directing it to regulate over-the-counter derivatives and swaps. Wall Street Transparency and Accountability Act of 2010, Pub. L. No. 111-203, tit. VII, 124 Stat. 1641 (codified as amended in scattered sections of 7 U.S.C. and 15 U.S.C.). The procedural provisions discussed here regarding the CFTC–NFA relationship were largely in place prior to Dodd-Frank and were not changed by that Act. See Fischer, supra note 7, at 73–78 (describing financial regulation before and after Dodd-Frank). First, it is difficult to identify even the procedures by which the NFA’s rules are developed. Second, agency oversight is so light-handed—both by statute and in practice—that the agency does not seem to have bargaining power to push the NFA toward openness. It is likely that considerable behind-the-scenes work takes place in developing the rules, but the lack of transparency makes it impossible to evaluate this pos­sibility either descriptively or normatively. 232 See Fischer, supra note 7, at 102 (“When the [agency–SRO] dialogue occurs outside the public eye, there is no way to determine whether oversight is actually meaningful or if the summary public approval of rule proposals that the study suggests reflects summary private approval as well.”). The standard of review for enforcement is likewise deferential, such that at the margins there is less opportunity to provide accountability. Finally, unlike the Dodd-Frank provisions for the SEC, those amendments did not establish any mech­anism for independent oversight of the CFTC–NFA relationship.

C. NERC and FERC

The FERC–NERC relationship, which was constituted under the Energy Policy Act (EPAct) of 2005, is the newest of the case studies. 233 16 U.S.C. § 824o (2012). As was true of financial SROs, NERC existed independently prior to 2005. Formed in 1962 as a voluntary industry association, NERC developed voluntary reliability standards for the increasingly interconnected electric grid—which now implicates issues related to national security and critical infrastructure protection. 234 See N. Am. Elec. Reliability Corp., History, at HIST-1-6 (Oct. 10, 2003) (on file with the Columbia Law Review) (setting forth a timeline). Although FERC relied heavily on NERC’s efforts over time, their relationship was rocky; as others have documen­ted, NERC viewed itself as having expertise superior to FERC and as a result resisted efforts to share data. 235 See Steve Isser, Electricity Restructuring in the United States: Markets and Policy from the 1978 Energy Act to the Present 151, 292 (2015) (providing examples of NERC’s resistance). This rocky history has likely contributed to some of the difficulties ironing out the SRO relationship described in this section. Nevertheless, the California energy crisis and the 2003 Northeast blackout—“the largest electrical power failure in the U.S. history” 236 Id. at 394. —prompted Congress to create a statutory framework mandating that reliability and cyber-security rules for the grid come from a FERC–SRO scheme. 237 For a detailed history of these events, see generally id. at 394–99 (describing Congress’s statutory response to the energy crisis).

Under the EPAct of 2005, NERC must establish rules of governance that “provide for reasonable notice and opportunity for public comment, due process, openness, and balancing of interests in developing reliability standards.” 238 16 U.S.C. § 824o(c)(2)(D). Should NERC wish to change its rules, it must submit them to FERC, which must provide for notice and comment and then make a finding that the rule is just and reasonable. 239 Id. § 824o(f). These governance rules are distinct from reliability rules, which are developed through a hybrid of procedures drawn from FERC requirements and the American National Standards Institute (ANSI) process (which NERC had used prior to becoming the reliability SRO). 240 See id. § 824o(d). To maintain ANSI accreditation, standard developers must comply with procedures governing a consensus development process. For details, see generally Am. Nat’l Standards Inst., ANSI Essential Requirements: Due Process Requirements for American National Standards (2016), http://share.ansi.org/shared%20
documents/Standards%20Activities/American%20National%20Standards/Procedures,%20Guides,%20and%20Forms/2016_ANSI_Essential_Requirements.pdf [http://perma.cc/
9VV4-UUDE].

Under this process, an industry-populated standards committee appoints teams that develop standards, define the standards’ scope and purpose, offer justifications, and post the standards for public com­ment. 241 Order Directing NERC to Propose Modification of Electric Reliability Organization Rules of Procedure, 130 FERC ¶ 61,203, para. 8 (Mar. 18, 2010), http://www.ferc.gov/
whats-new/comm-meet/2010/031810/E-10.pdf [http://perma.cc/8Y2Y-FE9M].
Thereafter, the standards committee votes whether to authorize a draft standard, and if so, it appoints a committee that drafts the stan­dard, submits it for comment, oversees field testing, and incorporates results into any revisions. 242 N. Am. Elec. Reliability Corp., Standard Processes Manual Version 3, at 14–20 (2013) [hereinafter NERC, Standard Processes], http://www.nerc.com/comm/SC/
Documents/Appendix_3A_StandardsProcessesManual.pdf [http://perma.cc/N2Y5-N6CY].
Next, the “ballot body”—consisting of hundreds of entities including utilities, organized markets, municipal generators, merchant generators, wholesale purchasers, citizen and con­sumer groups, and state utility commissions—votes on the standard. 243 Id. at 21–24. For the list of entities registered to the “ballot body,” see Registered Ballot Body, N. Am. Elec. Reliability Corp., http://standards.nerc.net/rbb.aspx [http://perma.cc/4G2Q-5AYK] (last visited Aug. 5, 2016). Notwithstanding the inclusive composition of the ballot body, the process is stacked heavily in favor of the industry—both because some sectors have fewer members than others and because some votes have more weight than others. For example, there are only a handful of consumer groups that are registered as members of NERC, compared to hundreds of electricity generators. 244 See Current Members, N. Am. Elec. Reliability Corp., http://www.nerc.net/eroregistration/currentmembers.aspx [http://perma.cc/9CXF-UMZK] (last visited Aug. 2, 2016) (listing current members of NERC including self-identified consumer groups). And their votes are weighted at thirty percent, compared to those of generators, which have full weight. 245 See, e.g., Ballot Results, Project 2014-01-DGR-PRC-005-5_Final_Ballot_March_2015, N. Am. Elec. Reliability Corp., (Mar. 2015), http://standards.nerc.net/BallotResults.aspx?BallotGUID=8e218a39-2650-403b-90c2-599cef88de87 [http://perma.cc/NA29-NXNW] (providing an example of balloting results and various voting weights). In other words, the voices of consumers and public interest groups are signifi­cantly diluted.

After the member vote, which must pass by sixty percent, NERC’s Board of Trustees votes on the proposal. 246 The full details are described in 130 FERC ¶ 61,203, paras. 8–11 (Mar. 18, 2010). The ANSI process required a two-thirds vote; FERC expressed concern that this requirement would result in reliability standards that were not sufficiently robust. Rules Concerning Certification of the Electric Reliability Organization, 71 Fed. Reg. 8662, 8691–92 (Feb. 17, 2006). Later, FERC argued that the standard permitted the ballot pool to effectively veto FERC directives. See 130 FERC ¶ 61,203, paras. 12–18 (describing such circumstances for one particular reliability standard). After a series of stand-offs and negotiations, NERC reduced the voting standard to sixty percent and gave its Board more authority to draft and approve standards. See generally Order on Compliance Filing, 134 FERC ¶ 61,216 (Mar. 17, 2011) (codified at 18 C.F.R. pt. 39), http://www.nerc.com/
FilingsOrders/us/FERCOrdersRules/Order_Compliance_filing_RoP_3.17.11.pdf [http://
perma.cc/JAA7-VJXD] (approving proposed changes). For questions still remaining, see Jonathan D. Schneider, NERC on a Wire, Pub. Util. Fortnightly 32, 36 (2013) (raising concerns about repeated FERC remands and timeliness, among others).
The Board may approve or reject a standard but may not modify it. 247 NERC, Standard Processes, supra note 242, at 24; see also John S. Moot, A Modest Proposal for Reforms of the FERC’s Reliability and Enforcement Programs, 33 Energy L.J. 475, 494 (2012) [hereinafter Moot, Modest Proposal] (describing THE procedures). One commentator has opined that this ANSI process is superior to FERC’s rulemaking process because it provides more opportunities for “participation, revision, formal voting, and the like.” 248 Moot, Defer, supra note 7, at 335. This is true as far as it goes, but as shown above, some sectors matter more than others in the number and weight of votes. 249 See supra notes 243–245 and accompanying text (discussing the ballot body). Further, the ANSI process is not suited for creating an administrative record. 250 The outcome of the ANSI process is a standard that is not open to judicial review. This stands in contrast to an administrative-rulemaking outcome, which includes not just a regulation but also its justification. See 5 U.S.C. § 553(c) (2012) (requiring a concise statement of basis and purpose to be published along with the rule); Hammond & Markell, supra note 16, at 322 (“The requirements, however, are far more robust . . . than the text of the APA might suggest.”). And finally, the process takes a significant amount of time, as is described in more detail below. 251 See 71 Fed. Reg. at 8687 (expressing concern “about the time it may take to develop a Reliability Standard under the ANSI-certified process”).

As in the other case studies, NERC must file its proposed rules with FERC. But unlike in those examples, here the statute codifies a substan­tive deference standard. Although FERC must consider whether the standards are “just, reasonable, not unduly discriminatory or preferen­tial, and in the public interest,” it must “give due weight to the technical expertise” of NERC on matters other than the effect of such standards on competition. 252 16 U.S.C. § 824o(d)(2) (2012). This standard has caused significant confusion for FERC, NERC, and the regulated community. 253 See 71 Fed. Reg. at 8690–91 (describing various proposals by commenters regarding how FERC should implement the “just, reasonable, not unduly discriminatory or preferential, and in the public interest” mandate). Of course, it mimics a common standard for judicial review of agency action, 254 See infra Part III (discussing cases in which courts applied the deferential standard of review based, in part, upon recognition of agency expertise). but such a standard does not necessarily translate effectively to a reviewing agency with its own expertise. 255 To the author’s knowledge, there is no legislative history to shed light on this provision.

If FERC disapproves a proposed rule, it must remand the matter to NERC; FERC lacks authority to redraft standards or create new ones. 256 16 U.S.C. § 824o(d)(4). This limitation has resulted in a significant amount of back and forth between FERC and NERC over proposed rules, one example of which is documented in the context of judicial review, discussed below. 257 See infra section III.B (discussing an example of the back and forth between FERC and NERC). Note that the back and forth ultimately resulted in FERC applying deference, but the example reveals that this particular SRO design is inefficient. One of the biggest challenges of this new arrangement seems to be delay. NERC’s standards-development process appears plagued by delays even greater than those associated with federal agencies’ major rulemakings: The development of NERC standards reportedly takes over three years on average and sometimes over five years. 258 Schneider, supra note 246, at 33. Moreover, NERC has revised its standards-development rules several times within the past decade. 259 FERC approved the first rules of procedure in 2006. See Rules Concerning Certification of the Electric Reliability Organization, 71 Fed. Reg. 8662, 8686 (Feb. 17, 2006) (codified at 18 C.F.R. pt. 39); Order Directing NERC to Propose Modification of Electric Reliability Organization Rules and Procedure, 130 FERC ¶ 61,203, para. 26 (Mar. 28, 2010), http://www.ferc.gov/whats-new/comm-meet/2010/031810/E-10.pdf [http://
perma.cc/8Y2Y-FE9M] (directing NERC to develop new procedures for circumstances when the ballot body refused to vote on a standard developed in response to a FERC order); Standards Process Input Grp., N. Am. Elec. Reliability Corp., Recommendations to Improve the NERC Standards Development Process 4, 14 (Apr. 2012), http://www.nerc.
com/pa/Stand/Standards%20Processes%20Manual%20revisions%20SPIG%20Recommen/Standards_Process_Input_Group_04.24.12_ver_8_FINAL.pdf [http://perma.cc/ZZY7-95SZ] (providing stakeholder recommendation for more interaction between FERC and NERC during standards-development process). An increased reliance on technical conferences seems to be emerging as one way to better promote FERC–NERC dialogue. Moot, Defer, supra note 7, at 332–33 (arguing technical conferences will promote communication about competing objectives, which in turn may encourage greater deference during agency review).

NERC has enforcement responsibilities similar to those of the other SROs above. 260 See 16 U.S.C. § 824o(e) (setting forth enforcement responsibilities). It has a detailed hearing-procedure document providing for an initial notice of violation, after which the respondent may seek a hearing before a hearing officer; the hearing is closed to the public. 261 Compliance & Certification Comm., N. Am. Elec. Reliability Corp., Appendix 4E: Compliance and Certification Committee Hearing Procedures, Hearing Procedures for Use in Appeals, and Mediation Procedures 8–12 (Apr. 2009), http://www.nerc.com/FilingsOrders/us/RuleOfProedureDL/Appendix_4E_CCC_Procedures_20131004.pdf [http://perma.cc/FCW8-JW92]. A hearing panel may issue a final order following the hearing officer’s recommendation, which is appealable to NERC. 262 Id. at 32–33. Thereafter, an aggrieved party may seek FERC’s review. 263 16 U.S.C. § 824o(e)(2). The statute does not specify the standard of review, but it permits FERC broader authority than in its review of rules. The agency may modify the penalty, for example, and it also has independent authority to “take such action as is necessary or appropriate” against NERC itself to ensure compliance with the reliabil­ity standards. 264 Id. § 824o(e)(5).

FERC and NERC share enforcement authority, but the statute does not provide guidance on dividing the workload. 265 Schneider, supra note 246, at 36. Early experience shows FERC leaving most enforcement to NERC and its regional relia­bility entities, but the agency has stepped in for high-profile incidents involving power outages. 266 Id.; see also Florida Blackout, 129 FERC ¶ 61,016 (Oct. 8, 2009), http://www.ferc.gov/industries/electric/indus-act/reliability/reliability-orders/IN08-5-000-03-05-2010.pdf [http://perma.cc/E7RZ-AMYC]. And although the two usually work together on enforcement matters, FERC sometimes acts independently. 267 Schneider, supra note 246, at 36 (noting that FERC issued a notice of alleged violation against Entergy without any apparent NERC involvement). Regarding enforcement under this scheme, then, it seems too early to tell how the division of authority and FERC’s oversight role will settle out.

However, commentators see inefficiencies at NERC due to its emphasis on enforcing many small violations rather than larger violations of more important standards. 268 Id. at 37–38. A NERC compliance filing emphasizes that, since 2011, NERC and its regional entities have prioritized enforcement of noncompliance matters that pose the greatest risk to reliability and have streamlined their processes for lower-risk matters. 269 N. Am. Elec. Reliability Corp., Informational Filing of the North American Electric Reliability Corporation in Response to the Federal Energy Regulatory Commission’s November 20, 2014 Order, Docket No. RR14-5-000, at 11 (2015), http://www.nerc.com/FilingsOrders/us/NERC%20Filings%20to%20FERC%20DL/ERO_FYPA_Info_Filing_11202015_RR14-5.pdf [http://perma.cc/UJ9X-XYPA].

Regarding transparency, NERC makes its notices of penalty and other enforcement actions public. 270 Id. at 13. Indeed, in researching this Article, the author found NERC’s website to be the most transparent and user friendly of the case study SROs. These reveal that reliability-standards violations usually settle. 271 Id. at 11–12 (providing data since December 2013). For example, in 2015, FERC and NERC jointly settled a major reliability matter related to an Arizona–Southern California blackout in 2011. 272 Press Release, Fed. Energy Regulatory Comm’n, FERC Approves Final Settlement in 2011 Southwest Blackout Case (May 26, 2015), http://www.ferc.gov/media/news-releases/2015/2015-2/05-26-15.asp#.V6DpdJMrKCQ [http://perma.cc/T2FV-M3V8]. The four different resulting settlements involved civil penalties totaling more than $37 million. 273 Id. FERC’s earlier Reports on Enforcement similarly did not reveal anything other than settlement of NERC reliability violations. Office of Enforcement, Fed. Energy Regulatory Comm’n, 2014 Report on Enforcement 6–12 (Nov. 20, 2014), http://www.
ferc.gov/legal/staff-reports/2014/11-20-14-enforcement.pdf [http://perma.cc/2VAW-TNRZ]; see also Office of Enforcement, Fed. Energy Regulatory Comm’n, 2010 Report on Enforcement 6 (Nov. 18, 2010), http://www.ferc.gov/legal/staff-reports/11-18-10-enforcement.pdf [http://
perma.cc/WF2K-EU9K] (reviewing about 1,300 violations in 190 Notices of Penalties, requesting information on ten, and reviewing one in which litigation was not involved); Office of Enforcement, Fed. Energy Regulatory Comm’n, 2011 Report on Enforcement 7–12 (Nov. 17, 2011), http://www.ferc.gov/legal/staff-reports/11-17-11-enforcement.pdf [http://perma.cc/
DBL2-RKKX] (describing market-manipulation litigation but not reliability litigation); Office of Enforcement, Fed. Energy Regulatory Comm’n, 2012 Report on Enforcement 9–10 (Nov. 15, 2012), http://www.ferc.gov/EventCalendar/Files/20121115105408-2012-Enforcement-Report-11-15-2012.pdf [http://perma.cc/BFH5-33PX] (same); Office of Enforcement, Fed. Energy Regulatory Comm’n, 2013 Report on Enforcement 8–13 (Nov. 21, 2013) [hereinafter FERC, 2013 Report], http://www.ferc.gov/legal/staff-reports/2013/11-21-13-enforcement.pdf [http:/
/perma.cc/4PNM-N3UG] (same).
This is unusual, as it appears that most enforcement actions do not receive FERC attention. A 2013 FERC report states that in that fiscal year, the Commission declined to review any of the notices of proposed violations that NERC filed, amounting to nearly 1,000 violations, among which the largest single potential penalty for which was nearly $1 million. 274 FERC, 2013 Report, supra note 273, at 27. FERC has been focused on market-manipulation issues that are not within the scope of the FERC–NERC relationship. Office of Enforcement, Fed. Energy Regulatory Comm’n, 2015 Report on Enforcement 2 (Nov. 19, 2015), http://www.ferc.gov/legal/staff-reports/2015/11-19-15-enforcement.pdf [http:
//perma.cc/6BLT-W8BN].

The EPAct of 2005 did not require any independent oversight of the FERC–NERC relationship, but FERC itself requires periodic reports from NERC as a condition of its SRO status. 275 Rules Concerning Certification of the Electric Reliability Organization, 71 Fed. Reg. 8662, 8679–80 (Feb. 17, 2006) (codified at 18 C.F.R. pt. 39). Moreover, FERC continues to actively push NERC for greater partnership. For example, FERC in 2015 proposed a rule requiring NERC to give it access to three NERC databases (on a nonpublic basis) so that FERC could assess the need for new or modified standards and “better understand NERC’s periodic reliability and adequacy assessments.” 276 Availability of Certain North American Electric Reliability Corporation Databases to the Commission, 80 Fed. Reg. 58,405, 58,405 (Sept. 29, 2015) (to be codified at 18 C.F.R. pt. 39). NERC’s response was less than enthusiastic: It proposed a different arrangement through which it would share some anonymized data but would thereafter develop NERC-managed ways for the Commission to access anonymized data in the future. 277 N. Am. Elec. Reliability Corp., Comments of the NERC in Response to Notice of Proposed Rulemaking 12–13 (Dec. 15, 2015), http://www.nerc.com/FilingsOrders/us/
NERC%20Filings%20to%20FERC%20DL/NERC_NOPR_Comments_RM15-25_12152015.
pdf [http://perma.cc/U2ZH-HQ9P].

 

***

 

The case studies presented in this section reveal important differences between the traditional story of SROs and what happens in practice. The nitty gritty of rules development is very different from the APA’s notice-and-comment model. 278 See supra sections II.A–.C (discussing procedures related to the APA). Participation is open not to all “interested persons” but rather to a defined set of industry members and, perhaps, to other stakeholders. Those other stakeholders are out­numbered, 279 See supra note 244 and accompanying text (noting only a handful of consumer groups are registered as members of NERC, compared to hundreds of electricity generators). and in some circumstances even underweighted, when it comes to counting votes. 280 See supra note 245 and accompanying text (highlighting that votes from consumers and public interest groups are diluted). The industry truly does develop rules to regulate itself, with little built into the SRO structure to provide a public interest counterbalance. Enforcement follows procedural due process norms, but the oversight agencies and outsiders alike have found reasons to criticize enforcement at their SROs.

As a matter of positive law, moreover, oversight agencies do not necessarily have full authority to approve, disapprove, or modify SROs’ rules. Instead, deference is built into the statutory scheme explicitly, practically, or both. 281 See supra sections II.A–.C (discussing SRO procedures and oversight-agency deference). The oversight agencies’ authorities are frequently more limited, and even those that do have the full scope of authority rarely reject SROs’ proposed rules. 282 See supra sections II.A–.C (providing examples). Similar observations adhere to the enforcement context, in which oversight agencies do not necessarily exercise de novo review and infrequently reject SROs’ enforcement outcomes. 283 See supra notes 225–232, 265–272 and accompanying text (discussing CFTC review of NFA enforcement actions and FERC oversight of NERC decisions). These observations directly contradict the prevailing narrative of SROs, revealing weaknesses in the assumptions underlying the private nondelegation and state action doctrines. Statutory pro­cedural constraints offer a partial, but incomplete, response. Overall, the SRO schemes are structured—whether formally by statute or informally by practice—such that the oversight agencies give deference to their SROs and the many departures from administrative law norms are hidden.

III. Double Deference: Judicial Review of SRO-Initiated Actions

If such deficiencies are obscured, perhaps it is not surprising that the theory and doctrine of administrative law take almost no account of this aspect of governance. Given the judicial role in checking administrative behavior, this relative lack of attention is troubling. It is possible, of course, that the absence of case law could indicate that the judicial forum is unnecessary (perhaps because stakeholders are satisfied) or irrelevant (perhaps because there are other means of oversight). Or the many reviewability doctrines could prevent particular types of petitioners or agency behaviors from entering the judicial forum. 284 Examples include finality, ripeness, exhaustion, standing, and zone of interests. These doctrines largely do not appear to present barriers any different from those in non-SRO contexts, though it is possible that they may have a compounding effect, as demonstrated below. 285 It seems theoretically possible that the stakeholder processes at the SROs—if they prevent public interest or other groups from participating—could create a waiver situation. To avoid incentivizing would-be commenters to bypass the ANSI process and wait for FERC review, FERC requires those raising concerns to explain how they presented their arguments to the SRO in the first instance. Rules Concerning Certification of the Electric Reliability Organization, 71 Fed. Reg. 8662, 8690 (Feb. 17, 2006) (codified at 18 C.F.R. pt. 39). Courts do not necessarily even acknowledge the SROs’ roles in the statutory scheme, though the occasional opinion takes SRO involvement as a given. 286 E.g., Charles Schwab & Co. v. Fin. Indus. Regulatory Auth. Inc., 861 F. Supp. 2d 1063, 1065 (N.D. Cal. 2012) (describing the scope of FINRA’s powers). The few opinions that more directly consider SRO processes are worth careful consideration.

The following sections group the decisions along procedural, substantive, and reviewability lines. First, as a procedural matter, courts have been willing to analogize the procedural constraints governing SROs to those governing agencies under the APA. 287 See infra section III.A (providing examples). Though convenient, the analysis suggests a lack of appreciation for the differences between SROs and agencies. Second, courts reviewing substantive decisions often exhibit two types of behavior. A number of courts simply equate the SRO with its oversight agency, giving no separate consideration to the relationship between the two. 288 See infra section III.B (highlighting the high level of deference given to SROs within the context of the SEC–FINRA and FERC–NERC relationships). And frequently, these courts extend even more deference to actions stemming from SRO–agency relationships than they would in the ordinary administrative law context. 289 See infra section III.B (discussing the court’s reasoning in Charles Schwab). Finally, reviewability barriers stand to compound problematic aspects of the SRO–agency relationship. 290 See infra section III.C (discussing reviewability). What emerges is double deference: judicial deference to agency actions that were deferential to SROs. As this Part concludes, this layered deference can lead to underenforcement of administrative law norms.

A. Procedures and APA Analogies

Although SROs must at least file their proposed rules with their oversight agencies, at least one statute contemplates circumstances that will not require agency review. 291 See, e.g., 15 U.S.C. § 78s(b)(3)(A) (2012) (permitting a FINRA rule to take immediate effect if it, among other things, “constitute[s] a stated policy, practice, or interpretation” or is “concerned solely with the administration of the [SRO]”). These circumstances—covering policies, interpretive statements, and SRO housekeeping—are similar to the exceptions from notice-and-comment rulemaking found in the APA. 292 Compare id. (exempting policy statements, interpretive rules, and SRO housekeeping rules), with 5 U.S.C. § 553(b)(A) (2012) (articulating notice-and-comment exceptions). And in deciding whether the SRO has properly invoked such exemp­tions, courts analogize to the body of law that has developed around the APA’s similar exemptions. Fiero v. Financial Industry Regulatory Authority, Inc., 293 660 F.3d 569 (2d Cir. 2011). for example, both illustrates this methodology and provides more context for how the SRO schemes work. Exercising its enforcement authority, FINRA fined and expelled certain members, who did not thereafter appeal to the SEC. 294 Id. at 572. When the members refused to pay the fine, FINRA brought an action in state court to recover the fine, relying on ordinary principles of contract law. 295 Id. The highest state court held that the state courts lacked subject matter jurisdiction because FINRA’s enforcement activities were exclusively within federal jurisdiction. 296 Id. at 573. Thereafter, the members sought declaratory relief in federal court, arguing FINRA had no authority to collect fines through the courts, and FINRA counterclaimed for the fines. 297 Id.

In the Second Circuit, FINRA argued that its authority stemmed from both the Exchange Act and a rule submitted to—and not disapproved by—the SEC. 298 Id. at 574. The court held that FINRA did not have Exchange Act authority. 299 Id. It reasoned that the Exchange Act’s silence on the matter was dispositive; the Act contains numerous specific provisions regarding actions in the federal courts, including express authority for the SEC—not FINRA—to seek judicial enforcement of penalties. 300 Id. More importantly, the court explained that when FINRA en­forces “its own rules promulgated pursuant to statutory or administrative authority, it is exercising the powers granted to it under the Exchange Act”—powers subject to divestment by the SEC, which did have authority to bring a judicial action. 301 Id. at 576.

Nor did FINRA have authority under its rule. 302 Id. at 578. Although the court could have easily disposed of the case given that the rule was inconsistent with the Exchange Act, it instead grounded its reasoning in procedure. 303 Id. FINRA had filed the rule with the SEC, but it used an ex­ception to the usual notice-and-comment procedures, whereby “[h]ouse-keeping” rules that do not “substantially affect the public interest or the protection of investors” are exempt. 304 Id. (internal quotation marks omitted) (quoting 121 Cong. Rec. 700 (1975) (statement of Sen. Williams)); see also 15 U.S.C. § 78s(b)(3)(A) (2012). Such rules become effective upon filing with the SEC if the SRO designates them as such. 305 15 U.S.C. § 78s(b)(3)(A). The court treated the issue much as one would expect for an issue involving the exceptions to the APA notice-and-comment procedures. 306 See 5 U.S.C. § 553(b)(A) (2012) (exempting, among other things, “rules of agency organization, procedure, or practice” and “general statements of policy”). It labeled rules that go through notice-and-comment “substantive” or “legislative,” in that such rules create new rights or duties, and it cited APA authority for the proposition. 307 Fiero, 660 F.3d at 578 (citing N.Y. State Elec. & Gas Corp. v. Saranac Power Partners, L.P., 267 F.3d 128, 131 (2d Cir. 2001)). Prior to FINRA’s rule, there was no existing authority for FINRA to bring enforcement actions in court; thus, the Second Circuit explained that this was not merely a policy change but a new substantive rule that affected the rights of members—and should have gone through the full rulemaking process. 308 Id. at 579.

The court’s substantive-rule analysis was light on reasoning and perhaps inapt; a mere policy change does not necessarily require notice and comment, as the Supreme Court recently affirmed in Perez v. Mortgage Bankers Ass’n. 309 135 S. Ct. 1199, 1203–04 (2015). The more important point is the court’s analogy—it used standard administrative law principles in a very different context, focusing on FINRA’s procedural choices (rather than those of the SEC). 310 Fiero, 660 F.3d at 571 (discussing FINRA’s role). Other courts have taken similar approaches, mimicking the judicial means of determining whether agencies properly relied on an exception from notice-and-comment rulemaking. 311 See, e.g., Gen. Bond & Share Co. v. SEC, 39 F.3d 1451, 1459–60 (10th Cir. 1994); cf. Am. Mining Cong. v. Mine Safety & Health Admin., 995 F.2d 1106, 1109 (D.C. Cir. 1993) (focusing analysis on whether, in the absence of the rule, the agency would have an adequate basis for enforcement). The analogy is admittedly helpful as an analytic tool. But the courts do not even consider whether it is appropriate. In the ordinary administrative law context, agencies’ increasing reliance on these exemptions has raised concern because exemptions represent a less participatory and deli­berative form of decisionmaking. 312 See, e.g., Nina A. Mendelson, Regulatory Beneficiaries and Informal Agency Policymaking, 92 Cornell L. Rev. 397, 432–33 (2007). This concern could be heightened in the SRO context, in which the procedures are even less in conformity with such norms. To be sure, there are good reasons for such exemptions in both contexts. 313 See id. at 408–10 (detailing the reasons agencies may prefer nonlegislative rules, including efficiency and housekeeping). The point is that closer attention to the differences between agencies and SROs is advisable.

B. Substantive Review, Equating with Agencies, and Deference to Expertise

In reviewing substantive matters, courts follow a similar path, often equating SROs and their oversight agencies. But here they add a new layer of concern, granting even more deference than in ordinary administrative law because of the SROs’ expertise. In Charles Schwab & Co. v. Financial Industry Regulatory Authority Inc., for example, FINRA brought an enforcement action against a member for failing to follow certain SEC-approved FINRA rules that prohibited class action waivers in account agreements. 314 861 F. Supp. 2d 1063, 1067–68 (N.D. Cal. 2012). Prior to the completion of FINRA’s process, the member brought an action in federal district court to invalidate the rule, arguing that it violated the Federal Arbitration Act. 315 Id. at 1067. Emphasizing the importance of FINRA’s expertise on enforcement matters and the detailed statutory appeals scheme, the court held that the member’s failure to exhaust administrative remedies barred his action. 316 Id. at 1069. Rather than just recognizing the SRO’s expertise, however, the court referred to FINRA and the SEC collectively as an agency, relying on case law outside of the SRO context for principles regarding exhaustion. 317 See id. at 1071–72. Specifically, it explained that the SEC–FINRA relationship was “very similar” to the relationship between administrative law judge (ALJ) proceedings and appeals-like review within an agency 318 The court analogized to the Mine Safety and Health Administration. Id. at 1071. —completely overlooking the many distinctions between ALJs 319 ALJs are agency “employees” for Appointments Clause purposes on the somewhat suspect reasoning of Landry v. FDIC, 204 F.3d 1125, 1132–34 (D.C. Cir. 2000). See generally Kent Barnett, Resolving the ALJ Quandary, 66 Vand. L. Rev. 797 (2013) (providing a comprehensive account of ALJs’ difficult constitutional status and developing a proposed remedy). and SROs. 320 Charles Schwab, 861 F. Supp. 2d at 1072. The courts’ insistence on exhaustion of administrative remedies in the SRO–agency context extends to SEC review of exchanges’ enforcement actions. See e.g., PennMont Sec. v. Frucher, 586 F.3d 242, 247 (3d Cir. 2009); see also Fiero v. Fin. Indus. Regulatory Auth. Inc., 660 F.3d 569, 571–72 (2d Cir. 2011) (describing FINRA enforcement procedures, including internal appeals to the National Adjudicatory Council and appeal to SEC).

A set of reliability standards from the FERC–NERC scheme likewise illustrates a high level of deference; it also provides another illustration of the dynamics between an SRO, its agency, and the federal courts. 321 For more on this, as well as another example with similar dynamics, see Moot, Modest Proposal, supra note 247, at 485–87 (discussing FERC’s deference to NERC’s vegetation-management standard). Following the EPAct of 2005, NERC developed its first set of reliability standards as an SRO, a number of which FERC approved. 322 Mandatory Reliability Standards for the Bulk-Power System, 72 Fed. Reg. 16,416, 16,425 (Apr. 4, 2007) (codified at 18 C.F.R. pt. 40). FERC expressed concerns, however, that NERC’s proposed definition of the bulk electric system—which involved deferring to regional councils—would leave gaps in coverage and undermine reliability. 323 Id. at 16,425–26. Following a series of compliance filings, 324 Revision to Electric Reliability Organization Definition of Bulk Electric System, 133 FERC ¶ 61,150, paras. 8–12 (Nov. 18, 2010) (codified at 18 C.F.R. pt. 40). FERC issued a notice of proposed rulemaking requiring NERC to revise the definition. 325 Revision to Electric Reliability Organization Definition of Bulk Electric System, Notice of Proposed Rulemaking, 75 Fed. Reg. 14,097, 14,099 (Mar. 24, 2010) (codified at 18 C.F.R. pt. 40). In its final rule—the remanding action—FERC described what it believed to be the “best way” to address these concerns, including eliminating the regional discretion, setting uniform criteria for determining which facilities were within the definition, and creating an exemption process. 326 Revision to Electric Reliability Organization Definition of Bulk Electric System, 75 Fed. Reg. 72,910, 72,910 (Nov. 26, 2010) (codified at 18 C.F.R. pt. 40).

In NERC’s comments on the proposed rule, it argued that FERC was circumventing the standards-development process by directing specific attributes of a rule that it lacked authority to write in the first place. 327 Id. at 72,913–14, paras. 19–20. FERC responded by emphasizing its supervisory role under the Federal Power Act and explaining that guidance was appropriate so that NERC would be able to adequately respond. 328 Id. at 72,915, para. 35. Of interest, FERC also specified that should NERC decline to adopt the agency’s recommendations, “it must explain in detail, and with a technical record sufficient enough for the Commission to make an informed decision, how its alternative addresses each of the . . . concerns in a manner that is as effective as, or more effective than, the Commission’s identified solution.” 329 Id. at 72,915, para 31.

NERC filed its proposed changes within the requested twelve-month timeframe, and FERC began notice and comment about six months thereafter. 330 Revisions to Electric Reliability Organization Definition of Bulk Electric System and Rules of Procedure, 77 Fed. Reg. 39,858 (proposed July 5, 2012) (codified at 18 C.F.R. pt. 40). FERC ultimately adopted the majority of the standards, 331 Revisions to Electric Reliability Organization Definition of Bulk Electric System and Rules of Procedure, 78 Fed. Reg. 804 (Jan. 4, 2013) (codified at 18 C.F.R. pt. 40). FERC directed modifications of one of the exclusions. Id. at 805, para. 1. which the New York Public Service Commission (New York) challenged in the Second Circuit. 332 New York v. Fed. Energy Regulatory Comm’n, 783 F.3d 946, 953 (2d Cir. 2015). New York sought review on two grounds: First, it argued that the definition exceeded the scope of FERC’s jurisdiction because it included some local electric distribution lines; 333 The Federal Power Act excludes from FERC jurisdiction “facilities used in local distribution.” 16 U.S.C. § 824(b)(1) (2012). and second, it argued the standards were arbitrary and capricious for the same reason. 334 New York, 783 F.3d at 953. The court applied a straightforward Chevron 335 The test announced in Chevron U.S.A. Inc. v. Natural Resources Defense Council, Inc. asks first whether the statute is clear, and if not, whether the agency’s interpretation is permissible. 467 U.S. 837, 842–43 (1984); see infra section IV.A.3 (engaging theory of Chevron). analysis to the first issue, noting first that the statute does not define “facilities used in local distribution” and reasoning second that the interpretation was permissible. 336 New York, 783 F.3d at 954–55. For example, NERC’s findings that the vast majority of jurisdictional facilities operate at higher voltages supported the low-voltage threshold used to identify local lines, and the standard was not determinative but subject to individualized adjustments. 337 Id. at 955. In considering whether the order was arbitrary and capricious, the court’s overview of standards emphasized not only traditional administrative law doctrine under APA section 706(2)(A) but also the specific statutory standard applicable to FERC, which provides that FERC’s fact-finding is conclusive if supported by substantial evidence. 338 Id. at 958; see also 16 U.S.C. § 825l(b).

The court was extraordinarily deferential—it emphasized the “serious consideration FERC and its designated agent, NERC, gave over a period of several years” to the criteria and described the “extensive array of factual material,” “scores of comments,” and FERC’s “reasoned explanations, spanning hundreds of pages.” 339 New York, 783 F.3d at 959. The court repeatedly noted the size of the factual record and “the agency’s industry expertise,” as well as the procedures under which FERC could provide an individual­ized determination. 340 Id. However, it said little about the actual technical issues or the details of the types of facilities that might or might not meet the jurisdictional threshold.

This passage merits several observations. First, the court’s deferential approach was especially evident; it relied more on the numerous iterations of work and the size of the record than on a close look at the agency’s work. In this particular case, the super-deferential approach is not especially troubling from the standpoint of ensuring a reasoned result because the record did indeed reflect extreme substantive care for the issues before the court. Moreover, the procedural history—in which FERC was not initially deferential to NERC and the entities engaged in substantial dialogic exchange—contributed to the robustness of the record. More broadly, however, the court’s lack of close review signals a preoccupation with the size of the record over its substance. This approach is worrisome because, as Professor Wendy Wagner has demon­strated, it incentivizes the “science charade.” 341 See Wendy E. Wagner, The Science Charade in Toxic Risk Regulation, 95 Colum. L. Rev. 1613, 1628 (1995). That is, when agencies know they will be rewarded with deference for large and technically complex records, they will amass such records, obscure the policy decisions underlying their actions and thereby undermine transparency and accountability. 342 See id. at 1674–77. With two levels of process—at the SRO and the agency—the record can become even more impenetrable if these incentives persist. Further, by failing to undertake a full analysis itself, the court missed an important signaling function that would have enabled external monitoring, as described in more detail below. 343 See infra section IV.A.2.

Second, just as in Charles Schwab, the court blurred the lines between FERC and NERC. Describing NERC once as FERC’s agent, 344 New York, 783 F.3d at 959 (“The record amply demonstrates the serious consideration FERC and its designated agent, NERC, gave over a period of several years to the standards and procedures.”). the court bundled the two together in referring to the agency’s “industry expertise” and FERC’s individualized determination process—which actually takes place before NERC and is subject to FERC review. 345 See supra section II.C (describing adjudicatory procedure at NERC). This aspect of the opinion is more difficult to assess normatively. First, the two have different roles and powers; NERC is not really FERC’s agent because FERC cannot rewrite standards itself. Second, the close association with industry obscured the many stakeholders in grid reliability; the New York Public Service Commission, after all, is a state rather than industry stakeholder. Taken further, the court’s logic can be viewed as an affront to the capture criticism that so often plagues SROs. 346 See supra note 4 and accompanying text (describing anticapture critiques). Of course, the court’s equating the two entities was likely meant to emphasize expertise. But this approach can invite the dangers noted above.

C. Reviewability

Reviewability doctrines ought to have the same impact on SRO schemes as on agency-only schemes. 347 In referring to reviewability doctrines, this Article includes but is not limited to the following sources of law: U.S. Const. art. III, § 2 (detailing the case or controversy requirement); 5 U.S.C. § 701(a) (2012) (providing exclusions from review); id. § 702 (describing the agency-action requirement); id. § 704 (setting forth zone-of-interests requirement, among other things). But when statutes are designed to impose reviewability limitations, the impact can magnify the concerns raised above. In NetCoalition v. SEC (NetCoalition II), for example, the D.C. Circuit held that it lacked jurisdiction to review the SEC’s refusal to suspend an SRO rule. 348 715 F.3d 342, 344 (D.C. Cir. 2013). There, several exchanges proposed changes to their rules for setting fees for acquiring market data, and two trade associations—which sought fewer barriers to transparency—requested that the SEC suspend the rules. 349 Id. This was not the first time the issue was before the court. Earlier, the D.C. Circuit held that an SEC order approving market data fee rules was arbitrary and capricious because it lacked sufficient reasoning. 350 NetCoalition v. SEC (NetCoalition I), 615 F.3d 525, 544 (D.C. Cir. 2010). At the time, the relevant statute required the SEC to approve changes to such rules before they became effective. 351 NetCoalition II, 715 F.3d at 344. Dodd-Frank, however, specified that such rules would take effect upon filing—unless the SEC suspended the rule, undertook notice and comment, and concluded that a suspension was necessary to further the purposes of the Exchange Act. 352 Id. (outlining the SEC’s approval process for SROs’ rule changes (citing 15 U.S.C. § 78s(b)(3)(C) (2006 & Supp. IV 2011))). The rule changes at issue in NetCoalition II fell within this category, and the SEC—which had neither conducted an administrative proceeding nor created a record explaining its failure to suspend the rules—argued the court lacked jurisdiction to review the challenge. 353 Id. at 346. The court agreed; it determined that the statute precluded review of a rule at the filing stage because the clear statutory text provided that such actions would not be reviewable. 354 Id. at 353 (citing 15 U.S.C. § 78s(b)(3)(C)) (providing action under this provi­sion is not reviewable).

The outcome in NetCoalition II hinges on statutory text rather than on administrative law doctrine per se. However, it is notable for revealing how statutory design and judicial review can interact in the SRO context. Here, the result of nonreviewability operates as another one-way ratchet favoring industry secrecy. Setting fees for information creates a barrier to transparency that cannot be protected without judicial review. 355 The court did note that the rule would be open to challenge in response to an enforcement action, meaning judicial review of the rule’s substance was not forever foreclosed. Id. at 352. But that would require the SRO or the SEC to initiate an action against a member; the trade associations with an interest in obtaining information would have to intervene if they wished to protect their interests. By crafting a statutory provision that rewards an oversight agency for inaction, Congress failed to further the transparency goals of Dodd-Frank and deepened such concerns regarding the SRO’s behavior generally.

D. A Critique of Double Deference

This study of SROs reveals that the ordinary administrative law paradigm is inapt for the SRO context. Although the constitutional legitimacy of SROs rests on the presumed authority of their oversight agencies, this presumption falls apart on a close look at the agencies’ actual powers. As is the case of FERC, the agency may lack authority to write rules itself or to modify the rules of the SRO. As happens at the CFTC and even the SEC, rules can become effective without any review by the agency at all. As is the practice at the CFTC and the SEC, any review is deferential, and a deferential stance is statutorily required at FERC. Enforcement fares little better, with deference either expected or required at the reviewing agencies.

In the ordinary administrative law context, deference by courts to agencies is often justified on the basis of the agency’s comparative expertise or political accountability. 356 See Chevron, U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 865 (1984) (recognizing the deference that courts give to administrative agencies because “[j]udges are not experts in the field”); see also Hammond, Deference Dilemma, supra note 27, at 1770–76 (developing accounts of expertise and political control); Elena Kagan, Presidential Administration, 114 Harv. L. Rev. 2245, 2269 (2001) (noting courts “shy away from such substantive review of agency outcomes, perhaps in recognition of their own inability to claim either a democratic pedigree or expert knowledge”). With respect to SROs, it is only fair to acknowledge the expertise of their members, which have day-to-day familiarity with the many issues falling within the SROs’ jurisdiction. But the political accountability is harder to identify. 357 Cf. Amtrak II, 821 F.3d 19, 30–31 (D.C. Cir. 2016) (rejecting political account­ability of Amtrak as means of avoiding due process violation). First, note that the oversight agencies in the case studies are independent agencies, which are at least somewhat insulated from presidential control by removal restrictions, their multimember structures, diverse political require­ments, and exemption from direct means of oversight like Executive Order 12,866 and its successors. 358 For a full description of these insulating mechanisms, see Hammond, Deference Dilemma, supra note 27, at 1777–78. Second, as already emphasized, the oversight agencies have less power to direct their SROs than is typically presumed.

Any accountability, therefore, must come from within the SROs’ processes themselves. But as previously shown, those procedures are not always models of democratic decisionmaking or transparency. The regulated industries have monopolies on their SROs’ procedures, whether by membership restrictions, simple numbers, or greater voting power. The procedures themselves are sometimes laudably transparent as in the NERC example—but other times hopelessly opaque, as in the NFA example.

Moreover, administrative law’s deference model takes place against an additional set of background norms: the expectation that the agency’s procedures are participatory, deliberative, and transparent. In the SRO context, blindly relying on this same set of norms to complement a deferential relationship between the oversight agency and the SRO is not appropriate. A fortiori, judicial review that layers on more deference completely obscures the import of procedural checks that maintain the legitimacy of the fourth branch.

IV. Toward a More Robust Model

Having identified the problematic aspects of double deference in the preceding Parts, this Part turns to potential ways of attacking the problem. The first section suggests how administrative law doctrine might better account for the role of SROs in the administrative scheme. The second section suggests institutional-design considerations for build­ing better SROs.

A. Situating SROs in Traditional Administrative Law

Before suggesting ways to account for SROs in administrative law, a possible criticism of this approach should be acknowledged. Any treatment of the judicial stance regarding administrative procedure must be mindful of the limitations courts face in policing that procedure. In particular, the landmark decision in Vermont Yankee Nuclear Power Corp. v. Natural Resources Defense Council put to rest the notion that courts may impose on agencies specific procedural requirements not found in the APA. 359 435 U.S. 519, 548 (1978). A number of scholars, however, have argued that the evolution of administrative law represents a continuing violation of Vermont Yankee. 360 The literature is vast. See, e.g., Jack M. Beermann & Gary Lawson, Reprocessing Vermont Yankee, 75 Geo. Wash. L. Rev. 856, 882–90 (2007) (arguing that a variety of doctrines ought to be reconsidered in light of Vermont Yankee); Richard J. Pierce, Jr., Waiting for Vermont Yankee II, 57 Admin. L. Rev. 669, 673–82 (2005) (criticizing presumptions in favor of formal adjudicatory procedures); Paul R. Verkuil, Judicial Review of Informal Rulemaking: Waiting for Vermont Yankee II, 55 Tul. L. Rev. 418, 419–21 (1981) [hereinafter Verkuil, Vermont Yankee II] (criticizing hard-look review in light of the Vermont Yankee decision). By way of brief example, the “concise” statement of “basis and purpose” that must accompany rules is anything but, 361 5 U.S.C. § 553(c) (2012); see also Gary Lawson, Federal Administrative Law 404 (7th ed. 2015) (calling such statements “monstrously long and complex”). and hard-look review has often been criticized for its tendency (perhaps theoretical) to promote ossification of agency rulemaking 362 Whether judicial review actually causes ossification is a matter of debate. Compare Richard J. Pierce, Jr., Seven Ways to Deossify Agency Rulemaking, 47 Admin. L. Rev. 59, 65 (1995) (“With the exception of a few agencies, the judicial branch is responsible for most of the ossification of the rulemaking process.”), with Mark Seidenfeld, Why Agencies Act: A Reassessment of the Ossification Critique of Judicial Review, 70 Ohio St. L.J. 251, 252 (2009) (providing a contrary view). For further discussion and sources, see also Hammond, Super Deference, supra note 28, at 750–51. as well as its arguable lack of statu­tory basis. 363 E.g., Beermann & Lawson, supra note 360, at 882 (“Hard-look review may or may not be a correct or even plausible interpretation of § 706(2)(A) . . . .”).

Acknowledging these important concerns, this section considers how existing judicial doctrine might better account for the SRO model’s normative shortcomings—distinct from imposing new procedures on agencies or SROs. 364 Although an analysis is beyond the scope of this Article, it seems likely that at least some oversight agencies would have sufficient plenary power to impose accountability-promoting procedures on SROs, avoiding the Vermont Yankee problem altogether. This approach is consistent with a variety of theories suggesting that judicial review promotes the legitimacy of administrative agencies. 365 See Hammond & Markell, supra note 16, at 326–27 (describing how judicial review promotes the legitimacy of administrative agencies). Given the lack of any meaningful standards promoting the constitutional legitimacy of SROs, the many ways they diverge from the expectations of procedural legitimacy, and the courts’ compounding of these concerns through the deference doctrines, an examination of judicial doctrine is at the very least a worthwhile enterprise.

1. The Chenery Cases. — The Chenery cases represent two bedrock principles of administrative law. First, courts review an agency’s action only on the basis of the agency’s rationale at the time of its decision. 366 See, e.g., SEC v. Chenery Corp. (Chenery I), 318 U.S. 80, 89–90 (1943) (“Since the Commission professed to decide the case before it according to settled judicial doctrines, its actions must be judged by the standards which the Commission itself invoked.”); see also Bowman Transp., Inc. v. Ark.-Best Freight Sys., Inc., 419 U.S. 281, 285–86 (1974) (“[W]e may not supply a reasoned basis for the agency’s action that the agency itself has not given.”). This distinguishes judicial review of agencies from legislative enactments because the latter can be upheld on any rational basis, provided there are no countervailing constitutional issues. 367 See Motor Vehicle Mfrs. Ass’n of the U.S. v. State Farm Mut. Auto. Ins., 463 U.S. 29, 43 n.9 (1983) (rejecting the proposition that review of agencies is akin to minimum rationality review for statutes under substantive due process). Indeed, courts’ rejection of the invitation to review agencies as they would legislatures provides a “structural check” on agencies and promotes their constitutional legiti­macy. 368 Kevin M. Stack, The Constitutional Foundations of Chenery, 116 Yale L.J. 952, 959 (2007) (“Chenery provides a structural check for the very presumptions of agency accountability, rationality, and expertise upon which Chevron deference is based.”); see also Emily Hammond Meazell, Deference and Dialogue in Administrative Law, 111 Colum. L. Rev. 1722, 1736 (2011) [hereinafter Hammond, Deference and Dialogue] (developing the history and constitutional import of the Chenery principle). As Professor Kevin Stack has argued, Chenery articulates a trade-off: Broad delegations of authority are constitutionally permissible, but in exchange, agencies must articulate a reasoned basis for their decisions for the courts’ review. 369 Stack, supra note 368, at 1000. Furthermore, the record requirement—famously articulated in Citizens to Preserve Overton Park, Inc. v. Volpe—facilitates the judicial review that Chenery I contemplates. 370 Citizens to Pres. Overton Park, Inc. v. Volpe, 401 U.S. 402, 417–21 (1971); see also Hammond, Super Deference, supra note 28, at 740, 761–63.

Similarly, in the SRO context the private nondelegation doctrine does not operate as a meaningful constraint. 371 See supra section I.C.1 (discussing the private nondelegation doctrine). Courts are confronted with the broad delegation of authority to oversight agencies compound­ed by further delegation to SROs. Part of the solution is for lawmakers to ensure that delegations to SROs have clear boundaries; the schemes presented in this Article are at least arguably more constrained regarding the scope of the SROs’ power.

But in reviewing agency actions originating at the SRO level, courts should be mindful that the entire record reaches back to the SRO itself. By emphasizing that the applicable standard of review includes the SRO record, the courts can accomplish several things. First, courts can enable their own review for reasonableness, which is considered in more detail in the next section.

Second, they promote ex ante legitimizing behavior at both the SRO and the oversight agency. The SRO should prepare a record in anticipation of the potential for judicial review—not just review by the oversight agency. This quasi-procedural act promotes transparency and guards against arbitrariness by necessitating a documented and reasoned decision. Furthermore, the SRO ought to document the participatory steps taken in developing the record. Observers of the administrative process are interested, for example, in whether the process was open to all interested parties and which parties participated. 372 This Article does not argue that courts ought to adjust their level of review in response to whether a particular party did or did not participate. Such matters might be relevant, for example, to determine whether a party has waived the right to participate in review. Instead, the author’s chief concern here is for accountability to others who provide checks on the administrative process, including Congress, the executive, journalists, and scholars. The agency should carefully consider any significant comments or concerns raised before both the SRO and the agency, particularly when there has been disagree­ment between members, the SRO, and/or the agency. By articulating the resulting action as its own decision based on the entirety of the record, the agency similarly engages in legitimizing behavior.

Third, when courts take care to examine the totality of the agency’s decision—here on the basis of the agency’s rationale that incorporates the SRO’s record—the courts themselves promote transparency and accountability. Other observers, such as Congress and the executive, journalists, and scholars, will be better able to assess for themselves the performance of the SRO model.

The second Chenery principle provides that an agency’s choice of procedures be within its “informed discretion.” 373 See SEC v. Chenery Corp. (Chenery II), 332 U.S. 194, 203 (1947). An important practical impact is that agencies are free to set policy through the adjudicatory process, notwithstanding that rulemaking has the benefits of notice, gen­eral applicability, and participatory decisionmaking. 374 Cf. id. at 202 (upholding the agency’s choice of adjudication, notwithstanding that “[t]he function of filling in the interstices of the Act should be performed, as much as possible, through this quasi-legislative promulgation of rules to be applied in the future”). In other words, whereas Chenery I speaks to judicial checks on agencies, Chenery II speaks to agency power. Even so, Chenery II informs the pragmatic debate about the benefits and drawbacks of rulemaking and adjudication. 375 Chenery II also offers a check on retroactivity, albeit a limited one. See id. at 203 (setting forth a test balancing the harm of retroactivity against the benefits of furthering the statutory scheme). How does this translate to the SRO context? One might argue that some of the statutory schemes limit SROs’ choice of procedures because the statutes themselves expressly detail the rulemaking and enforcement powers of the SROs. But that argument presumes a neat divide between rulemaking and enforcement that likely is not realistic. After all, an SRO may need to develop policy through the enforcement process when it encounters new facts in an adjudication over which it presides. 376 See id. at 201 (emphasizing the agency’s task was to resolve issues in the adjudication before it).

Consider again, however, the New York v. Federal Energy Regulatory Commission decision above, which involved delicate jurisdictional determinations for assessing the applicability of reliability standards. 377 783 F.3d 946, 950–53 (2d Cir. 2015). FERC and NERC engaged in considerable dialogue before NERC fashioned a rule that FERC ultimately approved. 378 Id. at 950–51. That rule provided a set of parameters by which an entity might presumptively be within the SRO–agency jurisdiction but which could be overcome by a series of specific showings. 379 Id. at 951–53. To be sure, the SRO made the initial jurisdictional determination. But the ability of entities to request individualized adjudications was an important feature of the scheme, alleviating jurisdictional concerns that might not have otherwise survived judicial review. In other words, Chenery II offers the insight that SRO rules may be fashioned to guard against arbitrariness, as well as avoid vulnerability on review, by preserving the flexibility for individualized decisionmaking when needed.

2. State Farm and Hard-Look Review. — When courts review agencies’ substantive decisions—informed by facts, policy, and experi­ence—they search for reasoned decisionmaking. This standard is tied to both the arbitrary and capricious and substantial evidence provisions of the APA, 380 See 5 U.S.C. § 706(2)(A) (2012) (articulating the arbitrary and capricious provision); id. § 706(2)(E) (articulating the substantial evidence standard); see also Ass’n of Data Processing Serv. Orgs. v. Bd. of Governors, 745 F.2d 677, 683 (D.C. Cir. 1984) (Scalia, J.) (stating that the arbitrary and capricious and substantial evidence standards are essentially the same). and it is notoriously variable in the degree of deference it affords. 381 See Sidney A. Shapiro & Richard E. Levy, Judicial Incentives and Indeterminacy in Substantive Review of Administrative Decisions, 44 Duke L.J. 1051, 1064–66 (1995) (describing the “proliferation of manipulable categories to which different degrees of deference apply”). For statutory schemes involving technical complexity and scientific uncertainty, courts often reference a “super deferential” stan­dard, which counsels that courts should be at their “most deferential” in such circumstances. 382 Balt. Gas & Elec. Co. v. Nat. Res. Def. Council, Inc., 462 U.S. 87, 103 (1983). See generally Hammond, Super Deference, supra note 28 (providing a comprehensive account and critique of super deference). This approach, which is often marked by judicial opinions notable for their unhelpful brevity, 383 See Hammond, Super Deference, supra note 28, at 766–69. stands in contrast to the “hard-look” varietal, most notably articulated in Motor Vehicle Manufacturers Ass’n. v. State Farm Mutual Automobile Insurance Co.:

Normally, an agency rule would be arbitrary and capricious if the agency has relied on factors which Congress has not intended it to consider, entirely failed to consider an important aspect of the problem, offered an explanation for its decision that runs counter to the evidence before the agency, or is so implausible that it could not be ascribed to a difference in view or the product of agency expertise. 384 463 U.S. 29, 43 (1983).

Hard-look review at its best can both promote accountability and guard against arbitrariness. It does this by signaling to agencies that their provided rationale will be deeply considered, which promotes ex ante legitimizing behaviors and provides a translation of specialized informa­tion for generalist observers of administrative law. 385 Hammond, Super Deference, supra note 28, at 778.

Given these attributes of hard-look review, courts should be especially diligent in applying it to the SRO context. When reviewing rules, this means considering and describing the record that Chenery I and Overton Park promote—including the procedures and analysis at the SRO. 386 An excellent example is the Court’s analysis of whether FERC’s actions were arbitrary and capricious in Federal Energy Regulatory Commission v. Electric Power Supply Association, 136 S. Ct. 760, 782–84 (2016). Evidence of dialogic activity between the oversight agency and SRO might be rewarded in this context because it helps demonstrate reasoned, careful analysis. 387 Cf. Hammond, Deference and Dialogue, supra note 368, at 1779–80 (suggesting when agencies are responsive on remand to courts’ dialogic overtures, they are more likely to withstand subsequent judicial review). By contrast, SRO rules that become effective without any agency involvement must stand without the benefit of that check, and SROs should thus develop fully reasoned analyses prior to submitting rules to their oversight agencies.

The State Farm hard-look lens may also offer insights for agencies that review SRO rules and adjudications, particularly for the statutory schemes that provide for some sort of agency deference. Indeed, FERC seems to have internalized this approach with respect to NERC rules; it has repeatedly emphasized that although NERC has authority to develop rules in the first instance, it should be prepared to explain why it has or has not followed FERC’s directions when rules are remanded. 388 E.g., Revision to Electric Reliability Organization Definition of Bulk Electric System, 75 Fed. Reg. 72,910, 72,913–14 (Nov. 26, 2010) (codified at 18 C.F.R. pt. 40) (recommending particular changes to definition and stating if SRO takes different approach, “it must explain in detail, and with a technical record sufficient enough” for review, how its alternative addresses the agency’s concerns); see also Moot, Defer, supra note 7, at 333 (advocating this approach). For adjudications, oversight agencies might also borrow from administrative law and be especially careful with SRO orders that depart from those proposed by the initial hearing officer, particularly if there are factual findings at issue. 389 See Universal Camera Corp. v. NLRB, 340 U.S. 474, 496 (1951) (recognizing evidence may be “less substantial when an impartial, experienced examiner who has observed the witnesses and lived with the case has drawn conclusions different from the [agency’s]”).

In advocating the hard-look standard, this author anticipates the concern that this approach already burdens agencies too much, leading to ossification as noted above. 390 See supra note 362 (comparing sources discussing the ossification question). Adherents to the ossification critique argue that it grinds agency rulemaking to a near-halt and incentivizes agencies to make rules through less participatory vehicles such as guidance documents. 391 See, e.g., Thomas O. McGarity, Some Thoughts on “Deossifying” the Rulemaking Process, 41 Duke L.J. 1385, 1386 (1992). Various studies have challenged the ossification critique, suggesting that it is not as problematic as some scholars theorize. 392 See Cary Coglianese, Empirical Analysis and Administrative Law, 2002 U. Ill. L. Rev. 1111, 1127–31 (“[A] retreat from rulemaking in the face of stringent judicial review is not nearly as clear as has been generally supposed.”); see also Anne Joseph O’Connell, Political Cycles of Rulemaking: An Empirical Portrait of the Modern Administrative State, 94 Va. L. Rev. 889, 923, 963–64 (2008) (discussing empirical findings suggesting the traditional regulatory process is not greatly ossified). Still, one takes the point and acknowledges that there may be heightened reasons for concern in the SRO context, in which a slower process undermines the efficiency, responsiveness, and nimbleness rationales that motivate SRO schemes. Ultimately, this Article takes the view that the benefits of hard-look review are worth the theoretical costs. Even more than for ordinary regulatory regimes, here judicial review ameliorates the deficiencies of the private nondelegation doctrine and lack of other checks on SROs. As already explained, courts fulfill important roles that promote the legitimacy of SROs, and to retreat from that role would be an abdication of Article III responsibility.

3. Chevron and Mead. — Courts reviewing agencies’ interpretations of their statutory mandates frequently make use of the two-step Chevron framework, 393 But this is not always true. See Fed. Energy Regulatory Comm’n v. Elec. Power Supply Ass’n, 136 S. Ct. 760, 773 n.5 (2016) (declining to address the agency’s alternative Chevron argument because the agency’s authority was clear); King v. Burwell, 135 S. Ct. 2480, 2488–89 (2015) (declining to afford Chevron deference to IRS interpretation of Affordable Care Act, reasoning Congress did not intend the agency to have interpretive authority and the matter was too important to find implicit delegation); 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, 1090 (2008) (discussing the results of empirical study showing the Court usually does not apply Chevron to Chevron-eligible cases); see also Emily Hammond & Richard J. Pierce, Jr., The Clean Power Plan: Testing the Limits of Administrative Law and the Electric Grid, 7 Geo. Wash. J. Energy & Envtl. L. 1, 8 (2016) (describing variant of Chevron as “Brand X avoidance” (internal quotation marks omitted)). which asks (a) whether Congress has clearly spoken and if not (b) whether the agency’s interpretation is permissible. 394 Chevron, U.S.A., Inc. v. Nat. Res. Def. Council, Inc., 467 U.S. 837, 842–43 (1984). Courts, scholars, and policymakers have been captivated by Chevron since its birth, and there is a rich literature debating both its positive application and its normative basis. 395 See Emily Hammond et al., Judicial Review of Statutory Issues Under the Chevron Doctrine, in A Guide to Judicial and Political Review of Federal Agencies 69–70 (Michael E. Herz et al. eds., 2d ed. 2015) (noting that “[t]here is no single widely accepted rationale for Chevron”). As an influential doctrine of administrative law, Chevron bears consideration in the SRO context. In addition, one of Chevron’s progeny, United States v. Mead Corp., 396 533 U.S. 218 (2001). offers insights into the relationship between agency procedures and the level of deference that are particularly helpful in the SRO context.

First, the Court in Mead rooted Chevron not only in agencies’ superior political accountability and expertise but also in an implied congressional intent to delegate interpretive authority to the agency tasked with administering the statute. 397 Id. at 229; see Barnett, Codifying Chevmore, supra note 49, at 15–16 (discussing literature addressing whether Congress has any intent as to interpretive primacy). But see Mark Seidenfeld, Chevron’s Foundation, 86 Notre Dame L. Rev. 273 (2011) (rejecting implied congressional intent rationale and detailing Article III rationale). Though at least two Justices have acknowledged that implied intent is a fiction, 398 See Stephen Breyer, Judicial Review of Questions of Law and Policy, 38 Admin. L. Rev. 363, 370 (1986); Antonin Scalia, Judicial Deference to Administrative Interpretations of Law, 1989 Duke L.J. 511, 517. that admission is not particularly problematic in the ordinary administrative law case. Indeed, the fiction of implied delegation serves as a signal to the legislature to state otherwise if it would prefer an agency not have interpretive author­ity. 399 See Scalia, supra note 398, at 517. Congress seems to have acquiesced to Chevron: 400 But see Lisa Schutz Bressman & Abbe R. Gluck, Statutory Interpretations from the Inside—An Empirical Study of Congressional Drafting, Delegation, and the Canons: Part II, 66 Stan. L. Rev. 725, 771–73 (2014) (providing a nuanced assessment of legislative drafters’ views of institutional interpretive roles). Recently, both houses of the legislature have considered bills purporting to abrogate Chevron. See Emily Hammond, Four Flaws with the Separation of Powers Restoration Act, 41 Admin. & Reg. L. News 4, 9–10 (2016) (describing and critiquing the bills). Only once has it specified a different deference regime. 401 See generally Barnett, Codifying Chevmore, supra note 49 (exploring Dodd-Frank provisions that clearly specify a standard of review other than Chevron).

But the fiction seems more troubling in the SRO context, in which Congress’s silence on interpretive authority is deafening and in which regulatory authority is shared. 402 This fact alone could undermine the case for Chevron deference because agencies typically must have sole authority to benefit from Chevron. See Rapaport v. U.S. Dep’t of the Treasury, Office of Thrift Supervision, 59 F.3d 212, 216 (D.C. Cir. 1995) (“We have already held in Wachtel that we owe no such deference to the OTS’s interpretation of § 1818 because that agency shares responsibility for the administration of the statute with at least three other agencies.”). One answer is that interpretive authority should rest solely with the oversight agency; if an SRO encounters any question regarding the meaning of a statutory provision, the SRO should seek the agency’s guidance rather than try to resolve the meaning in the first instance. 403 A similar issue may arise in the context of interpreting the agency’s interpreta­tions of its own regulations. See Auer v. Robbins, 519 U.S. 452, 461 (1997) (“Because the salary-basis test is a creation of the Secretary’s own regulations, his interpretation of it is, under our jurisprudence, controlling . . . .”). This approach may not be realistic in adjudications, which require the SRO to resolve the issues before it. But it may provide a useful reminder that the SRO and agency should coordinate during the rulemaking process so that the oversight agency can direct the SRO accordingly. In either type of action, maintaining the agency’s inter­pretive primacy requires meaningful oversight of proposed rules and enforcement orders.

This observation leads to United States v. Mead Corp., in which the Court signaled that the procedures an agency applies in developing an interpretation have a bearing on whether the interpretation is Chevron-eligible. 404 533 U.S. 218, 229–30 (2001). Relatively formal procedures that “foster . . . fairness and deliberation” are more likely to qualify for deference; this explains why rules developed through notice and comment or interpretations arising from formal adjudications enjoy such deference. 405 Id. at 230.

As Professor Lisa Bressman has explained, moreover, Mead offers a way for courts to facilitate oversight. 406 Lisa Schultz Bressman, Procedures as Politics in Administrative Law, 107 Colum. L. Rev. 1749, 1791 (2007). That is, procedures that are ad hoc, difficult to access, or otherwise lacking in participatory, deliberative, and transparency measures do not qualify for Chevron deference because they offer a reduced opportunity for external oversight. 407 Id. at 1793. These insights are particularly useful in the SRO context because—if the oversight agencies and SROs have facilitated a complete record as recommended above—courts have not one but two layers of procedure for consideration.

When an opaque or noninclusive SRO-rulemaking process con­cludes with a rule that becomes effective without any independent-agen­cy examination, the case is strongest for a lower level of deference. 408 As Mead made clear, Skidmore deference would apply. See Mead, 533 U.S. at 234–35; see also Skidmore v. Swift & Co., 323 U.S. 134, 139–40 (1944) (describing factors for lower courts’ consideration). On the other hand, a transparent, participatory process that reveals multiple opportunities for public engagement and full review by an oversight agency ought to qualify for Chevron deference. 409 For an example of this kind of analysis, see Kruse v. Wells Fargo Home Mortgage, Inc., 383 F.3d 49, 58–61 (2d Cir. 2004). That conclusion is even stronger when the record demonstrates open and dialogic behavior between the oversight agency and the SRO. Indeed, in such circum­stances, the private nondelegation doctrine’s presumption is correct: There is true agency oversight. Thus, deference in this context rewards legitimizing behavior that serves both constitutional and administrative law values.

B. Building Better SROs

There is significant opportunity for courts to deepen their analyses of SRO–agency actions within existing administrative law doctrine. But this Article has also demonstrated that there is great variety in the design choices underlying SRO schemes. Thus, this section briefly examines design considerations with an eye toward further promoting account­ability and guarding against arbitrariness. These options maintain the focus on SROs as a generalized feature of the administrative state; with some reluctance, this Article does not consider industry-specific propos­als that might improve how particular regulatory programs function. 410 Many such proposals are collected in the industry-specific works cited throughout this Article. For a thoughtful example, see Omarova, Wall Street, supra note 7, at 474 (arguing for redrawing financial regulatory boundaries between institutions dealing in OTC derivatives and complex financial instruments and those dealing in financial intermediation services). Section IV.B.1 engages large structural considerations inher­ent in the SRO scheme, including the structure of the oversight agency, the pres­ence of a nuclear option, and the procedures governing the SRO–agency relationship. Section IV.B.2 considers ways to facilitate external oversight, such as extending the reach of open-government laws and expanding the use of independent reviews. The final section mentions best practices for SROs and oversight agencies.

1. Structure of the Oversight Agency. — The three case studies pre­sented above involve independent oversight agencies, as distinguished from executive agencies. Although this is not a necessary feature of SRO models, it is a common one, so it bears mention in the context of institutional design. The theory of independent agencies holds that such agencies offer heightened benefits in the form of particularized expertise and insulation from presidential control. 411 See Hammond, Deference Dilemma, supra note 27, at 1777–79 (describing expertise and other rationales for creating independent agencies). As described above, 412 See supra text accompanying note 358 (describing this principle). the features of these agencies may help keep them on task and less susceptible to the shifting winds of politics. Their multimember construc­tion, moreover, enables the agency heads to benefit from deliberative decisionmaking and helps dampen any extreme views. 413 See Hammond, Deference Dilemma, supra note 27, at 1778–79. Further, inde­pendent agencies are at least theoretically less susceptible to capture than executive agencies are. 414 Rachel E. Barkow, Insulating Agencies: Avoiding Capture Through Institutional Design, 89 Tex. L. Rev. 15, 17 (2010). In these ways, an independent oversight agency can ameliorate some of the process-oriented flaws that may persist in SROs—as well as the concern that SROs are captured by definition.

Admittedly, the actual independence of these agencies is open to question. 415 See generally Neal Devins & David E. Lewis, Not-So Independent Agencies: Party Polarization and the Limits of Institutional Design, 88 B.U. L. Rev. 459 (2008) (describing how independent agencies have become politicized). But consider the alternative. Housing an SRO within an executive agency would most certainly open a Pandora’s box of political influence through direct presidential intervention, Office of Information and Regulatory Affairs oversight, and a variety of other mechanisms that would likely bog down SRO processes and could cause significant regula­tory disruption. 416 See generally Farber & O’Connell, supra note 37, at 1154–67 (noting that exec­utive control “strains the idea of policy delegation to a unique agency decision maker”). To the extent that an agency’s inde­pendence remains meaningful, housing SROs within such agencies likely offers the better choice, if only at the margins.

2. The Nuclear Option. — In most SRO regimes, SROs must qualify for their status and the oversight agencies retain authority to revoke that status. The Exchange Act, for example, provides that the SEC “may relieve any self-regulatory organization of any responsibility” under the statute or suspend or revoke its registration. 417 15 U.S.C. §§ 78s(g)(2)–(h) (2012). This type of threat—which Professor Brigham Daniels terms the “nuclear option”—can have persuasive force even if the oversight agency never uses it. 418 See Brigham Daniels, When Agencies Go Nuclear: A Game Theoretic Approach to the Biggest Sticks in an Agency’s Arsenal, 80 Geo. Wash. L. Rev. 442, 486 (2012). The threat, however, must have at least some credibility. This can be achieved in a variety of ways. For example, a legislative scheme or oversight agency itself could provide for incremental steps that stop short of the nuclear option but promote the oversight agency’s leverage over the SRO. 419 See id. at 485–86 (explaining that agency flexibility to take intermediate steps is important for reaching beneficial outcomes). These could include specific matters that may be withdrawn from SRO authority upon meeting certain conditions. 420 See, e.g., Omarova, Wall Street, supra note 7, at 475 (suggesting the credible threat of “targeted government intervention, such as a direct ban on complex financial products” as a potential regulatory incentive structure for improving SRO governance). In addition, it can be helpful for outside actors to have access to the nuclear option so that it does not reside only with the agency. 421 Daniels, supra note 418, at 482–83. An agency in all likelihood is unable as a matter of funding and staffing to take on an SRO’s entire scope of duties. Thus, structuring an SRO regime to enable members to file complaints or petitions directly to the agency would provide a trigger for the agency to bring its oversight function to bear more specifically on a given problem. A statute could also extend an oversight function to the courts, which could temporarily enjoin SROs from exercising their functions upon a certain (presumably very high) showing. These approaches have the benefit of providing a backstop against SRO arbitrariness in extreme circumstances without interfering with the day-to-day operations of SROs in the ordinary course. This maximizes the benefits of SROs while preserving an ultimate check should the need arise.

3. Procedures Governing the SRO–Agency Relationship. — As the case studies above reveal, variety is the hallmark of procedures governing the SRO–agency relationship. 422 See supra Part II (providing case studies). The benefits and costs of such procedures are explored above, but it is worth pausing here to offer a few additional thoughts. First, it should go without saying that legislators should carefully assess their options in choosing procedures because those choices have constitutional and administrative law implications. In particular, when SROs’ rules can become effective in the absence of agency review, it undermines a fundamental expectation of constitutional legitimacy. 423 See supra section I.C (examining the limited constraint that private nondelega­tion doctrine imposes on SROs). This is true to a lesser extent when an agency must be deferential to an SRO’s expertise; it sets up a scheme of reduced oversight and a decreased opportunity for the agency itself to police administrative law norms. Second, procedures can overwhelm both the SRO’s and the agency’s ability to take advantage of the efficiency benefits that motivate SRO schemes. Consider, for example, that the FERC–NERC scheme’s requirement of deference has been so challenging as to create a mini dialogue, resulting in both delays and regulatory uncertainty in the meantime. 424 See supra section II.C (explaining that uncertainty and delays are two of the biggest challenges for the FERC–NERC arrangement). The lesson is simple: The stakes are high, and proce­dures should be crafted with intention and care.

4. Facilitating External Oversight. — The combination of opaque pro­cedures, technical subject matter, and baked-in deference often found in SRO schemes makes external oversight extremely challenging. One response is to require external audits and reports, as Dodd-Frank did for the SEC. 425 See supra section II.A (noting that Dodd-Frank provides for mandatory review of the SEC–FINRA relationship). Alternatively, the agency itself may require periodic reports, as FERC asks of NERC. 426 See supra section II.C (discussing FERC requiring periodic reports from NERC as a condition of its SRO status).

But there are other more generalized methods of promoting oversight that are worth consideration. Some scholars, for example, have argued that expanded open-government laws would at least facilitate transparency within SROs. 427 See, e.g., Craig D. Feiser, Privatization and the Freedom of Information Act: An Analysis of Public Access to Private Entities Under Federal Law, 52 Fed. Comm. L.J. 21, 55–62 (1999) (suggesting an expanded FOIA definition of “agency” and “agency records”). Congress or the oversight agency could also create an independent ombudsman or advisory committee to assist in monitoring the SRO. 428 Omarova, Wall Street, supra note 7, at 488–89. Professor Omarova suggests independent councils comprised of academics, consumer advocates, and other “outsiders” but cautions that it would be difficult to navigate nondisclosure requirements and otherwise protect intellectual property. Id. at 489. Careful vetting and contracting, as well as the implementation of regulations, ought to be able to eliminate such concerns.

Either of these options—open government and independent over­sight—would involve difficult line-drawing and eligibility questions. 429 See id. at 489 (noting the “balancing of various policy interests” required in inde­pendent oversight). In the case of an oversight committee, it would be challenging to provide a meaningful role for true industry outsiders. For example, laypersons not intimately familiar with the industry would face credibility challenges, and their findings could be too easily dismissed. 430 The matter of public engagement on highly technical matters is a recurring challenge in the administrative law context generally. See Emily Hammond, Nuclear Power, Risk, and Retroactivity, 48 Vand. J. Transnat’l L. 1059, 1077–78 (2015) (describing agencies’ tendency to be particularly unresponsive to the public interest in specialized or technical areas). Further, SROs jealously guard their industry data; 431 Recall NERC’s considerable opposition to FERC’s request for access to NERC databases. Supra text accompanying notes 276–277; see also Omarova, Wall Street, supra note 7, at 489 (raising nondisclosure concerns). any oversight body would need to be carefully constrained using some mix of contract law and regulatory prohibitions to avoid the disclosure of sensitive data or intellectual property. Like the nuclear options, however, these approaches have the benefit of permitting the SROs to conduct their business without addi­tional procedural requirements.

The few other possibilities that are worth noting should probably be discarded as unworkable. For example, one option might be to extend the definition of “agency” under the APA to include SROs, which would open them to both procedural restrictions and judicial review. It is hard, however, to see the benefit of this approach, especially considering that finality requirements would mean that the oversight agency’s work would be part of the record anyway. The better approach is to use existing administrative law doctrine to capture the work of SROs, as suggested above. 432 See supra section IV.A (suggesting means by which administrative law can account for SROs). Another possibility might be to make SROs’ liability com­mensurate with that of the oversight agency for purposes of statutes like the Federal Tort Claims Act and Tucker Act, 433 Federal Torts Claims Act, Pub. L. No. 79-601, 60 Stat. 812 (codified as amended in scattered sections of 28 U.S.C. (2012)); Tucker Act, Pub. L. No. 97-164, 96 Stat. 39 (codified as amended in scattered sections of 28 U.S.C.). but the better approach would be to address such matters through savings clauses in the SRO–agency statutes. Even if these were workable approaches, however, they are only piecemeal. They are unsatisfying for constitutional and administrative law purposes because the relevant statutes target tort and contract matters. Such matters are surely important for government credibility, but they do not really address the concerns identified in this Article.

5. Best Practices. — It is worth emphasizing that oversight agencies and SROs can engage in legitimizing behavior sua sponte. 434 See generally Hammond & Markell, supra note 16 (presenting empirical results illustrating legitimizing behaviors at EPA in absence of judicial review). In fact, by originating participatory, deliberative, and transparent initiatives from within, both entities have the advantage of tailoring those initiatives to the needs of regulated industries and outside observers alike. Trans­parency measures are likely the easiest to accomplish. For example, some SROs have user-friendly websites that provide up-to-date information on rules, votes, and enforcement matters and outcomes. Oversight agencies could also dedicate web space specifically to their SROs so that actions stemming from such schemes could be easily accessed. When oversight agencies post rules for comment on their websites or Regulations.gov, they could also specifically flag the SRO rules for easy identification. Although it would require more resources, agencies might also conduct listening sessions on topics related to SRO performance, which would provide even those interested parties who are not members of the regulated industry with a means of participation in the SRO scheme. The possibilities for improved government are endless, of course. But as this Article has sought to demonstrate, there is much work that can be done in the SRO world.

Conclusion

This account of SROs in the world of administrative law has shown that there are important reasons to doubt the prevailing assumptions concerning this type of regulatory scheme. Because those assumptions underlie much of the policy, judicial, and scholarly literature on SROs, a full rethinking of SROs as regulatory actors is justified. This Article has begun that task by providing both a positive and theoretical account of SROs, situating them within the broader norms of administrative law. This analysis reveals a number of deficiencies. In particular, the combina­tion of oversight agencies’ deference to SROs and judicial deference to oversight agencies undermines both the constitutional and regulatory legitimacy of SROs. Making room for SROs in the theory of administra­tive law doctrine can at least partly address these short­comings. Further, a variety of institutional design choices and procedural innovations are available to buttress SROs’ legitimacy. These prescriptive approaches stand to better promote accountability and guard against arbitrariness not only for SROs but also for the modern regulatory state.